diff --git "a/checkpoint-16459/trainer_state.json" "b/checkpoint-16459/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-16459/trainer_state.json" @@ -0,0 +1,115247 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 16459, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.075703262652652e-05, + "grad_norm": 0.2272067368030548, + "learning_rate": 0.0, + "loss": 1.1955, + "step": 1 + }, + { + "epoch": 0.00012151406525305304, + "grad_norm": 32.889320373535156, + "learning_rate": 2.0000000000000003e-06, + "loss": 1.4223, + "step": 2 + }, + { + "epoch": 0.00018227109787957957, + "grad_norm": 0.25457024574279785, + "learning_rate": 4.000000000000001e-06, + "loss": 1.2148, + "step": 3 + }, + { + "epoch": 0.0002430281305061061, + "grad_norm": 0.32749560475349426, + "learning_rate": 6e-06, + "loss": 1.4133, + "step": 4 + }, + { + "epoch": 0.0003037851631326326, + "grad_norm": 0.3309006094932556, + "learning_rate": 8.000000000000001e-06, + "loss": 1.4186, + "step": 5 + }, + { + "epoch": 0.00036454219575915913, + "grad_norm": 0.23534928262233734, + "learning_rate": 1e-05, + "loss": 1.2312, + "step": 6 + }, + { + "epoch": 0.00042529922838568566, + "grad_norm": 0.6984177827835083, + "learning_rate": 1.2e-05, + "loss": 1.4328, + "step": 7 + }, + { + "epoch": 0.0004860562610122122, + "grad_norm": 0.2976743280887604, + "learning_rate": 1.4000000000000001e-05, + "loss": 1.2253, + "step": 8 + }, + { + "epoch": 0.0005468132936387387, + "grad_norm": 0.3609892427921295, + "learning_rate": 1.6000000000000003e-05, + "loss": 1.3907, + "step": 9 + }, + { + "epoch": 0.0006075703262652652, + "grad_norm": 0.34851765632629395, + "learning_rate": 1.8e-05, + "loss": 1.2117, + "step": 10 + }, + { + "epoch": 0.0006683273588917917, + "grad_norm": 0.34736311435699463, + "learning_rate": 2e-05, + "loss": 1.2304, + "step": 11 + }, + { + "epoch": 0.0007290843915183183, + "grad_norm": 0.4581914246082306, + "learning_rate": 2.2000000000000003e-05, + "loss": 1.2845, + "step": 12 + }, + { + "epoch": 0.0007898414241448448, + "grad_norm": 0.3641209006309509, + "learning_rate": 2.4e-05, + "loss": 1.2555, + "step": 13 + }, + { + "epoch": 0.0008505984567713713, + "grad_norm": 0.3929094672203064, + "learning_rate": 2.6000000000000002e-05, + "loss": 1.3666, + "step": 14 + }, + { + "epoch": 0.0009113554893978978, + "grad_norm": 6.762740135192871, + "learning_rate": 2.8000000000000003e-05, + "loss": 1.4476, + "step": 15 + }, + { + "epoch": 0.0009721125220244244, + "grad_norm": 0.4653647840023041, + "learning_rate": 3e-05, + "loss": 1.275, + "step": 16 + }, + { + "epoch": 0.0010328695546509509, + "grad_norm": 0.38922420144081116, + "learning_rate": 3.2000000000000005e-05, + "loss": 1.368, + "step": 17 + }, + { + "epoch": 0.0010936265872774774, + "grad_norm": 0.42700445652008057, + "learning_rate": 3.4000000000000007e-05, + "loss": 1.3997, + "step": 18 + }, + { + "epoch": 0.001154383619904004, + "grad_norm": 0.478219211101532, + "learning_rate": 3.6e-05, + "loss": 1.3368, + "step": 19 + }, + { + "epoch": 0.0012151406525305304, + "grad_norm": 0.41081514954566956, + "learning_rate": 3.8e-05, + "loss": 1.2107, + "step": 20 + }, + { + "epoch": 0.001275897685157057, + "grad_norm": 0.49198776483535767, + "learning_rate": 4e-05, + "loss": 1.359, + "step": 21 + }, + { + "epoch": 0.0013366547177835835, + "grad_norm": 0.3181796371936798, + "learning_rate": 4.2e-05, + "loss": 1.2263, + "step": 22 + }, + { + "epoch": 0.00139741175041011, + "grad_norm": 0.5175872445106506, + "learning_rate": 4.4000000000000006e-05, + "loss": 1.414, + "step": 23 + }, + { + "epoch": 0.0014581687830366365, + "grad_norm": 0.36023858189582825, + "learning_rate": 4.600000000000001e-05, + "loss": 1.2314, + "step": 24 + }, + { + "epoch": 0.001518925815663163, + "grad_norm": 0.39242467284202576, + "learning_rate": 4.8e-05, + "loss": 1.2258, + "step": 25 + }, + { + "epoch": 0.0015796828482896896, + "grad_norm": 0.5171683430671692, + "learning_rate": 5e-05, + "loss": 1.2817, + "step": 26 + }, + { + "epoch": 0.001640439880916216, + "grad_norm": 2.84808349609375, + "learning_rate": 5.2000000000000004e-05, + "loss": 1.2504, + "step": 27 + }, + { + "epoch": 0.0017011969135427426, + "grad_norm": 0.7344406843185425, + "learning_rate": 5.4000000000000005e-05, + "loss": 1.2022, + "step": 28 + }, + { + "epoch": 0.0017619539461692691, + "grad_norm": 0.3675571084022522, + "learning_rate": 5.6000000000000006e-05, + "loss": 1.3043, + "step": 29 + }, + { + "epoch": 0.0018227109787957957, + "grad_norm": 0.31444838643074036, + "learning_rate": 5.8e-05, + "loss": 1.1916, + "step": 30 + }, + { + "epoch": 0.0018834680114223222, + "grad_norm": 0.4948585331439972, + "learning_rate": 6e-05, + "loss": 1.1743, + "step": 31 + }, + { + "epoch": 0.0019442250440488487, + "grad_norm": 0.3522862195968628, + "learning_rate": 6.2e-05, + "loss": 1.2392, + "step": 32 + }, + { + "epoch": 0.002004982076675375, + "grad_norm": 2.315263032913208, + "learning_rate": 6.400000000000001e-05, + "loss": 1.304, + "step": 33 + }, + { + "epoch": 0.0020657391093019018, + "grad_norm": 0.5789252519607544, + "learning_rate": 6.6e-05, + "loss": 1.2751, + "step": 34 + }, + { + "epoch": 0.002126496141928428, + "grad_norm": 0.3697313368320465, + "learning_rate": 6.800000000000001e-05, + "loss": 1.1918, + "step": 35 + }, + { + "epoch": 0.002187253174554955, + "grad_norm": 22.630043029785156, + "learning_rate": 7e-05, + "loss": 1.4206, + "step": 36 + }, + { + "epoch": 0.002248010207181481, + "grad_norm": 0.5834069848060608, + "learning_rate": 7.2e-05, + "loss": 1.2229, + "step": 37 + }, + { + "epoch": 0.002308767239808008, + "grad_norm": 0.43790143728256226, + "learning_rate": 7.4e-05, + "loss": 1.2898, + "step": 38 + }, + { + "epoch": 0.002369524272434534, + "grad_norm": 0.5147876739501953, + "learning_rate": 7.6e-05, + "loss": 1.1915, + "step": 39 + }, + { + "epoch": 0.002430281305061061, + "grad_norm": 0.8219349384307861, + "learning_rate": 7.800000000000001e-05, + "loss": 1.4763, + "step": 40 + }, + { + "epoch": 0.002491038337687587, + "grad_norm": 0.42311155796051025, + "learning_rate": 8e-05, + "loss": 1.283, + "step": 41 + }, + { + "epoch": 0.002551795370314114, + "grad_norm": 1.8067470788955688, + "learning_rate": 8.2e-05, + "loss": 1.3324, + "step": 42 + }, + { + "epoch": 0.0026125524029406402, + "grad_norm": 0.765227198600769, + "learning_rate": 8.4e-05, + "loss": 1.3751, + "step": 43 + }, + { + "epoch": 0.002673309435567167, + "grad_norm": 0.34523752331733704, + "learning_rate": 8.6e-05, + "loss": 1.2064, + "step": 44 + }, + { + "epoch": 0.0027340664681936933, + "grad_norm": 0.5671051144599915, + "learning_rate": 8.800000000000001e-05, + "loss": 1.3814, + "step": 45 + }, + { + "epoch": 0.00279482350082022, + "grad_norm": 0.6613193154335022, + "learning_rate": 9e-05, + "loss": 1.4091, + "step": 46 + }, + { + "epoch": 0.0028555805334467463, + "grad_norm": 0.4120130240917206, + "learning_rate": 9.200000000000001e-05, + "loss": 1.2027, + "step": 47 + }, + { + "epoch": 0.002916337566073273, + "grad_norm": 0.6605786681175232, + "learning_rate": 9.4e-05, + "loss": 1.39, + "step": 48 + }, + { + "epoch": 0.0029770945986997994, + "grad_norm": 0.40838339924812317, + "learning_rate": 9.6e-05, + "loss": 1.2996, + "step": 49 + }, + { + "epoch": 0.003037851631326326, + "grad_norm": 0.45898669958114624, + "learning_rate": 9.8e-05, + "loss": 1.2495, + "step": 50 + }, + { + "epoch": 0.0030986086639528524, + "grad_norm": 0.6189008951187134, + "learning_rate": 0.0001, + "loss": 1.2385, + "step": 51 + }, + { + "epoch": 0.003159365696579379, + "grad_norm": 1.644277811050415, + "learning_rate": 9.999999908362017e-05, + "loss": 1.217, + "step": 52 + }, + { + "epoch": 0.0032201227292059055, + "grad_norm": 0.5897544622421265, + "learning_rate": 9.999999633448071e-05, + "loss": 1.3365, + "step": 53 + }, + { + "epoch": 0.003280879761832432, + "grad_norm": 0.47523191571235657, + "learning_rate": 9.999999175258171e-05, + "loss": 1.24, + "step": 54 + }, + { + "epoch": 0.0033416367944589585, + "grad_norm": 0.6498791575431824, + "learning_rate": 9.999998533792336e-05, + "loss": 1.2173, + "step": 55 + }, + { + "epoch": 0.0034023938270854853, + "grad_norm": 0.3443443477153778, + "learning_rate": 9.999997709050588e-05, + "loss": 1.2066, + "step": 56 + }, + { + "epoch": 0.0034631508597120116, + "grad_norm": 0.9542787075042725, + "learning_rate": 9.999996701032956e-05, + "loss": 1.2749, + "step": 57 + }, + { + "epoch": 0.0035239078923385383, + "grad_norm": 0.40234601497650146, + "learning_rate": 9.99999550973948e-05, + "loss": 1.1697, + "step": 58 + }, + { + "epoch": 0.0035846649249650646, + "grad_norm": 0.4386552572250366, + "learning_rate": 9.999994135170202e-05, + "loss": 1.2016, + "step": 59 + }, + { + "epoch": 0.0036454219575915913, + "grad_norm": 2.121547222137451, + "learning_rate": 9.999992577325172e-05, + "loss": 1.3354, + "step": 60 + }, + { + "epoch": 0.0037061789902181176, + "grad_norm": 1.1630247831344604, + "learning_rate": 9.999990836204447e-05, + "loss": 1.1821, + "step": 61 + }, + { + "epoch": 0.0037669360228446444, + "grad_norm": 0.432586669921875, + "learning_rate": 9.999988911808093e-05, + "loss": 1.3251, + "step": 62 + }, + { + "epoch": 0.0038276930554711707, + "grad_norm": 0.6474998593330383, + "learning_rate": 9.999986804136179e-05, + "loss": 1.2372, + "step": 63 + }, + { + "epoch": 0.0038884500880976974, + "grad_norm": 0.8090982437133789, + "learning_rate": 9.99998451318878e-05, + "loss": 1.2748, + "step": 64 + }, + { + "epoch": 0.003949207120724224, + "grad_norm": 0.5398616194725037, + "learning_rate": 9.999982038965985e-05, + "loss": 1.2084, + "step": 65 + }, + { + "epoch": 0.00400996415335075, + "grad_norm": 0.3661274015903473, + "learning_rate": 9.999979381467879e-05, + "loss": 1.1465, + "step": 66 + }, + { + "epoch": 0.004070721185977277, + "grad_norm": 3.7460553646087646, + "learning_rate": 9.999976540694564e-05, + "loss": 1.2355, + "step": 67 + }, + { + "epoch": 0.0041314782186038035, + "grad_norm": 0.3958328664302826, + "learning_rate": 9.999973516646142e-05, + "loss": 1.3393, + "step": 68 + }, + { + "epoch": 0.00419223525123033, + "grad_norm": 0.8217223882675171, + "learning_rate": 9.999970309322725e-05, + "loss": 1.2306, + "step": 69 + }, + { + "epoch": 0.004252992283856856, + "grad_norm": 0.45304906368255615, + "learning_rate": 9.999966918724429e-05, + "loss": 1.2571, + "step": 70 + }, + { + "epoch": 0.004313749316483383, + "grad_norm": 0.7765888571739197, + "learning_rate": 9.999963344851379e-05, + "loss": 1.3844, + "step": 71 + }, + { + "epoch": 0.00437450634910991, + "grad_norm": 0.5228551030158997, + "learning_rate": 9.999959587703706e-05, + "loss": 1.2533, + "step": 72 + }, + { + "epoch": 0.004435263381736436, + "grad_norm": 0.9935314059257507, + "learning_rate": 9.999955647281549e-05, + "loss": 1.1874, + "step": 73 + }, + { + "epoch": 0.004496020414362962, + "grad_norm": 1.5693347454071045, + "learning_rate": 9.999951523585051e-05, + "loss": 1.2142, + "step": 74 + }, + { + "epoch": 0.004556777446989489, + "grad_norm": 0.8428077101707458, + "learning_rate": 9.999947216614363e-05, + "loss": 1.1838, + "step": 75 + }, + { + "epoch": 0.004617534479616016, + "grad_norm": 0.5332551598548889, + "learning_rate": 9.999942726369644e-05, + "loss": 1.1809, + "step": 76 + }, + { + "epoch": 0.004678291512242542, + "grad_norm": 0.8460003733634949, + "learning_rate": 9.999938052851057e-05, + "loss": 1.237, + "step": 77 + }, + { + "epoch": 0.004739048544869068, + "grad_norm": 0.42547518014907837, + "learning_rate": 9.999933196058775e-05, + "loss": 1.2255, + "step": 78 + }, + { + "epoch": 0.0047998055774955955, + "grad_norm": 0.8895686864852905, + "learning_rate": 9.999928155992975e-05, + "loss": 1.3089, + "step": 79 + }, + { + "epoch": 0.004860562610122122, + "grad_norm": 0.4807729721069336, + "learning_rate": 9.99992293265384e-05, + "loss": 1.3781, + "step": 80 + }, + { + "epoch": 0.004921319642748648, + "grad_norm": 0.9749037623405457, + "learning_rate": 9.999917526041567e-05, + "loss": 1.3136, + "step": 81 + }, + { + "epoch": 0.004982076675375174, + "grad_norm": 0.5286920070648193, + "learning_rate": 9.999911936156348e-05, + "loss": 1.1842, + "step": 82 + }, + { + "epoch": 0.005042833708001702, + "grad_norm": 0.44175127148628235, + "learning_rate": 9.999906162998392e-05, + "loss": 1.2956, + "step": 83 + }, + { + "epoch": 0.005103590740628228, + "grad_norm": 1.4931577444076538, + "learning_rate": 9.999900206567906e-05, + "loss": 1.1799, + "step": 84 + }, + { + "epoch": 0.005164347773254754, + "grad_norm": 0.5636366605758667, + "learning_rate": 9.999894066865115e-05, + "loss": 1.2039, + "step": 85 + }, + { + "epoch": 0.0052251048058812805, + "grad_norm": 0.45222145318984985, + "learning_rate": 9.999887743890239e-05, + "loss": 1.3643, + "step": 86 + }, + { + "epoch": 0.005285861838507808, + "grad_norm": 0.5859764814376831, + "learning_rate": 9.999881237643512e-05, + "loss": 1.3885, + "step": 87 + }, + { + "epoch": 0.005346618871134334, + "grad_norm": 0.6274478435516357, + "learning_rate": 9.999874548125172e-05, + "loss": 1.223, + "step": 88 + }, + { + "epoch": 0.00540737590376086, + "grad_norm": 0.6620491743087769, + "learning_rate": 9.999867675335464e-05, + "loss": 1.1898, + "step": 89 + }, + { + "epoch": 0.005468132936387387, + "grad_norm": 0.6583855748176575, + "learning_rate": 9.999860619274639e-05, + "loss": 1.2295, + "step": 90 + }, + { + "epoch": 0.005528889969013914, + "grad_norm": 0.4854571223258972, + "learning_rate": 9.999853379942956e-05, + "loss": 1.3241, + "step": 91 + }, + { + "epoch": 0.00558964700164044, + "grad_norm": 0.3728140592575073, + "learning_rate": 9.999845957340684e-05, + "loss": 1.2128, + "step": 92 + }, + { + "epoch": 0.005650404034266966, + "grad_norm": 0.5496470928192139, + "learning_rate": 9.999838351468088e-05, + "loss": 1.2053, + "step": 93 + }, + { + "epoch": 0.005711161066893493, + "grad_norm": 0.36465147137641907, + "learning_rate": 9.999830562325453e-05, + "loss": 1.3135, + "step": 94 + }, + { + "epoch": 0.00577191809952002, + "grad_norm": 0.7408778071403503, + "learning_rate": 9.999822589913062e-05, + "loss": 1.2318, + "step": 95 + }, + { + "epoch": 0.005832675132146546, + "grad_norm": 0.4084360897541046, + "learning_rate": 9.999814434231207e-05, + "loss": 1.1696, + "step": 96 + }, + { + "epoch": 0.0058934321647730725, + "grad_norm": 0.4670582115650177, + "learning_rate": 9.999806095280189e-05, + "loss": 1.1783, + "step": 97 + }, + { + "epoch": 0.005954189197399599, + "grad_norm": 0.36866235733032227, + "learning_rate": 9.99979757306031e-05, + "loss": 1.1781, + "step": 98 + }, + { + "epoch": 0.006014946230026126, + "grad_norm": 0.36132994294166565, + "learning_rate": 9.999788867571887e-05, + "loss": 1.1912, + "step": 99 + }, + { + "epoch": 0.006075703262652652, + "grad_norm": 0.5671958923339844, + "learning_rate": 9.999779978815233e-05, + "loss": 1.219, + "step": 100 + }, + { + "epoch": 0.0061364602952791785, + "grad_norm": 0.3515869677066803, + "learning_rate": 9.99977090679068e-05, + "loss": 1.2068, + "step": 101 + }, + { + "epoch": 0.006197217327905705, + "grad_norm": 0.4869772493839264, + "learning_rate": 9.999761651498559e-05, + "loss": 1.2504, + "step": 102 + }, + { + "epoch": 0.006257974360532232, + "grad_norm": 0.9631686806678772, + "learning_rate": 9.999752212939206e-05, + "loss": 1.3453, + "step": 103 + }, + { + "epoch": 0.006318731393158758, + "grad_norm": 0.3553156554698944, + "learning_rate": 9.99974259111297e-05, + "loss": 1.2309, + "step": 104 + }, + { + "epoch": 0.006379488425785285, + "grad_norm": 0.34895628690719604, + "learning_rate": 9.999732786020203e-05, + "loss": 1.2026, + "step": 105 + }, + { + "epoch": 0.006440245458411811, + "grad_norm": 0.4009360671043396, + "learning_rate": 9.999722797661264e-05, + "loss": 1.2551, + "step": 106 + }, + { + "epoch": 0.006501002491038338, + "grad_norm": 0.3186067044734955, + "learning_rate": 9.99971262603652e-05, + "loss": 1.1718, + "step": 107 + }, + { + "epoch": 0.006561759523664864, + "grad_norm": 0.4459477365016937, + "learning_rate": 9.999702271146343e-05, + "loss": 1.2384, + "step": 108 + }, + { + "epoch": 0.006622516556291391, + "grad_norm": 0.3618723452091217, + "learning_rate": 9.999691732991112e-05, + "loss": 1.1969, + "step": 109 + }, + { + "epoch": 0.006683273588917917, + "grad_norm": 0.4399377405643463, + "learning_rate": 9.999681011571215e-05, + "loss": 1.1419, + "step": 110 + }, + { + "epoch": 0.006744030621544444, + "grad_norm": 0.3510799705982208, + "learning_rate": 9.999670106887045e-05, + "loss": 1.3376, + "step": 111 + }, + { + "epoch": 0.0068047876541709705, + "grad_norm": 0.4912393093109131, + "learning_rate": 9.999659018938998e-05, + "loss": 1.3611, + "step": 112 + }, + { + "epoch": 0.006865544686797497, + "grad_norm": 0.27913954854011536, + "learning_rate": 9.999647747727485e-05, + "loss": 1.1634, + "step": 113 + }, + { + "epoch": 0.006926301719424023, + "grad_norm": 0.34383776783943176, + "learning_rate": 9.999636293252919e-05, + "loss": 1.1658, + "step": 114 + }, + { + "epoch": 0.006987058752050549, + "grad_norm": 0.2893190383911133, + "learning_rate": 9.999624655515716e-05, + "loss": 1.1973, + "step": 115 + }, + { + "epoch": 0.007047815784677077, + "grad_norm": 0.3170207440853119, + "learning_rate": 9.999612834516305e-05, + "loss": 1.3758, + "step": 116 + }, + { + "epoch": 0.007108572817303603, + "grad_norm": 0.33819878101348877, + "learning_rate": 9.99960083025512e-05, + "loss": 1.401, + "step": 117 + }, + { + "epoch": 0.007169329849930129, + "grad_norm": 0.4710005819797516, + "learning_rate": 9.9995886427326e-05, + "loss": 1.1941, + "step": 118 + }, + { + "epoch": 0.0072300868825566555, + "grad_norm": 0.36113208532333374, + "learning_rate": 9.999576271949192e-05, + "loss": 1.2346, + "step": 119 + }, + { + "epoch": 0.007290843915183183, + "grad_norm": 0.4262344539165497, + "learning_rate": 9.999563717905349e-05, + "loss": 1.294, + "step": 120 + }, + { + "epoch": 0.007351600947809709, + "grad_norm": 0.38293981552124023, + "learning_rate": 9.999550980601531e-05, + "loss": 1.2922, + "step": 121 + }, + { + "epoch": 0.007412357980436235, + "grad_norm": 0.3802623450756073, + "learning_rate": 9.999538060038207e-05, + "loss": 1.2509, + "step": 122 + }, + { + "epoch": 0.007473115013062762, + "grad_norm": 0.5138214230537415, + "learning_rate": 9.999524956215848e-05, + "loss": 1.1748, + "step": 123 + }, + { + "epoch": 0.007533872045689289, + "grad_norm": 0.43086934089660645, + "learning_rate": 9.999511669134935e-05, + "loss": 1.1998, + "step": 124 + }, + { + "epoch": 0.007594629078315815, + "grad_norm": 0.49192190170288086, + "learning_rate": 9.999498198795955e-05, + "loss": 1.2272, + "step": 125 + }, + { + "epoch": 0.007655386110942341, + "grad_norm": 26.668746948242188, + "learning_rate": 9.999484545199405e-05, + "loss": 1.2224, + "step": 126 + }, + { + "epoch": 0.007716143143568868, + "grad_norm": 1.2917407751083374, + "learning_rate": 9.999470708345779e-05, + "loss": 1.1926, + "step": 127 + }, + { + "epoch": 0.007776900176195395, + "grad_norm": 4.33660364151001, + "learning_rate": 9.99945668823559e-05, + "loss": 1.196, + "step": 128 + }, + { + "epoch": 0.007837657208821921, + "grad_norm": 1.123703956604004, + "learning_rate": 9.999442484869348e-05, + "loss": 1.3309, + "step": 129 + }, + { + "epoch": 0.007898414241448447, + "grad_norm": 0.3729478418827057, + "learning_rate": 9.999428098247576e-05, + "loss": 1.3245, + "step": 130 + }, + { + "epoch": 0.007959171274074974, + "grad_norm": 1.2514550685882568, + "learning_rate": 9.999413528370803e-05, + "loss": 1.1774, + "step": 131 + }, + { + "epoch": 0.0080199283067015, + "grad_norm": 0.4410100281238556, + "learning_rate": 9.999398775239559e-05, + "loss": 1.1742, + "step": 132 + }, + { + "epoch": 0.008080685339328026, + "grad_norm": 0.5370081067085266, + "learning_rate": 9.999383838854385e-05, + "loss": 1.2059, + "step": 133 + }, + { + "epoch": 0.008141442371954554, + "grad_norm": 2.5626380443573, + "learning_rate": 9.999368719215831e-05, + "loss": 1.279, + "step": 134 + }, + { + "epoch": 0.00820219940458108, + "grad_norm": 1.4754201173782349, + "learning_rate": 9.999353416324452e-05, + "loss": 1.2805, + "step": 135 + }, + { + "epoch": 0.008262956437207607, + "grad_norm": 0.7902483940124512, + "learning_rate": 9.999337930180805e-05, + "loss": 1.2813, + "step": 136 + }, + { + "epoch": 0.008323713469834133, + "grad_norm": 0.9314702153205872, + "learning_rate": 9.999322260785461e-05, + "loss": 1.3383, + "step": 137 + }, + { + "epoch": 0.00838447050246066, + "grad_norm": 0.6036847233772278, + "learning_rate": 9.999306408138994e-05, + "loss": 1.2039, + "step": 138 + }, + { + "epoch": 0.008445227535087186, + "grad_norm": 0.6201358437538147, + "learning_rate": 9.999290372241982e-05, + "loss": 1.2012, + "step": 139 + }, + { + "epoch": 0.008505984567713712, + "grad_norm": 0.6081859469413757, + "learning_rate": 9.999274153095016e-05, + "loss": 1.2378, + "step": 140 + }, + { + "epoch": 0.008566741600340239, + "grad_norm": 0.5100321173667908, + "learning_rate": 9.99925775069869e-05, + "loss": 1.3004, + "step": 141 + }, + { + "epoch": 0.008627498632966767, + "grad_norm": 0.32394978404045105, + "learning_rate": 9.999241165053605e-05, + "loss": 1.2321, + "step": 142 + }, + { + "epoch": 0.008688255665593293, + "grad_norm": 0.5560949444770813, + "learning_rate": 9.999224396160367e-05, + "loss": 1.3599, + "step": 143 + }, + { + "epoch": 0.00874901269821982, + "grad_norm": 0.3612901270389557, + "learning_rate": 9.999207444019594e-05, + "loss": 1.2593, + "step": 144 + }, + { + "epoch": 0.008809769730846346, + "grad_norm": 0.5922803282737732, + "learning_rate": 9.999190308631906e-05, + "loss": 1.2144, + "step": 145 + }, + { + "epoch": 0.008870526763472872, + "grad_norm": 0.44897040724754333, + "learning_rate": 9.999172989997931e-05, + "loss": 1.2133, + "step": 146 + }, + { + "epoch": 0.008931283796099398, + "grad_norm": 0.3743618130683899, + "learning_rate": 9.999155488118302e-05, + "loss": 1.1651, + "step": 147 + }, + { + "epoch": 0.008992040828725924, + "grad_norm": 0.8505854606628418, + "learning_rate": 9.999137802993664e-05, + "loss": 1.1771, + "step": 148 + }, + { + "epoch": 0.00905279786135245, + "grad_norm": 0.6462696194648743, + "learning_rate": 9.999119934624663e-05, + "loss": 1.2366, + "step": 149 + }, + { + "epoch": 0.009113554893978979, + "grad_norm": 0.5867897272109985, + "learning_rate": 9.999101883011954e-05, + "loss": 1.2386, + "step": 150 + }, + { + "epoch": 0.009174311926605505, + "grad_norm": 0.4128705859184265, + "learning_rate": 9.9990836481562e-05, + "loss": 1.2075, + "step": 151 + }, + { + "epoch": 0.009235068959232031, + "grad_norm": 0.325204461812973, + "learning_rate": 9.99906523005807e-05, + "loss": 1.2498, + "step": 152 + }, + { + "epoch": 0.009295825991858558, + "grad_norm": 0.2922727167606354, + "learning_rate": 9.999046628718234e-05, + "loss": 1.2703, + "step": 153 + }, + { + "epoch": 0.009356583024485084, + "grad_norm": 1.0875980854034424, + "learning_rate": 9.99902784413738e-05, + "loss": 1.1914, + "step": 154 + }, + { + "epoch": 0.00941734005711161, + "grad_norm": 0.6689728498458862, + "learning_rate": 9.999008876316194e-05, + "loss": 1.2861, + "step": 155 + }, + { + "epoch": 0.009478097089738137, + "grad_norm": 0.42909765243530273, + "learning_rate": 9.99898972525537e-05, + "loss": 1.1622, + "step": 156 + }, + { + "epoch": 0.009538854122364663, + "grad_norm": 0.7490345239639282, + "learning_rate": 9.998970390955613e-05, + "loss": 1.2405, + "step": 157 + }, + { + "epoch": 0.009599611154991191, + "grad_norm": 0.4368993937969208, + "learning_rate": 9.99895087341763e-05, + "loss": 1.1538, + "step": 158 + }, + { + "epoch": 0.009660368187617717, + "grad_norm": 1.0807139873504639, + "learning_rate": 9.998931172642134e-05, + "loss": 1.1816, + "step": 159 + }, + { + "epoch": 0.009721125220244244, + "grad_norm": 0.46905410289764404, + "learning_rate": 9.998911288629852e-05, + "loss": 1.2833, + "step": 160 + }, + { + "epoch": 0.00978188225287077, + "grad_norm": 0.42148005962371826, + "learning_rate": 9.99889122138151e-05, + "loss": 1.1867, + "step": 161 + }, + { + "epoch": 0.009842639285497296, + "grad_norm": 0.3812882900238037, + "learning_rate": 9.998870970897844e-05, + "loss": 1.1441, + "step": 162 + }, + { + "epoch": 0.009903396318123822, + "grad_norm": 0.3105202913284302, + "learning_rate": 9.998850537179596e-05, + "loss": 1.1941, + "step": 163 + }, + { + "epoch": 0.009964153350750349, + "grad_norm": 0.3256356716156006, + "learning_rate": 9.998829920227514e-05, + "loss": 1.3024, + "step": 164 + }, + { + "epoch": 0.010024910383376875, + "grad_norm": 0.5735190510749817, + "learning_rate": 9.998809120042358e-05, + "loss": 1.1676, + "step": 165 + }, + { + "epoch": 0.010085667416003403, + "grad_norm": 0.540168285369873, + "learning_rate": 9.998788136624885e-05, + "loss": 1.2629, + "step": 166 + }, + { + "epoch": 0.01014642444862993, + "grad_norm": 0.5270628333091736, + "learning_rate": 9.998766969975868e-05, + "loss": 1.2002, + "step": 167 + }, + { + "epoch": 0.010207181481256456, + "grad_norm": 0.514396071434021, + "learning_rate": 9.99874562009608e-05, + "loss": 1.1502, + "step": 168 + }, + { + "epoch": 0.010267938513882982, + "grad_norm": 0.45396819710731506, + "learning_rate": 9.998724086986307e-05, + "loss": 1.1419, + "step": 169 + }, + { + "epoch": 0.010328695546509508, + "grad_norm": 0.3975193202495575, + "learning_rate": 9.998702370647335e-05, + "loss": 1.2583, + "step": 170 + }, + { + "epoch": 0.010389452579136035, + "grad_norm": 0.35191208124160767, + "learning_rate": 9.998680471079962e-05, + "loss": 1.2803, + "step": 171 + }, + { + "epoch": 0.010450209611762561, + "grad_norm": 0.37019944190979004, + "learning_rate": 9.99865838828499e-05, + "loss": 1.2539, + "step": 172 + }, + { + "epoch": 0.010510966644389087, + "grad_norm": 2.7112491130828857, + "learning_rate": 9.998636122263228e-05, + "loss": 1.1792, + "step": 173 + }, + { + "epoch": 0.010571723677015615, + "grad_norm": 0.3234882652759552, + "learning_rate": 9.998613673015494e-05, + "loss": 1.275, + "step": 174 + }, + { + "epoch": 0.010632480709642142, + "grad_norm": 0.2748982906341553, + "learning_rate": 9.998591040542608e-05, + "loss": 1.1987, + "step": 175 + }, + { + "epoch": 0.010693237742268668, + "grad_norm": 0.2726716995239258, + "learning_rate": 9.998568224845402e-05, + "loss": 1.2434, + "step": 176 + }, + { + "epoch": 0.010753994774895194, + "grad_norm": 0.35885581374168396, + "learning_rate": 9.998545225924711e-05, + "loss": 1.2138, + "step": 177 + }, + { + "epoch": 0.01081475180752172, + "grad_norm": 0.34916558861732483, + "learning_rate": 9.99852204378138e-05, + "loss": 1.2071, + "step": 178 + }, + { + "epoch": 0.010875508840148247, + "grad_norm": 9.863569259643555, + "learning_rate": 9.998498678416254e-05, + "loss": 1.1921, + "step": 179 + }, + { + "epoch": 0.010936265872774773, + "grad_norm": 0.4509379267692566, + "learning_rate": 9.998475129830196e-05, + "loss": 1.1722, + "step": 180 + }, + { + "epoch": 0.0109970229054013, + "grad_norm": 0.3903730511665344, + "learning_rate": 9.998451398024066e-05, + "loss": 1.2405, + "step": 181 + }, + { + "epoch": 0.011057779938027828, + "grad_norm": 0.3390004634857178, + "learning_rate": 9.998427482998732e-05, + "loss": 1.1666, + "step": 182 + }, + { + "epoch": 0.011118536970654354, + "grad_norm": 1.3930373191833496, + "learning_rate": 9.998403384755074e-05, + "loss": 1.1999, + "step": 183 + }, + { + "epoch": 0.01117929400328088, + "grad_norm": 0.3124197721481323, + "learning_rate": 9.998379103293973e-05, + "loss": 1.2067, + "step": 184 + }, + { + "epoch": 0.011240051035907406, + "grad_norm": 0.9586052298545837, + "learning_rate": 9.99835463861632e-05, + "loss": 1.1636, + "step": 185 + }, + { + "epoch": 0.011300808068533933, + "grad_norm": 0.4747290015220642, + "learning_rate": 9.998329990723011e-05, + "loss": 1.2245, + "step": 186 + }, + { + "epoch": 0.011361565101160459, + "grad_norm": 0.43976885080337524, + "learning_rate": 9.998305159614951e-05, + "loss": 1.211, + "step": 187 + }, + { + "epoch": 0.011422322133786985, + "grad_norm": 0.5567777156829834, + "learning_rate": 9.998280145293049e-05, + "loss": 1.3064, + "step": 188 + }, + { + "epoch": 0.011483079166413512, + "grad_norm": 0.592757523059845, + "learning_rate": 9.998254947758221e-05, + "loss": 1.2284, + "step": 189 + }, + { + "epoch": 0.01154383619904004, + "grad_norm": 0.6924203038215637, + "learning_rate": 9.998229567011393e-05, + "loss": 1.145, + "step": 190 + }, + { + "epoch": 0.011604593231666566, + "grad_norm": 0.6162521243095398, + "learning_rate": 9.998204003053492e-05, + "loss": 1.2998, + "step": 191 + }, + { + "epoch": 0.011665350264293092, + "grad_norm": 3.581974506378174, + "learning_rate": 9.99817825588546e-05, + "loss": 1.2817, + "step": 192 + }, + { + "epoch": 0.011726107296919619, + "grad_norm": 1.11673903465271, + "learning_rate": 9.998152325508236e-05, + "loss": 1.2531, + "step": 193 + }, + { + "epoch": 0.011786864329546145, + "grad_norm": 0.3923598825931549, + "learning_rate": 9.998126211922773e-05, + "loss": 1.3082, + "step": 194 + }, + { + "epoch": 0.011847621362172671, + "grad_norm": 0.9915650486946106, + "learning_rate": 9.998099915130029e-05, + "loss": 1.2604, + "step": 195 + }, + { + "epoch": 0.011908378394799198, + "grad_norm": 0.29419344663619995, + "learning_rate": 9.998073435130963e-05, + "loss": 1.1474, + "step": 196 + }, + { + "epoch": 0.011969135427425724, + "grad_norm": 1.095484972000122, + "learning_rate": 9.998046771926553e-05, + "loss": 1.3494, + "step": 197 + }, + { + "epoch": 0.012029892460052252, + "grad_norm": 0.5557772517204285, + "learning_rate": 9.99801992551777e-05, + "loss": 1.1667, + "step": 198 + }, + { + "epoch": 0.012090649492678778, + "grad_norm": 0.6271569728851318, + "learning_rate": 9.997992895905601e-05, + "loss": 1.2846, + "step": 199 + }, + { + "epoch": 0.012151406525305304, + "grad_norm": 0.6900817155838013, + "learning_rate": 9.997965683091037e-05, + "loss": 1.1911, + "step": 200 + }, + { + "epoch": 0.01221216355793183, + "grad_norm": 0.4427839517593384, + "learning_rate": 9.997938287075076e-05, + "loss": 1.2354, + "step": 201 + }, + { + "epoch": 0.012272920590558357, + "grad_norm": 0.8094660639762878, + "learning_rate": 9.997910707858718e-05, + "loss": 1.223, + "step": 202 + }, + { + "epoch": 0.012333677623184883, + "grad_norm": 0.42011329531669617, + "learning_rate": 9.997882945442978e-05, + "loss": 1.1965, + "step": 203 + }, + { + "epoch": 0.01239443465581141, + "grad_norm": 0.577828049659729, + "learning_rate": 9.997854999828873e-05, + "loss": 1.2149, + "step": 204 + }, + { + "epoch": 0.012455191688437936, + "grad_norm": 0.41236254572868347, + "learning_rate": 9.997826871017427e-05, + "loss": 1.2737, + "step": 205 + }, + { + "epoch": 0.012515948721064464, + "grad_norm": 1.4393216371536255, + "learning_rate": 9.99779855900967e-05, + "loss": 1.1964, + "step": 206 + }, + { + "epoch": 0.01257670575369099, + "grad_norm": 0.3379725515842438, + "learning_rate": 9.997770063806641e-05, + "loss": 1.1457, + "step": 207 + }, + { + "epoch": 0.012637462786317517, + "grad_norm": 0.6507219672203064, + "learning_rate": 9.997741385409385e-05, + "loss": 1.2841, + "step": 208 + }, + { + "epoch": 0.012698219818944043, + "grad_norm": 0.2724468410015106, + "learning_rate": 9.997712523818952e-05, + "loss": 1.1553, + "step": 209 + }, + { + "epoch": 0.01275897685157057, + "grad_norm": 0.69343501329422, + "learning_rate": 9.997683479036401e-05, + "loss": 1.1462, + "step": 210 + }, + { + "epoch": 0.012819733884197096, + "grad_norm": 0.2509300708770752, + "learning_rate": 9.997654251062795e-05, + "loss": 1.1805, + "step": 211 + }, + { + "epoch": 0.012880490916823622, + "grad_norm": 0.4491214156150818, + "learning_rate": 9.997624839899208e-05, + "loss": 1.1818, + "step": 212 + }, + { + "epoch": 0.012941247949450148, + "grad_norm": 0.24392662942409515, + "learning_rate": 9.997595245546716e-05, + "loss": 1.1732, + "step": 213 + }, + { + "epoch": 0.013002004982076676, + "grad_norm": 0.37815433740615845, + "learning_rate": 9.997565468006404e-05, + "loss": 1.2197, + "step": 214 + }, + { + "epoch": 0.013062762014703203, + "grad_norm": 0.2204202115535736, + "learning_rate": 9.997535507279363e-05, + "loss": 1.2281, + "step": 215 + }, + { + "epoch": 0.013123519047329729, + "grad_norm": 0.23933321237564087, + "learning_rate": 9.997505363366693e-05, + "loss": 1.1892, + "step": 216 + }, + { + "epoch": 0.013184276079956255, + "grad_norm": 0.23179109394550323, + "learning_rate": 9.997475036269497e-05, + "loss": 1.1732, + "step": 217 + }, + { + "epoch": 0.013245033112582781, + "grad_norm": 0.2972233295440674, + "learning_rate": 9.997444525988888e-05, + "loss": 1.2794, + "step": 218 + }, + { + "epoch": 0.013305790145209308, + "grad_norm": 0.3704680800437927, + "learning_rate": 9.997413832525984e-05, + "loss": 1.2789, + "step": 219 + }, + { + "epoch": 0.013366547177835834, + "grad_norm": 0.32269278168678284, + "learning_rate": 9.997382955881911e-05, + "loss": 1.2216, + "step": 220 + }, + { + "epoch": 0.01342730421046236, + "grad_norm": 0.4552355706691742, + "learning_rate": 9.997351896057799e-05, + "loss": 1.1501, + "step": 221 + }, + { + "epoch": 0.013488061243088888, + "grad_norm": 0.34602460265159607, + "learning_rate": 9.997320653054787e-05, + "loss": 1.1963, + "step": 222 + }, + { + "epoch": 0.013548818275715415, + "grad_norm": 0.3155742585659027, + "learning_rate": 9.99728922687402e-05, + "loss": 1.2413, + "step": 223 + }, + { + "epoch": 0.013609575308341941, + "grad_norm": 0.2310158908367157, + "learning_rate": 9.997257617516651e-05, + "loss": 1.2018, + "step": 224 + }, + { + "epoch": 0.013670332340968467, + "grad_norm": 0.5281551480293274, + "learning_rate": 9.997225824983838e-05, + "loss": 1.2586, + "step": 225 + }, + { + "epoch": 0.013731089373594994, + "grad_norm": 0.2650713324546814, + "learning_rate": 9.997193849276746e-05, + "loss": 1.1501, + "step": 226 + }, + { + "epoch": 0.01379184640622152, + "grad_norm": 0.8100301623344421, + "learning_rate": 9.997161690396549e-05, + "loss": 1.288, + "step": 227 + }, + { + "epoch": 0.013852603438848046, + "grad_norm": 0.43822601437568665, + "learning_rate": 9.997129348344423e-05, + "loss": 1.2075, + "step": 228 + }, + { + "epoch": 0.013913360471474573, + "grad_norm": 0.4594722390174866, + "learning_rate": 9.997096823121555e-05, + "loss": 1.2061, + "step": 229 + }, + { + "epoch": 0.013974117504101099, + "grad_norm": 0.31025299429893494, + "learning_rate": 9.997064114729138e-05, + "loss": 1.1726, + "step": 230 + }, + { + "epoch": 0.014034874536727627, + "grad_norm": 0.31103456020355225, + "learning_rate": 9.997031223168368e-05, + "loss": 1.2237, + "step": 231 + }, + { + "epoch": 0.014095631569354153, + "grad_norm": 0.35022541880607605, + "learning_rate": 9.996998148440454e-05, + "loss": 1.1849, + "step": 232 + }, + { + "epoch": 0.01415638860198068, + "grad_norm": 0.2876299321651459, + "learning_rate": 9.996964890546607e-05, + "loss": 1.1829, + "step": 233 + }, + { + "epoch": 0.014217145634607206, + "grad_norm": 0.5656312704086304, + "learning_rate": 9.996931449488045e-05, + "loss": 1.267, + "step": 234 + }, + { + "epoch": 0.014277902667233732, + "grad_norm": 0.47967249155044556, + "learning_rate": 9.996897825265996e-05, + "loss": 1.2922, + "step": 235 + }, + { + "epoch": 0.014338659699860258, + "grad_norm": 0.316143274307251, + "learning_rate": 9.99686401788169e-05, + "loss": 1.2052, + "step": 236 + }, + { + "epoch": 0.014399416732486785, + "grad_norm": 0.5354479551315308, + "learning_rate": 9.996830027336368e-05, + "loss": 1.2167, + "step": 237 + }, + { + "epoch": 0.014460173765113311, + "grad_norm": 0.3296726644039154, + "learning_rate": 9.996795853631278e-05, + "loss": 1.2922, + "step": 238 + }, + { + "epoch": 0.014520930797739839, + "grad_norm": 0.30068060755729675, + "learning_rate": 9.996761496767668e-05, + "loss": 1.1743, + "step": 239 + }, + { + "epoch": 0.014581687830366365, + "grad_norm": 0.37530872225761414, + "learning_rate": 9.9967269567468e-05, + "loss": 1.207, + "step": 240 + }, + { + "epoch": 0.014642444862992892, + "grad_norm": 0.46817246079444885, + "learning_rate": 9.996692233569938e-05, + "loss": 1.2515, + "step": 241 + }, + { + "epoch": 0.014703201895619418, + "grad_norm": 0.6425913572311401, + "learning_rate": 9.996657327238357e-05, + "loss": 1.1974, + "step": 242 + }, + { + "epoch": 0.014763958928245944, + "grad_norm": 0.2988743782043457, + "learning_rate": 9.996622237753336e-05, + "loss": 1.1921, + "step": 243 + }, + { + "epoch": 0.01482471596087247, + "grad_norm": 0.35703039169311523, + "learning_rate": 9.996586965116161e-05, + "loss": 1.3515, + "step": 244 + }, + { + "epoch": 0.014885472993498997, + "grad_norm": 1.2079401016235352, + "learning_rate": 9.996551509328127e-05, + "loss": 1.2463, + "step": 245 + }, + { + "epoch": 0.014946230026125523, + "grad_norm": 0.41708889603614807, + "learning_rate": 9.996515870390529e-05, + "loss": 1.1414, + "step": 246 + }, + { + "epoch": 0.015006987058752051, + "grad_norm": 0.29742732644081116, + "learning_rate": 9.996480048304677e-05, + "loss": 1.135, + "step": 247 + }, + { + "epoch": 0.015067744091378578, + "grad_norm": 0.29110848903656006, + "learning_rate": 9.996444043071883e-05, + "loss": 1.1411, + "step": 248 + }, + { + "epoch": 0.015128501124005104, + "grad_norm": 0.36033013463020325, + "learning_rate": 9.996407854693467e-05, + "loss": 1.2, + "step": 249 + }, + { + "epoch": 0.01518925815663163, + "grad_norm": 0.29062187671661377, + "learning_rate": 9.996371483170754e-05, + "loss": 1.1381, + "step": 250 + }, + { + "epoch": 0.015250015189258156, + "grad_norm": 0.6057248115539551, + "learning_rate": 9.99633492850508e-05, + "loss": 1.4161, + "step": 251 + }, + { + "epoch": 0.015310772221884683, + "grad_norm": 0.4413861930370331, + "learning_rate": 9.996298190697783e-05, + "loss": 1.2491, + "step": 252 + }, + { + "epoch": 0.015371529254511209, + "grad_norm": 0.3748348653316498, + "learning_rate": 9.996261269750211e-05, + "loss": 1.1694, + "step": 253 + }, + { + "epoch": 0.015432286287137735, + "grad_norm": 0.35884901881217957, + "learning_rate": 9.996224165663714e-05, + "loss": 1.2678, + "step": 254 + }, + { + "epoch": 0.015493043319764263, + "grad_norm": 0.388274222612381, + "learning_rate": 9.996186878439657e-05, + "loss": 1.3369, + "step": 255 + }, + { + "epoch": 0.01555380035239079, + "grad_norm": 0.39521563053131104, + "learning_rate": 9.996149408079403e-05, + "loss": 1.2749, + "step": 256 + }, + { + "epoch": 0.015614557385017316, + "grad_norm": 0.4442555010318756, + "learning_rate": 9.996111754584327e-05, + "loss": 1.2565, + "step": 257 + }, + { + "epoch": 0.015675314417643842, + "grad_norm": 0.4751284122467041, + "learning_rate": 9.996073917955808e-05, + "loss": 1.1461, + "step": 258 + }, + { + "epoch": 0.01573607145027037, + "grad_norm": 0.3552243709564209, + "learning_rate": 9.996035898195236e-05, + "loss": 1.143, + "step": 259 + }, + { + "epoch": 0.015796828482896895, + "grad_norm": 0.5472645163536072, + "learning_rate": 9.995997695304e-05, + "loss": 1.1465, + "step": 260 + }, + { + "epoch": 0.015857585515523423, + "grad_norm": 0.3277062177658081, + "learning_rate": 9.995959309283505e-05, + "loss": 1.3338, + "step": 261 + }, + { + "epoch": 0.015918342548149948, + "grad_norm": 0.24876654148101807, + "learning_rate": 9.995920740135154e-05, + "loss": 1.1851, + "step": 262 + }, + { + "epoch": 0.015979099580776476, + "grad_norm": 0.5001147985458374, + "learning_rate": 9.995881987860363e-05, + "loss": 1.1476, + "step": 263 + }, + { + "epoch": 0.016039856613403, + "grad_norm": 0.5432358980178833, + "learning_rate": 9.995843052460552e-05, + "loss": 1.2811, + "step": 264 + }, + { + "epoch": 0.016100613646029528, + "grad_norm": 0.4481656551361084, + "learning_rate": 9.995803933937149e-05, + "loss": 1.2494, + "step": 265 + }, + { + "epoch": 0.016161370678656053, + "grad_norm": 0.28406116366386414, + "learning_rate": 9.995764632291586e-05, + "loss": 1.1978, + "step": 266 + }, + { + "epoch": 0.01622212771128258, + "grad_norm": 4.454953193664551, + "learning_rate": 9.995725147525306e-05, + "loss": 1.2991, + "step": 267 + }, + { + "epoch": 0.01628288474390911, + "grad_norm": 1.0255284309387207, + "learning_rate": 9.995685479639754e-05, + "loss": 1.1767, + "step": 268 + }, + { + "epoch": 0.016343641776535633, + "grad_norm": 0.3760567605495453, + "learning_rate": 9.995645628636385e-05, + "loss": 1.2063, + "step": 269 + }, + { + "epoch": 0.01640439880916216, + "grad_norm": 0.8572304248809814, + "learning_rate": 9.995605594516661e-05, + "loss": 1.1805, + "step": 270 + }, + { + "epoch": 0.016465155841788686, + "grad_norm": 0.3504545986652374, + "learning_rate": 9.995565377282048e-05, + "loss": 1.2098, + "step": 271 + }, + { + "epoch": 0.016525912874415214, + "grad_norm": 0.7503202557563782, + "learning_rate": 9.995524976934018e-05, + "loss": 1.1981, + "step": 272 + }, + { + "epoch": 0.01658666990704174, + "grad_norm": 0.3303241729736328, + "learning_rate": 9.995484393474057e-05, + "loss": 1.1468, + "step": 273 + }, + { + "epoch": 0.016647426939668267, + "grad_norm": 0.9021998643875122, + "learning_rate": 9.995443626903651e-05, + "loss": 1.2755, + "step": 274 + }, + { + "epoch": 0.016708183972294795, + "grad_norm": 0.5612683296203613, + "learning_rate": 9.995402677224291e-05, + "loss": 1.1973, + "step": 275 + }, + { + "epoch": 0.01676894100492132, + "grad_norm": 0.5486313104629517, + "learning_rate": 9.995361544437481e-05, + "loss": 1.2768, + "step": 276 + }, + { + "epoch": 0.016829698037547847, + "grad_norm": 0.4837646186351776, + "learning_rate": 9.995320228544727e-05, + "loss": 1.5086, + "step": 277 + }, + { + "epoch": 0.016890455070174372, + "grad_norm": 0.5270864367485046, + "learning_rate": 9.995278729547546e-05, + "loss": 1.1768, + "step": 278 + }, + { + "epoch": 0.0169512121028009, + "grad_norm": 0.3636312186717987, + "learning_rate": 9.995237047447458e-05, + "loss": 1.3855, + "step": 279 + }, + { + "epoch": 0.017011969135427425, + "grad_norm": 0.400898277759552, + "learning_rate": 9.995195182245988e-05, + "loss": 1.2722, + "step": 280 + }, + { + "epoch": 0.017072726168053953, + "grad_norm": 0.4978293776512146, + "learning_rate": 9.995153133944676e-05, + "loss": 1.3489, + "step": 281 + }, + { + "epoch": 0.017133483200680477, + "grad_norm": 0.5199602246284485, + "learning_rate": 9.995110902545058e-05, + "loss": 1.2775, + "step": 282 + }, + { + "epoch": 0.017194240233307005, + "grad_norm": 0.7290297150611877, + "learning_rate": 9.995068488048685e-05, + "loss": 1.1803, + "step": 283 + }, + { + "epoch": 0.017254997265933533, + "grad_norm": 0.5594682097434998, + "learning_rate": 9.995025890457113e-05, + "loss": 1.1638, + "step": 284 + }, + { + "epoch": 0.017315754298560058, + "grad_norm": 0.5723556876182556, + "learning_rate": 9.994983109771898e-05, + "loss": 1.1706, + "step": 285 + }, + { + "epoch": 0.017376511331186586, + "grad_norm": 0.3864780068397522, + "learning_rate": 9.994940145994614e-05, + "loss": 1.1599, + "step": 286 + }, + { + "epoch": 0.01743726836381311, + "grad_norm": 0.8249163031578064, + "learning_rate": 9.994896999126832e-05, + "loss": 1.1309, + "step": 287 + }, + { + "epoch": 0.01749802539643964, + "grad_norm": 0.7230120897293091, + "learning_rate": 9.994853669170136e-05, + "loss": 1.173, + "step": 288 + }, + { + "epoch": 0.017558782429066163, + "grad_norm": 0.5831359624862671, + "learning_rate": 9.994810156126114e-05, + "loss": 1.166, + "step": 289 + }, + { + "epoch": 0.01761953946169269, + "grad_norm": 1.0810701847076416, + "learning_rate": 9.994766459996359e-05, + "loss": 1.2559, + "step": 290 + }, + { + "epoch": 0.01768029649431922, + "grad_norm": 0.38416218757629395, + "learning_rate": 9.994722580782475e-05, + "loss": 1.1677, + "step": 291 + }, + { + "epoch": 0.017741053526945744, + "grad_norm": 0.5766205787658691, + "learning_rate": 9.994678518486067e-05, + "loss": 1.1603, + "step": 292 + }, + { + "epoch": 0.01780181055957227, + "grad_norm": 0.37862247228622437, + "learning_rate": 9.994634273108755e-05, + "loss": 1.1506, + "step": 293 + }, + { + "epoch": 0.017862567592198796, + "grad_norm": 0.59559166431427, + "learning_rate": 9.994589844652157e-05, + "loss": 1.1593, + "step": 294 + }, + { + "epoch": 0.017923324624825324, + "grad_norm": 0.3908030390739441, + "learning_rate": 9.994545233117903e-05, + "loss": 1.3295, + "step": 295 + }, + { + "epoch": 0.01798408165745185, + "grad_norm": 0.8364783525466919, + "learning_rate": 9.994500438507628e-05, + "loss": 1.3471, + "step": 296 + }, + { + "epoch": 0.018044838690078377, + "grad_norm": 0.3317735493183136, + "learning_rate": 9.994455460822975e-05, + "loss": 1.1764, + "step": 297 + }, + { + "epoch": 0.0181055957227049, + "grad_norm": 0.3481360971927643, + "learning_rate": 9.994410300065591e-05, + "loss": 1.1275, + "step": 298 + }, + { + "epoch": 0.01816635275533143, + "grad_norm": 0.9130221009254456, + "learning_rate": 9.994364956237133e-05, + "loss": 1.3609, + "step": 299 + }, + { + "epoch": 0.018227109787957958, + "grad_norm": 0.33087942004203796, + "learning_rate": 9.994319429339259e-05, + "loss": 1.3995, + "step": 300 + }, + { + "epoch": 0.018287866820584482, + "grad_norm": 0.5412013530731201, + "learning_rate": 9.994273719373643e-05, + "loss": 1.167, + "step": 301 + }, + { + "epoch": 0.01834862385321101, + "grad_norm": 0.3879370093345642, + "learning_rate": 9.99422782634196e-05, + "loss": 1.1636, + "step": 302 + }, + { + "epoch": 0.018409380885837535, + "grad_norm": 0.3803066909313202, + "learning_rate": 9.994181750245887e-05, + "loss": 1.2159, + "step": 303 + }, + { + "epoch": 0.018470137918464063, + "grad_norm": 0.8665661215782166, + "learning_rate": 9.99413549108712e-05, + "loss": 1.1714, + "step": 304 + }, + { + "epoch": 0.018530894951090587, + "grad_norm": 0.8854460716247559, + "learning_rate": 9.99408904886735e-05, + "loss": 1.3281, + "step": 305 + }, + { + "epoch": 0.018591651983717115, + "grad_norm": 0.683215320110321, + "learning_rate": 9.99404242358828e-05, + "loss": 1.1933, + "step": 306 + }, + { + "epoch": 0.018652409016343643, + "grad_norm": 0.34080770611763, + "learning_rate": 9.993995615251622e-05, + "loss": 1.1587, + "step": 307 + }, + { + "epoch": 0.018713166048970168, + "grad_norm": 1.041289210319519, + "learning_rate": 9.993948623859086e-05, + "loss": 1.2302, + "step": 308 + }, + { + "epoch": 0.018773923081596696, + "grad_norm": 1.4010217189788818, + "learning_rate": 9.993901449412399e-05, + "loss": 1.3544, + "step": 309 + }, + { + "epoch": 0.01883468011422322, + "grad_norm": 0.6498396992683411, + "learning_rate": 9.993854091913291e-05, + "loss": 1.153, + "step": 310 + }, + { + "epoch": 0.01889543714684975, + "grad_norm": 0.33952218294143677, + "learning_rate": 9.993806551363494e-05, + "loss": 1.1899, + "step": 311 + }, + { + "epoch": 0.018956194179476273, + "grad_norm": 1.0371161699295044, + "learning_rate": 9.993758827764752e-05, + "loss": 1.4419, + "step": 312 + }, + { + "epoch": 0.0190169512121028, + "grad_norm": 0.31514158844947815, + "learning_rate": 9.993710921118816e-05, + "loss": 1.3453, + "step": 313 + }, + { + "epoch": 0.019077708244729326, + "grad_norm": 0.5691254734992981, + "learning_rate": 9.993662831427442e-05, + "loss": 1.1664, + "step": 314 + }, + { + "epoch": 0.019138465277355854, + "grad_norm": 0.7804281711578369, + "learning_rate": 9.99361455869239e-05, + "loss": 1.3042, + "step": 315 + }, + { + "epoch": 0.019199222309982382, + "grad_norm": 0.6667227149009705, + "learning_rate": 9.993566102915432e-05, + "loss": 1.1531, + "step": 316 + }, + { + "epoch": 0.019259979342608907, + "grad_norm": 1.1866753101348877, + "learning_rate": 9.993517464098344e-05, + "loss": 1.2688, + "step": 317 + }, + { + "epoch": 0.019320736375235435, + "grad_norm": 0.4700160026550293, + "learning_rate": 9.993468642242905e-05, + "loss": 1.1748, + "step": 318 + }, + { + "epoch": 0.01938149340786196, + "grad_norm": 0.5273417830467224, + "learning_rate": 9.99341963735091e-05, + "loss": 1.1609, + "step": 319 + }, + { + "epoch": 0.019442250440488487, + "grad_norm": 0.5841180086135864, + "learning_rate": 9.993370449424153e-05, + "loss": 1.2776, + "step": 320 + }, + { + "epoch": 0.01950300747311501, + "grad_norm": 0.8693719506263733, + "learning_rate": 9.993321078464437e-05, + "loss": 1.4404, + "step": 321 + }, + { + "epoch": 0.01956376450574154, + "grad_norm": 0.8929149508476257, + "learning_rate": 9.99327152447357e-05, + "loss": 1.1516, + "step": 322 + }, + { + "epoch": 0.019624521538368068, + "grad_norm": 0.5292776823043823, + "learning_rate": 9.993221787453373e-05, + "loss": 1.208, + "step": 323 + }, + { + "epoch": 0.019685278570994592, + "grad_norm": 0.6710240840911865, + "learning_rate": 9.993171867405664e-05, + "loss": 1.1868, + "step": 324 + }, + { + "epoch": 0.01974603560362112, + "grad_norm": 0.3856630325317383, + "learning_rate": 9.993121764332276e-05, + "loss": 1.1803, + "step": 325 + }, + { + "epoch": 0.019806792636247645, + "grad_norm": 0.444989413022995, + "learning_rate": 9.993071478235044e-05, + "loss": 1.2779, + "step": 326 + }, + { + "epoch": 0.019867549668874173, + "grad_norm": 0.40760594606399536, + "learning_rate": 9.993021009115812e-05, + "loss": 1.1767, + "step": 327 + }, + { + "epoch": 0.019928306701500698, + "grad_norm": 0.32412195205688477, + "learning_rate": 9.99297035697643e-05, + "loss": 1.1773, + "step": 328 + }, + { + "epoch": 0.019989063734127226, + "grad_norm": 0.33736076951026917, + "learning_rate": 9.992919521818756e-05, + "loss": 1.2875, + "step": 329 + }, + { + "epoch": 0.02004982076675375, + "grad_norm": 0.8887059688568115, + "learning_rate": 9.99286850364465e-05, + "loss": 1.1922, + "step": 330 + }, + { + "epoch": 0.020110577799380278, + "grad_norm": 0.2582882344722748, + "learning_rate": 9.992817302455983e-05, + "loss": 1.1738, + "step": 331 + }, + { + "epoch": 0.020171334832006806, + "grad_norm": 0.9457483291625977, + "learning_rate": 9.992765918254634e-05, + "loss": 1.1337, + "step": 332 + }, + { + "epoch": 0.02023209186463333, + "grad_norm": 0.7865509986877441, + "learning_rate": 9.992714351042485e-05, + "loss": 1.4879, + "step": 333 + }, + { + "epoch": 0.02029284889725986, + "grad_norm": 0.8098940849304199, + "learning_rate": 9.992662600821427e-05, + "loss": 1.2218, + "step": 334 + }, + { + "epoch": 0.020353605929886383, + "grad_norm": 0.3568423390388489, + "learning_rate": 9.992610667593355e-05, + "loss": 1.2915, + "step": 335 + }, + { + "epoch": 0.02041436296251291, + "grad_norm": 0.4988231956958771, + "learning_rate": 9.992558551360174e-05, + "loss": 1.2444, + "step": 336 + }, + { + "epoch": 0.020475119995139436, + "grad_norm": 0.5876823663711548, + "learning_rate": 9.992506252123796e-05, + "loss": 1.1641, + "step": 337 + }, + { + "epoch": 0.020535877027765964, + "grad_norm": 0.49272915720939636, + "learning_rate": 9.992453769886135e-05, + "loss": 1.1385, + "step": 338 + }, + { + "epoch": 0.02059663406039249, + "grad_norm": 0.5669564008712769, + "learning_rate": 9.992401104649116e-05, + "loss": 1.2466, + "step": 339 + }, + { + "epoch": 0.020657391093019017, + "grad_norm": 0.3532738983631134, + "learning_rate": 9.99234825641467e-05, + "loss": 1.3445, + "step": 340 + }, + { + "epoch": 0.020718148125645545, + "grad_norm": 0.3957774341106415, + "learning_rate": 9.992295225184733e-05, + "loss": 1.2417, + "step": 341 + }, + { + "epoch": 0.02077890515827207, + "grad_norm": 0.353468656539917, + "learning_rate": 9.99224201096125e-05, + "loss": 1.2773, + "step": 342 + }, + { + "epoch": 0.020839662190898597, + "grad_norm": 0.5320485234260559, + "learning_rate": 9.99218861374617e-05, + "loss": 1.282, + "step": 343 + }, + { + "epoch": 0.020900419223525122, + "grad_norm": 3.675830125808716, + "learning_rate": 9.992135033541454e-05, + "loss": 1.3127, + "step": 344 + }, + { + "epoch": 0.02096117625615165, + "grad_norm": 0.6867219805717468, + "learning_rate": 9.992081270349061e-05, + "loss": 1.1465, + "step": 345 + }, + { + "epoch": 0.021021933288778175, + "grad_norm": 0.39964741468429565, + "learning_rate": 9.992027324170965e-05, + "loss": 1.2446, + "step": 346 + }, + { + "epoch": 0.021082690321404703, + "grad_norm": 0.27902913093566895, + "learning_rate": 9.991973195009143e-05, + "loss": 1.1399, + "step": 347 + }, + { + "epoch": 0.02114344735403123, + "grad_norm": 0.8664928674697876, + "learning_rate": 9.991918882865576e-05, + "loss": 1.231, + "step": 348 + }, + { + "epoch": 0.021204204386657755, + "grad_norm": 5.707220077514648, + "learning_rate": 9.99186438774226e-05, + "loss": 1.1733, + "step": 349 + }, + { + "epoch": 0.021264961419284283, + "grad_norm": 0.44804102182388306, + "learning_rate": 9.991809709641189e-05, + "loss": 1.1569, + "step": 350 + }, + { + "epoch": 0.021325718451910808, + "grad_norm": 0.5047160983085632, + "learning_rate": 9.99175484856437e-05, + "loss": 1.1818, + "step": 351 + }, + { + "epoch": 0.021386475484537336, + "grad_norm": 0.6860536336898804, + "learning_rate": 9.99169980451381e-05, + "loss": 1.2578, + "step": 352 + }, + { + "epoch": 0.02144723251716386, + "grad_norm": 0.49656060338020325, + "learning_rate": 9.99164457749153e-05, + "loss": 1.2773, + "step": 353 + }, + { + "epoch": 0.02150798954979039, + "grad_norm": 0.48132970929145813, + "learning_rate": 9.991589167499553e-05, + "loss": 1.2462, + "step": 354 + }, + { + "epoch": 0.021568746582416913, + "grad_norm": 0.2523705065250397, + "learning_rate": 9.991533574539909e-05, + "loss": 1.125, + "step": 355 + }, + { + "epoch": 0.02162950361504344, + "grad_norm": 0.2974836230278015, + "learning_rate": 9.991477798614638e-05, + "loss": 1.1729, + "step": 356 + }, + { + "epoch": 0.02169026064766997, + "grad_norm": 0.2903299331665039, + "learning_rate": 9.991421839725783e-05, + "loss": 1.2566, + "step": 357 + }, + { + "epoch": 0.021751017680296494, + "grad_norm": 0.4438534080982208, + "learning_rate": 9.991365697875397e-05, + "loss": 1.1573, + "step": 358 + }, + { + "epoch": 0.021811774712923022, + "grad_norm": 1.53419828414917, + "learning_rate": 9.991309373065533e-05, + "loss": 1.1287, + "step": 359 + }, + { + "epoch": 0.021872531745549546, + "grad_norm": 0.31573331356048584, + "learning_rate": 9.991252865298262e-05, + "loss": 1.2379, + "step": 360 + }, + { + "epoch": 0.021933288778176074, + "grad_norm": 0.6188585162162781, + "learning_rate": 9.991196174575653e-05, + "loss": 1.3488, + "step": 361 + }, + { + "epoch": 0.0219940458108026, + "grad_norm": 0.3840904235839844, + "learning_rate": 9.991139300899784e-05, + "loss": 1.4367, + "step": 362 + }, + { + "epoch": 0.022054802843429127, + "grad_norm": 0.24351735413074493, + "learning_rate": 9.991082244272736e-05, + "loss": 1.2548, + "step": 363 + }, + { + "epoch": 0.022115559876055655, + "grad_norm": 0.30272170901298523, + "learning_rate": 9.991025004696605e-05, + "loss": 1.1426, + "step": 364 + }, + { + "epoch": 0.02217631690868218, + "grad_norm": 0.4914633631706238, + "learning_rate": 9.990967582173489e-05, + "loss": 1.2537, + "step": 365 + }, + { + "epoch": 0.022237073941308708, + "grad_norm": 0.43234190344810486, + "learning_rate": 9.99090997670549e-05, + "loss": 1.1673, + "step": 366 + }, + { + "epoch": 0.022297830973935232, + "grad_norm": 0.3534109592437744, + "learning_rate": 9.990852188294721e-05, + "loss": 1.1634, + "step": 367 + }, + { + "epoch": 0.02235858800656176, + "grad_norm": 0.5538597106933594, + "learning_rate": 9.990794216943302e-05, + "loss": 1.1448, + "step": 368 + }, + { + "epoch": 0.022419345039188285, + "grad_norm": 8.639116287231445, + "learning_rate": 9.990736062653354e-05, + "loss": 1.1871, + "step": 369 + }, + { + "epoch": 0.022480102071814813, + "grad_norm": 0.4889613389968872, + "learning_rate": 9.990677725427013e-05, + "loss": 1.2256, + "step": 370 + }, + { + "epoch": 0.022540859104441337, + "grad_norm": 0.5601900815963745, + "learning_rate": 9.990619205266414e-05, + "loss": 1.14, + "step": 371 + }, + { + "epoch": 0.022601616137067865, + "grad_norm": 0.3544400632381439, + "learning_rate": 9.990560502173704e-05, + "loss": 1.2737, + "step": 372 + }, + { + "epoch": 0.022662373169694393, + "grad_norm": 0.3993493616580963, + "learning_rate": 9.990501616151034e-05, + "loss": 1.2036, + "step": 373 + }, + { + "epoch": 0.022723130202320918, + "grad_norm": 0.2903212010860443, + "learning_rate": 9.990442547200562e-05, + "loss": 1.1663, + "step": 374 + }, + { + "epoch": 0.022783887234947446, + "grad_norm": 0.4948521554470062, + "learning_rate": 9.990383295324454e-05, + "loss": 1.5035, + "step": 375 + }, + { + "epoch": 0.02284464426757397, + "grad_norm": 0.2620209753513336, + "learning_rate": 9.990323860524882e-05, + "loss": 1.1383, + "step": 376 + }, + { + "epoch": 0.0229054013002005, + "grad_norm": 0.7125815749168396, + "learning_rate": 9.990264242804026e-05, + "loss": 1.3713, + "step": 377 + }, + { + "epoch": 0.022966158332827023, + "grad_norm": 0.30556267499923706, + "learning_rate": 9.990204442164066e-05, + "loss": 1.2459, + "step": 378 + }, + { + "epoch": 0.02302691536545355, + "grad_norm": 15.613347053527832, + "learning_rate": 9.9901444586072e-05, + "loss": 1.2393, + "step": 379 + }, + { + "epoch": 0.02308767239808008, + "grad_norm": 0.6168348789215088, + "learning_rate": 9.990084292135624e-05, + "loss": 1.1376, + "step": 380 + }, + { + "epoch": 0.023148429430706604, + "grad_norm": 0.3690885603427887, + "learning_rate": 9.990023942751543e-05, + "loss": 1.3218, + "step": 381 + }, + { + "epoch": 0.023209186463333132, + "grad_norm": 0.49662336707115173, + "learning_rate": 9.989963410457171e-05, + "loss": 1.2804, + "step": 382 + }, + { + "epoch": 0.023269943495959657, + "grad_norm": 0.3410033583641052, + "learning_rate": 9.989902695254725e-05, + "loss": 1.2679, + "step": 383 + }, + { + "epoch": 0.023330700528586185, + "grad_norm": 0.3017045855522156, + "learning_rate": 9.989841797146431e-05, + "loss": 1.2171, + "step": 384 + }, + { + "epoch": 0.02339145756121271, + "grad_norm": 0.4700063169002533, + "learning_rate": 9.989780716134522e-05, + "loss": 1.2637, + "step": 385 + }, + { + "epoch": 0.023452214593839237, + "grad_norm": 18.73973846435547, + "learning_rate": 9.989719452221235e-05, + "loss": 1.2375, + "step": 386 + }, + { + "epoch": 0.023512971626465762, + "grad_norm": 0.9934202432632446, + "learning_rate": 9.989658005408818e-05, + "loss": 1.2251, + "step": 387 + }, + { + "epoch": 0.02357372865909229, + "grad_norm": 0.27044376730918884, + "learning_rate": 9.989596375699522e-05, + "loss": 1.1668, + "step": 388 + }, + { + "epoch": 0.023634485691718818, + "grad_norm": 0.8887792229652405, + "learning_rate": 9.989534563095607e-05, + "loss": 1.3922, + "step": 389 + }, + { + "epoch": 0.023695242724345342, + "grad_norm": 0.4002656042575836, + "learning_rate": 9.989472567599338e-05, + "loss": 1.2772, + "step": 390 + }, + { + "epoch": 0.02375599975697187, + "grad_norm": 0.42063838243484497, + "learning_rate": 9.989410389212987e-05, + "loss": 1.2392, + "step": 391 + }, + { + "epoch": 0.023816756789598395, + "grad_norm": 0.6064113974571228, + "learning_rate": 9.989348027938834e-05, + "loss": 1.1548, + "step": 392 + }, + { + "epoch": 0.023877513822224923, + "grad_norm": 0.21485337615013123, + "learning_rate": 9.989285483779164e-05, + "loss": 1.3615, + "step": 393 + }, + { + "epoch": 0.023938270854851448, + "grad_norm": 0.9040740728378296, + "learning_rate": 9.989222756736271e-05, + "loss": 1.127, + "step": 394 + }, + { + "epoch": 0.023999027887477976, + "grad_norm": 0.32483619451522827, + "learning_rate": 9.989159846812454e-05, + "loss": 1.2115, + "step": 395 + }, + { + "epoch": 0.024059784920104504, + "grad_norm": 0.5586081743240356, + "learning_rate": 9.989096754010018e-05, + "loss": 1.1597, + "step": 396 + }, + { + "epoch": 0.02412054195273103, + "grad_norm": 0.2920112907886505, + "learning_rate": 9.989033478331276e-05, + "loss": 1.2421, + "step": 397 + }, + { + "epoch": 0.024181298985357556, + "grad_norm": 0.48237845301628113, + "learning_rate": 9.988970019778547e-05, + "loss": 1.2306, + "step": 398 + }, + { + "epoch": 0.02424205601798408, + "grad_norm": 0.5133035778999329, + "learning_rate": 9.988906378354158e-05, + "loss": 1.171, + "step": 399 + }, + { + "epoch": 0.02430281305061061, + "grad_norm": 0.39558055996894836, + "learning_rate": 9.988842554060441e-05, + "loss": 1.1469, + "step": 400 + }, + { + "epoch": 0.024363570083237134, + "grad_norm": 0.44907623529434204, + "learning_rate": 9.988778546899738e-05, + "loss": 1.1255, + "step": 401 + }, + { + "epoch": 0.02442432711586366, + "grad_norm": 0.28002673387527466, + "learning_rate": 9.98871435687439e-05, + "loss": 1.1768, + "step": 402 + }, + { + "epoch": 0.024485084148490186, + "grad_norm": 0.3769404888153076, + "learning_rate": 9.988649983986755e-05, + "loss": 1.1825, + "step": 403 + }, + { + "epoch": 0.024545841181116714, + "grad_norm": 0.300797700881958, + "learning_rate": 9.98858542823919e-05, + "loss": 1.2668, + "step": 404 + }, + { + "epoch": 0.024606598213743242, + "grad_norm": 0.29228487610816956, + "learning_rate": 9.98852068963406e-05, + "loss": 1.1443, + "step": 405 + }, + { + "epoch": 0.024667355246369767, + "grad_norm": 0.38730016350746155, + "learning_rate": 9.988455768173741e-05, + "loss": 1.2278, + "step": 406 + }, + { + "epoch": 0.024728112278996295, + "grad_norm": 0.2510862648487091, + "learning_rate": 9.988390663860613e-05, + "loss": 1.2796, + "step": 407 + }, + { + "epoch": 0.02478886931162282, + "grad_norm": 0.34693482518196106, + "learning_rate": 9.988325376697059e-05, + "loss": 1.1435, + "step": 408 + }, + { + "epoch": 0.024849626344249347, + "grad_norm": 0.29950496554374695, + "learning_rate": 9.988259906685475e-05, + "loss": 1.1527, + "step": 409 + }, + { + "epoch": 0.024910383376875872, + "grad_norm": 0.24924340844154358, + "learning_rate": 9.98819425382826e-05, + "loss": 1.2589, + "step": 410 + }, + { + "epoch": 0.0249711404095024, + "grad_norm": 4.91303014755249, + "learning_rate": 9.988128418127822e-05, + "loss": 1.1588, + "step": 411 + }, + { + "epoch": 0.025031897442128928, + "grad_norm": 0.6393350958824158, + "learning_rate": 9.98806239958657e-05, + "loss": 1.2209, + "step": 412 + }, + { + "epoch": 0.025092654474755453, + "grad_norm": 0.3684506416320801, + "learning_rate": 9.987996198206928e-05, + "loss": 1.101, + "step": 413 + }, + { + "epoch": 0.02515341150738198, + "grad_norm": 1.5797674655914307, + "learning_rate": 9.987929813991321e-05, + "loss": 1.2611, + "step": 414 + }, + { + "epoch": 0.025214168540008505, + "grad_norm": 0.3158626854419708, + "learning_rate": 9.987863246942181e-05, + "loss": 1.1646, + "step": 415 + }, + { + "epoch": 0.025274925572635033, + "grad_norm": 0.2776279151439667, + "learning_rate": 9.987796497061952e-05, + "loss": 1.2361, + "step": 416 + }, + { + "epoch": 0.025335682605261558, + "grad_norm": 0.35253334045410156, + "learning_rate": 9.987729564353078e-05, + "loss": 1.3057, + "step": 417 + }, + { + "epoch": 0.025396439637888086, + "grad_norm": 0.43254759907722473, + "learning_rate": 9.987662448818011e-05, + "loss": 1.1399, + "step": 418 + }, + { + "epoch": 0.02545719667051461, + "grad_norm": 0.2638023793697357, + "learning_rate": 9.987595150459215e-05, + "loss": 1.235, + "step": 419 + }, + { + "epoch": 0.02551795370314114, + "grad_norm": 0.6823543310165405, + "learning_rate": 9.987527669279152e-05, + "loss": 1.1606, + "step": 420 + }, + { + "epoch": 0.025578710735767667, + "grad_norm": 0.2420654594898224, + "learning_rate": 9.9874600052803e-05, + "loss": 1.2463, + "step": 421 + }, + { + "epoch": 0.02563946776839419, + "grad_norm": 0.2917928993701935, + "learning_rate": 9.987392158465138e-05, + "loss": 1.1461, + "step": 422 + }, + { + "epoch": 0.02570022480102072, + "grad_norm": 0.3107789158821106, + "learning_rate": 9.987324128836151e-05, + "loss": 1.1695, + "step": 423 + }, + { + "epoch": 0.025760981833647244, + "grad_norm": 0.4063679575920105, + "learning_rate": 9.987255916395835e-05, + "loss": 1.2358, + "step": 424 + }, + { + "epoch": 0.025821738866273772, + "grad_norm": 0.6911201477050781, + "learning_rate": 9.987187521146688e-05, + "loss": 1.2368, + "step": 425 + }, + { + "epoch": 0.025882495898900296, + "grad_norm": 0.21983657777309418, + "learning_rate": 9.98711894309122e-05, + "loss": 1.1325, + "step": 426 + }, + { + "epoch": 0.025943252931526824, + "grad_norm": 0.22269152104854584, + "learning_rate": 9.987050182231943e-05, + "loss": 1.1578, + "step": 427 + }, + { + "epoch": 0.026004009964153352, + "grad_norm": 0.35398516058921814, + "learning_rate": 9.986981238571376e-05, + "loss": 1.2242, + "step": 428 + }, + { + "epoch": 0.026064766996779877, + "grad_norm": 0.3698769211769104, + "learning_rate": 9.986912112112049e-05, + "loss": 1.1194, + "step": 429 + }, + { + "epoch": 0.026125524029406405, + "grad_norm": 0.2556109130382538, + "learning_rate": 9.986842802856494e-05, + "loss": 1.185, + "step": 430 + }, + { + "epoch": 0.02618628106203293, + "grad_norm": 6.222453594207764, + "learning_rate": 9.986773310807254e-05, + "loss": 1.2771, + "step": 431 + }, + { + "epoch": 0.026247038094659458, + "grad_norm": 0.45141589641571045, + "learning_rate": 9.986703635966872e-05, + "loss": 1.1841, + "step": 432 + }, + { + "epoch": 0.026307795127285982, + "grad_norm": 0.3502272665500641, + "learning_rate": 9.986633778337905e-05, + "loss": 1.2674, + "step": 433 + }, + { + "epoch": 0.02636855215991251, + "grad_norm": 0.2743486762046814, + "learning_rate": 9.986563737922914e-05, + "loss": 1.1431, + "step": 434 + }, + { + "epoch": 0.026429309192539035, + "grad_norm": 0.8279253244400024, + "learning_rate": 9.986493514724464e-05, + "loss": 1.2559, + "step": 435 + }, + { + "epoch": 0.026490066225165563, + "grad_norm": 1.013606309890747, + "learning_rate": 9.98642310874513e-05, + "loss": 1.2181, + "step": 436 + }, + { + "epoch": 0.02655082325779209, + "grad_norm": 0.744698166847229, + "learning_rate": 9.986352519987494e-05, + "loss": 1.2706, + "step": 437 + }, + { + "epoch": 0.026611580290418616, + "grad_norm": 0.34134721755981445, + "learning_rate": 9.986281748454143e-05, + "loss": 1.1662, + "step": 438 + }, + { + "epoch": 0.026672337323045144, + "grad_norm": 0.3625911772251129, + "learning_rate": 9.986210794147671e-05, + "loss": 1.172, + "step": 439 + }, + { + "epoch": 0.026733094355671668, + "grad_norm": 0.22426804900169373, + "learning_rate": 9.986139657070677e-05, + "loss": 1.1757, + "step": 440 + }, + { + "epoch": 0.026793851388298196, + "grad_norm": 0.40342411398887634, + "learning_rate": 9.98606833722577e-05, + "loss": 1.1583, + "step": 441 + }, + { + "epoch": 0.02685460842092472, + "grad_norm": 0.26724132895469666, + "learning_rate": 9.985996834615566e-05, + "loss": 1.1289, + "step": 442 + }, + { + "epoch": 0.02691536545355125, + "grad_norm": 0.31080037355422974, + "learning_rate": 9.985925149242683e-05, + "loss": 1.2452, + "step": 443 + }, + { + "epoch": 0.026976122486177777, + "grad_norm": 0.7767373323440552, + "learning_rate": 9.98585328110975e-05, + "loss": 1.3516, + "step": 444 + }, + { + "epoch": 0.0270368795188043, + "grad_norm": 0.27356621623039246, + "learning_rate": 9.985781230219402e-05, + "loss": 1.2305, + "step": 445 + }, + { + "epoch": 0.02709763655143083, + "grad_norm": 0.4051680564880371, + "learning_rate": 9.985708996574278e-05, + "loss": 1.304, + "step": 446 + }, + { + "epoch": 0.027158393584057354, + "grad_norm": 0.3881681561470032, + "learning_rate": 9.985636580177026e-05, + "loss": 1.4459, + "step": 447 + }, + { + "epoch": 0.027219150616683882, + "grad_norm": 0.32349786162376404, + "learning_rate": 9.985563981030304e-05, + "loss": 1.1459, + "step": 448 + }, + { + "epoch": 0.027279907649310407, + "grad_norm": 0.4432174265384674, + "learning_rate": 9.98549119913677e-05, + "loss": 1.2107, + "step": 449 + }, + { + "epoch": 0.027340664681936935, + "grad_norm": 0.314494252204895, + "learning_rate": 9.985418234499094e-05, + "loss": 1.1604, + "step": 450 + }, + { + "epoch": 0.02740142171456346, + "grad_norm": 0.3270628750324249, + "learning_rate": 9.985345087119946e-05, + "loss": 1.1441, + "step": 451 + }, + { + "epoch": 0.027462178747189987, + "grad_norm": 0.41795921325683594, + "learning_rate": 9.98527175700201e-05, + "loss": 1.2755, + "step": 452 + }, + { + "epoch": 0.027522935779816515, + "grad_norm": 0.36719265580177307, + "learning_rate": 9.985198244147978e-05, + "loss": 1.1922, + "step": 453 + }, + { + "epoch": 0.02758369281244304, + "grad_norm": 0.2682739198207855, + "learning_rate": 9.985124548560536e-05, + "loss": 1.1593, + "step": 454 + }, + { + "epoch": 0.027644449845069568, + "grad_norm": 0.7295563220977783, + "learning_rate": 9.985050670242393e-05, + "loss": 1.2708, + "step": 455 + }, + { + "epoch": 0.027705206877696092, + "grad_norm": 0.2742316722869873, + "learning_rate": 9.984976609196253e-05, + "loss": 1.2488, + "step": 456 + }, + { + "epoch": 0.02776596391032262, + "grad_norm": 0.5238128900527954, + "learning_rate": 9.984902365424833e-05, + "loss": 1.1487, + "step": 457 + }, + { + "epoch": 0.027826720942949145, + "grad_norm": 0.2640613317489624, + "learning_rate": 9.984827938930852e-05, + "loss": 1.2264, + "step": 458 + }, + { + "epoch": 0.027887477975575673, + "grad_norm": 0.3218802213668823, + "learning_rate": 9.98475332971704e-05, + "loss": 1.2221, + "step": 459 + }, + { + "epoch": 0.027948235008202198, + "grad_norm": 0.31852445006370544, + "learning_rate": 9.98467853778613e-05, + "loss": 1.1192, + "step": 460 + }, + { + "epoch": 0.028008992040828726, + "grad_norm": 1.1100140810012817, + "learning_rate": 9.984603563140866e-05, + "loss": 1.1632, + "step": 461 + }, + { + "epoch": 0.028069749073455254, + "grad_norm": 0.46260789036750793, + "learning_rate": 9.984528405783994e-05, + "loss": 1.2879, + "step": 462 + }, + { + "epoch": 0.02813050610608178, + "grad_norm": 0.28880906105041504, + "learning_rate": 9.984453065718271e-05, + "loss": 1.1488, + "step": 463 + }, + { + "epoch": 0.028191263138708306, + "grad_norm": 0.408460795879364, + "learning_rate": 9.984377542946458e-05, + "loss": 1.3038, + "step": 464 + }, + { + "epoch": 0.02825202017133483, + "grad_norm": 12.436212539672852, + "learning_rate": 9.98430183747132e-05, + "loss": 1.182, + "step": 465 + }, + { + "epoch": 0.02831277720396136, + "grad_norm": 0.3148921728134155, + "learning_rate": 9.984225949295636e-05, + "loss": 1.1628, + "step": 466 + }, + { + "epoch": 0.028373534236587884, + "grad_norm": 0.32460933923721313, + "learning_rate": 9.984149878422186e-05, + "loss": 1.1354, + "step": 467 + }, + { + "epoch": 0.02843429126921441, + "grad_norm": 0.27630913257598877, + "learning_rate": 9.98407362485376e-05, + "loss": 1.1832, + "step": 468 + }, + { + "epoch": 0.02849504830184094, + "grad_norm": 0.278171569108963, + "learning_rate": 9.98399718859315e-05, + "loss": 1.2161, + "step": 469 + }, + { + "epoch": 0.028555805334467464, + "grad_norm": 0.260118693113327, + "learning_rate": 9.983920569643161e-05, + "loss": 1.1749, + "step": 470 + }, + { + "epoch": 0.028616562367093992, + "grad_norm": 1.030730962753296, + "learning_rate": 9.9838437680066e-05, + "loss": 1.2154, + "step": 471 + }, + { + "epoch": 0.028677319399720517, + "grad_norm": 0.3792092204093933, + "learning_rate": 9.983766783686284e-05, + "loss": 1.2945, + "step": 472 + }, + { + "epoch": 0.028738076432347045, + "grad_norm": 0.46523478627204895, + "learning_rate": 9.983689616685031e-05, + "loss": 1.3603, + "step": 473 + }, + { + "epoch": 0.02879883346497357, + "grad_norm": 0.5205716490745544, + "learning_rate": 9.983612267005671e-05, + "loss": 1.2439, + "step": 474 + }, + { + "epoch": 0.028859590497600097, + "grad_norm": 0.25445348024368286, + "learning_rate": 9.983534734651042e-05, + "loss": 1.1737, + "step": 475 + }, + { + "epoch": 0.028920347530226622, + "grad_norm": 0.4947931468486786, + "learning_rate": 9.983457019623983e-05, + "loss": 1.1644, + "step": 476 + }, + { + "epoch": 0.02898110456285315, + "grad_norm": 0.26506200432777405, + "learning_rate": 9.983379121927345e-05, + "loss": 1.1825, + "step": 477 + }, + { + "epoch": 0.029041861595479678, + "grad_norm": 0.3688337206840515, + "learning_rate": 9.98330104156398e-05, + "loss": 1.2443, + "step": 478 + }, + { + "epoch": 0.029102618628106203, + "grad_norm": 0.2743681073188782, + "learning_rate": 9.983222778536754e-05, + "loss": 1.2086, + "step": 479 + }, + { + "epoch": 0.02916337566073273, + "grad_norm": 8.567276954650879, + "learning_rate": 9.983144332848533e-05, + "loss": 1.288, + "step": 480 + }, + { + "epoch": 0.029224132693359255, + "grad_norm": 0.4259561598300934, + "learning_rate": 9.983065704502193e-05, + "loss": 1.2686, + "step": 481 + }, + { + "epoch": 0.029284889725985783, + "grad_norm": 0.309903085231781, + "learning_rate": 9.982986893500617e-05, + "loss": 1.1829, + "step": 482 + }, + { + "epoch": 0.029345646758612308, + "grad_norm": 8.346363067626953, + "learning_rate": 9.982907899846692e-05, + "loss": 1.1195, + "step": 483 + }, + { + "epoch": 0.029406403791238836, + "grad_norm": 0.487554669380188, + "learning_rate": 9.982828723543316e-05, + "loss": 1.3601, + "step": 484 + }, + { + "epoch": 0.029467160823865364, + "grad_norm": 0.3033584654331207, + "learning_rate": 9.98274936459339e-05, + "loss": 1.2768, + "step": 485 + }, + { + "epoch": 0.02952791785649189, + "grad_norm": 0.805773913860321, + "learning_rate": 9.982669822999822e-05, + "loss": 1.1538, + "step": 486 + }, + { + "epoch": 0.029588674889118417, + "grad_norm": 0.3356128931045532, + "learning_rate": 9.982590098765528e-05, + "loss": 1.3401, + "step": 487 + }, + { + "epoch": 0.02964943192174494, + "grad_norm": 1.0753117799758911, + "learning_rate": 9.982510191893433e-05, + "loss": 1.229, + "step": 488 + }, + { + "epoch": 0.02971018895437147, + "grad_norm": 0.3465622067451477, + "learning_rate": 9.982430102386462e-05, + "loss": 1.1872, + "step": 489 + }, + { + "epoch": 0.029770945986997994, + "grad_norm": 0.9397584199905396, + "learning_rate": 9.982349830247553e-05, + "loss": 1.2456, + "step": 490 + }, + { + "epoch": 0.029831703019624522, + "grad_norm": 0.3869839012622833, + "learning_rate": 9.982269375479648e-05, + "loss": 1.1547, + "step": 491 + }, + { + "epoch": 0.029892460052251046, + "grad_norm": 0.8682321906089783, + "learning_rate": 9.982188738085697e-05, + "loss": 1.291, + "step": 492 + }, + { + "epoch": 0.029953217084877574, + "grad_norm": 0.46015435457229614, + "learning_rate": 9.982107918068652e-05, + "loss": 1.3159, + "step": 493 + }, + { + "epoch": 0.030013974117504102, + "grad_norm": 0.4634031355381012, + "learning_rate": 9.98202691543148e-05, + "loss": 1.1673, + "step": 494 + }, + { + "epoch": 0.030074731150130627, + "grad_norm": 0.4604499936103821, + "learning_rate": 9.98194573017715e-05, + "loss": 1.1372, + "step": 495 + }, + { + "epoch": 0.030135488182757155, + "grad_norm": 0.25906088948249817, + "learning_rate": 9.981864362308634e-05, + "loss": 1.1178, + "step": 496 + }, + { + "epoch": 0.03019624521538368, + "grad_norm": 0.30098816752433777, + "learning_rate": 9.981782811828918e-05, + "loss": 1.1726, + "step": 497 + }, + { + "epoch": 0.030257002248010208, + "grad_norm": 0.24130785465240479, + "learning_rate": 9.981701078740988e-05, + "loss": 1.1796, + "step": 498 + }, + { + "epoch": 0.030317759280636732, + "grad_norm": 0.5331498384475708, + "learning_rate": 9.981619163047843e-05, + "loss": 1.2687, + "step": 499 + }, + { + "epoch": 0.03037851631326326, + "grad_norm": 0.3563857674598694, + "learning_rate": 9.981537064752486e-05, + "loss": 1.3309, + "step": 500 + }, + { + "epoch": 0.03043927334588979, + "grad_norm": 0.2574976086616516, + "learning_rate": 9.981454783857925e-05, + "loss": 1.2464, + "step": 501 + }, + { + "epoch": 0.030500030378516313, + "grad_norm": 0.2564973831176758, + "learning_rate": 9.981372320367175e-05, + "loss": 1.1492, + "step": 502 + }, + { + "epoch": 0.03056078741114284, + "grad_norm": 0.40071558952331543, + "learning_rate": 9.98128967428326e-05, + "loss": 1.1865, + "step": 503 + }, + { + "epoch": 0.030621544443769366, + "grad_norm": 0.5712208151817322, + "learning_rate": 9.981206845609209e-05, + "loss": 1.4144, + "step": 504 + }, + { + "epoch": 0.030682301476395894, + "grad_norm": 0.3333223760128021, + "learning_rate": 9.981123834348058e-05, + "loss": 1.2817, + "step": 505 + }, + { + "epoch": 0.030743058509022418, + "grad_norm": 0.48780977725982666, + "learning_rate": 9.98104064050285e-05, + "loss": 1.1402, + "step": 506 + }, + { + "epoch": 0.030803815541648946, + "grad_norm": 4.371847152709961, + "learning_rate": 9.980957264076635e-05, + "loss": 1.2464, + "step": 507 + }, + { + "epoch": 0.03086457257427547, + "grad_norm": 0.3993298411369324, + "learning_rate": 9.980873705072468e-05, + "loss": 1.2447, + "step": 508 + }, + { + "epoch": 0.030925329606902, + "grad_norm": 0.5067611932754517, + "learning_rate": 9.980789963493415e-05, + "loss": 1.2781, + "step": 509 + }, + { + "epoch": 0.030986086639528527, + "grad_norm": 0.7356559634208679, + "learning_rate": 9.98070603934254e-05, + "loss": 1.1915, + "step": 510 + }, + { + "epoch": 0.03104684367215505, + "grad_norm": 0.3917902410030365, + "learning_rate": 9.980621932622924e-05, + "loss": 1.1211, + "step": 511 + }, + { + "epoch": 0.03110760070478158, + "grad_norm": 0.3990289270877838, + "learning_rate": 9.980537643337648e-05, + "loss": 1.164, + "step": 512 + }, + { + "epoch": 0.031168357737408104, + "grad_norm": 0.372371643781662, + "learning_rate": 9.980453171489802e-05, + "loss": 1.1983, + "step": 513 + }, + { + "epoch": 0.031229114770034632, + "grad_norm": 0.3391153812408447, + "learning_rate": 9.980368517082481e-05, + "loss": 1.1633, + "step": 514 + }, + { + "epoch": 0.03128987180266116, + "grad_norm": 0.304869145154953, + "learning_rate": 9.980283680118792e-05, + "loss": 1.4087, + "step": 515 + }, + { + "epoch": 0.031350628835287685, + "grad_norm": 0.4680231809616089, + "learning_rate": 9.98019866060184e-05, + "loss": 1.2693, + "step": 516 + }, + { + "epoch": 0.03141138586791421, + "grad_norm": 0.42668598890304565, + "learning_rate": 9.980113458534744e-05, + "loss": 1.1394, + "step": 517 + }, + { + "epoch": 0.03147214290054074, + "grad_norm": 0.4277767539024353, + "learning_rate": 9.980028073920627e-05, + "loss": 1.1594, + "step": 518 + }, + { + "epoch": 0.03153289993316726, + "grad_norm": 3.0939595699310303, + "learning_rate": 9.979942506762617e-05, + "loss": 1.324, + "step": 519 + }, + { + "epoch": 0.03159365696579379, + "grad_norm": 0.3940166234970093, + "learning_rate": 9.979856757063853e-05, + "loss": 1.1717, + "step": 520 + }, + { + "epoch": 0.03165441399842032, + "grad_norm": 0.6198388338088989, + "learning_rate": 9.979770824827477e-05, + "loss": 1.1376, + "step": 521 + }, + { + "epoch": 0.031715171031046846, + "grad_norm": 0.48856398463249207, + "learning_rate": 9.979684710056639e-05, + "loss": 1.1905, + "step": 522 + }, + { + "epoch": 0.03177592806367337, + "grad_norm": 0.7191376090049744, + "learning_rate": 9.979598412754493e-05, + "loss": 1.1719, + "step": 523 + }, + { + "epoch": 0.031836685096299895, + "grad_norm": 0.33660340309143066, + "learning_rate": 9.979511932924209e-05, + "loss": 1.235, + "step": 524 + }, + { + "epoch": 0.03189744212892642, + "grad_norm": 0.2960328757762909, + "learning_rate": 9.979425270568949e-05, + "loss": 1.1624, + "step": 525 + }, + { + "epoch": 0.03195819916155295, + "grad_norm": 0.5520493984222412, + "learning_rate": 9.979338425691894e-05, + "loss": 1.2754, + "step": 526 + }, + { + "epoch": 0.03201895619417948, + "grad_norm": 3.723741054534912, + "learning_rate": 9.979251398296228e-05, + "loss": 1.2858, + "step": 527 + }, + { + "epoch": 0.032079713226806, + "grad_norm": 9.237574577331543, + "learning_rate": 9.979164188385138e-05, + "loss": 1.1406, + "step": 528 + }, + { + "epoch": 0.03214047025943253, + "grad_norm": 1.2526761293411255, + "learning_rate": 9.979076795961821e-05, + "loss": 1.1874, + "step": 529 + }, + { + "epoch": 0.032201227292059056, + "grad_norm": 0.5664454102516174, + "learning_rate": 9.978989221029484e-05, + "loss": 1.1463, + "step": 530 + }, + { + "epoch": 0.032261984324685584, + "grad_norm": 1.0137161016464233, + "learning_rate": 9.978901463591335e-05, + "loss": 1.1405, + "step": 531 + }, + { + "epoch": 0.032322741357312106, + "grad_norm": 0.299500972032547, + "learning_rate": 9.978813523650589e-05, + "loss": 1.154, + "step": 532 + }, + { + "epoch": 0.032383498389938634, + "grad_norm": 0.7356293797492981, + "learning_rate": 9.978725401210471e-05, + "loss": 1.1776, + "step": 533 + }, + { + "epoch": 0.03244425542256516, + "grad_norm": 0.33569473028182983, + "learning_rate": 9.97863709627421e-05, + "loss": 1.2617, + "step": 534 + }, + { + "epoch": 0.03250501245519169, + "grad_norm": 0.5644264817237854, + "learning_rate": 9.978548608845046e-05, + "loss": 1.1513, + "step": 535 + }, + { + "epoch": 0.03256576948781822, + "grad_norm": 2.1869406700134277, + "learning_rate": 9.97845993892622e-05, + "loss": 1.2151, + "step": 536 + }, + { + "epoch": 0.03262652652044474, + "grad_norm": 0.40671515464782715, + "learning_rate": 9.978371086520983e-05, + "loss": 1.1732, + "step": 537 + }, + { + "epoch": 0.03268728355307127, + "grad_norm": 0.6579546332359314, + "learning_rate": 9.97828205163259e-05, + "loss": 1.172, + "step": 538 + }, + { + "epoch": 0.032748040585697795, + "grad_norm": 0.5245298147201538, + "learning_rate": 9.978192834264306e-05, + "loss": 1.1672, + "step": 539 + }, + { + "epoch": 0.03280879761832432, + "grad_norm": 0.4068654179573059, + "learning_rate": 9.978103434419403e-05, + "loss": 1.1934, + "step": 540 + }, + { + "epoch": 0.03286955465095085, + "grad_norm": 0.2756313979625702, + "learning_rate": 9.978013852101155e-05, + "loss": 1.1583, + "step": 541 + }, + { + "epoch": 0.03293031168357737, + "grad_norm": 1.071942925453186, + "learning_rate": 9.977924087312849e-05, + "loss": 1.1911, + "step": 542 + }, + { + "epoch": 0.0329910687162039, + "grad_norm": 0.39561423659324646, + "learning_rate": 9.977834140057771e-05, + "loss": 1.2859, + "step": 543 + }, + { + "epoch": 0.03305182574883043, + "grad_norm": 0.34798678755760193, + "learning_rate": 9.977744010339222e-05, + "loss": 1.2883, + "step": 544 + }, + { + "epoch": 0.033112582781456956, + "grad_norm": 0.42213043570518494, + "learning_rate": 9.977653698160503e-05, + "loss": 1.1116, + "step": 545 + }, + { + "epoch": 0.03317333981408348, + "grad_norm": 0.40069815516471863, + "learning_rate": 9.977563203524926e-05, + "loss": 1.353, + "step": 546 + }, + { + "epoch": 0.033234096846710005, + "grad_norm": 4.614850044250488, + "learning_rate": 9.977472526435809e-05, + "loss": 1.2518, + "step": 547 + }, + { + "epoch": 0.03329485387933653, + "grad_norm": 0.7935022711753845, + "learning_rate": 9.977381666896472e-05, + "loss": 1.1575, + "step": 548 + }, + { + "epoch": 0.03335561091196306, + "grad_norm": 0.44215452671051025, + "learning_rate": 9.977290624910249e-05, + "loss": 1.1742, + "step": 549 + }, + { + "epoch": 0.03341636794458959, + "grad_norm": 1.6043121814727783, + "learning_rate": 9.977199400480477e-05, + "loss": 1.1584, + "step": 550 + }, + { + "epoch": 0.03347712497721611, + "grad_norm": 1.0440819263458252, + "learning_rate": 9.977107993610497e-05, + "loss": 1.2099, + "step": 551 + }, + { + "epoch": 0.03353788200984264, + "grad_norm": 0.36986932158470154, + "learning_rate": 9.977016404303663e-05, + "loss": 1.2515, + "step": 552 + }, + { + "epoch": 0.03359863904246917, + "grad_norm": 1.6186177730560303, + "learning_rate": 9.97692463256333e-05, + "loss": 1.1697, + "step": 553 + }, + { + "epoch": 0.033659396075095695, + "grad_norm": 1.04796302318573, + "learning_rate": 9.976832678392862e-05, + "loss": 1.3574, + "step": 554 + }, + { + "epoch": 0.033720153107722216, + "grad_norm": 0.7352060079574585, + "learning_rate": 9.976740541795632e-05, + "loss": 1.2869, + "step": 555 + }, + { + "epoch": 0.033780910140348744, + "grad_norm": 1.3824442625045776, + "learning_rate": 9.976648222775013e-05, + "loss": 1.1284, + "step": 556 + }, + { + "epoch": 0.03384166717297527, + "grad_norm": 0.525844931602478, + "learning_rate": 9.976555721334392e-05, + "loss": 1.2068, + "step": 557 + }, + { + "epoch": 0.0339024242056018, + "grad_norm": 0.8211742639541626, + "learning_rate": 9.976463037477162e-05, + "loss": 1.1974, + "step": 558 + }, + { + "epoch": 0.03396318123822833, + "grad_norm": 0.7520949840545654, + "learning_rate": 9.976370171206714e-05, + "loss": 1.2751, + "step": 559 + }, + { + "epoch": 0.03402393827085485, + "grad_norm": 0.3710823655128479, + "learning_rate": 9.976277122526457e-05, + "loss": 1.2268, + "step": 560 + }, + { + "epoch": 0.03408469530348138, + "grad_norm": 0.8229123950004578, + "learning_rate": 9.976183891439801e-05, + "loss": 1.3209, + "step": 561 + }, + { + "epoch": 0.034145452336107905, + "grad_norm": 0.3961959183216095, + "learning_rate": 9.976090477950161e-05, + "loss": 1.195, + "step": 562 + }, + { + "epoch": 0.03420620936873443, + "grad_norm": 0.6174319982528687, + "learning_rate": 9.975996882060963e-05, + "loss": 1.1849, + "step": 563 + }, + { + "epoch": 0.034266966401360954, + "grad_norm": 0.3079189658164978, + "learning_rate": 9.975903103775638e-05, + "loss": 1.2714, + "step": 564 + }, + { + "epoch": 0.03432772343398748, + "grad_norm": 0.4284653067588806, + "learning_rate": 9.975809143097622e-05, + "loss": 1.135, + "step": 565 + }, + { + "epoch": 0.03438848046661401, + "grad_norm": 0.40955662727355957, + "learning_rate": 9.975715000030361e-05, + "loss": 1.2577, + "step": 566 + }, + { + "epoch": 0.03444923749924054, + "grad_norm": 0.4995199143886566, + "learning_rate": 9.975620674577303e-05, + "loss": 1.1287, + "step": 567 + }, + { + "epoch": 0.034509994531867066, + "grad_norm": 0.4478966295719147, + "learning_rate": 9.97552616674191e-05, + "loss": 1.3107, + "step": 568 + }, + { + "epoch": 0.03457075156449359, + "grad_norm": 0.3018648624420166, + "learning_rate": 9.975431476527642e-05, + "loss": 1.1867, + "step": 569 + }, + { + "epoch": 0.034631508597120116, + "grad_norm": 0.6003732085227966, + "learning_rate": 9.975336603937972e-05, + "loss": 1.1405, + "step": 570 + }, + { + "epoch": 0.034692265629746644, + "grad_norm": 0.3611820638179779, + "learning_rate": 9.975241548976377e-05, + "loss": 1.1721, + "step": 571 + }, + { + "epoch": 0.03475302266237317, + "grad_norm": 0.4594098627567291, + "learning_rate": 9.975146311646342e-05, + "loss": 1.2864, + "step": 572 + }, + { + "epoch": 0.03481377969499969, + "grad_norm": 0.272799551486969, + "learning_rate": 9.975050891951356e-05, + "loss": 1.1958, + "step": 573 + }, + { + "epoch": 0.03487453672762622, + "grad_norm": 0.30093836784362793, + "learning_rate": 9.97495528989492e-05, + "loss": 1.1663, + "step": 574 + }, + { + "epoch": 0.03493529376025275, + "grad_norm": 1.0804803371429443, + "learning_rate": 9.974859505480534e-05, + "loss": 1.1686, + "step": 575 + }, + { + "epoch": 0.03499605079287928, + "grad_norm": 4.020408630371094, + "learning_rate": 9.974763538711713e-05, + "loss": 1.1657, + "step": 576 + }, + { + "epoch": 0.035056807825505805, + "grad_norm": 0.3319571614265442, + "learning_rate": 9.97466738959197e-05, + "loss": 1.1572, + "step": 577 + }, + { + "epoch": 0.035117564858132326, + "grad_norm": 0.31526005268096924, + "learning_rate": 9.974571058124836e-05, + "loss": 1.1912, + "step": 578 + }, + { + "epoch": 0.035178321890758854, + "grad_norm": 0.36514735221862793, + "learning_rate": 9.974474544313835e-05, + "loss": 1.2383, + "step": 579 + }, + { + "epoch": 0.03523907892338538, + "grad_norm": 0.5009825229644775, + "learning_rate": 9.974377848162511e-05, + "loss": 1.2577, + "step": 580 + }, + { + "epoch": 0.03529983595601191, + "grad_norm": 0.4094410240650177, + "learning_rate": 9.974280969674403e-05, + "loss": 1.1673, + "step": 581 + }, + { + "epoch": 0.03536059298863844, + "grad_norm": 0.6641716361045837, + "learning_rate": 9.974183908853065e-05, + "loss": 1.1876, + "step": 582 + }, + { + "epoch": 0.03542135002126496, + "grad_norm": 2.715134620666504, + "learning_rate": 9.974086665702055e-05, + "loss": 1.2225, + "step": 583 + }, + { + "epoch": 0.03548210705389149, + "grad_norm": 0.7420315146446228, + "learning_rate": 9.973989240224936e-05, + "loss": 1.2741, + "step": 584 + }, + { + "epoch": 0.035542864086518015, + "grad_norm": 0.27612122893333435, + "learning_rate": 9.97389163242528e-05, + "loss": 1.1604, + "step": 585 + }, + { + "epoch": 0.03560362111914454, + "grad_norm": 0.6098658442497253, + "learning_rate": 9.973793842306667e-05, + "loss": 1.171, + "step": 586 + }, + { + "epoch": 0.035664378151771065, + "grad_norm": 0.2526607811450958, + "learning_rate": 9.973695869872676e-05, + "loss": 1.1509, + "step": 587 + }, + { + "epoch": 0.03572513518439759, + "grad_norm": 0.3331722617149353, + "learning_rate": 9.973597715126904e-05, + "loss": 1.1167, + "step": 588 + }, + { + "epoch": 0.03578589221702412, + "grad_norm": 0.22618432343006134, + "learning_rate": 9.973499378072945e-05, + "loss": 1.1308, + "step": 589 + }, + { + "epoch": 0.03584664924965065, + "grad_norm": 0.24366286396980286, + "learning_rate": 9.973400858714406e-05, + "loss": 1.1187, + "step": 590 + }, + { + "epoch": 0.03590740628227718, + "grad_norm": 0.49789854884147644, + "learning_rate": 9.973302157054897e-05, + "loss": 1.3979, + "step": 591 + }, + { + "epoch": 0.0359681633149037, + "grad_norm": 0.3018556833267212, + "learning_rate": 9.973203273098035e-05, + "loss": 1.1762, + "step": 592 + }, + { + "epoch": 0.036028920347530226, + "grad_norm": 0.311050146818161, + "learning_rate": 9.973104206847446e-05, + "loss": 1.2813, + "step": 593 + }, + { + "epoch": 0.036089677380156754, + "grad_norm": 0.2821758985519409, + "learning_rate": 9.973004958306761e-05, + "loss": 1.1528, + "step": 594 + }, + { + "epoch": 0.03615043441278328, + "grad_norm": 0.25873619318008423, + "learning_rate": 9.972905527479621e-05, + "loss": 1.1504, + "step": 595 + }, + { + "epoch": 0.0362111914454098, + "grad_norm": 0.3195224702358246, + "learning_rate": 9.972805914369665e-05, + "loss": 1.1685, + "step": 596 + }, + { + "epoch": 0.03627194847803633, + "grad_norm": 0.8943705558776855, + "learning_rate": 9.972706118980546e-05, + "loss": 1.138, + "step": 597 + }, + { + "epoch": 0.03633270551066286, + "grad_norm": 0.26881542801856995, + "learning_rate": 9.972606141315924e-05, + "loss": 1.1684, + "step": 598 + }, + { + "epoch": 0.03639346254328939, + "grad_norm": 0.4361538589000702, + "learning_rate": 9.972505981379464e-05, + "loss": 1.1828, + "step": 599 + }, + { + "epoch": 0.036454219575915915, + "grad_norm": 0.3007811903953552, + "learning_rate": 9.972405639174834e-05, + "loss": 1.2164, + "step": 600 + }, + { + "epoch": 0.036514976608542436, + "grad_norm": 0.3201947808265686, + "learning_rate": 9.972305114705715e-05, + "loss": 1.2573, + "step": 601 + }, + { + "epoch": 0.036575733641168964, + "grad_norm": 1.156196117401123, + "learning_rate": 9.972204407975792e-05, + "loss": 1.3427, + "step": 602 + }, + { + "epoch": 0.03663649067379549, + "grad_norm": 0.4037075936794281, + "learning_rate": 9.972103518988753e-05, + "loss": 1.3905, + "step": 603 + }, + { + "epoch": 0.03669724770642202, + "grad_norm": 0.6921052932739258, + "learning_rate": 9.9720024477483e-05, + "loss": 1.1348, + "step": 604 + }, + { + "epoch": 0.03675800473904854, + "grad_norm": 0.7323834896087646, + "learning_rate": 9.971901194258135e-05, + "loss": 1.2508, + "step": 605 + }, + { + "epoch": 0.03681876177167507, + "grad_norm": 0.4509468972682953, + "learning_rate": 9.971799758521972e-05, + "loss": 1.1974, + "step": 606 + }, + { + "epoch": 0.0368795188043016, + "grad_norm": 0.47222015261650085, + "learning_rate": 9.971698140543528e-05, + "loss": 1.1539, + "step": 607 + }, + { + "epoch": 0.036940275836928126, + "grad_norm": 0.3922358751296997, + "learning_rate": 9.971596340326526e-05, + "loss": 1.2571, + "step": 608 + }, + { + "epoch": 0.037001032869554654, + "grad_norm": 0.847470223903656, + "learning_rate": 9.971494357874701e-05, + "loss": 1.2547, + "step": 609 + }, + { + "epoch": 0.037061789902181175, + "grad_norm": 0.34950578212738037, + "learning_rate": 9.971392193191789e-05, + "loss": 1.1422, + "step": 610 + }, + { + "epoch": 0.0371225469348077, + "grad_norm": 2.7385778427124023, + "learning_rate": 9.971289846281535e-05, + "loss": 1.2829, + "step": 611 + }, + { + "epoch": 0.03718330396743423, + "grad_norm": 0.774427592754364, + "learning_rate": 9.97118731714769e-05, + "loss": 1.3312, + "step": 612 + }, + { + "epoch": 0.03724406100006076, + "grad_norm": 1.9668570756912231, + "learning_rate": 9.971084605794013e-05, + "loss": 1.1589, + "step": 613 + }, + { + "epoch": 0.03730481803268729, + "grad_norm": 8.825408935546875, + "learning_rate": 9.97098171222427e-05, + "loss": 1.2234, + "step": 614 + }, + { + "epoch": 0.03736557506531381, + "grad_norm": 1.2516340017318726, + "learning_rate": 9.970878636442231e-05, + "loss": 1.19, + "step": 615 + }, + { + "epoch": 0.037426332097940336, + "grad_norm": 0.35028496384620667, + "learning_rate": 9.970775378451674e-05, + "loss": 1.1605, + "step": 616 + }, + { + "epoch": 0.037487089130566864, + "grad_norm": 0.9051156640052795, + "learning_rate": 9.970671938256386e-05, + "loss": 1.4697, + "step": 617 + }, + { + "epoch": 0.03754784616319339, + "grad_norm": 0.9893403053283691, + "learning_rate": 9.970568315860156e-05, + "loss": 1.2041, + "step": 618 + }, + { + "epoch": 0.03760860319581991, + "grad_norm": 0.6288272738456726, + "learning_rate": 9.970464511266784e-05, + "loss": 1.2631, + "step": 619 + }, + { + "epoch": 0.03766936022844644, + "grad_norm": 0.9357312321662903, + "learning_rate": 9.970360524480076e-05, + "loss": 1.147, + "step": 620 + }, + { + "epoch": 0.03773011726107297, + "grad_norm": 0.48195454478263855, + "learning_rate": 9.97025635550384e-05, + "loss": 1.2646, + "step": 621 + }, + { + "epoch": 0.0377908742936995, + "grad_norm": 1.2346327304840088, + "learning_rate": 9.970152004341897e-05, + "loss": 1.1435, + "step": 622 + }, + { + "epoch": 0.037851631326326025, + "grad_norm": 0.6847200989723206, + "learning_rate": 9.970047470998071e-05, + "loss": 1.2298, + "step": 623 + }, + { + "epoch": 0.037912388358952546, + "grad_norm": 1.1393777132034302, + "learning_rate": 9.969942755476198e-05, + "loss": 1.2686, + "step": 624 + }, + { + "epoch": 0.037973145391579075, + "grad_norm": 0.6302562355995178, + "learning_rate": 9.969837857780109e-05, + "loss": 1.4583, + "step": 625 + }, + { + "epoch": 0.0380339024242056, + "grad_norm": 0.28724125027656555, + "learning_rate": 9.969732777913653e-05, + "loss": 1.1581, + "step": 626 + }, + { + "epoch": 0.03809465945683213, + "grad_norm": 0.5562950372695923, + "learning_rate": 9.969627515880682e-05, + "loss": 1.1385, + "step": 627 + }, + { + "epoch": 0.03815541648945865, + "grad_norm": 0.2949165403842926, + "learning_rate": 9.969522071685054e-05, + "loss": 1.1118, + "step": 628 + }, + { + "epoch": 0.03821617352208518, + "grad_norm": 0.3058724105358124, + "learning_rate": 9.969416445330634e-05, + "loss": 1.1736, + "step": 629 + }, + { + "epoch": 0.03827693055471171, + "grad_norm": 0.48314663767814636, + "learning_rate": 9.969310636821294e-05, + "loss": 1.2542, + "step": 630 + }, + { + "epoch": 0.038337687587338236, + "grad_norm": 0.4123871326446533, + "learning_rate": 9.96920464616091e-05, + "loss": 1.1803, + "step": 631 + }, + { + "epoch": 0.038398444619964764, + "grad_norm": 0.5740299224853516, + "learning_rate": 9.96909847335337e-05, + "loss": 1.1688, + "step": 632 + }, + { + "epoch": 0.038459201652591285, + "grad_norm": 0.6740185022354126, + "learning_rate": 9.968992118402567e-05, + "loss": 1.2553, + "step": 633 + }, + { + "epoch": 0.03851995868521781, + "grad_norm": 0.5766704082489014, + "learning_rate": 9.968885581312397e-05, + "loss": 1.1804, + "step": 634 + }, + { + "epoch": 0.03858071571784434, + "grad_norm": 0.4964434504508972, + "learning_rate": 9.968778862086763e-05, + "loss": 1.2448, + "step": 635 + }, + { + "epoch": 0.03864147275047087, + "grad_norm": 0.2984093725681305, + "learning_rate": 9.968671960729581e-05, + "loss": 1.1238, + "step": 636 + }, + { + "epoch": 0.03870222978309739, + "grad_norm": 0.3555094003677368, + "learning_rate": 9.968564877244768e-05, + "loss": 1.2324, + "step": 637 + }, + { + "epoch": 0.03876298681572392, + "grad_norm": 0.3393813967704773, + "learning_rate": 9.968457611636248e-05, + "loss": 1.2495, + "step": 638 + }, + { + "epoch": 0.038823743848350446, + "grad_norm": 0.33393850922584534, + "learning_rate": 9.968350163907956e-05, + "loss": 1.0918, + "step": 639 + }, + { + "epoch": 0.038884500880976974, + "grad_norm": 0.4749682545661926, + "learning_rate": 9.968242534063827e-05, + "loss": 1.1882, + "step": 640 + }, + { + "epoch": 0.0389452579136035, + "grad_norm": 0.4337617754936218, + "learning_rate": 9.968134722107807e-05, + "loss": 1.3418, + "step": 641 + }, + { + "epoch": 0.03900601494623002, + "grad_norm": 0.3234958052635193, + "learning_rate": 9.968026728043849e-05, + "loss": 1.1197, + "step": 642 + }, + { + "epoch": 0.03906677197885655, + "grad_norm": 0.5257587432861328, + "learning_rate": 9.96791855187591e-05, + "loss": 1.3474, + "step": 643 + }, + { + "epoch": 0.03912752901148308, + "grad_norm": 0.30168628692626953, + "learning_rate": 9.967810193607958e-05, + "loss": 1.1354, + "step": 644 + }, + { + "epoch": 0.03918828604410961, + "grad_norm": 0.3834643065929413, + "learning_rate": 9.967701653243963e-05, + "loss": 1.2864, + "step": 645 + }, + { + "epoch": 0.039249043076736136, + "grad_norm": 0.23010212182998657, + "learning_rate": 9.967592930787902e-05, + "loss": 1.1586, + "step": 646 + }, + { + "epoch": 0.03930980010936266, + "grad_norm": 0.2839571535587311, + "learning_rate": 9.967484026243763e-05, + "loss": 1.1588, + "step": 647 + }, + { + "epoch": 0.039370557141989185, + "grad_norm": 0.418510764837265, + "learning_rate": 9.967374939615537e-05, + "loss": 1.1823, + "step": 648 + }, + { + "epoch": 0.03943131417461571, + "grad_norm": 0.3012124300003052, + "learning_rate": 9.967265670907222e-05, + "loss": 1.2296, + "step": 649 + }, + { + "epoch": 0.03949207120724224, + "grad_norm": 0.3510946035385132, + "learning_rate": 9.967156220122824e-05, + "loss": 1.1891, + "step": 650 + }, + { + "epoch": 0.03955282823986876, + "grad_norm": 0.44404536485671997, + "learning_rate": 9.967046587266354e-05, + "loss": 1.296, + "step": 651 + }, + { + "epoch": 0.03961358527249529, + "grad_norm": 1.2017494440078735, + "learning_rate": 9.966936772341832e-05, + "loss": 1.2568, + "step": 652 + }, + { + "epoch": 0.03967434230512182, + "grad_norm": 0.5485511422157288, + "learning_rate": 9.966826775353281e-05, + "loss": 1.3018, + "step": 653 + }, + { + "epoch": 0.039735099337748346, + "grad_norm": 0.2625437378883362, + "learning_rate": 9.966716596304737e-05, + "loss": 1.177, + "step": 654 + }, + { + "epoch": 0.039795856370374874, + "grad_norm": 0.29605385661125183, + "learning_rate": 9.966606235200234e-05, + "loss": 1.2052, + "step": 655 + }, + { + "epoch": 0.039856613403001395, + "grad_norm": 2.1790478229522705, + "learning_rate": 9.96649569204382e-05, + "loss": 1.2445, + "step": 656 + }, + { + "epoch": 0.03991737043562792, + "grad_norm": 0.4488444924354553, + "learning_rate": 9.966384966839547e-05, + "loss": 1.2119, + "step": 657 + }, + { + "epoch": 0.03997812746825445, + "grad_norm": 0.3017423450946808, + "learning_rate": 9.966274059591472e-05, + "loss": 1.1891, + "step": 658 + }, + { + "epoch": 0.04003888450088098, + "grad_norm": 0.6103899478912354, + "learning_rate": 9.966162970303663e-05, + "loss": 1.1812, + "step": 659 + }, + { + "epoch": 0.0400996415335075, + "grad_norm": 0.2901724874973297, + "learning_rate": 9.96605169898019e-05, + "loss": 1.1522, + "step": 660 + }, + { + "epoch": 0.04016039856613403, + "grad_norm": 0.32875317335128784, + "learning_rate": 9.965940245625131e-05, + "loss": 1.1567, + "step": 661 + }, + { + "epoch": 0.040221155598760557, + "grad_norm": 0.3900388479232788, + "learning_rate": 9.965828610242574e-05, + "loss": 1.1884, + "step": 662 + }, + { + "epoch": 0.040281912631387085, + "grad_norm": 0.4106425940990448, + "learning_rate": 9.965716792836606e-05, + "loss": 1.3581, + "step": 663 + }, + { + "epoch": 0.04034266966401361, + "grad_norm": 0.3474547266960144, + "learning_rate": 9.965604793411331e-05, + "loss": 1.3391, + "step": 664 + }, + { + "epoch": 0.040403426696640134, + "grad_norm": 1.639144778251648, + "learning_rate": 9.965492611970854e-05, + "loss": 1.2061, + "step": 665 + }, + { + "epoch": 0.04046418372926666, + "grad_norm": 0.28106215596199036, + "learning_rate": 9.965380248519284e-05, + "loss": 1.1788, + "step": 666 + }, + { + "epoch": 0.04052494076189319, + "grad_norm": 0.34009480476379395, + "learning_rate": 9.965267703060743e-05, + "loss": 1.2803, + "step": 667 + }, + { + "epoch": 0.04058569779451972, + "grad_norm": 0.3992701768875122, + "learning_rate": 9.965154975599352e-05, + "loss": 1.2979, + "step": 668 + }, + { + "epoch": 0.04064645482714624, + "grad_norm": 0.2819100618362427, + "learning_rate": 9.965042066139248e-05, + "loss": 1.1539, + "step": 669 + }, + { + "epoch": 0.04070721185977277, + "grad_norm": 0.44505077600479126, + "learning_rate": 9.964928974684565e-05, + "loss": 1.1422, + "step": 670 + }, + { + "epoch": 0.040767968892399295, + "grad_norm": 0.5367493033409119, + "learning_rate": 9.964815701239452e-05, + "loss": 1.3082, + "step": 671 + }, + { + "epoch": 0.04082872592502582, + "grad_norm": 1.2104253768920898, + "learning_rate": 9.964702245808059e-05, + "loss": 1.0962, + "step": 672 + }, + { + "epoch": 0.04088948295765235, + "grad_norm": 0.5617942810058594, + "learning_rate": 9.964588608394547e-05, + "loss": 1.346, + "step": 673 + }, + { + "epoch": 0.04095023999027887, + "grad_norm": 0.2873060405254364, + "learning_rate": 9.96447478900308e-05, + "loss": 1.1686, + "step": 674 + }, + { + "epoch": 0.0410109970229054, + "grad_norm": 0.5462177991867065, + "learning_rate": 9.964360787637827e-05, + "loss": 1.2267, + "step": 675 + }, + { + "epoch": 0.04107175405553193, + "grad_norm": 0.5941920876502991, + "learning_rate": 9.964246604302974e-05, + "loss": 1.1238, + "step": 676 + }, + { + "epoch": 0.041132511088158456, + "grad_norm": 0.5403715372085571, + "learning_rate": 9.964132239002699e-05, + "loss": 1.2144, + "step": 677 + }, + { + "epoch": 0.04119326812078498, + "grad_norm": 0.8025391697883606, + "learning_rate": 9.964017691741198e-05, + "loss": 1.1718, + "step": 678 + }, + { + "epoch": 0.041254025153411505, + "grad_norm": 0.5789414644241333, + "learning_rate": 9.96390296252267e-05, + "loss": 1.12, + "step": 679 + }, + { + "epoch": 0.041314782186038033, + "grad_norm": 0.7309096455574036, + "learning_rate": 9.963788051351318e-05, + "loss": 1.326, + "step": 680 + }, + { + "epoch": 0.04137553921866456, + "grad_norm": 0.4682430028915405, + "learning_rate": 9.963672958231356e-05, + "loss": 1.2722, + "step": 681 + }, + { + "epoch": 0.04143629625129109, + "grad_norm": 0.5169833898544312, + "learning_rate": 9.963557683167003e-05, + "loss": 1.1328, + "step": 682 + }, + { + "epoch": 0.04149705328391761, + "grad_norm": 1.668455719947815, + "learning_rate": 9.963442226162482e-05, + "loss": 1.1684, + "step": 683 + }, + { + "epoch": 0.04155781031654414, + "grad_norm": 0.3803899884223938, + "learning_rate": 9.963326587222028e-05, + "loss": 1.153, + "step": 684 + }, + { + "epoch": 0.04161856734917067, + "grad_norm": 0.47263482213020325, + "learning_rate": 9.96321076634988e-05, + "loss": 1.1971, + "step": 685 + }, + { + "epoch": 0.041679324381797195, + "grad_norm": 0.5909217596054077, + "learning_rate": 9.96309476355028e-05, + "loss": 1.1999, + "step": 686 + }, + { + "epoch": 0.04174008141442372, + "grad_norm": 0.3885141909122467, + "learning_rate": 9.962978578827483e-05, + "loss": 1.1736, + "step": 687 + }, + { + "epoch": 0.041800838447050244, + "grad_norm": 0.40036067366600037, + "learning_rate": 9.962862212185746e-05, + "loss": 1.3672, + "step": 688 + }, + { + "epoch": 0.04186159547967677, + "grad_norm": 0.4191473126411438, + "learning_rate": 9.962745663629335e-05, + "loss": 1.2271, + "step": 689 + }, + { + "epoch": 0.0419223525123033, + "grad_norm": 0.2534301280975342, + "learning_rate": 9.962628933162524e-05, + "loss": 1.2463, + "step": 690 + }, + { + "epoch": 0.04198310954492983, + "grad_norm": 0.2598084509372711, + "learning_rate": 9.962512020789588e-05, + "loss": 1.1717, + "step": 691 + }, + { + "epoch": 0.04204386657755635, + "grad_norm": 0.3403211534023285, + "learning_rate": 9.962394926514817e-05, + "loss": 1.1852, + "step": 692 + }, + { + "epoch": 0.04210462361018288, + "grad_norm": 0.38530513644218445, + "learning_rate": 9.9622776503425e-05, + "loss": 1.2658, + "step": 693 + }, + { + "epoch": 0.042165380642809405, + "grad_norm": 0.4242812395095825, + "learning_rate": 9.962160192276935e-05, + "loss": 1.1336, + "step": 694 + }, + { + "epoch": 0.04222613767543593, + "grad_norm": 0.3795033097267151, + "learning_rate": 9.962042552322431e-05, + "loss": 1.1454, + "step": 695 + }, + { + "epoch": 0.04228689470806246, + "grad_norm": 0.36581653356552124, + "learning_rate": 9.961924730483297e-05, + "loss": 1.2173, + "step": 696 + }, + { + "epoch": 0.04234765174068898, + "grad_norm": 0.8262035250663757, + "learning_rate": 9.961806726763854e-05, + "loss": 1.2861, + "step": 697 + }, + { + "epoch": 0.04240840877331551, + "grad_norm": 0.859083354473114, + "learning_rate": 9.961688541168424e-05, + "loss": 1.1219, + "step": 698 + }, + { + "epoch": 0.04246916580594204, + "grad_norm": 0.30236339569091797, + "learning_rate": 9.961570173701344e-05, + "loss": 1.1677, + "step": 699 + }, + { + "epoch": 0.04252992283856857, + "grad_norm": 0.41853728890419006, + "learning_rate": 9.961451624366949e-05, + "loss": 1.141, + "step": 700 + }, + { + "epoch": 0.04259067987119509, + "grad_norm": 0.39060354232788086, + "learning_rate": 9.961332893169586e-05, + "loss": 1.3677, + "step": 701 + }, + { + "epoch": 0.042651436903821616, + "grad_norm": 0.4406738877296448, + "learning_rate": 9.961213980113606e-05, + "loss": 1.1904, + "step": 702 + }, + { + "epoch": 0.042712193936448144, + "grad_norm": 0.42465704679489136, + "learning_rate": 9.96109488520337e-05, + "loss": 1.152, + "step": 703 + }, + { + "epoch": 0.04277295096907467, + "grad_norm": 0.47877779603004456, + "learning_rate": 9.960975608443241e-05, + "loss": 1.2074, + "step": 704 + }, + { + "epoch": 0.0428337080017012, + "grad_norm": 0.48993995785713196, + "learning_rate": 9.960856149837592e-05, + "loss": 1.1793, + "step": 705 + }, + { + "epoch": 0.04289446503432772, + "grad_norm": 0.35581985116004944, + "learning_rate": 9.960736509390804e-05, + "loss": 1.1589, + "step": 706 + }, + { + "epoch": 0.04295522206695425, + "grad_norm": 0.29639628529548645, + "learning_rate": 9.960616687107258e-05, + "loss": 1.1402, + "step": 707 + }, + { + "epoch": 0.04301597909958078, + "grad_norm": 0.9274160861968994, + "learning_rate": 9.96049668299135e-05, + "loss": 1.1916, + "step": 708 + }, + { + "epoch": 0.043076736132207305, + "grad_norm": 0.3238775134086609, + "learning_rate": 9.960376497047476e-05, + "loss": 1.1985, + "step": 709 + }, + { + "epoch": 0.043137493164833826, + "grad_norm": 0.4327608048915863, + "learning_rate": 9.960256129280043e-05, + "loss": 1.2165, + "step": 710 + }, + { + "epoch": 0.043198250197460354, + "grad_norm": 0.3544516861438751, + "learning_rate": 9.960135579693462e-05, + "loss": 1.1291, + "step": 711 + }, + { + "epoch": 0.04325900723008688, + "grad_norm": 0.2925226390361786, + "learning_rate": 9.960014848292154e-05, + "loss": 1.1159, + "step": 712 + }, + { + "epoch": 0.04331976426271341, + "grad_norm": 0.4761642515659332, + "learning_rate": 9.959893935080542e-05, + "loss": 1.1259, + "step": 713 + }, + { + "epoch": 0.04338052129533994, + "grad_norm": 0.34526893496513367, + "learning_rate": 9.95977284006306e-05, + "loss": 1.2255, + "step": 714 + }, + { + "epoch": 0.04344127832796646, + "grad_norm": 0.5785567164421082, + "learning_rate": 9.959651563244146e-05, + "loss": 1.1606, + "step": 715 + }, + { + "epoch": 0.04350203536059299, + "grad_norm": 0.3859511911869049, + "learning_rate": 9.959530104628245e-05, + "loss": 1.1521, + "step": 716 + }, + { + "epoch": 0.043562792393219515, + "grad_norm": 0.3595500588417053, + "learning_rate": 9.959408464219809e-05, + "loss": 1.2585, + "step": 717 + }, + { + "epoch": 0.043623549425846044, + "grad_norm": 0.7069131731987, + "learning_rate": 9.959286642023297e-05, + "loss": 1.1612, + "step": 718 + }, + { + "epoch": 0.04368430645847257, + "grad_norm": 1.7231045961380005, + "learning_rate": 9.959164638043175e-05, + "loss": 1.1519, + "step": 719 + }, + { + "epoch": 0.04374506349109909, + "grad_norm": 0.3725694715976715, + "learning_rate": 9.959042452283914e-05, + "loss": 1.3835, + "step": 720 + }, + { + "epoch": 0.04380582052372562, + "grad_norm": 0.4322760999202728, + "learning_rate": 9.958920084749993e-05, + "loss": 1.2474, + "step": 721 + }, + { + "epoch": 0.04386657755635215, + "grad_norm": 0.7434590458869934, + "learning_rate": 9.958797535445899e-05, + "loss": 1.1417, + "step": 722 + }, + { + "epoch": 0.04392733458897868, + "grad_norm": 2.5536746978759766, + "learning_rate": 9.958674804376123e-05, + "loss": 1.1188, + "step": 723 + }, + { + "epoch": 0.0439880916216052, + "grad_norm": 3.2335243225097656, + "learning_rate": 9.958551891545161e-05, + "loss": 1.2444, + "step": 724 + }, + { + "epoch": 0.044048848654231726, + "grad_norm": 1.1662545204162598, + "learning_rate": 9.958428796957523e-05, + "loss": 1.1705, + "step": 725 + }, + { + "epoch": 0.044109605686858254, + "grad_norm": 0.34551477432250977, + "learning_rate": 9.95830552061772e-05, + "loss": 1.1448, + "step": 726 + }, + { + "epoch": 0.04417036271948478, + "grad_norm": 1.8899147510528564, + "learning_rate": 9.958182062530268e-05, + "loss": 1.1491, + "step": 727 + }, + { + "epoch": 0.04423111975211131, + "grad_norm": 0.9002741575241089, + "learning_rate": 9.958058422699695e-05, + "loss": 1.1197, + "step": 728 + }, + { + "epoch": 0.04429187678473783, + "grad_norm": 0.32681119441986084, + "learning_rate": 9.957934601130529e-05, + "loss": 1.353, + "step": 729 + }, + { + "epoch": 0.04435263381736436, + "grad_norm": 0.6409501433372498, + "learning_rate": 9.957810597827316e-05, + "loss": 1.2756, + "step": 730 + }, + { + "epoch": 0.04441339084999089, + "grad_norm": 0.5005955100059509, + "learning_rate": 9.957686412794595e-05, + "loss": 1.1063, + "step": 731 + }, + { + "epoch": 0.044474147882617415, + "grad_norm": 0.5368000864982605, + "learning_rate": 9.957562046036921e-05, + "loss": 1.276, + "step": 732 + }, + { + "epoch": 0.044534904915243936, + "grad_norm": 1.8605934381484985, + "learning_rate": 9.957437497558852e-05, + "loss": 1.1718, + "step": 733 + }, + { + "epoch": 0.044595661947870464, + "grad_norm": 0.6098967790603638, + "learning_rate": 9.957312767364953e-05, + "loss": 1.302, + "step": 734 + }, + { + "epoch": 0.04465641898049699, + "grad_norm": 0.3161158561706543, + "learning_rate": 9.957187855459798e-05, + "loss": 1.169, + "step": 735 + }, + { + "epoch": 0.04471717601312352, + "grad_norm": 1.0314669609069824, + "learning_rate": 9.957062761847962e-05, + "loss": 1.3151, + "step": 736 + }, + { + "epoch": 0.04477793304575005, + "grad_norm": 0.46221718192100525, + "learning_rate": 9.956937486534033e-05, + "loss": 1.1821, + "step": 737 + }, + { + "epoch": 0.04483869007837657, + "grad_norm": 0.3953547775745392, + "learning_rate": 9.956812029522602e-05, + "loss": 1.1715, + "step": 738 + }, + { + "epoch": 0.0448994471110031, + "grad_norm": 0.7425771951675415, + "learning_rate": 9.956686390818268e-05, + "loss": 1.3491, + "step": 739 + }, + { + "epoch": 0.044960204143629626, + "grad_norm": 0.44215551018714905, + "learning_rate": 9.956560570425635e-05, + "loss": 1.3949, + "step": 740 + }, + { + "epoch": 0.045020961176256154, + "grad_norm": 0.5746815800666809, + "learning_rate": 9.956434568349318e-05, + "loss": 1.1903, + "step": 741 + }, + { + "epoch": 0.045081718208882675, + "grad_norm": 0.37910205125808716, + "learning_rate": 9.956308384593933e-05, + "loss": 1.1265, + "step": 742 + }, + { + "epoch": 0.0451424752415092, + "grad_norm": 0.6894158720970154, + "learning_rate": 9.956182019164106e-05, + "loss": 1.2741, + "step": 743 + }, + { + "epoch": 0.04520323227413573, + "grad_norm": 0.3763534724712372, + "learning_rate": 9.956055472064469e-05, + "loss": 1.1401, + "step": 744 + }, + { + "epoch": 0.04526398930676226, + "grad_norm": 0.6466929316520691, + "learning_rate": 9.955928743299662e-05, + "loss": 1.203, + "step": 745 + }, + { + "epoch": 0.04532474633938879, + "grad_norm": 0.3915845453739166, + "learning_rate": 9.955801832874328e-05, + "loss": 1.1563, + "step": 746 + }, + { + "epoch": 0.04538550337201531, + "grad_norm": 10.527915000915527, + "learning_rate": 9.955674740793119e-05, + "loss": 1.2764, + "step": 747 + }, + { + "epoch": 0.045446260404641836, + "grad_norm": 1.3537558317184448, + "learning_rate": 9.955547467060696e-05, + "loss": 1.3087, + "step": 748 + }, + { + "epoch": 0.045507017437268364, + "grad_norm": 0.377202570438385, + "learning_rate": 9.955420011681722e-05, + "loss": 1.1839, + "step": 749 + }, + { + "epoch": 0.04556777446989489, + "grad_norm": 0.704920768737793, + "learning_rate": 9.955292374660869e-05, + "loss": 1.1507, + "step": 750 + }, + { + "epoch": 0.04562853150252142, + "grad_norm": 0.6823822855949402, + "learning_rate": 9.955164556002819e-05, + "loss": 1.1149, + "step": 751 + }, + { + "epoch": 0.04568928853514794, + "grad_norm": 0.7192385196685791, + "learning_rate": 9.955036555712252e-05, + "loss": 1.2125, + "step": 752 + }, + { + "epoch": 0.04575004556777447, + "grad_norm": 0.7568918466567993, + "learning_rate": 9.954908373793861e-05, + "loss": 1.5913, + "step": 753 + }, + { + "epoch": 0.045810802600401, + "grad_norm": 0.5486075282096863, + "learning_rate": 9.954780010252347e-05, + "loss": 1.1207, + "step": 754 + }, + { + "epoch": 0.045871559633027525, + "grad_norm": 1.3861589431762695, + "learning_rate": 9.954651465092414e-05, + "loss": 1.1366, + "step": 755 + }, + { + "epoch": 0.04593231666565405, + "grad_norm": 1.0610893964767456, + "learning_rate": 9.954522738318776e-05, + "loss": 1.2659, + "step": 756 + }, + { + "epoch": 0.045993073698280575, + "grad_norm": 2.5721888542175293, + "learning_rate": 9.954393829936146e-05, + "loss": 1.1177, + "step": 757 + }, + { + "epoch": 0.0460538307309071, + "grad_norm": 1.484799861907959, + "learning_rate": 9.954264739949255e-05, + "loss": 1.2274, + "step": 758 + }, + { + "epoch": 0.04611458776353363, + "grad_norm": 1.4131289720535278, + "learning_rate": 9.954135468362831e-05, + "loss": 1.214, + "step": 759 + }, + { + "epoch": 0.04617534479616016, + "grad_norm": 0.5996922254562378, + "learning_rate": 9.954006015181614e-05, + "loss": 1.2802, + "step": 760 + }, + { + "epoch": 0.04623610182878668, + "grad_norm": 1.9764797687530518, + "learning_rate": 9.95387638041035e-05, + "loss": 1.1091, + "step": 761 + }, + { + "epoch": 0.04629685886141321, + "grad_norm": 0.9029185771942139, + "learning_rate": 9.953746564053788e-05, + "loss": 1.1653, + "step": 762 + }, + { + "epoch": 0.046357615894039736, + "grad_norm": 0.8886936902999878, + "learning_rate": 9.953616566116689e-05, + "loss": 1.2422, + "step": 763 + }, + { + "epoch": 0.046418372926666264, + "grad_norm": 0.9418110251426697, + "learning_rate": 9.953486386603816e-05, + "loss": 1.1108, + "step": 764 + }, + { + "epoch": 0.046479129959292785, + "grad_norm": 0.406977117061615, + "learning_rate": 9.953356025519945e-05, + "loss": 1.1958, + "step": 765 + }, + { + "epoch": 0.04653988699191931, + "grad_norm": 0.9886865019798279, + "learning_rate": 9.953225482869849e-05, + "loss": 1.2174, + "step": 766 + }, + { + "epoch": 0.04660064402454584, + "grad_norm": 0.5407412648200989, + "learning_rate": 9.953094758658315e-05, + "loss": 1.1729, + "step": 767 + }, + { + "epoch": 0.04666140105717237, + "grad_norm": 0.9849939346313477, + "learning_rate": 9.952963852890137e-05, + "loss": 1.1107, + "step": 768 + }, + { + "epoch": 0.0467221580897989, + "grad_norm": 3.656435251235962, + "learning_rate": 9.95283276557011e-05, + "loss": 1.2562, + "step": 769 + }, + { + "epoch": 0.04678291512242542, + "grad_norm": 0.5425155758857727, + "learning_rate": 9.952701496703043e-05, + "loss": 1.2146, + "step": 770 + }, + { + "epoch": 0.046843672155051946, + "grad_norm": 1.5659294128417969, + "learning_rate": 9.952570046293743e-05, + "loss": 1.1968, + "step": 771 + }, + { + "epoch": 0.046904429187678474, + "grad_norm": 0.5599942207336426, + "learning_rate": 9.952438414347032e-05, + "loss": 1.1785, + "step": 772 + }, + { + "epoch": 0.046965186220305, + "grad_norm": 0.2474890649318695, + "learning_rate": 9.952306600867734e-05, + "loss": 1.1714, + "step": 773 + }, + { + "epoch": 0.047025943252931524, + "grad_norm": 0.5029478669166565, + "learning_rate": 9.952174605860678e-05, + "loss": 1.2196, + "step": 774 + }, + { + "epoch": 0.04708670028555805, + "grad_norm": 0.23033669590950012, + "learning_rate": 9.952042429330705e-05, + "loss": 1.1361, + "step": 775 + }, + { + "epoch": 0.04714745731818458, + "grad_norm": 0.23688176274299622, + "learning_rate": 9.951910071282662e-05, + "loss": 1.145, + "step": 776 + }, + { + "epoch": 0.04720821435081111, + "grad_norm": 0.39027896523475647, + "learning_rate": 9.951777531721396e-05, + "loss": 1.1485, + "step": 777 + }, + { + "epoch": 0.047268971383437636, + "grad_norm": 0.5398600697517395, + "learning_rate": 9.951644810651767e-05, + "loss": 1.2471, + "step": 778 + }, + { + "epoch": 0.04732972841606416, + "grad_norm": 0.39004793763160706, + "learning_rate": 9.951511908078644e-05, + "loss": 1.3301, + "step": 779 + }, + { + "epoch": 0.047390485448690685, + "grad_norm": 0.3112984895706177, + "learning_rate": 9.951378824006891e-05, + "loss": 1.227, + "step": 780 + }, + { + "epoch": 0.04745124248131721, + "grad_norm": 0.32644209265708923, + "learning_rate": 9.951245558441391e-05, + "loss": 1.2124, + "step": 781 + }, + { + "epoch": 0.04751199951394374, + "grad_norm": 0.28608182072639465, + "learning_rate": 9.951112111387029e-05, + "loss": 1.1586, + "step": 782 + }, + { + "epoch": 0.04757275654657027, + "grad_norm": 0.4290551543235779, + "learning_rate": 9.950978482848693e-05, + "loss": 1.2921, + "step": 783 + }, + { + "epoch": 0.04763351357919679, + "grad_norm": 0.29653456807136536, + "learning_rate": 9.950844672831288e-05, + "loss": 1.2175, + "step": 784 + }, + { + "epoch": 0.04769427061182332, + "grad_norm": 0.2684912085533142, + "learning_rate": 9.950710681339712e-05, + "loss": 1.1152, + "step": 785 + }, + { + "epoch": 0.047755027644449846, + "grad_norm": 0.21391050517559052, + "learning_rate": 9.95057650837888e-05, + "loss": 1.1659, + "step": 786 + }, + { + "epoch": 0.047815784677076374, + "grad_norm": 0.3373187482357025, + "learning_rate": 9.950442153953707e-05, + "loss": 1.1886, + "step": 787 + }, + { + "epoch": 0.047876541709702895, + "grad_norm": 0.2740505039691925, + "learning_rate": 9.950307618069122e-05, + "loss": 1.2081, + "step": 788 + }, + { + "epoch": 0.04793729874232942, + "grad_norm": 0.25902989506721497, + "learning_rate": 9.950172900730053e-05, + "loss": 1.3046, + "step": 789 + }, + { + "epoch": 0.04799805577495595, + "grad_norm": 0.3308719992637634, + "learning_rate": 9.950038001941441e-05, + "loss": 1.2488, + "step": 790 + }, + { + "epoch": 0.04805881280758248, + "grad_norm": 0.4035930335521698, + "learning_rate": 9.949902921708228e-05, + "loss": 1.2213, + "step": 791 + }, + { + "epoch": 0.04811956984020901, + "grad_norm": 0.46484047174453735, + "learning_rate": 9.949767660035368e-05, + "loss": 1.153, + "step": 792 + }, + { + "epoch": 0.04818032687283553, + "grad_norm": 0.2411298155784607, + "learning_rate": 9.949632216927815e-05, + "loss": 1.15, + "step": 793 + }, + { + "epoch": 0.04824108390546206, + "grad_norm": 0.35771939158439636, + "learning_rate": 9.949496592390539e-05, + "loss": 1.2881, + "step": 794 + }, + { + "epoch": 0.048301840938088585, + "grad_norm": 0.31029146909713745, + "learning_rate": 9.949360786428508e-05, + "loss": 1.1128, + "step": 795 + }, + { + "epoch": 0.04836259797071511, + "grad_norm": 0.36905092000961304, + "learning_rate": 9.949224799046699e-05, + "loss": 1.3203, + "step": 796 + }, + { + "epoch": 0.048423355003341634, + "grad_norm": 0.2597476840019226, + "learning_rate": 9.9490886302501e-05, + "loss": 1.1288, + "step": 797 + }, + { + "epoch": 0.04848411203596816, + "grad_norm": 1.012656807899475, + "learning_rate": 9.948952280043701e-05, + "loss": 1.1572, + "step": 798 + }, + { + "epoch": 0.04854486906859469, + "grad_norm": 0.33429670333862305, + "learning_rate": 9.948815748432499e-05, + "loss": 1.191, + "step": 799 + }, + { + "epoch": 0.04860562610122122, + "grad_norm": 0.3133918344974518, + "learning_rate": 9.948679035421499e-05, + "loss": 1.104, + "step": 800 + }, + { + "epoch": 0.048666383133847746, + "grad_norm": 0.24302445352077484, + "learning_rate": 9.948542141015714e-05, + "loss": 1.2348, + "step": 801 + }, + { + "epoch": 0.04872714016647427, + "grad_norm": 0.3677830100059509, + "learning_rate": 9.948405065220156e-05, + "loss": 1.1159, + "step": 802 + }, + { + "epoch": 0.048787897199100795, + "grad_norm": 0.5479230284690857, + "learning_rate": 9.948267808039857e-05, + "loss": 1.2109, + "step": 803 + }, + { + "epoch": 0.04884865423172732, + "grad_norm": 0.2904531955718994, + "learning_rate": 9.948130369479844e-05, + "loss": 1.1275, + "step": 804 + }, + { + "epoch": 0.04890941126435385, + "grad_norm": 0.4609297513961792, + "learning_rate": 9.947992749545158e-05, + "loss": 1.1101, + "step": 805 + }, + { + "epoch": 0.04897016829698037, + "grad_norm": 0.8064852952957153, + "learning_rate": 9.94785494824084e-05, + "loss": 1.3596, + "step": 806 + }, + { + "epoch": 0.0490309253296069, + "grad_norm": 0.6158781051635742, + "learning_rate": 9.947716965571941e-05, + "loss": 1.1169, + "step": 807 + }, + { + "epoch": 0.04909168236223343, + "grad_norm": 0.3816830813884735, + "learning_rate": 9.947578801543522e-05, + "loss": 1.1654, + "step": 808 + }, + { + "epoch": 0.049152439394859956, + "grad_norm": 0.4013568162918091, + "learning_rate": 9.947440456160645e-05, + "loss": 1.1445, + "step": 809 + }, + { + "epoch": 0.049213196427486484, + "grad_norm": 0.44469669461250305, + "learning_rate": 9.947301929428384e-05, + "loss": 1.161, + "step": 810 + }, + { + "epoch": 0.049273953460113006, + "grad_norm": 0.5190894603729248, + "learning_rate": 9.94716322135181e-05, + "loss": 1.1706, + "step": 811 + }, + { + "epoch": 0.049334710492739534, + "grad_norm": 0.3834163546562195, + "learning_rate": 9.947024331936015e-05, + "loss": 1.1215, + "step": 812 + }, + { + "epoch": 0.04939546752536606, + "grad_norm": 0.3233342170715332, + "learning_rate": 9.946885261186086e-05, + "loss": 1.1363, + "step": 813 + }, + { + "epoch": 0.04945622455799259, + "grad_norm": 0.2785968780517578, + "learning_rate": 9.946746009107121e-05, + "loss": 1.1276, + "step": 814 + }, + { + "epoch": 0.04951698159061911, + "grad_norm": 0.4092848002910614, + "learning_rate": 9.946606575704227e-05, + "loss": 1.1537, + "step": 815 + }, + { + "epoch": 0.04957773862324564, + "grad_norm": 0.5956786870956421, + "learning_rate": 9.946466960982512e-05, + "loss": 1.2371, + "step": 816 + }, + { + "epoch": 0.04963849565587217, + "grad_norm": 0.33586862683296204, + "learning_rate": 9.946327164947094e-05, + "loss": 1.1243, + "step": 817 + }, + { + "epoch": 0.049699252688498695, + "grad_norm": 0.2830551564693451, + "learning_rate": 9.946187187603097e-05, + "loss": 1.166, + "step": 818 + }, + { + "epoch": 0.04976000972112522, + "grad_norm": 0.37107354402542114, + "learning_rate": 9.946047028955651e-05, + "loss": 1.2259, + "step": 819 + }, + { + "epoch": 0.049820766753751744, + "grad_norm": 0.3013699948787689, + "learning_rate": 9.945906689009898e-05, + "loss": 1.2132, + "step": 820 + }, + { + "epoch": 0.04988152378637827, + "grad_norm": 0.2775358557701111, + "learning_rate": 9.94576616777098e-05, + "loss": 1.1449, + "step": 821 + }, + { + "epoch": 0.0499422808190048, + "grad_norm": 0.29960718750953674, + "learning_rate": 9.945625465244044e-05, + "loss": 1.1834, + "step": 822 + }, + { + "epoch": 0.05000303785163133, + "grad_norm": 0.21503199636936188, + "learning_rate": 9.945484581434253e-05, + "loss": 1.17, + "step": 823 + }, + { + "epoch": 0.050063794884257856, + "grad_norm": 0.28748559951782227, + "learning_rate": 9.945343516346768e-05, + "loss": 1.1555, + "step": 824 + }, + { + "epoch": 0.05012455191688438, + "grad_norm": 0.2709907591342926, + "learning_rate": 9.94520226998676e-05, + "loss": 1.1744, + "step": 825 + }, + { + "epoch": 0.050185308949510905, + "grad_norm": 0.23551982641220093, + "learning_rate": 9.945060842359408e-05, + "loss": 1.1707, + "step": 826 + }, + { + "epoch": 0.05024606598213743, + "grad_norm": 0.24281173944473267, + "learning_rate": 9.944919233469894e-05, + "loss": 1.2363, + "step": 827 + }, + { + "epoch": 0.05030682301476396, + "grad_norm": 1.0215262174606323, + "learning_rate": 9.944777443323409e-05, + "loss": 1.2947, + "step": 828 + }, + { + "epoch": 0.05036758004739048, + "grad_norm": 0.265689879655838, + "learning_rate": 9.944635471925153e-05, + "loss": 1.1499, + "step": 829 + }, + { + "epoch": 0.05042833708001701, + "grad_norm": 0.6228410005569458, + "learning_rate": 9.944493319280326e-05, + "loss": 1.1823, + "step": 830 + }, + { + "epoch": 0.05048909411264354, + "grad_norm": 0.2640187740325928, + "learning_rate": 9.944350985394142e-05, + "loss": 1.1492, + "step": 831 + }, + { + "epoch": 0.05054985114527007, + "grad_norm": 0.5398237705230713, + "learning_rate": 9.944208470271816e-05, + "loss": 1.1408, + "step": 832 + }, + { + "epoch": 0.050610608177896595, + "grad_norm": 0.2894194722175598, + "learning_rate": 9.944065773918574e-05, + "loss": 1.1729, + "step": 833 + }, + { + "epoch": 0.050671365210523116, + "grad_norm": 0.4743664562702179, + "learning_rate": 9.943922896339644e-05, + "loss": 1.2626, + "step": 834 + }, + { + "epoch": 0.050732122243149644, + "grad_norm": 0.21232442557811737, + "learning_rate": 9.943779837540266e-05, + "loss": 1.1729, + "step": 835 + }, + { + "epoch": 0.05079287927577617, + "grad_norm": 0.3184720277786255, + "learning_rate": 9.94363659752568e-05, + "loss": 1.1598, + "step": 836 + }, + { + "epoch": 0.0508536363084027, + "grad_norm": 0.4539019465446472, + "learning_rate": 9.943493176301142e-05, + "loss": 1.1198, + "step": 837 + }, + { + "epoch": 0.05091439334102922, + "grad_norm": 0.5775179266929626, + "learning_rate": 9.943349573871905e-05, + "loss": 1.4587, + "step": 838 + }, + { + "epoch": 0.05097515037365575, + "grad_norm": 0.27564606070518494, + "learning_rate": 9.943205790243233e-05, + "loss": 1.1059, + "step": 839 + }, + { + "epoch": 0.05103590740628228, + "grad_norm": 0.261208713054657, + "learning_rate": 9.943061825420398e-05, + "loss": 1.1841, + "step": 840 + }, + { + "epoch": 0.051096664438908805, + "grad_norm": 0.33368104696273804, + "learning_rate": 9.942917679408676e-05, + "loss": 1.2307, + "step": 841 + }, + { + "epoch": 0.05115742147153533, + "grad_norm": 0.33767732977867126, + "learning_rate": 9.94277335221335e-05, + "loss": 1.1516, + "step": 842 + }, + { + "epoch": 0.051218178504161854, + "grad_norm": 0.3082011044025421, + "learning_rate": 9.942628843839714e-05, + "loss": 1.1161, + "step": 843 + }, + { + "epoch": 0.05127893553678838, + "grad_norm": 0.6501439213752747, + "learning_rate": 9.94248415429306e-05, + "loss": 1.1426, + "step": 844 + }, + { + "epoch": 0.05133969256941491, + "grad_norm": 0.4692160487174988, + "learning_rate": 9.942339283578694e-05, + "loss": 1.1063, + "step": 845 + }, + { + "epoch": 0.05140044960204144, + "grad_norm": 0.4904324412345886, + "learning_rate": 9.942194231701929e-05, + "loss": 1.3452, + "step": 846 + }, + { + "epoch": 0.05146120663466796, + "grad_norm": 0.571820080280304, + "learning_rate": 9.942048998668077e-05, + "loss": 1.2471, + "step": 847 + }, + { + "epoch": 0.05152196366729449, + "grad_norm": 0.40043818950653076, + "learning_rate": 9.941903584482463e-05, + "loss": 1.2489, + "step": 848 + }, + { + "epoch": 0.051582720699921016, + "grad_norm": 0.6650826930999756, + "learning_rate": 9.941757989150419e-05, + "loss": 1.2643, + "step": 849 + }, + { + "epoch": 0.051643477732547544, + "grad_norm": 0.29371532797813416, + "learning_rate": 9.94161221267728e-05, + "loss": 1.1233, + "step": 850 + }, + { + "epoch": 0.05170423476517407, + "grad_norm": 0.5212189555168152, + "learning_rate": 9.94146625506839e-05, + "loss": 1.3177, + "step": 851 + }, + { + "epoch": 0.05176499179780059, + "grad_norm": 0.34916383028030396, + "learning_rate": 9.9413201163291e-05, + "loss": 1.2988, + "step": 852 + }, + { + "epoch": 0.05182574883042712, + "grad_norm": 0.4646431505680084, + "learning_rate": 9.941173796464767e-05, + "loss": 1.1696, + "step": 853 + }, + { + "epoch": 0.05188650586305365, + "grad_norm": 0.3617874085903168, + "learning_rate": 9.941027295480752e-05, + "loss": 1.127, + "step": 854 + }, + { + "epoch": 0.05194726289568018, + "grad_norm": 0.31283125281333923, + "learning_rate": 9.940880613382427e-05, + "loss": 1.2337, + "step": 855 + }, + { + "epoch": 0.052008019928306705, + "grad_norm": 0.5063871145248413, + "learning_rate": 9.940733750175166e-05, + "loss": 1.2672, + "step": 856 + }, + { + "epoch": 0.052068776960933226, + "grad_norm": 0.6898863315582275, + "learning_rate": 9.940586705864357e-05, + "loss": 1.2564, + "step": 857 + }, + { + "epoch": 0.052129533993559754, + "grad_norm": 0.7902609705924988, + "learning_rate": 9.940439480455386e-05, + "loss": 1.1371, + "step": 858 + }, + { + "epoch": 0.05219029102618628, + "grad_norm": 0.5012494921684265, + "learning_rate": 9.940292073953652e-05, + "loss": 1.1278, + "step": 859 + }, + { + "epoch": 0.05225104805881281, + "grad_norm": 0.27230304479599, + "learning_rate": 9.940144486364556e-05, + "loss": 1.1015, + "step": 860 + }, + { + "epoch": 0.05231180509143933, + "grad_norm": 0.6863483190536499, + "learning_rate": 9.93999671769351e-05, + "loss": 1.1131, + "step": 861 + }, + { + "epoch": 0.05237256212406586, + "grad_norm": 0.29172226786613464, + "learning_rate": 9.939848767945927e-05, + "loss": 1.1589, + "step": 862 + }, + { + "epoch": 0.05243331915669239, + "grad_norm": 0.4777410924434662, + "learning_rate": 9.939700637127235e-05, + "loss": 1.1289, + "step": 863 + }, + { + "epoch": 0.052494076189318915, + "grad_norm": 0.3544274568557739, + "learning_rate": 9.93955232524286e-05, + "loss": 1.0822, + "step": 864 + }, + { + "epoch": 0.05255483322194544, + "grad_norm": 0.3159794807434082, + "learning_rate": 9.93940383229824e-05, + "loss": 1.219, + "step": 865 + }, + { + "epoch": 0.052615590254571964, + "grad_norm": 0.45264145731925964, + "learning_rate": 9.939255158298819e-05, + "loss": 1.1699, + "step": 866 + }, + { + "epoch": 0.05267634728719849, + "grad_norm": 0.28136083483695984, + "learning_rate": 9.939106303250044e-05, + "loss": 1.1484, + "step": 867 + }, + { + "epoch": 0.05273710431982502, + "grad_norm": 0.537706732749939, + "learning_rate": 9.938957267157374e-05, + "loss": 1.174, + "step": 868 + }, + { + "epoch": 0.05279786135245155, + "grad_norm": 5.582092761993408, + "learning_rate": 9.93880805002627e-05, + "loss": 1.1147, + "step": 869 + }, + { + "epoch": 0.05285861838507807, + "grad_norm": 0.3526427447795868, + "learning_rate": 9.938658651862201e-05, + "loss": 1.287, + "step": 870 + }, + { + "epoch": 0.0529193754177046, + "grad_norm": 0.2831250727176666, + "learning_rate": 9.938509072670646e-05, + "loss": 1.1733, + "step": 871 + }, + { + "epoch": 0.052980132450331126, + "grad_norm": 0.3057091534137726, + "learning_rate": 9.938359312457087e-05, + "loss": 1.2841, + "step": 872 + }, + { + "epoch": 0.053040889482957654, + "grad_norm": 0.21608521044254303, + "learning_rate": 9.938209371227011e-05, + "loss": 1.1197, + "step": 873 + }, + { + "epoch": 0.05310164651558418, + "grad_norm": 0.2625390887260437, + "learning_rate": 9.938059248985916e-05, + "loss": 1.1487, + "step": 874 + }, + { + "epoch": 0.0531624035482107, + "grad_norm": 0.2997075021266937, + "learning_rate": 9.937908945739307e-05, + "loss": 1.1192, + "step": 875 + }, + { + "epoch": 0.05322316058083723, + "grad_norm": 0.2858152389526367, + "learning_rate": 9.93775846149269e-05, + "loss": 1.2175, + "step": 876 + }, + { + "epoch": 0.05328391761346376, + "grad_norm": 0.35043519735336304, + "learning_rate": 9.937607796251582e-05, + "loss": 1.2055, + "step": 877 + }, + { + "epoch": 0.05334467464609029, + "grad_norm": 0.26519519090652466, + "learning_rate": 9.937456950021504e-05, + "loss": 1.1256, + "step": 878 + }, + { + "epoch": 0.05340543167871681, + "grad_norm": 0.28305256366729736, + "learning_rate": 9.937305922807989e-05, + "loss": 1.1359, + "step": 879 + }, + { + "epoch": 0.053466188711343336, + "grad_norm": 0.3161332309246063, + "learning_rate": 9.937154714616572e-05, + "loss": 1.2266, + "step": 880 + }, + { + "epoch": 0.053526945743969864, + "grad_norm": 0.2984045743942261, + "learning_rate": 9.937003325452792e-05, + "loss": 1.2022, + "step": 881 + }, + { + "epoch": 0.05358770277659639, + "grad_norm": 0.26703837513923645, + "learning_rate": 9.936851755322203e-05, + "loss": 1.1711, + "step": 882 + }, + { + "epoch": 0.05364845980922292, + "grad_norm": 0.48756542801856995, + "learning_rate": 9.936700004230358e-05, + "loss": 1.0909, + "step": 883 + }, + { + "epoch": 0.05370921684184944, + "grad_norm": 0.32010743021965027, + "learning_rate": 9.93654807218282e-05, + "loss": 1.203, + "step": 884 + }, + { + "epoch": 0.05376997387447597, + "grad_norm": 0.4005170166492462, + "learning_rate": 9.936395959185158e-05, + "loss": 1.2612, + "step": 885 + }, + { + "epoch": 0.0538307309071025, + "grad_norm": 0.272266149520874, + "learning_rate": 9.936243665242948e-05, + "loss": 1.1433, + "step": 886 + }, + { + "epoch": 0.053891487939729026, + "grad_norm": 0.24361053109169006, + "learning_rate": 9.936091190361772e-05, + "loss": 1.1316, + "step": 887 + }, + { + "epoch": 0.053952244972355554, + "grad_norm": 0.44918161630630493, + "learning_rate": 9.93593853454722e-05, + "loss": 1.2746, + "step": 888 + }, + { + "epoch": 0.054013002004982075, + "grad_norm": 0.25405973196029663, + "learning_rate": 9.935785697804887e-05, + "loss": 1.2022, + "step": 889 + }, + { + "epoch": 0.0540737590376086, + "grad_norm": 1.5784674882888794, + "learning_rate": 9.935632680140374e-05, + "loss": 1.1215, + "step": 890 + }, + { + "epoch": 0.05413451607023513, + "grad_norm": 0.3585794270038605, + "learning_rate": 9.935479481559291e-05, + "loss": 1.2935, + "step": 891 + }, + { + "epoch": 0.05419527310286166, + "grad_norm": 0.2580166161060333, + "learning_rate": 9.935326102067255e-05, + "loss": 1.1407, + "step": 892 + }, + { + "epoch": 0.05425603013548818, + "grad_norm": 0.3839646279811859, + "learning_rate": 9.935172541669886e-05, + "loss": 1.1009, + "step": 893 + }, + { + "epoch": 0.05431678716811471, + "grad_norm": 0.23919793963432312, + "learning_rate": 9.935018800372813e-05, + "loss": 1.1166, + "step": 894 + }, + { + "epoch": 0.054377544200741236, + "grad_norm": 0.3730957508087158, + "learning_rate": 9.934864878181673e-05, + "loss": 1.1934, + "step": 895 + }, + { + "epoch": 0.054438301233367764, + "grad_norm": 0.5280106663703918, + "learning_rate": 9.934710775102107e-05, + "loss": 1.2226, + "step": 896 + }, + { + "epoch": 0.05449905826599429, + "grad_norm": 0.3558480143547058, + "learning_rate": 9.934556491139764e-05, + "loss": 1.2481, + "step": 897 + }, + { + "epoch": 0.05455981529862081, + "grad_norm": 0.34314075112342834, + "learning_rate": 9.934402026300298e-05, + "loss": 1.1252, + "step": 898 + }, + { + "epoch": 0.05462057233124734, + "grad_norm": 0.6996009945869446, + "learning_rate": 9.934247380589371e-05, + "loss": 1.2446, + "step": 899 + }, + { + "epoch": 0.05468132936387387, + "grad_norm": 0.5187616944313049, + "learning_rate": 9.934092554012655e-05, + "loss": 1.1259, + "step": 900 + }, + { + "epoch": 0.0547420863965004, + "grad_norm": 0.5063722729682922, + "learning_rate": 9.933937546575822e-05, + "loss": 1.1394, + "step": 901 + }, + { + "epoch": 0.05480284342912692, + "grad_norm": 0.25389522314071655, + "learning_rate": 9.933782358284553e-05, + "loss": 1.1562, + "step": 902 + }, + { + "epoch": 0.054863600461753446, + "grad_norm": 0.442928671836853, + "learning_rate": 9.93362698914454e-05, + "loss": 1.1347, + "step": 903 + }, + { + "epoch": 0.054924357494379974, + "grad_norm": 0.6543065309524536, + "learning_rate": 9.933471439161475e-05, + "loss": 1.1712, + "step": 904 + }, + { + "epoch": 0.0549851145270065, + "grad_norm": 0.3114290237426758, + "learning_rate": 9.93331570834106e-05, + "loss": 1.1145, + "step": 905 + }, + { + "epoch": 0.05504587155963303, + "grad_norm": 0.6922655701637268, + "learning_rate": 9.933159796689005e-05, + "loss": 1.271, + "step": 906 + }, + { + "epoch": 0.05510662859225955, + "grad_norm": 0.2505572736263275, + "learning_rate": 9.933003704211024e-05, + "loss": 1.2036, + "step": 907 + }, + { + "epoch": 0.05516738562488608, + "grad_norm": 0.31724056601524353, + "learning_rate": 9.93284743091284e-05, + "loss": 1.1347, + "step": 908 + }, + { + "epoch": 0.05522814265751261, + "grad_norm": 0.305523544549942, + "learning_rate": 9.932690976800177e-05, + "loss": 1.2089, + "step": 909 + }, + { + "epoch": 0.055288899690139136, + "grad_norm": 0.24788978695869446, + "learning_rate": 9.932534341878774e-05, + "loss": 1.4071, + "step": 910 + }, + { + "epoch": 0.05534965672276566, + "grad_norm": 0.40088146924972534, + "learning_rate": 9.932377526154373e-05, + "loss": 1.2492, + "step": 911 + }, + { + "epoch": 0.055410413755392185, + "grad_norm": 0.24124351143836975, + "learning_rate": 9.932220529632719e-05, + "loss": 1.0832, + "step": 912 + }, + { + "epoch": 0.05547117078801871, + "grad_norm": 0.4385591149330139, + "learning_rate": 9.932063352319567e-05, + "loss": 1.1338, + "step": 913 + }, + { + "epoch": 0.05553192782064524, + "grad_norm": 0.31110692024230957, + "learning_rate": 9.931905994220681e-05, + "loss": 1.2647, + "step": 914 + }, + { + "epoch": 0.05559268485327177, + "grad_norm": 0.29110389947891235, + "learning_rate": 9.931748455341828e-05, + "loss": 1.1266, + "step": 915 + }, + { + "epoch": 0.05565344188589829, + "grad_norm": 0.2130296528339386, + "learning_rate": 9.93159073568878e-05, + "loss": 1.1221, + "step": 916 + }, + { + "epoch": 0.05571419891852482, + "grad_norm": 0.307987779378891, + "learning_rate": 9.931432835267322e-05, + "loss": 1.1944, + "step": 917 + }, + { + "epoch": 0.055774955951151346, + "grad_norm": 0.26360923051834106, + "learning_rate": 9.931274754083239e-05, + "loss": 1.1284, + "step": 918 + }, + { + "epoch": 0.055835712983777874, + "grad_norm": 0.2246204912662506, + "learning_rate": 9.931116492142328e-05, + "loss": 1.14, + "step": 919 + }, + { + "epoch": 0.055896470016404395, + "grad_norm": 0.3153011202812195, + "learning_rate": 9.930958049450389e-05, + "loss": 1.1907, + "step": 920 + }, + { + "epoch": 0.05595722704903092, + "grad_norm": 0.3226519823074341, + "learning_rate": 9.930799426013229e-05, + "loss": 1.1629, + "step": 921 + }, + { + "epoch": 0.05601798408165745, + "grad_norm": 4.871769905090332, + "learning_rate": 9.930640621836662e-05, + "loss": 1.1251, + "step": 922 + }, + { + "epoch": 0.05607874111428398, + "grad_norm": 0.2369663566350937, + "learning_rate": 9.93048163692651e-05, + "loss": 1.1338, + "step": 923 + }, + { + "epoch": 0.05613949814691051, + "grad_norm": 0.21475675702095032, + "learning_rate": 9.930322471288603e-05, + "loss": 1.1108, + "step": 924 + }, + { + "epoch": 0.05620025517953703, + "grad_norm": 0.7375205159187317, + "learning_rate": 9.93016312492877e-05, + "loss": 1.1796, + "step": 925 + }, + { + "epoch": 0.05626101221216356, + "grad_norm": 0.2866756021976471, + "learning_rate": 9.930003597852855e-05, + "loss": 1.1342, + "step": 926 + }, + { + "epoch": 0.056321769244790085, + "grad_norm": 0.3238368034362793, + "learning_rate": 9.929843890066705e-05, + "loss": 1.2786, + "step": 927 + }, + { + "epoch": 0.05638252627741661, + "grad_norm": 0.27854958176612854, + "learning_rate": 9.929684001576176e-05, + "loss": 1.2254, + "step": 928 + }, + { + "epoch": 0.05644328331004314, + "grad_norm": 0.7819719314575195, + "learning_rate": 9.929523932387126e-05, + "loss": 1.104, + "step": 929 + }, + { + "epoch": 0.05650404034266966, + "grad_norm": 0.6707141995429993, + "learning_rate": 9.929363682505424e-05, + "loss": 1.2267, + "step": 930 + }, + { + "epoch": 0.05656479737529619, + "grad_norm": 0.33406177163124084, + "learning_rate": 9.929203251936942e-05, + "loss": 1.1584, + "step": 931 + }, + { + "epoch": 0.05662555440792272, + "grad_norm": 1.2318133115768433, + "learning_rate": 9.929042640687565e-05, + "loss": 1.152, + "step": 932 + }, + { + "epoch": 0.056686311440549246, + "grad_norm": 0.32251033186912537, + "learning_rate": 9.928881848763174e-05, + "loss": 1.2664, + "step": 933 + }, + { + "epoch": 0.05674706847317577, + "grad_norm": 0.2779635488986969, + "learning_rate": 9.928720876169668e-05, + "loss": 1.1253, + "step": 934 + }, + { + "epoch": 0.056807825505802295, + "grad_norm": 1.045859694480896, + "learning_rate": 9.928559722912944e-05, + "loss": 1.1455, + "step": 935 + }, + { + "epoch": 0.05686858253842882, + "grad_norm": 0.2645396292209625, + "learning_rate": 9.928398388998911e-05, + "loss": 1.1332, + "step": 936 + }, + { + "epoch": 0.05692933957105535, + "grad_norm": 0.23510730266571045, + "learning_rate": 9.928236874433485e-05, + "loss": 1.1471, + "step": 937 + }, + { + "epoch": 0.05699009660368188, + "grad_norm": 0.33817562460899353, + "learning_rate": 9.92807517922258e-05, + "loss": 1.205, + "step": 938 + }, + { + "epoch": 0.0570508536363084, + "grad_norm": 0.33143338561058044, + "learning_rate": 9.92791330337213e-05, + "loss": 1.1001, + "step": 939 + }, + { + "epoch": 0.05711161066893493, + "grad_norm": 0.5831192135810852, + "learning_rate": 9.927751246888064e-05, + "loss": 1.4342, + "step": 940 + }, + { + "epoch": 0.057172367701561456, + "grad_norm": 0.23700013756752014, + "learning_rate": 9.927589009776322e-05, + "loss": 1.1112, + "step": 941 + }, + { + "epoch": 0.057233124734187985, + "grad_norm": 0.28254929184913635, + "learning_rate": 9.927426592042854e-05, + "loss": 1.2005, + "step": 942 + }, + { + "epoch": 0.057293881766814506, + "grad_norm": 0.49469679594039917, + "learning_rate": 9.927263993693612e-05, + "loss": 1.1614, + "step": 943 + }, + { + "epoch": 0.057354638799441034, + "grad_norm": 0.4577910900115967, + "learning_rate": 9.927101214734555e-05, + "loss": 1.2383, + "step": 944 + }, + { + "epoch": 0.05741539583206756, + "grad_norm": 0.3149077296257019, + "learning_rate": 9.926938255171652e-05, + "loss": 1.2962, + "step": 945 + }, + { + "epoch": 0.05747615286469409, + "grad_norm": 0.3340935707092285, + "learning_rate": 9.926775115010873e-05, + "loss": 1.2067, + "step": 946 + }, + { + "epoch": 0.05753690989732062, + "grad_norm": 0.33628931641578674, + "learning_rate": 9.9266117942582e-05, + "loss": 1.1518, + "step": 947 + }, + { + "epoch": 0.05759766692994714, + "grad_norm": 0.34515145421028137, + "learning_rate": 9.926448292919622e-05, + "loss": 1.3152, + "step": 948 + }, + { + "epoch": 0.05765842396257367, + "grad_norm": 0.49664726853370667, + "learning_rate": 9.926284611001126e-05, + "loss": 1.2834, + "step": 949 + }, + { + "epoch": 0.057719180995200195, + "grad_norm": 0.9339893460273743, + "learning_rate": 9.926120748508719e-05, + "loss": 1.4244, + "step": 950 + }, + { + "epoch": 0.05777993802782672, + "grad_norm": 0.457669734954834, + "learning_rate": 9.925956705448401e-05, + "loss": 1.1875, + "step": 951 + }, + { + "epoch": 0.057840695060453244, + "grad_norm": 0.5318709015846252, + "learning_rate": 9.925792481826188e-05, + "loss": 1.3121, + "step": 952 + }, + { + "epoch": 0.05790145209307977, + "grad_norm": 0.27497851848602295, + "learning_rate": 9.9256280776481e-05, + "loss": 1.1328, + "step": 953 + }, + { + "epoch": 0.0579622091257063, + "grad_norm": 0.39902961254119873, + "learning_rate": 9.925463492920163e-05, + "loss": 1.1382, + "step": 954 + }, + { + "epoch": 0.05802296615833283, + "grad_norm": 0.2557663023471832, + "learning_rate": 9.92529872764841e-05, + "loss": 1.1682, + "step": 955 + }, + { + "epoch": 0.058083723190959356, + "grad_norm": 0.25504904985427856, + "learning_rate": 9.925133781838878e-05, + "loss": 1.1393, + "step": 956 + }, + { + "epoch": 0.05814448022358588, + "grad_norm": 0.26297393441200256, + "learning_rate": 9.924968655497617e-05, + "loss": 1.1485, + "step": 957 + }, + { + "epoch": 0.058205237256212405, + "grad_norm": 0.2493596076965332, + "learning_rate": 9.924803348630678e-05, + "loss": 1.2115, + "step": 958 + }, + { + "epoch": 0.05826599428883893, + "grad_norm": 0.34783464670181274, + "learning_rate": 9.924637861244121e-05, + "loss": 1.4121, + "step": 959 + }, + { + "epoch": 0.05832675132146546, + "grad_norm": 7.837698936462402, + "learning_rate": 9.92447219334401e-05, + "loss": 1.2931, + "step": 960 + }, + { + "epoch": 0.05838750835409199, + "grad_norm": 0.3985854983329773, + "learning_rate": 9.924306344936418e-05, + "loss": 1.123, + "step": 961 + }, + { + "epoch": 0.05844826538671851, + "grad_norm": 1.068037748336792, + "learning_rate": 9.924140316027426e-05, + "loss": 1.3082, + "step": 962 + }, + { + "epoch": 0.05850902241934504, + "grad_norm": 0.582751989364624, + "learning_rate": 9.92397410662312e-05, + "loss": 1.1889, + "step": 963 + }, + { + "epoch": 0.05856977945197157, + "grad_norm": 0.3472442924976349, + "learning_rate": 9.92380771672959e-05, + "loss": 1.1925, + "step": 964 + }, + { + "epoch": 0.058630536484598095, + "grad_norm": 0.36476099491119385, + "learning_rate": 9.923641146352938e-05, + "loss": 1.2555, + "step": 965 + }, + { + "epoch": 0.058691293517224616, + "grad_norm": 0.38190335035324097, + "learning_rate": 9.923474395499265e-05, + "loss": 1.2831, + "step": 966 + }, + { + "epoch": 0.058752050549851144, + "grad_norm": 0.3635016977787018, + "learning_rate": 9.923307464174689e-05, + "loss": 1.1483, + "step": 967 + }, + { + "epoch": 0.05881280758247767, + "grad_norm": 0.44516876339912415, + "learning_rate": 9.923140352385325e-05, + "loss": 1.1004, + "step": 968 + }, + { + "epoch": 0.0588735646151042, + "grad_norm": 0.21110209822654724, + "learning_rate": 9.922973060137301e-05, + "loss": 1.1749, + "step": 969 + }, + { + "epoch": 0.05893432164773073, + "grad_norm": 0.9567240476608276, + "learning_rate": 9.922805587436748e-05, + "loss": 1.1312, + "step": 970 + }, + { + "epoch": 0.05899507868035725, + "grad_norm": 0.592278242111206, + "learning_rate": 9.922637934289802e-05, + "loss": 1.1406, + "step": 971 + }, + { + "epoch": 0.05905583571298378, + "grad_norm": 0.569057047367096, + "learning_rate": 9.922470100702615e-05, + "loss": 1.1548, + "step": 972 + }, + { + "epoch": 0.059116592745610305, + "grad_norm": 0.3540876507759094, + "learning_rate": 9.922302086681332e-05, + "loss": 1.1628, + "step": 973 + }, + { + "epoch": 0.05917734977823683, + "grad_norm": 0.4148215651512146, + "learning_rate": 9.922133892232115e-05, + "loss": 1.1232, + "step": 974 + }, + { + "epoch": 0.059238106810863354, + "grad_norm": 0.3766629695892334, + "learning_rate": 9.921965517361129e-05, + "loss": 1.1592, + "step": 975 + }, + { + "epoch": 0.05929886384348988, + "grad_norm": 0.8496180772781372, + "learning_rate": 9.921796962074545e-05, + "loss": 1.1192, + "step": 976 + }, + { + "epoch": 0.05935962087611641, + "grad_norm": 0.5288809537887573, + "learning_rate": 9.921628226378543e-05, + "loss": 1.1196, + "step": 977 + }, + { + "epoch": 0.05942037790874294, + "grad_norm": 0.7967683672904968, + "learning_rate": 9.921459310279306e-05, + "loss": 1.2785, + "step": 978 + }, + { + "epoch": 0.059481134941369466, + "grad_norm": 1.4156494140625, + "learning_rate": 9.921290213783027e-05, + "loss": 1.2301, + "step": 979 + }, + { + "epoch": 0.05954189197399599, + "grad_norm": 1.2619818449020386, + "learning_rate": 9.921120936895904e-05, + "loss": 1.1592, + "step": 980 + }, + { + "epoch": 0.059602649006622516, + "grad_norm": 0.5984318256378174, + "learning_rate": 9.920951479624143e-05, + "loss": 1.2059, + "step": 981 + }, + { + "epoch": 0.059663406039249044, + "grad_norm": 0.7489466667175293, + "learning_rate": 9.920781841973955e-05, + "loss": 1.1923, + "step": 982 + }, + { + "epoch": 0.05972416307187557, + "grad_norm": 0.517080545425415, + "learning_rate": 9.920612023951554e-05, + "loss": 1.1189, + "step": 983 + }, + { + "epoch": 0.05978492010450209, + "grad_norm": 0.5091676115989685, + "learning_rate": 9.920442025563171e-05, + "loss": 1.2382, + "step": 984 + }, + { + "epoch": 0.05984567713712862, + "grad_norm": 0.3141508996486664, + "learning_rate": 9.920271846815032e-05, + "loss": 1.1486, + "step": 985 + }, + { + "epoch": 0.05990643416975515, + "grad_norm": 0.3604765236377716, + "learning_rate": 9.92010148771338e-05, + "loss": 1.2423, + "step": 986 + }, + { + "epoch": 0.05996719120238168, + "grad_norm": 0.527212917804718, + "learning_rate": 9.919930948264456e-05, + "loss": 1.1487, + "step": 987 + }, + { + "epoch": 0.060027948235008205, + "grad_norm": 0.24980883300304413, + "learning_rate": 9.91976022847451e-05, + "loss": 1.1104, + "step": 988 + }, + { + "epoch": 0.060088705267634726, + "grad_norm": 0.4576713740825653, + "learning_rate": 9.919589328349805e-05, + "loss": 1.2279, + "step": 989 + }, + { + "epoch": 0.060149462300261254, + "grad_norm": 0.39471200108528137, + "learning_rate": 9.919418247896602e-05, + "loss": 1.3409, + "step": 990 + }, + { + "epoch": 0.06021021933288778, + "grad_norm": 0.23512251675128937, + "learning_rate": 9.919246987121169e-05, + "loss": 1.1638, + "step": 991 + }, + { + "epoch": 0.06027097636551431, + "grad_norm": 0.22036854922771454, + "learning_rate": 9.91907554602979e-05, + "loss": 1.0745, + "step": 992 + }, + { + "epoch": 0.06033173339814084, + "grad_norm": 0.33566814661026, + "learning_rate": 9.918903924628744e-05, + "loss": 1.1628, + "step": 993 + }, + { + "epoch": 0.06039249043076736, + "grad_norm": 2.704174518585205, + "learning_rate": 9.918732122924326e-05, + "loss": 1.208, + "step": 994 + }, + { + "epoch": 0.06045324746339389, + "grad_norm": 0.25995364785194397, + "learning_rate": 9.91856014092283e-05, + "loss": 1.1346, + "step": 995 + }, + { + "epoch": 0.060514004496020415, + "grad_norm": 0.29886284470558167, + "learning_rate": 9.91838797863056e-05, + "loss": 1.1976, + "step": 996 + }, + { + "epoch": 0.06057476152864694, + "grad_norm": 0.5589959621429443, + "learning_rate": 9.91821563605383e-05, + "loss": 1.104, + "step": 997 + }, + { + "epoch": 0.060635518561273465, + "grad_norm": 0.7914170026779175, + "learning_rate": 9.918043113198955e-05, + "loss": 1.3317, + "step": 998 + }, + { + "epoch": 0.06069627559389999, + "grad_norm": 0.27024585008621216, + "learning_rate": 9.917870410072258e-05, + "loss": 1.231, + "step": 999 + }, + { + "epoch": 0.06075703262652652, + "grad_norm": 0.42895710468292236, + "learning_rate": 9.917697526680072e-05, + "loss": 1.1647, + "step": 1000 + }, + { + "epoch": 0.06081778965915305, + "grad_norm": 0.20736272633075714, + "learning_rate": 9.917524463028731e-05, + "loss": 1.1133, + "step": 1001 + }, + { + "epoch": 0.06087854669177958, + "grad_norm": 0.2284463793039322, + "learning_rate": 9.91735121912458e-05, + "loss": 1.1467, + "step": 1002 + }, + { + "epoch": 0.0609393037244061, + "grad_norm": 0.6051624417304993, + "learning_rate": 9.91717779497397e-05, + "loss": 1.1571, + "step": 1003 + }, + { + "epoch": 0.061000060757032626, + "grad_norm": 0.23619496822357178, + "learning_rate": 9.917004190583259e-05, + "loss": 1.0975, + "step": 1004 + }, + { + "epoch": 0.061060817789659154, + "grad_norm": 0.5558038353919983, + "learning_rate": 9.916830405958808e-05, + "loss": 1.1497, + "step": 1005 + }, + { + "epoch": 0.06112157482228568, + "grad_norm": 0.5257781147956848, + "learning_rate": 9.916656441106989e-05, + "loss": 1.2697, + "step": 1006 + }, + { + "epoch": 0.0611823318549122, + "grad_norm": 0.2926448881626129, + "learning_rate": 9.916482296034175e-05, + "loss": 1.1957, + "step": 1007 + }, + { + "epoch": 0.06124308888753873, + "grad_norm": 0.36588895320892334, + "learning_rate": 9.916307970746754e-05, + "loss": 1.2183, + "step": 1008 + }, + { + "epoch": 0.06130384592016526, + "grad_norm": 0.24381615221500397, + "learning_rate": 9.916133465251114e-05, + "loss": 1.1083, + "step": 1009 + }, + { + "epoch": 0.06136460295279179, + "grad_norm": 0.4629080593585968, + "learning_rate": 9.915958779553652e-05, + "loss": 1.1365, + "step": 1010 + }, + { + "epoch": 0.061425359985418315, + "grad_norm": 0.3053455352783203, + "learning_rate": 9.91578391366077e-05, + "loss": 1.2418, + "step": 1011 + }, + { + "epoch": 0.061486117018044836, + "grad_norm": 0.33790913224220276, + "learning_rate": 9.915608867578878e-05, + "loss": 1.1186, + "step": 1012 + }, + { + "epoch": 0.061546874050671364, + "grad_norm": 0.25527098774909973, + "learning_rate": 9.915433641314394e-05, + "loss": 1.162, + "step": 1013 + }, + { + "epoch": 0.06160763108329789, + "grad_norm": 0.3676719069480896, + "learning_rate": 9.915258234873739e-05, + "loss": 1.1656, + "step": 1014 + }, + { + "epoch": 0.06166838811592442, + "grad_norm": 0.21692173182964325, + "learning_rate": 9.915082648263343e-05, + "loss": 1.1224, + "step": 1015 + }, + { + "epoch": 0.06172914514855094, + "grad_norm": 0.2588801980018616, + "learning_rate": 9.914906881489643e-05, + "loss": 1.272, + "step": 1016 + }, + { + "epoch": 0.06178990218117747, + "grad_norm": 0.2632429301738739, + "learning_rate": 9.914730934559083e-05, + "loss": 1.103, + "step": 1017 + }, + { + "epoch": 0.061850659213804, + "grad_norm": 0.3593953549861908, + "learning_rate": 9.914554807478108e-05, + "loss": 1.1473, + "step": 1018 + }, + { + "epoch": 0.061911416246430526, + "grad_norm": 0.2675398588180542, + "learning_rate": 9.914378500253176e-05, + "loss": 1.2444, + "step": 1019 + }, + { + "epoch": 0.061972173279057054, + "grad_norm": 0.24753326177597046, + "learning_rate": 9.914202012890751e-05, + "loss": 1.1478, + "step": 1020 + }, + { + "epoch": 0.062032930311683575, + "grad_norm": 0.2554571032524109, + "learning_rate": 9.914025345397304e-05, + "loss": 1.1338, + "step": 1021 + }, + { + "epoch": 0.0620936873443101, + "grad_norm": 0.25916150212287903, + "learning_rate": 9.913848497779306e-05, + "loss": 1.1358, + "step": 1022 + }, + { + "epoch": 0.06215444437693663, + "grad_norm": 0.288501501083374, + "learning_rate": 9.91367147004324e-05, + "loss": 1.1834, + "step": 1023 + }, + { + "epoch": 0.06221520140956316, + "grad_norm": 0.5686520934104919, + "learning_rate": 9.913494262195597e-05, + "loss": 1.2223, + "step": 1024 + }, + { + "epoch": 0.06227595844218968, + "grad_norm": 0.8924007415771484, + "learning_rate": 9.913316874242874e-05, + "loss": 1.1762, + "step": 1025 + }, + { + "epoch": 0.06233671547481621, + "grad_norm": 0.43462514877319336, + "learning_rate": 9.91313930619157e-05, + "loss": 1.2431, + "step": 1026 + }, + { + "epoch": 0.062397472507442736, + "grad_norm": 0.37436816096305847, + "learning_rate": 9.912961558048195e-05, + "loss": 1.139, + "step": 1027 + }, + { + "epoch": 0.062458229540069264, + "grad_norm": 0.43720147013664246, + "learning_rate": 9.912783629819264e-05, + "loss": 1.2857, + "step": 1028 + }, + { + "epoch": 0.06251898657269579, + "grad_norm": 0.6083366870880127, + "learning_rate": 9.912605521511301e-05, + "loss": 1.3171, + "step": 1029 + }, + { + "epoch": 0.06257974360532231, + "grad_norm": 3.9070780277252197, + "learning_rate": 9.912427233130832e-05, + "loss": 1.1901, + "step": 1030 + }, + { + "epoch": 0.06264050063794885, + "grad_norm": 0.6451282501220703, + "learning_rate": 9.912248764684393e-05, + "loss": 1.1871, + "step": 1031 + }, + { + "epoch": 0.06270125767057537, + "grad_norm": 0.42376747727394104, + "learning_rate": 9.912070116178526e-05, + "loss": 1.0873, + "step": 1032 + }, + { + "epoch": 0.06276201470320189, + "grad_norm": 0.3612758219242096, + "learning_rate": 9.91189128761978e-05, + "loss": 1.1476, + "step": 1033 + }, + { + "epoch": 0.06282277173582843, + "grad_norm": 0.4746432304382324, + "learning_rate": 9.911712279014709e-05, + "loss": 1.1682, + "step": 1034 + }, + { + "epoch": 0.06288352876845495, + "grad_norm": 0.21555911004543304, + "learning_rate": 9.911533090369876e-05, + "loss": 1.3292, + "step": 1035 + }, + { + "epoch": 0.06294428580108148, + "grad_norm": 0.3674027919769287, + "learning_rate": 9.911353721691848e-05, + "loss": 1.1569, + "step": 1036 + }, + { + "epoch": 0.063005042833708, + "grad_norm": 0.21336917579174042, + "learning_rate": 9.911174172987199e-05, + "loss": 1.1183, + "step": 1037 + }, + { + "epoch": 0.06306579986633452, + "grad_norm": 0.29123494029045105, + "learning_rate": 9.910994444262512e-05, + "loss": 1.2318, + "step": 1038 + }, + { + "epoch": 0.06312655689896106, + "grad_norm": 0.19978398084640503, + "learning_rate": 9.910814535524372e-05, + "loss": 1.1266, + "step": 1039 + }, + { + "epoch": 0.06318731393158758, + "grad_norm": 2.1680283546447754, + "learning_rate": 9.910634446779379e-05, + "loss": 1.1288, + "step": 1040 + }, + { + "epoch": 0.0632480709642141, + "grad_norm": 0.2583906650543213, + "learning_rate": 9.91045417803413e-05, + "loss": 1.1554, + "step": 1041 + }, + { + "epoch": 0.06330882799684064, + "grad_norm": 0.4378766119480133, + "learning_rate": 9.910273729295235e-05, + "loss": 1.1133, + "step": 1042 + }, + { + "epoch": 0.06336958502946716, + "grad_norm": 0.25887638330459595, + "learning_rate": 9.910093100569305e-05, + "loss": 1.1943, + "step": 1043 + }, + { + "epoch": 0.06343034206209369, + "grad_norm": 20.409955978393555, + "learning_rate": 9.909912291862966e-05, + "loss": 1.1126, + "step": 1044 + }, + { + "epoch": 0.06349109909472021, + "grad_norm": 0.36200186610221863, + "learning_rate": 9.909731303182841e-05, + "loss": 1.1319, + "step": 1045 + }, + { + "epoch": 0.06355185612734673, + "grad_norm": 0.40067556500434875, + "learning_rate": 9.909550134535566e-05, + "loss": 1.136, + "step": 1046 + }, + { + "epoch": 0.06361261315997327, + "grad_norm": 0.5778839588165283, + "learning_rate": 9.909368785927783e-05, + "loss": 1.4697, + "step": 1047 + }, + { + "epoch": 0.06367337019259979, + "grad_norm": 0.392610639333725, + "learning_rate": 9.909187257366137e-05, + "loss": 1.1973, + "step": 1048 + }, + { + "epoch": 0.06373412722522633, + "grad_norm": 0.2898315191268921, + "learning_rate": 9.909005548857284e-05, + "loss": 1.1465, + "step": 1049 + }, + { + "epoch": 0.06379488425785285, + "grad_norm": 0.3183806836605072, + "learning_rate": 9.908823660407884e-05, + "loss": 1.1192, + "step": 1050 + }, + { + "epoch": 0.06385564129047937, + "grad_norm": 0.4587804079055786, + "learning_rate": 9.908641592024603e-05, + "loss": 1.2059, + "step": 1051 + }, + { + "epoch": 0.0639163983231059, + "grad_norm": 0.317844957113266, + "learning_rate": 9.908459343714116e-05, + "loss": 1.145, + "step": 1052 + }, + { + "epoch": 0.06397715535573242, + "grad_norm": 0.4280650317668915, + "learning_rate": 9.908276915483104e-05, + "loss": 1.1027, + "step": 1053 + }, + { + "epoch": 0.06403791238835896, + "grad_norm": 0.5388709306716919, + "learning_rate": 9.908094307338251e-05, + "loss": 1.2059, + "step": 1054 + }, + { + "epoch": 0.06409866942098548, + "grad_norm": 0.6156328916549683, + "learning_rate": 9.907911519286254e-05, + "loss": 1.1506, + "step": 1055 + }, + { + "epoch": 0.064159426453612, + "grad_norm": 0.5336007475852966, + "learning_rate": 9.90772855133381e-05, + "loss": 1.1181, + "step": 1056 + }, + { + "epoch": 0.06422018348623854, + "grad_norm": 1.1483440399169922, + "learning_rate": 9.907545403487629e-05, + "loss": 1.2459, + "step": 1057 + }, + { + "epoch": 0.06428094051886506, + "grad_norm": 0.919160008430481, + "learning_rate": 9.907362075754422e-05, + "loss": 1.1359, + "step": 1058 + }, + { + "epoch": 0.06434169755149159, + "grad_norm": 0.5912401676177979, + "learning_rate": 9.907178568140909e-05, + "loss": 1.1345, + "step": 1059 + }, + { + "epoch": 0.06440245458411811, + "grad_norm": 0.2012874186038971, + "learning_rate": 9.906994880653818e-05, + "loss": 1.1277, + "step": 1060 + }, + { + "epoch": 0.06446321161674463, + "grad_norm": 0.38084134459495544, + "learning_rate": 9.90681101329988e-05, + "loss": 1.3282, + "step": 1061 + }, + { + "epoch": 0.06452396864937117, + "grad_norm": 0.37889721989631653, + "learning_rate": 9.906626966085837e-05, + "loss": 1.1228, + "step": 1062 + }, + { + "epoch": 0.06458472568199769, + "grad_norm": 0.7509076595306396, + "learning_rate": 9.906442739018433e-05, + "loss": 1.1452, + "step": 1063 + }, + { + "epoch": 0.06464548271462421, + "grad_norm": 0.28603652119636536, + "learning_rate": 9.906258332104424e-05, + "loss": 1.1307, + "step": 1064 + }, + { + "epoch": 0.06470623974725075, + "grad_norm": 0.3042452037334442, + "learning_rate": 9.906073745350565e-05, + "loss": 1.2067, + "step": 1065 + }, + { + "epoch": 0.06476699677987727, + "grad_norm": 0.34387460350990295, + "learning_rate": 9.905888978763627e-05, + "loss": 1.1801, + "step": 1066 + }, + { + "epoch": 0.0648277538125038, + "grad_norm": 0.3155616521835327, + "learning_rate": 9.905704032350378e-05, + "loss": 1.1804, + "step": 1067 + }, + { + "epoch": 0.06488851084513032, + "grad_norm": 0.29974400997161865, + "learning_rate": 9.9055189061176e-05, + "loss": 1.1386, + "step": 1068 + }, + { + "epoch": 0.06494926787775684, + "grad_norm": 0.24127930402755737, + "learning_rate": 9.905333600072079e-05, + "loss": 1.1107, + "step": 1069 + }, + { + "epoch": 0.06501002491038338, + "grad_norm": 5.226344108581543, + "learning_rate": 9.905148114220606e-05, + "loss": 1.2696, + "step": 1070 + }, + { + "epoch": 0.0650707819430099, + "grad_norm": 0.5176244378089905, + "learning_rate": 9.904962448569981e-05, + "loss": 1.1506, + "step": 1071 + }, + { + "epoch": 0.06513153897563644, + "grad_norm": 0.26255178451538086, + "learning_rate": 9.904776603127009e-05, + "loss": 1.1497, + "step": 1072 + }, + { + "epoch": 0.06519229600826296, + "grad_norm": 0.4290417730808258, + "learning_rate": 9.904590577898503e-05, + "loss": 1.1632, + "step": 1073 + }, + { + "epoch": 0.06525305304088948, + "grad_norm": 0.24741199612617493, + "learning_rate": 9.904404372891281e-05, + "loss": 1.0932, + "step": 1074 + }, + { + "epoch": 0.06531381007351601, + "grad_norm": 1.419188141822815, + "learning_rate": 9.90421798811217e-05, + "loss": 1.1602, + "step": 1075 + }, + { + "epoch": 0.06537456710614253, + "grad_norm": 0.46958374977111816, + "learning_rate": 9.904031423567998e-05, + "loss": 1.1322, + "step": 1076 + }, + { + "epoch": 0.06543532413876907, + "grad_norm": 0.21372784674167633, + "learning_rate": 9.903844679265608e-05, + "loss": 1.1069, + "step": 1077 + }, + { + "epoch": 0.06549608117139559, + "grad_norm": 0.3436409831047058, + "learning_rate": 9.903657755211844e-05, + "loss": 1.2349, + "step": 1078 + }, + { + "epoch": 0.06555683820402211, + "grad_norm": 0.3673514127731323, + "learning_rate": 9.903470651413555e-05, + "loss": 1.2395, + "step": 1079 + }, + { + "epoch": 0.06561759523664865, + "grad_norm": 0.3007248044013977, + "learning_rate": 9.903283367877603e-05, + "loss": 1.1648, + "step": 1080 + }, + { + "epoch": 0.06567835226927517, + "grad_norm": 0.32089921832084656, + "learning_rate": 9.903095904610851e-05, + "loss": 1.1609, + "step": 1081 + }, + { + "epoch": 0.0657391093019017, + "grad_norm": 0.8967872262001038, + "learning_rate": 9.902908261620172e-05, + "loss": 1.1531, + "step": 1082 + }, + { + "epoch": 0.06579986633452822, + "grad_norm": 0.608903169631958, + "learning_rate": 9.902720438912441e-05, + "loss": 1.339, + "step": 1083 + }, + { + "epoch": 0.06586062336715474, + "grad_norm": 0.36996495723724365, + "learning_rate": 9.902532436494547e-05, + "loss": 1.1063, + "step": 1084 + }, + { + "epoch": 0.06592138039978128, + "grad_norm": 0.31174659729003906, + "learning_rate": 9.902344254373377e-05, + "loss": 1.1699, + "step": 1085 + }, + { + "epoch": 0.0659821374324078, + "grad_norm": 1.4040777683258057, + "learning_rate": 9.902155892555831e-05, + "loss": 1.2781, + "step": 1086 + }, + { + "epoch": 0.06604289446503432, + "grad_norm": 0.2816430628299713, + "learning_rate": 9.901967351048813e-05, + "loss": 1.1989, + "step": 1087 + }, + { + "epoch": 0.06610365149766086, + "grad_norm": 0.5177039504051208, + "learning_rate": 9.901778629859235e-05, + "loss": 1.1749, + "step": 1088 + }, + { + "epoch": 0.06616440853028738, + "grad_norm": 0.29074007272720337, + "learning_rate": 9.901589728994014e-05, + "loss": 1.2069, + "step": 1089 + }, + { + "epoch": 0.06622516556291391, + "grad_norm": 0.3605645000934601, + "learning_rate": 9.901400648460074e-05, + "loss": 1.1949, + "step": 1090 + }, + { + "epoch": 0.06628592259554043, + "grad_norm": 0.42420080304145813, + "learning_rate": 9.901211388264345e-05, + "loss": 1.1805, + "step": 1091 + }, + { + "epoch": 0.06634667962816695, + "grad_norm": 0.3319961428642273, + "learning_rate": 9.901021948413765e-05, + "loss": 1.1434, + "step": 1092 + }, + { + "epoch": 0.06640743666079349, + "grad_norm": 0.3663233816623688, + "learning_rate": 9.90083232891528e-05, + "loss": 1.1441, + "step": 1093 + }, + { + "epoch": 0.06646819369342001, + "grad_norm": 0.45285674929618835, + "learning_rate": 9.900642529775836e-05, + "loss": 1.1275, + "step": 1094 + }, + { + "epoch": 0.06652895072604655, + "grad_norm": 0.3910238742828369, + "learning_rate": 9.900452551002395e-05, + "loss": 1.4843, + "step": 1095 + }, + { + "epoch": 0.06658970775867307, + "grad_norm": 0.4343225061893463, + "learning_rate": 9.900262392601918e-05, + "loss": 1.2555, + "step": 1096 + }, + { + "epoch": 0.06665046479129959, + "grad_norm": 0.662329912185669, + "learning_rate": 9.900072054581375e-05, + "loss": 1.15, + "step": 1097 + }, + { + "epoch": 0.06671122182392612, + "grad_norm": 0.3869973421096802, + "learning_rate": 9.899881536947744e-05, + "loss": 1.3056, + "step": 1098 + }, + { + "epoch": 0.06677197885655264, + "grad_norm": 0.548058271408081, + "learning_rate": 9.899690839708008e-05, + "loss": 1.2469, + "step": 1099 + }, + { + "epoch": 0.06683273588917918, + "grad_norm": 0.29039615392684937, + "learning_rate": 9.899499962869157e-05, + "loss": 1.2345, + "step": 1100 + }, + { + "epoch": 0.0668934929218057, + "grad_norm": 0.5887446999549866, + "learning_rate": 9.899308906438189e-05, + "loss": 1.2276, + "step": 1101 + }, + { + "epoch": 0.06695424995443222, + "grad_norm": 0.6240502595901489, + "learning_rate": 9.899117670422104e-05, + "loss": 1.1269, + "step": 1102 + }, + { + "epoch": 0.06701500698705876, + "grad_norm": 0.5116984248161316, + "learning_rate": 9.898926254827916e-05, + "loss": 1.2276, + "step": 1103 + }, + { + "epoch": 0.06707576401968528, + "grad_norm": 0.657288134098053, + "learning_rate": 9.898734659662638e-05, + "loss": 1.1391, + "step": 1104 + }, + { + "epoch": 0.0671365210523118, + "grad_norm": 0.4482358396053314, + "learning_rate": 9.898542884933295e-05, + "loss": 1.1686, + "step": 1105 + }, + { + "epoch": 0.06719727808493833, + "grad_norm": 0.6678813695907593, + "learning_rate": 9.898350930646916e-05, + "loss": 1.2276, + "step": 1106 + }, + { + "epoch": 0.06725803511756485, + "grad_norm": 0.3689097464084625, + "learning_rate": 9.898158796810535e-05, + "loss": 1.141, + "step": 1107 + }, + { + "epoch": 0.06731879215019139, + "grad_norm": 0.37625017762184143, + "learning_rate": 9.897966483431198e-05, + "loss": 1.1693, + "step": 1108 + }, + { + "epoch": 0.06737954918281791, + "grad_norm": 0.5635063052177429, + "learning_rate": 9.897773990515952e-05, + "loss": 1.2039, + "step": 1109 + }, + { + "epoch": 0.06744030621544443, + "grad_norm": 0.2624109387397766, + "learning_rate": 9.897581318071853e-05, + "loss": 1.1171, + "step": 1110 + }, + { + "epoch": 0.06750106324807097, + "grad_norm": 0.35183796286582947, + "learning_rate": 9.897388466105966e-05, + "loss": 1.2073, + "step": 1111 + }, + { + "epoch": 0.06756182028069749, + "grad_norm": 0.5033414959907532, + "learning_rate": 9.897195434625356e-05, + "loss": 1.2092, + "step": 1112 + }, + { + "epoch": 0.06762257731332402, + "grad_norm": 0.5837149620056152, + "learning_rate": 9.897002223637103e-05, + "loss": 1.1279, + "step": 1113 + }, + { + "epoch": 0.06768333434595054, + "grad_norm": 0.33090782165527344, + "learning_rate": 9.896808833148285e-05, + "loss": 1.1263, + "step": 1114 + }, + { + "epoch": 0.06774409137857706, + "grad_norm": 0.29453933238983154, + "learning_rate": 9.896615263165994e-05, + "loss": 1.1052, + "step": 1115 + }, + { + "epoch": 0.0678048484112036, + "grad_norm": 0.5820531845092773, + "learning_rate": 9.896421513697323e-05, + "loss": 1.1732, + "step": 1116 + }, + { + "epoch": 0.06786560544383012, + "grad_norm": 0.7342537045478821, + "learning_rate": 9.896227584749375e-05, + "loss": 1.3774, + "step": 1117 + }, + { + "epoch": 0.06792636247645666, + "grad_norm": 0.31015297770500183, + "learning_rate": 9.89603347632926e-05, + "loss": 1.1419, + "step": 1118 + }, + { + "epoch": 0.06798711950908318, + "grad_norm": 0.5793747305870056, + "learning_rate": 9.89583918844409e-05, + "loss": 1.2077, + "step": 1119 + }, + { + "epoch": 0.0680478765417097, + "grad_norm": 0.36365461349487305, + "learning_rate": 9.895644721100989e-05, + "loss": 1.3128, + "step": 1120 + }, + { + "epoch": 0.06810863357433623, + "grad_norm": 0.44211041927337646, + "learning_rate": 9.895450074307084e-05, + "loss": 1.0978, + "step": 1121 + }, + { + "epoch": 0.06816939060696275, + "grad_norm": 0.36834749579429626, + "learning_rate": 9.89525524806951e-05, + "loss": 1.1553, + "step": 1122 + }, + { + "epoch": 0.06823014763958929, + "grad_norm": 0.28230908513069153, + "learning_rate": 9.895060242395408e-05, + "loss": 1.1409, + "step": 1123 + }, + { + "epoch": 0.06829090467221581, + "grad_norm": 1.5769546031951904, + "learning_rate": 9.894865057291928e-05, + "loss": 1.1625, + "step": 1124 + }, + { + "epoch": 0.06835166170484233, + "grad_norm": 0.29780131578445435, + "learning_rate": 9.894669692766223e-05, + "loss": 1.2598, + "step": 1125 + }, + { + "epoch": 0.06841241873746887, + "grad_norm": 0.31186074018478394, + "learning_rate": 9.894474148825454e-05, + "loss": 1.1401, + "step": 1126 + }, + { + "epoch": 0.06847317577009539, + "grad_norm": 0.567206859588623, + "learning_rate": 9.89427842547679e-05, + "loss": 1.2253, + "step": 1127 + }, + { + "epoch": 0.06853393280272191, + "grad_norm": 0.31455615162849426, + "learning_rate": 9.894082522727403e-05, + "loss": 1.1444, + "step": 1128 + }, + { + "epoch": 0.06859468983534844, + "grad_norm": 1.0472739934921265, + "learning_rate": 9.893886440584476e-05, + "loss": 1.2112, + "step": 1129 + }, + { + "epoch": 0.06865544686797496, + "grad_norm": 0.5503324866294861, + "learning_rate": 9.893690179055194e-05, + "loss": 1.0981, + "step": 1130 + }, + { + "epoch": 0.0687162039006015, + "grad_norm": 0.3120340406894684, + "learning_rate": 9.893493738146754e-05, + "loss": 1.25, + "step": 1131 + }, + { + "epoch": 0.06877696093322802, + "grad_norm": 0.6391900181770325, + "learning_rate": 9.893297117866355e-05, + "loss": 1.1456, + "step": 1132 + }, + { + "epoch": 0.06883771796585454, + "grad_norm": 0.263694703578949, + "learning_rate": 9.893100318221204e-05, + "loss": 1.15, + "step": 1133 + }, + { + "epoch": 0.06889847499848108, + "grad_norm": 0.7932913303375244, + "learning_rate": 9.892903339218515e-05, + "loss": 1.1191, + "step": 1134 + }, + { + "epoch": 0.0689592320311076, + "grad_norm": 0.3033735752105713, + "learning_rate": 9.892706180865509e-05, + "loss": 1.1517, + "step": 1135 + }, + { + "epoch": 0.06901998906373413, + "grad_norm": 0.4355921149253845, + "learning_rate": 9.89250884316941e-05, + "loss": 1.2179, + "step": 1136 + }, + { + "epoch": 0.06908074609636065, + "grad_norm": 0.25827935338020325, + "learning_rate": 9.892311326137455e-05, + "loss": 1.1252, + "step": 1137 + }, + { + "epoch": 0.06914150312898718, + "grad_norm": 0.5501392483711243, + "learning_rate": 9.892113629776884e-05, + "loss": 1.2907, + "step": 1138 + }, + { + "epoch": 0.06920226016161371, + "grad_norm": 0.4540093243122101, + "learning_rate": 9.891915754094941e-05, + "loss": 1.126, + "step": 1139 + }, + { + "epoch": 0.06926301719424023, + "grad_norm": 0.2849162817001343, + "learning_rate": 9.891717699098882e-05, + "loss": 1.3377, + "step": 1140 + }, + { + "epoch": 0.06932377422686677, + "grad_norm": 0.27934205532073975, + "learning_rate": 9.891519464795961e-05, + "loss": 1.1084, + "step": 1141 + }, + { + "epoch": 0.06938453125949329, + "grad_norm": 0.35017848014831543, + "learning_rate": 9.891321051193452e-05, + "loss": 1.2137, + "step": 1142 + }, + { + "epoch": 0.06944528829211981, + "grad_norm": 0.7662164568901062, + "learning_rate": 9.891122458298626e-05, + "loss": 1.2956, + "step": 1143 + }, + { + "epoch": 0.06950604532474634, + "grad_norm": 0.45953139662742615, + "learning_rate": 9.890923686118758e-05, + "loss": 1.0901, + "step": 1144 + }, + { + "epoch": 0.06956680235737286, + "grad_norm": 0.33483320474624634, + "learning_rate": 9.890724734661137e-05, + "loss": 1.1515, + "step": 1145 + }, + { + "epoch": 0.06962755938999939, + "grad_norm": 0.32463687658309937, + "learning_rate": 9.890525603933057e-05, + "loss": 1.1867, + "step": 1146 + }, + { + "epoch": 0.06968831642262592, + "grad_norm": 0.3187297582626343, + "learning_rate": 9.890326293941813e-05, + "loss": 1.0984, + "step": 1147 + }, + { + "epoch": 0.06974907345525244, + "grad_norm": 0.7814391255378723, + "learning_rate": 9.890126804694715e-05, + "loss": 1.3098, + "step": 1148 + }, + { + "epoch": 0.06980983048787898, + "grad_norm": 0.3355152904987335, + "learning_rate": 9.889927136199074e-05, + "loss": 1.3312, + "step": 1149 + }, + { + "epoch": 0.0698705875205055, + "grad_norm": 0.6966271996498108, + "learning_rate": 9.889727288462208e-05, + "loss": 1.1618, + "step": 1150 + }, + { + "epoch": 0.06993134455313202, + "grad_norm": 0.31015637516975403, + "learning_rate": 9.889527261491444e-05, + "loss": 1.1329, + "step": 1151 + }, + { + "epoch": 0.06999210158575855, + "grad_norm": 0.5841100811958313, + "learning_rate": 9.889327055294112e-05, + "loss": 1.3558, + "step": 1152 + }, + { + "epoch": 0.07005285861838507, + "grad_norm": 0.30163145065307617, + "learning_rate": 9.889126669877553e-05, + "loss": 1.1085, + "step": 1153 + }, + { + "epoch": 0.07011361565101161, + "grad_norm": 1.042690396308899, + "learning_rate": 9.88892610524911e-05, + "loss": 1.1298, + "step": 1154 + }, + { + "epoch": 0.07017437268363813, + "grad_norm": 0.39691975712776184, + "learning_rate": 9.888725361416136e-05, + "loss": 1.3098, + "step": 1155 + }, + { + "epoch": 0.07023512971626465, + "grad_norm": 0.3143743872642517, + "learning_rate": 9.888524438385988e-05, + "loss": 1.1937, + "step": 1156 + }, + { + "epoch": 0.07029588674889119, + "grad_norm": 0.2704367935657501, + "learning_rate": 9.888323336166033e-05, + "loss": 1.2537, + "step": 1157 + }, + { + "epoch": 0.07035664378151771, + "grad_norm": 0.2279391586780548, + "learning_rate": 9.88812205476364e-05, + "loss": 1.0964, + "step": 1158 + }, + { + "epoch": 0.07041740081414424, + "grad_norm": 0.3239094018936157, + "learning_rate": 9.88792059418619e-05, + "loss": 1.1916, + "step": 1159 + }, + { + "epoch": 0.07047815784677076, + "grad_norm": 0.3912469744682312, + "learning_rate": 9.887718954441064e-05, + "loss": 1.2866, + "step": 1160 + }, + { + "epoch": 0.07053891487939729, + "grad_norm": 0.24745208024978638, + "learning_rate": 9.887517135535656e-05, + "loss": 1.0948, + "step": 1161 + }, + { + "epoch": 0.07059967191202382, + "grad_norm": 0.21561789512634277, + "learning_rate": 9.887315137477362e-05, + "loss": 1.1644, + "step": 1162 + }, + { + "epoch": 0.07066042894465034, + "grad_norm": 0.3207648992538452, + "learning_rate": 9.887112960273587e-05, + "loss": 1.1154, + "step": 1163 + }, + { + "epoch": 0.07072118597727688, + "grad_norm": 0.2158115953207016, + "learning_rate": 9.886910603931743e-05, + "loss": 1.0976, + "step": 1164 + }, + { + "epoch": 0.0707819430099034, + "grad_norm": 0.2824103534221649, + "learning_rate": 9.886708068459244e-05, + "loss": 1.1908, + "step": 1165 + }, + { + "epoch": 0.07084270004252992, + "grad_norm": 0.5944103598594666, + "learning_rate": 9.886505353863518e-05, + "loss": 1.1574, + "step": 1166 + }, + { + "epoch": 0.07090345707515645, + "grad_norm": 0.3784528076648712, + "learning_rate": 9.886302460151992e-05, + "loss": 1.2257, + "step": 1167 + }, + { + "epoch": 0.07096421410778297, + "grad_norm": 0.45542946457862854, + "learning_rate": 9.886099387332107e-05, + "loss": 1.4517, + "step": 1168 + }, + { + "epoch": 0.0710249711404095, + "grad_norm": 0.31356704235076904, + "learning_rate": 9.885896135411304e-05, + "loss": 1.1931, + "step": 1169 + }, + { + "epoch": 0.07108572817303603, + "grad_norm": 0.419672429561615, + "learning_rate": 9.885692704397032e-05, + "loss": 1.2209, + "step": 1170 + }, + { + "epoch": 0.07114648520566255, + "grad_norm": 0.7691190838813782, + "learning_rate": 9.885489094296751e-05, + "loss": 1.2641, + "step": 1171 + }, + { + "epoch": 0.07120724223828909, + "grad_norm": 0.2246992588043213, + "learning_rate": 9.885285305117922e-05, + "loss": 1.2269, + "step": 1172 + }, + { + "epoch": 0.07126799927091561, + "grad_norm": 0.37041518092155457, + "learning_rate": 9.885081336868018e-05, + "loss": 1.2188, + "step": 1173 + }, + { + "epoch": 0.07132875630354213, + "grad_norm": 1.391851544380188, + "learning_rate": 9.884877189554511e-05, + "loss": 1.1768, + "step": 1174 + }, + { + "epoch": 0.07138951333616866, + "grad_norm": 0.34428805112838745, + "learning_rate": 9.884672863184888e-05, + "loss": 1.2878, + "step": 1175 + }, + { + "epoch": 0.07145027036879519, + "grad_norm": 0.26112544536590576, + "learning_rate": 9.884468357766636e-05, + "loss": 1.1894, + "step": 1176 + }, + { + "epoch": 0.07151102740142172, + "grad_norm": 0.3130583167076111, + "learning_rate": 9.884263673307253e-05, + "loss": 1.1949, + "step": 1177 + }, + { + "epoch": 0.07157178443404824, + "grad_norm": 0.23426231741905212, + "learning_rate": 9.884058809814239e-05, + "loss": 1.1768, + "step": 1178 + }, + { + "epoch": 0.07163254146667476, + "grad_norm": 0.34472355246543884, + "learning_rate": 9.883853767295108e-05, + "loss": 1.1986, + "step": 1179 + }, + { + "epoch": 0.0716932984993013, + "grad_norm": 1.8434522151947021, + "learning_rate": 9.883648545757372e-05, + "loss": 1.1627, + "step": 1180 + }, + { + "epoch": 0.07175405553192782, + "grad_norm": 0.81297767162323, + "learning_rate": 9.883443145208553e-05, + "loss": 1.2132, + "step": 1181 + }, + { + "epoch": 0.07181481256455435, + "grad_norm": 0.3747221529483795, + "learning_rate": 9.883237565656182e-05, + "loss": 1.1826, + "step": 1182 + }, + { + "epoch": 0.07187556959718087, + "grad_norm": 0.32811105251312256, + "learning_rate": 9.883031807107797e-05, + "loss": 1.2439, + "step": 1183 + }, + { + "epoch": 0.0719363266298074, + "grad_norm": 0.6419550776481628, + "learning_rate": 9.882825869570934e-05, + "loss": 1.243, + "step": 1184 + }, + { + "epoch": 0.07199708366243393, + "grad_norm": 0.3182179629802704, + "learning_rate": 9.882619753053146e-05, + "loss": 1.2044, + "step": 1185 + }, + { + "epoch": 0.07205784069506045, + "grad_norm": 0.3979061543941498, + "learning_rate": 9.882413457561988e-05, + "loss": 1.2288, + "step": 1186 + }, + { + "epoch": 0.07211859772768699, + "grad_norm": 2.1301016807556152, + "learning_rate": 9.88220698310502e-05, + "loss": 1.2316, + "step": 1187 + }, + { + "epoch": 0.07217935476031351, + "grad_norm": 0.29915323853492737, + "learning_rate": 9.88200032968981e-05, + "loss": 1.0985, + "step": 1188 + }, + { + "epoch": 0.07224011179294003, + "grad_norm": 0.4142237603664398, + "learning_rate": 9.881793497323937e-05, + "loss": 1.2663, + "step": 1189 + }, + { + "epoch": 0.07230086882556656, + "grad_norm": 0.34139931201934814, + "learning_rate": 9.881586486014979e-05, + "loss": 1.1094, + "step": 1190 + }, + { + "epoch": 0.07236162585819308, + "grad_norm": 0.2735024690628052, + "learning_rate": 9.881379295770524e-05, + "loss": 1.2134, + "step": 1191 + }, + { + "epoch": 0.0724223828908196, + "grad_norm": 1.2520874738693237, + "learning_rate": 9.881171926598167e-05, + "loss": 1.1803, + "step": 1192 + }, + { + "epoch": 0.07248313992344614, + "grad_norm": 0.28545674681663513, + "learning_rate": 9.88096437850551e-05, + "loss": 1.1361, + "step": 1193 + }, + { + "epoch": 0.07254389695607266, + "grad_norm": 0.3079105615615845, + "learning_rate": 9.880756651500162e-05, + "loss": 1.1155, + "step": 1194 + }, + { + "epoch": 0.0726046539886992, + "grad_norm": 0.21594907343387604, + "learning_rate": 9.880548745589733e-05, + "loss": 1.0963, + "step": 1195 + }, + { + "epoch": 0.07266541102132572, + "grad_norm": 0.35872331261634827, + "learning_rate": 9.880340660781848e-05, + "loss": 1.2725, + "step": 1196 + }, + { + "epoch": 0.07272616805395224, + "grad_norm": 0.47961264848709106, + "learning_rate": 9.880132397084132e-05, + "loss": 1.1074, + "step": 1197 + }, + { + "epoch": 0.07278692508657877, + "grad_norm": 0.4087565839290619, + "learning_rate": 9.87992395450422e-05, + "loss": 1.1775, + "step": 1198 + }, + { + "epoch": 0.0728476821192053, + "grad_norm": 0.7602114677429199, + "learning_rate": 9.87971533304975e-05, + "loss": 1.1932, + "step": 1199 + }, + { + "epoch": 0.07290843915183183, + "grad_norm": 0.8478974103927612, + "learning_rate": 9.879506532728375e-05, + "loss": 1.203, + "step": 1200 + }, + { + "epoch": 0.07296919618445835, + "grad_norm": 0.5022792816162109, + "learning_rate": 9.879297553547742e-05, + "loss": 1.1336, + "step": 1201 + }, + { + "epoch": 0.07302995321708487, + "grad_norm": 0.5166787505149841, + "learning_rate": 9.879088395515516e-05, + "loss": 1.1801, + "step": 1202 + }, + { + "epoch": 0.07309071024971141, + "grad_norm": 3.4764270782470703, + "learning_rate": 9.87887905863936e-05, + "loss": 1.1866, + "step": 1203 + }, + { + "epoch": 0.07315146728233793, + "grad_norm": 0.7526102662086487, + "learning_rate": 9.878669542926951e-05, + "loss": 1.2185, + "step": 1204 + }, + { + "epoch": 0.07321222431496446, + "grad_norm": 0.4030632972717285, + "learning_rate": 9.878459848385965e-05, + "loss": 1.2656, + "step": 1205 + }, + { + "epoch": 0.07327298134759098, + "grad_norm": 0.5551868677139282, + "learning_rate": 9.87824997502409e-05, + "loss": 1.1906, + "step": 1206 + }, + { + "epoch": 0.0733337383802175, + "grad_norm": 0.5404157638549805, + "learning_rate": 9.87803992284902e-05, + "loss": 1.1356, + "step": 1207 + }, + { + "epoch": 0.07339449541284404, + "grad_norm": 0.2693815231323242, + "learning_rate": 9.877829691868454e-05, + "loss": 1.1282, + "step": 1208 + }, + { + "epoch": 0.07345525244547056, + "grad_norm": 0.5651394128799438, + "learning_rate": 9.877619282090099e-05, + "loss": 1.1095, + "step": 1209 + }, + { + "epoch": 0.07351600947809708, + "grad_norm": 0.298004150390625, + "learning_rate": 9.877408693521663e-05, + "loss": 1.1654, + "step": 1210 + }, + { + "epoch": 0.07357676651072362, + "grad_norm": 0.3621234595775604, + "learning_rate": 9.877197926170872e-05, + "loss": 1.1269, + "step": 1211 + }, + { + "epoch": 0.07363752354335014, + "grad_norm": 0.33750227093696594, + "learning_rate": 9.876986980045446e-05, + "loss": 1.2283, + "step": 1212 + }, + { + "epoch": 0.07369828057597667, + "grad_norm": 0.262653112411499, + "learning_rate": 9.876775855153122e-05, + "loss": 1.2667, + "step": 1213 + }, + { + "epoch": 0.0737590376086032, + "grad_norm": 0.5421192049980164, + "learning_rate": 9.876564551501635e-05, + "loss": 1.1687, + "step": 1214 + }, + { + "epoch": 0.07381979464122972, + "grad_norm": 0.27910009026527405, + "learning_rate": 9.87635306909873e-05, + "loss": 1.2723, + "step": 1215 + }, + { + "epoch": 0.07388055167385625, + "grad_norm": 0.7909706830978394, + "learning_rate": 9.876141407952163e-05, + "loss": 1.2036, + "step": 1216 + }, + { + "epoch": 0.07394130870648277, + "grad_norm": 0.3347342014312744, + "learning_rate": 9.87592956806969e-05, + "loss": 1.2557, + "step": 1217 + }, + { + "epoch": 0.07400206573910931, + "grad_norm": 0.36630088090896606, + "learning_rate": 9.875717549459075e-05, + "loss": 1.1393, + "step": 1218 + }, + { + "epoch": 0.07406282277173583, + "grad_norm": 1.0792884826660156, + "learning_rate": 9.875505352128091e-05, + "loss": 1.0984, + "step": 1219 + }, + { + "epoch": 0.07412357980436235, + "grad_norm": 0.24648475646972656, + "learning_rate": 9.875292976084516e-05, + "loss": 1.1221, + "step": 1220 + }, + { + "epoch": 0.07418433683698888, + "grad_norm": 0.6548019647598267, + "learning_rate": 9.875080421336135e-05, + "loss": 1.1156, + "step": 1221 + }, + { + "epoch": 0.0742450938696154, + "grad_norm": 0.31058618426322937, + "learning_rate": 9.874867687890739e-05, + "loss": 1.2353, + "step": 1222 + }, + { + "epoch": 0.07430585090224194, + "grad_norm": 0.6895356774330139, + "learning_rate": 9.874654775756125e-05, + "loss": 1.183, + "step": 1223 + }, + { + "epoch": 0.07436660793486846, + "grad_norm": 0.2270854264497757, + "learning_rate": 9.874441684940099e-05, + "loss": 1.1598, + "step": 1224 + }, + { + "epoch": 0.07442736496749498, + "grad_norm": 0.5635459423065186, + "learning_rate": 9.874228415450469e-05, + "loss": 1.1394, + "step": 1225 + }, + { + "epoch": 0.07448812200012152, + "grad_norm": 0.6923201680183411, + "learning_rate": 9.874014967295055e-05, + "loss": 1.4865, + "step": 1226 + }, + { + "epoch": 0.07454887903274804, + "grad_norm": 0.7228917479515076, + "learning_rate": 9.873801340481681e-05, + "loss": 1.2178, + "step": 1227 + }, + { + "epoch": 0.07460963606537457, + "grad_norm": 0.49757659435272217, + "learning_rate": 9.873587535018177e-05, + "loss": 1.1624, + "step": 1228 + }, + { + "epoch": 0.0746703930980011, + "grad_norm": 0.4263117015361786, + "learning_rate": 9.87337355091238e-05, + "loss": 1.2197, + "step": 1229 + }, + { + "epoch": 0.07473115013062762, + "grad_norm": 0.46037915349006653, + "learning_rate": 9.873159388172132e-05, + "loss": 1.2191, + "step": 1230 + }, + { + "epoch": 0.07479190716325415, + "grad_norm": 0.228056862950325, + "learning_rate": 9.872945046805284e-05, + "loss": 1.1569, + "step": 1231 + }, + { + "epoch": 0.07485266419588067, + "grad_norm": 0.3085692822933197, + "learning_rate": 9.872730526819696e-05, + "loss": 1.1449, + "step": 1232 + }, + { + "epoch": 0.0749134212285072, + "grad_norm": 0.5713579058647156, + "learning_rate": 9.872515828223228e-05, + "loss": 1.2831, + "step": 1233 + }, + { + "epoch": 0.07497417826113373, + "grad_norm": 0.42978593707084656, + "learning_rate": 9.87230095102375e-05, + "loss": 1.1193, + "step": 1234 + }, + { + "epoch": 0.07503493529376025, + "grad_norm": 0.6379752159118652, + "learning_rate": 9.872085895229138e-05, + "loss": 1.2921, + "step": 1235 + }, + { + "epoch": 0.07509569232638678, + "grad_norm": 0.2620867192745209, + "learning_rate": 9.871870660847276e-05, + "loss": 1.202, + "step": 1236 + }, + { + "epoch": 0.0751564493590133, + "grad_norm": 0.6096946597099304, + "learning_rate": 9.871655247886053e-05, + "loss": 1.2193, + "step": 1237 + }, + { + "epoch": 0.07521720639163983, + "grad_norm": 0.3008558750152588, + "learning_rate": 9.871439656353367e-05, + "loss": 1.2169, + "step": 1238 + }, + { + "epoch": 0.07527796342426636, + "grad_norm": 0.42608126997947693, + "learning_rate": 9.871223886257117e-05, + "loss": 1.3143, + "step": 1239 + }, + { + "epoch": 0.07533872045689288, + "grad_norm": 0.44430163502693176, + "learning_rate": 9.871007937605214e-05, + "loss": 1.1076, + "step": 1240 + }, + { + "epoch": 0.07539947748951942, + "grad_norm": 0.37956732511520386, + "learning_rate": 9.870791810405575e-05, + "loss": 1.1712, + "step": 1241 + }, + { + "epoch": 0.07546023452214594, + "grad_norm": 0.41033270955085754, + "learning_rate": 9.870575504666119e-05, + "loss": 1.2025, + "step": 1242 + }, + { + "epoch": 0.07552099155477246, + "grad_norm": 3.8582603931427, + "learning_rate": 9.870359020394779e-05, + "loss": 1.11, + "step": 1243 + }, + { + "epoch": 0.075581748587399, + "grad_norm": 0.26687827706336975, + "learning_rate": 9.870142357599485e-05, + "loss": 1.151, + "step": 1244 + }, + { + "epoch": 0.07564250562002552, + "grad_norm": 10.206640243530273, + "learning_rate": 9.869925516288182e-05, + "loss": 1.1268, + "step": 1245 + }, + { + "epoch": 0.07570326265265205, + "grad_norm": 1.0739787817001343, + "learning_rate": 9.869708496468819e-05, + "loss": 1.2076, + "step": 1246 + }, + { + "epoch": 0.07576401968527857, + "grad_norm": 0.5768694281578064, + "learning_rate": 9.869491298149348e-05, + "loss": 1.2212, + "step": 1247 + }, + { + "epoch": 0.07582477671790509, + "grad_norm": 0.779432475566864, + "learning_rate": 9.869273921337734e-05, + "loss": 1.223, + "step": 1248 + }, + { + "epoch": 0.07588553375053163, + "grad_norm": 0.5604224801063538, + "learning_rate": 9.869056366041942e-05, + "loss": 1.117, + "step": 1249 + }, + { + "epoch": 0.07594629078315815, + "grad_norm": 0.2856954336166382, + "learning_rate": 9.868838632269947e-05, + "loss": 1.176, + "step": 1250 + }, + { + "epoch": 0.07600704781578467, + "grad_norm": 0.5719025731086731, + "learning_rate": 9.868620720029734e-05, + "loss": 1.2333, + "step": 1251 + }, + { + "epoch": 0.0760678048484112, + "grad_norm": 0.269532710313797, + "learning_rate": 9.868402629329284e-05, + "loss": 1.1534, + "step": 1252 + }, + { + "epoch": 0.07612856188103773, + "grad_norm": 0.2904431223869324, + "learning_rate": 9.868184360176596e-05, + "loss": 1.1419, + "step": 1253 + }, + { + "epoch": 0.07618931891366426, + "grad_norm": 0.4057616591453552, + "learning_rate": 9.86796591257967e-05, + "loss": 1.1233, + "step": 1254 + }, + { + "epoch": 0.07625007594629078, + "grad_norm": 0.3209756910800934, + "learning_rate": 9.867747286546512e-05, + "loss": 1.2188, + "step": 1255 + }, + { + "epoch": 0.0763108329789173, + "grad_norm": 0.32646259665489197, + "learning_rate": 9.867528482085135e-05, + "loss": 1.1764, + "step": 1256 + }, + { + "epoch": 0.07637159001154384, + "grad_norm": 2.517519950866699, + "learning_rate": 9.867309499203562e-05, + "loss": 1.1418, + "step": 1257 + }, + { + "epoch": 0.07643234704417036, + "grad_norm": 0.2584417164325714, + "learning_rate": 9.867090337909819e-05, + "loss": 1.1556, + "step": 1258 + }, + { + "epoch": 0.0764931040767969, + "grad_norm": 0.5838896036148071, + "learning_rate": 9.866870998211937e-05, + "loss": 1.1106, + "step": 1259 + }, + { + "epoch": 0.07655386110942342, + "grad_norm": 0.3076235353946686, + "learning_rate": 9.866651480117958e-05, + "loss": 1.2422, + "step": 1260 + }, + { + "epoch": 0.07661461814204994, + "grad_norm": 0.37714332342147827, + "learning_rate": 9.86643178363593e-05, + "loss": 1.1268, + "step": 1261 + }, + { + "epoch": 0.07667537517467647, + "grad_norm": 0.24863176047801971, + "learning_rate": 9.866211908773903e-05, + "loss": 1.213, + "step": 1262 + }, + { + "epoch": 0.07673613220730299, + "grad_norm": 6.387319087982178, + "learning_rate": 9.865991855539939e-05, + "loss": 1.1435, + "step": 1263 + }, + { + "epoch": 0.07679688923992953, + "grad_norm": 0.5338149070739746, + "learning_rate": 9.865771623942102e-05, + "loss": 1.1332, + "step": 1264 + }, + { + "epoch": 0.07685764627255605, + "grad_norm": 0.2633802890777588, + "learning_rate": 9.865551213988467e-05, + "loss": 1.2064, + "step": 1265 + }, + { + "epoch": 0.07691840330518257, + "grad_norm": 0.23981378972530365, + "learning_rate": 9.86533062568711e-05, + "loss": 1.2107, + "step": 1266 + }, + { + "epoch": 0.0769791603378091, + "grad_norm": 0.2614775002002716, + "learning_rate": 9.86510985904612e-05, + "loss": 1.1451, + "step": 1267 + }, + { + "epoch": 0.07703991737043563, + "grad_norm": 0.1951579749584198, + "learning_rate": 9.864888914073586e-05, + "loss": 1.1261, + "step": 1268 + }, + { + "epoch": 0.07710067440306216, + "grad_norm": 0.28182828426361084, + "learning_rate": 9.864667790777609e-05, + "loss": 1.1233, + "step": 1269 + }, + { + "epoch": 0.07716143143568868, + "grad_norm": 0.21723072230815887, + "learning_rate": 9.864446489166294e-05, + "loss": 1.1308, + "step": 1270 + }, + { + "epoch": 0.0772221884683152, + "grad_norm": 0.5454809665679932, + "learning_rate": 9.864225009247751e-05, + "loss": 1.1432, + "step": 1271 + }, + { + "epoch": 0.07728294550094174, + "grad_norm": 0.25778231024742126, + "learning_rate": 9.864003351030104e-05, + "loss": 1.1298, + "step": 1272 + }, + { + "epoch": 0.07734370253356826, + "grad_norm": 0.4455838203430176, + "learning_rate": 9.863781514521471e-05, + "loss": 1.0719, + "step": 1273 + }, + { + "epoch": 0.07740445956619478, + "grad_norm": 0.3059065043926239, + "learning_rate": 9.863559499729988e-05, + "loss": 1.1993, + "step": 1274 + }, + { + "epoch": 0.07746521659882132, + "grad_norm": 2.9546594619750977, + "learning_rate": 9.86333730666379e-05, + "loss": 1.1217, + "step": 1275 + }, + { + "epoch": 0.07752597363144784, + "grad_norm": 0.27482664585113525, + "learning_rate": 9.863114935331023e-05, + "loss": 1.2382, + "step": 1276 + }, + { + "epoch": 0.07758673066407437, + "grad_norm": 0.2941252291202545, + "learning_rate": 9.86289238573984e-05, + "loss": 1.2675, + "step": 1277 + }, + { + "epoch": 0.07764748769670089, + "grad_norm": 0.5004340410232544, + "learning_rate": 9.862669657898395e-05, + "loss": 1.3137, + "step": 1278 + }, + { + "epoch": 0.07770824472932741, + "grad_norm": 0.3673279583454132, + "learning_rate": 9.862446751814854e-05, + "loss": 1.2285, + "step": 1279 + }, + { + "epoch": 0.07776900176195395, + "grad_norm": 0.66520094871521, + "learning_rate": 9.862223667497388e-05, + "loss": 1.1288, + "step": 1280 + }, + { + "epoch": 0.07782975879458047, + "grad_norm": 0.6351696252822876, + "learning_rate": 9.862000404954172e-05, + "loss": 1.2026, + "step": 1281 + }, + { + "epoch": 0.077890515827207, + "grad_norm": 0.5101736783981323, + "learning_rate": 9.861776964193392e-05, + "loss": 1.3496, + "step": 1282 + }, + { + "epoch": 0.07795127285983353, + "grad_norm": 3.463336944580078, + "learning_rate": 9.861553345223238e-05, + "loss": 1.1037, + "step": 1283 + }, + { + "epoch": 0.07801202989246005, + "grad_norm": 0.3180168867111206, + "learning_rate": 9.861329548051908e-05, + "loss": 1.1425, + "step": 1284 + }, + { + "epoch": 0.07807278692508658, + "grad_norm": 0.2676180303096771, + "learning_rate": 9.861105572687602e-05, + "loss": 1.1245, + "step": 1285 + }, + { + "epoch": 0.0781335439577131, + "grad_norm": 0.19991973042488098, + "learning_rate": 9.860881419138531e-05, + "loss": 1.1433, + "step": 1286 + }, + { + "epoch": 0.07819430099033964, + "grad_norm": 0.2770824432373047, + "learning_rate": 9.860657087412912e-05, + "loss": 1.1466, + "step": 1287 + }, + { + "epoch": 0.07825505802296616, + "grad_norm": 0.2503451406955719, + "learning_rate": 9.86043257751897e-05, + "loss": 1.132, + "step": 1288 + }, + { + "epoch": 0.07831581505559268, + "grad_norm": 0.2607448995113373, + "learning_rate": 9.860207889464929e-05, + "loss": 1.2087, + "step": 1289 + }, + { + "epoch": 0.07837657208821922, + "grad_norm": 0.22834083437919617, + "learning_rate": 9.859983023259031e-05, + "loss": 1.2142, + "step": 1290 + }, + { + "epoch": 0.07843732912084574, + "grad_norm": 0.4763070046901703, + "learning_rate": 9.859757978909516e-05, + "loss": 1.2432, + "step": 1291 + }, + { + "epoch": 0.07849808615347227, + "grad_norm": 0.29238930344581604, + "learning_rate": 9.859532756424632e-05, + "loss": 1.2293, + "step": 1292 + }, + { + "epoch": 0.07855884318609879, + "grad_norm": 0.2369741052389145, + "learning_rate": 9.859307355812637e-05, + "loss": 1.1908, + "step": 1293 + }, + { + "epoch": 0.07861960021872531, + "grad_norm": 0.2785097360610962, + "learning_rate": 9.859081777081789e-05, + "loss": 1.2194, + "step": 1294 + }, + { + "epoch": 0.07868035725135185, + "grad_norm": 0.22421719133853912, + "learning_rate": 9.858856020240362e-05, + "loss": 1.0738, + "step": 1295 + }, + { + "epoch": 0.07874111428397837, + "grad_norm": 0.2898969352245331, + "learning_rate": 9.858630085296625e-05, + "loss": 1.1808, + "step": 1296 + }, + { + "epoch": 0.07880187131660489, + "grad_norm": 0.3319620192050934, + "learning_rate": 9.858403972258865e-05, + "loss": 1.1604, + "step": 1297 + }, + { + "epoch": 0.07886262834923143, + "grad_norm": 0.3657490313053131, + "learning_rate": 9.858177681135368e-05, + "loss": 1.1438, + "step": 1298 + }, + { + "epoch": 0.07892338538185795, + "grad_norm": 0.20954523980617523, + "learning_rate": 9.85795121193443e-05, + "loss": 1.092, + "step": 1299 + }, + { + "epoch": 0.07898414241448448, + "grad_norm": 0.31321367621421814, + "learning_rate": 9.85772456466435e-05, + "loss": 1.1001, + "step": 1300 + }, + { + "epoch": 0.079044899447111, + "grad_norm": 4.647886753082275, + "learning_rate": 9.85749773933344e-05, + "loss": 1.1021, + "step": 1301 + }, + { + "epoch": 0.07910565647973752, + "grad_norm": 0.49679285287857056, + "learning_rate": 9.857270735950008e-05, + "loss": 1.2046, + "step": 1302 + }, + { + "epoch": 0.07916641351236406, + "grad_norm": 0.22540220618247986, + "learning_rate": 9.857043554522379e-05, + "loss": 1.1531, + "step": 1303 + }, + { + "epoch": 0.07922717054499058, + "grad_norm": 0.24935844540596008, + "learning_rate": 9.85681619505888e-05, + "loss": 1.1336, + "step": 1304 + }, + { + "epoch": 0.07928792757761711, + "grad_norm": 0.43808746337890625, + "learning_rate": 9.856588657567845e-05, + "loss": 1.1185, + "step": 1305 + }, + { + "epoch": 0.07934868461024364, + "grad_norm": 0.1898421049118042, + "learning_rate": 9.856360942057615e-05, + "loss": 1.1347, + "step": 1306 + }, + { + "epoch": 0.07940944164287016, + "grad_norm": 0.8042020797729492, + "learning_rate": 9.856133048536533e-05, + "loss": 1.1241, + "step": 1307 + }, + { + "epoch": 0.07947019867549669, + "grad_norm": 0.39315828680992126, + "learning_rate": 9.855904977012958e-05, + "loss": 1.2585, + "step": 1308 + }, + { + "epoch": 0.07953095570812321, + "grad_norm": 0.5951108336448669, + "learning_rate": 9.855676727495246e-05, + "loss": 1.066, + "step": 1309 + }, + { + "epoch": 0.07959171274074975, + "grad_norm": 0.2079402655363083, + "learning_rate": 9.855448299991765e-05, + "loss": 1.1418, + "step": 1310 + }, + { + "epoch": 0.07965246977337627, + "grad_norm": 0.29990267753601074, + "learning_rate": 9.855219694510889e-05, + "loss": 1.1312, + "step": 1311 + }, + { + "epoch": 0.07971322680600279, + "grad_norm": 3.048137903213501, + "learning_rate": 9.854990911060995e-05, + "loss": 1.0883, + "step": 1312 + }, + { + "epoch": 0.07977398383862933, + "grad_norm": 0.20971716940402985, + "learning_rate": 9.854761949650473e-05, + "loss": 1.1205, + "step": 1313 + }, + { + "epoch": 0.07983474087125585, + "grad_norm": 0.43340668082237244, + "learning_rate": 9.854532810287711e-05, + "loss": 1.2288, + "step": 1314 + }, + { + "epoch": 0.07989549790388237, + "grad_norm": 0.4614243507385254, + "learning_rate": 9.854303492981111e-05, + "loss": 1.1112, + "step": 1315 + }, + { + "epoch": 0.0799562549365089, + "grad_norm": 0.2724321186542511, + "learning_rate": 9.854073997739079e-05, + "loss": 1.1779, + "step": 1316 + }, + { + "epoch": 0.08001701196913542, + "grad_norm": 0.32499074935913086, + "learning_rate": 9.853844324570027e-05, + "loss": 1.2012, + "step": 1317 + }, + { + "epoch": 0.08007776900176196, + "grad_norm": 0.38562822341918945, + "learning_rate": 9.853614473482371e-05, + "loss": 1.3039, + "step": 1318 + }, + { + "epoch": 0.08013852603438848, + "grad_norm": 2.3515474796295166, + "learning_rate": 9.85338444448454e-05, + "loss": 1.1603, + "step": 1319 + }, + { + "epoch": 0.080199283067015, + "grad_norm": 1.032150149345398, + "learning_rate": 9.853154237584965e-05, + "loss": 1.1381, + "step": 1320 + }, + { + "epoch": 0.08026004009964154, + "grad_norm": 0.3963473439216614, + "learning_rate": 9.852923852792081e-05, + "loss": 1.0825, + "step": 1321 + }, + { + "epoch": 0.08032079713226806, + "grad_norm": 0.469703733921051, + "learning_rate": 9.852693290114335e-05, + "loss": 1.216, + "step": 1322 + }, + { + "epoch": 0.08038155416489459, + "grad_norm": 0.29840418696403503, + "learning_rate": 9.85246254956018e-05, + "loss": 1.1965, + "step": 1323 + }, + { + "epoch": 0.08044231119752111, + "grad_norm": 0.3972066342830658, + "learning_rate": 9.852231631138071e-05, + "loss": 1.0938, + "step": 1324 + }, + { + "epoch": 0.08050306823014763, + "grad_norm": 0.45785295963287354, + "learning_rate": 9.852000534856474e-05, + "loss": 1.2022, + "step": 1325 + }, + { + "epoch": 0.08056382526277417, + "grad_norm": 0.7046598196029663, + "learning_rate": 9.851769260723862e-05, + "loss": 1.1734, + "step": 1326 + }, + { + "epoch": 0.08062458229540069, + "grad_norm": 0.2530883252620697, + "learning_rate": 9.851537808748706e-05, + "loss": 1.1538, + "step": 1327 + }, + { + "epoch": 0.08068533932802723, + "grad_norm": 0.29121655225753784, + "learning_rate": 9.851306178939497e-05, + "loss": 1.1029, + "step": 1328 + }, + { + "epoch": 0.08074609636065375, + "grad_norm": 0.3323487341403961, + "learning_rate": 9.85107437130472e-05, + "loss": 1.1681, + "step": 1329 + }, + { + "epoch": 0.08080685339328027, + "grad_norm": 0.22726240754127502, + "learning_rate": 9.850842385852876e-05, + "loss": 1.0913, + "step": 1330 + }, + { + "epoch": 0.0808676104259068, + "grad_norm": 0.2669685482978821, + "learning_rate": 9.850610222592465e-05, + "loss": 1.3229, + "step": 1331 + }, + { + "epoch": 0.08092836745853332, + "grad_norm": 0.2949540615081787, + "learning_rate": 9.850377881531999e-05, + "loss": 1.1495, + "step": 1332 + }, + { + "epoch": 0.08098912449115986, + "grad_norm": 0.2769593298435211, + "learning_rate": 9.850145362679994e-05, + "loss": 1.1316, + "step": 1333 + }, + { + "epoch": 0.08104988152378638, + "grad_norm": 0.21406446397304535, + "learning_rate": 9.849912666044975e-05, + "loss": 1.1633, + "step": 1334 + }, + { + "epoch": 0.0811106385564129, + "grad_norm": 0.4471951127052307, + "learning_rate": 9.84967979163547e-05, + "loss": 1.4437, + "step": 1335 + }, + { + "epoch": 0.08117139558903944, + "grad_norm": 0.2742786705493927, + "learning_rate": 9.849446739460012e-05, + "loss": 1.0841, + "step": 1336 + }, + { + "epoch": 0.08123215262166596, + "grad_norm": 0.21352770924568176, + "learning_rate": 9.849213509527148e-05, + "loss": 1.1042, + "step": 1337 + }, + { + "epoch": 0.08129290965429248, + "grad_norm": 0.24987784028053284, + "learning_rate": 9.848980101845426e-05, + "loss": 1.1397, + "step": 1338 + }, + { + "epoch": 0.08135366668691901, + "grad_norm": 0.4437633156776428, + "learning_rate": 9.848746516423399e-05, + "loss": 1.2028, + "step": 1339 + }, + { + "epoch": 0.08141442371954553, + "grad_norm": 0.36336585879325867, + "learning_rate": 9.848512753269635e-05, + "loss": 1.2516, + "step": 1340 + }, + { + "epoch": 0.08147518075217207, + "grad_norm": 0.5716739892959595, + "learning_rate": 9.848278812392696e-05, + "loss": 1.1256, + "step": 1341 + }, + { + "epoch": 0.08153593778479859, + "grad_norm": 0.24191294610500336, + "learning_rate": 9.848044693801159e-05, + "loss": 1.0759, + "step": 1342 + }, + { + "epoch": 0.08159669481742511, + "grad_norm": 0.22706559300422668, + "learning_rate": 9.847810397503609e-05, + "loss": 1.1336, + "step": 1343 + }, + { + "epoch": 0.08165745185005165, + "grad_norm": 0.2629750370979309, + "learning_rate": 9.84757592350863e-05, + "loss": 1.0977, + "step": 1344 + }, + { + "epoch": 0.08171820888267817, + "grad_norm": 0.7989439964294434, + "learning_rate": 9.84734127182482e-05, + "loss": 1.1354, + "step": 1345 + }, + { + "epoch": 0.0817789659153047, + "grad_norm": 0.28090718388557434, + "learning_rate": 9.847106442460778e-05, + "loss": 1.1205, + "step": 1346 + }, + { + "epoch": 0.08183972294793122, + "grad_norm": 0.38990846276283264, + "learning_rate": 9.846871435425112e-05, + "loss": 1.2242, + "step": 1347 + }, + { + "epoch": 0.08190047998055774, + "grad_norm": 0.47051331400871277, + "learning_rate": 9.846636250726438e-05, + "loss": 1.1747, + "step": 1348 + }, + { + "epoch": 0.08196123701318428, + "grad_norm": 0.23000651597976685, + "learning_rate": 9.846400888373374e-05, + "loss": 1.1295, + "step": 1349 + }, + { + "epoch": 0.0820219940458108, + "grad_norm": 0.43397194147109985, + "learning_rate": 9.84616534837455e-05, + "loss": 1.1149, + "step": 1350 + }, + { + "epoch": 0.08208275107843734, + "grad_norm": 0.2523467242717743, + "learning_rate": 9.845929630738598e-05, + "loss": 1.1301, + "step": 1351 + }, + { + "epoch": 0.08214350811106386, + "grad_norm": 0.46510210633277893, + "learning_rate": 9.845693735474158e-05, + "loss": 1.1265, + "step": 1352 + }, + { + "epoch": 0.08220426514369038, + "grad_norm": 0.31882843375205994, + "learning_rate": 9.845457662589877e-05, + "loss": 1.2207, + "step": 1353 + }, + { + "epoch": 0.08226502217631691, + "grad_norm": 0.49581679701805115, + "learning_rate": 9.845221412094409e-05, + "loss": 1.1895, + "step": 1354 + }, + { + "epoch": 0.08232577920894343, + "grad_norm": 5.497439384460449, + "learning_rate": 9.844984983996415e-05, + "loss": 1.1372, + "step": 1355 + }, + { + "epoch": 0.08238653624156995, + "grad_norm": 0.41685107350349426, + "learning_rate": 9.84474837830456e-05, + "loss": 1.1536, + "step": 1356 + }, + { + "epoch": 0.08244729327419649, + "grad_norm": 0.4489719569683075, + "learning_rate": 9.844511595027515e-05, + "loss": 1.1778, + "step": 1357 + }, + { + "epoch": 0.08250805030682301, + "grad_norm": 0.4206383228302002, + "learning_rate": 9.844274634173962e-05, + "loss": 1.1353, + "step": 1358 + }, + { + "epoch": 0.08256880733944955, + "grad_norm": 0.18028047680854797, + "learning_rate": 9.844037495752584e-05, + "loss": 1.1265, + "step": 1359 + }, + { + "epoch": 0.08262956437207607, + "grad_norm": 0.3056849241256714, + "learning_rate": 9.843800179772078e-05, + "loss": 1.2089, + "step": 1360 + }, + { + "epoch": 0.08269032140470259, + "grad_norm": 0.4130638837814331, + "learning_rate": 9.843562686241139e-05, + "loss": 1.219, + "step": 1361 + }, + { + "epoch": 0.08275107843732912, + "grad_norm": 0.2548668384552002, + "learning_rate": 9.843325015168474e-05, + "loss": 1.2724, + "step": 1362 + }, + { + "epoch": 0.08281183546995564, + "grad_norm": 0.2973799705505371, + "learning_rate": 9.843087166562795e-05, + "loss": 1.3369, + "step": 1363 + }, + { + "epoch": 0.08287259250258218, + "grad_norm": 0.339207261800766, + "learning_rate": 9.842849140432817e-05, + "loss": 1.2598, + "step": 1364 + }, + { + "epoch": 0.0829333495352087, + "grad_norm": 0.23756246268749237, + "learning_rate": 9.84261093678727e-05, + "loss": 1.2516, + "step": 1365 + }, + { + "epoch": 0.08299410656783522, + "grad_norm": 0.5455178022384644, + "learning_rate": 9.842372555634882e-05, + "loss": 1.1631, + "step": 1366 + }, + { + "epoch": 0.08305486360046176, + "grad_norm": 0.18757346272468567, + "learning_rate": 9.842133996984393e-05, + "loss": 1.1013, + "step": 1367 + }, + { + "epoch": 0.08311562063308828, + "grad_norm": 0.4960692226886749, + "learning_rate": 9.841895260844546e-05, + "loss": 1.128, + "step": 1368 + }, + { + "epoch": 0.08317637766571481, + "grad_norm": 0.6121457815170288, + "learning_rate": 9.841656347224091e-05, + "loss": 1.3478, + "step": 1369 + }, + { + "epoch": 0.08323713469834133, + "grad_norm": 0.256935715675354, + "learning_rate": 9.841417256131788e-05, + "loss": 1.1196, + "step": 1370 + }, + { + "epoch": 0.08329789173096785, + "grad_norm": 1.7397311925888062, + "learning_rate": 9.8411779875764e-05, + "loss": 1.1275, + "step": 1371 + }, + { + "epoch": 0.08335864876359439, + "grad_norm": 0.2929978668689728, + "learning_rate": 9.840938541566698e-05, + "loss": 1.2762, + "step": 1372 + }, + { + "epoch": 0.08341940579622091, + "grad_norm": 0.23026397824287415, + "learning_rate": 9.840698918111455e-05, + "loss": 1.1183, + "step": 1373 + }, + { + "epoch": 0.08348016282884745, + "grad_norm": 0.36054399609565735, + "learning_rate": 9.84045911721946e-05, + "loss": 1.1969, + "step": 1374 + }, + { + "epoch": 0.08354091986147397, + "grad_norm": 0.314739853143692, + "learning_rate": 9.8402191388995e-05, + "loss": 1.2639, + "step": 1375 + }, + { + "epoch": 0.08360167689410049, + "grad_norm": 0.21742641925811768, + "learning_rate": 9.83997898316037e-05, + "loss": 1.1588, + "step": 1376 + }, + { + "epoch": 0.08366243392672702, + "grad_norm": 0.2955823540687561, + "learning_rate": 9.839738650010877e-05, + "loss": 1.0895, + "step": 1377 + }, + { + "epoch": 0.08372319095935354, + "grad_norm": 0.249719500541687, + "learning_rate": 9.839498139459827e-05, + "loss": 1.2044, + "step": 1378 + }, + { + "epoch": 0.08378394799198007, + "grad_norm": 0.3078293800354004, + "learning_rate": 9.839257451516037e-05, + "loss": 1.4073, + "step": 1379 + }, + { + "epoch": 0.0838447050246066, + "grad_norm": 0.29615670442581177, + "learning_rate": 9.839016586188331e-05, + "loss": 1.1354, + "step": 1380 + }, + { + "epoch": 0.08390546205723312, + "grad_norm": 0.29500436782836914, + "learning_rate": 9.838775543485537e-05, + "loss": 1.1906, + "step": 1381 + }, + { + "epoch": 0.08396621908985966, + "grad_norm": 0.4208049774169922, + "learning_rate": 9.838534323416489e-05, + "loss": 1.2147, + "step": 1382 + }, + { + "epoch": 0.08402697612248618, + "grad_norm": 0.27508795261383057, + "learning_rate": 9.838292925990029e-05, + "loss": 1.1243, + "step": 1383 + }, + { + "epoch": 0.0840877331551127, + "grad_norm": 0.27086108922958374, + "learning_rate": 9.83805135121501e-05, + "loss": 1.1731, + "step": 1384 + }, + { + "epoch": 0.08414849018773923, + "grad_norm": 0.3610636293888092, + "learning_rate": 9.837809599100281e-05, + "loss": 1.2295, + "step": 1385 + }, + { + "epoch": 0.08420924722036575, + "grad_norm": 0.24157598614692688, + "learning_rate": 9.837567669654706e-05, + "loss": 1.0977, + "step": 1386 + }, + { + "epoch": 0.08427000425299229, + "grad_norm": 0.41961565613746643, + "learning_rate": 9.837325562887155e-05, + "loss": 1.1691, + "step": 1387 + }, + { + "epoch": 0.08433076128561881, + "grad_norm": 0.7909747958183289, + "learning_rate": 9.837083278806497e-05, + "loss": 1.097, + "step": 1388 + }, + { + "epoch": 0.08439151831824533, + "grad_norm": 1.8897919654846191, + "learning_rate": 9.836840817421619e-05, + "loss": 1.123, + "step": 1389 + }, + { + "epoch": 0.08445227535087187, + "grad_norm": 1.8377037048339844, + "learning_rate": 9.836598178741405e-05, + "loss": 1.1944, + "step": 1390 + }, + { + "epoch": 0.08451303238349839, + "grad_norm": 0.4132292568683624, + "learning_rate": 9.83635536277475e-05, + "loss": 1.1056, + "step": 1391 + }, + { + "epoch": 0.08457378941612492, + "grad_norm": 0.47743967175483704, + "learning_rate": 9.836112369530555e-05, + "loss": 1.1684, + "step": 1392 + }, + { + "epoch": 0.08463454644875144, + "grad_norm": 0.32137900590896606, + "learning_rate": 9.835869199017725e-05, + "loss": 1.0632, + "step": 1393 + }, + { + "epoch": 0.08469530348137796, + "grad_norm": 0.23343518376350403, + "learning_rate": 9.835625851245177e-05, + "loss": 1.1765, + "step": 1394 + }, + { + "epoch": 0.0847560605140045, + "grad_norm": 0.28076237440109253, + "learning_rate": 9.835382326221826e-05, + "loss": 1.2066, + "step": 1395 + }, + { + "epoch": 0.08481681754663102, + "grad_norm": 3.974865436553955, + "learning_rate": 9.835138623956603e-05, + "loss": 1.2106, + "step": 1396 + }, + { + "epoch": 0.08487757457925756, + "grad_norm": 3.0839321613311768, + "learning_rate": 9.834894744458438e-05, + "loss": 1.268, + "step": 1397 + }, + { + "epoch": 0.08493833161188408, + "grad_norm": 0.44023963809013367, + "learning_rate": 9.834650687736272e-05, + "loss": 1.157, + "step": 1398 + }, + { + "epoch": 0.0849990886445106, + "grad_norm": 0.2471620887517929, + "learning_rate": 9.83440645379905e-05, + "loss": 1.2124, + "step": 1399 + }, + { + "epoch": 0.08505984567713713, + "grad_norm": 0.4825039803981781, + "learning_rate": 9.834162042655726e-05, + "loss": 1.1166, + "step": 1400 + }, + { + "epoch": 0.08512060270976365, + "grad_norm": 0.27850329875946045, + "learning_rate": 9.833917454315258e-05, + "loss": 1.1234, + "step": 1401 + }, + { + "epoch": 0.08518135974239018, + "grad_norm": 0.5909184813499451, + "learning_rate": 9.83367268878661e-05, + "loss": 1.0834, + "step": 1402 + }, + { + "epoch": 0.08524211677501671, + "grad_norm": 0.20529113709926605, + "learning_rate": 9.833427746078756e-05, + "loss": 1.0991, + "step": 1403 + }, + { + "epoch": 0.08530287380764323, + "grad_norm": 0.3067261874675751, + "learning_rate": 9.833182626200675e-05, + "loss": 1.121, + "step": 1404 + }, + { + "epoch": 0.08536363084026977, + "grad_norm": 0.21722310781478882, + "learning_rate": 9.83293732916135e-05, + "loss": 1.0957, + "step": 1405 + }, + { + "epoch": 0.08542438787289629, + "grad_norm": 0.34592217206954956, + "learning_rate": 9.832691854969771e-05, + "loss": 1.2032, + "step": 1406 + }, + { + "epoch": 0.08548514490552281, + "grad_norm": 0.2721937894821167, + "learning_rate": 9.83244620363494e-05, + "loss": 1.1249, + "step": 1407 + }, + { + "epoch": 0.08554590193814934, + "grad_norm": 0.23150306940078735, + "learning_rate": 9.832200375165858e-05, + "loss": 1.2101, + "step": 1408 + }, + { + "epoch": 0.08560665897077586, + "grad_norm": 0.5178032517433167, + "learning_rate": 9.831954369571538e-05, + "loss": 1.1523, + "step": 1409 + }, + { + "epoch": 0.0856674160034024, + "grad_norm": 0.2342749536037445, + "learning_rate": 9.831708186860996e-05, + "loss": 1.1564, + "step": 1410 + }, + { + "epoch": 0.08572817303602892, + "grad_norm": 0.4153948128223419, + "learning_rate": 9.831461827043257e-05, + "loss": 1.1216, + "step": 1411 + }, + { + "epoch": 0.08578893006865544, + "grad_norm": 0.2572597861289978, + "learning_rate": 9.831215290127351e-05, + "loss": 1.2497, + "step": 1412 + }, + { + "epoch": 0.08584968710128198, + "grad_norm": 0.29278334975242615, + "learning_rate": 9.830968576122315e-05, + "loss": 1.4024, + "step": 1413 + }, + { + "epoch": 0.0859104441339085, + "grad_norm": 0.392046719789505, + "learning_rate": 9.830721685037191e-05, + "loss": 1.178, + "step": 1414 + }, + { + "epoch": 0.08597120116653503, + "grad_norm": 0.19422471523284912, + "learning_rate": 9.830474616881029e-05, + "loss": 1.1023, + "step": 1415 + }, + { + "epoch": 0.08603195819916155, + "grad_norm": 0.3753969669342041, + "learning_rate": 9.830227371662888e-05, + "loss": 1.0595, + "step": 1416 + }, + { + "epoch": 0.08609271523178808, + "grad_norm": 0.2701476812362671, + "learning_rate": 9.829979949391828e-05, + "loss": 1.1742, + "step": 1417 + }, + { + "epoch": 0.08615347226441461, + "grad_norm": 0.5247951745986938, + "learning_rate": 9.82973235007692e-05, + "loss": 1.3615, + "step": 1418 + }, + { + "epoch": 0.08621422929704113, + "grad_norm": 0.5167084336280823, + "learning_rate": 9.82948457372724e-05, + "loss": 1.1396, + "step": 1419 + }, + { + "epoch": 0.08627498632966765, + "grad_norm": 0.3301272690296173, + "learning_rate": 9.829236620351868e-05, + "loss": 1.1239, + "step": 1420 + }, + { + "epoch": 0.08633574336229419, + "grad_norm": 0.7128092646598816, + "learning_rate": 9.828988489959894e-05, + "loss": 1.1051, + "step": 1421 + }, + { + "epoch": 0.08639650039492071, + "grad_norm": 0.26560088992118835, + "learning_rate": 9.828740182560415e-05, + "loss": 1.2102, + "step": 1422 + }, + { + "epoch": 0.08645725742754724, + "grad_norm": 0.6973952054977417, + "learning_rate": 9.828491698162531e-05, + "loss": 1.1231, + "step": 1423 + }, + { + "epoch": 0.08651801446017376, + "grad_norm": 0.308383047580719, + "learning_rate": 9.828243036775348e-05, + "loss": 1.1222, + "step": 1424 + }, + { + "epoch": 0.08657877149280029, + "grad_norm": 0.32966095209121704, + "learning_rate": 9.827994198407986e-05, + "loss": 1.1384, + "step": 1425 + }, + { + "epoch": 0.08663952852542682, + "grad_norm": 0.37722039222717285, + "learning_rate": 9.827745183069563e-05, + "loss": 1.1636, + "step": 1426 + }, + { + "epoch": 0.08670028555805334, + "grad_norm": 0.29763877391815186, + "learning_rate": 9.827495990769207e-05, + "loss": 1.2365, + "step": 1427 + }, + { + "epoch": 0.08676104259067988, + "grad_norm": 0.44911399483680725, + "learning_rate": 9.827246621516051e-05, + "loss": 1.197, + "step": 1428 + }, + { + "epoch": 0.0868217996233064, + "grad_norm": 0.2792888283729553, + "learning_rate": 9.826997075319238e-05, + "loss": 1.107, + "step": 1429 + }, + { + "epoch": 0.08688255665593292, + "grad_norm": 0.24899256229400635, + "learning_rate": 9.826747352187915e-05, + "loss": 1.1512, + "step": 1430 + }, + { + "epoch": 0.08694331368855945, + "grad_norm": 0.3737781345844269, + "learning_rate": 9.826497452131234e-05, + "loss": 1.1666, + "step": 1431 + }, + { + "epoch": 0.08700407072118597, + "grad_norm": 0.3070584237575531, + "learning_rate": 9.826247375158355e-05, + "loss": 1.1832, + "step": 1432 + }, + { + "epoch": 0.08706482775381251, + "grad_norm": 0.3636440932750702, + "learning_rate": 9.825997121278448e-05, + "loss": 1.1367, + "step": 1433 + }, + { + "epoch": 0.08712558478643903, + "grad_norm": 0.19736309349536896, + "learning_rate": 9.825746690500681e-05, + "loss": 1.0871, + "step": 1434 + }, + { + "epoch": 0.08718634181906555, + "grad_norm": 0.3926192820072174, + "learning_rate": 9.825496082834238e-05, + "loss": 1.1163, + "step": 1435 + }, + { + "epoch": 0.08724709885169209, + "grad_norm": 0.2951951026916504, + "learning_rate": 9.825245298288301e-05, + "loss": 1.1477, + "step": 1436 + }, + { + "epoch": 0.08730785588431861, + "grad_norm": 0.2625080645084381, + "learning_rate": 9.824994336872068e-05, + "loss": 1.1737, + "step": 1437 + }, + { + "epoch": 0.08736861291694514, + "grad_norm": 0.3721047043800354, + "learning_rate": 9.824743198594734e-05, + "loss": 1.1118, + "step": 1438 + }, + { + "epoch": 0.08742936994957166, + "grad_norm": 0.8178937435150146, + "learning_rate": 9.824491883465504e-05, + "loss": 1.2435, + "step": 1439 + }, + { + "epoch": 0.08749012698219819, + "grad_norm": 0.3319724202156067, + "learning_rate": 9.824240391493593e-05, + "loss": 1.1362, + "step": 1440 + }, + { + "epoch": 0.08755088401482472, + "grad_norm": 1.6131713390350342, + "learning_rate": 9.823988722688218e-05, + "loss": 1.1934, + "step": 1441 + }, + { + "epoch": 0.08761164104745124, + "grad_norm": 0.21765705943107605, + "learning_rate": 9.823736877058602e-05, + "loss": 1.26, + "step": 1442 + }, + { + "epoch": 0.08767239808007776, + "grad_norm": 0.43235087394714355, + "learning_rate": 9.82348485461398e-05, + "loss": 1.0884, + "step": 1443 + }, + { + "epoch": 0.0877331551127043, + "grad_norm": 0.3108364939689636, + "learning_rate": 9.823232655363588e-05, + "loss": 1.2788, + "step": 1444 + }, + { + "epoch": 0.08779391214533082, + "grad_norm": 0.33811429142951965, + "learning_rate": 9.822980279316671e-05, + "loss": 1.2496, + "step": 1445 + }, + { + "epoch": 0.08785466917795735, + "grad_norm": 0.28995388746261597, + "learning_rate": 9.82272772648248e-05, + "loss": 1.2742, + "step": 1446 + }, + { + "epoch": 0.08791542621058387, + "grad_norm": 0.9235050678253174, + "learning_rate": 9.82247499687027e-05, + "loss": 1.1366, + "step": 1447 + }, + { + "epoch": 0.0879761832432104, + "grad_norm": 0.27627459168434143, + "learning_rate": 9.822222090489307e-05, + "loss": 1.2742, + "step": 1448 + }, + { + "epoch": 0.08803694027583693, + "grad_norm": 0.45420557260513306, + "learning_rate": 9.821969007348863e-05, + "loss": 1.2169, + "step": 1449 + }, + { + "epoch": 0.08809769730846345, + "grad_norm": 0.5346761345863342, + "learning_rate": 9.821715747458213e-05, + "loss": 1.2053, + "step": 1450 + }, + { + "epoch": 0.08815845434108999, + "grad_norm": 0.2524607181549072, + "learning_rate": 9.821462310826639e-05, + "loss": 1.2202, + "step": 1451 + }, + { + "epoch": 0.08821921137371651, + "grad_norm": 4.679330825805664, + "learning_rate": 9.821208697463434e-05, + "loss": 1.208, + "step": 1452 + }, + { + "epoch": 0.08827996840634303, + "grad_norm": 0.3004501461982727, + "learning_rate": 9.820954907377891e-05, + "loss": 1.1468, + "step": 1453 + }, + { + "epoch": 0.08834072543896956, + "grad_norm": 0.3175809979438782, + "learning_rate": 9.820700940579312e-05, + "loss": 1.1181, + "step": 1454 + }, + { + "epoch": 0.08840148247159609, + "grad_norm": 0.3236725926399231, + "learning_rate": 9.82044679707701e-05, + "loss": 1.2076, + "step": 1455 + }, + { + "epoch": 0.08846223950422262, + "grad_norm": 0.4570675194263458, + "learning_rate": 9.820192476880299e-05, + "loss": 1.2613, + "step": 1456 + }, + { + "epoch": 0.08852299653684914, + "grad_norm": 0.25527098774909973, + "learning_rate": 9.819937979998501e-05, + "loss": 1.1354, + "step": 1457 + }, + { + "epoch": 0.08858375356947566, + "grad_norm": 0.24131177365779877, + "learning_rate": 9.819683306440945e-05, + "loss": 1.1026, + "step": 1458 + }, + { + "epoch": 0.0886445106021022, + "grad_norm": 0.6098653078079224, + "learning_rate": 9.819428456216966e-05, + "loss": 1.2523, + "step": 1459 + }, + { + "epoch": 0.08870526763472872, + "grad_norm": 0.3119412660598755, + "learning_rate": 9.819173429335904e-05, + "loss": 1.1528, + "step": 1460 + }, + { + "epoch": 0.08876602466735524, + "grad_norm": 12.466018676757812, + "learning_rate": 9.81891822580711e-05, + "loss": 1.1338, + "step": 1461 + }, + { + "epoch": 0.08882678169998177, + "grad_norm": 0.40344536304473877, + "learning_rate": 9.818662845639936e-05, + "loss": 1.2172, + "step": 1462 + }, + { + "epoch": 0.0888875387326083, + "grad_norm": 0.3731597065925598, + "learning_rate": 9.818407288843743e-05, + "loss": 1.397, + "step": 1463 + }, + { + "epoch": 0.08894829576523483, + "grad_norm": 0.2263525128364563, + "learning_rate": 9.818151555427901e-05, + "loss": 1.1597, + "step": 1464 + }, + { + "epoch": 0.08900905279786135, + "grad_norm": 0.28693854808807373, + "learning_rate": 9.817895645401783e-05, + "loss": 1.2211, + "step": 1465 + }, + { + "epoch": 0.08906980983048787, + "grad_norm": 0.26339980959892273, + "learning_rate": 9.817639558774768e-05, + "loss": 1.1217, + "step": 1466 + }, + { + "epoch": 0.08913056686311441, + "grad_norm": 0.26475346088409424, + "learning_rate": 9.817383295556242e-05, + "loss": 1.2341, + "step": 1467 + }, + { + "epoch": 0.08919132389574093, + "grad_norm": 0.21460701525211334, + "learning_rate": 9.817126855755601e-05, + "loss": 1.1264, + "step": 1468 + }, + { + "epoch": 0.08925208092836746, + "grad_norm": 0.4389999806880951, + "learning_rate": 9.816870239382245e-05, + "loss": 1.1525, + "step": 1469 + }, + { + "epoch": 0.08931283796099398, + "grad_norm": 0.32700854539871216, + "learning_rate": 9.81661344644558e-05, + "loss": 1.2082, + "step": 1470 + }, + { + "epoch": 0.0893735949936205, + "grad_norm": 0.47564396262168884, + "learning_rate": 9.816356476955015e-05, + "loss": 1.1642, + "step": 1471 + }, + { + "epoch": 0.08943435202624704, + "grad_norm": 0.2876659333705902, + "learning_rate": 9.816099330919975e-05, + "loss": 1.1161, + "step": 1472 + }, + { + "epoch": 0.08949510905887356, + "grad_norm": 0.34097447991371155, + "learning_rate": 9.815842008349882e-05, + "loss": 1.2021, + "step": 1473 + }, + { + "epoch": 0.0895558660915001, + "grad_norm": 0.3062654137611389, + "learning_rate": 9.815584509254169e-05, + "loss": 1.1295, + "step": 1474 + }, + { + "epoch": 0.08961662312412662, + "grad_norm": 0.2728365361690521, + "learning_rate": 9.815326833642278e-05, + "loss": 1.0976, + "step": 1475 + }, + { + "epoch": 0.08967738015675314, + "grad_norm": 0.2941268980503082, + "learning_rate": 9.815068981523647e-05, + "loss": 1.2189, + "step": 1476 + }, + { + "epoch": 0.08973813718937967, + "grad_norm": 9.694623947143555, + "learning_rate": 9.814810952907735e-05, + "loss": 1.2262, + "step": 1477 + }, + { + "epoch": 0.0897988942220062, + "grad_norm": 0.6052282452583313, + "learning_rate": 9.814552747803996e-05, + "loss": 1.1537, + "step": 1478 + }, + { + "epoch": 0.08985965125463273, + "grad_norm": 0.3128054141998291, + "learning_rate": 9.814294366221895e-05, + "loss": 1.1925, + "step": 1479 + }, + { + "epoch": 0.08992040828725925, + "grad_norm": 0.4922245144844055, + "learning_rate": 9.814035808170902e-05, + "loss": 1.1989, + "step": 1480 + }, + { + "epoch": 0.08998116531988577, + "grad_norm": 0.5280200839042664, + "learning_rate": 9.813777073660498e-05, + "loss": 1.194, + "step": 1481 + }, + { + "epoch": 0.09004192235251231, + "grad_norm": 0.2638612687587738, + "learning_rate": 9.813518162700164e-05, + "loss": 1.1592, + "step": 1482 + }, + { + "epoch": 0.09010267938513883, + "grad_norm": 0.2799600660800934, + "learning_rate": 9.81325907529939e-05, + "loss": 1.1387, + "step": 1483 + }, + { + "epoch": 0.09016343641776535, + "grad_norm": 0.3849027752876282, + "learning_rate": 9.812999811467676e-05, + "loss": 1.1892, + "step": 1484 + }, + { + "epoch": 0.09022419345039188, + "grad_norm": 0.411874383687973, + "learning_rate": 9.812740371214523e-05, + "loss": 1.0993, + "step": 1485 + }, + { + "epoch": 0.0902849504830184, + "grad_norm": 0.4891902506351471, + "learning_rate": 9.812480754549442e-05, + "loss": 1.1091, + "step": 1486 + }, + { + "epoch": 0.09034570751564494, + "grad_norm": 0.3104701042175293, + "learning_rate": 9.812220961481949e-05, + "loss": 1.1827, + "step": 1487 + }, + { + "epoch": 0.09040646454827146, + "grad_norm": 1.3639897108078003, + "learning_rate": 9.811960992021565e-05, + "loss": 1.1285, + "step": 1488 + }, + { + "epoch": 0.09046722158089798, + "grad_norm": 0.48198938369750977, + "learning_rate": 9.811700846177823e-05, + "loss": 1.3099, + "step": 1489 + }, + { + "epoch": 0.09052797861352452, + "grad_norm": 0.25473663210868835, + "learning_rate": 9.811440523960254e-05, + "loss": 1.1985, + "step": 1490 + }, + { + "epoch": 0.09058873564615104, + "grad_norm": 0.26125696301460266, + "learning_rate": 9.811180025378404e-05, + "loss": 1.1014, + "step": 1491 + }, + { + "epoch": 0.09064949267877757, + "grad_norm": 0.40752044320106506, + "learning_rate": 9.810919350441818e-05, + "loss": 1.1414, + "step": 1492 + }, + { + "epoch": 0.0907102497114041, + "grad_norm": 0.3353760242462158, + "learning_rate": 9.810658499160056e-05, + "loss": 1.2519, + "step": 1493 + }, + { + "epoch": 0.09077100674403062, + "grad_norm": 1.0922770500183105, + "learning_rate": 9.810397471542676e-05, + "loss": 1.2237, + "step": 1494 + }, + { + "epoch": 0.09083176377665715, + "grad_norm": 0.2544972002506256, + "learning_rate": 9.810136267599246e-05, + "loss": 1.128, + "step": 1495 + }, + { + "epoch": 0.09089252080928367, + "grad_norm": 0.4165824353694916, + "learning_rate": 9.809874887339343e-05, + "loss": 1.1838, + "step": 1496 + }, + { + "epoch": 0.09095327784191021, + "grad_norm": 0.4201692044734955, + "learning_rate": 9.809613330772544e-05, + "loss": 1.1882, + "step": 1497 + }, + { + "epoch": 0.09101403487453673, + "grad_norm": 0.4648939073085785, + "learning_rate": 9.809351597908441e-05, + "loss": 1.1246, + "step": 1498 + }, + { + "epoch": 0.09107479190716325, + "grad_norm": 0.34819141030311584, + "learning_rate": 9.809089688756623e-05, + "loss": 1.1316, + "step": 1499 + }, + { + "epoch": 0.09113554893978978, + "grad_norm": 0.2317804992198944, + "learning_rate": 9.808827603326695e-05, + "loss": 1.1127, + "step": 1500 + }, + { + "epoch": 0.0911963059724163, + "grad_norm": 0.26921936869621277, + "learning_rate": 9.808565341628261e-05, + "loss": 1.1776, + "step": 1501 + }, + { + "epoch": 0.09125706300504284, + "grad_norm": 6.363933563232422, + "learning_rate": 9.808302903670935e-05, + "loss": 1.1166, + "step": 1502 + }, + { + "epoch": 0.09131782003766936, + "grad_norm": 0.41766443848609924, + "learning_rate": 9.808040289464336e-05, + "loss": 1.1006, + "step": 1503 + }, + { + "epoch": 0.09137857707029588, + "grad_norm": 0.4737500548362732, + "learning_rate": 9.807777499018092e-05, + "loss": 1.1341, + "step": 1504 + }, + { + "epoch": 0.09143933410292242, + "grad_norm": 1.43143630027771, + "learning_rate": 9.807514532341832e-05, + "loss": 1.0887, + "step": 1505 + }, + { + "epoch": 0.09150009113554894, + "grad_norm": 0.27003738284111023, + "learning_rate": 9.8072513894452e-05, + "loss": 1.1486, + "step": 1506 + }, + { + "epoch": 0.09156084816817546, + "grad_norm": 0.29585084319114685, + "learning_rate": 9.806988070337839e-05, + "loss": 1.2192, + "step": 1507 + }, + { + "epoch": 0.091621605200802, + "grad_norm": 0.269220232963562, + "learning_rate": 9.8067245750294e-05, + "loss": 1.1987, + "step": 1508 + }, + { + "epoch": 0.09168236223342852, + "grad_norm": 0.33832868933677673, + "learning_rate": 9.806460903529544e-05, + "loss": 1.1802, + "step": 1509 + }, + { + "epoch": 0.09174311926605505, + "grad_norm": 0.24337616562843323, + "learning_rate": 9.806197055847934e-05, + "loss": 1.1568, + "step": 1510 + }, + { + "epoch": 0.09180387629868157, + "grad_norm": 0.2861745357513428, + "learning_rate": 9.805933031994241e-05, + "loss": 1.2769, + "step": 1511 + }, + { + "epoch": 0.0918646333313081, + "grad_norm": 0.2774497866630554, + "learning_rate": 9.805668831978145e-05, + "loss": 1.2352, + "step": 1512 + }, + { + "epoch": 0.09192539036393463, + "grad_norm": 0.29891932010650635, + "learning_rate": 9.805404455809328e-05, + "loss": 1.134, + "step": 1513 + }, + { + "epoch": 0.09198614739656115, + "grad_norm": 0.3469502925872803, + "learning_rate": 9.805139903497482e-05, + "loss": 1.3215, + "step": 1514 + }, + { + "epoch": 0.09204690442918768, + "grad_norm": 0.36886611580848694, + "learning_rate": 9.804875175052304e-05, + "loss": 1.1094, + "step": 1515 + }, + { + "epoch": 0.0921076614618142, + "grad_norm": 0.4239455461502075, + "learning_rate": 9.804610270483498e-05, + "loss": 1.0778, + "step": 1516 + }, + { + "epoch": 0.09216841849444073, + "grad_norm": 0.2676193118095398, + "learning_rate": 9.804345189800774e-05, + "loss": 1.301, + "step": 1517 + }, + { + "epoch": 0.09222917552706726, + "grad_norm": 0.26064276695251465, + "learning_rate": 9.804079933013847e-05, + "loss": 1.2016, + "step": 1518 + }, + { + "epoch": 0.09228993255969378, + "grad_norm": 0.29170241951942444, + "learning_rate": 9.803814500132445e-05, + "loss": 1.1722, + "step": 1519 + }, + { + "epoch": 0.09235068959232032, + "grad_norm": 0.2798483967781067, + "learning_rate": 9.80354889116629e-05, + "loss": 1.0775, + "step": 1520 + }, + { + "epoch": 0.09241144662494684, + "grad_norm": 0.3079824447631836, + "learning_rate": 9.803283106125123e-05, + "loss": 1.1775, + "step": 1521 + }, + { + "epoch": 0.09247220365757336, + "grad_norm": 0.26351845264434814, + "learning_rate": 9.803017145018687e-05, + "loss": 1.1427, + "step": 1522 + }, + { + "epoch": 0.0925329606901999, + "grad_norm": 0.9863926768302917, + "learning_rate": 9.802751007856728e-05, + "loss": 1.3279, + "step": 1523 + }, + { + "epoch": 0.09259371772282642, + "grad_norm": 0.3570411801338196, + "learning_rate": 9.802484694649001e-05, + "loss": 1.1668, + "step": 1524 + }, + { + "epoch": 0.09265447475545294, + "grad_norm": 0.2354201227426529, + "learning_rate": 9.802218205405272e-05, + "loss": 1.124, + "step": 1525 + }, + { + "epoch": 0.09271523178807947, + "grad_norm": 0.2799304127693176, + "learning_rate": 9.801951540135304e-05, + "loss": 1.16, + "step": 1526 + }, + { + "epoch": 0.09277598882070599, + "grad_norm": 0.21745845675468445, + "learning_rate": 9.801684698848874e-05, + "loss": 1.1658, + "step": 1527 + }, + { + "epoch": 0.09283674585333253, + "grad_norm": 0.9136560559272766, + "learning_rate": 9.801417681555765e-05, + "loss": 1.2822, + "step": 1528 + }, + { + "epoch": 0.09289750288595905, + "grad_norm": 0.6795973777770996, + "learning_rate": 9.801150488265762e-05, + "loss": 1.2942, + "step": 1529 + }, + { + "epoch": 0.09295825991858557, + "grad_norm": 0.3220146596431732, + "learning_rate": 9.80088311898866e-05, + "loss": 1.1424, + "step": 1530 + }, + { + "epoch": 0.0930190169512121, + "grad_norm": 0.5186130404472351, + "learning_rate": 9.80061557373426e-05, + "loss": 1.2201, + "step": 1531 + }, + { + "epoch": 0.09307977398383863, + "grad_norm": 0.8507881164550781, + "learning_rate": 9.800347852512366e-05, + "loss": 1.0876, + "step": 1532 + }, + { + "epoch": 0.09314053101646516, + "grad_norm": 0.3460976481437683, + "learning_rate": 9.800079955332795e-05, + "loss": 1.1183, + "step": 1533 + }, + { + "epoch": 0.09320128804909168, + "grad_norm": 0.48681315779685974, + "learning_rate": 9.799811882205366e-05, + "loss": 1.1259, + "step": 1534 + }, + { + "epoch": 0.0932620450817182, + "grad_norm": 0.4568346440792084, + "learning_rate": 9.799543633139904e-05, + "loss": 1.2374, + "step": 1535 + }, + { + "epoch": 0.09332280211434474, + "grad_norm": 0.3432723581790924, + "learning_rate": 9.799275208146242e-05, + "loss": 1.1645, + "step": 1536 + }, + { + "epoch": 0.09338355914697126, + "grad_norm": 0.36032453179359436, + "learning_rate": 9.79900660723422e-05, + "loss": 1.3103, + "step": 1537 + }, + { + "epoch": 0.0934443161795978, + "grad_norm": 0.3071368932723999, + "learning_rate": 9.798737830413683e-05, + "loss": 1.1678, + "step": 1538 + }, + { + "epoch": 0.09350507321222432, + "grad_norm": 0.29380348324775696, + "learning_rate": 9.798468877694484e-05, + "loss": 1.0484, + "step": 1539 + }, + { + "epoch": 0.09356583024485084, + "grad_norm": 0.24329933524131775, + "learning_rate": 9.79819974908648e-05, + "loss": 1.1629, + "step": 1540 + }, + { + "epoch": 0.09362658727747737, + "grad_norm": 0.3831419348716736, + "learning_rate": 9.797930444599537e-05, + "loss": 1.1316, + "step": 1541 + }, + { + "epoch": 0.09368734431010389, + "grad_norm": 0.6009194254875183, + "learning_rate": 9.797660964243527e-05, + "loss": 1.2261, + "step": 1542 + }, + { + "epoch": 0.09374810134273043, + "grad_norm": 0.45703133940696716, + "learning_rate": 9.797391308028325e-05, + "loss": 1.2158, + "step": 1543 + }, + { + "epoch": 0.09380885837535695, + "grad_norm": 0.28675854206085205, + "learning_rate": 9.797121475963819e-05, + "loss": 1.1557, + "step": 1544 + }, + { + "epoch": 0.09386961540798347, + "grad_norm": 0.3204266130924225, + "learning_rate": 9.796851468059897e-05, + "loss": 1.1213, + "step": 1545 + }, + { + "epoch": 0.09393037244061, + "grad_norm": 0.28548192977905273, + "learning_rate": 9.796581284326458e-05, + "loss": 1.1812, + "step": 1546 + }, + { + "epoch": 0.09399112947323653, + "grad_norm": 0.33858296275138855, + "learning_rate": 9.796310924773404e-05, + "loss": 1.0751, + "step": 1547 + }, + { + "epoch": 0.09405188650586305, + "grad_norm": 0.3599202632904053, + "learning_rate": 9.796040389410645e-05, + "loss": 1.0916, + "step": 1548 + }, + { + "epoch": 0.09411264353848958, + "grad_norm": 0.38406550884246826, + "learning_rate": 9.795769678248101e-05, + "loss": 1.1875, + "step": 1549 + }, + { + "epoch": 0.0941734005711161, + "grad_norm": 0.5666332840919495, + "learning_rate": 9.795498791295692e-05, + "loss": 1.2814, + "step": 1550 + }, + { + "epoch": 0.09423415760374264, + "grad_norm": 0.4843580722808838, + "learning_rate": 9.795227728563346e-05, + "loss": 1.1178, + "step": 1551 + }, + { + "epoch": 0.09429491463636916, + "grad_norm": 0.5978760123252869, + "learning_rate": 9.794956490061002e-05, + "loss": 1.1287, + "step": 1552 + }, + { + "epoch": 0.09435567166899568, + "grad_norm": 0.6704818606376648, + "learning_rate": 9.794685075798602e-05, + "loss": 1.1849, + "step": 1553 + }, + { + "epoch": 0.09441642870162222, + "grad_norm": 0.3926982581615448, + "learning_rate": 9.794413485786092e-05, + "loss": 1.115, + "step": 1554 + }, + { + "epoch": 0.09447718573424874, + "grad_norm": 0.27134189009666443, + "learning_rate": 9.79414172003343e-05, + "loss": 1.1088, + "step": 1555 + }, + { + "epoch": 0.09453794276687527, + "grad_norm": 0.3986111283302307, + "learning_rate": 9.793869778550575e-05, + "loss": 1.3152, + "step": 1556 + }, + { + "epoch": 0.09459869979950179, + "grad_norm": 0.2637803852558136, + "learning_rate": 9.793597661347498e-05, + "loss": 1.1664, + "step": 1557 + }, + { + "epoch": 0.09465945683212831, + "grad_norm": 3.9087822437286377, + "learning_rate": 9.793325368434172e-05, + "loss": 1.228, + "step": 1558 + }, + { + "epoch": 0.09472021386475485, + "grad_norm": 0.47382786870002747, + "learning_rate": 9.793052899820579e-05, + "loss": 1.1099, + "step": 1559 + }, + { + "epoch": 0.09478097089738137, + "grad_norm": 0.9069790244102478, + "learning_rate": 9.792780255516704e-05, + "loss": 1.1218, + "step": 1560 + }, + { + "epoch": 0.0948417279300079, + "grad_norm": 0.7107006311416626, + "learning_rate": 9.792507435532543e-05, + "loss": 1.2855, + "step": 1561 + }, + { + "epoch": 0.09490248496263443, + "grad_norm": 0.2888360619544983, + "learning_rate": 9.792234439878095e-05, + "loss": 1.2169, + "step": 1562 + }, + { + "epoch": 0.09496324199526095, + "grad_norm": 3.4105563163757324, + "learning_rate": 9.791961268563368e-05, + "loss": 1.2199, + "step": 1563 + }, + { + "epoch": 0.09502399902788748, + "grad_norm": 0.9619441032409668, + "learning_rate": 9.791687921598374e-05, + "loss": 1.1351, + "step": 1564 + }, + { + "epoch": 0.095084756060514, + "grad_norm": 0.43657463788986206, + "learning_rate": 9.791414398993133e-05, + "loss": 1.0869, + "step": 1565 + }, + { + "epoch": 0.09514551309314054, + "grad_norm": 1.0621010065078735, + "learning_rate": 9.791140700757671e-05, + "loss": 1.2738, + "step": 1566 + }, + { + "epoch": 0.09520627012576706, + "grad_norm": 0.2834378480911255, + "learning_rate": 9.790866826902021e-05, + "loss": 1.0652, + "step": 1567 + }, + { + "epoch": 0.09526702715839358, + "grad_norm": 0.678031861782074, + "learning_rate": 9.790592777436221e-05, + "loss": 1.1223, + "step": 1568 + }, + { + "epoch": 0.09532778419102012, + "grad_norm": 0.809569776058197, + "learning_rate": 9.790318552370317e-05, + "loss": 1.2859, + "step": 1569 + }, + { + "epoch": 0.09538854122364664, + "grad_norm": 0.5147567391395569, + "learning_rate": 9.79004415171436e-05, + "loss": 1.1964, + "step": 1570 + }, + { + "epoch": 0.09544929825627316, + "grad_norm": 0.7337372303009033, + "learning_rate": 9.78976957547841e-05, + "loss": 1.2243, + "step": 1571 + }, + { + "epoch": 0.09551005528889969, + "grad_norm": 0.5374388694763184, + "learning_rate": 9.789494823672529e-05, + "loss": 1.1082, + "step": 1572 + }, + { + "epoch": 0.09557081232152621, + "grad_norm": 0.7596743702888489, + "learning_rate": 9.78921989630679e-05, + "loss": 1.1206, + "step": 1573 + }, + { + "epoch": 0.09563156935415275, + "grad_norm": 0.7003148794174194, + "learning_rate": 9.78894479339127e-05, + "loss": 1.1171, + "step": 1574 + }, + { + "epoch": 0.09569232638677927, + "grad_norm": 0.5913709402084351, + "learning_rate": 9.788669514936052e-05, + "loss": 1.25, + "step": 1575 + }, + { + "epoch": 0.09575308341940579, + "grad_norm": 0.822468638420105, + "learning_rate": 9.788394060951229e-05, + "loss": 1.1024, + "step": 1576 + }, + { + "epoch": 0.09581384045203233, + "grad_norm": 0.30357545614242554, + "learning_rate": 9.788118431446894e-05, + "loss": 1.1399, + "step": 1577 + }, + { + "epoch": 0.09587459748465885, + "grad_norm": 2.112104654312134, + "learning_rate": 9.787842626433154e-05, + "loss": 1.1581, + "step": 1578 + }, + { + "epoch": 0.09593535451728538, + "grad_norm": 1.3856312036514282, + "learning_rate": 9.787566645920116e-05, + "loss": 1.0691, + "step": 1579 + }, + { + "epoch": 0.0959961115499119, + "grad_norm": 0.30305448174476624, + "learning_rate": 9.787290489917899e-05, + "loss": 1.2203, + "step": 1580 + }, + { + "epoch": 0.09605686858253842, + "grad_norm": 1.169049620628357, + "learning_rate": 9.787014158436623e-05, + "loss": 1.1089, + "step": 1581 + }, + { + "epoch": 0.09611762561516496, + "grad_norm": 0.6660367846488953, + "learning_rate": 9.786737651486416e-05, + "loss": 1.2146, + "step": 1582 + }, + { + "epoch": 0.09617838264779148, + "grad_norm": 0.6673032641410828, + "learning_rate": 9.786460969077417e-05, + "loss": 1.2728, + "step": 1583 + }, + { + "epoch": 0.09623913968041801, + "grad_norm": 0.7762752175331116, + "learning_rate": 9.786184111219765e-05, + "loss": 1.0944, + "step": 1584 + }, + { + "epoch": 0.09629989671304454, + "grad_norm": 0.23842112720012665, + "learning_rate": 9.78590707792361e-05, + "loss": 1.2884, + "step": 1585 + }, + { + "epoch": 0.09636065374567106, + "grad_norm": 8.788839340209961, + "learning_rate": 9.785629869199104e-05, + "loss": 1.1375, + "step": 1586 + }, + { + "epoch": 0.09642141077829759, + "grad_norm": 1.700287103652954, + "learning_rate": 9.785352485056413e-05, + "loss": 1.1029, + "step": 1587 + }, + { + "epoch": 0.09648216781092411, + "grad_norm": 4.368428707122803, + "learning_rate": 9.7850749255057e-05, + "loss": 1.2745, + "step": 1588 + }, + { + "epoch": 0.09654292484355063, + "grad_norm": 0.49327319860458374, + "learning_rate": 9.784797190557142e-05, + "loss": 1.2051, + "step": 1589 + }, + { + "epoch": 0.09660368187617717, + "grad_norm": 0.9215505719184875, + "learning_rate": 9.784519280220916e-05, + "loss": 1.0808, + "step": 1590 + }, + { + "epoch": 0.09666443890880369, + "grad_norm": 0.2542787194252014, + "learning_rate": 9.784241194507212e-05, + "loss": 1.1336, + "step": 1591 + }, + { + "epoch": 0.09672519594143023, + "grad_norm": 0.3423195779323578, + "learning_rate": 9.783962933426223e-05, + "loss": 1.1235, + "step": 1592 + }, + { + "epoch": 0.09678595297405675, + "grad_norm": 0.2905953526496887, + "learning_rate": 9.783684496988147e-05, + "loss": 1.1091, + "step": 1593 + }, + { + "epoch": 0.09684671000668327, + "grad_norm": 0.3346598148345947, + "learning_rate": 9.783405885203192e-05, + "loss": 1.1598, + "step": 1594 + }, + { + "epoch": 0.0969074670393098, + "grad_norm": 0.37749120593070984, + "learning_rate": 9.783127098081569e-05, + "loss": 1.2355, + "step": 1595 + }, + { + "epoch": 0.09696822407193632, + "grad_norm": 0.24154165387153625, + "learning_rate": 9.782848135633497e-05, + "loss": 1.2002, + "step": 1596 + }, + { + "epoch": 0.09702898110456286, + "grad_norm": 0.2090793401002884, + "learning_rate": 9.782568997869203e-05, + "loss": 1.1565, + "step": 1597 + }, + { + "epoch": 0.09708973813718938, + "grad_norm": 0.39691615104675293, + "learning_rate": 9.782289684798917e-05, + "loss": 1.2095, + "step": 1598 + }, + { + "epoch": 0.0971504951698159, + "grad_norm": 0.2281525433063507, + "learning_rate": 9.78201019643288e-05, + "loss": 1.1239, + "step": 1599 + }, + { + "epoch": 0.09721125220244244, + "grad_norm": 0.30582958459854126, + "learning_rate": 9.781730532781333e-05, + "loss": 1.1778, + "step": 1600 + }, + { + "epoch": 0.09727200923506896, + "grad_norm": 0.33836454153060913, + "learning_rate": 9.781450693854531e-05, + "loss": 1.1324, + "step": 1601 + }, + { + "epoch": 0.09733276626769549, + "grad_norm": 1.0696678161621094, + "learning_rate": 9.781170679662728e-05, + "loss": 1.148, + "step": 1602 + }, + { + "epoch": 0.09739352330032201, + "grad_norm": 0.5562969446182251, + "learning_rate": 9.78089049021619e-05, + "loss": 1.2595, + "step": 1603 + }, + { + "epoch": 0.09745428033294853, + "grad_norm": 0.27862802147865295, + "learning_rate": 9.780610125525187e-05, + "loss": 1.2136, + "step": 1604 + }, + { + "epoch": 0.09751503736557507, + "grad_norm": 0.45213550329208374, + "learning_rate": 9.780329585599995e-05, + "loss": 1.0892, + "step": 1605 + }, + { + "epoch": 0.09757579439820159, + "grad_norm": 0.22858665883541107, + "learning_rate": 9.7800488704509e-05, + "loss": 1.0486, + "step": 1606 + }, + { + "epoch": 0.09763655143082813, + "grad_norm": 0.2135869562625885, + "learning_rate": 9.779767980088189e-05, + "loss": 1.1405, + "step": 1607 + }, + { + "epoch": 0.09769730846345465, + "grad_norm": 0.25731176137924194, + "learning_rate": 9.779486914522157e-05, + "loss": 1.1991, + "step": 1608 + }, + { + "epoch": 0.09775806549608117, + "grad_norm": 0.5220553278923035, + "learning_rate": 9.779205673763109e-05, + "loss": 1.3185, + "step": 1609 + }, + { + "epoch": 0.0978188225287077, + "grad_norm": 0.2193889170885086, + "learning_rate": 9.778924257821353e-05, + "loss": 1.0938, + "step": 1610 + }, + { + "epoch": 0.09787957956133422, + "grad_norm": 0.41876065731048584, + "learning_rate": 9.778642666707206e-05, + "loss": 1.1373, + "step": 1611 + }, + { + "epoch": 0.09794033659396074, + "grad_norm": 0.31580841541290283, + "learning_rate": 9.778360900430987e-05, + "loss": 1.2198, + "step": 1612 + }, + { + "epoch": 0.09800109362658728, + "grad_norm": 0.26740050315856934, + "learning_rate": 9.778078959003027e-05, + "loss": 1.1351, + "step": 1613 + }, + { + "epoch": 0.0980618506592138, + "grad_norm": 0.7678807377815247, + "learning_rate": 9.777796842433658e-05, + "loss": 1.3623, + "step": 1614 + }, + { + "epoch": 0.09812260769184034, + "grad_norm": 0.22142338752746582, + "learning_rate": 9.777514550733224e-05, + "loss": 1.0931, + "step": 1615 + }, + { + "epoch": 0.09818336472446686, + "grad_norm": 1.0666821002960205, + "learning_rate": 9.777232083912069e-05, + "loss": 1.2193, + "step": 1616 + }, + { + "epoch": 0.09824412175709338, + "grad_norm": 0.2140878289937973, + "learning_rate": 9.776949441980549e-05, + "loss": 1.1295, + "step": 1617 + }, + { + "epoch": 0.09830487878971991, + "grad_norm": 0.2419116199016571, + "learning_rate": 9.776666624949023e-05, + "loss": 1.1836, + "step": 1618 + }, + { + "epoch": 0.09836563582234643, + "grad_norm": 0.3117300570011139, + "learning_rate": 9.77638363282786e-05, + "loss": 1.2045, + "step": 1619 + }, + { + "epoch": 0.09842639285497297, + "grad_norm": 0.3185001313686371, + "learning_rate": 9.77610046562743e-05, + "loss": 1.1772, + "step": 1620 + }, + { + "epoch": 0.09848714988759949, + "grad_norm": 0.3699362576007843, + "learning_rate": 9.775817123358116e-05, + "loss": 1.2378, + "step": 1621 + }, + { + "epoch": 0.09854790692022601, + "grad_norm": 2.33440899848938, + "learning_rate": 9.7755336060303e-05, + "loss": 1.1965, + "step": 1622 + }, + { + "epoch": 0.09860866395285255, + "grad_norm": 5.425417423248291, + "learning_rate": 9.775249913654378e-05, + "loss": 1.3076, + "step": 1623 + }, + { + "epoch": 0.09866942098547907, + "grad_norm": 0.8368164300918579, + "learning_rate": 9.774966046240748e-05, + "loss": 1.2967, + "step": 1624 + }, + { + "epoch": 0.0987301780181056, + "grad_norm": 0.2417559176683426, + "learning_rate": 9.774682003799813e-05, + "loss": 1.1162, + "step": 1625 + }, + { + "epoch": 0.09879093505073212, + "grad_norm": 0.42883631587028503, + "learning_rate": 9.774397786341988e-05, + "loss": 1.1637, + "step": 1626 + }, + { + "epoch": 0.09885169208335864, + "grad_norm": 0.5557936429977417, + "learning_rate": 9.774113393877688e-05, + "loss": 1.0748, + "step": 1627 + }, + { + "epoch": 0.09891244911598518, + "grad_norm": 0.35464271903038025, + "learning_rate": 9.77382882641734e-05, + "loss": 1.2331, + "step": 1628 + }, + { + "epoch": 0.0989732061486117, + "grad_norm": 0.25659364461898804, + "learning_rate": 9.773544083971372e-05, + "loss": 1.2158, + "step": 1629 + }, + { + "epoch": 0.09903396318123822, + "grad_norm": 0.3499470353126526, + "learning_rate": 9.773259166550224e-05, + "loss": 1.1521, + "step": 1630 + }, + { + "epoch": 0.09909472021386476, + "grad_norm": 0.2286265790462494, + "learning_rate": 9.772974074164338e-05, + "loss": 1.1521, + "step": 1631 + }, + { + "epoch": 0.09915547724649128, + "grad_norm": 0.36820587515830994, + "learning_rate": 9.772688806824164e-05, + "loss": 1.1542, + "step": 1632 + }, + { + "epoch": 0.09921623427911781, + "grad_norm": 6.930588245391846, + "learning_rate": 9.772403364540161e-05, + "loss": 1.2179, + "step": 1633 + }, + { + "epoch": 0.09927699131174433, + "grad_norm": 0.5297050476074219, + "learning_rate": 9.77211774732279e-05, + "loss": 1.1366, + "step": 1634 + }, + { + "epoch": 0.09933774834437085, + "grad_norm": 0.22831743955612183, + "learning_rate": 9.77183195518252e-05, + "loss": 1.1106, + "step": 1635 + }, + { + "epoch": 0.09939850537699739, + "grad_norm": 0.4506749212741852, + "learning_rate": 9.771545988129827e-05, + "loss": 1.32, + "step": 1636 + }, + { + "epoch": 0.09945926240962391, + "grad_norm": 0.37910062074661255, + "learning_rate": 9.771259846175193e-05, + "loss": 1.1941, + "step": 1637 + }, + { + "epoch": 0.09952001944225045, + "grad_norm": 0.30395469069480896, + "learning_rate": 9.770973529329108e-05, + "loss": 1.0616, + "step": 1638 + }, + { + "epoch": 0.09958077647487697, + "grad_norm": 0.6610385179519653, + "learning_rate": 9.770687037602066e-05, + "loss": 1.0793, + "step": 1639 + }, + { + "epoch": 0.09964153350750349, + "grad_norm": 0.2210586667060852, + "learning_rate": 9.770400371004568e-05, + "loss": 1.1435, + "step": 1640 + }, + { + "epoch": 0.09970229054013002, + "grad_norm": 2.1089587211608887, + "learning_rate": 9.770113529547123e-05, + "loss": 1.2461, + "step": 1641 + }, + { + "epoch": 0.09976304757275654, + "grad_norm": 0.4501532018184662, + "learning_rate": 9.769826513240244e-05, + "loss": 1.2204, + "step": 1642 + }, + { + "epoch": 0.09982380460538308, + "grad_norm": 2.594223737716675, + "learning_rate": 9.769539322094453e-05, + "loss": 1.3475, + "step": 1643 + }, + { + "epoch": 0.0998845616380096, + "grad_norm": 0.3882279396057129, + "learning_rate": 9.769251956120276e-05, + "loss": 1.2396, + "step": 1644 + }, + { + "epoch": 0.09994531867063612, + "grad_norm": 0.24910221993923187, + "learning_rate": 9.768964415328245e-05, + "loss": 1.1389, + "step": 1645 + }, + { + "epoch": 0.10000607570326266, + "grad_norm": 0.20524665713310242, + "learning_rate": 9.768676699728903e-05, + "loss": 1.0906, + "step": 1646 + }, + { + "epoch": 0.10006683273588918, + "grad_norm": 0.21005727350711823, + "learning_rate": 9.768388809332796e-05, + "loss": 1.108, + "step": 1647 + }, + { + "epoch": 0.10012758976851571, + "grad_norm": 0.24129657447338104, + "learning_rate": 9.768100744150473e-05, + "loss": 1.2497, + "step": 1648 + }, + { + "epoch": 0.10018834680114223, + "grad_norm": 0.2664303481578827, + "learning_rate": 9.767812504192497e-05, + "loss": 1.0571, + "step": 1649 + }, + { + "epoch": 0.10024910383376875, + "grad_norm": 0.5384125113487244, + "learning_rate": 9.767524089469432e-05, + "loss": 1.2707, + "step": 1650 + }, + { + "epoch": 0.10030986086639529, + "grad_norm": 0.35119715332984924, + "learning_rate": 9.76723549999185e-05, + "loss": 1.0686, + "step": 1651 + }, + { + "epoch": 0.10037061789902181, + "grad_norm": 0.23146125674247742, + "learning_rate": 9.766946735770328e-05, + "loss": 1.0982, + "step": 1652 + }, + { + "epoch": 0.10043137493164833, + "grad_norm": 0.21282026171684265, + "learning_rate": 9.766657796815453e-05, + "loss": 1.0649, + "step": 1653 + }, + { + "epoch": 0.10049213196427487, + "grad_norm": 0.37460172176361084, + "learning_rate": 9.766368683137815e-05, + "loss": 1.2775, + "step": 1654 + }, + { + "epoch": 0.10055288899690139, + "grad_norm": 0.19907572865486145, + "learning_rate": 9.766079394748012e-05, + "loss": 1.1302, + "step": 1655 + }, + { + "epoch": 0.10061364602952792, + "grad_norm": 0.5494251251220703, + "learning_rate": 9.765789931656646e-05, + "loss": 1.3188, + "step": 1656 + }, + { + "epoch": 0.10067440306215444, + "grad_norm": 0.41927623748779297, + "learning_rate": 9.765500293874331e-05, + "loss": 1.3667, + "step": 1657 + }, + { + "epoch": 0.10073516009478097, + "grad_norm": 0.32295456528663635, + "learning_rate": 9.765210481411681e-05, + "loss": 1.1622, + "step": 1658 + }, + { + "epoch": 0.1007959171274075, + "grad_norm": 0.4438265264034271, + "learning_rate": 9.764920494279319e-05, + "loss": 1.135, + "step": 1659 + }, + { + "epoch": 0.10085667416003402, + "grad_norm": 0.42376741766929626, + "learning_rate": 9.764630332487876e-05, + "loss": 1.1317, + "step": 1660 + }, + { + "epoch": 0.10091743119266056, + "grad_norm": 0.35175731778144836, + "learning_rate": 9.764339996047987e-05, + "loss": 1.3031, + "step": 1661 + }, + { + "epoch": 0.10097818822528708, + "grad_norm": 0.3289128541946411, + "learning_rate": 9.764049484970293e-05, + "loss": 1.2934, + "step": 1662 + }, + { + "epoch": 0.1010389452579136, + "grad_norm": 0.2322559505701065, + "learning_rate": 9.763758799265446e-05, + "loss": 1.1207, + "step": 1663 + }, + { + "epoch": 0.10109970229054013, + "grad_norm": 0.19285151362419128, + "learning_rate": 9.763467938944098e-05, + "loss": 1.0876, + "step": 1664 + }, + { + "epoch": 0.10116045932316665, + "grad_norm": 0.7607999444007874, + "learning_rate": 9.763176904016913e-05, + "loss": 1.1637, + "step": 1665 + }, + { + "epoch": 0.10122121635579319, + "grad_norm": 0.2592668831348419, + "learning_rate": 9.762885694494558e-05, + "loss": 1.1406, + "step": 1666 + }, + { + "epoch": 0.10128197338841971, + "grad_norm": 2.214916467666626, + "learning_rate": 9.762594310387707e-05, + "loss": 1.1318, + "step": 1667 + }, + { + "epoch": 0.10134273042104623, + "grad_norm": 0.27004650235176086, + "learning_rate": 9.762302751707039e-05, + "loss": 1.0708, + "step": 1668 + }, + { + "epoch": 0.10140348745367277, + "grad_norm": 0.2551700472831726, + "learning_rate": 9.762011018463246e-05, + "loss": 1.1195, + "step": 1669 + }, + { + "epoch": 0.10146424448629929, + "grad_norm": 0.27055346965789795, + "learning_rate": 9.761719110667019e-05, + "loss": 1.1105, + "step": 1670 + }, + { + "epoch": 0.10152500151892582, + "grad_norm": 0.33498701453208923, + "learning_rate": 9.761427028329056e-05, + "loss": 1.2212, + "step": 1671 + }, + { + "epoch": 0.10158575855155234, + "grad_norm": 0.2240496575832367, + "learning_rate": 9.761134771460064e-05, + "loss": 1.0487, + "step": 1672 + }, + { + "epoch": 0.10164651558417886, + "grad_norm": 0.41090425848960876, + "learning_rate": 9.760842340070758e-05, + "loss": 1.0746, + "step": 1673 + }, + { + "epoch": 0.1017072726168054, + "grad_norm": 0.40921294689178467, + "learning_rate": 9.760549734171856e-05, + "loss": 1.1987, + "step": 1674 + }, + { + "epoch": 0.10176802964943192, + "grad_norm": 0.27647605538368225, + "learning_rate": 9.760256953774083e-05, + "loss": 1.1325, + "step": 1675 + }, + { + "epoch": 0.10182878668205844, + "grad_norm": 0.48200803995132446, + "learning_rate": 9.759963998888171e-05, + "loss": 1.1165, + "step": 1676 + }, + { + "epoch": 0.10188954371468498, + "grad_norm": 0.26628968119621277, + "learning_rate": 9.759670869524858e-05, + "loss": 1.1922, + "step": 1677 + }, + { + "epoch": 0.1019503007473115, + "grad_norm": 0.29594507813453674, + "learning_rate": 9.75937756569489e-05, + "loss": 1.2677, + "step": 1678 + }, + { + "epoch": 0.10201105777993803, + "grad_norm": 0.4247731566429138, + "learning_rate": 9.759084087409017e-05, + "loss": 1.1735, + "step": 1679 + }, + { + "epoch": 0.10207181481256455, + "grad_norm": 0.2708252966403961, + "learning_rate": 9.758790434677998e-05, + "loss": 1.1549, + "step": 1680 + }, + { + "epoch": 0.10213257184519108, + "grad_norm": 1.6024590730667114, + "learning_rate": 9.758496607512594e-05, + "loss": 1.2997, + "step": 1681 + }, + { + "epoch": 0.10219332887781761, + "grad_norm": 0.5131689310073853, + "learning_rate": 9.758202605923579e-05, + "loss": 1.1872, + "step": 1682 + }, + { + "epoch": 0.10225408591044413, + "grad_norm": 0.257293701171875, + "learning_rate": 9.757908429921725e-05, + "loss": 1.2008, + "step": 1683 + }, + { + "epoch": 0.10231484294307067, + "grad_norm": 1.0415771007537842, + "learning_rate": 9.757614079517821e-05, + "loss": 1.1228, + "step": 1684 + }, + { + "epoch": 0.10237559997569719, + "grad_norm": 0.2493562251329422, + "learning_rate": 9.75731955472265e-05, + "loss": 1.1507, + "step": 1685 + }, + { + "epoch": 0.10243635700832371, + "grad_norm": 0.27078208327293396, + "learning_rate": 9.757024855547014e-05, + "loss": 1.1395, + "step": 1686 + }, + { + "epoch": 0.10249711404095024, + "grad_norm": 0.4143364131450653, + "learning_rate": 9.756729982001712e-05, + "loss": 1.2472, + "step": 1687 + }, + { + "epoch": 0.10255787107357676, + "grad_norm": 0.3801232576370239, + "learning_rate": 9.756434934097552e-05, + "loss": 1.1841, + "step": 1688 + }, + { + "epoch": 0.1026186281062033, + "grad_norm": 0.3787832260131836, + "learning_rate": 9.756139711845349e-05, + "loss": 1.2005, + "step": 1689 + }, + { + "epoch": 0.10267938513882982, + "grad_norm": 0.6781951189041138, + "learning_rate": 9.755844315255928e-05, + "loss": 1.3485, + "step": 1690 + }, + { + "epoch": 0.10274014217145634, + "grad_norm": 0.5911668539047241, + "learning_rate": 9.755548744340112e-05, + "loss": 1.0902, + "step": 1691 + }, + { + "epoch": 0.10280089920408288, + "grad_norm": 0.22285832464694977, + "learning_rate": 9.75525299910874e-05, + "loss": 1.1043, + "step": 1692 + }, + { + "epoch": 0.1028616562367094, + "grad_norm": 0.41297417879104614, + "learning_rate": 9.754957079572649e-05, + "loss": 1.1488, + "step": 1693 + }, + { + "epoch": 0.10292241326933592, + "grad_norm": 0.23495450615882874, + "learning_rate": 9.754660985742688e-05, + "loss": 1.2504, + "step": 1694 + }, + { + "epoch": 0.10298317030196245, + "grad_norm": 0.3729107677936554, + "learning_rate": 9.754364717629707e-05, + "loss": 1.066, + "step": 1695 + }, + { + "epoch": 0.10304392733458898, + "grad_norm": 0.3694894015789032, + "learning_rate": 9.75406827524457e-05, + "loss": 1.2013, + "step": 1696 + }, + { + "epoch": 0.10310468436721551, + "grad_norm": 0.21680964529514313, + "learning_rate": 9.753771658598141e-05, + "loss": 1.1464, + "step": 1697 + }, + { + "epoch": 0.10316544139984203, + "grad_norm": 0.35377949476242065, + "learning_rate": 9.753474867701294e-05, + "loss": 1.1251, + "step": 1698 + }, + { + "epoch": 0.10322619843246855, + "grad_norm": 0.2200571447610855, + "learning_rate": 9.753177902564905e-05, + "loss": 1.1777, + "step": 1699 + }, + { + "epoch": 0.10328695546509509, + "grad_norm": 0.26050424575805664, + "learning_rate": 9.752880763199864e-05, + "loss": 1.119, + "step": 1700 + }, + { + "epoch": 0.10334771249772161, + "grad_norm": 0.23730768263339996, + "learning_rate": 9.752583449617058e-05, + "loss": 1.1861, + "step": 1701 + }, + { + "epoch": 0.10340846953034814, + "grad_norm": 0.24562667310237885, + "learning_rate": 9.752285961827388e-05, + "loss": 1.1404, + "step": 1702 + }, + { + "epoch": 0.10346922656297466, + "grad_norm": 0.20798423886299133, + "learning_rate": 9.751988299841756e-05, + "loss": 1.1198, + "step": 1703 + }, + { + "epoch": 0.10352998359560119, + "grad_norm": 0.19898158311843872, + "learning_rate": 9.751690463671075e-05, + "loss": 1.0825, + "step": 1704 + }, + { + "epoch": 0.10359074062822772, + "grad_norm": 0.3932807147502899, + "learning_rate": 9.751392453326262e-05, + "loss": 1.1323, + "step": 1705 + }, + { + "epoch": 0.10365149766085424, + "grad_norm": 0.23124247789382935, + "learning_rate": 9.751094268818239e-05, + "loss": 1.1539, + "step": 1706 + }, + { + "epoch": 0.10371225469348078, + "grad_norm": 0.8822628855705261, + "learning_rate": 9.750795910157939e-05, + "loss": 1.1245, + "step": 1707 + }, + { + "epoch": 0.1037730117261073, + "grad_norm": 0.27742400765419006, + "learning_rate": 9.750497377356296e-05, + "loss": 1.1801, + "step": 1708 + }, + { + "epoch": 0.10383376875873382, + "grad_norm": 0.4829384386539459, + "learning_rate": 9.750198670424254e-05, + "loss": 1.1875, + "step": 1709 + }, + { + "epoch": 0.10389452579136035, + "grad_norm": 0.38290831446647644, + "learning_rate": 9.749899789372759e-05, + "loss": 1.195, + "step": 1710 + }, + { + "epoch": 0.10395528282398687, + "grad_norm": 0.42251408100128174, + "learning_rate": 9.749600734212771e-05, + "loss": 1.1189, + "step": 1711 + }, + { + "epoch": 0.10401603985661341, + "grad_norm": 1.1835583448410034, + "learning_rate": 9.74930150495525e-05, + "loss": 1.0961, + "step": 1712 + }, + { + "epoch": 0.10407679688923993, + "grad_norm": 0.21068885922431946, + "learning_rate": 9.749002101611164e-05, + "loss": 1.1448, + "step": 1713 + }, + { + "epoch": 0.10413755392186645, + "grad_norm": 3.1366095542907715, + "learning_rate": 9.748702524191489e-05, + "loss": 1.1587, + "step": 1714 + }, + { + "epoch": 0.10419831095449299, + "grad_norm": 0.8137587308883667, + "learning_rate": 9.748402772707204e-05, + "loss": 1.0996, + "step": 1715 + }, + { + "epoch": 0.10425906798711951, + "grad_norm": 0.2750990390777588, + "learning_rate": 9.748102847169296e-05, + "loss": 1.1115, + "step": 1716 + }, + { + "epoch": 0.10431982501974603, + "grad_norm": 0.6004108786582947, + "learning_rate": 9.747802747588763e-05, + "loss": 1.275, + "step": 1717 + }, + { + "epoch": 0.10438058205237256, + "grad_norm": 1.5282827615737915, + "learning_rate": 9.747502473976602e-05, + "loss": 1.327, + "step": 1718 + }, + { + "epoch": 0.10444133908499909, + "grad_norm": 1.2948552370071411, + "learning_rate": 9.74720202634382e-05, + "loss": 1.1172, + "step": 1719 + }, + { + "epoch": 0.10450209611762562, + "grad_norm": 0.8784529566764832, + "learning_rate": 9.74690140470143e-05, + "loss": 1.2251, + "step": 1720 + }, + { + "epoch": 0.10456285315025214, + "grad_norm": 0.5034112930297852, + "learning_rate": 9.746600609060451e-05, + "loss": 1.1276, + "step": 1721 + }, + { + "epoch": 0.10462361018287866, + "grad_norm": 0.5372108817100525, + "learning_rate": 9.746299639431909e-05, + "loss": 1.1555, + "step": 1722 + }, + { + "epoch": 0.1046843672155052, + "grad_norm": 0.3265959322452545, + "learning_rate": 9.745998495826836e-05, + "loss": 1.1565, + "step": 1723 + }, + { + "epoch": 0.10474512424813172, + "grad_norm": 0.4325443208217621, + "learning_rate": 9.745697178256273e-05, + "loss": 1.1154, + "step": 1724 + }, + { + "epoch": 0.10480588128075825, + "grad_norm": 0.6536804437637329, + "learning_rate": 9.74539568673126e-05, + "loss": 1.2605, + "step": 1725 + }, + { + "epoch": 0.10486663831338477, + "grad_norm": 0.2933708131313324, + "learning_rate": 9.745094021262852e-05, + "loss": 1.0795, + "step": 1726 + }, + { + "epoch": 0.1049273953460113, + "grad_norm": 0.6307336688041687, + "learning_rate": 9.744792181862106e-05, + "loss": 1.2304, + "step": 1727 + }, + { + "epoch": 0.10498815237863783, + "grad_norm": 0.39969977736473083, + "learning_rate": 9.744490168540083e-05, + "loss": 1.178, + "step": 1728 + }, + { + "epoch": 0.10504890941126435, + "grad_norm": 0.4638993442058563, + "learning_rate": 9.74418798130786e-05, + "loss": 1.2375, + "step": 1729 + }, + { + "epoch": 0.10510966644389089, + "grad_norm": 0.6738314032554626, + "learning_rate": 9.743885620176506e-05, + "loss": 1.1273, + "step": 1730 + }, + { + "epoch": 0.10517042347651741, + "grad_norm": 0.4268384575843811, + "learning_rate": 9.743583085157109e-05, + "loss": 1.2204, + "step": 1731 + }, + { + "epoch": 0.10523118050914393, + "grad_norm": 0.7052971124649048, + "learning_rate": 9.743280376260758e-05, + "loss": 1.223, + "step": 1732 + }, + { + "epoch": 0.10529193754177046, + "grad_norm": 0.4514221251010895, + "learning_rate": 9.742977493498546e-05, + "loss": 1.1037, + "step": 1733 + }, + { + "epoch": 0.10535269457439699, + "grad_norm": 0.4445760250091553, + "learning_rate": 9.742674436881578e-05, + "loss": 1.0986, + "step": 1734 + }, + { + "epoch": 0.1054134516070235, + "grad_norm": 0.3338969051837921, + "learning_rate": 9.742371206420962e-05, + "loss": 1.1914, + "step": 1735 + }, + { + "epoch": 0.10547420863965004, + "grad_norm": 0.29008111357688904, + "learning_rate": 9.742067802127812e-05, + "loss": 1.1056, + "step": 1736 + }, + { + "epoch": 0.10553496567227656, + "grad_norm": 0.3837815821170807, + "learning_rate": 9.74176422401325e-05, + "loss": 1.2195, + "step": 1737 + }, + { + "epoch": 0.1055957227049031, + "grad_norm": 0.4793948829174042, + "learning_rate": 9.741460472088404e-05, + "loss": 1.221, + "step": 1738 + }, + { + "epoch": 0.10565647973752962, + "grad_norm": 0.390137255191803, + "learning_rate": 9.74115654636441e-05, + "loss": 1.1535, + "step": 1739 + }, + { + "epoch": 0.10571723677015614, + "grad_norm": 0.34765323996543884, + "learning_rate": 9.740852446852403e-05, + "loss": 1.1011, + "step": 1740 + }, + { + "epoch": 0.10577799380278267, + "grad_norm": 0.34302106499671936, + "learning_rate": 9.740548173563536e-05, + "loss": 1.1106, + "step": 1741 + }, + { + "epoch": 0.1058387508354092, + "grad_norm": 0.8245735168457031, + "learning_rate": 9.740243726508957e-05, + "loss": 1.1042, + "step": 1742 + }, + { + "epoch": 0.10589950786803573, + "grad_norm": 2.8590786457061768, + "learning_rate": 9.739939105699829e-05, + "loss": 1.0953, + "step": 1743 + }, + { + "epoch": 0.10596026490066225, + "grad_norm": 0.340989887714386, + "learning_rate": 9.739634311147318e-05, + "loss": 1.1132, + "step": 1744 + }, + { + "epoch": 0.10602102193328877, + "grad_norm": 0.21168555319309235, + "learning_rate": 9.739329342862593e-05, + "loss": 1.166, + "step": 1745 + }, + { + "epoch": 0.10608177896591531, + "grad_norm": 0.3137030303478241, + "learning_rate": 9.739024200856836e-05, + "loss": 1.2394, + "step": 1746 + }, + { + "epoch": 0.10614253599854183, + "grad_norm": 0.2891295850276947, + "learning_rate": 9.738718885141231e-05, + "loss": 1.0894, + "step": 1747 + }, + { + "epoch": 0.10620329303116836, + "grad_norm": 0.21782030165195465, + "learning_rate": 9.73841339572697e-05, + "loss": 1.17, + "step": 1748 + }, + { + "epoch": 0.10626405006379488, + "grad_norm": 0.22449402511119843, + "learning_rate": 9.738107732625249e-05, + "loss": 1.1511, + "step": 1749 + }, + { + "epoch": 0.1063248070964214, + "grad_norm": 0.37193745374679565, + "learning_rate": 9.737801895847273e-05, + "loss": 1.1387, + "step": 1750 + }, + { + "epoch": 0.10638556412904794, + "grad_norm": 0.3650052547454834, + "learning_rate": 9.737495885404253e-05, + "loss": 1.1579, + "step": 1751 + }, + { + "epoch": 0.10644632116167446, + "grad_norm": 1.0710258483886719, + "learning_rate": 9.737189701307405e-05, + "loss": 1.1141, + "step": 1752 + }, + { + "epoch": 0.106507078194301, + "grad_norm": 0.19307558238506317, + "learning_rate": 9.736883343567955e-05, + "loss": 1.1029, + "step": 1753 + }, + { + "epoch": 0.10656783522692752, + "grad_norm": 0.2696276009082794, + "learning_rate": 9.736576812197129e-05, + "loss": 1.2935, + "step": 1754 + }, + { + "epoch": 0.10662859225955404, + "grad_norm": 0.3459651470184326, + "learning_rate": 9.736270107206164e-05, + "loss": 1.374, + "step": 1755 + }, + { + "epoch": 0.10668934929218057, + "grad_norm": 0.2929039001464844, + "learning_rate": 9.735963228606303e-05, + "loss": 1.0937, + "step": 1756 + }, + { + "epoch": 0.1067501063248071, + "grad_norm": 0.22432143986225128, + "learning_rate": 9.735656176408793e-05, + "loss": 1.1241, + "step": 1757 + }, + { + "epoch": 0.10681086335743362, + "grad_norm": 0.42875826358795166, + "learning_rate": 9.735348950624895e-05, + "loss": 1.085, + "step": 1758 + }, + { + "epoch": 0.10687162039006015, + "grad_norm": 0.2821659743785858, + "learning_rate": 9.735041551265862e-05, + "loss": 1.1571, + "step": 1759 + }, + { + "epoch": 0.10693237742268667, + "grad_norm": 0.27997884154319763, + "learning_rate": 9.734733978342965e-05, + "loss": 1.1244, + "step": 1760 + }, + { + "epoch": 0.10699313445531321, + "grad_norm": 0.27221789956092834, + "learning_rate": 9.734426231867482e-05, + "loss": 1.1395, + "step": 1761 + }, + { + "epoch": 0.10705389148793973, + "grad_norm": 0.30700260400772095, + "learning_rate": 9.734118311850688e-05, + "loss": 1.1704, + "step": 1762 + }, + { + "epoch": 0.10711464852056625, + "grad_norm": 1.1230804920196533, + "learning_rate": 9.733810218303873e-05, + "loss": 1.3641, + "step": 1763 + }, + { + "epoch": 0.10717540555319278, + "grad_norm": 1.050394892692566, + "learning_rate": 9.733501951238329e-05, + "loss": 1.2294, + "step": 1764 + }, + { + "epoch": 0.1072361625858193, + "grad_norm": 0.2548660337924957, + "learning_rate": 9.733193510665358e-05, + "loss": 1.0893, + "step": 1765 + }, + { + "epoch": 0.10729691961844584, + "grad_norm": 0.3848778307437897, + "learning_rate": 9.732884896596262e-05, + "loss": 1.124, + "step": 1766 + }, + { + "epoch": 0.10735767665107236, + "grad_norm": 0.19525152444839478, + "learning_rate": 9.732576109042355e-05, + "loss": 1.1078, + "step": 1767 + }, + { + "epoch": 0.10741843368369888, + "grad_norm": 0.6491109132766724, + "learning_rate": 9.732267148014955e-05, + "loss": 1.232, + "step": 1768 + }, + { + "epoch": 0.10747919071632542, + "grad_norm": 0.3085942268371582, + "learning_rate": 9.73195801352539e-05, + "loss": 1.1651, + "step": 1769 + }, + { + "epoch": 0.10753994774895194, + "grad_norm": 0.745150089263916, + "learning_rate": 9.731648705584989e-05, + "loss": 1.142, + "step": 1770 + }, + { + "epoch": 0.10760070478157847, + "grad_norm": 0.3338721692562103, + "learning_rate": 9.73133922420509e-05, + "loss": 1.2908, + "step": 1771 + }, + { + "epoch": 0.107661461814205, + "grad_norm": 0.3524227738380432, + "learning_rate": 9.731029569397037e-05, + "loss": 1.211, + "step": 1772 + }, + { + "epoch": 0.10772221884683152, + "grad_norm": 0.2828766107559204, + "learning_rate": 9.730719741172181e-05, + "loss": 1.2333, + "step": 1773 + }, + { + "epoch": 0.10778297587945805, + "grad_norm": 0.35445529222488403, + "learning_rate": 9.730409739541877e-05, + "loss": 1.3872, + "step": 1774 + }, + { + "epoch": 0.10784373291208457, + "grad_norm": 0.9497116208076477, + "learning_rate": 9.730099564517491e-05, + "loss": 1.1311, + "step": 1775 + }, + { + "epoch": 0.10790448994471111, + "grad_norm": 0.46369385719299316, + "learning_rate": 9.729789216110392e-05, + "loss": 1.1014, + "step": 1776 + }, + { + "epoch": 0.10796524697733763, + "grad_norm": 0.3550243377685547, + "learning_rate": 9.729478694331954e-05, + "loss": 1.1933, + "step": 1777 + }, + { + "epoch": 0.10802600400996415, + "grad_norm": 0.4471387565135956, + "learning_rate": 9.72916799919356e-05, + "loss": 1.2576, + "step": 1778 + }, + { + "epoch": 0.10808676104259068, + "grad_norm": 0.8278563618659973, + "learning_rate": 9.7288571307066e-05, + "loss": 1.2366, + "step": 1779 + }, + { + "epoch": 0.1081475180752172, + "grad_norm": 0.7949388027191162, + "learning_rate": 9.728546088882466e-05, + "loss": 1.1399, + "step": 1780 + }, + { + "epoch": 0.10820827510784373, + "grad_norm": 1.026513695716858, + "learning_rate": 9.728234873732562e-05, + "loss": 1.1155, + "step": 1781 + }, + { + "epoch": 0.10826903214047026, + "grad_norm": 0.256692111492157, + "learning_rate": 9.727923485268296e-05, + "loss": 1.1723, + "step": 1782 + }, + { + "epoch": 0.10832978917309678, + "grad_norm": 0.9380750060081482, + "learning_rate": 9.727611923501079e-05, + "loss": 1.2443, + "step": 1783 + }, + { + "epoch": 0.10839054620572332, + "grad_norm": 1.3167489767074585, + "learning_rate": 9.727300188442334e-05, + "loss": 1.1862, + "step": 1784 + }, + { + "epoch": 0.10845130323834984, + "grad_norm": 0.9672444462776184, + "learning_rate": 9.726988280103485e-05, + "loss": 1.1074, + "step": 1785 + }, + { + "epoch": 0.10851206027097636, + "grad_norm": 0.5226255655288696, + "learning_rate": 9.726676198495968e-05, + "loss": 1.2446, + "step": 1786 + }, + { + "epoch": 0.1085728173036029, + "grad_norm": 0.5178629755973816, + "learning_rate": 9.726363943631223e-05, + "loss": 1.132, + "step": 1787 + }, + { + "epoch": 0.10863357433622942, + "grad_norm": 0.4548768401145935, + "learning_rate": 9.726051515520692e-05, + "loss": 1.0825, + "step": 1788 + }, + { + "epoch": 0.10869433136885595, + "grad_norm": 0.3194658160209656, + "learning_rate": 9.725738914175831e-05, + "loss": 1.1774, + "step": 1789 + }, + { + "epoch": 0.10875508840148247, + "grad_norm": 0.6261642575263977, + "learning_rate": 9.725426139608094e-05, + "loss": 1.1318, + "step": 1790 + }, + { + "epoch": 0.108815845434109, + "grad_norm": 0.2752445936203003, + "learning_rate": 9.72511319182895e-05, + "loss": 1.2402, + "step": 1791 + }, + { + "epoch": 0.10887660246673553, + "grad_norm": 1.870620608329773, + "learning_rate": 9.724800070849869e-05, + "loss": 1.1922, + "step": 1792 + }, + { + "epoch": 0.10893735949936205, + "grad_norm": 1.0097769498825073, + "learning_rate": 9.724486776682328e-05, + "loss": 1.3976, + "step": 1793 + }, + { + "epoch": 0.10899811653198858, + "grad_norm": 1.7846604585647583, + "learning_rate": 9.72417330933781e-05, + "loss": 1.2002, + "step": 1794 + }, + { + "epoch": 0.1090588735646151, + "grad_norm": 0.3644181489944458, + "learning_rate": 9.723859668827809e-05, + "loss": 1.1357, + "step": 1795 + }, + { + "epoch": 0.10911963059724163, + "grad_norm": 0.5321322083473206, + "learning_rate": 9.723545855163816e-05, + "loss": 1.2843, + "step": 1796 + }, + { + "epoch": 0.10918038762986816, + "grad_norm": 2.3697009086608887, + "learning_rate": 9.723231868357337e-05, + "loss": 1.1634, + "step": 1797 + }, + { + "epoch": 0.10924114466249468, + "grad_norm": 0.2717841565608978, + "learning_rate": 9.722917708419881e-05, + "loss": 1.1253, + "step": 1798 + }, + { + "epoch": 0.1093019016951212, + "grad_norm": 0.28151172399520874, + "learning_rate": 9.722603375362963e-05, + "loss": 1.1275, + "step": 1799 + }, + { + "epoch": 0.10936265872774774, + "grad_norm": 0.2906649708747864, + "learning_rate": 9.722288869198106e-05, + "loss": 1.1923, + "step": 1800 + }, + { + "epoch": 0.10942341576037426, + "grad_norm": 0.38771432638168335, + "learning_rate": 9.721974189936837e-05, + "loss": 1.2277, + "step": 1801 + }, + { + "epoch": 0.1094841727930008, + "grad_norm": 1.0388009548187256, + "learning_rate": 9.721659337590693e-05, + "loss": 1.1209, + "step": 1802 + }, + { + "epoch": 0.10954492982562732, + "grad_norm": 0.7597936987876892, + "learning_rate": 9.721344312171213e-05, + "loss": 1.2356, + "step": 1803 + }, + { + "epoch": 0.10960568685825384, + "grad_norm": 0.3214944899082184, + "learning_rate": 9.721029113689943e-05, + "loss": 1.0857, + "step": 1804 + }, + { + "epoch": 0.10966644389088037, + "grad_norm": 0.3037773370742798, + "learning_rate": 9.72071374215844e-05, + "loss": 1.1337, + "step": 1805 + }, + { + "epoch": 0.10972720092350689, + "grad_norm": 0.5074493885040283, + "learning_rate": 9.72039819758826e-05, + "loss": 1.1133, + "step": 1806 + }, + { + "epoch": 0.10978795795613343, + "grad_norm": 0.31327539682388306, + "learning_rate": 9.720082479990974e-05, + "loss": 1.2214, + "step": 1807 + }, + { + "epoch": 0.10984871498875995, + "grad_norm": 0.34689489006996155, + "learning_rate": 9.719766589378152e-05, + "loss": 1.0817, + "step": 1808 + }, + { + "epoch": 0.10990947202138647, + "grad_norm": 0.2772051692008972, + "learning_rate": 9.719450525761373e-05, + "loss": 1.2037, + "step": 1809 + }, + { + "epoch": 0.109970229054013, + "grad_norm": 1.5915595293045044, + "learning_rate": 9.719134289152224e-05, + "loss": 1.1669, + "step": 1810 + }, + { + "epoch": 0.11003098608663953, + "grad_norm": 0.2747737169265747, + "learning_rate": 9.718817879562293e-05, + "loss": 1.1094, + "step": 1811 + }, + { + "epoch": 0.11009174311926606, + "grad_norm": 0.4412619471549988, + "learning_rate": 9.718501297003181e-05, + "loss": 1.2573, + "step": 1812 + }, + { + "epoch": 0.11015250015189258, + "grad_norm": 0.36585792899131775, + "learning_rate": 9.718184541486492e-05, + "loss": 1.1013, + "step": 1813 + }, + { + "epoch": 0.1102132571845191, + "grad_norm": 0.2986741065979004, + "learning_rate": 9.717867613023838e-05, + "loss": 1.1194, + "step": 1814 + }, + { + "epoch": 0.11027401421714564, + "grad_norm": 2.6618616580963135, + "learning_rate": 9.717550511626834e-05, + "loss": 1.1559, + "step": 1815 + }, + { + "epoch": 0.11033477124977216, + "grad_norm": 0.7550770044326782, + "learning_rate": 9.717233237307104e-05, + "loss": 1.2479, + "step": 1816 + }, + { + "epoch": 0.1103955282823987, + "grad_norm": 0.45871394872665405, + "learning_rate": 9.716915790076278e-05, + "loss": 1.2472, + "step": 1817 + }, + { + "epoch": 0.11045628531502522, + "grad_norm": 0.28900033235549927, + "learning_rate": 9.71659816994599e-05, + "loss": 1.1444, + "step": 1818 + }, + { + "epoch": 0.11051704234765174, + "grad_norm": 0.28553080558776855, + "learning_rate": 9.716280376927887e-05, + "loss": 1.1098, + "step": 1819 + }, + { + "epoch": 0.11057779938027827, + "grad_norm": 1.7440497875213623, + "learning_rate": 9.715962411033615e-05, + "loss": 1.1322, + "step": 1820 + }, + { + "epoch": 0.11063855641290479, + "grad_norm": 0.2957719564437866, + "learning_rate": 9.715644272274828e-05, + "loss": 1.2311, + "step": 1821 + }, + { + "epoch": 0.11069931344553131, + "grad_norm": 0.7647997736930847, + "learning_rate": 9.715325960663188e-05, + "loss": 1.4009, + "step": 1822 + }, + { + "epoch": 0.11076007047815785, + "grad_norm": 0.2915436625480652, + "learning_rate": 9.715007476210364e-05, + "loss": 1.1998, + "step": 1823 + }, + { + "epoch": 0.11082082751078437, + "grad_norm": 0.3427944481372833, + "learning_rate": 9.714688818928031e-05, + "loss": 1.092, + "step": 1824 + }, + { + "epoch": 0.1108815845434109, + "grad_norm": 0.3064861297607422, + "learning_rate": 9.714369988827867e-05, + "loss": 1.107, + "step": 1825 + }, + { + "epoch": 0.11094234157603743, + "grad_norm": 0.20696088671684265, + "learning_rate": 9.71405098592156e-05, + "loss": 1.1019, + "step": 1826 + }, + { + "epoch": 0.11100309860866395, + "grad_norm": 0.23702514171600342, + "learning_rate": 9.713731810220803e-05, + "loss": 1.0889, + "step": 1827 + }, + { + "epoch": 0.11106385564129048, + "grad_norm": 0.28070566058158875, + "learning_rate": 9.713412461737296e-05, + "loss": 1.166, + "step": 1828 + }, + { + "epoch": 0.111124612673917, + "grad_norm": 2.0678529739379883, + "learning_rate": 9.713092940482743e-05, + "loss": 1.1255, + "step": 1829 + }, + { + "epoch": 0.11118536970654354, + "grad_norm": 0.45067664980888367, + "learning_rate": 9.712773246468859e-05, + "loss": 1.2696, + "step": 1830 + }, + { + "epoch": 0.11124612673917006, + "grad_norm": 0.47404322028160095, + "learning_rate": 9.712453379707357e-05, + "loss": 1.1521, + "step": 1831 + }, + { + "epoch": 0.11130688377179658, + "grad_norm": 3.8739099502563477, + "learning_rate": 9.712133340209969e-05, + "loss": 1.2746, + "step": 1832 + }, + { + "epoch": 0.11136764080442312, + "grad_norm": 0.2852421998977661, + "learning_rate": 9.711813127988422e-05, + "loss": 1.08, + "step": 1833 + }, + { + "epoch": 0.11142839783704964, + "grad_norm": 0.25971418619155884, + "learning_rate": 9.711492743054453e-05, + "loss": 1.1509, + "step": 1834 + }, + { + "epoch": 0.11148915486967617, + "grad_norm": 0.2614261209964752, + "learning_rate": 9.711172185419808e-05, + "loss": 1.1331, + "step": 1835 + }, + { + "epoch": 0.11154991190230269, + "grad_norm": 0.6091518998146057, + "learning_rate": 9.710851455096235e-05, + "loss": 1.1394, + "step": 1836 + }, + { + "epoch": 0.11161066893492921, + "grad_norm": 1.1787514686584473, + "learning_rate": 9.71053055209549e-05, + "loss": 1.0872, + "step": 1837 + }, + { + "epoch": 0.11167142596755575, + "grad_norm": 0.34711092710494995, + "learning_rate": 9.71020947642934e-05, + "loss": 1.2034, + "step": 1838 + }, + { + "epoch": 0.11173218300018227, + "grad_norm": 0.30840054154396057, + "learning_rate": 9.70988822810955e-05, + "loss": 1.0995, + "step": 1839 + }, + { + "epoch": 0.11179294003280879, + "grad_norm": 0.3172672390937805, + "learning_rate": 9.709566807147895e-05, + "loss": 1.1807, + "step": 1840 + }, + { + "epoch": 0.11185369706543533, + "grad_norm": 0.2639411687850952, + "learning_rate": 9.70924521355616e-05, + "loss": 1.0709, + "step": 1841 + }, + { + "epoch": 0.11191445409806185, + "grad_norm": 0.20194025337696075, + "learning_rate": 9.70892344734613e-05, + "loss": 1.1454, + "step": 1842 + }, + { + "epoch": 0.11197521113068838, + "grad_norm": 0.45515477657318115, + "learning_rate": 9.708601508529602e-05, + "loss": 1.1138, + "step": 1843 + }, + { + "epoch": 0.1120359681633149, + "grad_norm": 0.2960728704929352, + "learning_rate": 9.708279397118376e-05, + "loss": 1.291, + "step": 1844 + }, + { + "epoch": 0.11209672519594142, + "grad_norm": 0.520592451095581, + "learning_rate": 9.707957113124257e-05, + "loss": 1.122, + "step": 1845 + }, + { + "epoch": 0.11215748222856796, + "grad_norm": 0.3725055754184723, + "learning_rate": 9.707634656559061e-05, + "loss": 1.1656, + "step": 1846 + }, + { + "epoch": 0.11221823926119448, + "grad_norm": 0.46529340744018555, + "learning_rate": 9.707312027434608e-05, + "loss": 1.2306, + "step": 1847 + }, + { + "epoch": 0.11227899629382102, + "grad_norm": 0.7003339529037476, + "learning_rate": 9.706989225762721e-05, + "loss": 1.1594, + "step": 1848 + }, + { + "epoch": 0.11233975332644754, + "grad_norm": 0.3014855980873108, + "learning_rate": 9.706666251555234e-05, + "loss": 1.2913, + "step": 1849 + }, + { + "epoch": 0.11240051035907406, + "grad_norm": 0.41318583488464355, + "learning_rate": 9.706343104823987e-05, + "loss": 1.2034, + "step": 1850 + }, + { + "epoch": 0.11246126739170059, + "grad_norm": 0.5326007008552551, + "learning_rate": 9.706019785580823e-05, + "loss": 1.1719, + "step": 1851 + }, + { + "epoch": 0.11252202442432711, + "grad_norm": 0.22465647757053375, + "learning_rate": 9.705696293837594e-05, + "loss": 1.2079, + "step": 1852 + }, + { + "epoch": 0.11258278145695365, + "grad_norm": 0.5792282223701477, + "learning_rate": 9.705372629606158e-05, + "loss": 1.1351, + "step": 1853 + }, + { + "epoch": 0.11264353848958017, + "grad_norm": 0.22790828347206116, + "learning_rate": 9.705048792898378e-05, + "loss": 1.2086, + "step": 1854 + }, + { + "epoch": 0.11270429552220669, + "grad_norm": 0.20958641171455383, + "learning_rate": 9.704724783726127e-05, + "loss": 1.1364, + "step": 1855 + }, + { + "epoch": 0.11276505255483323, + "grad_norm": 0.8518720865249634, + "learning_rate": 9.704400602101278e-05, + "loss": 1.1526, + "step": 1856 + }, + { + "epoch": 0.11282580958745975, + "grad_norm": 0.2794401943683624, + "learning_rate": 9.704076248035717e-05, + "loss": 1.2833, + "step": 1857 + }, + { + "epoch": 0.11288656662008628, + "grad_norm": 0.2753097712993622, + "learning_rate": 9.703751721541329e-05, + "loss": 1.1152, + "step": 1858 + }, + { + "epoch": 0.1129473236527128, + "grad_norm": 0.2931939363479614, + "learning_rate": 9.703427022630015e-05, + "loss": 1.2159, + "step": 1859 + }, + { + "epoch": 0.11300808068533932, + "grad_norm": 1.1634893417358398, + "learning_rate": 9.703102151313674e-05, + "loss": 1.2302, + "step": 1860 + }, + { + "epoch": 0.11306883771796586, + "grad_norm": 0.2761184871196747, + "learning_rate": 9.702777107604215e-05, + "loss": 1.1973, + "step": 1861 + }, + { + "epoch": 0.11312959475059238, + "grad_norm": 0.3362026512622833, + "learning_rate": 9.70245189151355e-05, + "loss": 1.091, + "step": 1862 + }, + { + "epoch": 0.1131903517832189, + "grad_norm": 0.21765007078647614, + "learning_rate": 9.702126503053603e-05, + "loss": 1.295, + "step": 1863 + }, + { + "epoch": 0.11325110881584544, + "grad_norm": 0.2185797393321991, + "learning_rate": 9.701800942236301e-05, + "loss": 1.1112, + "step": 1864 + }, + { + "epoch": 0.11331186584847196, + "grad_norm": 0.3310438394546509, + "learning_rate": 9.701475209073576e-05, + "loss": 1.1696, + "step": 1865 + }, + { + "epoch": 0.11337262288109849, + "grad_norm": 0.23154328763484955, + "learning_rate": 9.701149303577368e-05, + "loss": 1.2049, + "step": 1866 + }, + { + "epoch": 0.11343337991372501, + "grad_norm": 0.24431203305721283, + "learning_rate": 9.700823225759623e-05, + "loss": 1.114, + "step": 1867 + }, + { + "epoch": 0.11349413694635153, + "grad_norm": 1.5276261568069458, + "learning_rate": 9.700496975632295e-05, + "loss": 1.173, + "step": 1868 + }, + { + "epoch": 0.11355489397897807, + "grad_norm": 0.23430532217025757, + "learning_rate": 9.700170553207341e-05, + "loss": 1.1427, + "step": 1869 + }, + { + "epoch": 0.11361565101160459, + "grad_norm": 0.2724741995334625, + "learning_rate": 9.699843958496729e-05, + "loss": 1.155, + "step": 1870 + }, + { + "epoch": 0.11367640804423113, + "grad_norm": 0.24015015363693237, + "learning_rate": 9.699517191512426e-05, + "loss": 1.1652, + "step": 1871 + }, + { + "epoch": 0.11373716507685765, + "grad_norm": 0.309568852186203, + "learning_rate": 9.699190252266414e-05, + "loss": 1.1282, + "step": 1872 + }, + { + "epoch": 0.11379792210948417, + "grad_norm": 1.172499179840088, + "learning_rate": 9.698863140770673e-05, + "loss": 1.2369, + "step": 1873 + }, + { + "epoch": 0.1138586791421107, + "grad_norm": 0.19172875583171844, + "learning_rate": 9.698535857037197e-05, + "loss": 1.1245, + "step": 1874 + }, + { + "epoch": 0.11391943617473722, + "grad_norm": 0.22102876007556915, + "learning_rate": 9.698208401077979e-05, + "loss": 1.1306, + "step": 1875 + }, + { + "epoch": 0.11398019320736376, + "grad_norm": 0.328581303358078, + "learning_rate": 9.697880772905026e-05, + "loss": 1.3456, + "step": 1876 + }, + { + "epoch": 0.11404095023999028, + "grad_norm": 0.37319880723953247, + "learning_rate": 9.697552972530344e-05, + "loss": 1.3909, + "step": 1877 + }, + { + "epoch": 0.1141017072726168, + "grad_norm": 0.23405173420906067, + "learning_rate": 9.697224999965948e-05, + "loss": 1.3061, + "step": 1878 + }, + { + "epoch": 0.11416246430524334, + "grad_norm": 0.17760202288627625, + "learning_rate": 9.696896855223865e-05, + "loss": 1.0873, + "step": 1879 + }, + { + "epoch": 0.11422322133786986, + "grad_norm": 0.2311343103647232, + "learning_rate": 9.696568538316117e-05, + "loss": 1.1567, + "step": 1880 + }, + { + "epoch": 0.11428397837049639, + "grad_norm": 0.29994258284568787, + "learning_rate": 9.696240049254743e-05, + "loss": 1.1735, + "step": 1881 + }, + { + "epoch": 0.11434473540312291, + "grad_norm": 0.34590792655944824, + "learning_rate": 9.695911388051783e-05, + "loss": 1.1059, + "step": 1882 + }, + { + "epoch": 0.11440549243574943, + "grad_norm": 0.34476161003112793, + "learning_rate": 9.695582554719283e-05, + "loss": 1.0728, + "step": 1883 + }, + { + "epoch": 0.11446624946837597, + "grad_norm": 1.0319693088531494, + "learning_rate": 9.695253549269296e-05, + "loss": 1.2117, + "step": 1884 + }, + { + "epoch": 0.11452700650100249, + "grad_norm": 0.3396998345851898, + "learning_rate": 9.694924371713883e-05, + "loss": 1.1678, + "step": 1885 + }, + { + "epoch": 0.11458776353362901, + "grad_norm": 0.27898073196411133, + "learning_rate": 9.694595022065109e-05, + "loss": 1.2677, + "step": 1886 + }, + { + "epoch": 0.11464852056625555, + "grad_norm": 0.3002793490886688, + "learning_rate": 9.694265500335046e-05, + "loss": 1.2615, + "step": 1887 + }, + { + "epoch": 0.11470927759888207, + "grad_norm": 0.334608256816864, + "learning_rate": 9.693935806535776e-05, + "loss": 1.1611, + "step": 1888 + }, + { + "epoch": 0.1147700346315086, + "grad_norm": 0.24664580821990967, + "learning_rate": 9.693605940679381e-05, + "loss": 1.1363, + "step": 1889 + }, + { + "epoch": 0.11483079166413512, + "grad_norm": 0.4067285656929016, + "learning_rate": 9.693275902777953e-05, + "loss": 1.1353, + "step": 1890 + }, + { + "epoch": 0.11489154869676164, + "grad_norm": 0.6787147521972656, + "learning_rate": 9.692945692843589e-05, + "loss": 1.2572, + "step": 1891 + }, + { + "epoch": 0.11495230572938818, + "grad_norm": 0.3914863169193268, + "learning_rate": 9.692615310888394e-05, + "loss": 1.1494, + "step": 1892 + }, + { + "epoch": 0.1150130627620147, + "grad_norm": 0.42598459124565125, + "learning_rate": 9.692284756924477e-05, + "loss": 1.1594, + "step": 1893 + }, + { + "epoch": 0.11507381979464124, + "grad_norm": 0.39596012234687805, + "learning_rate": 9.691954030963957e-05, + "loss": 1.2423, + "step": 1894 + }, + { + "epoch": 0.11513457682726776, + "grad_norm": 0.7186443209648132, + "learning_rate": 9.691623133018955e-05, + "loss": 1.3405, + "step": 1895 + }, + { + "epoch": 0.11519533385989428, + "grad_norm": 0.338821679353714, + "learning_rate": 9.691292063101598e-05, + "loss": 1.1531, + "step": 1896 + }, + { + "epoch": 0.11525609089252081, + "grad_norm": 5.212579250335693, + "learning_rate": 9.690960821224025e-05, + "loss": 1.1129, + "step": 1897 + }, + { + "epoch": 0.11531684792514733, + "grad_norm": 0.644722580909729, + "learning_rate": 9.690629407398376e-05, + "loss": 1.2905, + "step": 1898 + }, + { + "epoch": 0.11537760495777387, + "grad_norm": 0.36073824763298035, + "learning_rate": 9.6902978216368e-05, + "loss": 1.284, + "step": 1899 + }, + { + "epoch": 0.11543836199040039, + "grad_norm": 0.28805962204933167, + "learning_rate": 9.68996606395145e-05, + "loss": 1.1534, + "step": 1900 + }, + { + "epoch": 0.11549911902302691, + "grad_norm": 0.5544807314872742, + "learning_rate": 9.689634134354489e-05, + "loss": 1.0874, + "step": 1901 + }, + { + "epoch": 0.11555987605565345, + "grad_norm": 0.4203668236732483, + "learning_rate": 9.68930203285808e-05, + "loss": 1.1302, + "step": 1902 + }, + { + "epoch": 0.11562063308827997, + "grad_norm": 0.466437429189682, + "learning_rate": 9.688969759474399e-05, + "loss": 1.2035, + "step": 1903 + }, + { + "epoch": 0.11568139012090649, + "grad_norm": 0.22199200093746185, + "learning_rate": 9.688637314215625e-05, + "loss": 1.1021, + "step": 1904 + }, + { + "epoch": 0.11574214715353302, + "grad_norm": 0.3709351718425751, + "learning_rate": 9.688304697093947e-05, + "loss": 1.1737, + "step": 1905 + }, + { + "epoch": 0.11580290418615954, + "grad_norm": 2.3607397079467773, + "learning_rate": 9.68797190812155e-05, + "loss": 1.1182, + "step": 1906 + }, + { + "epoch": 0.11586366121878608, + "grad_norm": 0.3466080129146576, + "learning_rate": 9.687638947310638e-05, + "loss": 1.181, + "step": 1907 + }, + { + "epoch": 0.1159244182514126, + "grad_norm": 0.3125613331794739, + "learning_rate": 9.687305814673415e-05, + "loss": 1.1265, + "step": 1908 + }, + { + "epoch": 0.11598517528403912, + "grad_norm": 1.341249704360962, + "learning_rate": 9.686972510222091e-05, + "loss": 1.2113, + "step": 1909 + }, + { + "epoch": 0.11604593231666566, + "grad_norm": 0.5886200070381165, + "learning_rate": 9.686639033968883e-05, + "loss": 1.1083, + "step": 1910 + }, + { + "epoch": 0.11610668934929218, + "grad_norm": 0.25061628222465515, + "learning_rate": 9.686305385926016e-05, + "loss": 1.1712, + "step": 1911 + }, + { + "epoch": 0.11616744638191871, + "grad_norm": 0.44334354996681213, + "learning_rate": 9.685971566105719e-05, + "loss": 1.0951, + "step": 1912 + }, + { + "epoch": 0.11622820341454523, + "grad_norm": 0.3861365020275116, + "learning_rate": 9.685637574520228e-05, + "loss": 1.1107, + "step": 1913 + }, + { + "epoch": 0.11628896044717175, + "grad_norm": 0.22862835228443146, + "learning_rate": 9.685303411181788e-05, + "loss": 1.1865, + "step": 1914 + }, + { + "epoch": 0.11634971747979829, + "grad_norm": 0.3258768916130066, + "learning_rate": 9.684969076102643e-05, + "loss": 1.1127, + "step": 1915 + }, + { + "epoch": 0.11641047451242481, + "grad_norm": 0.3555976152420044, + "learning_rate": 9.684634569295053e-05, + "loss": 1.1958, + "step": 1916 + }, + { + "epoch": 0.11647123154505135, + "grad_norm": 0.37971553206443787, + "learning_rate": 9.684299890771275e-05, + "loss": 1.1023, + "step": 1917 + }, + { + "epoch": 0.11653198857767787, + "grad_norm": 0.4318792521953583, + "learning_rate": 9.683965040543582e-05, + "loss": 1.2279, + "step": 1918 + }, + { + "epoch": 0.11659274561030439, + "grad_norm": 0.45337846875190735, + "learning_rate": 9.683630018624243e-05, + "loss": 1.0929, + "step": 1919 + }, + { + "epoch": 0.11665350264293092, + "grad_norm": 0.2752816677093506, + "learning_rate": 9.683294825025541e-05, + "loss": 1.234, + "step": 1920 + }, + { + "epoch": 0.11671425967555744, + "grad_norm": 1.4375146627426147, + "learning_rate": 9.682959459759763e-05, + "loss": 1.1867, + "step": 1921 + }, + { + "epoch": 0.11677501670818398, + "grad_norm": 0.2892298102378845, + "learning_rate": 9.6826239228392e-05, + "loss": 1.1352, + "step": 1922 + }, + { + "epoch": 0.1168357737408105, + "grad_norm": 0.21238848567008972, + "learning_rate": 9.682288214276152e-05, + "loss": 1.148, + "step": 1923 + }, + { + "epoch": 0.11689653077343702, + "grad_norm": 0.27681073546409607, + "learning_rate": 9.681952334082923e-05, + "loss": 1.0977, + "step": 1924 + }, + { + "epoch": 0.11695728780606356, + "grad_norm": 0.20225673913955688, + "learning_rate": 9.681616282271828e-05, + "loss": 1.0788, + "step": 1925 + }, + { + "epoch": 0.11701804483869008, + "grad_norm": 0.2752569019794464, + "learning_rate": 9.681280058855183e-05, + "loss": 1.1007, + "step": 1926 + }, + { + "epoch": 0.1170788018713166, + "grad_norm": 0.3313400447368622, + "learning_rate": 9.680943663845312e-05, + "loss": 1.124, + "step": 1927 + }, + { + "epoch": 0.11713955890394313, + "grad_norm": 0.3881179988384247, + "learning_rate": 9.680607097254546e-05, + "loss": 1.1532, + "step": 1928 + }, + { + "epoch": 0.11720031593656965, + "grad_norm": 0.20384211838245392, + "learning_rate": 9.680270359095222e-05, + "loss": 1.0988, + "step": 1929 + }, + { + "epoch": 0.11726107296919619, + "grad_norm": 0.4139309227466583, + "learning_rate": 9.679933449379684e-05, + "loss": 1.3122, + "step": 1930 + }, + { + "epoch": 0.11732183000182271, + "grad_norm": 0.36846089363098145, + "learning_rate": 9.67959636812028e-05, + "loss": 1.1505, + "step": 1931 + }, + { + "epoch": 0.11738258703444923, + "grad_norm": 0.3128921687602997, + "learning_rate": 9.679259115329369e-05, + "loss": 1.0985, + "step": 1932 + }, + { + "epoch": 0.11744334406707577, + "grad_norm": 0.4877334237098694, + "learning_rate": 9.678921691019308e-05, + "loss": 1.2669, + "step": 1933 + }, + { + "epoch": 0.11750410109970229, + "grad_norm": 0.24451249837875366, + "learning_rate": 9.678584095202468e-05, + "loss": 1.174, + "step": 1934 + }, + { + "epoch": 0.11756485813232882, + "grad_norm": 0.43930745124816895, + "learning_rate": 9.678246327891224e-05, + "loss": 1.1241, + "step": 1935 + }, + { + "epoch": 0.11762561516495534, + "grad_norm": 0.272484689950943, + "learning_rate": 9.677908389097959e-05, + "loss": 1.0982, + "step": 1936 + }, + { + "epoch": 0.11768637219758186, + "grad_norm": 0.27668496966362, + "learning_rate": 9.677570278835055e-05, + "loss": 1.1208, + "step": 1937 + }, + { + "epoch": 0.1177471292302084, + "grad_norm": 0.38551968336105347, + "learning_rate": 9.677231997114909e-05, + "loss": 1.135, + "step": 1938 + }, + { + "epoch": 0.11780788626283492, + "grad_norm": 0.23142309486865997, + "learning_rate": 9.676893543949921e-05, + "loss": 1.1498, + "step": 1939 + }, + { + "epoch": 0.11786864329546146, + "grad_norm": 0.2698286771774292, + "learning_rate": 9.676554919352494e-05, + "loss": 1.1403, + "step": 1940 + }, + { + "epoch": 0.11792940032808798, + "grad_norm": 0.5868292450904846, + "learning_rate": 9.676216123335045e-05, + "loss": 1.3097, + "step": 1941 + }, + { + "epoch": 0.1179901573607145, + "grad_norm": 0.22429725527763367, + "learning_rate": 9.675877155909988e-05, + "loss": 1.1228, + "step": 1942 + }, + { + "epoch": 0.11805091439334103, + "grad_norm": 0.41946396231651306, + "learning_rate": 9.675538017089752e-05, + "loss": 1.1063, + "step": 1943 + }, + { + "epoch": 0.11811167142596755, + "grad_norm": 0.4610346555709839, + "learning_rate": 9.675198706886765e-05, + "loss": 1.2005, + "step": 1944 + }, + { + "epoch": 0.11817242845859408, + "grad_norm": 0.5070992112159729, + "learning_rate": 9.674859225313465e-05, + "loss": 1.1759, + "step": 1945 + }, + { + "epoch": 0.11823318549122061, + "grad_norm": 3.2501020431518555, + "learning_rate": 9.674519572382297e-05, + "loss": 1.209, + "step": 1946 + }, + { + "epoch": 0.11829394252384713, + "grad_norm": 0.3523193299770355, + "learning_rate": 9.674179748105712e-05, + "loss": 1.3248, + "step": 1947 + }, + { + "epoch": 0.11835469955647367, + "grad_norm": 0.3355488181114197, + "learning_rate": 9.673839752496163e-05, + "loss": 1.3259, + "step": 1948 + }, + { + "epoch": 0.11841545658910019, + "grad_norm": 0.43970683217048645, + "learning_rate": 9.673499585566114e-05, + "loss": 1.078, + "step": 1949 + }, + { + "epoch": 0.11847621362172671, + "grad_norm": 0.23309676349163055, + "learning_rate": 9.673159247328036e-05, + "loss": 1.1296, + "step": 1950 + }, + { + "epoch": 0.11853697065435324, + "grad_norm": 0.3487436771392822, + "learning_rate": 9.672818737794403e-05, + "loss": 1.1229, + "step": 1951 + }, + { + "epoch": 0.11859772768697976, + "grad_norm": 0.23054072260856628, + "learning_rate": 9.672478056977694e-05, + "loss": 1.1397, + "step": 1952 + }, + { + "epoch": 0.1186584847196063, + "grad_norm": 0.294677734375, + "learning_rate": 9.672137204890398e-05, + "loss": 1.1861, + "step": 1953 + }, + { + "epoch": 0.11871924175223282, + "grad_norm": 0.2330155223608017, + "learning_rate": 9.671796181545012e-05, + "loss": 1.1254, + "step": 1954 + }, + { + "epoch": 0.11877999878485934, + "grad_norm": 0.19884991645812988, + "learning_rate": 9.671454986954035e-05, + "loss": 1.1103, + "step": 1955 + }, + { + "epoch": 0.11884075581748588, + "grad_norm": 0.47316795587539673, + "learning_rate": 9.671113621129969e-05, + "loss": 1.1312, + "step": 1956 + }, + { + "epoch": 0.1189015128501124, + "grad_norm": 0.27472180128097534, + "learning_rate": 9.670772084085331e-05, + "loss": 1.2778, + "step": 1957 + }, + { + "epoch": 0.11896226988273893, + "grad_norm": 0.2202560007572174, + "learning_rate": 9.670430375832641e-05, + "loss": 1.1405, + "step": 1958 + }, + { + "epoch": 0.11902302691536545, + "grad_norm": 0.26746514439582825, + "learning_rate": 9.670088496384422e-05, + "loss": 1.0537, + "step": 1959 + }, + { + "epoch": 0.11908378394799198, + "grad_norm": 0.20914916694164276, + "learning_rate": 9.669746445753207e-05, + "loss": 1.1292, + "step": 1960 + }, + { + "epoch": 0.11914454098061851, + "grad_norm": 0.24406176805496216, + "learning_rate": 9.669404223951533e-05, + "loss": 1.1185, + "step": 1961 + }, + { + "epoch": 0.11920529801324503, + "grad_norm": 0.6328793168067932, + "learning_rate": 9.669061830991943e-05, + "loss": 1.138, + "step": 1962 + }, + { + "epoch": 0.11926605504587157, + "grad_norm": 0.24986669421195984, + "learning_rate": 9.668719266886992e-05, + "loss": 1.1289, + "step": 1963 + }, + { + "epoch": 0.11932681207849809, + "grad_norm": 0.413897305727005, + "learning_rate": 9.668376531649233e-05, + "loss": 1.1402, + "step": 1964 + }, + { + "epoch": 0.11938756911112461, + "grad_norm": 0.2828940451145172, + "learning_rate": 9.668033625291229e-05, + "loss": 1.2596, + "step": 1965 + }, + { + "epoch": 0.11944832614375114, + "grad_norm": 0.322428435087204, + "learning_rate": 9.667690547825549e-05, + "loss": 1.1281, + "step": 1966 + }, + { + "epoch": 0.11950908317637766, + "grad_norm": 0.31709563732147217, + "learning_rate": 9.667347299264772e-05, + "loss": 1.1115, + "step": 1967 + }, + { + "epoch": 0.11956984020900419, + "grad_norm": 3.076204776763916, + "learning_rate": 9.667003879621477e-05, + "loss": 1.1656, + "step": 1968 + }, + { + "epoch": 0.11963059724163072, + "grad_norm": 0.38741374015808105, + "learning_rate": 9.666660288908251e-05, + "loss": 1.0864, + "step": 1969 + }, + { + "epoch": 0.11969135427425724, + "grad_norm": 0.24754831194877625, + "learning_rate": 9.666316527137693e-05, + "loss": 1.0593, + "step": 1970 + }, + { + "epoch": 0.11975211130688378, + "grad_norm": 0.3259872794151306, + "learning_rate": 9.665972594322399e-05, + "loss": 1.0942, + "step": 1971 + }, + { + "epoch": 0.1198128683395103, + "grad_norm": 2.7607169151306152, + "learning_rate": 9.665628490474977e-05, + "loss": 1.225, + "step": 1972 + }, + { + "epoch": 0.11987362537213682, + "grad_norm": 0.8714908957481384, + "learning_rate": 9.665284215608041e-05, + "loss": 1.1102, + "step": 1973 + }, + { + "epoch": 0.11993438240476335, + "grad_norm": 0.282874196767807, + "learning_rate": 9.664939769734212e-05, + "loss": 1.2858, + "step": 1974 + }, + { + "epoch": 0.11999513943738987, + "grad_norm": 0.21523600816726685, + "learning_rate": 9.664595152866113e-05, + "loss": 1.1025, + "step": 1975 + }, + { + "epoch": 0.12005589647001641, + "grad_norm": 0.2753016948699951, + "learning_rate": 9.664250365016376e-05, + "loss": 1.2193, + "step": 1976 + }, + { + "epoch": 0.12011665350264293, + "grad_norm": 0.35401204228401184, + "learning_rate": 9.663905406197642e-05, + "loss": 1.2395, + "step": 1977 + }, + { + "epoch": 0.12017741053526945, + "grad_norm": 0.22871582210063934, + "learning_rate": 9.663560276422553e-05, + "loss": 1.0599, + "step": 1978 + }, + { + "epoch": 0.12023816756789599, + "grad_norm": 2.556197166442871, + "learning_rate": 9.66321497570376e-05, + "loss": 1.1471, + "step": 1979 + }, + { + "epoch": 0.12029892460052251, + "grad_norm": 0.31443190574645996, + "learning_rate": 9.662869504053923e-05, + "loss": 1.2239, + "step": 1980 + }, + { + "epoch": 0.12035968163314904, + "grad_norm": 0.7996320128440857, + "learning_rate": 9.6625238614857e-05, + "loss": 1.2748, + "step": 1981 + }, + { + "epoch": 0.12042043866577556, + "grad_norm": 2.061707019805908, + "learning_rate": 9.662178048011765e-05, + "loss": 1.1297, + "step": 1982 + }, + { + "epoch": 0.12048119569840209, + "grad_norm": 0.3946833312511444, + "learning_rate": 9.661832063644794e-05, + "loss": 1.1215, + "step": 1983 + }, + { + "epoch": 0.12054195273102862, + "grad_norm": 0.2606329917907715, + "learning_rate": 9.661485908397465e-05, + "loss": 1.1116, + "step": 1984 + }, + { + "epoch": 0.12060270976365514, + "grad_norm": 0.2944069504737854, + "learning_rate": 9.66113958228247e-05, + "loss": 1.0935, + "step": 1985 + }, + { + "epoch": 0.12066346679628168, + "grad_norm": 0.3792330324649811, + "learning_rate": 9.660793085312503e-05, + "loss": 1.3213, + "step": 1986 + }, + { + "epoch": 0.1207242238289082, + "grad_norm": 0.5132965445518494, + "learning_rate": 9.660446417500265e-05, + "loss": 1.1851, + "step": 1987 + }, + { + "epoch": 0.12078498086153472, + "grad_norm": 0.30100464820861816, + "learning_rate": 9.660099578858461e-05, + "loss": 1.1185, + "step": 1988 + }, + { + "epoch": 0.12084573789416125, + "grad_norm": 0.5011243224143982, + "learning_rate": 9.659752569399807e-05, + "loss": 1.0766, + "step": 1989 + }, + { + "epoch": 0.12090649492678777, + "grad_norm": 0.469701886177063, + "learning_rate": 9.659405389137021e-05, + "loss": 1.1105, + "step": 1990 + }, + { + "epoch": 0.1209672519594143, + "grad_norm": 0.6148624420166016, + "learning_rate": 9.659058038082833e-05, + "loss": 1.128, + "step": 1991 + }, + { + "epoch": 0.12102800899204083, + "grad_norm": 0.47404032945632935, + "learning_rate": 9.65871051624997e-05, + "loss": 1.2835, + "step": 1992 + }, + { + "epoch": 0.12108876602466735, + "grad_norm": 0.5288172364234924, + "learning_rate": 9.658362823651172e-05, + "loss": 1.195, + "step": 1993 + }, + { + "epoch": 0.12114952305729389, + "grad_norm": 1.320677638053894, + "learning_rate": 9.658014960299185e-05, + "loss": 1.1647, + "step": 1994 + }, + { + "epoch": 0.12121028008992041, + "grad_norm": 0.6657614707946777, + "learning_rate": 9.657666926206758e-05, + "loss": 1.1889, + "step": 1995 + }, + { + "epoch": 0.12127103712254693, + "grad_norm": 0.597414493560791, + "learning_rate": 9.65731872138665e-05, + "loss": 1.1742, + "step": 1996 + }, + { + "epoch": 0.12133179415517346, + "grad_norm": 0.5814148187637329, + "learning_rate": 9.656970345851625e-05, + "loss": 1.3342, + "step": 1997 + }, + { + "epoch": 0.12139255118779999, + "grad_norm": 1.0570162534713745, + "learning_rate": 9.656621799614452e-05, + "loss": 1.1174, + "step": 1998 + }, + { + "epoch": 0.12145330822042652, + "grad_norm": 0.5145533680915833, + "learning_rate": 9.656273082687905e-05, + "loss": 1.1138, + "step": 1999 + }, + { + "epoch": 0.12151406525305304, + "grad_norm": 0.8125649094581604, + "learning_rate": 9.655924195084768e-05, + "loss": 1.1168, + "step": 2000 + }, + { + "epoch": 0.12157482228567956, + "grad_norm": 0.31321588158607483, + "learning_rate": 9.655575136817831e-05, + "loss": 1.3553, + "step": 2001 + }, + { + "epoch": 0.1216355793183061, + "grad_norm": 1.0420430898666382, + "learning_rate": 9.655225907899888e-05, + "loss": 1.0833, + "step": 2002 + }, + { + "epoch": 0.12169633635093262, + "grad_norm": 0.4602111876010895, + "learning_rate": 9.654876508343738e-05, + "loss": 1.0537, + "step": 2003 + }, + { + "epoch": 0.12175709338355915, + "grad_norm": 0.6072948575019836, + "learning_rate": 9.654526938162191e-05, + "loss": 1.1621, + "step": 2004 + }, + { + "epoch": 0.12181785041618567, + "grad_norm": 0.2676675319671631, + "learning_rate": 9.65417719736806e-05, + "loss": 1.1522, + "step": 2005 + }, + { + "epoch": 0.1218786074488122, + "grad_norm": 0.5492773056030273, + "learning_rate": 9.653827285974163e-05, + "loss": 1.1491, + "step": 2006 + }, + { + "epoch": 0.12193936448143873, + "grad_norm": 0.4248824417591095, + "learning_rate": 9.653477203993327e-05, + "loss": 1.1971, + "step": 2007 + }, + { + "epoch": 0.12200012151406525, + "grad_norm": 0.4388446807861328, + "learning_rate": 9.653126951438385e-05, + "loss": 1.163, + "step": 2008 + }, + { + "epoch": 0.12206087854669177, + "grad_norm": 0.5561811923980713, + "learning_rate": 9.652776528322175e-05, + "loss": 1.1322, + "step": 2009 + }, + { + "epoch": 0.12212163557931831, + "grad_norm": 3.7818498611450195, + "learning_rate": 9.652425934657542e-05, + "loss": 1.2122, + "step": 2010 + }, + { + "epoch": 0.12218239261194483, + "grad_norm": 0.4403693377971649, + "learning_rate": 9.652075170457337e-05, + "loss": 1.1231, + "step": 2011 + }, + { + "epoch": 0.12224314964457136, + "grad_norm": 0.2745683789253235, + "learning_rate": 9.651724235734418e-05, + "loss": 1.16, + "step": 2012 + }, + { + "epoch": 0.12230390667719789, + "grad_norm": 0.24837814271450043, + "learning_rate": 9.651373130501649e-05, + "loss": 1.2036, + "step": 2013 + }, + { + "epoch": 0.1223646637098244, + "grad_norm": 0.28888681530952454, + "learning_rate": 9.651021854771897e-05, + "loss": 1.1355, + "step": 2014 + }, + { + "epoch": 0.12242542074245094, + "grad_norm": 0.5621148347854614, + "learning_rate": 9.650670408558041e-05, + "loss": 1.2016, + "step": 2015 + }, + { + "epoch": 0.12248617777507746, + "grad_norm": 0.7050667405128479, + "learning_rate": 9.65031879187296e-05, + "loss": 1.2677, + "step": 2016 + }, + { + "epoch": 0.122546934807704, + "grad_norm": 0.2840939462184906, + "learning_rate": 9.649967004729548e-05, + "loss": 1.1152, + "step": 2017 + }, + { + "epoch": 0.12260769184033052, + "grad_norm": 0.7255191206932068, + "learning_rate": 9.649615047140695e-05, + "loss": 1.2016, + "step": 2018 + }, + { + "epoch": 0.12266844887295704, + "grad_norm": 0.30125904083251953, + "learning_rate": 9.649262919119303e-05, + "loss": 1.1578, + "step": 2019 + }, + { + "epoch": 0.12272920590558357, + "grad_norm": 0.2972518503665924, + "learning_rate": 9.64891062067828e-05, + "loss": 1.1293, + "step": 2020 + }, + { + "epoch": 0.1227899629382101, + "grad_norm": 0.3766010105609894, + "learning_rate": 9.64855815183054e-05, + "loss": 1.1728, + "step": 2021 + }, + { + "epoch": 0.12285071997083663, + "grad_norm": 0.3118601143360138, + "learning_rate": 9.648205512589003e-05, + "loss": 1.0827, + "step": 2022 + }, + { + "epoch": 0.12291147700346315, + "grad_norm": 0.29879453778266907, + "learning_rate": 9.647852702966596e-05, + "loss": 1.1083, + "step": 2023 + }, + { + "epoch": 0.12297223403608967, + "grad_norm": 0.354612797498703, + "learning_rate": 9.647499722976247e-05, + "loss": 1.1696, + "step": 2024 + }, + { + "epoch": 0.12303299106871621, + "grad_norm": 0.4042516350746155, + "learning_rate": 9.647146572630899e-05, + "loss": 1.1062, + "step": 2025 + }, + { + "epoch": 0.12309374810134273, + "grad_norm": 0.7183785438537598, + "learning_rate": 9.646793251943493e-05, + "loss": 1.1505, + "step": 2026 + }, + { + "epoch": 0.12315450513396926, + "grad_norm": 0.28505176305770874, + "learning_rate": 9.646439760926984e-05, + "loss": 1.0847, + "step": 2027 + }, + { + "epoch": 0.12321526216659578, + "grad_norm": 0.3930915892124176, + "learning_rate": 9.646086099594326e-05, + "loss": 1.1798, + "step": 2028 + }, + { + "epoch": 0.1232760191992223, + "grad_norm": 0.5896264910697937, + "learning_rate": 9.645732267958485e-05, + "loss": 1.3354, + "step": 2029 + }, + { + "epoch": 0.12333677623184884, + "grad_norm": 0.3955848217010498, + "learning_rate": 9.645378266032429e-05, + "loss": 1.2877, + "step": 2030 + }, + { + "epoch": 0.12339753326447536, + "grad_norm": 3.0918405055999756, + "learning_rate": 9.645024093829136e-05, + "loss": 1.1003, + "step": 2031 + }, + { + "epoch": 0.12345829029710188, + "grad_norm": 0.28922566771507263, + "learning_rate": 9.644669751361587e-05, + "loss": 1.152, + "step": 2032 + }, + { + "epoch": 0.12351904732972842, + "grad_norm": 0.24025261402130127, + "learning_rate": 9.644315238642769e-05, + "loss": 1.0796, + "step": 2033 + }, + { + "epoch": 0.12357980436235494, + "grad_norm": 0.2681881785392761, + "learning_rate": 9.643960555685679e-05, + "loss": 1.0991, + "step": 2034 + }, + { + "epoch": 0.12364056139498147, + "grad_norm": 0.3252476751804352, + "learning_rate": 9.643605702503318e-05, + "loss": 1.1441, + "step": 2035 + }, + { + "epoch": 0.123701318427608, + "grad_norm": 0.49251851439476013, + "learning_rate": 9.643250679108691e-05, + "loss": 1.1207, + "step": 2036 + }, + { + "epoch": 0.12376207546023452, + "grad_norm": 0.26207244396209717, + "learning_rate": 9.642895485514813e-05, + "loss": 1.15, + "step": 2037 + }, + { + "epoch": 0.12382283249286105, + "grad_norm": 0.4089462459087372, + "learning_rate": 9.642540121734705e-05, + "loss": 1.1226, + "step": 2038 + }, + { + "epoch": 0.12388358952548757, + "grad_norm": 0.6951386332511902, + "learning_rate": 9.64218458778139e-05, + "loss": 1.1828, + "step": 2039 + }, + { + "epoch": 0.12394434655811411, + "grad_norm": 0.5540273189544678, + "learning_rate": 9.641828883667902e-05, + "loss": 1.1914, + "step": 2040 + }, + { + "epoch": 0.12400510359074063, + "grad_norm": 0.668719470500946, + "learning_rate": 9.641473009407278e-05, + "loss": 1.1336, + "step": 2041 + }, + { + "epoch": 0.12406586062336715, + "grad_norm": 1.3161060810089111, + "learning_rate": 9.641116965012565e-05, + "loss": 1.1363, + "step": 2042 + }, + { + "epoch": 0.12412661765599368, + "grad_norm": 0.8786684274673462, + "learning_rate": 9.640760750496812e-05, + "loss": 1.3284, + "step": 2043 + }, + { + "epoch": 0.1241873746886202, + "grad_norm": 0.283647358417511, + "learning_rate": 9.640404365873077e-05, + "loss": 1.0955, + "step": 2044 + }, + { + "epoch": 0.12424813172124674, + "grad_norm": 0.7797661423683167, + "learning_rate": 9.640047811154424e-05, + "loss": 1.2619, + "step": 2045 + }, + { + "epoch": 0.12430888875387326, + "grad_norm": 0.42274177074432373, + "learning_rate": 9.639691086353919e-05, + "loss": 1.1142, + "step": 2046 + }, + { + "epoch": 0.12436964578649978, + "grad_norm": 0.3193400204181671, + "learning_rate": 9.639334191484642e-05, + "loss": 1.1461, + "step": 2047 + }, + { + "epoch": 0.12443040281912632, + "grad_norm": 0.5529247522354126, + "learning_rate": 9.638977126559675e-05, + "loss": 1.1334, + "step": 2048 + }, + { + "epoch": 0.12449115985175284, + "grad_norm": 0.3457753658294678, + "learning_rate": 9.638619891592102e-05, + "loss": 1.1648, + "step": 2049 + }, + { + "epoch": 0.12455191688437936, + "grad_norm": 0.29722633957862854, + "learning_rate": 9.638262486595022e-05, + "loss": 1.095, + "step": 2050 + }, + { + "epoch": 0.1246126739170059, + "grad_norm": 0.5841728448867798, + "learning_rate": 9.637904911581532e-05, + "loss": 1.1808, + "step": 2051 + }, + { + "epoch": 0.12467343094963242, + "grad_norm": 0.46711453795433044, + "learning_rate": 9.637547166564741e-05, + "loss": 1.1985, + "step": 2052 + }, + { + "epoch": 0.12473418798225895, + "grad_norm": 0.5162690281867981, + "learning_rate": 9.637189251557766e-05, + "loss": 1.1144, + "step": 2053 + }, + { + "epoch": 0.12479494501488547, + "grad_norm": 0.37609705328941345, + "learning_rate": 9.636831166573719e-05, + "loss": 1.0825, + "step": 2054 + }, + { + "epoch": 0.124855702047512, + "grad_norm": 1.7572757005691528, + "learning_rate": 9.63647291162573e-05, + "loss": 1.1012, + "step": 2055 + }, + { + "epoch": 0.12491645908013853, + "grad_norm": 0.8972417116165161, + "learning_rate": 9.63611448672693e-05, + "loss": 1.1159, + "step": 2056 + }, + { + "epoch": 0.12497721611276505, + "grad_norm": 0.5035214424133301, + "learning_rate": 9.635755891890459e-05, + "loss": 1.1438, + "step": 2057 + }, + { + "epoch": 0.12503797314539158, + "grad_norm": 0.21978595852851868, + "learning_rate": 9.635397127129458e-05, + "loss": 1.1044, + "step": 2058 + }, + { + "epoch": 0.12509873017801812, + "grad_norm": 0.5355223417282104, + "learning_rate": 9.63503819245708e-05, + "loss": 1.2621, + "step": 2059 + }, + { + "epoch": 0.12515948721064463, + "grad_norm": 0.438656210899353, + "learning_rate": 9.634679087886481e-05, + "loss": 1.222, + "step": 2060 + }, + { + "epoch": 0.12522024424327116, + "grad_norm": 0.2514871060848236, + "learning_rate": 9.634319813430826e-05, + "loss": 1.1106, + "step": 2061 + }, + { + "epoch": 0.1252810012758977, + "grad_norm": 0.22231470048427582, + "learning_rate": 9.633960369103279e-05, + "loss": 1.1185, + "step": 2062 + }, + { + "epoch": 0.1253417583085242, + "grad_norm": 0.43077242374420166, + "learning_rate": 9.63360075491702e-05, + "loss": 1.2281, + "step": 2063 + }, + { + "epoch": 0.12540251534115074, + "grad_norm": 0.3033542335033417, + "learning_rate": 9.63324097088523e-05, + "loss": 1.243, + "step": 2064 + }, + { + "epoch": 0.12546327237377727, + "grad_norm": 0.4114849269390106, + "learning_rate": 9.632881017021098e-05, + "loss": 1.1471, + "step": 2065 + }, + { + "epoch": 0.12552402940640378, + "grad_norm": 0.5144845247268677, + "learning_rate": 9.632520893337815e-05, + "loss": 1.1538, + "step": 2066 + }, + { + "epoch": 0.12558478643903032, + "grad_norm": 0.41076716780662537, + "learning_rate": 9.632160599848585e-05, + "loss": 1.116, + "step": 2067 + }, + { + "epoch": 0.12564554347165685, + "grad_norm": 0.23057472705841064, + "learning_rate": 9.63180013656661e-05, + "loss": 1.2042, + "step": 2068 + }, + { + "epoch": 0.12570630050428336, + "grad_norm": 0.3216788172721863, + "learning_rate": 9.631439503505108e-05, + "loss": 1.2381, + "step": 2069 + }, + { + "epoch": 0.1257670575369099, + "grad_norm": 0.3786807954311371, + "learning_rate": 9.631078700677296e-05, + "loss": 1.1965, + "step": 2070 + }, + { + "epoch": 0.12582781456953643, + "grad_norm": 0.2500993311405182, + "learning_rate": 9.6307177280964e-05, + "loss": 1.0986, + "step": 2071 + }, + { + "epoch": 0.12588857160216296, + "grad_norm": 0.3342118561267853, + "learning_rate": 9.630356585775647e-05, + "loss": 1.2994, + "step": 2072 + }, + { + "epoch": 0.12594932863478947, + "grad_norm": 0.28680527210235596, + "learning_rate": 9.62999527372828e-05, + "loss": 1.1038, + "step": 2073 + }, + { + "epoch": 0.126010085667416, + "grad_norm": 0.47306519746780396, + "learning_rate": 9.629633791967541e-05, + "loss": 1.074, + "step": 2074 + }, + { + "epoch": 0.12607084270004254, + "grad_norm": 2.291471242904663, + "learning_rate": 9.629272140506682e-05, + "loss": 1.0972, + "step": 2075 + }, + { + "epoch": 0.12613159973266905, + "grad_norm": 0.44187307357788086, + "learning_rate": 9.628910319358956e-05, + "loss": 1.2263, + "step": 2076 + }, + { + "epoch": 0.12619235676529558, + "grad_norm": 0.2546493113040924, + "learning_rate": 9.62854832853763e-05, + "loss": 1.1026, + "step": 2077 + }, + { + "epoch": 0.12625311379792212, + "grad_norm": 0.34413769841194153, + "learning_rate": 9.628186168055968e-05, + "loss": 1.1626, + "step": 2078 + }, + { + "epoch": 0.12631387083054862, + "grad_norm": 0.30373620986938477, + "learning_rate": 9.627823837927247e-05, + "loss": 1.1324, + "step": 2079 + }, + { + "epoch": 0.12637462786317516, + "grad_norm": 0.8798522353172302, + "learning_rate": 9.627461338164751e-05, + "loss": 1.0694, + "step": 2080 + }, + { + "epoch": 0.1264353848958017, + "grad_norm": 0.4222569465637207, + "learning_rate": 9.627098668781763e-05, + "loss": 1.2088, + "step": 2081 + }, + { + "epoch": 0.1264961419284282, + "grad_norm": 0.19593752920627594, + "learning_rate": 9.62673582979158e-05, + "loss": 1.1302, + "step": 2082 + }, + { + "epoch": 0.12655689896105474, + "grad_norm": 0.42246896028518677, + "learning_rate": 9.626372821207501e-05, + "loss": 1.1254, + "step": 2083 + }, + { + "epoch": 0.12661765599368127, + "grad_norm": 0.38738641142845154, + "learning_rate": 9.62600964304283e-05, + "loss": 1.2869, + "step": 2084 + }, + { + "epoch": 0.1266784130263078, + "grad_norm": 0.40146419405937195, + "learning_rate": 9.625646295310884e-05, + "loss": 1.0547, + "step": 2085 + }, + { + "epoch": 0.12673917005893431, + "grad_norm": 0.30133286118507385, + "learning_rate": 9.625282778024978e-05, + "loss": 1.2655, + "step": 2086 + }, + { + "epoch": 0.12679992709156085, + "grad_norm": 0.3317384719848633, + "learning_rate": 9.624919091198437e-05, + "loss": 1.1755, + "step": 2087 + }, + { + "epoch": 0.12686068412418738, + "grad_norm": 0.31135112047195435, + "learning_rate": 9.624555234844595e-05, + "loss": 1.0938, + "step": 2088 + }, + { + "epoch": 0.1269214411568139, + "grad_norm": 0.4897431433200836, + "learning_rate": 9.624191208976785e-05, + "loss": 1.1528, + "step": 2089 + }, + { + "epoch": 0.12698219818944043, + "grad_norm": 0.262713760137558, + "learning_rate": 9.623827013608351e-05, + "loss": 1.0736, + "step": 2090 + }, + { + "epoch": 0.12704295522206696, + "grad_norm": 0.39854153990745544, + "learning_rate": 9.623462648752648e-05, + "loss": 1.1864, + "step": 2091 + }, + { + "epoch": 0.12710371225469347, + "grad_norm": 0.22994931042194366, + "learning_rate": 9.623098114423024e-05, + "loss": 1.1835, + "step": 2092 + }, + { + "epoch": 0.12716446928732, + "grad_norm": 2.9490737915039062, + "learning_rate": 9.622733410632847e-05, + "loss": 1.1231, + "step": 2093 + }, + { + "epoch": 0.12722522631994654, + "grad_norm": 0.27409034967422485, + "learning_rate": 9.622368537395483e-05, + "loss": 1.1354, + "step": 2094 + }, + { + "epoch": 0.12728598335257307, + "grad_norm": 1.776922583580017, + "learning_rate": 9.622003494724307e-05, + "loss": 1.0703, + "step": 2095 + }, + { + "epoch": 0.12734674038519958, + "grad_norm": 0.46753358840942383, + "learning_rate": 9.621638282632698e-05, + "loss": 1.2442, + "step": 2096 + }, + { + "epoch": 0.12740749741782612, + "grad_norm": 0.28803837299346924, + "learning_rate": 9.621272901134046e-05, + "loss": 1.2062, + "step": 2097 + }, + { + "epoch": 0.12746825445045265, + "grad_norm": 0.4772378206253052, + "learning_rate": 9.620907350241742e-05, + "loss": 1.198, + "step": 2098 + }, + { + "epoch": 0.12752901148307916, + "grad_norm": 0.2363538295030594, + "learning_rate": 9.620541629969187e-05, + "loss": 1.1182, + "step": 2099 + }, + { + "epoch": 0.1275897685157057, + "grad_norm": 0.5278410315513611, + "learning_rate": 9.620175740329784e-05, + "loss": 1.2149, + "step": 2100 + }, + { + "epoch": 0.12765052554833223, + "grad_norm": 2.6147265434265137, + "learning_rate": 9.619809681336947e-05, + "loss": 1.0758, + "step": 2101 + }, + { + "epoch": 0.12771128258095873, + "grad_norm": 10.027637481689453, + "learning_rate": 9.619443453004092e-05, + "loss": 1.2425, + "step": 2102 + }, + { + "epoch": 0.12777203961358527, + "grad_norm": 1.025498867034912, + "learning_rate": 9.619077055344645e-05, + "loss": 1.187, + "step": 2103 + }, + { + "epoch": 0.1278327966462118, + "grad_norm": 0.8241715431213379, + "learning_rate": 9.618710488372035e-05, + "loss": 1.4068, + "step": 2104 + }, + { + "epoch": 0.1278935536788383, + "grad_norm": 0.3839017152786255, + "learning_rate": 9.618343752099701e-05, + "loss": 1.2756, + "step": 2105 + }, + { + "epoch": 0.12795431071146485, + "grad_norm": 0.5076162219047546, + "learning_rate": 9.617976846541084e-05, + "loss": 1.1222, + "step": 2106 + }, + { + "epoch": 0.12801506774409138, + "grad_norm": 0.8946338295936584, + "learning_rate": 9.617609771709632e-05, + "loss": 1.1835, + "step": 2107 + }, + { + "epoch": 0.12807582477671792, + "grad_norm": 0.6206286549568176, + "learning_rate": 9.617242527618802e-05, + "loss": 1.2043, + "step": 2108 + }, + { + "epoch": 0.12813658180934442, + "grad_norm": 0.7014588117599487, + "learning_rate": 9.616875114282053e-05, + "loss": 1.1608, + "step": 2109 + }, + { + "epoch": 0.12819733884197096, + "grad_norm": 0.22046838700771332, + "learning_rate": 9.616507531712855e-05, + "loss": 1.1383, + "step": 2110 + }, + { + "epoch": 0.1282580958745975, + "grad_norm": 0.7123265862464905, + "learning_rate": 9.616139779924682e-05, + "loss": 1.3398, + "step": 2111 + }, + { + "epoch": 0.128318852907224, + "grad_norm": 0.7035486698150635, + "learning_rate": 9.615771858931013e-05, + "loss": 1.0837, + "step": 2112 + }, + { + "epoch": 0.12837960993985054, + "grad_norm": 0.4782705307006836, + "learning_rate": 9.615403768745332e-05, + "loss": 1.3172, + "step": 2113 + }, + { + "epoch": 0.12844036697247707, + "grad_norm": 0.4483402669429779, + "learning_rate": 9.615035509381136e-05, + "loss": 1.171, + "step": 2114 + }, + { + "epoch": 0.12850112400510358, + "grad_norm": 0.8622254133224487, + "learning_rate": 9.614667080851922e-05, + "loss": 1.1128, + "step": 2115 + }, + { + "epoch": 0.1285618810377301, + "grad_norm": 0.3492850363254547, + "learning_rate": 9.614298483171192e-05, + "loss": 1.2447, + "step": 2116 + }, + { + "epoch": 0.12862263807035665, + "grad_norm": 0.6977007985115051, + "learning_rate": 9.61392971635246e-05, + "loss": 1.2216, + "step": 2117 + }, + { + "epoch": 0.12868339510298318, + "grad_norm": 0.3378121554851532, + "learning_rate": 9.613560780409243e-05, + "loss": 1.152, + "step": 2118 + }, + { + "epoch": 0.1287441521356097, + "grad_norm": 2.3020408153533936, + "learning_rate": 9.613191675355063e-05, + "loss": 1.1221, + "step": 2119 + }, + { + "epoch": 0.12880490916823623, + "grad_norm": 0.8442638516426086, + "learning_rate": 9.612822401203452e-05, + "loss": 1.1877, + "step": 2120 + }, + { + "epoch": 0.12886566620086276, + "grad_norm": 0.25912898778915405, + "learning_rate": 9.612452957967943e-05, + "loss": 1.0798, + "step": 2121 + }, + { + "epoch": 0.12892642323348927, + "grad_norm": 0.837409257888794, + "learning_rate": 9.612083345662078e-05, + "loss": 1.0602, + "step": 2122 + }, + { + "epoch": 0.1289871802661158, + "grad_norm": 0.4536583721637726, + "learning_rate": 9.611713564299408e-05, + "loss": 1.2194, + "step": 2123 + }, + { + "epoch": 0.12904793729874234, + "grad_norm": 0.5204725861549377, + "learning_rate": 9.611343613893487e-05, + "loss": 1.2341, + "step": 2124 + }, + { + "epoch": 0.12910869433136885, + "grad_norm": 0.904131293296814, + "learning_rate": 9.610973494457872e-05, + "loss": 1.2108, + "step": 2125 + }, + { + "epoch": 0.12916945136399538, + "grad_norm": 2.7763569355010986, + "learning_rate": 9.610603206006133e-05, + "loss": 1.1954, + "step": 2126 + }, + { + "epoch": 0.12923020839662192, + "grad_norm": 0.6752125024795532, + "learning_rate": 9.610232748551844e-05, + "loss": 1.3056, + "step": 2127 + }, + { + "epoch": 0.12929096542924842, + "grad_norm": 0.9078549146652222, + "learning_rate": 9.60986212210858e-05, + "loss": 1.1873, + "step": 2128 + }, + { + "epoch": 0.12935172246187496, + "grad_norm": 0.9426804780960083, + "learning_rate": 9.60949132668993e-05, + "loss": 1.2619, + "step": 2129 + }, + { + "epoch": 0.1294124794945015, + "grad_norm": 0.6227080225944519, + "learning_rate": 9.609120362309485e-05, + "loss": 1.0848, + "step": 2130 + }, + { + "epoch": 0.12947323652712803, + "grad_norm": 0.5800155997276306, + "learning_rate": 9.608749228980841e-05, + "loss": 1.2415, + "step": 2131 + }, + { + "epoch": 0.12953399355975453, + "grad_norm": 0.3226926326751709, + "learning_rate": 9.608377926717604e-05, + "loss": 1.1381, + "step": 2132 + }, + { + "epoch": 0.12959475059238107, + "grad_norm": 0.5377199053764343, + "learning_rate": 9.608006455533383e-05, + "loss": 1.2876, + "step": 2133 + }, + { + "epoch": 0.1296555076250076, + "grad_norm": 0.24919915199279785, + "learning_rate": 9.607634815441795e-05, + "loss": 1.1407, + "step": 2134 + }, + { + "epoch": 0.1297162646576341, + "grad_norm": 0.3594720959663391, + "learning_rate": 9.607263006456462e-05, + "loss": 1.126, + "step": 2135 + }, + { + "epoch": 0.12977702169026065, + "grad_norm": 0.43963423371315, + "learning_rate": 9.606891028591013e-05, + "loss": 1.083, + "step": 2136 + }, + { + "epoch": 0.12983777872288718, + "grad_norm": 0.46740561723709106, + "learning_rate": 9.606518881859083e-05, + "loss": 1.1769, + "step": 2137 + }, + { + "epoch": 0.1298985357555137, + "grad_norm": 0.4893267750740051, + "learning_rate": 9.606146566274312e-05, + "loss": 1.1168, + "step": 2138 + }, + { + "epoch": 0.12995929278814022, + "grad_norm": 0.43979552388191223, + "learning_rate": 9.60577408185035e-05, + "loss": 1.163, + "step": 2139 + }, + { + "epoch": 0.13002004982076676, + "grad_norm": 0.26166394352912903, + "learning_rate": 9.605401428600846e-05, + "loss": 1.0559, + "step": 2140 + }, + { + "epoch": 0.1300808068533933, + "grad_norm": 0.6359075903892517, + "learning_rate": 9.605028606539466e-05, + "loss": 1.3181, + "step": 2141 + }, + { + "epoch": 0.1301415638860198, + "grad_norm": 0.5433409810066223, + "learning_rate": 9.60465561567987e-05, + "loss": 1.234, + "step": 2142 + }, + { + "epoch": 0.13020232091864634, + "grad_norm": 1.0502241849899292, + "learning_rate": 9.604282456035733e-05, + "loss": 1.3229, + "step": 2143 + }, + { + "epoch": 0.13026307795127287, + "grad_norm": 0.681891143321991, + "learning_rate": 9.603909127620732e-05, + "loss": 1.1649, + "step": 2144 + }, + { + "epoch": 0.13032383498389938, + "grad_norm": 0.2804941236972809, + "learning_rate": 9.603535630448552e-05, + "loss": 1.0915, + "step": 2145 + }, + { + "epoch": 0.1303845920165259, + "grad_norm": 0.5088815689086914, + "learning_rate": 9.603161964532885e-05, + "loss": 1.1521, + "step": 2146 + }, + { + "epoch": 0.13044534904915245, + "grad_norm": 0.8997474312782288, + "learning_rate": 9.602788129887426e-05, + "loss": 1.2739, + "step": 2147 + }, + { + "epoch": 0.13050610608177896, + "grad_norm": 0.3703928291797638, + "learning_rate": 9.602414126525876e-05, + "loss": 1.3187, + "step": 2148 + }, + { + "epoch": 0.1305668631144055, + "grad_norm": 3.3672919273376465, + "learning_rate": 9.60203995446195e-05, + "loss": 1.0797, + "step": 2149 + }, + { + "epoch": 0.13062762014703203, + "grad_norm": 0.6692112684249878, + "learning_rate": 9.601665613709357e-05, + "loss": 1.2237, + "step": 2150 + }, + { + "epoch": 0.13068837717965853, + "grad_norm": 0.46392446756362915, + "learning_rate": 9.601291104281823e-05, + "loss": 1.114, + "step": 2151 + }, + { + "epoch": 0.13074913421228507, + "grad_norm": 0.5911328196525574, + "learning_rate": 9.600916426193074e-05, + "loss": 1.1327, + "step": 2152 + }, + { + "epoch": 0.1308098912449116, + "grad_norm": 0.5822901129722595, + "learning_rate": 9.600541579456844e-05, + "loss": 1.1284, + "step": 2153 + }, + { + "epoch": 0.13087064827753814, + "grad_norm": 0.36543387174606323, + "learning_rate": 9.600166564086874e-05, + "loss": 1.0993, + "step": 2154 + }, + { + "epoch": 0.13093140531016464, + "grad_norm": 1.4713478088378906, + "learning_rate": 9.599791380096908e-05, + "loss": 1.1267, + "step": 2155 + }, + { + "epoch": 0.13099216234279118, + "grad_norm": 0.34431546926498413, + "learning_rate": 9.5994160275007e-05, + "loss": 1.2385, + "step": 2156 + }, + { + "epoch": 0.13105291937541771, + "grad_norm": 0.9987990260124207, + "learning_rate": 9.59904050631201e-05, + "loss": 1.1339, + "step": 2157 + }, + { + "epoch": 0.13111367640804422, + "grad_norm": 0.5638148784637451, + "learning_rate": 9.5986648165446e-05, + "loss": 1.2649, + "step": 2158 + }, + { + "epoch": 0.13117443344067076, + "grad_norm": 0.6502277255058289, + "learning_rate": 9.598288958212242e-05, + "loss": 1.1107, + "step": 2159 + }, + { + "epoch": 0.1312351904732973, + "grad_norm": 0.39166396856307983, + "learning_rate": 9.597912931328714e-05, + "loss": 1.2855, + "step": 2160 + }, + { + "epoch": 0.1312959475059238, + "grad_norm": 0.33873245120048523, + "learning_rate": 9.597536735907797e-05, + "loss": 1.1435, + "step": 2161 + }, + { + "epoch": 0.13135670453855033, + "grad_norm": 0.3533531427383423, + "learning_rate": 9.597160371963286e-05, + "loss": 1.2459, + "step": 2162 + }, + { + "epoch": 0.13141746157117687, + "grad_norm": 0.29473355412483215, + "learning_rate": 9.596783839508969e-05, + "loss": 1.231, + "step": 2163 + }, + { + "epoch": 0.1314782186038034, + "grad_norm": 0.3215120732784271, + "learning_rate": 9.596407138558653e-05, + "loss": 1.0716, + "step": 2164 + }, + { + "epoch": 0.1315389756364299, + "grad_norm": 0.8879538774490356, + "learning_rate": 9.596030269126146e-05, + "loss": 1.1547, + "step": 2165 + }, + { + "epoch": 0.13159973266905645, + "grad_norm": 0.557893693447113, + "learning_rate": 9.595653231225261e-05, + "loss": 1.2268, + "step": 2166 + }, + { + "epoch": 0.13166048970168298, + "grad_norm": 1.1506541967391968, + "learning_rate": 9.595276024869817e-05, + "loss": 1.1509, + "step": 2167 + }, + { + "epoch": 0.1317212467343095, + "grad_norm": 0.2225322723388672, + "learning_rate": 9.594898650073644e-05, + "loss": 1.1107, + "step": 2168 + }, + { + "epoch": 0.13178200376693602, + "grad_norm": 0.17786553502082825, + "learning_rate": 9.594521106850571e-05, + "loss": 1.071, + "step": 2169 + }, + { + "epoch": 0.13184276079956256, + "grad_norm": 0.27344810962677, + "learning_rate": 9.594143395214441e-05, + "loss": 1.1604, + "step": 2170 + }, + { + "epoch": 0.13190351783218907, + "grad_norm": 2.2722160816192627, + "learning_rate": 9.593765515179095e-05, + "loss": 1.1419, + "step": 2171 + }, + { + "epoch": 0.1319642748648156, + "grad_norm": 0.2781543433666229, + "learning_rate": 9.593387466758386e-05, + "loss": 1.1222, + "step": 2172 + }, + { + "epoch": 0.13202503189744214, + "grad_norm": 0.26901504397392273, + "learning_rate": 9.593009249966172e-05, + "loss": 1.0881, + "step": 2173 + }, + { + "epoch": 0.13208578893006864, + "grad_norm": 0.19517002999782562, + "learning_rate": 9.592630864816317e-05, + "loss": 1.1118, + "step": 2174 + }, + { + "epoch": 0.13214654596269518, + "grad_norm": 0.28816160559654236, + "learning_rate": 9.59225231132269e-05, + "loss": 1.1206, + "step": 2175 + }, + { + "epoch": 0.1322073029953217, + "grad_norm": 0.29141685366630554, + "learning_rate": 9.591873589499166e-05, + "loss": 1.2177, + "step": 2176 + }, + { + "epoch": 0.13226806002794825, + "grad_norm": 0.20551984012126923, + "learning_rate": 9.591494699359627e-05, + "loss": 1.0626, + "step": 2177 + }, + { + "epoch": 0.13232881706057475, + "grad_norm": 0.29510751366615295, + "learning_rate": 9.591115640917963e-05, + "loss": 1.3861, + "step": 2178 + }, + { + "epoch": 0.1323895740932013, + "grad_norm": 3.546431541442871, + "learning_rate": 9.590736414188069e-05, + "loss": 1.2754, + "step": 2179 + }, + { + "epoch": 0.13245033112582782, + "grad_norm": 0.2874051332473755, + "learning_rate": 9.590357019183843e-05, + "loss": 1.1322, + "step": 2180 + }, + { + "epoch": 0.13251108815845433, + "grad_norm": 0.20549120008945465, + "learning_rate": 9.589977455919193e-05, + "loss": 1.1013, + "step": 2181 + }, + { + "epoch": 0.13257184519108087, + "grad_norm": 0.2770366072654724, + "learning_rate": 9.589597724408032e-05, + "loss": 1.1163, + "step": 2182 + }, + { + "epoch": 0.1326326022237074, + "grad_norm": 0.24166569113731384, + "learning_rate": 9.589217824664281e-05, + "loss": 1.0944, + "step": 2183 + }, + { + "epoch": 0.1326933592563339, + "grad_norm": 0.43672075867652893, + "learning_rate": 9.588837756701861e-05, + "loss": 1.2312, + "step": 2184 + }, + { + "epoch": 0.13275411628896044, + "grad_norm": 0.23559635877609253, + "learning_rate": 9.588457520534708e-05, + "loss": 1.1368, + "step": 2185 + }, + { + "epoch": 0.13281487332158698, + "grad_norm": 0.2878612279891968, + "learning_rate": 9.588077116176756e-05, + "loss": 1.1772, + "step": 2186 + }, + { + "epoch": 0.1328756303542135, + "grad_norm": 0.4838666319847107, + "learning_rate": 9.587696543641953e-05, + "loss": 1.2777, + "step": 2187 + }, + { + "epoch": 0.13293638738684002, + "grad_norm": 0.2695195972919464, + "learning_rate": 9.587315802944245e-05, + "loss": 1.2124, + "step": 2188 + }, + { + "epoch": 0.13299714441946656, + "grad_norm": 0.5042257905006409, + "learning_rate": 9.58693489409759e-05, + "loss": 1.1258, + "step": 2189 + }, + { + "epoch": 0.1330579014520931, + "grad_norm": 0.2447011023759842, + "learning_rate": 9.586553817115948e-05, + "loss": 1.082, + "step": 2190 + }, + { + "epoch": 0.1331186584847196, + "grad_norm": 0.23643170297145844, + "learning_rate": 9.586172572013291e-05, + "loss": 1.1386, + "step": 2191 + }, + { + "epoch": 0.13317941551734613, + "grad_norm": 0.28526461124420166, + "learning_rate": 9.585791158803592e-05, + "loss": 1.1733, + "step": 2192 + }, + { + "epoch": 0.13324017254997267, + "grad_norm": 0.35178592801094055, + "learning_rate": 9.58540957750083e-05, + "loss": 1.1491, + "step": 2193 + }, + { + "epoch": 0.13330092958259918, + "grad_norm": 0.23286591470241547, + "learning_rate": 9.585027828118995e-05, + "loss": 1.1351, + "step": 2194 + }, + { + "epoch": 0.1333616866152257, + "grad_norm": 0.737048864364624, + "learning_rate": 9.584645910672078e-05, + "loss": 1.094, + "step": 2195 + }, + { + "epoch": 0.13342244364785225, + "grad_norm": 0.36242058873176575, + "learning_rate": 9.58426382517408e-05, + "loss": 1.1289, + "step": 2196 + }, + { + "epoch": 0.13348320068047875, + "grad_norm": 0.3016394078731537, + "learning_rate": 9.583881571639005e-05, + "loss": 1.2083, + "step": 2197 + }, + { + "epoch": 0.1335439577131053, + "grad_norm": 0.599119246006012, + "learning_rate": 9.583499150080865e-05, + "loss": 1.1313, + "step": 2198 + }, + { + "epoch": 0.13360471474573182, + "grad_norm": 0.21455225348472595, + "learning_rate": 9.583116560513678e-05, + "loss": 1.0772, + "step": 2199 + }, + { + "epoch": 0.13366547177835836, + "grad_norm": 0.4866546094417572, + "learning_rate": 9.582733802951467e-05, + "loss": 1.1044, + "step": 2200 + }, + { + "epoch": 0.13372622881098487, + "grad_norm": 0.46806877851486206, + "learning_rate": 9.582350877408262e-05, + "loss": 1.1194, + "step": 2201 + }, + { + "epoch": 0.1337869858436114, + "grad_norm": 0.291044682264328, + "learning_rate": 9.5819677838981e-05, + "loss": 1.1643, + "step": 2202 + }, + { + "epoch": 0.13384774287623794, + "grad_norm": 0.31724461913108826, + "learning_rate": 9.581584522435024e-05, + "loss": 1.0819, + "step": 2203 + }, + { + "epoch": 0.13390849990886444, + "grad_norm": 0.38432320952415466, + "learning_rate": 9.581201093033082e-05, + "loss": 1.1819, + "step": 2204 + }, + { + "epoch": 0.13396925694149098, + "grad_norm": 0.36087366938591003, + "learning_rate": 9.580817495706327e-05, + "loss": 1.0916, + "step": 2205 + }, + { + "epoch": 0.1340300139741175, + "grad_norm": 1.5061930418014526, + "learning_rate": 9.580433730468823e-05, + "loss": 1.1589, + "step": 2206 + }, + { + "epoch": 0.13409077100674402, + "grad_norm": 0.22357815504074097, + "learning_rate": 9.580049797334634e-05, + "loss": 1.1477, + "step": 2207 + }, + { + "epoch": 0.13415152803937055, + "grad_norm": 0.35472071170806885, + "learning_rate": 9.579665696317835e-05, + "loss": 1.2887, + "step": 2208 + }, + { + "epoch": 0.1342122850719971, + "grad_norm": 1.228430151939392, + "learning_rate": 9.579281427432504e-05, + "loss": 1.1396, + "step": 2209 + }, + { + "epoch": 0.1342730421046236, + "grad_norm": 0.28522780537605286, + "learning_rate": 9.578896990692727e-05, + "loss": 1.1825, + "step": 2210 + }, + { + "epoch": 0.13433379913725013, + "grad_norm": 0.24901317059993744, + "learning_rate": 9.578512386112596e-05, + "loss": 1.0975, + "step": 2211 + }, + { + "epoch": 0.13439455616987667, + "grad_norm": 0.46600842475891113, + "learning_rate": 9.578127613706209e-05, + "loss": 1.1027, + "step": 2212 + }, + { + "epoch": 0.1344553132025032, + "grad_norm": 0.2915239930152893, + "learning_rate": 9.577742673487668e-05, + "loss": 1.196, + "step": 2213 + }, + { + "epoch": 0.1345160702351297, + "grad_norm": 0.5074911117553711, + "learning_rate": 9.577357565471086e-05, + "loss": 1.1142, + "step": 2214 + }, + { + "epoch": 0.13457682726775624, + "grad_norm": 0.2162403166294098, + "learning_rate": 9.576972289670578e-05, + "loss": 1.1174, + "step": 2215 + }, + { + "epoch": 0.13463758430038278, + "grad_norm": 0.2914368808269501, + "learning_rate": 9.576586846100264e-05, + "loss": 1.1341, + "step": 2216 + }, + { + "epoch": 0.13469834133300929, + "grad_norm": 0.8893709778785706, + "learning_rate": 9.576201234774275e-05, + "loss": 1.0927, + "step": 2217 + }, + { + "epoch": 0.13475909836563582, + "grad_norm": 0.20071116089820862, + "learning_rate": 9.575815455706748e-05, + "loss": 1.0787, + "step": 2218 + }, + { + "epoch": 0.13481985539826236, + "grad_norm": 0.49274778366088867, + "learning_rate": 9.575429508911817e-05, + "loss": 1.2033, + "step": 2219 + }, + { + "epoch": 0.13488061243088886, + "grad_norm": 0.33693727850914, + "learning_rate": 9.575043394403635e-05, + "loss": 1.2642, + "step": 2220 + }, + { + "epoch": 0.1349413694635154, + "grad_norm": 0.18772336840629578, + "learning_rate": 9.574657112196354e-05, + "loss": 1.1091, + "step": 2221 + }, + { + "epoch": 0.13500212649614193, + "grad_norm": 2.229919672012329, + "learning_rate": 9.57427066230413e-05, + "loss": 1.0775, + "step": 2222 + }, + { + "epoch": 0.13506288352876847, + "grad_norm": 0.35690152645111084, + "learning_rate": 9.573884044741133e-05, + "loss": 1.1256, + "step": 2223 + }, + { + "epoch": 0.13512364056139498, + "grad_norm": 0.22040393948554993, + "learning_rate": 9.57349725952153e-05, + "loss": 1.139, + "step": 2224 + }, + { + "epoch": 0.1351843975940215, + "grad_norm": 0.41975006461143494, + "learning_rate": 9.573110306659504e-05, + "loss": 1.2138, + "step": 2225 + }, + { + "epoch": 0.13524515462664805, + "grad_norm": 0.3909105956554413, + "learning_rate": 9.572723186169232e-05, + "loss": 1.1234, + "step": 2226 + }, + { + "epoch": 0.13530591165927455, + "grad_norm": 0.22478719055652618, + "learning_rate": 9.57233589806491e-05, + "loss": 1.0891, + "step": 2227 + }, + { + "epoch": 0.1353666686919011, + "grad_norm": 0.22070099413394928, + "learning_rate": 9.57194844236073e-05, + "loss": 1.077, + "step": 2228 + }, + { + "epoch": 0.13542742572452762, + "grad_norm": 0.32302504777908325, + "learning_rate": 9.571560819070898e-05, + "loss": 1.192, + "step": 2229 + }, + { + "epoch": 0.13548818275715413, + "grad_norm": 0.20182038843631744, + "learning_rate": 9.57117302820962e-05, + "loss": 1.1725, + "step": 2230 + }, + { + "epoch": 0.13554893978978066, + "grad_norm": 0.4659487009048462, + "learning_rate": 9.57078506979111e-05, + "loss": 1.121, + "step": 2231 + }, + { + "epoch": 0.1356096968224072, + "grad_norm": 0.2527261972427368, + "learning_rate": 9.570396943829592e-05, + "loss": 1.1731, + "step": 2232 + }, + { + "epoch": 0.1356704538550337, + "grad_norm": 0.32121726870536804, + "learning_rate": 9.570008650339288e-05, + "loss": 1.1213, + "step": 2233 + }, + { + "epoch": 0.13573121088766024, + "grad_norm": 0.37967798113822937, + "learning_rate": 9.569620189334435e-05, + "loss": 1.2954, + "step": 2234 + }, + { + "epoch": 0.13579196792028678, + "grad_norm": 0.19558396935462952, + "learning_rate": 9.56923156082927e-05, + "loss": 1.114, + "step": 2235 + }, + { + "epoch": 0.1358527249529133, + "grad_norm": 0.32540005445480347, + "learning_rate": 9.568842764838041e-05, + "loss": 1.0942, + "step": 2236 + }, + { + "epoch": 0.13591348198553982, + "grad_norm": 0.30359238386154175, + "learning_rate": 9.568453801374995e-05, + "loss": 1.2996, + "step": 2237 + }, + { + "epoch": 0.13597423901816635, + "grad_norm": 0.6339631080627441, + "learning_rate": 9.568064670454393e-05, + "loss": 1.0678, + "step": 2238 + }, + { + "epoch": 0.1360349960507929, + "grad_norm": 0.27422013878822327, + "learning_rate": 9.567675372090497e-05, + "loss": 1.139, + "step": 2239 + }, + { + "epoch": 0.1360957530834194, + "grad_norm": 0.3529973328113556, + "learning_rate": 9.567285906297577e-05, + "loss": 1.2865, + "step": 2240 + }, + { + "epoch": 0.13615651011604593, + "grad_norm": 1.2537630796432495, + "learning_rate": 9.56689627308991e-05, + "loss": 1.2045, + "step": 2241 + }, + { + "epoch": 0.13621726714867247, + "grad_norm": 0.5994548797607422, + "learning_rate": 9.566506472481778e-05, + "loss": 1.1065, + "step": 2242 + }, + { + "epoch": 0.13627802418129897, + "grad_norm": 1.985215187072754, + "learning_rate": 9.566116504487466e-05, + "loss": 1.1796, + "step": 2243 + }, + { + "epoch": 0.1363387812139255, + "grad_norm": 0.39160412549972534, + "learning_rate": 9.565726369121273e-05, + "loss": 1.118, + "step": 2244 + }, + { + "epoch": 0.13639953824655204, + "grad_norm": 0.2566390931606293, + "learning_rate": 9.565336066397496e-05, + "loss": 1.0957, + "step": 2245 + }, + { + "epoch": 0.13646029527917858, + "grad_norm": 0.44709840416908264, + "learning_rate": 9.564945596330446e-05, + "loss": 1.3552, + "step": 2246 + }, + { + "epoch": 0.13652105231180509, + "grad_norm": 0.3959415555000305, + "learning_rate": 9.564554958934431e-05, + "loss": 1.2443, + "step": 2247 + }, + { + "epoch": 0.13658180934443162, + "grad_norm": 0.2654580771923065, + "learning_rate": 9.56416415422377e-05, + "loss": 1.1212, + "step": 2248 + }, + { + "epoch": 0.13664256637705816, + "grad_norm": 0.41380831599235535, + "learning_rate": 9.563773182212791e-05, + "loss": 1.1667, + "step": 2249 + }, + { + "epoch": 0.13670332340968466, + "grad_norm": 0.3116408586502075, + "learning_rate": 9.563382042915826e-05, + "loss": 1.1065, + "step": 2250 + }, + { + "epoch": 0.1367640804423112, + "grad_norm": 2.370426893234253, + "learning_rate": 9.562990736347207e-05, + "loss": 1.1833, + "step": 2251 + }, + { + "epoch": 0.13682483747493773, + "grad_norm": 0.45543670654296875, + "learning_rate": 9.562599262521281e-05, + "loss": 1.1099, + "step": 2252 + }, + { + "epoch": 0.13688559450756424, + "grad_norm": 0.39558133482933044, + "learning_rate": 9.562207621452398e-05, + "loss": 1.2975, + "step": 2253 + }, + { + "epoch": 0.13694635154019077, + "grad_norm": 0.5960584282875061, + "learning_rate": 9.561815813154913e-05, + "loss": 1.1221, + "step": 2254 + }, + { + "epoch": 0.1370071085728173, + "grad_norm": 0.31104594469070435, + "learning_rate": 9.561423837643186e-05, + "loss": 1.1316, + "step": 2255 + }, + { + "epoch": 0.13706786560544382, + "grad_norm": 0.9412122368812561, + "learning_rate": 9.561031694931589e-05, + "loss": 1.1108, + "step": 2256 + }, + { + "epoch": 0.13712862263807035, + "grad_norm": 0.3803810775279999, + "learning_rate": 9.560639385034491e-05, + "loss": 1.1156, + "step": 2257 + }, + { + "epoch": 0.1371893796706969, + "grad_norm": 0.2515859603881836, + "learning_rate": 9.560246907966276e-05, + "loss": 1.0899, + "step": 2258 + }, + { + "epoch": 0.13725013670332342, + "grad_norm": 1.1129956245422363, + "learning_rate": 9.55985426374133e-05, + "loss": 1.0837, + "step": 2259 + }, + { + "epoch": 0.13731089373594993, + "grad_norm": 0.4452212452888489, + "learning_rate": 9.559461452374043e-05, + "loss": 1.1324, + "step": 2260 + }, + { + "epoch": 0.13737165076857646, + "grad_norm": 0.3553762137889862, + "learning_rate": 9.559068473878816e-05, + "loss": 1.0399, + "step": 2261 + }, + { + "epoch": 0.137432407801203, + "grad_norm": 0.44725340604782104, + "learning_rate": 9.558675328270053e-05, + "loss": 1.2255, + "step": 2262 + }, + { + "epoch": 0.1374931648338295, + "grad_norm": 0.5255569219589233, + "learning_rate": 9.558282015562165e-05, + "loss": 1.0988, + "step": 2263 + }, + { + "epoch": 0.13755392186645604, + "grad_norm": 0.23519571125507355, + "learning_rate": 9.557888535769568e-05, + "loss": 1.0964, + "step": 2264 + }, + { + "epoch": 0.13761467889908258, + "grad_norm": 0.3283027708530426, + "learning_rate": 9.557494888906685e-05, + "loss": 1.1026, + "step": 2265 + }, + { + "epoch": 0.13767543593170908, + "grad_norm": 0.2479170709848404, + "learning_rate": 9.557101074987946e-05, + "loss": 1.126, + "step": 2266 + }, + { + "epoch": 0.13773619296433562, + "grad_norm": 0.22774219512939453, + "learning_rate": 9.556707094027786e-05, + "loss": 1.1703, + "step": 2267 + }, + { + "epoch": 0.13779694999696215, + "grad_norm": 0.9517698287963867, + "learning_rate": 9.556312946040649e-05, + "loss": 1.3336, + "step": 2268 + }, + { + "epoch": 0.1378577070295887, + "grad_norm": 0.2158520668745041, + "learning_rate": 9.555918631040979e-05, + "loss": 1.1216, + "step": 2269 + }, + { + "epoch": 0.1379184640622152, + "grad_norm": 0.215272456407547, + "learning_rate": 9.55552414904323e-05, + "loss": 1.1408, + "step": 2270 + }, + { + "epoch": 0.13797922109484173, + "grad_norm": 0.44052061438560486, + "learning_rate": 9.555129500061864e-05, + "loss": 1.1433, + "step": 2271 + }, + { + "epoch": 0.13803997812746827, + "grad_norm": 0.25354281067848206, + "learning_rate": 9.554734684111345e-05, + "loss": 1.2066, + "step": 2272 + }, + { + "epoch": 0.13810073516009477, + "grad_norm": 0.32312583923339844, + "learning_rate": 9.554339701206145e-05, + "loss": 1.2472, + "step": 2273 + }, + { + "epoch": 0.1381614921927213, + "grad_norm": 1.4273040294647217, + "learning_rate": 9.553944551360745e-05, + "loss": 1.1652, + "step": 2274 + }, + { + "epoch": 0.13822224922534784, + "grad_norm": 0.9198738932609558, + "learning_rate": 9.553549234589624e-05, + "loss": 1.3033, + "step": 2275 + }, + { + "epoch": 0.13828300625797435, + "grad_norm": 0.29373809695243835, + "learning_rate": 9.553153750907279e-05, + "loss": 1.1147, + "step": 2276 + }, + { + "epoch": 0.13834376329060089, + "grad_norm": 0.3678988218307495, + "learning_rate": 9.552758100328202e-05, + "loss": 1.1929, + "step": 2277 + }, + { + "epoch": 0.13840452032322742, + "grad_norm": 0.3079800605773926, + "learning_rate": 9.552362282866896e-05, + "loss": 1.0762, + "step": 2278 + }, + { + "epoch": 0.13846527735585393, + "grad_norm": 0.24822096526622772, + "learning_rate": 9.55196629853787e-05, + "loss": 1.1433, + "step": 2279 + }, + { + "epoch": 0.13852603438848046, + "grad_norm": 0.34546127915382385, + "learning_rate": 9.551570147355642e-05, + "loss": 1.2324, + "step": 2280 + }, + { + "epoch": 0.138586791421107, + "grad_norm": 0.2798556685447693, + "learning_rate": 9.55117382933473e-05, + "loss": 1.1784, + "step": 2281 + }, + { + "epoch": 0.13864754845373353, + "grad_norm": 1.5265111923217773, + "learning_rate": 9.55077734448966e-05, + "loss": 1.2166, + "step": 2282 + }, + { + "epoch": 0.13870830548636004, + "grad_norm": 0.7261504530906677, + "learning_rate": 9.550380692834967e-05, + "loss": 1.1279, + "step": 2283 + }, + { + "epoch": 0.13876906251898657, + "grad_norm": 0.2804091274738312, + "learning_rate": 9.549983874385191e-05, + "loss": 1.1426, + "step": 2284 + }, + { + "epoch": 0.1388298195516131, + "grad_norm": 0.3043665885925293, + "learning_rate": 9.549586889154875e-05, + "loss": 1.1262, + "step": 2285 + }, + { + "epoch": 0.13889057658423962, + "grad_norm": 0.7021775841712952, + "learning_rate": 9.549189737158574e-05, + "loss": 1.0911, + "step": 2286 + }, + { + "epoch": 0.13895133361686615, + "grad_norm": 0.5193327069282532, + "learning_rate": 9.548792418410844e-05, + "loss": 1.1868, + "step": 2287 + }, + { + "epoch": 0.1390120906494927, + "grad_norm": 2.50764536857605, + "learning_rate": 9.548394932926248e-05, + "loss": 1.2353, + "step": 2288 + }, + { + "epoch": 0.1390728476821192, + "grad_norm": 1.069471001625061, + "learning_rate": 9.547997280719355e-05, + "loss": 1.2042, + "step": 2289 + }, + { + "epoch": 0.13913360471474573, + "grad_norm": 0.26147180795669556, + "learning_rate": 9.547599461804743e-05, + "loss": 1.1772, + "step": 2290 + }, + { + "epoch": 0.13919436174737226, + "grad_norm": 0.24329783022403717, + "learning_rate": 9.547201476196996e-05, + "loss": 1.1292, + "step": 2291 + }, + { + "epoch": 0.13925511877999877, + "grad_norm": 0.205004021525383, + "learning_rate": 9.546803323910698e-05, + "loss": 1.0776, + "step": 2292 + }, + { + "epoch": 0.1393158758126253, + "grad_norm": 0.3156979978084564, + "learning_rate": 9.546405004960445e-05, + "loss": 1.0596, + "step": 2293 + }, + { + "epoch": 0.13937663284525184, + "grad_norm": 1.8628566265106201, + "learning_rate": 9.546006519360839e-05, + "loss": 1.0926, + "step": 2294 + }, + { + "epoch": 0.13943738987787838, + "grad_norm": 0.2612345814704895, + "learning_rate": 9.545607867126484e-05, + "loss": 1.1799, + "step": 2295 + }, + { + "epoch": 0.13949814691050488, + "grad_norm": 0.2186928391456604, + "learning_rate": 9.545209048271997e-05, + "loss": 1.1231, + "step": 2296 + }, + { + "epoch": 0.13955890394313142, + "grad_norm": 0.2838176488876343, + "learning_rate": 9.54481006281199e-05, + "loss": 1.2351, + "step": 2297 + }, + { + "epoch": 0.13961966097575795, + "grad_norm": 0.1982366293668747, + "learning_rate": 9.544410910761094e-05, + "loss": 1.1598, + "step": 2298 + }, + { + "epoch": 0.13968041800838446, + "grad_norm": 0.19656279683113098, + "learning_rate": 9.544011592133936e-05, + "loss": 1.0744, + "step": 2299 + }, + { + "epoch": 0.139741175041011, + "grad_norm": 0.3489125669002533, + "learning_rate": 9.543612106945157e-05, + "loss": 1.081, + "step": 2300 + }, + { + "epoch": 0.13980193207363753, + "grad_norm": 0.836559534072876, + "learning_rate": 9.543212455209396e-05, + "loss": 1.0798, + "step": 2301 + }, + { + "epoch": 0.13986268910626404, + "grad_norm": 0.24588853120803833, + "learning_rate": 9.542812636941306e-05, + "loss": 1.1039, + "step": 2302 + }, + { + "epoch": 0.13992344613889057, + "grad_norm": 1.5301426649093628, + "learning_rate": 9.542412652155539e-05, + "loss": 1.1047, + "step": 2303 + }, + { + "epoch": 0.1399842031715171, + "grad_norm": 0.2257785201072693, + "learning_rate": 9.542012500866759e-05, + "loss": 1.1092, + "step": 2304 + }, + { + "epoch": 0.14004496020414364, + "grad_norm": 0.5840712785720825, + "learning_rate": 9.541612183089632e-05, + "loss": 1.2012, + "step": 2305 + }, + { + "epoch": 0.14010571723677015, + "grad_norm": 1.0040967464447021, + "learning_rate": 9.541211698838834e-05, + "loss": 1.0606, + "step": 2306 + }, + { + "epoch": 0.14016647426939668, + "grad_norm": 0.3399321436882019, + "learning_rate": 9.540811048129041e-05, + "loss": 1.1485, + "step": 2307 + }, + { + "epoch": 0.14022723130202322, + "grad_norm": 0.3565303683280945, + "learning_rate": 9.540410230974943e-05, + "loss": 1.4051, + "step": 2308 + }, + { + "epoch": 0.14028798833464973, + "grad_norm": 0.30013400316238403, + "learning_rate": 9.54000924739123e-05, + "loss": 1.1917, + "step": 2309 + }, + { + "epoch": 0.14034874536727626, + "grad_norm": 0.32694026827812195, + "learning_rate": 9.539608097392602e-05, + "loss": 1.1477, + "step": 2310 + }, + { + "epoch": 0.1404095023999028, + "grad_norm": 0.5325272083282471, + "learning_rate": 9.539206780993761e-05, + "loss": 1.1865, + "step": 2311 + }, + { + "epoch": 0.1404702594325293, + "grad_norm": 0.33460378646850586, + "learning_rate": 9.538805298209416e-05, + "loss": 1.1364, + "step": 2312 + }, + { + "epoch": 0.14053101646515584, + "grad_norm": 0.3484179377555847, + "learning_rate": 9.538403649054288e-05, + "loss": 1.0848, + "step": 2313 + }, + { + "epoch": 0.14059177349778237, + "grad_norm": 0.4126632809638977, + "learning_rate": 9.538001833543095e-05, + "loss": 1.1471, + "step": 2314 + }, + { + "epoch": 0.14065253053040888, + "grad_norm": 0.41456538438796997, + "learning_rate": 9.537599851690569e-05, + "loss": 1.1267, + "step": 2315 + }, + { + "epoch": 0.14071328756303542, + "grad_norm": 0.5023922920227051, + "learning_rate": 9.537197703511443e-05, + "loss": 1.1365, + "step": 2316 + }, + { + "epoch": 0.14077404459566195, + "grad_norm": 0.4450730085372925, + "learning_rate": 9.536795389020457e-05, + "loss": 1.2871, + "step": 2317 + }, + { + "epoch": 0.1408348016282885, + "grad_norm": 0.7221823334693909, + "learning_rate": 9.536392908232361e-05, + "loss": 1.1133, + "step": 2318 + }, + { + "epoch": 0.140895558660915, + "grad_norm": 0.726874828338623, + "learning_rate": 9.535990261161904e-05, + "loss": 1.2013, + "step": 2319 + }, + { + "epoch": 0.14095631569354153, + "grad_norm": 0.5453711748123169, + "learning_rate": 9.535587447823849e-05, + "loss": 1.2047, + "step": 2320 + }, + { + "epoch": 0.14101707272616806, + "grad_norm": 0.2919989824295044, + "learning_rate": 9.535184468232958e-05, + "loss": 1.1354, + "step": 2321 + }, + { + "epoch": 0.14107782975879457, + "grad_norm": 0.29179686307907104, + "learning_rate": 9.534781322404005e-05, + "loss": 1.1736, + "step": 2322 + }, + { + "epoch": 0.1411385867914211, + "grad_norm": 0.2980515956878662, + "learning_rate": 9.534378010351766e-05, + "loss": 1.0663, + "step": 2323 + }, + { + "epoch": 0.14119934382404764, + "grad_norm": 0.3906993865966797, + "learning_rate": 9.533974532091023e-05, + "loss": 1.0915, + "step": 2324 + }, + { + "epoch": 0.14126010085667415, + "grad_norm": 0.3440910875797272, + "learning_rate": 9.53357088763657e-05, + "loss": 1.1644, + "step": 2325 + }, + { + "epoch": 0.14132085788930068, + "grad_norm": 0.4226357042789459, + "learning_rate": 9.533167077003198e-05, + "loss": 1.0523, + "step": 2326 + }, + { + "epoch": 0.14138161492192722, + "grad_norm": 0.2944369912147522, + "learning_rate": 9.532763100205709e-05, + "loss": 1.1802, + "step": 2327 + }, + { + "epoch": 0.14144237195455375, + "grad_norm": 0.7065515518188477, + "learning_rate": 9.532358957258915e-05, + "loss": 1.2199, + "step": 2328 + }, + { + "epoch": 0.14150312898718026, + "grad_norm": 0.2822442054748535, + "learning_rate": 9.531954648177625e-05, + "loss": 1.0817, + "step": 2329 + }, + { + "epoch": 0.1415638860198068, + "grad_norm": 0.6117932796478271, + "learning_rate": 9.531550172976664e-05, + "loss": 1.2473, + "step": 2330 + }, + { + "epoch": 0.14162464305243333, + "grad_norm": 0.60774827003479, + "learning_rate": 9.531145531670853e-05, + "loss": 1.1248, + "step": 2331 + }, + { + "epoch": 0.14168540008505984, + "grad_norm": 0.4836391508579254, + "learning_rate": 9.530740724275028e-05, + "loss": 1.2029, + "step": 2332 + }, + { + "epoch": 0.14174615711768637, + "grad_norm": 0.5419015884399414, + "learning_rate": 9.530335750804025e-05, + "loss": 1.2447, + "step": 2333 + }, + { + "epoch": 0.1418069141503129, + "grad_norm": 0.3192337155342102, + "learning_rate": 9.52993061127269e-05, + "loss": 1.1027, + "step": 2334 + }, + { + "epoch": 0.14186767118293941, + "grad_norm": 0.6591618061065674, + "learning_rate": 9.529525305695874e-05, + "loss": 1.1899, + "step": 2335 + }, + { + "epoch": 0.14192842821556595, + "grad_norm": 0.2896754741668701, + "learning_rate": 9.529119834088432e-05, + "loss": 1.0976, + "step": 2336 + }, + { + "epoch": 0.14198918524819248, + "grad_norm": 0.23174209892749786, + "learning_rate": 9.528714196465224e-05, + "loss": 1.2125, + "step": 2337 + }, + { + "epoch": 0.142049942280819, + "grad_norm": 0.333446741104126, + "learning_rate": 9.528308392841126e-05, + "loss": 1.3078, + "step": 2338 + }, + { + "epoch": 0.14211069931344553, + "grad_norm": 0.23320063948631287, + "learning_rate": 9.527902423231005e-05, + "loss": 1.1999, + "step": 2339 + }, + { + "epoch": 0.14217145634607206, + "grad_norm": 0.3463149964809418, + "learning_rate": 9.527496287649748e-05, + "loss": 1.1503, + "step": 2340 + }, + { + "epoch": 0.1422322133786986, + "grad_norm": 0.32801148295402527, + "learning_rate": 9.527089986112239e-05, + "loss": 1.1919, + "step": 2341 + }, + { + "epoch": 0.1422929704113251, + "grad_norm": 0.30774742364883423, + "learning_rate": 9.52668351863337e-05, + "loss": 1.1929, + "step": 2342 + }, + { + "epoch": 0.14235372744395164, + "grad_norm": 0.3742309510707855, + "learning_rate": 9.526276885228044e-05, + "loss": 1.1008, + "step": 2343 + }, + { + "epoch": 0.14241448447657817, + "grad_norm": 0.3455180525779724, + "learning_rate": 9.525870085911162e-05, + "loss": 1.3977, + "step": 2344 + }, + { + "epoch": 0.14247524150920468, + "grad_norm": 0.2631663382053375, + "learning_rate": 9.525463120697637e-05, + "loss": 1.0667, + "step": 2345 + }, + { + "epoch": 0.14253599854183122, + "grad_norm": 0.28304222226142883, + "learning_rate": 9.525055989602388e-05, + "loss": 1.3099, + "step": 2346 + }, + { + "epoch": 0.14259675557445775, + "grad_norm": 3.1334152221679688, + "learning_rate": 9.524648692640336e-05, + "loss": 1.1185, + "step": 2347 + }, + { + "epoch": 0.14265751260708426, + "grad_norm": 0.3407087028026581, + "learning_rate": 9.524241229826411e-05, + "loss": 1.1178, + "step": 2348 + }, + { + "epoch": 0.1427182696397108, + "grad_norm": 0.7797468304634094, + "learning_rate": 9.523833601175551e-05, + "loss": 1.0931, + "step": 2349 + }, + { + "epoch": 0.14277902667233733, + "grad_norm": 0.2778628170490265, + "learning_rate": 9.523425806702695e-05, + "loss": 1.1084, + "step": 2350 + }, + { + "epoch": 0.14283978370496386, + "grad_norm": 0.23804964125156403, + "learning_rate": 9.523017846422793e-05, + "loss": 1.1044, + "step": 2351 + }, + { + "epoch": 0.14290054073759037, + "grad_norm": 0.44853854179382324, + "learning_rate": 9.522609720350795e-05, + "loss": 1.317, + "step": 2352 + }, + { + "epoch": 0.1429612977702169, + "grad_norm": 0.24140501022338867, + "learning_rate": 9.522201428501664e-05, + "loss": 1.0917, + "step": 2353 + }, + { + "epoch": 0.14302205480284344, + "grad_norm": 0.40807491540908813, + "learning_rate": 9.521792970890366e-05, + "loss": 1.1978, + "step": 2354 + }, + { + "epoch": 0.14308281183546995, + "grad_norm": 0.28740110993385315, + "learning_rate": 9.521384347531874e-05, + "loss": 1.2265, + "step": 2355 + }, + { + "epoch": 0.14314356886809648, + "grad_norm": 0.1885412633419037, + "learning_rate": 9.520975558441164e-05, + "loss": 1.0723, + "step": 2356 + }, + { + "epoch": 0.14320432590072302, + "grad_norm": 0.31568068265914917, + "learning_rate": 9.520566603633221e-05, + "loss": 1.0841, + "step": 2357 + }, + { + "epoch": 0.14326508293334952, + "grad_norm": 0.23744343221187592, + "learning_rate": 9.520157483123034e-05, + "loss": 1.1876, + "step": 2358 + }, + { + "epoch": 0.14332583996597606, + "grad_norm": 0.29597485065460205, + "learning_rate": 9.519748196925601e-05, + "loss": 1.2222, + "step": 2359 + }, + { + "epoch": 0.1433865969986026, + "grad_norm": 1.6374118328094482, + "learning_rate": 9.519338745055926e-05, + "loss": 1.0676, + "step": 2360 + }, + { + "epoch": 0.1434473540312291, + "grad_norm": 0.5812447667121887, + "learning_rate": 9.518929127529016e-05, + "loss": 1.1388, + "step": 2361 + }, + { + "epoch": 0.14350811106385564, + "grad_norm": 0.22026678919792175, + "learning_rate": 9.518519344359882e-05, + "loss": 1.1516, + "step": 2362 + }, + { + "epoch": 0.14356886809648217, + "grad_norm": 0.29614517092704773, + "learning_rate": 9.51810939556355e-05, + "loss": 1.1885, + "step": 2363 + }, + { + "epoch": 0.1436296251291087, + "grad_norm": 3.8940482139587402, + "learning_rate": 9.517699281155046e-05, + "loss": 1.1183, + "step": 2364 + }, + { + "epoch": 0.14369038216173521, + "grad_norm": 2.639572858810425, + "learning_rate": 9.5172890011494e-05, + "loss": 1.2733, + "step": 2365 + }, + { + "epoch": 0.14375113919436175, + "grad_norm": 0.2860769033432007, + "learning_rate": 9.516878555561652e-05, + "loss": 1.0693, + "step": 2366 + }, + { + "epoch": 0.14381189622698828, + "grad_norm": 7.738712787628174, + "learning_rate": 9.516467944406848e-05, + "loss": 1.1449, + "step": 2367 + }, + { + "epoch": 0.1438726532596148, + "grad_norm": 0.29405665397644043, + "learning_rate": 9.516057167700039e-05, + "loss": 1.1305, + "step": 2368 + }, + { + "epoch": 0.14393341029224133, + "grad_norm": 0.3594772219657898, + "learning_rate": 9.515646225456282e-05, + "loss": 1.1268, + "step": 2369 + }, + { + "epoch": 0.14399416732486786, + "grad_norm": 0.36056041717529297, + "learning_rate": 9.51523511769064e-05, + "loss": 1.1363, + "step": 2370 + }, + { + "epoch": 0.14405492435749437, + "grad_norm": 0.6050369739532471, + "learning_rate": 9.51482384441818e-05, + "loss": 1.0485, + "step": 2371 + }, + { + "epoch": 0.1441156813901209, + "grad_norm": 0.4982859194278717, + "learning_rate": 9.51441240565398e-05, + "loss": 1.1746, + "step": 2372 + }, + { + "epoch": 0.14417643842274744, + "grad_norm": 0.32035139203071594, + "learning_rate": 9.514000801413122e-05, + "loss": 1.0844, + "step": 2373 + }, + { + "epoch": 0.14423719545537397, + "grad_norm": 0.3074304461479187, + "learning_rate": 9.513589031710691e-05, + "loss": 1.1608, + "step": 2374 + }, + { + "epoch": 0.14429795248800048, + "grad_norm": 0.3757053017616272, + "learning_rate": 9.513177096561781e-05, + "loss": 1.1114, + "step": 2375 + }, + { + "epoch": 0.14435870952062702, + "grad_norm": 0.26606789231300354, + "learning_rate": 9.512764995981493e-05, + "loss": 1.1339, + "step": 2376 + }, + { + "epoch": 0.14441946655325355, + "grad_norm": 0.361607164144516, + "learning_rate": 9.512352729984932e-05, + "loss": 1.1079, + "step": 2377 + }, + { + "epoch": 0.14448022358588006, + "grad_norm": 0.7943695187568665, + "learning_rate": 9.51194029858721e-05, + "loss": 1.1214, + "step": 2378 + }, + { + "epoch": 0.1445409806185066, + "grad_norm": 0.23717911541461945, + "learning_rate": 9.511527701803442e-05, + "loss": 1.0975, + "step": 2379 + }, + { + "epoch": 0.14460173765113313, + "grad_norm": 0.3201199769973755, + "learning_rate": 9.511114939648755e-05, + "loss": 1.2079, + "step": 2380 + }, + { + "epoch": 0.14466249468375963, + "grad_norm": 0.3122469484806061, + "learning_rate": 9.51070201213828e-05, + "loss": 1.0981, + "step": 2381 + }, + { + "epoch": 0.14472325171638617, + "grad_norm": 0.38328269124031067, + "learning_rate": 9.510288919287149e-05, + "loss": 1.195, + "step": 2382 + }, + { + "epoch": 0.1447840087490127, + "grad_norm": 0.49258527159690857, + "learning_rate": 9.509875661110506e-05, + "loss": 1.1233, + "step": 2383 + }, + { + "epoch": 0.1448447657816392, + "grad_norm": 0.22695033252239227, + "learning_rate": 9.509462237623499e-05, + "loss": 1.1418, + "step": 2384 + }, + { + "epoch": 0.14490552281426575, + "grad_norm": 0.22431959211826324, + "learning_rate": 9.509048648841281e-05, + "loss": 1.116, + "step": 2385 + }, + { + "epoch": 0.14496627984689228, + "grad_norm": 0.35765689611434937, + "learning_rate": 9.508634894779014e-05, + "loss": 1.0864, + "step": 2386 + }, + { + "epoch": 0.14502703687951882, + "grad_norm": 0.2634720504283905, + "learning_rate": 9.508220975451864e-05, + "loss": 1.1404, + "step": 2387 + }, + { + "epoch": 0.14508779391214532, + "grad_norm": 0.5153269171714783, + "learning_rate": 9.507806890875002e-05, + "loss": 1.1801, + "step": 2388 + }, + { + "epoch": 0.14514855094477186, + "grad_norm": 0.2498011589050293, + "learning_rate": 9.507392641063607e-05, + "loss": 1.1942, + "step": 2389 + }, + { + "epoch": 0.1452093079773984, + "grad_norm": 0.3493437170982361, + "learning_rate": 9.506978226032864e-05, + "loss": 1.1307, + "step": 2390 + }, + { + "epoch": 0.1452700650100249, + "grad_norm": 0.22119544446468353, + "learning_rate": 9.506563645797963e-05, + "loss": 1.1062, + "step": 2391 + }, + { + "epoch": 0.14533082204265144, + "grad_norm": 0.34596630930900574, + "learning_rate": 9.5061489003741e-05, + "loss": 1.141, + "step": 2392 + }, + { + "epoch": 0.14539157907527797, + "grad_norm": 0.24335211515426636, + "learning_rate": 9.505733989776478e-05, + "loss": 1.1113, + "step": 2393 + }, + { + "epoch": 0.14545233610790448, + "grad_norm": 0.31806132197380066, + "learning_rate": 9.505318914020307e-05, + "loss": 1.0612, + "step": 2394 + }, + { + "epoch": 0.145513093140531, + "grad_norm": 0.3140270709991455, + "learning_rate": 9.504903673120799e-05, + "loss": 1.1284, + "step": 2395 + }, + { + "epoch": 0.14557385017315755, + "grad_norm": 0.4527260661125183, + "learning_rate": 9.504488267093175e-05, + "loss": 1.0694, + "step": 2396 + }, + { + "epoch": 0.14563460720578406, + "grad_norm": 0.3877547085285187, + "learning_rate": 9.504072695952664e-05, + "loss": 1.1693, + "step": 2397 + }, + { + "epoch": 0.1456953642384106, + "grad_norm": 0.9170753359794617, + "learning_rate": 9.503656959714499e-05, + "loss": 1.191, + "step": 2398 + }, + { + "epoch": 0.14575612127103713, + "grad_norm": 0.6479384899139404, + "learning_rate": 9.503241058393917e-05, + "loss": 1.1045, + "step": 2399 + }, + { + "epoch": 0.14581687830366366, + "grad_norm": 0.5442968010902405, + "learning_rate": 9.502824992006163e-05, + "loss": 1.1509, + "step": 2400 + }, + { + "epoch": 0.14587763533629017, + "grad_norm": 11.330225944519043, + "learning_rate": 9.502408760566489e-05, + "loss": 1.4284, + "step": 2401 + }, + { + "epoch": 0.1459383923689167, + "grad_norm": 0.5091323852539062, + "learning_rate": 9.501992364090151e-05, + "loss": 1.1568, + "step": 2402 + }, + { + "epoch": 0.14599914940154324, + "grad_norm": 0.46708250045776367, + "learning_rate": 9.501575802592413e-05, + "loss": 1.1207, + "step": 2403 + }, + { + "epoch": 0.14605990643416975, + "grad_norm": 0.2684347331523895, + "learning_rate": 9.501159076088543e-05, + "loss": 1.1925, + "step": 2404 + }, + { + "epoch": 0.14612066346679628, + "grad_norm": 0.49111559987068176, + "learning_rate": 9.500742184593818e-05, + "loss": 1.1676, + "step": 2405 + }, + { + "epoch": 0.14618142049942282, + "grad_norm": 0.28784361481666565, + "learning_rate": 9.500325128123517e-05, + "loss": 1.1631, + "step": 2406 + }, + { + "epoch": 0.14624217753204932, + "grad_norm": 1.695853590965271, + "learning_rate": 9.49990790669293e-05, + "loss": 1.1794, + "step": 2407 + }, + { + "epoch": 0.14630293456467586, + "grad_norm": 0.368839293718338, + "learning_rate": 9.499490520317349e-05, + "loss": 1.1264, + "step": 2408 + }, + { + "epoch": 0.1463636915973024, + "grad_norm": 0.22555385529994965, + "learning_rate": 9.499072969012073e-05, + "loss": 1.0842, + "step": 2409 + }, + { + "epoch": 0.14642444862992893, + "grad_norm": 0.21837903559207916, + "learning_rate": 9.498655252792406e-05, + "loss": 1.0932, + "step": 2410 + }, + { + "epoch": 0.14648520566255543, + "grad_norm": 0.23410211503505707, + "learning_rate": 9.498237371673663e-05, + "loss": 1.0737, + "step": 2411 + }, + { + "epoch": 0.14654596269518197, + "grad_norm": 0.19407926499843597, + "learning_rate": 9.497819325671159e-05, + "loss": 1.0791, + "step": 2412 + }, + { + "epoch": 0.1466067197278085, + "grad_norm": 0.27732452750205994, + "learning_rate": 9.497401114800219e-05, + "loss": 1.2407, + "step": 2413 + }, + { + "epoch": 0.146667476760435, + "grad_norm": 0.23629742860794067, + "learning_rate": 9.49698273907617e-05, + "loss": 1.0569, + "step": 2414 + }, + { + "epoch": 0.14672823379306155, + "grad_norm": 0.3383291959762573, + "learning_rate": 9.496564198514351e-05, + "loss": 1.2741, + "step": 2415 + }, + { + "epoch": 0.14678899082568808, + "grad_norm": 0.4506233036518097, + "learning_rate": 9.496145493130102e-05, + "loss": 1.1006, + "step": 2416 + }, + { + "epoch": 0.1468497478583146, + "grad_norm": 0.9940173029899597, + "learning_rate": 9.49572662293877e-05, + "loss": 1.0559, + "step": 2417 + }, + { + "epoch": 0.14691050489094112, + "grad_norm": 0.3138459324836731, + "learning_rate": 9.495307587955711e-05, + "loss": 1.0839, + "step": 2418 + }, + { + "epoch": 0.14697126192356766, + "grad_norm": 0.1834966093301773, + "learning_rate": 9.494888388196282e-05, + "loss": 1.0888, + "step": 2419 + }, + { + "epoch": 0.14703201895619417, + "grad_norm": 0.47984644770622253, + "learning_rate": 9.494469023675851e-05, + "loss": 1.1485, + "step": 2420 + }, + { + "epoch": 0.1470927759888207, + "grad_norm": 0.18648222088813782, + "learning_rate": 9.49404949440979e-05, + "loss": 1.0789, + "step": 2421 + }, + { + "epoch": 0.14715353302144724, + "grad_norm": 2.830824136734009, + "learning_rate": 9.493629800413474e-05, + "loss": 1.3078, + "step": 2422 + }, + { + "epoch": 0.14721429005407377, + "grad_norm": 0.47586968541145325, + "learning_rate": 9.49320994170229e-05, + "loss": 1.2105, + "step": 2423 + }, + { + "epoch": 0.14727504708670028, + "grad_norm": 0.47054439783096313, + "learning_rate": 9.492789918291627e-05, + "loss": 1.1804, + "step": 2424 + }, + { + "epoch": 0.1473358041193268, + "grad_norm": 0.21035854518413544, + "learning_rate": 9.492369730196882e-05, + "loss": 1.0556, + "step": 2425 + }, + { + "epoch": 0.14739656115195335, + "grad_norm": 0.7703952789306641, + "learning_rate": 9.491949377433456e-05, + "loss": 1.1083, + "step": 2426 + }, + { + "epoch": 0.14745731818457986, + "grad_norm": 0.3565000295639038, + "learning_rate": 9.491528860016756e-05, + "loss": 1.1516, + "step": 2427 + }, + { + "epoch": 0.1475180752172064, + "grad_norm": 0.2610003650188446, + "learning_rate": 9.491108177962199e-05, + "loss": 1.1242, + "step": 2428 + }, + { + "epoch": 0.14757883224983293, + "grad_norm": 0.2539733946323395, + "learning_rate": 9.490687331285202e-05, + "loss": 1.0916, + "step": 2429 + }, + { + "epoch": 0.14763958928245943, + "grad_norm": 0.349924772977829, + "learning_rate": 9.490266320001195e-05, + "loss": 1.0691, + "step": 2430 + }, + { + "epoch": 0.14770034631508597, + "grad_norm": 1.8721426725387573, + "learning_rate": 9.489845144125607e-05, + "loss": 1.1172, + "step": 2431 + }, + { + "epoch": 0.1477611033477125, + "grad_norm": 0.349567711353302, + "learning_rate": 9.489423803673877e-05, + "loss": 1.0891, + "step": 2432 + }, + { + "epoch": 0.14782186038033904, + "grad_norm": 0.26168742775917053, + "learning_rate": 9.489002298661449e-05, + "loss": 1.18, + "step": 2433 + }, + { + "epoch": 0.14788261741296554, + "grad_norm": 0.5303388833999634, + "learning_rate": 9.488580629103776e-05, + "loss": 1.0796, + "step": 2434 + }, + { + "epoch": 0.14794337444559208, + "grad_norm": 0.2392454892396927, + "learning_rate": 9.488158795016311e-05, + "loss": 1.102, + "step": 2435 + }, + { + "epoch": 0.14800413147821861, + "grad_norm": 0.25127843022346497, + "learning_rate": 9.487736796414519e-05, + "loss": 1.1343, + "step": 2436 + }, + { + "epoch": 0.14806488851084512, + "grad_norm": 0.50868821144104, + "learning_rate": 9.487314633313867e-05, + "loss": 1.2151, + "step": 2437 + }, + { + "epoch": 0.14812564554347166, + "grad_norm": 0.2242830991744995, + "learning_rate": 9.486892305729828e-05, + "loss": 1.1292, + "step": 2438 + }, + { + "epoch": 0.1481864025760982, + "grad_norm": 0.3668231666088104, + "learning_rate": 9.486469813677885e-05, + "loss": 1.0819, + "step": 2439 + }, + { + "epoch": 0.1482471596087247, + "grad_norm": 0.2902078330516815, + "learning_rate": 9.486047157173526e-05, + "loss": 1.0536, + "step": 2440 + }, + { + "epoch": 0.14830791664135123, + "grad_norm": 0.4646584093570709, + "learning_rate": 9.485624336232239e-05, + "loss": 1.1238, + "step": 2441 + }, + { + "epoch": 0.14836867367397777, + "grad_norm": 0.2634126842021942, + "learning_rate": 9.485201350869525e-05, + "loss": 1.1286, + "step": 2442 + }, + { + "epoch": 0.14842943070660428, + "grad_norm": 0.3723178803920746, + "learning_rate": 9.48477820110089e-05, + "loss": 1.1674, + "step": 2443 + }, + { + "epoch": 0.1484901877392308, + "grad_norm": 0.21849285066127777, + "learning_rate": 9.484354886941842e-05, + "loss": 1.1018, + "step": 2444 + }, + { + "epoch": 0.14855094477185735, + "grad_norm": 0.2241188883781433, + "learning_rate": 9.483931408407898e-05, + "loss": 1.1321, + "step": 2445 + }, + { + "epoch": 0.14861170180448388, + "grad_norm": 0.3175460994243622, + "learning_rate": 9.483507765514583e-05, + "loss": 1.2208, + "step": 2446 + }, + { + "epoch": 0.1486724588371104, + "grad_norm": 0.4757917523384094, + "learning_rate": 9.483083958277423e-05, + "loss": 1.2422, + "step": 2447 + }, + { + "epoch": 0.14873321586973692, + "grad_norm": 0.3011987507343292, + "learning_rate": 9.482659986711955e-05, + "loss": 1.1363, + "step": 2448 + }, + { + "epoch": 0.14879397290236346, + "grad_norm": 0.9036540389060974, + "learning_rate": 9.482235850833719e-05, + "loss": 1.1045, + "step": 2449 + }, + { + "epoch": 0.14885472993498997, + "grad_norm": 0.3020455837249756, + "learning_rate": 9.48181155065826e-05, + "loss": 1.1268, + "step": 2450 + }, + { + "epoch": 0.1489154869676165, + "grad_norm": 1.1768310070037842, + "learning_rate": 9.481387086201132e-05, + "loss": 1.0964, + "step": 2451 + }, + { + "epoch": 0.14897624400024304, + "grad_norm": 0.2857222259044647, + "learning_rate": 9.480962457477896e-05, + "loss": 1.0848, + "step": 2452 + }, + { + "epoch": 0.14903700103286954, + "grad_norm": 1.7870455980300903, + "learning_rate": 9.480537664504115e-05, + "loss": 1.1352, + "step": 2453 + }, + { + "epoch": 0.14909775806549608, + "grad_norm": 0.5158678293228149, + "learning_rate": 9.480112707295359e-05, + "loss": 1.16, + "step": 2454 + }, + { + "epoch": 0.1491585150981226, + "grad_norm": 0.24269725382328033, + "learning_rate": 9.479687585867205e-05, + "loss": 1.1532, + "step": 2455 + }, + { + "epoch": 0.14921927213074915, + "grad_norm": 0.4435185194015503, + "learning_rate": 9.479262300235238e-05, + "loss": 1.0816, + "step": 2456 + }, + { + "epoch": 0.14928002916337565, + "grad_norm": 0.7389569878578186, + "learning_rate": 9.478836850415045e-05, + "loss": 1.2218, + "step": 2457 + }, + { + "epoch": 0.1493407861960022, + "grad_norm": 0.46674802899360657, + "learning_rate": 9.478411236422221e-05, + "loss": 1.0901, + "step": 2458 + }, + { + "epoch": 0.14940154322862872, + "grad_norm": 0.31880855560302734, + "learning_rate": 9.477985458272369e-05, + "loss": 1.209, + "step": 2459 + }, + { + "epoch": 0.14946230026125523, + "grad_norm": 0.5213747620582581, + "learning_rate": 9.477559515981092e-05, + "loss": 1.204, + "step": 2460 + }, + { + "epoch": 0.14952305729388177, + "grad_norm": 0.21345965564250946, + "learning_rate": 9.477133409564007e-05, + "loss": 1.2422, + "step": 2461 + }, + { + "epoch": 0.1495838143265083, + "grad_norm": 0.2577822208404541, + "learning_rate": 9.476707139036731e-05, + "loss": 1.127, + "step": 2462 + }, + { + "epoch": 0.1496445713591348, + "grad_norm": 0.17826706171035767, + "learning_rate": 9.47628070441489e-05, + "loss": 1.0603, + "step": 2463 + }, + { + "epoch": 0.14970532839176134, + "grad_norm": 0.36057189106941223, + "learning_rate": 9.475854105714115e-05, + "loss": 1.116, + "step": 2464 + }, + { + "epoch": 0.14976608542438788, + "grad_norm": 0.23256990313529968, + "learning_rate": 9.475427342950042e-05, + "loss": 1.1584, + "step": 2465 + }, + { + "epoch": 0.1498268424570144, + "grad_norm": 0.37565556168556213, + "learning_rate": 9.475000416138314e-05, + "loss": 1.0853, + "step": 2466 + }, + { + "epoch": 0.14988759948964092, + "grad_norm": 0.2615987956523895, + "learning_rate": 9.474573325294582e-05, + "loss": 1.0886, + "step": 2467 + }, + { + "epoch": 0.14994835652226746, + "grad_norm": 0.3069855868816376, + "learning_rate": 9.4741460704345e-05, + "loss": 1.1777, + "step": 2468 + }, + { + "epoch": 0.150009113554894, + "grad_norm": 0.4225909411907196, + "learning_rate": 9.473718651573726e-05, + "loss": 1.1548, + "step": 2469 + }, + { + "epoch": 0.1500698705875205, + "grad_norm": 0.28703489899635315, + "learning_rate": 9.473291068727933e-05, + "loss": 1.1102, + "step": 2470 + }, + { + "epoch": 0.15013062762014703, + "grad_norm": 0.5458278059959412, + "learning_rate": 9.472863321912791e-05, + "loss": 1.2952, + "step": 2471 + }, + { + "epoch": 0.15019138465277357, + "grad_norm": 1.0638123750686646, + "learning_rate": 9.472435411143978e-05, + "loss": 1.1278, + "step": 2472 + }, + { + "epoch": 0.15025214168540008, + "grad_norm": 0.5020226240158081, + "learning_rate": 9.47200733643718e-05, + "loss": 1.2574, + "step": 2473 + }, + { + "epoch": 0.1503128987180266, + "grad_norm": 0.8277014493942261, + "learning_rate": 9.471579097808092e-05, + "loss": 1.1589, + "step": 2474 + }, + { + "epoch": 0.15037365575065315, + "grad_norm": 0.20234432816505432, + "learning_rate": 9.471150695272406e-05, + "loss": 1.1147, + "step": 2475 + }, + { + "epoch": 0.15043441278327965, + "grad_norm": 0.4093621075153351, + "learning_rate": 9.470722128845826e-05, + "loss": 1.0679, + "step": 2476 + }, + { + "epoch": 0.1504951698159062, + "grad_norm": 1.0158871412277222, + "learning_rate": 9.470293398544064e-05, + "loss": 1.2335, + "step": 2477 + }, + { + "epoch": 0.15055592684853272, + "grad_norm": 0.30244317650794983, + "learning_rate": 9.469864504382832e-05, + "loss": 1.125, + "step": 2478 + }, + { + "epoch": 0.15061668388115926, + "grad_norm": 0.3834805190563202, + "learning_rate": 9.469435446377852e-05, + "loss": 1.131, + "step": 2479 + }, + { + "epoch": 0.15067744091378577, + "grad_norm": 0.565513014793396, + "learning_rate": 9.469006224544854e-05, + "loss": 1.1359, + "step": 2480 + }, + { + "epoch": 0.1507381979464123, + "grad_norm": 2.740659713745117, + "learning_rate": 9.468576838899567e-05, + "loss": 1.0788, + "step": 2481 + }, + { + "epoch": 0.15079895497903884, + "grad_norm": 0.31828320026397705, + "learning_rate": 9.468147289457732e-05, + "loss": 1.1914, + "step": 2482 + }, + { + "epoch": 0.15085971201166534, + "grad_norm": 0.31970521807670593, + "learning_rate": 9.467717576235095e-05, + "loss": 1.2543, + "step": 2483 + }, + { + "epoch": 0.15092046904429188, + "grad_norm": 0.3736356198787689, + "learning_rate": 9.467287699247409e-05, + "loss": 1.2629, + "step": 2484 + }, + { + "epoch": 0.1509812260769184, + "grad_norm": 0.2861364185810089, + "learning_rate": 9.466857658510427e-05, + "loss": 1.1576, + "step": 2485 + }, + { + "epoch": 0.15104198310954492, + "grad_norm": 0.48578718304634094, + "learning_rate": 9.466427454039914e-05, + "loss": 1.0846, + "step": 2486 + }, + { + "epoch": 0.15110274014217145, + "grad_norm": 0.28584396839141846, + "learning_rate": 9.465997085851639e-05, + "loss": 1.15, + "step": 2487 + }, + { + "epoch": 0.151163497174798, + "grad_norm": 0.7971346378326416, + "learning_rate": 9.465566553961378e-05, + "loss": 1.1781, + "step": 2488 + }, + { + "epoch": 0.1512242542074245, + "grad_norm": 0.25013306736946106, + "learning_rate": 9.465135858384911e-05, + "loss": 1.0797, + "step": 2489 + }, + { + "epoch": 0.15128501124005103, + "grad_norm": 0.24690698087215424, + "learning_rate": 9.464704999138027e-05, + "loss": 1.2386, + "step": 2490 + }, + { + "epoch": 0.15134576827267757, + "grad_norm": 0.29810598492622375, + "learning_rate": 9.464273976236517e-05, + "loss": 1.1106, + "step": 2491 + }, + { + "epoch": 0.1514065253053041, + "grad_norm": 0.2450282871723175, + "learning_rate": 9.463842789696184e-05, + "loss": 1.082, + "step": 2492 + }, + { + "epoch": 0.1514672823379306, + "grad_norm": 0.24212051928043365, + "learning_rate": 9.463411439532828e-05, + "loss": 1.1312, + "step": 2493 + }, + { + "epoch": 0.15152803937055714, + "grad_norm": 0.21999149024486542, + "learning_rate": 9.462979925762266e-05, + "loss": 1.215, + "step": 2494 + }, + { + "epoch": 0.15158879640318368, + "grad_norm": 0.2244645208120346, + "learning_rate": 9.462548248400309e-05, + "loss": 1.1624, + "step": 2495 + }, + { + "epoch": 0.15164955343581019, + "grad_norm": 0.2349688708782196, + "learning_rate": 9.462116407462784e-05, + "loss": 1.1165, + "step": 2496 + }, + { + "epoch": 0.15171031046843672, + "grad_norm": 0.2561975121498108, + "learning_rate": 9.46168440296552e-05, + "loss": 1.1101, + "step": 2497 + }, + { + "epoch": 0.15177106750106326, + "grad_norm": 0.2725282609462738, + "learning_rate": 9.461252234924351e-05, + "loss": 1.1595, + "step": 2498 + }, + { + "epoch": 0.15183182453368976, + "grad_norm": 0.33647653460502625, + "learning_rate": 9.460819903355121e-05, + "loss": 1.3346, + "step": 2499 + }, + { + "epoch": 0.1518925815663163, + "grad_norm": 0.2569330334663391, + "learning_rate": 9.460387408273675e-05, + "loss": 1.1704, + "step": 2500 + }, + { + "epoch": 0.15195333859894283, + "grad_norm": 0.6744318604469299, + "learning_rate": 9.459954749695865e-05, + "loss": 1.1305, + "step": 2501 + }, + { + "epoch": 0.15201409563156934, + "grad_norm": 0.5805237293243408, + "learning_rate": 9.459521927637551e-05, + "loss": 1.095, + "step": 2502 + }, + { + "epoch": 0.15207485266419588, + "grad_norm": 1.8826560974121094, + "learning_rate": 9.4590889421146e-05, + "loss": 1.1563, + "step": 2503 + }, + { + "epoch": 0.1521356096968224, + "grad_norm": 0.5640044808387756, + "learning_rate": 9.458655793142883e-05, + "loss": 1.1851, + "step": 2504 + }, + { + "epoch": 0.15219636672944895, + "grad_norm": 0.5330599546432495, + "learning_rate": 9.458222480738275e-05, + "loss": 1.1553, + "step": 2505 + }, + { + "epoch": 0.15225712376207545, + "grad_norm": 0.539725661277771, + "learning_rate": 9.45778900491666e-05, + "loss": 1.2014, + "step": 2506 + }, + { + "epoch": 0.152317880794702, + "grad_norm": 0.5315496325492859, + "learning_rate": 9.457355365693927e-05, + "loss": 1.1909, + "step": 2507 + }, + { + "epoch": 0.15237863782732852, + "grad_norm": 0.23454418778419495, + "learning_rate": 9.456921563085972e-05, + "loss": 1.1186, + "step": 2508 + }, + { + "epoch": 0.15243939485995503, + "grad_norm": 0.7672280073165894, + "learning_rate": 9.456487597108694e-05, + "loss": 1.0755, + "step": 2509 + }, + { + "epoch": 0.15250015189258156, + "grad_norm": 0.25444382429122925, + "learning_rate": 9.456053467778003e-05, + "loss": 1.0688, + "step": 2510 + }, + { + "epoch": 0.1525609089252081, + "grad_norm": 0.7714465856552124, + "learning_rate": 9.455619175109811e-05, + "loss": 1.1571, + "step": 2511 + }, + { + "epoch": 0.1526216659578346, + "grad_norm": 0.21766959130764008, + "learning_rate": 9.455184719120038e-05, + "loss": 1.1275, + "step": 2512 + }, + { + "epoch": 0.15268242299046114, + "grad_norm": 0.9938517212867737, + "learning_rate": 9.454750099824606e-05, + "loss": 1.0999, + "step": 2513 + }, + { + "epoch": 0.15274318002308768, + "grad_norm": 0.29452961683273315, + "learning_rate": 9.454315317239448e-05, + "loss": 1.1052, + "step": 2514 + }, + { + "epoch": 0.1528039370557142, + "grad_norm": 1.9244537353515625, + "learning_rate": 9.453880371380503e-05, + "loss": 1.1167, + "step": 2515 + }, + { + "epoch": 0.15286469408834072, + "grad_norm": 0.7682761549949646, + "learning_rate": 9.453445262263711e-05, + "loss": 1.1112, + "step": 2516 + }, + { + "epoch": 0.15292545112096725, + "grad_norm": 0.2688964307308197, + "learning_rate": 9.453009989905022e-05, + "loss": 1.1217, + "step": 2517 + }, + { + "epoch": 0.1529862081535938, + "grad_norm": 0.8672596216201782, + "learning_rate": 9.452574554320391e-05, + "loss": 1.1878, + "step": 2518 + }, + { + "epoch": 0.1530469651862203, + "grad_norm": 0.3098766505718231, + "learning_rate": 9.452138955525779e-05, + "loss": 1.1722, + "step": 2519 + }, + { + "epoch": 0.15310772221884683, + "grad_norm": 0.7696993350982666, + "learning_rate": 9.451703193537153e-05, + "loss": 1.1948, + "step": 2520 + }, + { + "epoch": 0.15316847925147337, + "grad_norm": 0.23606593906879425, + "learning_rate": 9.451267268370486e-05, + "loss": 1.1278, + "step": 2521 + }, + { + "epoch": 0.15322923628409987, + "grad_norm": 0.6184613704681396, + "learning_rate": 9.450831180041757e-05, + "loss": 1.1712, + "step": 2522 + }, + { + "epoch": 0.1532899933167264, + "grad_norm": 0.28219518065452576, + "learning_rate": 9.450394928566951e-05, + "loss": 1.0765, + "step": 2523 + }, + { + "epoch": 0.15335075034935294, + "grad_norm": 0.23568551242351532, + "learning_rate": 9.449958513962059e-05, + "loss": 1.0952, + "step": 2524 + }, + { + "epoch": 0.15341150738197945, + "grad_norm": 0.9417229890823364, + "learning_rate": 9.449521936243075e-05, + "loss": 1.2211, + "step": 2525 + }, + { + "epoch": 0.15347226441460599, + "grad_norm": 0.2840830087661743, + "learning_rate": 9.449085195426007e-05, + "loss": 1.1589, + "step": 2526 + }, + { + "epoch": 0.15353302144723252, + "grad_norm": 0.4772968292236328, + "learning_rate": 9.448648291526861e-05, + "loss": 1.1096, + "step": 2527 + }, + { + "epoch": 0.15359377847985906, + "grad_norm": 2.2898738384246826, + "learning_rate": 9.448211224561651e-05, + "loss": 1.0835, + "step": 2528 + }, + { + "epoch": 0.15365453551248556, + "grad_norm": 0.3046601116657257, + "learning_rate": 9.447773994546399e-05, + "loss": 1.1139, + "step": 2529 + }, + { + "epoch": 0.1537152925451121, + "grad_norm": 0.6479328274726868, + "learning_rate": 9.44733660149713e-05, + "loss": 1.2734, + "step": 2530 + }, + { + "epoch": 0.15377604957773863, + "grad_norm": 0.9155849814414978, + "learning_rate": 9.446899045429881e-05, + "loss": 1.2907, + "step": 2531 + }, + { + "epoch": 0.15383680661036514, + "grad_norm": 0.29011771082878113, + "learning_rate": 9.446461326360687e-05, + "loss": 1.1588, + "step": 2532 + }, + { + "epoch": 0.15389756364299167, + "grad_norm": 3.068434000015259, + "learning_rate": 9.446023444305592e-05, + "loss": 1.3313, + "step": 2533 + }, + { + "epoch": 0.1539583206756182, + "grad_norm": 0.6821197271347046, + "learning_rate": 9.44558539928065e-05, + "loss": 1.0951, + "step": 2534 + }, + { + "epoch": 0.15401907770824472, + "grad_norm": 0.34131988883018494, + "learning_rate": 9.445147191301917e-05, + "loss": 1.1472, + "step": 2535 + }, + { + "epoch": 0.15407983474087125, + "grad_norm": 0.6731616854667664, + "learning_rate": 9.444708820385451e-05, + "loss": 1.1414, + "step": 2536 + }, + { + "epoch": 0.1541405917734978, + "grad_norm": 0.42329713702201843, + "learning_rate": 9.444270286547328e-05, + "loss": 1.0887, + "step": 2537 + }, + { + "epoch": 0.15420134880612432, + "grad_norm": 0.6714568138122559, + "learning_rate": 9.443831589803616e-05, + "loss": 1.2241, + "step": 2538 + }, + { + "epoch": 0.15426210583875083, + "grad_norm": 0.7552052736282349, + "learning_rate": 9.443392730170398e-05, + "loss": 1.1352, + "step": 2539 + }, + { + "epoch": 0.15432286287137736, + "grad_norm": 0.5308777689933777, + "learning_rate": 9.442953707663763e-05, + "loss": 1.2078, + "step": 2540 + }, + { + "epoch": 0.1543836199040039, + "grad_norm": 1.0084806680679321, + "learning_rate": 9.442514522299797e-05, + "loss": 1.1006, + "step": 2541 + }, + { + "epoch": 0.1544443769366304, + "grad_norm": 0.2573261857032776, + "learning_rate": 9.442075174094606e-05, + "loss": 1.086, + "step": 2542 + }, + { + "epoch": 0.15450513396925694, + "grad_norm": 1.5209696292877197, + "learning_rate": 9.441635663064289e-05, + "loss": 1.0337, + "step": 2543 + }, + { + "epoch": 0.15456589100188348, + "grad_norm": 0.4975840449333191, + "learning_rate": 9.441195989224958e-05, + "loss": 1.1145, + "step": 2544 + }, + { + "epoch": 0.15462664803450998, + "grad_norm": 0.4960063099861145, + "learning_rate": 9.440756152592731e-05, + "loss": 1.1478, + "step": 2545 + }, + { + "epoch": 0.15468740506713652, + "grad_norm": 2.0245907306671143, + "learning_rate": 9.440316153183729e-05, + "loss": 1.2396, + "step": 2546 + }, + { + "epoch": 0.15474816209976305, + "grad_norm": 0.7554837465286255, + "learning_rate": 9.43987599101408e-05, + "loss": 1.3512, + "step": 2547 + }, + { + "epoch": 0.15480891913238956, + "grad_norm": 0.688527524471283, + "learning_rate": 9.439435666099916e-05, + "loss": 1.0993, + "step": 2548 + }, + { + "epoch": 0.1548696761650161, + "grad_norm": 0.5203417539596558, + "learning_rate": 9.438995178457381e-05, + "loss": 1.2889, + "step": 2549 + }, + { + "epoch": 0.15493043319764263, + "grad_norm": 0.5554254055023193, + "learning_rate": 9.438554528102621e-05, + "loss": 1.1265, + "step": 2550 + }, + { + "epoch": 0.15499119023026917, + "grad_norm": 0.28404244780540466, + "learning_rate": 9.438113715051785e-05, + "loss": 1.1438, + "step": 2551 + }, + { + "epoch": 0.15505194726289567, + "grad_norm": 0.7331209182739258, + "learning_rate": 9.437672739321034e-05, + "loss": 1.084, + "step": 2552 + }, + { + "epoch": 0.1551127042955222, + "grad_norm": 0.21935121715068817, + "learning_rate": 9.43723160092653e-05, + "loss": 1.0835, + "step": 2553 + }, + { + "epoch": 0.15517346132814874, + "grad_norm": 0.6086776852607727, + "learning_rate": 9.436790299884445e-05, + "loss": 1.1855, + "step": 2554 + }, + { + "epoch": 0.15523421836077525, + "grad_norm": 0.42407557368278503, + "learning_rate": 9.436348836210953e-05, + "loss": 1.1438, + "step": 2555 + }, + { + "epoch": 0.15529497539340179, + "grad_norm": 0.3682560324668884, + "learning_rate": 9.435907209922238e-05, + "loss": 1.0572, + "step": 2556 + }, + { + "epoch": 0.15535573242602832, + "grad_norm": 0.43461874127388, + "learning_rate": 9.435465421034486e-05, + "loss": 1.5366, + "step": 2557 + }, + { + "epoch": 0.15541648945865483, + "grad_norm": 0.31885766983032227, + "learning_rate": 9.435023469563892e-05, + "loss": 1.1862, + "step": 2558 + }, + { + "epoch": 0.15547724649128136, + "grad_norm": 0.4323097765445709, + "learning_rate": 9.434581355526655e-05, + "loss": 1.1444, + "step": 2559 + }, + { + "epoch": 0.1555380035239079, + "grad_norm": 0.27753692865371704, + "learning_rate": 9.434139078938982e-05, + "loss": 1.208, + "step": 2560 + }, + { + "epoch": 0.15559876055653443, + "grad_norm": 0.5659012198448181, + "learning_rate": 9.433696639817082e-05, + "loss": 1.1136, + "step": 2561 + }, + { + "epoch": 0.15565951758916094, + "grad_norm": 0.3612508177757263, + "learning_rate": 9.433254038177178e-05, + "loss": 1.1779, + "step": 2562 + }, + { + "epoch": 0.15572027462178747, + "grad_norm": 0.32047387957572937, + "learning_rate": 9.432811274035486e-05, + "loss": 1.1525, + "step": 2563 + }, + { + "epoch": 0.155781031654414, + "grad_norm": 0.40015602111816406, + "learning_rate": 9.432368347408244e-05, + "loss": 1.0643, + "step": 2564 + }, + { + "epoch": 0.15584178868704052, + "grad_norm": 0.24259313941001892, + "learning_rate": 9.431925258311682e-05, + "loss": 1.1737, + "step": 2565 + }, + { + "epoch": 0.15590254571966705, + "grad_norm": 0.39857929944992065, + "learning_rate": 9.431482006762042e-05, + "loss": 1.1258, + "step": 2566 + }, + { + "epoch": 0.1559633027522936, + "grad_norm": 0.2899479568004608, + "learning_rate": 9.431038592775575e-05, + "loss": 1.1274, + "step": 2567 + }, + { + "epoch": 0.1560240597849201, + "grad_norm": 0.35384660959243774, + "learning_rate": 9.430595016368528e-05, + "loss": 1.2616, + "step": 2568 + }, + { + "epoch": 0.15608481681754663, + "grad_norm": 0.49837592244148254, + "learning_rate": 9.430151277557167e-05, + "loss": 1.1754, + "step": 2569 + }, + { + "epoch": 0.15614557385017316, + "grad_norm": 0.28873783349990845, + "learning_rate": 9.429707376357753e-05, + "loss": 1.1807, + "step": 2570 + }, + { + "epoch": 0.15620633088279967, + "grad_norm": 0.3139941692352295, + "learning_rate": 9.42926331278656e-05, + "loss": 1.1041, + "step": 2571 + }, + { + "epoch": 0.1562670879154262, + "grad_norm": 0.38745471835136414, + "learning_rate": 9.428819086859865e-05, + "loss": 1.1554, + "step": 2572 + }, + { + "epoch": 0.15632784494805274, + "grad_norm": 0.34701013565063477, + "learning_rate": 9.428374698593949e-05, + "loss": 1.2448, + "step": 2573 + }, + { + "epoch": 0.15638860198067928, + "grad_norm": 0.25233176350593567, + "learning_rate": 9.427930148005102e-05, + "loss": 1.1831, + "step": 2574 + }, + { + "epoch": 0.15644935901330578, + "grad_norm": 0.2578613758087158, + "learning_rate": 9.42748543510962e-05, + "loss": 1.1221, + "step": 2575 + }, + { + "epoch": 0.15651011604593232, + "grad_norm": 0.28400206565856934, + "learning_rate": 9.427040559923804e-05, + "loss": 1.0966, + "step": 2576 + }, + { + "epoch": 0.15657087307855885, + "grad_norm": 2.314439058303833, + "learning_rate": 9.426595522463961e-05, + "loss": 1.0723, + "step": 2577 + }, + { + "epoch": 0.15663163011118536, + "grad_norm": 0.30299174785614014, + "learning_rate": 9.426150322746404e-05, + "loss": 1.0931, + "step": 2578 + }, + { + "epoch": 0.1566923871438119, + "grad_norm": 0.20927217602729797, + "learning_rate": 9.425704960787449e-05, + "loss": 1.0832, + "step": 2579 + }, + { + "epoch": 0.15675314417643843, + "grad_norm": 0.30179864168167114, + "learning_rate": 9.425259436603425e-05, + "loss": 1.1089, + "step": 2580 + }, + { + "epoch": 0.15681390120906494, + "grad_norm": 0.47571635246276855, + "learning_rate": 9.42481375021066e-05, + "loss": 1.1589, + "step": 2581 + }, + { + "epoch": 0.15687465824169147, + "grad_norm": 0.24007441103458405, + "learning_rate": 9.424367901625491e-05, + "loss": 1.101, + "step": 2582 + }, + { + "epoch": 0.156935415274318, + "grad_norm": 5.019867897033691, + "learning_rate": 9.423921890864263e-05, + "loss": 1.1422, + "step": 2583 + }, + { + "epoch": 0.15699617230694454, + "grad_norm": 0.2471000999212265, + "learning_rate": 9.423475717943321e-05, + "loss": 1.1537, + "step": 2584 + }, + { + "epoch": 0.15705692933957105, + "grad_norm": 0.2308974415063858, + "learning_rate": 9.423029382879023e-05, + "loss": 1.1125, + "step": 2585 + }, + { + "epoch": 0.15711768637219758, + "grad_norm": 0.25561994314193726, + "learning_rate": 9.422582885687728e-05, + "loss": 1.1768, + "step": 2586 + }, + { + "epoch": 0.15717844340482412, + "grad_norm": 1.740403413772583, + "learning_rate": 9.422136226385802e-05, + "loss": 1.1715, + "step": 2587 + }, + { + "epoch": 0.15723920043745063, + "grad_norm": 0.47867506742477417, + "learning_rate": 9.42168940498962e-05, + "loss": 1.3043, + "step": 2588 + }, + { + "epoch": 0.15729995747007716, + "grad_norm": 0.6371132731437683, + "learning_rate": 9.421242421515554e-05, + "loss": 1.0636, + "step": 2589 + }, + { + "epoch": 0.1573607145027037, + "grad_norm": 0.1899743676185608, + "learning_rate": 9.420795275979997e-05, + "loss": 1.1028, + "step": 2590 + }, + { + "epoch": 0.1574214715353302, + "grad_norm": 0.3206506669521332, + "learning_rate": 9.420347968399332e-05, + "loss": 1.0837, + "step": 2591 + }, + { + "epoch": 0.15748222856795674, + "grad_norm": 0.1950460821390152, + "learning_rate": 9.419900498789958e-05, + "loss": 1.1047, + "step": 2592 + }, + { + "epoch": 0.15754298560058327, + "grad_norm": 0.19482262432575226, + "learning_rate": 9.419452867168277e-05, + "loss": 1.1235, + "step": 2593 + }, + { + "epoch": 0.15760374263320978, + "grad_norm": 0.14973847568035126, + "learning_rate": 9.419005073550697e-05, + "loss": 1.2643, + "step": 2594 + }, + { + "epoch": 0.15766449966583632, + "grad_norm": 0.1692671924829483, + "learning_rate": 9.418557117953632e-05, + "loss": 1.0649, + "step": 2595 + }, + { + "epoch": 0.15772525669846285, + "grad_norm": 0.1721692681312561, + "learning_rate": 9.418109000393501e-05, + "loss": 1.0333, + "step": 2596 + }, + { + "epoch": 0.1577860137310894, + "grad_norm": 0.2623113989830017, + "learning_rate": 9.417660720886732e-05, + "loss": 1.1939, + "step": 2597 + }, + { + "epoch": 0.1578467707637159, + "grad_norm": 0.2686692476272583, + "learning_rate": 9.417212279449754e-05, + "loss": 1.0594, + "step": 2598 + }, + { + "epoch": 0.15790752779634243, + "grad_norm": 0.26072388887405396, + "learning_rate": 9.416763676099007e-05, + "loss": 1.2215, + "step": 2599 + }, + { + "epoch": 0.15796828482896896, + "grad_norm": 0.3139735758304596, + "learning_rate": 9.416314910850931e-05, + "loss": 1.1881, + "step": 2600 + }, + { + "epoch": 0.15802904186159547, + "grad_norm": 0.23548336327075958, + "learning_rate": 9.415865983721982e-05, + "loss": 1.1003, + "step": 2601 + }, + { + "epoch": 0.158089798894222, + "grad_norm": 0.20403902232646942, + "learning_rate": 9.41541689472861e-05, + "loss": 1.1643, + "step": 2602 + }, + { + "epoch": 0.15815055592684854, + "grad_norm": 0.2097340226173401, + "learning_rate": 9.414967643887279e-05, + "loss": 1.1321, + "step": 2603 + }, + { + "epoch": 0.15821131295947505, + "grad_norm": 0.21385569870471954, + "learning_rate": 9.414518231214455e-05, + "loss": 1.1319, + "step": 2604 + }, + { + "epoch": 0.15827206999210158, + "grad_norm": 0.3653961718082428, + "learning_rate": 9.414068656726612e-05, + "loss": 1.1058, + "step": 2605 + }, + { + "epoch": 0.15833282702472812, + "grad_norm": 0.26645660400390625, + "learning_rate": 9.41361892044023e-05, + "loss": 1.124, + "step": 2606 + }, + { + "epoch": 0.15839358405735462, + "grad_norm": 0.2077946811914444, + "learning_rate": 9.413169022371793e-05, + "loss": 1.043, + "step": 2607 + }, + { + "epoch": 0.15845434108998116, + "grad_norm": 0.2906290888786316, + "learning_rate": 9.412718962537792e-05, + "loss": 1.0577, + "step": 2608 + }, + { + "epoch": 0.1585150981226077, + "grad_norm": 0.2705436646938324, + "learning_rate": 9.412268740954724e-05, + "loss": 1.2098, + "step": 2609 + }, + { + "epoch": 0.15857585515523423, + "grad_norm": 0.2686663866043091, + "learning_rate": 9.411818357639093e-05, + "loss": 1.1962, + "step": 2610 + }, + { + "epoch": 0.15863661218786074, + "grad_norm": 0.9315751791000366, + "learning_rate": 9.411367812607408e-05, + "loss": 1.1997, + "step": 2611 + }, + { + "epoch": 0.15869736922048727, + "grad_norm": 0.5988966822624207, + "learning_rate": 9.410917105876183e-05, + "loss": 1.2765, + "step": 2612 + }, + { + "epoch": 0.1587581262531138, + "grad_norm": 0.3456157445907593, + "learning_rate": 9.410466237461937e-05, + "loss": 1.1121, + "step": 2613 + }, + { + "epoch": 0.15881888328574031, + "grad_norm": 0.27668699622154236, + "learning_rate": 9.4100152073812e-05, + "loss": 1.1761, + "step": 2614 + }, + { + "epoch": 0.15887964031836685, + "grad_norm": 0.27273792028427124, + "learning_rate": 9.409564015650502e-05, + "loss": 1.1861, + "step": 2615 + }, + { + "epoch": 0.15894039735099338, + "grad_norm": 1.874498963356018, + "learning_rate": 9.409112662286385e-05, + "loss": 1.2041, + "step": 2616 + }, + { + "epoch": 0.1590011543836199, + "grad_norm": 0.3487119674682617, + "learning_rate": 9.408661147305389e-05, + "loss": 1.188, + "step": 2617 + }, + { + "epoch": 0.15906191141624643, + "grad_norm": 0.22690211236476898, + "learning_rate": 9.408209470724068e-05, + "loss": 1.0749, + "step": 2618 + }, + { + "epoch": 0.15912266844887296, + "grad_norm": 0.32784268260002136, + "learning_rate": 9.407757632558975e-05, + "loss": 1.2013, + "step": 2619 + }, + { + "epoch": 0.1591834254814995, + "grad_norm": 0.33431586623191833, + "learning_rate": 9.407305632826675e-05, + "loss": 1.2275, + "step": 2620 + }, + { + "epoch": 0.159244182514126, + "grad_norm": 0.18178267776966095, + "learning_rate": 9.406853471543735e-05, + "loss": 1.0661, + "step": 2621 + }, + { + "epoch": 0.15930493954675254, + "grad_norm": 0.567473292350769, + "learning_rate": 9.40640114872673e-05, + "loss": 1.0682, + "step": 2622 + }, + { + "epoch": 0.15936569657937907, + "grad_norm": 0.23264090716838837, + "learning_rate": 9.405948664392239e-05, + "loss": 1.0818, + "step": 2623 + }, + { + "epoch": 0.15942645361200558, + "grad_norm": 0.38584303855895996, + "learning_rate": 9.405496018556847e-05, + "loss": 1.0729, + "step": 2624 + }, + { + "epoch": 0.15948721064463212, + "grad_norm": 0.3485680818557739, + "learning_rate": 9.405043211237147e-05, + "loss": 1.1031, + "step": 2625 + }, + { + "epoch": 0.15954796767725865, + "grad_norm": 0.44877228140830994, + "learning_rate": 9.404590242449738e-05, + "loss": 1.2355, + "step": 2626 + }, + { + "epoch": 0.15960872470988516, + "grad_norm": 1.8843427896499634, + "learning_rate": 9.40413711221122e-05, + "loss": 1.0995, + "step": 2627 + }, + { + "epoch": 0.1596694817425117, + "grad_norm": 0.29903000593185425, + "learning_rate": 9.403683820538208e-05, + "loss": 1.2161, + "step": 2628 + }, + { + "epoch": 0.15973023877513823, + "grad_norm": 0.42043474316596985, + "learning_rate": 9.403230367447311e-05, + "loss": 1.197, + "step": 2629 + }, + { + "epoch": 0.15979099580776474, + "grad_norm": 0.35084953904151917, + "learning_rate": 9.402776752955155e-05, + "loss": 1.2171, + "step": 2630 + }, + { + "epoch": 0.15985175284039127, + "grad_norm": 0.19534771144390106, + "learning_rate": 9.402322977078366e-05, + "loss": 1.1284, + "step": 2631 + }, + { + "epoch": 0.1599125098730178, + "grad_norm": 0.2793023884296417, + "learning_rate": 9.401869039833577e-05, + "loss": 1.1543, + "step": 2632 + }, + { + "epoch": 0.15997326690564434, + "grad_norm": 0.18426819145679474, + "learning_rate": 9.40141494123743e-05, + "loss": 1.0825, + "step": 2633 + }, + { + "epoch": 0.16003402393827085, + "grad_norm": 0.5413832664489746, + "learning_rate": 9.400960681306566e-05, + "loss": 1.4225, + "step": 2634 + }, + { + "epoch": 0.16009478097089738, + "grad_norm": 0.19359029829502106, + "learning_rate": 9.400506260057635e-05, + "loss": 1.1517, + "step": 2635 + }, + { + "epoch": 0.16015553800352392, + "grad_norm": 0.19823139905929565, + "learning_rate": 9.4000516775073e-05, + "loss": 1.07, + "step": 2636 + }, + { + "epoch": 0.16021629503615042, + "grad_norm": 0.25896403193473816, + "learning_rate": 9.399596933672217e-05, + "loss": 1.123, + "step": 2637 + }, + { + "epoch": 0.16027705206877696, + "grad_norm": 0.24006807804107666, + "learning_rate": 9.399142028569061e-05, + "loss": 1.0775, + "step": 2638 + }, + { + "epoch": 0.1603378091014035, + "grad_norm": 0.49177098274230957, + "learning_rate": 9.398686962214502e-05, + "loss": 1.2406, + "step": 2639 + }, + { + "epoch": 0.16039856613403, + "grad_norm": 0.19349293410778046, + "learning_rate": 9.398231734625222e-05, + "loss": 1.0592, + "step": 2640 + }, + { + "epoch": 0.16045932316665654, + "grad_norm": 1.7689653635025024, + "learning_rate": 9.39777634581791e-05, + "loss": 1.0872, + "step": 2641 + }, + { + "epoch": 0.16052008019928307, + "grad_norm": 0.4315688908100128, + "learning_rate": 9.397320795809254e-05, + "loss": 1.3716, + "step": 2642 + }, + { + "epoch": 0.1605808372319096, + "grad_norm": 0.4280950427055359, + "learning_rate": 9.396865084615954e-05, + "loss": 1.1031, + "step": 2643 + }, + { + "epoch": 0.16064159426453611, + "grad_norm": 0.24947115778923035, + "learning_rate": 9.396409212254715e-05, + "loss": 1.1301, + "step": 2644 + }, + { + "epoch": 0.16070235129716265, + "grad_norm": 0.3653484284877777, + "learning_rate": 9.395953178742246e-05, + "loss": 1.1847, + "step": 2645 + }, + { + "epoch": 0.16076310832978918, + "grad_norm": 0.18685197830200195, + "learning_rate": 9.395496984095264e-05, + "loss": 1.09, + "step": 2646 + }, + { + "epoch": 0.1608238653624157, + "grad_norm": 0.3702411353588104, + "learning_rate": 9.39504062833049e-05, + "loss": 1.2629, + "step": 2647 + }, + { + "epoch": 0.16088462239504223, + "grad_norm": 0.3324156105518341, + "learning_rate": 9.394584111464653e-05, + "loss": 1.1504, + "step": 2648 + }, + { + "epoch": 0.16094537942766876, + "grad_norm": 0.3112645745277405, + "learning_rate": 9.394127433514485e-05, + "loss": 1.2476, + "step": 2649 + }, + { + "epoch": 0.16100613646029527, + "grad_norm": 0.2611483335494995, + "learning_rate": 9.393670594496727e-05, + "loss": 1.1362, + "step": 2650 + }, + { + "epoch": 0.1610668934929218, + "grad_norm": 0.2793232798576355, + "learning_rate": 9.393213594428124e-05, + "loss": 1.1172, + "step": 2651 + }, + { + "epoch": 0.16112765052554834, + "grad_norm": 0.3248705267906189, + "learning_rate": 9.392756433325426e-05, + "loss": 1.0561, + "step": 2652 + }, + { + "epoch": 0.16118840755817485, + "grad_norm": 0.546266496181488, + "learning_rate": 9.392299111205396e-05, + "loss": 1.322, + "step": 2653 + }, + { + "epoch": 0.16124916459080138, + "grad_norm": 0.6148990988731384, + "learning_rate": 9.39184162808479e-05, + "loss": 1.1502, + "step": 2654 + }, + { + "epoch": 0.16130992162342792, + "grad_norm": 0.36716222763061523, + "learning_rate": 9.39138398398038e-05, + "loss": 1.1841, + "step": 2655 + }, + { + "epoch": 0.16137067865605445, + "grad_norm": 0.372345894575119, + "learning_rate": 9.390926178908943e-05, + "loss": 1.2263, + "step": 2656 + }, + { + "epoch": 0.16143143568868096, + "grad_norm": 0.716346263885498, + "learning_rate": 9.390468212887255e-05, + "loss": 1.0976, + "step": 2657 + }, + { + "epoch": 0.1614921927213075, + "grad_norm": 0.33051377534866333, + "learning_rate": 9.39001008593211e-05, + "loss": 1.0952, + "step": 2658 + }, + { + "epoch": 0.16155294975393403, + "grad_norm": 0.6288651823997498, + "learning_rate": 9.389551798060294e-05, + "loss": 1.0874, + "step": 2659 + }, + { + "epoch": 0.16161370678656053, + "grad_norm": 0.29638445377349854, + "learning_rate": 9.38909334928861e-05, + "loss": 1.1426, + "step": 2660 + }, + { + "epoch": 0.16167446381918707, + "grad_norm": 0.590007483959198, + "learning_rate": 9.388634739633859e-05, + "loss": 1.0932, + "step": 2661 + }, + { + "epoch": 0.1617352208518136, + "grad_norm": 0.4825347065925598, + "learning_rate": 9.388175969112856e-05, + "loss": 1.105, + "step": 2662 + }, + { + "epoch": 0.1617959778844401, + "grad_norm": 0.3040347397327423, + "learning_rate": 9.387717037742412e-05, + "loss": 1.1562, + "step": 2663 + }, + { + "epoch": 0.16185673491706665, + "grad_norm": 0.6424562335014343, + "learning_rate": 9.387257945539354e-05, + "loss": 1.0914, + "step": 2664 + }, + { + "epoch": 0.16191749194969318, + "grad_norm": 0.7367885112762451, + "learning_rate": 9.386798692520507e-05, + "loss": 1.1068, + "step": 2665 + }, + { + "epoch": 0.16197824898231972, + "grad_norm": 2.050274133682251, + "learning_rate": 9.386339278702704e-05, + "loss": 1.154, + "step": 2666 + }, + { + "epoch": 0.16203900601494622, + "grad_norm": 0.927492618560791, + "learning_rate": 9.385879704102789e-05, + "loss": 1.0913, + "step": 2667 + }, + { + "epoch": 0.16209976304757276, + "grad_norm": 0.23280343413352966, + "learning_rate": 9.385419968737605e-05, + "loss": 1.1199, + "step": 2668 + }, + { + "epoch": 0.1621605200801993, + "grad_norm": 0.7448911070823669, + "learning_rate": 9.384960072624005e-05, + "loss": 1.1712, + "step": 2669 + }, + { + "epoch": 0.1622212771128258, + "grad_norm": 0.373996764421463, + "learning_rate": 9.384500015778846e-05, + "loss": 1.1538, + "step": 2670 + }, + { + "epoch": 0.16228203414545234, + "grad_norm": 0.6237447261810303, + "learning_rate": 9.384039798218991e-05, + "loss": 1.1494, + "step": 2671 + }, + { + "epoch": 0.16234279117807887, + "grad_norm": 0.4030022621154785, + "learning_rate": 9.38357941996131e-05, + "loss": 1.119, + "step": 2672 + }, + { + "epoch": 0.16240354821070538, + "grad_norm": 0.2548151910305023, + "learning_rate": 9.383118881022679e-05, + "loss": 1.1229, + "step": 2673 + }, + { + "epoch": 0.1624643052433319, + "grad_norm": 0.5915089845657349, + "learning_rate": 9.382658181419977e-05, + "loss": 1.2519, + "step": 2674 + }, + { + "epoch": 0.16252506227595845, + "grad_norm": 0.5456612706184387, + "learning_rate": 9.382197321170092e-05, + "loss": 1.1287, + "step": 2675 + }, + { + "epoch": 0.16258581930858496, + "grad_norm": 0.2995673418045044, + "learning_rate": 9.381736300289917e-05, + "loss": 1.1638, + "step": 2676 + }, + { + "epoch": 0.1626465763412115, + "grad_norm": 0.5792114734649658, + "learning_rate": 9.38127511879635e-05, + "loss": 1.1315, + "step": 2677 + }, + { + "epoch": 0.16270733337383803, + "grad_norm": 0.23783735930919647, + "learning_rate": 9.380813776706298e-05, + "loss": 1.1694, + "step": 2678 + }, + { + "epoch": 0.16276809040646456, + "grad_norm": 0.47661513090133667, + "learning_rate": 9.380352274036672e-05, + "loss": 1.2579, + "step": 2679 + }, + { + "epoch": 0.16282884743909107, + "grad_norm": 0.30699706077575684, + "learning_rate": 9.379890610804385e-05, + "loss": 1.2152, + "step": 2680 + }, + { + "epoch": 0.1628896044717176, + "grad_norm": 0.24369481205940247, + "learning_rate": 9.379428787026361e-05, + "loss": 1.1929, + "step": 2681 + }, + { + "epoch": 0.16295036150434414, + "grad_norm": 0.44185954332351685, + "learning_rate": 9.378966802719528e-05, + "loss": 1.0927, + "step": 2682 + }, + { + "epoch": 0.16301111853697065, + "grad_norm": 0.41857442259788513, + "learning_rate": 9.378504657900821e-05, + "loss": 1.0763, + "step": 2683 + }, + { + "epoch": 0.16307187556959718, + "grad_norm": 0.2775743305683136, + "learning_rate": 9.37804235258718e-05, + "loss": 1.1465, + "step": 2684 + }, + { + "epoch": 0.16313263260222371, + "grad_norm": 0.5739840865135193, + "learning_rate": 9.377579886795549e-05, + "loss": 1.1895, + "step": 2685 + }, + { + "epoch": 0.16319338963485022, + "grad_norm": 0.5433043837547302, + "learning_rate": 9.37711726054288e-05, + "loss": 1.2278, + "step": 2686 + }, + { + "epoch": 0.16325414666747676, + "grad_norm": 0.3524523377418518, + "learning_rate": 9.376654473846134e-05, + "loss": 1.1787, + "step": 2687 + }, + { + "epoch": 0.1633149037001033, + "grad_norm": 0.3072892129421234, + "learning_rate": 9.376191526722272e-05, + "loss": 1.2476, + "step": 2688 + }, + { + "epoch": 0.16337566073272983, + "grad_norm": 0.35175472497940063, + "learning_rate": 9.375728419188264e-05, + "loss": 1.1881, + "step": 2689 + }, + { + "epoch": 0.16343641776535633, + "grad_norm": 0.24125882983207703, + "learning_rate": 9.375265151261083e-05, + "loss": 1.1276, + "step": 2690 + }, + { + "epoch": 0.16349717479798287, + "grad_norm": 0.2587316930294037, + "learning_rate": 9.374801722957714e-05, + "loss": 1.0868, + "step": 2691 + }, + { + "epoch": 0.1635579318306094, + "grad_norm": 0.2258414775133133, + "learning_rate": 9.374338134295141e-05, + "loss": 1.0683, + "step": 2692 + }, + { + "epoch": 0.1636186888632359, + "grad_norm": 0.47443661093711853, + "learning_rate": 9.373874385290358e-05, + "loss": 1.1676, + "step": 2693 + }, + { + "epoch": 0.16367944589586245, + "grad_norm": 0.37387844920158386, + "learning_rate": 9.373410475960364e-05, + "loss": 1.0917, + "step": 2694 + }, + { + "epoch": 0.16374020292848898, + "grad_norm": 0.5431914925575256, + "learning_rate": 9.372946406322164e-05, + "loss": 1.0998, + "step": 2695 + }, + { + "epoch": 0.1638009599611155, + "grad_norm": 0.31997984647750854, + "learning_rate": 9.37248217639277e-05, + "loss": 1.1111, + "step": 2696 + }, + { + "epoch": 0.16386171699374202, + "grad_norm": 0.28344979882240295, + "learning_rate": 9.372017786189195e-05, + "loss": 1.236, + "step": 2697 + }, + { + "epoch": 0.16392247402636856, + "grad_norm": 0.5752473473548889, + "learning_rate": 9.371553235728463e-05, + "loss": 1.2153, + "step": 2698 + }, + { + "epoch": 0.16398323105899507, + "grad_norm": 0.2367956042289734, + "learning_rate": 9.371088525027602e-05, + "loss": 1.1434, + "step": 2699 + }, + { + "epoch": 0.1640439880916216, + "grad_norm": 0.3402453362941742, + "learning_rate": 9.370623654103646e-05, + "loss": 1.1935, + "step": 2700 + }, + { + "epoch": 0.16410474512424814, + "grad_norm": 0.7115744948387146, + "learning_rate": 9.370158622973637e-05, + "loss": 1.1943, + "step": 2701 + }, + { + "epoch": 0.16416550215687467, + "grad_norm": 0.35338589549064636, + "learning_rate": 9.369693431654617e-05, + "loss": 1.2106, + "step": 2702 + }, + { + "epoch": 0.16422625918950118, + "grad_norm": 0.32895776629447937, + "learning_rate": 9.36922808016364e-05, + "loss": 1.1851, + "step": 2703 + }, + { + "epoch": 0.1642870162221277, + "grad_norm": 0.2714262008666992, + "learning_rate": 9.368762568517763e-05, + "loss": 1.0835, + "step": 2704 + }, + { + "epoch": 0.16434777325475425, + "grad_norm": 0.49229857325553894, + "learning_rate": 9.36829689673405e-05, + "loss": 1.0889, + "step": 2705 + }, + { + "epoch": 0.16440853028738076, + "grad_norm": 0.24390706419944763, + "learning_rate": 9.367831064829571e-05, + "loss": 1.1263, + "step": 2706 + }, + { + "epoch": 0.1644692873200073, + "grad_norm": 0.35617780685424805, + "learning_rate": 9.367365072821397e-05, + "loss": 1.1755, + "step": 2707 + }, + { + "epoch": 0.16453004435263383, + "grad_norm": 0.22776710987091064, + "learning_rate": 9.366898920726615e-05, + "loss": 1.1324, + "step": 2708 + }, + { + "epoch": 0.16459080138526033, + "grad_norm": 0.3052212595939636, + "learning_rate": 9.366432608562308e-05, + "loss": 1.0715, + "step": 2709 + }, + { + "epoch": 0.16465155841788687, + "grad_norm": 0.2809009253978729, + "learning_rate": 9.365966136345569e-05, + "loss": 1.1186, + "step": 2710 + }, + { + "epoch": 0.1647123154505134, + "grad_norm": 0.3502916097640991, + "learning_rate": 9.365499504093498e-05, + "loss": 1.2184, + "step": 2711 + }, + { + "epoch": 0.1647730724831399, + "grad_norm": 0.43182381987571716, + "learning_rate": 9.365032711823198e-05, + "loss": 1.415, + "step": 2712 + }, + { + "epoch": 0.16483382951576644, + "grad_norm": 0.4340287446975708, + "learning_rate": 9.364565759551782e-05, + "loss": 1.1728, + "step": 2713 + }, + { + "epoch": 0.16489458654839298, + "grad_norm": 0.413770854473114, + "learning_rate": 9.364098647296364e-05, + "loss": 1.1131, + "step": 2714 + }, + { + "epoch": 0.16495534358101951, + "grad_norm": 0.29897448420524597, + "learning_rate": 9.363631375074065e-05, + "loss": 1.1306, + "step": 2715 + }, + { + "epoch": 0.16501610061364602, + "grad_norm": 0.3501521348953247, + "learning_rate": 9.363163942902016e-05, + "loss": 1.2144, + "step": 2716 + }, + { + "epoch": 0.16507685764627256, + "grad_norm": 0.6310165524482727, + "learning_rate": 9.362696350797348e-05, + "loss": 1.1169, + "step": 2717 + }, + { + "epoch": 0.1651376146788991, + "grad_norm": 0.3033841550350189, + "learning_rate": 9.362228598777203e-05, + "loss": 1.2129, + "step": 2718 + }, + { + "epoch": 0.1651983717115256, + "grad_norm": 0.5143870711326599, + "learning_rate": 9.361760686858725e-05, + "loss": 1.1778, + "step": 2719 + }, + { + "epoch": 0.16525912874415213, + "grad_norm": 0.403216153383255, + "learning_rate": 9.361292615059065e-05, + "loss": 1.16, + "step": 2720 + }, + { + "epoch": 0.16531988577677867, + "grad_norm": 0.37667316198349, + "learning_rate": 9.360824383395383e-05, + "loss": 1.1534, + "step": 2721 + }, + { + "epoch": 0.16538064280940518, + "grad_norm": 0.2899409830570221, + "learning_rate": 9.360355991884839e-05, + "loss": 1.0433, + "step": 2722 + }, + { + "epoch": 0.1654413998420317, + "grad_norm": 0.3253422975540161, + "learning_rate": 9.359887440544605e-05, + "loss": 1.297, + "step": 2723 + }, + { + "epoch": 0.16550215687465825, + "grad_norm": 0.47208738327026367, + "learning_rate": 9.359418729391852e-05, + "loss": 1.255, + "step": 2724 + }, + { + "epoch": 0.16556291390728478, + "grad_norm": 0.2546895742416382, + "learning_rate": 9.358949858443764e-05, + "loss": 1.1213, + "step": 2725 + }, + { + "epoch": 0.1656236709399113, + "grad_norm": 0.3175262212753296, + "learning_rate": 9.358480827717526e-05, + "loss": 1.2932, + "step": 2726 + }, + { + "epoch": 0.16568442797253782, + "grad_norm": 0.24299059808254242, + "learning_rate": 9.358011637230331e-05, + "loss": 1.1156, + "step": 2727 + }, + { + "epoch": 0.16574518500516436, + "grad_norm": 0.41144052147865295, + "learning_rate": 9.357542286999377e-05, + "loss": 1.1416, + "step": 2728 + }, + { + "epoch": 0.16580594203779087, + "grad_norm": 0.21732720732688904, + "learning_rate": 9.357072777041869e-05, + "loss": 1.0887, + "step": 2729 + }, + { + "epoch": 0.1658666990704174, + "grad_norm": 0.3629020154476166, + "learning_rate": 9.356603107375015e-05, + "loss": 1.3212, + "step": 2730 + }, + { + "epoch": 0.16592745610304394, + "grad_norm": 0.49388447403907776, + "learning_rate": 9.356133278016033e-05, + "loss": 1.1857, + "step": 2731 + }, + { + "epoch": 0.16598821313567044, + "grad_norm": 0.20743955671787262, + "learning_rate": 9.355663288982141e-05, + "loss": 1.0904, + "step": 2732 + }, + { + "epoch": 0.16604897016829698, + "grad_norm": 0.6185721158981323, + "learning_rate": 9.355193140290573e-05, + "loss": 1.3248, + "step": 2733 + }, + { + "epoch": 0.1661097272009235, + "grad_norm": 1.125495433807373, + "learning_rate": 9.354722831958556e-05, + "loss": 1.0882, + "step": 2734 + }, + { + "epoch": 0.16617048423355002, + "grad_norm": 0.40673282742500305, + "learning_rate": 9.354252364003333e-05, + "loss": 1.2203, + "step": 2735 + }, + { + "epoch": 0.16623124126617655, + "grad_norm": 0.2969594895839691, + "learning_rate": 9.353781736442146e-05, + "loss": 1.0794, + "step": 2736 + }, + { + "epoch": 0.1662919982988031, + "grad_norm": 0.2784421145915985, + "learning_rate": 9.353310949292251e-05, + "loss": 1.1456, + "step": 2737 + }, + { + "epoch": 0.16635275533142962, + "grad_norm": 0.3885911703109741, + "learning_rate": 9.3528400025709e-05, + "loss": 1.1163, + "step": 2738 + }, + { + "epoch": 0.16641351236405613, + "grad_norm": 0.25742724537849426, + "learning_rate": 9.352368896295356e-05, + "loss": 1.1034, + "step": 2739 + }, + { + "epoch": 0.16647426939668267, + "grad_norm": 0.41958558559417725, + "learning_rate": 9.351897630482891e-05, + "loss": 1.2566, + "step": 2740 + }, + { + "epoch": 0.1665350264293092, + "grad_norm": 0.21760554611682892, + "learning_rate": 9.351426205150777e-05, + "loss": 1.1396, + "step": 2741 + }, + { + "epoch": 0.1665957834619357, + "grad_norm": 0.39776769280433655, + "learning_rate": 9.350954620316293e-05, + "loss": 1.1219, + "step": 2742 + }, + { + "epoch": 0.16665654049456224, + "grad_norm": 0.333066463470459, + "learning_rate": 9.350482875996726e-05, + "loss": 1.2102, + "step": 2743 + }, + { + "epoch": 0.16671729752718878, + "grad_norm": 0.36990055441856384, + "learning_rate": 9.350010972209371e-05, + "loss": 1.3059, + "step": 2744 + }, + { + "epoch": 0.1667780545598153, + "grad_norm": 0.2662367522716522, + "learning_rate": 9.349538908971521e-05, + "loss": 1.2067, + "step": 2745 + }, + { + "epoch": 0.16683881159244182, + "grad_norm": 0.5546712875366211, + "learning_rate": 9.349066686300483e-05, + "loss": 1.2174, + "step": 2746 + }, + { + "epoch": 0.16689956862506836, + "grad_norm": 0.3266717493534088, + "learning_rate": 9.348594304213562e-05, + "loss": 1.3251, + "step": 2747 + }, + { + "epoch": 0.1669603256576949, + "grad_norm": 0.3384968936443329, + "learning_rate": 9.348121762728079e-05, + "loss": 1.1398, + "step": 2748 + }, + { + "epoch": 0.1670210826903214, + "grad_norm": 0.24762506783008575, + "learning_rate": 9.347649061861352e-05, + "loss": 1.225, + "step": 2749 + }, + { + "epoch": 0.16708183972294793, + "grad_norm": 0.42243295907974243, + "learning_rate": 9.347176201630707e-05, + "loss": 1.1294, + "step": 2750 + }, + { + "epoch": 0.16714259675557447, + "grad_norm": 0.2884136736392975, + "learning_rate": 9.346703182053479e-05, + "loss": 1.1948, + "step": 2751 + }, + { + "epoch": 0.16720335378820098, + "grad_norm": 0.3167388439178467, + "learning_rate": 9.346230003147005e-05, + "loss": 1.2118, + "step": 2752 + }, + { + "epoch": 0.1672641108208275, + "grad_norm": 0.6310069561004639, + "learning_rate": 9.345756664928631e-05, + "loss": 1.0985, + "step": 2753 + }, + { + "epoch": 0.16732486785345405, + "grad_norm": 0.23422326147556305, + "learning_rate": 9.345283167415706e-05, + "loss": 1.1165, + "step": 2754 + }, + { + "epoch": 0.16738562488608055, + "grad_norm": 0.3113587200641632, + "learning_rate": 9.344809510625586e-05, + "loss": 1.0596, + "step": 2755 + }, + { + "epoch": 0.1674463819187071, + "grad_norm": 0.23958683013916016, + "learning_rate": 9.344335694575634e-05, + "loss": 1.1085, + "step": 2756 + }, + { + "epoch": 0.16750713895133362, + "grad_norm": 0.42176344990730286, + "learning_rate": 9.343861719283217e-05, + "loss": 1.0797, + "step": 2757 + }, + { + "epoch": 0.16756789598396013, + "grad_norm": 0.27492448687553406, + "learning_rate": 9.34338758476571e-05, + "loss": 1.1297, + "step": 2758 + }, + { + "epoch": 0.16762865301658667, + "grad_norm": 2.1752285957336426, + "learning_rate": 9.342913291040491e-05, + "loss": 1.1403, + "step": 2759 + }, + { + "epoch": 0.1676894100492132, + "grad_norm": 0.2248554825782776, + "learning_rate": 9.342438838124945e-05, + "loss": 1.1045, + "step": 2760 + }, + { + "epoch": 0.16775016708183974, + "grad_norm": 0.38420942425727844, + "learning_rate": 9.341964226036466e-05, + "loss": 1.0931, + "step": 2761 + }, + { + "epoch": 0.16781092411446624, + "grad_norm": 0.23379847407341003, + "learning_rate": 9.341489454792447e-05, + "loss": 1.2129, + "step": 2762 + }, + { + "epoch": 0.16787168114709278, + "grad_norm": 0.399139940738678, + "learning_rate": 9.341014524410295e-05, + "loss": 1.1905, + "step": 2763 + }, + { + "epoch": 0.1679324381797193, + "grad_norm": 0.2786547541618347, + "learning_rate": 9.340539434907414e-05, + "loss": 1.1145, + "step": 2764 + }, + { + "epoch": 0.16799319521234582, + "grad_norm": 0.3586018979549408, + "learning_rate": 9.340064186301221e-05, + "loss": 1.0734, + "step": 2765 + }, + { + "epoch": 0.16805395224497235, + "grad_norm": 0.4905034601688385, + "learning_rate": 9.339588778609138e-05, + "loss": 1.0542, + "step": 2766 + }, + { + "epoch": 0.1681147092775989, + "grad_norm": 0.5317531228065491, + "learning_rate": 9.33911321184859e-05, + "loss": 1.2154, + "step": 2767 + }, + { + "epoch": 0.1681754663102254, + "grad_norm": 0.5478570461273193, + "learning_rate": 9.338637486037007e-05, + "loss": 1.2746, + "step": 2768 + }, + { + "epoch": 0.16823622334285193, + "grad_norm": 0.41557973623275757, + "learning_rate": 9.338161601191828e-05, + "loss": 1.202, + "step": 2769 + }, + { + "epoch": 0.16829698037547847, + "grad_norm": 0.2911107838153839, + "learning_rate": 9.337685557330497e-05, + "loss": 1.0766, + "step": 2770 + }, + { + "epoch": 0.168357737408105, + "grad_norm": 0.41177603602409363, + "learning_rate": 9.337209354470467e-05, + "loss": 1.1168, + "step": 2771 + }, + { + "epoch": 0.1684184944407315, + "grad_norm": 0.26611047983169556, + "learning_rate": 9.336732992629186e-05, + "loss": 1.1863, + "step": 2772 + }, + { + "epoch": 0.16847925147335804, + "grad_norm": 0.25866058468818665, + "learning_rate": 9.33625647182412e-05, + "loss": 1.0871, + "step": 2773 + }, + { + "epoch": 0.16854000850598458, + "grad_norm": 0.36982694268226624, + "learning_rate": 9.335779792072734e-05, + "loss": 1.0661, + "step": 2774 + }, + { + "epoch": 0.16860076553861109, + "grad_norm": 0.31414881348609924, + "learning_rate": 9.335302953392501e-05, + "loss": 1.2386, + "step": 2775 + }, + { + "epoch": 0.16866152257123762, + "grad_norm": 0.4192929267883301, + "learning_rate": 9.334825955800904e-05, + "loss": 1.0546, + "step": 2776 + }, + { + "epoch": 0.16872227960386416, + "grad_norm": 0.23979736864566803, + "learning_rate": 9.33434879931542e-05, + "loss": 1.2159, + "step": 2777 + }, + { + "epoch": 0.16878303663649066, + "grad_norm": 0.38037747144699097, + "learning_rate": 9.333871483953544e-05, + "loss": 1.1706, + "step": 2778 + }, + { + "epoch": 0.1688437936691172, + "grad_norm": 0.40348020195961, + "learning_rate": 9.333394009732772e-05, + "loss": 1.3237, + "step": 2779 + }, + { + "epoch": 0.16890455070174373, + "grad_norm": 0.3569577932357788, + "learning_rate": 9.332916376670603e-05, + "loss": 1.126, + "step": 2780 + }, + { + "epoch": 0.16896530773437024, + "grad_norm": 0.25449296832084656, + "learning_rate": 9.332438584784549e-05, + "loss": 1.1387, + "step": 2781 + }, + { + "epoch": 0.16902606476699678, + "grad_norm": 0.29519376158714294, + "learning_rate": 9.33196063409212e-05, + "loss": 1.0842, + "step": 2782 + }, + { + "epoch": 0.1690868217996233, + "grad_norm": 0.29107919335365295, + "learning_rate": 9.331482524610837e-05, + "loss": 1.0711, + "step": 2783 + }, + { + "epoch": 0.16914757883224985, + "grad_norm": 0.2481674700975418, + "learning_rate": 9.331004256358224e-05, + "loss": 1.0905, + "step": 2784 + }, + { + "epoch": 0.16920833586487635, + "grad_norm": 0.3577122688293457, + "learning_rate": 9.330525829351813e-05, + "loss": 1.0995, + "step": 2785 + }, + { + "epoch": 0.1692690928975029, + "grad_norm": 0.3262396454811096, + "learning_rate": 9.330047243609142e-05, + "loss": 1.2579, + "step": 2786 + }, + { + "epoch": 0.16932984993012942, + "grad_norm": 0.5815495252609253, + "learning_rate": 9.32956849914775e-05, + "loss": 1.1419, + "step": 2787 + }, + { + "epoch": 0.16939060696275593, + "grad_norm": 0.24938301742076874, + "learning_rate": 9.329089595985189e-05, + "loss": 1.2277, + "step": 2788 + }, + { + "epoch": 0.16945136399538246, + "grad_norm": 2.030596971511841, + "learning_rate": 9.328610534139013e-05, + "loss": 1.1832, + "step": 2789 + }, + { + "epoch": 0.169512121028009, + "grad_norm": 0.3923100233078003, + "learning_rate": 9.32813131362678e-05, + "loss": 1.1597, + "step": 2790 + }, + { + "epoch": 0.1695728780606355, + "grad_norm": 0.5162453055381775, + "learning_rate": 9.327651934466057e-05, + "loss": 1.0595, + "step": 2791 + }, + { + "epoch": 0.16963363509326204, + "grad_norm": 0.34237298369407654, + "learning_rate": 9.327172396674417e-05, + "loss": 1.1135, + "step": 2792 + }, + { + "epoch": 0.16969439212588858, + "grad_norm": 0.6583300232887268, + "learning_rate": 9.326692700269434e-05, + "loss": 1.0907, + "step": 2793 + }, + { + "epoch": 0.1697551491585151, + "grad_norm": 0.2087724655866623, + "learning_rate": 9.326212845268695e-05, + "loss": 1.0831, + "step": 2794 + }, + { + "epoch": 0.16981590619114162, + "grad_norm": 0.7012311816215515, + "learning_rate": 9.325732831689791e-05, + "loss": 1.0581, + "step": 2795 + }, + { + "epoch": 0.16987666322376815, + "grad_norm": 0.22481393814086914, + "learning_rate": 9.325252659550309e-05, + "loss": 1.0777, + "step": 2796 + }, + { + "epoch": 0.1699374202563947, + "grad_norm": 0.43668508529663086, + "learning_rate": 9.324772328867858e-05, + "loss": 1.0437, + "step": 2797 + }, + { + "epoch": 0.1699981772890212, + "grad_norm": 0.25170382857322693, + "learning_rate": 9.32429183966004e-05, + "loss": 1.0732, + "step": 2798 + }, + { + "epoch": 0.17005893432164773, + "grad_norm": 0.18107229471206665, + "learning_rate": 9.323811191944469e-05, + "loss": 1.0672, + "step": 2799 + }, + { + "epoch": 0.17011969135427427, + "grad_norm": 0.23534062504768372, + "learning_rate": 9.323330385738762e-05, + "loss": 1.1136, + "step": 2800 + }, + { + "epoch": 0.17018044838690077, + "grad_norm": 0.27315929532051086, + "learning_rate": 9.322849421060546e-05, + "loss": 1.1967, + "step": 2801 + }, + { + "epoch": 0.1702412054195273, + "grad_norm": 0.23659847676753998, + "learning_rate": 9.322368297927448e-05, + "loss": 1.1158, + "step": 2802 + }, + { + "epoch": 0.17030196245215384, + "grad_norm": 0.22341887652873993, + "learning_rate": 9.321887016357104e-05, + "loss": 1.1189, + "step": 2803 + }, + { + "epoch": 0.17036271948478035, + "grad_norm": 0.2514699101448059, + "learning_rate": 9.321405576367157e-05, + "loss": 1.3526, + "step": 2804 + }, + { + "epoch": 0.17042347651740689, + "grad_norm": 0.3664356768131256, + "learning_rate": 9.320923977975253e-05, + "loss": 1.2214, + "step": 2805 + }, + { + "epoch": 0.17048423355003342, + "grad_norm": 0.25590214133262634, + "learning_rate": 9.320442221199044e-05, + "loss": 1.2329, + "step": 2806 + }, + { + "epoch": 0.17054499058265996, + "grad_norm": 0.21339033544063568, + "learning_rate": 9.319960306056192e-05, + "loss": 1.1004, + "step": 2807 + }, + { + "epoch": 0.17060574761528646, + "grad_norm": 0.4351475238800049, + "learning_rate": 9.319478232564359e-05, + "loss": 1.0728, + "step": 2808 + }, + { + "epoch": 0.170666504647913, + "grad_norm": 0.2654882073402405, + "learning_rate": 9.318996000741215e-05, + "loss": 1.2275, + "step": 2809 + }, + { + "epoch": 0.17072726168053953, + "grad_norm": 0.3951309025287628, + "learning_rate": 9.31851361060444e-05, + "loss": 1.1608, + "step": 2810 + }, + { + "epoch": 0.17078801871316604, + "grad_norm": 5.332469940185547, + "learning_rate": 9.318031062171715e-05, + "loss": 1.3282, + "step": 2811 + }, + { + "epoch": 0.17084877574579257, + "grad_norm": 0.6825553178787231, + "learning_rate": 9.317548355460725e-05, + "loss": 1.3574, + "step": 2812 + }, + { + "epoch": 0.1709095327784191, + "grad_norm": 0.3772541582584381, + "learning_rate": 9.317065490489165e-05, + "loss": 1.1576, + "step": 2813 + }, + { + "epoch": 0.17097028981104562, + "grad_norm": 0.4450722634792328, + "learning_rate": 9.316582467274735e-05, + "loss": 1.2052, + "step": 2814 + }, + { + "epoch": 0.17103104684367215, + "grad_norm": 0.44607067108154297, + "learning_rate": 9.316099285835141e-05, + "loss": 1.148, + "step": 2815 + }, + { + "epoch": 0.1710918038762987, + "grad_norm": 0.36093851923942566, + "learning_rate": 9.315615946188093e-05, + "loss": 1.1512, + "step": 2816 + }, + { + "epoch": 0.1711525609089252, + "grad_norm": 0.23870272934436798, + "learning_rate": 9.315132448351309e-05, + "loss": 1.1027, + "step": 2817 + }, + { + "epoch": 0.17121331794155173, + "grad_norm": 0.231075718998909, + "learning_rate": 9.314648792342511e-05, + "loss": 1.0934, + "step": 2818 + }, + { + "epoch": 0.17127407497417826, + "grad_norm": 0.27368393540382385, + "learning_rate": 9.314164978179427e-05, + "loss": 1.1438, + "step": 2819 + }, + { + "epoch": 0.1713348320068048, + "grad_norm": 0.5192923545837402, + "learning_rate": 9.313681005879791e-05, + "loss": 1.0786, + "step": 2820 + }, + { + "epoch": 0.1713955890394313, + "grad_norm": 0.19180211424827576, + "learning_rate": 9.313196875461345e-05, + "loss": 1.1173, + "step": 2821 + }, + { + "epoch": 0.17145634607205784, + "grad_norm": 0.3187803328037262, + "learning_rate": 9.312712586941834e-05, + "loss": 1.156, + "step": 2822 + }, + { + "epoch": 0.17151710310468438, + "grad_norm": 0.24049100279808044, + "learning_rate": 9.312228140339009e-05, + "loss": 1.0789, + "step": 2823 + }, + { + "epoch": 0.17157786013731088, + "grad_norm": 0.2890015244483948, + "learning_rate": 9.311743535670628e-05, + "loss": 1.1025, + "step": 2824 + }, + { + "epoch": 0.17163861716993742, + "grad_norm": 0.3852143883705139, + "learning_rate": 9.311258772954454e-05, + "loss": 1.0849, + "step": 2825 + }, + { + "epoch": 0.17169937420256395, + "grad_norm": 0.4183599650859833, + "learning_rate": 9.310773852208256e-05, + "loss": 1.1048, + "step": 2826 + }, + { + "epoch": 0.17176013123519046, + "grad_norm": 0.32630330324172974, + "learning_rate": 9.310288773449811e-05, + "loss": 1.1104, + "step": 2827 + }, + { + "epoch": 0.171820888267817, + "grad_norm": 0.2336236983537674, + "learning_rate": 9.309803536696897e-05, + "loss": 1.2383, + "step": 2828 + }, + { + "epoch": 0.17188164530044353, + "grad_norm": 0.30658724904060364, + "learning_rate": 9.3093181419673e-05, + "loss": 1.1726, + "step": 2829 + }, + { + "epoch": 0.17194240233307007, + "grad_norm": 1.1383273601531982, + "learning_rate": 9.308832589278815e-05, + "loss": 1.114, + "step": 2830 + }, + { + "epoch": 0.17200315936569657, + "grad_norm": 2.222982168197632, + "learning_rate": 9.308346878649238e-05, + "loss": 1.1288, + "step": 2831 + }, + { + "epoch": 0.1720639163983231, + "grad_norm": 0.49243050813674927, + "learning_rate": 9.307861010096372e-05, + "loss": 1.0674, + "step": 2832 + }, + { + "epoch": 0.17212467343094964, + "grad_norm": 0.31965532898902893, + "learning_rate": 9.30737498363803e-05, + "loss": 1.1382, + "step": 2833 + }, + { + "epoch": 0.17218543046357615, + "grad_norm": 0.3648076057434082, + "learning_rate": 9.306888799292027e-05, + "loss": 1.08, + "step": 2834 + }, + { + "epoch": 0.17224618749620269, + "grad_norm": 0.25439128279685974, + "learning_rate": 9.306402457076178e-05, + "loss": 1.1499, + "step": 2835 + }, + { + "epoch": 0.17230694452882922, + "grad_norm": 0.7025170922279358, + "learning_rate": 9.305915957008317e-05, + "loss": 1.086, + "step": 2836 + }, + { + "epoch": 0.17236770156145573, + "grad_norm": 0.31440097093582153, + "learning_rate": 9.305429299106275e-05, + "loss": 1.0499, + "step": 2837 + }, + { + "epoch": 0.17242845859408226, + "grad_norm": 0.20028644800186157, + "learning_rate": 9.304942483387889e-05, + "loss": 1.0782, + "step": 2838 + }, + { + "epoch": 0.1724892156267088, + "grad_norm": 0.2834944427013397, + "learning_rate": 9.304455509871004e-05, + "loss": 1.1311, + "step": 2839 + }, + { + "epoch": 0.1725499726593353, + "grad_norm": 0.8304158449172974, + "learning_rate": 9.303968378573471e-05, + "loss": 1.1376, + "step": 2840 + }, + { + "epoch": 0.17261072969196184, + "grad_norm": 3.8430137634277344, + "learning_rate": 9.303481089513144e-05, + "loss": 1.3148, + "step": 2841 + }, + { + "epoch": 0.17267148672458837, + "grad_norm": 0.49445876479148865, + "learning_rate": 9.302993642707887e-05, + "loss": 1.1772, + "step": 2842 + }, + { + "epoch": 0.1727322437572149, + "grad_norm": 0.5228264331817627, + "learning_rate": 9.302506038175565e-05, + "loss": 1.2395, + "step": 2843 + }, + { + "epoch": 0.17279300078984142, + "grad_norm": 0.48553332686424255, + "learning_rate": 9.302018275934053e-05, + "loss": 1.083, + "step": 2844 + }, + { + "epoch": 0.17285375782246795, + "grad_norm": 0.3930256962776184, + "learning_rate": 9.30153035600123e-05, + "loss": 1.164, + "step": 2845 + }, + { + "epoch": 0.1729145148550945, + "grad_norm": 0.2888900339603424, + "learning_rate": 9.30104227839498e-05, + "loss": 1.1815, + "step": 2846 + }, + { + "epoch": 0.172975271887721, + "grad_norm": 0.3630005419254303, + "learning_rate": 9.300554043133194e-05, + "loss": 1.0782, + "step": 2847 + }, + { + "epoch": 0.17303602892034753, + "grad_norm": 0.22247537970542908, + "learning_rate": 9.300065650233767e-05, + "loss": 1.0576, + "step": 2848 + }, + { + "epoch": 0.17309678595297406, + "grad_norm": 0.35226649045944214, + "learning_rate": 9.299577099714605e-05, + "loss": 1.2762, + "step": 2849 + }, + { + "epoch": 0.17315754298560057, + "grad_norm": 0.3527758717536926, + "learning_rate": 9.299088391593611e-05, + "loss": 1.1019, + "step": 2850 + }, + { + "epoch": 0.1732183000182271, + "grad_norm": 0.25990408658981323, + "learning_rate": 9.298599525888703e-05, + "loss": 1.1481, + "step": 2851 + }, + { + "epoch": 0.17327905705085364, + "grad_norm": 0.5389834642410278, + "learning_rate": 9.298110502617796e-05, + "loss": 1.1469, + "step": 2852 + }, + { + "epoch": 0.17333981408348018, + "grad_norm": 0.6207941174507141, + "learning_rate": 9.29762132179882e-05, + "loss": 1.1334, + "step": 2853 + }, + { + "epoch": 0.17340057111610668, + "grad_norm": 0.30199700593948364, + "learning_rate": 9.297131983449702e-05, + "loss": 1.2193, + "step": 2854 + }, + { + "epoch": 0.17346132814873322, + "grad_norm": 0.7605002522468567, + "learning_rate": 9.296642487588381e-05, + "loss": 1.1658, + "step": 2855 + }, + { + "epoch": 0.17352208518135975, + "grad_norm": 0.27793237566947937, + "learning_rate": 9.296152834232801e-05, + "loss": 1.0837, + "step": 2856 + }, + { + "epoch": 0.17358284221398626, + "grad_norm": 0.31688594818115234, + "learning_rate": 9.295663023400907e-05, + "loss": 1.2042, + "step": 2857 + }, + { + "epoch": 0.1736435992466128, + "grad_norm": 0.5309053659439087, + "learning_rate": 9.295173055110653e-05, + "loss": 1.3967, + "step": 2858 + }, + { + "epoch": 0.17370435627923933, + "grad_norm": 0.37057194113731384, + "learning_rate": 9.294682929380003e-05, + "loss": 1.2784, + "step": 2859 + }, + { + "epoch": 0.17376511331186584, + "grad_norm": 0.21573063731193542, + "learning_rate": 9.29419264622692e-05, + "loss": 1.0518, + "step": 2860 + }, + { + "epoch": 0.17382587034449237, + "grad_norm": 0.6183753609657288, + "learning_rate": 9.293702205669373e-05, + "loss": 1.2555, + "step": 2861 + }, + { + "epoch": 0.1738866273771189, + "grad_norm": 0.2507856786251068, + "learning_rate": 9.293211607725342e-05, + "loss": 1.2423, + "step": 2862 + }, + { + "epoch": 0.17394738440974541, + "grad_norm": 0.2787857949733734, + "learning_rate": 9.292720852412811e-05, + "loss": 1.1646, + "step": 2863 + }, + { + "epoch": 0.17400814144237195, + "grad_norm": 0.4763575494289398, + "learning_rate": 9.292229939749768e-05, + "loss": 1.1459, + "step": 2864 + }, + { + "epoch": 0.17406889847499848, + "grad_norm": 1.0242886543273926, + "learning_rate": 9.291738869754207e-05, + "loss": 1.1457, + "step": 2865 + }, + { + "epoch": 0.17412965550762502, + "grad_norm": 0.3670434355735779, + "learning_rate": 9.291247642444127e-05, + "loss": 1.1062, + "step": 2866 + }, + { + "epoch": 0.17419041254025153, + "grad_norm": 0.2926396131515503, + "learning_rate": 9.290756257837535e-05, + "loss": 1.2677, + "step": 2867 + }, + { + "epoch": 0.17425116957287806, + "grad_norm": 0.5605287551879883, + "learning_rate": 9.290264715952444e-05, + "loss": 1.2003, + "step": 2868 + }, + { + "epoch": 0.1743119266055046, + "grad_norm": 2.9236443042755127, + "learning_rate": 9.28977301680687e-05, + "loss": 1.1502, + "step": 2869 + }, + { + "epoch": 0.1743726836381311, + "grad_norm": 0.3267732560634613, + "learning_rate": 9.289281160418837e-05, + "loss": 1.3392, + "step": 2870 + }, + { + "epoch": 0.17443344067075764, + "grad_norm": 1.9855542182922363, + "learning_rate": 9.288789146806374e-05, + "loss": 1.0728, + "step": 2871 + }, + { + "epoch": 0.17449419770338417, + "grad_norm": 0.2774808704853058, + "learning_rate": 9.288296975987515e-05, + "loss": 1.3619, + "step": 2872 + }, + { + "epoch": 0.17455495473601068, + "grad_norm": 1.1166397333145142, + "learning_rate": 9.287804647980303e-05, + "loss": 1.2764, + "step": 2873 + }, + { + "epoch": 0.17461571176863722, + "grad_norm": 0.23068572580814362, + "learning_rate": 9.287312162802782e-05, + "loss": 1.12, + "step": 2874 + }, + { + "epoch": 0.17467646880126375, + "grad_norm": 0.32616013288497925, + "learning_rate": 9.286819520473004e-05, + "loss": 1.1586, + "step": 2875 + }, + { + "epoch": 0.1747372258338903, + "grad_norm": 0.46845772862434387, + "learning_rate": 9.286326721009029e-05, + "loss": 1.1784, + "step": 2876 + }, + { + "epoch": 0.1747979828665168, + "grad_norm": 0.3170464038848877, + "learning_rate": 9.28583376442892e-05, + "loss": 1.1866, + "step": 2877 + }, + { + "epoch": 0.17485873989914333, + "grad_norm": 0.5311518311500549, + "learning_rate": 9.285340650750745e-05, + "loss": 1.2586, + "step": 2878 + }, + { + "epoch": 0.17491949693176986, + "grad_norm": 0.20377646386623383, + "learning_rate": 9.284847379992578e-05, + "loss": 1.0328, + "step": 2879 + }, + { + "epoch": 0.17498025396439637, + "grad_norm": 0.40234893560409546, + "learning_rate": 9.284353952172506e-05, + "loss": 1.2917, + "step": 2880 + }, + { + "epoch": 0.1750410109970229, + "grad_norm": 0.34197208285331726, + "learning_rate": 9.28386036730861e-05, + "loss": 1.1396, + "step": 2881 + }, + { + "epoch": 0.17510176802964944, + "grad_norm": 0.2687784433364868, + "learning_rate": 9.283366625418984e-05, + "loss": 1.1648, + "step": 2882 + }, + { + "epoch": 0.17516252506227595, + "grad_norm": 0.6162838935852051, + "learning_rate": 9.282872726521726e-05, + "loss": 1.1564, + "step": 2883 + }, + { + "epoch": 0.17522328209490248, + "grad_norm": 0.3748330771923065, + "learning_rate": 9.282378670634941e-05, + "loss": 1.061, + "step": 2884 + }, + { + "epoch": 0.17528403912752902, + "grad_norm": 0.39206159114837646, + "learning_rate": 9.281884457776739e-05, + "loss": 1.1314, + "step": 2885 + }, + { + "epoch": 0.17534479616015552, + "grad_norm": 0.3421037793159485, + "learning_rate": 9.281390087965232e-05, + "loss": 1.14, + "step": 2886 + }, + { + "epoch": 0.17540555319278206, + "grad_norm": 0.5817000269889832, + "learning_rate": 9.280895561218546e-05, + "loss": 1.1166, + "step": 2887 + }, + { + "epoch": 0.1754663102254086, + "grad_norm": 0.3381463289260864, + "learning_rate": 9.280400877554805e-05, + "loss": 1.0452, + "step": 2888 + }, + { + "epoch": 0.17552706725803513, + "grad_norm": 0.1933777779340744, + "learning_rate": 9.279906036992143e-05, + "loss": 1.0709, + "step": 2889 + }, + { + "epoch": 0.17558782429066164, + "grad_norm": 0.5630345344543457, + "learning_rate": 9.279411039548697e-05, + "loss": 1.1052, + "step": 2890 + }, + { + "epoch": 0.17564858132328817, + "grad_norm": 0.4031135141849518, + "learning_rate": 9.278915885242612e-05, + "loss": 1.1975, + "step": 2891 + }, + { + "epoch": 0.1757093383559147, + "grad_norm": 0.290926992893219, + "learning_rate": 9.278420574092039e-05, + "loss": 1.1006, + "step": 2892 + }, + { + "epoch": 0.17577009538854121, + "grad_norm": 0.23710793256759644, + "learning_rate": 9.277925106115133e-05, + "loss": 1.0556, + "step": 2893 + }, + { + "epoch": 0.17583085242116775, + "grad_norm": 0.4211599826812744, + "learning_rate": 9.277429481330054e-05, + "loss": 1.1653, + "step": 2894 + }, + { + "epoch": 0.17589160945379428, + "grad_norm": 0.3209962844848633, + "learning_rate": 9.276933699754973e-05, + "loss": 1.1204, + "step": 2895 + }, + { + "epoch": 0.1759523664864208, + "grad_norm": 0.2031252682209015, + "learning_rate": 9.276437761408058e-05, + "loss": 1.1035, + "step": 2896 + }, + { + "epoch": 0.17601312351904733, + "grad_norm": 0.3221298158168793, + "learning_rate": 9.27594166630749e-05, + "loss": 1.098, + "step": 2897 + }, + { + "epoch": 0.17607388055167386, + "grad_norm": 0.3921942710876465, + "learning_rate": 9.275445414471455e-05, + "loss": 1.3156, + "step": 2898 + }, + { + "epoch": 0.1761346375843004, + "grad_norm": 0.24889561533927917, + "learning_rate": 9.274949005918143e-05, + "loss": 1.1212, + "step": 2899 + }, + { + "epoch": 0.1761953946169269, + "grad_norm": 0.2660517394542694, + "learning_rate": 9.274452440665748e-05, + "loss": 1.0551, + "step": 2900 + }, + { + "epoch": 0.17625615164955344, + "grad_norm": 0.1939934939146042, + "learning_rate": 9.273955718732472e-05, + "loss": 1.1071, + "step": 2901 + }, + { + "epoch": 0.17631690868217997, + "grad_norm": 0.2715827524662018, + "learning_rate": 9.273458840136523e-05, + "loss": 1.1498, + "step": 2902 + }, + { + "epoch": 0.17637766571480648, + "grad_norm": 0.3262174129486084, + "learning_rate": 9.272961804896114e-05, + "loss": 1.1526, + "step": 2903 + }, + { + "epoch": 0.17643842274743302, + "grad_norm": 0.28970614075660706, + "learning_rate": 9.272464613029466e-05, + "loss": 1.1255, + "step": 2904 + }, + { + "epoch": 0.17649917978005955, + "grad_norm": 0.355563759803772, + "learning_rate": 9.271967264554799e-05, + "loss": 1.1557, + "step": 2905 + }, + { + "epoch": 0.17655993681268606, + "grad_norm": 0.30853351950645447, + "learning_rate": 9.271469759490348e-05, + "loss": 1.1829, + "step": 2906 + }, + { + "epoch": 0.1766206938453126, + "grad_norm": 0.3171469569206238, + "learning_rate": 9.270972097854346e-05, + "loss": 1.0895, + "step": 2907 + }, + { + "epoch": 0.17668145087793913, + "grad_norm": 0.6228137612342834, + "learning_rate": 9.270474279665037e-05, + "loss": 1.3167, + "step": 2908 + }, + { + "epoch": 0.17674220791056564, + "grad_norm": 1.7970753908157349, + "learning_rate": 9.269976304940669e-05, + "loss": 1.0707, + "step": 2909 + }, + { + "epoch": 0.17680296494319217, + "grad_norm": 0.2563970983028412, + "learning_rate": 9.269478173699492e-05, + "loss": 1.0796, + "step": 2910 + }, + { + "epoch": 0.1768637219758187, + "grad_norm": 0.1868768036365509, + "learning_rate": 9.268979885959767e-05, + "loss": 1.1227, + "step": 2911 + }, + { + "epoch": 0.17692447900844524, + "grad_norm": 0.2906378209590912, + "learning_rate": 9.26848144173976e-05, + "loss": 1.2465, + "step": 2912 + }, + { + "epoch": 0.17698523604107175, + "grad_norm": 5.847995758056641, + "learning_rate": 9.267982841057743e-05, + "loss": 1.2095, + "step": 2913 + }, + { + "epoch": 0.17704599307369828, + "grad_norm": 0.28656110167503357, + "learning_rate": 9.267484083931987e-05, + "loss": 1.1712, + "step": 2914 + }, + { + "epoch": 0.17710675010632482, + "grad_norm": 2.1816444396972656, + "learning_rate": 9.266985170380777e-05, + "loss": 1.13, + "step": 2915 + }, + { + "epoch": 0.17716750713895132, + "grad_norm": 0.3123765289783478, + "learning_rate": 9.266486100422402e-05, + "loss": 1.0832, + "step": 2916 + }, + { + "epoch": 0.17722826417157786, + "grad_norm": 0.22642475366592407, + "learning_rate": 9.265986874075155e-05, + "loss": 1.164, + "step": 2917 + }, + { + "epoch": 0.1772890212042044, + "grad_norm": 0.3309307396411896, + "learning_rate": 9.265487491357334e-05, + "loss": 1.2249, + "step": 2918 + }, + { + "epoch": 0.1773497782368309, + "grad_norm": 0.2169215828180313, + "learning_rate": 9.264987952287243e-05, + "loss": 1.0856, + "step": 2919 + }, + { + "epoch": 0.17741053526945744, + "grad_norm": 0.2982223629951477, + "learning_rate": 9.264488256883194e-05, + "loss": 1.0666, + "step": 2920 + }, + { + "epoch": 0.17747129230208397, + "grad_norm": 0.5391450524330139, + "learning_rate": 9.263988405163506e-05, + "loss": 1.316, + "step": 2921 + }, + { + "epoch": 0.17753204933471048, + "grad_norm": 0.3179141581058502, + "learning_rate": 9.263488397146498e-05, + "loss": 1.1097, + "step": 2922 + }, + { + "epoch": 0.177592806367337, + "grad_norm": 0.3158731460571289, + "learning_rate": 9.262988232850497e-05, + "loss": 1.1983, + "step": 2923 + }, + { + "epoch": 0.17765356339996355, + "grad_norm": 0.2191188782453537, + "learning_rate": 9.262487912293841e-05, + "loss": 1.1403, + "step": 2924 + }, + { + "epoch": 0.17771432043259008, + "grad_norm": 0.3418218493461609, + "learning_rate": 9.261987435494866e-05, + "loss": 1.2283, + "step": 2925 + }, + { + "epoch": 0.1777750774652166, + "grad_norm": 0.2579060196876526, + "learning_rate": 9.261486802471916e-05, + "loss": 1.1242, + "step": 2926 + }, + { + "epoch": 0.17783583449784313, + "grad_norm": 0.3849816620349884, + "learning_rate": 9.260986013243345e-05, + "loss": 1.1333, + "step": 2927 + }, + { + "epoch": 0.17789659153046966, + "grad_norm": 0.23828557133674622, + "learning_rate": 9.260485067827508e-05, + "loss": 1.0551, + "step": 2928 + }, + { + "epoch": 0.17795734856309617, + "grad_norm": 0.26404693722724915, + "learning_rate": 9.259983966242767e-05, + "loss": 1.1097, + "step": 2929 + }, + { + "epoch": 0.1780181055957227, + "grad_norm": 0.2667824625968933, + "learning_rate": 9.25948270850749e-05, + "loss": 1.2016, + "step": 2930 + }, + { + "epoch": 0.17807886262834924, + "grad_norm": 0.9972683191299438, + "learning_rate": 9.258981294640051e-05, + "loss": 1.199, + "step": 2931 + }, + { + "epoch": 0.17813961966097575, + "grad_norm": 0.20327286422252655, + "learning_rate": 9.25847972465883e-05, + "loss": 1.0904, + "step": 2932 + }, + { + "epoch": 0.17820037669360228, + "grad_norm": 0.22860532999038696, + "learning_rate": 9.25797799858221e-05, + "loss": 1.0835, + "step": 2933 + }, + { + "epoch": 0.17826113372622882, + "grad_norm": 0.4043506979942322, + "learning_rate": 9.257476116428586e-05, + "loss": 1.1605, + "step": 2934 + }, + { + "epoch": 0.17832189075885535, + "grad_norm": 0.3272720277309418, + "learning_rate": 9.256974078216349e-05, + "loss": 1.1074, + "step": 2935 + }, + { + "epoch": 0.17838264779148186, + "grad_norm": 0.41846179962158203, + "learning_rate": 9.256471883963906e-05, + "loss": 1.0393, + "step": 2936 + }, + { + "epoch": 0.1784434048241084, + "grad_norm": 0.7277179956436157, + "learning_rate": 9.255969533689664e-05, + "loss": 1.0552, + "step": 2937 + }, + { + "epoch": 0.17850416185673493, + "grad_norm": 0.37653717398643494, + "learning_rate": 9.255467027412034e-05, + "loss": 1.1695, + "step": 2938 + }, + { + "epoch": 0.17856491888936143, + "grad_norm": 0.361298531293869, + "learning_rate": 9.25496436514944e-05, + "loss": 1.1194, + "step": 2939 + }, + { + "epoch": 0.17862567592198797, + "grad_norm": 0.22349484264850616, + "learning_rate": 9.254461546920303e-05, + "loss": 1.0908, + "step": 2940 + }, + { + "epoch": 0.1786864329546145, + "grad_norm": 0.4283367991447449, + "learning_rate": 9.253958572743058e-05, + "loss": 1.152, + "step": 2941 + }, + { + "epoch": 0.178747189987241, + "grad_norm": 0.22911860048770905, + "learning_rate": 9.253455442636137e-05, + "loss": 1.1286, + "step": 2942 + }, + { + "epoch": 0.17880794701986755, + "grad_norm": 0.269394189119339, + "learning_rate": 9.252952156617985e-05, + "loss": 1.2017, + "step": 2943 + }, + { + "epoch": 0.17886870405249408, + "grad_norm": 0.314388245344162, + "learning_rate": 9.25244871470705e-05, + "loss": 1.276, + "step": 2944 + }, + { + "epoch": 0.1789294610851206, + "grad_norm": 0.19677217304706573, + "learning_rate": 9.251945116921786e-05, + "loss": 1.0816, + "step": 2945 + }, + { + "epoch": 0.17899021811774712, + "grad_norm": 0.21860292553901672, + "learning_rate": 9.25144136328065e-05, + "loss": 1.1207, + "step": 2946 + }, + { + "epoch": 0.17905097515037366, + "grad_norm": 0.2552696466445923, + "learning_rate": 9.250937453802109e-05, + "loss": 1.26, + "step": 2947 + }, + { + "epoch": 0.1791117321830002, + "grad_norm": 0.22064392268657684, + "learning_rate": 9.250433388504635e-05, + "loss": 1.1131, + "step": 2948 + }, + { + "epoch": 0.1791724892156267, + "grad_norm": 0.38646045327186584, + "learning_rate": 9.249929167406703e-05, + "loss": 1.2657, + "step": 2949 + }, + { + "epoch": 0.17923324624825324, + "grad_norm": 0.4657926559448242, + "learning_rate": 9.249424790526797e-05, + "loss": 1.2344, + "step": 2950 + }, + { + "epoch": 0.17929400328087977, + "grad_norm": 0.3978821039199829, + "learning_rate": 9.248920257883404e-05, + "loss": 1.2145, + "step": 2951 + }, + { + "epoch": 0.17935476031350628, + "grad_norm": 0.22265848517417908, + "learning_rate": 9.248415569495016e-05, + "loss": 1.0392, + "step": 2952 + }, + { + "epoch": 0.1794155173461328, + "grad_norm": 0.3151405453681946, + "learning_rate": 9.247910725380136e-05, + "loss": 1.1574, + "step": 2953 + }, + { + "epoch": 0.17947627437875935, + "grad_norm": 0.3059755265712738, + "learning_rate": 9.247405725557266e-05, + "loss": 1.2356, + "step": 2954 + }, + { + "epoch": 0.17953703141138586, + "grad_norm": 0.24969476461410522, + "learning_rate": 9.246900570044917e-05, + "loss": 1.1101, + "step": 2955 + }, + { + "epoch": 0.1795977884440124, + "grad_norm": 0.44281208515167236, + "learning_rate": 9.246395258861609e-05, + "loss": 1.1616, + "step": 2956 + }, + { + "epoch": 0.17965854547663893, + "grad_norm": 0.3311314284801483, + "learning_rate": 9.245889792025862e-05, + "loss": 1.0997, + "step": 2957 + }, + { + "epoch": 0.17971930250926546, + "grad_norm": 0.6681373715400696, + "learning_rate": 9.245384169556202e-05, + "loss": 1.1331, + "step": 2958 + }, + { + "epoch": 0.17978005954189197, + "grad_norm": 0.24887549877166748, + "learning_rate": 9.244878391471166e-05, + "loss": 1.2267, + "step": 2959 + }, + { + "epoch": 0.1798408165745185, + "grad_norm": 0.2625834047794342, + "learning_rate": 9.244372457789291e-05, + "loss": 1.1115, + "step": 2960 + }, + { + "epoch": 0.17990157360714504, + "grad_norm": 0.22148437798023224, + "learning_rate": 9.243866368529124e-05, + "loss": 1.1747, + "step": 2961 + }, + { + "epoch": 0.17996233063977154, + "grad_norm": 0.40260833501815796, + "learning_rate": 9.243360123709215e-05, + "loss": 1.1395, + "step": 2962 + }, + { + "epoch": 0.18002308767239808, + "grad_norm": 0.3440132141113281, + "learning_rate": 9.242853723348119e-05, + "loss": 1.2023, + "step": 2963 + }, + { + "epoch": 0.18008384470502461, + "grad_norm": 0.4685153663158417, + "learning_rate": 9.242347167464402e-05, + "loss": 1.2064, + "step": 2964 + }, + { + "epoch": 0.18014460173765112, + "grad_norm": 0.33053314685821533, + "learning_rate": 9.241840456076628e-05, + "loss": 1.0985, + "step": 2965 + }, + { + "epoch": 0.18020535877027766, + "grad_norm": 0.2586310803890228, + "learning_rate": 9.241333589203372e-05, + "loss": 1.2273, + "step": 2966 + }, + { + "epoch": 0.1802661158029042, + "grad_norm": 0.3619564175605774, + "learning_rate": 9.240826566863214e-05, + "loss": 1.1026, + "step": 2967 + }, + { + "epoch": 0.1803268728355307, + "grad_norm": 0.19196383655071259, + "learning_rate": 9.240319389074739e-05, + "loss": 1.0915, + "step": 2968 + }, + { + "epoch": 0.18038762986815723, + "grad_norm": 0.24559327960014343, + "learning_rate": 9.239812055856536e-05, + "loss": 1.1749, + "step": 2969 + }, + { + "epoch": 0.18044838690078377, + "grad_norm": 0.312445729970932, + "learning_rate": 9.239304567227204e-05, + "loss": 1.0852, + "step": 2970 + }, + { + "epoch": 0.1805091439334103, + "grad_norm": 0.2483806312084198, + "learning_rate": 9.238796923205343e-05, + "loss": 1.1702, + "step": 2971 + }, + { + "epoch": 0.1805699009660368, + "grad_norm": 0.21753054857254028, + "learning_rate": 9.238289123809561e-05, + "loss": 1.1271, + "step": 2972 + }, + { + "epoch": 0.18063065799866335, + "grad_norm": 0.2005295753479004, + "learning_rate": 9.237781169058473e-05, + "loss": 1.2198, + "step": 2973 + }, + { + "epoch": 0.18069141503128988, + "grad_norm": 0.8546209931373596, + "learning_rate": 9.237273058970694e-05, + "loss": 1.1189, + "step": 2974 + }, + { + "epoch": 0.1807521720639164, + "grad_norm": 0.25500378012657166, + "learning_rate": 9.236764793564857e-05, + "loss": 1.1567, + "step": 2975 + }, + { + "epoch": 0.18081292909654292, + "grad_norm": 0.37952733039855957, + "learning_rate": 9.236256372859583e-05, + "loss": 1.2749, + "step": 2976 + }, + { + "epoch": 0.18087368612916946, + "grad_norm": 0.33086255192756653, + "learning_rate": 9.235747796873514e-05, + "loss": 1.1031, + "step": 2977 + }, + { + "epoch": 0.18093444316179597, + "grad_norm": 0.4124637842178345, + "learning_rate": 9.235239065625292e-05, + "loss": 1.0559, + "step": 2978 + }, + { + "epoch": 0.1809952001944225, + "grad_norm": 0.2770545780658722, + "learning_rate": 9.234730179133564e-05, + "loss": 1.0666, + "step": 2979 + }, + { + "epoch": 0.18105595722704904, + "grad_norm": 0.261209636926651, + "learning_rate": 9.234221137416981e-05, + "loss": 1.075, + "step": 2980 + }, + { + "epoch": 0.18111671425967557, + "grad_norm": 0.28440988063812256, + "learning_rate": 9.233711940494204e-05, + "loss": 1.1545, + "step": 2981 + }, + { + "epoch": 0.18117747129230208, + "grad_norm": 0.34159722924232483, + "learning_rate": 9.233202588383897e-05, + "loss": 1.28, + "step": 2982 + }, + { + "epoch": 0.1812382283249286, + "grad_norm": 0.25685015320777893, + "learning_rate": 9.232693081104734e-05, + "loss": 1.1554, + "step": 2983 + }, + { + "epoch": 0.18129898535755515, + "grad_norm": 0.2930513322353363, + "learning_rate": 9.232183418675384e-05, + "loss": 1.1302, + "step": 2984 + }, + { + "epoch": 0.18135974239018166, + "grad_norm": 0.2983180284500122, + "learning_rate": 9.231673601114534e-05, + "loss": 1.0514, + "step": 2985 + }, + { + "epoch": 0.1814204994228082, + "grad_norm": 0.2708021402359009, + "learning_rate": 9.231163628440871e-05, + "loss": 1.2268, + "step": 2986 + }, + { + "epoch": 0.18148125645543473, + "grad_norm": 0.25981590151786804, + "learning_rate": 9.230653500673086e-05, + "loss": 1.0528, + "step": 2987 + }, + { + "epoch": 0.18154201348806123, + "grad_norm": 0.3440215587615967, + "learning_rate": 9.230143217829881e-05, + "loss": 1.246, + "step": 2988 + }, + { + "epoch": 0.18160277052068777, + "grad_norm": 0.2571810483932495, + "learning_rate": 9.229632779929957e-05, + "loss": 1.0907, + "step": 2989 + }, + { + "epoch": 0.1816635275533143, + "grad_norm": 0.3652714192867279, + "learning_rate": 9.229122186992026e-05, + "loss": 1.0445, + "step": 2990 + }, + { + "epoch": 0.1817242845859408, + "grad_norm": 0.38130810856819153, + "learning_rate": 9.228611439034804e-05, + "loss": 1.3221, + "step": 2991 + }, + { + "epoch": 0.18178504161856734, + "grad_norm": 0.24126467108726501, + "learning_rate": 9.228100536077013e-05, + "loss": 1.1244, + "step": 2992 + }, + { + "epoch": 0.18184579865119388, + "grad_norm": 0.27137261629104614, + "learning_rate": 9.22758947813738e-05, + "loss": 1.1247, + "step": 2993 + }, + { + "epoch": 0.18190655568382041, + "grad_norm": 0.20742538571357727, + "learning_rate": 9.227078265234637e-05, + "loss": 1.1118, + "step": 2994 + }, + { + "epoch": 0.18196731271644692, + "grad_norm": 0.6276535987854004, + "learning_rate": 9.226566897387522e-05, + "loss": 1.1214, + "step": 2995 + }, + { + "epoch": 0.18202806974907346, + "grad_norm": 0.2775959372520447, + "learning_rate": 9.226055374614781e-05, + "loss": 1.1517, + "step": 2996 + }, + { + "epoch": 0.1820888267817, + "grad_norm": 0.29811933636665344, + "learning_rate": 9.225543696935164e-05, + "loss": 1.0855, + "step": 2997 + }, + { + "epoch": 0.1821495838143265, + "grad_norm": 0.19156180322170258, + "learning_rate": 9.225031864367425e-05, + "loss": 1.1077, + "step": 2998 + }, + { + "epoch": 0.18221034084695303, + "grad_norm": 0.40932497382164, + "learning_rate": 9.224519876930327e-05, + "loss": 1.2219, + "step": 2999 + }, + { + "epoch": 0.18227109787957957, + "grad_norm": 0.19413401186466217, + "learning_rate": 9.224007734642637e-05, + "loss": 1.0704, + "step": 3000 + }, + { + "epoch": 0.18233185491220608, + "grad_norm": 0.2166336476802826, + "learning_rate": 9.223495437523127e-05, + "loss": 1.12, + "step": 3001 + }, + { + "epoch": 0.1823926119448326, + "grad_norm": 0.30116361379623413, + "learning_rate": 9.222982985590574e-05, + "loss": 1.2718, + "step": 3002 + }, + { + "epoch": 0.18245336897745915, + "grad_norm": 0.312965989112854, + "learning_rate": 9.222470378863763e-05, + "loss": 1.1041, + "step": 3003 + }, + { + "epoch": 0.18251412601008568, + "grad_norm": 0.24075277149677277, + "learning_rate": 9.221957617361485e-05, + "loss": 1.1691, + "step": 3004 + }, + { + "epoch": 0.1825748830427122, + "grad_norm": 0.33824020624160767, + "learning_rate": 9.221444701102534e-05, + "loss": 1.1248, + "step": 3005 + }, + { + "epoch": 0.18263564007533872, + "grad_norm": 0.33058738708496094, + "learning_rate": 9.220931630105712e-05, + "loss": 1.0937, + "step": 3006 + }, + { + "epoch": 0.18269639710796526, + "grad_norm": 0.21481086313724518, + "learning_rate": 9.220418404389826e-05, + "loss": 1.0838, + "step": 3007 + }, + { + "epoch": 0.18275715414059177, + "grad_norm": 0.1858397275209427, + "learning_rate": 9.219905023973687e-05, + "loss": 1.0693, + "step": 3008 + }, + { + "epoch": 0.1828179111732183, + "grad_norm": 0.2046642005443573, + "learning_rate": 9.219391488876112e-05, + "loss": 1.1526, + "step": 3009 + }, + { + "epoch": 0.18287866820584484, + "grad_norm": 0.2653827667236328, + "learning_rate": 9.218877799115928e-05, + "loss": 1.1103, + "step": 3010 + }, + { + "epoch": 0.18293942523847134, + "grad_norm": 0.22898223996162415, + "learning_rate": 9.218363954711964e-05, + "loss": 1.2014, + "step": 3011 + }, + { + "epoch": 0.18300018227109788, + "grad_norm": 0.332379549741745, + "learning_rate": 9.217849955683051e-05, + "loss": 1.2709, + "step": 3012 + }, + { + "epoch": 0.1830609393037244, + "grad_norm": 0.3244803547859192, + "learning_rate": 9.217335802048035e-05, + "loss": 1.1606, + "step": 3013 + }, + { + "epoch": 0.18312169633635092, + "grad_norm": 0.4434811472892761, + "learning_rate": 9.21682149382576e-05, + "loss": 1.1825, + "step": 3014 + }, + { + "epoch": 0.18318245336897745, + "grad_norm": 0.24811863899230957, + "learning_rate": 9.216307031035077e-05, + "loss": 1.0913, + "step": 3015 + }, + { + "epoch": 0.183243210401604, + "grad_norm": 1.922195315361023, + "learning_rate": 9.215792413694844e-05, + "loss": 1.2169, + "step": 3016 + }, + { + "epoch": 0.18330396743423052, + "grad_norm": 0.6606126427650452, + "learning_rate": 9.215277641823925e-05, + "loss": 1.1585, + "step": 3017 + }, + { + "epoch": 0.18336472446685703, + "grad_norm": 0.20095914602279663, + "learning_rate": 9.214762715441192e-05, + "loss": 1.0844, + "step": 3018 + }, + { + "epoch": 0.18342548149948357, + "grad_norm": 0.2916877567768097, + "learning_rate": 9.214247634565514e-05, + "loss": 1.1175, + "step": 3019 + }, + { + "epoch": 0.1834862385321101, + "grad_norm": 0.5368306636810303, + "learning_rate": 9.213732399215776e-05, + "loss": 1.2638, + "step": 3020 + }, + { + "epoch": 0.1835469955647366, + "grad_norm": 0.24616336822509766, + "learning_rate": 9.213217009410862e-05, + "loss": 1.0865, + "step": 3021 + }, + { + "epoch": 0.18360775259736314, + "grad_norm": 0.19987960159778595, + "learning_rate": 9.212701465169664e-05, + "loss": 1.1453, + "step": 3022 + }, + { + "epoch": 0.18366850962998968, + "grad_norm": 0.18338149785995483, + "learning_rate": 9.21218576651108e-05, + "loss": 1.1499, + "step": 3023 + }, + { + "epoch": 0.1837292666626162, + "grad_norm": 0.22008994221687317, + "learning_rate": 9.211669913454013e-05, + "loss": 1.0804, + "step": 3024 + }, + { + "epoch": 0.18379002369524272, + "grad_norm": 0.24982674419879913, + "learning_rate": 9.21115390601737e-05, + "loss": 1.1745, + "step": 3025 + }, + { + "epoch": 0.18385078072786926, + "grad_norm": 0.20403137803077698, + "learning_rate": 9.210637744220067e-05, + "loss": 1.0542, + "step": 3026 + }, + { + "epoch": 0.18391153776049576, + "grad_norm": 0.37922266125679016, + "learning_rate": 9.210121428081022e-05, + "loss": 1.3125, + "step": 3027 + }, + { + "epoch": 0.1839722947931223, + "grad_norm": 0.359588623046875, + "learning_rate": 9.209604957619164e-05, + "loss": 1.0706, + "step": 3028 + }, + { + "epoch": 0.18403305182574883, + "grad_norm": 0.23511454463005066, + "learning_rate": 9.209088332853424e-05, + "loss": 1.1375, + "step": 3029 + }, + { + "epoch": 0.18409380885837537, + "grad_norm": 0.2682068943977356, + "learning_rate": 9.208571553802734e-05, + "loss": 1.1186, + "step": 3030 + }, + { + "epoch": 0.18415456589100188, + "grad_norm": 0.35418057441711426, + "learning_rate": 9.208054620486043e-05, + "loss": 1.1697, + "step": 3031 + }, + { + "epoch": 0.1842153229236284, + "grad_norm": 0.2783893048763275, + "learning_rate": 9.207537532922296e-05, + "loss": 1.2095, + "step": 3032 + }, + { + "epoch": 0.18427607995625495, + "grad_norm": 0.39824506640434265, + "learning_rate": 9.207020291130448e-05, + "loss": 1.2493, + "step": 3033 + }, + { + "epoch": 0.18433683698888145, + "grad_norm": 0.2556644678115845, + "learning_rate": 9.206502895129456e-05, + "loss": 1.1094, + "step": 3034 + }, + { + "epoch": 0.184397594021508, + "grad_norm": 0.21846148371696472, + "learning_rate": 9.205985344938288e-05, + "loss": 1.1063, + "step": 3035 + }, + { + "epoch": 0.18445835105413452, + "grad_norm": 0.25165486335754395, + "learning_rate": 9.205467640575916e-05, + "loss": 1.1748, + "step": 3036 + }, + { + "epoch": 0.18451910808676103, + "grad_norm": 0.20205509662628174, + "learning_rate": 9.204949782061311e-05, + "loss": 1.04, + "step": 3037 + }, + { + "epoch": 0.18457986511938757, + "grad_norm": 0.2803945541381836, + "learning_rate": 9.204431769413462e-05, + "loss": 1.2595, + "step": 3038 + }, + { + "epoch": 0.1846406221520141, + "grad_norm": 0.20643943548202515, + "learning_rate": 9.203913602651354e-05, + "loss": 1.1042, + "step": 3039 + }, + { + "epoch": 0.18470137918464063, + "grad_norm": 0.20411239564418793, + "learning_rate": 9.203395281793979e-05, + "loss": 1.0748, + "step": 3040 + }, + { + "epoch": 0.18476213621726714, + "grad_norm": 0.38612937927246094, + "learning_rate": 9.202876806860338e-05, + "loss": 1.0963, + "step": 3041 + }, + { + "epoch": 0.18482289324989368, + "grad_norm": 0.2762918174266815, + "learning_rate": 9.202358177869435e-05, + "loss": 1.2309, + "step": 3042 + }, + { + "epoch": 0.1848836502825202, + "grad_norm": 0.922309398651123, + "learning_rate": 9.20183939484028e-05, + "loss": 1.1853, + "step": 3043 + }, + { + "epoch": 0.18494440731514672, + "grad_norm": 0.3634815216064453, + "learning_rate": 9.201320457791891e-05, + "loss": 1.0824, + "step": 3044 + }, + { + "epoch": 0.18500516434777325, + "grad_norm": 0.35319578647613525, + "learning_rate": 9.200801366743289e-05, + "loss": 1.1241, + "step": 3045 + }, + { + "epoch": 0.1850659213803998, + "grad_norm": 0.35610246658325195, + "learning_rate": 9.200282121713499e-05, + "loss": 1.2011, + "step": 3046 + }, + { + "epoch": 0.1851266784130263, + "grad_norm": 0.240784153342247, + "learning_rate": 9.199762722721557e-05, + "loss": 1.0768, + "step": 3047 + }, + { + "epoch": 0.18518743544565283, + "grad_norm": 3.135982036590576, + "learning_rate": 9.1992431697865e-05, + "loss": 1.3197, + "step": 3048 + }, + { + "epoch": 0.18524819247827937, + "grad_norm": 0.3077400326728821, + "learning_rate": 9.198723462927375e-05, + "loss": 1.0704, + "step": 3049 + }, + { + "epoch": 0.18530894951090587, + "grad_norm": 0.19032785296440125, + "learning_rate": 9.198203602163227e-05, + "loss": 1.0687, + "step": 3050 + }, + { + "epoch": 0.1853697065435324, + "grad_norm": 0.26754164695739746, + "learning_rate": 9.197683587513116e-05, + "loss": 1.1909, + "step": 3051 + }, + { + "epoch": 0.18543046357615894, + "grad_norm": 0.2467351108789444, + "learning_rate": 9.197163418996101e-05, + "loss": 1.1021, + "step": 3052 + }, + { + "epoch": 0.18549122060878548, + "grad_norm": 0.23067933320999146, + "learning_rate": 9.19664309663125e-05, + "loss": 1.0898, + "step": 3053 + }, + { + "epoch": 0.18555197764141199, + "grad_norm": 0.412874311208725, + "learning_rate": 9.196122620437635e-05, + "loss": 1.0749, + "step": 3054 + }, + { + "epoch": 0.18561273467403852, + "grad_norm": 0.257097452878952, + "learning_rate": 9.195601990434335e-05, + "loss": 1.2153, + "step": 3055 + }, + { + "epoch": 0.18567349170666506, + "grad_norm": 0.32998475432395935, + "learning_rate": 9.195081206640433e-05, + "loss": 1.1933, + "step": 3056 + }, + { + "epoch": 0.18573424873929156, + "grad_norm": 0.3984575867652893, + "learning_rate": 9.194560269075016e-05, + "loss": 1.0371, + "step": 3057 + }, + { + "epoch": 0.1857950057719181, + "grad_norm": 0.19496065378189087, + "learning_rate": 9.194039177757184e-05, + "loss": 1.0841, + "step": 3058 + }, + { + "epoch": 0.18585576280454463, + "grad_norm": 0.3898847699165344, + "learning_rate": 9.193517932706036e-05, + "loss": 1.0971, + "step": 3059 + }, + { + "epoch": 0.18591651983717114, + "grad_norm": 0.48681244254112244, + "learning_rate": 9.192996533940675e-05, + "loss": 1.12, + "step": 3060 + }, + { + "epoch": 0.18597727686979768, + "grad_norm": 0.33356973528862, + "learning_rate": 9.192474981480216e-05, + "loss": 1.1273, + "step": 3061 + }, + { + "epoch": 0.1860380339024242, + "grad_norm": 0.28480425477027893, + "learning_rate": 9.191953275343779e-05, + "loss": 1.3238, + "step": 3062 + }, + { + "epoch": 0.18609879093505075, + "grad_norm": 0.204045370221138, + "learning_rate": 9.191431415550482e-05, + "loss": 1.0937, + "step": 3063 + }, + { + "epoch": 0.18615954796767725, + "grad_norm": 0.20480507612228394, + "learning_rate": 9.190909402119456e-05, + "loss": 1.1093, + "step": 3064 + }, + { + "epoch": 0.1862203050003038, + "grad_norm": 7.3999924659729, + "learning_rate": 9.190387235069837e-05, + "loss": 1.1756, + "step": 3065 + }, + { + "epoch": 0.18628106203293032, + "grad_norm": 0.23585659265518188, + "learning_rate": 9.189864914420763e-05, + "loss": 1.0623, + "step": 3066 + }, + { + "epoch": 0.18634181906555683, + "grad_norm": 0.2400611788034439, + "learning_rate": 9.189342440191382e-05, + "loss": 1.1219, + "step": 3067 + }, + { + "epoch": 0.18640257609818336, + "grad_norm": 1.0079388618469238, + "learning_rate": 9.188819812400842e-05, + "loss": 1.1897, + "step": 3068 + }, + { + "epoch": 0.1864633331308099, + "grad_norm": 0.48840805888175964, + "learning_rate": 9.188297031068304e-05, + "loss": 1.0651, + "step": 3069 + }, + { + "epoch": 0.1865240901634364, + "grad_norm": 0.22186236083507538, + "learning_rate": 9.187774096212926e-05, + "loss": 1.1873, + "step": 3070 + }, + { + "epoch": 0.18658484719606294, + "grad_norm": 0.22673046588897705, + "learning_rate": 9.187251007853883e-05, + "loss": 1.2941, + "step": 3071 + }, + { + "epoch": 0.18664560422868948, + "grad_norm": 0.5850257873535156, + "learning_rate": 9.186727766010343e-05, + "loss": 1.1787, + "step": 3072 + }, + { + "epoch": 0.18670636126131598, + "grad_norm": 0.276504248380661, + "learning_rate": 9.186204370701486e-05, + "loss": 1.1486, + "step": 3073 + }, + { + "epoch": 0.18676711829394252, + "grad_norm": 0.3929198980331421, + "learning_rate": 9.185680821946501e-05, + "loss": 1.3315, + "step": 3074 + }, + { + "epoch": 0.18682787532656905, + "grad_norm": 1.8744144439697266, + "learning_rate": 9.185157119764575e-05, + "loss": 1.1204, + "step": 3075 + }, + { + "epoch": 0.1868886323591956, + "grad_norm": 0.6167188882827759, + "learning_rate": 9.184633264174907e-05, + "loss": 1.2433, + "step": 3076 + }, + { + "epoch": 0.1869493893918221, + "grad_norm": 0.3212830424308777, + "learning_rate": 9.184109255196696e-05, + "loss": 1.1149, + "step": 3077 + }, + { + "epoch": 0.18701014642444863, + "grad_norm": 0.30843648314476013, + "learning_rate": 9.183585092849153e-05, + "loss": 1.1425, + "step": 3078 + }, + { + "epoch": 0.18707090345707517, + "grad_norm": 0.3646651804447174, + "learning_rate": 9.183060777151489e-05, + "loss": 1.0822, + "step": 3079 + }, + { + "epoch": 0.18713166048970167, + "grad_norm": 0.5849332809448242, + "learning_rate": 9.182536308122925e-05, + "loss": 1.0993, + "step": 3080 + }, + { + "epoch": 0.1871924175223282, + "grad_norm": 0.32188504934310913, + "learning_rate": 9.182011685782683e-05, + "loss": 1.0704, + "step": 3081 + }, + { + "epoch": 0.18725317455495474, + "grad_norm": 0.23076030611991882, + "learning_rate": 9.181486910149994e-05, + "loss": 1.0792, + "step": 3082 + }, + { + "epoch": 0.18731393158758125, + "grad_norm": 0.25014469027519226, + "learning_rate": 9.180961981244094e-05, + "loss": 1.1064, + "step": 3083 + }, + { + "epoch": 0.18737468862020779, + "grad_norm": 0.2624911069869995, + "learning_rate": 9.180436899084225e-05, + "loss": 1.1761, + "step": 3084 + }, + { + "epoch": 0.18743544565283432, + "grad_norm": 0.2671333849430084, + "learning_rate": 9.179911663689633e-05, + "loss": 1.2024, + "step": 3085 + }, + { + "epoch": 0.18749620268546086, + "grad_norm": 0.25764450430870056, + "learning_rate": 9.179386275079571e-05, + "loss": 1.2435, + "step": 3086 + }, + { + "epoch": 0.18755695971808736, + "grad_norm": 0.19135116040706635, + "learning_rate": 9.178860733273298e-05, + "loss": 1.1005, + "step": 3087 + }, + { + "epoch": 0.1876177167507139, + "grad_norm": 0.19494393467903137, + "learning_rate": 9.178335038290075e-05, + "loss": 1.1071, + "step": 3088 + }, + { + "epoch": 0.18767847378334043, + "grad_norm": 0.2702561914920807, + "learning_rate": 9.177809190149176e-05, + "loss": 1.1844, + "step": 3089 + }, + { + "epoch": 0.18773923081596694, + "grad_norm": 0.27724677324295044, + "learning_rate": 9.177283188869873e-05, + "loss": 1.0776, + "step": 3090 + }, + { + "epoch": 0.18779998784859347, + "grad_norm": 0.2412506341934204, + "learning_rate": 9.176757034471447e-05, + "loss": 1.1612, + "step": 3091 + }, + { + "epoch": 0.18786074488122, + "grad_norm": 0.3440745770931244, + "learning_rate": 9.176230726973185e-05, + "loss": 1.1487, + "step": 3092 + }, + { + "epoch": 0.18792150191384652, + "grad_norm": 0.46961691975593567, + "learning_rate": 9.175704266394378e-05, + "loss": 1.1377, + "step": 3093 + }, + { + "epoch": 0.18798225894647305, + "grad_norm": 0.18857328593730927, + "learning_rate": 9.175177652754323e-05, + "loss": 1.0811, + "step": 3094 + }, + { + "epoch": 0.1880430159790996, + "grad_norm": 0.28688573837280273, + "learning_rate": 9.174650886072325e-05, + "loss": 1.1707, + "step": 3095 + }, + { + "epoch": 0.1881037730117261, + "grad_norm": 0.24852831661701202, + "learning_rate": 9.174123966367691e-05, + "loss": 1.2269, + "step": 3096 + }, + { + "epoch": 0.18816453004435263, + "grad_norm": 0.294095516204834, + "learning_rate": 9.173596893659738e-05, + "loss": 1.2465, + "step": 3097 + }, + { + "epoch": 0.18822528707697916, + "grad_norm": 0.4267716705799103, + "learning_rate": 9.173069667967785e-05, + "loss": 1.1071, + "step": 3098 + }, + { + "epoch": 0.1882860441096057, + "grad_norm": 0.2354598045349121, + "learning_rate": 9.172542289311153e-05, + "loss": 1.0729, + "step": 3099 + }, + { + "epoch": 0.1883468011422322, + "grad_norm": 0.28527241945266724, + "learning_rate": 9.172014757709179e-05, + "loss": 1.1577, + "step": 3100 + }, + { + "epoch": 0.18840755817485874, + "grad_norm": 0.25236281752586365, + "learning_rate": 9.171487073181198e-05, + "loss": 1.1125, + "step": 3101 + }, + { + "epoch": 0.18846831520748528, + "grad_norm": 0.21823987364768982, + "learning_rate": 9.17095923574655e-05, + "loss": 1.0506, + "step": 3102 + }, + { + "epoch": 0.18852907224011178, + "grad_norm": 0.27860915660858154, + "learning_rate": 9.170431245424587e-05, + "loss": 1.095, + "step": 3103 + }, + { + "epoch": 0.18858982927273832, + "grad_norm": 0.2313152253627777, + "learning_rate": 9.169903102234659e-05, + "loss": 1.1888, + "step": 3104 + }, + { + "epoch": 0.18865058630536485, + "grad_norm": 0.2475203275680542, + "learning_rate": 9.169374806196127e-05, + "loss": 1.1679, + "step": 3105 + }, + { + "epoch": 0.18871134333799136, + "grad_norm": 0.2492544800043106, + "learning_rate": 9.168846357328357e-05, + "loss": 1.1274, + "step": 3106 + }, + { + "epoch": 0.1887721003706179, + "grad_norm": 0.3743326663970947, + "learning_rate": 9.168317755650717e-05, + "loss": 1.3735, + "step": 3107 + }, + { + "epoch": 0.18883285740324443, + "grad_norm": 0.33873629570007324, + "learning_rate": 9.167789001182584e-05, + "loss": 1.1942, + "step": 3108 + }, + { + "epoch": 0.18889361443587097, + "grad_norm": 0.21263690292835236, + "learning_rate": 9.16726009394334e-05, + "loss": 1.0972, + "step": 3109 + }, + { + "epoch": 0.18895437146849747, + "grad_norm": 0.20669060945510864, + "learning_rate": 9.166731033952372e-05, + "loss": 1.107, + "step": 3110 + }, + { + "epoch": 0.189015128501124, + "grad_norm": 0.19627468287944794, + "learning_rate": 9.166201821229072e-05, + "loss": 1.0292, + "step": 3111 + }, + { + "epoch": 0.18907588553375054, + "grad_norm": 0.4125955402851105, + "learning_rate": 9.165672455792838e-05, + "loss": 1.0659, + "step": 3112 + }, + { + "epoch": 0.18913664256637705, + "grad_norm": 0.37949925661087036, + "learning_rate": 9.165142937663076e-05, + "loss": 1.1208, + "step": 3113 + }, + { + "epoch": 0.18919739959900359, + "grad_norm": 0.4148748815059662, + "learning_rate": 9.164613266859193e-05, + "loss": 1.2435, + "step": 3114 + }, + { + "epoch": 0.18925815663163012, + "grad_norm": 0.3271282911300659, + "learning_rate": 9.164083443400609e-05, + "loss": 1.0518, + "step": 3115 + }, + { + "epoch": 0.18931891366425663, + "grad_norm": 1.2755006551742554, + "learning_rate": 9.163553467306738e-05, + "loss": 1.2035, + "step": 3116 + }, + { + "epoch": 0.18937967069688316, + "grad_norm": 0.23444201052188873, + "learning_rate": 9.163023338597011e-05, + "loss": 1.1189, + "step": 3117 + }, + { + "epoch": 0.1894404277295097, + "grad_norm": 0.32523980736732483, + "learning_rate": 9.16249305729086e-05, + "loss": 1.1672, + "step": 3118 + }, + { + "epoch": 0.1895011847621362, + "grad_norm": 0.22614625096321106, + "learning_rate": 9.16196262340772e-05, + "loss": 1.0813, + "step": 3119 + }, + { + "epoch": 0.18956194179476274, + "grad_norm": 0.2449781745672226, + "learning_rate": 9.161432036967036e-05, + "loss": 1.0758, + "step": 3120 + }, + { + "epoch": 0.18962269882738927, + "grad_norm": 0.39479583501815796, + "learning_rate": 9.160901297988256e-05, + "loss": 1.1555, + "step": 3121 + }, + { + "epoch": 0.1896834558600158, + "grad_norm": 0.40087559819221497, + "learning_rate": 9.160370406490834e-05, + "loss": 1.111, + "step": 3122 + }, + { + "epoch": 0.18974421289264232, + "grad_norm": 0.35269734263420105, + "learning_rate": 9.159839362494232e-05, + "loss": 1.2063, + "step": 3123 + }, + { + "epoch": 0.18980496992526885, + "grad_norm": 1.941612720489502, + "learning_rate": 9.159308166017915e-05, + "loss": 1.1286, + "step": 3124 + }, + { + "epoch": 0.1898657269578954, + "grad_norm": 0.4391666650772095, + "learning_rate": 9.158776817081351e-05, + "loss": 1.1463, + "step": 3125 + }, + { + "epoch": 0.1899264839905219, + "grad_norm": 0.26761773228645325, + "learning_rate": 9.158245315704019e-05, + "loss": 1.0995, + "step": 3126 + }, + { + "epoch": 0.18998724102314843, + "grad_norm": 0.2000022530555725, + "learning_rate": 9.157713661905402e-05, + "loss": 1.081, + "step": 3127 + }, + { + "epoch": 0.19004799805577496, + "grad_norm": 0.3460334539413452, + "learning_rate": 9.157181855704988e-05, + "loss": 1.1264, + "step": 3128 + }, + { + "epoch": 0.19010875508840147, + "grad_norm": 11.452260971069336, + "learning_rate": 9.15664989712227e-05, + "loss": 1.1138, + "step": 3129 + }, + { + "epoch": 0.190169512121028, + "grad_norm": 0.306286484003067, + "learning_rate": 9.156117786176743e-05, + "loss": 1.1919, + "step": 3130 + }, + { + "epoch": 0.19023026915365454, + "grad_norm": 0.2920655906200409, + "learning_rate": 9.155585522887917e-05, + "loss": 1.1816, + "step": 3131 + }, + { + "epoch": 0.19029102618628108, + "grad_norm": 0.2891019582748413, + "learning_rate": 9.155053107275301e-05, + "loss": 1.0875, + "step": 3132 + }, + { + "epoch": 0.19035178321890758, + "grad_norm": 0.19459807872772217, + "learning_rate": 9.15452053935841e-05, + "loss": 1.0916, + "step": 3133 + }, + { + "epoch": 0.19041254025153412, + "grad_norm": 0.20118562877178192, + "learning_rate": 9.153987819156766e-05, + "loss": 1.0879, + "step": 3134 + }, + { + "epoch": 0.19047329728416065, + "grad_norm": 0.36979541182518005, + "learning_rate": 9.153454946689895e-05, + "loss": 1.0483, + "step": 3135 + }, + { + "epoch": 0.19053405431678716, + "grad_norm": 0.22860446572303772, + "learning_rate": 9.152921921977329e-05, + "loss": 1.1553, + "step": 3136 + }, + { + "epoch": 0.1905948113494137, + "grad_norm": 0.30773770809173584, + "learning_rate": 9.15238874503861e-05, + "loss": 1.136, + "step": 3137 + }, + { + "epoch": 0.19065556838204023, + "grad_norm": 0.31998884677886963, + "learning_rate": 9.151855415893276e-05, + "loss": 1.0933, + "step": 3138 + }, + { + "epoch": 0.19071632541466674, + "grad_norm": 0.19996775686740875, + "learning_rate": 9.15132193456088e-05, + "loss": 1.0725, + "step": 3139 + }, + { + "epoch": 0.19077708244729327, + "grad_norm": 0.3007640838623047, + "learning_rate": 9.150788301060977e-05, + "loss": 1.1903, + "step": 3140 + }, + { + "epoch": 0.1908378394799198, + "grad_norm": 0.2727598249912262, + "learning_rate": 9.150254515413126e-05, + "loss": 1.2078, + "step": 3141 + }, + { + "epoch": 0.19089859651254631, + "grad_norm": 0.35017552971839905, + "learning_rate": 9.149720577636893e-05, + "loss": 1.105, + "step": 3142 + }, + { + "epoch": 0.19095935354517285, + "grad_norm": 0.215254545211792, + "learning_rate": 9.14918648775185e-05, + "loss": 1.1391, + "step": 3143 + }, + { + "epoch": 0.19102011057779938, + "grad_norm": 0.3343048691749573, + "learning_rate": 9.148652245777576e-05, + "loss": 1.0772, + "step": 3144 + }, + { + "epoch": 0.19108086761042592, + "grad_norm": 0.22141067683696747, + "learning_rate": 9.14811785173365e-05, + "loss": 1.3398, + "step": 3145 + }, + { + "epoch": 0.19114162464305243, + "grad_norm": 0.2885318398475647, + "learning_rate": 9.147583305639663e-05, + "loss": 1.1084, + "step": 3146 + }, + { + "epoch": 0.19120238167567896, + "grad_norm": 0.33269333839416504, + "learning_rate": 9.14704860751521e-05, + "loss": 1.0424, + "step": 3147 + }, + { + "epoch": 0.1912631387083055, + "grad_norm": 0.30590111017227173, + "learning_rate": 9.146513757379886e-05, + "loss": 1.1113, + "step": 3148 + }, + { + "epoch": 0.191323895740932, + "grad_norm": 0.5303155779838562, + "learning_rate": 9.145978755253299e-05, + "loss": 1.2139, + "step": 3149 + }, + { + "epoch": 0.19138465277355854, + "grad_norm": 0.27111369371414185, + "learning_rate": 9.14544360115506e-05, + "loss": 1.0601, + "step": 3150 + }, + { + "epoch": 0.19144540980618507, + "grad_norm": 0.42402365803718567, + "learning_rate": 9.144908295104786e-05, + "loss": 1.1374, + "step": 3151 + }, + { + "epoch": 0.19150616683881158, + "grad_norm": 0.29252055287361145, + "learning_rate": 9.144372837122095e-05, + "loss": 1.1008, + "step": 3152 + }, + { + "epoch": 0.19156692387143812, + "grad_norm": 0.23134812712669373, + "learning_rate": 9.143837227226617e-05, + "loss": 1.2369, + "step": 3153 + }, + { + "epoch": 0.19162768090406465, + "grad_norm": 0.34692153334617615, + "learning_rate": 9.143301465437986e-05, + "loss": 1.0754, + "step": 3154 + }, + { + "epoch": 0.19168843793669116, + "grad_norm": 0.27057382464408875, + "learning_rate": 9.142765551775838e-05, + "loss": 1.1755, + "step": 3155 + }, + { + "epoch": 0.1917491949693177, + "grad_norm": 0.35990026593208313, + "learning_rate": 9.142229486259817e-05, + "loss": 1.2807, + "step": 3156 + }, + { + "epoch": 0.19180995200194423, + "grad_norm": 0.4851399064064026, + "learning_rate": 9.141693268909575e-05, + "loss": 1.0918, + "step": 3157 + }, + { + "epoch": 0.19187070903457076, + "grad_norm": 0.2555850148200989, + "learning_rate": 9.141156899744764e-05, + "loss": 1.1802, + "step": 3158 + }, + { + "epoch": 0.19193146606719727, + "grad_norm": 0.4629479646682739, + "learning_rate": 9.140620378785047e-05, + "loss": 1.1838, + "step": 3159 + }, + { + "epoch": 0.1919922230998238, + "grad_norm": 0.388776570558548, + "learning_rate": 9.14008370605009e-05, + "loss": 1.1509, + "step": 3160 + }, + { + "epoch": 0.19205298013245034, + "grad_norm": 0.20831067860126495, + "learning_rate": 9.139546881559564e-05, + "loss": 1.0976, + "step": 3161 + }, + { + "epoch": 0.19211373716507685, + "grad_norm": 0.3283165395259857, + "learning_rate": 9.139009905333147e-05, + "loss": 1.0526, + "step": 3162 + }, + { + "epoch": 0.19217449419770338, + "grad_norm": 0.22886575758457184, + "learning_rate": 9.138472777390523e-05, + "loss": 1.2076, + "step": 3163 + }, + { + "epoch": 0.19223525123032992, + "grad_norm": 1.1828349828720093, + "learning_rate": 9.137935497751378e-05, + "loss": 1.1845, + "step": 3164 + }, + { + "epoch": 0.19229600826295642, + "grad_norm": 0.41144421696662903, + "learning_rate": 9.137398066435409e-05, + "loss": 1.1239, + "step": 3165 + }, + { + "epoch": 0.19235676529558296, + "grad_norm": 0.21680109202861786, + "learning_rate": 9.136860483462312e-05, + "loss": 1.0721, + "step": 3166 + }, + { + "epoch": 0.1924175223282095, + "grad_norm": 0.31411242485046387, + "learning_rate": 9.136322748851796e-05, + "loss": 1.0688, + "step": 3167 + }, + { + "epoch": 0.19247827936083603, + "grad_norm": 0.28844329714775085, + "learning_rate": 9.135784862623569e-05, + "loss": 1.0583, + "step": 3168 + }, + { + "epoch": 0.19253903639346254, + "grad_norm": 0.7784070372581482, + "learning_rate": 9.13524682479735e-05, + "loss": 1.1365, + "step": 3169 + }, + { + "epoch": 0.19259979342608907, + "grad_norm": 0.27679479122161865, + "learning_rate": 9.134708635392858e-05, + "loss": 1.1991, + "step": 3170 + }, + { + "epoch": 0.1926605504587156, + "grad_norm": 0.2529323101043701, + "learning_rate": 9.134170294429822e-05, + "loss": 1.1389, + "step": 3171 + }, + { + "epoch": 0.19272130749134211, + "grad_norm": 0.49346330761909485, + "learning_rate": 9.133631801927975e-05, + "loss": 1.1419, + "step": 3172 + }, + { + "epoch": 0.19278206452396865, + "grad_norm": 0.2782764434814453, + "learning_rate": 9.133093157907055e-05, + "loss": 1.1762, + "step": 3173 + }, + { + "epoch": 0.19284282155659518, + "grad_norm": 0.28708550333976746, + "learning_rate": 9.132554362386807e-05, + "loss": 1.2153, + "step": 3174 + }, + { + "epoch": 0.1929035785892217, + "grad_norm": 0.22017507255077362, + "learning_rate": 9.132015415386979e-05, + "loss": 1.046, + "step": 3175 + }, + { + "epoch": 0.19296433562184823, + "grad_norm": 0.5357812643051147, + "learning_rate": 9.131476316927327e-05, + "loss": 1.2833, + "step": 3176 + }, + { + "epoch": 0.19302509265447476, + "grad_norm": 0.3946494460105896, + "learning_rate": 9.130937067027615e-05, + "loss": 1.2065, + "step": 3177 + }, + { + "epoch": 0.19308584968710127, + "grad_norm": 0.2677405774593353, + "learning_rate": 9.130397665707603e-05, + "loss": 1.1568, + "step": 3178 + }, + { + "epoch": 0.1931466067197278, + "grad_norm": 0.20055626332759857, + "learning_rate": 9.129858112987066e-05, + "loss": 1.0562, + "step": 3179 + }, + { + "epoch": 0.19320736375235434, + "grad_norm": 0.18907460570335388, + "learning_rate": 9.129318408885784e-05, + "loss": 1.0975, + "step": 3180 + }, + { + "epoch": 0.19326812078498087, + "grad_norm": 0.46790555119514465, + "learning_rate": 9.128778553423536e-05, + "loss": 1.1316, + "step": 3181 + }, + { + "epoch": 0.19332887781760738, + "grad_norm": 0.4143708050251007, + "learning_rate": 9.128238546620114e-05, + "loss": 1.2114, + "step": 3182 + }, + { + "epoch": 0.19338963485023392, + "grad_norm": 0.23438739776611328, + "learning_rate": 9.127698388495307e-05, + "loss": 1.094, + "step": 3183 + }, + { + "epoch": 0.19345039188286045, + "grad_norm": 0.2291349470615387, + "learning_rate": 9.12715807906892e-05, + "loss": 1.1829, + "step": 3184 + }, + { + "epoch": 0.19351114891548696, + "grad_norm": 0.3957231044769287, + "learning_rate": 9.126617618360756e-05, + "loss": 1.2351, + "step": 3185 + }, + { + "epoch": 0.1935719059481135, + "grad_norm": 0.20947127044200897, + "learning_rate": 9.126077006390623e-05, + "loss": 1.0753, + "step": 3186 + }, + { + "epoch": 0.19363266298074003, + "grad_norm": 0.21727082133293152, + "learning_rate": 9.125536243178342e-05, + "loss": 1.1708, + "step": 3187 + }, + { + "epoch": 0.19369342001336654, + "grad_norm": 0.31573453545570374, + "learning_rate": 9.124995328743732e-05, + "loss": 1.2129, + "step": 3188 + }, + { + "epoch": 0.19375417704599307, + "grad_norm": 0.25833621621131897, + "learning_rate": 9.124454263106622e-05, + "loss": 1.1221, + "step": 3189 + }, + { + "epoch": 0.1938149340786196, + "grad_norm": 0.2905177175998688, + "learning_rate": 9.123913046286841e-05, + "loss": 1.2084, + "step": 3190 + }, + { + "epoch": 0.19387569111124614, + "grad_norm": 0.30853450298309326, + "learning_rate": 9.123371678304232e-05, + "loss": 1.102, + "step": 3191 + }, + { + "epoch": 0.19393644814387265, + "grad_norm": 0.36089417338371277, + "learning_rate": 9.122830159178636e-05, + "loss": 1.1366, + "step": 3192 + }, + { + "epoch": 0.19399720517649918, + "grad_norm": 0.2385888397693634, + "learning_rate": 9.122288488929906e-05, + "loss": 1.1033, + "step": 3193 + }, + { + "epoch": 0.19405796220912572, + "grad_norm": 0.3437388837337494, + "learning_rate": 9.121746667577893e-05, + "loss": 1.0997, + "step": 3194 + }, + { + "epoch": 0.19411871924175222, + "grad_norm": 0.2662302553653717, + "learning_rate": 9.12120469514246e-05, + "loss": 1.1157, + "step": 3195 + }, + { + "epoch": 0.19417947627437876, + "grad_norm": 0.3347611725330353, + "learning_rate": 9.12066257164347e-05, + "loss": 1.1859, + "step": 3196 + }, + { + "epoch": 0.1942402333070053, + "grad_norm": 0.26880061626434326, + "learning_rate": 9.120120297100799e-05, + "loss": 1.1716, + "step": 3197 + }, + { + "epoch": 0.1943009903396318, + "grad_norm": 0.37834835052490234, + "learning_rate": 9.119577871534322e-05, + "loss": 1.0714, + "step": 3198 + }, + { + "epoch": 0.19436174737225834, + "grad_norm": 0.2542210817337036, + "learning_rate": 9.119035294963921e-05, + "loss": 1.1337, + "step": 3199 + }, + { + "epoch": 0.19442250440488487, + "grad_norm": 0.3667239844799042, + "learning_rate": 9.118492567409486e-05, + "loss": 1.1847, + "step": 3200 + }, + { + "epoch": 0.19448326143751138, + "grad_norm": 0.266865998506546, + "learning_rate": 9.117949688890909e-05, + "loss": 1.2782, + "step": 3201 + }, + { + "epoch": 0.1945440184701379, + "grad_norm": 0.27282825112342834, + "learning_rate": 9.11740665942809e-05, + "loss": 1.0828, + "step": 3202 + }, + { + "epoch": 0.19460477550276445, + "grad_norm": 0.4177188575267792, + "learning_rate": 9.116863479040936e-05, + "loss": 1.1004, + "step": 3203 + }, + { + "epoch": 0.19466553253539098, + "grad_norm": 0.27978628873825073, + "learning_rate": 9.116320147749354e-05, + "loss": 1.1218, + "step": 3204 + }, + { + "epoch": 0.1947262895680175, + "grad_norm": 0.2817535698413849, + "learning_rate": 9.11577666557326e-05, + "loss": 1.0811, + "step": 3205 + }, + { + "epoch": 0.19478704660064403, + "grad_norm": 0.2331988364458084, + "learning_rate": 9.115233032532578e-05, + "loss": 1.0732, + "step": 3206 + }, + { + "epoch": 0.19484780363327056, + "grad_norm": 0.22885434329509735, + "learning_rate": 9.114689248647235e-05, + "loss": 1.1464, + "step": 3207 + }, + { + "epoch": 0.19490856066589707, + "grad_norm": 0.26414093375205994, + "learning_rate": 9.114145313937161e-05, + "loss": 1.0937, + "step": 3208 + }, + { + "epoch": 0.1949693176985236, + "grad_norm": 1.745571255683899, + "learning_rate": 9.113601228422294e-05, + "loss": 1.2242, + "step": 3209 + }, + { + "epoch": 0.19503007473115014, + "grad_norm": 0.2982587218284607, + "learning_rate": 9.11305699212258e-05, + "loss": 1.1938, + "step": 3210 + }, + { + "epoch": 0.19509083176377665, + "grad_norm": 0.2844155430793762, + "learning_rate": 9.112512605057968e-05, + "loss": 1.107, + "step": 3211 + }, + { + "epoch": 0.19515158879640318, + "grad_norm": 3.100975513458252, + "learning_rate": 9.111968067248411e-05, + "loss": 1.2136, + "step": 3212 + }, + { + "epoch": 0.19521234582902972, + "grad_norm": 0.512143611907959, + "learning_rate": 9.111423378713868e-05, + "loss": 1.0786, + "step": 3213 + }, + { + "epoch": 0.19527310286165625, + "grad_norm": 0.18578726053237915, + "learning_rate": 9.110878539474308e-05, + "loss": 1.1016, + "step": 3214 + }, + { + "epoch": 0.19533385989428276, + "grad_norm": 3.6386947631835938, + "learning_rate": 9.1103335495497e-05, + "loss": 1.1097, + "step": 3215 + }, + { + "epoch": 0.1953946169269093, + "grad_norm": 0.44069069623947144, + "learning_rate": 9.10978840896002e-05, + "loss": 1.0815, + "step": 3216 + }, + { + "epoch": 0.19545537395953583, + "grad_norm": 0.24008500576019287, + "learning_rate": 9.109243117725254e-05, + "loss": 1.1957, + "step": 3217 + }, + { + "epoch": 0.19551613099216233, + "grad_norm": 0.569942831993103, + "learning_rate": 9.108697675865385e-05, + "loss": 1.3636, + "step": 3218 + }, + { + "epoch": 0.19557688802478887, + "grad_norm": 0.3215605318546295, + "learning_rate": 9.10815208340041e-05, + "loss": 1.0987, + "step": 3219 + }, + { + "epoch": 0.1956376450574154, + "grad_norm": 0.3446838855743408, + "learning_rate": 9.107606340350325e-05, + "loss": 1.1586, + "step": 3220 + }, + { + "epoch": 0.1956984020900419, + "grad_norm": 0.4048527479171753, + "learning_rate": 9.107060446735138e-05, + "loss": 1.0753, + "step": 3221 + }, + { + "epoch": 0.19575915912266845, + "grad_norm": 0.1986507624387741, + "learning_rate": 9.106514402574854e-05, + "loss": 1.0764, + "step": 3222 + }, + { + "epoch": 0.19581991615529498, + "grad_norm": 0.3455974757671356, + "learning_rate": 9.105968207889492e-05, + "loss": 1.0981, + "step": 3223 + }, + { + "epoch": 0.1958806731879215, + "grad_norm": 0.23602989315986633, + "learning_rate": 9.105421862699071e-05, + "loss": 1.1412, + "step": 3224 + }, + { + "epoch": 0.19594143022054802, + "grad_norm": 0.3418799638748169, + "learning_rate": 9.104875367023619e-05, + "loss": 1.1508, + "step": 3225 + }, + { + "epoch": 0.19600218725317456, + "grad_norm": 0.26207971572875977, + "learning_rate": 9.104328720883166e-05, + "loss": 1.1493, + "step": 3226 + }, + { + "epoch": 0.1960629442858011, + "grad_norm": 0.18124482035636902, + "learning_rate": 9.103781924297751e-05, + "loss": 1.1395, + "step": 3227 + }, + { + "epoch": 0.1961237013184276, + "grad_norm": 0.43913325667381287, + "learning_rate": 9.103234977287416e-05, + "loss": 1.1472, + "step": 3228 + }, + { + "epoch": 0.19618445835105414, + "grad_norm": 0.3542608618736267, + "learning_rate": 9.10268787987221e-05, + "loss": 1.1214, + "step": 3229 + }, + { + "epoch": 0.19624521538368067, + "grad_norm": 0.46739068627357483, + "learning_rate": 9.102140632072187e-05, + "loss": 1.0299, + "step": 3230 + }, + { + "epoch": 0.19630597241630718, + "grad_norm": 0.30446046590805054, + "learning_rate": 9.101593233907405e-05, + "loss": 1.0933, + "step": 3231 + }, + { + "epoch": 0.1963667294489337, + "grad_norm": 0.4346044361591339, + "learning_rate": 9.101045685397931e-05, + "loss": 1.091, + "step": 3232 + }, + { + "epoch": 0.19642748648156025, + "grad_norm": 0.20626798272132874, + "learning_rate": 9.100497986563833e-05, + "loss": 1.1629, + "step": 3233 + }, + { + "epoch": 0.19648824351418676, + "grad_norm": 0.37170013785362244, + "learning_rate": 9.099950137425193e-05, + "loss": 1.3227, + "step": 3234 + }, + { + "epoch": 0.1965490005468133, + "grad_norm": 0.24555674195289612, + "learning_rate": 9.099402138002084e-05, + "loss": 1.0562, + "step": 3235 + }, + { + "epoch": 0.19660975757943983, + "grad_norm": 0.22307699918746948, + "learning_rate": 9.098853988314599e-05, + "loss": 1.0809, + "step": 3236 + }, + { + "epoch": 0.19667051461206636, + "grad_norm": 0.5041522979736328, + "learning_rate": 9.098305688382829e-05, + "loss": 1.193, + "step": 3237 + }, + { + "epoch": 0.19673127164469287, + "grad_norm": 0.20439326763153076, + "learning_rate": 9.097757238226872e-05, + "loss": 1.1372, + "step": 3238 + }, + { + "epoch": 0.1967920286773194, + "grad_norm": 0.37200728058815, + "learning_rate": 9.097208637866829e-05, + "loss": 1.1745, + "step": 3239 + }, + { + "epoch": 0.19685278570994594, + "grad_norm": 0.2528643012046814, + "learning_rate": 9.096659887322813e-05, + "loss": 1.0928, + "step": 3240 + }, + { + "epoch": 0.19691354274257244, + "grad_norm": 0.2649160325527191, + "learning_rate": 9.096110986614938e-05, + "loss": 1.0552, + "step": 3241 + }, + { + "epoch": 0.19697429977519898, + "grad_norm": 0.30476313829421997, + "learning_rate": 9.095561935763321e-05, + "loss": 1.1394, + "step": 3242 + }, + { + "epoch": 0.19703505680782551, + "grad_norm": 1.00332510471344, + "learning_rate": 9.095012734788093e-05, + "loss": 1.1824, + "step": 3243 + }, + { + "epoch": 0.19709581384045202, + "grad_norm": 0.220684215426445, + "learning_rate": 9.094463383709379e-05, + "loss": 1.1368, + "step": 3244 + }, + { + "epoch": 0.19715657087307856, + "grad_norm": 0.22012288868427277, + "learning_rate": 9.093913882547318e-05, + "loss": 1.0934, + "step": 3245 + }, + { + "epoch": 0.1972173279057051, + "grad_norm": 0.24159879982471466, + "learning_rate": 9.093364231322052e-05, + "loss": 1.0691, + "step": 3246 + }, + { + "epoch": 0.1972780849383316, + "grad_norm": 0.18609967827796936, + "learning_rate": 9.09281443005373e-05, + "loss": 1.066, + "step": 3247 + }, + { + "epoch": 0.19733884197095813, + "grad_norm": 0.8121808767318726, + "learning_rate": 9.092264478762505e-05, + "loss": 1.0854, + "step": 3248 + }, + { + "epoch": 0.19739959900358467, + "grad_norm": 0.28250765800476074, + "learning_rate": 9.091714377468532e-05, + "loss": 1.1142, + "step": 3249 + }, + { + "epoch": 0.1974603560362112, + "grad_norm": 0.2897456884384155, + "learning_rate": 9.09116412619198e-05, + "loss": 1.1171, + "step": 3250 + }, + { + "epoch": 0.1975211130688377, + "grad_norm": 1.195672869682312, + "learning_rate": 9.090613724953014e-05, + "loss": 1.2771, + "step": 3251 + }, + { + "epoch": 0.19758187010146425, + "grad_norm": 0.2923084497451782, + "learning_rate": 9.090063173771813e-05, + "loss": 1.187, + "step": 3252 + }, + { + "epoch": 0.19764262713409078, + "grad_norm": 0.1628628671169281, + "learning_rate": 9.089512472668554e-05, + "loss": 1.0646, + "step": 3253 + }, + { + "epoch": 0.1977033841667173, + "grad_norm": 0.23243358731269836, + "learning_rate": 9.088961621663425e-05, + "loss": 1.1851, + "step": 3254 + }, + { + "epoch": 0.19776414119934382, + "grad_norm": 0.2904801368713379, + "learning_rate": 9.08841062077662e-05, + "loss": 1.233, + "step": 3255 + }, + { + "epoch": 0.19782489823197036, + "grad_norm": 0.340282142162323, + "learning_rate": 9.08785947002833e-05, + "loss": 1.1009, + "step": 3256 + }, + { + "epoch": 0.19788565526459687, + "grad_norm": 0.23964887857437134, + "learning_rate": 9.087308169438765e-05, + "loss": 1.1251, + "step": 3257 + }, + { + "epoch": 0.1979464122972234, + "grad_norm": 0.3045327067375183, + "learning_rate": 9.086756719028124e-05, + "loss": 1.1879, + "step": 3258 + }, + { + "epoch": 0.19800716932984994, + "grad_norm": 0.4541415274143219, + "learning_rate": 9.086205118816628e-05, + "loss": 1.2373, + "step": 3259 + }, + { + "epoch": 0.19806792636247644, + "grad_norm": 0.21064196527004242, + "learning_rate": 9.085653368824494e-05, + "loss": 1.1119, + "step": 3260 + }, + { + "epoch": 0.19812868339510298, + "grad_norm": 0.33118903636932373, + "learning_rate": 9.085101469071945e-05, + "loss": 1.1069, + "step": 3261 + }, + { + "epoch": 0.1981894404277295, + "grad_norm": 0.18136803805828094, + "learning_rate": 9.084549419579213e-05, + "loss": 1.0736, + "step": 3262 + }, + { + "epoch": 0.19825019746035605, + "grad_norm": 0.39444753527641296, + "learning_rate": 9.083997220366533e-05, + "loss": 1.1084, + "step": 3263 + }, + { + "epoch": 0.19831095449298256, + "grad_norm": 0.19552722573280334, + "learning_rate": 9.083444871454144e-05, + "loss": 1.1144, + "step": 3264 + }, + { + "epoch": 0.1983717115256091, + "grad_norm": 0.24402844905853271, + "learning_rate": 9.082892372862295e-05, + "loss": 1.1226, + "step": 3265 + }, + { + "epoch": 0.19843246855823563, + "grad_norm": 2.442267417907715, + "learning_rate": 9.082339724611236e-05, + "loss": 1.0912, + "step": 3266 + }, + { + "epoch": 0.19849322559086213, + "grad_norm": 0.21277213096618652, + "learning_rate": 9.081786926721226e-05, + "loss": 1.0718, + "step": 3267 + }, + { + "epoch": 0.19855398262348867, + "grad_norm": 0.2457491159439087, + "learning_rate": 9.081233979212526e-05, + "loss": 1.1836, + "step": 3268 + }, + { + "epoch": 0.1986147396561152, + "grad_norm": 0.2352142035961151, + "learning_rate": 9.080680882105406e-05, + "loss": 1.1158, + "step": 3269 + }, + { + "epoch": 0.1986754966887417, + "grad_norm": 0.24459496140480042, + "learning_rate": 9.080127635420142e-05, + "loss": 1.1375, + "step": 3270 + }, + { + "epoch": 0.19873625372136824, + "grad_norm": 0.24818889796733856, + "learning_rate": 9.079574239177008e-05, + "loss": 1.1363, + "step": 3271 + }, + { + "epoch": 0.19879701075399478, + "grad_norm": 0.29332593083381653, + "learning_rate": 9.079020693396292e-05, + "loss": 1.1418, + "step": 3272 + }, + { + "epoch": 0.19885776778662131, + "grad_norm": 0.276141494512558, + "learning_rate": 9.078466998098285e-05, + "loss": 1.1497, + "step": 3273 + }, + { + "epoch": 0.19891852481924782, + "grad_norm": 0.30076923966407776, + "learning_rate": 9.077913153303282e-05, + "loss": 1.2404, + "step": 3274 + }, + { + "epoch": 0.19897928185187436, + "grad_norm": 0.3659590780735016, + "learning_rate": 9.077359159031582e-05, + "loss": 1.139, + "step": 3275 + }, + { + "epoch": 0.1990400388845009, + "grad_norm": 0.33483052253723145, + "learning_rate": 9.076805015303496e-05, + "loss": 1.0118, + "step": 3276 + }, + { + "epoch": 0.1991007959171274, + "grad_norm": 0.37975969910621643, + "learning_rate": 9.076250722139333e-05, + "loss": 1.1603, + "step": 3277 + }, + { + "epoch": 0.19916155294975393, + "grad_norm": 0.8754175305366516, + "learning_rate": 9.075696279559414e-05, + "loss": 1.2416, + "step": 3278 + }, + { + "epoch": 0.19922230998238047, + "grad_norm": 0.3032798767089844, + "learning_rate": 9.075141687584057e-05, + "loss": 1.1858, + "step": 3279 + }, + { + "epoch": 0.19928306701500698, + "grad_norm": 1.1042546033859253, + "learning_rate": 9.074586946233595e-05, + "loss": 1.177, + "step": 3280 + }, + { + "epoch": 0.1993438240476335, + "grad_norm": 0.5889238715171814, + "learning_rate": 9.074032055528361e-05, + "loss": 1.2127, + "step": 3281 + }, + { + "epoch": 0.19940458108026005, + "grad_norm": 2.3201022148132324, + "learning_rate": 9.073477015488696e-05, + "loss": 1.1461, + "step": 3282 + }, + { + "epoch": 0.19946533811288655, + "grad_norm": 0.6564784049987793, + "learning_rate": 9.072921826134942e-05, + "loss": 1.1243, + "step": 3283 + }, + { + "epoch": 0.1995260951455131, + "grad_norm": 0.22137127816677094, + "learning_rate": 9.072366487487451e-05, + "loss": 1.0724, + "step": 3284 + }, + { + "epoch": 0.19958685217813962, + "grad_norm": 0.5857661962509155, + "learning_rate": 9.07181099956658e-05, + "loss": 1.1103, + "step": 3285 + }, + { + "epoch": 0.19964760921076616, + "grad_norm": 1.1090264320373535, + "learning_rate": 9.071255362392688e-05, + "loss": 1.1181, + "step": 3286 + }, + { + "epoch": 0.19970836624339267, + "grad_norm": 0.6247833371162415, + "learning_rate": 9.070699575986146e-05, + "loss": 1.1664, + "step": 3287 + }, + { + "epoch": 0.1997691232760192, + "grad_norm": 0.7372910380363464, + "learning_rate": 9.070143640367323e-05, + "loss": 1.18, + "step": 3288 + }, + { + "epoch": 0.19982988030864574, + "grad_norm": 2.5752205848693848, + "learning_rate": 9.069587555556598e-05, + "loss": 1.2979, + "step": 3289 + }, + { + "epoch": 0.19989063734127224, + "grad_norm": 0.2157212495803833, + "learning_rate": 9.069031321574354e-05, + "loss": 1.132, + "step": 3290 + }, + { + "epoch": 0.19995139437389878, + "grad_norm": 0.2926340103149414, + "learning_rate": 9.068474938440983e-05, + "loss": 1.1097, + "step": 3291 + }, + { + "epoch": 0.2000121514065253, + "grad_norm": 0.20449599623680115, + "learning_rate": 9.067918406176875e-05, + "loss": 1.1005, + "step": 3292 + }, + { + "epoch": 0.20007290843915182, + "grad_norm": 0.1954326033592224, + "learning_rate": 9.06736172480243e-05, + "loss": 1.0959, + "step": 3293 + }, + { + "epoch": 0.20013366547177835, + "grad_norm": 0.48577409982681274, + "learning_rate": 9.066804894338056e-05, + "loss": 1.049, + "step": 3294 + }, + { + "epoch": 0.2001944225044049, + "grad_norm": 0.302401065826416, + "learning_rate": 9.066247914804164e-05, + "loss": 1.0979, + "step": 3295 + }, + { + "epoch": 0.20025517953703142, + "grad_norm": 0.41936248540878296, + "learning_rate": 9.065690786221168e-05, + "loss": 1.2231, + "step": 3296 + }, + { + "epoch": 0.20031593656965793, + "grad_norm": 0.4368925094604492, + "learning_rate": 9.06513350860949e-05, + "loss": 1.084, + "step": 3297 + }, + { + "epoch": 0.20037669360228447, + "grad_norm": 0.32463338971138, + "learning_rate": 9.064576081989557e-05, + "loss": 1.0849, + "step": 3298 + }, + { + "epoch": 0.200437450634911, + "grad_norm": 0.36803242564201355, + "learning_rate": 9.064018506381802e-05, + "loss": 1.1139, + "step": 3299 + }, + { + "epoch": 0.2004982076675375, + "grad_norm": 0.18876750767230988, + "learning_rate": 9.063460781806663e-05, + "loss": 1.1221, + "step": 3300 + }, + { + "epoch": 0.20055896470016404, + "grad_norm": 0.42081791162490845, + "learning_rate": 9.062902908284585e-05, + "loss": 1.1733, + "step": 3301 + }, + { + "epoch": 0.20061972173279058, + "grad_norm": 0.2598825991153717, + "learning_rate": 9.062344885836014e-05, + "loss": 1.1346, + "step": 3302 + }, + { + "epoch": 0.2006804787654171, + "grad_norm": 0.32085493206977844, + "learning_rate": 9.061786714481406e-05, + "loss": 1.1303, + "step": 3303 + }, + { + "epoch": 0.20074123579804362, + "grad_norm": 0.5949311256408691, + "learning_rate": 9.061228394241222e-05, + "loss": 1.1257, + "step": 3304 + }, + { + "epoch": 0.20080199283067016, + "grad_norm": 0.18799662590026855, + "learning_rate": 9.060669925135925e-05, + "loss": 1.0513, + "step": 3305 + }, + { + "epoch": 0.20086274986329666, + "grad_norm": 0.6351726055145264, + "learning_rate": 9.060111307185988e-05, + "loss": 1.0666, + "step": 3306 + }, + { + "epoch": 0.2009235068959232, + "grad_norm": 0.207907035946846, + "learning_rate": 9.059552540411886e-05, + "loss": 1.1664, + "step": 3307 + }, + { + "epoch": 0.20098426392854973, + "grad_norm": 0.5979374647140503, + "learning_rate": 9.0589936248341e-05, + "loss": 1.1024, + "step": 3308 + }, + { + "epoch": 0.20104502096117627, + "grad_norm": 0.43916672468185425, + "learning_rate": 9.058434560473119e-05, + "loss": 1.0741, + "step": 3309 + }, + { + "epoch": 0.20110577799380278, + "grad_norm": 0.5928214192390442, + "learning_rate": 9.057875347349434e-05, + "loss": 1.1246, + "step": 3310 + }, + { + "epoch": 0.2011665350264293, + "grad_norm": 0.6521312594413757, + "learning_rate": 9.057315985483545e-05, + "loss": 1.1211, + "step": 3311 + }, + { + "epoch": 0.20122729205905585, + "grad_norm": 0.5383350253105164, + "learning_rate": 9.056756474895954e-05, + "loss": 1.0997, + "step": 3312 + }, + { + "epoch": 0.20128804909168235, + "grad_norm": 0.561617910861969, + "learning_rate": 9.05619681560717e-05, + "loss": 1.0881, + "step": 3313 + }, + { + "epoch": 0.2013488061243089, + "grad_norm": 0.4649290442466736, + "learning_rate": 9.055637007637707e-05, + "loss": 1.1672, + "step": 3314 + }, + { + "epoch": 0.20140956315693542, + "grad_norm": 0.20864000916481018, + "learning_rate": 9.055077051008087e-05, + "loss": 1.0756, + "step": 3315 + }, + { + "epoch": 0.20147032018956193, + "grad_norm": 0.4433615207672119, + "learning_rate": 9.054516945738833e-05, + "loss": 1.1923, + "step": 3316 + }, + { + "epoch": 0.20153107722218846, + "grad_norm": 0.2718067765235901, + "learning_rate": 9.053956691850476e-05, + "loss": 1.0875, + "step": 3317 + }, + { + "epoch": 0.201591834254815, + "grad_norm": 0.6589035987854004, + "learning_rate": 9.053396289363553e-05, + "loss": 1.4642, + "step": 3318 + }, + { + "epoch": 0.20165259128744153, + "grad_norm": 0.5452963709831238, + "learning_rate": 9.052835738298606e-05, + "loss": 1.1139, + "step": 3319 + }, + { + "epoch": 0.20171334832006804, + "grad_norm": 1.057050347328186, + "learning_rate": 9.052275038676183e-05, + "loss": 1.0597, + "step": 3320 + }, + { + "epoch": 0.20177410535269458, + "grad_norm": 0.42550352215766907, + "learning_rate": 9.051714190516834e-05, + "loss": 1.2634, + "step": 3321 + }, + { + "epoch": 0.2018348623853211, + "grad_norm": 0.5254426002502441, + "learning_rate": 9.051153193841118e-05, + "loss": 1.3359, + "step": 3322 + }, + { + "epoch": 0.20189561941794762, + "grad_norm": 0.7363297939300537, + "learning_rate": 9.050592048669597e-05, + "loss": 1.093, + "step": 3323 + }, + { + "epoch": 0.20195637645057415, + "grad_norm": 0.25182509422302246, + "learning_rate": 9.050030755022844e-05, + "loss": 1.1061, + "step": 3324 + }, + { + "epoch": 0.2020171334832007, + "grad_norm": 0.6574466824531555, + "learning_rate": 9.04946931292143e-05, + "loss": 1.0842, + "step": 3325 + }, + { + "epoch": 0.2020778905158272, + "grad_norm": 0.33787721395492554, + "learning_rate": 9.048907722385935e-05, + "loss": 1.2065, + "step": 3326 + }, + { + "epoch": 0.20213864754845373, + "grad_norm": 0.3194088637828827, + "learning_rate": 9.048345983436946e-05, + "loss": 1.1397, + "step": 3327 + }, + { + "epoch": 0.20219940458108027, + "grad_norm": 0.3715285062789917, + "learning_rate": 9.047784096095052e-05, + "loss": 1.1624, + "step": 3328 + }, + { + "epoch": 0.20226016161370677, + "grad_norm": 0.2945951223373413, + "learning_rate": 9.047222060380849e-05, + "loss": 1.0725, + "step": 3329 + }, + { + "epoch": 0.2023209186463333, + "grad_norm": 2.4057018756866455, + "learning_rate": 9.046659876314938e-05, + "loss": 1.0883, + "step": 3330 + }, + { + "epoch": 0.20238167567895984, + "grad_norm": 0.6218315362930298, + "learning_rate": 9.046097543917928e-05, + "loss": 1.1845, + "step": 3331 + }, + { + "epoch": 0.20244243271158638, + "grad_norm": 0.2737712562084198, + "learning_rate": 9.045535063210432e-05, + "loss": 1.1756, + "step": 3332 + }, + { + "epoch": 0.20250318974421289, + "grad_norm": 0.9794842004776001, + "learning_rate": 9.044972434213064e-05, + "loss": 1.1445, + "step": 3333 + }, + { + "epoch": 0.20256394677683942, + "grad_norm": 1.646140217781067, + "learning_rate": 9.04440965694645e-05, + "loss": 1.2334, + "step": 3334 + }, + { + "epoch": 0.20262470380946596, + "grad_norm": 0.37521985173225403, + "learning_rate": 9.04384673143122e-05, + "loss": 1.1453, + "step": 3335 + }, + { + "epoch": 0.20268546084209246, + "grad_norm": 0.19489142298698425, + "learning_rate": 9.043283657688005e-05, + "loss": 1.0962, + "step": 3336 + }, + { + "epoch": 0.202746217874719, + "grad_norm": 0.27855822443962097, + "learning_rate": 9.042720435737447e-05, + "loss": 1.1576, + "step": 3337 + }, + { + "epoch": 0.20280697490734553, + "grad_norm": 0.31201520562171936, + "learning_rate": 9.042157065600188e-05, + "loss": 1.0983, + "step": 3338 + }, + { + "epoch": 0.20286773193997204, + "grad_norm": 0.304785817861557, + "learning_rate": 9.041593547296883e-05, + "loss": 1.0852, + "step": 3339 + }, + { + "epoch": 0.20292848897259858, + "grad_norm": 0.2534112334251404, + "learning_rate": 9.041029880848183e-05, + "loss": 1.2224, + "step": 3340 + }, + { + "epoch": 0.2029892460052251, + "grad_norm": 1.238376259803772, + "learning_rate": 9.040466066274753e-05, + "loss": 1.1446, + "step": 3341 + }, + { + "epoch": 0.20305000303785165, + "grad_norm": 0.2306281179189682, + "learning_rate": 9.039902103597258e-05, + "loss": 1.1176, + "step": 3342 + }, + { + "epoch": 0.20311076007047815, + "grad_norm": 0.2823100686073303, + "learning_rate": 9.03933799283637e-05, + "loss": 1.1759, + "step": 3343 + }, + { + "epoch": 0.2031715171031047, + "grad_norm": 0.1960209459066391, + "learning_rate": 9.038773734012768e-05, + "loss": 1.0917, + "step": 3344 + }, + { + "epoch": 0.20323227413573122, + "grad_norm": 0.4982382655143738, + "learning_rate": 9.038209327147133e-05, + "loss": 1.2267, + "step": 3345 + }, + { + "epoch": 0.20329303116835773, + "grad_norm": 4.838447570800781, + "learning_rate": 9.037644772260155e-05, + "loss": 1.1091, + "step": 3346 + }, + { + "epoch": 0.20335378820098426, + "grad_norm": 1.0698766708374023, + "learning_rate": 9.037080069372527e-05, + "loss": 1.1104, + "step": 3347 + }, + { + "epoch": 0.2034145452336108, + "grad_norm": 0.36323273181915283, + "learning_rate": 9.03651521850495e-05, + "loss": 1.1395, + "step": 3348 + }, + { + "epoch": 0.2034753022662373, + "grad_norm": 0.5239294767379761, + "learning_rate": 9.035950219678126e-05, + "loss": 1.1097, + "step": 3349 + }, + { + "epoch": 0.20353605929886384, + "grad_norm": 0.7276628613471985, + "learning_rate": 9.035385072912768e-05, + "loss": 1.2858, + "step": 3350 + }, + { + "epoch": 0.20359681633149038, + "grad_norm": 0.23691920936107635, + "learning_rate": 9.034819778229588e-05, + "loss": 1.2557, + "step": 3351 + }, + { + "epoch": 0.20365757336411688, + "grad_norm": 0.1887449324131012, + "learning_rate": 9.034254335649311e-05, + "loss": 1.0932, + "step": 3352 + }, + { + "epoch": 0.20371833039674342, + "grad_norm": 0.26820170879364014, + "learning_rate": 9.033688745192661e-05, + "loss": 1.0826, + "step": 3353 + }, + { + "epoch": 0.20377908742936995, + "grad_norm": 0.2283105105161667, + "learning_rate": 9.03312300688037e-05, + "loss": 1.0933, + "step": 3354 + }, + { + "epoch": 0.2038398444619965, + "grad_norm": 0.26691800355911255, + "learning_rate": 9.032557120733176e-05, + "loss": 1.1915, + "step": 3355 + }, + { + "epoch": 0.203900601494623, + "grad_norm": 0.3651144206523895, + "learning_rate": 9.031991086771823e-05, + "loss": 1.1832, + "step": 3356 + }, + { + "epoch": 0.20396135852724953, + "grad_norm": 0.1835484802722931, + "learning_rate": 9.031424905017054e-05, + "loss": 1.0982, + "step": 3357 + }, + { + "epoch": 0.20402211555987607, + "grad_norm": 0.2541102468967438, + "learning_rate": 9.030858575489628e-05, + "loss": 1.1309, + "step": 3358 + }, + { + "epoch": 0.20408287259250257, + "grad_norm": 0.3518550395965576, + "learning_rate": 9.030292098210302e-05, + "loss": 1.0887, + "step": 3359 + }, + { + "epoch": 0.2041436296251291, + "grad_norm": 0.3566979765892029, + "learning_rate": 9.029725473199839e-05, + "loss": 1.1743, + "step": 3360 + }, + { + "epoch": 0.20420438665775564, + "grad_norm": 0.22790130972862244, + "learning_rate": 9.02915870047901e-05, + "loss": 1.0912, + "step": 3361 + }, + { + "epoch": 0.20426514369038215, + "grad_norm": 0.2512471675872803, + "learning_rate": 9.02859178006859e-05, + "loss": 1.1631, + "step": 3362 + }, + { + "epoch": 0.20432590072300869, + "grad_norm": 0.25767388939857483, + "learning_rate": 9.02802471198936e-05, + "loss": 1.0913, + "step": 3363 + }, + { + "epoch": 0.20438665775563522, + "grad_norm": 0.1904771327972412, + "learning_rate": 9.027457496262104e-05, + "loss": 1.1049, + "step": 3364 + }, + { + "epoch": 0.20444741478826173, + "grad_norm": 0.282078355550766, + "learning_rate": 9.026890132907618e-05, + "loss": 1.181, + "step": 3365 + }, + { + "epoch": 0.20450817182088826, + "grad_norm": 0.4109525680541992, + "learning_rate": 9.026322621946692e-05, + "loss": 1.0929, + "step": 3366 + }, + { + "epoch": 0.2045689288535148, + "grad_norm": 1.8262817859649658, + "learning_rate": 9.025754963400134e-05, + "loss": 1.1388, + "step": 3367 + }, + { + "epoch": 0.20462968588614133, + "grad_norm": 0.23712901771068573, + "learning_rate": 9.02518715728875e-05, + "loss": 1.1135, + "step": 3368 + }, + { + "epoch": 0.20469044291876784, + "grad_norm": 0.27330780029296875, + "learning_rate": 9.024619203633353e-05, + "loss": 1.1098, + "step": 3369 + }, + { + "epoch": 0.20475119995139437, + "grad_norm": 0.2928329110145569, + "learning_rate": 9.02405110245476e-05, + "loss": 1.1481, + "step": 3370 + }, + { + "epoch": 0.2048119569840209, + "grad_norm": 0.4196322560310364, + "learning_rate": 9.023482853773797e-05, + "loss": 1.0743, + "step": 3371 + }, + { + "epoch": 0.20487271401664742, + "grad_norm": 0.2326882928609848, + "learning_rate": 9.022914457611293e-05, + "loss": 1.0906, + "step": 3372 + }, + { + "epoch": 0.20493347104927395, + "grad_norm": 0.2831681966781616, + "learning_rate": 9.02234591398808e-05, + "loss": 1.1106, + "step": 3373 + }, + { + "epoch": 0.2049942280819005, + "grad_norm": 0.3011842966079712, + "learning_rate": 9.021777222925001e-05, + "loss": 1.0992, + "step": 3374 + }, + { + "epoch": 0.205054985114527, + "grad_norm": 1.237701654434204, + "learning_rate": 9.021208384442901e-05, + "loss": 1.1038, + "step": 3375 + }, + { + "epoch": 0.20511574214715353, + "grad_norm": 0.3185805380344391, + "learning_rate": 9.020639398562629e-05, + "loss": 1.1417, + "step": 3376 + }, + { + "epoch": 0.20517649917978006, + "grad_norm": 0.2854292690753937, + "learning_rate": 9.020070265305043e-05, + "loss": 1.0845, + "step": 3377 + }, + { + "epoch": 0.2052372562124066, + "grad_norm": 1.1912065744400024, + "learning_rate": 9.019500984691004e-05, + "loss": 1.3709, + "step": 3378 + }, + { + "epoch": 0.2052980132450331, + "grad_norm": 0.2931707501411438, + "learning_rate": 9.018931556741381e-05, + "loss": 1.1643, + "step": 3379 + }, + { + "epoch": 0.20535877027765964, + "grad_norm": 0.4930371344089508, + "learning_rate": 9.018361981477045e-05, + "loss": 1.0882, + "step": 3380 + }, + { + "epoch": 0.20541952731028618, + "grad_norm": 0.38935717940330505, + "learning_rate": 9.017792258918873e-05, + "loss": 1.1312, + "step": 3381 + }, + { + "epoch": 0.20548028434291268, + "grad_norm": 0.5490107536315918, + "learning_rate": 9.01722238908775e-05, + "loss": 1.1904, + "step": 3382 + }, + { + "epoch": 0.20554104137553922, + "grad_norm": 0.25143784284591675, + "learning_rate": 9.016652372004562e-05, + "loss": 1.0775, + "step": 3383 + }, + { + "epoch": 0.20560179840816575, + "grad_norm": 0.4033779799938202, + "learning_rate": 9.016082207690207e-05, + "loss": 1.1177, + "step": 3384 + }, + { + "epoch": 0.20566255544079226, + "grad_norm": 0.2509956657886505, + "learning_rate": 9.01551189616558e-05, + "loss": 1.1384, + "step": 3385 + }, + { + "epoch": 0.2057233124734188, + "grad_norm": 0.33178776502609253, + "learning_rate": 9.01494143745159e-05, + "loss": 1.1923, + "step": 3386 + }, + { + "epoch": 0.20578406950604533, + "grad_norm": 0.205919548869133, + "learning_rate": 9.014370831569144e-05, + "loss": 1.1218, + "step": 3387 + }, + { + "epoch": 0.20584482653867184, + "grad_norm": 0.20552746951580048, + "learning_rate": 9.01380007853916e-05, + "loss": 1.0706, + "step": 3388 + }, + { + "epoch": 0.20590558357129837, + "grad_norm": 0.40637433528900146, + "learning_rate": 9.01322917838256e-05, + "loss": 1.074, + "step": 3389 + }, + { + "epoch": 0.2059663406039249, + "grad_norm": 0.21393392980098724, + "learning_rate": 9.012658131120266e-05, + "loss": 1.102, + "step": 3390 + }, + { + "epoch": 0.20602709763655144, + "grad_norm": 0.35514283180236816, + "learning_rate": 9.012086936773214e-05, + "loss": 1.0724, + "step": 3391 + }, + { + "epoch": 0.20608785466917795, + "grad_norm": 0.37697187066078186, + "learning_rate": 9.011515595362342e-05, + "loss": 1.0764, + "step": 3392 + }, + { + "epoch": 0.20614861170180449, + "grad_norm": 0.42121294140815735, + "learning_rate": 9.010944106908586e-05, + "loss": 1.0954, + "step": 3393 + }, + { + "epoch": 0.20620936873443102, + "grad_norm": 0.40075138211250305, + "learning_rate": 9.010372471432902e-05, + "loss": 1.0517, + "step": 3394 + }, + { + "epoch": 0.20627012576705753, + "grad_norm": 0.3212359845638275, + "learning_rate": 9.009800688956237e-05, + "loss": 1.1488, + "step": 3395 + }, + { + "epoch": 0.20633088279968406, + "grad_norm": 0.6188732981681824, + "learning_rate": 9.009228759499555e-05, + "loss": 1.0811, + "step": 3396 + }, + { + "epoch": 0.2063916398323106, + "grad_norm": 0.4885496199131012, + "learning_rate": 9.008656683083816e-05, + "loss": 1.0455, + "step": 3397 + }, + { + "epoch": 0.2064523968649371, + "grad_norm": 0.6051378846168518, + "learning_rate": 9.008084459729993e-05, + "loss": 1.1111, + "step": 3398 + }, + { + "epoch": 0.20651315389756364, + "grad_norm": 0.5744614005088806, + "learning_rate": 9.007512089459058e-05, + "loss": 1.1756, + "step": 3399 + }, + { + "epoch": 0.20657391093019017, + "grad_norm": 0.2092771977186203, + "learning_rate": 9.006939572291994e-05, + "loss": 1.0452, + "step": 3400 + }, + { + "epoch": 0.2066346679628167, + "grad_norm": 0.602838397026062, + "learning_rate": 9.006366908249784e-05, + "loss": 1.1033, + "step": 3401 + }, + { + "epoch": 0.20669542499544322, + "grad_norm": 0.2793847620487213, + "learning_rate": 9.005794097353422e-05, + "loss": 1.0772, + "step": 3402 + }, + { + "epoch": 0.20675618202806975, + "grad_norm": 0.37822622060775757, + "learning_rate": 9.005221139623903e-05, + "loss": 1.1558, + "step": 3403 + }, + { + "epoch": 0.2068169390606963, + "grad_norm": 1.472434639930725, + "learning_rate": 9.004648035082228e-05, + "loss": 1.1511, + "step": 3404 + }, + { + "epoch": 0.2068776960933228, + "grad_norm": 0.327202707529068, + "learning_rate": 9.004074783749405e-05, + "loss": 1.2315, + "step": 3405 + }, + { + "epoch": 0.20693845312594933, + "grad_norm": 0.24670813977718353, + "learning_rate": 9.003501385646449e-05, + "loss": 1.1808, + "step": 3406 + }, + { + "epoch": 0.20699921015857586, + "grad_norm": 0.5312100648880005, + "learning_rate": 9.002927840794374e-05, + "loss": 1.0783, + "step": 3407 + }, + { + "epoch": 0.20705996719120237, + "grad_norm": 0.25706616044044495, + "learning_rate": 9.002354149214205e-05, + "loss": 1.1331, + "step": 3408 + }, + { + "epoch": 0.2071207242238289, + "grad_norm": 0.2662416696548462, + "learning_rate": 9.001780310926972e-05, + "loss": 1.1249, + "step": 3409 + }, + { + "epoch": 0.20718148125645544, + "grad_norm": 2.5181188583374023, + "learning_rate": 9.001206325953709e-05, + "loss": 1.059, + "step": 3410 + }, + { + "epoch": 0.20724223828908195, + "grad_norm": 0.37585198879241943, + "learning_rate": 9.000632194315453e-05, + "loss": 1.1578, + "step": 3411 + }, + { + "epoch": 0.20730299532170848, + "grad_norm": 0.4133990705013275, + "learning_rate": 9.00005791603325e-05, + "loss": 1.1402, + "step": 3412 + }, + { + "epoch": 0.20736375235433502, + "grad_norm": 0.3630717098712921, + "learning_rate": 8.999483491128153e-05, + "loss": 1.0641, + "step": 3413 + }, + { + "epoch": 0.20742450938696155, + "grad_norm": 0.1877926141023636, + "learning_rate": 8.998908919621214e-05, + "loss": 1.1142, + "step": 3414 + }, + { + "epoch": 0.20748526641958806, + "grad_norm": 0.5594463348388672, + "learning_rate": 8.998334201533495e-05, + "loss": 1.2017, + "step": 3415 + }, + { + "epoch": 0.2075460234522146, + "grad_norm": 0.3596123158931732, + "learning_rate": 8.997759336886066e-05, + "loss": 1.1907, + "step": 3416 + }, + { + "epoch": 0.20760678048484113, + "grad_norm": 0.36783456802368164, + "learning_rate": 8.997184325699994e-05, + "loss": 1.0854, + "step": 3417 + }, + { + "epoch": 0.20766753751746764, + "grad_norm": 0.35031986236572266, + "learning_rate": 8.996609167996357e-05, + "loss": 1.1967, + "step": 3418 + }, + { + "epoch": 0.20772829455009417, + "grad_norm": 0.35371699929237366, + "learning_rate": 8.99603386379624e-05, + "loss": 1.0625, + "step": 3419 + }, + { + "epoch": 0.2077890515827207, + "grad_norm": 0.35674139857292175, + "learning_rate": 8.995458413120727e-05, + "loss": 1.1851, + "step": 3420 + }, + { + "epoch": 0.20784980861534721, + "grad_norm": 0.32445228099823, + "learning_rate": 8.994882815990917e-05, + "loss": 1.153, + "step": 3421 + }, + { + "epoch": 0.20791056564797375, + "grad_norm": 0.673724353313446, + "learning_rate": 8.994307072427902e-05, + "loss": 1.079, + "step": 3422 + }, + { + "epoch": 0.20797132268060028, + "grad_norm": 0.3673085868358612, + "learning_rate": 8.993731182452791e-05, + "loss": 1.3594, + "step": 3423 + }, + { + "epoch": 0.20803207971322682, + "grad_norm": 0.19975057244300842, + "learning_rate": 8.993155146086691e-05, + "loss": 1.0885, + "step": 3424 + }, + { + "epoch": 0.20809283674585333, + "grad_norm": 0.19591695070266724, + "learning_rate": 8.992578963350717e-05, + "loss": 1.1563, + "step": 3425 + }, + { + "epoch": 0.20815359377847986, + "grad_norm": 0.2949843108654022, + "learning_rate": 8.99200263426599e-05, + "loss": 1.1687, + "step": 3426 + }, + { + "epoch": 0.2082143508111064, + "grad_norm": 0.2903730571269989, + "learning_rate": 8.991426158853635e-05, + "loss": 1.0736, + "step": 3427 + }, + { + "epoch": 0.2082751078437329, + "grad_norm": 0.2379426807165146, + "learning_rate": 8.990849537134782e-05, + "loss": 1.1155, + "step": 3428 + }, + { + "epoch": 0.20833586487635944, + "grad_norm": 0.43215957283973694, + "learning_rate": 8.990272769130567e-05, + "loss": 1.2331, + "step": 3429 + }, + { + "epoch": 0.20839662190898597, + "grad_norm": 2.558924436569214, + "learning_rate": 8.989695854862132e-05, + "loss": 1.0956, + "step": 3430 + }, + { + "epoch": 0.20845737894161248, + "grad_norm": 0.3600039482116699, + "learning_rate": 8.989118794350625e-05, + "loss": 1.2509, + "step": 3431 + }, + { + "epoch": 0.20851813597423902, + "grad_norm": 0.4987107813358307, + "learning_rate": 8.9885415876172e-05, + "loss": 1.1149, + "step": 3432 + }, + { + "epoch": 0.20857889300686555, + "grad_norm": 0.3423027992248535, + "learning_rate": 8.987964234683009e-05, + "loss": 1.0478, + "step": 3433 + }, + { + "epoch": 0.20863965003949206, + "grad_norm": 0.32997873425483704, + "learning_rate": 8.98738673556922e-05, + "loss": 1.1, + "step": 3434 + }, + { + "epoch": 0.2087004070721186, + "grad_norm": 0.24387899041175842, + "learning_rate": 8.986809090296998e-05, + "loss": 1.1611, + "step": 3435 + }, + { + "epoch": 0.20876116410474513, + "grad_norm": 0.23385708034038544, + "learning_rate": 8.98623129888752e-05, + "loss": 1.0829, + "step": 3436 + }, + { + "epoch": 0.20882192113737166, + "grad_norm": 0.5016626119613647, + "learning_rate": 8.985653361361961e-05, + "loss": 1.0945, + "step": 3437 + }, + { + "epoch": 0.20888267816999817, + "grad_norm": 0.5924230813980103, + "learning_rate": 8.98507527774151e-05, + "loss": 1.2836, + "step": 3438 + }, + { + "epoch": 0.2089434352026247, + "grad_norm": 0.4242420494556427, + "learning_rate": 8.98449704804735e-05, + "loss": 1.0687, + "step": 3439 + }, + { + "epoch": 0.20900419223525124, + "grad_norm": 3.821106433868408, + "learning_rate": 8.983918672300687e-05, + "loss": 1.0972, + "step": 3440 + }, + { + "epoch": 0.20906494926787775, + "grad_norm": 0.6792571544647217, + "learning_rate": 8.98334015052271e-05, + "loss": 1.2186, + "step": 3441 + }, + { + "epoch": 0.20912570630050428, + "grad_norm": 0.25056201219558716, + "learning_rate": 8.982761482734632e-05, + "loss": 1.0928, + "step": 3442 + }, + { + "epoch": 0.20918646333313082, + "grad_norm": 0.32500478625297546, + "learning_rate": 8.982182668957661e-05, + "loss": 1.0814, + "step": 3443 + }, + { + "epoch": 0.20924722036575732, + "grad_norm": 0.9380246996879578, + "learning_rate": 8.981603709213014e-05, + "loss": 1.1438, + "step": 3444 + }, + { + "epoch": 0.20930797739838386, + "grad_norm": 0.2966766357421875, + "learning_rate": 8.981024603521914e-05, + "loss": 1.2158, + "step": 3445 + }, + { + "epoch": 0.2093687344310104, + "grad_norm": 0.5393595099449158, + "learning_rate": 8.980445351905588e-05, + "loss": 1.0708, + "step": 3446 + }, + { + "epoch": 0.20942949146363693, + "grad_norm": 0.2098262757062912, + "learning_rate": 8.979865954385269e-05, + "loss": 1.245, + "step": 3447 + }, + { + "epoch": 0.20949024849626344, + "grad_norm": 0.8116827011108398, + "learning_rate": 8.979286410982194e-05, + "loss": 1.1252, + "step": 3448 + }, + { + "epoch": 0.20955100552888997, + "grad_norm": 0.2955787181854248, + "learning_rate": 8.978706721717605e-05, + "loss": 1.1385, + "step": 3449 + }, + { + "epoch": 0.2096117625615165, + "grad_norm": 0.2282090038061142, + "learning_rate": 8.978126886612754e-05, + "loss": 1.1674, + "step": 3450 + }, + { + "epoch": 0.20967251959414301, + "grad_norm": 0.2110520899295807, + "learning_rate": 8.977546905688891e-05, + "loss": 1.0898, + "step": 3451 + }, + { + "epoch": 0.20973327662676955, + "grad_norm": 0.2905969023704529, + "learning_rate": 8.976966778967279e-05, + "loss": 1.0904, + "step": 3452 + }, + { + "epoch": 0.20979403365939608, + "grad_norm": 0.5926627516746521, + "learning_rate": 8.976386506469178e-05, + "loss": 1.2544, + "step": 3453 + }, + { + "epoch": 0.2098547906920226, + "grad_norm": 0.3120969831943512, + "learning_rate": 8.975806088215864e-05, + "loss": 1.1529, + "step": 3454 + }, + { + "epoch": 0.20991554772464913, + "grad_norm": 0.2321310192346573, + "learning_rate": 8.975225524228609e-05, + "loss": 1.112, + "step": 3455 + }, + { + "epoch": 0.20997630475727566, + "grad_norm": 0.36554744839668274, + "learning_rate": 8.974644814528694e-05, + "loss": 1.1213, + "step": 3456 + }, + { + "epoch": 0.21003706178990217, + "grad_norm": 0.2901652455329895, + "learning_rate": 8.974063959137405e-05, + "loss": 1.0675, + "step": 3457 + }, + { + "epoch": 0.2100978188225287, + "grad_norm": 0.21345216035842896, + "learning_rate": 8.973482958076032e-05, + "loss": 1.0905, + "step": 3458 + }, + { + "epoch": 0.21015857585515524, + "grad_norm": 0.2896096408367157, + "learning_rate": 8.972901811365875e-05, + "loss": 1.081, + "step": 3459 + }, + { + "epoch": 0.21021933288778177, + "grad_norm": 0.24613705277442932, + "learning_rate": 8.972320519028232e-05, + "loss": 1.2019, + "step": 3460 + }, + { + "epoch": 0.21028008992040828, + "grad_norm": 0.1897294819355011, + "learning_rate": 8.971739081084414e-05, + "loss": 1.0612, + "step": 3461 + }, + { + "epoch": 0.21034084695303482, + "grad_norm": 0.6065770983695984, + "learning_rate": 8.971157497555733e-05, + "loss": 1.2315, + "step": 3462 + }, + { + "epoch": 0.21040160398566135, + "grad_norm": 0.28365975618362427, + "learning_rate": 8.970575768463505e-05, + "loss": 1.2905, + "step": 3463 + }, + { + "epoch": 0.21046236101828786, + "grad_norm": 0.8950469493865967, + "learning_rate": 8.969993893829054e-05, + "loss": 1.1229, + "step": 3464 + }, + { + "epoch": 0.2105231180509144, + "grad_norm": 0.5066483616828918, + "learning_rate": 8.969411873673712e-05, + "loss": 1.1418, + "step": 3465 + }, + { + "epoch": 0.21058387508354093, + "grad_norm": 0.46005985140800476, + "learning_rate": 8.968829708018809e-05, + "loss": 1.1784, + "step": 3466 + }, + { + "epoch": 0.21064463211616744, + "grad_norm": 0.8842597603797913, + "learning_rate": 8.968247396885685e-05, + "loss": 1.0757, + "step": 3467 + }, + { + "epoch": 0.21070538914879397, + "grad_norm": 0.22984270751476288, + "learning_rate": 8.967664940295685e-05, + "loss": 1.1105, + "step": 3468 + }, + { + "epoch": 0.2107661461814205, + "grad_norm": 0.2960885465145111, + "learning_rate": 8.967082338270163e-05, + "loss": 1.1534, + "step": 3469 + }, + { + "epoch": 0.210826903214047, + "grad_norm": 0.36603888869285583, + "learning_rate": 8.966499590830468e-05, + "loss": 1.1275, + "step": 3470 + }, + { + "epoch": 0.21088766024667355, + "grad_norm": 0.482968807220459, + "learning_rate": 8.965916697997966e-05, + "loss": 1.1522, + "step": 3471 + }, + { + "epoch": 0.21094841727930008, + "grad_norm": 0.3588212728500366, + "learning_rate": 8.96533365979402e-05, + "loss": 1.0648, + "step": 3472 + }, + { + "epoch": 0.21100917431192662, + "grad_norm": 0.673191487789154, + "learning_rate": 8.964750476240003e-05, + "loss": 1.198, + "step": 3473 + }, + { + "epoch": 0.21106993134455312, + "grad_norm": 0.23914748430252075, + "learning_rate": 8.964167147357289e-05, + "loss": 1.0738, + "step": 3474 + }, + { + "epoch": 0.21113068837717966, + "grad_norm": 0.612608790397644, + "learning_rate": 8.963583673167263e-05, + "loss": 1.1855, + "step": 3475 + }, + { + "epoch": 0.2111914454098062, + "grad_norm": 0.37512755393981934, + "learning_rate": 8.96300005369131e-05, + "loss": 1.1308, + "step": 3476 + }, + { + "epoch": 0.2112522024424327, + "grad_norm": 0.310783714056015, + "learning_rate": 8.962416288950827e-05, + "loss": 1.3249, + "step": 3477 + }, + { + "epoch": 0.21131295947505924, + "grad_norm": 0.8673312664031982, + "learning_rate": 8.961832378967207e-05, + "loss": 1.0795, + "step": 3478 + }, + { + "epoch": 0.21137371650768577, + "grad_norm": 3.193424701690674, + "learning_rate": 8.961248323761855e-05, + "loss": 1.0732, + "step": 3479 + }, + { + "epoch": 0.21143447354031228, + "grad_norm": 0.5715492963790894, + "learning_rate": 8.960664123356182e-05, + "loss": 1.1333, + "step": 3480 + }, + { + "epoch": 0.2114952305729388, + "grad_norm": 0.36726000905036926, + "learning_rate": 8.960079777771598e-05, + "loss": 1.1181, + "step": 3481 + }, + { + "epoch": 0.21155598760556535, + "grad_norm": 0.37686166167259216, + "learning_rate": 8.959495287029525e-05, + "loss": 1.045, + "step": 3482 + }, + { + "epoch": 0.21161674463819188, + "grad_norm": 1.1148327589035034, + "learning_rate": 8.958910651151387e-05, + "loss": 1.0535, + "step": 3483 + }, + { + "epoch": 0.2116775016708184, + "grad_norm": 2.4224467277526855, + "learning_rate": 8.958325870158613e-05, + "loss": 1.085, + "step": 3484 + }, + { + "epoch": 0.21173825870344493, + "grad_norm": 0.3158526122570038, + "learning_rate": 8.95774094407264e-05, + "loss": 1.0784, + "step": 3485 + }, + { + "epoch": 0.21179901573607146, + "grad_norm": 0.44088003039360046, + "learning_rate": 8.957155872914907e-05, + "loss": 1.0613, + "step": 3486 + }, + { + "epoch": 0.21185977276869797, + "grad_norm": 0.33493566513061523, + "learning_rate": 8.956570656706862e-05, + "loss": 1.1196, + "step": 3487 + }, + { + "epoch": 0.2119205298013245, + "grad_norm": 0.21833741664886475, + "learning_rate": 8.955985295469953e-05, + "loss": 1.094, + "step": 3488 + }, + { + "epoch": 0.21198128683395104, + "grad_norm": 0.37176820635795593, + "learning_rate": 8.95539978922564e-05, + "loss": 1.0676, + "step": 3489 + }, + { + "epoch": 0.21204204386657755, + "grad_norm": 0.19479458034038544, + "learning_rate": 8.954814137995382e-05, + "loss": 1.1096, + "step": 3490 + }, + { + "epoch": 0.21210280089920408, + "grad_norm": 0.3835144340991974, + "learning_rate": 8.954228341800648e-05, + "loss": 1.0683, + "step": 3491 + }, + { + "epoch": 0.21216355793183062, + "grad_norm": 0.1991848647594452, + "learning_rate": 8.953642400662909e-05, + "loss": 1.0974, + "step": 3492 + }, + { + "epoch": 0.21222431496445712, + "grad_norm": 0.214741051197052, + "learning_rate": 8.953056314603644e-05, + "loss": 1.1628, + "step": 3493 + }, + { + "epoch": 0.21228507199708366, + "grad_norm": 0.23162704706192017, + "learning_rate": 8.952470083644336e-05, + "loss": 1.1141, + "step": 3494 + }, + { + "epoch": 0.2123458290297102, + "grad_norm": 0.27646422386169434, + "learning_rate": 8.951883707806472e-05, + "loss": 1.0939, + "step": 3495 + }, + { + "epoch": 0.21240658606233673, + "grad_norm": 0.25253263115882874, + "learning_rate": 8.951297187111547e-05, + "loss": 1.0876, + "step": 3496 + }, + { + "epoch": 0.21246734309496323, + "grad_norm": 0.274149626493454, + "learning_rate": 8.95071052158106e-05, + "loss": 1.1265, + "step": 3497 + }, + { + "epoch": 0.21252810012758977, + "grad_norm": 0.5955387353897095, + "learning_rate": 8.950123711236517e-05, + "loss": 1.0988, + "step": 3498 + }, + { + "epoch": 0.2125888571602163, + "grad_norm": 0.22578568756580353, + "learning_rate": 8.949536756099424e-05, + "loss": 1.1543, + "step": 3499 + }, + { + "epoch": 0.2126496141928428, + "grad_norm": 0.41070738434791565, + "learning_rate": 8.948949656191299e-05, + "loss": 1.1193, + "step": 3500 + }, + { + "epoch": 0.21271037122546935, + "grad_norm": 0.2455478310585022, + "learning_rate": 8.94836241153366e-05, + "loss": 1.0749, + "step": 3501 + }, + { + "epoch": 0.21277112825809588, + "grad_norm": 0.40537217259407043, + "learning_rate": 8.947775022148033e-05, + "loss": 1.0743, + "step": 3502 + }, + { + "epoch": 0.2128318852907224, + "grad_norm": 0.2949778735637665, + "learning_rate": 8.94718748805595e-05, + "loss": 1.3076, + "step": 3503 + }, + { + "epoch": 0.21289264232334892, + "grad_norm": 0.32014939188957214, + "learning_rate": 8.946599809278949e-05, + "loss": 1.152, + "step": 3504 + }, + { + "epoch": 0.21295339935597546, + "grad_norm": 0.2701198160648346, + "learning_rate": 8.946011985838565e-05, + "loss": 1.3215, + "step": 3505 + }, + { + "epoch": 0.213014156388602, + "grad_norm": 0.5408681035041809, + "learning_rate": 8.945424017756352e-05, + "loss": 1.124, + "step": 3506 + }, + { + "epoch": 0.2130749134212285, + "grad_norm": 0.20604859292507172, + "learning_rate": 8.944835905053858e-05, + "loss": 1.0566, + "step": 3507 + }, + { + "epoch": 0.21313567045385504, + "grad_norm": 0.26530587673187256, + "learning_rate": 8.944247647752642e-05, + "loss": 1.1644, + "step": 3508 + }, + { + "epoch": 0.21319642748648157, + "grad_norm": 0.26491791009902954, + "learning_rate": 8.943659245874266e-05, + "loss": 1.2045, + "step": 3509 + }, + { + "epoch": 0.21325718451910808, + "grad_norm": 0.2437717765569687, + "learning_rate": 8.943070699440295e-05, + "loss": 1.3636, + "step": 3510 + }, + { + "epoch": 0.2133179415517346, + "grad_norm": 0.3257567882537842, + "learning_rate": 8.942482008472309e-05, + "loss": 1.1522, + "step": 3511 + }, + { + "epoch": 0.21337869858436115, + "grad_norm": 0.35287317633628845, + "learning_rate": 8.941893172991882e-05, + "loss": 1.0742, + "step": 3512 + }, + { + "epoch": 0.21343945561698766, + "grad_norm": 0.3571230471134186, + "learning_rate": 8.941304193020599e-05, + "loss": 1.3511, + "step": 3513 + }, + { + "epoch": 0.2135002126496142, + "grad_norm": 0.32294270396232605, + "learning_rate": 8.940715068580047e-05, + "loss": 1.3056, + "step": 3514 + }, + { + "epoch": 0.21356096968224073, + "grad_norm": 0.4447548985481262, + "learning_rate": 8.940125799691826e-05, + "loss": 1.0791, + "step": 3515 + }, + { + "epoch": 0.21362172671486723, + "grad_norm": 0.1982898712158203, + "learning_rate": 8.93953638637753e-05, + "loss": 1.0573, + "step": 3516 + }, + { + "epoch": 0.21368248374749377, + "grad_norm": 1.8522236347198486, + "learning_rate": 8.938946828658766e-05, + "loss": 1.0678, + "step": 3517 + }, + { + "epoch": 0.2137432407801203, + "grad_norm": 0.4071902334690094, + "learning_rate": 8.938357126557144e-05, + "loss": 1.279, + "step": 3518 + }, + { + "epoch": 0.21380399781274684, + "grad_norm": 0.28447458148002625, + "learning_rate": 8.937767280094281e-05, + "loss": 1.0888, + "step": 3519 + }, + { + "epoch": 0.21386475484537334, + "grad_norm": 0.26805147528648376, + "learning_rate": 8.937177289291798e-05, + "loss": 1.0791, + "step": 3520 + }, + { + "epoch": 0.21392551187799988, + "grad_norm": 0.5727316737174988, + "learning_rate": 8.93658715417132e-05, + "loss": 1.0918, + "step": 3521 + }, + { + "epoch": 0.21398626891062641, + "grad_norm": 1.8367565870285034, + "learning_rate": 8.935996874754478e-05, + "loss": 1.2231, + "step": 3522 + }, + { + "epoch": 0.21404702594325292, + "grad_norm": 0.7219190001487732, + "learning_rate": 8.935406451062911e-05, + "loss": 1.2823, + "step": 3523 + }, + { + "epoch": 0.21410778297587946, + "grad_norm": 0.34251299500465393, + "learning_rate": 8.93481588311826e-05, + "loss": 1.1028, + "step": 3524 + }, + { + "epoch": 0.214168540008506, + "grad_norm": 2.0956454277038574, + "learning_rate": 8.934225170942172e-05, + "loss": 1.0809, + "step": 3525 + }, + { + "epoch": 0.2142292970411325, + "grad_norm": 0.7784513831138611, + "learning_rate": 8.933634314556299e-05, + "loss": 1.1988, + "step": 3526 + }, + { + "epoch": 0.21429005407375903, + "grad_norm": 2.3493189811706543, + "learning_rate": 8.9330433139823e-05, + "loss": 1.0806, + "step": 3527 + }, + { + "epoch": 0.21435081110638557, + "grad_norm": 0.4207421541213989, + "learning_rate": 8.932452169241838e-05, + "loss": 1.1711, + "step": 3528 + }, + { + "epoch": 0.2144115681390121, + "grad_norm": 0.5475471019744873, + "learning_rate": 8.931860880356583e-05, + "loss": 1.1401, + "step": 3529 + }, + { + "epoch": 0.2144723251716386, + "grad_norm": 0.21160103380680084, + "learning_rate": 8.931269447348206e-05, + "loss": 1.1118, + "step": 3530 + }, + { + "epoch": 0.21453308220426515, + "grad_norm": 0.3001171946525574, + "learning_rate": 8.930677870238388e-05, + "loss": 1.136, + "step": 3531 + }, + { + "epoch": 0.21459383923689168, + "grad_norm": 0.6058387756347656, + "learning_rate": 8.930086149048813e-05, + "loss": 1.0512, + "step": 3532 + }, + { + "epoch": 0.2146545962695182, + "grad_norm": 0.28012603521347046, + "learning_rate": 8.92949428380117e-05, + "loss": 1.2613, + "step": 3533 + }, + { + "epoch": 0.21471535330214472, + "grad_norm": 0.6427878737449646, + "learning_rate": 8.928902274517154e-05, + "loss": 1.1411, + "step": 3534 + }, + { + "epoch": 0.21477611033477126, + "grad_norm": 1.0942015647888184, + "learning_rate": 8.928310121218469e-05, + "loss": 1.1278, + "step": 3535 + }, + { + "epoch": 0.21483686736739777, + "grad_norm": 0.3067569136619568, + "learning_rate": 8.927717823926813e-05, + "loss": 1.1397, + "step": 3536 + }, + { + "epoch": 0.2148976244000243, + "grad_norm": 1.1183362007141113, + "learning_rate": 8.927125382663903e-05, + "loss": 1.1269, + "step": 3537 + }, + { + "epoch": 0.21495838143265084, + "grad_norm": 0.4447917938232422, + "learning_rate": 8.926532797451452e-05, + "loss": 1.171, + "step": 3538 + }, + { + "epoch": 0.21501913846527734, + "grad_norm": 0.30020472407341003, + "learning_rate": 8.925940068311183e-05, + "loss": 1.2653, + "step": 3539 + }, + { + "epoch": 0.21507989549790388, + "grad_norm": 0.3023083806037903, + "learning_rate": 8.925347195264823e-05, + "loss": 1.0684, + "step": 3540 + }, + { + "epoch": 0.2151406525305304, + "grad_norm": 0.3037755787372589, + "learning_rate": 8.924754178334102e-05, + "loss": 1.071, + "step": 3541 + }, + { + "epoch": 0.21520140956315695, + "grad_norm": 0.19217005372047424, + "learning_rate": 8.924161017540758e-05, + "loss": 1.0876, + "step": 3542 + }, + { + "epoch": 0.21526216659578346, + "grad_norm": 0.2573149800300598, + "learning_rate": 8.923567712906534e-05, + "loss": 1.1454, + "step": 3543 + }, + { + "epoch": 0.21532292362841, + "grad_norm": 0.28273525834083557, + "learning_rate": 8.922974264453175e-05, + "loss": 1.1328, + "step": 3544 + }, + { + "epoch": 0.21538368066103653, + "grad_norm": 0.2546294033527374, + "learning_rate": 8.922380672202437e-05, + "loss": 1.214, + "step": 3545 + }, + { + "epoch": 0.21544443769366303, + "grad_norm": 0.36929014325141907, + "learning_rate": 8.921786936176077e-05, + "loss": 1.1906, + "step": 3546 + }, + { + "epoch": 0.21550519472628957, + "grad_norm": 0.29610875248908997, + "learning_rate": 8.92119305639586e-05, + "loss": 1.1845, + "step": 3547 + }, + { + "epoch": 0.2155659517589161, + "grad_norm": 0.279045969247818, + "learning_rate": 8.920599032883554e-05, + "loss": 1.1017, + "step": 3548 + }, + { + "epoch": 0.2156267087915426, + "grad_norm": 0.7364672422409058, + "learning_rate": 8.92000486566093e-05, + "loss": 1.084, + "step": 3549 + }, + { + "epoch": 0.21568746582416914, + "grad_norm": 0.2714690864086151, + "learning_rate": 8.919410554749772e-05, + "loss": 1.1239, + "step": 3550 + }, + { + "epoch": 0.21574822285679568, + "grad_norm": 0.24270565807819366, + "learning_rate": 8.918816100171862e-05, + "loss": 1.1421, + "step": 3551 + }, + { + "epoch": 0.21580897988942221, + "grad_norm": 0.4487621784210205, + "learning_rate": 8.91822150194899e-05, + "loss": 1.1091, + "step": 3552 + }, + { + "epoch": 0.21586973692204872, + "grad_norm": 0.2807360589504242, + "learning_rate": 8.917626760102951e-05, + "loss": 1.172, + "step": 3553 + }, + { + "epoch": 0.21593049395467526, + "grad_norm": 0.19047701358795166, + "learning_rate": 8.917031874655546e-05, + "loss": 1.0945, + "step": 3554 + }, + { + "epoch": 0.2159912509873018, + "grad_norm": 0.21109971404075623, + "learning_rate": 8.916436845628582e-05, + "loss": 1.1838, + "step": 3555 + }, + { + "epoch": 0.2160520080199283, + "grad_norm": 0.26162368059158325, + "learning_rate": 8.915841673043868e-05, + "loss": 1.1588, + "step": 3556 + }, + { + "epoch": 0.21611276505255483, + "grad_norm": 0.275624543428421, + "learning_rate": 8.91524635692322e-05, + "loss": 1.166, + "step": 3557 + }, + { + "epoch": 0.21617352208518137, + "grad_norm": 0.21419569849967957, + "learning_rate": 8.91465089728846e-05, + "loss": 1.1408, + "step": 3558 + }, + { + "epoch": 0.21623427911780788, + "grad_norm": 0.3190023899078369, + "learning_rate": 8.914055294161414e-05, + "loss": 1.1208, + "step": 3559 + }, + { + "epoch": 0.2162950361504344, + "grad_norm": 0.29606959223747253, + "learning_rate": 8.913459547563917e-05, + "loss": 1.3099, + "step": 3560 + }, + { + "epoch": 0.21635579318306095, + "grad_norm": 0.3781929910182953, + "learning_rate": 8.9128636575178e-05, + "loss": 1.2893, + "step": 3561 + }, + { + "epoch": 0.21641655021568745, + "grad_norm": 0.19794784486293793, + "learning_rate": 8.912267624044913e-05, + "loss": 1.0905, + "step": 3562 + }, + { + "epoch": 0.216477307248314, + "grad_norm": 0.43172550201416016, + "learning_rate": 8.911671447167099e-05, + "loss": 1.2475, + "step": 3563 + }, + { + "epoch": 0.21653806428094052, + "grad_norm": 0.1837187260389328, + "learning_rate": 8.91107512690621e-05, + "loss": 1.0814, + "step": 3564 + }, + { + "epoch": 0.21659882131356706, + "grad_norm": 0.22071275115013123, + "learning_rate": 8.910478663284108e-05, + "loss": 1.1634, + "step": 3565 + }, + { + "epoch": 0.21665957834619357, + "grad_norm": 0.21257944405078888, + "learning_rate": 8.909882056322653e-05, + "loss": 1.1655, + "step": 3566 + }, + { + "epoch": 0.2167203353788201, + "grad_norm": 0.47290828824043274, + "learning_rate": 8.909285306043717e-05, + "loss": 1.1211, + "step": 3567 + }, + { + "epoch": 0.21678109241144664, + "grad_norm": 0.20498418807983398, + "learning_rate": 8.908688412469173e-05, + "loss": 1.0372, + "step": 3568 + }, + { + "epoch": 0.21684184944407314, + "grad_norm": 0.19669866561889648, + "learning_rate": 8.9080913756209e-05, + "loss": 1.1041, + "step": 3569 + }, + { + "epoch": 0.21690260647669968, + "grad_norm": 0.19793730974197388, + "learning_rate": 8.90749419552078e-05, + "loss": 1.0615, + "step": 3570 + }, + { + "epoch": 0.2169633635093262, + "grad_norm": 0.20443733036518097, + "learning_rate": 8.906896872190706e-05, + "loss": 1.1499, + "step": 3571 + }, + { + "epoch": 0.21702412054195272, + "grad_norm": 0.4876909852027893, + "learning_rate": 8.906299405652573e-05, + "loss": 1.1765, + "step": 3572 + }, + { + "epoch": 0.21708487757457925, + "grad_norm": 0.18095681071281433, + "learning_rate": 8.90570179592828e-05, + "loss": 1.0809, + "step": 3573 + }, + { + "epoch": 0.2171456346072058, + "grad_norm": 0.3138459026813507, + "learning_rate": 8.905104043039731e-05, + "loss": 1.079, + "step": 3574 + }, + { + "epoch": 0.2172063916398323, + "grad_norm": 0.30645492672920227, + "learning_rate": 8.90450614700884e-05, + "loss": 1.0941, + "step": 3575 + }, + { + "epoch": 0.21726714867245883, + "grad_norm": 0.8269155025482178, + "learning_rate": 8.903908107857521e-05, + "loss": 1.1351, + "step": 3576 + }, + { + "epoch": 0.21732790570508537, + "grad_norm": 0.2424241304397583, + "learning_rate": 8.903309925607696e-05, + "loss": 1.0921, + "step": 3577 + }, + { + "epoch": 0.2173886627377119, + "grad_norm": 0.4220335781574249, + "learning_rate": 8.90271160028129e-05, + "loss": 1.3899, + "step": 3578 + }, + { + "epoch": 0.2174494197703384, + "grad_norm": 0.21445181965827942, + "learning_rate": 8.902113131900238e-05, + "loss": 1.1438, + "step": 3579 + }, + { + "epoch": 0.21751017680296494, + "grad_norm": 0.19127045571804047, + "learning_rate": 8.901514520486473e-05, + "loss": 1.0467, + "step": 3580 + }, + { + "epoch": 0.21757093383559148, + "grad_norm": 0.2837027609348297, + "learning_rate": 8.90091576606194e-05, + "loss": 1.2572, + "step": 3581 + }, + { + "epoch": 0.217631690868218, + "grad_norm": 0.23541265726089478, + "learning_rate": 8.900316868648586e-05, + "loss": 1.1717, + "step": 3582 + }, + { + "epoch": 0.21769244790084452, + "grad_norm": 0.2644394338130951, + "learning_rate": 8.899717828268361e-05, + "loss": 1.1016, + "step": 3583 + }, + { + "epoch": 0.21775320493347106, + "grad_norm": 0.22277748584747314, + "learning_rate": 8.899118644943227e-05, + "loss": 1.1182, + "step": 3584 + }, + { + "epoch": 0.21781396196609756, + "grad_norm": 0.3112041652202606, + "learning_rate": 8.898519318695146e-05, + "loss": 1.0472, + "step": 3585 + }, + { + "epoch": 0.2178747189987241, + "grad_norm": 0.23184004426002502, + "learning_rate": 8.897919849546086e-05, + "loss": 1.092, + "step": 3586 + }, + { + "epoch": 0.21793547603135063, + "grad_norm": 0.3312663435935974, + "learning_rate": 8.897320237518019e-05, + "loss": 1.0525, + "step": 3587 + }, + { + "epoch": 0.21799623306397717, + "grad_norm": 0.16587431728839874, + "learning_rate": 8.896720482632925e-05, + "loss": 1.0489, + "step": 3588 + }, + { + "epoch": 0.21805699009660368, + "grad_norm": 0.2613990306854248, + "learning_rate": 8.896120584912791e-05, + "loss": 1.0856, + "step": 3589 + }, + { + "epoch": 0.2181177471292302, + "grad_norm": 0.2676975131034851, + "learning_rate": 8.895520544379601e-05, + "loss": 1.1909, + "step": 3590 + }, + { + "epoch": 0.21817850416185675, + "grad_norm": 0.21448172628879547, + "learning_rate": 8.894920361055354e-05, + "loss": 1.1185, + "step": 3591 + }, + { + "epoch": 0.21823926119448325, + "grad_norm": 0.2731534540653229, + "learning_rate": 8.894320034962047e-05, + "loss": 1.2378, + "step": 3592 + }, + { + "epoch": 0.2183000182271098, + "grad_norm": 0.21493865549564362, + "learning_rate": 8.893719566121688e-05, + "loss": 1.195, + "step": 3593 + }, + { + "epoch": 0.21836077525973632, + "grad_norm": 0.24433006346225739, + "learning_rate": 8.893118954556286e-05, + "loss": 1.0828, + "step": 3594 + }, + { + "epoch": 0.21842153229236283, + "grad_norm": 0.21704885363578796, + "learning_rate": 8.892518200287854e-05, + "loss": 1.0762, + "step": 3595 + }, + { + "epoch": 0.21848228932498936, + "grad_norm": 0.1808609515428543, + "learning_rate": 8.891917303338416e-05, + "loss": 1.0441, + "step": 3596 + }, + { + "epoch": 0.2185430463576159, + "grad_norm": 0.49666911363601685, + "learning_rate": 8.891316263729997e-05, + "loss": 1.107, + "step": 3597 + }, + { + "epoch": 0.2186038033902424, + "grad_norm": 0.2628198266029358, + "learning_rate": 8.890715081484628e-05, + "loss": 1.1592, + "step": 3598 + }, + { + "epoch": 0.21866456042286894, + "grad_norm": 0.8344945907592773, + "learning_rate": 8.890113756624347e-05, + "loss": 1.1643, + "step": 3599 + }, + { + "epoch": 0.21872531745549548, + "grad_norm": 0.3708471953868866, + "learning_rate": 8.889512289171194e-05, + "loss": 1.1652, + "step": 3600 + }, + { + "epoch": 0.218786074488122, + "grad_norm": 0.48135897517204285, + "learning_rate": 8.888910679147215e-05, + "loss": 1.0767, + "step": 3601 + }, + { + "epoch": 0.21884683152074852, + "grad_norm": 0.3165457546710968, + "learning_rate": 8.888308926574463e-05, + "loss": 1.2208, + "step": 3602 + }, + { + "epoch": 0.21890758855337505, + "grad_norm": 0.4605223834514618, + "learning_rate": 8.887707031474996e-05, + "loss": 1.1789, + "step": 3603 + }, + { + "epoch": 0.2189683455860016, + "grad_norm": 0.6766998171806335, + "learning_rate": 8.887104993870878e-05, + "loss": 1.1492, + "step": 3604 + }, + { + "epoch": 0.2190291026186281, + "grad_norm": 0.5501927137374878, + "learning_rate": 8.886502813784173e-05, + "loss": 1.2039, + "step": 3605 + }, + { + "epoch": 0.21908985965125463, + "grad_norm": 2.7024359703063965, + "learning_rate": 8.885900491236957e-05, + "loss": 1.1142, + "step": 3606 + }, + { + "epoch": 0.21915061668388117, + "grad_norm": 1.3017866611480713, + "learning_rate": 8.885298026251307e-05, + "loss": 1.149, + "step": 3607 + }, + { + "epoch": 0.21921137371650767, + "grad_norm": 0.649744987487793, + "learning_rate": 8.884695418849307e-05, + "loss": 1.0978, + "step": 3608 + }, + { + "epoch": 0.2192721307491342, + "grad_norm": 5.623937129974365, + "learning_rate": 8.884092669053046e-05, + "loss": 1.159, + "step": 3609 + }, + { + "epoch": 0.21933288778176074, + "grad_norm": 1.0707684755325317, + "learning_rate": 8.883489776884617e-05, + "loss": 1.2088, + "step": 3610 + }, + { + "epoch": 0.21939364481438728, + "grad_norm": 0.4993567168712616, + "learning_rate": 8.88288674236612e-05, + "loss": 1.1187, + "step": 3611 + }, + { + "epoch": 0.21945440184701379, + "grad_norm": 3.359678030014038, + "learning_rate": 8.882283565519658e-05, + "loss": 1.0735, + "step": 3612 + }, + { + "epoch": 0.21951515887964032, + "grad_norm": 1.483595371246338, + "learning_rate": 8.881680246367343e-05, + "loss": 1.1498, + "step": 3613 + }, + { + "epoch": 0.21957591591226686, + "grad_norm": 2.1215660572052, + "learning_rate": 8.881076784931287e-05, + "loss": 1.0614, + "step": 3614 + }, + { + "epoch": 0.21963667294489336, + "grad_norm": 2.6376383304595947, + "learning_rate": 8.880473181233611e-05, + "loss": 1.0648, + "step": 3615 + }, + { + "epoch": 0.2196974299775199, + "grad_norm": 1.2618935108184814, + "learning_rate": 8.879869435296442e-05, + "loss": 1.2008, + "step": 3616 + }, + { + "epoch": 0.21975818701014643, + "grad_norm": 0.42415690422058105, + "learning_rate": 8.879265547141909e-05, + "loss": 1.1138, + "step": 3617 + }, + { + "epoch": 0.21981894404277294, + "grad_norm": 0.24902956187725067, + "learning_rate": 8.878661516792144e-05, + "loss": 1.0761, + "step": 3618 + }, + { + "epoch": 0.21987970107539948, + "grad_norm": 0.5424796342849731, + "learning_rate": 8.878057344269295e-05, + "loss": 1.1859, + "step": 3619 + }, + { + "epoch": 0.219940458108026, + "grad_norm": 0.5163174271583557, + "learning_rate": 8.877453029595504e-05, + "loss": 1.0443, + "step": 3620 + }, + { + "epoch": 0.22000121514065252, + "grad_norm": 0.46897539496421814, + "learning_rate": 8.876848572792922e-05, + "loss": 1.0879, + "step": 3621 + }, + { + "epoch": 0.22006197217327905, + "grad_norm": 0.4106047749519348, + "learning_rate": 8.876243973883706e-05, + "loss": 1.2926, + "step": 3622 + }, + { + "epoch": 0.2201227292059056, + "grad_norm": 0.24506781995296478, + "learning_rate": 8.875639232890018e-05, + "loss": 1.1538, + "step": 3623 + }, + { + "epoch": 0.22018348623853212, + "grad_norm": 0.2511557936668396, + "learning_rate": 8.875034349834023e-05, + "loss": 1.0769, + "step": 3624 + }, + { + "epoch": 0.22024424327115863, + "grad_norm": 0.2335241436958313, + "learning_rate": 8.874429324737898e-05, + "loss": 1.1256, + "step": 3625 + }, + { + "epoch": 0.22030500030378516, + "grad_norm": 0.31653398275375366, + "learning_rate": 8.873824157623816e-05, + "loss": 1.2307, + "step": 3626 + }, + { + "epoch": 0.2203657573364117, + "grad_norm": 0.27423879504203796, + "learning_rate": 8.87321884851396e-05, + "loss": 1.218, + "step": 3627 + }, + { + "epoch": 0.2204265143690382, + "grad_norm": 0.33643218874931335, + "learning_rate": 8.872613397430519e-05, + "loss": 1.1686, + "step": 3628 + }, + { + "epoch": 0.22048727140166474, + "grad_norm": 0.5712282657623291, + "learning_rate": 8.872007804395685e-05, + "loss": 1.1896, + "step": 3629 + }, + { + "epoch": 0.22054802843429128, + "grad_norm": 0.607809841632843, + "learning_rate": 8.871402069431659e-05, + "loss": 1.3216, + "step": 3630 + }, + { + "epoch": 0.22060878546691778, + "grad_norm": 0.3159072995185852, + "learning_rate": 8.870796192560639e-05, + "loss": 1.0888, + "step": 3631 + }, + { + "epoch": 0.22066954249954432, + "grad_norm": 0.23734909296035767, + "learning_rate": 8.870190173804837e-05, + "loss": 1.0625, + "step": 3632 + }, + { + "epoch": 0.22073029953217085, + "grad_norm": 0.8097267746925354, + "learning_rate": 8.869584013186465e-05, + "loss": 1.0729, + "step": 3633 + }, + { + "epoch": 0.2207910565647974, + "grad_norm": 1.0590852499008179, + "learning_rate": 8.868977710727744e-05, + "loss": 1.2094, + "step": 3634 + }, + { + "epoch": 0.2208518135974239, + "grad_norm": 0.24392752349376678, + "learning_rate": 8.8683712664509e-05, + "loss": 1.1544, + "step": 3635 + }, + { + "epoch": 0.22091257063005043, + "grad_norm": 0.3893691301345825, + "learning_rate": 8.867764680378155e-05, + "loss": 1.0782, + "step": 3636 + }, + { + "epoch": 0.22097332766267697, + "grad_norm": 0.38787147402763367, + "learning_rate": 8.867157952531751e-05, + "loss": 1.3316, + "step": 3637 + }, + { + "epoch": 0.22103408469530347, + "grad_norm": 0.18337516486644745, + "learning_rate": 8.866551082933923e-05, + "loss": 1.0938, + "step": 3638 + }, + { + "epoch": 0.22109484172793, + "grad_norm": 0.324688583612442, + "learning_rate": 8.865944071606918e-05, + "loss": 1.1195, + "step": 3639 + }, + { + "epoch": 0.22115559876055654, + "grad_norm": 0.2607549726963043, + "learning_rate": 8.865336918572986e-05, + "loss": 1.1938, + "step": 3640 + }, + { + "epoch": 0.22121635579318305, + "grad_norm": 0.3139565587043762, + "learning_rate": 8.864729623854382e-05, + "loss": 1.1209, + "step": 3641 + }, + { + "epoch": 0.22127711282580959, + "grad_norm": 0.4976213276386261, + "learning_rate": 8.864122187473367e-05, + "loss": 1.1227, + "step": 3642 + }, + { + "epoch": 0.22133786985843612, + "grad_norm": 0.2381664663553238, + "learning_rate": 8.863514609452205e-05, + "loss": 1.1226, + "step": 3643 + }, + { + "epoch": 0.22139862689106263, + "grad_norm": 0.6007063388824463, + "learning_rate": 8.86290688981317e-05, + "loss": 1.1261, + "step": 3644 + }, + { + "epoch": 0.22145938392368916, + "grad_norm": 0.23112396895885468, + "learning_rate": 8.862299028578536e-05, + "loss": 1.0673, + "step": 3645 + }, + { + "epoch": 0.2215201409563157, + "grad_norm": 0.20958848297595978, + "learning_rate": 8.861691025770583e-05, + "loss": 1.1371, + "step": 3646 + }, + { + "epoch": 0.22158089798894223, + "grad_norm": 0.22145237028598785, + "learning_rate": 8.861082881411599e-05, + "loss": 1.1776, + "step": 3647 + }, + { + "epoch": 0.22164165502156874, + "grad_norm": 2.636857509613037, + "learning_rate": 8.860474595523878e-05, + "loss": 1.1126, + "step": 3648 + }, + { + "epoch": 0.22170241205419527, + "grad_norm": 0.27267101407051086, + "learning_rate": 8.859866168129714e-05, + "loss": 1.0709, + "step": 3649 + }, + { + "epoch": 0.2217631690868218, + "grad_norm": 0.2658604085445404, + "learning_rate": 8.859257599251407e-05, + "loss": 1.1108, + "step": 3650 + }, + { + "epoch": 0.22182392611944832, + "grad_norm": 0.2087506800889969, + "learning_rate": 8.85864888891127e-05, + "loss": 1.1381, + "step": 3651 + }, + { + "epoch": 0.22188468315207485, + "grad_norm": 0.21732908487319946, + "learning_rate": 8.858040037131609e-05, + "loss": 1.0904, + "step": 3652 + }, + { + "epoch": 0.2219454401847014, + "grad_norm": 0.4136199951171875, + "learning_rate": 8.857431043934747e-05, + "loss": 1.0683, + "step": 3653 + }, + { + "epoch": 0.2220061972173279, + "grad_norm": 0.46964842081069946, + "learning_rate": 8.856821909343004e-05, + "loss": 1.1442, + "step": 3654 + }, + { + "epoch": 0.22206695424995443, + "grad_norm": 0.36826029419898987, + "learning_rate": 8.856212633378707e-05, + "loss": 1.0833, + "step": 3655 + }, + { + "epoch": 0.22212771128258096, + "grad_norm": 0.3161435127258301, + "learning_rate": 8.855603216064192e-05, + "loss": 1.1453, + "step": 3656 + }, + { + "epoch": 0.2221884683152075, + "grad_norm": 0.5760843753814697, + "learning_rate": 8.854993657421796e-05, + "loss": 1.5548, + "step": 3657 + }, + { + "epoch": 0.222249225347834, + "grad_norm": 0.2688594460487366, + "learning_rate": 8.854383957473863e-05, + "loss": 1.1632, + "step": 3658 + }, + { + "epoch": 0.22230998238046054, + "grad_norm": 0.25979748368263245, + "learning_rate": 8.85377411624274e-05, + "loss": 1.1132, + "step": 3659 + }, + { + "epoch": 0.22237073941308708, + "grad_norm": 0.27163466811180115, + "learning_rate": 8.85316413375078e-05, + "loss": 1.1865, + "step": 3660 + }, + { + "epoch": 0.22243149644571358, + "grad_norm": 0.19699271023273468, + "learning_rate": 8.852554010020346e-05, + "loss": 1.0461, + "step": 3661 + }, + { + "epoch": 0.22249225347834012, + "grad_norm": 0.26535409688949585, + "learning_rate": 8.8519437450738e-05, + "loss": 1.0758, + "step": 3662 + }, + { + "epoch": 0.22255301051096665, + "grad_norm": 0.4278830587863922, + "learning_rate": 8.85133333893351e-05, + "loss": 1.0516, + "step": 3663 + }, + { + "epoch": 0.22261376754359316, + "grad_norm": 0.2898333668708801, + "learning_rate": 8.850722791621853e-05, + "loss": 1.377, + "step": 3664 + }, + { + "epoch": 0.2226745245762197, + "grad_norm": 0.36190763115882874, + "learning_rate": 8.850112103161207e-05, + "loss": 1.1601, + "step": 3665 + }, + { + "epoch": 0.22273528160884623, + "grad_norm": 0.44738680124282837, + "learning_rate": 8.849501273573959e-05, + "loss": 1.2817, + "step": 3666 + }, + { + "epoch": 0.22279603864147274, + "grad_norm": 0.5276013612747192, + "learning_rate": 8.848890302882497e-05, + "loss": 1.0751, + "step": 3667 + }, + { + "epoch": 0.22285679567409927, + "grad_norm": 0.4858834743499756, + "learning_rate": 8.848279191109217e-05, + "loss": 1.2921, + "step": 3668 + }, + { + "epoch": 0.2229175527067258, + "grad_norm": 0.28615033626556396, + "learning_rate": 8.84766793827652e-05, + "loss": 1.1922, + "step": 3669 + }, + { + "epoch": 0.22297830973935234, + "grad_norm": 0.4913071095943451, + "learning_rate": 8.847056544406811e-05, + "loss": 1.1207, + "step": 3670 + }, + { + "epoch": 0.22303906677197885, + "grad_norm": 0.21914042532444, + "learning_rate": 8.8464450095225e-05, + "loss": 1.0738, + "step": 3671 + }, + { + "epoch": 0.22309982380460538, + "grad_norm": 0.26202788949012756, + "learning_rate": 8.845833333646005e-05, + "loss": 1.0973, + "step": 3672 + }, + { + "epoch": 0.22316058083723192, + "grad_norm": 0.19486472010612488, + "learning_rate": 8.845221516799745e-05, + "loss": 1.0756, + "step": 3673 + }, + { + "epoch": 0.22322133786985843, + "grad_norm": 0.9182816743850708, + "learning_rate": 8.844609559006145e-05, + "loss": 1.1279, + "step": 3674 + }, + { + "epoch": 0.22328209490248496, + "grad_norm": 0.2652653753757477, + "learning_rate": 8.843997460287641e-05, + "loss": 1.0545, + "step": 3675 + }, + { + "epoch": 0.2233428519351115, + "grad_norm": 0.2346404492855072, + "learning_rate": 8.843385220666667e-05, + "loss": 1.0753, + "step": 3676 + }, + { + "epoch": 0.223403608967738, + "grad_norm": 0.17741109430789948, + "learning_rate": 8.842772840165663e-05, + "loss": 1.0875, + "step": 3677 + }, + { + "epoch": 0.22346436600036454, + "grad_norm": 0.1890181601047516, + "learning_rate": 8.84216031880708e-05, + "loss": 1.1179, + "step": 3678 + }, + { + "epoch": 0.22352512303299107, + "grad_norm": 0.26611095666885376, + "learning_rate": 8.841547656613367e-05, + "loss": 1.1965, + "step": 3679 + }, + { + "epoch": 0.22358588006561758, + "grad_norm": 0.22856132686138153, + "learning_rate": 8.840934853606981e-05, + "loss": 1.0561, + "step": 3680 + }, + { + "epoch": 0.22364663709824412, + "grad_norm": 0.31674808263778687, + "learning_rate": 8.840321909810386e-05, + "loss": 1.15, + "step": 3681 + }, + { + "epoch": 0.22370739413087065, + "grad_norm": 0.5436544418334961, + "learning_rate": 8.839708825246049e-05, + "loss": 1.1774, + "step": 3682 + }, + { + "epoch": 0.2237681511634972, + "grad_norm": 0.4072345495223999, + "learning_rate": 8.839095599936444e-05, + "loss": 1.3528, + "step": 3683 + }, + { + "epoch": 0.2238289081961237, + "grad_norm": 0.4397291839122772, + "learning_rate": 8.838482233904046e-05, + "loss": 1.0729, + "step": 3684 + }, + { + "epoch": 0.22388966522875023, + "grad_norm": 0.44826629757881165, + "learning_rate": 8.837868727171342e-05, + "loss": 1.2101, + "step": 3685 + }, + { + "epoch": 0.22395042226137676, + "grad_norm": 0.3518873453140259, + "learning_rate": 8.837255079760815e-05, + "loss": 1.1694, + "step": 3686 + }, + { + "epoch": 0.22401117929400327, + "grad_norm": 0.6268365383148193, + "learning_rate": 8.836641291694964e-05, + "loss": 1.1227, + "step": 3687 + }, + { + "epoch": 0.2240719363266298, + "grad_norm": 0.26220810413360596, + "learning_rate": 8.836027362996285e-05, + "loss": 1.1415, + "step": 3688 + }, + { + "epoch": 0.22413269335925634, + "grad_norm": 0.38209739327430725, + "learning_rate": 8.83541329368728e-05, + "loss": 1.0958, + "step": 3689 + }, + { + "epoch": 0.22419345039188285, + "grad_norm": 0.23123860359191895, + "learning_rate": 8.834799083790459e-05, + "loss": 1.048, + "step": 3690 + }, + { + "epoch": 0.22425420742450938, + "grad_norm": 0.19853684306144714, + "learning_rate": 8.834184733328337e-05, + "loss": 1.1334, + "step": 3691 + }, + { + "epoch": 0.22431496445713592, + "grad_norm": 0.3083157241344452, + "learning_rate": 8.833570242323432e-05, + "loss": 1.0705, + "step": 3692 + }, + { + "epoch": 0.22437572148976245, + "grad_norm": 0.3771040141582489, + "learning_rate": 8.83295561079827e-05, + "loss": 1.234, + "step": 3693 + }, + { + "epoch": 0.22443647852238896, + "grad_norm": 0.2504826784133911, + "learning_rate": 8.832340838775378e-05, + "loss": 1.1835, + "step": 3694 + }, + { + "epoch": 0.2244972355550155, + "grad_norm": 0.1967342346906662, + "learning_rate": 8.831725926277293e-05, + "loss": 1.1256, + "step": 3695 + }, + { + "epoch": 0.22455799258764203, + "grad_norm": 0.3082854151725769, + "learning_rate": 8.831110873326552e-05, + "loss": 1.1718, + "step": 3696 + }, + { + "epoch": 0.22461874962026854, + "grad_norm": 0.24681229889392853, + "learning_rate": 8.830495679945703e-05, + "loss": 1.1347, + "step": 3697 + }, + { + "epoch": 0.22467950665289507, + "grad_norm": 0.22830981016159058, + "learning_rate": 8.829880346157294e-05, + "loss": 1.1214, + "step": 3698 + }, + { + "epoch": 0.2247402636855216, + "grad_norm": 0.19750316441059113, + "learning_rate": 8.82926487198388e-05, + "loss": 1.093, + "step": 3699 + }, + { + "epoch": 0.22480102071814811, + "grad_norm": 0.2924378216266632, + "learning_rate": 8.82864925744802e-05, + "loss": 1.0567, + "step": 3700 + }, + { + "epoch": 0.22486177775077465, + "grad_norm": 0.23816899955272675, + "learning_rate": 8.828033502572285e-05, + "loss": 1.1042, + "step": 3701 + }, + { + "epoch": 0.22492253478340118, + "grad_norm": 0.23094728589057922, + "learning_rate": 8.82741760737924e-05, + "loss": 1.1639, + "step": 3702 + }, + { + "epoch": 0.2249832918160277, + "grad_norm": 0.19948410987854004, + "learning_rate": 8.826801571891463e-05, + "loss": 1.0665, + "step": 3703 + }, + { + "epoch": 0.22504404884865423, + "grad_norm": 1.9790393114089966, + "learning_rate": 8.826185396131536e-05, + "loss": 1.0959, + "step": 3704 + }, + { + "epoch": 0.22510480588128076, + "grad_norm": 0.1894340068101883, + "learning_rate": 8.825569080122043e-05, + "loss": 1.0246, + "step": 3705 + }, + { + "epoch": 0.2251655629139073, + "grad_norm": 0.26310989260673523, + "learning_rate": 8.824952623885576e-05, + "loss": 1.0828, + "step": 3706 + }, + { + "epoch": 0.2252263199465338, + "grad_norm": 0.23267211019992828, + "learning_rate": 8.82433602744473e-05, + "loss": 1.0759, + "step": 3707 + }, + { + "epoch": 0.22528707697916034, + "grad_norm": 0.21230845153331757, + "learning_rate": 8.823719290822108e-05, + "loss": 1.1449, + "step": 3708 + }, + { + "epoch": 0.22534783401178687, + "grad_norm": 0.30772846937179565, + "learning_rate": 8.823102414040317e-05, + "loss": 1.1778, + "step": 3709 + }, + { + "epoch": 0.22540859104441338, + "grad_norm": 1.2115049362182617, + "learning_rate": 8.822485397121968e-05, + "loss": 1.1228, + "step": 3710 + }, + { + "epoch": 0.22546934807703992, + "grad_norm": 0.24158276617527008, + "learning_rate": 8.821868240089676e-05, + "loss": 1.1417, + "step": 3711 + }, + { + "epoch": 0.22553010510966645, + "grad_norm": 0.3266376852989197, + "learning_rate": 8.821250942966069e-05, + "loss": 1.146, + "step": 3712 + }, + { + "epoch": 0.22559086214229296, + "grad_norm": 0.2102629542350769, + "learning_rate": 8.820633505773766e-05, + "loss": 1.095, + "step": 3713 + }, + { + "epoch": 0.2256516191749195, + "grad_norm": 0.3125240206718445, + "learning_rate": 8.820015928535406e-05, + "loss": 1.1252, + "step": 3714 + }, + { + "epoch": 0.22571237620754603, + "grad_norm": 0.24530869722366333, + "learning_rate": 8.819398211273623e-05, + "loss": 1.0932, + "step": 3715 + }, + { + "epoch": 0.22577313324017256, + "grad_norm": 0.23530815541744232, + "learning_rate": 8.81878035401106e-05, + "loss": 1.18, + "step": 3716 + }, + { + "epoch": 0.22583389027279907, + "grad_norm": 0.17486156523227692, + "learning_rate": 8.818162356770366e-05, + "loss": 1.0789, + "step": 3717 + }, + { + "epoch": 0.2258946473054256, + "grad_norm": 0.24670076370239258, + "learning_rate": 8.81754421957419e-05, + "loss": 1.1076, + "step": 3718 + }, + { + "epoch": 0.22595540433805214, + "grad_norm": 0.37098145484924316, + "learning_rate": 8.816925942445196e-05, + "loss": 1.0587, + "step": 3719 + }, + { + "epoch": 0.22601616137067865, + "grad_norm": 0.48902156949043274, + "learning_rate": 8.816307525406042e-05, + "loss": 1.2533, + "step": 3720 + }, + { + "epoch": 0.22607691840330518, + "grad_norm": 0.4928256869316101, + "learning_rate": 8.815688968479399e-05, + "loss": 1.1106, + "step": 3721 + }, + { + "epoch": 0.22613767543593172, + "grad_norm": 0.24221043288707733, + "learning_rate": 8.815070271687937e-05, + "loss": 1.1722, + "step": 3722 + }, + { + "epoch": 0.22619843246855822, + "grad_norm": 0.6080340147018433, + "learning_rate": 8.814451435054339e-05, + "loss": 1.1185, + "step": 3723 + }, + { + "epoch": 0.22625918950118476, + "grad_norm": 0.30628731846809387, + "learning_rate": 8.813832458601285e-05, + "loss": 1.1162, + "step": 3724 + }, + { + "epoch": 0.2263199465338113, + "grad_norm": 0.2866918444633484, + "learning_rate": 8.813213342351466e-05, + "loss": 1.232, + "step": 3725 + }, + { + "epoch": 0.2263807035664378, + "grad_norm": 0.36669132113456726, + "learning_rate": 8.812594086327574e-05, + "loss": 1.17, + "step": 3726 + }, + { + "epoch": 0.22644146059906434, + "grad_norm": 0.3828333914279938, + "learning_rate": 8.811974690552311e-05, + "loss": 1.1905, + "step": 3727 + }, + { + "epoch": 0.22650221763169087, + "grad_norm": 0.190073624253273, + "learning_rate": 8.811355155048377e-05, + "loss": 1.1215, + "step": 3728 + }, + { + "epoch": 0.2265629746643174, + "grad_norm": 0.46124276518821716, + "learning_rate": 8.810735479838483e-05, + "loss": 1.077, + "step": 3729 + }, + { + "epoch": 0.22662373169694391, + "grad_norm": 0.5575589537620544, + "learning_rate": 8.810115664945343e-05, + "loss": 1.1886, + "step": 3730 + }, + { + "epoch": 0.22668448872957045, + "grad_norm": 0.4777982831001282, + "learning_rate": 8.809495710391678e-05, + "loss": 1.2552, + "step": 3731 + }, + { + "epoch": 0.22674524576219698, + "grad_norm": 0.5398675203323364, + "learning_rate": 8.808875616200211e-05, + "loss": 1.0996, + "step": 3732 + }, + { + "epoch": 0.2268060027948235, + "grad_norm": 0.896020770072937, + "learning_rate": 8.808255382393672e-05, + "loss": 1.1285, + "step": 3733 + }, + { + "epoch": 0.22686675982745003, + "grad_norm": 0.5593417286872864, + "learning_rate": 8.807635008994796e-05, + "loss": 1.1415, + "step": 3734 + }, + { + "epoch": 0.22692751686007656, + "grad_norm": 0.5072895288467407, + "learning_rate": 8.807014496026322e-05, + "loss": 1.205, + "step": 3735 + }, + { + "epoch": 0.22698827389270307, + "grad_norm": 0.2821612060070038, + "learning_rate": 8.806393843510996e-05, + "loss": 1.1578, + "step": 3736 + }, + { + "epoch": 0.2270490309253296, + "grad_norm": 0.2629121243953705, + "learning_rate": 8.805773051471568e-05, + "loss": 1.0743, + "step": 3737 + }, + { + "epoch": 0.22710978795795614, + "grad_norm": 0.7700924277305603, + "learning_rate": 8.805152119930796e-05, + "loss": 1.1325, + "step": 3738 + }, + { + "epoch": 0.22717054499058267, + "grad_norm": 0.21515251696109772, + "learning_rate": 8.804531048911433e-05, + "loss": 1.1338, + "step": 3739 + }, + { + "epoch": 0.22723130202320918, + "grad_norm": 0.33193105459213257, + "learning_rate": 8.803909838436251e-05, + "loss": 1.1791, + "step": 3740 + }, + { + "epoch": 0.22729205905583572, + "grad_norm": 0.5152245163917542, + "learning_rate": 8.803288488528019e-05, + "loss": 1.0923, + "step": 3741 + }, + { + "epoch": 0.22735281608846225, + "grad_norm": 0.30317363142967224, + "learning_rate": 8.802666999209512e-05, + "loss": 1.0566, + "step": 3742 + }, + { + "epoch": 0.22741357312108876, + "grad_norm": 0.38615602254867554, + "learning_rate": 8.802045370503512e-05, + "loss": 1.0836, + "step": 3743 + }, + { + "epoch": 0.2274743301537153, + "grad_norm": 0.27767929434776306, + "learning_rate": 8.801423602432801e-05, + "loss": 1.1666, + "step": 3744 + }, + { + "epoch": 0.22753508718634183, + "grad_norm": 0.25688058137893677, + "learning_rate": 8.800801695020176e-05, + "loss": 1.1946, + "step": 3745 + }, + { + "epoch": 0.22759584421896834, + "grad_norm": 0.36349722743034363, + "learning_rate": 8.800179648288429e-05, + "loss": 1.0854, + "step": 3746 + }, + { + "epoch": 0.22765660125159487, + "grad_norm": 0.3987286686897278, + "learning_rate": 8.799557462260361e-05, + "loss": 1.0644, + "step": 3747 + }, + { + "epoch": 0.2277173582842214, + "grad_norm": 0.2504458725452423, + "learning_rate": 8.798935136958782e-05, + "loss": 1.0449, + "step": 3748 + }, + { + "epoch": 0.2277781153168479, + "grad_norm": 0.274185448884964, + "learning_rate": 8.798312672406501e-05, + "loss": 1.1827, + "step": 3749 + }, + { + "epoch": 0.22783887234947445, + "grad_norm": 0.8240941762924194, + "learning_rate": 8.797690068626334e-05, + "loss": 1.3477, + "step": 3750 + }, + { + "epoch": 0.22789962938210098, + "grad_norm": 0.24660734832286835, + "learning_rate": 8.797067325641104e-05, + "loss": 1.1597, + "step": 3751 + }, + { + "epoch": 0.22796038641472752, + "grad_norm": 0.43585994839668274, + "learning_rate": 8.796444443473637e-05, + "loss": 1.1642, + "step": 3752 + }, + { + "epoch": 0.22802114344735402, + "grad_norm": 0.2636988162994385, + "learning_rate": 8.795821422146766e-05, + "loss": 1.0418, + "step": 3753 + }, + { + "epoch": 0.22808190047998056, + "grad_norm": 0.43805158138275146, + "learning_rate": 8.795198261683326e-05, + "loss": 1.2813, + "step": 3754 + }, + { + "epoch": 0.2281426575126071, + "grad_norm": 0.3826257884502411, + "learning_rate": 8.794574962106161e-05, + "loss": 1.1893, + "step": 3755 + }, + { + "epoch": 0.2282034145452336, + "grad_norm": 0.2315654158592224, + "learning_rate": 8.793951523438116e-05, + "loss": 1.1048, + "step": 3756 + }, + { + "epoch": 0.22826417157786014, + "grad_norm": 0.3920183777809143, + "learning_rate": 8.793327945702046e-05, + "loss": 1.0479, + "step": 3757 + }, + { + "epoch": 0.22832492861048667, + "grad_norm": 0.22065088152885437, + "learning_rate": 8.792704228920805e-05, + "loss": 1.0493, + "step": 3758 + }, + { + "epoch": 0.22838568564311318, + "grad_norm": 0.5010303854942322, + "learning_rate": 8.79208037311726e-05, + "loss": 1.1296, + "step": 3759 + }, + { + "epoch": 0.2284464426757397, + "grad_norm": 0.31396186351776123, + "learning_rate": 8.791456378314273e-05, + "loss": 1.1631, + "step": 3760 + }, + { + "epoch": 0.22850719970836625, + "grad_norm": 0.3177626132965088, + "learning_rate": 8.790832244534721e-05, + "loss": 1.1979, + "step": 3761 + }, + { + "epoch": 0.22856795674099278, + "grad_norm": 0.3218107223510742, + "learning_rate": 8.790207971801481e-05, + "loss": 1.313, + "step": 3762 + }, + { + "epoch": 0.2286287137736193, + "grad_norm": 0.7122443914413452, + "learning_rate": 8.789583560137435e-05, + "loss": 1.1739, + "step": 3763 + }, + { + "epoch": 0.22868947080624583, + "grad_norm": 0.2781575322151184, + "learning_rate": 8.78895900956547e-05, + "loss": 1.0793, + "step": 3764 + }, + { + "epoch": 0.22875022783887236, + "grad_norm": 0.3846098780632019, + "learning_rate": 8.78833432010848e-05, + "loss": 1.1741, + "step": 3765 + }, + { + "epoch": 0.22881098487149887, + "grad_norm": 4.030361652374268, + "learning_rate": 8.787709491789364e-05, + "loss": 1.1015, + "step": 3766 + }, + { + "epoch": 0.2288717419041254, + "grad_norm": 0.430060476064682, + "learning_rate": 8.787084524631025e-05, + "loss": 1.1273, + "step": 3767 + }, + { + "epoch": 0.22893249893675194, + "grad_norm": 0.5940700769424438, + "learning_rate": 8.786459418656369e-05, + "loss": 1.2135, + "step": 3768 + }, + { + "epoch": 0.22899325596937845, + "grad_norm": 0.3549436032772064, + "learning_rate": 8.785834173888315e-05, + "loss": 1.1719, + "step": 3769 + }, + { + "epoch": 0.22905401300200498, + "grad_norm": 0.322826087474823, + "learning_rate": 8.785208790349773e-05, + "loss": 1.1588, + "step": 3770 + }, + { + "epoch": 0.22911477003463152, + "grad_norm": 0.2573994994163513, + "learning_rate": 8.784583268063673e-05, + "loss": 1.1582, + "step": 3771 + }, + { + "epoch": 0.22917552706725802, + "grad_norm": 0.2851214110851288, + "learning_rate": 8.783957607052941e-05, + "loss": 1.0962, + "step": 3772 + }, + { + "epoch": 0.22923628409988456, + "grad_norm": 0.35574695467948914, + "learning_rate": 8.783331807340514e-05, + "loss": 1.2712, + "step": 3773 + }, + { + "epoch": 0.2292970411325111, + "grad_norm": 0.3030840754508972, + "learning_rate": 8.782705868949326e-05, + "loss": 1.261, + "step": 3774 + }, + { + "epoch": 0.22935779816513763, + "grad_norm": 0.3104516267776489, + "learning_rate": 8.782079791902322e-05, + "loss": 1.0712, + "step": 3775 + }, + { + "epoch": 0.22941855519776413, + "grad_norm": 0.5127542614936829, + "learning_rate": 8.781453576222455e-05, + "loss": 1.102, + "step": 3776 + }, + { + "epoch": 0.22947931223039067, + "grad_norm": 0.27611440420150757, + "learning_rate": 8.780827221932675e-05, + "loss": 1.1334, + "step": 3777 + }, + { + "epoch": 0.2295400692630172, + "grad_norm": 0.352556049823761, + "learning_rate": 8.780200729055943e-05, + "loss": 1.1898, + "step": 3778 + }, + { + "epoch": 0.2296008262956437, + "grad_norm": 0.388899028301239, + "learning_rate": 8.77957409761522e-05, + "loss": 1.167, + "step": 3779 + }, + { + "epoch": 0.22966158332827025, + "grad_norm": 0.36045005917549133, + "learning_rate": 8.778947327633481e-05, + "loss": 1.2404, + "step": 3780 + }, + { + "epoch": 0.22972234036089678, + "grad_norm": 0.6858901381492615, + "learning_rate": 8.778320419133697e-05, + "loss": 1.1366, + "step": 3781 + }, + { + "epoch": 0.2297830973935233, + "grad_norm": 0.24919074773788452, + "learning_rate": 8.777693372138846e-05, + "loss": 1.1502, + "step": 3782 + }, + { + "epoch": 0.22984385442614982, + "grad_norm": 0.3068561255931854, + "learning_rate": 8.777066186671915e-05, + "loss": 1.1057, + "step": 3783 + }, + { + "epoch": 0.22990461145877636, + "grad_norm": 0.31032705307006836, + "learning_rate": 8.776438862755893e-05, + "loss": 1.142, + "step": 3784 + }, + { + "epoch": 0.22996536849140287, + "grad_norm": 0.305415540933609, + "learning_rate": 8.775811400413774e-05, + "loss": 1.1496, + "step": 3785 + }, + { + "epoch": 0.2300261255240294, + "grad_norm": 0.43146246671676636, + "learning_rate": 8.775183799668559e-05, + "loss": 1.1578, + "step": 3786 + }, + { + "epoch": 0.23008688255665594, + "grad_norm": 2.1299641132354736, + "learning_rate": 8.774556060543253e-05, + "loss": 1.1624, + "step": 3787 + }, + { + "epoch": 0.23014763958928247, + "grad_norm": 0.2741229236125946, + "learning_rate": 8.773928183060862e-05, + "loss": 1.1451, + "step": 3788 + }, + { + "epoch": 0.23020839662190898, + "grad_norm": 0.21930907666683197, + "learning_rate": 8.773300167244407e-05, + "loss": 1.0813, + "step": 3789 + }, + { + "epoch": 0.2302691536545355, + "grad_norm": 0.27755874395370483, + "learning_rate": 8.772672013116903e-05, + "loss": 1.0959, + "step": 3790 + }, + { + "epoch": 0.23032991068716205, + "grad_norm": 0.25883546471595764, + "learning_rate": 8.772043720701378e-05, + "loss": 1.0756, + "step": 3791 + }, + { + "epoch": 0.23039066771978856, + "grad_norm": 0.29824262857437134, + "learning_rate": 8.771415290020862e-05, + "loss": 1.1922, + "step": 3792 + }, + { + "epoch": 0.2304514247524151, + "grad_norm": 0.5571320652961731, + "learning_rate": 8.770786721098387e-05, + "loss": 1.2694, + "step": 3793 + }, + { + "epoch": 0.23051218178504163, + "grad_norm": 0.20247036218643188, + "learning_rate": 8.770158013956999e-05, + "loss": 1.0888, + "step": 3794 + }, + { + "epoch": 0.23057293881766813, + "grad_norm": 0.22159387171268463, + "learning_rate": 8.769529168619739e-05, + "loss": 1.0642, + "step": 3795 + }, + { + "epoch": 0.23063369585029467, + "grad_norm": 0.4019584357738495, + "learning_rate": 8.768900185109658e-05, + "loss": 1.1206, + "step": 3796 + }, + { + "epoch": 0.2306944528829212, + "grad_norm": 0.21337725222110748, + "learning_rate": 8.768271063449812e-05, + "loss": 1.0711, + "step": 3797 + }, + { + "epoch": 0.23075520991554774, + "grad_norm": 0.30635541677474976, + "learning_rate": 8.767641803663262e-05, + "loss": 1.1846, + "step": 3798 + }, + { + "epoch": 0.23081596694817424, + "grad_norm": 0.3011818528175354, + "learning_rate": 8.767012405773074e-05, + "loss": 1.2325, + "step": 3799 + }, + { + "epoch": 0.23087672398080078, + "grad_norm": 0.3766297698020935, + "learning_rate": 8.766382869802318e-05, + "loss": 1.3281, + "step": 3800 + }, + { + "epoch": 0.23093748101342731, + "grad_norm": 0.6781142950057983, + "learning_rate": 8.76575319577407e-05, + "loss": 1.4112, + "step": 3801 + }, + { + "epoch": 0.23099823804605382, + "grad_norm": 0.2550434172153473, + "learning_rate": 8.76512338371141e-05, + "loss": 1.0662, + "step": 3802 + }, + { + "epoch": 0.23105899507868036, + "grad_norm": 0.21415716409683228, + "learning_rate": 8.764493433637426e-05, + "loss": 1.2287, + "step": 3803 + }, + { + "epoch": 0.2311197521113069, + "grad_norm": 0.43630123138427734, + "learning_rate": 8.763863345575208e-05, + "loss": 1.2675, + "step": 3804 + }, + { + "epoch": 0.2311805091439334, + "grad_norm": 0.25498470664024353, + "learning_rate": 8.763233119547849e-05, + "loss": 1.0939, + "step": 3805 + }, + { + "epoch": 0.23124126617655993, + "grad_norm": 0.47066742181777954, + "learning_rate": 8.762602755578454e-05, + "loss": 1.1876, + "step": 3806 + }, + { + "epoch": 0.23130202320918647, + "grad_norm": 0.5968990921974182, + "learning_rate": 8.761972253690127e-05, + "loss": 1.0811, + "step": 3807 + }, + { + "epoch": 0.23136278024181298, + "grad_norm": 0.3829292058944702, + "learning_rate": 8.76134161390598e-05, + "loss": 1.1679, + "step": 3808 + }, + { + "epoch": 0.2314235372744395, + "grad_norm": 0.2810189723968506, + "learning_rate": 8.76071083624913e-05, + "loss": 1.1109, + "step": 3809 + }, + { + "epoch": 0.23148429430706605, + "grad_norm": 0.2997722327709198, + "learning_rate": 8.760079920742698e-05, + "loss": 1.1188, + "step": 3810 + }, + { + "epoch": 0.23154505133969258, + "grad_norm": 0.24355408549308777, + "learning_rate": 8.759448867409807e-05, + "loss": 1.0298, + "step": 3811 + }, + { + "epoch": 0.2316058083723191, + "grad_norm": 0.40661489963531494, + "learning_rate": 8.758817676273593e-05, + "loss": 1.1892, + "step": 3812 + }, + { + "epoch": 0.23166656540494562, + "grad_norm": 0.273549348115921, + "learning_rate": 8.758186347357191e-05, + "loss": 1.2262, + "step": 3813 + }, + { + "epoch": 0.23172732243757216, + "grad_norm": 0.1962147057056427, + "learning_rate": 8.75755488068374e-05, + "loss": 1.1518, + "step": 3814 + }, + { + "epoch": 0.23178807947019867, + "grad_norm": 0.4252879321575165, + "learning_rate": 8.75692327627639e-05, + "loss": 1.0894, + "step": 3815 + }, + { + "epoch": 0.2318488365028252, + "grad_norm": 0.48274168372154236, + "learning_rate": 8.756291534158292e-05, + "loss": 1.0787, + "step": 3816 + }, + { + "epoch": 0.23190959353545174, + "grad_norm": 1.3644229173660278, + "learning_rate": 8.755659654352599e-05, + "loss": 1.0887, + "step": 3817 + }, + { + "epoch": 0.23197035056807824, + "grad_norm": 0.7444817423820496, + "learning_rate": 8.755027636882478e-05, + "loss": 1.0656, + "step": 3818 + }, + { + "epoch": 0.23203110760070478, + "grad_norm": 0.25976666808128357, + "learning_rate": 8.754395481771091e-05, + "loss": 1.1329, + "step": 3819 + }, + { + "epoch": 0.2320918646333313, + "grad_norm": 0.4177325963973999, + "learning_rate": 8.753763189041614e-05, + "loss": 1.0594, + "step": 3820 + }, + { + "epoch": 0.23215262166595785, + "grad_norm": 0.5769382119178772, + "learning_rate": 8.75313075871722e-05, + "loss": 1.176, + "step": 3821 + }, + { + "epoch": 0.23221337869858436, + "grad_norm": 0.2755463123321533, + "learning_rate": 8.752498190821095e-05, + "loss": 1.0824, + "step": 3822 + }, + { + "epoch": 0.2322741357312109, + "grad_norm": 0.32161498069763184, + "learning_rate": 8.75186548537642e-05, + "loss": 1.0886, + "step": 3823 + }, + { + "epoch": 0.23233489276383743, + "grad_norm": 0.2344479113817215, + "learning_rate": 8.751232642406392e-05, + "loss": 1.1258, + "step": 3824 + }, + { + "epoch": 0.23239564979646393, + "grad_norm": 0.6590091586112976, + "learning_rate": 8.750599661934206e-05, + "loss": 1.1926, + "step": 3825 + }, + { + "epoch": 0.23245640682909047, + "grad_norm": 0.22046126425266266, + "learning_rate": 8.749966543983066e-05, + "loss": 1.1236, + "step": 3826 + }, + { + "epoch": 0.232517163861717, + "grad_norm": 0.2521543502807617, + "learning_rate": 8.749333288576177e-05, + "loss": 1.2596, + "step": 3827 + }, + { + "epoch": 0.2325779208943435, + "grad_norm": 0.31845536828041077, + "learning_rate": 8.748699895736752e-05, + "loss": 1.067, + "step": 3828 + }, + { + "epoch": 0.23263867792697004, + "grad_norm": 0.27097252011299133, + "learning_rate": 8.748066365488006e-05, + "loss": 1.1074, + "step": 3829 + }, + { + "epoch": 0.23269943495959658, + "grad_norm": 0.23012784123420715, + "learning_rate": 8.747432697853164e-05, + "loss": 1.1024, + "step": 3830 + }, + { + "epoch": 0.2327601919922231, + "grad_norm": 0.2217753827571869, + "learning_rate": 8.74679889285545e-05, + "loss": 1.1185, + "step": 3831 + }, + { + "epoch": 0.23282094902484962, + "grad_norm": 0.285765677690506, + "learning_rate": 8.746164950518101e-05, + "loss": 1.2425, + "step": 3832 + }, + { + "epoch": 0.23288170605747616, + "grad_norm": 0.3414607048034668, + "learning_rate": 8.745530870864351e-05, + "loss": 1.0627, + "step": 3833 + }, + { + "epoch": 0.2329424630901027, + "grad_norm": 0.23624590039253235, + "learning_rate": 8.744896653917443e-05, + "loss": 1.0412, + "step": 3834 + }, + { + "epoch": 0.2330032201227292, + "grad_norm": 0.3137994110584259, + "learning_rate": 8.744262299700623e-05, + "loss": 1.0776, + "step": 3835 + }, + { + "epoch": 0.23306397715535573, + "grad_norm": 0.2694247364997864, + "learning_rate": 8.743627808237146e-05, + "loss": 1.2517, + "step": 3836 + }, + { + "epoch": 0.23312473418798227, + "grad_norm": 0.19919265806674957, + "learning_rate": 8.742993179550267e-05, + "loss": 1.1995, + "step": 3837 + }, + { + "epoch": 0.23318549122060878, + "grad_norm": 0.27168017625808716, + "learning_rate": 8.74235841366325e-05, + "loss": 1.2579, + "step": 3838 + }, + { + "epoch": 0.2332462482532353, + "grad_norm": 0.264413446187973, + "learning_rate": 8.741723510599363e-05, + "loss": 1.0398, + "step": 3839 + }, + { + "epoch": 0.23330700528586185, + "grad_norm": 0.199133038520813, + "learning_rate": 8.741088470381877e-05, + "loss": 1.0641, + "step": 3840 + }, + { + "epoch": 0.23336776231848835, + "grad_norm": 0.376117080450058, + "learning_rate": 8.740453293034069e-05, + "loss": 1.0764, + "step": 3841 + }, + { + "epoch": 0.2334285193511149, + "grad_norm": 0.20632115006446838, + "learning_rate": 8.739817978579223e-05, + "loss": 1.0669, + "step": 3842 + }, + { + "epoch": 0.23348927638374142, + "grad_norm": 0.23657606542110443, + "learning_rate": 8.739182527040626e-05, + "loss": 1.1161, + "step": 3843 + }, + { + "epoch": 0.23355003341636796, + "grad_norm": 0.3815009891986847, + "learning_rate": 8.738546938441571e-05, + "loss": 1.3162, + "step": 3844 + }, + { + "epoch": 0.23361079044899447, + "grad_norm": 0.2095230221748352, + "learning_rate": 8.737911212805355e-05, + "loss": 1.1744, + "step": 3845 + }, + { + "epoch": 0.233671547481621, + "grad_norm": 0.5384068489074707, + "learning_rate": 8.737275350155281e-05, + "loss": 1.243, + "step": 3846 + }, + { + "epoch": 0.23373230451424754, + "grad_norm": 0.35408055782318115, + "learning_rate": 8.736639350514658e-05, + "loss": 1.1085, + "step": 3847 + }, + { + "epoch": 0.23379306154687404, + "grad_norm": 0.4657049775123596, + "learning_rate": 8.736003213906795e-05, + "loss": 1.0717, + "step": 3848 + }, + { + "epoch": 0.23385381857950058, + "grad_norm": 0.27329879999160767, + "learning_rate": 8.735366940355014e-05, + "loss": 1.0939, + "step": 3849 + }, + { + "epoch": 0.2339145756121271, + "grad_norm": 0.260672926902771, + "learning_rate": 8.734730529882634e-05, + "loss": 1.1249, + "step": 3850 + }, + { + "epoch": 0.23397533264475362, + "grad_norm": 0.5476902723312378, + "learning_rate": 8.734093982512986e-05, + "loss": 1.2177, + "step": 3851 + }, + { + "epoch": 0.23403608967738015, + "grad_norm": 0.24789705872535706, + "learning_rate": 8.733457298269401e-05, + "loss": 1.1405, + "step": 3852 + }, + { + "epoch": 0.2340968467100067, + "grad_norm": 0.23033031821250916, + "learning_rate": 8.732820477175217e-05, + "loss": 1.1705, + "step": 3853 + }, + { + "epoch": 0.2341576037426332, + "grad_norm": 0.24374619126319885, + "learning_rate": 8.732183519253776e-05, + "loss": 1.1894, + "step": 3854 + }, + { + "epoch": 0.23421836077525973, + "grad_norm": 0.22070418298244476, + "learning_rate": 8.731546424528428e-05, + "loss": 1.1359, + "step": 3855 + }, + { + "epoch": 0.23427911780788627, + "grad_norm": 0.2108503133058548, + "learning_rate": 8.730909193022523e-05, + "loss": 1.0419, + "step": 3856 + }, + { + "epoch": 0.2343398748405128, + "grad_norm": 0.3153754472732544, + "learning_rate": 8.730271824759421e-05, + "loss": 1.1768, + "step": 3857 + }, + { + "epoch": 0.2344006318731393, + "grad_norm": 3.3047661781311035, + "learning_rate": 8.729634319762486e-05, + "loss": 1.2883, + "step": 3858 + }, + { + "epoch": 0.23446138890576584, + "grad_norm": 0.259928435087204, + "learning_rate": 8.728996678055081e-05, + "loss": 1.0732, + "step": 3859 + }, + { + "epoch": 0.23452214593839238, + "grad_norm": 0.31056806445121765, + "learning_rate": 8.728358899660585e-05, + "loss": 1.0764, + "step": 3860 + }, + { + "epoch": 0.2345829029710189, + "grad_norm": 4.2688307762146, + "learning_rate": 8.727720984602371e-05, + "loss": 1.0685, + "step": 3861 + }, + { + "epoch": 0.23464366000364542, + "grad_norm": 0.44061970710754395, + "learning_rate": 8.727082932903823e-05, + "loss": 1.0627, + "step": 3862 + }, + { + "epoch": 0.23470441703627196, + "grad_norm": 0.1837361603975296, + "learning_rate": 8.72644474458833e-05, + "loss": 1.0816, + "step": 3863 + }, + { + "epoch": 0.23476517406889846, + "grad_norm": 3.8701095581054688, + "learning_rate": 8.725806419679287e-05, + "loss": 1.1018, + "step": 3864 + }, + { + "epoch": 0.234825931101525, + "grad_norm": 0.5309779047966003, + "learning_rate": 8.725167958200088e-05, + "loss": 1.0894, + "step": 3865 + }, + { + "epoch": 0.23488668813415153, + "grad_norm": 0.4665716290473938, + "learning_rate": 8.724529360174137e-05, + "loss": 1.1669, + "step": 3866 + }, + { + "epoch": 0.23494744516677807, + "grad_norm": 0.9781427979469299, + "learning_rate": 8.723890625624844e-05, + "loss": 1.2483, + "step": 3867 + }, + { + "epoch": 0.23500820219940458, + "grad_norm": 0.5543884038925171, + "learning_rate": 8.723251754575619e-05, + "loss": 1.2423, + "step": 3868 + }, + { + "epoch": 0.2350689592320311, + "grad_norm": 0.4816453456878662, + "learning_rate": 8.72261274704988e-05, + "loss": 1.0559, + "step": 3869 + }, + { + "epoch": 0.23512971626465765, + "grad_norm": 0.7280380129814148, + "learning_rate": 8.721973603071053e-05, + "loss": 1.0894, + "step": 3870 + }, + { + "epoch": 0.23519047329728415, + "grad_norm": 0.30582621693611145, + "learning_rate": 8.721334322662564e-05, + "loss": 1.2293, + "step": 3871 + }, + { + "epoch": 0.2352512303299107, + "grad_norm": 1.2320599555969238, + "learning_rate": 8.720694905847844e-05, + "loss": 1.1316, + "step": 3872 + }, + { + "epoch": 0.23531198736253722, + "grad_norm": 0.5318282246589661, + "learning_rate": 8.720055352650337e-05, + "loss": 1.1363, + "step": 3873 + }, + { + "epoch": 0.23537274439516373, + "grad_norm": 0.5668740272521973, + "learning_rate": 8.719415663093479e-05, + "loss": 1.1101, + "step": 3874 + }, + { + "epoch": 0.23543350142779026, + "grad_norm": 0.7241581678390503, + "learning_rate": 8.71877583720072e-05, + "loss": 1.1366, + "step": 3875 + }, + { + "epoch": 0.2354942584604168, + "grad_norm": 0.8203892111778259, + "learning_rate": 8.718135874995516e-05, + "loss": 1.1159, + "step": 3876 + }, + { + "epoch": 0.2355550154930433, + "grad_norm": 0.1724739372730255, + "learning_rate": 8.717495776501323e-05, + "loss": 1.0637, + "step": 3877 + }, + { + "epoch": 0.23561577252566984, + "grad_norm": 0.6078988313674927, + "learning_rate": 8.716855541741603e-05, + "loss": 1.0956, + "step": 3878 + }, + { + "epoch": 0.23567652955829638, + "grad_norm": 0.8598876595497131, + "learning_rate": 8.716215170739825e-05, + "loss": 1.1599, + "step": 3879 + }, + { + "epoch": 0.2357372865909229, + "grad_norm": 0.27505892515182495, + "learning_rate": 8.715574663519462e-05, + "loss": 1.3042, + "step": 3880 + }, + { + "epoch": 0.23579804362354942, + "grad_norm": 0.7888695001602173, + "learning_rate": 8.71493402010399e-05, + "loss": 1.2485, + "step": 3881 + }, + { + "epoch": 0.23585880065617595, + "grad_norm": 0.2219749242067337, + "learning_rate": 8.714293240516894e-05, + "loss": 1.1504, + "step": 3882 + }, + { + "epoch": 0.2359195576888025, + "grad_norm": 0.6046619415283203, + "learning_rate": 8.713652324781662e-05, + "loss": 1.0878, + "step": 3883 + }, + { + "epoch": 0.235980314721429, + "grad_norm": 0.6110017895698547, + "learning_rate": 8.713011272921785e-05, + "loss": 1.3096, + "step": 3884 + }, + { + "epoch": 0.23604107175405553, + "grad_norm": 0.3499814569950104, + "learning_rate": 8.712370084960764e-05, + "loss": 1.104, + "step": 3885 + }, + { + "epoch": 0.23610182878668207, + "grad_norm": 0.7304849028587341, + "learning_rate": 8.711728760922097e-05, + "loss": 1.043, + "step": 3886 + }, + { + "epoch": 0.23616258581930857, + "grad_norm": 1.1178996562957764, + "learning_rate": 8.711087300829297e-05, + "loss": 1.1621, + "step": 3887 + }, + { + "epoch": 0.2362233428519351, + "grad_norm": 2.051408290863037, + "learning_rate": 8.710445704705874e-05, + "loss": 1.038, + "step": 3888 + }, + { + "epoch": 0.23628409988456164, + "grad_norm": 0.8726368546485901, + "learning_rate": 8.709803972575346e-05, + "loss": 1.1027, + "step": 3889 + }, + { + "epoch": 0.23634485691718815, + "grad_norm": 0.2506411373615265, + "learning_rate": 8.709162104461238e-05, + "loss": 1.0742, + "step": 3890 + }, + { + "epoch": 0.23640561394981469, + "grad_norm": 0.5322414040565491, + "learning_rate": 8.708520100387074e-05, + "loss": 1.1126, + "step": 3891 + }, + { + "epoch": 0.23646637098244122, + "grad_norm": 0.46142417192459106, + "learning_rate": 8.707877960376392e-05, + "loss": 1.097, + "step": 3892 + }, + { + "epoch": 0.23652712801506776, + "grad_norm": 0.4300399124622345, + "learning_rate": 8.707235684452723e-05, + "loss": 1.0887, + "step": 3893 + }, + { + "epoch": 0.23658788504769426, + "grad_norm": 0.564732015132904, + "learning_rate": 8.706593272639615e-05, + "loss": 1.0519, + "step": 3894 + }, + { + "epoch": 0.2366486420803208, + "grad_norm": 0.2832649350166321, + "learning_rate": 8.705950724960614e-05, + "loss": 1.0694, + "step": 3895 + }, + { + "epoch": 0.23670939911294733, + "grad_norm": 0.3859423100948334, + "learning_rate": 8.705308041439272e-05, + "loss": 1.1119, + "step": 3896 + }, + { + "epoch": 0.23677015614557384, + "grad_norm": 0.33983322978019714, + "learning_rate": 8.70466522209915e-05, + "loss": 1.0727, + "step": 3897 + }, + { + "epoch": 0.23683091317820038, + "grad_norm": 0.27681493759155273, + "learning_rate": 8.704022266963807e-05, + "loss": 1.2187, + "step": 3898 + }, + { + "epoch": 0.2368916702108269, + "grad_norm": 0.5306503772735596, + "learning_rate": 8.703379176056811e-05, + "loss": 1.166, + "step": 3899 + }, + { + "epoch": 0.23695242724345342, + "grad_norm": 0.2714451551437378, + "learning_rate": 8.702735949401736e-05, + "loss": 1.1567, + "step": 3900 + }, + { + "epoch": 0.23701318427607995, + "grad_norm": 0.32279518246650696, + "learning_rate": 8.702092587022159e-05, + "loss": 1.0998, + "step": 3901 + }, + { + "epoch": 0.2370739413087065, + "grad_norm": 0.4036634564399719, + "learning_rate": 8.701449088941662e-05, + "loss": 1.0392, + "step": 3902 + }, + { + "epoch": 0.23713469834133302, + "grad_norm": 0.1997005045413971, + "learning_rate": 8.700805455183835e-05, + "loss": 1.1646, + "step": 3903 + }, + { + "epoch": 0.23719545537395953, + "grad_norm": 0.3059917092323303, + "learning_rate": 8.700161685772267e-05, + "loss": 1.0944, + "step": 3904 + }, + { + "epoch": 0.23725621240658606, + "grad_norm": 0.26258715987205505, + "learning_rate": 8.699517780730557e-05, + "loss": 1.1057, + "step": 3905 + }, + { + "epoch": 0.2373169694392126, + "grad_norm": 0.43534669280052185, + "learning_rate": 8.698873740082309e-05, + "loss": 1.1708, + "step": 3906 + }, + { + "epoch": 0.2373777264718391, + "grad_norm": 0.31681427359580994, + "learning_rate": 8.698229563851129e-05, + "loss": 1.1193, + "step": 3907 + }, + { + "epoch": 0.23743848350446564, + "grad_norm": 0.3385699689388275, + "learning_rate": 8.697585252060629e-05, + "loss": 1.1264, + "step": 3908 + }, + { + "epoch": 0.23749924053709218, + "grad_norm": 0.22170230746269226, + "learning_rate": 8.696940804734426e-05, + "loss": 1.0815, + "step": 3909 + }, + { + "epoch": 0.23755999756971868, + "grad_norm": 0.3702619969844818, + "learning_rate": 8.696296221896144e-05, + "loss": 1.0943, + "step": 3910 + }, + { + "epoch": 0.23762075460234522, + "grad_norm": 0.1804535835981369, + "learning_rate": 8.695651503569409e-05, + "loss": 1.114, + "step": 3911 + }, + { + "epoch": 0.23768151163497175, + "grad_norm": 0.25025054812431335, + "learning_rate": 8.695006649777854e-05, + "loss": 1.1147, + "step": 3912 + }, + { + "epoch": 0.23774226866759826, + "grad_norm": 1.3843361139297485, + "learning_rate": 8.694361660545116e-05, + "loss": 1.1281, + "step": 3913 + }, + { + "epoch": 0.2378030257002248, + "grad_norm": 0.23322568833827972, + "learning_rate": 8.693716535894837e-05, + "loss": 1.071, + "step": 3914 + }, + { + "epoch": 0.23786378273285133, + "grad_norm": 0.33052903413772583, + "learning_rate": 8.693071275850666e-05, + "loss": 1.1594, + "step": 3915 + }, + { + "epoch": 0.23792453976547787, + "grad_norm": 0.23573338985443115, + "learning_rate": 8.692425880436253e-05, + "loss": 1.1172, + "step": 3916 + }, + { + "epoch": 0.23798529679810437, + "grad_norm": 0.17324313521385193, + "learning_rate": 8.691780349675255e-05, + "loss": 1.0639, + "step": 3917 + }, + { + "epoch": 0.2380460538307309, + "grad_norm": 0.28432953357696533, + "learning_rate": 8.691134683591334e-05, + "loss": 1.1005, + "step": 3918 + }, + { + "epoch": 0.23810681086335744, + "grad_norm": 0.202653706073761, + "learning_rate": 8.690488882208159e-05, + "loss": 1.1282, + "step": 3919 + }, + { + "epoch": 0.23816756789598395, + "grad_norm": 0.2898375988006592, + "learning_rate": 8.6898429455494e-05, + "loss": 1.1326, + "step": 3920 + }, + { + "epoch": 0.23822832492861049, + "grad_norm": 0.17361988127231598, + "learning_rate": 8.689196873638734e-05, + "loss": 1.1137, + "step": 3921 + }, + { + "epoch": 0.23828908196123702, + "grad_norm": 0.187607541680336, + "learning_rate": 8.688550666499844e-05, + "loss": 1.0952, + "step": 3922 + }, + { + "epoch": 0.23834983899386353, + "grad_norm": 0.3016396760940552, + "learning_rate": 8.687904324156417e-05, + "loss": 1.1308, + "step": 3923 + }, + { + "epoch": 0.23841059602649006, + "grad_norm": 0.2506144344806671, + "learning_rate": 8.687257846632142e-05, + "loss": 1.0706, + "step": 3924 + }, + { + "epoch": 0.2384713530591166, + "grad_norm": 0.2849816083908081, + "learning_rate": 8.686611233950721e-05, + "loss": 1.1386, + "step": 3925 + }, + { + "epoch": 0.23853211009174313, + "grad_norm": 0.43760111927986145, + "learning_rate": 8.685964486135852e-05, + "loss": 1.0831, + "step": 3926 + }, + { + "epoch": 0.23859286712436964, + "grad_norm": 0.19355927407741547, + "learning_rate": 8.68531760321124e-05, + "loss": 1.0462, + "step": 3927 + }, + { + "epoch": 0.23865362415699617, + "grad_norm": 0.5281374454498291, + "learning_rate": 8.684670585200601e-05, + "loss": 1.1166, + "step": 3928 + }, + { + "epoch": 0.2387143811896227, + "grad_norm": 0.21112392842769623, + "learning_rate": 8.684023432127648e-05, + "loss": 1.186, + "step": 3929 + }, + { + "epoch": 0.23877513822224922, + "grad_norm": 0.3644089102745056, + "learning_rate": 8.683376144016106e-05, + "loss": 1.085, + "step": 3930 + }, + { + "epoch": 0.23883589525487575, + "grad_norm": 1.5819870233535767, + "learning_rate": 8.682728720889697e-05, + "loss": 1.0981, + "step": 3931 + }, + { + "epoch": 0.2388966522875023, + "grad_norm": 0.3216436207294464, + "learning_rate": 8.682081162772156e-05, + "loss": 1.0846, + "step": 3932 + }, + { + "epoch": 0.2389574093201288, + "grad_norm": 0.2139171063899994, + "learning_rate": 8.681433469687219e-05, + "loss": 1.1339, + "step": 3933 + }, + { + "epoch": 0.23901816635275533, + "grad_norm": 0.36131635308265686, + "learning_rate": 8.680785641658626e-05, + "loss": 1.1164, + "step": 3934 + }, + { + "epoch": 0.23907892338538186, + "grad_norm": 0.29613709449768066, + "learning_rate": 8.680137678710125e-05, + "loss": 1.0687, + "step": 3935 + }, + { + "epoch": 0.23913968041800837, + "grad_norm": 0.2938118577003479, + "learning_rate": 8.679489580865463e-05, + "loss": 1.1245, + "step": 3936 + }, + { + "epoch": 0.2392004374506349, + "grad_norm": 0.2208794802427292, + "learning_rate": 8.6788413481484e-05, + "loss": 1.107, + "step": 3937 + }, + { + "epoch": 0.23926119448326144, + "grad_norm": 0.197188600897789, + "learning_rate": 8.678192980582697e-05, + "loss": 1.0963, + "step": 3938 + }, + { + "epoch": 0.23932195151588798, + "grad_norm": 0.18471983075141907, + "learning_rate": 8.677544478192121e-05, + "loss": 1.1626, + "step": 3939 + }, + { + "epoch": 0.23938270854851448, + "grad_norm": 0.17389152944087982, + "learning_rate": 8.676895841000441e-05, + "loss": 1.066, + "step": 3940 + }, + { + "epoch": 0.23944346558114102, + "grad_norm": 0.1828600913286209, + "learning_rate": 8.676247069031431e-05, + "loss": 1.0361, + "step": 3941 + }, + { + "epoch": 0.23950422261376755, + "grad_norm": 0.29423803091049194, + "learning_rate": 8.675598162308875e-05, + "loss": 1.2581, + "step": 3942 + }, + { + "epoch": 0.23956497964639406, + "grad_norm": 0.23067732155323029, + "learning_rate": 8.67494912085656e-05, + "loss": 1.1615, + "step": 3943 + }, + { + "epoch": 0.2396257366790206, + "grad_norm": 0.2594285309314728, + "learning_rate": 8.674299944698271e-05, + "loss": 1.0693, + "step": 3944 + }, + { + "epoch": 0.23968649371164713, + "grad_norm": 0.20736142992973328, + "learning_rate": 8.67365063385781e-05, + "loss": 1.1127, + "step": 3945 + }, + { + "epoch": 0.23974725074427364, + "grad_norm": 0.22839221358299255, + "learning_rate": 8.673001188358974e-05, + "loss": 1.0438, + "step": 3946 + }, + { + "epoch": 0.23980800777690017, + "grad_norm": 0.18721593916416168, + "learning_rate": 8.672351608225568e-05, + "loss": 1.101, + "step": 3947 + }, + { + "epoch": 0.2398687648095267, + "grad_norm": 0.25217029452323914, + "learning_rate": 8.671701893481405e-05, + "loss": 1.1886, + "step": 3948 + }, + { + "epoch": 0.23992952184215324, + "grad_norm": 0.4437046945095062, + "learning_rate": 8.6710520441503e-05, + "loss": 1.2038, + "step": 3949 + }, + { + "epoch": 0.23999027887477975, + "grad_norm": 0.5329896807670593, + "learning_rate": 8.670402060256073e-05, + "loss": 1.101, + "step": 3950 + }, + { + "epoch": 0.24005103590740628, + "grad_norm": 0.23877553641796112, + "learning_rate": 8.669751941822549e-05, + "loss": 1.0771, + "step": 3951 + }, + { + "epoch": 0.24011179294003282, + "grad_norm": 0.20105764269828796, + "learning_rate": 8.669101688873558e-05, + "loss": 1.0873, + "step": 3952 + }, + { + "epoch": 0.24017254997265933, + "grad_norm": 0.2936873137950897, + "learning_rate": 8.668451301432934e-05, + "loss": 1.0862, + "step": 3953 + }, + { + "epoch": 0.24023330700528586, + "grad_norm": 0.2818056046962738, + "learning_rate": 8.66780077952452e-05, + "loss": 1.1109, + "step": 3954 + }, + { + "epoch": 0.2402940640379124, + "grad_norm": 0.221087247133255, + "learning_rate": 8.667150123172158e-05, + "loss": 1.0935, + "step": 3955 + }, + { + "epoch": 0.2403548210705389, + "grad_norm": 0.9349890947341919, + "learning_rate": 8.6664993323997e-05, + "loss": 1.2343, + "step": 3956 + }, + { + "epoch": 0.24041557810316544, + "grad_norm": 0.2955245077610016, + "learning_rate": 8.665848407231001e-05, + "loss": 1.0911, + "step": 3957 + }, + { + "epoch": 0.24047633513579197, + "grad_norm": 3.3875460624694824, + "learning_rate": 8.66519734768992e-05, + "loss": 1.1253, + "step": 3958 + }, + { + "epoch": 0.24053709216841848, + "grad_norm": 0.2986474335193634, + "learning_rate": 8.664546153800322e-05, + "loss": 1.1039, + "step": 3959 + }, + { + "epoch": 0.24059784920104502, + "grad_norm": 0.24428454041481018, + "learning_rate": 8.663894825586073e-05, + "loss": 1.1194, + "step": 3960 + }, + { + "epoch": 0.24065860623367155, + "grad_norm": 0.20833313465118408, + "learning_rate": 8.663243363071056e-05, + "loss": 1.1227, + "step": 3961 + }, + { + "epoch": 0.2407193632662981, + "grad_norm": 0.25138452649116516, + "learning_rate": 8.662591766279142e-05, + "loss": 1.2123, + "step": 3962 + }, + { + "epoch": 0.2407801202989246, + "grad_norm": 0.24809257686138153, + "learning_rate": 8.66194003523422e-05, + "loss": 1.0653, + "step": 3963 + }, + { + "epoch": 0.24084087733155113, + "grad_norm": 0.239301860332489, + "learning_rate": 8.661288169960177e-05, + "loss": 1.14, + "step": 3964 + }, + { + "epoch": 0.24090163436417766, + "grad_norm": 0.20981092751026154, + "learning_rate": 8.660636170480912e-05, + "loss": 1.1346, + "step": 3965 + }, + { + "epoch": 0.24096239139680417, + "grad_norm": 0.3041893243789673, + "learning_rate": 8.659984036820317e-05, + "loss": 1.0663, + "step": 3966 + }, + { + "epoch": 0.2410231484294307, + "grad_norm": 0.26527395844459534, + "learning_rate": 8.659331769002301e-05, + "loss": 1.2204, + "step": 3967 + }, + { + "epoch": 0.24108390546205724, + "grad_norm": 0.9073504209518433, + "learning_rate": 8.65867936705077e-05, + "loss": 1.233, + "step": 3968 + }, + { + "epoch": 0.24114466249468375, + "grad_norm": 0.22109931707382202, + "learning_rate": 8.658026830989642e-05, + "loss": 1.1144, + "step": 3969 + }, + { + "epoch": 0.24120541952731028, + "grad_norm": 0.36727145314216614, + "learning_rate": 8.657374160842833e-05, + "loss": 1.0454, + "step": 3970 + }, + { + "epoch": 0.24126617655993682, + "grad_norm": 0.32622721791267395, + "learning_rate": 8.656721356634267e-05, + "loss": 1.1708, + "step": 3971 + }, + { + "epoch": 0.24132693359256335, + "grad_norm": 0.5424013137817383, + "learning_rate": 8.656068418387873e-05, + "loss": 1.1075, + "step": 3972 + }, + { + "epoch": 0.24138769062518986, + "grad_norm": 0.29868772625923157, + "learning_rate": 8.655415346127582e-05, + "loss": 1.163, + "step": 3973 + }, + { + "epoch": 0.2414484476578164, + "grad_norm": 0.33693525195121765, + "learning_rate": 8.654762139877338e-05, + "loss": 1.0963, + "step": 3974 + }, + { + "epoch": 0.24150920469044293, + "grad_norm": 0.20950333774089813, + "learning_rate": 8.654108799661081e-05, + "loss": 1.0886, + "step": 3975 + }, + { + "epoch": 0.24156996172306944, + "grad_norm": 0.21750472486019135, + "learning_rate": 8.653455325502757e-05, + "loss": 1.0768, + "step": 3976 + }, + { + "epoch": 0.24163071875569597, + "grad_norm": 0.3137146234512329, + "learning_rate": 8.652801717426324e-05, + "loss": 1.1348, + "step": 3977 + }, + { + "epoch": 0.2416914757883225, + "grad_norm": 0.4214821457862854, + "learning_rate": 8.652147975455738e-05, + "loss": 1.2989, + "step": 3978 + }, + { + "epoch": 0.24175223282094901, + "grad_norm": 0.7343482375144958, + "learning_rate": 8.651494099614963e-05, + "loss": 1.1498, + "step": 3979 + }, + { + "epoch": 0.24181298985357555, + "grad_norm": 0.18380630016326904, + "learning_rate": 8.650840089927964e-05, + "loss": 1.0628, + "step": 3980 + }, + { + "epoch": 0.24187374688620208, + "grad_norm": 0.2583775222301483, + "learning_rate": 8.650185946418716e-05, + "loss": 1.1154, + "step": 3981 + }, + { + "epoch": 0.2419345039188286, + "grad_norm": 0.4193519949913025, + "learning_rate": 8.649531669111198e-05, + "loss": 1.0552, + "step": 3982 + }, + { + "epoch": 0.24199526095145513, + "grad_norm": 0.20230033993721008, + "learning_rate": 8.648877258029389e-05, + "loss": 1.0871, + "step": 3983 + }, + { + "epoch": 0.24205601798408166, + "grad_norm": 0.4743443727493286, + "learning_rate": 8.64822271319728e-05, + "loss": 1.1637, + "step": 3984 + }, + { + "epoch": 0.2421167750167082, + "grad_norm": 0.28776419162750244, + "learning_rate": 8.647568034638862e-05, + "loss": 1.1456, + "step": 3985 + }, + { + "epoch": 0.2421775320493347, + "grad_norm": 0.6662983894348145, + "learning_rate": 8.646913222378134e-05, + "loss": 1.0831, + "step": 3986 + }, + { + "epoch": 0.24223828908196124, + "grad_norm": 0.3407902121543884, + "learning_rate": 8.646258276439096e-05, + "loss": 1.0541, + "step": 3987 + }, + { + "epoch": 0.24229904611458777, + "grad_norm": 0.21117804944515228, + "learning_rate": 8.645603196845756e-05, + "loss": 1.0779, + "step": 3988 + }, + { + "epoch": 0.24235980314721428, + "grad_norm": 0.37179431319236755, + "learning_rate": 8.644947983622125e-05, + "loss": 1.0687, + "step": 3989 + }, + { + "epoch": 0.24242056017984082, + "grad_norm": 0.154281347990036, + "learning_rate": 8.644292636792221e-05, + "loss": 1.077, + "step": 3990 + }, + { + "epoch": 0.24248131721246735, + "grad_norm": 0.2919796407222748, + "learning_rate": 8.64363715638007e-05, + "loss": 1.0883, + "step": 3991 + }, + { + "epoch": 0.24254207424509386, + "grad_norm": 0.30052173137664795, + "learning_rate": 8.64298154240969e-05, + "loss": 1.154, + "step": 3992 + }, + { + "epoch": 0.2426028312777204, + "grad_norm": 0.22354765236377716, + "learning_rate": 8.64232579490512e-05, + "loss": 1.108, + "step": 3993 + }, + { + "epoch": 0.24266358831034693, + "grad_norm": 0.31550291180610657, + "learning_rate": 8.641669913890394e-05, + "loss": 1.1737, + "step": 3994 + }, + { + "epoch": 0.24272434534297344, + "grad_norm": 0.48148852586746216, + "learning_rate": 8.64101389938955e-05, + "loss": 1.0979, + "step": 3995 + }, + { + "epoch": 0.24278510237559997, + "grad_norm": 0.2621382772922516, + "learning_rate": 8.640357751426642e-05, + "loss": 1.1397, + "step": 3996 + }, + { + "epoch": 0.2428458594082265, + "grad_norm": 0.1898399442434311, + "learning_rate": 8.639701470025714e-05, + "loss": 1.1013, + "step": 3997 + }, + { + "epoch": 0.24290661644085304, + "grad_norm": 0.2752940356731415, + "learning_rate": 8.639045055210827e-05, + "loss": 1.1054, + "step": 3998 + }, + { + "epoch": 0.24296737347347955, + "grad_norm": 0.3360599875450134, + "learning_rate": 8.638388507006039e-05, + "loss": 1.1622, + "step": 3999 + }, + { + "epoch": 0.24302813050610608, + "grad_norm": 0.19555874168872833, + "learning_rate": 8.637731825435418e-05, + "loss": 1.1263, + "step": 4000 + }, + { + "epoch": 0.24308888753873262, + "grad_norm": 0.5630113482475281, + "learning_rate": 8.637075010523032e-05, + "loss": 1.114, + "step": 4001 + }, + { + "epoch": 0.24314964457135912, + "grad_norm": 2.356717348098755, + "learning_rate": 8.636418062292962e-05, + "loss": 1.1061, + "step": 4002 + }, + { + "epoch": 0.24321040160398566, + "grad_norm": 0.731472909450531, + "learning_rate": 8.635760980769281e-05, + "loss": 1.1573, + "step": 4003 + }, + { + "epoch": 0.2432711586366122, + "grad_norm": 0.3826003670692444, + "learning_rate": 8.635103765976081e-05, + "loss": 1.0831, + "step": 4004 + }, + { + "epoch": 0.2433319156692387, + "grad_norm": 0.2556750774383545, + "learning_rate": 8.634446417937449e-05, + "loss": 1.1658, + "step": 4005 + }, + { + "epoch": 0.24339267270186524, + "grad_norm": 0.2815265655517578, + "learning_rate": 8.63378893667748e-05, + "loss": 1.1412, + "step": 4006 + }, + { + "epoch": 0.24345342973449177, + "grad_norm": 0.4167288541793823, + "learning_rate": 8.633131322220276e-05, + "loss": 1.08, + "step": 4007 + }, + { + "epoch": 0.2435141867671183, + "grad_norm": 0.3338865637779236, + "learning_rate": 8.632473574589941e-05, + "loss": 1.0909, + "step": 4008 + }, + { + "epoch": 0.24357494379974481, + "grad_norm": 0.24205607175827026, + "learning_rate": 8.631815693810586e-05, + "loss": 1.1394, + "step": 4009 + }, + { + "epoch": 0.24363570083237135, + "grad_norm": 2.4270808696746826, + "learning_rate": 8.631157679906323e-05, + "loss": 1.1507, + "step": 4010 + }, + { + "epoch": 0.24369645786499788, + "grad_norm": 0.22711487114429474, + "learning_rate": 8.630499532901275e-05, + "loss": 1.1668, + "step": 4011 + }, + { + "epoch": 0.2437572148976244, + "grad_norm": 0.3731253743171692, + "learning_rate": 8.629841252819564e-05, + "loss": 1.1662, + "step": 4012 + }, + { + "epoch": 0.24381797193025093, + "grad_norm": 0.273828387260437, + "learning_rate": 8.629182839685321e-05, + "loss": 1.1178, + "step": 4013 + }, + { + "epoch": 0.24387872896287746, + "grad_norm": 0.7459601163864136, + "learning_rate": 8.628524293522679e-05, + "loss": 1.1681, + "step": 4014 + }, + { + "epoch": 0.24393948599550397, + "grad_norm": 0.20066064596176147, + "learning_rate": 8.627865614355777e-05, + "loss": 1.0237, + "step": 4015 + }, + { + "epoch": 0.2440002430281305, + "grad_norm": 0.2522144019603729, + "learning_rate": 8.62720680220876e-05, + "loss": 1.1377, + "step": 4016 + }, + { + "epoch": 0.24406100006075704, + "grad_norm": 0.2461540699005127, + "learning_rate": 8.626547857105777e-05, + "loss": 1.0516, + "step": 4017 + }, + { + "epoch": 0.24412175709338355, + "grad_norm": 0.22894492745399475, + "learning_rate": 8.62588877907098e-05, + "loss": 1.1479, + "step": 4018 + }, + { + "epoch": 0.24418251412601008, + "grad_norm": 0.43170854449272156, + "learning_rate": 8.625229568128531e-05, + "loss": 1.2412, + "step": 4019 + }, + { + "epoch": 0.24424327115863662, + "grad_norm": 0.1729089319705963, + "learning_rate": 8.624570224302591e-05, + "loss": 1.0635, + "step": 4020 + }, + { + "epoch": 0.24430402819126315, + "grad_norm": 0.2000235617160797, + "learning_rate": 8.623910747617328e-05, + "loss": 1.0674, + "step": 4021 + }, + { + "epoch": 0.24436478522388966, + "grad_norm": 0.3221941590309143, + "learning_rate": 8.623251138096916e-05, + "loss": 1.1291, + "step": 4022 + }, + { + "epoch": 0.2444255422565162, + "grad_norm": 2.275629758834839, + "learning_rate": 8.622591395765533e-05, + "loss": 1.0994, + "step": 4023 + }, + { + "epoch": 0.24448629928914273, + "grad_norm": 0.5258769392967224, + "learning_rate": 8.621931520647363e-05, + "loss": 1.13, + "step": 4024 + }, + { + "epoch": 0.24454705632176924, + "grad_norm": 0.2967522144317627, + "learning_rate": 8.621271512766593e-05, + "loss": 1.2532, + "step": 4025 + }, + { + "epoch": 0.24460781335439577, + "grad_norm": 0.38757386803627014, + "learning_rate": 8.620611372147414e-05, + "loss": 1.1047, + "step": 4026 + }, + { + "epoch": 0.2446685703870223, + "grad_norm": 4.310270309448242, + "learning_rate": 8.619951098814028e-05, + "loss": 1.0946, + "step": 4027 + }, + { + "epoch": 0.2447293274196488, + "grad_norm": 0.44539642333984375, + "learning_rate": 8.619290692790633e-05, + "loss": 1.1972, + "step": 4028 + }, + { + "epoch": 0.24479008445227535, + "grad_norm": 0.3264528810977936, + "learning_rate": 8.618630154101438e-05, + "loss": 1.1745, + "step": 4029 + }, + { + "epoch": 0.24485084148490188, + "grad_norm": 0.9619507193565369, + "learning_rate": 8.617969482770656e-05, + "loss": 1.0574, + "step": 4030 + }, + { + "epoch": 0.24491159851752842, + "grad_norm": 0.5961844325065613, + "learning_rate": 8.617308678822502e-05, + "loss": 1.1731, + "step": 4031 + }, + { + "epoch": 0.24497235555015492, + "grad_norm": 0.6532117128372192, + "learning_rate": 8.6166477422812e-05, + "loss": 1.1282, + "step": 4032 + }, + { + "epoch": 0.24503311258278146, + "grad_norm": 0.5110911726951599, + "learning_rate": 8.615986673170977e-05, + "loss": 1.0906, + "step": 4033 + }, + { + "epoch": 0.245093869615408, + "grad_norm": 0.4157450199127197, + "learning_rate": 8.615325471516063e-05, + "loss": 1.1374, + "step": 4034 + }, + { + "epoch": 0.2451546266480345, + "grad_norm": 0.6731694340705872, + "learning_rate": 8.614664137340694e-05, + "loss": 1.0809, + "step": 4035 + }, + { + "epoch": 0.24521538368066104, + "grad_norm": 0.26499301195144653, + "learning_rate": 8.614002670669114e-05, + "loss": 1.3586, + "step": 4036 + }, + { + "epoch": 0.24527614071328757, + "grad_norm": 0.6510512232780457, + "learning_rate": 8.613341071525565e-05, + "loss": 1.1771, + "step": 4037 + }, + { + "epoch": 0.24533689774591408, + "grad_norm": 0.289264053106308, + "learning_rate": 8.612679339934302e-05, + "loss": 1.0913, + "step": 4038 + }, + { + "epoch": 0.2453976547785406, + "grad_norm": 0.21807226538658142, + "learning_rate": 8.61201747591958e-05, + "loss": 1.2176, + "step": 4039 + }, + { + "epoch": 0.24545841181116715, + "grad_norm": 0.31609657406806946, + "learning_rate": 8.611355479505658e-05, + "loss": 1.0555, + "step": 4040 + }, + { + "epoch": 0.24551916884379366, + "grad_norm": 0.46881070733070374, + "learning_rate": 8.610693350716806e-05, + "loss": 1.1465, + "step": 4041 + }, + { + "epoch": 0.2455799258764202, + "grad_norm": 0.7690351605415344, + "learning_rate": 8.610031089577289e-05, + "loss": 1.0889, + "step": 4042 + }, + { + "epoch": 0.24564068290904673, + "grad_norm": 0.6717157959938049, + "learning_rate": 8.609368696111383e-05, + "loss": 1.009, + "step": 4043 + }, + { + "epoch": 0.24570143994167326, + "grad_norm": 0.21355541050434113, + "learning_rate": 8.608706170343371e-05, + "loss": 1.0713, + "step": 4044 + }, + { + "epoch": 0.24576219697429977, + "grad_norm": 0.3016953766345978, + "learning_rate": 8.608043512297537e-05, + "loss": 1.1373, + "step": 4045 + }, + { + "epoch": 0.2458229540069263, + "grad_norm": 0.3814987540245056, + "learning_rate": 8.607380721998171e-05, + "loss": 1.0562, + "step": 4046 + }, + { + "epoch": 0.24588371103955284, + "grad_norm": 0.7475008964538574, + "learning_rate": 8.606717799469568e-05, + "loss": 1.128, + "step": 4047 + }, + { + "epoch": 0.24594446807217935, + "grad_norm": 0.4798276722431183, + "learning_rate": 8.606054744736026e-05, + "loss": 1.1311, + "step": 4048 + }, + { + "epoch": 0.24600522510480588, + "grad_norm": 0.25417768955230713, + "learning_rate": 8.605391557821849e-05, + "loss": 1.1023, + "step": 4049 + }, + { + "epoch": 0.24606598213743242, + "grad_norm": 0.3219625651836395, + "learning_rate": 8.604728238751349e-05, + "loss": 1.2442, + "step": 4050 + }, + { + "epoch": 0.24612673917005892, + "grad_norm": 3.745697259902954, + "learning_rate": 8.604064787548839e-05, + "loss": 1.1162, + "step": 4051 + }, + { + "epoch": 0.24618749620268546, + "grad_norm": 0.47977200150489807, + "learning_rate": 8.603401204238637e-05, + "loss": 1.1589, + "step": 4052 + }, + { + "epoch": 0.246248253235312, + "grad_norm": 0.2328408807516098, + "learning_rate": 8.602737488845067e-05, + "loss": 1.1497, + "step": 4053 + }, + { + "epoch": 0.24630901026793853, + "grad_norm": 0.29289308190345764, + "learning_rate": 8.602073641392458e-05, + "loss": 1.033, + "step": 4054 + }, + { + "epoch": 0.24636976730056503, + "grad_norm": 0.35585033893585205, + "learning_rate": 8.601409661905144e-05, + "loss": 1.1675, + "step": 4055 + }, + { + "epoch": 0.24643052433319157, + "grad_norm": 0.20216651260852814, + "learning_rate": 8.600745550407464e-05, + "loss": 1.1245, + "step": 4056 + }, + { + "epoch": 0.2464912813658181, + "grad_norm": 0.3123777210712433, + "learning_rate": 8.600081306923758e-05, + "loss": 1.1455, + "step": 4057 + }, + { + "epoch": 0.2465520383984446, + "grad_norm": 0.40729108452796936, + "learning_rate": 8.599416931478374e-05, + "loss": 1.2636, + "step": 4058 + }, + { + "epoch": 0.24661279543107115, + "grad_norm": 0.2026059329509735, + "learning_rate": 8.598752424095669e-05, + "loss": 1.0687, + "step": 4059 + }, + { + "epoch": 0.24667355246369768, + "grad_norm": 1.872135043144226, + "learning_rate": 8.598087784799998e-05, + "loss": 1.2284, + "step": 4060 + }, + { + "epoch": 0.2467343094963242, + "grad_norm": 0.43025729060173035, + "learning_rate": 8.597423013615724e-05, + "loss": 1.0445, + "step": 4061 + }, + { + "epoch": 0.24679506652895072, + "grad_norm": 0.30069974064826965, + "learning_rate": 8.596758110567214e-05, + "loss": 1.1746, + "step": 4062 + }, + { + "epoch": 0.24685582356157726, + "grad_norm": 0.5396690368652344, + "learning_rate": 8.59609307567884e-05, + "loss": 1.0782, + "step": 4063 + }, + { + "epoch": 0.24691658059420377, + "grad_norm": 0.23962604999542236, + "learning_rate": 8.595427908974979e-05, + "loss": 1.1662, + "step": 4064 + }, + { + "epoch": 0.2469773376268303, + "grad_norm": 0.2537461817264557, + "learning_rate": 8.594762610480012e-05, + "loss": 1.212, + "step": 4065 + }, + { + "epoch": 0.24703809465945684, + "grad_norm": 0.2212119996547699, + "learning_rate": 8.594097180218329e-05, + "loss": 1.1254, + "step": 4066 + }, + { + "epoch": 0.24709885169208337, + "grad_norm": 0.34347203373908997, + "learning_rate": 8.593431618214318e-05, + "loss": 1.219, + "step": 4067 + }, + { + "epoch": 0.24715960872470988, + "grad_norm": 0.3240489363670349, + "learning_rate": 8.592765924492377e-05, + "loss": 1.1885, + "step": 4068 + }, + { + "epoch": 0.2472203657573364, + "grad_norm": 2.8406341075897217, + "learning_rate": 8.592100099076905e-05, + "loss": 1.0443, + "step": 4069 + }, + { + "epoch": 0.24728112278996295, + "grad_norm": 0.5181599855422974, + "learning_rate": 8.59143414199231e-05, + "loss": 1.2346, + "step": 4070 + }, + { + "epoch": 0.24734187982258946, + "grad_norm": 0.3141525387763977, + "learning_rate": 8.590768053263005e-05, + "loss": 1.1608, + "step": 4071 + }, + { + "epoch": 0.247402636855216, + "grad_norm": 0.5005430579185486, + "learning_rate": 8.5901018329134e-05, + "loss": 1.1885, + "step": 4072 + }, + { + "epoch": 0.24746339388784253, + "grad_norm": 0.2558469772338867, + "learning_rate": 8.589435480967918e-05, + "loss": 1.2118, + "step": 4073 + }, + { + "epoch": 0.24752415092046903, + "grad_norm": 0.2807365357875824, + "learning_rate": 8.588768997450987e-05, + "loss": 1.2097, + "step": 4074 + }, + { + "epoch": 0.24758490795309557, + "grad_norm": 0.7362846732139587, + "learning_rate": 8.588102382387033e-05, + "loss": 1.1115, + "step": 4075 + }, + { + "epoch": 0.2476456649857221, + "grad_norm": 0.26292684674263, + "learning_rate": 8.587435635800494e-05, + "loss": 1.1728, + "step": 4076 + }, + { + "epoch": 0.24770642201834864, + "grad_norm": 0.4258117079734802, + "learning_rate": 8.586768757715806e-05, + "loss": 1.045, + "step": 4077 + }, + { + "epoch": 0.24776717905097514, + "grad_norm": 0.2380322962999344, + "learning_rate": 8.586101748157418e-05, + "loss": 1.0641, + "step": 4078 + }, + { + "epoch": 0.24782793608360168, + "grad_norm": 0.6344050168991089, + "learning_rate": 8.585434607149776e-05, + "loss": 1.102, + "step": 4079 + }, + { + "epoch": 0.24788869311622821, + "grad_norm": 0.3072490394115448, + "learning_rate": 8.584767334717337e-05, + "loss": 1.0954, + "step": 4080 + }, + { + "epoch": 0.24794945014885472, + "grad_norm": 0.18298567831516266, + "learning_rate": 8.584099930884557e-05, + "loss": 1.1408, + "step": 4081 + }, + { + "epoch": 0.24801020718148126, + "grad_norm": 0.47453510761260986, + "learning_rate": 8.583432395675901e-05, + "loss": 1.0601, + "step": 4082 + }, + { + "epoch": 0.2480709642141078, + "grad_norm": 0.1774015575647354, + "learning_rate": 8.582764729115838e-05, + "loss": 1.0732, + "step": 4083 + }, + { + "epoch": 0.2481317212467343, + "grad_norm": 0.2853648066520691, + "learning_rate": 8.582096931228842e-05, + "loss": 1.2544, + "step": 4084 + }, + { + "epoch": 0.24819247827936083, + "grad_norm": 0.31408658623695374, + "learning_rate": 8.581429002039392e-05, + "loss": 1.227, + "step": 4085 + }, + { + "epoch": 0.24825323531198737, + "grad_norm": 0.3116123080253601, + "learning_rate": 8.580760941571967e-05, + "loss": 1.1375, + "step": 4086 + }, + { + "epoch": 0.24831399234461388, + "grad_norm": 0.24364641308784485, + "learning_rate": 8.580092749851058e-05, + "loss": 1.1155, + "step": 4087 + }, + { + "epoch": 0.2483747493772404, + "grad_norm": 0.20923365652561188, + "learning_rate": 8.579424426901159e-05, + "loss": 1.1095, + "step": 4088 + }, + { + "epoch": 0.24843550640986695, + "grad_norm": 1.4079455137252808, + "learning_rate": 8.578755972746763e-05, + "loss": 1.2461, + "step": 4089 + }, + { + "epoch": 0.24849626344249348, + "grad_norm": 0.17929084599018097, + "learning_rate": 8.578087387412377e-05, + "loss": 1.0362, + "step": 4090 + }, + { + "epoch": 0.24855702047512, + "grad_norm": 0.2408154159784317, + "learning_rate": 8.577418670922506e-05, + "loss": 1.1262, + "step": 4091 + }, + { + "epoch": 0.24861777750774652, + "grad_norm": 0.2008085697889328, + "learning_rate": 8.576749823301662e-05, + "loss": 1.0992, + "step": 4092 + }, + { + "epoch": 0.24867853454037306, + "grad_norm": 0.22592371702194214, + "learning_rate": 8.576080844574362e-05, + "loss": 1.1142, + "step": 4093 + }, + { + "epoch": 0.24873929157299957, + "grad_norm": 0.20659062266349792, + "learning_rate": 8.575411734765128e-05, + "loss": 1.0441, + "step": 4094 + }, + { + "epoch": 0.2488000486056261, + "grad_norm": 0.278988778591156, + "learning_rate": 8.574742493898484e-05, + "loss": 1.0995, + "step": 4095 + }, + { + "epoch": 0.24886080563825264, + "grad_norm": 0.38181138038635254, + "learning_rate": 8.574073121998964e-05, + "loss": 1.1241, + "step": 4096 + }, + { + "epoch": 0.24892156267087914, + "grad_norm": 0.1932530701160431, + "learning_rate": 8.573403619091103e-05, + "loss": 1.061, + "step": 4097 + }, + { + "epoch": 0.24898231970350568, + "grad_norm": 0.3438214659690857, + "learning_rate": 8.572733985199442e-05, + "loss": 1.1344, + "step": 4098 + }, + { + "epoch": 0.2490430767361322, + "grad_norm": 0.25462281703948975, + "learning_rate": 8.572064220348525e-05, + "loss": 1.2021, + "step": 4099 + }, + { + "epoch": 0.24910383376875872, + "grad_norm": 0.3071964979171753, + "learning_rate": 8.571394324562905e-05, + "loss": 1.062, + "step": 4100 + }, + { + "epoch": 0.24916459080138526, + "grad_norm": 0.7796054482460022, + "learning_rate": 8.570724297867135e-05, + "loss": 1.0987, + "step": 4101 + }, + { + "epoch": 0.2492253478340118, + "grad_norm": 0.25127848982810974, + "learning_rate": 8.570054140285775e-05, + "loss": 1.1635, + "step": 4102 + }, + { + "epoch": 0.24928610486663833, + "grad_norm": 0.285133421421051, + "learning_rate": 8.56938385184339e-05, + "loss": 1.2892, + "step": 4103 + }, + { + "epoch": 0.24934686189926483, + "grad_norm": 0.19972917437553406, + "learning_rate": 8.568713432564552e-05, + "loss": 1.1025, + "step": 4104 + }, + { + "epoch": 0.24940761893189137, + "grad_norm": 0.2590099275112152, + "learning_rate": 8.56804288247383e-05, + "loss": 1.0887, + "step": 4105 + }, + { + "epoch": 0.2494683759645179, + "grad_norm": 0.2210921198129654, + "learning_rate": 8.56737220159581e-05, + "loss": 1.152, + "step": 4106 + }, + { + "epoch": 0.2495291329971444, + "grad_norm": 0.2531004250049591, + "learning_rate": 8.56670138995507e-05, + "loss": 1.1356, + "step": 4107 + }, + { + "epoch": 0.24958989002977094, + "grad_norm": 0.3316219747066498, + "learning_rate": 8.566030447576203e-05, + "loss": 1.1589, + "step": 4108 + }, + { + "epoch": 0.24965064706239748, + "grad_norm": 0.341658353805542, + "learning_rate": 8.565359374483799e-05, + "loss": 1.2161, + "step": 4109 + }, + { + "epoch": 0.249711404095024, + "grad_norm": 1.6548233032226562, + "learning_rate": 8.56468817070246e-05, + "loss": 1.1412, + "step": 4110 + }, + { + "epoch": 0.24977216112765052, + "grad_norm": 0.34155818819999695, + "learning_rate": 8.564016836256785e-05, + "loss": 1.2819, + "step": 4111 + }, + { + "epoch": 0.24983291816027706, + "grad_norm": 0.2702336609363556, + "learning_rate": 8.563345371171385e-05, + "loss": 1.1029, + "step": 4112 + }, + { + "epoch": 0.2498936751929036, + "grad_norm": 0.3895370662212372, + "learning_rate": 8.56267377547087e-05, + "loss": 1.0643, + "step": 4113 + }, + { + "epoch": 0.2499544322255301, + "grad_norm": 0.3319454491138458, + "learning_rate": 8.562002049179862e-05, + "loss": 1.1695, + "step": 4114 + }, + { + "epoch": 0.25001518925815663, + "grad_norm": 0.2394411563873291, + "learning_rate": 8.561330192322978e-05, + "loss": 1.1334, + "step": 4115 + }, + { + "epoch": 0.25007594629078317, + "grad_norm": 0.4622597098350525, + "learning_rate": 8.560658204924849e-05, + "loss": 1.1024, + "step": 4116 + }, + { + "epoch": 0.2501367033234097, + "grad_norm": 0.25531700253486633, + "learning_rate": 8.559986087010104e-05, + "loss": 1.1451, + "step": 4117 + }, + { + "epoch": 0.25019746035603624, + "grad_norm": 0.529337465763092, + "learning_rate": 8.559313838603381e-05, + "loss": 1.207, + "step": 4118 + }, + { + "epoch": 0.2502582173886627, + "grad_norm": 0.24293553829193115, + "learning_rate": 8.558641459729321e-05, + "loss": 1.1317, + "step": 4119 + }, + { + "epoch": 0.25031897442128925, + "grad_norm": 0.3784746825695038, + "learning_rate": 8.557968950412571e-05, + "loss": 1.0825, + "step": 4120 + }, + { + "epoch": 0.2503797314539158, + "grad_norm": 0.26305362582206726, + "learning_rate": 8.557296310677783e-05, + "loss": 1.1362, + "step": 4121 + }, + { + "epoch": 0.2504404884865423, + "grad_norm": 0.1871000975370407, + "learning_rate": 8.556623540549608e-05, + "loss": 1.0405, + "step": 4122 + }, + { + "epoch": 0.25050124551916886, + "grad_norm": 0.24662111699581146, + "learning_rate": 8.555950640052713e-05, + "loss": 1.0892, + "step": 4123 + }, + { + "epoch": 0.2505620025517954, + "grad_norm": 0.29056432843208313, + "learning_rate": 8.555277609211757e-05, + "loss": 1.0678, + "step": 4124 + }, + { + "epoch": 0.2506227595844219, + "grad_norm": 0.23873057961463928, + "learning_rate": 8.554604448051413e-05, + "loss": 1.046, + "step": 4125 + }, + { + "epoch": 0.2506835166170484, + "grad_norm": 0.5368660688400269, + "learning_rate": 8.553931156596358e-05, + "loss": 1.2684, + "step": 4126 + }, + { + "epoch": 0.25074427364967494, + "grad_norm": 0.30210354924201965, + "learning_rate": 8.553257734871267e-05, + "loss": 1.0464, + "step": 4127 + }, + { + "epoch": 0.2508050306823015, + "grad_norm": 0.2462279200553894, + "learning_rate": 8.55258418290083e-05, + "loss": 1.0856, + "step": 4128 + }, + { + "epoch": 0.250865787714928, + "grad_norm": 0.2989644408226013, + "learning_rate": 8.551910500709732e-05, + "loss": 1.1535, + "step": 4129 + }, + { + "epoch": 0.25092654474755455, + "grad_norm": 0.4827086627483368, + "learning_rate": 8.551236688322666e-05, + "loss": 1.2858, + "step": 4130 + }, + { + "epoch": 0.2509873017801811, + "grad_norm": 0.18739399313926697, + "learning_rate": 8.550562745764334e-05, + "loss": 1.0957, + "step": 4131 + }, + { + "epoch": 0.25104805881280756, + "grad_norm": 0.20033054053783417, + "learning_rate": 8.549888673059439e-05, + "loss": 1.0922, + "step": 4132 + }, + { + "epoch": 0.2511088158454341, + "grad_norm": 0.2685336172580719, + "learning_rate": 8.549214470232688e-05, + "loss": 1.1787, + "step": 4133 + }, + { + "epoch": 0.25116957287806063, + "grad_norm": 0.30085259675979614, + "learning_rate": 8.548540137308795e-05, + "loss": 1.125, + "step": 4134 + }, + { + "epoch": 0.25123032991068717, + "grad_norm": 3.25395131111145, + "learning_rate": 8.547865674312476e-05, + "loss": 1.0955, + "step": 4135 + }, + { + "epoch": 0.2512910869433137, + "grad_norm": 0.2485281527042389, + "learning_rate": 8.547191081268456e-05, + "loss": 1.1069, + "step": 4136 + }, + { + "epoch": 0.25135184397594024, + "grad_norm": 0.24271564185619354, + "learning_rate": 8.546516358201462e-05, + "loss": 1.1121, + "step": 4137 + }, + { + "epoch": 0.2514126010085667, + "grad_norm": 0.6528868675231934, + "learning_rate": 8.545841505136224e-05, + "loss": 1.3524, + "step": 4138 + }, + { + "epoch": 0.25147335804119325, + "grad_norm": 0.20853766798973083, + "learning_rate": 8.545166522097481e-05, + "loss": 1.1267, + "step": 4139 + }, + { + "epoch": 0.2515341150738198, + "grad_norm": 3.1683387756347656, + "learning_rate": 8.544491409109973e-05, + "loss": 1.1459, + "step": 4140 + }, + { + "epoch": 0.2515948721064463, + "grad_norm": 0.5986893773078918, + "learning_rate": 8.543816166198449e-05, + "loss": 1.1257, + "step": 4141 + }, + { + "epoch": 0.25165562913907286, + "grad_norm": 0.6423225402832031, + "learning_rate": 8.543140793387657e-05, + "loss": 1.1638, + "step": 4142 + }, + { + "epoch": 0.2517163861716994, + "grad_norm": 0.4325331747531891, + "learning_rate": 8.542465290702353e-05, + "loss": 1.1281, + "step": 4143 + }, + { + "epoch": 0.2517771432043259, + "grad_norm": 0.3461361229419708, + "learning_rate": 8.541789658167301e-05, + "loss": 1.0152, + "step": 4144 + }, + { + "epoch": 0.2518379002369524, + "grad_norm": 0.3811987042427063, + "learning_rate": 8.541113895807264e-05, + "loss": 1.1852, + "step": 4145 + }, + { + "epoch": 0.25189865726957894, + "grad_norm": 0.6824424862861633, + "learning_rate": 8.540438003647013e-05, + "loss": 1.0496, + "step": 4146 + }, + { + "epoch": 0.2519594143022055, + "grad_norm": 0.29640993475914, + "learning_rate": 8.539761981711321e-05, + "loss": 1.0616, + "step": 4147 + }, + { + "epoch": 0.252020171334832, + "grad_norm": 0.5454568862915039, + "learning_rate": 8.539085830024971e-05, + "loss": 1.1728, + "step": 4148 + }, + { + "epoch": 0.25208092836745855, + "grad_norm": 0.3840097188949585, + "learning_rate": 8.538409548612746e-05, + "loss": 1.2395, + "step": 4149 + }, + { + "epoch": 0.2521416854000851, + "grad_norm": 0.7966206073760986, + "learning_rate": 8.537733137499435e-05, + "loss": 1.1789, + "step": 4150 + }, + { + "epoch": 0.25220244243271156, + "grad_norm": 0.8933805227279663, + "learning_rate": 8.537056596709832e-05, + "loss": 1.1289, + "step": 4151 + }, + { + "epoch": 0.2522631994653381, + "grad_norm": 0.17207057774066925, + "learning_rate": 8.536379926268734e-05, + "loss": 1.093, + "step": 4152 + }, + { + "epoch": 0.25232395649796463, + "grad_norm": 0.6583653092384338, + "learning_rate": 8.535703126200948e-05, + "loss": 1.1056, + "step": 4153 + }, + { + "epoch": 0.25238471353059116, + "grad_norm": 0.3321758508682251, + "learning_rate": 8.535026196531279e-05, + "loss": 1.1331, + "step": 4154 + }, + { + "epoch": 0.2524454705632177, + "grad_norm": 0.3148443102836609, + "learning_rate": 8.534349137284543e-05, + "loss": 1.2253, + "step": 4155 + }, + { + "epoch": 0.25250622759584423, + "grad_norm": 0.20556001365184784, + "learning_rate": 8.533671948485555e-05, + "loss": 1.0951, + "step": 4156 + }, + { + "epoch": 0.25256698462847077, + "grad_norm": 0.2917083501815796, + "learning_rate": 8.53299463015914e-05, + "loss": 1.0695, + "step": 4157 + }, + { + "epoch": 0.25262774166109725, + "grad_norm": 0.3155859708786011, + "learning_rate": 8.532317182330123e-05, + "loss": 1.2938, + "step": 4158 + }, + { + "epoch": 0.2526884986937238, + "grad_norm": 1.084568738937378, + "learning_rate": 8.531639605023339e-05, + "loss": 1.175, + "step": 4159 + }, + { + "epoch": 0.2527492557263503, + "grad_norm": 0.30736738443374634, + "learning_rate": 8.53096189826362e-05, + "loss": 1.0902, + "step": 4160 + }, + { + "epoch": 0.25281001275897685, + "grad_norm": 0.29184088110923767, + "learning_rate": 8.53028406207581e-05, + "loss": 1.1369, + "step": 4161 + }, + { + "epoch": 0.2528707697916034, + "grad_norm": 0.25988054275512695, + "learning_rate": 8.529606096484757e-05, + "loss": 1.071, + "step": 4162 + }, + { + "epoch": 0.2529315268242299, + "grad_norm": 0.22810271382331848, + "learning_rate": 8.528928001515309e-05, + "loss": 1.1943, + "step": 4163 + }, + { + "epoch": 0.2529922838568564, + "grad_norm": 0.24224255979061127, + "learning_rate": 8.528249777192324e-05, + "loss": 1.1248, + "step": 4164 + }, + { + "epoch": 0.25305304088948294, + "grad_norm": 0.22306948900222778, + "learning_rate": 8.527571423540662e-05, + "loss": 1.0594, + "step": 4165 + }, + { + "epoch": 0.2531137979221095, + "grad_norm": 0.34177395701408386, + "learning_rate": 8.526892940585185e-05, + "loss": 1.2169, + "step": 4166 + }, + { + "epoch": 0.253174554954736, + "grad_norm": 0.27090179920196533, + "learning_rate": 8.526214328350767e-05, + "loss": 1.0673, + "step": 4167 + }, + { + "epoch": 0.25323531198736254, + "grad_norm": 0.7239984273910522, + "learning_rate": 8.525535586862281e-05, + "loss": 1.1709, + "step": 4168 + }, + { + "epoch": 0.2532960690199891, + "grad_norm": 0.29093125462532043, + "learning_rate": 8.524856716144607e-05, + "loss": 1.0851, + "step": 4169 + }, + { + "epoch": 0.2533568260526156, + "grad_norm": 0.24587218463420868, + "learning_rate": 8.524177716222628e-05, + "loss": 1.102, + "step": 4170 + }, + { + "epoch": 0.2534175830852421, + "grad_norm": 0.2142520248889923, + "learning_rate": 8.523498587121235e-05, + "loss": 1.0898, + "step": 4171 + }, + { + "epoch": 0.25347834011786863, + "grad_norm": 0.374002069234848, + "learning_rate": 8.52281932886532e-05, + "loss": 1.2348, + "step": 4172 + }, + { + "epoch": 0.25353909715049516, + "grad_norm": 0.23794032633304596, + "learning_rate": 8.52213994147978e-05, + "loss": 1.1646, + "step": 4173 + }, + { + "epoch": 0.2535998541831217, + "grad_norm": 2.918766498565674, + "learning_rate": 8.52146042498952e-05, + "loss": 1.1013, + "step": 4174 + }, + { + "epoch": 0.25366061121574823, + "grad_norm": 0.4217281639575958, + "learning_rate": 8.520780779419448e-05, + "loss": 1.0747, + "step": 4175 + }, + { + "epoch": 0.25372136824837477, + "grad_norm": 0.35247811675071716, + "learning_rate": 8.520101004794476e-05, + "loss": 1.2429, + "step": 4176 + }, + { + "epoch": 0.2537821252810013, + "grad_norm": 0.49471721053123474, + "learning_rate": 8.519421101139522e-05, + "loss": 1.1295, + "step": 4177 + }, + { + "epoch": 0.2538428823136278, + "grad_norm": 0.21667112410068512, + "learning_rate": 8.518741068479506e-05, + "loss": 1.1549, + "step": 4178 + }, + { + "epoch": 0.2539036393462543, + "grad_norm": 0.23236200213432312, + "learning_rate": 8.518060906839356e-05, + "loss": 1.1906, + "step": 4179 + }, + { + "epoch": 0.25396439637888085, + "grad_norm": 0.2592597007751465, + "learning_rate": 8.517380616244005e-05, + "loss": 1.1564, + "step": 4180 + }, + { + "epoch": 0.2540251534115074, + "grad_norm": 0.24538828432559967, + "learning_rate": 8.516700196718385e-05, + "loss": 1.1499, + "step": 4181 + }, + { + "epoch": 0.2540859104441339, + "grad_norm": 0.37052246928215027, + "learning_rate": 8.516019648287441e-05, + "loss": 1.1159, + "step": 4182 + }, + { + "epoch": 0.25414666747676046, + "grad_norm": 0.38620540499687195, + "learning_rate": 8.515338970976117e-05, + "loss": 1.0983, + "step": 4183 + }, + { + "epoch": 0.25420742450938694, + "grad_norm": 0.2608968913555145, + "learning_rate": 8.514658164809364e-05, + "loss": 1.0782, + "step": 4184 + }, + { + "epoch": 0.25426818154201347, + "grad_norm": 0.3777622580528259, + "learning_rate": 8.513977229812135e-05, + "loss": 1.0782, + "step": 4185 + }, + { + "epoch": 0.25432893857464, + "grad_norm": 0.26426827907562256, + "learning_rate": 8.513296166009395e-05, + "loss": 1.0881, + "step": 4186 + }, + { + "epoch": 0.25438969560726654, + "grad_norm": 0.3776608407497406, + "learning_rate": 8.512614973426101e-05, + "loss": 1.1066, + "step": 4187 + }, + { + "epoch": 0.2544504526398931, + "grad_norm": 0.2883710563182831, + "learning_rate": 8.511933652087227e-05, + "loss": 1.1016, + "step": 4188 + }, + { + "epoch": 0.2545112096725196, + "grad_norm": 0.6933721899986267, + "learning_rate": 8.511252202017747e-05, + "loss": 1.099, + "step": 4189 + }, + { + "epoch": 0.25457196670514615, + "grad_norm": 0.2624293565750122, + "learning_rate": 8.51057062324264e-05, + "loss": 1.2166, + "step": 4190 + }, + { + "epoch": 0.2546327237377726, + "grad_norm": 0.31153953075408936, + "learning_rate": 8.509888915786886e-05, + "loss": 1.1644, + "step": 4191 + }, + { + "epoch": 0.25469348077039916, + "grad_norm": 0.22108764946460724, + "learning_rate": 8.509207079675478e-05, + "loss": 1.2681, + "step": 4192 + }, + { + "epoch": 0.2547542378030257, + "grad_norm": 0.5502873659133911, + "learning_rate": 8.508525114933404e-05, + "loss": 1.0848, + "step": 4193 + }, + { + "epoch": 0.25481499483565223, + "grad_norm": 0.2335219532251358, + "learning_rate": 8.507843021585665e-05, + "loss": 1.0856, + "step": 4194 + }, + { + "epoch": 0.25487575186827877, + "grad_norm": 0.2066269963979721, + "learning_rate": 8.50716079965726e-05, + "loss": 1.0883, + "step": 4195 + }, + { + "epoch": 0.2549365089009053, + "grad_norm": 0.1894065886735916, + "learning_rate": 8.5064784491732e-05, + "loss": 1.0686, + "step": 4196 + }, + { + "epoch": 0.2549972659335318, + "grad_norm": 6.0065813064575195, + "learning_rate": 8.505795970158493e-05, + "loss": 1.0914, + "step": 4197 + }, + { + "epoch": 0.2550580229661583, + "grad_norm": 0.4737032949924469, + "learning_rate": 8.505113362638158e-05, + "loss": 1.1723, + "step": 4198 + }, + { + "epoch": 0.25511877999878485, + "grad_norm": 0.32969895005226135, + "learning_rate": 8.504430626637215e-05, + "loss": 1.1799, + "step": 4199 + }, + { + "epoch": 0.2551795370314114, + "grad_norm": 0.40604618191719055, + "learning_rate": 8.503747762180691e-05, + "loss": 1.1158, + "step": 4200 + }, + { + "epoch": 0.2552402940640379, + "grad_norm": 0.1970643699169159, + "learning_rate": 8.503064769293614e-05, + "loss": 1.1305, + "step": 4201 + }, + { + "epoch": 0.25530105109666446, + "grad_norm": 0.24687586724758148, + "learning_rate": 8.50238164800102e-05, + "loss": 1.1162, + "step": 4202 + }, + { + "epoch": 0.255361808129291, + "grad_norm": 0.20285314321517944, + "learning_rate": 8.501698398327951e-05, + "loss": 1.0883, + "step": 4203 + }, + { + "epoch": 0.25542256516191747, + "grad_norm": 0.21740952134132385, + "learning_rate": 8.501015020299452e-05, + "loss": 1.3822, + "step": 4204 + }, + { + "epoch": 0.255483322194544, + "grad_norm": 0.2961682677268982, + "learning_rate": 8.500331513940569e-05, + "loss": 1.163, + "step": 4205 + }, + { + "epoch": 0.25554407922717054, + "grad_norm": 0.16597232222557068, + "learning_rate": 8.499647879276359e-05, + "loss": 1.078, + "step": 4206 + }, + { + "epoch": 0.2556048362597971, + "grad_norm": 0.24658124148845673, + "learning_rate": 8.498964116331876e-05, + "loss": 1.1343, + "step": 4207 + }, + { + "epoch": 0.2556655932924236, + "grad_norm": 0.3427563011646271, + "learning_rate": 8.49828022513219e-05, + "loss": 1.2419, + "step": 4208 + }, + { + "epoch": 0.25572635032505014, + "grad_norm": 0.18742281198501587, + "learning_rate": 8.497596205702366e-05, + "loss": 1.1225, + "step": 4209 + }, + { + "epoch": 0.2557871073576766, + "grad_norm": 0.4612957239151001, + "learning_rate": 8.496912058067476e-05, + "loss": 1.2071, + "step": 4210 + }, + { + "epoch": 0.25584786439030316, + "grad_norm": 0.25626954436302185, + "learning_rate": 8.496227782252599e-05, + "loss": 1.0771, + "step": 4211 + }, + { + "epoch": 0.2559086214229297, + "grad_norm": 0.28671959042549133, + "learning_rate": 8.495543378282818e-05, + "loss": 1.0371, + "step": 4212 + }, + { + "epoch": 0.25596937845555623, + "grad_norm": 0.19839106500148773, + "learning_rate": 8.494858846183217e-05, + "loss": 1.0516, + "step": 4213 + }, + { + "epoch": 0.25603013548818276, + "grad_norm": 0.1880330741405487, + "learning_rate": 8.494174185978889e-05, + "loss": 1.1261, + "step": 4214 + }, + { + "epoch": 0.2560908925208093, + "grad_norm": 1.288488507270813, + "learning_rate": 8.493489397694932e-05, + "loss": 1.1841, + "step": 4215 + }, + { + "epoch": 0.25615164955343583, + "grad_norm": 0.23587387800216675, + "learning_rate": 8.492804481356446e-05, + "loss": 1.2136, + "step": 4216 + }, + { + "epoch": 0.2562124065860623, + "grad_norm": 0.22789743542671204, + "learning_rate": 8.492119436988538e-05, + "loss": 1.0554, + "step": 4217 + }, + { + "epoch": 0.25627316361868885, + "grad_norm": 0.24875977635383606, + "learning_rate": 8.491434264616313e-05, + "loss": 1.1454, + "step": 4218 + }, + { + "epoch": 0.2563339206513154, + "grad_norm": 0.5325719118118286, + "learning_rate": 8.490748964264894e-05, + "loss": 1.1786, + "step": 4219 + }, + { + "epoch": 0.2563946776839419, + "grad_norm": 0.2714542746543884, + "learning_rate": 8.490063535959396e-05, + "loss": 1.1782, + "step": 4220 + }, + { + "epoch": 0.25645543471656845, + "grad_norm": 0.23506374657154083, + "learning_rate": 8.489377979724944e-05, + "loss": 1.1001, + "step": 4221 + }, + { + "epoch": 0.256516191749195, + "grad_norm": 0.32083916664123535, + "learning_rate": 8.488692295586668e-05, + "loss": 1.0533, + "step": 4222 + }, + { + "epoch": 0.2565769487818215, + "grad_norm": 0.18636618554592133, + "learning_rate": 8.4880064835697e-05, + "loss": 1.0879, + "step": 4223 + }, + { + "epoch": 0.256637705814448, + "grad_norm": 0.23444344103336334, + "learning_rate": 8.487320543699183e-05, + "loss": 1.1243, + "step": 4224 + }, + { + "epoch": 0.25669846284707454, + "grad_norm": 0.3113109767436981, + "learning_rate": 8.486634476000255e-05, + "loss": 1.2599, + "step": 4225 + }, + { + "epoch": 0.2567592198797011, + "grad_norm": 0.24309055507183075, + "learning_rate": 8.485948280498067e-05, + "loss": 1.0308, + "step": 4226 + }, + { + "epoch": 0.2568199769123276, + "grad_norm": 0.25935932993888855, + "learning_rate": 8.485261957217771e-05, + "loss": 1.1473, + "step": 4227 + }, + { + "epoch": 0.25688073394495414, + "grad_norm": 0.7025704383850098, + "learning_rate": 8.484575506184525e-05, + "loss": 1.1136, + "step": 4228 + }, + { + "epoch": 0.2569414909775807, + "grad_norm": 0.21199765801429749, + "learning_rate": 8.483888927423489e-05, + "loss": 1.0928, + "step": 4229 + }, + { + "epoch": 0.25700224801020716, + "grad_norm": 6.800530433654785, + "learning_rate": 8.483202220959831e-05, + "loss": 1.049, + "step": 4230 + }, + { + "epoch": 0.2570630050428337, + "grad_norm": 0.30008864402770996, + "learning_rate": 8.482515386818724e-05, + "loss": 1.0782, + "step": 4231 + }, + { + "epoch": 0.2571237620754602, + "grad_norm": 0.4060353934764862, + "learning_rate": 8.48182842502534e-05, + "loss": 1.1326, + "step": 4232 + }, + { + "epoch": 0.25718451910808676, + "grad_norm": 0.2504235506057739, + "learning_rate": 8.481141335604863e-05, + "loss": 1.0857, + "step": 4233 + }, + { + "epoch": 0.2572452761407133, + "grad_norm": 0.39608508348464966, + "learning_rate": 8.480454118582479e-05, + "loss": 1.0581, + "step": 4234 + }, + { + "epoch": 0.25730603317333983, + "grad_norm": 0.2246747612953186, + "learning_rate": 8.479766773983374e-05, + "loss": 1.0543, + "step": 4235 + }, + { + "epoch": 0.25736679020596637, + "grad_norm": 0.573861837387085, + "learning_rate": 8.479079301832746e-05, + "loss": 1.1842, + "step": 4236 + }, + { + "epoch": 0.25742754723859285, + "grad_norm": 0.2523806691169739, + "learning_rate": 8.478391702155795e-05, + "loss": 1.1018, + "step": 4237 + }, + { + "epoch": 0.2574883042712194, + "grad_norm": 0.22414204478263855, + "learning_rate": 8.477703974977722e-05, + "loss": 1.1863, + "step": 4238 + }, + { + "epoch": 0.2575490613038459, + "grad_norm": 0.3768017590045929, + "learning_rate": 8.47701612032374e-05, + "loss": 1.0187, + "step": 4239 + }, + { + "epoch": 0.25760981833647245, + "grad_norm": 0.2964487671852112, + "learning_rate": 8.476328138219058e-05, + "loss": 1.155, + "step": 4240 + }, + { + "epoch": 0.257670575369099, + "grad_norm": 0.2785825729370117, + "learning_rate": 8.475640028688896e-05, + "loss": 1.2409, + "step": 4241 + }, + { + "epoch": 0.2577313324017255, + "grad_norm": 0.2292880415916443, + "learning_rate": 8.474951791758478e-05, + "loss": 1.0631, + "step": 4242 + }, + { + "epoch": 0.257792089434352, + "grad_norm": 0.21183456480503082, + "learning_rate": 8.474263427453031e-05, + "loss": 1.1244, + "step": 4243 + }, + { + "epoch": 0.25785284646697854, + "grad_norm": 0.25534364581108093, + "learning_rate": 8.473574935797786e-05, + "loss": 1.2437, + "step": 4244 + }, + { + "epoch": 0.25791360349960507, + "grad_norm": 0.27569276094436646, + "learning_rate": 8.472886316817979e-05, + "loss": 1.148, + "step": 4245 + }, + { + "epoch": 0.2579743605322316, + "grad_norm": 0.374663382768631, + "learning_rate": 8.472197570538853e-05, + "loss": 1.061, + "step": 4246 + }, + { + "epoch": 0.25803511756485814, + "grad_norm": 0.18445877730846405, + "learning_rate": 8.471508696985655e-05, + "loss": 1.0931, + "step": 4247 + }, + { + "epoch": 0.2580958745974847, + "grad_norm": 0.31282830238342285, + "learning_rate": 8.470819696183634e-05, + "loss": 1.2476, + "step": 4248 + }, + { + "epoch": 0.2581566316301112, + "grad_norm": 0.29129475355148315, + "learning_rate": 8.470130568158047e-05, + "loss": 1.1607, + "step": 4249 + }, + { + "epoch": 0.2582173886627377, + "grad_norm": 0.19582155346870422, + "learning_rate": 8.469441312934153e-05, + "loss": 1.102, + "step": 4250 + }, + { + "epoch": 0.2582781456953642, + "grad_norm": 1.0393133163452148, + "learning_rate": 8.468751930537215e-05, + "loss": 1.0536, + "step": 4251 + }, + { + "epoch": 0.25833890272799076, + "grad_norm": 0.2689116299152374, + "learning_rate": 8.468062420992506e-05, + "loss": 1.121, + "step": 4252 + }, + { + "epoch": 0.2583996597606173, + "grad_norm": 0.29417991638183594, + "learning_rate": 8.467372784325299e-05, + "loss": 1.1714, + "step": 4253 + }, + { + "epoch": 0.25846041679324383, + "grad_norm": 0.8359894156455994, + "learning_rate": 8.466683020560873e-05, + "loss": 1.192, + "step": 4254 + }, + { + "epoch": 0.25852117382587037, + "grad_norm": 0.32633134722709656, + "learning_rate": 8.465993129724508e-05, + "loss": 1.1639, + "step": 4255 + }, + { + "epoch": 0.25858193085849684, + "grad_norm": 0.261666476726532, + "learning_rate": 8.465303111841498e-05, + "loss": 1.2164, + "step": 4256 + }, + { + "epoch": 0.2586426878911234, + "grad_norm": 0.2949962615966797, + "learning_rate": 8.464612966937131e-05, + "loss": 1.2298, + "step": 4257 + }, + { + "epoch": 0.2587034449237499, + "grad_norm": 0.4787708818912506, + "learning_rate": 8.463922695036706e-05, + "loss": 1.0874, + "step": 4258 + }, + { + "epoch": 0.25876420195637645, + "grad_norm": 0.19603420794010162, + "learning_rate": 8.463232296165526e-05, + "loss": 1.0467, + "step": 4259 + }, + { + "epoch": 0.258824958989003, + "grad_norm": 0.26096540689468384, + "learning_rate": 8.462541770348895e-05, + "loss": 1.205, + "step": 4260 + }, + { + "epoch": 0.2588857160216295, + "grad_norm": 0.44197505712509155, + "learning_rate": 8.461851117612127e-05, + "loss": 1.0886, + "step": 4261 + }, + { + "epoch": 0.25894647305425605, + "grad_norm": 0.18796958029270172, + "learning_rate": 8.461160337980538e-05, + "loss": 1.1281, + "step": 4262 + }, + { + "epoch": 0.25900723008688253, + "grad_norm": 0.4147983193397522, + "learning_rate": 8.460469431479448e-05, + "loss": 1.1303, + "step": 4263 + }, + { + "epoch": 0.25906798711950907, + "grad_norm": 0.3305838108062744, + "learning_rate": 8.45977839813418e-05, + "loss": 1.0795, + "step": 4264 + }, + { + "epoch": 0.2591287441521356, + "grad_norm": 0.25396838784217834, + "learning_rate": 8.459087237970068e-05, + "loss": 1.1274, + "step": 4265 + }, + { + "epoch": 0.25918950118476214, + "grad_norm": 0.24851106107234955, + "learning_rate": 8.458395951012444e-05, + "loss": 1.1232, + "step": 4266 + }, + { + "epoch": 0.2592502582173887, + "grad_norm": 0.26165053248405457, + "learning_rate": 8.457704537286647e-05, + "loss": 1.1185, + "step": 4267 + }, + { + "epoch": 0.2593110152500152, + "grad_norm": 1.5206787586212158, + "learning_rate": 8.457012996818023e-05, + "loss": 1.1821, + "step": 4268 + }, + { + "epoch": 0.2593717722826417, + "grad_norm": 0.29504266381263733, + "learning_rate": 8.45632132963192e-05, + "loss": 1.2088, + "step": 4269 + }, + { + "epoch": 0.2594325293152682, + "grad_norm": 0.22433483600616455, + "learning_rate": 8.45562953575369e-05, + "loss": 1.0334, + "step": 4270 + }, + { + "epoch": 0.25949328634789476, + "grad_norm": 0.414593368768692, + "learning_rate": 8.454937615208692e-05, + "loss": 1.1436, + "step": 4271 + }, + { + "epoch": 0.2595540433805213, + "grad_norm": 0.17690570652484894, + "learning_rate": 8.454245568022285e-05, + "loss": 1.0647, + "step": 4272 + }, + { + "epoch": 0.25961480041314783, + "grad_norm": 1.0235302448272705, + "learning_rate": 8.453553394219841e-05, + "loss": 1.1133, + "step": 4273 + }, + { + "epoch": 0.25967555744577436, + "grad_norm": 0.28203049302101135, + "learning_rate": 8.45286109382673e-05, + "loss": 1.1264, + "step": 4274 + }, + { + "epoch": 0.2597363144784009, + "grad_norm": 0.1850062608718872, + "learning_rate": 8.452168666868327e-05, + "loss": 1.0932, + "step": 4275 + }, + { + "epoch": 0.2597970715110274, + "grad_norm": 0.18996860086917877, + "learning_rate": 8.451476113370016e-05, + "loss": 1.0751, + "step": 4276 + }, + { + "epoch": 0.2598578285436539, + "grad_norm": 0.27383437752723694, + "learning_rate": 8.450783433357182e-05, + "loss": 1.1344, + "step": 4277 + }, + { + "epoch": 0.25991858557628045, + "grad_norm": 0.32678601145744324, + "learning_rate": 8.45009062685521e-05, + "loss": 1.1909, + "step": 4278 + }, + { + "epoch": 0.259979342608907, + "grad_norm": 0.35783901810646057, + "learning_rate": 8.449397693889502e-05, + "loss": 1.1084, + "step": 4279 + }, + { + "epoch": 0.2600400996415335, + "grad_norm": 0.22366228699684143, + "learning_rate": 8.448704634485454e-05, + "loss": 1.1131, + "step": 4280 + }, + { + "epoch": 0.26010085667416005, + "grad_norm": 0.638050377368927, + "learning_rate": 8.448011448668472e-05, + "loss": 1.0933, + "step": 4281 + }, + { + "epoch": 0.2601616137067866, + "grad_norm": 0.3927302658557892, + "learning_rate": 8.447318136463962e-05, + "loss": 1.1327, + "step": 4282 + }, + { + "epoch": 0.26022237073941307, + "grad_norm": 0.46872028708457947, + "learning_rate": 8.446624697897342e-05, + "loss": 1.0748, + "step": 4283 + }, + { + "epoch": 0.2602831277720396, + "grad_norm": 0.15257960557937622, + "learning_rate": 8.445931132994025e-05, + "loss": 1.0805, + "step": 4284 + }, + { + "epoch": 0.26034388480466614, + "grad_norm": 0.3433448076248169, + "learning_rate": 8.445237441779437e-05, + "loss": 1.1135, + "step": 4285 + }, + { + "epoch": 0.26040464183729267, + "grad_norm": 0.17797091603279114, + "learning_rate": 8.444543624279006e-05, + "loss": 1.0633, + "step": 4286 + }, + { + "epoch": 0.2604653988699192, + "grad_norm": 0.27488309144973755, + "learning_rate": 8.443849680518161e-05, + "loss": 1.0079, + "step": 4287 + }, + { + "epoch": 0.26052615590254574, + "grad_norm": 0.28464433550834656, + "learning_rate": 8.44315561052234e-05, + "loss": 1.1039, + "step": 4288 + }, + { + "epoch": 0.2605869129351722, + "grad_norm": 0.21960724890232086, + "learning_rate": 8.442461414316986e-05, + "loss": 1.0924, + "step": 4289 + }, + { + "epoch": 0.26064766996779876, + "grad_norm": 0.21383512020111084, + "learning_rate": 8.441767091927543e-05, + "loss": 1.0924, + "step": 4290 + }, + { + "epoch": 0.2607084270004253, + "grad_norm": 0.3452387750148773, + "learning_rate": 8.441072643379461e-05, + "loss": 1.1751, + "step": 4291 + }, + { + "epoch": 0.2607691840330518, + "grad_norm": 0.5509575605392456, + "learning_rate": 8.440378068698196e-05, + "loss": 1.1602, + "step": 4292 + }, + { + "epoch": 0.26082994106567836, + "grad_norm": 0.2336532324552536, + "learning_rate": 8.43968336790921e-05, + "loss": 1.1943, + "step": 4293 + }, + { + "epoch": 0.2608906980983049, + "grad_norm": 0.3090567886829376, + "learning_rate": 8.438988541037964e-05, + "loss": 1.2657, + "step": 4294 + }, + { + "epoch": 0.26095145513093143, + "grad_norm": 0.3274080753326416, + "learning_rate": 8.438293588109928e-05, + "loss": 1.378, + "step": 4295 + }, + { + "epoch": 0.2610122121635579, + "grad_norm": 0.18711645901203156, + "learning_rate": 8.437598509150575e-05, + "loss": 1.0749, + "step": 4296 + }, + { + "epoch": 0.26107296919618445, + "grad_norm": 0.20669077336788177, + "learning_rate": 8.436903304185386e-05, + "loss": 1.0294, + "step": 4297 + }, + { + "epoch": 0.261133726228811, + "grad_norm": 8.442023277282715, + "learning_rate": 8.436207973239842e-05, + "loss": 1.0866, + "step": 4298 + }, + { + "epoch": 0.2611944832614375, + "grad_norm": 0.21571186184883118, + "learning_rate": 8.43551251633943e-05, + "loss": 1.0882, + "step": 4299 + }, + { + "epoch": 0.26125524029406405, + "grad_norm": 0.2751650810241699, + "learning_rate": 8.434816933509641e-05, + "loss": 1.1717, + "step": 4300 + }, + { + "epoch": 0.2613159973266906, + "grad_norm": 0.17793066799640656, + "learning_rate": 8.434121224775977e-05, + "loss": 1.0849, + "step": 4301 + }, + { + "epoch": 0.26137675435931707, + "grad_norm": 0.2410004585981369, + "learning_rate": 8.433425390163932e-05, + "loss": 1.1462, + "step": 4302 + }, + { + "epoch": 0.2614375113919436, + "grad_norm": 0.332926869392395, + "learning_rate": 8.432729429699017e-05, + "loss": 1.2889, + "step": 4303 + }, + { + "epoch": 0.26149826842457013, + "grad_norm": 0.20868287980556488, + "learning_rate": 8.432033343406742e-05, + "loss": 1.1218, + "step": 4304 + }, + { + "epoch": 0.26155902545719667, + "grad_norm": 0.17644573748111725, + "learning_rate": 8.431337131312622e-05, + "loss": 1.0741, + "step": 4305 + }, + { + "epoch": 0.2616197824898232, + "grad_norm": 0.2919235825538635, + "learning_rate": 8.430640793442175e-05, + "loss": 1.0339, + "step": 4306 + }, + { + "epoch": 0.26168053952244974, + "grad_norm": 0.22068335115909576, + "learning_rate": 8.429944329820927e-05, + "loss": 1.0875, + "step": 4307 + }, + { + "epoch": 0.2617412965550763, + "grad_norm": 0.1841055005788803, + "learning_rate": 8.429247740474406e-05, + "loss": 1.1186, + "step": 4308 + }, + { + "epoch": 0.26180205358770275, + "grad_norm": 0.3049737215042114, + "learning_rate": 8.428551025428146e-05, + "loss": 1.1279, + "step": 4309 + }, + { + "epoch": 0.2618628106203293, + "grad_norm": 0.19936512410640717, + "learning_rate": 8.427854184707686e-05, + "loss": 1.0757, + "step": 4310 + }, + { + "epoch": 0.2619235676529558, + "grad_norm": 0.2518964111804962, + "learning_rate": 8.427157218338568e-05, + "loss": 1.2067, + "step": 4311 + }, + { + "epoch": 0.26198432468558236, + "grad_norm": 0.37787115573883057, + "learning_rate": 8.426460126346341e-05, + "loss": 1.2324, + "step": 4312 + }, + { + "epoch": 0.2620450817182089, + "grad_norm": 0.2554229497909546, + "learning_rate": 8.425762908756553e-05, + "loss": 1.0798, + "step": 4313 + }, + { + "epoch": 0.26210583875083543, + "grad_norm": 0.5193288326263428, + "learning_rate": 8.425065565594767e-05, + "loss": 1.0656, + "step": 4314 + }, + { + "epoch": 0.2621665957834619, + "grad_norm": 1.606097936630249, + "learning_rate": 8.424368096886538e-05, + "loss": 1.1992, + "step": 4315 + }, + { + "epoch": 0.26222735281608844, + "grad_norm": 0.35156965255737305, + "learning_rate": 8.423670502657435e-05, + "loss": 1.0466, + "step": 4316 + }, + { + "epoch": 0.262288109848715, + "grad_norm": 0.24291445314884186, + "learning_rate": 8.422972782933029e-05, + "loss": 1.2506, + "step": 4317 + }, + { + "epoch": 0.2623488668813415, + "grad_norm": 0.23549553751945496, + "learning_rate": 8.422274937738893e-05, + "loss": 1.0727, + "step": 4318 + }, + { + "epoch": 0.26240962391396805, + "grad_norm": 0.1667470633983612, + "learning_rate": 8.421576967100607e-05, + "loss": 1.077, + "step": 4319 + }, + { + "epoch": 0.2624703809465946, + "grad_norm": 0.2424613982439041, + "learning_rate": 8.420878871043757e-05, + "loss": 1.1314, + "step": 4320 + }, + { + "epoch": 0.2625311379792211, + "grad_norm": 0.1981724351644516, + "learning_rate": 8.420180649593929e-05, + "loss": 1.1085, + "step": 4321 + }, + { + "epoch": 0.2625918950118476, + "grad_norm": 0.36126285791397095, + "learning_rate": 8.419482302776721e-05, + "loss": 1.1671, + "step": 4322 + }, + { + "epoch": 0.26265265204447413, + "grad_norm": 0.20216041803359985, + "learning_rate": 8.418783830617727e-05, + "loss": 1.1681, + "step": 4323 + }, + { + "epoch": 0.26271340907710067, + "grad_norm": 0.4142190217971802, + "learning_rate": 8.418085233142551e-05, + "loss": 1.079, + "step": 4324 + }, + { + "epoch": 0.2627741661097272, + "grad_norm": 0.4255751967430115, + "learning_rate": 8.4173865103768e-05, + "loss": 1.0658, + "step": 4325 + }, + { + "epoch": 0.26283492314235374, + "grad_norm": 0.3624703288078308, + "learning_rate": 8.416687662346085e-05, + "loss": 1.1247, + "step": 4326 + }, + { + "epoch": 0.2628956801749803, + "grad_norm": 0.5399686098098755, + "learning_rate": 8.415988689076025e-05, + "loss": 1.111, + "step": 4327 + }, + { + "epoch": 0.2629564372076068, + "grad_norm": 0.22710169851779938, + "learning_rate": 8.415289590592237e-05, + "loss": 1.1071, + "step": 4328 + }, + { + "epoch": 0.2630171942402333, + "grad_norm": 0.3340166211128235, + "learning_rate": 8.41459036692035e-05, + "loss": 1.0599, + "step": 4329 + }, + { + "epoch": 0.2630779512728598, + "grad_norm": 0.26834020018577576, + "learning_rate": 8.413891018085993e-05, + "loss": 1.1779, + "step": 4330 + }, + { + "epoch": 0.26313870830548636, + "grad_norm": 0.35126960277557373, + "learning_rate": 8.413191544114803e-05, + "loss": 1.0928, + "step": 4331 + }, + { + "epoch": 0.2631994653381129, + "grad_norm": 0.49582040309906006, + "learning_rate": 8.412491945032414e-05, + "loss": 1.0681, + "step": 4332 + }, + { + "epoch": 0.2632602223707394, + "grad_norm": 0.1998249590396881, + "learning_rate": 8.411792220864474e-05, + "loss": 1.0556, + "step": 4333 + }, + { + "epoch": 0.26332097940336596, + "grad_norm": 0.31229686737060547, + "learning_rate": 8.41109237163663e-05, + "loss": 1.0992, + "step": 4334 + }, + { + "epoch": 0.26338173643599244, + "grad_norm": 0.2491099238395691, + "learning_rate": 8.410392397374538e-05, + "loss": 1.0788, + "step": 4335 + }, + { + "epoch": 0.263442493468619, + "grad_norm": 0.4382314085960388, + "learning_rate": 8.409692298103852e-05, + "loss": 1.0591, + "step": 4336 + }, + { + "epoch": 0.2635032505012455, + "grad_norm": 0.2499891221523285, + "learning_rate": 8.408992073850237e-05, + "loss": 1.182, + "step": 4337 + }, + { + "epoch": 0.26356400753387205, + "grad_norm": 0.21687695384025574, + "learning_rate": 8.408291724639357e-05, + "loss": 1.1483, + "step": 4338 + }, + { + "epoch": 0.2636247645664986, + "grad_norm": 0.4915340542793274, + "learning_rate": 8.407591250496885e-05, + "loss": 1.0649, + "step": 4339 + }, + { + "epoch": 0.2636855215991251, + "grad_norm": 0.18667708337306976, + "learning_rate": 8.406890651448499e-05, + "loss": 1.0589, + "step": 4340 + }, + { + "epoch": 0.26374627863175165, + "grad_norm": 4.36269998550415, + "learning_rate": 8.406189927519876e-05, + "loss": 1.1752, + "step": 4341 + }, + { + "epoch": 0.26380703566437813, + "grad_norm": 0.33686190843582153, + "learning_rate": 8.405489078736704e-05, + "loss": 1.104, + "step": 4342 + }, + { + "epoch": 0.26386779269700467, + "grad_norm": 0.29295504093170166, + "learning_rate": 8.40478810512467e-05, + "loss": 1.0362, + "step": 4343 + }, + { + "epoch": 0.2639285497296312, + "grad_norm": 0.24595855176448822, + "learning_rate": 8.404087006709471e-05, + "loss": 1.1191, + "step": 4344 + }, + { + "epoch": 0.26398930676225774, + "grad_norm": 0.2646639347076416, + "learning_rate": 8.403385783516804e-05, + "loss": 1.2243, + "step": 4345 + }, + { + "epoch": 0.26405006379488427, + "grad_norm": 0.23265108466148376, + "learning_rate": 8.402684435572374e-05, + "loss": 1.1493, + "step": 4346 + }, + { + "epoch": 0.2641108208275108, + "grad_norm": 0.25843653082847595, + "learning_rate": 8.401982962901887e-05, + "loss": 1.0626, + "step": 4347 + }, + { + "epoch": 0.2641715778601373, + "grad_norm": 0.18480026721954346, + "learning_rate": 8.401281365531058e-05, + "loss": 1.0603, + "step": 4348 + }, + { + "epoch": 0.2642323348927638, + "grad_norm": 0.2599584758281708, + "learning_rate": 8.400579643485602e-05, + "loss": 1.0977, + "step": 4349 + }, + { + "epoch": 0.26429309192539036, + "grad_norm": 0.32290536165237427, + "learning_rate": 8.399877796791244e-05, + "loss": 1.2481, + "step": 4350 + }, + { + "epoch": 0.2643538489580169, + "grad_norm": 0.2697908282279968, + "learning_rate": 8.399175825473706e-05, + "loss": 1.1819, + "step": 4351 + }, + { + "epoch": 0.2644146059906434, + "grad_norm": 0.3472760319709778, + "learning_rate": 8.398473729558723e-05, + "loss": 1.1714, + "step": 4352 + }, + { + "epoch": 0.26447536302326996, + "grad_norm": 0.2620764672756195, + "learning_rate": 8.397771509072028e-05, + "loss": 1.091, + "step": 4353 + }, + { + "epoch": 0.2645361200558965, + "grad_norm": 0.4871889352798462, + "learning_rate": 8.39706916403936e-05, + "loss": 1.1135, + "step": 4354 + }, + { + "epoch": 0.264596877088523, + "grad_norm": 0.19167040288448334, + "learning_rate": 8.396366694486466e-05, + "loss": 1.106, + "step": 4355 + }, + { + "epoch": 0.2646576341211495, + "grad_norm": 0.2969217896461487, + "learning_rate": 8.395664100439096e-05, + "loss": 1.0232, + "step": 4356 + }, + { + "epoch": 0.26471839115377604, + "grad_norm": 0.33228251338005066, + "learning_rate": 8.394961381922999e-05, + "loss": 1.1068, + "step": 4357 + }, + { + "epoch": 0.2647791481864026, + "grad_norm": 0.32561129331588745, + "learning_rate": 8.394258538963937e-05, + "loss": 1.1309, + "step": 4358 + }, + { + "epoch": 0.2648399052190291, + "grad_norm": 0.3207565248012543, + "learning_rate": 8.393555571587673e-05, + "loss": 1.0377, + "step": 4359 + }, + { + "epoch": 0.26490066225165565, + "grad_norm": 0.20054785907268524, + "learning_rate": 8.392852479819974e-05, + "loss": 1.137, + "step": 4360 + }, + { + "epoch": 0.26496141928428213, + "grad_norm": 0.2658354341983795, + "learning_rate": 8.392149263686611e-05, + "loss": 1.0979, + "step": 4361 + }, + { + "epoch": 0.26502217631690866, + "grad_norm": 0.36198854446411133, + "learning_rate": 8.391445923213361e-05, + "loss": 1.0933, + "step": 4362 + }, + { + "epoch": 0.2650829333495352, + "grad_norm": 14.417988777160645, + "learning_rate": 8.390742458426007e-05, + "loss": 1.2295, + "step": 4363 + }, + { + "epoch": 0.26514369038216173, + "grad_norm": 0.570966362953186, + "learning_rate": 8.39003886935033e-05, + "loss": 1.1695, + "step": 4364 + }, + { + "epoch": 0.26520444741478827, + "grad_norm": 0.23233571648597717, + "learning_rate": 8.389335156012124e-05, + "loss": 1.1126, + "step": 4365 + }, + { + "epoch": 0.2652652044474148, + "grad_norm": 0.7579413056373596, + "learning_rate": 8.388631318437183e-05, + "loss": 1.0859, + "step": 4366 + }, + { + "epoch": 0.26532596148004134, + "grad_norm": 0.22977612912654877, + "learning_rate": 8.387927356651306e-05, + "loss": 1.1497, + "step": 4367 + }, + { + "epoch": 0.2653867185126678, + "grad_norm": 0.41820210218429565, + "learning_rate": 8.387223270680296e-05, + "loss": 1.0749, + "step": 4368 + }, + { + "epoch": 0.26544747554529435, + "grad_norm": 0.5477020144462585, + "learning_rate": 8.386519060549962e-05, + "loss": 1.1062, + "step": 4369 + }, + { + "epoch": 0.2655082325779209, + "grad_norm": 0.31084322929382324, + "learning_rate": 8.38581472628612e-05, + "loss": 1.1104, + "step": 4370 + }, + { + "epoch": 0.2655689896105474, + "grad_norm": 0.49966466426849365, + "learning_rate": 8.385110267914582e-05, + "loss": 1.1404, + "step": 4371 + }, + { + "epoch": 0.26562974664317396, + "grad_norm": 0.3427087366580963, + "learning_rate": 8.384405685461174e-05, + "loss": 1.1574, + "step": 4372 + }, + { + "epoch": 0.2656905036758005, + "grad_norm": 0.5143511295318604, + "learning_rate": 8.383700978951722e-05, + "loss": 1.1574, + "step": 4373 + }, + { + "epoch": 0.265751260708427, + "grad_norm": 0.5515393018722534, + "learning_rate": 8.382996148412055e-05, + "loss": 1.0723, + "step": 4374 + }, + { + "epoch": 0.2658120177410535, + "grad_norm": 0.22138310968875885, + "learning_rate": 8.38229119386801e-05, + "loss": 1.1601, + "step": 4375 + }, + { + "epoch": 0.26587277477368004, + "grad_norm": 0.5872916579246521, + "learning_rate": 8.38158611534543e-05, + "loss": 1.1063, + "step": 4376 + }, + { + "epoch": 0.2659335318063066, + "grad_norm": 0.4621509909629822, + "learning_rate": 8.380880912870155e-05, + "loss": 1.0842, + "step": 4377 + }, + { + "epoch": 0.2659942888389331, + "grad_norm": 0.4543784558773041, + "learning_rate": 8.380175586468038e-05, + "loss": 1.099, + "step": 4378 + }, + { + "epoch": 0.26605504587155965, + "grad_norm": 0.6176244020462036, + "learning_rate": 8.37947013616493e-05, + "loss": 1.0805, + "step": 4379 + }, + { + "epoch": 0.2661158029041862, + "grad_norm": 0.1799081563949585, + "learning_rate": 8.378764561986691e-05, + "loss": 0.9934, + "step": 4380 + }, + { + "epoch": 0.26617655993681266, + "grad_norm": 0.45475855469703674, + "learning_rate": 8.378058863959186e-05, + "loss": 1.1279, + "step": 4381 + }, + { + "epoch": 0.2662373169694392, + "grad_norm": 0.394123911857605, + "learning_rate": 8.377353042108277e-05, + "loss": 1.0986, + "step": 4382 + }, + { + "epoch": 0.26629807400206573, + "grad_norm": 0.2000361829996109, + "learning_rate": 8.376647096459843e-05, + "loss": 1.0742, + "step": 4383 + }, + { + "epoch": 0.26635883103469227, + "grad_norm": 0.5201019644737244, + "learning_rate": 8.375941027039755e-05, + "loss": 1.0507, + "step": 4384 + }, + { + "epoch": 0.2664195880673188, + "grad_norm": 0.195346400141716, + "learning_rate": 8.375234833873898e-05, + "loss": 1.0689, + "step": 4385 + }, + { + "epoch": 0.26648034509994534, + "grad_norm": 0.38381296396255493, + "learning_rate": 8.374528516988154e-05, + "loss": 1.0421, + "step": 4386 + }, + { + "epoch": 0.26654110213257187, + "grad_norm": 0.29690536856651306, + "learning_rate": 8.373822076408415e-05, + "loss": 1.0599, + "step": 4387 + }, + { + "epoch": 0.26660185916519835, + "grad_norm": 0.19969159364700317, + "learning_rate": 8.373115512160576e-05, + "loss": 1.159, + "step": 4388 + }, + { + "epoch": 0.2666626161978249, + "grad_norm": 3.2348456382751465, + "learning_rate": 8.372408824270534e-05, + "loss": 1.1338, + "step": 4389 + }, + { + "epoch": 0.2667233732304514, + "grad_norm": 0.4804811179637909, + "learning_rate": 8.371702012764198e-05, + "loss": 1.1437, + "step": 4390 + }, + { + "epoch": 0.26678413026307796, + "grad_norm": 0.21917611360549927, + "learning_rate": 8.370995077667471e-05, + "loss": 1.0686, + "step": 4391 + }, + { + "epoch": 0.2668448872957045, + "grad_norm": 0.4135713279247284, + "learning_rate": 8.370288019006269e-05, + "loss": 1.0766, + "step": 4392 + }, + { + "epoch": 0.266905644328331, + "grad_norm": 0.3015177845954895, + "learning_rate": 8.369580836806508e-05, + "loss": 1.0967, + "step": 4393 + }, + { + "epoch": 0.2669664013609575, + "grad_norm": 0.3408491015434265, + "learning_rate": 8.368873531094109e-05, + "loss": 1.2322, + "step": 4394 + }, + { + "epoch": 0.26702715839358404, + "grad_norm": 0.23241235315799713, + "learning_rate": 8.368166101894999e-05, + "loss": 1.1435, + "step": 4395 + }, + { + "epoch": 0.2670879154262106, + "grad_norm": 0.3057917654514313, + "learning_rate": 8.36745854923511e-05, + "loss": 1.166, + "step": 4396 + }, + { + "epoch": 0.2671486724588371, + "grad_norm": 0.2352297157049179, + "learning_rate": 8.366750873140379e-05, + "loss": 1.1263, + "step": 4397 + }, + { + "epoch": 0.26720942949146365, + "grad_norm": 6.783035755157471, + "learning_rate": 8.366043073636743e-05, + "loss": 1.1894, + "step": 4398 + }, + { + "epoch": 0.2672701865240902, + "grad_norm": 0.3271315395832062, + "learning_rate": 8.365335150750147e-05, + "loss": 1.12, + "step": 4399 + }, + { + "epoch": 0.2673309435567167, + "grad_norm": 0.2918102443218231, + "learning_rate": 8.36462710450654e-05, + "loss": 1.0892, + "step": 4400 + }, + { + "epoch": 0.2673917005893432, + "grad_norm": 0.2087276577949524, + "learning_rate": 8.363918934931875e-05, + "loss": 1.1161, + "step": 4401 + }, + { + "epoch": 0.26745245762196973, + "grad_norm": 0.19637629389762878, + "learning_rate": 8.363210642052113e-05, + "loss": 1.0399, + "step": 4402 + }, + { + "epoch": 0.26751321465459627, + "grad_norm": 0.30010461807250977, + "learning_rate": 8.362502225893216e-05, + "loss": 1.0752, + "step": 4403 + }, + { + "epoch": 0.2675739716872228, + "grad_norm": 0.21788573265075684, + "learning_rate": 8.361793686481149e-05, + "loss": 1.2208, + "step": 4404 + }, + { + "epoch": 0.26763472871984934, + "grad_norm": 0.30013507604599, + "learning_rate": 8.361085023841884e-05, + "loss": 1.2609, + "step": 4405 + }, + { + "epoch": 0.26769548575247587, + "grad_norm": 0.2899112403392792, + "learning_rate": 8.360376238001397e-05, + "loss": 1.1493, + "step": 4406 + }, + { + "epoch": 0.26775624278510235, + "grad_norm": 0.19916804134845734, + "learning_rate": 8.359667328985672e-05, + "loss": 1.0945, + "step": 4407 + }, + { + "epoch": 0.2678169998177289, + "grad_norm": 0.21261560916900635, + "learning_rate": 8.358958296820689e-05, + "loss": 1.1082, + "step": 4408 + }, + { + "epoch": 0.2678777568503554, + "grad_norm": 0.4338446259498596, + "learning_rate": 8.358249141532442e-05, + "loss": 1.0905, + "step": 4409 + }, + { + "epoch": 0.26793851388298195, + "grad_norm": 0.21074606478214264, + "learning_rate": 8.357539863146922e-05, + "loss": 1.1419, + "step": 4410 + }, + { + "epoch": 0.2679992709156085, + "grad_norm": 0.39377039670944214, + "learning_rate": 8.356830461690131e-05, + "loss": 1.0397, + "step": 4411 + }, + { + "epoch": 0.268060027948235, + "grad_norm": 0.24027006328105927, + "learning_rate": 8.35612093718807e-05, + "loss": 1.0611, + "step": 4412 + }, + { + "epoch": 0.26812078498086156, + "grad_norm": 0.3314197063446045, + "learning_rate": 8.355411289666749e-05, + "loss": 1.0934, + "step": 4413 + }, + { + "epoch": 0.26818154201348804, + "grad_norm": 0.19897693395614624, + "learning_rate": 8.354701519152175e-05, + "loss": 1.1058, + "step": 4414 + }, + { + "epoch": 0.2682422990461146, + "grad_norm": 0.25227728486061096, + "learning_rate": 8.35399162567037e-05, + "loss": 1.0667, + "step": 4415 + }, + { + "epoch": 0.2683030560787411, + "grad_norm": 0.43013399839401245, + "learning_rate": 8.353281609247355e-05, + "loss": 1.204, + "step": 4416 + }, + { + "epoch": 0.26836381311136764, + "grad_norm": 0.20363515615463257, + "learning_rate": 8.352571469909152e-05, + "loss": 1.0807, + "step": 4417 + }, + { + "epoch": 0.2684245701439942, + "grad_norm": 0.329195499420166, + "learning_rate": 8.351861207681795e-05, + "loss": 1.1642, + "step": 4418 + }, + { + "epoch": 0.2684853271766207, + "grad_norm": 0.5914226174354553, + "learning_rate": 8.351150822591318e-05, + "loss": 1.176, + "step": 4419 + }, + { + "epoch": 0.2685460842092472, + "grad_norm": 0.27818551659584045, + "learning_rate": 8.350440314663759e-05, + "loss": 1.0943, + "step": 4420 + }, + { + "epoch": 0.26860684124187373, + "grad_norm": 0.3945620656013489, + "learning_rate": 8.349729683925163e-05, + "loss": 1.2065, + "step": 4421 + }, + { + "epoch": 0.26866759827450026, + "grad_norm": 0.1966278851032257, + "learning_rate": 8.349018930401577e-05, + "loss": 1.092, + "step": 4422 + }, + { + "epoch": 0.2687283553071268, + "grad_norm": 0.21751713752746582, + "learning_rate": 8.348308054119058e-05, + "loss": 1.1605, + "step": 4423 + }, + { + "epoch": 0.26878911233975333, + "grad_norm": 0.2212274819612503, + "learning_rate": 8.347597055103658e-05, + "loss": 1.1126, + "step": 4424 + }, + { + "epoch": 0.26884986937237987, + "grad_norm": 2.9583353996276855, + "learning_rate": 8.346885933381439e-05, + "loss": 1.1589, + "step": 4425 + }, + { + "epoch": 0.2689106264050064, + "grad_norm": 0.34951403737068176, + "learning_rate": 8.346174688978471e-05, + "loss": 1.1574, + "step": 4426 + }, + { + "epoch": 0.2689713834376329, + "grad_norm": 0.24047352373600006, + "learning_rate": 8.345463321920823e-05, + "loss": 1.0528, + "step": 4427 + }, + { + "epoch": 0.2690321404702594, + "grad_norm": 0.2411516308784485, + "learning_rate": 8.34475183223457e-05, + "loss": 1.0654, + "step": 4428 + }, + { + "epoch": 0.26909289750288595, + "grad_norm": 0.3529675304889679, + "learning_rate": 8.344040219945793e-05, + "loss": 1.1128, + "step": 4429 + }, + { + "epoch": 0.2691536545355125, + "grad_norm": 0.22589266300201416, + "learning_rate": 8.343328485080574e-05, + "loss": 1.0863, + "step": 4430 + }, + { + "epoch": 0.269214411568139, + "grad_norm": 0.35135719180107117, + "learning_rate": 8.342616627665003e-05, + "loss": 1.1592, + "step": 4431 + }, + { + "epoch": 0.26927516860076556, + "grad_norm": 0.3361550569534302, + "learning_rate": 8.341904647725174e-05, + "loss": 1.1825, + "step": 4432 + }, + { + "epoch": 0.2693359256333921, + "grad_norm": 0.32101860642433167, + "learning_rate": 8.341192545287185e-05, + "loss": 1.1401, + "step": 4433 + }, + { + "epoch": 0.26939668266601857, + "grad_norm": 0.2533859312534332, + "learning_rate": 8.340480320377136e-05, + "loss": 1.0993, + "step": 4434 + }, + { + "epoch": 0.2694574396986451, + "grad_norm": 0.21875229477882385, + "learning_rate": 8.339767973021135e-05, + "loss": 1.101, + "step": 4435 + }, + { + "epoch": 0.26951819673127164, + "grad_norm": 0.24688014388084412, + "learning_rate": 8.339055503245294e-05, + "loss": 1.1043, + "step": 4436 + }, + { + "epoch": 0.2695789537638982, + "grad_norm": 0.19598980247974396, + "learning_rate": 8.338342911075729e-05, + "loss": 1.1378, + "step": 4437 + }, + { + "epoch": 0.2696397107965247, + "grad_norm": 0.1976533830165863, + "learning_rate": 8.33763019653856e-05, + "loss": 1.1499, + "step": 4438 + }, + { + "epoch": 0.26970046782915125, + "grad_norm": 0.32871246337890625, + "learning_rate": 8.336917359659909e-05, + "loss": 1.1565, + "step": 4439 + }, + { + "epoch": 0.2697612248617777, + "grad_norm": 0.21482804417610168, + "learning_rate": 8.336204400465908e-05, + "loss": 1.0315, + "step": 4440 + }, + { + "epoch": 0.26982198189440426, + "grad_norm": 0.5966326594352722, + "learning_rate": 8.335491318982689e-05, + "loss": 1.3054, + "step": 4441 + }, + { + "epoch": 0.2698827389270308, + "grad_norm": 0.21577760577201843, + "learning_rate": 8.334778115236392e-05, + "loss": 1.1041, + "step": 4442 + }, + { + "epoch": 0.26994349595965733, + "grad_norm": 0.2320018857717514, + "learning_rate": 8.334064789253157e-05, + "loss": 1.1062, + "step": 4443 + }, + { + "epoch": 0.27000425299228387, + "grad_norm": 0.17997534573078156, + "learning_rate": 8.333351341059134e-05, + "loss": 1.0765, + "step": 4444 + }, + { + "epoch": 0.2700650100249104, + "grad_norm": 0.33960995078086853, + "learning_rate": 8.332637770680474e-05, + "loss": 1.0646, + "step": 4445 + }, + { + "epoch": 0.27012576705753694, + "grad_norm": 0.24968475103378296, + "learning_rate": 8.331924078143329e-05, + "loss": 1.1461, + "step": 4446 + }, + { + "epoch": 0.2701865240901634, + "grad_norm": 0.5295467376708984, + "learning_rate": 8.331210263473867e-05, + "loss": 1.1431, + "step": 4447 + }, + { + "epoch": 0.27024728112278995, + "grad_norm": 1.67926025390625, + "learning_rate": 8.330496326698247e-05, + "loss": 1.0577, + "step": 4448 + }, + { + "epoch": 0.2703080381554165, + "grad_norm": 0.39247840642929077, + "learning_rate": 8.329782267842641e-05, + "loss": 1.1531, + "step": 4449 + }, + { + "epoch": 0.270368795188043, + "grad_norm": 0.44849616289138794, + "learning_rate": 8.329068086933221e-05, + "loss": 1.0747, + "step": 4450 + }, + { + "epoch": 0.27042955222066956, + "grad_norm": 0.356730192899704, + "learning_rate": 8.328353783996168e-05, + "loss": 1.0966, + "step": 4451 + }, + { + "epoch": 0.2704903092532961, + "grad_norm": 0.43844807147979736, + "learning_rate": 8.327639359057663e-05, + "loss": 1.0472, + "step": 4452 + }, + { + "epoch": 0.27055106628592257, + "grad_norm": 2.6905150413513184, + "learning_rate": 8.326924812143894e-05, + "loss": 1.1129, + "step": 4453 + }, + { + "epoch": 0.2706118233185491, + "grad_norm": 0.4950474202632904, + "learning_rate": 8.326210143281055e-05, + "loss": 1.0577, + "step": 4454 + }, + { + "epoch": 0.27067258035117564, + "grad_norm": 0.2950094938278198, + "learning_rate": 8.325495352495339e-05, + "loss": 1.0446, + "step": 4455 + }, + { + "epoch": 0.2707333373838022, + "grad_norm": 4.69804573059082, + "learning_rate": 8.324780439812946e-05, + "loss": 1.1252, + "step": 4456 + }, + { + "epoch": 0.2707940944164287, + "grad_norm": 0.5593882203102112, + "learning_rate": 8.324065405260085e-05, + "loss": 1.1366, + "step": 4457 + }, + { + "epoch": 0.27085485144905525, + "grad_norm": 0.22021150588989258, + "learning_rate": 8.323350248862964e-05, + "loss": 1.0988, + "step": 4458 + }, + { + "epoch": 0.2709156084816818, + "grad_norm": 0.319269061088562, + "learning_rate": 8.322634970647797e-05, + "loss": 1.1774, + "step": 4459 + }, + { + "epoch": 0.27097636551430826, + "grad_norm": 1.1788930892944336, + "learning_rate": 8.321919570640803e-05, + "loss": 1.3237, + "step": 4460 + }, + { + "epoch": 0.2710371225469348, + "grad_norm": 0.2121584117412567, + "learning_rate": 8.321204048868206e-05, + "loss": 1.0996, + "step": 4461 + }, + { + "epoch": 0.27109787957956133, + "grad_norm": 0.27695637941360474, + "learning_rate": 8.32048840535623e-05, + "loss": 1.1145, + "step": 4462 + }, + { + "epoch": 0.27115863661218786, + "grad_norm": 0.47333312034606934, + "learning_rate": 8.319772640131113e-05, + "loss": 1.2801, + "step": 4463 + }, + { + "epoch": 0.2712193936448144, + "grad_norm": 0.2464013546705246, + "learning_rate": 8.319056753219086e-05, + "loss": 1.1489, + "step": 4464 + }, + { + "epoch": 0.27128015067744093, + "grad_norm": 0.19197922945022583, + "learning_rate": 8.318340744646393e-05, + "loss": 1.2153, + "step": 4465 + }, + { + "epoch": 0.2713409077100674, + "grad_norm": 0.4292735755443573, + "learning_rate": 8.317624614439278e-05, + "loss": 1.3433, + "step": 4466 + }, + { + "epoch": 0.27140166474269395, + "grad_norm": 0.38255488872528076, + "learning_rate": 8.316908362623991e-05, + "loss": 1.0753, + "step": 4467 + }, + { + "epoch": 0.2714624217753205, + "grad_norm": 0.2850903868675232, + "learning_rate": 8.31619198922679e-05, + "loss": 1.2107, + "step": 4468 + }, + { + "epoch": 0.271523178807947, + "grad_norm": 0.31674012541770935, + "learning_rate": 8.315475494273927e-05, + "loss": 1.1508, + "step": 4469 + }, + { + "epoch": 0.27158393584057355, + "grad_norm": 0.24269595742225647, + "learning_rate": 8.31475887779167e-05, + "loss": 1.1452, + "step": 4470 + }, + { + "epoch": 0.2716446928732001, + "grad_norm": 0.19250309467315674, + "learning_rate": 8.314042139806287e-05, + "loss": 1.0395, + "step": 4471 + }, + { + "epoch": 0.2717054499058266, + "grad_norm": 0.170707106590271, + "learning_rate": 8.313325280344046e-05, + "loss": 1.062, + "step": 4472 + }, + { + "epoch": 0.2717662069384531, + "grad_norm": 0.2630065977573395, + "learning_rate": 8.312608299431229e-05, + "loss": 1.075, + "step": 4473 + }, + { + "epoch": 0.27182696397107964, + "grad_norm": 1.1534496545791626, + "learning_rate": 8.311891197094113e-05, + "loss": 1.1498, + "step": 4474 + }, + { + "epoch": 0.2718877210037062, + "grad_norm": 0.29185751080513, + "learning_rate": 8.311173973358985e-05, + "loss": 1.0869, + "step": 4475 + }, + { + "epoch": 0.2719484780363327, + "grad_norm": 0.16828122735023499, + "learning_rate": 8.310456628252136e-05, + "loss": 1.0625, + "step": 4476 + }, + { + "epoch": 0.27200923506895924, + "grad_norm": 0.285775363445282, + "learning_rate": 8.309739161799859e-05, + "loss": 1.0919, + "step": 4477 + }, + { + "epoch": 0.2720699921015858, + "grad_norm": 0.23441657423973083, + "learning_rate": 8.309021574028453e-05, + "loss": 1.1188, + "step": 4478 + }, + { + "epoch": 0.27213074913421226, + "grad_norm": 0.20591187477111816, + "learning_rate": 8.30830386496422e-05, + "loss": 1.0619, + "step": 4479 + }, + { + "epoch": 0.2721915061668388, + "grad_norm": 0.31597620248794556, + "learning_rate": 8.307586034633472e-05, + "loss": 1.2292, + "step": 4480 + }, + { + "epoch": 0.2722522631994653, + "grad_norm": 0.30900150537490845, + "learning_rate": 8.306868083062518e-05, + "loss": 1.0914, + "step": 4481 + }, + { + "epoch": 0.27231302023209186, + "grad_norm": 0.2688179612159729, + "learning_rate": 8.306150010277675e-05, + "loss": 1.1122, + "step": 4482 + }, + { + "epoch": 0.2723737772647184, + "grad_norm": 0.17576207220554352, + "learning_rate": 8.305431816305262e-05, + "loss": 1.1027, + "step": 4483 + }, + { + "epoch": 0.27243453429734493, + "grad_norm": 0.4719553589820862, + "learning_rate": 8.30471350117161e-05, + "loss": 1.0398, + "step": 4484 + }, + { + "epoch": 0.27249529132997147, + "grad_norm": 0.19829913973808289, + "learning_rate": 8.303995064903045e-05, + "loss": 1.0574, + "step": 4485 + }, + { + "epoch": 0.27255604836259795, + "grad_norm": 0.3735748529434204, + "learning_rate": 8.303276507525902e-05, + "loss": 1.0643, + "step": 4486 + }, + { + "epoch": 0.2726168053952245, + "grad_norm": 0.22488151490688324, + "learning_rate": 8.30255782906652e-05, + "loss": 1.1137, + "step": 4487 + }, + { + "epoch": 0.272677562427851, + "grad_norm": 0.28803712129592896, + "learning_rate": 8.301839029551243e-05, + "loss": 1.0945, + "step": 4488 + }, + { + "epoch": 0.27273831946047755, + "grad_norm": 0.2904054820537567, + "learning_rate": 8.301120109006418e-05, + "loss": 1.1321, + "step": 4489 + }, + { + "epoch": 0.2727990764931041, + "grad_norm": 0.22809189558029175, + "learning_rate": 8.300401067458396e-05, + "loss": 1.0777, + "step": 4490 + }, + { + "epoch": 0.2728598335257306, + "grad_norm": 0.31189483404159546, + "learning_rate": 8.299681904933536e-05, + "loss": 1.1666, + "step": 4491 + }, + { + "epoch": 0.27292059055835716, + "grad_norm": 0.23239701986312866, + "learning_rate": 8.298962621458197e-05, + "loss": 1.0493, + "step": 4492 + }, + { + "epoch": 0.27298134759098364, + "grad_norm": 0.4474465847015381, + "learning_rate": 8.298243217058747e-05, + "loss": 1.0671, + "step": 4493 + }, + { + "epoch": 0.27304210462361017, + "grad_norm": 0.25847166776657104, + "learning_rate": 8.297523691761554e-05, + "loss": 1.0451, + "step": 4494 + }, + { + "epoch": 0.2731028616562367, + "grad_norm": 0.3619801104068756, + "learning_rate": 8.296804045592992e-05, + "loss": 1.0723, + "step": 4495 + }, + { + "epoch": 0.27316361868886324, + "grad_norm": 0.40111687779426575, + "learning_rate": 8.296084278579442e-05, + "loss": 1.3051, + "step": 4496 + }, + { + "epoch": 0.2732243757214898, + "grad_norm": 0.29694414138793945, + "learning_rate": 8.295364390747284e-05, + "loss": 1.0709, + "step": 4497 + }, + { + "epoch": 0.2732851327541163, + "grad_norm": 0.43788018822669983, + "learning_rate": 8.294644382122906e-05, + "loss": 1.2631, + "step": 4498 + }, + { + "epoch": 0.2733458897867428, + "grad_norm": 0.36168351769447327, + "learning_rate": 8.293924252732704e-05, + "loss": 1.105, + "step": 4499 + }, + { + "epoch": 0.2734066468193693, + "grad_norm": 0.6962228417396545, + "learning_rate": 8.29320400260307e-05, + "loss": 1.263, + "step": 4500 + }, + { + "epoch": 0.27346740385199586, + "grad_norm": 0.3127216100692749, + "learning_rate": 8.292483631760408e-05, + "loss": 1.2386, + "step": 4501 + }, + { + "epoch": 0.2735281608846224, + "grad_norm": 0.27952879667282104, + "learning_rate": 8.291763140231122e-05, + "loss": 1.1337, + "step": 4502 + }, + { + "epoch": 0.27358891791724893, + "grad_norm": 0.38213732838630676, + "learning_rate": 8.291042528041622e-05, + "loss": 1.113, + "step": 4503 + }, + { + "epoch": 0.27364967494987547, + "grad_norm": 0.27743270993232727, + "learning_rate": 8.290321795218321e-05, + "loss": 1.0594, + "step": 4504 + }, + { + "epoch": 0.273710431982502, + "grad_norm": 0.25722596049308777, + "learning_rate": 8.289600941787639e-05, + "loss": 1.08, + "step": 4505 + }, + { + "epoch": 0.2737711890151285, + "grad_norm": 0.33142557740211487, + "learning_rate": 8.288879967775997e-05, + "loss": 1.0838, + "step": 4506 + }, + { + "epoch": 0.273831946047755, + "grad_norm": 0.41427505016326904, + "learning_rate": 8.288158873209826e-05, + "loss": 1.2089, + "step": 4507 + }, + { + "epoch": 0.27389270308038155, + "grad_norm": 0.2597549855709076, + "learning_rate": 8.287437658115556e-05, + "loss": 1.1131, + "step": 4508 + }, + { + "epoch": 0.2739534601130081, + "grad_norm": 0.28753677010536194, + "learning_rate": 8.286716322519624e-05, + "loss": 1.0716, + "step": 4509 + }, + { + "epoch": 0.2740142171456346, + "grad_norm": 0.21665723621845245, + "learning_rate": 8.285994866448468e-05, + "loss": 1.0744, + "step": 4510 + }, + { + "epoch": 0.27407497417826115, + "grad_norm": 0.28974324464797974, + "learning_rate": 8.285273289928535e-05, + "loss": 1.1968, + "step": 4511 + }, + { + "epoch": 0.27413573121088763, + "grad_norm": 0.19824141263961792, + "learning_rate": 8.284551592986277e-05, + "loss": 1.1261, + "step": 4512 + }, + { + "epoch": 0.27419648824351417, + "grad_norm": 0.7094298601150513, + "learning_rate": 8.283829775648143e-05, + "loss": 1.1475, + "step": 4513 + }, + { + "epoch": 0.2742572452761407, + "grad_norm": 0.3282495439052582, + "learning_rate": 8.283107837940593e-05, + "loss": 1.103, + "step": 4514 + }, + { + "epoch": 0.27431800230876724, + "grad_norm": 0.26868101954460144, + "learning_rate": 8.282385779890092e-05, + "loss": 1.1025, + "step": 4515 + }, + { + "epoch": 0.2743787593413938, + "grad_norm": 0.4487287700176239, + "learning_rate": 8.281663601523107e-05, + "loss": 1.2435, + "step": 4516 + }, + { + "epoch": 0.2744395163740203, + "grad_norm": 0.25463542342185974, + "learning_rate": 8.280941302866108e-05, + "loss": 1.1108, + "step": 4517 + }, + { + "epoch": 0.27450027340664684, + "grad_norm": 0.33673956990242004, + "learning_rate": 8.28021888394557e-05, + "loss": 1.1322, + "step": 4518 + }, + { + "epoch": 0.2745610304392733, + "grad_norm": 0.48676127195358276, + "learning_rate": 8.279496344787975e-05, + "loss": 1.0467, + "step": 4519 + }, + { + "epoch": 0.27462178747189986, + "grad_norm": 0.3328222632408142, + "learning_rate": 8.278773685419808e-05, + "loss": 1.1637, + "step": 4520 + }, + { + "epoch": 0.2746825445045264, + "grad_norm": 0.39478740096092224, + "learning_rate": 8.278050905867559e-05, + "loss": 1.1036, + "step": 4521 + }, + { + "epoch": 0.27474330153715293, + "grad_norm": 0.3095814287662506, + "learning_rate": 8.277328006157718e-05, + "loss": 1.1272, + "step": 4522 + }, + { + "epoch": 0.27480405856977946, + "grad_norm": 0.4269719123840332, + "learning_rate": 8.276604986316785e-05, + "loss": 1.0833, + "step": 4523 + }, + { + "epoch": 0.274864815602406, + "grad_norm": 0.34222328662872314, + "learning_rate": 8.275881846371265e-05, + "loss": 1.054, + "step": 4524 + }, + { + "epoch": 0.2749255726350325, + "grad_norm": 0.1977594941854477, + "learning_rate": 8.275158586347662e-05, + "loss": 1.0709, + "step": 4525 + }, + { + "epoch": 0.274986329667659, + "grad_norm": 0.507420539855957, + "learning_rate": 8.274435206272489e-05, + "loss": 1.0439, + "step": 4526 + }, + { + "epoch": 0.27504708670028555, + "grad_norm": 0.19630765914916992, + "learning_rate": 8.27371170617226e-05, + "loss": 1.0722, + "step": 4527 + }, + { + "epoch": 0.2751078437329121, + "grad_norm": 0.311144083738327, + "learning_rate": 8.272988086073494e-05, + "loss": 1.1275, + "step": 4528 + }, + { + "epoch": 0.2751686007655386, + "grad_norm": 0.3327949345111847, + "learning_rate": 8.272264346002719e-05, + "loss": 1.2369, + "step": 4529 + }, + { + "epoch": 0.27522935779816515, + "grad_norm": 0.32370978593826294, + "learning_rate": 8.271540485986461e-05, + "loss": 1.1795, + "step": 4530 + }, + { + "epoch": 0.2752901148307917, + "grad_norm": 0.29816627502441406, + "learning_rate": 8.270816506051253e-05, + "loss": 1.1365, + "step": 4531 + }, + { + "epoch": 0.27535087186341817, + "grad_norm": 1.914139747619629, + "learning_rate": 8.270092406223635e-05, + "loss": 1.1116, + "step": 4532 + }, + { + "epoch": 0.2754116288960447, + "grad_norm": 0.18972396850585938, + "learning_rate": 8.269368186530149e-05, + "loss": 1.1159, + "step": 4533 + }, + { + "epoch": 0.27547238592867124, + "grad_norm": 0.2889735698699951, + "learning_rate": 8.26864384699734e-05, + "loss": 1.1701, + "step": 4534 + }, + { + "epoch": 0.2755331429612978, + "grad_norm": 0.310587078332901, + "learning_rate": 8.267919387651757e-05, + "loss": 1.1019, + "step": 4535 + }, + { + "epoch": 0.2755938999939243, + "grad_norm": 0.254253089427948, + "learning_rate": 8.267194808519958e-05, + "loss": 1.0592, + "step": 4536 + }, + { + "epoch": 0.27565465702655084, + "grad_norm": 0.2537367641925812, + "learning_rate": 8.266470109628501e-05, + "loss": 1.0481, + "step": 4537 + }, + { + "epoch": 0.2757154140591774, + "grad_norm": 0.18268471956253052, + "learning_rate": 8.265745291003953e-05, + "loss": 1.0428, + "step": 4538 + }, + { + "epoch": 0.27577617109180386, + "grad_norm": 0.2010815143585205, + "learning_rate": 8.265020352672877e-05, + "loss": 1.035, + "step": 4539 + }, + { + "epoch": 0.2758369281244304, + "grad_norm": 0.15907078981399536, + "learning_rate": 8.26429529466185e-05, + "loss": 1.0916, + "step": 4540 + }, + { + "epoch": 0.2758976851570569, + "grad_norm": 0.3099391758441925, + "learning_rate": 8.26357011699745e-05, + "loss": 1.0273, + "step": 4541 + }, + { + "epoch": 0.27595844218968346, + "grad_norm": 0.19727499783039093, + "learning_rate": 8.262844819706255e-05, + "loss": 1.0987, + "step": 4542 + }, + { + "epoch": 0.27601919922231, + "grad_norm": 0.16243308782577515, + "learning_rate": 8.262119402814852e-05, + "loss": 1.0868, + "step": 4543 + }, + { + "epoch": 0.27607995625493653, + "grad_norm": 1.352102279663086, + "learning_rate": 8.261393866349832e-05, + "loss": 1.263, + "step": 4544 + }, + { + "epoch": 0.276140713287563, + "grad_norm": 0.22592325508594513, + "learning_rate": 8.260668210337789e-05, + "loss": 1.169, + "step": 4545 + }, + { + "epoch": 0.27620147032018955, + "grad_norm": 0.1809045821428299, + "learning_rate": 8.259942434805324e-05, + "loss": 1.071, + "step": 4546 + }, + { + "epoch": 0.2762622273528161, + "grad_norm": 0.2381470799446106, + "learning_rate": 8.259216539779038e-05, + "loss": 1.0926, + "step": 4547 + }, + { + "epoch": 0.2763229843854426, + "grad_norm": 1.194684386253357, + "learning_rate": 8.25849052528554e-05, + "loss": 1.0723, + "step": 4548 + }, + { + "epoch": 0.27638374141806915, + "grad_norm": 0.2497381716966629, + "learning_rate": 8.25776439135144e-05, + "loss": 1.0648, + "step": 4549 + }, + { + "epoch": 0.2764444984506957, + "grad_norm": 0.612889289855957, + "learning_rate": 8.25703813800336e-05, + "loss": 1.1093, + "step": 4550 + }, + { + "epoch": 0.2765052554833222, + "grad_norm": 0.22292213141918182, + "learning_rate": 8.256311765267916e-05, + "loss": 1.155, + "step": 4551 + }, + { + "epoch": 0.2765660125159487, + "grad_norm": 0.24881669878959656, + "learning_rate": 8.255585273171735e-05, + "loss": 1.0091, + "step": 4552 + }, + { + "epoch": 0.27662676954857524, + "grad_norm": 0.3226565420627594, + "learning_rate": 8.254858661741444e-05, + "loss": 1.1966, + "step": 4553 + }, + { + "epoch": 0.27668752658120177, + "grad_norm": 0.20807965099811554, + "learning_rate": 8.254131931003683e-05, + "loss": 1.0756, + "step": 4554 + }, + { + "epoch": 0.2767482836138283, + "grad_norm": 0.26362279057502747, + "learning_rate": 8.253405080985087e-05, + "loss": 1.1042, + "step": 4555 + }, + { + "epoch": 0.27680904064645484, + "grad_norm": 0.5383837819099426, + "learning_rate": 8.252678111712296e-05, + "loss": 1.3065, + "step": 4556 + }, + { + "epoch": 0.2768697976790814, + "grad_norm": 0.33916106820106506, + "learning_rate": 8.251951023211962e-05, + "loss": 1.1349, + "step": 4557 + }, + { + "epoch": 0.27693055471170785, + "grad_norm": 3.300945281982422, + "learning_rate": 8.251223815510734e-05, + "loss": 1.1283, + "step": 4558 + }, + { + "epoch": 0.2769913117443344, + "grad_norm": 0.5464043021202087, + "learning_rate": 8.250496488635268e-05, + "loss": 1.1048, + "step": 4559 + }, + { + "epoch": 0.2770520687769609, + "grad_norm": 0.24270865321159363, + "learning_rate": 8.249769042612227e-05, + "loss": 1.0849, + "step": 4560 + }, + { + "epoch": 0.27711282580958746, + "grad_norm": 0.46245354413986206, + "learning_rate": 8.249041477468273e-05, + "loss": 1.2, + "step": 4561 + }, + { + "epoch": 0.277173582842214, + "grad_norm": 0.2853115200996399, + "learning_rate": 8.248313793230074e-05, + "loss": 1.0691, + "step": 4562 + }, + { + "epoch": 0.27723433987484053, + "grad_norm": 0.3142154812812805, + "learning_rate": 8.247585989924306e-05, + "loss": 1.3461, + "step": 4563 + }, + { + "epoch": 0.27729509690746706, + "grad_norm": 0.3482964336872101, + "learning_rate": 8.246858067577648e-05, + "loss": 1.114, + "step": 4564 + }, + { + "epoch": 0.27735585394009354, + "grad_norm": 0.23432257771492004, + "learning_rate": 8.246130026216776e-05, + "loss": 1.1552, + "step": 4565 + }, + { + "epoch": 0.2774166109727201, + "grad_norm": 0.27266794443130493, + "learning_rate": 8.245401865868383e-05, + "loss": 1.1407, + "step": 4566 + }, + { + "epoch": 0.2774773680053466, + "grad_norm": 0.2638716995716095, + "learning_rate": 8.244673586559157e-05, + "loss": 1.1058, + "step": 4567 + }, + { + "epoch": 0.27753812503797315, + "grad_norm": 0.410335898399353, + "learning_rate": 8.243945188315791e-05, + "loss": 1.0986, + "step": 4568 + }, + { + "epoch": 0.2775988820705997, + "grad_norm": 0.7254387140274048, + "learning_rate": 8.243216671164991e-05, + "loss": 1.1177, + "step": 4569 + }, + { + "epoch": 0.2776596391032262, + "grad_norm": 0.2305910289287567, + "learning_rate": 8.242488035133454e-05, + "loss": 1.1272, + "step": 4570 + }, + { + "epoch": 0.2777203961358527, + "grad_norm": 0.35898032784461975, + "learning_rate": 8.241759280247893e-05, + "loss": 1.0516, + "step": 4571 + }, + { + "epoch": 0.27778115316847923, + "grad_norm": 0.28244301676750183, + "learning_rate": 8.241030406535017e-05, + "loss": 1.0495, + "step": 4572 + }, + { + "epoch": 0.27784191020110577, + "grad_norm": 0.23469945788383484, + "learning_rate": 8.240301414021548e-05, + "loss": 1.092, + "step": 4573 + }, + { + "epoch": 0.2779026672337323, + "grad_norm": 0.34002864360809326, + "learning_rate": 8.239572302734202e-05, + "loss": 1.6133, + "step": 4574 + }, + { + "epoch": 0.27796342426635884, + "grad_norm": 0.23743776977062225, + "learning_rate": 8.238843072699708e-05, + "loss": 1.0695, + "step": 4575 + }, + { + "epoch": 0.2780241812989854, + "grad_norm": 0.258198082447052, + "learning_rate": 8.238113723944795e-05, + "loss": 1.1424, + "step": 4576 + }, + { + "epoch": 0.2780849383316119, + "grad_norm": 0.3876579999923706, + "learning_rate": 8.237384256496197e-05, + "loss": 1.0621, + "step": 4577 + }, + { + "epoch": 0.2781456953642384, + "grad_norm": 0.20569178462028503, + "learning_rate": 8.236654670380655e-05, + "loss": 1.1371, + "step": 4578 + }, + { + "epoch": 0.2782064523968649, + "grad_norm": 0.31127026677131653, + "learning_rate": 8.235924965624909e-05, + "loss": 1.0436, + "step": 4579 + }, + { + "epoch": 0.27826720942949146, + "grad_norm": 0.17023280262947083, + "learning_rate": 8.235195142255707e-05, + "loss": 1.0861, + "step": 4580 + }, + { + "epoch": 0.278327966462118, + "grad_norm": 0.2348984032869339, + "learning_rate": 8.234465200299803e-05, + "loss": 1.2061, + "step": 4581 + }, + { + "epoch": 0.2783887234947445, + "grad_norm": 0.2197936624288559, + "learning_rate": 8.233735139783952e-05, + "loss": 1.1354, + "step": 4582 + }, + { + "epoch": 0.27844948052737106, + "grad_norm": 0.1964321881532669, + "learning_rate": 8.233004960734913e-05, + "loss": 1.1066, + "step": 4583 + }, + { + "epoch": 0.27851023755999754, + "grad_norm": 0.23386959731578827, + "learning_rate": 8.232274663179453e-05, + "loss": 1.2055, + "step": 4584 + }, + { + "epoch": 0.2785709945926241, + "grad_norm": 0.1769588738679886, + "learning_rate": 8.231544247144341e-05, + "loss": 1.0931, + "step": 4585 + }, + { + "epoch": 0.2786317516252506, + "grad_norm": 0.16374368965625763, + "learning_rate": 8.23081371265635e-05, + "loss": 1.0524, + "step": 4586 + }, + { + "epoch": 0.27869250865787715, + "grad_norm": 4.657360553741455, + "learning_rate": 8.230083059742257e-05, + "loss": 1.1016, + "step": 4587 + }, + { + "epoch": 0.2787532656905037, + "grad_norm": 0.19765813648700714, + "learning_rate": 8.229352288428846e-05, + "loss": 1.032, + "step": 4588 + }, + { + "epoch": 0.2788140227231302, + "grad_norm": 0.5915250778198242, + "learning_rate": 8.228621398742902e-05, + "loss": 1.1775, + "step": 4589 + }, + { + "epoch": 0.27887477975575675, + "grad_norm": 8.53027057647705, + "learning_rate": 8.227890390711216e-05, + "loss": 1.1496, + "step": 4590 + }, + { + "epoch": 0.27893553678838323, + "grad_norm": 0.2438729852437973, + "learning_rate": 8.227159264360586e-05, + "loss": 1.0588, + "step": 4591 + }, + { + "epoch": 0.27899629382100977, + "grad_norm": 0.21710507571697235, + "learning_rate": 8.226428019717807e-05, + "loss": 1.1071, + "step": 4592 + }, + { + "epoch": 0.2790570508536363, + "grad_norm": 0.4122526943683624, + "learning_rate": 8.225696656809688e-05, + "loss": 1.0854, + "step": 4593 + }, + { + "epoch": 0.27911780788626284, + "grad_norm": 0.200723797082901, + "learning_rate": 8.224965175663032e-05, + "loss": 1.109, + "step": 4594 + }, + { + "epoch": 0.27917856491888937, + "grad_norm": 0.24379725754261017, + "learning_rate": 8.224233576304655e-05, + "loss": 1.0798, + "step": 4595 + }, + { + "epoch": 0.2792393219515159, + "grad_norm": 0.22564616799354553, + "learning_rate": 8.223501858761372e-05, + "loss": 1.056, + "step": 4596 + }, + { + "epoch": 0.27930007898414244, + "grad_norm": 0.17423412203788757, + "learning_rate": 8.222770023060006e-05, + "loss": 1.0859, + "step": 4597 + }, + { + "epoch": 0.2793608360167689, + "grad_norm": 0.1881750226020813, + "learning_rate": 8.222038069227383e-05, + "loss": 1.0971, + "step": 4598 + }, + { + "epoch": 0.27942159304939546, + "grad_norm": 0.25505539774894714, + "learning_rate": 8.221305997290329e-05, + "loss": 1.0453, + "step": 4599 + }, + { + "epoch": 0.279482350082022, + "grad_norm": 0.25756052136421204, + "learning_rate": 8.220573807275685e-05, + "loss": 1.1496, + "step": 4600 + }, + { + "epoch": 0.2795431071146485, + "grad_norm": 0.205243319272995, + "learning_rate": 8.219841499210283e-05, + "loss": 1.0764, + "step": 4601 + }, + { + "epoch": 0.27960386414727506, + "grad_norm": 0.19651411473751068, + "learning_rate": 8.219109073120967e-05, + "loss": 1.084, + "step": 4602 + }, + { + "epoch": 0.2796646211799016, + "grad_norm": 0.21760444343090057, + "learning_rate": 8.218376529034589e-05, + "loss": 1.1536, + "step": 4603 + }, + { + "epoch": 0.2797253782125281, + "grad_norm": 0.2487654834985733, + "learning_rate": 8.217643866977995e-05, + "loss": 1.1286, + "step": 4604 + }, + { + "epoch": 0.2797861352451546, + "grad_norm": 0.175584077835083, + "learning_rate": 8.216911086978046e-05, + "loss": 1.1006, + "step": 4605 + }, + { + "epoch": 0.27984689227778115, + "grad_norm": 0.25617557764053345, + "learning_rate": 8.216178189061595e-05, + "loss": 1.2272, + "step": 4606 + }, + { + "epoch": 0.2799076493104077, + "grad_norm": 4.585201263427734, + "learning_rate": 8.215445173255514e-05, + "loss": 1.0963, + "step": 4607 + }, + { + "epoch": 0.2799684063430342, + "grad_norm": 0.2532254457473755, + "learning_rate": 8.214712039586667e-05, + "loss": 1.179, + "step": 4608 + }, + { + "epoch": 0.28002916337566075, + "grad_norm": 0.33015140891075134, + "learning_rate": 8.21397878808193e-05, + "loss": 1.2262, + "step": 4609 + }, + { + "epoch": 0.2800899204082873, + "grad_norm": 0.6041632294654846, + "learning_rate": 8.213245418768179e-05, + "loss": 1.2257, + "step": 4610 + }, + { + "epoch": 0.28015067744091376, + "grad_norm": 0.19492700695991516, + "learning_rate": 8.212511931672296e-05, + "loss": 1.0697, + "step": 4611 + }, + { + "epoch": 0.2802114344735403, + "grad_norm": 1.4801768064498901, + "learning_rate": 8.211778326821166e-05, + "loss": 1.1988, + "step": 4612 + }, + { + "epoch": 0.28027219150616683, + "grad_norm": 0.5917450785636902, + "learning_rate": 8.211044604241681e-05, + "loss": 1.1953, + "step": 4613 + }, + { + "epoch": 0.28033294853879337, + "grad_norm": 0.274695485830307, + "learning_rate": 8.210310763960737e-05, + "loss": 1.15, + "step": 4614 + }, + { + "epoch": 0.2803937055714199, + "grad_norm": 0.699866533279419, + "learning_rate": 8.20957680600523e-05, + "loss": 1.18, + "step": 4615 + }, + { + "epoch": 0.28045446260404644, + "grad_norm": 0.3854803740978241, + "learning_rate": 8.208842730402065e-05, + "loss": 1.1978, + "step": 4616 + }, + { + "epoch": 0.2805152196366729, + "grad_norm": 0.254964679479599, + "learning_rate": 8.208108537178149e-05, + "loss": 1.118, + "step": 4617 + }, + { + "epoch": 0.28057597666929945, + "grad_norm": 0.3284163773059845, + "learning_rate": 8.207374226360396e-05, + "loss": 1.1841, + "step": 4618 + }, + { + "epoch": 0.280636733701926, + "grad_norm": 0.3168507218360901, + "learning_rate": 8.206639797975719e-05, + "loss": 1.0914, + "step": 4619 + }, + { + "epoch": 0.2806974907345525, + "grad_norm": 0.25356167554855347, + "learning_rate": 8.205905252051041e-05, + "loss": 1.0888, + "step": 4620 + }, + { + "epoch": 0.28075824776717906, + "grad_norm": 0.3373039662837982, + "learning_rate": 8.205170588613286e-05, + "loss": 1.1509, + "step": 4621 + }, + { + "epoch": 0.2808190047998056, + "grad_norm": 0.2134825587272644, + "learning_rate": 8.204435807689385e-05, + "loss": 1.0686, + "step": 4622 + }, + { + "epoch": 0.28087976183243213, + "grad_norm": 0.2237076312303543, + "learning_rate": 8.203700909306269e-05, + "loss": 1.0998, + "step": 4623 + }, + { + "epoch": 0.2809405188650586, + "grad_norm": 0.2600153386592865, + "learning_rate": 8.202965893490878e-05, + "loss": 1.0584, + "step": 4624 + }, + { + "epoch": 0.28100127589768514, + "grad_norm": 0.2137063592672348, + "learning_rate": 8.202230760270151e-05, + "loss": 1.0981, + "step": 4625 + }, + { + "epoch": 0.2810620329303117, + "grad_norm": 0.34476083517074585, + "learning_rate": 8.201495509671037e-05, + "loss": 1.0638, + "step": 4626 + }, + { + "epoch": 0.2811227899629382, + "grad_norm": 2.700437545776367, + "learning_rate": 8.200760141720486e-05, + "loss": 1.2365, + "step": 4627 + }, + { + "epoch": 0.28118354699556475, + "grad_norm": 0.22395522892475128, + "learning_rate": 8.200024656445455e-05, + "loss": 1.0571, + "step": 4628 + }, + { + "epoch": 0.2812443040281913, + "grad_norm": 0.24478545784950256, + "learning_rate": 8.1992890538729e-05, + "loss": 1.1215, + "step": 4629 + }, + { + "epoch": 0.28130506106081776, + "grad_norm": 0.5497540235519409, + "learning_rate": 8.198553334029786e-05, + "loss": 1.0706, + "step": 4630 + }, + { + "epoch": 0.2813658180934443, + "grad_norm": 0.3321070969104767, + "learning_rate": 8.197817496943084e-05, + "loss": 1.0748, + "step": 4631 + }, + { + "epoch": 0.28142657512607083, + "grad_norm": 0.20654618740081787, + "learning_rate": 8.197081542639762e-05, + "loss": 1.0401, + "step": 4632 + }, + { + "epoch": 0.28148733215869737, + "grad_norm": 0.2842791974544525, + "learning_rate": 8.196345471146797e-05, + "loss": 1.1375, + "step": 4633 + }, + { + "epoch": 0.2815480891913239, + "grad_norm": 0.3148439824581146, + "learning_rate": 8.195609282491172e-05, + "loss": 1.0667, + "step": 4634 + }, + { + "epoch": 0.28160884622395044, + "grad_norm": 0.4401419460773468, + "learning_rate": 8.194872976699871e-05, + "loss": 1.169, + "step": 4635 + }, + { + "epoch": 0.281669603256577, + "grad_norm": 0.3144529163837433, + "learning_rate": 8.194136553799883e-05, + "loss": 1.0769, + "step": 4636 + }, + { + "epoch": 0.28173036028920345, + "grad_norm": 0.33955028653144836, + "learning_rate": 8.193400013818202e-05, + "loss": 1.2141, + "step": 4637 + }, + { + "epoch": 0.28179111732183, + "grad_norm": 0.27602940797805786, + "learning_rate": 8.192663356781827e-05, + "loss": 1.2168, + "step": 4638 + }, + { + "epoch": 0.2818518743544565, + "grad_norm": 0.24206721782684326, + "learning_rate": 8.19192658271776e-05, + "loss": 1.1036, + "step": 4639 + }, + { + "epoch": 0.28191263138708306, + "grad_norm": 0.28041771054267883, + "learning_rate": 8.191189691653006e-05, + "loss": 1.083, + "step": 4640 + }, + { + "epoch": 0.2819733884197096, + "grad_norm": 0.2534532845020294, + "learning_rate": 8.190452683614577e-05, + "loss": 1.1309, + "step": 4641 + }, + { + "epoch": 0.2820341454523361, + "grad_norm": 1.6230865716934204, + "learning_rate": 8.18971555862949e-05, + "loss": 1.0938, + "step": 4642 + }, + { + "epoch": 0.28209490248496266, + "grad_norm": 0.33333122730255127, + "learning_rate": 8.188978316724761e-05, + "loss": 1.1001, + "step": 4643 + }, + { + "epoch": 0.28215565951758914, + "grad_norm": 0.24876262247562408, + "learning_rate": 8.188240957927416e-05, + "loss": 1.21, + "step": 4644 + }, + { + "epoch": 0.2822164165502157, + "grad_norm": 0.19993427395820618, + "learning_rate": 8.187503482264482e-05, + "loss": 1.3678, + "step": 4645 + }, + { + "epoch": 0.2822771735828422, + "grad_norm": 0.2232554405927658, + "learning_rate": 8.18676588976299e-05, + "loss": 1.169, + "step": 4646 + }, + { + "epoch": 0.28233793061546875, + "grad_norm": 0.4347221255302429, + "learning_rate": 8.186028180449981e-05, + "loss": 1.0744, + "step": 4647 + }, + { + "epoch": 0.2823986876480953, + "grad_norm": 0.6534870266914368, + "learning_rate": 8.185290354352493e-05, + "loss": 1.3305, + "step": 4648 + }, + { + "epoch": 0.2824594446807218, + "grad_norm": 0.3760615289211273, + "learning_rate": 8.18455241149757e-05, + "loss": 1.1328, + "step": 4649 + }, + { + "epoch": 0.2825202017133483, + "grad_norm": 0.4175221621990204, + "learning_rate": 8.183814351912265e-05, + "loss": 1.1343, + "step": 4650 + }, + { + "epoch": 0.28258095874597483, + "grad_norm": 5.1945953369140625, + "learning_rate": 8.183076175623628e-05, + "loss": 1.2592, + "step": 4651 + }, + { + "epoch": 0.28264171577860137, + "grad_norm": 0.8030882477760315, + "learning_rate": 8.182337882658717e-05, + "loss": 1.1369, + "step": 4652 + }, + { + "epoch": 0.2827024728112279, + "grad_norm": 0.29600220918655396, + "learning_rate": 8.181599473044598e-05, + "loss": 1.2705, + "step": 4653 + }, + { + "epoch": 0.28276322984385444, + "grad_norm": 1.0098615884780884, + "learning_rate": 8.180860946808336e-05, + "loss": 1.5526, + "step": 4654 + }, + { + "epoch": 0.28282398687648097, + "grad_norm": 0.6748232841491699, + "learning_rate": 8.180122303976999e-05, + "loss": 1.184, + "step": 4655 + }, + { + "epoch": 0.2828847439091075, + "grad_norm": 0.7741841077804565, + "learning_rate": 8.179383544577664e-05, + "loss": 1.2726, + "step": 4656 + }, + { + "epoch": 0.282945500941734, + "grad_norm": 0.25124672055244446, + "learning_rate": 8.178644668637412e-05, + "loss": 1.0615, + "step": 4657 + }, + { + "epoch": 0.2830062579743605, + "grad_norm": 0.3545131981372833, + "learning_rate": 8.177905676183325e-05, + "loss": 1.1536, + "step": 4658 + }, + { + "epoch": 0.28306701500698705, + "grad_norm": 0.31539592146873474, + "learning_rate": 8.177166567242491e-05, + "loss": 1.1575, + "step": 4659 + }, + { + "epoch": 0.2831277720396136, + "grad_norm": 0.4407671391963959, + "learning_rate": 8.176427341842001e-05, + "loss": 1.1529, + "step": 4660 + }, + { + "epoch": 0.2831885290722401, + "grad_norm": 0.4772582948207855, + "learning_rate": 8.175688000008956e-05, + "loss": 1.1145, + "step": 4661 + }, + { + "epoch": 0.28324928610486666, + "grad_norm": 0.29201602935791016, + "learning_rate": 8.17494854177045e-05, + "loss": 1.1758, + "step": 4662 + }, + { + "epoch": 0.28331004313749314, + "grad_norm": 0.412558376789093, + "learning_rate": 8.174208967153593e-05, + "loss": 1.3043, + "step": 4663 + }, + { + "epoch": 0.2833708001701197, + "grad_norm": 0.27653542160987854, + "learning_rate": 8.173469276185495e-05, + "loss": 1.1261, + "step": 4664 + }, + { + "epoch": 0.2834315572027462, + "grad_norm": 0.523465096950531, + "learning_rate": 8.172729468893264e-05, + "loss": 1.1248, + "step": 4665 + }, + { + "epoch": 0.28349231423537274, + "grad_norm": 0.4232143759727478, + "learning_rate": 8.171989545304022e-05, + "loss": 1.3377, + "step": 4666 + }, + { + "epoch": 0.2835530712679993, + "grad_norm": 0.265605628490448, + "learning_rate": 8.171249505444891e-05, + "loss": 1.2004, + "step": 4667 + }, + { + "epoch": 0.2836138283006258, + "grad_norm": 0.2792421281337738, + "learning_rate": 8.170509349342996e-05, + "loss": 1.0819, + "step": 4668 + }, + { + "epoch": 0.28367458533325235, + "grad_norm": 0.46600598096847534, + "learning_rate": 8.169769077025466e-05, + "loss": 1.0589, + "step": 4669 + }, + { + "epoch": 0.28373534236587883, + "grad_norm": 0.3838799297809601, + "learning_rate": 8.16902868851944e-05, + "loss": 1.1078, + "step": 4670 + }, + { + "epoch": 0.28379609939850536, + "grad_norm": 0.48313459753990173, + "learning_rate": 8.168288183852055e-05, + "loss": 1.0562, + "step": 4671 + }, + { + "epoch": 0.2838568564311319, + "grad_norm": 0.574755072593689, + "learning_rate": 8.167547563050452e-05, + "loss": 1.2856, + "step": 4672 + }, + { + "epoch": 0.28391761346375843, + "grad_norm": 0.7972866892814636, + "learning_rate": 8.166806826141784e-05, + "loss": 1.0864, + "step": 4673 + }, + { + "epoch": 0.28397837049638497, + "grad_norm": 0.25257012248039246, + "learning_rate": 8.166065973153199e-05, + "loss": 1.1356, + "step": 4674 + }, + { + "epoch": 0.2840391275290115, + "grad_norm": 0.28807583451271057, + "learning_rate": 8.165325004111852e-05, + "loss": 1.1442, + "step": 4675 + }, + { + "epoch": 0.284099884561638, + "grad_norm": 0.5628198385238647, + "learning_rate": 8.164583919044906e-05, + "loss": 1.0737, + "step": 4676 + }, + { + "epoch": 0.2841606415942645, + "grad_norm": 0.1939159780740738, + "learning_rate": 8.163842717979525e-05, + "loss": 1.0827, + "step": 4677 + }, + { + "epoch": 0.28422139862689105, + "grad_norm": 0.6005893349647522, + "learning_rate": 8.163101400942877e-05, + "loss": 1.1081, + "step": 4678 + }, + { + "epoch": 0.2842821556595176, + "grad_norm": 0.23524098098278046, + "learning_rate": 8.162359967962134e-05, + "loss": 1.116, + "step": 4679 + }, + { + "epoch": 0.2843429126921441, + "grad_norm": 0.5827543139457703, + "learning_rate": 8.161618419064478e-05, + "loss": 1.1405, + "step": 4680 + }, + { + "epoch": 0.28440366972477066, + "grad_norm": 0.2714623808860779, + "learning_rate": 8.160876754277087e-05, + "loss": 1.0937, + "step": 4681 + }, + { + "epoch": 0.2844644267573972, + "grad_norm": 0.5354325175285339, + "learning_rate": 8.160134973627147e-05, + "loss": 1.1466, + "step": 4682 + }, + { + "epoch": 0.2845251837900237, + "grad_norm": 0.25990355014801025, + "learning_rate": 8.159393077141849e-05, + "loss": 1.0867, + "step": 4683 + }, + { + "epoch": 0.2845859408226502, + "grad_norm": 0.21582327783107758, + "learning_rate": 8.158651064848387e-05, + "loss": 1.2649, + "step": 4684 + }, + { + "epoch": 0.28464669785527674, + "grad_norm": 0.32816874980926514, + "learning_rate": 8.157908936773961e-05, + "loss": 1.0955, + "step": 4685 + }, + { + "epoch": 0.2847074548879033, + "grad_norm": 0.2591687738895416, + "learning_rate": 8.15716669294577e-05, + "loss": 1.1754, + "step": 4686 + }, + { + "epoch": 0.2847682119205298, + "grad_norm": 0.2678835093975067, + "learning_rate": 8.156424333391025e-05, + "loss": 1.1982, + "step": 4687 + }, + { + "epoch": 0.28482896895315635, + "grad_norm": 0.1716891974210739, + "learning_rate": 8.155681858136937e-05, + "loss": 1.0233, + "step": 4688 + }, + { + "epoch": 0.2848897259857828, + "grad_norm": 0.16800156235694885, + "learning_rate": 8.154939267210719e-05, + "loss": 1.046, + "step": 4689 + }, + { + "epoch": 0.28495048301840936, + "grad_norm": 0.27414199709892273, + "learning_rate": 8.154196560639592e-05, + "loss": 1.1261, + "step": 4690 + }, + { + "epoch": 0.2850112400510359, + "grad_norm": 0.17501957714557648, + "learning_rate": 8.153453738450782e-05, + "loss": 1.0777, + "step": 4691 + }, + { + "epoch": 0.28507199708366243, + "grad_norm": 0.3050796091556549, + "learning_rate": 8.152710800671515e-05, + "loss": 1.1229, + "step": 4692 + }, + { + "epoch": 0.28513275411628897, + "grad_norm": 0.22571823000907898, + "learning_rate": 8.151967747329024e-05, + "loss": 1.1594, + "step": 4693 + }, + { + "epoch": 0.2851935111489155, + "grad_norm": 0.478141725063324, + "learning_rate": 8.151224578450545e-05, + "loss": 1.1875, + "step": 4694 + }, + { + "epoch": 0.28525426818154204, + "grad_norm": 0.35870644450187683, + "learning_rate": 8.150481294063321e-05, + "loss": 1.0978, + "step": 4695 + }, + { + "epoch": 0.2853150252141685, + "grad_norm": 0.2505946457386017, + "learning_rate": 8.149737894194596e-05, + "loss": 1.0811, + "step": 4696 + }, + { + "epoch": 0.28537578224679505, + "grad_norm": 1.882879376411438, + "learning_rate": 8.14899437887162e-05, + "loss": 1.1895, + "step": 4697 + }, + { + "epoch": 0.2854365392794216, + "grad_norm": 0.5569814443588257, + "learning_rate": 8.148250748121646e-05, + "loss": 1.1473, + "step": 4698 + }, + { + "epoch": 0.2854972963120481, + "grad_norm": 0.4845321476459503, + "learning_rate": 8.147507001971933e-05, + "loss": 1.2276, + "step": 4699 + }, + { + "epoch": 0.28555805334467466, + "grad_norm": 0.45004796981811523, + "learning_rate": 8.14676314044974e-05, + "loss": 1.1053, + "step": 4700 + }, + { + "epoch": 0.2856188103773012, + "grad_norm": 0.27117130160331726, + "learning_rate": 8.146019163582339e-05, + "loss": 1.1783, + "step": 4701 + }, + { + "epoch": 0.2856795674099277, + "grad_norm": 0.38146039843559265, + "learning_rate": 8.145275071396996e-05, + "loss": 1.1832, + "step": 4702 + }, + { + "epoch": 0.2857403244425542, + "grad_norm": 0.2563825249671936, + "learning_rate": 8.144530863920987e-05, + "loss": 1.2185, + "step": 4703 + }, + { + "epoch": 0.28580108147518074, + "grad_norm": 0.3222849667072296, + "learning_rate": 8.143786541181591e-05, + "loss": 1.0424, + "step": 4704 + }, + { + "epoch": 0.2858618385078073, + "grad_norm": 0.20929233729839325, + "learning_rate": 8.143042103206094e-05, + "loss": 1.015, + "step": 4705 + }, + { + "epoch": 0.2859225955404338, + "grad_norm": 0.18352508544921875, + "learning_rate": 8.142297550021777e-05, + "loss": 1.0557, + "step": 4706 + }, + { + "epoch": 0.28598335257306035, + "grad_norm": 0.3157120645046234, + "learning_rate": 8.14155288165594e-05, + "loss": 1.0619, + "step": 4707 + }, + { + "epoch": 0.2860441096056869, + "grad_norm": 0.1993403434753418, + "learning_rate": 8.140808098135872e-05, + "loss": 1.0674, + "step": 4708 + }, + { + "epoch": 0.28610486663831336, + "grad_norm": 0.18266800045967102, + "learning_rate": 8.140063199488878e-05, + "loss": 1.0829, + "step": 4709 + }, + { + "epoch": 0.2861656236709399, + "grad_norm": 0.32904791831970215, + "learning_rate": 8.139318185742259e-05, + "loss": 1.2158, + "step": 4710 + }, + { + "epoch": 0.28622638070356643, + "grad_norm": 0.3400256335735321, + "learning_rate": 8.138573056923328e-05, + "loss": 1.1389, + "step": 4711 + }, + { + "epoch": 0.28628713773619296, + "grad_norm": 0.767938494682312, + "learning_rate": 8.137827813059392e-05, + "loss": 1.1, + "step": 4712 + }, + { + "epoch": 0.2863478947688195, + "grad_norm": 0.2562240958213806, + "learning_rate": 8.137082454177771e-05, + "loss": 1.2268, + "step": 4713 + }, + { + "epoch": 0.28640865180144603, + "grad_norm": 0.3525850176811218, + "learning_rate": 8.136336980305787e-05, + "loss": 1.0524, + "step": 4714 + }, + { + "epoch": 0.28646940883407257, + "grad_norm": 0.35473915934562683, + "learning_rate": 8.135591391470766e-05, + "loss": 1.224, + "step": 4715 + }, + { + "epoch": 0.28653016586669905, + "grad_norm": 0.2544901967048645, + "learning_rate": 8.134845687700035e-05, + "loss": 1.1479, + "step": 4716 + }, + { + "epoch": 0.2865909228993256, + "grad_norm": 0.26960378885269165, + "learning_rate": 8.13409986902093e-05, + "loss": 1.0741, + "step": 4717 + }, + { + "epoch": 0.2866516799319521, + "grad_norm": 0.25985801219940186, + "learning_rate": 8.133353935460788e-05, + "loss": 1.1539, + "step": 4718 + }, + { + "epoch": 0.28671243696457865, + "grad_norm": 0.26099732518196106, + "learning_rate": 8.132607887046952e-05, + "loss": 1.134, + "step": 4719 + }, + { + "epoch": 0.2867731939972052, + "grad_norm": 0.24528297781944275, + "learning_rate": 8.131861723806769e-05, + "loss": 1.2319, + "step": 4720 + }, + { + "epoch": 0.2868339510298317, + "grad_norm": 0.24983979761600494, + "learning_rate": 8.131115445767588e-05, + "loss": 1.0634, + "step": 4721 + }, + { + "epoch": 0.2868947080624582, + "grad_norm": 0.40290379524230957, + "learning_rate": 8.130369052956765e-05, + "loss": 1.0965, + "step": 4722 + }, + { + "epoch": 0.28695546509508474, + "grad_norm": 0.353474885225296, + "learning_rate": 8.129622545401661e-05, + "loss": 1.0558, + "step": 4723 + }, + { + "epoch": 0.2870162221277113, + "grad_norm": 0.2520107328891754, + "learning_rate": 8.128875923129635e-05, + "loss": 1.0567, + "step": 4724 + }, + { + "epoch": 0.2870769791603378, + "grad_norm": 1.327829122543335, + "learning_rate": 8.12812918616806e-05, + "loss": 1.0893, + "step": 4725 + }, + { + "epoch": 0.28713773619296434, + "grad_norm": 6.429571151733398, + "learning_rate": 8.127382334544302e-05, + "loss": 1.0596, + "step": 4726 + }, + { + "epoch": 0.2871984932255909, + "grad_norm": 0.4918801784515381, + "learning_rate": 8.126635368285742e-05, + "loss": 1.2275, + "step": 4727 + }, + { + "epoch": 0.2872592502582174, + "grad_norm": 0.3049384355545044, + "learning_rate": 8.125888287419759e-05, + "loss": 1.1507, + "step": 4728 + }, + { + "epoch": 0.2873200072908439, + "grad_norm": 0.22929777204990387, + "learning_rate": 8.125141091973736e-05, + "loss": 1.2415, + "step": 4729 + }, + { + "epoch": 0.28738076432347043, + "grad_norm": 0.3162703514099121, + "learning_rate": 8.124393781975062e-05, + "loss": 1.3832, + "step": 4730 + }, + { + "epoch": 0.28744152135609696, + "grad_norm": 0.26970574259757996, + "learning_rate": 8.123646357451129e-05, + "loss": 1.0367, + "step": 4731 + }, + { + "epoch": 0.2875022783887235, + "grad_norm": 0.4018248915672302, + "learning_rate": 8.122898818429336e-05, + "loss": 1.2973, + "step": 4732 + }, + { + "epoch": 0.28756303542135003, + "grad_norm": 0.5206171870231628, + "learning_rate": 8.122151164937083e-05, + "loss": 1.2566, + "step": 4733 + }, + { + "epoch": 0.28762379245397657, + "grad_norm": 0.35473230481147766, + "learning_rate": 8.121403397001774e-05, + "loss": 1.0765, + "step": 4734 + }, + { + "epoch": 0.28768454948660305, + "grad_norm": 0.24050723016262054, + "learning_rate": 8.120655514650824e-05, + "loss": 1.0765, + "step": 4735 + }, + { + "epoch": 0.2877453065192296, + "grad_norm": 0.3031902611255646, + "learning_rate": 8.11990751791164e-05, + "loss": 1.134, + "step": 4736 + }, + { + "epoch": 0.2878060635518561, + "grad_norm": 0.5883634090423584, + "learning_rate": 8.119159406811643e-05, + "loss": 1.081, + "step": 4737 + }, + { + "epoch": 0.28786682058448265, + "grad_norm": 0.5956151485443115, + "learning_rate": 8.118411181378256e-05, + "loss": 1.1048, + "step": 4738 + }, + { + "epoch": 0.2879275776171092, + "grad_norm": 1.155529499053955, + "learning_rate": 8.117662841638903e-05, + "loss": 1.3776, + "step": 4739 + }, + { + "epoch": 0.2879883346497357, + "grad_norm": 0.23624932765960693, + "learning_rate": 8.116914387621018e-05, + "loss": 1.1357, + "step": 4740 + }, + { + "epoch": 0.28804909168236226, + "grad_norm": 0.3633232116699219, + "learning_rate": 8.116165819352031e-05, + "loss": 1.1071, + "step": 4741 + }, + { + "epoch": 0.28810984871498874, + "grad_norm": 0.26505330204963684, + "learning_rate": 8.115417136859385e-05, + "loss": 1.1852, + "step": 4742 + }, + { + "epoch": 0.28817060574761527, + "grad_norm": 0.24976864457130432, + "learning_rate": 8.114668340170522e-05, + "loss": 1.0823, + "step": 4743 + }, + { + "epoch": 0.2882313627802418, + "grad_norm": 0.3006606698036194, + "learning_rate": 8.113919429312888e-05, + "loss": 1.1433, + "step": 4744 + }, + { + "epoch": 0.28829211981286834, + "grad_norm": 0.19041763246059418, + "learning_rate": 8.113170404313937e-05, + "loss": 1.0594, + "step": 4745 + }, + { + "epoch": 0.2883528768454949, + "grad_norm": 0.5029729008674622, + "learning_rate": 8.112421265201121e-05, + "loss": 1.1594, + "step": 4746 + }, + { + "epoch": 0.2884136338781214, + "grad_norm": 0.37129664421081543, + "learning_rate": 8.111672012001904e-05, + "loss": 1.2629, + "step": 4747 + }, + { + "epoch": 0.28847439091074795, + "grad_norm": 0.3520868122577667, + "learning_rate": 8.110922644743746e-05, + "loss": 1.1053, + "step": 4748 + }, + { + "epoch": 0.2885351479433744, + "grad_norm": 0.30296894907951355, + "learning_rate": 8.110173163454118e-05, + "loss": 1.1553, + "step": 4749 + }, + { + "epoch": 0.28859590497600096, + "grad_norm": 0.4728729724884033, + "learning_rate": 8.109423568160491e-05, + "loss": 1.0998, + "step": 4750 + }, + { + "epoch": 0.2886566620086275, + "grad_norm": 0.34350451827049255, + "learning_rate": 8.108673858890342e-05, + "loss": 1.1108, + "step": 4751 + }, + { + "epoch": 0.28871741904125403, + "grad_norm": 0.2862670123577118, + "learning_rate": 8.107924035671153e-05, + "loss": 1.1896, + "step": 4752 + }, + { + "epoch": 0.28877817607388057, + "grad_norm": 0.2585833668708801, + "learning_rate": 8.107174098530406e-05, + "loss": 1.1275, + "step": 4753 + }, + { + "epoch": 0.2888389331065071, + "grad_norm": 0.4934181571006775, + "learning_rate": 8.106424047495593e-05, + "loss": 1.1856, + "step": 4754 + }, + { + "epoch": 0.2888996901391336, + "grad_norm": 0.17318494617938995, + "learning_rate": 8.105673882594207e-05, + "loss": 1.0259, + "step": 4755 + }, + { + "epoch": 0.2889604471717601, + "grad_norm": 0.21970781683921814, + "learning_rate": 8.104923603853743e-05, + "loss": 1.0783, + "step": 4756 + }, + { + "epoch": 0.28902120420438665, + "grad_norm": 0.2062210589647293, + "learning_rate": 8.104173211301705e-05, + "loss": 1.0786, + "step": 4757 + }, + { + "epoch": 0.2890819612370132, + "grad_norm": 0.47149160504341125, + "learning_rate": 8.103422704965597e-05, + "loss": 1.0316, + "step": 4758 + }, + { + "epoch": 0.2891427182696397, + "grad_norm": 0.35853391885757446, + "learning_rate": 8.10267208487293e-05, + "loss": 1.3533, + "step": 4759 + }, + { + "epoch": 0.28920347530226626, + "grad_norm": 0.5158296823501587, + "learning_rate": 8.10192135105122e-05, + "loss": 1.3309, + "step": 4760 + }, + { + "epoch": 0.2892642323348928, + "grad_norm": 0.2478027045726776, + "learning_rate": 8.10117050352798e-05, + "loss": 1.0868, + "step": 4761 + }, + { + "epoch": 0.28932498936751927, + "grad_norm": 0.6252611875534058, + "learning_rate": 8.100419542330738e-05, + "loss": 1.067, + "step": 4762 + }, + { + "epoch": 0.2893857464001458, + "grad_norm": 0.30082762241363525, + "learning_rate": 8.099668467487018e-05, + "loss": 1.058, + "step": 4763 + }, + { + "epoch": 0.28944650343277234, + "grad_norm": 0.2391790747642517, + "learning_rate": 8.09891727902435e-05, + "loss": 1.1532, + "step": 4764 + }, + { + "epoch": 0.2895072604653989, + "grad_norm": 0.17011988162994385, + "learning_rate": 8.098165976970273e-05, + "loss": 1.0567, + "step": 4765 + }, + { + "epoch": 0.2895680174980254, + "grad_norm": 0.42123278975486755, + "learning_rate": 8.097414561352321e-05, + "loss": 1.0341, + "step": 4766 + }, + { + "epoch": 0.28962877453065194, + "grad_norm": 0.26598915457725525, + "learning_rate": 8.09666303219804e-05, + "loss": 1.1164, + "step": 4767 + }, + { + "epoch": 0.2896895315632784, + "grad_norm": 0.2022605985403061, + "learning_rate": 8.095911389534977e-05, + "loss": 1.0845, + "step": 4768 + }, + { + "epoch": 0.28975028859590496, + "grad_norm": 0.29814863204956055, + "learning_rate": 8.095159633390683e-05, + "loss": 1.0849, + "step": 4769 + }, + { + "epoch": 0.2898110456285315, + "grad_norm": 0.19839178025722504, + "learning_rate": 8.094407763792714e-05, + "loss": 1.0847, + "step": 4770 + }, + { + "epoch": 0.28987180266115803, + "grad_norm": 0.26245805621147156, + "learning_rate": 8.093655780768632e-05, + "loss": 1.1953, + "step": 4771 + }, + { + "epoch": 0.28993255969378456, + "grad_norm": 0.19956426322460175, + "learning_rate": 8.092903684345999e-05, + "loss": 1.0852, + "step": 4772 + }, + { + "epoch": 0.2899933167264111, + "grad_norm": 0.36077210307121277, + "learning_rate": 8.092151474552381e-05, + "loss": 1.0625, + "step": 4773 + }, + { + "epoch": 0.29005407375903763, + "grad_norm": 0.2334013432264328, + "learning_rate": 8.091399151415355e-05, + "loss": 1.0596, + "step": 4774 + }, + { + "epoch": 0.2901148307916641, + "grad_norm": 0.20021478831768036, + "learning_rate": 8.090646714962497e-05, + "loss": 1.0568, + "step": 4775 + }, + { + "epoch": 0.29017558782429065, + "grad_norm": 0.395464152097702, + "learning_rate": 8.089894165221386e-05, + "loss": 1.0981, + "step": 4776 + }, + { + "epoch": 0.2902363448569172, + "grad_norm": 0.23848798871040344, + "learning_rate": 8.089141502219603e-05, + "loss": 1.0483, + "step": 4777 + }, + { + "epoch": 0.2902971018895437, + "grad_norm": 0.15365374088287354, + "learning_rate": 8.088388725984745e-05, + "loss": 1.049, + "step": 4778 + }, + { + "epoch": 0.29035785892217025, + "grad_norm": 7.788456916809082, + "learning_rate": 8.0876358365444e-05, + "loss": 1.1027, + "step": 4779 + }, + { + "epoch": 0.2904186159547968, + "grad_norm": 0.30950480699539185, + "learning_rate": 8.086882833926168e-05, + "loss": 1.153, + "step": 4780 + }, + { + "epoch": 0.29047937298742327, + "grad_norm": 0.3053435683250427, + "learning_rate": 8.086129718157648e-05, + "loss": 1.1653, + "step": 4781 + }, + { + "epoch": 0.2905401300200498, + "grad_norm": 3.6193840503692627, + "learning_rate": 8.085376489266447e-05, + "loss": 1.0252, + "step": 4782 + }, + { + "epoch": 0.29060088705267634, + "grad_norm": 0.4975748658180237, + "learning_rate": 8.084623147280173e-05, + "loss": 1.2016, + "step": 4783 + }, + { + "epoch": 0.2906616440853029, + "grad_norm": 0.31479156017303467, + "learning_rate": 8.083869692226442e-05, + "loss": 1.1583, + "step": 4784 + }, + { + "epoch": 0.2907224011179294, + "grad_norm": 0.24141883850097656, + "learning_rate": 8.083116124132873e-05, + "loss": 1.1741, + "step": 4785 + }, + { + "epoch": 0.29078315815055594, + "grad_norm": 0.43555358052253723, + "learning_rate": 8.082362443027085e-05, + "loss": 1.1195, + "step": 4786 + }, + { + "epoch": 0.2908439151831825, + "grad_norm": 0.24714143574237823, + "learning_rate": 8.081608648936707e-05, + "loss": 1.2388, + "step": 4787 + }, + { + "epoch": 0.29090467221580896, + "grad_norm": 1.014924168586731, + "learning_rate": 8.080854741889367e-05, + "loss": 1.1174, + "step": 4788 + }, + { + "epoch": 0.2909654292484355, + "grad_norm": 0.4160824716091156, + "learning_rate": 8.080100721912702e-05, + "loss": 1.0306, + "step": 4789 + }, + { + "epoch": 0.291026186281062, + "grad_norm": 0.24467824399471283, + "learning_rate": 8.079346589034347e-05, + "loss": 1.0815, + "step": 4790 + }, + { + "epoch": 0.29108694331368856, + "grad_norm": 0.3311941921710968, + "learning_rate": 8.07859234328195e-05, + "loss": 1.0689, + "step": 4791 + }, + { + "epoch": 0.2911477003463151, + "grad_norm": 0.393278032541275, + "learning_rate": 8.077837984683155e-05, + "loss": 1.0623, + "step": 4792 + }, + { + "epoch": 0.29120845737894163, + "grad_norm": 0.2150166928768158, + "learning_rate": 8.077083513265613e-05, + "loss": 1.1447, + "step": 4793 + }, + { + "epoch": 0.2912692144115681, + "grad_norm": 0.4615291953086853, + "learning_rate": 8.076328929056982e-05, + "loss": 1.1243, + "step": 4794 + }, + { + "epoch": 0.29132997144419465, + "grad_norm": 0.46372532844543457, + "learning_rate": 8.075574232084918e-05, + "loss": 1.0755, + "step": 4795 + }, + { + "epoch": 0.2913907284768212, + "grad_norm": 0.26397913694381714, + "learning_rate": 8.074819422377086e-05, + "loss": 1.1276, + "step": 4796 + }, + { + "epoch": 0.2914514855094477, + "grad_norm": 1.4219920635223389, + "learning_rate": 8.074064499961153e-05, + "loss": 1.0795, + "step": 4797 + }, + { + "epoch": 0.29151224254207425, + "grad_norm": 2.0002365112304688, + "learning_rate": 8.073309464864794e-05, + "loss": 1.2136, + "step": 4798 + }, + { + "epoch": 0.2915729995747008, + "grad_norm": 0.6222421526908875, + "learning_rate": 8.072554317115681e-05, + "loss": 1.0664, + "step": 4799 + }, + { + "epoch": 0.2916337566073273, + "grad_norm": 1.6377849578857422, + "learning_rate": 8.071799056741495e-05, + "loss": 1.1107, + "step": 4800 + }, + { + "epoch": 0.2916945136399538, + "grad_norm": 2.1757493019104004, + "learning_rate": 8.071043683769922e-05, + "loss": 1.1095, + "step": 4801 + }, + { + "epoch": 0.29175527067258034, + "grad_norm": 0.5298777222633362, + "learning_rate": 8.070288198228648e-05, + "loss": 1.1174, + "step": 4802 + }, + { + "epoch": 0.29181602770520687, + "grad_norm": 0.25302112102508545, + "learning_rate": 8.069532600145368e-05, + "loss": 1.2088, + "step": 4803 + }, + { + "epoch": 0.2918767847378334, + "grad_norm": 0.4449176490306854, + "learning_rate": 8.068776889547777e-05, + "loss": 1.1457, + "step": 4804 + }, + { + "epoch": 0.29193754177045994, + "grad_norm": 0.3400658071041107, + "learning_rate": 8.068021066463575e-05, + "loss": 1.0406, + "step": 4805 + }, + { + "epoch": 0.2919982988030865, + "grad_norm": 0.27357742190361023, + "learning_rate": 8.067265130920468e-05, + "loss": 1.0776, + "step": 4806 + }, + { + "epoch": 0.292059055835713, + "grad_norm": 0.7307799458503723, + "learning_rate": 8.066509082946166e-05, + "loss": 1.1328, + "step": 4807 + }, + { + "epoch": 0.2921198128683395, + "grad_norm": 0.2022358775138855, + "learning_rate": 8.06575292256838e-05, + "loss": 1.0974, + "step": 4808 + }, + { + "epoch": 0.292180569900966, + "grad_norm": 0.27088725566864014, + "learning_rate": 8.064996649814827e-05, + "loss": 1.1611, + "step": 4809 + }, + { + "epoch": 0.29224132693359256, + "grad_norm": 0.284970223903656, + "learning_rate": 8.06424026471323e-05, + "loss": 1.0632, + "step": 4810 + }, + { + "epoch": 0.2923020839662191, + "grad_norm": 0.1866750419139862, + "learning_rate": 8.063483767291313e-05, + "loss": 1.1064, + "step": 4811 + }, + { + "epoch": 0.29236284099884563, + "grad_norm": 0.2803458571434021, + "learning_rate": 8.062727157576808e-05, + "loss": 1.1063, + "step": 4812 + }, + { + "epoch": 0.29242359803147217, + "grad_norm": 0.37271830439567566, + "learning_rate": 8.061970435597447e-05, + "loss": 1.1023, + "step": 4813 + }, + { + "epoch": 0.29248435506409864, + "grad_norm": 0.20964722335338593, + "learning_rate": 8.061213601380966e-05, + "loss": 1.0884, + "step": 4814 + }, + { + "epoch": 0.2925451120967252, + "grad_norm": 0.16280953586101532, + "learning_rate": 8.06045665495511e-05, + "loss": 1.0167, + "step": 4815 + }, + { + "epoch": 0.2926058691293517, + "grad_norm": 0.29231753945350647, + "learning_rate": 8.059699596347625e-05, + "loss": 1.2014, + "step": 4816 + }, + { + "epoch": 0.29266662616197825, + "grad_norm": 0.6116070747375488, + "learning_rate": 8.058942425586258e-05, + "loss": 1.0969, + "step": 4817 + }, + { + "epoch": 0.2927273831946048, + "grad_norm": 0.16750992834568024, + "learning_rate": 8.058185142698765e-05, + "loss": 1.0697, + "step": 4818 + }, + { + "epoch": 0.2927881402272313, + "grad_norm": 0.21943433582782745, + "learning_rate": 8.057427747712905e-05, + "loss": 1.1011, + "step": 4819 + }, + { + "epoch": 0.29284889725985785, + "grad_norm": 0.23373888432979584, + "learning_rate": 8.05667024065644e-05, + "loss": 1.0951, + "step": 4820 + }, + { + "epoch": 0.29290965429248433, + "grad_norm": 0.2343854010105133, + "learning_rate": 8.055912621557136e-05, + "loss": 1.1453, + "step": 4821 + }, + { + "epoch": 0.29297041132511087, + "grad_norm": 0.17664922773838043, + "learning_rate": 8.055154890442766e-05, + "loss": 1.0206, + "step": 4822 + }, + { + "epoch": 0.2930311683577374, + "grad_norm": 0.27996671199798584, + "learning_rate": 8.0543970473411e-05, + "loss": 1.0157, + "step": 4823 + }, + { + "epoch": 0.29309192539036394, + "grad_norm": 0.3980349600315094, + "learning_rate": 8.053639092279922e-05, + "loss": 1.1145, + "step": 4824 + }, + { + "epoch": 0.2931526824229905, + "grad_norm": 0.21316516399383545, + "learning_rate": 8.052881025287014e-05, + "loss": 1.0753, + "step": 4825 + }, + { + "epoch": 0.293213439455617, + "grad_norm": 0.17730183899402618, + "learning_rate": 8.052122846390159e-05, + "loss": 1.0559, + "step": 4826 + }, + { + "epoch": 0.2932741964882435, + "grad_norm": 0.2058178037405014, + "learning_rate": 8.051364555617151e-05, + "loss": 1.122, + "step": 4827 + }, + { + "epoch": 0.29333495352087, + "grad_norm": 0.2155120074748993, + "learning_rate": 8.050606152995788e-05, + "loss": 1.1714, + "step": 4828 + }, + { + "epoch": 0.29339571055349656, + "grad_norm": 0.33687013387680054, + "learning_rate": 8.049847638553864e-05, + "loss": 1.0339, + "step": 4829 + }, + { + "epoch": 0.2934564675861231, + "grad_norm": 0.2360881119966507, + "learning_rate": 8.049089012319188e-05, + "loss": 1.1989, + "step": 4830 + }, + { + "epoch": 0.29351722461874963, + "grad_norm": 0.30361685156822205, + "learning_rate": 8.048330274319562e-05, + "loss": 1.1157, + "step": 4831 + }, + { + "epoch": 0.29357798165137616, + "grad_norm": 0.17342445254325867, + "learning_rate": 8.047571424582803e-05, + "loss": 1.0857, + "step": 4832 + }, + { + "epoch": 0.2936387386840027, + "grad_norm": 0.21961171925067902, + "learning_rate": 8.046812463136724e-05, + "loss": 1.0504, + "step": 4833 + }, + { + "epoch": 0.2936994957166292, + "grad_norm": 0.26273876428604126, + "learning_rate": 8.046053390009144e-05, + "loss": 1.1295, + "step": 4834 + }, + { + "epoch": 0.2937602527492557, + "grad_norm": 0.18286018073558807, + "learning_rate": 8.045294205227889e-05, + "loss": 1.0861, + "step": 4835 + }, + { + "epoch": 0.29382100978188225, + "grad_norm": 0.1811809539794922, + "learning_rate": 8.044534908820786e-05, + "loss": 1.0573, + "step": 4836 + }, + { + "epoch": 0.2938817668145088, + "grad_norm": 0.300636887550354, + "learning_rate": 8.043775500815668e-05, + "loss": 1.2696, + "step": 4837 + }, + { + "epoch": 0.2939425238471353, + "grad_norm": 5.600392818450928, + "learning_rate": 8.04301598124037e-05, + "loss": 1.1711, + "step": 4838 + }, + { + "epoch": 0.29400328087976185, + "grad_norm": 0.271645188331604, + "learning_rate": 8.042256350122734e-05, + "loss": 1.2045, + "step": 4839 + }, + { + "epoch": 0.29406403791238833, + "grad_norm": 0.32567843794822693, + "learning_rate": 8.041496607490602e-05, + "loss": 1.1491, + "step": 4840 + }, + { + "epoch": 0.29412479494501487, + "grad_norm": 1.7831114530563354, + "learning_rate": 8.040736753371825e-05, + "loss": 1.1191, + "step": 4841 + }, + { + "epoch": 0.2941855519776414, + "grad_norm": 0.27481865882873535, + "learning_rate": 8.039976787794255e-05, + "loss": 1.0585, + "step": 4842 + }, + { + "epoch": 0.29424630901026794, + "grad_norm": 0.2196209728717804, + "learning_rate": 8.039216710785746e-05, + "loss": 1.0669, + "step": 4843 + }, + { + "epoch": 0.29430706604289447, + "grad_norm": 6.92055082321167, + "learning_rate": 8.038456522374163e-05, + "loss": 1.3323, + "step": 4844 + }, + { + "epoch": 0.294367823075521, + "grad_norm": 0.24645161628723145, + "learning_rate": 8.037696222587368e-05, + "loss": 1.0473, + "step": 4845 + }, + { + "epoch": 0.29442858010814754, + "grad_norm": 0.23090581595897675, + "learning_rate": 8.036935811453231e-05, + "loss": 1.0358, + "step": 4846 + }, + { + "epoch": 0.294489337140774, + "grad_norm": 0.2889900207519531, + "learning_rate": 8.036175288999625e-05, + "loss": 1.2067, + "step": 4847 + }, + { + "epoch": 0.29455009417340056, + "grad_norm": 0.4082173705101013, + "learning_rate": 8.035414655254427e-05, + "loss": 1.238, + "step": 4848 + }, + { + "epoch": 0.2946108512060271, + "grad_norm": 0.3699961304664612, + "learning_rate": 8.034653910245517e-05, + "loss": 1.0779, + "step": 4849 + }, + { + "epoch": 0.2946716082386536, + "grad_norm": 0.23439577221870422, + "learning_rate": 8.033893054000783e-05, + "loss": 1.1637, + "step": 4850 + }, + { + "epoch": 0.29473236527128016, + "grad_norm": 0.49045947194099426, + "learning_rate": 8.033132086548112e-05, + "loss": 1.1999, + "step": 4851 + }, + { + "epoch": 0.2947931223039067, + "grad_norm": 0.44720524549484253, + "learning_rate": 8.032371007915397e-05, + "loss": 1.1874, + "step": 4852 + }, + { + "epoch": 0.29485387933653323, + "grad_norm": 0.2704732418060303, + "learning_rate": 8.031609818130538e-05, + "loss": 1.0458, + "step": 4853 + }, + { + "epoch": 0.2949146363691597, + "grad_norm": 0.34996867179870605, + "learning_rate": 8.030848517221434e-05, + "loss": 1.0729, + "step": 4854 + }, + { + "epoch": 0.29497539340178625, + "grad_norm": 0.24919576942920685, + "learning_rate": 8.030087105215993e-05, + "loss": 1.0753, + "step": 4855 + }, + { + "epoch": 0.2950361504344128, + "grad_norm": 0.5158717632293701, + "learning_rate": 8.029325582142121e-05, + "loss": 1.1161, + "step": 4856 + }, + { + "epoch": 0.2950969074670393, + "grad_norm": 0.45708325505256653, + "learning_rate": 8.028563948027737e-05, + "loss": 1.1288, + "step": 4857 + }, + { + "epoch": 0.29515766449966585, + "grad_norm": 0.22103480994701385, + "learning_rate": 8.027802202900756e-05, + "loss": 1.0398, + "step": 4858 + }, + { + "epoch": 0.2952184215322924, + "grad_norm": 1.8068064451217651, + "learning_rate": 8.027040346789099e-05, + "loss": 1.1268, + "step": 4859 + }, + { + "epoch": 0.29527917856491886, + "grad_norm": 0.4310981035232544, + "learning_rate": 8.026278379720693e-05, + "loss": 1.0423, + "step": 4860 + }, + { + "epoch": 0.2953399355975454, + "grad_norm": 0.21926051378250122, + "learning_rate": 8.025516301723468e-05, + "loss": 1.0748, + "step": 4861 + }, + { + "epoch": 0.29540069263017193, + "grad_norm": 0.2934091091156006, + "learning_rate": 8.02475411282536e-05, + "loss": 1.0575, + "step": 4862 + }, + { + "epoch": 0.29546144966279847, + "grad_norm": 0.3048541247844696, + "learning_rate": 8.023991813054302e-05, + "loss": 1.0575, + "step": 4863 + }, + { + "epoch": 0.295522206695425, + "grad_norm": 0.22716474533081055, + "learning_rate": 8.023229402438241e-05, + "loss": 1.1623, + "step": 4864 + }, + { + "epoch": 0.29558296372805154, + "grad_norm": 0.2674742341041565, + "learning_rate": 8.022466881005121e-05, + "loss": 1.0464, + "step": 4865 + }, + { + "epoch": 0.2956437207606781, + "grad_norm": 0.22621771693229675, + "learning_rate": 8.021704248782895e-05, + "loss": 1.1497, + "step": 4866 + }, + { + "epoch": 0.29570447779330455, + "grad_norm": 0.588484525680542, + "learning_rate": 8.020941505799515e-05, + "loss": 1.3153, + "step": 4867 + }, + { + "epoch": 0.2957652348259311, + "grad_norm": 0.21318784356117249, + "learning_rate": 8.02017865208294e-05, + "loss": 1.0964, + "step": 4868 + }, + { + "epoch": 0.2958259918585576, + "grad_norm": 0.28407055139541626, + "learning_rate": 8.019415687661134e-05, + "loss": 1.0834, + "step": 4869 + }, + { + "epoch": 0.29588674889118416, + "grad_norm": 0.3952322006225586, + "learning_rate": 8.01865261256206e-05, + "loss": 1.1139, + "step": 4870 + }, + { + "epoch": 0.2959475059238107, + "grad_norm": 0.3263978958129883, + "learning_rate": 8.017889426813692e-05, + "loss": 1.1167, + "step": 4871 + }, + { + "epoch": 0.29600826295643723, + "grad_norm": 0.865190327167511, + "learning_rate": 8.017126130444004e-05, + "loss": 1.1577, + "step": 4872 + }, + { + "epoch": 0.2960690199890637, + "grad_norm": 0.2038256675004959, + "learning_rate": 8.016362723480974e-05, + "loss": 1.0447, + "step": 4873 + }, + { + "epoch": 0.29612977702169024, + "grad_norm": 0.8161500096321106, + "learning_rate": 8.015599205952586e-05, + "loss": 1.1054, + "step": 4874 + }, + { + "epoch": 0.2961905340543168, + "grad_norm": 0.3879222571849823, + "learning_rate": 8.014835577886826e-05, + "loss": 1.0902, + "step": 4875 + }, + { + "epoch": 0.2962512910869433, + "grad_norm": 0.29953858256340027, + "learning_rate": 8.014071839311686e-05, + "loss": 1.1637, + "step": 4876 + }, + { + "epoch": 0.29631204811956985, + "grad_norm": 0.3429078757762909, + "learning_rate": 8.013307990255159e-05, + "loss": 1.0678, + "step": 4877 + }, + { + "epoch": 0.2963728051521964, + "grad_norm": 0.5738452076911926, + "learning_rate": 8.012544030745244e-05, + "loss": 1.1966, + "step": 4878 + }, + { + "epoch": 0.2964335621848229, + "grad_norm": 0.20697364211082458, + "learning_rate": 8.011779960809947e-05, + "loss": 1.1056, + "step": 4879 + }, + { + "epoch": 0.2964943192174494, + "grad_norm": 0.5073276162147522, + "learning_rate": 8.011015780477274e-05, + "loss": 1.091, + "step": 4880 + }, + { + "epoch": 0.29655507625007593, + "grad_norm": 0.2145596444606781, + "learning_rate": 8.010251489775234e-05, + "loss": 1.088, + "step": 4881 + }, + { + "epoch": 0.29661583328270247, + "grad_norm": 0.2333339899778366, + "learning_rate": 8.009487088731845e-05, + "loss": 1.1971, + "step": 4882 + }, + { + "epoch": 0.296676590315329, + "grad_norm": 0.36890196800231934, + "learning_rate": 8.008722577375122e-05, + "loss": 1.1287, + "step": 4883 + }, + { + "epoch": 0.29673734734795554, + "grad_norm": 0.2669888436794281, + "learning_rate": 8.007957955733094e-05, + "loss": 1.181, + "step": 4884 + }, + { + "epoch": 0.2967981043805821, + "grad_norm": 1.2394779920578003, + "learning_rate": 8.007193223833786e-05, + "loss": 1.1026, + "step": 4885 + }, + { + "epoch": 0.29685886141320855, + "grad_norm": 0.31223371624946594, + "learning_rate": 8.006428381705228e-05, + "loss": 1.1187, + "step": 4886 + }, + { + "epoch": 0.2969196184458351, + "grad_norm": 0.1892867386341095, + "learning_rate": 8.005663429375457e-05, + "loss": 1.062, + "step": 4887 + }, + { + "epoch": 0.2969803754784616, + "grad_norm": 0.45092037320137024, + "learning_rate": 8.004898366872512e-05, + "loss": 1.1009, + "step": 4888 + }, + { + "epoch": 0.29704113251108816, + "grad_norm": 0.22004619240760803, + "learning_rate": 8.004133194224436e-05, + "loss": 1.0785, + "step": 4889 + }, + { + "epoch": 0.2971018895437147, + "grad_norm": 0.25986236333847046, + "learning_rate": 8.003367911459279e-05, + "loss": 1.1416, + "step": 4890 + }, + { + "epoch": 0.2971626465763412, + "grad_norm": 0.31051480770111084, + "learning_rate": 8.002602518605089e-05, + "loss": 1.0882, + "step": 4891 + }, + { + "epoch": 0.29722340360896776, + "grad_norm": 0.2984970510005951, + "learning_rate": 8.001837015689924e-05, + "loss": 1.183, + "step": 4892 + }, + { + "epoch": 0.29728416064159424, + "grad_norm": 0.27492889761924744, + "learning_rate": 8.001071402741842e-05, + "loss": 1.082, + "step": 4893 + }, + { + "epoch": 0.2973449176742208, + "grad_norm": 0.2532314658164978, + "learning_rate": 8.000305679788908e-05, + "loss": 1.0729, + "step": 4894 + }, + { + "epoch": 0.2974056747068473, + "grad_norm": 0.19243381917476654, + "learning_rate": 7.99953984685919e-05, + "loss": 1.0976, + "step": 4895 + }, + { + "epoch": 0.29746643173947385, + "grad_norm": 0.3168008625507355, + "learning_rate": 7.998773903980759e-05, + "loss": 1.2505, + "step": 4896 + }, + { + "epoch": 0.2975271887721004, + "grad_norm": 0.27109095454216003, + "learning_rate": 7.99800785118169e-05, + "loss": 1.0834, + "step": 4897 + }, + { + "epoch": 0.2975879458047269, + "grad_norm": 0.2351675182580948, + "learning_rate": 7.997241688490062e-05, + "loss": 1.0673, + "step": 4898 + }, + { + "epoch": 0.2976487028373534, + "grad_norm": 0.23139260709285736, + "learning_rate": 7.996475415933964e-05, + "loss": 1.1515, + "step": 4899 + }, + { + "epoch": 0.29770945986997993, + "grad_norm": 0.22361722588539124, + "learning_rate": 7.99570903354148e-05, + "loss": 1.066, + "step": 4900 + }, + { + "epoch": 0.29777021690260647, + "grad_norm": 0.19363203644752502, + "learning_rate": 7.9949425413407e-05, + "loss": 1.1163, + "step": 4901 + }, + { + "epoch": 0.297830973935233, + "grad_norm": 0.2517297565937042, + "learning_rate": 7.994175939359725e-05, + "loss": 1.2217, + "step": 4902 + }, + { + "epoch": 0.29789173096785954, + "grad_norm": 0.21075789630413055, + "learning_rate": 7.993409227626651e-05, + "loss": 1.1088, + "step": 4903 + }, + { + "epoch": 0.29795248800048607, + "grad_norm": 0.21663813292980194, + "learning_rate": 7.992642406169584e-05, + "loss": 1.0532, + "step": 4904 + }, + { + "epoch": 0.2980132450331126, + "grad_norm": 0.29269886016845703, + "learning_rate": 7.991875475016629e-05, + "loss": 1.0949, + "step": 4905 + }, + { + "epoch": 0.2980740020657391, + "grad_norm": 0.2963198721408844, + "learning_rate": 7.991108434195902e-05, + "loss": 1.0822, + "step": 4906 + }, + { + "epoch": 0.2981347590983656, + "grad_norm": 0.22851325571537018, + "learning_rate": 7.990341283735516e-05, + "loss": 1.1389, + "step": 4907 + }, + { + "epoch": 0.29819551613099216, + "grad_norm": 0.18272531032562256, + "learning_rate": 7.989574023663592e-05, + "loss": 1.0359, + "step": 4908 + }, + { + "epoch": 0.2982562731636187, + "grad_norm": 0.1861983984708786, + "learning_rate": 7.988806654008256e-05, + "loss": 1.0616, + "step": 4909 + }, + { + "epoch": 0.2983170301962452, + "grad_norm": 0.30188316106796265, + "learning_rate": 7.988039174797633e-05, + "loss": 1.1545, + "step": 4910 + }, + { + "epoch": 0.29837778722887176, + "grad_norm": 0.26625677943229675, + "learning_rate": 7.987271586059857e-05, + "loss": 1.1799, + "step": 4911 + }, + { + "epoch": 0.2984385442614983, + "grad_norm": 0.3447107970714569, + "learning_rate": 7.986503887823064e-05, + "loss": 1.0791, + "step": 4912 + }, + { + "epoch": 0.2984993012941248, + "grad_norm": 0.27067360281944275, + "learning_rate": 7.985736080115393e-05, + "loss": 1.1665, + "step": 4913 + }, + { + "epoch": 0.2985600583267513, + "grad_norm": 0.21835681796073914, + "learning_rate": 7.98496816296499e-05, + "loss": 1.0334, + "step": 4914 + }, + { + "epoch": 0.29862081535937784, + "grad_norm": 8.168757438659668, + "learning_rate": 7.984200136399999e-05, + "loss": 1.0668, + "step": 4915 + }, + { + "epoch": 0.2986815723920044, + "grad_norm": 0.3131493628025055, + "learning_rate": 7.983432000448578e-05, + "loss": 1.0832, + "step": 4916 + }, + { + "epoch": 0.2987423294246309, + "grad_norm": 0.24023108184337616, + "learning_rate": 7.982663755138879e-05, + "loss": 1.0636, + "step": 4917 + }, + { + "epoch": 0.29880308645725745, + "grad_norm": 0.37717965245246887, + "learning_rate": 7.981895400499065e-05, + "loss": 1.0659, + "step": 4918 + }, + { + "epoch": 0.29886384348988393, + "grad_norm": 0.5195465087890625, + "learning_rate": 7.981126936557298e-05, + "loss": 1.2569, + "step": 4919 + }, + { + "epoch": 0.29892460052251046, + "grad_norm": 0.23798975348472595, + "learning_rate": 7.980358363341747e-05, + "loss": 1.1225, + "step": 4920 + }, + { + "epoch": 0.298985357555137, + "grad_norm": 0.4029634892940521, + "learning_rate": 7.979589680880584e-05, + "loss": 1.1843, + "step": 4921 + }, + { + "epoch": 0.29904611458776353, + "grad_norm": 0.19362621009349823, + "learning_rate": 7.978820889201985e-05, + "loss": 1.0602, + "step": 4922 + }, + { + "epoch": 0.29910687162039007, + "grad_norm": 0.42026543617248535, + "learning_rate": 7.97805198833413e-05, + "loss": 1.1549, + "step": 4923 + }, + { + "epoch": 0.2991676286530166, + "grad_norm": 0.5220346450805664, + "learning_rate": 7.977282978305205e-05, + "loss": 1.1579, + "step": 4924 + }, + { + "epoch": 0.29922838568564314, + "grad_norm": 0.3725937008857727, + "learning_rate": 7.976513859143395e-05, + "loss": 1.0356, + "step": 4925 + }, + { + "epoch": 0.2992891427182696, + "grad_norm": 0.6083593964576721, + "learning_rate": 7.975744630876896e-05, + "loss": 1.0832, + "step": 4926 + }, + { + "epoch": 0.29934989975089615, + "grad_norm": 0.19677482545375824, + "learning_rate": 7.974975293533902e-05, + "loss": 1.0962, + "step": 4927 + }, + { + "epoch": 0.2994106567835227, + "grad_norm": 0.2961776554584503, + "learning_rate": 7.974205847142613e-05, + "loss": 1.1324, + "step": 4928 + }, + { + "epoch": 0.2994714138161492, + "grad_norm": 0.202584907412529, + "learning_rate": 7.973436291731234e-05, + "loss": 1.0812, + "step": 4929 + }, + { + "epoch": 0.29953217084877576, + "grad_norm": 3.134655475616455, + "learning_rate": 7.972666627327974e-05, + "loss": 1.0521, + "step": 4930 + }, + { + "epoch": 0.2995929278814023, + "grad_norm": 0.25369200110435486, + "learning_rate": 7.971896853961042e-05, + "loss": 1.1277, + "step": 4931 + }, + { + "epoch": 0.2996536849140288, + "grad_norm": 0.5255468487739563, + "learning_rate": 7.971126971658659e-05, + "loss": 1.3923, + "step": 4932 + }, + { + "epoch": 0.2997144419466553, + "grad_norm": 0.1718311905860901, + "learning_rate": 7.970356980449041e-05, + "loss": 1.0568, + "step": 4933 + }, + { + "epoch": 0.29977519897928184, + "grad_norm": 0.22870288789272308, + "learning_rate": 7.969586880360413e-05, + "loss": 1.0894, + "step": 4934 + }, + { + "epoch": 0.2998359560119084, + "grad_norm": 0.40918076038360596, + "learning_rate": 7.968816671421005e-05, + "loss": 1.2372, + "step": 4935 + }, + { + "epoch": 0.2998967130445349, + "grad_norm": 0.2263970673084259, + "learning_rate": 7.968046353659048e-05, + "loss": 1.2608, + "step": 4936 + }, + { + "epoch": 0.29995747007716145, + "grad_norm": 0.19780585169792175, + "learning_rate": 7.967275927102777e-05, + "loss": 1.0464, + "step": 4937 + }, + { + "epoch": 0.300018227109788, + "grad_norm": 0.2243894785642624, + "learning_rate": 7.966505391780436e-05, + "loss": 1.1545, + "step": 4938 + }, + { + "epoch": 0.30007898414241446, + "grad_norm": 0.16121649742126465, + "learning_rate": 7.965734747720264e-05, + "loss": 1.0932, + "step": 4939 + }, + { + "epoch": 0.300139741175041, + "grad_norm": 0.1909124106168747, + "learning_rate": 7.964963994950513e-05, + "loss": 1.0806, + "step": 4940 + }, + { + "epoch": 0.30020049820766753, + "grad_norm": 0.7291968464851379, + "learning_rate": 7.964193133499432e-05, + "loss": 1.043, + "step": 4941 + }, + { + "epoch": 0.30026125524029407, + "grad_norm": 0.23494552075862885, + "learning_rate": 7.963422163395281e-05, + "loss": 1.1377, + "step": 4942 + }, + { + "epoch": 0.3003220122729206, + "grad_norm": 0.16675245761871338, + "learning_rate": 7.962651084666316e-05, + "loss": 1.049, + "step": 4943 + }, + { + "epoch": 0.30038276930554714, + "grad_norm": 0.15246737003326416, + "learning_rate": 7.961879897340803e-05, + "loss": 1.0122, + "step": 4944 + }, + { + "epoch": 0.3004435263381736, + "grad_norm": 0.198641836643219, + "learning_rate": 7.96110860144701e-05, + "loss": 1.1274, + "step": 4945 + }, + { + "epoch": 0.30050428337080015, + "grad_norm": 0.15419761836528778, + "learning_rate": 7.96033719701321e-05, + "loss": 0.9904, + "step": 4946 + }, + { + "epoch": 0.3005650404034267, + "grad_norm": 0.2493741661310196, + "learning_rate": 7.959565684067675e-05, + "loss": 1.2688, + "step": 4947 + }, + { + "epoch": 0.3006257974360532, + "grad_norm": 0.15781021118164062, + "learning_rate": 7.95879406263869e-05, + "loss": 1.0744, + "step": 4948 + }, + { + "epoch": 0.30068655446867976, + "grad_norm": 0.19979479908943176, + "learning_rate": 7.958022332754537e-05, + "loss": 1.0714, + "step": 4949 + }, + { + "epoch": 0.3007473115013063, + "grad_norm": 0.26432523131370544, + "learning_rate": 7.957250494443504e-05, + "loss": 1.1539, + "step": 4950 + }, + { + "epoch": 0.3008080685339328, + "grad_norm": 0.1792435199022293, + "learning_rate": 7.956478547733881e-05, + "loss": 1.0167, + "step": 4951 + }, + { + "epoch": 0.3008688255665593, + "grad_norm": 0.8181754946708679, + "learning_rate": 7.955706492653966e-05, + "loss": 1.1018, + "step": 4952 + }, + { + "epoch": 0.30092958259918584, + "grad_norm": 0.18530167639255524, + "learning_rate": 7.954934329232058e-05, + "loss": 1.0959, + "step": 4953 + }, + { + "epoch": 0.3009903396318124, + "grad_norm": 0.7599412798881531, + "learning_rate": 7.954162057496462e-05, + "loss": 1.1771, + "step": 4954 + }, + { + "epoch": 0.3010510966644389, + "grad_norm": 0.2918197214603424, + "learning_rate": 7.953389677475482e-05, + "loss": 1.1608, + "step": 4955 + }, + { + "epoch": 0.30111185369706545, + "grad_norm": 0.2729848623275757, + "learning_rate": 7.952617189197435e-05, + "loss": 1.0875, + "step": 4956 + }, + { + "epoch": 0.301172610729692, + "grad_norm": 0.18601828813552856, + "learning_rate": 7.951844592690634e-05, + "loss": 1.0624, + "step": 4957 + }, + { + "epoch": 0.3012333677623185, + "grad_norm": 0.2701513469219208, + "learning_rate": 7.951071887983397e-05, + "loss": 1.1257, + "step": 4958 + }, + { + "epoch": 0.301294124794945, + "grad_norm": 0.1926315873861313, + "learning_rate": 7.950299075104052e-05, + "loss": 1.0737, + "step": 4959 + }, + { + "epoch": 0.30135488182757153, + "grad_norm": 0.2479248344898224, + "learning_rate": 7.949526154080922e-05, + "loss": 1.178, + "step": 4960 + }, + { + "epoch": 0.30141563886019807, + "grad_norm": 0.34194740653038025, + "learning_rate": 7.948753124942343e-05, + "loss": 1.0736, + "step": 4961 + }, + { + "epoch": 0.3014763958928246, + "grad_norm": 0.3293874263763428, + "learning_rate": 7.947979987716647e-05, + "loss": 1.1912, + "step": 4962 + }, + { + "epoch": 0.30153715292545114, + "grad_norm": 0.270099937915802, + "learning_rate": 7.947206742432174e-05, + "loss": 1.1693, + "step": 4963 + }, + { + "epoch": 0.30159790995807767, + "grad_norm": 0.3271844983100891, + "learning_rate": 7.946433389117268e-05, + "loss": 1.0654, + "step": 4964 + }, + { + "epoch": 0.30165866699070415, + "grad_norm": 0.18504393100738525, + "learning_rate": 7.945659927800277e-05, + "loss": 1.0636, + "step": 4965 + }, + { + "epoch": 0.3017194240233307, + "grad_norm": 0.2591862678527832, + "learning_rate": 7.944886358509554e-05, + "loss": 1.0529, + "step": 4966 + }, + { + "epoch": 0.3017801810559572, + "grad_norm": 0.18392758071422577, + "learning_rate": 7.94411268127345e-05, + "loss": 1.031, + "step": 4967 + }, + { + "epoch": 0.30184093808858375, + "grad_norm": 0.2856574058532715, + "learning_rate": 7.943338896120328e-05, + "loss": 1.2172, + "step": 4968 + }, + { + "epoch": 0.3019016951212103, + "grad_norm": 0.20470723509788513, + "learning_rate": 7.942565003078548e-05, + "loss": 1.1029, + "step": 4969 + }, + { + "epoch": 0.3019624521538368, + "grad_norm": 0.243682861328125, + "learning_rate": 7.94179100217648e-05, + "loss": 1.1191, + "step": 4970 + }, + { + "epoch": 0.30202320918646336, + "grad_norm": 0.24615466594696045, + "learning_rate": 7.941016893442495e-05, + "loss": 1.1864, + "step": 4971 + }, + { + "epoch": 0.30208396621908984, + "grad_norm": 0.25825944542884827, + "learning_rate": 7.940242676904965e-05, + "loss": 1.3935, + "step": 4972 + }, + { + "epoch": 0.3021447232517164, + "grad_norm": 0.2408507615327835, + "learning_rate": 7.939468352592274e-05, + "loss": 1.1241, + "step": 4973 + }, + { + "epoch": 0.3022054802843429, + "grad_norm": 0.28525951504707336, + "learning_rate": 7.938693920532801e-05, + "loss": 1.0779, + "step": 4974 + }, + { + "epoch": 0.30226623731696944, + "grad_norm": 0.16390888392925262, + "learning_rate": 7.937919380754934e-05, + "loss": 1.0236, + "step": 4975 + }, + { + "epoch": 0.302326994349596, + "grad_norm": 0.2186303734779358, + "learning_rate": 7.937144733287065e-05, + "loss": 1.0452, + "step": 4976 + }, + { + "epoch": 0.3023877513822225, + "grad_norm": 0.3011877238750458, + "learning_rate": 7.936369978157589e-05, + "loss": 1.187, + "step": 4977 + }, + { + "epoch": 0.302448508414849, + "grad_norm": 0.17117010056972504, + "learning_rate": 7.935595115394902e-05, + "loss": 1.064, + "step": 4978 + }, + { + "epoch": 0.30250926544747553, + "grad_norm": 0.15970885753631592, + "learning_rate": 7.93482014502741e-05, + "loss": 1.0452, + "step": 4979 + }, + { + "epoch": 0.30257002248010206, + "grad_norm": 0.21643060445785522, + "learning_rate": 7.934045067083517e-05, + "loss": 1.0609, + "step": 4980 + }, + { + "epoch": 0.3026307795127286, + "grad_norm": 0.18936458230018616, + "learning_rate": 7.933269881591635e-05, + "loss": 1.095, + "step": 4981 + }, + { + "epoch": 0.30269153654535513, + "grad_norm": 0.23942099511623383, + "learning_rate": 7.932494588580179e-05, + "loss": 1.0967, + "step": 4982 + }, + { + "epoch": 0.30275229357798167, + "grad_norm": 0.3439854383468628, + "learning_rate": 7.931719188077567e-05, + "loss": 1.1222, + "step": 4983 + }, + { + "epoch": 0.3028130506106082, + "grad_norm": 0.23856957256793976, + "learning_rate": 7.930943680112222e-05, + "loss": 1.1796, + "step": 4984 + }, + { + "epoch": 0.3028738076432347, + "grad_norm": 8.016382217407227, + "learning_rate": 7.930168064712569e-05, + "loss": 1.0792, + "step": 4985 + }, + { + "epoch": 0.3029345646758612, + "grad_norm": 0.9487180709838867, + "learning_rate": 7.92939234190704e-05, + "loss": 1.1607, + "step": 4986 + }, + { + "epoch": 0.30299532170848775, + "grad_norm": 0.3822665214538574, + "learning_rate": 7.928616511724068e-05, + "loss": 1.1783, + "step": 4987 + }, + { + "epoch": 0.3030560787411143, + "grad_norm": 0.35022422671318054, + "learning_rate": 7.927840574192092e-05, + "loss": 1.1461, + "step": 4988 + }, + { + "epoch": 0.3031168357737408, + "grad_norm": 0.43493959307670593, + "learning_rate": 7.927064529339555e-05, + "loss": 1.1431, + "step": 4989 + }, + { + "epoch": 0.30317759280636736, + "grad_norm": 0.4426456093788147, + "learning_rate": 7.926288377194899e-05, + "loss": 1.2305, + "step": 4990 + }, + { + "epoch": 0.30323834983899384, + "grad_norm": 0.6046739816665649, + "learning_rate": 7.925512117786578e-05, + "loss": 1.1182, + "step": 4991 + }, + { + "epoch": 0.30329910687162037, + "grad_norm": 0.5481765270233154, + "learning_rate": 7.924735751143045e-05, + "loss": 1.2475, + "step": 4992 + }, + { + "epoch": 0.3033598639042469, + "grad_norm": 0.9899928569793701, + "learning_rate": 7.923959277292759e-05, + "loss": 1.193, + "step": 4993 + }, + { + "epoch": 0.30342062093687344, + "grad_norm": 0.3839530348777771, + "learning_rate": 7.923182696264177e-05, + "loss": 1.0791, + "step": 4994 + }, + { + "epoch": 0.3034813779695, + "grad_norm": 0.3279713988304138, + "learning_rate": 7.922406008085771e-05, + "loss": 1.2202, + "step": 4995 + }, + { + "epoch": 0.3035421350021265, + "grad_norm": 0.25487226247787476, + "learning_rate": 7.921629212786008e-05, + "loss": 1.1488, + "step": 4996 + }, + { + "epoch": 0.30360289203475305, + "grad_norm": 0.23226816952228546, + "learning_rate": 7.92085231039336e-05, + "loss": 1.1273, + "step": 4997 + }, + { + "epoch": 0.3036636490673795, + "grad_norm": 0.3342590928077698, + "learning_rate": 7.920075300936308e-05, + "loss": 1.1915, + "step": 4998 + }, + { + "epoch": 0.30372440610000606, + "grad_norm": 0.23901338875293732, + "learning_rate": 7.919298184443331e-05, + "loss": 1.0448, + "step": 4999 + }, + { + "epoch": 0.3037851631326326, + "grad_norm": 1.035646677017212, + "learning_rate": 7.918520960942913e-05, + "loss": 1.2407, + "step": 5000 + }, + { + "epoch": 0.30384592016525913, + "grad_norm": 0.30226588249206543, + "learning_rate": 7.917743630463546e-05, + "loss": 1.1395, + "step": 5001 + }, + { + "epoch": 0.30390667719788567, + "grad_norm": 0.30129021406173706, + "learning_rate": 7.916966193033723e-05, + "loss": 1.1233, + "step": 5002 + }, + { + "epoch": 0.3039674342305122, + "grad_norm": 0.28231826424598694, + "learning_rate": 7.916188648681939e-05, + "loss": 1.1962, + "step": 5003 + }, + { + "epoch": 0.3040281912631387, + "grad_norm": 0.2995006740093231, + "learning_rate": 7.915410997436697e-05, + "loss": 1.0772, + "step": 5004 + }, + { + "epoch": 0.3040889482957652, + "grad_norm": 0.2836805284023285, + "learning_rate": 7.914633239326502e-05, + "loss": 1.0969, + "step": 5005 + }, + { + "epoch": 0.30414970532839175, + "grad_norm": 6.130701065063477, + "learning_rate": 7.913855374379862e-05, + "loss": 1.2708, + "step": 5006 + }, + { + "epoch": 0.3042104623610183, + "grad_norm": 0.5131632089614868, + "learning_rate": 7.913077402625289e-05, + "loss": 1.0968, + "step": 5007 + }, + { + "epoch": 0.3042712193936448, + "grad_norm": 0.4194967448711395, + "learning_rate": 7.912299324091301e-05, + "loss": 1.1847, + "step": 5008 + }, + { + "epoch": 0.30433197642627136, + "grad_norm": 0.3844158351421356, + "learning_rate": 7.911521138806419e-05, + "loss": 1.0683, + "step": 5009 + }, + { + "epoch": 0.3043927334588979, + "grad_norm": 0.3503633141517639, + "learning_rate": 7.910742846799169e-05, + "loss": 1.0719, + "step": 5010 + }, + { + "epoch": 0.30445349049152437, + "grad_norm": 0.4003380835056305, + "learning_rate": 7.909964448098073e-05, + "loss": 1.0575, + "step": 5011 + }, + { + "epoch": 0.3045142475241509, + "grad_norm": 0.444274365901947, + "learning_rate": 7.90918594273167e-05, + "loss": 1.3055, + "step": 5012 + }, + { + "epoch": 0.30457500455677744, + "grad_norm": 0.24903325736522675, + "learning_rate": 7.908407330728493e-05, + "loss": 1.0533, + "step": 5013 + }, + { + "epoch": 0.304635761589404, + "grad_norm": 1.3028353452682495, + "learning_rate": 7.907628612117084e-05, + "loss": 1.2449, + "step": 5014 + }, + { + "epoch": 0.3046965186220305, + "grad_norm": 0.2387785017490387, + "learning_rate": 7.906849786925987e-05, + "loss": 1.1353, + "step": 5015 + }, + { + "epoch": 0.30475727565465704, + "grad_norm": 0.3817557990550995, + "learning_rate": 7.906070855183747e-05, + "loss": 1.1789, + "step": 5016 + }, + { + "epoch": 0.3048180326872836, + "grad_norm": 0.6513555645942688, + "learning_rate": 7.90529181691892e-05, + "loss": 1.2433, + "step": 5017 + }, + { + "epoch": 0.30487878971991006, + "grad_norm": 0.2681402862071991, + "learning_rate": 7.904512672160058e-05, + "loss": 1.1394, + "step": 5018 + }, + { + "epoch": 0.3049395467525366, + "grad_norm": 0.48683372139930725, + "learning_rate": 7.903733420935723e-05, + "loss": 1.0701, + "step": 5019 + }, + { + "epoch": 0.30500030378516313, + "grad_norm": 0.19611535966396332, + "learning_rate": 7.90295406327448e-05, + "loss": 1.0826, + "step": 5020 + }, + { + "epoch": 0.30506106081778966, + "grad_norm": 0.7331171035766602, + "learning_rate": 7.902174599204892e-05, + "loss": 1.1151, + "step": 5021 + }, + { + "epoch": 0.3051218178504162, + "grad_norm": 0.4019535481929779, + "learning_rate": 7.901395028755536e-05, + "loss": 1.3106, + "step": 5022 + }, + { + "epoch": 0.30518257488304273, + "grad_norm": 0.25834062695503235, + "learning_rate": 7.900615351954982e-05, + "loss": 1.1374, + "step": 5023 + }, + { + "epoch": 0.3052433319156692, + "grad_norm": 0.3149908185005188, + "learning_rate": 7.899835568831813e-05, + "loss": 1.0381, + "step": 5024 + }, + { + "epoch": 0.30530408894829575, + "grad_norm": 0.4119406044483185, + "learning_rate": 7.89905567941461e-05, + "loss": 1.1122, + "step": 5025 + }, + { + "epoch": 0.3053648459809223, + "grad_norm": 0.2397211790084839, + "learning_rate": 7.898275683731962e-05, + "loss": 1.121, + "step": 5026 + }, + { + "epoch": 0.3054256030135488, + "grad_norm": 0.39136889576911926, + "learning_rate": 7.897495581812457e-05, + "loss": 1.0347, + "step": 5027 + }, + { + "epoch": 0.30548636004617535, + "grad_norm": 0.2746813893318176, + "learning_rate": 7.89671537368469e-05, + "loss": 1.1424, + "step": 5028 + }, + { + "epoch": 0.3055471170788019, + "grad_norm": 1.182684302330017, + "learning_rate": 7.895935059377264e-05, + "loss": 1.3109, + "step": 5029 + }, + { + "epoch": 0.3056078741114284, + "grad_norm": 0.6201323866844177, + "learning_rate": 7.895154638918778e-05, + "loss": 1.1416, + "step": 5030 + }, + { + "epoch": 0.3056686311440549, + "grad_norm": 0.3139815330505371, + "learning_rate": 7.894374112337839e-05, + "loss": 1.141, + "step": 5031 + }, + { + "epoch": 0.30572938817668144, + "grad_norm": 0.20476464927196503, + "learning_rate": 7.893593479663057e-05, + "loss": 1.0389, + "step": 5032 + }, + { + "epoch": 0.305790145209308, + "grad_norm": 0.3290420472621918, + "learning_rate": 7.892812740923045e-05, + "loss": 1.0539, + "step": 5033 + }, + { + "epoch": 0.3058509022419345, + "grad_norm": 0.20500969886779785, + "learning_rate": 7.892031896146425e-05, + "loss": 1.0631, + "step": 5034 + }, + { + "epoch": 0.30591165927456104, + "grad_norm": 0.16665057837963104, + "learning_rate": 7.891250945361817e-05, + "loss": 1.0794, + "step": 5035 + }, + { + "epoch": 0.3059724163071876, + "grad_norm": 0.19782572984695435, + "learning_rate": 7.890469888597847e-05, + "loss": 1.0357, + "step": 5036 + }, + { + "epoch": 0.30603317333981406, + "grad_norm": 0.2886766493320465, + "learning_rate": 7.889688725883143e-05, + "loss": 1.08, + "step": 5037 + }, + { + "epoch": 0.3060939303724406, + "grad_norm": 0.16725771129131317, + "learning_rate": 7.88890745724634e-05, + "loss": 1.0933, + "step": 5038 + }, + { + "epoch": 0.3061546874050671, + "grad_norm": 0.29976433515548706, + "learning_rate": 7.888126082716077e-05, + "loss": 1.0861, + "step": 5039 + }, + { + "epoch": 0.30621544443769366, + "grad_norm": 0.20861996710300446, + "learning_rate": 7.887344602320994e-05, + "loss": 1.087, + "step": 5040 + }, + { + "epoch": 0.3062762014703202, + "grad_norm": 0.3871104121208191, + "learning_rate": 7.886563016089737e-05, + "loss": 1.1858, + "step": 5041 + }, + { + "epoch": 0.30633695850294673, + "grad_norm": 0.2581096589565277, + "learning_rate": 7.885781324050953e-05, + "loss": 1.1666, + "step": 5042 + }, + { + "epoch": 0.30639771553557327, + "grad_norm": 0.23690351843833923, + "learning_rate": 7.884999526233297e-05, + "loss": 1.0795, + "step": 5043 + }, + { + "epoch": 0.30645847256819975, + "grad_norm": 0.18148718774318695, + "learning_rate": 7.884217622665426e-05, + "loss": 1.0854, + "step": 5044 + }, + { + "epoch": 0.3065192296008263, + "grad_norm": 0.2581735849380493, + "learning_rate": 7.883435613376003e-05, + "loss": 1.1206, + "step": 5045 + }, + { + "epoch": 0.3065799866334528, + "grad_norm": 0.6960493922233582, + "learning_rate": 7.882653498393688e-05, + "loss": 1.1955, + "step": 5046 + }, + { + "epoch": 0.30664074366607935, + "grad_norm": 0.29692861437797546, + "learning_rate": 7.88187127774715e-05, + "loss": 1.139, + "step": 5047 + }, + { + "epoch": 0.3067015006987059, + "grad_norm": 0.2609367072582245, + "learning_rate": 7.881088951465066e-05, + "loss": 1.0608, + "step": 5048 + }, + { + "epoch": 0.3067622577313324, + "grad_norm": 0.22074176371097565, + "learning_rate": 7.880306519576109e-05, + "loss": 1.129, + "step": 5049 + }, + { + "epoch": 0.3068230147639589, + "grad_norm": 0.19651450216770172, + "learning_rate": 7.87952398210896e-05, + "loss": 1.0186, + "step": 5050 + }, + { + "epoch": 0.30688377179658544, + "grad_norm": 0.17749401926994324, + "learning_rate": 7.878741339092302e-05, + "loss": 1.0667, + "step": 5051 + }, + { + "epoch": 0.30694452882921197, + "grad_norm": 0.24965369701385498, + "learning_rate": 7.877958590554824e-05, + "loss": 1.2572, + "step": 5052 + }, + { + "epoch": 0.3070052858618385, + "grad_norm": 0.2053968459367752, + "learning_rate": 7.877175736525217e-05, + "loss": 1.0656, + "step": 5053 + }, + { + "epoch": 0.30706604289446504, + "grad_norm": 0.2848794162273407, + "learning_rate": 7.87639277703218e-05, + "loss": 1.3795, + "step": 5054 + }, + { + "epoch": 0.3071267999270916, + "grad_norm": 0.21953065693378448, + "learning_rate": 7.875609712104407e-05, + "loss": 1.0995, + "step": 5055 + }, + { + "epoch": 0.3071875569597181, + "grad_norm": 0.3174144923686981, + "learning_rate": 7.874826541770604e-05, + "loss": 1.0603, + "step": 5056 + }, + { + "epoch": 0.3072483139923446, + "grad_norm": 0.19155822694301605, + "learning_rate": 7.874043266059479e-05, + "loss": 1.1048, + "step": 5057 + }, + { + "epoch": 0.3073090710249711, + "grad_norm": 0.4409440755844116, + "learning_rate": 7.873259884999744e-05, + "loss": 1.0071, + "step": 5058 + }, + { + "epoch": 0.30736982805759766, + "grad_norm": 0.18718333542346954, + "learning_rate": 7.87247639862011e-05, + "loss": 1.0175, + "step": 5059 + }, + { + "epoch": 0.3074305850902242, + "grad_norm": 0.22756560146808624, + "learning_rate": 7.871692806949301e-05, + "loss": 1.0555, + "step": 5060 + }, + { + "epoch": 0.30749134212285073, + "grad_norm": 0.20601484179496765, + "learning_rate": 7.870909110016036e-05, + "loss": 1.0734, + "step": 5061 + }, + { + "epoch": 0.30755209915547727, + "grad_norm": 0.2214791178703308, + "learning_rate": 7.870125307849042e-05, + "loss": 1.0399, + "step": 5062 + }, + { + "epoch": 0.3076128561881038, + "grad_norm": 0.15091925859451294, + "learning_rate": 7.86934140047705e-05, + "loss": 1.0628, + "step": 5063 + }, + { + "epoch": 0.3076736132207303, + "grad_norm": 0.3641943037509918, + "learning_rate": 7.868557387928796e-05, + "loss": 1.2, + "step": 5064 + }, + { + "epoch": 0.3077343702533568, + "grad_norm": 0.4969852864742279, + "learning_rate": 7.867773270233014e-05, + "loss": 1.0782, + "step": 5065 + }, + { + "epoch": 0.30779512728598335, + "grad_norm": 0.3811797797679901, + "learning_rate": 7.866989047418452e-05, + "loss": 1.264, + "step": 5066 + }, + { + "epoch": 0.3078558843186099, + "grad_norm": 0.49702659249305725, + "learning_rate": 7.86620471951385e-05, + "loss": 1.0576, + "step": 5067 + }, + { + "epoch": 0.3079166413512364, + "grad_norm": 0.42418020963668823, + "learning_rate": 7.865420286547959e-05, + "loss": 1.0751, + "step": 5068 + }, + { + "epoch": 0.30797739838386295, + "grad_norm": 0.4082507789134979, + "learning_rate": 7.864635748549534e-05, + "loss": 1.0856, + "step": 5069 + }, + { + "epoch": 0.30803815541648943, + "grad_norm": 0.4407390058040619, + "learning_rate": 7.863851105547333e-05, + "loss": 1.0817, + "step": 5070 + }, + { + "epoch": 0.30809891244911597, + "grad_norm": 0.23118513822555542, + "learning_rate": 7.863066357570115e-05, + "loss": 1.0399, + "step": 5071 + }, + { + "epoch": 0.3081596694817425, + "grad_norm": 0.5499794483184814, + "learning_rate": 7.862281504646647e-05, + "loss": 1.0201, + "step": 5072 + }, + { + "epoch": 0.30822042651436904, + "grad_norm": 0.30706626176834106, + "learning_rate": 7.861496546805697e-05, + "loss": 1.1775, + "step": 5073 + }, + { + "epoch": 0.3082811835469956, + "grad_norm": 0.41222333908081055, + "learning_rate": 7.860711484076038e-05, + "loss": 1.1985, + "step": 5074 + }, + { + "epoch": 0.3083419405796221, + "grad_norm": 0.564678430557251, + "learning_rate": 7.859926316486448e-05, + "loss": 1.1262, + "step": 5075 + }, + { + "epoch": 0.30840269761224864, + "grad_norm": 0.3298918604850769, + "learning_rate": 7.859141044065704e-05, + "loss": 1.1248, + "step": 5076 + }, + { + "epoch": 0.3084634546448751, + "grad_norm": 0.412176251411438, + "learning_rate": 7.858355666842594e-05, + "loss": 1.0623, + "step": 5077 + }, + { + "epoch": 0.30852421167750166, + "grad_norm": 0.4415786862373352, + "learning_rate": 7.857570184845903e-05, + "loss": 1.1275, + "step": 5078 + }, + { + "epoch": 0.3085849687101282, + "grad_norm": 0.24321508407592773, + "learning_rate": 7.856784598104426e-05, + "loss": 1.2308, + "step": 5079 + }, + { + "epoch": 0.30864572574275473, + "grad_norm": 0.4931944012641907, + "learning_rate": 7.855998906646956e-05, + "loss": 1.1388, + "step": 5080 + }, + { + "epoch": 0.30870648277538126, + "grad_norm": 0.3036195635795593, + "learning_rate": 7.855213110502293e-05, + "loss": 1.1333, + "step": 5081 + }, + { + "epoch": 0.3087672398080078, + "grad_norm": 0.29873010516166687, + "learning_rate": 7.854427209699244e-05, + "loss": 1.1314, + "step": 5082 + }, + { + "epoch": 0.3088279968406343, + "grad_norm": 0.3395376205444336, + "learning_rate": 7.853641204266614e-05, + "loss": 1.1144, + "step": 5083 + }, + { + "epoch": 0.3088887538732608, + "grad_norm": 0.3218289017677307, + "learning_rate": 7.852855094233211e-05, + "loss": 1.1287, + "step": 5084 + }, + { + "epoch": 0.30894951090588735, + "grad_norm": 0.22814008593559265, + "learning_rate": 7.852068879627855e-05, + "loss": 1.1888, + "step": 5085 + }, + { + "epoch": 0.3090102679385139, + "grad_norm": 0.6513504385948181, + "learning_rate": 7.851282560479363e-05, + "loss": 1.0927, + "step": 5086 + }, + { + "epoch": 0.3090710249711404, + "grad_norm": 0.31061989068984985, + "learning_rate": 7.850496136816558e-05, + "loss": 1.0985, + "step": 5087 + }, + { + "epoch": 0.30913178200376695, + "grad_norm": 0.3138844668865204, + "learning_rate": 7.849709608668265e-05, + "loss": 1.2117, + "step": 5088 + }, + { + "epoch": 0.3091925390363935, + "grad_norm": 0.8921366930007935, + "learning_rate": 7.848922976063317e-05, + "loss": 1.047, + "step": 5089 + }, + { + "epoch": 0.30925329606901997, + "grad_norm": 1.410239815711975, + "learning_rate": 7.848136239030546e-05, + "loss": 1.1153, + "step": 5090 + }, + { + "epoch": 0.3093140531016465, + "grad_norm": 0.743571937084198, + "learning_rate": 7.84734939759879e-05, + "loss": 1.1637, + "step": 5091 + }, + { + "epoch": 0.30937481013427304, + "grad_norm": 0.5028643012046814, + "learning_rate": 7.846562451796892e-05, + "loss": 1.0689, + "step": 5092 + }, + { + "epoch": 0.30943556716689957, + "grad_norm": 0.5440186262130737, + "learning_rate": 7.845775401653696e-05, + "loss": 1.0633, + "step": 5093 + }, + { + "epoch": 0.3094963241995261, + "grad_norm": 0.45079725980758667, + "learning_rate": 7.844988247198055e-05, + "loss": 1.2119, + "step": 5094 + }, + { + "epoch": 0.30955708123215264, + "grad_norm": 0.20431235432624817, + "learning_rate": 7.844200988458818e-05, + "loss": 1.1412, + "step": 5095 + }, + { + "epoch": 0.3096178382647791, + "grad_norm": 0.2653803825378418, + "learning_rate": 7.843413625464843e-05, + "loss": 1.0503, + "step": 5096 + }, + { + "epoch": 0.30967859529740566, + "grad_norm": 0.3558289408683777, + "learning_rate": 7.842626158244994e-05, + "loss": 1.0725, + "step": 5097 + }, + { + "epoch": 0.3097393523300322, + "grad_norm": 0.21236690878868103, + "learning_rate": 7.841838586828134e-05, + "loss": 1.0287, + "step": 5098 + }, + { + "epoch": 0.3098001093626587, + "grad_norm": 0.23955917358398438, + "learning_rate": 7.841050911243129e-05, + "loss": 1.0996, + "step": 5099 + }, + { + "epoch": 0.30986086639528526, + "grad_norm": 0.3626970648765564, + "learning_rate": 7.840263131518856e-05, + "loss": 1.1771, + "step": 5100 + }, + { + "epoch": 0.3099216234279118, + "grad_norm": 0.3867068588733673, + "learning_rate": 7.839475247684187e-05, + "loss": 1.04, + "step": 5101 + }, + { + "epoch": 0.30998238046053833, + "grad_norm": 0.2842971086502075, + "learning_rate": 7.838687259768004e-05, + "loss": 1.149, + "step": 5102 + }, + { + "epoch": 0.3100431374931648, + "grad_norm": 0.22530293464660645, + "learning_rate": 7.837899167799194e-05, + "loss": 1.088, + "step": 5103 + }, + { + "epoch": 0.31010389452579135, + "grad_norm": 0.20787085592746735, + "learning_rate": 7.837110971806638e-05, + "loss": 1.1205, + "step": 5104 + }, + { + "epoch": 0.3101646515584179, + "grad_norm": 0.19384881854057312, + "learning_rate": 7.836322671819232e-05, + "loss": 1.1149, + "step": 5105 + }, + { + "epoch": 0.3102254085910444, + "grad_norm": 0.26110291481018066, + "learning_rate": 7.83553426786587e-05, + "loss": 1.1618, + "step": 5106 + }, + { + "epoch": 0.31028616562367095, + "grad_norm": 0.19485929608345032, + "learning_rate": 7.834745759975453e-05, + "loss": 1.0783, + "step": 5107 + }, + { + "epoch": 0.3103469226562975, + "grad_norm": 0.4204399585723877, + "learning_rate": 7.83395714817688e-05, + "loss": 1.1199, + "step": 5108 + }, + { + "epoch": 0.31040767968892397, + "grad_norm": 0.21291494369506836, + "learning_rate": 7.83316843249906e-05, + "loss": 1.1327, + "step": 5109 + }, + { + "epoch": 0.3104684367215505, + "grad_norm": 1.6521141529083252, + "learning_rate": 7.832379612970905e-05, + "loss": 1.1577, + "step": 5110 + }, + { + "epoch": 0.31052919375417704, + "grad_norm": 0.38527724146842957, + "learning_rate": 7.831590689621326e-05, + "loss": 1.0524, + "step": 5111 + }, + { + "epoch": 0.31058995078680357, + "grad_norm": 0.22910213470458984, + "learning_rate": 7.830801662479244e-05, + "loss": 1.249, + "step": 5112 + }, + { + "epoch": 0.3106507078194301, + "grad_norm": 0.8584737181663513, + "learning_rate": 7.830012531573581e-05, + "loss": 1.0769, + "step": 5113 + }, + { + "epoch": 0.31071146485205664, + "grad_norm": 0.2121630609035492, + "learning_rate": 7.829223296933259e-05, + "loss": 1.1706, + "step": 5114 + }, + { + "epoch": 0.3107722218846832, + "grad_norm": 0.2052018940448761, + "learning_rate": 7.828433958587213e-05, + "loss": 1.1482, + "step": 5115 + }, + { + "epoch": 0.31083297891730965, + "grad_norm": 0.3341614603996277, + "learning_rate": 7.827644516564371e-05, + "loss": 1.1638, + "step": 5116 + }, + { + "epoch": 0.3108937359499362, + "grad_norm": 0.2674962282180786, + "learning_rate": 7.826854970893674e-05, + "loss": 1.2353, + "step": 5117 + }, + { + "epoch": 0.3109544929825627, + "grad_norm": 0.2116563469171524, + "learning_rate": 7.826065321604061e-05, + "loss": 1.0716, + "step": 5118 + }, + { + "epoch": 0.31101525001518926, + "grad_norm": 0.40958085656166077, + "learning_rate": 7.825275568724478e-05, + "loss": 1.1159, + "step": 5119 + }, + { + "epoch": 0.3110760070478158, + "grad_norm": 0.1608574390411377, + "learning_rate": 7.824485712283873e-05, + "loss": 1.1033, + "step": 5120 + }, + { + "epoch": 0.31113676408044233, + "grad_norm": 1.2929497957229614, + "learning_rate": 7.823695752311198e-05, + "loss": 1.0645, + "step": 5121 + }, + { + "epoch": 0.31119752111306886, + "grad_norm": 0.2876337170600891, + "learning_rate": 7.82290568883541e-05, + "loss": 1.0308, + "step": 5122 + }, + { + "epoch": 0.31125827814569534, + "grad_norm": 0.48659127950668335, + "learning_rate": 7.822115521885466e-05, + "loss": 1.1345, + "step": 5123 + }, + { + "epoch": 0.3113190351783219, + "grad_norm": 0.21452446281909943, + "learning_rate": 7.821325251490335e-05, + "loss": 1.0606, + "step": 5124 + }, + { + "epoch": 0.3113797922109484, + "grad_norm": 0.28587761521339417, + "learning_rate": 7.82053487767898e-05, + "loss": 1.0634, + "step": 5125 + }, + { + "epoch": 0.31144054924357495, + "grad_norm": 0.21157759428024292, + "learning_rate": 7.819744400480373e-05, + "loss": 1.093, + "step": 5126 + }, + { + "epoch": 0.3115013062762015, + "grad_norm": 0.4360100328922272, + "learning_rate": 7.818953819923491e-05, + "loss": 1.2171, + "step": 5127 + }, + { + "epoch": 0.311562063308828, + "grad_norm": 0.371588796377182, + "learning_rate": 7.818163136037314e-05, + "loss": 1.1501, + "step": 5128 + }, + { + "epoch": 0.3116228203414545, + "grad_norm": 0.4839996099472046, + "learning_rate": 7.817372348850818e-05, + "loss": 1.0765, + "step": 5129 + }, + { + "epoch": 0.31168357737408103, + "grad_norm": 0.23760634660720825, + "learning_rate": 7.816581458392998e-05, + "loss": 1.1737, + "step": 5130 + }, + { + "epoch": 0.31174433440670757, + "grad_norm": 0.24858391284942627, + "learning_rate": 7.815790464692838e-05, + "loss": 1.1482, + "step": 5131 + }, + { + "epoch": 0.3118050914393341, + "grad_norm": 0.24542641639709473, + "learning_rate": 7.814999367779336e-05, + "loss": 1.0869, + "step": 5132 + }, + { + "epoch": 0.31186584847196064, + "grad_norm": 0.21299444139003754, + "learning_rate": 7.814208167681486e-05, + "loss": 1.0824, + "step": 5133 + }, + { + "epoch": 0.3119266055045872, + "grad_norm": 2.76069712638855, + "learning_rate": 7.813416864428294e-05, + "loss": 1.0371, + "step": 5134 + }, + { + "epoch": 0.3119873625372137, + "grad_norm": 0.2692389488220215, + "learning_rate": 7.81262545804876e-05, + "loss": 1.1515, + "step": 5135 + }, + { + "epoch": 0.3120481195698402, + "grad_norm": 0.286941260099411, + "learning_rate": 7.811833948571898e-05, + "loss": 1.265, + "step": 5136 + }, + { + "epoch": 0.3121088766024667, + "grad_norm": 0.1871623694896698, + "learning_rate": 7.81104233602672e-05, + "loss": 1.1053, + "step": 5137 + }, + { + "epoch": 0.31216963363509326, + "grad_norm": 0.28715887665748596, + "learning_rate": 7.810250620442239e-05, + "loss": 1.1844, + "step": 5138 + }, + { + "epoch": 0.3122303906677198, + "grad_norm": 0.20306207239627838, + "learning_rate": 7.80945880184748e-05, + "loss": 1.0934, + "step": 5139 + }, + { + "epoch": 0.3122911477003463, + "grad_norm": 0.29322513937950134, + "learning_rate": 7.808666880271466e-05, + "loss": 1.2059, + "step": 5140 + }, + { + "epoch": 0.31235190473297286, + "grad_norm": 0.17915090918540955, + "learning_rate": 7.807874855743223e-05, + "loss": 1.0416, + "step": 5141 + }, + { + "epoch": 0.31241266176559934, + "grad_norm": 0.22026482224464417, + "learning_rate": 7.807082728291786e-05, + "loss": 1.0604, + "step": 5142 + }, + { + "epoch": 0.3124734187982259, + "grad_norm": 0.4311763048171997, + "learning_rate": 7.806290497946189e-05, + "loss": 1.3291, + "step": 5143 + }, + { + "epoch": 0.3125341758308524, + "grad_norm": 0.6587535738945007, + "learning_rate": 7.805498164735469e-05, + "loss": 1.105, + "step": 5144 + }, + { + "epoch": 0.31259493286347895, + "grad_norm": 0.4846359193325043, + "learning_rate": 7.804705728688675e-05, + "loss": 1.1579, + "step": 5145 + }, + { + "epoch": 0.3126556898961055, + "grad_norm": 0.40906253457069397, + "learning_rate": 7.803913189834849e-05, + "loss": 1.3569, + "step": 5146 + }, + { + "epoch": 0.312716446928732, + "grad_norm": 0.19532684981822968, + "learning_rate": 7.803120548203042e-05, + "loss": 1.0668, + "step": 5147 + }, + { + "epoch": 0.31277720396135855, + "grad_norm": 0.27177271246910095, + "learning_rate": 7.80232780382231e-05, + "loss": 1.1363, + "step": 5148 + }, + { + "epoch": 0.31283796099398503, + "grad_norm": 0.25876736640930176, + "learning_rate": 7.80153495672171e-05, + "loss": 1.0712, + "step": 5149 + }, + { + "epoch": 0.31289871802661157, + "grad_norm": 0.17532879114151, + "learning_rate": 7.800742006930306e-05, + "loss": 1.0901, + "step": 5150 + }, + { + "epoch": 0.3129594750592381, + "grad_norm": 1.9612362384796143, + "learning_rate": 7.799948954477163e-05, + "loss": 1.07, + "step": 5151 + }, + { + "epoch": 0.31302023209186464, + "grad_norm": 0.2579033374786377, + "learning_rate": 7.799155799391349e-05, + "loss": 1.0138, + "step": 5152 + }, + { + "epoch": 0.31308098912449117, + "grad_norm": 0.31394633650779724, + "learning_rate": 7.798362541701938e-05, + "loss": 1.1101, + "step": 5153 + }, + { + "epoch": 0.3131417461571177, + "grad_norm": 0.23297218978405, + "learning_rate": 7.797569181438007e-05, + "loss": 1.1012, + "step": 5154 + }, + { + "epoch": 0.3132025031897442, + "grad_norm": 0.24324153363704681, + "learning_rate": 7.796775718628636e-05, + "loss": 1.1039, + "step": 5155 + }, + { + "epoch": 0.3132632602223707, + "grad_norm": 0.5499903559684753, + "learning_rate": 7.795982153302911e-05, + "loss": 1.1253, + "step": 5156 + }, + { + "epoch": 0.31332401725499726, + "grad_norm": 0.20277142524719238, + "learning_rate": 7.79518848548992e-05, + "loss": 1.0875, + "step": 5157 + }, + { + "epoch": 0.3133847742876238, + "grad_norm": 0.2957229018211365, + "learning_rate": 7.794394715218755e-05, + "loss": 1.1484, + "step": 5158 + }, + { + "epoch": 0.3134455313202503, + "grad_norm": 0.9821813702583313, + "learning_rate": 7.793600842518512e-05, + "loss": 1.1494, + "step": 5159 + }, + { + "epoch": 0.31350628835287686, + "grad_norm": 0.2527599036693573, + "learning_rate": 7.79280686741829e-05, + "loss": 1.1474, + "step": 5160 + }, + { + "epoch": 0.3135670453855034, + "grad_norm": 0.6673265695571899, + "learning_rate": 7.792012789947192e-05, + "loss": 1.1467, + "step": 5161 + }, + { + "epoch": 0.3136278024181299, + "grad_norm": 0.3175158202648163, + "learning_rate": 7.791218610134323e-05, + "loss": 1.1108, + "step": 5162 + }, + { + "epoch": 0.3136885594507564, + "grad_norm": 0.29613250494003296, + "learning_rate": 7.7904243280088e-05, + "loss": 1.0781, + "step": 5163 + }, + { + "epoch": 0.31374931648338295, + "grad_norm": 0.41838839650154114, + "learning_rate": 7.789629943599732e-05, + "loss": 1.1291, + "step": 5164 + }, + { + "epoch": 0.3138100735160095, + "grad_norm": 0.25021111965179443, + "learning_rate": 7.78883545693624e-05, + "loss": 1.2102, + "step": 5165 + }, + { + "epoch": 0.313870830548636, + "grad_norm": 0.37901967763900757, + "learning_rate": 7.788040868047445e-05, + "loss": 1.0721, + "step": 5166 + }, + { + "epoch": 0.31393158758126255, + "grad_norm": 0.24078921973705292, + "learning_rate": 7.787246176962474e-05, + "loss": 1.1094, + "step": 5167 + }, + { + "epoch": 0.3139923446138891, + "grad_norm": 0.2133210450410843, + "learning_rate": 7.786451383710455e-05, + "loss": 1.1609, + "step": 5168 + }, + { + "epoch": 0.31405310164651556, + "grad_norm": 0.23776042461395264, + "learning_rate": 7.78565648832052e-05, + "loss": 1.1235, + "step": 5169 + }, + { + "epoch": 0.3141138586791421, + "grad_norm": 0.25585445761680603, + "learning_rate": 7.784861490821811e-05, + "loss": 1.0845, + "step": 5170 + }, + { + "epoch": 0.31417461571176863, + "grad_norm": 0.2946337163448334, + "learning_rate": 7.784066391243465e-05, + "loss": 1.1299, + "step": 5171 + }, + { + "epoch": 0.31423537274439517, + "grad_norm": 0.19392918050289154, + "learning_rate": 7.783271189614626e-05, + "loss": 1.0817, + "step": 5172 + }, + { + "epoch": 0.3142961297770217, + "grad_norm": 0.23942556977272034, + "learning_rate": 7.782475885964444e-05, + "loss": 1.0837, + "step": 5173 + }, + { + "epoch": 0.31435688680964824, + "grad_norm": 0.4521375596523285, + "learning_rate": 7.78168048032207e-05, + "loss": 1.116, + "step": 5174 + }, + { + "epoch": 0.3144176438422747, + "grad_norm": 0.27296268939971924, + "learning_rate": 7.780884972716662e-05, + "loss": 1.0544, + "step": 5175 + }, + { + "epoch": 0.31447840087490125, + "grad_norm": 0.20544856786727905, + "learning_rate": 7.780089363177376e-05, + "loss": 1.0008, + "step": 5176 + }, + { + "epoch": 0.3145391579075278, + "grad_norm": 0.31177660822868347, + "learning_rate": 7.779293651733379e-05, + "loss": 1.1223, + "step": 5177 + }, + { + "epoch": 0.3145999149401543, + "grad_norm": 0.22274397313594818, + "learning_rate": 7.778497838413834e-05, + "loss": 1.1276, + "step": 5178 + }, + { + "epoch": 0.31466067197278086, + "grad_norm": 0.1892411857843399, + "learning_rate": 7.777701923247916e-05, + "loss": 1.0791, + "step": 5179 + }, + { + "epoch": 0.3147214290054074, + "grad_norm": 0.5216176509857178, + "learning_rate": 7.776905906264795e-05, + "loss": 1.112, + "step": 5180 + }, + { + "epoch": 0.31478218603803393, + "grad_norm": 0.5806528925895691, + "learning_rate": 7.776109787493651e-05, + "loss": 1.1314, + "step": 5181 + }, + { + "epoch": 0.3148429430706604, + "grad_norm": 0.15210238099098206, + "learning_rate": 7.775313566963667e-05, + "loss": 1.1007, + "step": 5182 + }, + { + "epoch": 0.31490370010328694, + "grad_norm": 0.20401963591575623, + "learning_rate": 7.77451724470403e-05, + "loss": 1.0179, + "step": 5183 + }, + { + "epoch": 0.3149644571359135, + "grad_norm": 0.20245835185050964, + "learning_rate": 7.773720820743922e-05, + "loss": 1.1061, + "step": 5184 + }, + { + "epoch": 0.31502521416854, + "grad_norm": 2.099736452102661, + "learning_rate": 7.772924295112545e-05, + "loss": 1.0943, + "step": 5185 + }, + { + "epoch": 0.31508597120116655, + "grad_norm": 0.3350406587123871, + "learning_rate": 7.772127667839092e-05, + "loss": 1.0747, + "step": 5186 + }, + { + "epoch": 0.3151467282337931, + "grad_norm": 0.21987709403038025, + "learning_rate": 7.771330938952761e-05, + "loss": 1.0862, + "step": 5187 + }, + { + "epoch": 0.31520748526641956, + "grad_norm": 0.24611034989356995, + "learning_rate": 7.770534108482761e-05, + "loss": 1.077, + "step": 5188 + }, + { + "epoch": 0.3152682422990461, + "grad_norm": 0.19164875149726868, + "learning_rate": 7.769737176458296e-05, + "loss": 1.1547, + "step": 5189 + }, + { + "epoch": 0.31532899933167263, + "grad_norm": 0.1765301376581192, + "learning_rate": 7.768940142908581e-05, + "loss": 1.0406, + "step": 5190 + }, + { + "epoch": 0.31538975636429917, + "grad_norm": 0.21545591950416565, + "learning_rate": 7.768143007862829e-05, + "loss": 1.0936, + "step": 5191 + }, + { + "epoch": 0.3154505133969257, + "grad_norm": 12.144241333007812, + "learning_rate": 7.767345771350261e-05, + "loss": 1.1707, + "step": 5192 + }, + { + "epoch": 0.31551127042955224, + "grad_norm": 0.2648518979549408, + "learning_rate": 7.766548433400096e-05, + "loss": 1.1594, + "step": 5193 + }, + { + "epoch": 0.3155720274621788, + "grad_norm": 1.2412129640579224, + "learning_rate": 7.765750994041566e-05, + "loss": 1.0817, + "step": 5194 + }, + { + "epoch": 0.31563278449480525, + "grad_norm": 0.2827819287776947, + "learning_rate": 7.764953453303898e-05, + "loss": 1.1514, + "step": 5195 + }, + { + "epoch": 0.3156935415274318, + "grad_norm": 0.2071688324213028, + "learning_rate": 7.764155811216327e-05, + "loss": 1.0291, + "step": 5196 + }, + { + "epoch": 0.3157542985600583, + "grad_norm": 0.41227173805236816, + "learning_rate": 7.76335806780809e-05, + "loss": 1.2817, + "step": 5197 + }, + { + "epoch": 0.31581505559268486, + "grad_norm": 0.3671801686286926, + "learning_rate": 7.76256022310843e-05, + "loss": 1.156, + "step": 5198 + }, + { + "epoch": 0.3158758126253114, + "grad_norm": 0.2600345015525818, + "learning_rate": 7.761762277146589e-05, + "loss": 1.189, + "step": 5199 + }, + { + "epoch": 0.3159365696579379, + "grad_norm": 0.358178049325943, + "learning_rate": 7.76096422995182e-05, + "loss": 1.1517, + "step": 5200 + }, + { + "epoch": 0.3159973266905644, + "grad_norm": 0.4109429121017456, + "learning_rate": 7.760166081553373e-05, + "loss": 1.1552, + "step": 5201 + }, + { + "epoch": 0.31605808372319094, + "grad_norm": 0.2728237807750702, + "learning_rate": 7.759367831980504e-05, + "loss": 1.0697, + "step": 5202 + }, + { + "epoch": 0.3161188407558175, + "grad_norm": 0.24336020648479462, + "learning_rate": 7.758569481262475e-05, + "loss": 1.2094, + "step": 5203 + }, + { + "epoch": 0.316179597788444, + "grad_norm": 0.40323176980018616, + "learning_rate": 7.757771029428547e-05, + "loss": 1.2048, + "step": 5204 + }, + { + "epoch": 0.31624035482107055, + "grad_norm": 0.2383427917957306, + "learning_rate": 7.756972476507989e-05, + "loss": 1.107, + "step": 5205 + }, + { + "epoch": 0.3163011118536971, + "grad_norm": 0.20246252417564392, + "learning_rate": 7.756173822530071e-05, + "loss": 1.1349, + "step": 5206 + }, + { + "epoch": 0.3163618688863236, + "grad_norm": 0.20169281959533691, + "learning_rate": 7.755375067524071e-05, + "loss": 1.0801, + "step": 5207 + }, + { + "epoch": 0.3164226259189501, + "grad_norm": 0.23224472999572754, + "learning_rate": 7.754576211519264e-05, + "loss": 1.1792, + "step": 5208 + }, + { + "epoch": 0.31648338295157663, + "grad_norm": 0.21000228822231293, + "learning_rate": 7.753777254544934e-05, + "loss": 1.1277, + "step": 5209 + }, + { + "epoch": 0.31654413998420317, + "grad_norm": 0.2066906988620758, + "learning_rate": 7.752978196630368e-05, + "loss": 1.1383, + "step": 5210 + }, + { + "epoch": 0.3166048970168297, + "grad_norm": 1.7213609218597412, + "learning_rate": 7.752179037804851e-05, + "loss": 1.0922, + "step": 5211 + }, + { + "epoch": 0.31666565404945624, + "grad_norm": 0.3722764849662781, + "learning_rate": 7.75137977809768e-05, + "loss": 1.076, + "step": 5212 + }, + { + "epoch": 0.31672641108208277, + "grad_norm": 0.21817952394485474, + "learning_rate": 7.750580417538153e-05, + "loss": 1.1718, + "step": 5213 + }, + { + "epoch": 0.31678716811470925, + "grad_norm": 0.2928653359413147, + "learning_rate": 7.749780956155568e-05, + "loss": 1.0606, + "step": 5214 + }, + { + "epoch": 0.3168479251473358, + "grad_norm": 0.19641238451004028, + "learning_rate": 7.74898139397923e-05, + "loss": 1.039, + "step": 5215 + }, + { + "epoch": 0.3169086821799623, + "grad_norm": 0.3434431850910187, + "learning_rate": 7.748181731038449e-05, + "loss": 1.1708, + "step": 5216 + }, + { + "epoch": 0.31696943921258885, + "grad_norm": 0.28805726766586304, + "learning_rate": 7.747381967362536e-05, + "loss": 1.0947, + "step": 5217 + }, + { + "epoch": 0.3170301962452154, + "grad_norm": 0.25769177079200745, + "learning_rate": 7.746582102980804e-05, + "loss": 1.1819, + "step": 5218 + }, + { + "epoch": 0.3170909532778419, + "grad_norm": 0.2461240142583847, + "learning_rate": 7.745782137922574e-05, + "loss": 1.128, + "step": 5219 + }, + { + "epoch": 0.31715171031046846, + "grad_norm": 0.27344274520874023, + "learning_rate": 7.744982072217171e-05, + "loss": 1.1366, + "step": 5220 + }, + { + "epoch": 0.31721246734309494, + "grad_norm": 0.6878777742385864, + "learning_rate": 7.74418190589392e-05, + "loss": 1.0829, + "step": 5221 + }, + { + "epoch": 0.3172732243757215, + "grad_norm": 0.17862841486930847, + "learning_rate": 7.743381638982148e-05, + "loss": 1.051, + "step": 5222 + }, + { + "epoch": 0.317333981408348, + "grad_norm": 2.0692288875579834, + "learning_rate": 7.742581271511192e-05, + "loss": 1.0847, + "step": 5223 + }, + { + "epoch": 0.31739473844097454, + "grad_norm": 0.20793701708316803, + "learning_rate": 7.741780803510391e-05, + "loss": 1.1527, + "step": 5224 + }, + { + "epoch": 0.3174554954736011, + "grad_norm": 0.26439404487609863, + "learning_rate": 7.740980235009084e-05, + "loss": 1.1469, + "step": 5225 + }, + { + "epoch": 0.3175162525062276, + "grad_norm": 0.16334249079227448, + "learning_rate": 7.740179566036617e-05, + "loss": 1.0685, + "step": 5226 + }, + { + "epoch": 0.31757700953885415, + "grad_norm": 0.17709150910377502, + "learning_rate": 7.739378796622338e-05, + "loss": 1.0519, + "step": 5227 + }, + { + "epoch": 0.31763776657148063, + "grad_norm": 0.19160524010658264, + "learning_rate": 7.738577926795601e-05, + "loss": 1.0989, + "step": 5228 + }, + { + "epoch": 0.31769852360410716, + "grad_norm": 0.20437699556350708, + "learning_rate": 7.737776956585759e-05, + "loss": 1.1257, + "step": 5229 + }, + { + "epoch": 0.3177592806367337, + "grad_norm": 0.2606956958770752, + "learning_rate": 7.736975886022173e-05, + "loss": 1.0172, + "step": 5230 + }, + { + "epoch": 0.31782003766936023, + "grad_norm": 0.20450285077095032, + "learning_rate": 7.736174715134208e-05, + "loss": 1.0722, + "step": 5231 + }, + { + "epoch": 0.31788079470198677, + "grad_norm": 0.38969486951828003, + "learning_rate": 7.735373443951229e-05, + "loss": 1.0761, + "step": 5232 + }, + { + "epoch": 0.3179415517346133, + "grad_norm": 0.3062594532966614, + "learning_rate": 7.734572072502609e-05, + "loss": 1.158, + "step": 5233 + }, + { + "epoch": 0.3180023087672398, + "grad_norm": 1.2661969661712646, + "learning_rate": 7.733770600817721e-05, + "loss": 1.089, + "step": 5234 + }, + { + "epoch": 0.3180630657998663, + "grad_norm": 0.36462631821632385, + "learning_rate": 7.732969028925943e-05, + "loss": 1.0699, + "step": 5235 + }, + { + "epoch": 0.31812382283249285, + "grad_norm": 0.22891183197498322, + "learning_rate": 7.732167356856655e-05, + "loss": 1.2308, + "step": 5236 + }, + { + "epoch": 0.3181845798651194, + "grad_norm": 0.26444292068481445, + "learning_rate": 7.731365584639248e-05, + "loss": 1.1636, + "step": 5237 + }, + { + "epoch": 0.3182453368977459, + "grad_norm": 0.871006429195404, + "learning_rate": 7.730563712303104e-05, + "loss": 1.1093, + "step": 5238 + }, + { + "epoch": 0.31830609393037246, + "grad_norm": 0.817584216594696, + "learning_rate": 7.729761739877622e-05, + "loss": 1.0678, + "step": 5239 + }, + { + "epoch": 0.318366850962999, + "grad_norm": 0.2338828444480896, + "learning_rate": 7.728959667392195e-05, + "loss": 1.1214, + "step": 5240 + }, + { + "epoch": 0.3184276079956255, + "grad_norm": 0.18441972136497498, + "learning_rate": 7.728157494876223e-05, + "loss": 1.0756, + "step": 5241 + }, + { + "epoch": 0.318488365028252, + "grad_norm": 0.974688708782196, + "learning_rate": 7.727355222359112e-05, + "loss": 1.1685, + "step": 5242 + }, + { + "epoch": 0.31854912206087854, + "grad_norm": 0.925870954990387, + "learning_rate": 7.726552849870267e-05, + "loss": 1.202, + "step": 5243 + }, + { + "epoch": 0.3186098790935051, + "grad_norm": 0.3290261924266815, + "learning_rate": 7.7257503774391e-05, + "loss": 1.2505, + "step": 5244 + }, + { + "epoch": 0.3186706361261316, + "grad_norm": 0.4470921754837036, + "learning_rate": 7.724947805095027e-05, + "loss": 1.0144, + "step": 5245 + }, + { + "epoch": 0.31873139315875815, + "grad_norm": 0.1670859009027481, + "learning_rate": 7.724145132867466e-05, + "loss": 1.0608, + "step": 5246 + }, + { + "epoch": 0.3187921501913846, + "grad_norm": 0.2043517529964447, + "learning_rate": 7.723342360785837e-05, + "loss": 1.0513, + "step": 5247 + }, + { + "epoch": 0.31885290722401116, + "grad_norm": 0.3656461238861084, + "learning_rate": 7.722539488879569e-05, + "loss": 1.2799, + "step": 5248 + }, + { + "epoch": 0.3189136642566377, + "grad_norm": 0.3164917528629303, + "learning_rate": 7.721736517178089e-05, + "loss": 1.0747, + "step": 5249 + }, + { + "epoch": 0.31897442128926423, + "grad_norm": 0.21212932467460632, + "learning_rate": 7.72093344571083e-05, + "loss": 1.0839, + "step": 5250 + }, + { + "epoch": 0.31903517832189077, + "grad_norm": 0.20034246146678925, + "learning_rate": 7.72013027450723e-05, + "loss": 1.0347, + "step": 5251 + }, + { + "epoch": 0.3190959353545173, + "grad_norm": 0.2337179332971573, + "learning_rate": 7.71932700359673e-05, + "loss": 1.1808, + "step": 5252 + }, + { + "epoch": 0.31915669238714384, + "grad_norm": 0.2311897724866867, + "learning_rate": 7.718523633008772e-05, + "loss": 1.1587, + "step": 5253 + }, + { + "epoch": 0.3192174494197703, + "grad_norm": 0.2944612503051758, + "learning_rate": 7.717720162772806e-05, + "loss": 1.1117, + "step": 5254 + }, + { + "epoch": 0.31927820645239685, + "grad_norm": 0.701381266117096, + "learning_rate": 7.716916592918281e-05, + "loss": 1.2422, + "step": 5255 + }, + { + "epoch": 0.3193389634850234, + "grad_norm": 0.2760903537273407, + "learning_rate": 7.716112923474654e-05, + "loss": 1.1552, + "step": 5256 + }, + { + "epoch": 0.3193997205176499, + "grad_norm": 0.41636353731155396, + "learning_rate": 7.715309154471382e-05, + "loss": 1.271, + "step": 5257 + }, + { + "epoch": 0.31946047755027646, + "grad_norm": 0.2917998433113098, + "learning_rate": 7.714505285937928e-05, + "loss": 1.0108, + "step": 5258 + }, + { + "epoch": 0.319521234582903, + "grad_norm": 0.3211327791213989, + "learning_rate": 7.713701317903759e-05, + "loss": 1.1595, + "step": 5259 + }, + { + "epoch": 0.31958199161552947, + "grad_norm": 0.45367148518562317, + "learning_rate": 7.712897250398343e-05, + "loss": 1.0782, + "step": 5260 + }, + { + "epoch": 0.319642748648156, + "grad_norm": 0.25913360714912415, + "learning_rate": 7.712093083451153e-05, + "loss": 1.1743, + "step": 5261 + }, + { + "epoch": 0.31970350568078254, + "grad_norm": 0.30565401911735535, + "learning_rate": 7.711288817091669e-05, + "loss": 1.0949, + "step": 5262 + }, + { + "epoch": 0.3197642627134091, + "grad_norm": 0.2851046621799469, + "learning_rate": 7.710484451349367e-05, + "loss": 1.0957, + "step": 5263 + }, + { + "epoch": 0.3198250197460356, + "grad_norm": 0.23404426872730255, + "learning_rate": 7.709679986253735e-05, + "loss": 1.0803, + "step": 5264 + }, + { + "epoch": 0.31988577677866215, + "grad_norm": 10.596319198608398, + "learning_rate": 7.708875421834258e-05, + "loss": 1.0643, + "step": 5265 + }, + { + "epoch": 0.3199465338112887, + "grad_norm": 0.5411157011985779, + "learning_rate": 7.70807075812043e-05, + "loss": 1.2387, + "step": 5266 + }, + { + "epoch": 0.32000729084391516, + "grad_norm": 0.17123252153396606, + "learning_rate": 7.707265995141744e-05, + "loss": 1.1124, + "step": 5267 + }, + { + "epoch": 0.3200680478765417, + "grad_norm": 0.239118292927742, + "learning_rate": 7.7064611329277e-05, + "loss": 1.1048, + "step": 5268 + }, + { + "epoch": 0.32012880490916823, + "grad_norm": 0.3169063329696655, + "learning_rate": 7.705656171507802e-05, + "loss": 1.0552, + "step": 5269 + }, + { + "epoch": 0.32018956194179476, + "grad_norm": 0.21243150532245636, + "learning_rate": 7.70485111091155e-05, + "loss": 1.1468, + "step": 5270 + }, + { + "epoch": 0.3202503189744213, + "grad_norm": 0.3541998267173767, + "learning_rate": 7.704045951168461e-05, + "loss": 1.098, + "step": 5271 + }, + { + "epoch": 0.32031107600704783, + "grad_norm": 0.2529548704624176, + "learning_rate": 7.703240692308044e-05, + "loss": 1.1787, + "step": 5272 + }, + { + "epoch": 0.32037183303967437, + "grad_norm": 0.24251440167427063, + "learning_rate": 7.702435334359817e-05, + "loss": 1.1555, + "step": 5273 + }, + { + "epoch": 0.32043259007230085, + "grad_norm": 0.29362478852272034, + "learning_rate": 7.7016298773533e-05, + "loss": 1.0919, + "step": 5274 + }, + { + "epoch": 0.3204933471049274, + "grad_norm": 3.6648120880126953, + "learning_rate": 7.700824321318019e-05, + "loss": 1.0862, + "step": 5275 + }, + { + "epoch": 0.3205541041375539, + "grad_norm": 0.24022914469242096, + "learning_rate": 7.7000186662835e-05, + "loss": 1.1075, + "step": 5276 + }, + { + "epoch": 0.32061486117018045, + "grad_norm": 0.33304280042648315, + "learning_rate": 7.699212912279274e-05, + "loss": 1.0333, + "step": 5277 + }, + { + "epoch": 0.320675618202807, + "grad_norm": 0.555258572101593, + "learning_rate": 7.698407059334877e-05, + "loss": 1.2503, + "step": 5278 + }, + { + "epoch": 0.3207363752354335, + "grad_norm": 0.32734206318855286, + "learning_rate": 7.697601107479846e-05, + "loss": 1.1709, + "step": 5279 + }, + { + "epoch": 0.32079713226806, + "grad_norm": 0.22176554799079895, + "learning_rate": 7.696795056743727e-05, + "loss": 1.1616, + "step": 5280 + }, + { + "epoch": 0.32085788930068654, + "grad_norm": 0.333980530500412, + "learning_rate": 7.695988907156066e-05, + "loss": 1.2435, + "step": 5281 + }, + { + "epoch": 0.3209186463333131, + "grad_norm": 3.6714234352111816, + "learning_rate": 7.695182658746407e-05, + "loss": 1.1436, + "step": 5282 + }, + { + "epoch": 0.3209794033659396, + "grad_norm": 0.35980966687202454, + "learning_rate": 7.694376311544307e-05, + "loss": 1.1879, + "step": 5283 + }, + { + "epoch": 0.32104016039856614, + "grad_norm": 0.3152298331260681, + "learning_rate": 7.693569865579325e-05, + "loss": 1.1265, + "step": 5284 + }, + { + "epoch": 0.3211009174311927, + "grad_norm": 0.2907601594924927, + "learning_rate": 7.692763320881015e-05, + "loss": 1.2445, + "step": 5285 + }, + { + "epoch": 0.3211616744638192, + "grad_norm": 0.34247633814811707, + "learning_rate": 7.691956677478948e-05, + "loss": 1.2353, + "step": 5286 + }, + { + "epoch": 0.3212224314964457, + "grad_norm": 0.5046535730361938, + "learning_rate": 7.691149935402689e-05, + "loss": 1.0486, + "step": 5287 + }, + { + "epoch": 0.32128318852907223, + "grad_norm": 0.25887230038642883, + "learning_rate": 7.690343094681807e-05, + "loss": 1.1557, + "step": 5288 + }, + { + "epoch": 0.32134394556169876, + "grad_norm": 0.416591078042984, + "learning_rate": 7.68953615534588e-05, + "loss": 0.9994, + "step": 5289 + }, + { + "epoch": 0.3214047025943253, + "grad_norm": 0.25751540064811707, + "learning_rate": 7.688729117424484e-05, + "loss": 1.0393, + "step": 5290 + }, + { + "epoch": 0.32146545962695183, + "grad_norm": 0.447074294090271, + "learning_rate": 7.687921980947206e-05, + "loss": 1.0647, + "step": 5291 + }, + { + "epoch": 0.32152621665957837, + "grad_norm": 0.47063034772872925, + "learning_rate": 7.687114745943624e-05, + "loss": 1.2857, + "step": 5292 + }, + { + "epoch": 0.32158697369220485, + "grad_norm": 0.28080108761787415, + "learning_rate": 7.686307412443334e-05, + "loss": 1.163, + "step": 5293 + }, + { + "epoch": 0.3216477307248314, + "grad_norm": 0.44447919726371765, + "learning_rate": 7.685499980475927e-05, + "loss": 1.1069, + "step": 5294 + }, + { + "epoch": 0.3217084877574579, + "grad_norm": 0.2320144921541214, + "learning_rate": 7.684692450070996e-05, + "loss": 1.0529, + "step": 5295 + }, + { + "epoch": 0.32176924479008445, + "grad_norm": 0.24997901916503906, + "learning_rate": 7.683884821258147e-05, + "loss": 1.0344, + "step": 5296 + }, + { + "epoch": 0.321830001822711, + "grad_norm": 2.0971434116363525, + "learning_rate": 7.68307709406698e-05, + "loss": 1.056, + "step": 5297 + }, + { + "epoch": 0.3218907588553375, + "grad_norm": 0.22199587523937225, + "learning_rate": 7.682269268527104e-05, + "loss": 1.0546, + "step": 5298 + }, + { + "epoch": 0.32195151588796406, + "grad_norm": 0.35019007325172424, + "learning_rate": 7.681461344668131e-05, + "loss": 1.1988, + "step": 5299 + }, + { + "epoch": 0.32201227292059054, + "grad_norm": 0.3319653272628784, + "learning_rate": 7.680653322519671e-05, + "loss": 1.0489, + "step": 5300 + }, + { + "epoch": 0.32207302995321707, + "grad_norm": 0.22761480510234833, + "learning_rate": 7.679845202111347e-05, + "loss": 1.1788, + "step": 5301 + }, + { + "epoch": 0.3221337869858436, + "grad_norm": 0.37962260842323303, + "learning_rate": 7.67903698347278e-05, + "loss": 1.1205, + "step": 5302 + }, + { + "epoch": 0.32219454401847014, + "grad_norm": 0.19670340418815613, + "learning_rate": 7.678228666633592e-05, + "loss": 1.0474, + "step": 5303 + }, + { + "epoch": 0.3222553010510967, + "grad_norm": 0.6671786904335022, + "learning_rate": 7.677420251623418e-05, + "loss": 1.1288, + "step": 5304 + }, + { + "epoch": 0.3223160580837232, + "grad_norm": 0.20764663815498352, + "learning_rate": 7.676611738471882e-05, + "loss": 1.0858, + "step": 5305 + }, + { + "epoch": 0.3223768151163497, + "grad_norm": 0.2705416977405548, + "learning_rate": 7.675803127208629e-05, + "loss": 1.051, + "step": 5306 + }, + { + "epoch": 0.3224375721489762, + "grad_norm": 0.18956951797008514, + "learning_rate": 7.674994417863295e-05, + "loss": 1.0901, + "step": 5307 + }, + { + "epoch": 0.32249832918160276, + "grad_norm": 0.6854101419448853, + "learning_rate": 7.674185610465523e-05, + "loss": 1.1872, + "step": 5308 + }, + { + "epoch": 0.3225590862142293, + "grad_norm": 0.23262162506580353, + "learning_rate": 7.673376705044962e-05, + "loss": 1.0796, + "step": 5309 + }, + { + "epoch": 0.32261984324685583, + "grad_norm": 0.22977684438228607, + "learning_rate": 7.67256770163126e-05, + "loss": 1.0805, + "step": 5310 + }, + { + "epoch": 0.32268060027948237, + "grad_norm": 0.2594844698905945, + "learning_rate": 7.671758600254071e-05, + "loss": 1.1586, + "step": 5311 + }, + { + "epoch": 0.3227413573121089, + "grad_norm": 1.9189934730529785, + "learning_rate": 7.670949400943054e-05, + "loss": 1.1065, + "step": 5312 + }, + { + "epoch": 0.3228021143447354, + "grad_norm": 0.305322527885437, + "learning_rate": 7.670140103727871e-05, + "loss": 1.2237, + "step": 5313 + }, + { + "epoch": 0.3228628713773619, + "grad_norm": 0.215011864900589, + "learning_rate": 7.66933070863819e-05, + "loss": 1.093, + "step": 5314 + }, + { + "epoch": 0.32292362840998845, + "grad_norm": 0.24171243607997894, + "learning_rate": 7.668521215703672e-05, + "loss": 1.0555, + "step": 5315 + }, + { + "epoch": 0.322984385442615, + "grad_norm": 0.20692235231399536, + "learning_rate": 7.667711624953993e-05, + "loss": 1.0623, + "step": 5316 + }, + { + "epoch": 0.3230451424752415, + "grad_norm": 0.2545945644378662, + "learning_rate": 7.66690193641883e-05, + "loss": 1.0308, + "step": 5317 + }, + { + "epoch": 0.32310589950786806, + "grad_norm": 0.39406195282936096, + "learning_rate": 7.66609215012786e-05, + "loss": 1.1188, + "step": 5318 + }, + { + "epoch": 0.32316665654049453, + "grad_norm": 1.1065839529037476, + "learning_rate": 7.665282266110768e-05, + "loss": 1.0721, + "step": 5319 + }, + { + "epoch": 0.32322741357312107, + "grad_norm": 0.28751951456069946, + "learning_rate": 7.664472284397239e-05, + "loss": 1.1097, + "step": 5320 + }, + { + "epoch": 0.3232881706057476, + "grad_norm": 0.5218733549118042, + "learning_rate": 7.663662205016964e-05, + "loss": 1.1094, + "step": 5321 + }, + { + "epoch": 0.32334892763837414, + "grad_norm": 0.36970657110214233, + "learning_rate": 7.662852027999636e-05, + "loss": 1.2394, + "step": 5322 + }, + { + "epoch": 0.3234096846710007, + "grad_norm": 1.4099067449569702, + "learning_rate": 7.662041753374949e-05, + "loss": 1.0483, + "step": 5323 + }, + { + "epoch": 0.3234704417036272, + "grad_norm": 0.22628851234912872, + "learning_rate": 7.661231381172611e-05, + "loss": 1.0699, + "step": 5324 + }, + { + "epoch": 0.32353119873625374, + "grad_norm": 0.25925686955451965, + "learning_rate": 7.660420911422321e-05, + "loss": 1.2392, + "step": 5325 + }, + { + "epoch": 0.3235919557688802, + "grad_norm": 0.21859438717365265, + "learning_rate": 7.659610344153787e-05, + "loss": 1.2233, + "step": 5326 + }, + { + "epoch": 0.32365271280150676, + "grad_norm": 0.19066980481147766, + "learning_rate": 7.658799679396724e-05, + "loss": 1.0977, + "step": 5327 + }, + { + "epoch": 0.3237134698341333, + "grad_norm": 0.2220623940229416, + "learning_rate": 7.657988917180843e-05, + "loss": 1.1401, + "step": 5328 + }, + { + "epoch": 0.32377422686675983, + "grad_norm": 0.6856718063354492, + "learning_rate": 7.657178057535866e-05, + "loss": 1.4118, + "step": 5329 + }, + { + "epoch": 0.32383498389938636, + "grad_norm": 0.44474828243255615, + "learning_rate": 7.656367100491512e-05, + "loss": 1.0863, + "step": 5330 + }, + { + "epoch": 0.3238957409320129, + "grad_norm": 0.19305351376533508, + "learning_rate": 7.65555604607751e-05, + "loss": 1.1031, + "step": 5331 + }, + { + "epoch": 0.32395649796463943, + "grad_norm": 0.37008732557296753, + "learning_rate": 7.654744894323586e-05, + "loss": 1.1106, + "step": 5332 + }, + { + "epoch": 0.3240172549972659, + "grad_norm": 0.2740049362182617, + "learning_rate": 7.653933645259475e-05, + "loss": 1.1353, + "step": 5333 + }, + { + "epoch": 0.32407801202989245, + "grad_norm": 0.32392963767051697, + "learning_rate": 7.653122298914915e-05, + "loss": 1.094, + "step": 5334 + }, + { + "epoch": 0.324138769062519, + "grad_norm": 0.2343919277191162, + "learning_rate": 7.652310855319643e-05, + "loss": 1.1411, + "step": 5335 + }, + { + "epoch": 0.3241995260951455, + "grad_norm": 0.42748695611953735, + "learning_rate": 7.651499314503403e-05, + "loss": 1.1092, + "step": 5336 + }, + { + "epoch": 0.32426028312777205, + "grad_norm": 0.22072601318359375, + "learning_rate": 7.650687676495945e-05, + "loss": 1.0723, + "step": 5337 + }, + { + "epoch": 0.3243210401603986, + "grad_norm": 0.21634657680988312, + "learning_rate": 7.649875941327018e-05, + "loss": 1.0632, + "step": 5338 + }, + { + "epoch": 0.32438179719302507, + "grad_norm": 0.2249966561794281, + "learning_rate": 7.649064109026374e-05, + "loss": 1.1345, + "step": 5339 + }, + { + "epoch": 0.3244425542256516, + "grad_norm": 0.28644341230392456, + "learning_rate": 7.648252179623775e-05, + "loss": 1.1274, + "step": 5340 + }, + { + "epoch": 0.32450331125827814, + "grad_norm": 0.19570733606815338, + "learning_rate": 7.647440153148979e-05, + "loss": 1.1315, + "step": 5341 + }, + { + "epoch": 0.3245640682909047, + "grad_norm": 0.20028486847877502, + "learning_rate": 7.646628029631754e-05, + "loss": 1.1329, + "step": 5342 + }, + { + "epoch": 0.3246248253235312, + "grad_norm": 0.48429426550865173, + "learning_rate": 7.645815809101866e-05, + "loss": 1.3149, + "step": 5343 + }, + { + "epoch": 0.32468558235615774, + "grad_norm": 0.17645619809627533, + "learning_rate": 7.645003491589089e-05, + "loss": 1.0582, + "step": 5344 + }, + { + "epoch": 0.3247463393887843, + "grad_norm": 0.2790333330631256, + "learning_rate": 7.644191077123197e-05, + "loss": 1.2112, + "step": 5345 + }, + { + "epoch": 0.32480709642141076, + "grad_norm": 0.9742593169212341, + "learning_rate": 7.64337856573397e-05, + "loss": 1.189, + "step": 5346 + }, + { + "epoch": 0.3248678534540373, + "grad_norm": 0.2916494607925415, + "learning_rate": 7.64256595745119e-05, + "loss": 1.0817, + "step": 5347 + }, + { + "epoch": 0.3249286104866638, + "grad_norm": 0.25182244181632996, + "learning_rate": 7.641753252304646e-05, + "loss": 1.1205, + "step": 5348 + }, + { + "epoch": 0.32498936751929036, + "grad_norm": 0.24194931983947754, + "learning_rate": 7.640940450324124e-05, + "loss": 1.1919, + "step": 5349 + }, + { + "epoch": 0.3250501245519169, + "grad_norm": 0.48090827465057373, + "learning_rate": 7.640127551539419e-05, + "loss": 1.1142, + "step": 5350 + }, + { + "epoch": 0.32511088158454343, + "grad_norm": 0.25265631079673767, + "learning_rate": 7.63931455598033e-05, + "loss": 1.1897, + "step": 5351 + }, + { + "epoch": 0.3251716386171699, + "grad_norm": 0.22122783958911896, + "learning_rate": 7.638501463676655e-05, + "loss": 1.1024, + "step": 5352 + }, + { + "epoch": 0.32523239564979645, + "grad_norm": 0.19860023260116577, + "learning_rate": 7.6376882746582e-05, + "loss": 1.0931, + "step": 5353 + }, + { + "epoch": 0.325293152682423, + "grad_norm": 0.19961878657341003, + "learning_rate": 7.63687498895477e-05, + "loss": 1.1074, + "step": 5354 + }, + { + "epoch": 0.3253539097150495, + "grad_norm": 0.27491065859794617, + "learning_rate": 7.636061606596178e-05, + "loss": 1.0655, + "step": 5355 + }, + { + "epoch": 0.32541466674767605, + "grad_norm": 0.1800784468650818, + "learning_rate": 7.635248127612238e-05, + "loss": 1.0611, + "step": 5356 + }, + { + "epoch": 0.3254754237803026, + "grad_norm": 0.34458163380622864, + "learning_rate": 7.634434552032768e-05, + "loss": 1.1013, + "step": 5357 + }, + { + "epoch": 0.3255361808129291, + "grad_norm": 0.8056153655052185, + "learning_rate": 7.63362087988759e-05, + "loss": 1.0488, + "step": 5358 + }, + { + "epoch": 0.3255969378455556, + "grad_norm": 0.21309220790863037, + "learning_rate": 7.632807111206531e-05, + "loss": 1.1189, + "step": 5359 + }, + { + "epoch": 0.32565769487818214, + "grad_norm": 0.15306927263736725, + "learning_rate": 7.631993246019418e-05, + "loss": 1.0681, + "step": 5360 + }, + { + "epoch": 0.32571845191080867, + "grad_norm": 0.17501188814640045, + "learning_rate": 7.631179284356085e-05, + "loss": 1.076, + "step": 5361 + }, + { + "epoch": 0.3257792089434352, + "grad_norm": 0.2582508623600006, + "learning_rate": 7.630365226246365e-05, + "loss": 1.0556, + "step": 5362 + }, + { + "epoch": 0.32583996597606174, + "grad_norm": 0.23859822750091553, + "learning_rate": 7.6295510717201e-05, + "loss": 1.1533, + "step": 5363 + }, + { + "epoch": 0.3259007230086883, + "grad_norm": 0.2562359869480133, + "learning_rate": 7.628736820807132e-05, + "loss": 1.1124, + "step": 5364 + }, + { + "epoch": 0.32596148004131476, + "grad_norm": 0.2441672831773758, + "learning_rate": 7.627922473537307e-05, + "loss": 1.1458, + "step": 5365 + }, + { + "epoch": 0.3260222370739413, + "grad_norm": 0.19998250901699066, + "learning_rate": 7.627108029940477e-05, + "loss": 1.0775, + "step": 5366 + }, + { + "epoch": 0.3260829941065678, + "grad_norm": 0.24868734180927277, + "learning_rate": 7.626293490046493e-05, + "loss": 1.1831, + "step": 5367 + }, + { + "epoch": 0.32614375113919436, + "grad_norm": 0.30341723561286926, + "learning_rate": 7.625478853885215e-05, + "loss": 1.0389, + "step": 5368 + }, + { + "epoch": 0.3262045081718209, + "grad_norm": 0.18511201441287994, + "learning_rate": 7.6246641214865e-05, + "loss": 1.0271, + "step": 5369 + }, + { + "epoch": 0.32626526520444743, + "grad_norm": 0.19850307703018188, + "learning_rate": 7.623849292880216e-05, + "loss": 1.1177, + "step": 5370 + }, + { + "epoch": 0.32632602223707396, + "grad_norm": 0.2819867432117462, + "learning_rate": 7.623034368096229e-05, + "loss": 1.0064, + "step": 5371 + }, + { + "epoch": 0.32638677926970044, + "grad_norm": 0.20224255323410034, + "learning_rate": 7.62221934716441e-05, + "loss": 1.0566, + "step": 5372 + }, + { + "epoch": 0.326447536302327, + "grad_norm": 1.1778730154037476, + "learning_rate": 7.621404230114632e-05, + "loss": 1.0608, + "step": 5373 + }, + { + "epoch": 0.3265082933349535, + "grad_norm": 0.24865086376667023, + "learning_rate": 7.62058901697678e-05, + "loss": 1.0908, + "step": 5374 + }, + { + "epoch": 0.32656905036758005, + "grad_norm": 0.20016466081142426, + "learning_rate": 7.619773707780727e-05, + "loss": 1.075, + "step": 5375 + }, + { + "epoch": 0.3266298074002066, + "grad_norm": 0.1663936823606491, + "learning_rate": 7.618958302556363e-05, + "loss": 1.0895, + "step": 5376 + }, + { + "epoch": 0.3266905644328331, + "grad_norm": 1.247120976448059, + "learning_rate": 7.618142801333576e-05, + "loss": 1.1749, + "step": 5377 + }, + { + "epoch": 0.32675132146545965, + "grad_norm": 0.5096200704574585, + "learning_rate": 7.61732720414226e-05, + "loss": 1.2171, + "step": 5378 + }, + { + "epoch": 0.32681207849808613, + "grad_norm": 0.23748070001602173, + "learning_rate": 7.616511511012308e-05, + "loss": 1.0663, + "step": 5379 + }, + { + "epoch": 0.32687283553071267, + "grad_norm": 0.26105207204818726, + "learning_rate": 7.615695721973621e-05, + "loss": 1.112, + "step": 5380 + }, + { + "epoch": 0.3269335925633392, + "grad_norm": 1.9382989406585693, + "learning_rate": 7.614879837056103e-05, + "loss": 1.0285, + "step": 5381 + }, + { + "epoch": 0.32699434959596574, + "grad_norm": 0.21667708456516266, + "learning_rate": 7.614063856289658e-05, + "loss": 1.0756, + "step": 5382 + }, + { + "epoch": 0.3270551066285923, + "grad_norm": 0.20560555160045624, + "learning_rate": 7.613247779704197e-05, + "loss": 1.0617, + "step": 5383 + }, + { + "epoch": 0.3271158636612188, + "grad_norm": 0.19508345425128937, + "learning_rate": 7.612431607329635e-05, + "loss": 1.0577, + "step": 5384 + }, + { + "epoch": 0.3271766206938453, + "grad_norm": 0.2453996241092682, + "learning_rate": 7.611615339195886e-05, + "loss": 1.0805, + "step": 5385 + }, + { + "epoch": 0.3272373777264718, + "grad_norm": 0.5964996218681335, + "learning_rate": 7.610798975332871e-05, + "loss": 1.1166, + "step": 5386 + }, + { + "epoch": 0.32729813475909836, + "grad_norm": 0.3189411461353302, + "learning_rate": 7.609982515770516e-05, + "loss": 1.1938, + "step": 5387 + }, + { + "epoch": 0.3273588917917249, + "grad_norm": 0.23660995066165924, + "learning_rate": 7.609165960538748e-05, + "loss": 1.1624, + "step": 5388 + }, + { + "epoch": 0.32741964882435143, + "grad_norm": 0.2508578598499298, + "learning_rate": 7.608349309667498e-05, + "loss": 1.2191, + "step": 5389 + }, + { + "epoch": 0.32748040585697796, + "grad_norm": 0.24006661772727966, + "learning_rate": 7.607532563186696e-05, + "loss": 1.1297, + "step": 5390 + }, + { + "epoch": 0.3275411628896045, + "grad_norm": 0.23372696340084076, + "learning_rate": 7.606715721126288e-05, + "loss": 1.0831, + "step": 5391 + }, + { + "epoch": 0.327601919922231, + "grad_norm": 0.18675574660301208, + "learning_rate": 7.60589878351621e-05, + "loss": 1.0956, + "step": 5392 + }, + { + "epoch": 0.3276626769548575, + "grad_norm": 1.1280027627944946, + "learning_rate": 7.605081750386407e-05, + "loss": 1.137, + "step": 5393 + }, + { + "epoch": 0.32772343398748405, + "grad_norm": 0.31657490134239197, + "learning_rate": 7.604264621766833e-05, + "loss": 1.1182, + "step": 5394 + }, + { + "epoch": 0.3277841910201106, + "grad_norm": 0.16286912560462952, + "learning_rate": 7.603447397687432e-05, + "loss": 1.0698, + "step": 5395 + }, + { + "epoch": 0.3278449480527371, + "grad_norm": 2.87026047706604, + "learning_rate": 7.602630078178164e-05, + "loss": 1.0514, + "step": 5396 + }, + { + "epoch": 0.32790570508536365, + "grad_norm": 0.212200328707695, + "learning_rate": 7.601812663268988e-05, + "loss": 1.0462, + "step": 5397 + }, + { + "epoch": 0.32796646211799013, + "grad_norm": 0.1862175613641739, + "learning_rate": 7.600995152989865e-05, + "loss": 1.0784, + "step": 5398 + }, + { + "epoch": 0.32802721915061667, + "grad_norm": 0.2752746343612671, + "learning_rate": 7.600177547370763e-05, + "loss": 1.1307, + "step": 5399 + }, + { + "epoch": 0.3280879761832432, + "grad_norm": 0.19000092148780823, + "learning_rate": 7.599359846441651e-05, + "loss": 1.077, + "step": 5400 + }, + { + "epoch": 0.32814873321586974, + "grad_norm": 0.24259427189826965, + "learning_rate": 7.5985420502325e-05, + "loss": 1.0704, + "step": 5401 + }, + { + "epoch": 0.32820949024849627, + "grad_norm": 0.17529001832008362, + "learning_rate": 7.597724158773289e-05, + "loss": 1.0727, + "step": 5402 + }, + { + "epoch": 0.3282702472811228, + "grad_norm": 0.2693842351436615, + "learning_rate": 7.596906172093995e-05, + "loss": 1.0949, + "step": 5403 + }, + { + "epoch": 0.32833100431374934, + "grad_norm": 0.32453179359436035, + "learning_rate": 7.596088090224604e-05, + "loss": 1.1321, + "step": 5404 + }, + { + "epoch": 0.3283917613463758, + "grad_norm": 2.8437135219573975, + "learning_rate": 7.595269913195104e-05, + "loss": 1.1395, + "step": 5405 + }, + { + "epoch": 0.32845251837900236, + "grad_norm": 0.48856520652770996, + "learning_rate": 7.594451641035482e-05, + "loss": 1.0726, + "step": 5406 + }, + { + "epoch": 0.3285132754116289, + "grad_norm": 0.4369836747646332, + "learning_rate": 7.593633273775734e-05, + "loss": 1.0859, + "step": 5407 + }, + { + "epoch": 0.3285740324442554, + "grad_norm": 0.3082618713378906, + "learning_rate": 7.592814811445856e-05, + "loss": 1.1261, + "step": 5408 + }, + { + "epoch": 0.32863478947688196, + "grad_norm": 0.3599332869052887, + "learning_rate": 7.591996254075851e-05, + "loss": 1.0675, + "step": 5409 + }, + { + "epoch": 0.3286955465095085, + "grad_norm": 9.183272361755371, + "learning_rate": 7.591177601695722e-05, + "loss": 1.0812, + "step": 5410 + }, + { + "epoch": 0.328756303542135, + "grad_norm": 0.45782169699668884, + "learning_rate": 7.59035885433548e-05, + "loss": 1.0528, + "step": 5411 + }, + { + "epoch": 0.3288170605747615, + "grad_norm": 0.20099540054798126, + "learning_rate": 7.58954001202513e-05, + "loss": 1.0612, + "step": 5412 + }, + { + "epoch": 0.32887781760738805, + "grad_norm": 0.19425509870052338, + "learning_rate": 7.588721074794691e-05, + "loss": 1.0635, + "step": 5413 + }, + { + "epoch": 0.3289385746400146, + "grad_norm": 0.2860231399536133, + "learning_rate": 7.58790204267418e-05, + "loss": 1.1287, + "step": 5414 + }, + { + "epoch": 0.3289993316726411, + "grad_norm": 3.791538953781128, + "learning_rate": 7.58708291569362e-05, + "loss": 1.1112, + "step": 5415 + }, + { + "epoch": 0.32906008870526765, + "grad_norm": 0.7450807690620422, + "learning_rate": 7.586263693883036e-05, + "loss": 1.0657, + "step": 5416 + }, + { + "epoch": 0.3291208457378942, + "grad_norm": 0.26344379782676697, + "learning_rate": 7.585444377272456e-05, + "loss": 1.1785, + "step": 5417 + }, + { + "epoch": 0.32918160277052066, + "grad_norm": 0.16956184804439545, + "learning_rate": 7.584624965891913e-05, + "loss": 1.1022, + "step": 5418 + }, + { + "epoch": 0.3292423598031472, + "grad_norm": 0.24892961978912354, + "learning_rate": 7.583805459771443e-05, + "loss": 1.0999, + "step": 5419 + }, + { + "epoch": 0.32930311683577373, + "grad_norm": 0.28406572341918945, + "learning_rate": 7.582985858941083e-05, + "loss": 1.0455, + "step": 5420 + }, + { + "epoch": 0.32936387386840027, + "grad_norm": 0.2557935416698456, + "learning_rate": 7.582166163430877e-05, + "loss": 1.034, + "step": 5421 + }, + { + "epoch": 0.3294246309010268, + "grad_norm": 0.23036473989486694, + "learning_rate": 7.581346373270873e-05, + "loss": 1.1323, + "step": 5422 + }, + { + "epoch": 0.32948538793365334, + "grad_norm": 0.214420348405838, + "learning_rate": 7.580526488491116e-05, + "loss": 1.0166, + "step": 5423 + }, + { + "epoch": 0.3295461449662798, + "grad_norm": 1.8969358205795288, + "learning_rate": 7.579706509121663e-05, + "loss": 1.3916, + "step": 5424 + }, + { + "epoch": 0.32960690199890635, + "grad_norm": 0.22537456452846527, + "learning_rate": 7.57888643519257e-05, + "loss": 1.0505, + "step": 5425 + }, + { + "epoch": 0.3296676590315329, + "grad_norm": 0.2674037218093872, + "learning_rate": 7.578066266733897e-05, + "loss": 1.0969, + "step": 5426 + }, + { + "epoch": 0.3297284160641594, + "grad_norm": 0.1871471256017685, + "learning_rate": 7.577246003775704e-05, + "loss": 1.0819, + "step": 5427 + }, + { + "epoch": 0.32978917309678596, + "grad_norm": 0.26442214846611023, + "learning_rate": 7.576425646348062e-05, + "loss": 1.0404, + "step": 5428 + }, + { + "epoch": 0.3298499301294125, + "grad_norm": 0.18481558561325073, + "learning_rate": 7.575605194481039e-05, + "loss": 1.0364, + "step": 5429 + }, + { + "epoch": 0.32991068716203903, + "grad_norm": 0.29945889115333557, + "learning_rate": 7.574784648204711e-05, + "loss": 1.2723, + "step": 5430 + }, + { + "epoch": 0.3299714441946655, + "grad_norm": 0.21873514354228973, + "learning_rate": 7.573964007549155e-05, + "loss": 1.1329, + "step": 5431 + }, + { + "epoch": 0.33003220122729204, + "grad_norm": 0.2477567344903946, + "learning_rate": 7.573143272544449e-05, + "loss": 1.0353, + "step": 5432 + }, + { + "epoch": 0.3300929582599186, + "grad_norm": 0.17394210398197174, + "learning_rate": 7.572322443220678e-05, + "loss": 1.0556, + "step": 5433 + }, + { + "epoch": 0.3301537152925451, + "grad_norm": 0.4207371175289154, + "learning_rate": 7.571501519607933e-05, + "loss": 1.0389, + "step": 5434 + }, + { + "epoch": 0.33021447232517165, + "grad_norm": 0.26431000232696533, + "learning_rate": 7.570680501736303e-05, + "loss": 1.0858, + "step": 5435 + }, + { + "epoch": 0.3302752293577982, + "grad_norm": 0.7328451871871948, + "learning_rate": 7.56985938963588e-05, + "loss": 1.1315, + "step": 5436 + }, + { + "epoch": 0.3303359863904247, + "grad_norm": 0.4436350464820862, + "learning_rate": 7.569038183336765e-05, + "loss": 1.2578, + "step": 5437 + }, + { + "epoch": 0.3303967434230512, + "grad_norm": 0.2691571116447449, + "learning_rate": 7.56821688286906e-05, + "loss": 1.1593, + "step": 5438 + }, + { + "epoch": 0.33045750045567773, + "grad_norm": 0.3349302411079407, + "learning_rate": 7.567395488262868e-05, + "loss": 1.1897, + "step": 5439 + }, + { + "epoch": 0.33051825748830427, + "grad_norm": 0.317264199256897, + "learning_rate": 7.566573999548297e-05, + "loss": 1.0314, + "step": 5440 + }, + { + "epoch": 0.3305790145209308, + "grad_norm": 0.5453688502311707, + "learning_rate": 7.565752416755461e-05, + "loss": 1.096, + "step": 5441 + }, + { + "epoch": 0.33063977155355734, + "grad_norm": 0.38669684529304504, + "learning_rate": 7.564930739914473e-05, + "loss": 1.1343, + "step": 5442 + }, + { + "epoch": 0.3307005285861839, + "grad_norm": 0.3520776033401489, + "learning_rate": 7.564108969055455e-05, + "loss": 1.1993, + "step": 5443 + }, + { + "epoch": 0.33076128561881035, + "grad_norm": 0.26440250873565674, + "learning_rate": 7.563287104208526e-05, + "loss": 1.1365, + "step": 5444 + }, + { + "epoch": 0.3308220426514369, + "grad_norm": 3.2212727069854736, + "learning_rate": 7.562465145403813e-05, + "loss": 1.1118, + "step": 5445 + }, + { + "epoch": 0.3308827996840634, + "grad_norm": 0.5784273147583008, + "learning_rate": 7.561643092671444e-05, + "loss": 1.1024, + "step": 5446 + }, + { + "epoch": 0.33094355671668996, + "grad_norm": 0.2402225136756897, + "learning_rate": 7.560820946041554e-05, + "loss": 1.0488, + "step": 5447 + }, + { + "epoch": 0.3310043137493165, + "grad_norm": 0.361889123916626, + "learning_rate": 7.559998705544277e-05, + "loss": 1.093, + "step": 5448 + }, + { + "epoch": 0.331065070781943, + "grad_norm": 0.324540913105011, + "learning_rate": 7.559176371209751e-05, + "loss": 1.0251, + "step": 5449 + }, + { + "epoch": 0.33112582781456956, + "grad_norm": 0.2523747980594635, + "learning_rate": 7.558353943068122e-05, + "loss": 1.1881, + "step": 5450 + }, + { + "epoch": 0.33118658484719604, + "grad_norm": 0.5838483572006226, + "learning_rate": 7.557531421149534e-05, + "loss": 1.1305, + "step": 5451 + }, + { + "epoch": 0.3312473418798226, + "grad_norm": 0.17959848046302795, + "learning_rate": 7.556708805484138e-05, + "loss": 1.0234, + "step": 5452 + }, + { + "epoch": 0.3313080989124491, + "grad_norm": 0.3179149925708771, + "learning_rate": 7.555886096102084e-05, + "loss": 1.1699, + "step": 5453 + }, + { + "epoch": 0.33136885594507565, + "grad_norm": 0.40076711773872375, + "learning_rate": 7.555063293033533e-05, + "loss": 1.1375, + "step": 5454 + }, + { + "epoch": 0.3314296129777022, + "grad_norm": 0.23310671746730804, + "learning_rate": 7.554240396308643e-05, + "loss": 1.0573, + "step": 5455 + }, + { + "epoch": 0.3314903700103287, + "grad_norm": 0.23253920674324036, + "learning_rate": 7.553417405957579e-05, + "loss": 1.2095, + "step": 5456 + }, + { + "epoch": 0.3315511270429552, + "grad_norm": 0.43200939893722534, + "learning_rate": 7.552594322010504e-05, + "loss": 1.1532, + "step": 5457 + }, + { + "epoch": 0.33161188407558173, + "grad_norm": 0.16281524300575256, + "learning_rate": 7.551771144497592e-05, + "loss": 1.0819, + "step": 5458 + }, + { + "epoch": 0.33167264110820827, + "grad_norm": 0.32548272609710693, + "learning_rate": 7.550947873449016e-05, + "loss": 1.1975, + "step": 5459 + }, + { + "epoch": 0.3317333981408348, + "grad_norm": 0.4708641767501831, + "learning_rate": 7.550124508894951e-05, + "loss": 1.0779, + "step": 5460 + }, + { + "epoch": 0.33179415517346134, + "grad_norm": 0.15611782670021057, + "learning_rate": 7.54930105086558e-05, + "loss": 1.0853, + "step": 5461 + }, + { + "epoch": 0.33185491220608787, + "grad_norm": 0.25309041142463684, + "learning_rate": 7.548477499391087e-05, + "loss": 1.1209, + "step": 5462 + }, + { + "epoch": 0.3319156692387144, + "grad_norm": 0.31104180216789246, + "learning_rate": 7.547653854501657e-05, + "loss": 1.099, + "step": 5463 + }, + { + "epoch": 0.3319764262713409, + "grad_norm": 0.16987203061580658, + "learning_rate": 7.546830116227484e-05, + "loss": 1.1114, + "step": 5464 + }, + { + "epoch": 0.3320371833039674, + "grad_norm": 0.2852203845977783, + "learning_rate": 7.54600628459876e-05, + "loss": 1.1121, + "step": 5465 + }, + { + "epoch": 0.33209794033659396, + "grad_norm": 0.2020912766456604, + "learning_rate": 7.545182359645684e-05, + "loss": 1.1125, + "step": 5466 + }, + { + "epoch": 0.3321586973692205, + "grad_norm": 0.6708133220672607, + "learning_rate": 7.544358341398457e-05, + "loss": 1.1816, + "step": 5467 + }, + { + "epoch": 0.332219454401847, + "grad_norm": 0.19098488986492157, + "learning_rate": 7.543534229887281e-05, + "loss": 1.0851, + "step": 5468 + }, + { + "epoch": 0.33228021143447356, + "grad_norm": 0.2439623922109604, + "learning_rate": 7.54271002514237e-05, + "loss": 1.3647, + "step": 5469 + }, + { + "epoch": 0.33234096846710004, + "grad_norm": 0.2515058219432831, + "learning_rate": 7.541885727193928e-05, + "loss": 1.1212, + "step": 5470 + }, + { + "epoch": 0.3324017254997266, + "grad_norm": 0.2202656865119934, + "learning_rate": 7.541061336072175e-05, + "loss": 1.0916, + "step": 5471 + }, + { + "epoch": 0.3324624825323531, + "grad_norm": 0.29517635703086853, + "learning_rate": 7.540236851807327e-05, + "loss": 1.0824, + "step": 5472 + }, + { + "epoch": 0.33252323956497964, + "grad_norm": 0.41751599311828613, + "learning_rate": 7.539412274429605e-05, + "loss": 1.0747, + "step": 5473 + }, + { + "epoch": 0.3325839965976062, + "grad_norm": 0.2361290007829666, + "learning_rate": 7.538587603969238e-05, + "loss": 1.0526, + "step": 5474 + }, + { + "epoch": 0.3326447536302327, + "grad_norm": 0.5817698836326599, + "learning_rate": 7.53776284045645e-05, + "loss": 1.0083, + "step": 5475 + }, + { + "epoch": 0.33270551066285925, + "grad_norm": 0.6068136096000671, + "learning_rate": 7.536937983921474e-05, + "loss": 1.1291, + "step": 5476 + }, + { + "epoch": 0.33276626769548573, + "grad_norm": 1.0519905090332031, + "learning_rate": 7.536113034394547e-05, + "loss": 1.0557, + "step": 5477 + }, + { + "epoch": 0.33282702472811226, + "grad_norm": 0.397795706987381, + "learning_rate": 7.535287991905906e-05, + "loss": 1.0813, + "step": 5478 + }, + { + "epoch": 0.3328877817607388, + "grad_norm": 0.9957154393196106, + "learning_rate": 7.534462856485793e-05, + "loss": 1.1204, + "step": 5479 + }, + { + "epoch": 0.33294853879336533, + "grad_norm": 1.0611766576766968, + "learning_rate": 7.533637628164455e-05, + "loss": 1.1566, + "step": 5480 + }, + { + "epoch": 0.33300929582599187, + "grad_norm": 0.19772228598594666, + "learning_rate": 7.532812306972139e-05, + "loss": 1.0916, + "step": 5481 + }, + { + "epoch": 0.3330700528586184, + "grad_norm": 0.47837135195732117, + "learning_rate": 7.531986892939101e-05, + "loss": 1.0749, + "step": 5482 + }, + { + "epoch": 0.33313080989124494, + "grad_norm": 0.6841157674789429, + "learning_rate": 7.531161386095591e-05, + "loss": 1.0983, + "step": 5483 + }, + { + "epoch": 0.3331915669238714, + "grad_norm": 0.4009842276573181, + "learning_rate": 7.530335786471871e-05, + "loss": 1.219, + "step": 5484 + }, + { + "epoch": 0.33325232395649795, + "grad_norm": 0.5157473683357239, + "learning_rate": 7.529510094098207e-05, + "loss": 1.0799, + "step": 5485 + }, + { + "epoch": 0.3333130809891245, + "grad_norm": 0.29484841227531433, + "learning_rate": 7.528684309004859e-05, + "loss": 1.2112, + "step": 5486 + }, + { + "epoch": 0.333373838021751, + "grad_norm": 0.4728744626045227, + "learning_rate": 7.527858431222098e-05, + "loss": 1.1002, + "step": 5487 + }, + { + "epoch": 0.33343459505437756, + "grad_norm": 0.24175560474395752, + "learning_rate": 7.527032460780201e-05, + "loss": 1.0798, + "step": 5488 + }, + { + "epoch": 0.3334953520870041, + "grad_norm": 0.16613490879535675, + "learning_rate": 7.526206397709437e-05, + "loss": 1.1807, + "step": 5489 + }, + { + "epoch": 0.3335561091196306, + "grad_norm": 0.45182284712791443, + "learning_rate": 7.525380242040092e-05, + "loss": 1.0333, + "step": 5490 + }, + { + "epoch": 0.3336168661522571, + "grad_norm": 0.20438359677791595, + "learning_rate": 7.524553993802443e-05, + "loss": 1.2023, + "step": 5491 + }, + { + "epoch": 0.33367762318488364, + "grad_norm": 0.27111056447029114, + "learning_rate": 7.523727653026783e-05, + "loss": 1.1297, + "step": 5492 + }, + { + "epoch": 0.3337383802175102, + "grad_norm": 0.1997470110654831, + "learning_rate": 7.522901219743396e-05, + "loss": 1.1126, + "step": 5493 + }, + { + "epoch": 0.3337991372501367, + "grad_norm": 0.23613186180591583, + "learning_rate": 7.522074693982577e-05, + "loss": 1.1271, + "step": 5494 + }, + { + "epoch": 0.33385989428276325, + "grad_norm": 0.2127677947282791, + "learning_rate": 7.521248075774623e-05, + "loss": 1.0758, + "step": 5495 + }, + { + "epoch": 0.3339206513153898, + "grad_norm": 0.5567682981491089, + "learning_rate": 7.520421365149834e-05, + "loss": 1.1842, + "step": 5496 + }, + { + "epoch": 0.33398140834801626, + "grad_norm": 0.17385333776474, + "learning_rate": 7.519594562138511e-05, + "loss": 1.0198, + "step": 5497 + }, + { + "epoch": 0.3340421653806428, + "grad_norm": 0.44032323360443115, + "learning_rate": 7.518767666770965e-05, + "loss": 1.0542, + "step": 5498 + }, + { + "epoch": 0.33410292241326933, + "grad_norm": 0.16658762097358704, + "learning_rate": 7.5179406790775e-05, + "loss": 1.0504, + "step": 5499 + }, + { + "epoch": 0.33416367944589587, + "grad_norm": 0.2646464705467224, + "learning_rate": 7.517113599088434e-05, + "loss": 1.0133, + "step": 5500 + }, + { + "epoch": 0.3342244364785224, + "grad_norm": 0.18440215289592743, + "learning_rate": 7.516286426834084e-05, + "loss": 1.0325, + "step": 5501 + }, + { + "epoch": 0.33428519351114894, + "grad_norm": 0.23189516365528107, + "learning_rate": 7.515459162344767e-05, + "loss": 1.1642, + "step": 5502 + }, + { + "epoch": 0.3343459505437754, + "grad_norm": 3.810866117477417, + "learning_rate": 7.51463180565081e-05, + "loss": 1.129, + "step": 5503 + }, + { + "epoch": 0.33440670757640195, + "grad_norm": 0.3608061969280243, + "learning_rate": 7.513804356782536e-05, + "loss": 1.0454, + "step": 5504 + }, + { + "epoch": 0.3344674646090285, + "grad_norm": 0.19507181644439697, + "learning_rate": 7.512976815770279e-05, + "loss": 1.0476, + "step": 5505 + }, + { + "epoch": 0.334528221641655, + "grad_norm": 0.263811856508255, + "learning_rate": 7.512149182644371e-05, + "loss": 1.0598, + "step": 5506 + }, + { + "epoch": 0.33458897867428156, + "grad_norm": 0.16296547651290894, + "learning_rate": 7.511321457435145e-05, + "loss": 1.0501, + "step": 5507 + }, + { + "epoch": 0.3346497357069081, + "grad_norm": 0.36544615030288696, + "learning_rate": 7.51049364017295e-05, + "loss": 1.1833, + "step": 5508 + }, + { + "epoch": 0.3347104927395346, + "grad_norm": 0.49034029245376587, + "learning_rate": 7.509665730888125e-05, + "loss": 1.2058, + "step": 5509 + }, + { + "epoch": 0.3347712497721611, + "grad_norm": 0.2619384527206421, + "learning_rate": 7.508837729611016e-05, + "loss": 1.0955, + "step": 5510 + }, + { + "epoch": 0.33483200680478764, + "grad_norm": 0.573919415473938, + "learning_rate": 7.508009636371976e-05, + "loss": 1.0382, + "step": 5511 + }, + { + "epoch": 0.3348927638374142, + "grad_norm": 0.17120429873466492, + "learning_rate": 7.507181451201357e-05, + "loss": 1.1009, + "step": 5512 + }, + { + "epoch": 0.3349535208700407, + "grad_norm": 0.5637596845626831, + "learning_rate": 7.506353174129518e-05, + "loss": 1.198, + "step": 5513 + }, + { + "epoch": 0.33501427790266725, + "grad_norm": 0.5281242728233337, + "learning_rate": 7.505524805186821e-05, + "loss": 1.0468, + "step": 5514 + }, + { + "epoch": 0.3350750349352938, + "grad_norm": 1.283241868019104, + "learning_rate": 7.504696344403626e-05, + "loss": 1.0823, + "step": 5515 + }, + { + "epoch": 0.33513579196792026, + "grad_norm": 0.256847620010376, + "learning_rate": 7.503867791810302e-05, + "loss": 1.202, + "step": 5516 + }, + { + "epoch": 0.3351965490005468, + "grad_norm": 0.7701228857040405, + "learning_rate": 7.503039147437218e-05, + "loss": 1.2079, + "step": 5517 + }, + { + "epoch": 0.33525730603317333, + "grad_norm": 0.38645246624946594, + "learning_rate": 7.502210411314755e-05, + "loss": 1.042, + "step": 5518 + }, + { + "epoch": 0.33531806306579987, + "grad_norm": 0.552549421787262, + "learning_rate": 7.501381583473284e-05, + "loss": 1.0673, + "step": 5519 + }, + { + "epoch": 0.3353788200984264, + "grad_norm": 0.4692688286304474, + "learning_rate": 7.500552663943186e-05, + "loss": 1.1346, + "step": 5520 + }, + { + "epoch": 0.33543957713105294, + "grad_norm": 0.28835394978523254, + "learning_rate": 7.499723652754847e-05, + "loss": 1.1466, + "step": 5521 + }, + { + "epoch": 0.33550033416367947, + "grad_norm": 0.2520778477191925, + "learning_rate": 7.498894549938656e-05, + "loss": 1.057, + "step": 5522 + }, + { + "epoch": 0.33556109119630595, + "grad_norm": 0.21005330979824066, + "learning_rate": 7.498065355525002e-05, + "loss": 1.0739, + "step": 5523 + }, + { + "epoch": 0.3356218482289325, + "grad_norm": 0.7380683422088623, + "learning_rate": 7.497236069544278e-05, + "loss": 1.1288, + "step": 5524 + }, + { + "epoch": 0.335682605261559, + "grad_norm": 0.1789902299642563, + "learning_rate": 7.496406692026884e-05, + "loss": 1.0441, + "step": 5525 + }, + { + "epoch": 0.33574336229418555, + "grad_norm": 0.21991556882858276, + "learning_rate": 7.49557722300322e-05, + "loss": 1.1108, + "step": 5526 + }, + { + "epoch": 0.3358041193268121, + "grad_norm": 0.43132907152175903, + "learning_rate": 7.494747662503691e-05, + "loss": 1.0749, + "step": 5527 + }, + { + "epoch": 0.3358648763594386, + "grad_norm": 0.18612077832221985, + "learning_rate": 7.493918010558704e-05, + "loss": 1.0594, + "step": 5528 + }, + { + "epoch": 0.3359256333920651, + "grad_norm": 0.17746864259243011, + "learning_rate": 7.49308826719867e-05, + "loss": 1.0725, + "step": 5529 + }, + { + "epoch": 0.33598639042469164, + "grad_norm": 0.3089925944805145, + "learning_rate": 7.492258432454002e-05, + "loss": 1.1219, + "step": 5530 + }, + { + "epoch": 0.3360471474573182, + "grad_norm": 0.21356301009655, + "learning_rate": 7.491428506355122e-05, + "loss": 1.0792, + "step": 5531 + }, + { + "epoch": 0.3361079044899447, + "grad_norm": 0.17830733954906464, + "learning_rate": 7.490598488932449e-05, + "loss": 1.0957, + "step": 5532 + }, + { + "epoch": 0.33616866152257124, + "grad_norm": 0.28326064348220825, + "learning_rate": 7.489768380216405e-05, + "loss": 1.2217, + "step": 5533 + }, + { + "epoch": 0.3362294185551978, + "grad_norm": 0.21209542453289032, + "learning_rate": 7.48893818023742e-05, + "loss": 1.008, + "step": 5534 + }, + { + "epoch": 0.3362901755878243, + "grad_norm": 1.0577402114868164, + "learning_rate": 7.488107889025926e-05, + "loss": 1.0896, + "step": 5535 + }, + { + "epoch": 0.3363509326204508, + "grad_norm": 0.3982503414154053, + "learning_rate": 7.487277506612356e-05, + "loss": 1.1928, + "step": 5536 + }, + { + "epoch": 0.33641168965307733, + "grad_norm": 0.24695736169815063, + "learning_rate": 7.486447033027148e-05, + "loss": 1.1234, + "step": 5537 + }, + { + "epoch": 0.33647244668570386, + "grad_norm": 0.30757272243499756, + "learning_rate": 7.485616468300743e-05, + "loss": 1.1714, + "step": 5538 + }, + { + "epoch": 0.3365332037183304, + "grad_norm": 0.24509373307228088, + "learning_rate": 7.484785812463587e-05, + "loss": 1.1896, + "step": 5539 + }, + { + "epoch": 0.33659396075095693, + "grad_norm": 0.21111565828323364, + "learning_rate": 7.483955065546126e-05, + "loss": 1.1389, + "step": 5540 + }, + { + "epoch": 0.33665471778358347, + "grad_norm": 0.26708856225013733, + "learning_rate": 7.483124227578811e-05, + "loss": 1.2307, + "step": 5541 + }, + { + "epoch": 0.33671547481621, + "grad_norm": 0.22500869631767273, + "learning_rate": 7.482293298592098e-05, + "loss": 1.0786, + "step": 5542 + }, + { + "epoch": 0.3367762318488365, + "grad_norm": 0.2700563073158264, + "learning_rate": 7.481462278616444e-05, + "loss": 1.0469, + "step": 5543 + }, + { + "epoch": 0.336836988881463, + "grad_norm": 0.2349618524312973, + "learning_rate": 7.480631167682312e-05, + "loss": 1.2211, + "step": 5544 + }, + { + "epoch": 0.33689774591408955, + "grad_norm": 0.3942452371120453, + "learning_rate": 7.479799965820164e-05, + "loss": 1.0782, + "step": 5545 + }, + { + "epoch": 0.3369585029467161, + "grad_norm": 0.23508605360984802, + "learning_rate": 7.478968673060469e-05, + "loss": 1.1876, + "step": 5546 + }, + { + "epoch": 0.3370192599793426, + "grad_norm": 0.20002005994319916, + "learning_rate": 7.478137289433699e-05, + "loss": 1.0311, + "step": 5547 + }, + { + "epoch": 0.33708001701196916, + "grad_norm": 0.20272500813007355, + "learning_rate": 7.477305814970327e-05, + "loss": 1.0604, + "step": 5548 + }, + { + "epoch": 0.33714077404459564, + "grad_norm": 1.2619998455047607, + "learning_rate": 7.476474249700831e-05, + "loss": 1.155, + "step": 5549 + }, + { + "epoch": 0.33720153107722217, + "grad_norm": 0.36180660128593445, + "learning_rate": 7.475642593655694e-05, + "loss": 1.1749, + "step": 5550 + }, + { + "epoch": 0.3372622881098487, + "grad_norm": 0.3042593002319336, + "learning_rate": 7.474810846865396e-05, + "loss": 1.1152, + "step": 5551 + }, + { + "epoch": 0.33732304514247524, + "grad_norm": 0.28500834107398987, + "learning_rate": 7.473979009360433e-05, + "loss": 1.0704, + "step": 5552 + }, + { + "epoch": 0.3373838021751018, + "grad_norm": 0.22955618798732758, + "learning_rate": 7.473147081171287e-05, + "loss": 1.1514, + "step": 5553 + }, + { + "epoch": 0.3374445592077283, + "grad_norm": 0.3436807692050934, + "learning_rate": 7.472315062328459e-05, + "loss": 1.2207, + "step": 5554 + }, + { + "epoch": 0.33750531624035485, + "grad_norm": 0.23744896054267883, + "learning_rate": 7.471482952862444e-05, + "loss": 1.0789, + "step": 5555 + }, + { + "epoch": 0.3375660732729813, + "grad_norm": 0.22982428967952728, + "learning_rate": 7.470650752803743e-05, + "loss": 1.0552, + "step": 5556 + }, + { + "epoch": 0.33762683030560786, + "grad_norm": 0.4424126446247101, + "learning_rate": 7.46981846218286e-05, + "loss": 1.2159, + "step": 5557 + }, + { + "epoch": 0.3376875873382344, + "grad_norm": 0.18888282775878906, + "learning_rate": 7.468986081030307e-05, + "loss": 1.0789, + "step": 5558 + }, + { + "epoch": 0.33774834437086093, + "grad_norm": 0.49860331416130066, + "learning_rate": 7.468153609376591e-05, + "loss": 1.0899, + "step": 5559 + }, + { + "epoch": 0.33780910140348747, + "grad_norm": 0.3263925611972809, + "learning_rate": 7.467321047252226e-05, + "loss": 1.158, + "step": 5560 + }, + { + "epoch": 0.337869858436114, + "grad_norm": 0.20108215510845184, + "learning_rate": 7.466488394687731e-05, + "loss": 1.0581, + "step": 5561 + }, + { + "epoch": 0.3379306154687405, + "grad_norm": 0.5139026641845703, + "learning_rate": 7.465655651713627e-05, + "loss": 1.1, + "step": 5562 + }, + { + "epoch": 0.337991372501367, + "grad_norm": 0.18616123497486115, + "learning_rate": 7.464822818360441e-05, + "loss": 1.0886, + "step": 5563 + }, + { + "epoch": 0.33805212953399355, + "grad_norm": 3.09488844871521, + "learning_rate": 7.463989894658696e-05, + "loss": 1.2352, + "step": 5564 + }, + { + "epoch": 0.3381128865666201, + "grad_norm": 0.4342299997806549, + "learning_rate": 7.463156880638926e-05, + "loss": 1.0448, + "step": 5565 + }, + { + "epoch": 0.3381736435992466, + "grad_norm": 0.29608485102653503, + "learning_rate": 7.462323776331664e-05, + "loss": 1.1559, + "step": 5566 + }, + { + "epoch": 0.33823440063187316, + "grad_norm": 0.19445081055164337, + "learning_rate": 7.461490581767448e-05, + "loss": 1.0889, + "step": 5567 + }, + { + "epoch": 0.3382951576644997, + "grad_norm": 0.37279465794563293, + "learning_rate": 7.460657296976819e-05, + "loss": 1.0573, + "step": 5568 + }, + { + "epoch": 0.33835591469712617, + "grad_norm": 0.22310063242912292, + "learning_rate": 7.459823921990321e-05, + "loss": 1.064, + "step": 5569 + }, + { + "epoch": 0.3384166717297527, + "grad_norm": 0.21040183305740356, + "learning_rate": 7.458990456838501e-05, + "loss": 1.0721, + "step": 5570 + }, + { + "epoch": 0.33847742876237924, + "grad_norm": 1.125022530555725, + "learning_rate": 7.458156901551912e-05, + "loss": 1.0487, + "step": 5571 + }, + { + "epoch": 0.3385381857950058, + "grad_norm": 0.23865027725696564, + "learning_rate": 7.457323256161106e-05, + "loss": 1.1458, + "step": 5572 + }, + { + "epoch": 0.3385989428276323, + "grad_norm": 0.249007910490036, + "learning_rate": 7.456489520696642e-05, + "loss": 1.0471, + "step": 5573 + }, + { + "epoch": 0.33865969986025884, + "grad_norm": 0.2564275562763214, + "learning_rate": 7.455655695189077e-05, + "loss": 1.1062, + "step": 5574 + }, + { + "epoch": 0.3387204568928853, + "grad_norm": 0.22975635528564453, + "learning_rate": 7.454821779668981e-05, + "loss": 1.0589, + "step": 5575 + }, + { + "epoch": 0.33878121392551186, + "grad_norm": 0.2332553118467331, + "learning_rate": 7.453987774166917e-05, + "loss": 1.1087, + "step": 5576 + }, + { + "epoch": 0.3388419709581384, + "grad_norm": 0.20201462507247925, + "learning_rate": 7.453153678713455e-05, + "loss": 1.136, + "step": 5577 + }, + { + "epoch": 0.33890272799076493, + "grad_norm": 0.27439427375793457, + "learning_rate": 7.452319493339173e-05, + "loss": 1.265, + "step": 5578 + }, + { + "epoch": 0.33896348502339146, + "grad_norm": 0.3693367540836334, + "learning_rate": 7.451485218074646e-05, + "loss": 1.0795, + "step": 5579 + }, + { + "epoch": 0.339024242056018, + "grad_norm": 0.21314868330955505, + "learning_rate": 7.450650852950452e-05, + "loss": 1.0691, + "step": 5580 + }, + { + "epoch": 0.33908499908864453, + "grad_norm": 0.26669618487358093, + "learning_rate": 7.44981639799718e-05, + "loss": 1.1583, + "step": 5581 + }, + { + "epoch": 0.339145756121271, + "grad_norm": 0.3302687704563141, + "learning_rate": 7.448981853245414e-05, + "loss": 1.3221, + "step": 5582 + }, + { + "epoch": 0.33920651315389755, + "grad_norm": 0.32206854224205017, + "learning_rate": 7.448147218725744e-05, + "loss": 1.1773, + "step": 5583 + }, + { + "epoch": 0.3392672701865241, + "grad_norm": 3.680023670196533, + "learning_rate": 7.447312494468764e-05, + "loss": 1.0762, + "step": 5584 + }, + { + "epoch": 0.3393280272191506, + "grad_norm": 0.3142934739589691, + "learning_rate": 7.446477680505073e-05, + "loss": 1.1257, + "step": 5585 + }, + { + "epoch": 0.33938878425177715, + "grad_norm": 0.2706977427005768, + "learning_rate": 7.445642776865269e-05, + "loss": 1.0655, + "step": 5586 + }, + { + "epoch": 0.3394495412844037, + "grad_norm": 0.2008800059556961, + "learning_rate": 7.444807783579955e-05, + "loss": 1.052, + "step": 5587 + }, + { + "epoch": 0.3395102983170302, + "grad_norm": 0.2373208850622177, + "learning_rate": 7.44397270067974e-05, + "loss": 1.1126, + "step": 5588 + }, + { + "epoch": 0.3395710553496567, + "grad_norm": 0.2877863049507141, + "learning_rate": 7.443137528195234e-05, + "loss": 1.0873, + "step": 5589 + }, + { + "epoch": 0.33963181238228324, + "grad_norm": 0.2483014166355133, + "learning_rate": 7.442302266157048e-05, + "loss": 1.0842, + "step": 5590 + }, + { + "epoch": 0.3396925694149098, + "grad_norm": 0.40851476788520813, + "learning_rate": 7.441466914595803e-05, + "loss": 1.1002, + "step": 5591 + }, + { + "epoch": 0.3397533264475363, + "grad_norm": 0.22570998966693878, + "learning_rate": 7.440631473542115e-05, + "loss": 1.2116, + "step": 5592 + }, + { + "epoch": 0.33981408348016284, + "grad_norm": 0.35261374711990356, + "learning_rate": 7.439795943026608e-05, + "loss": 1.0527, + "step": 5593 + }, + { + "epoch": 0.3398748405127894, + "grad_norm": 0.3454292118549347, + "learning_rate": 7.43896032307991e-05, + "loss": 1.1017, + "step": 5594 + }, + { + "epoch": 0.33993559754541586, + "grad_norm": 0.18269822001457214, + "learning_rate": 7.438124613732648e-05, + "loss": 1.0237, + "step": 5595 + }, + { + "epoch": 0.3399963545780424, + "grad_norm": 0.29915931820869446, + "learning_rate": 7.43728881501546e-05, + "loss": 1.1324, + "step": 5596 + }, + { + "epoch": 0.3400571116106689, + "grad_norm": 0.3109831213951111, + "learning_rate": 7.436452926958978e-05, + "loss": 1.1221, + "step": 5597 + }, + { + "epoch": 0.34011786864329546, + "grad_norm": 0.17742733657360077, + "learning_rate": 7.435616949593842e-05, + "loss": 1.0762, + "step": 5598 + }, + { + "epoch": 0.340178625675922, + "grad_norm": 0.2405356764793396, + "learning_rate": 7.434780882950698e-05, + "loss": 1.0424, + "step": 5599 + }, + { + "epoch": 0.34023938270854853, + "grad_norm": 0.34910479187965393, + "learning_rate": 7.433944727060187e-05, + "loss": 1.1934, + "step": 5600 + }, + { + "epoch": 0.34030013974117507, + "grad_norm": 0.5067254900932312, + "learning_rate": 7.433108481952965e-05, + "loss": 1.2571, + "step": 5601 + }, + { + "epoch": 0.34036089677380155, + "grad_norm": 0.23744887113571167, + "learning_rate": 7.432272147659678e-05, + "loss": 1.0956, + "step": 5602 + }, + { + "epoch": 0.3404216538064281, + "grad_norm": 0.4821052551269531, + "learning_rate": 7.431435724210988e-05, + "loss": 1.1067, + "step": 5603 + }, + { + "epoch": 0.3404824108390546, + "grad_norm": 0.22347436845302582, + "learning_rate": 7.43059921163755e-05, + "loss": 1.0838, + "step": 5604 + }, + { + "epoch": 0.34054316787168115, + "grad_norm": 0.3074796199798584, + "learning_rate": 7.42976260997003e-05, + "loss": 1.0669, + "step": 5605 + }, + { + "epoch": 0.3406039249043077, + "grad_norm": 0.17580543458461761, + "learning_rate": 7.428925919239089e-05, + "loss": 1.0832, + "step": 5606 + }, + { + "epoch": 0.3406646819369342, + "grad_norm": 0.3393591344356537, + "learning_rate": 7.4280891394754e-05, + "loss": 1.1939, + "step": 5607 + }, + { + "epoch": 0.3407254389695607, + "grad_norm": 0.17790570855140686, + "learning_rate": 7.427252270709635e-05, + "loss": 1.0606, + "step": 5608 + }, + { + "epoch": 0.34078619600218724, + "grad_norm": 0.42695674300193787, + "learning_rate": 7.426415312972469e-05, + "loss": 1.083, + "step": 5609 + }, + { + "epoch": 0.34084695303481377, + "grad_norm": 0.18511222302913666, + "learning_rate": 7.425578266294579e-05, + "loss": 1.1041, + "step": 5610 + }, + { + "epoch": 0.3409077100674403, + "grad_norm": 0.24028939008712769, + "learning_rate": 7.42474113070665e-05, + "loss": 1.12, + "step": 5611 + }, + { + "epoch": 0.34096846710006684, + "grad_norm": 0.25160080194473267, + "learning_rate": 7.423903906239367e-05, + "loss": 1.0706, + "step": 5612 + }, + { + "epoch": 0.3410292241326934, + "grad_norm": 0.19215840101242065, + "learning_rate": 7.423066592923417e-05, + "loss": 1.0943, + "step": 5613 + }, + { + "epoch": 0.3410899811653199, + "grad_norm": 0.2835266590118408, + "learning_rate": 7.422229190789492e-05, + "loss": 1.0582, + "step": 5614 + }, + { + "epoch": 0.3411507381979464, + "grad_norm": 0.1954669952392578, + "learning_rate": 7.421391699868288e-05, + "loss": 1.0309, + "step": 5615 + }, + { + "epoch": 0.3412114952305729, + "grad_norm": 0.27827927470207214, + "learning_rate": 7.420554120190504e-05, + "loss": 1.0975, + "step": 5616 + }, + { + "epoch": 0.34127225226319946, + "grad_norm": 0.31608474254608154, + "learning_rate": 7.41971645178684e-05, + "loss": 1.0752, + "step": 5617 + }, + { + "epoch": 0.341333009295826, + "grad_norm": 0.42969006299972534, + "learning_rate": 7.418878694688002e-05, + "loss": 1.1258, + "step": 5618 + }, + { + "epoch": 0.34139376632845253, + "grad_norm": 0.4891491234302521, + "learning_rate": 7.418040848924698e-05, + "loss": 1.1807, + "step": 5619 + }, + { + "epoch": 0.34145452336107907, + "grad_norm": 0.3236261010169983, + "learning_rate": 7.41720291452764e-05, + "loss": 1.1522, + "step": 5620 + }, + { + "epoch": 0.34151528039370554, + "grad_norm": 0.2415841817855835, + "learning_rate": 7.416364891527541e-05, + "loss": 1.0756, + "step": 5621 + }, + { + "epoch": 0.3415760374263321, + "grad_norm": 0.39893993735313416, + "learning_rate": 7.415526779955121e-05, + "loss": 1.0745, + "step": 5622 + }, + { + "epoch": 0.3416367944589586, + "grad_norm": 0.17209921777248383, + "learning_rate": 7.414688579841098e-05, + "loss": 1.0984, + "step": 5623 + }, + { + "epoch": 0.34169755149158515, + "grad_norm": 0.24523979425430298, + "learning_rate": 7.4138502912162e-05, + "loss": 1.111, + "step": 5624 + }, + { + "epoch": 0.3417583085242117, + "grad_norm": 0.4194061756134033, + "learning_rate": 7.413011914111152e-05, + "loss": 1.0921, + "step": 5625 + }, + { + "epoch": 0.3418190655568382, + "grad_norm": 0.4137375056743622, + "learning_rate": 7.412173448556687e-05, + "loss": 1.2066, + "step": 5626 + }, + { + "epoch": 0.34187982258946475, + "grad_norm": 0.665217399597168, + "learning_rate": 7.411334894583538e-05, + "loss": 1.1385, + "step": 5627 + }, + { + "epoch": 0.34194057962209123, + "grad_norm": 0.2416275441646576, + "learning_rate": 7.41049625222244e-05, + "loss": 1.0569, + "step": 5628 + }, + { + "epoch": 0.34200133665471777, + "grad_norm": 0.6429489254951477, + "learning_rate": 7.409657521504137e-05, + "loss": 1.1169, + "step": 5629 + }, + { + "epoch": 0.3420620936873443, + "grad_norm": 0.19447322189807892, + "learning_rate": 7.408818702459374e-05, + "loss": 1.0888, + "step": 5630 + }, + { + "epoch": 0.34212285071997084, + "grad_norm": 0.3485240936279297, + "learning_rate": 7.407979795118892e-05, + "loss": 1.205, + "step": 5631 + }, + { + "epoch": 0.3421836077525974, + "grad_norm": 0.31084737181663513, + "learning_rate": 7.407140799513449e-05, + "loss": 1.0961, + "step": 5632 + }, + { + "epoch": 0.3422443647852239, + "grad_norm": 0.23880264163017273, + "learning_rate": 7.406301715673793e-05, + "loss": 1.0623, + "step": 5633 + }, + { + "epoch": 0.3423051218178504, + "grad_norm": 0.469424307346344, + "learning_rate": 7.405462543630681e-05, + "loss": 1.1642, + "step": 5634 + }, + { + "epoch": 0.3423658788504769, + "grad_norm": 0.734106719493866, + "learning_rate": 7.404623283414876e-05, + "loss": 1.0731, + "step": 5635 + }, + { + "epoch": 0.34242663588310346, + "grad_norm": 0.26658308506011963, + "learning_rate": 7.40378393505714e-05, + "loss": 1.0385, + "step": 5636 + }, + { + "epoch": 0.34248739291573, + "grad_norm": 1.2295684814453125, + "learning_rate": 7.402944498588239e-05, + "loss": 1.1509, + "step": 5637 + }, + { + "epoch": 0.34254814994835653, + "grad_norm": 0.389790803194046, + "learning_rate": 7.402104974038943e-05, + "loss": 1.1748, + "step": 5638 + }, + { + "epoch": 0.34260890698098306, + "grad_norm": 0.2092919796705246, + "learning_rate": 7.401265361440026e-05, + "loss": 1.1386, + "step": 5639 + }, + { + "epoch": 0.3426696640136096, + "grad_norm": 0.45629552006721497, + "learning_rate": 7.40042566082226e-05, + "loss": 1.0648, + "step": 5640 + }, + { + "epoch": 0.3427304210462361, + "grad_norm": 0.3841695785522461, + "learning_rate": 7.39958587221643e-05, + "loss": 1.1984, + "step": 5641 + }, + { + "epoch": 0.3427911780788626, + "grad_norm": 0.3087284564971924, + "learning_rate": 7.398745995653316e-05, + "loss": 1.0946, + "step": 5642 + }, + { + "epoch": 0.34285193511148915, + "grad_norm": 0.45674625039100647, + "learning_rate": 7.397906031163702e-05, + "loss": 1.1217, + "step": 5643 + }, + { + "epoch": 0.3429126921441157, + "grad_norm": 0.2547856569290161, + "learning_rate": 7.397065978778379e-05, + "loss": 1.1532, + "step": 5644 + }, + { + "epoch": 0.3429734491767422, + "grad_norm": 0.29508382081985474, + "learning_rate": 7.396225838528141e-05, + "loss": 1.0711, + "step": 5645 + }, + { + "epoch": 0.34303420620936875, + "grad_norm": 0.46495160460472107, + "learning_rate": 7.395385610443782e-05, + "loss": 1.1731, + "step": 5646 + }, + { + "epoch": 0.3430949632419953, + "grad_norm": 0.1505955159664154, + "learning_rate": 7.394545294556098e-05, + "loss": 1.023, + "step": 5647 + }, + { + "epoch": 0.34315572027462177, + "grad_norm": 0.44654780626296997, + "learning_rate": 7.393704890895895e-05, + "loss": 1.0931, + "step": 5648 + }, + { + "epoch": 0.3432164773072483, + "grad_norm": 0.24531032145023346, + "learning_rate": 7.392864399493976e-05, + "loss": 1.2158, + "step": 5649 + }, + { + "epoch": 0.34327723433987484, + "grad_norm": 0.35129088163375854, + "learning_rate": 7.39202382038115e-05, + "loss": 1.2294, + "step": 5650 + }, + { + "epoch": 0.34333799137250137, + "grad_norm": 0.41323909163475037, + "learning_rate": 7.39118315358823e-05, + "loss": 1.1303, + "step": 5651 + }, + { + "epoch": 0.3433987484051279, + "grad_norm": 0.45980751514434814, + "learning_rate": 7.390342399146027e-05, + "loss": 1.1056, + "step": 5652 + }, + { + "epoch": 0.34345950543775444, + "grad_norm": 0.21634501218795776, + "learning_rate": 7.389501557085362e-05, + "loss": 1.0693, + "step": 5653 + }, + { + "epoch": 0.3435202624703809, + "grad_norm": 0.5432146787643433, + "learning_rate": 7.388660627437055e-05, + "loss": 1.2046, + "step": 5654 + }, + { + "epoch": 0.34358101950300746, + "grad_norm": 0.34360048174858093, + "learning_rate": 7.387819610231932e-05, + "loss": 1.1411, + "step": 5655 + }, + { + "epoch": 0.343641776535634, + "grad_norm": 1.770283818244934, + "learning_rate": 7.386978505500822e-05, + "loss": 1.0809, + "step": 5656 + }, + { + "epoch": 0.3437025335682605, + "grad_norm": 0.43656134605407715, + "learning_rate": 7.386137313274548e-05, + "loss": 1.1216, + "step": 5657 + }, + { + "epoch": 0.34376329060088706, + "grad_norm": 0.4038173258304596, + "learning_rate": 7.385296033583953e-05, + "loss": 1.1627, + "step": 5658 + }, + { + "epoch": 0.3438240476335136, + "grad_norm": 0.6463785767555237, + "learning_rate": 7.38445466645987e-05, + "loss": 1.0634, + "step": 5659 + }, + { + "epoch": 0.34388480466614013, + "grad_norm": 0.3278791606426239, + "learning_rate": 7.383613211933141e-05, + "loss": 1.0861, + "step": 5660 + }, + { + "epoch": 0.3439455616987666, + "grad_norm": 0.21886733174324036, + "learning_rate": 7.382771670034608e-05, + "loss": 1.1171, + "step": 5661 + }, + { + "epoch": 0.34400631873139315, + "grad_norm": 0.6112619638442993, + "learning_rate": 7.381930040795119e-05, + "loss": 1.2697, + "step": 5662 + }, + { + "epoch": 0.3440670757640197, + "grad_norm": 0.3182641863822937, + "learning_rate": 7.381088324245525e-05, + "loss": 1.0819, + "step": 5663 + }, + { + "epoch": 0.3441278327966462, + "grad_norm": 0.37121114134788513, + "learning_rate": 7.380246520416678e-05, + "loss": 1.0897, + "step": 5664 + }, + { + "epoch": 0.34418858982927275, + "grad_norm": 0.20945681631565094, + "learning_rate": 7.379404629339433e-05, + "loss": 1.091, + "step": 5665 + }, + { + "epoch": 0.3442493468618993, + "grad_norm": 0.7312408685684204, + "learning_rate": 7.378562651044653e-05, + "loss": 1.1417, + "step": 5666 + }, + { + "epoch": 0.34431010389452577, + "grad_norm": 0.2518223524093628, + "learning_rate": 7.377720585563199e-05, + "loss": 1.1415, + "step": 5667 + }, + { + "epoch": 0.3443708609271523, + "grad_norm": 0.29222944378852844, + "learning_rate": 7.376878432925938e-05, + "loss": 1.1638, + "step": 5668 + }, + { + "epoch": 0.34443161795977884, + "grad_norm": 0.2349882870912552, + "learning_rate": 7.376036193163738e-05, + "loss": 1.0364, + "step": 5669 + }, + { + "epoch": 0.34449237499240537, + "grad_norm": 0.5807378888130188, + "learning_rate": 7.375193866307471e-05, + "loss": 1.1553, + "step": 5670 + }, + { + "epoch": 0.3445531320250319, + "grad_norm": 0.2327040433883667, + "learning_rate": 7.374351452388016e-05, + "loss": 1.1214, + "step": 5671 + }, + { + "epoch": 0.34461388905765844, + "grad_norm": 0.5924752950668335, + "learning_rate": 7.373508951436247e-05, + "loss": 1.2714, + "step": 5672 + }, + { + "epoch": 0.344674646090285, + "grad_norm": 0.18569479882717133, + "learning_rate": 7.37266636348305e-05, + "loss": 1.0526, + "step": 5673 + }, + { + "epoch": 0.34473540312291145, + "grad_norm": 0.17636126279830933, + "learning_rate": 7.371823688559308e-05, + "loss": 1.0603, + "step": 5674 + }, + { + "epoch": 0.344796160155538, + "grad_norm": 0.31182244420051575, + "learning_rate": 7.37098092669591e-05, + "loss": 1.1194, + "step": 5675 + }, + { + "epoch": 0.3448569171881645, + "grad_norm": 0.29238808155059814, + "learning_rate": 7.37013807792375e-05, + "loss": 1.0851, + "step": 5676 + }, + { + "epoch": 0.34491767422079106, + "grad_norm": 0.3122333288192749, + "learning_rate": 7.36929514227372e-05, + "loss": 1.0587, + "step": 5677 + }, + { + "epoch": 0.3449784312534176, + "grad_norm": 1.4509057998657227, + "learning_rate": 7.368452119776716e-05, + "loss": 1.0596, + "step": 5678 + }, + { + "epoch": 0.34503918828604413, + "grad_norm": 5.834701061248779, + "learning_rate": 7.367609010463646e-05, + "loss": 1.0322, + "step": 5679 + }, + { + "epoch": 0.3450999453186706, + "grad_norm": 0.3240301012992859, + "learning_rate": 7.366765814365406e-05, + "loss": 1.0799, + "step": 5680 + }, + { + "epoch": 0.34516070235129714, + "grad_norm": 0.8705154061317444, + "learning_rate": 7.36592253151291e-05, + "loss": 1.1721, + "step": 5681 + }, + { + "epoch": 0.3452214593839237, + "grad_norm": 0.41325345635414124, + "learning_rate": 7.365079161937066e-05, + "loss": 1.167, + "step": 5682 + }, + { + "epoch": 0.3452822164165502, + "grad_norm": 0.3813403844833374, + "learning_rate": 7.364235705668787e-05, + "loss": 1.1023, + "step": 5683 + }, + { + "epoch": 0.34534297344917675, + "grad_norm": 0.2294529378414154, + "learning_rate": 7.363392162738992e-05, + "loss": 1.1249, + "step": 5684 + }, + { + "epoch": 0.3454037304818033, + "grad_norm": 0.20475754141807556, + "learning_rate": 7.362548533178599e-05, + "loss": 1.0551, + "step": 5685 + }, + { + "epoch": 0.3454644875144298, + "grad_norm": 0.2383648157119751, + "learning_rate": 7.361704817018535e-05, + "loss": 1.2264, + "step": 5686 + }, + { + "epoch": 0.3455252445470563, + "grad_norm": 0.19728489220142365, + "learning_rate": 7.360861014289724e-05, + "loss": 1.15, + "step": 5687 + }, + { + "epoch": 0.34558600157968283, + "grad_norm": 0.32222267985343933, + "learning_rate": 7.360017125023096e-05, + "loss": 1.1708, + "step": 5688 + }, + { + "epoch": 0.34564675861230937, + "grad_norm": 0.27936238050460815, + "learning_rate": 7.359173149249584e-05, + "loss": 1.2761, + "step": 5689 + }, + { + "epoch": 0.3457075156449359, + "grad_norm": 0.23428481817245483, + "learning_rate": 7.358329087000124e-05, + "loss": 1.0564, + "step": 5690 + }, + { + "epoch": 0.34576827267756244, + "grad_norm": 0.20199131965637207, + "learning_rate": 7.357484938305655e-05, + "loss": 1.09, + "step": 5691 + }, + { + "epoch": 0.345829029710189, + "grad_norm": 0.27093037962913513, + "learning_rate": 7.356640703197119e-05, + "loss": 1.1562, + "step": 5692 + }, + { + "epoch": 0.3458897867428155, + "grad_norm": 0.21575915813446045, + "learning_rate": 7.355796381705465e-05, + "loss": 1.1381, + "step": 5693 + }, + { + "epoch": 0.345950543775442, + "grad_norm": 0.32112282514572144, + "learning_rate": 7.354951973861639e-05, + "loss": 1.1443, + "step": 5694 + }, + { + "epoch": 0.3460113008080685, + "grad_norm": 0.275910347700119, + "learning_rate": 7.354107479696591e-05, + "loss": 1.112, + "step": 5695 + }, + { + "epoch": 0.34607205784069506, + "grad_norm": 0.4309602677822113, + "learning_rate": 7.35326289924128e-05, + "loss": 1.0578, + "step": 5696 + }, + { + "epoch": 0.3461328148733216, + "grad_norm": 0.23620781302452087, + "learning_rate": 7.352418232526663e-05, + "loss": 1.136, + "step": 5697 + }, + { + "epoch": 0.3461935719059481, + "grad_norm": 0.2078590989112854, + "learning_rate": 7.3515734795837e-05, + "loss": 1.0708, + "step": 5698 + }, + { + "epoch": 0.34625432893857466, + "grad_norm": 0.2000613510608673, + "learning_rate": 7.350728640443358e-05, + "loss": 1.0619, + "step": 5699 + }, + { + "epoch": 0.34631508597120114, + "grad_norm": 0.16540998220443726, + "learning_rate": 7.3498837151366e-05, + "loss": 1.0621, + "step": 5700 + }, + { + "epoch": 0.3463758430038277, + "grad_norm": 0.6190958023071289, + "learning_rate": 7.349038703694401e-05, + "loss": 1.0864, + "step": 5701 + }, + { + "epoch": 0.3464366000364542, + "grad_norm": 0.2357926368713379, + "learning_rate": 7.348193606147738e-05, + "loss": 1.1585, + "step": 5702 + }, + { + "epoch": 0.34649735706908075, + "grad_norm": 0.20251639187335968, + "learning_rate": 7.347348422527581e-05, + "loss": 1.0561, + "step": 5703 + }, + { + "epoch": 0.3465581141017073, + "grad_norm": 0.21775290369987488, + "learning_rate": 7.346503152864916e-05, + "loss": 1.0903, + "step": 5704 + }, + { + "epoch": 0.3466188711343338, + "grad_norm": 0.21620871126651764, + "learning_rate": 7.345657797190723e-05, + "loss": 1.0648, + "step": 5705 + }, + { + "epoch": 0.34667962816696035, + "grad_norm": 0.18312202394008636, + "learning_rate": 7.344812355535989e-05, + "loss": 1.1003, + "step": 5706 + }, + { + "epoch": 0.34674038519958683, + "grad_norm": 0.1562747210264206, + "learning_rate": 7.343966827931707e-05, + "loss": 1.0826, + "step": 5707 + }, + { + "epoch": 0.34680114223221337, + "grad_norm": 0.41605180501937866, + "learning_rate": 7.343121214408867e-05, + "loss": 1.1733, + "step": 5708 + }, + { + "epoch": 0.3468618992648399, + "grad_norm": 0.3232215940952301, + "learning_rate": 7.342275514998466e-05, + "loss": 1.0322, + "step": 5709 + }, + { + "epoch": 0.34692265629746644, + "grad_norm": 0.37585213780403137, + "learning_rate": 7.341429729731503e-05, + "loss": 1.1859, + "step": 5710 + }, + { + "epoch": 0.34698341333009297, + "grad_norm": 0.670525848865509, + "learning_rate": 7.34058385863898e-05, + "loss": 1.2311, + "step": 5711 + }, + { + "epoch": 0.3470441703627195, + "grad_norm": 0.4025414288043976, + "learning_rate": 7.339737901751905e-05, + "loss": 1.2488, + "step": 5712 + }, + { + "epoch": 0.347104927395346, + "grad_norm": 0.33200472593307495, + "learning_rate": 7.338891859101285e-05, + "loss": 1.2745, + "step": 5713 + }, + { + "epoch": 0.3471656844279725, + "grad_norm": 0.20677070319652557, + "learning_rate": 7.338045730718129e-05, + "loss": 1.1316, + "step": 5714 + }, + { + "epoch": 0.34722644146059906, + "grad_norm": 0.2655167281627655, + "learning_rate": 7.337199516633458e-05, + "loss": 1.1309, + "step": 5715 + }, + { + "epoch": 0.3472871984932256, + "grad_norm": 0.23249642550945282, + "learning_rate": 7.336353216878285e-05, + "loss": 1.0515, + "step": 5716 + }, + { + "epoch": 0.3473479555258521, + "grad_norm": 0.2561902403831482, + "learning_rate": 7.335506831483634e-05, + "loss": 1.2291, + "step": 5717 + }, + { + "epoch": 0.34740871255847866, + "grad_norm": 0.1944292038679123, + "learning_rate": 7.334660360480527e-05, + "loss": 1.0919, + "step": 5718 + }, + { + "epoch": 0.3474694695911052, + "grad_norm": 0.27357015013694763, + "learning_rate": 7.333813803899996e-05, + "loss": 1.1596, + "step": 5719 + }, + { + "epoch": 0.3475302266237317, + "grad_norm": 0.25469493865966797, + "learning_rate": 7.332967161773068e-05, + "loss": 1.1602, + "step": 5720 + }, + { + "epoch": 0.3475909836563582, + "grad_norm": 0.2988016903400421, + "learning_rate": 7.332120434130778e-05, + "loss": 1.1694, + "step": 5721 + }, + { + "epoch": 0.34765174068898475, + "grad_norm": 0.21968673169612885, + "learning_rate": 7.33127362100416e-05, + "loss": 1.126, + "step": 5722 + }, + { + "epoch": 0.3477124977216113, + "grad_norm": 0.2018594890832901, + "learning_rate": 7.33042672242426e-05, + "loss": 1.1269, + "step": 5723 + }, + { + "epoch": 0.3477732547542378, + "grad_norm": 0.3044150471687317, + "learning_rate": 7.329579738422116e-05, + "loss": 1.0821, + "step": 5724 + }, + { + "epoch": 0.34783401178686435, + "grad_norm": 0.49157774448394775, + "learning_rate": 7.328732669028777e-05, + "loss": 1.1564, + "step": 5725 + }, + { + "epoch": 0.34789476881949083, + "grad_norm": 0.21208806335926056, + "learning_rate": 7.327885514275294e-05, + "loss": 1.0307, + "step": 5726 + }, + { + "epoch": 0.34795552585211736, + "grad_norm": 0.5574318170547485, + "learning_rate": 7.327038274192714e-05, + "loss": 1.0583, + "step": 5727 + }, + { + "epoch": 0.3480162828847439, + "grad_norm": 0.16888654232025146, + "learning_rate": 7.326190948812097e-05, + "loss": 1.0877, + "step": 5728 + }, + { + "epoch": 0.34807703991737043, + "grad_norm": 14.913818359375, + "learning_rate": 7.325343538164503e-05, + "loss": 1.0685, + "step": 5729 + }, + { + "epoch": 0.34813779694999697, + "grad_norm": 0.7267302870750427, + "learning_rate": 7.32449604228099e-05, + "loss": 1.0148, + "step": 5730 + }, + { + "epoch": 0.3481985539826235, + "grad_norm": 0.3143238127231598, + "learning_rate": 7.323648461192627e-05, + "loss": 1.1399, + "step": 5731 + }, + { + "epoch": 0.34825931101525004, + "grad_norm": 0.6576926708221436, + "learning_rate": 7.32280079493048e-05, + "loss": 1.128, + "step": 5732 + }, + { + "epoch": 0.3483200680478765, + "grad_norm": 0.3847138285636902, + "learning_rate": 7.321953043525622e-05, + "loss": 1.0667, + "step": 5733 + }, + { + "epoch": 0.34838082508050305, + "grad_norm": 0.38284698128700256, + "learning_rate": 7.321105207009124e-05, + "loss": 1.1227, + "step": 5734 + }, + { + "epoch": 0.3484415821131296, + "grad_norm": 0.5855658054351807, + "learning_rate": 7.320257285412066e-05, + "loss": 1.1068, + "step": 5735 + }, + { + "epoch": 0.3485023391457561, + "grad_norm": 0.2504933178424835, + "learning_rate": 7.319409278765532e-05, + "loss": 1.2599, + "step": 5736 + }, + { + "epoch": 0.34856309617838266, + "grad_norm": 0.6202361583709717, + "learning_rate": 7.318561187100598e-05, + "loss": 1.2084, + "step": 5737 + }, + { + "epoch": 0.3486238532110092, + "grad_norm": 0.4402351677417755, + "learning_rate": 7.317713010448359e-05, + "loss": 1.0848, + "step": 5738 + }, + { + "epoch": 0.3486846102436357, + "grad_norm": 0.21430347859859467, + "learning_rate": 7.316864748839901e-05, + "loss": 1.0275, + "step": 5739 + }, + { + "epoch": 0.3487453672762622, + "grad_norm": 0.24353273212909698, + "learning_rate": 7.316016402306315e-05, + "loss": 1.0574, + "step": 5740 + }, + { + "epoch": 0.34880612430888874, + "grad_norm": 0.4429553747177124, + "learning_rate": 7.315167970878705e-05, + "loss": 1.0977, + "step": 5741 + }, + { + "epoch": 0.3488668813415153, + "grad_norm": 0.5605708956718445, + "learning_rate": 7.314319454588162e-05, + "loss": 1.2228, + "step": 5742 + }, + { + "epoch": 0.3489276383741418, + "grad_norm": 0.5582001209259033, + "learning_rate": 7.313470853465794e-05, + "loss": 1.1998, + "step": 5743 + }, + { + "epoch": 0.34898839540676835, + "grad_norm": 0.22867664694786072, + "learning_rate": 7.312622167542702e-05, + "loss": 1.1105, + "step": 5744 + }, + { + "epoch": 0.3490491524393949, + "grad_norm": 0.175890251994133, + "learning_rate": 7.311773396849999e-05, + "loss": 1.0535, + "step": 5745 + }, + { + "epoch": 0.34910990947202136, + "grad_norm": 0.2645072638988495, + "learning_rate": 7.310924541418796e-05, + "loss": 1.1308, + "step": 5746 + }, + { + "epoch": 0.3491706665046479, + "grad_norm": 0.48679685592651367, + "learning_rate": 7.310075601280206e-05, + "loss": 1.3309, + "step": 5747 + }, + { + "epoch": 0.34923142353727443, + "grad_norm": 0.32934266328811646, + "learning_rate": 7.309226576465349e-05, + "loss": 1.25, + "step": 5748 + }, + { + "epoch": 0.34929218056990097, + "grad_norm": 0.6751758456230164, + "learning_rate": 7.308377467005347e-05, + "loss": 1.2105, + "step": 5749 + }, + { + "epoch": 0.3493529376025275, + "grad_norm": 0.4389433264732361, + "learning_rate": 7.307528272931319e-05, + "loss": 1.0682, + "step": 5750 + }, + { + "epoch": 0.34941369463515404, + "grad_norm": 0.3540611267089844, + "learning_rate": 7.306678994274398e-05, + "loss": 1.0604, + "step": 5751 + }, + { + "epoch": 0.3494744516677806, + "grad_norm": 0.3279232382774353, + "learning_rate": 7.305829631065714e-05, + "loss": 1.1235, + "step": 5752 + }, + { + "epoch": 0.34953520870040705, + "grad_norm": 1.1791020631790161, + "learning_rate": 7.304980183336397e-05, + "loss": 1.1165, + "step": 5753 + }, + { + "epoch": 0.3495959657330336, + "grad_norm": 0.44660109281539917, + "learning_rate": 7.304130651117588e-05, + "loss": 1.1062, + "step": 5754 + }, + { + "epoch": 0.3496567227656601, + "grad_norm": 0.29831475019454956, + "learning_rate": 7.303281034440421e-05, + "loss": 1.0975, + "step": 5755 + }, + { + "epoch": 0.34971747979828666, + "grad_norm": 0.2996464967727661, + "learning_rate": 7.302431333336048e-05, + "loss": 1.0902, + "step": 5756 + }, + { + "epoch": 0.3497782368309132, + "grad_norm": 0.23401302099227905, + "learning_rate": 7.301581547835606e-05, + "loss": 1.1889, + "step": 5757 + }, + { + "epoch": 0.3498389938635397, + "grad_norm": 0.2532077729701996, + "learning_rate": 7.300731677970246e-05, + "loss": 1.1122, + "step": 5758 + }, + { + "epoch": 0.3498997508961662, + "grad_norm": 0.4395678639411926, + "learning_rate": 7.299881723771124e-05, + "loss": 1.217, + "step": 5759 + }, + { + "epoch": 0.34996050792879274, + "grad_norm": 0.2599908709526062, + "learning_rate": 7.299031685269392e-05, + "loss": 1.0447, + "step": 5760 + }, + { + "epoch": 0.3500212649614193, + "grad_norm": 0.5497962236404419, + "learning_rate": 7.298181562496209e-05, + "loss": 1.1516, + "step": 5761 + }, + { + "epoch": 0.3500820219940458, + "grad_norm": 0.20130595564842224, + "learning_rate": 7.297331355482737e-05, + "loss": 1.0625, + "step": 5762 + }, + { + "epoch": 0.35014277902667235, + "grad_norm": 0.16679467260837555, + "learning_rate": 7.296481064260141e-05, + "loss": 1.0531, + "step": 5763 + }, + { + "epoch": 0.3502035360592989, + "grad_norm": 0.5226178765296936, + "learning_rate": 7.295630688859587e-05, + "loss": 1.0786, + "step": 5764 + }, + { + "epoch": 0.3502642930919254, + "grad_norm": 0.385495662689209, + "learning_rate": 7.294780229312246e-05, + "loss": 1.225, + "step": 5765 + }, + { + "epoch": 0.3503250501245519, + "grad_norm": 0.3197796940803528, + "learning_rate": 7.293929685649292e-05, + "loss": 1.061, + "step": 5766 + }, + { + "epoch": 0.35038580715717843, + "grad_norm": 0.23956212401390076, + "learning_rate": 7.293079057901902e-05, + "loss": 1.1524, + "step": 5767 + }, + { + "epoch": 0.35044656418980497, + "grad_norm": 0.19215701520442963, + "learning_rate": 7.292228346101255e-05, + "loss": 1.0872, + "step": 5768 + }, + { + "epoch": 0.3505073212224315, + "grad_norm": 1.1353996992111206, + "learning_rate": 7.291377550278537e-05, + "loss": 1.1224, + "step": 5769 + }, + { + "epoch": 0.35056807825505804, + "grad_norm": 0.5410329103469849, + "learning_rate": 7.290526670464933e-05, + "loss": 1.3085, + "step": 5770 + }, + { + "epoch": 0.35062883528768457, + "grad_norm": 1.792519211769104, + "learning_rate": 7.289675706691628e-05, + "loss": 1.1042, + "step": 5771 + }, + { + "epoch": 0.35068959232031105, + "grad_norm": 0.2121838480234146, + "learning_rate": 7.288824658989818e-05, + "loss": 1.0519, + "step": 5772 + }, + { + "epoch": 0.3507503493529376, + "grad_norm": 0.18687193095684052, + "learning_rate": 7.287973527390699e-05, + "loss": 1.0954, + "step": 5773 + }, + { + "epoch": 0.3508111063855641, + "grad_norm": 0.9593226313591003, + "learning_rate": 7.287122311925467e-05, + "loss": 1.1209, + "step": 5774 + }, + { + "epoch": 0.35087186341819065, + "grad_norm": 0.2516026496887207, + "learning_rate": 7.286271012625326e-05, + "loss": 1.0659, + "step": 5775 + }, + { + "epoch": 0.3509326204508172, + "grad_norm": 0.1833799034357071, + "learning_rate": 7.285419629521477e-05, + "loss": 1.0803, + "step": 5776 + }, + { + "epoch": 0.3509933774834437, + "grad_norm": 0.24927709996700287, + "learning_rate": 7.284568162645131e-05, + "loss": 1.0435, + "step": 5777 + }, + { + "epoch": 0.35105413451607026, + "grad_norm": 0.5010703206062317, + "learning_rate": 7.283716612027497e-05, + "loss": 1.2671, + "step": 5778 + }, + { + "epoch": 0.35111489154869674, + "grad_norm": 0.18309448659420013, + "learning_rate": 7.282864977699789e-05, + "loss": 1.0303, + "step": 5779 + }, + { + "epoch": 0.3511756485813233, + "grad_norm": 0.2068338394165039, + "learning_rate": 7.282013259693225e-05, + "loss": 1.0489, + "step": 5780 + }, + { + "epoch": 0.3512364056139498, + "grad_norm": 0.16653107106685638, + "learning_rate": 7.281161458039022e-05, + "loss": 1.0847, + "step": 5781 + }, + { + "epoch": 0.35129716264657634, + "grad_norm": 2.4543557167053223, + "learning_rate": 7.280309572768407e-05, + "loss": 1.126, + "step": 5782 + }, + { + "epoch": 0.3513579196792029, + "grad_norm": 0.3572819232940674, + "learning_rate": 7.279457603912603e-05, + "loss": 1.2925, + "step": 5783 + }, + { + "epoch": 0.3514186767118294, + "grad_norm": 0.1606816053390503, + "learning_rate": 7.27860555150284e-05, + "loss": 1.0301, + "step": 5784 + }, + { + "epoch": 0.3514794337444559, + "grad_norm": 0.19758851826190948, + "learning_rate": 7.277753415570349e-05, + "loss": 1.1224, + "step": 5785 + }, + { + "epoch": 0.35154019077708243, + "grad_norm": 0.1966302990913391, + "learning_rate": 7.276901196146367e-05, + "loss": 1.1739, + "step": 5786 + }, + { + "epoch": 0.35160094780970896, + "grad_norm": 0.2936049997806549, + "learning_rate": 7.276048893262132e-05, + "loss": 1.0675, + "step": 5787 + }, + { + "epoch": 0.3516617048423355, + "grad_norm": 0.16774167120456696, + "learning_rate": 7.275196506948885e-05, + "loss": 1.0344, + "step": 5788 + }, + { + "epoch": 0.35172246187496203, + "grad_norm": 0.31567466259002686, + "learning_rate": 7.274344037237868e-05, + "loss": 1.2644, + "step": 5789 + }, + { + "epoch": 0.35178321890758857, + "grad_norm": 0.26671212911605835, + "learning_rate": 7.273491484160334e-05, + "loss": 1.0691, + "step": 5790 + }, + { + "epoch": 0.3518439759402151, + "grad_norm": 0.25029870867729187, + "learning_rate": 7.272638847747528e-05, + "loss": 1.0496, + "step": 5791 + }, + { + "epoch": 0.3519047329728416, + "grad_norm": 0.3433615565299988, + "learning_rate": 7.271786128030706e-05, + "loss": 1.2082, + "step": 5792 + }, + { + "epoch": 0.3519654900054681, + "grad_norm": 0.15879976749420166, + "learning_rate": 7.270933325041125e-05, + "loss": 1.0387, + "step": 5793 + }, + { + "epoch": 0.35202624703809465, + "grad_norm": 0.24053643643856049, + "learning_rate": 7.270080438810043e-05, + "loss": 1.0618, + "step": 5794 + }, + { + "epoch": 0.3520870040707212, + "grad_norm": 0.253104567527771, + "learning_rate": 7.269227469368725e-05, + "loss": 1.0793, + "step": 5795 + }, + { + "epoch": 0.3521477611033477, + "grad_norm": 0.2387893795967102, + "learning_rate": 7.268374416748434e-05, + "loss": 1.1204, + "step": 5796 + }, + { + "epoch": 0.35220851813597426, + "grad_norm": 0.3321709632873535, + "learning_rate": 7.267521280980442e-05, + "loss": 1.1095, + "step": 5797 + }, + { + "epoch": 0.3522692751686008, + "grad_norm": 0.5780216455459595, + "learning_rate": 7.266668062096018e-05, + "loss": 1.3212, + "step": 5798 + }, + { + "epoch": 0.3523300322012273, + "grad_norm": 0.40298137068748474, + "learning_rate": 7.265814760126439e-05, + "loss": 1.1505, + "step": 5799 + }, + { + "epoch": 0.3523907892338538, + "grad_norm": 0.21227653324604034, + "learning_rate": 7.26496137510298e-05, + "loss": 1.0499, + "step": 5800 + }, + { + "epoch": 0.35245154626648034, + "grad_norm": 4.4540510177612305, + "learning_rate": 7.264107907056928e-05, + "loss": 1.1228, + "step": 5801 + }, + { + "epoch": 0.3525123032991069, + "grad_norm": 0.26766711473464966, + "learning_rate": 7.263254356019558e-05, + "loss": 1.0717, + "step": 5802 + }, + { + "epoch": 0.3525730603317334, + "grad_norm": 0.3056172728538513, + "learning_rate": 7.262400722022165e-05, + "loss": 1.168, + "step": 5803 + }, + { + "epoch": 0.35263381736435995, + "grad_norm": 0.34795325994491577, + "learning_rate": 7.261547005096036e-05, + "loss": 1.1165, + "step": 5804 + }, + { + "epoch": 0.3526945743969864, + "grad_norm": 0.20817117393016815, + "learning_rate": 7.260693205272463e-05, + "loss": 1.123, + "step": 5805 + }, + { + "epoch": 0.35275533142961296, + "grad_norm": 0.22067409753799438, + "learning_rate": 7.259839322582747e-05, + "loss": 1.0391, + "step": 5806 + }, + { + "epoch": 0.3528160884622395, + "grad_norm": 1.9492067098617554, + "learning_rate": 7.25898535705818e-05, + "loss": 1.047, + "step": 5807 + }, + { + "epoch": 0.35287684549486603, + "grad_norm": 0.31844770908355713, + "learning_rate": 7.258131308730069e-05, + "loss": 1.1309, + "step": 5808 + }, + { + "epoch": 0.35293760252749257, + "grad_norm": 0.28544968366622925, + "learning_rate": 7.25727717762972e-05, + "loss": 1.1542, + "step": 5809 + }, + { + "epoch": 0.3529983595601191, + "grad_norm": 0.3531951904296875, + "learning_rate": 7.256422963788438e-05, + "loss": 1.2349, + "step": 5810 + }, + { + "epoch": 0.35305911659274564, + "grad_norm": 0.35208800435066223, + "learning_rate": 7.255568667237537e-05, + "loss": 1.2358, + "step": 5811 + }, + { + "epoch": 0.3531198736253721, + "grad_norm": 0.1911199986934662, + "learning_rate": 7.25471428800833e-05, + "loss": 1.1177, + "step": 5812 + }, + { + "epoch": 0.35318063065799865, + "grad_norm": 0.3050733804702759, + "learning_rate": 7.253859826132137e-05, + "loss": 1.1788, + "step": 5813 + }, + { + "epoch": 0.3532413876906252, + "grad_norm": 0.36744076013565063, + "learning_rate": 7.253005281640275e-05, + "loss": 1.2287, + "step": 5814 + }, + { + "epoch": 0.3533021447232517, + "grad_norm": 0.2238994836807251, + "learning_rate": 7.252150654564067e-05, + "loss": 1.0854, + "step": 5815 + }, + { + "epoch": 0.35336290175587826, + "grad_norm": 0.4832395017147064, + "learning_rate": 7.251295944934843e-05, + "loss": 1.2188, + "step": 5816 + }, + { + "epoch": 0.3534236587885048, + "grad_norm": 1.470828652381897, + "learning_rate": 7.250441152783931e-05, + "loss": 1.1859, + "step": 5817 + }, + { + "epoch": 0.35348441582113127, + "grad_norm": 0.36085245013237, + "learning_rate": 7.249586278142664e-05, + "loss": 1.1068, + "step": 5818 + }, + { + "epoch": 0.3535451728537578, + "grad_norm": 0.3683243691921234, + "learning_rate": 7.248731321042377e-05, + "loss": 1.1232, + "step": 5819 + }, + { + "epoch": 0.35360592988638434, + "grad_norm": 0.2512708604335785, + "learning_rate": 7.247876281514408e-05, + "loss": 1.1016, + "step": 5820 + }, + { + "epoch": 0.3536666869190109, + "grad_norm": 0.3547723591327667, + "learning_rate": 7.2470211595901e-05, + "loss": 1.131, + "step": 5821 + }, + { + "epoch": 0.3537274439516374, + "grad_norm": 0.3073154091835022, + "learning_rate": 7.246165955300796e-05, + "loss": 1.0544, + "step": 5822 + }, + { + "epoch": 0.35378820098426395, + "grad_norm": 0.2123650163412094, + "learning_rate": 7.245310668677848e-05, + "loss": 1.0479, + "step": 5823 + }, + { + "epoch": 0.3538489580168905, + "grad_norm": 7.976345539093018, + "learning_rate": 7.2444552997526e-05, + "loss": 1.0973, + "step": 5824 + }, + { + "epoch": 0.35390971504951696, + "grad_norm": 8.218181610107422, + "learning_rate": 7.24359984855641e-05, + "loss": 1.1469, + "step": 5825 + }, + { + "epoch": 0.3539704720821435, + "grad_norm": 0.523672342300415, + "learning_rate": 7.242744315120634e-05, + "loss": 1.0995, + "step": 5826 + }, + { + "epoch": 0.35403122911477003, + "grad_norm": 0.7876652479171753, + "learning_rate": 7.241888699476632e-05, + "loss": 1.1745, + "step": 5827 + }, + { + "epoch": 0.35409198614739656, + "grad_norm": 0.641087532043457, + "learning_rate": 7.241033001655765e-05, + "loss": 1.0153, + "step": 5828 + }, + { + "epoch": 0.3541527431800231, + "grad_norm": 0.4092211425304413, + "learning_rate": 7.240177221689401e-05, + "loss": 1.0737, + "step": 5829 + }, + { + "epoch": 0.35421350021264963, + "grad_norm": 0.3513365089893341, + "learning_rate": 7.239321359608908e-05, + "loss": 1.0613, + "step": 5830 + }, + { + "epoch": 0.3542742572452761, + "grad_norm": 0.5521125197410583, + "learning_rate": 7.238465415445659e-05, + "loss": 1.0354, + "step": 5831 + }, + { + "epoch": 0.35433501427790265, + "grad_norm": 0.2168331891298294, + "learning_rate": 7.237609389231025e-05, + "loss": 1.1413, + "step": 5832 + }, + { + "epoch": 0.3543957713105292, + "grad_norm": 0.4371474087238312, + "learning_rate": 7.23675328099639e-05, + "loss": 1.1125, + "step": 5833 + }, + { + "epoch": 0.3544565283431557, + "grad_norm": 1.0865225791931152, + "learning_rate": 7.235897090773127e-05, + "loss": 1.08, + "step": 5834 + }, + { + "epoch": 0.35451728537578225, + "grad_norm": 0.2314506471157074, + "learning_rate": 7.235040818592627e-05, + "loss": 1.2207, + "step": 5835 + }, + { + "epoch": 0.3545780424084088, + "grad_norm": 0.4576619267463684, + "learning_rate": 7.234184464486271e-05, + "loss": 1.0877, + "step": 5836 + }, + { + "epoch": 0.3546387994410353, + "grad_norm": 0.1701354831457138, + "learning_rate": 7.233328028485454e-05, + "loss": 1.0075, + "step": 5837 + }, + { + "epoch": 0.3546995564736618, + "grad_norm": 0.2704429030418396, + "learning_rate": 7.232471510621565e-05, + "loss": 1.0385, + "step": 5838 + }, + { + "epoch": 0.35476031350628834, + "grad_norm": 0.42151540517807007, + "learning_rate": 7.231614910926002e-05, + "loss": 1.0903, + "step": 5839 + }, + { + "epoch": 0.3548210705389149, + "grad_norm": 0.7713186740875244, + "learning_rate": 7.230758229430164e-05, + "loss": 1.4484, + "step": 5840 + }, + { + "epoch": 0.3548818275715414, + "grad_norm": 0.511639416217804, + "learning_rate": 7.229901466165449e-05, + "loss": 1.119, + "step": 5841 + }, + { + "epoch": 0.35494258460416794, + "grad_norm": 0.5013548135757446, + "learning_rate": 7.229044621163267e-05, + "loss": 1.0386, + "step": 5842 + }, + { + "epoch": 0.3550033416367945, + "grad_norm": 0.2581771910190582, + "learning_rate": 7.228187694455023e-05, + "loss": 1.1074, + "step": 5843 + }, + { + "epoch": 0.35506409866942096, + "grad_norm": 0.6767067313194275, + "learning_rate": 7.227330686072129e-05, + "loss": 1.1004, + "step": 5844 + }, + { + "epoch": 0.3551248557020475, + "grad_norm": 0.34328657388687134, + "learning_rate": 7.226473596045997e-05, + "loss": 1.0537, + "step": 5845 + }, + { + "epoch": 0.355185612734674, + "grad_norm": 0.42628318071365356, + "learning_rate": 7.225616424408045e-05, + "loss": 1.1226, + "step": 5846 + }, + { + "epoch": 0.35524636976730056, + "grad_norm": 0.5616112947463989, + "learning_rate": 7.224759171189694e-05, + "loss": 1.1032, + "step": 5847 + }, + { + "epoch": 0.3553071267999271, + "grad_norm": 0.25972259044647217, + "learning_rate": 7.223901836422366e-05, + "loss": 1.0868, + "step": 5848 + }, + { + "epoch": 0.35536788383255363, + "grad_norm": 0.39373013377189636, + "learning_rate": 7.223044420137484e-05, + "loss": 1.0559, + "step": 5849 + }, + { + "epoch": 0.35542864086518017, + "grad_norm": 0.5329267382621765, + "learning_rate": 7.22218692236648e-05, + "loss": 1.2472, + "step": 5850 + }, + { + "epoch": 0.35548939789780665, + "grad_norm": 0.2477867752313614, + "learning_rate": 7.221329343140784e-05, + "loss": 1.1207, + "step": 5851 + }, + { + "epoch": 0.3555501549304332, + "grad_norm": 0.4808276891708374, + "learning_rate": 7.220471682491834e-05, + "loss": 1.1379, + "step": 5852 + }, + { + "epoch": 0.3556109119630597, + "grad_norm": 0.20636846125125885, + "learning_rate": 7.219613940451064e-05, + "loss": 1.0569, + "step": 5853 + }, + { + "epoch": 0.35567166899568625, + "grad_norm": 0.39592546224594116, + "learning_rate": 7.218756117049915e-05, + "loss": 1.1166, + "step": 5854 + }, + { + "epoch": 0.3557324260283128, + "grad_norm": 0.4000478684902191, + "learning_rate": 7.217898212319833e-05, + "loss": 1.1843, + "step": 5855 + }, + { + "epoch": 0.3557931830609393, + "grad_norm": 0.46339693665504456, + "learning_rate": 7.217040226292262e-05, + "loss": 1.1277, + "step": 5856 + }, + { + "epoch": 0.35585394009356586, + "grad_norm": 0.4452223479747772, + "learning_rate": 7.216182158998654e-05, + "loss": 1.1386, + "step": 5857 + }, + { + "epoch": 0.35591469712619234, + "grad_norm": 0.2495124191045761, + "learning_rate": 7.215324010470462e-05, + "loss": 1.0725, + "step": 5858 + }, + { + "epoch": 0.35597545415881887, + "grad_norm": 0.24482768774032593, + "learning_rate": 7.214465780739137e-05, + "loss": 1.0437, + "step": 5859 + }, + { + "epoch": 0.3560362111914454, + "grad_norm": 0.2162971794605255, + "learning_rate": 7.213607469836143e-05, + "loss": 1.1215, + "step": 5860 + }, + { + "epoch": 0.35609696822407194, + "grad_norm": 0.29403597116470337, + "learning_rate": 7.212749077792937e-05, + "loss": 1.0237, + "step": 5861 + }, + { + "epoch": 0.3561577252566985, + "grad_norm": 0.216622993350029, + "learning_rate": 7.211890604640989e-05, + "loss": 1.1807, + "step": 5862 + }, + { + "epoch": 0.356218482289325, + "grad_norm": 0.4531051814556122, + "learning_rate": 7.211032050411763e-05, + "loss": 1.1399, + "step": 5863 + }, + { + "epoch": 0.3562792393219515, + "grad_norm": 0.21599365770816803, + "learning_rate": 7.210173415136727e-05, + "loss": 1.1081, + "step": 5864 + }, + { + "epoch": 0.356339996354578, + "grad_norm": 2.140284538269043, + "learning_rate": 7.209314698847359e-05, + "loss": 1.2474, + "step": 5865 + }, + { + "epoch": 0.35640075338720456, + "grad_norm": 0.410271555185318, + "learning_rate": 7.208455901575133e-05, + "loss": 1.2414, + "step": 5866 + }, + { + "epoch": 0.3564615104198311, + "grad_norm": 0.4960329234600067, + "learning_rate": 7.20759702335153e-05, + "loss": 1.1495, + "step": 5867 + }, + { + "epoch": 0.35652226745245763, + "grad_norm": 0.38220521807670593, + "learning_rate": 7.206738064208031e-05, + "loss": 1.0514, + "step": 5868 + }, + { + "epoch": 0.35658302448508417, + "grad_norm": 0.39340558648109436, + "learning_rate": 7.205879024176121e-05, + "loss": 1.0941, + "step": 5869 + }, + { + "epoch": 0.3566437815177107, + "grad_norm": 0.19361025094985962, + "learning_rate": 7.205019903287291e-05, + "loss": 1.0854, + "step": 5870 + }, + { + "epoch": 0.3567045385503372, + "grad_norm": 0.3317127823829651, + "learning_rate": 7.204160701573028e-05, + "loss": 1.3019, + "step": 5871 + }, + { + "epoch": 0.3567652955829637, + "grad_norm": 0.24956105649471283, + "learning_rate": 7.203301419064828e-05, + "loss": 1.1167, + "step": 5872 + }, + { + "epoch": 0.35682605261559025, + "grad_norm": 0.24158872663974762, + "learning_rate": 7.202442055794192e-05, + "loss": 1.1002, + "step": 5873 + }, + { + "epoch": 0.3568868096482168, + "grad_norm": 0.237736776471138, + "learning_rate": 7.201582611792614e-05, + "loss": 1.0654, + "step": 5874 + }, + { + "epoch": 0.3569475666808433, + "grad_norm": 0.24741022288799286, + "learning_rate": 7.2007230870916e-05, + "loss": 1.1483, + "step": 5875 + }, + { + "epoch": 0.35700832371346986, + "grad_norm": 0.3427708148956299, + "learning_rate": 7.199863481722656e-05, + "loss": 1.2358, + "step": 5876 + }, + { + "epoch": 0.35706908074609633, + "grad_norm": 0.3130645155906677, + "learning_rate": 7.199003795717292e-05, + "loss": 1.0503, + "step": 5877 + }, + { + "epoch": 0.35712983777872287, + "grad_norm": 0.19013427197933197, + "learning_rate": 7.198144029107018e-05, + "loss": 1.091, + "step": 5878 + }, + { + "epoch": 0.3571905948113494, + "grad_norm": 0.424010694026947, + "learning_rate": 7.19728418192335e-05, + "loss": 1.0916, + "step": 5879 + }, + { + "epoch": 0.35725135184397594, + "grad_norm": 0.2775021493434906, + "learning_rate": 7.196424254197804e-05, + "loss": 1.1451, + "step": 5880 + }, + { + "epoch": 0.3573121088766025, + "grad_norm": 0.5044656991958618, + "learning_rate": 7.195564245961904e-05, + "loss": 1.1323, + "step": 5881 + }, + { + "epoch": 0.357372865909229, + "grad_norm": 0.18651695549488068, + "learning_rate": 7.194704157247172e-05, + "loss": 1.0547, + "step": 5882 + }, + { + "epoch": 0.35743362294185554, + "grad_norm": 0.4497082531452179, + "learning_rate": 7.193843988085135e-05, + "loss": 1.1833, + "step": 5883 + }, + { + "epoch": 0.357494379974482, + "grad_norm": 0.2812330424785614, + "learning_rate": 7.192983738507322e-05, + "loss": 1.0942, + "step": 5884 + }, + { + "epoch": 0.35755513700710856, + "grad_norm": 0.20607328414916992, + "learning_rate": 7.192123408545266e-05, + "loss": 1.0988, + "step": 5885 + }, + { + "epoch": 0.3576158940397351, + "grad_norm": 0.30754441022872925, + "learning_rate": 7.191262998230504e-05, + "loss": 1.2414, + "step": 5886 + }, + { + "epoch": 0.35767665107236163, + "grad_norm": 0.3936502933502197, + "learning_rate": 7.190402507594571e-05, + "loss": 1.154, + "step": 5887 + }, + { + "epoch": 0.35773740810498816, + "grad_norm": 0.4713932275772095, + "learning_rate": 7.189541936669013e-05, + "loss": 1.1586, + "step": 5888 + }, + { + "epoch": 0.3577981651376147, + "grad_norm": 0.46565139293670654, + "learning_rate": 7.188681285485371e-05, + "loss": 1.0671, + "step": 5889 + }, + { + "epoch": 0.3578589221702412, + "grad_norm": 2.7555177211761475, + "learning_rate": 7.187820554075193e-05, + "loss": 1.2267, + "step": 5890 + }, + { + "epoch": 0.3579196792028677, + "grad_norm": 0.23018629848957062, + "learning_rate": 7.186959742470031e-05, + "loss": 1.1352, + "step": 5891 + }, + { + "epoch": 0.35798043623549425, + "grad_norm": 0.4853644073009491, + "learning_rate": 7.186098850701435e-05, + "loss": 1.1465, + "step": 5892 + }, + { + "epoch": 0.3580411932681208, + "grad_norm": 0.2814134359359741, + "learning_rate": 7.185237878800966e-05, + "loss": 1.2432, + "step": 5893 + }, + { + "epoch": 0.3581019503007473, + "grad_norm": 0.14376969635486603, + "learning_rate": 7.184376826800178e-05, + "loss": 1.0605, + "step": 5894 + }, + { + "epoch": 0.35816270733337385, + "grad_norm": 0.23665255308151245, + "learning_rate": 7.183515694730635e-05, + "loss": 1.273, + "step": 5895 + }, + { + "epoch": 0.3582234643660004, + "grad_norm": 0.25462818145751953, + "learning_rate": 7.182654482623903e-05, + "loss": 1.1282, + "step": 5896 + }, + { + "epoch": 0.35828422139862687, + "grad_norm": 0.18486633896827698, + "learning_rate": 7.181793190511549e-05, + "loss": 1.093, + "step": 5897 + }, + { + "epoch": 0.3583449784312534, + "grad_norm": 0.351715087890625, + "learning_rate": 7.180931818425142e-05, + "loss": 1.2094, + "step": 5898 + }, + { + "epoch": 0.35840573546387994, + "grad_norm": 0.24343159794807434, + "learning_rate": 7.180070366396261e-05, + "loss": 1.1651, + "step": 5899 + }, + { + "epoch": 0.3584664924965065, + "grad_norm": 0.22065769135951996, + "learning_rate": 7.179208834456477e-05, + "loss": 1.0331, + "step": 5900 + }, + { + "epoch": 0.358527249529133, + "grad_norm": 0.3812246024608612, + "learning_rate": 7.178347222637373e-05, + "loss": 1.0883, + "step": 5901 + }, + { + "epoch": 0.35858800656175954, + "grad_norm": 0.24011486768722534, + "learning_rate": 7.17748553097053e-05, + "loss": 1.0238, + "step": 5902 + }, + { + "epoch": 0.3586487635943861, + "grad_norm": 0.5111322402954102, + "learning_rate": 7.176623759487534e-05, + "loss": 1.0765, + "step": 5903 + }, + { + "epoch": 0.35870952062701256, + "grad_norm": 0.20378683507442474, + "learning_rate": 7.175761908219974e-05, + "loss": 1.0871, + "step": 5904 + }, + { + "epoch": 0.3587702776596391, + "grad_norm": 0.37354815006256104, + "learning_rate": 7.174899977199439e-05, + "loss": 1.2218, + "step": 5905 + }, + { + "epoch": 0.3588310346922656, + "grad_norm": 0.21111074090003967, + "learning_rate": 7.174037966457525e-05, + "loss": 1.1209, + "step": 5906 + }, + { + "epoch": 0.35889179172489216, + "grad_norm": 0.1863575428724289, + "learning_rate": 7.173175876025831e-05, + "loss": 1.0759, + "step": 5907 + }, + { + "epoch": 0.3589525487575187, + "grad_norm": 0.36061999201774597, + "learning_rate": 7.172313705935955e-05, + "loss": 1.0576, + "step": 5908 + }, + { + "epoch": 0.35901330579014523, + "grad_norm": 0.20433875918388367, + "learning_rate": 7.171451456219499e-05, + "loss": 1.0779, + "step": 5909 + }, + { + "epoch": 0.3590740628227717, + "grad_norm": 0.2103491574525833, + "learning_rate": 7.17058912690807e-05, + "loss": 1.133, + "step": 5910 + }, + { + "epoch": 0.35913481985539825, + "grad_norm": 0.44911956787109375, + "learning_rate": 7.169726718033278e-05, + "loss": 1.2741, + "step": 5911 + }, + { + "epoch": 0.3591955768880248, + "grad_norm": 0.27871084213256836, + "learning_rate": 7.168864229626734e-05, + "loss": 1.0822, + "step": 5912 + }, + { + "epoch": 0.3592563339206513, + "grad_norm": 0.34126710891723633, + "learning_rate": 7.168001661720051e-05, + "loss": 1.0405, + "step": 5913 + }, + { + "epoch": 0.35931709095327785, + "grad_norm": 0.3373902440071106, + "learning_rate": 7.16713901434485e-05, + "loss": 1.1915, + "step": 5914 + }, + { + "epoch": 0.3593778479859044, + "grad_norm": 0.18636579811573029, + "learning_rate": 7.166276287532748e-05, + "loss": 1.1255, + "step": 5915 + }, + { + "epoch": 0.3594386050185309, + "grad_norm": 0.4257730543613434, + "learning_rate": 7.16541348131537e-05, + "loss": 1.0656, + "step": 5916 + }, + { + "epoch": 0.3594993620511574, + "grad_norm": 0.30280590057373047, + "learning_rate": 7.164550595724344e-05, + "loss": 1.103, + "step": 5917 + }, + { + "epoch": 0.35956011908378394, + "grad_norm": 1.0134764909744263, + "learning_rate": 7.163687630791294e-05, + "loss": 1.0949, + "step": 5918 + }, + { + "epoch": 0.35962087611641047, + "grad_norm": 0.3119845390319824, + "learning_rate": 7.162824586547858e-05, + "loss": 1.1007, + "step": 5919 + }, + { + "epoch": 0.359681633149037, + "grad_norm": 0.2806399166584015, + "learning_rate": 7.161961463025669e-05, + "loss": 1.1485, + "step": 5920 + }, + { + "epoch": 0.35974239018166354, + "grad_norm": 0.3692055642604828, + "learning_rate": 7.161098260256363e-05, + "loss": 1.1336, + "step": 5921 + }, + { + "epoch": 0.3598031472142901, + "grad_norm": 0.2712574005126953, + "learning_rate": 7.160234978271585e-05, + "loss": 1.1776, + "step": 5922 + }, + { + "epoch": 0.35986390424691655, + "grad_norm": 0.2192717045545578, + "learning_rate": 7.159371617102975e-05, + "loss": 1.07, + "step": 5923 + }, + { + "epoch": 0.3599246612795431, + "grad_norm": 0.3257123827934265, + "learning_rate": 7.15850817678218e-05, + "loss": 1.1299, + "step": 5924 + }, + { + "epoch": 0.3599854183121696, + "grad_norm": 0.3663870096206665, + "learning_rate": 7.157644657340852e-05, + "loss": 1.0902, + "step": 5925 + }, + { + "epoch": 0.36004617534479616, + "grad_norm": 0.25223222374916077, + "learning_rate": 7.15678105881064e-05, + "loss": 1.1436, + "step": 5926 + }, + { + "epoch": 0.3601069323774227, + "grad_norm": 0.29753226041793823, + "learning_rate": 7.155917381223205e-05, + "loss": 1.1359, + "step": 5927 + }, + { + "epoch": 0.36016768941004923, + "grad_norm": 0.7590225338935852, + "learning_rate": 7.1550536246102e-05, + "loss": 1.0915, + "step": 5928 + }, + { + "epoch": 0.36022844644267576, + "grad_norm": 0.27148616313934326, + "learning_rate": 7.154189789003286e-05, + "loss": 1.0694, + "step": 5929 + }, + { + "epoch": 0.36028920347530224, + "grad_norm": 0.5021152496337891, + "learning_rate": 7.153325874434131e-05, + "loss": 1.0359, + "step": 5930 + }, + { + "epoch": 0.3603499605079288, + "grad_norm": 0.17821384966373444, + "learning_rate": 7.152461880934399e-05, + "loss": 1.0915, + "step": 5931 + }, + { + "epoch": 0.3604107175405553, + "grad_norm": 0.2173178791999817, + "learning_rate": 7.151597808535761e-05, + "loss": 1.0437, + "step": 5932 + }, + { + "epoch": 0.36047147457318185, + "grad_norm": 0.2564549446105957, + "learning_rate": 7.150733657269889e-05, + "loss": 1.0667, + "step": 5933 + }, + { + "epoch": 0.3605322316058084, + "grad_norm": 0.9256630539894104, + "learning_rate": 7.14986942716846e-05, + "loss": 1.1235, + "step": 5934 + }, + { + "epoch": 0.3605929886384349, + "grad_norm": 5.113297462463379, + "learning_rate": 7.14900511826315e-05, + "loss": 1.0885, + "step": 5935 + }, + { + "epoch": 0.3606537456710614, + "grad_norm": 1.4798799753189087, + "learning_rate": 7.148140730585644e-05, + "loss": 1.0828, + "step": 5936 + }, + { + "epoch": 0.36071450270368793, + "grad_norm": 0.927744448184967, + "learning_rate": 7.147276264167623e-05, + "loss": 1.1513, + "step": 5937 + }, + { + "epoch": 0.36077525973631447, + "grad_norm": 0.32261425256729126, + "learning_rate": 7.146411719040775e-05, + "loss": 1.0806, + "step": 5938 + }, + { + "epoch": 0.360836016768941, + "grad_norm": 0.2541137933731079, + "learning_rate": 7.145547095236791e-05, + "loss": 1.0449, + "step": 5939 + }, + { + "epoch": 0.36089677380156754, + "grad_norm": 0.27033689618110657, + "learning_rate": 7.144682392787365e-05, + "loss": 1.099, + "step": 5940 + }, + { + "epoch": 0.3609575308341941, + "grad_norm": 0.3376336395740509, + "learning_rate": 7.14381761172419e-05, + "loss": 1.203, + "step": 5941 + }, + { + "epoch": 0.3610182878668206, + "grad_norm": 0.23571033775806427, + "learning_rate": 7.142952752078965e-05, + "loss": 1.0574, + "step": 5942 + }, + { + "epoch": 0.3610790448994471, + "grad_norm": 0.3035906255245209, + "learning_rate": 7.142087813883395e-05, + "loss": 1.2241, + "step": 5943 + }, + { + "epoch": 0.3611398019320736, + "grad_norm": 0.1998484581708908, + "learning_rate": 7.14122279716918e-05, + "loss": 1.0628, + "step": 5944 + }, + { + "epoch": 0.36120055896470016, + "grad_norm": 0.9037440419197083, + "learning_rate": 7.140357701968031e-05, + "loss": 1.1042, + "step": 5945 + }, + { + "epoch": 0.3612613159973267, + "grad_norm": 0.22878238558769226, + "learning_rate": 7.139492528311657e-05, + "loss": 1.1057, + "step": 5946 + }, + { + "epoch": 0.36132207302995323, + "grad_norm": 0.27864599227905273, + "learning_rate": 7.138627276231768e-05, + "loss": 1.3209, + "step": 5947 + }, + { + "epoch": 0.36138283006257976, + "grad_norm": 0.41246727108955383, + "learning_rate": 7.137761945760086e-05, + "loss": 1.1532, + "step": 5948 + }, + { + "epoch": 0.36144358709520624, + "grad_norm": 0.46069201827049255, + "learning_rate": 7.136896536928326e-05, + "loss": 1.1771, + "step": 5949 + }, + { + "epoch": 0.3615043441278328, + "grad_norm": 1.4408278465270996, + "learning_rate": 7.13603104976821e-05, + "loss": 1.0958, + "step": 5950 + }, + { + "epoch": 0.3615651011604593, + "grad_norm": 0.369762659072876, + "learning_rate": 7.135165484311462e-05, + "loss": 1.2635, + "step": 5951 + }, + { + "epoch": 0.36162585819308585, + "grad_norm": 0.23664060235023499, + "learning_rate": 7.134299840589811e-05, + "loss": 1.0923, + "step": 5952 + }, + { + "epoch": 0.3616866152257124, + "grad_norm": 0.27583807706832886, + "learning_rate": 7.133434118634986e-05, + "loss": 1.0835, + "step": 5953 + }, + { + "epoch": 0.3617473722583389, + "grad_norm": 0.41345030069351196, + "learning_rate": 7.132568318478722e-05, + "loss": 1.2189, + "step": 5954 + }, + { + "epoch": 0.36180812929096545, + "grad_norm": 0.37629038095474243, + "learning_rate": 7.131702440152755e-05, + "loss": 1.0856, + "step": 5955 + }, + { + "epoch": 0.36186888632359193, + "grad_norm": 0.2190614491701126, + "learning_rate": 7.130836483688821e-05, + "loss": 1.0628, + "step": 5956 + }, + { + "epoch": 0.36192964335621847, + "grad_norm": 0.3250643312931061, + "learning_rate": 7.129970449118666e-05, + "loss": 1.0763, + "step": 5957 + }, + { + "epoch": 0.361990400388845, + "grad_norm": 3.925217866897583, + "learning_rate": 7.129104336474031e-05, + "loss": 1.1414, + "step": 5958 + }, + { + "epoch": 0.36205115742147154, + "grad_norm": 0.26151242852211, + "learning_rate": 7.128238145786665e-05, + "loss": 1.1153, + "step": 5959 + }, + { + "epoch": 0.36211191445409807, + "grad_norm": 0.3245666027069092, + "learning_rate": 7.12737187708832e-05, + "loss": 1.0244, + "step": 5960 + }, + { + "epoch": 0.3621726714867246, + "grad_norm": 0.25276991724967957, + "learning_rate": 7.126505530410746e-05, + "loss": 1.1878, + "step": 5961 + }, + { + "epoch": 0.36223342851935114, + "grad_norm": 0.4960314631462097, + "learning_rate": 7.125639105785699e-05, + "loss": 1.1492, + "step": 5962 + }, + { + "epoch": 0.3622941855519776, + "grad_norm": 0.4006102979183197, + "learning_rate": 7.124772603244943e-05, + "loss": 1.1552, + "step": 5963 + }, + { + "epoch": 0.36235494258460416, + "grad_norm": 0.23062191903591156, + "learning_rate": 7.123906022820238e-05, + "loss": 1.0996, + "step": 5964 + }, + { + "epoch": 0.3624156996172307, + "grad_norm": 0.2751602828502655, + "learning_rate": 7.123039364543342e-05, + "loss": 1.0953, + "step": 5965 + }, + { + "epoch": 0.3624764566498572, + "grad_norm": 0.14705193042755127, + "learning_rate": 7.122172628446031e-05, + "loss": 0.9846, + "step": 5966 + }, + { + "epoch": 0.36253721368248376, + "grad_norm": 0.24618010222911835, + "learning_rate": 7.12130581456007e-05, + "loss": 1.2015, + "step": 5967 + }, + { + "epoch": 0.3625979707151103, + "grad_norm": 0.2509921193122864, + "learning_rate": 7.120438922917236e-05, + "loss": 1.1812, + "step": 5968 + }, + { + "epoch": 0.3626587277477368, + "grad_norm": 0.2760959267616272, + "learning_rate": 7.119571953549305e-05, + "loss": 1.2072, + "step": 5969 + }, + { + "epoch": 0.3627194847803633, + "grad_norm": 0.1874544620513916, + "learning_rate": 7.118704906488052e-05, + "loss": 1.059, + "step": 5970 + }, + { + "epoch": 0.36278024181298985, + "grad_norm": 0.206515833735466, + "learning_rate": 7.117837781765261e-05, + "loss": 1.1062, + "step": 5971 + }, + { + "epoch": 0.3628409988456164, + "grad_norm": 0.24718137085437775, + "learning_rate": 7.116970579412717e-05, + "loss": 1.1791, + "step": 5972 + }, + { + "epoch": 0.3629017558782429, + "grad_norm": 0.276242196559906, + "learning_rate": 7.116103299462207e-05, + "loss": 1.1569, + "step": 5973 + }, + { + "epoch": 0.36296251291086945, + "grad_norm": 0.3344092071056366, + "learning_rate": 7.115235941945524e-05, + "loss": 1.1682, + "step": 5974 + }, + { + "epoch": 0.363023269943496, + "grad_norm": 0.21671617031097412, + "learning_rate": 7.114368506894456e-05, + "loss": 1.0983, + "step": 5975 + }, + { + "epoch": 0.36308402697612246, + "grad_norm": 2.0468785762786865, + "learning_rate": 7.113500994340801e-05, + "loss": 1.1847, + "step": 5976 + }, + { + "epoch": 0.363144784008749, + "grad_norm": 0.24501799046993256, + "learning_rate": 7.112633404316361e-05, + "loss": 1.0694, + "step": 5977 + }, + { + "epoch": 0.36320554104137553, + "grad_norm": 0.23404110968112946, + "learning_rate": 7.111765736852934e-05, + "loss": 1.1925, + "step": 5978 + }, + { + "epoch": 0.36326629807400207, + "grad_norm": 0.3558104932308197, + "learning_rate": 7.110897991982326e-05, + "loss": 1.1789, + "step": 5979 + }, + { + "epoch": 0.3633270551066286, + "grad_norm": 0.44063472747802734, + "learning_rate": 7.110030169736344e-05, + "loss": 1.2016, + "step": 5980 + }, + { + "epoch": 0.36338781213925514, + "grad_norm": 0.22934550046920776, + "learning_rate": 7.109162270146797e-05, + "loss": 1.1747, + "step": 5981 + }, + { + "epoch": 0.3634485691718816, + "grad_norm": 0.39520901441574097, + "learning_rate": 7.108294293245501e-05, + "loss": 1.1648, + "step": 5982 + }, + { + "epoch": 0.36350932620450815, + "grad_norm": 0.22019106149673462, + "learning_rate": 7.107426239064269e-05, + "loss": 1.0378, + "step": 5983 + }, + { + "epoch": 0.3635700832371347, + "grad_norm": 0.2645670473575592, + "learning_rate": 7.106558107634923e-05, + "loss": 1.2474, + "step": 5984 + }, + { + "epoch": 0.3636308402697612, + "grad_norm": 0.2817346453666687, + "learning_rate": 7.10568989898928e-05, + "loss": 1.0515, + "step": 5985 + }, + { + "epoch": 0.36369159730238776, + "grad_norm": 0.3286506235599518, + "learning_rate": 7.104821613159167e-05, + "loss": 1.2429, + "step": 5986 + }, + { + "epoch": 0.3637523543350143, + "grad_norm": 0.3681623041629791, + "learning_rate": 7.103953250176413e-05, + "loss": 1.1024, + "step": 5987 + }, + { + "epoch": 0.36381311136764083, + "grad_norm": 0.5349355936050415, + "learning_rate": 7.103084810072843e-05, + "loss": 1.2029, + "step": 5988 + }, + { + "epoch": 0.3638738684002673, + "grad_norm": 0.2615949809551239, + "learning_rate": 7.102216292880295e-05, + "loss": 1.1163, + "step": 5989 + }, + { + "epoch": 0.36393462543289384, + "grad_norm": 0.27036231756210327, + "learning_rate": 7.101347698630601e-05, + "loss": 1.1309, + "step": 5990 + }, + { + "epoch": 0.3639953824655204, + "grad_norm": 0.17866019904613495, + "learning_rate": 7.100479027355603e-05, + "loss": 1.0632, + "step": 5991 + }, + { + "epoch": 0.3640561394981469, + "grad_norm": 0.7059449553489685, + "learning_rate": 7.099610279087138e-05, + "loss": 1.1916, + "step": 5992 + }, + { + "epoch": 0.36411689653077345, + "grad_norm": 0.2810089588165283, + "learning_rate": 7.098741453857053e-05, + "loss": 1.1673, + "step": 5993 + }, + { + "epoch": 0.3641776535634, + "grad_norm": 0.2594994902610779, + "learning_rate": 7.097872551697195e-05, + "loss": 1.0454, + "step": 5994 + }, + { + "epoch": 0.36423841059602646, + "grad_norm": 0.18744207918643951, + "learning_rate": 7.097003572639413e-05, + "loss": 1.1085, + "step": 5995 + }, + { + "epoch": 0.364299167628653, + "grad_norm": 0.2545723617076874, + "learning_rate": 7.096134516715557e-05, + "loss": 1.0958, + "step": 5996 + }, + { + "epoch": 0.36435992466127953, + "grad_norm": 0.19261644780635834, + "learning_rate": 7.09526538395749e-05, + "loss": 1.0993, + "step": 5997 + }, + { + "epoch": 0.36442068169390607, + "grad_norm": 0.4278690814971924, + "learning_rate": 7.094396174397061e-05, + "loss": 1.0969, + "step": 5998 + }, + { + "epoch": 0.3644814387265326, + "grad_norm": 0.20948807895183563, + "learning_rate": 7.093526888066137e-05, + "loss": 1.1433, + "step": 5999 + }, + { + "epoch": 0.36454219575915914, + "grad_norm": 0.2865559160709381, + "learning_rate": 7.09265752499658e-05, + "loss": 1.2196, + "step": 6000 + }, + { + "epoch": 0.3646029527917857, + "grad_norm": 0.20242883265018463, + "learning_rate": 7.091788085220259e-05, + "loss": 1.1175, + "step": 6001 + }, + { + "epoch": 0.36466370982441215, + "grad_norm": 0.2483610063791275, + "learning_rate": 7.09091856876904e-05, + "loss": 1.1109, + "step": 6002 + }, + { + "epoch": 0.3647244668570387, + "grad_norm": 0.2752256989479065, + "learning_rate": 7.090048975674796e-05, + "loss": 1.2346, + "step": 6003 + }, + { + "epoch": 0.3647852238896652, + "grad_norm": 0.20909246802330017, + "learning_rate": 7.089179305969403e-05, + "loss": 1.0809, + "step": 6004 + }, + { + "epoch": 0.36484598092229176, + "grad_norm": 0.20199701189994812, + "learning_rate": 7.088309559684739e-05, + "loss": 1.037, + "step": 6005 + }, + { + "epoch": 0.3649067379549183, + "grad_norm": 0.351783812046051, + "learning_rate": 7.087439736852684e-05, + "loss": 1.0217, + "step": 6006 + }, + { + "epoch": 0.3649674949875448, + "grad_norm": 6.01610803604126, + "learning_rate": 7.086569837505122e-05, + "loss": 1.1648, + "step": 6007 + }, + { + "epoch": 0.36502825202017136, + "grad_norm": 0.1928926408290863, + "learning_rate": 7.08569986167394e-05, + "loss": 1.1257, + "step": 6008 + }, + { + "epoch": 0.36508900905279784, + "grad_norm": 1.3307048082351685, + "learning_rate": 7.084829809391026e-05, + "loss": 1.1204, + "step": 6009 + }, + { + "epoch": 0.3651497660854244, + "grad_norm": 0.22861313819885254, + "learning_rate": 7.083959680688273e-05, + "loss": 1.0963, + "step": 6010 + }, + { + "epoch": 0.3652105231180509, + "grad_norm": 0.17970292270183563, + "learning_rate": 7.083089475597574e-05, + "loss": 1.0898, + "step": 6011 + }, + { + "epoch": 0.36527128015067745, + "grad_norm": 0.1525614857673645, + "learning_rate": 7.082219194150828e-05, + "loss": 1.0415, + "step": 6012 + }, + { + "epoch": 0.365332037183304, + "grad_norm": 0.23378543555736542, + "learning_rate": 7.081348836379935e-05, + "loss": 1.0506, + "step": 6013 + }, + { + "epoch": 0.3653927942159305, + "grad_norm": 0.1888275444507599, + "learning_rate": 7.080478402316798e-05, + "loss": 1.0823, + "step": 6014 + }, + { + "epoch": 0.365453551248557, + "grad_norm": 0.2924087643623352, + "learning_rate": 7.079607891993323e-05, + "loss": 1.1094, + "step": 6015 + }, + { + "epoch": 0.36551430828118353, + "grad_norm": 8.997772216796875, + "learning_rate": 7.07873730544142e-05, + "loss": 1.0719, + "step": 6016 + }, + { + "epoch": 0.36557506531381007, + "grad_norm": 0.31928735971450806, + "learning_rate": 7.077866642692998e-05, + "loss": 1.1074, + "step": 6017 + }, + { + "epoch": 0.3656358223464366, + "grad_norm": 0.1892041265964508, + "learning_rate": 7.076995903779972e-05, + "loss": 1.1745, + "step": 6018 + }, + { + "epoch": 0.36569657937906314, + "grad_norm": 0.18936464190483093, + "learning_rate": 7.07612508873426e-05, + "loss": 1.026, + "step": 6019 + }, + { + "epoch": 0.36575733641168967, + "grad_norm": 0.17004993557929993, + "learning_rate": 7.075254197587782e-05, + "loss": 1.0246, + "step": 6020 + }, + { + "epoch": 0.3658180934443162, + "grad_norm": 0.2953880727291107, + "learning_rate": 7.074383230372461e-05, + "loss": 1.0659, + "step": 6021 + }, + { + "epoch": 0.3658788504769427, + "grad_norm": 0.23347707092761993, + "learning_rate": 7.073512187120221e-05, + "loss": 1.1085, + "step": 6022 + }, + { + "epoch": 0.3659396075095692, + "grad_norm": 1.2416003942489624, + "learning_rate": 7.072641067862991e-05, + "loss": 1.2566, + "step": 6023 + }, + { + "epoch": 0.36600036454219576, + "grad_norm": 0.7075762748718262, + "learning_rate": 7.071769872632701e-05, + "loss": 1.3222, + "step": 6024 + }, + { + "epoch": 0.3660611215748223, + "grad_norm": 0.2512654662132263, + "learning_rate": 7.070898601461287e-05, + "loss": 1.0669, + "step": 6025 + }, + { + "epoch": 0.3661218786074488, + "grad_norm": 0.2368144392967224, + "learning_rate": 7.070027254380684e-05, + "loss": 1.0419, + "step": 6026 + }, + { + "epoch": 0.36618263564007536, + "grad_norm": 0.17379184067249298, + "learning_rate": 7.069155831422832e-05, + "loss": 1.0614, + "step": 6027 + }, + { + "epoch": 0.36624339267270184, + "grad_norm": 0.2701704204082489, + "learning_rate": 7.068284332619673e-05, + "loss": 1.0684, + "step": 6028 + }, + { + "epoch": 0.3663041497053284, + "grad_norm": 0.38034093379974365, + "learning_rate": 7.067412758003154e-05, + "loss": 1.1072, + "step": 6029 + }, + { + "epoch": 0.3663649067379549, + "grad_norm": 0.24150525033473969, + "learning_rate": 7.066541107605217e-05, + "loss": 1.1622, + "step": 6030 + }, + { + "epoch": 0.36642566377058144, + "grad_norm": 0.3052763044834137, + "learning_rate": 7.065669381457819e-05, + "loss": 1.128, + "step": 6031 + }, + { + "epoch": 0.366486420803208, + "grad_norm": 0.41426753997802734, + "learning_rate": 7.064797579592909e-05, + "loss": 1.047, + "step": 6032 + }, + { + "epoch": 0.3665471778358345, + "grad_norm": 0.3082740902900696, + "learning_rate": 7.063925702042446e-05, + "loss": 1.0689, + "step": 6033 + }, + { + "epoch": 0.36660793486846105, + "grad_norm": 0.29588383436203003, + "learning_rate": 7.063053748838386e-05, + "loss": 1.1062, + "step": 6034 + }, + { + "epoch": 0.36666869190108753, + "grad_norm": 0.2586400508880615, + "learning_rate": 7.062181720012693e-05, + "loss": 1.1812, + "step": 6035 + }, + { + "epoch": 0.36672944893371406, + "grad_norm": 0.33517301082611084, + "learning_rate": 7.06130961559733e-05, + "loss": 1.0711, + "step": 6036 + }, + { + "epoch": 0.3667902059663406, + "grad_norm": 0.4143451452255249, + "learning_rate": 7.060437435624265e-05, + "loss": 1.0562, + "step": 6037 + }, + { + "epoch": 0.36685096299896713, + "grad_norm": 0.3315509855747223, + "learning_rate": 7.059565180125466e-05, + "loss": 1.1939, + "step": 6038 + }, + { + "epoch": 0.36691172003159367, + "grad_norm": 0.23199841380119324, + "learning_rate": 7.05869284913291e-05, + "loss": 1.0496, + "step": 6039 + }, + { + "epoch": 0.3669724770642202, + "grad_norm": 0.2603539228439331, + "learning_rate": 7.057820442678567e-05, + "loss": 1.2617, + "step": 6040 + }, + { + "epoch": 0.3670332340968467, + "grad_norm": 2.0048351287841797, + "learning_rate": 7.05694796079442e-05, + "loss": 1.0773, + "step": 6041 + }, + { + "epoch": 0.3670939911294732, + "grad_norm": 0.3135457932949066, + "learning_rate": 7.056075403512448e-05, + "loss": 1.2205, + "step": 6042 + }, + { + "epoch": 0.36715474816209975, + "grad_norm": 0.23415252566337585, + "learning_rate": 7.055202770864633e-05, + "loss": 1.0754, + "step": 6043 + }, + { + "epoch": 0.3672155051947263, + "grad_norm": 0.18994826078414917, + "learning_rate": 7.054330062882964e-05, + "loss": 1.0619, + "step": 6044 + }, + { + "epoch": 0.3672762622273528, + "grad_norm": 0.30746370553970337, + "learning_rate": 7.05345727959943e-05, + "loss": 1.2823, + "step": 6045 + }, + { + "epoch": 0.36733701925997936, + "grad_norm": 0.19453735649585724, + "learning_rate": 7.05258442104602e-05, + "loss": 1.1042, + "step": 6046 + }, + { + "epoch": 0.3673977762926059, + "grad_norm": 0.4125863015651703, + "learning_rate": 7.051711487254734e-05, + "loss": 1.0974, + "step": 6047 + }, + { + "epoch": 0.3674585333252324, + "grad_norm": 0.24339830875396729, + "learning_rate": 7.050838478257566e-05, + "loss": 1.0368, + "step": 6048 + }, + { + "epoch": 0.3675192903578589, + "grad_norm": 0.38127291202545166, + "learning_rate": 7.049965394086518e-05, + "loss": 1.0487, + "step": 6049 + }, + { + "epoch": 0.36758004739048544, + "grad_norm": 0.2161187082529068, + "learning_rate": 7.04909223477359e-05, + "loss": 1.0785, + "step": 6050 + }, + { + "epoch": 0.367640804423112, + "grad_norm": 0.26314160227775574, + "learning_rate": 7.048219000350792e-05, + "loss": 1.1976, + "step": 6051 + }, + { + "epoch": 0.3677015614557385, + "grad_norm": 0.43870389461517334, + "learning_rate": 7.04734569085013e-05, + "loss": 1.015, + "step": 6052 + }, + { + "epoch": 0.36776231848836505, + "grad_norm": 1.343644142150879, + "learning_rate": 7.046472306303616e-05, + "loss": 1.1026, + "step": 6053 + }, + { + "epoch": 0.3678230755209915, + "grad_norm": 0.44411394000053406, + "learning_rate": 7.045598846743264e-05, + "loss": 1.2173, + "step": 6054 + }, + { + "epoch": 0.36788383255361806, + "grad_norm": 3.50353741645813, + "learning_rate": 7.04472531220109e-05, + "loss": 1.3747, + "step": 6055 + }, + { + "epoch": 0.3679445895862446, + "grad_norm": 0.4374399185180664, + "learning_rate": 7.043851702709115e-05, + "loss": 1.1524, + "step": 6056 + }, + { + "epoch": 0.36800534661887113, + "grad_norm": 0.2078254371881485, + "learning_rate": 7.042978018299361e-05, + "loss": 1.1075, + "step": 6057 + }, + { + "epoch": 0.36806610365149767, + "grad_norm": 0.5986235737800598, + "learning_rate": 7.042104259003853e-05, + "loss": 1.1644, + "step": 6058 + }, + { + "epoch": 0.3681268606841242, + "grad_norm": 0.2968313694000244, + "learning_rate": 7.041230424854619e-05, + "loss": 1.0914, + "step": 6059 + }, + { + "epoch": 0.36818761771675074, + "grad_norm": 0.34527862071990967, + "learning_rate": 7.04035651588369e-05, + "loss": 1.058, + "step": 6060 + }, + { + "epoch": 0.3682483747493772, + "grad_norm": 0.41134896874427795, + "learning_rate": 7.039482532123096e-05, + "loss": 1.119, + "step": 6061 + }, + { + "epoch": 0.36830913178200375, + "grad_norm": 1.033369541168213, + "learning_rate": 7.038608473604877e-05, + "loss": 1.1896, + "step": 6062 + }, + { + "epoch": 0.3683698888146303, + "grad_norm": 0.22340096533298492, + "learning_rate": 7.037734340361069e-05, + "loss": 1.1188, + "step": 6063 + }, + { + "epoch": 0.3684306458472568, + "grad_norm": 0.33844539523124695, + "learning_rate": 7.036860132423718e-05, + "loss": 1.3323, + "step": 6064 + }, + { + "epoch": 0.36849140287988336, + "grad_norm": 0.3055606484413147, + "learning_rate": 7.035985849824863e-05, + "loss": 1.0768, + "step": 6065 + }, + { + "epoch": 0.3685521599125099, + "grad_norm": 0.23188263177871704, + "learning_rate": 7.035111492596552e-05, + "loss": 1.0983, + "step": 6066 + }, + { + "epoch": 0.3686129169451364, + "grad_norm": 0.6486999988555908, + "learning_rate": 7.034237060770838e-05, + "loss": 1.1418, + "step": 6067 + }, + { + "epoch": 0.3686736739777629, + "grad_norm": 0.2795666754245758, + "learning_rate": 7.03336255437977e-05, + "loss": 1.0891, + "step": 6068 + }, + { + "epoch": 0.36873443101038944, + "grad_norm": 0.2658239006996155, + "learning_rate": 7.032487973455406e-05, + "loss": 1.1063, + "step": 6069 + }, + { + "epoch": 0.368795188043016, + "grad_norm": 0.4116172194480896, + "learning_rate": 7.031613318029802e-05, + "loss": 1.0481, + "step": 6070 + }, + { + "epoch": 0.3688559450756425, + "grad_norm": 0.29068508744239807, + "learning_rate": 7.030738588135018e-05, + "loss": 1.062, + "step": 6071 + }, + { + "epoch": 0.36891670210826905, + "grad_norm": 0.23946480453014374, + "learning_rate": 7.02986378380312e-05, + "loss": 1.0447, + "step": 6072 + }, + { + "epoch": 0.3689774591408956, + "grad_norm": 0.31897827982902527, + "learning_rate": 7.028988905066171e-05, + "loss": 1.0746, + "step": 6073 + }, + { + "epoch": 0.36903821617352206, + "grad_norm": 0.42437320947647095, + "learning_rate": 7.028113951956242e-05, + "loss": 1.032, + "step": 6074 + }, + { + "epoch": 0.3690989732061486, + "grad_norm": 0.1866168975830078, + "learning_rate": 7.027238924505404e-05, + "loss": 1.0419, + "step": 6075 + }, + { + "epoch": 0.36915973023877513, + "grad_norm": 0.21309222280979156, + "learning_rate": 7.026363822745732e-05, + "loss": 1.1156, + "step": 6076 + }, + { + "epoch": 0.36922048727140167, + "grad_norm": 0.4237158000469208, + "learning_rate": 7.025488646709302e-05, + "loss": 1.1144, + "step": 6077 + }, + { + "epoch": 0.3692812443040282, + "grad_norm": 0.3203827738761902, + "learning_rate": 7.024613396428195e-05, + "loss": 1.1876, + "step": 6078 + }, + { + "epoch": 0.36934200133665473, + "grad_norm": 0.26472795009613037, + "learning_rate": 7.023738071934491e-05, + "loss": 1.0673, + "step": 6079 + }, + { + "epoch": 0.36940275836928127, + "grad_norm": 0.24500873684883118, + "learning_rate": 7.022862673260278e-05, + "loss": 1.0616, + "step": 6080 + }, + { + "epoch": 0.36946351540190775, + "grad_norm": 0.19023045897483826, + "learning_rate": 7.021987200437642e-05, + "loss": 1.087, + "step": 6081 + }, + { + "epoch": 0.3695242724345343, + "grad_norm": 0.38321763277053833, + "learning_rate": 7.021111653498675e-05, + "loss": 1.0354, + "step": 6082 + }, + { + "epoch": 0.3695850294671608, + "grad_norm": 0.3026789128780365, + "learning_rate": 7.02023603247547e-05, + "loss": 1.0985, + "step": 6083 + }, + { + "epoch": 0.36964578649978735, + "grad_norm": 0.27910709381103516, + "learning_rate": 7.019360337400122e-05, + "loss": 1.3649, + "step": 6084 + }, + { + "epoch": 0.3697065435324139, + "grad_norm": 0.4078048765659332, + "learning_rate": 7.01848456830473e-05, + "loss": 1.0893, + "step": 6085 + }, + { + "epoch": 0.3697673005650404, + "grad_norm": 0.26925551891326904, + "learning_rate": 7.017608725221397e-05, + "loss": 1.1433, + "step": 6086 + }, + { + "epoch": 0.3698280575976669, + "grad_norm": 0.17030160129070282, + "learning_rate": 7.016732808182224e-05, + "loss": 1.0552, + "step": 6087 + }, + { + "epoch": 0.36988881463029344, + "grad_norm": 0.29098063707351685, + "learning_rate": 7.015856817219324e-05, + "loss": 1.2475, + "step": 6088 + }, + { + "epoch": 0.36994957166292, + "grad_norm": 0.2074175477027893, + "learning_rate": 7.014980752364799e-05, + "loss": 1.0563, + "step": 6089 + }, + { + "epoch": 0.3700103286955465, + "grad_norm": 0.2527397871017456, + "learning_rate": 7.014104613650766e-05, + "loss": 1.1466, + "step": 6090 + }, + { + "epoch": 0.37007108572817304, + "grad_norm": 0.21011464297771454, + "learning_rate": 7.013228401109341e-05, + "loss": 1.0425, + "step": 6091 + }, + { + "epoch": 0.3701318427607996, + "grad_norm": 0.6035009026527405, + "learning_rate": 7.012352114772638e-05, + "loss": 1.046, + "step": 6092 + }, + { + "epoch": 0.3701925997934261, + "grad_norm": 0.24817050993442535, + "learning_rate": 7.011475754672781e-05, + "loss": 1.2857, + "step": 6093 + }, + { + "epoch": 0.3702533568260526, + "grad_norm": 0.32921817898750305, + "learning_rate": 7.010599320841889e-05, + "loss": 1.0102, + "step": 6094 + }, + { + "epoch": 0.37031411385867913, + "grad_norm": 0.6056867837905884, + "learning_rate": 7.009722813312092e-05, + "loss": 1.1324, + "step": 6095 + }, + { + "epoch": 0.37037487089130566, + "grad_norm": 0.17226547002792358, + "learning_rate": 7.008846232115515e-05, + "loss": 1.1034, + "step": 6096 + }, + { + "epoch": 0.3704356279239322, + "grad_norm": 0.2278028130531311, + "learning_rate": 7.007969577284292e-05, + "loss": 1.1585, + "step": 6097 + }, + { + "epoch": 0.37049638495655873, + "grad_norm": 0.35627973079681396, + "learning_rate": 7.007092848850559e-05, + "loss": 1.0665, + "step": 6098 + }, + { + "epoch": 0.37055714198918527, + "grad_norm": 0.26589345932006836, + "learning_rate": 7.006216046846444e-05, + "loss": 1.0943, + "step": 6099 + }, + { + "epoch": 0.37061789902181175, + "grad_norm": 0.18078437447547913, + "learning_rate": 7.005339171304095e-05, + "loss": 1.0543, + "step": 6100 + }, + { + "epoch": 0.3706786560544383, + "grad_norm": 0.21096277236938477, + "learning_rate": 7.004462222255653e-05, + "loss": 1.0944, + "step": 6101 + }, + { + "epoch": 0.3707394130870648, + "grad_norm": 0.1661684215068817, + "learning_rate": 7.003585199733258e-05, + "loss": 1.0703, + "step": 6102 + }, + { + "epoch": 0.37080017011969135, + "grad_norm": 0.17436473071575165, + "learning_rate": 7.00270810376906e-05, + "loss": 1.0658, + "step": 6103 + }, + { + "epoch": 0.3708609271523179, + "grad_norm": 0.1906927525997162, + "learning_rate": 7.00183093439521e-05, + "loss": 1.0885, + "step": 6104 + }, + { + "epoch": 0.3709216841849444, + "grad_norm": 0.2379797399044037, + "learning_rate": 7.000953691643861e-05, + "loss": 1.1498, + "step": 6105 + }, + { + "epoch": 0.37098244121757096, + "grad_norm": 0.1962262988090515, + "learning_rate": 7.000076375547168e-05, + "loss": 1.0427, + "step": 6106 + }, + { + "epoch": 0.37104319825019744, + "grad_norm": 0.6283131241798401, + "learning_rate": 6.999198986137287e-05, + "loss": 1.1828, + "step": 6107 + }, + { + "epoch": 0.37110395528282397, + "grad_norm": 0.20437495410442352, + "learning_rate": 6.998321523446383e-05, + "loss": 1.0494, + "step": 6108 + }, + { + "epoch": 0.3711647123154505, + "grad_norm": 0.38200441002845764, + "learning_rate": 6.997443987506616e-05, + "loss": 1.2785, + "step": 6109 + }, + { + "epoch": 0.37122546934807704, + "grad_norm": 0.2615712881088257, + "learning_rate": 6.996566378350152e-05, + "loss": 1.1172, + "step": 6110 + }, + { + "epoch": 0.3712862263807036, + "grad_norm": 0.21745048463344574, + "learning_rate": 6.995688696009164e-05, + "loss": 1.2084, + "step": 6111 + }, + { + "epoch": 0.3713469834133301, + "grad_norm": 0.24408693611621857, + "learning_rate": 6.994810940515819e-05, + "loss": 1.191, + "step": 6112 + }, + { + "epoch": 0.37140774044595665, + "grad_norm": 1.8709906339645386, + "learning_rate": 6.993933111902296e-05, + "loss": 1.2071, + "step": 6113 + }, + { + "epoch": 0.3714684974785831, + "grad_norm": 0.39469924569129944, + "learning_rate": 6.993055210200768e-05, + "loss": 1.084, + "step": 6114 + }, + { + "epoch": 0.37152925451120966, + "grad_norm": 0.288495808839798, + "learning_rate": 6.992177235443416e-05, + "loss": 1.2103, + "step": 6115 + }, + { + "epoch": 0.3715900115438362, + "grad_norm": 0.2579781413078308, + "learning_rate": 6.991299187662421e-05, + "loss": 1.204, + "step": 6116 + }, + { + "epoch": 0.37165076857646273, + "grad_norm": 0.22606410086154938, + "learning_rate": 6.990421066889971e-05, + "loss": 1.0494, + "step": 6117 + }, + { + "epoch": 0.37171152560908927, + "grad_norm": 0.3533797860145569, + "learning_rate": 6.989542873158252e-05, + "loss": 1.1245, + "step": 6118 + }, + { + "epoch": 0.3717722826417158, + "grad_norm": 0.3024650812149048, + "learning_rate": 6.988664606499454e-05, + "loss": 1.1657, + "step": 6119 + }, + { + "epoch": 0.3718330396743423, + "grad_norm": 0.2895429730415344, + "learning_rate": 6.98778626694577e-05, + "loss": 1.0881, + "step": 6120 + }, + { + "epoch": 0.3718937967069688, + "grad_norm": 0.20209649205207825, + "learning_rate": 6.986907854529397e-05, + "loss": 1.0108, + "step": 6121 + }, + { + "epoch": 0.37195455373959535, + "grad_norm": 0.30950626730918884, + "learning_rate": 6.986029369282533e-05, + "loss": 1.1737, + "step": 6122 + }, + { + "epoch": 0.3720153107722219, + "grad_norm": 0.1779547780752182, + "learning_rate": 6.985150811237377e-05, + "loss": 1.0346, + "step": 6123 + }, + { + "epoch": 0.3720760678048484, + "grad_norm": 0.2577000558376312, + "learning_rate": 6.984272180426134e-05, + "loss": 1.0966, + "step": 6124 + }, + { + "epoch": 0.37213682483747496, + "grad_norm": 0.6275244951248169, + "learning_rate": 6.983393476881013e-05, + "loss": 1.0795, + "step": 6125 + }, + { + "epoch": 0.3721975818701015, + "grad_norm": 0.15838930010795593, + "learning_rate": 6.98251470063422e-05, + "loss": 1.0251, + "step": 6126 + }, + { + "epoch": 0.37225833890272797, + "grad_norm": 0.4600166380405426, + "learning_rate": 6.981635851717966e-05, + "loss": 1.2267, + "step": 6127 + }, + { + "epoch": 0.3723190959353545, + "grad_norm": 0.24281449615955353, + "learning_rate": 6.980756930164468e-05, + "loss": 1.193, + "step": 6128 + }, + { + "epoch": 0.37237985296798104, + "grad_norm": 0.2159792184829712, + "learning_rate": 6.979877936005942e-05, + "loss": 1.1377, + "step": 6129 + }, + { + "epoch": 0.3724406100006076, + "grad_norm": 0.6401416063308716, + "learning_rate": 6.978998869274606e-05, + "loss": 1.0761, + "step": 6130 + }, + { + "epoch": 0.3725013670332341, + "grad_norm": 0.2548671066761017, + "learning_rate": 6.978119730002686e-05, + "loss": 1.0253, + "step": 6131 + }, + { + "epoch": 0.37256212406586064, + "grad_norm": 0.22326402366161346, + "learning_rate": 6.977240518222403e-05, + "loss": 1.0734, + "step": 6132 + }, + { + "epoch": 0.3726228810984871, + "grad_norm": 0.6491445302963257, + "learning_rate": 6.976361233965988e-05, + "loss": 1.1321, + "step": 6133 + }, + { + "epoch": 0.37268363813111366, + "grad_norm": 0.2918967604637146, + "learning_rate": 6.97548187726567e-05, + "loss": 1.1722, + "step": 6134 + }, + { + "epoch": 0.3727443951637402, + "grad_norm": 0.2472890168428421, + "learning_rate": 6.974602448153683e-05, + "loss": 1.0811, + "step": 6135 + }, + { + "epoch": 0.37280515219636673, + "grad_norm": 0.21059446036815643, + "learning_rate": 6.973722946662258e-05, + "loss": 1.1057, + "step": 6136 + }, + { + "epoch": 0.37286590922899326, + "grad_norm": 0.21334360539913177, + "learning_rate": 6.972843372823638e-05, + "loss": 1.1592, + "step": 6137 + }, + { + "epoch": 0.3729266662616198, + "grad_norm": 0.2115211933851242, + "learning_rate": 6.971963726670065e-05, + "loss": 1.0825, + "step": 6138 + }, + { + "epoch": 0.37298742329424633, + "grad_norm": 0.20887506008148193, + "learning_rate": 6.97108400823378e-05, + "loss": 1.0965, + "step": 6139 + }, + { + "epoch": 0.3730481803268728, + "grad_norm": 0.2530198097229004, + "learning_rate": 6.970204217547027e-05, + "loss": 1.165, + "step": 6140 + }, + { + "epoch": 0.37310893735949935, + "grad_norm": 0.20424573123455048, + "learning_rate": 6.96932435464206e-05, + "loss": 1.0704, + "step": 6141 + }, + { + "epoch": 0.3731696943921259, + "grad_norm": 0.2597198188304901, + "learning_rate": 6.968444419551128e-05, + "loss": 1.1469, + "step": 6142 + }, + { + "epoch": 0.3732304514247524, + "grad_norm": 0.22950716316699982, + "learning_rate": 6.967564412306485e-05, + "loss": 1.0, + "step": 6143 + }, + { + "epoch": 0.37329120845737895, + "grad_norm": 0.22117146849632263, + "learning_rate": 6.966684332940388e-05, + "loss": 1.1219, + "step": 6144 + }, + { + "epoch": 0.3733519654900055, + "grad_norm": 0.24216118454933167, + "learning_rate": 6.965804181485097e-05, + "loss": 1.0985, + "step": 6145 + }, + { + "epoch": 0.37341272252263197, + "grad_norm": 0.2078687846660614, + "learning_rate": 6.964923957972872e-05, + "loss": 1.0558, + "step": 6146 + }, + { + "epoch": 0.3734734795552585, + "grad_norm": 0.18559493124485016, + "learning_rate": 6.964043662435981e-05, + "loss": 1.121, + "step": 6147 + }, + { + "epoch": 0.37353423658788504, + "grad_norm": 0.15101692080497742, + "learning_rate": 6.963163294906691e-05, + "loss": 1.0317, + "step": 6148 + }, + { + "epoch": 0.3735949936205116, + "grad_norm": 0.24683165550231934, + "learning_rate": 6.962282855417269e-05, + "loss": 1.1251, + "step": 6149 + }, + { + "epoch": 0.3736557506531381, + "grad_norm": 0.3027634918689728, + "learning_rate": 6.96140234399999e-05, + "loss": 1.0782, + "step": 6150 + }, + { + "epoch": 0.37371650768576464, + "grad_norm": 0.48501235246658325, + "learning_rate": 6.960521760687129e-05, + "loss": 1.0767, + "step": 6151 + }, + { + "epoch": 0.3737772647183912, + "grad_norm": 0.20950312912464142, + "learning_rate": 6.959641105510966e-05, + "loss": 1.0175, + "step": 6152 + }, + { + "epoch": 0.37383802175101766, + "grad_norm": 0.32944241166114807, + "learning_rate": 6.958760378503776e-05, + "loss": 1.0518, + "step": 6153 + }, + { + "epoch": 0.3738987787836442, + "grad_norm": 0.2912905514240265, + "learning_rate": 6.957879579697848e-05, + "loss": 1.1919, + "step": 6154 + }, + { + "epoch": 0.3739595358162707, + "grad_norm": 0.2776426672935486, + "learning_rate": 6.956998709125467e-05, + "loss": 1.1111, + "step": 6155 + }, + { + "epoch": 0.37402029284889726, + "grad_norm": 2.230964422225952, + "learning_rate": 6.956117766818915e-05, + "loss": 1.0557, + "step": 6156 + }, + { + "epoch": 0.3740810498815238, + "grad_norm": 0.284554123878479, + "learning_rate": 6.955236752810493e-05, + "loss": 1.0788, + "step": 6157 + }, + { + "epoch": 0.37414180691415033, + "grad_norm": 0.30509552359580994, + "learning_rate": 6.954355667132489e-05, + "loss": 1.2342, + "step": 6158 + }, + { + "epoch": 0.3742025639467768, + "grad_norm": 0.4198652505874634, + "learning_rate": 6.953474509817197e-05, + "loss": 1.13, + "step": 6159 + }, + { + "epoch": 0.37426332097940335, + "grad_norm": 0.2295794039964676, + "learning_rate": 6.952593280896922e-05, + "loss": 1.0649, + "step": 6160 + }, + { + "epoch": 0.3743240780120299, + "grad_norm": 0.3075230121612549, + "learning_rate": 6.951711980403963e-05, + "loss": 1.1023, + "step": 6161 + }, + { + "epoch": 0.3743848350446564, + "grad_norm": 0.5523928999900818, + "learning_rate": 6.950830608370623e-05, + "loss": 1.1376, + "step": 6162 + }, + { + "epoch": 0.37444559207728295, + "grad_norm": 0.3659617304801941, + "learning_rate": 6.94994916482921e-05, + "loss": 1.2391, + "step": 6163 + }, + { + "epoch": 0.3745063491099095, + "grad_norm": 0.4639756977558136, + "learning_rate": 6.949067649812033e-05, + "loss": 1.0492, + "step": 6164 + }, + { + "epoch": 0.374567106142536, + "grad_norm": 0.5739428400993347, + "learning_rate": 6.948186063351407e-05, + "loss": 1.3835, + "step": 6165 + }, + { + "epoch": 0.3746278631751625, + "grad_norm": 0.23870936036109924, + "learning_rate": 6.947304405479644e-05, + "loss": 1.0881, + "step": 6166 + }, + { + "epoch": 0.37468862020778904, + "grad_norm": 0.538027286529541, + "learning_rate": 6.946422676229058e-05, + "loss": 1.2374, + "step": 6167 + }, + { + "epoch": 0.37474937724041557, + "grad_norm": 0.20781545341014862, + "learning_rate": 6.945540875631977e-05, + "loss": 1.1127, + "step": 6168 + }, + { + "epoch": 0.3748101342730421, + "grad_norm": 0.21300722658634186, + "learning_rate": 6.944659003720716e-05, + "loss": 1.0897, + "step": 6169 + }, + { + "epoch": 0.37487089130566864, + "grad_norm": 0.2649407684803009, + "learning_rate": 6.943777060527605e-05, + "loss": 1.1037, + "step": 6170 + }, + { + "epoch": 0.3749316483382952, + "grad_norm": 0.18946529924869537, + "learning_rate": 6.94289504608497e-05, + "loss": 1.0229, + "step": 6171 + }, + { + "epoch": 0.3749924053709217, + "grad_norm": 1.2179205417633057, + "learning_rate": 6.94201296042514e-05, + "loss": 1.0429, + "step": 6172 + }, + { + "epoch": 0.3750531624035482, + "grad_norm": 0.2897929847240448, + "learning_rate": 6.941130803580451e-05, + "loss": 1.0815, + "step": 6173 + }, + { + "epoch": 0.3751139194361747, + "grad_norm": 1.185583472251892, + "learning_rate": 6.940248575583236e-05, + "loss": 1.0255, + "step": 6174 + }, + { + "epoch": 0.37517467646880126, + "grad_norm": 0.17818598449230194, + "learning_rate": 6.939366276465836e-05, + "loss": 1.0729, + "step": 6175 + }, + { + "epoch": 0.3752354335014278, + "grad_norm": 0.29268550872802734, + "learning_rate": 6.938483906260589e-05, + "loss": 1.1088, + "step": 6176 + }, + { + "epoch": 0.37529619053405433, + "grad_norm": 0.18495263159275055, + "learning_rate": 6.93760146499984e-05, + "loss": 1.0603, + "step": 6177 + }, + { + "epoch": 0.37535694756668087, + "grad_norm": 0.203537717461586, + "learning_rate": 6.936718952715936e-05, + "loss": 1.0968, + "step": 6178 + }, + { + "epoch": 0.37541770459930734, + "grad_norm": 0.32517480850219727, + "learning_rate": 6.935836369441223e-05, + "loss": 1.2005, + "step": 6179 + }, + { + "epoch": 0.3754784616319339, + "grad_norm": 0.2983331084251404, + "learning_rate": 6.934953715208054e-05, + "loss": 1.2482, + "step": 6180 + }, + { + "epoch": 0.3755392186645604, + "grad_norm": 0.2072860598564148, + "learning_rate": 6.934070990048784e-05, + "loss": 1.054, + "step": 6181 + }, + { + "epoch": 0.37559997569718695, + "grad_norm": 0.22101300954818726, + "learning_rate": 6.933188193995766e-05, + "loss": 1.0832, + "step": 6182 + }, + { + "epoch": 0.3756607327298135, + "grad_norm": 0.2372554987668991, + "learning_rate": 6.932305327081361e-05, + "loss": 1.1826, + "step": 6183 + }, + { + "epoch": 0.37572148976244, + "grad_norm": 0.2059735357761383, + "learning_rate": 6.931422389337932e-05, + "loss": 1.1071, + "step": 6184 + }, + { + "epoch": 0.37578224679506655, + "grad_norm": 0.43729692697525024, + "learning_rate": 6.93053938079784e-05, + "loss": 1.1005, + "step": 6185 + }, + { + "epoch": 0.37584300382769303, + "grad_norm": 0.64780193567276, + "learning_rate": 6.929656301493455e-05, + "loss": 1.073, + "step": 6186 + }, + { + "epoch": 0.37590376086031957, + "grad_norm": 0.18860197067260742, + "learning_rate": 6.928773151457147e-05, + "loss": 1.0417, + "step": 6187 + }, + { + "epoch": 0.3759645178929461, + "grad_norm": 0.30241718888282776, + "learning_rate": 6.927889930721286e-05, + "loss": 1.176, + "step": 6188 + }, + { + "epoch": 0.37602527492557264, + "grad_norm": 0.37454894185066223, + "learning_rate": 6.927006639318245e-05, + "loss": 1.2163, + "step": 6189 + }, + { + "epoch": 0.3760860319581992, + "grad_norm": 0.7262923717498779, + "learning_rate": 6.926123277280403e-05, + "loss": 1.0863, + "step": 6190 + }, + { + "epoch": 0.3761467889908257, + "grad_norm": 0.22545242309570312, + "learning_rate": 6.925239844640141e-05, + "loss": 1.0857, + "step": 6191 + }, + { + "epoch": 0.3762075460234522, + "grad_norm": 0.22957634925842285, + "learning_rate": 6.92435634142984e-05, + "loss": 1.104, + "step": 6192 + }, + { + "epoch": 0.3762683030560787, + "grad_norm": 0.5483096241950989, + "learning_rate": 6.923472767681886e-05, + "loss": 1.1253, + "step": 6193 + }, + { + "epoch": 0.37632906008870526, + "grad_norm": 0.20304445922374725, + "learning_rate": 6.922589123428665e-05, + "loss": 1.0979, + "step": 6194 + }, + { + "epoch": 0.3763898171213318, + "grad_norm": 0.23823319375514984, + "learning_rate": 6.92170540870257e-05, + "loss": 1.0548, + "step": 6195 + }, + { + "epoch": 0.37645057415395833, + "grad_norm": 0.2750590145587921, + "learning_rate": 6.920821623535989e-05, + "loss": 1.1669, + "step": 6196 + }, + { + "epoch": 0.37651133118658486, + "grad_norm": 0.19844850897789001, + "learning_rate": 6.919937767961322e-05, + "loss": 1.0556, + "step": 6197 + }, + { + "epoch": 0.3765720882192114, + "grad_norm": 0.24682916700839996, + "learning_rate": 6.919053842010964e-05, + "loss": 1.1588, + "step": 6198 + }, + { + "epoch": 0.3766328452518379, + "grad_norm": 0.38083910942077637, + "learning_rate": 6.918169845717318e-05, + "loss": 1.1103, + "step": 6199 + }, + { + "epoch": 0.3766936022844644, + "grad_norm": 0.35714900493621826, + "learning_rate": 6.917285779112784e-05, + "loss": 1.0785, + "step": 6200 + }, + { + "epoch": 0.37675435931709095, + "grad_norm": 0.23043179512023926, + "learning_rate": 6.916401642229769e-05, + "loss": 1.0888, + "step": 6201 + }, + { + "epoch": 0.3768151163497175, + "grad_norm": 0.23609089851379395, + "learning_rate": 6.915517435100683e-05, + "loss": 1.1059, + "step": 6202 + }, + { + "epoch": 0.376875873382344, + "grad_norm": 1.3536678552627563, + "learning_rate": 6.914633157757932e-05, + "loss": 1.1083, + "step": 6203 + }, + { + "epoch": 0.37693663041497055, + "grad_norm": 0.23241961002349854, + "learning_rate": 6.913748810233935e-05, + "loss": 1.0987, + "step": 6204 + }, + { + "epoch": 0.37699738744759703, + "grad_norm": 0.2243051528930664, + "learning_rate": 6.912864392561105e-05, + "loss": 1.1236, + "step": 6205 + }, + { + "epoch": 0.37705814448022357, + "grad_norm": 0.46866920590400696, + "learning_rate": 6.911979904771862e-05, + "loss": 1.1418, + "step": 6206 + }, + { + "epoch": 0.3771189015128501, + "grad_norm": 0.19177576899528503, + "learning_rate": 6.911095346898624e-05, + "loss": 1.0227, + "step": 6207 + }, + { + "epoch": 0.37717965854547664, + "grad_norm": 4.680105209350586, + "learning_rate": 6.910210718973818e-05, + "loss": 1.2334, + "step": 6208 + }, + { + "epoch": 0.37724041557810317, + "grad_norm": 0.8148078918457031, + "learning_rate": 6.90932602102987e-05, + "loss": 1.1223, + "step": 6209 + }, + { + "epoch": 0.3773011726107297, + "grad_norm": 0.24385613203048706, + "learning_rate": 6.908441253099205e-05, + "loss": 1.1492, + "step": 6210 + }, + { + "epoch": 0.37736192964335624, + "grad_norm": 0.23646089434623718, + "learning_rate": 6.907556415214258e-05, + "loss": 1.0915, + "step": 6211 + }, + { + "epoch": 0.3774226866759827, + "grad_norm": 0.19970782101154327, + "learning_rate": 6.906671507407463e-05, + "loss": 1.08, + "step": 6212 + }, + { + "epoch": 0.37748344370860926, + "grad_norm": 0.24710312485694885, + "learning_rate": 6.905786529711254e-05, + "loss": 1.0693, + "step": 6213 + }, + { + "epoch": 0.3775442007412358, + "grad_norm": 0.23810215294361115, + "learning_rate": 6.90490148215807e-05, + "loss": 1.209, + "step": 6214 + }, + { + "epoch": 0.3776049577738623, + "grad_norm": 0.18675287067890167, + "learning_rate": 6.904016364780358e-05, + "loss": 1.0995, + "step": 6215 + }, + { + "epoch": 0.37766571480648886, + "grad_norm": 0.38546550273895264, + "learning_rate": 6.903131177610554e-05, + "loss": 1.0708, + "step": 6216 + }, + { + "epoch": 0.3777264718391154, + "grad_norm": 0.19149380922317505, + "learning_rate": 6.902245920681112e-05, + "loss": 1.0665, + "step": 6217 + }, + { + "epoch": 0.37778722887174193, + "grad_norm": 0.2784052789211273, + "learning_rate": 6.901360594024477e-05, + "loss": 1.1826, + "step": 6218 + }, + { + "epoch": 0.3778479859043684, + "grad_norm": 0.27127376198768616, + "learning_rate": 6.900475197673101e-05, + "loss": 1.0915, + "step": 6219 + }, + { + "epoch": 0.37790874293699495, + "grad_norm": 0.31531715393066406, + "learning_rate": 6.89958973165944e-05, + "loss": 1.1207, + "step": 6220 + }, + { + "epoch": 0.3779694999696215, + "grad_norm": 0.2799668312072754, + "learning_rate": 6.898704196015949e-05, + "loss": 1.0948, + "step": 6221 + }, + { + "epoch": 0.378030257002248, + "grad_norm": 0.21449260413646698, + "learning_rate": 6.897818590775092e-05, + "loss": 1.1767, + "step": 6222 + }, + { + "epoch": 0.37809101403487455, + "grad_norm": 0.24828341603279114, + "learning_rate": 6.896932915969325e-05, + "loss": 1.153, + "step": 6223 + }, + { + "epoch": 0.3781517710675011, + "grad_norm": 0.742504894733429, + "learning_rate": 6.896047171631114e-05, + "loss": 1.3078, + "step": 6224 + }, + { + "epoch": 0.37821252810012757, + "grad_norm": 0.731936514377594, + "learning_rate": 6.895161357792931e-05, + "loss": 1.1172, + "step": 6225 + }, + { + "epoch": 0.3782732851327541, + "grad_norm": 0.3543374836444855, + "learning_rate": 6.89427547448724e-05, + "loss": 1.0764, + "step": 6226 + }, + { + "epoch": 0.37833404216538064, + "grad_norm": 0.25636300444602966, + "learning_rate": 6.893389521746515e-05, + "loss": 1.0926, + "step": 6227 + }, + { + "epoch": 0.37839479919800717, + "grad_norm": 0.2150675505399704, + "learning_rate": 6.892503499603233e-05, + "loss": 1.0417, + "step": 6228 + }, + { + "epoch": 0.3784555562306337, + "grad_norm": 0.2801735997200012, + "learning_rate": 6.891617408089866e-05, + "loss": 1.1289, + "step": 6229 + }, + { + "epoch": 0.37851631326326024, + "grad_norm": 0.2639968991279602, + "learning_rate": 6.8907312472389e-05, + "loss": 1.0429, + "step": 6230 + }, + { + "epoch": 0.3785770702958868, + "grad_norm": 0.25557443499565125, + "learning_rate": 6.889845017082815e-05, + "loss": 1.112, + "step": 6231 + }, + { + "epoch": 0.37863782732851325, + "grad_norm": 0.25097954273223877, + "learning_rate": 6.888958717654095e-05, + "loss": 1.0718, + "step": 6232 + }, + { + "epoch": 0.3786985843611398, + "grad_norm": 0.23674561083316803, + "learning_rate": 6.888072348985227e-05, + "loss": 1.0359, + "step": 6233 + }, + { + "epoch": 0.3787593413937663, + "grad_norm": 0.2909475564956665, + "learning_rate": 6.887185911108702e-05, + "loss": 1.1009, + "step": 6234 + }, + { + "epoch": 0.37882009842639286, + "grad_norm": 0.2609412372112274, + "learning_rate": 6.886299404057015e-05, + "loss": 1.0924, + "step": 6235 + }, + { + "epoch": 0.3788808554590194, + "grad_norm": 0.3491469621658325, + "learning_rate": 6.885412827862657e-05, + "loss": 1.2001, + "step": 6236 + }, + { + "epoch": 0.37894161249164593, + "grad_norm": 0.24456825852394104, + "learning_rate": 6.884526182558126e-05, + "loss": 1.1644, + "step": 6237 + }, + { + "epoch": 0.3790023695242724, + "grad_norm": 0.21998301148414612, + "learning_rate": 6.883639468175927e-05, + "loss": 1.0628, + "step": 6238 + }, + { + "epoch": 0.37906312655689894, + "grad_norm": 0.32298099994659424, + "learning_rate": 6.882752684748556e-05, + "loss": 1.1316, + "step": 6239 + }, + { + "epoch": 0.3791238835895255, + "grad_norm": 0.21436940133571625, + "learning_rate": 6.881865832308523e-05, + "loss": 1.1191, + "step": 6240 + }, + { + "epoch": 0.379184640622152, + "grad_norm": 0.32846590876579285, + "learning_rate": 6.880978910888334e-05, + "loss": 1.1163, + "step": 6241 + }, + { + "epoch": 0.37924539765477855, + "grad_norm": 0.18213006854057312, + "learning_rate": 6.8800919205205e-05, + "loss": 1.0981, + "step": 6242 + }, + { + "epoch": 0.3793061546874051, + "grad_norm": 0.20189440250396729, + "learning_rate": 6.879204861237532e-05, + "loss": 1.0626, + "step": 6243 + }, + { + "epoch": 0.3793669117200316, + "grad_norm": 0.15365155041217804, + "learning_rate": 6.878317733071947e-05, + "loss": 1.0369, + "step": 6244 + }, + { + "epoch": 0.3794276687526581, + "grad_norm": 0.3027987480163574, + "learning_rate": 6.877430536056263e-05, + "loss": 1.0385, + "step": 6245 + }, + { + "epoch": 0.37948842578528463, + "grad_norm": 0.2907388508319855, + "learning_rate": 6.876543270222999e-05, + "loss": 1.1892, + "step": 6246 + }, + { + "epoch": 0.37954918281791117, + "grad_norm": 0.15780092775821686, + "learning_rate": 6.875655935604678e-05, + "loss": 1.0808, + "step": 6247 + }, + { + "epoch": 0.3796099398505377, + "grad_norm": 0.16876766085624695, + "learning_rate": 6.874768532233828e-05, + "loss": 1.0622, + "step": 6248 + }, + { + "epoch": 0.37967069688316424, + "grad_norm": 0.3144146800041199, + "learning_rate": 6.873881060142973e-05, + "loss": 1.2024, + "step": 6249 + }, + { + "epoch": 0.3797314539157908, + "grad_norm": 0.17376616597175598, + "learning_rate": 6.872993519364648e-05, + "loss": 1.0687, + "step": 6250 + }, + { + "epoch": 0.37979221094841725, + "grad_norm": 0.18693332374095917, + "learning_rate": 6.872105909931382e-05, + "loss": 1.075, + "step": 6251 + }, + { + "epoch": 0.3798529679810438, + "grad_norm": 0.23095564544200897, + "learning_rate": 6.871218231875713e-05, + "loss": 1.0883, + "step": 6252 + }, + { + "epoch": 0.3799137250136703, + "grad_norm": 0.2414688616991043, + "learning_rate": 6.870330485230176e-05, + "loss": 1.202, + "step": 6253 + }, + { + "epoch": 0.37997448204629686, + "grad_norm": 0.2178136557340622, + "learning_rate": 6.869442670027315e-05, + "loss": 1.0391, + "step": 6254 + }, + { + "epoch": 0.3800352390789234, + "grad_norm": 0.18792198598384857, + "learning_rate": 6.868554786299672e-05, + "loss": 1.1166, + "step": 6255 + }, + { + "epoch": 0.3800959961115499, + "grad_norm": 0.20471571385860443, + "learning_rate": 6.867666834079791e-05, + "loss": 1.039, + "step": 6256 + }, + { + "epoch": 0.38015675314417646, + "grad_norm": 0.19588330388069153, + "learning_rate": 6.86677881340022e-05, + "loss": 1.0846, + "step": 6257 + }, + { + "epoch": 0.38021751017680294, + "grad_norm": 0.16898991167545319, + "learning_rate": 6.865890724293515e-05, + "loss": 1.0516, + "step": 6258 + }, + { + "epoch": 0.3802782672094295, + "grad_norm": 0.23112809658050537, + "learning_rate": 6.865002566792223e-05, + "loss": 1.1391, + "step": 6259 + }, + { + "epoch": 0.380339024242056, + "grad_norm": 0.1884627789258957, + "learning_rate": 6.864114340928898e-05, + "loss": 1.0418, + "step": 6260 + }, + { + "epoch": 0.38039978127468255, + "grad_norm": 0.15469366312026978, + "learning_rate": 6.863226046736105e-05, + "loss": 1.0967, + "step": 6261 + }, + { + "epoch": 0.3804605383073091, + "grad_norm": 0.1870676726102829, + "learning_rate": 6.862337684246401e-05, + "loss": 1.0778, + "step": 6262 + }, + { + "epoch": 0.3805212953399356, + "grad_norm": 0.1789902299642563, + "learning_rate": 6.861449253492348e-05, + "loss": 1.1204, + "step": 6263 + }, + { + "epoch": 0.38058205237256215, + "grad_norm": 0.16738058626651764, + "learning_rate": 6.860560754506513e-05, + "loss": 1.0585, + "step": 6264 + }, + { + "epoch": 0.38064280940518863, + "grad_norm": 0.5214715600013733, + "learning_rate": 6.859672187321464e-05, + "loss": 1.1679, + "step": 6265 + }, + { + "epoch": 0.38070356643781517, + "grad_norm": 0.3082159757614136, + "learning_rate": 6.858783551969773e-05, + "loss": 1.0764, + "step": 6266 + }, + { + "epoch": 0.3807643234704417, + "grad_norm": 0.35823553800582886, + "learning_rate": 6.85789484848401e-05, + "loss": 1.1288, + "step": 6267 + }, + { + "epoch": 0.38082508050306824, + "grad_norm": 0.2691412568092346, + "learning_rate": 6.857006076896751e-05, + "loss": 1.0782, + "step": 6268 + }, + { + "epoch": 0.38088583753569477, + "grad_norm": 0.3738453984260559, + "learning_rate": 6.856117237240577e-05, + "loss": 1.2158, + "step": 6269 + }, + { + "epoch": 0.3809465945683213, + "grad_norm": 1.347652792930603, + "learning_rate": 6.855228329548067e-05, + "loss": 1.0907, + "step": 6270 + }, + { + "epoch": 0.3810073516009478, + "grad_norm": 0.5322484374046326, + "learning_rate": 6.854339353851804e-05, + "loss": 1.236, + "step": 6271 + }, + { + "epoch": 0.3810681086335743, + "grad_norm": 0.1588245928287506, + "learning_rate": 6.853450310184375e-05, + "loss": 1.0376, + "step": 6272 + }, + { + "epoch": 0.38112886566620086, + "grad_norm": 0.21055133640766144, + "learning_rate": 6.852561198578364e-05, + "loss": 1.0971, + "step": 6273 + }, + { + "epoch": 0.3811896226988274, + "grad_norm": 0.2923809885978699, + "learning_rate": 6.851672019066364e-05, + "loss": 1.24, + "step": 6274 + }, + { + "epoch": 0.3812503797314539, + "grad_norm": 0.19614262878894806, + "learning_rate": 6.85078277168097e-05, + "loss": 1.0508, + "step": 6275 + }, + { + "epoch": 0.38131113676408046, + "grad_norm": 0.15569236874580383, + "learning_rate": 6.849893456454775e-05, + "loss": 1.0739, + "step": 6276 + }, + { + "epoch": 0.381371893796707, + "grad_norm": 0.15835842490196228, + "learning_rate": 6.849004073420377e-05, + "loss": 1.0675, + "step": 6277 + }, + { + "epoch": 0.3814326508293335, + "grad_norm": 0.335661917924881, + "learning_rate": 6.848114622610377e-05, + "loss": 1.2325, + "step": 6278 + }, + { + "epoch": 0.38149340786196, + "grad_norm": 0.2634800672531128, + "learning_rate": 6.847225104057382e-05, + "loss": 1.1016, + "step": 6279 + }, + { + "epoch": 0.38155416489458654, + "grad_norm": 7.165130138397217, + "learning_rate": 6.84633551779399e-05, + "loss": 1.1966, + "step": 6280 + }, + { + "epoch": 0.3816149219272131, + "grad_norm": 0.2787732481956482, + "learning_rate": 6.845445863852814e-05, + "loss": 1.1561, + "step": 6281 + }, + { + "epoch": 0.3816756789598396, + "grad_norm": 0.20111484825611115, + "learning_rate": 6.844556142266464e-05, + "loss": 1.0986, + "step": 6282 + }, + { + "epoch": 0.38173643599246615, + "grad_norm": 1.384251594543457, + "learning_rate": 6.843666353067549e-05, + "loss": 1.0702, + "step": 6283 + }, + { + "epoch": 0.38179719302509263, + "grad_norm": 0.2577054500579834, + "learning_rate": 6.84277649628869e-05, + "loss": 1.1166, + "step": 6284 + }, + { + "epoch": 0.38185795005771916, + "grad_norm": 0.28675204515457153, + "learning_rate": 6.841886571962501e-05, + "loss": 1.098, + "step": 6285 + }, + { + "epoch": 0.3819187070903457, + "grad_norm": 0.20656707882881165, + "learning_rate": 6.840996580121606e-05, + "loss": 1.083, + "step": 6286 + }, + { + "epoch": 0.38197946412297223, + "grad_norm": 0.1893717646598816, + "learning_rate": 6.840106520798624e-05, + "loss": 1.0728, + "step": 6287 + }, + { + "epoch": 0.38204022115559877, + "grad_norm": 0.31077486276626587, + "learning_rate": 6.839216394026182e-05, + "loss": 1.0588, + "step": 6288 + }, + { + "epoch": 0.3821009781882253, + "grad_norm": 0.311691552400589, + "learning_rate": 6.838326199836906e-05, + "loss": 1.1007, + "step": 6289 + }, + { + "epoch": 0.38216173522085184, + "grad_norm": 0.20039303600788116, + "learning_rate": 6.83743593826343e-05, + "loss": 1.0774, + "step": 6290 + }, + { + "epoch": 0.3822224922534783, + "grad_norm": 0.18968652188777924, + "learning_rate": 6.836545609338383e-05, + "loss": 1.0587, + "step": 6291 + }, + { + "epoch": 0.38228324928610485, + "grad_norm": 0.4978015720844269, + "learning_rate": 6.835655213094403e-05, + "loss": 1.2125, + "step": 6292 + }, + { + "epoch": 0.3823440063187314, + "grad_norm": 0.23968303203582764, + "learning_rate": 6.834764749564125e-05, + "loss": 1.136, + "step": 6293 + }, + { + "epoch": 0.3824047633513579, + "grad_norm": 2.357583999633789, + "learning_rate": 6.833874218780192e-05, + "loss": 1.0849, + "step": 6294 + }, + { + "epoch": 0.38246552038398446, + "grad_norm": 0.25504496693611145, + "learning_rate": 6.832983620775244e-05, + "loss": 1.1216, + "step": 6295 + }, + { + "epoch": 0.382526277416611, + "grad_norm": 0.18713578581809998, + "learning_rate": 6.832092955581926e-05, + "loss": 1.0744, + "step": 6296 + }, + { + "epoch": 0.3825870344492375, + "grad_norm": 0.219355046749115, + "learning_rate": 6.831202223232889e-05, + "loss": 1.0479, + "step": 6297 + }, + { + "epoch": 0.382647791481864, + "grad_norm": 0.2766493558883667, + "learning_rate": 6.830311423760779e-05, + "loss": 1.0508, + "step": 6298 + }, + { + "epoch": 0.38270854851449054, + "grad_norm": 0.20154567062854767, + "learning_rate": 6.829420557198251e-05, + "loss": 1.043, + "step": 6299 + }, + { + "epoch": 0.3827693055471171, + "grad_norm": 0.22238002717494965, + "learning_rate": 6.828529623577957e-05, + "loss": 1.1754, + "step": 6300 + }, + { + "epoch": 0.3828300625797436, + "grad_norm": 0.2255709171295166, + "learning_rate": 6.827638622932558e-05, + "loss": 1.0781, + "step": 6301 + }, + { + "epoch": 0.38289081961237015, + "grad_norm": 0.18991853296756744, + "learning_rate": 6.826747555294711e-05, + "loss": 1.0574, + "step": 6302 + }, + { + "epoch": 0.3829515766449967, + "grad_norm": 0.33614668250083923, + "learning_rate": 6.825856420697081e-05, + "loss": 1.1505, + "step": 6303 + }, + { + "epoch": 0.38301233367762316, + "grad_norm": 0.21880415081977844, + "learning_rate": 6.824965219172328e-05, + "loss": 1.0209, + "step": 6304 + }, + { + "epoch": 0.3830730907102497, + "grad_norm": 0.18376559019088745, + "learning_rate": 6.824073950753124e-05, + "loss": 1.0742, + "step": 6305 + }, + { + "epoch": 0.38313384774287623, + "grad_norm": 0.25688204169273376, + "learning_rate": 6.823182615472137e-05, + "loss": 1.0705, + "step": 6306 + }, + { + "epoch": 0.38319460477550277, + "grad_norm": 5.014655590057373, + "learning_rate": 6.822291213362038e-05, + "loss": 1.1874, + "step": 6307 + }, + { + "epoch": 0.3832553618081293, + "grad_norm": 0.43115514516830444, + "learning_rate": 6.821399744455503e-05, + "loss": 1.1026, + "step": 6308 + }, + { + "epoch": 0.38331611884075584, + "grad_norm": 0.3493623435497284, + "learning_rate": 6.820508208785207e-05, + "loss": 1.0747, + "step": 6309 + }, + { + "epoch": 0.3833768758733823, + "grad_norm": 0.17909878492355347, + "learning_rate": 6.819616606383832e-05, + "loss": 1.0739, + "step": 6310 + }, + { + "epoch": 0.38343763290600885, + "grad_norm": 0.22200603783130646, + "learning_rate": 6.818724937284059e-05, + "loss": 1.075, + "step": 6311 + }, + { + "epoch": 0.3834983899386354, + "grad_norm": 0.23221872746944427, + "learning_rate": 6.817833201518571e-05, + "loss": 1.0788, + "step": 6312 + }, + { + "epoch": 0.3835591469712619, + "grad_norm": 0.32129982113838196, + "learning_rate": 6.816941399120056e-05, + "loss": 1.0939, + "step": 6313 + }, + { + "epoch": 0.38361990400388846, + "grad_norm": 0.28064998984336853, + "learning_rate": 6.816049530121202e-05, + "loss": 1.1366, + "step": 6314 + }, + { + "epoch": 0.383680661036515, + "grad_norm": 0.3087383508682251, + "learning_rate": 6.815157594554702e-05, + "loss": 1.0953, + "step": 6315 + }, + { + "epoch": 0.3837414180691415, + "grad_norm": 0.7785828709602356, + "learning_rate": 6.81426559245325e-05, + "loss": 1.1346, + "step": 6316 + }, + { + "epoch": 0.383802175101768, + "grad_norm": 0.2852405309677124, + "learning_rate": 6.81337352384954e-05, + "loss": 1.1989, + "step": 6317 + }, + { + "epoch": 0.38386293213439454, + "grad_norm": 0.2999477684497833, + "learning_rate": 6.812481388776275e-05, + "loss": 1.0894, + "step": 6318 + }, + { + "epoch": 0.3839236891670211, + "grad_norm": 0.21821171045303345, + "learning_rate": 6.811589187266155e-05, + "loss": 1.1647, + "step": 6319 + }, + { + "epoch": 0.3839844461996476, + "grad_norm": 0.2937355637550354, + "learning_rate": 6.81069691935188e-05, + "loss": 1.0328, + "step": 6320 + }, + { + "epoch": 0.38404520323227415, + "grad_norm": 1.4060155153274536, + "learning_rate": 6.809804585066162e-05, + "loss": 1.2378, + "step": 6321 + }, + { + "epoch": 0.3841059602649007, + "grad_norm": 0.2041044682264328, + "learning_rate": 6.808912184441707e-05, + "loss": 1.0819, + "step": 6322 + }, + { + "epoch": 0.3841667172975272, + "grad_norm": 0.2247789204120636, + "learning_rate": 6.808019717511226e-05, + "loss": 1.0663, + "step": 6323 + }, + { + "epoch": 0.3842274743301537, + "grad_norm": 0.18973910808563232, + "learning_rate": 6.807127184307433e-05, + "loss": 1.0592, + "step": 6324 + }, + { + "epoch": 0.38428823136278023, + "grad_norm": 0.18199588358402252, + "learning_rate": 6.806234584863043e-05, + "loss": 1.0554, + "step": 6325 + }, + { + "epoch": 0.38434898839540677, + "grad_norm": 0.1774917095899582, + "learning_rate": 6.805341919210775e-05, + "loss": 1.048, + "step": 6326 + }, + { + "epoch": 0.3844097454280333, + "grad_norm": 0.20697571337223053, + "learning_rate": 6.80444918738335e-05, + "loss": 1.1619, + "step": 6327 + }, + { + "epoch": 0.38447050246065984, + "grad_norm": 0.3373805582523346, + "learning_rate": 6.803556389413491e-05, + "loss": 1.0604, + "step": 6328 + }, + { + "epoch": 0.38453125949328637, + "grad_norm": 0.2412053346633911, + "learning_rate": 6.802663525333926e-05, + "loss": 1.1069, + "step": 6329 + }, + { + "epoch": 0.38459201652591285, + "grad_norm": 0.2644733190536499, + "learning_rate": 6.801770595177379e-05, + "loss": 1.0363, + "step": 6330 + }, + { + "epoch": 0.3846527735585394, + "grad_norm": 0.29324495792388916, + "learning_rate": 6.800877598976583e-05, + "loss": 1.1816, + "step": 6331 + }, + { + "epoch": 0.3847135305911659, + "grad_norm": 0.33521199226379395, + "learning_rate": 6.79998453676427e-05, + "loss": 1.1379, + "step": 6332 + }, + { + "epoch": 0.38477428762379245, + "grad_norm": 0.3780655264854431, + "learning_rate": 6.799091408573177e-05, + "loss": 1.1157, + "step": 6333 + }, + { + "epoch": 0.384835044656419, + "grad_norm": 0.20314742624759674, + "learning_rate": 6.79819821443604e-05, + "loss": 1.1082, + "step": 6334 + }, + { + "epoch": 0.3848958016890455, + "grad_norm": 0.18882369995117188, + "learning_rate": 6.7973049543856e-05, + "loss": 1.09, + "step": 6335 + }, + { + "epoch": 0.38495655872167206, + "grad_norm": 0.1587405502796173, + "learning_rate": 6.7964116284546e-05, + "loss": 1.0472, + "step": 6336 + }, + { + "epoch": 0.38501731575429854, + "grad_norm": 0.24982228875160217, + "learning_rate": 6.795518236675784e-05, + "loss": 1.2145, + "step": 6337 + }, + { + "epoch": 0.3850780727869251, + "grad_norm": 0.187753826379776, + "learning_rate": 6.7946247790819e-05, + "loss": 1.0614, + "step": 6338 + }, + { + "epoch": 0.3851388298195516, + "grad_norm": 0.18271668255329132, + "learning_rate": 6.793731255705699e-05, + "loss": 1.1018, + "step": 6339 + }, + { + "epoch": 0.38519958685217814, + "grad_norm": 0.17044109106063843, + "learning_rate": 6.79283766657993e-05, + "loss": 1.0511, + "step": 6340 + }, + { + "epoch": 0.3852603438848047, + "grad_norm": 0.911301851272583, + "learning_rate": 6.79194401173735e-05, + "loss": 1.0561, + "step": 6341 + }, + { + "epoch": 0.3853211009174312, + "grad_norm": 0.3517626225948334, + "learning_rate": 6.791050291210716e-05, + "loss": 1.0683, + "step": 6342 + }, + { + "epoch": 0.3853818579500577, + "grad_norm": 0.2430453598499298, + "learning_rate": 6.79015650503279e-05, + "loss": 1.1547, + "step": 6343 + }, + { + "epoch": 0.38544261498268423, + "grad_norm": 0.3260055482387543, + "learning_rate": 6.78926265323633e-05, + "loss": 1.0088, + "step": 6344 + }, + { + "epoch": 0.38550337201531076, + "grad_norm": 0.3974256217479706, + "learning_rate": 6.788368735854099e-05, + "loss": 1.0551, + "step": 6345 + }, + { + "epoch": 0.3855641290479373, + "grad_norm": 0.4663706123828888, + "learning_rate": 6.787474752918869e-05, + "loss": 1.0656, + "step": 6346 + }, + { + "epoch": 0.38562488608056383, + "grad_norm": 0.5852028131484985, + "learning_rate": 6.786580704463406e-05, + "loss": 1.2639, + "step": 6347 + }, + { + "epoch": 0.38568564311319037, + "grad_norm": 0.2230263352394104, + "learning_rate": 6.785686590520481e-05, + "loss": 1.1163, + "step": 6348 + }, + { + "epoch": 0.3857464001458169, + "grad_norm": 0.28689587116241455, + "learning_rate": 6.78479241112287e-05, + "loss": 1.2859, + "step": 6349 + }, + { + "epoch": 0.3858071571784434, + "grad_norm": 1.090317726135254, + "learning_rate": 6.783898166303346e-05, + "loss": 1.1629, + "step": 6350 + }, + { + "epoch": 0.3858679142110699, + "grad_norm": 0.19127705693244934, + "learning_rate": 6.783003856094692e-05, + "loss": 1.0754, + "step": 6351 + }, + { + "epoch": 0.38592867124369645, + "grad_norm": 0.2574733793735504, + "learning_rate": 6.782109480529689e-05, + "loss": 1.0928, + "step": 6352 + }, + { + "epoch": 0.385989428276323, + "grad_norm": 0.25125718116760254, + "learning_rate": 6.781215039641114e-05, + "loss": 1.0488, + "step": 6353 + }, + { + "epoch": 0.3860501853089495, + "grad_norm": 0.35906079411506653, + "learning_rate": 6.78032053346176e-05, + "loss": 1.0567, + "step": 6354 + }, + { + "epoch": 0.38611094234157606, + "grad_norm": 0.2846589982509613, + "learning_rate": 6.779425962024411e-05, + "loss": 1.244, + "step": 6355 + }, + { + "epoch": 0.38617169937420254, + "grad_norm": 0.19998668134212494, + "learning_rate": 6.77853132536186e-05, + "loss": 1.0638, + "step": 6356 + }, + { + "epoch": 0.38623245640682907, + "grad_norm": 3.2779223918914795, + "learning_rate": 6.777636623506901e-05, + "loss": 1.0695, + "step": 6357 + }, + { + "epoch": 0.3862932134394556, + "grad_norm": 0.22329023480415344, + "learning_rate": 6.776741856492326e-05, + "loss": 1.1748, + "step": 6358 + }, + { + "epoch": 0.38635397047208214, + "grad_norm": 0.825016438961029, + "learning_rate": 6.775847024350936e-05, + "loss": 1.1076, + "step": 6359 + }, + { + "epoch": 0.3864147275047087, + "grad_norm": 0.18653784692287445, + "learning_rate": 6.77495212711553e-05, + "loss": 1.1876, + "step": 6360 + }, + { + "epoch": 0.3864754845373352, + "grad_norm": 0.20956452190876007, + "learning_rate": 6.77405716481891e-05, + "loss": 1.079, + "step": 6361 + }, + { + "epoch": 0.38653624156996175, + "grad_norm": 0.30178210139274597, + "learning_rate": 6.773162137493884e-05, + "loss": 1.089, + "step": 6362 + }, + { + "epoch": 0.3865969986025882, + "grad_norm": 0.22494235634803772, + "learning_rate": 6.772267045173254e-05, + "loss": 1.0437, + "step": 6363 + }, + { + "epoch": 0.38665775563521476, + "grad_norm": 0.24685817956924438, + "learning_rate": 6.771371887889835e-05, + "loss": 1.098, + "step": 6364 + }, + { + "epoch": 0.3867185126678413, + "grad_norm": 0.2786250114440918, + "learning_rate": 6.770476665676435e-05, + "loss": 1.0999, + "step": 6365 + }, + { + "epoch": 0.38677926970046783, + "grad_norm": 0.24210099875926971, + "learning_rate": 6.769581378565873e-05, + "loss": 1.1805, + "step": 6366 + }, + { + "epoch": 0.38684002673309437, + "grad_norm": 0.2363266795873642, + "learning_rate": 6.768686026590963e-05, + "loss": 1.072, + "step": 6367 + }, + { + "epoch": 0.3869007837657209, + "grad_norm": 0.5668714642524719, + "learning_rate": 6.767790609784523e-05, + "loss": 1.1718, + "step": 6368 + }, + { + "epoch": 0.38696154079834744, + "grad_norm": 0.27067887783050537, + "learning_rate": 6.766895128179378e-05, + "loss": 1.054, + "step": 6369 + }, + { + "epoch": 0.3870222978309739, + "grad_norm": 0.3380316197872162, + "learning_rate": 6.765999581808351e-05, + "loss": 1.0646, + "step": 6370 + }, + { + "epoch": 0.38708305486360045, + "grad_norm": 1.6876212358474731, + "learning_rate": 6.765103970704266e-05, + "loss": 1.1784, + "step": 6371 + }, + { + "epoch": 0.387143811896227, + "grad_norm": 0.24330714344978333, + "learning_rate": 6.764208294899956e-05, + "loss": 1.0657, + "step": 6372 + }, + { + "epoch": 0.3872045689288535, + "grad_norm": 0.38808006048202515, + "learning_rate": 6.76331255442825e-05, + "loss": 1.0637, + "step": 6373 + }, + { + "epoch": 0.38726532596148006, + "grad_norm": 0.16064061224460602, + "learning_rate": 6.762416749321978e-05, + "loss": 1.0954, + "step": 6374 + }, + { + "epoch": 0.3873260829941066, + "grad_norm": 0.25251638889312744, + "learning_rate": 6.761520879613982e-05, + "loss": 1.1237, + "step": 6375 + }, + { + "epoch": 0.38738684002673307, + "grad_norm": 0.3030086159706116, + "learning_rate": 6.760624945337097e-05, + "loss": 1.1632, + "step": 6376 + }, + { + "epoch": 0.3874475970593596, + "grad_norm": 0.6720753908157349, + "learning_rate": 6.759728946524165e-05, + "loss": 1.094, + "step": 6377 + }, + { + "epoch": 0.38750835409198614, + "grad_norm": 0.3215318024158478, + "learning_rate": 6.758832883208029e-05, + "loss": 1.0334, + "step": 6378 + }, + { + "epoch": 0.3875691111246127, + "grad_norm": 0.26222652196884155, + "learning_rate": 6.757936755421532e-05, + "loss": 1.1069, + "step": 6379 + }, + { + "epoch": 0.3876298681572392, + "grad_norm": 0.2161160260438919, + "learning_rate": 6.757040563197525e-05, + "loss": 1.0705, + "step": 6380 + }, + { + "epoch": 0.38769062518986575, + "grad_norm": 2.251253366470337, + "learning_rate": 6.756144306568856e-05, + "loss": 1.2539, + "step": 6381 + }, + { + "epoch": 0.3877513822224923, + "grad_norm": 0.22988536953926086, + "learning_rate": 6.755247985568379e-05, + "loss": 1.087, + "step": 6382 + }, + { + "epoch": 0.38781213925511876, + "grad_norm": 0.2181500345468521, + "learning_rate": 6.754351600228945e-05, + "loss": 1.1407, + "step": 6383 + }, + { + "epoch": 0.3878728962877453, + "grad_norm": 0.3095181882381439, + "learning_rate": 6.753455150583415e-05, + "loss": 1.1364, + "step": 6384 + }, + { + "epoch": 0.38793365332037183, + "grad_norm": 0.20954351127147675, + "learning_rate": 6.752558636664648e-05, + "loss": 1.1431, + "step": 6385 + }, + { + "epoch": 0.38799441035299836, + "grad_norm": 0.20905303955078125, + "learning_rate": 6.751662058505506e-05, + "loss": 1.0529, + "step": 6386 + }, + { + "epoch": 0.3880551673856249, + "grad_norm": 0.3908860981464386, + "learning_rate": 6.750765416138851e-05, + "loss": 1.215, + "step": 6387 + }, + { + "epoch": 0.38811592441825143, + "grad_norm": 0.18732436001300812, + "learning_rate": 6.749868709597553e-05, + "loss": 1.027, + "step": 6388 + }, + { + "epoch": 0.3881766814508779, + "grad_norm": 0.2447345107793808, + "learning_rate": 6.748971938914479e-05, + "loss": 1.1253, + "step": 6389 + }, + { + "epoch": 0.38823743848350445, + "grad_norm": 0.35996878147125244, + "learning_rate": 6.7480751041225e-05, + "loss": 1.1639, + "step": 6390 + }, + { + "epoch": 0.388298195516131, + "grad_norm": 0.253505140542984, + "learning_rate": 6.74717820525449e-05, + "loss": 1.2247, + "step": 6391 + }, + { + "epoch": 0.3883589525487575, + "grad_norm": 0.7387507557868958, + "learning_rate": 6.746281242343325e-05, + "loss": 1.1715, + "step": 6392 + }, + { + "epoch": 0.38841970958138405, + "grad_norm": 0.31757861375808716, + "learning_rate": 6.745384215421884e-05, + "loss": 1.0493, + "step": 6393 + }, + { + "epoch": 0.3884804666140106, + "grad_norm": 0.16202113032341003, + "learning_rate": 6.744487124523047e-05, + "loss": 1.0779, + "step": 6394 + }, + { + "epoch": 0.3885412236466371, + "grad_norm": 0.30141791701316833, + "learning_rate": 6.743589969679697e-05, + "loss": 1.1526, + "step": 6395 + }, + { + "epoch": 0.3886019806792636, + "grad_norm": 0.212767094373703, + "learning_rate": 6.742692750924722e-05, + "loss": 1.0413, + "step": 6396 + }, + { + "epoch": 0.38866273771189014, + "grad_norm": 0.31104981899261475, + "learning_rate": 6.741795468291003e-05, + "loss": 1.1268, + "step": 6397 + }, + { + "epoch": 0.3887234947445167, + "grad_norm": 0.2536655068397522, + "learning_rate": 6.740898121811437e-05, + "loss": 1.0539, + "step": 6398 + }, + { + "epoch": 0.3887842517771432, + "grad_norm": 0.18488703668117523, + "learning_rate": 6.740000711518915e-05, + "loss": 1.0937, + "step": 6399 + }, + { + "epoch": 0.38884500880976974, + "grad_norm": 0.24771010875701904, + "learning_rate": 6.739103237446328e-05, + "loss": 1.1574, + "step": 6400 + }, + { + "epoch": 0.3889057658423963, + "grad_norm": 0.21560527384281158, + "learning_rate": 6.738205699626578e-05, + "loss": 1.0227, + "step": 6401 + }, + { + "epoch": 0.38896652287502276, + "grad_norm": 0.2638882100582123, + "learning_rate": 6.737308098092561e-05, + "loss": 1.1582, + "step": 6402 + }, + { + "epoch": 0.3890272799076493, + "grad_norm": 1.0715861320495605, + "learning_rate": 6.736410432877179e-05, + "loss": 1.1655, + "step": 6403 + }, + { + "epoch": 0.3890880369402758, + "grad_norm": 0.2793117165565491, + "learning_rate": 6.735512704013339e-05, + "loss": 1.1893, + "step": 6404 + }, + { + "epoch": 0.38914879397290236, + "grad_norm": 0.20608031749725342, + "learning_rate": 6.734614911533944e-05, + "loss": 1.1738, + "step": 6405 + }, + { + "epoch": 0.3892095510055289, + "grad_norm": 0.16853372752666473, + "learning_rate": 6.733717055471904e-05, + "loss": 1.0497, + "step": 6406 + }, + { + "epoch": 0.38927030803815543, + "grad_norm": 0.22243182361125946, + "learning_rate": 6.732819135860129e-05, + "loss": 1.1619, + "step": 6407 + }, + { + "epoch": 0.38933106507078197, + "grad_norm": 0.20129846036434174, + "learning_rate": 6.731921152731535e-05, + "loss": 1.0623, + "step": 6408 + }, + { + "epoch": 0.38939182210340845, + "grad_norm": 0.4767175614833832, + "learning_rate": 6.731023106119038e-05, + "loss": 1.1563, + "step": 6409 + }, + { + "epoch": 0.389452579136035, + "grad_norm": 0.19893300533294678, + "learning_rate": 6.730124996055552e-05, + "loss": 1.1967, + "step": 6410 + }, + { + "epoch": 0.3895133361686615, + "grad_norm": 0.28141626715660095, + "learning_rate": 6.729226822574e-05, + "loss": 1.1166, + "step": 6411 + }, + { + "epoch": 0.38957409320128805, + "grad_norm": 0.30152684450149536, + "learning_rate": 6.728328585707306e-05, + "loss": 1.0942, + "step": 6412 + }, + { + "epoch": 0.3896348502339146, + "grad_norm": 0.21385829150676727, + "learning_rate": 6.727430285488393e-05, + "loss": 1.0892, + "step": 6413 + }, + { + "epoch": 0.3896956072665411, + "grad_norm": 0.2695333957672119, + "learning_rate": 6.726531921950189e-05, + "loss": 1.1952, + "step": 6414 + }, + { + "epoch": 0.3897563642991676, + "grad_norm": 0.15710245072841644, + "learning_rate": 6.725633495125623e-05, + "loss": 1.0401, + "step": 6415 + }, + { + "epoch": 0.38981712133179414, + "grad_norm": 0.2224552184343338, + "learning_rate": 6.724735005047629e-05, + "loss": 1.0852, + "step": 6416 + }, + { + "epoch": 0.38987787836442067, + "grad_norm": 0.32308387756347656, + "learning_rate": 6.72383645174914e-05, + "loss": 1.3042, + "step": 6417 + }, + { + "epoch": 0.3899386353970472, + "grad_norm": 5.668177127838135, + "learning_rate": 6.72293783526309e-05, + "loss": 1.0747, + "step": 6418 + }, + { + "epoch": 0.38999939242967374, + "grad_norm": 0.459460586309433, + "learning_rate": 6.722039155622424e-05, + "loss": 1.1121, + "step": 6419 + }, + { + "epoch": 0.3900601494623003, + "grad_norm": 1.109707236289978, + "learning_rate": 6.721140412860077e-05, + "loss": 1.0948, + "step": 6420 + }, + { + "epoch": 0.3901209064949268, + "grad_norm": 0.25619563460350037, + "learning_rate": 6.720241607008997e-05, + "loss": 1.0562, + "step": 6421 + }, + { + "epoch": 0.3901816635275533, + "grad_norm": 0.21115005016326904, + "learning_rate": 6.719342738102128e-05, + "loss": 1.0394, + "step": 6422 + }, + { + "epoch": 0.3902424205601798, + "grad_norm": 1.3127421140670776, + "learning_rate": 6.71844380617242e-05, + "loss": 1.1071, + "step": 6423 + }, + { + "epoch": 0.39030317759280636, + "grad_norm": 0.31879696249961853, + "learning_rate": 6.717544811252821e-05, + "loss": 1.1726, + "step": 6424 + }, + { + "epoch": 0.3903639346254329, + "grad_norm": 0.15854428708553314, + "learning_rate": 6.716645753376284e-05, + "loss": 1.0293, + "step": 6425 + }, + { + "epoch": 0.39042469165805943, + "grad_norm": 0.18014802038669586, + "learning_rate": 6.715746632575767e-05, + "loss": 1.041, + "step": 6426 + }, + { + "epoch": 0.39048544869068597, + "grad_norm": 0.21742500364780426, + "learning_rate": 6.714847448884225e-05, + "loss": 1.0376, + "step": 6427 + }, + { + "epoch": 0.3905462057233125, + "grad_norm": 0.8158523440361023, + "learning_rate": 6.713948202334617e-05, + "loss": 1.1267, + "step": 6428 + }, + { + "epoch": 0.390606962755939, + "grad_norm": 0.24622945487499237, + "learning_rate": 6.713048892959908e-05, + "loss": 1.1186, + "step": 6429 + }, + { + "epoch": 0.3906677197885655, + "grad_norm": 0.30093732476234436, + "learning_rate": 6.71214952079306e-05, + "loss": 1.1419, + "step": 6430 + }, + { + "epoch": 0.39072847682119205, + "grad_norm": 0.29259830713272095, + "learning_rate": 6.711250085867039e-05, + "loss": 1.1732, + "step": 6431 + }, + { + "epoch": 0.3907892338538186, + "grad_norm": 0.26576000452041626, + "learning_rate": 6.710350588214817e-05, + "loss": 1.1522, + "step": 6432 + }, + { + "epoch": 0.3908499908864451, + "grad_norm": 0.19261577725410461, + "learning_rate": 6.709451027869364e-05, + "loss": 1.0781, + "step": 6433 + }, + { + "epoch": 0.39091074791907165, + "grad_norm": 0.18543879687786102, + "learning_rate": 6.708551404863653e-05, + "loss": 1.0383, + "step": 6434 + }, + { + "epoch": 0.39097150495169813, + "grad_norm": 0.28963205218315125, + "learning_rate": 6.707651719230658e-05, + "loss": 1.0656, + "step": 6435 + }, + { + "epoch": 0.39103226198432467, + "grad_norm": 0.1532381772994995, + "learning_rate": 6.70675197100336e-05, + "loss": 1.0311, + "step": 6436 + }, + { + "epoch": 0.3910930190169512, + "grad_norm": 0.22672246396541595, + "learning_rate": 6.70585216021474e-05, + "loss": 1.1446, + "step": 6437 + }, + { + "epoch": 0.39115377604957774, + "grad_norm": 3.197603225708008, + "learning_rate": 6.704952286897778e-05, + "loss": 1.0372, + "step": 6438 + }, + { + "epoch": 0.3912145330822043, + "grad_norm": 0.27045390009880066, + "learning_rate": 6.704052351085462e-05, + "loss": 1.0854, + "step": 6439 + }, + { + "epoch": 0.3912752901148308, + "grad_norm": 0.20071037113666534, + "learning_rate": 6.703152352810775e-05, + "loss": 1.05, + "step": 6440 + }, + { + "epoch": 0.39133604714745734, + "grad_norm": 0.22197884321212769, + "learning_rate": 6.70225229210671e-05, + "loss": 1.1629, + "step": 6441 + }, + { + "epoch": 0.3913968041800838, + "grad_norm": 0.2129591703414917, + "learning_rate": 6.70135216900626e-05, + "loss": 1.0163, + "step": 6442 + }, + { + "epoch": 0.39145756121271036, + "grad_norm": 0.2719408869743347, + "learning_rate": 6.700451983542415e-05, + "loss": 1.1348, + "step": 6443 + }, + { + "epoch": 0.3915183182453369, + "grad_norm": 0.20657086372375488, + "learning_rate": 6.699551735748176e-05, + "loss": 1.077, + "step": 6444 + }, + { + "epoch": 0.39157907527796343, + "grad_norm": 0.27585330605506897, + "learning_rate": 6.698651425656539e-05, + "loss": 1.2298, + "step": 6445 + }, + { + "epoch": 0.39163983231058996, + "grad_norm": 0.20878803730010986, + "learning_rate": 6.697751053300505e-05, + "loss": 1.1432, + "step": 6446 + }, + { + "epoch": 0.3917005893432165, + "grad_norm": 0.2670849561691284, + "learning_rate": 6.696850618713079e-05, + "loss": 1.1927, + "step": 6447 + }, + { + "epoch": 0.391761346375843, + "grad_norm": 0.16683627665042877, + "learning_rate": 6.695950121927265e-05, + "loss": 1.0629, + "step": 6448 + }, + { + "epoch": 0.3918221034084695, + "grad_norm": 1.2232494354248047, + "learning_rate": 6.695049562976072e-05, + "loss": 1.3266, + "step": 6449 + }, + { + "epoch": 0.39188286044109605, + "grad_norm": 0.31205451488494873, + "learning_rate": 6.694148941892509e-05, + "loss": 1.0757, + "step": 6450 + }, + { + "epoch": 0.3919436174737226, + "grad_norm": 0.4389975070953369, + "learning_rate": 6.693248258709588e-05, + "loss": 1.1915, + "step": 6451 + }, + { + "epoch": 0.3920043745063491, + "grad_norm": 0.21663759648799896, + "learning_rate": 6.692347513460326e-05, + "loss": 1.0152, + "step": 6452 + }, + { + "epoch": 0.39206513153897565, + "grad_norm": 0.16130897402763367, + "learning_rate": 6.69144670617774e-05, + "loss": 1.0697, + "step": 6453 + }, + { + "epoch": 0.3921258885716022, + "grad_norm": 0.2647676169872284, + "learning_rate": 6.690545836894845e-05, + "loss": 1.2673, + "step": 6454 + }, + { + "epoch": 0.39218664560422867, + "grad_norm": 0.20604512095451355, + "learning_rate": 6.689644905644668e-05, + "loss": 1.0363, + "step": 6455 + }, + { + "epoch": 0.3922474026368552, + "grad_norm": 0.26505059003829956, + "learning_rate": 6.688743912460229e-05, + "loss": 1.0859, + "step": 6456 + }, + { + "epoch": 0.39230815966948174, + "grad_norm": 0.23641453683376312, + "learning_rate": 6.687842857374556e-05, + "loss": 1.1877, + "step": 6457 + }, + { + "epoch": 0.3923689167021083, + "grad_norm": 0.25958356261253357, + "learning_rate": 6.686941740420678e-05, + "loss": 1.1923, + "step": 6458 + }, + { + "epoch": 0.3924296737347348, + "grad_norm": 0.3640066385269165, + "learning_rate": 6.686040561631623e-05, + "loss": 1.1205, + "step": 6459 + }, + { + "epoch": 0.39249043076736134, + "grad_norm": 0.21976006031036377, + "learning_rate": 6.685139321040426e-05, + "loss": 1.1277, + "step": 6460 + }, + { + "epoch": 0.3925511877999878, + "grad_norm": 0.27792975306510925, + "learning_rate": 6.684238018680121e-05, + "loss": 1.0938, + "step": 6461 + }, + { + "epoch": 0.39261194483261436, + "grad_norm": 2.917229652404785, + "learning_rate": 6.683336654583745e-05, + "loss": 1.0475, + "step": 6462 + }, + { + "epoch": 0.3926727018652409, + "grad_norm": 0.2829456627368927, + "learning_rate": 6.682435228784341e-05, + "loss": 1.0522, + "step": 6463 + }, + { + "epoch": 0.3927334588978674, + "grad_norm": 0.2245984971523285, + "learning_rate": 6.681533741314945e-05, + "loss": 1.1496, + "step": 6464 + }, + { + "epoch": 0.39279421593049396, + "grad_norm": 0.34320905804634094, + "learning_rate": 6.680632192208609e-05, + "loss": 1.1192, + "step": 6465 + }, + { + "epoch": 0.3928549729631205, + "grad_norm": 0.360634982585907, + "learning_rate": 6.679730581498372e-05, + "loss": 1.1658, + "step": 6466 + }, + { + "epoch": 0.39291572999574703, + "grad_norm": 0.26463451981544495, + "learning_rate": 6.678828909217286e-05, + "loss": 1.1123, + "step": 6467 + }, + { + "epoch": 0.3929764870283735, + "grad_norm": 0.38197797536849976, + "learning_rate": 6.677927175398403e-05, + "loss": 1.0132, + "step": 6468 + }, + { + "epoch": 0.39303724406100005, + "grad_norm": 0.15202993154525757, + "learning_rate": 6.677025380074775e-05, + "loss": 1.1559, + "step": 6469 + }, + { + "epoch": 0.3930980010936266, + "grad_norm": 0.3712639808654785, + "learning_rate": 6.676123523279457e-05, + "loss": 1.0779, + "step": 6470 + }, + { + "epoch": 0.3931587581262531, + "grad_norm": 0.2633703947067261, + "learning_rate": 6.675221605045507e-05, + "loss": 1.2234, + "step": 6471 + }, + { + "epoch": 0.39321951515887965, + "grad_norm": 0.5182471871376038, + "learning_rate": 6.674319625405985e-05, + "loss": 1.0731, + "step": 6472 + }, + { + "epoch": 0.3932802721915062, + "grad_norm": 0.8649107217788696, + "learning_rate": 6.673417584393955e-05, + "loss": 1.1544, + "step": 6473 + }, + { + "epoch": 0.3933410292241327, + "grad_norm": 0.2366161197423935, + "learning_rate": 6.67251548204248e-05, + "loss": 1.0912, + "step": 6474 + }, + { + "epoch": 0.3934017862567592, + "grad_norm": 0.29931312799453735, + "learning_rate": 6.671613318384624e-05, + "loss": 1.1146, + "step": 6475 + }, + { + "epoch": 0.39346254328938574, + "grad_norm": 0.379379540681839, + "learning_rate": 6.670711093453461e-05, + "loss": 1.0732, + "step": 6476 + }, + { + "epoch": 0.39352330032201227, + "grad_norm": 0.20447179675102234, + "learning_rate": 6.669808807282058e-05, + "loss": 1.0564, + "step": 6477 + }, + { + "epoch": 0.3935840573546388, + "grad_norm": 0.3506558835506439, + "learning_rate": 6.668906459903493e-05, + "loss": 1.0575, + "step": 6478 + }, + { + "epoch": 0.39364481438726534, + "grad_norm": 0.33056285977363586, + "learning_rate": 6.668004051350839e-05, + "loss": 1.2248, + "step": 6479 + }, + { + "epoch": 0.3937055714198919, + "grad_norm": 0.27900615334510803, + "learning_rate": 6.667101581657172e-05, + "loss": 1.1771, + "step": 6480 + }, + { + "epoch": 0.39376632845251835, + "grad_norm": 0.3498802185058594, + "learning_rate": 6.666199050855575e-05, + "loss": 1.0537, + "step": 6481 + }, + { + "epoch": 0.3938270854851449, + "grad_norm": 0.24188032746315002, + "learning_rate": 6.665296458979131e-05, + "loss": 1.092, + "step": 6482 + }, + { + "epoch": 0.3938878425177714, + "grad_norm": 0.1928771734237671, + "learning_rate": 6.664393806060922e-05, + "loss": 1.0796, + "step": 6483 + }, + { + "epoch": 0.39394859955039796, + "grad_norm": 0.33135178685188293, + "learning_rate": 6.663491092134038e-05, + "loss": 1.0545, + "step": 6484 + }, + { + "epoch": 0.3940093565830245, + "grad_norm": 0.18423447012901306, + "learning_rate": 6.662588317231564e-05, + "loss": 1.1537, + "step": 6485 + }, + { + "epoch": 0.39407011361565103, + "grad_norm": 0.3939998149871826, + "learning_rate": 6.661685481386596e-05, + "loss": 1.1828, + "step": 6486 + }, + { + "epoch": 0.39413087064827756, + "grad_norm": 0.2723671793937683, + "learning_rate": 6.660782584632223e-05, + "loss": 1.1729, + "step": 6487 + }, + { + "epoch": 0.39419162768090404, + "grad_norm": 0.27051278948783875, + "learning_rate": 6.659879627001545e-05, + "loss": 1.1829, + "step": 6488 + }, + { + "epoch": 0.3942523847135306, + "grad_norm": 0.3542154133319855, + "learning_rate": 6.658976608527658e-05, + "loss": 1.0245, + "step": 6489 + }, + { + "epoch": 0.3943131417461571, + "grad_norm": 0.23470456898212433, + "learning_rate": 6.658073529243664e-05, + "loss": 1.1564, + "step": 6490 + }, + { + "epoch": 0.39437389877878365, + "grad_norm": 0.19932225346565247, + "learning_rate": 6.657170389182663e-05, + "loss": 1.0703, + "step": 6491 + }, + { + "epoch": 0.3944346558114102, + "grad_norm": 0.5831389427185059, + "learning_rate": 6.656267188377762e-05, + "loss": 1.0588, + "step": 6492 + }, + { + "epoch": 0.3944954128440367, + "grad_norm": 0.3360936939716339, + "learning_rate": 6.655363926862066e-05, + "loss": 1.1181, + "step": 6493 + }, + { + "epoch": 0.3945561698766632, + "grad_norm": 0.2753875255584717, + "learning_rate": 6.654460604668687e-05, + "loss": 1.1565, + "step": 6494 + }, + { + "epoch": 0.39461692690928973, + "grad_norm": 0.48314201831817627, + "learning_rate": 6.653557221830732e-05, + "loss": 1.2348, + "step": 6495 + }, + { + "epoch": 0.39467768394191627, + "grad_norm": 0.2670505940914154, + "learning_rate": 6.652653778381319e-05, + "loss": 1.0483, + "step": 6496 + }, + { + "epoch": 0.3947384409745428, + "grad_norm": 0.20062226057052612, + "learning_rate": 6.651750274353563e-05, + "loss": 1.0757, + "step": 6497 + }, + { + "epoch": 0.39479919800716934, + "grad_norm": 1.248395562171936, + "learning_rate": 6.650846709780581e-05, + "loss": 1.0683, + "step": 6498 + }, + { + "epoch": 0.3948599550397959, + "grad_norm": 0.5881316661834717, + "learning_rate": 6.649943084695493e-05, + "loss": 1.0675, + "step": 6499 + }, + { + "epoch": 0.3949207120724224, + "grad_norm": 1.3627214431762695, + "learning_rate": 6.649039399131423e-05, + "loss": 1.056, + "step": 6500 + }, + { + "epoch": 0.3949814691050489, + "grad_norm": 0.31135034561157227, + "learning_rate": 6.648135653121494e-05, + "loss": 1.1009, + "step": 6501 + }, + { + "epoch": 0.3950422261376754, + "grad_norm": 0.17680184543132782, + "learning_rate": 6.647231846698835e-05, + "loss": 1.1004, + "step": 6502 + }, + { + "epoch": 0.39510298317030196, + "grad_norm": 0.1913098394870758, + "learning_rate": 6.646327979896574e-05, + "loss": 1.0798, + "step": 6503 + }, + { + "epoch": 0.3951637402029285, + "grad_norm": 0.18262144923210144, + "learning_rate": 6.645424052747843e-05, + "loss": 1.0526, + "step": 6504 + }, + { + "epoch": 0.39522449723555503, + "grad_norm": 0.19773466885089874, + "learning_rate": 6.644520065285775e-05, + "loss": 1.0397, + "step": 6505 + }, + { + "epoch": 0.39528525426818156, + "grad_norm": 0.14859731495380402, + "learning_rate": 6.643616017543506e-05, + "loss": 1.0367, + "step": 6506 + }, + { + "epoch": 0.39534601130080804, + "grad_norm": 0.1895281821489334, + "learning_rate": 6.642711909554174e-05, + "loss": 1.0577, + "step": 6507 + }, + { + "epoch": 0.3954067683334346, + "grad_norm": 0.25603267550468445, + "learning_rate": 6.64180774135092e-05, + "loss": 1.0525, + "step": 6508 + }, + { + "epoch": 0.3954675253660611, + "grad_norm": 0.28024548292160034, + "learning_rate": 6.640903512966885e-05, + "loss": 1.0779, + "step": 6509 + }, + { + "epoch": 0.39552828239868765, + "grad_norm": 0.2207445353269577, + "learning_rate": 6.639999224435217e-05, + "loss": 1.1105, + "step": 6510 + }, + { + "epoch": 0.3955890394313142, + "grad_norm": 0.23242032527923584, + "learning_rate": 6.639094875789058e-05, + "loss": 1.1124, + "step": 6511 + }, + { + "epoch": 0.3956497964639407, + "grad_norm": 0.2855518162250519, + "learning_rate": 6.638190467061559e-05, + "loss": 1.2743, + "step": 6512 + }, + { + "epoch": 0.39571055349656725, + "grad_norm": 0.5975963473320007, + "learning_rate": 6.637285998285874e-05, + "loss": 1.0776, + "step": 6513 + }, + { + "epoch": 0.39577131052919373, + "grad_norm": 0.21736150979995728, + "learning_rate": 6.636381469495152e-05, + "loss": 1.0939, + "step": 6514 + }, + { + "epoch": 0.39583206756182027, + "grad_norm": 2.118527889251709, + "learning_rate": 6.635476880722553e-05, + "loss": 1.0229, + "step": 6515 + }, + { + "epoch": 0.3958928245944468, + "grad_norm": 0.2877878248691559, + "learning_rate": 6.634572232001232e-05, + "loss": 1.0775, + "step": 6516 + }, + { + "epoch": 0.39595358162707334, + "grad_norm": 0.22778743505477905, + "learning_rate": 6.633667523364349e-05, + "loss": 1.1323, + "step": 6517 + }, + { + "epoch": 0.39601433865969987, + "grad_norm": 0.30745941400527954, + "learning_rate": 6.63276275484507e-05, + "loss": 1.0843, + "step": 6518 + }, + { + "epoch": 0.3960750956923264, + "grad_norm": 1.860317587852478, + "learning_rate": 6.631857926476552e-05, + "loss": 1.0941, + "step": 6519 + }, + { + "epoch": 0.3961358527249529, + "grad_norm": 0.23120824992656708, + "learning_rate": 6.63095303829197e-05, + "loss": 1.1606, + "step": 6520 + }, + { + "epoch": 0.3961966097575794, + "grad_norm": 3.6885416507720947, + "learning_rate": 6.630048090324488e-05, + "loss": 1.1184, + "step": 6521 + }, + { + "epoch": 0.39625736679020596, + "grad_norm": 0.4488631784915924, + "learning_rate": 6.629143082607278e-05, + "loss": 1.0594, + "step": 6522 + }, + { + "epoch": 0.3963181238228325, + "grad_norm": 0.1940697878599167, + "learning_rate": 6.628238015173514e-05, + "loss": 1.1172, + "step": 6523 + }, + { + "epoch": 0.396378880855459, + "grad_norm": 0.2685048282146454, + "learning_rate": 6.627332888056369e-05, + "loss": 1.0966, + "step": 6524 + }, + { + "epoch": 0.39643963788808556, + "grad_norm": 0.3317084312438965, + "learning_rate": 6.626427701289025e-05, + "loss": 1.1496, + "step": 6525 + }, + { + "epoch": 0.3965003949207121, + "grad_norm": 0.18556047976016998, + "learning_rate": 6.625522454904657e-05, + "loss": 1.0891, + "step": 6526 + }, + { + "epoch": 0.3965611519533386, + "grad_norm": 0.35271117091178894, + "learning_rate": 6.62461714893645e-05, + "loss": 1.0956, + "step": 6527 + }, + { + "epoch": 0.3966219089859651, + "grad_norm": 0.3039400279521942, + "learning_rate": 6.623711783417587e-05, + "loss": 1.0479, + "step": 6528 + }, + { + "epoch": 0.39668266601859165, + "grad_norm": 0.24666444957256317, + "learning_rate": 6.622806358381255e-05, + "loss": 1.1442, + "step": 6529 + }, + { + "epoch": 0.3967434230512182, + "grad_norm": 0.6116878986358643, + "learning_rate": 6.621900873860645e-05, + "loss": 1.1024, + "step": 6530 + }, + { + "epoch": 0.3968041800838447, + "grad_norm": 0.33929240703582764, + "learning_rate": 6.620995329888942e-05, + "loss": 1.0951, + "step": 6531 + }, + { + "epoch": 0.39686493711647125, + "grad_norm": 0.20404334366321564, + "learning_rate": 6.620089726499341e-05, + "loss": 1.1032, + "step": 6532 + }, + { + "epoch": 0.3969256941490978, + "grad_norm": 0.504788339138031, + "learning_rate": 6.61918406372504e-05, + "loss": 1.1831, + "step": 6533 + }, + { + "epoch": 0.39698645118172426, + "grad_norm": 0.4639085531234741, + "learning_rate": 6.618278341599233e-05, + "loss": 1.0345, + "step": 6534 + }, + { + "epoch": 0.3970472082143508, + "grad_norm": 0.5196904540061951, + "learning_rate": 6.61737256015512e-05, + "loss": 1.057, + "step": 6535 + }, + { + "epoch": 0.39710796524697733, + "grad_norm": 6.176527500152588, + "learning_rate": 6.616466719425905e-05, + "loss": 1.0903, + "step": 6536 + }, + { + "epoch": 0.39716872227960387, + "grad_norm": 0.2886123061180115, + "learning_rate": 6.61556081944479e-05, + "loss": 1.1359, + "step": 6537 + }, + { + "epoch": 0.3972294793122304, + "grad_norm": 0.2945880591869354, + "learning_rate": 6.61465486024498e-05, + "loss": 1.0824, + "step": 6538 + }, + { + "epoch": 0.39729023634485694, + "grad_norm": 0.2940830886363983, + "learning_rate": 6.613748841859683e-05, + "loss": 1.1936, + "step": 6539 + }, + { + "epoch": 0.3973509933774834, + "grad_norm": 0.1936923861503601, + "learning_rate": 6.612842764322113e-05, + "loss": 1.0268, + "step": 6540 + }, + { + "epoch": 0.39741175041010995, + "grad_norm": 0.25084957480430603, + "learning_rate": 6.611936627665477e-05, + "loss": 1.1362, + "step": 6541 + }, + { + "epoch": 0.3974725074427365, + "grad_norm": 0.15377183258533478, + "learning_rate": 6.611030431922993e-05, + "loss": 1.0691, + "step": 6542 + }, + { + "epoch": 0.397533264475363, + "grad_norm": 0.2854984402656555, + "learning_rate": 6.610124177127879e-05, + "loss": 1.0029, + "step": 6543 + }, + { + "epoch": 0.39759402150798956, + "grad_norm": 0.19456151127815247, + "learning_rate": 6.609217863313349e-05, + "loss": 1.057, + "step": 6544 + }, + { + "epoch": 0.3976547785406161, + "grad_norm": 0.2655160129070282, + "learning_rate": 6.608311490512629e-05, + "loss": 1.0365, + "step": 6545 + }, + { + "epoch": 0.39771553557324263, + "grad_norm": 0.23341482877731323, + "learning_rate": 6.60740505875894e-05, + "loss": 1.2575, + "step": 6546 + }, + { + "epoch": 0.3977762926058691, + "grad_norm": 0.38992664217948914, + "learning_rate": 6.606498568085508e-05, + "loss": 1.0455, + "step": 6547 + }, + { + "epoch": 0.39783704963849564, + "grad_norm": 0.1601528823375702, + "learning_rate": 6.60559201852556e-05, + "loss": 1.092, + "step": 6548 + }, + { + "epoch": 0.3978978066711222, + "grad_norm": 0.28493934869766235, + "learning_rate": 6.604685410112327e-05, + "loss": 1.1137, + "step": 6549 + }, + { + "epoch": 0.3979585637037487, + "grad_norm": 0.20269769430160522, + "learning_rate": 6.60377874287904e-05, + "loss": 1.0965, + "step": 6550 + }, + { + "epoch": 0.39801932073637525, + "grad_norm": 0.26552116870880127, + "learning_rate": 6.602872016858933e-05, + "loss": 1.0566, + "step": 6551 + }, + { + "epoch": 0.3980800777690018, + "grad_norm": 0.22329623997211456, + "learning_rate": 6.601965232085241e-05, + "loss": 1.082, + "step": 6552 + }, + { + "epoch": 0.39814083480162826, + "grad_norm": 0.36417409777641296, + "learning_rate": 6.601058388591206e-05, + "loss": 1.1598, + "step": 6553 + }, + { + "epoch": 0.3982015918342548, + "grad_norm": 0.15952885150909424, + "learning_rate": 6.600151486410066e-05, + "loss": 0.9954, + "step": 6554 + }, + { + "epoch": 0.39826234886688133, + "grad_norm": 0.1855550855398178, + "learning_rate": 6.599244525575062e-05, + "loss": 1.0657, + "step": 6555 + }, + { + "epoch": 0.39832310589950787, + "grad_norm": 0.3037247955799103, + "learning_rate": 6.598337506119443e-05, + "loss": 1.1075, + "step": 6556 + }, + { + "epoch": 0.3983838629321344, + "grad_norm": 0.3026576042175293, + "learning_rate": 6.597430428076451e-05, + "loss": 1.144, + "step": 6557 + }, + { + "epoch": 0.39844461996476094, + "grad_norm": 8.997053146362305, + "learning_rate": 6.596523291479342e-05, + "loss": 1.1201, + "step": 6558 + }, + { + "epoch": 0.3985053769973875, + "grad_norm": 0.7445966601371765, + "learning_rate": 6.59561609636136e-05, + "loss": 1.2124, + "step": 6559 + }, + { + "epoch": 0.39856613403001395, + "grad_norm": 0.21606764197349548, + "learning_rate": 6.594708842755762e-05, + "loss": 1.1255, + "step": 6560 + }, + { + "epoch": 0.3986268910626405, + "grad_norm": 0.44044992327690125, + "learning_rate": 6.593801530695804e-05, + "loss": 1.1415, + "step": 6561 + }, + { + "epoch": 0.398687648095267, + "grad_norm": 2.0879759788513184, + "learning_rate": 6.592894160214743e-05, + "loss": 1.1465, + "step": 6562 + }, + { + "epoch": 0.39874840512789356, + "grad_norm": 0.30447283387184143, + "learning_rate": 6.591986731345839e-05, + "loss": 1.1925, + "step": 6563 + }, + { + "epoch": 0.3988091621605201, + "grad_norm": 0.3468186855316162, + "learning_rate": 6.591079244122352e-05, + "loss": 1.0048, + "step": 6564 + }, + { + "epoch": 0.3988699191931466, + "grad_norm": 0.28498584032058716, + "learning_rate": 6.590171698577547e-05, + "loss": 1.1161, + "step": 6565 + }, + { + "epoch": 0.3989306762257731, + "grad_norm": 0.3815024197101593, + "learning_rate": 6.589264094744693e-05, + "loss": 1.0953, + "step": 6566 + }, + { + "epoch": 0.39899143325839964, + "grad_norm": 0.2131989300251007, + "learning_rate": 6.588356432657057e-05, + "loss": 1.128, + "step": 6567 + }, + { + "epoch": 0.3990521902910262, + "grad_norm": 0.4072997570037842, + "learning_rate": 6.587448712347907e-05, + "loss": 1.0871, + "step": 6568 + }, + { + "epoch": 0.3991129473236527, + "grad_norm": 0.15983837842941284, + "learning_rate": 6.586540933850518e-05, + "loss": 1.1068, + "step": 6569 + }, + { + "epoch": 0.39917370435627925, + "grad_norm": 0.20638173818588257, + "learning_rate": 6.585633097198165e-05, + "loss": 1.076, + "step": 6570 + }, + { + "epoch": 0.3992344613889058, + "grad_norm": 0.21482278406620026, + "learning_rate": 6.584725202424125e-05, + "loss": 1.1558, + "step": 6571 + }, + { + "epoch": 0.3992952184215323, + "grad_norm": 0.34638452529907227, + "learning_rate": 6.583817249561675e-05, + "loss": 1.1315, + "step": 6572 + }, + { + "epoch": 0.3993559754541588, + "grad_norm": 0.21774446964263916, + "learning_rate": 6.582909238644098e-05, + "loss": 1.0779, + "step": 6573 + }, + { + "epoch": 0.39941673248678533, + "grad_norm": 0.2853248715400696, + "learning_rate": 6.582001169704676e-05, + "loss": 1.0575, + "step": 6574 + }, + { + "epoch": 0.39947748951941187, + "grad_norm": 0.2660810947418213, + "learning_rate": 6.581093042776699e-05, + "loss": 1.1072, + "step": 6575 + }, + { + "epoch": 0.3995382465520384, + "grad_norm": 0.2002153843641281, + "learning_rate": 6.580184857893446e-05, + "loss": 1.0451, + "step": 6576 + }, + { + "epoch": 0.39959900358466494, + "grad_norm": 0.17404139041900635, + "learning_rate": 6.579276615088214e-05, + "loss": 1.0471, + "step": 6577 + }, + { + "epoch": 0.39965976061729147, + "grad_norm": 4.905503749847412, + "learning_rate": 6.578368314394292e-05, + "loss": 1.085, + "step": 6578 + }, + { + "epoch": 0.399720517649918, + "grad_norm": 0.40136656165122986, + "learning_rate": 6.577459955844976e-05, + "loss": 1.1278, + "step": 6579 + }, + { + "epoch": 0.3997812746825445, + "grad_norm": 0.2077597677707672, + "learning_rate": 6.576551539473559e-05, + "loss": 1.083, + "step": 6580 + }, + { + "epoch": 0.399842031715171, + "grad_norm": 0.3461739122867584, + "learning_rate": 6.575643065313339e-05, + "loss": 1.0772, + "step": 6581 + }, + { + "epoch": 0.39990278874779756, + "grad_norm": 0.22769854962825775, + "learning_rate": 6.574734533397621e-05, + "loss": 1.0949, + "step": 6582 + }, + { + "epoch": 0.3999635457804241, + "grad_norm": 0.21439218521118164, + "learning_rate": 6.573825943759702e-05, + "loss": 1.0739, + "step": 6583 + }, + { + "epoch": 0.4000243028130506, + "grad_norm": 0.6491912007331848, + "learning_rate": 6.57291729643289e-05, + "loss": 1.0359, + "step": 6584 + }, + { + "epoch": 0.40008505984567716, + "grad_norm": 0.24847102165222168, + "learning_rate": 6.57200859145049e-05, + "loss": 1.1696, + "step": 6585 + }, + { + "epoch": 0.40014581687830364, + "grad_norm": 0.5888840556144714, + "learning_rate": 6.57109982884581e-05, + "loss": 1.1772, + "step": 6586 + }, + { + "epoch": 0.4002065739109302, + "grad_norm": 0.49292904138565063, + "learning_rate": 6.570191008652165e-05, + "loss": 1.0747, + "step": 6587 + }, + { + "epoch": 0.4002673309435567, + "grad_norm": 0.2689090669155121, + "learning_rate": 6.569282130902862e-05, + "loss": 1.2028, + "step": 6588 + }, + { + "epoch": 0.40032808797618324, + "grad_norm": 0.4863063097000122, + "learning_rate": 6.568373195631221e-05, + "loss": 1.0848, + "step": 6589 + }, + { + "epoch": 0.4003888450088098, + "grad_norm": 0.19241324067115784, + "learning_rate": 6.567464202870559e-05, + "loss": 1.0467, + "step": 6590 + }, + { + "epoch": 0.4004496020414363, + "grad_norm": 0.2618734836578369, + "learning_rate": 6.566555152654189e-05, + "loss": 1.0588, + "step": 6591 + }, + { + "epoch": 0.40051035907406285, + "grad_norm": 0.21361418068408966, + "learning_rate": 6.56564604501544e-05, + "loss": 1.0638, + "step": 6592 + }, + { + "epoch": 0.40057111610668933, + "grad_norm": 0.24492493271827698, + "learning_rate": 6.564736879987632e-05, + "loss": 1.0817, + "step": 6593 + }, + { + "epoch": 0.40063187313931586, + "grad_norm": 0.3205677270889282, + "learning_rate": 6.563827657604092e-05, + "loss": 1.147, + "step": 6594 + }, + { + "epoch": 0.4006926301719424, + "grad_norm": 0.4201907813549042, + "learning_rate": 6.562918377898146e-05, + "loss": 1.1555, + "step": 6595 + }, + { + "epoch": 0.40075338720456893, + "grad_norm": 0.24224303662776947, + "learning_rate": 6.562009040903124e-05, + "loss": 1.0142, + "step": 6596 + }, + { + "epoch": 0.40081414423719547, + "grad_norm": 2.781158447265625, + "learning_rate": 6.561099646652361e-05, + "loss": 1.1098, + "step": 6597 + }, + { + "epoch": 0.400874901269822, + "grad_norm": 0.3630264699459076, + "learning_rate": 6.560190195179188e-05, + "loss": 1.0513, + "step": 6598 + }, + { + "epoch": 0.4009356583024485, + "grad_norm": 0.24826371669769287, + "learning_rate": 6.55928068651694e-05, + "loss": 1.1529, + "step": 6599 + }, + { + "epoch": 0.400996415335075, + "grad_norm": 0.25876083970069885, + "learning_rate": 6.55837112069896e-05, + "loss": 1.0621, + "step": 6600 + }, + { + "epoch": 0.40105717236770155, + "grad_norm": 0.21266372501850128, + "learning_rate": 6.557461497758583e-05, + "loss": 1.096, + "step": 6601 + }, + { + "epoch": 0.4011179294003281, + "grad_norm": 0.37797823548316956, + "learning_rate": 6.556551817729153e-05, + "loss": 1.1549, + "step": 6602 + }, + { + "epoch": 0.4011786864329546, + "grad_norm": 0.19926127791404724, + "learning_rate": 6.555642080644018e-05, + "loss": 1.0719, + "step": 6603 + }, + { + "epoch": 0.40123944346558116, + "grad_norm": 0.19590777158737183, + "learning_rate": 6.55473228653652e-05, + "loss": 1.1631, + "step": 6604 + }, + { + "epoch": 0.4013002004982077, + "grad_norm": 0.26935631036758423, + "learning_rate": 6.55382243544001e-05, + "loss": 1.0355, + "step": 6605 + }, + { + "epoch": 0.4013609575308342, + "grad_norm": 0.9733561277389526, + "learning_rate": 6.552912527387838e-05, + "loss": 1.0949, + "step": 6606 + }, + { + "epoch": 0.4014217145634607, + "grad_norm": 0.17518796026706696, + "learning_rate": 6.552002562413358e-05, + "loss": 1.062, + "step": 6607 + }, + { + "epoch": 0.40148247159608724, + "grad_norm": 0.2723778188228607, + "learning_rate": 6.551092540549923e-05, + "loss": 1.0526, + "step": 6608 + }, + { + "epoch": 0.4015432286287138, + "grad_norm": 0.16333264112472534, + "learning_rate": 6.550182461830892e-05, + "loss": 1.0794, + "step": 6609 + }, + { + "epoch": 0.4016039856613403, + "grad_norm": 0.2515769898891449, + "learning_rate": 6.549272326289623e-05, + "loss": 1.1861, + "step": 6610 + }, + { + "epoch": 0.40166474269396685, + "grad_norm": 0.29475200176239014, + "learning_rate": 6.548362133959478e-05, + "loss": 1.1018, + "step": 6611 + }, + { + "epoch": 0.4017254997265933, + "grad_norm": 0.21763141453266144, + "learning_rate": 6.547451884873817e-05, + "loss": 1.131, + "step": 6612 + }, + { + "epoch": 0.40178625675921986, + "grad_norm": 0.31892532110214233, + "learning_rate": 6.54654157906601e-05, + "loss": 1.0919, + "step": 6613 + }, + { + "epoch": 0.4018470137918464, + "grad_norm": 0.18061256408691406, + "learning_rate": 6.545631216569423e-05, + "loss": 1.1106, + "step": 6614 + }, + { + "epoch": 0.40190777082447293, + "grad_norm": 0.16695570945739746, + "learning_rate": 6.544720797417426e-05, + "loss": 1.0687, + "step": 6615 + }, + { + "epoch": 0.40196852785709947, + "grad_norm": 0.22369176149368286, + "learning_rate": 6.543810321643386e-05, + "loss": 1.0769, + "step": 6616 + }, + { + "epoch": 0.402029284889726, + "grad_norm": 0.688895583152771, + "learning_rate": 6.542899789280684e-05, + "loss": 1.0656, + "step": 6617 + }, + { + "epoch": 0.40209004192235254, + "grad_norm": 0.49183815717697144, + "learning_rate": 6.541989200362689e-05, + "loss": 1.1012, + "step": 6618 + }, + { + "epoch": 0.402150798954979, + "grad_norm": 0.27219653129577637, + "learning_rate": 6.541078554922784e-05, + "loss": 1.1326, + "step": 6619 + }, + { + "epoch": 0.40221155598760555, + "grad_norm": 0.2710092067718506, + "learning_rate": 6.540167852994346e-05, + "loss": 1.0777, + "step": 6620 + }, + { + "epoch": 0.4022723130202321, + "grad_norm": 2.8927762508392334, + "learning_rate": 6.539257094610757e-05, + "loss": 1.0499, + "step": 6621 + }, + { + "epoch": 0.4023330700528586, + "grad_norm": 0.4115152955055237, + "learning_rate": 6.538346279805401e-05, + "loss": 1.1184, + "step": 6622 + }, + { + "epoch": 0.40239382708548516, + "grad_norm": 0.19886744022369385, + "learning_rate": 6.537435408611667e-05, + "loss": 1.1708, + "step": 6623 + }, + { + "epoch": 0.4024545841181117, + "grad_norm": 0.30183348059654236, + "learning_rate": 6.53652448106294e-05, + "loss": 1.0269, + "step": 6624 + }, + { + "epoch": 0.40251534115073817, + "grad_norm": 0.377632200717926, + "learning_rate": 6.53561349719261e-05, + "loss": 1.1511, + "step": 6625 + }, + { + "epoch": 0.4025760981833647, + "grad_norm": 0.21191415190696716, + "learning_rate": 6.534702457034071e-05, + "loss": 1.1216, + "step": 6626 + }, + { + "epoch": 0.40263685521599124, + "grad_norm": 0.35215067863464355, + "learning_rate": 6.533791360620717e-05, + "loss": 1.0614, + "step": 6627 + }, + { + "epoch": 0.4026976122486178, + "grad_norm": 0.24649515748023987, + "learning_rate": 6.532880207985945e-05, + "loss": 1.1314, + "step": 6628 + }, + { + "epoch": 0.4027583692812443, + "grad_norm": 0.20147667825222015, + "learning_rate": 6.531968999163151e-05, + "loss": 1.0647, + "step": 6629 + }, + { + "epoch": 0.40281912631387085, + "grad_norm": 0.844028890132904, + "learning_rate": 6.53105773418574e-05, + "loss": 1.2403, + "step": 6630 + }, + { + "epoch": 0.4028798833464974, + "grad_norm": 0.26139578223228455, + "learning_rate": 6.53014641308711e-05, + "loss": 1.0781, + "step": 6631 + }, + { + "epoch": 0.40294064037912386, + "grad_norm": 0.27062833309173584, + "learning_rate": 6.529235035900667e-05, + "loss": 1.0688, + "step": 6632 + }, + { + "epoch": 0.4030013974117504, + "grad_norm": 0.25070080161094666, + "learning_rate": 6.528323602659817e-05, + "loss": 1.0982, + "step": 6633 + }, + { + "epoch": 0.40306215444437693, + "grad_norm": 0.2962891161441803, + "learning_rate": 6.527412113397973e-05, + "loss": 1.1395, + "step": 6634 + }, + { + "epoch": 0.40312291147700346, + "grad_norm": 0.21380262076854706, + "learning_rate": 6.52650056814854e-05, + "loss": 1.043, + "step": 6635 + }, + { + "epoch": 0.40318366850963, + "grad_norm": 0.25312748551368713, + "learning_rate": 6.525588966944936e-05, + "loss": 1.0764, + "step": 6636 + }, + { + "epoch": 0.40324442554225653, + "grad_norm": 0.39826807379722595, + "learning_rate": 6.524677309820571e-05, + "loss": 1.0616, + "step": 6637 + }, + { + "epoch": 0.40330518257488307, + "grad_norm": 0.23450417816638947, + "learning_rate": 6.523765596808866e-05, + "loss": 1.0835, + "step": 6638 + }, + { + "epoch": 0.40336593960750955, + "grad_norm": 0.6194207072257996, + "learning_rate": 6.52285382794324e-05, + "loss": 1.001, + "step": 6639 + }, + { + "epoch": 0.4034266966401361, + "grad_norm": 0.17375154793262482, + "learning_rate": 6.52194200325711e-05, + "loss": 1.0259, + "step": 6640 + }, + { + "epoch": 0.4034874536727626, + "grad_norm": 0.38126394152641296, + "learning_rate": 6.521030122783903e-05, + "loss": 1.0465, + "step": 6641 + }, + { + "epoch": 0.40354821070538915, + "grad_norm": 0.6500436067581177, + "learning_rate": 6.520118186557043e-05, + "loss": 1.0566, + "step": 6642 + }, + { + "epoch": 0.4036089677380157, + "grad_norm": 0.16747702658176422, + "learning_rate": 6.519206194609954e-05, + "loss": 1.0625, + "step": 6643 + }, + { + "epoch": 0.4036697247706422, + "grad_norm": 0.31581762433052063, + "learning_rate": 6.518294146976073e-05, + "loss": 1.0308, + "step": 6644 + }, + { + "epoch": 0.4037304818032687, + "grad_norm": 0.250774085521698, + "learning_rate": 6.517382043688823e-05, + "loss": 1.0848, + "step": 6645 + }, + { + "epoch": 0.40379123883589524, + "grad_norm": 0.19357405602931976, + "learning_rate": 6.516469884781644e-05, + "loss": 1.0776, + "step": 6646 + }, + { + "epoch": 0.4038519958685218, + "grad_norm": 0.616610586643219, + "learning_rate": 6.515557670287966e-05, + "loss": 1.1319, + "step": 6647 + }, + { + "epoch": 0.4039127529011483, + "grad_norm": 0.4080381691455841, + "learning_rate": 6.514645400241229e-05, + "loss": 1.0765, + "step": 6648 + }, + { + "epoch": 0.40397350993377484, + "grad_norm": 0.4306723475456238, + "learning_rate": 6.513733074674872e-05, + "loss": 1.3083, + "step": 6649 + }, + { + "epoch": 0.4040342669664014, + "grad_norm": 0.24498018622398376, + "learning_rate": 6.512820693622337e-05, + "loss": 1.163, + "step": 6650 + }, + { + "epoch": 0.4040950239990279, + "grad_norm": 0.3446204662322998, + "learning_rate": 6.511908257117067e-05, + "loss": 1.1468, + "step": 6651 + }, + { + "epoch": 0.4041557810316544, + "grad_norm": 0.21964560449123383, + "learning_rate": 6.510995765192508e-05, + "loss": 1.0484, + "step": 6652 + }, + { + "epoch": 0.40421653806428093, + "grad_norm": 0.3722042441368103, + "learning_rate": 6.510083217882106e-05, + "loss": 1.2235, + "step": 6653 + }, + { + "epoch": 0.40427729509690746, + "grad_norm": 0.2237611711025238, + "learning_rate": 6.509170615219313e-05, + "loss": 1.1267, + "step": 6654 + }, + { + "epoch": 0.404338052129534, + "grad_norm": 0.4426112174987793, + "learning_rate": 6.508257957237578e-05, + "loss": 1.2724, + "step": 6655 + }, + { + "epoch": 0.40439880916216053, + "grad_norm": 0.17184481024742126, + "learning_rate": 6.507345243970356e-05, + "loss": 1.0292, + "step": 6656 + }, + { + "epoch": 0.40445956619478707, + "grad_norm": 1.3717663288116455, + "learning_rate": 6.506432475451106e-05, + "loss": 1.0541, + "step": 6657 + }, + { + "epoch": 0.40452032322741355, + "grad_norm": 0.1891658753156662, + "learning_rate": 6.505519651713277e-05, + "loss": 1.0681, + "step": 6658 + }, + { + "epoch": 0.4045810802600401, + "grad_norm": 0.26333919167518616, + "learning_rate": 6.504606772790338e-05, + "loss": 1.2012, + "step": 6659 + }, + { + "epoch": 0.4046418372926666, + "grad_norm": 0.2559639811515808, + "learning_rate": 6.503693838715748e-05, + "loss": 1.0348, + "step": 6660 + }, + { + "epoch": 0.40470259432529315, + "grad_norm": 0.22521044313907623, + "learning_rate": 6.502780849522965e-05, + "loss": 1.1341, + "step": 6661 + }, + { + "epoch": 0.4047633513579197, + "grad_norm": 0.2567588686943054, + "learning_rate": 6.501867805245463e-05, + "loss": 1.068, + "step": 6662 + }, + { + "epoch": 0.4048241083905462, + "grad_norm": 0.20376884937286377, + "learning_rate": 6.500954705916706e-05, + "loss": 1.1644, + "step": 6663 + }, + { + "epoch": 0.40488486542317276, + "grad_norm": 0.1929156482219696, + "learning_rate": 6.500041551570163e-05, + "loss": 1.1083, + "step": 6664 + }, + { + "epoch": 0.40494562245579924, + "grad_norm": 0.2436862587928772, + "learning_rate": 6.499128342239307e-05, + "loss": 1.205, + "step": 6665 + }, + { + "epoch": 0.40500637948842577, + "grad_norm": 0.23891015350818634, + "learning_rate": 6.498215077957612e-05, + "loss": 1.1489, + "step": 6666 + }, + { + "epoch": 0.4050671365210523, + "grad_norm": 0.2901690900325775, + "learning_rate": 6.497301758758557e-05, + "loss": 1.2176, + "step": 6667 + }, + { + "epoch": 0.40512789355367884, + "grad_norm": 0.25642985105514526, + "learning_rate": 6.496388384675613e-05, + "loss": 1.1278, + "step": 6668 + }, + { + "epoch": 0.4051886505863054, + "grad_norm": 0.33963361382484436, + "learning_rate": 6.495474955742263e-05, + "loss": 1.1456, + "step": 6669 + }, + { + "epoch": 0.4052494076189319, + "grad_norm": 0.19503577053546906, + "learning_rate": 6.494561471991991e-05, + "loss": 1.0774, + "step": 6670 + }, + { + "epoch": 0.4053101646515584, + "grad_norm": 0.16205629706382751, + "learning_rate": 6.493647933458278e-05, + "loss": 1.0775, + "step": 6671 + }, + { + "epoch": 0.4053709216841849, + "grad_norm": 0.22484925389289856, + "learning_rate": 6.492734340174614e-05, + "loss": 1.0472, + "step": 6672 + }, + { + "epoch": 0.40543167871681146, + "grad_norm": 0.15797729790210724, + "learning_rate": 6.491820692174481e-05, + "loss": 1.0098, + "step": 6673 + }, + { + "epoch": 0.405492435749438, + "grad_norm": 0.18715055286884308, + "learning_rate": 6.490906989491375e-05, + "loss": 1.0592, + "step": 6674 + }, + { + "epoch": 0.40555319278206453, + "grad_norm": 0.6883099675178528, + "learning_rate": 6.489993232158783e-05, + "loss": 1.1596, + "step": 6675 + }, + { + "epoch": 0.40561394981469107, + "grad_norm": 0.250351220369339, + "learning_rate": 6.489079420210203e-05, + "loss": 1.2482, + "step": 6676 + }, + { + "epoch": 0.4056747068473176, + "grad_norm": 0.2168281227350235, + "learning_rate": 6.488165553679128e-05, + "loss": 1.1465, + "step": 6677 + }, + { + "epoch": 0.4057354638799441, + "grad_norm": 0.20468685030937195, + "learning_rate": 6.487251632599056e-05, + "loss": 1.1701, + "step": 6678 + }, + { + "epoch": 0.4057962209125706, + "grad_norm": 0.22917494177818298, + "learning_rate": 6.486337657003489e-05, + "loss": 1.1122, + "step": 6679 + }, + { + "epoch": 0.40585697794519715, + "grad_norm": 0.5877894759178162, + "learning_rate": 6.485423626925928e-05, + "loss": 1.136, + "step": 6680 + }, + { + "epoch": 0.4059177349778237, + "grad_norm": 0.22052550315856934, + "learning_rate": 6.484509542399876e-05, + "loss": 1.0174, + "step": 6681 + }, + { + "epoch": 0.4059784920104502, + "grad_norm": 2.845457077026367, + "learning_rate": 6.48359540345884e-05, + "loss": 1.0799, + "step": 6682 + }, + { + "epoch": 0.40603924904307676, + "grad_norm": 0.27754032611846924, + "learning_rate": 6.482681210136328e-05, + "loss": 1.2168, + "step": 6683 + }, + { + "epoch": 0.4061000060757033, + "grad_norm": 0.18512116372585297, + "learning_rate": 6.48176696246585e-05, + "loss": 1.1198, + "step": 6684 + }, + { + "epoch": 0.40616076310832977, + "grad_norm": 1.2609513998031616, + "learning_rate": 6.480852660480918e-05, + "loss": 1.1598, + "step": 6685 + }, + { + "epoch": 0.4062215201409563, + "grad_norm": 0.17133130133152008, + "learning_rate": 6.479938304215043e-05, + "loss": 1.0673, + "step": 6686 + }, + { + "epoch": 0.40628227717358284, + "grad_norm": 0.18530414998531342, + "learning_rate": 6.479023893701745e-05, + "loss": 1.0757, + "step": 6687 + }, + { + "epoch": 0.4063430342062094, + "grad_norm": 0.798298716545105, + "learning_rate": 6.478109428974542e-05, + "loss": 1.1138, + "step": 6688 + }, + { + "epoch": 0.4064037912388359, + "grad_norm": 0.31745538115501404, + "learning_rate": 6.47719491006695e-05, + "loss": 0.9951, + "step": 6689 + }, + { + "epoch": 0.40646454827146244, + "grad_norm": 0.2509850263595581, + "learning_rate": 6.476280337012494e-05, + "loss": 1.1168, + "step": 6690 + }, + { + "epoch": 0.4065253053040889, + "grad_norm": 0.2813272178173065, + "learning_rate": 6.475365709844697e-05, + "loss": 1.0689, + "step": 6691 + }, + { + "epoch": 0.40658606233671546, + "grad_norm": 0.29280996322631836, + "learning_rate": 6.474451028597083e-05, + "loss": 0.9874, + "step": 6692 + }, + { + "epoch": 0.406646819369342, + "grad_norm": 0.20532162487506866, + "learning_rate": 6.473536293303182e-05, + "loss": 1.0515, + "step": 6693 + }, + { + "epoch": 0.40670757640196853, + "grad_norm": 0.3780338168144226, + "learning_rate": 6.472621503996525e-05, + "loss": 1.1018, + "step": 6694 + }, + { + "epoch": 0.40676833343459506, + "grad_norm": 0.22907578945159912, + "learning_rate": 6.471706660710641e-05, + "loss": 1.1517, + "step": 6695 + }, + { + "epoch": 0.4068290904672216, + "grad_norm": 0.27387797832489014, + "learning_rate": 6.470791763479065e-05, + "loss": 1.0849, + "step": 6696 + }, + { + "epoch": 0.40688984749984813, + "grad_norm": 0.7252191305160522, + "learning_rate": 6.469876812335333e-05, + "loss": 1.1838, + "step": 6697 + }, + { + "epoch": 0.4069506045324746, + "grad_norm": 0.2331368774175644, + "learning_rate": 6.468961807312982e-05, + "loss": 1.1424, + "step": 6698 + }, + { + "epoch": 0.40701136156510115, + "grad_norm": 0.3179691433906555, + "learning_rate": 6.468046748445552e-05, + "loss": 1.049, + "step": 6699 + }, + { + "epoch": 0.4070721185977277, + "grad_norm": 0.3126300871372223, + "learning_rate": 6.467131635766584e-05, + "loss": 1.1215, + "step": 6700 + }, + { + "epoch": 0.4071328756303542, + "grad_norm": 0.26539427042007446, + "learning_rate": 6.466216469309624e-05, + "loss": 1.0819, + "step": 6701 + }, + { + "epoch": 0.40719363266298075, + "grad_norm": 1.617948293685913, + "learning_rate": 6.465301249108214e-05, + "loss": 1.0492, + "step": 6702 + }, + { + "epoch": 0.4072543896956073, + "grad_norm": 2.7478201389312744, + "learning_rate": 6.464385975195905e-05, + "loss": 1.1033, + "step": 6703 + }, + { + "epoch": 0.40731514672823377, + "grad_norm": 0.3685939610004425, + "learning_rate": 6.463470647606244e-05, + "loss": 1.1065, + "step": 6704 + }, + { + "epoch": 0.4073759037608603, + "grad_norm": 0.26526594161987305, + "learning_rate": 6.462555266372784e-05, + "loss": 1.1374, + "step": 6705 + }, + { + "epoch": 0.40743666079348684, + "grad_norm": 0.37030771374702454, + "learning_rate": 6.461639831529078e-05, + "loss": 1.0706, + "step": 6706 + }, + { + "epoch": 0.4074974178261134, + "grad_norm": 0.17250211536884308, + "learning_rate": 6.460724343108681e-05, + "loss": 1.0332, + "step": 6707 + }, + { + "epoch": 0.4075581748587399, + "grad_norm": 0.3151414692401886, + "learning_rate": 6.459808801145151e-05, + "loss": 1.1008, + "step": 6708 + }, + { + "epoch": 0.40761893189136644, + "grad_norm": 0.25966957211494446, + "learning_rate": 6.458893205672047e-05, + "loss": 1.0348, + "step": 6709 + }, + { + "epoch": 0.407679688923993, + "grad_norm": 0.2645031213760376, + "learning_rate": 6.457977556722931e-05, + "loss": 1.0641, + "step": 6710 + }, + { + "epoch": 0.40774044595661946, + "grad_norm": 0.3007969856262207, + "learning_rate": 6.457061854331368e-05, + "loss": 1.0673, + "step": 6711 + }, + { + "epoch": 0.407801202989246, + "grad_norm": 0.35739219188690186, + "learning_rate": 6.456146098530918e-05, + "loss": 1.2546, + "step": 6712 + }, + { + "epoch": 0.4078619600218725, + "grad_norm": 0.2460625171661377, + "learning_rate": 6.455230289355151e-05, + "loss": 1.0232, + "step": 6713 + }, + { + "epoch": 0.40792271705449906, + "grad_norm": 0.5621457695960999, + "learning_rate": 6.454314426837639e-05, + "loss": 1.1204, + "step": 6714 + }, + { + "epoch": 0.4079834740871256, + "grad_norm": 0.20863153040409088, + "learning_rate": 6.453398511011948e-05, + "loss": 1.1357, + "step": 6715 + }, + { + "epoch": 0.40804423111975213, + "grad_norm": 0.3403131067752838, + "learning_rate": 6.452482541911656e-05, + "loss": 1.1289, + "step": 6716 + }, + { + "epoch": 0.4081049881523786, + "grad_norm": 0.19946803152561188, + "learning_rate": 6.451566519570336e-05, + "loss": 1.025, + "step": 6717 + }, + { + "epoch": 0.40816574518500515, + "grad_norm": 0.18478591740131378, + "learning_rate": 6.450650444021561e-05, + "loss": 1.1716, + "step": 6718 + }, + { + "epoch": 0.4082265022176317, + "grad_norm": 0.32393935322761536, + "learning_rate": 6.449734315298917e-05, + "loss": 1.0821, + "step": 6719 + }, + { + "epoch": 0.4082872592502582, + "grad_norm": 0.2728693187236786, + "learning_rate": 6.448818133435978e-05, + "loss": 1.0641, + "step": 6720 + }, + { + "epoch": 0.40834801628288475, + "grad_norm": 0.2620105445384979, + "learning_rate": 6.447901898466335e-05, + "loss": 1.103, + "step": 6721 + }, + { + "epoch": 0.4084087733155113, + "grad_norm": 0.26071810722351074, + "learning_rate": 6.446985610423565e-05, + "loss": 1.1357, + "step": 6722 + }, + { + "epoch": 0.4084695303481378, + "grad_norm": 0.19963276386260986, + "learning_rate": 6.446069269341257e-05, + "loss": 1.1289, + "step": 6723 + }, + { + "epoch": 0.4085302873807643, + "grad_norm": 0.18390987813472748, + "learning_rate": 6.445152875253002e-05, + "loss": 1.0645, + "step": 6724 + }, + { + "epoch": 0.40859104441339084, + "grad_norm": 0.22196181118488312, + "learning_rate": 6.44423642819239e-05, + "loss": 1.1129, + "step": 6725 + }, + { + "epoch": 0.40865180144601737, + "grad_norm": 0.2822682857513428, + "learning_rate": 6.443319928193011e-05, + "loss": 1.05, + "step": 6726 + }, + { + "epoch": 0.4087125584786439, + "grad_norm": 0.15827298164367676, + "learning_rate": 6.442403375288462e-05, + "loss": 1.0432, + "step": 6727 + }, + { + "epoch": 0.40877331551127044, + "grad_norm": 0.3720375597476959, + "learning_rate": 6.441486769512336e-05, + "loss": 1.0892, + "step": 6728 + }, + { + "epoch": 0.408834072543897, + "grad_norm": 2.064282178878784, + "learning_rate": 6.440570110898236e-05, + "loss": 1.1702, + "step": 6729 + }, + { + "epoch": 0.40889482957652346, + "grad_norm": 0.3302643597126007, + "learning_rate": 6.43965339947976e-05, + "loss": 1.2387, + "step": 6730 + }, + { + "epoch": 0.40895558660915, + "grad_norm": 0.2584555149078369, + "learning_rate": 6.43873663529051e-05, + "loss": 1.1412, + "step": 6731 + }, + { + "epoch": 0.4090163436417765, + "grad_norm": 0.6520850658416748, + "learning_rate": 6.43781981836409e-05, + "loss": 1.1586, + "step": 6732 + }, + { + "epoch": 0.40907710067440306, + "grad_norm": 0.2644966244697571, + "learning_rate": 6.436902948734107e-05, + "loss": 1.0596, + "step": 6733 + }, + { + "epoch": 0.4091378577070296, + "grad_norm": 0.2285577803850174, + "learning_rate": 6.43598602643417e-05, + "loss": 1.0264, + "step": 6734 + }, + { + "epoch": 0.40919861473965613, + "grad_norm": 0.3187740445137024, + "learning_rate": 6.435069051497886e-05, + "loss": 1.1387, + "step": 6735 + }, + { + "epoch": 0.40925937177228267, + "grad_norm": 0.1811455935239792, + "learning_rate": 6.434152023958869e-05, + "loss": 1.0098, + "step": 6736 + }, + { + "epoch": 0.40932012880490914, + "grad_norm": 0.19887258112430573, + "learning_rate": 6.433234943850733e-05, + "loss": 1.0133, + "step": 6737 + }, + { + "epoch": 0.4093808858375357, + "grad_norm": 0.20278601348400116, + "learning_rate": 6.432317811207091e-05, + "loss": 1.1323, + "step": 6738 + }, + { + "epoch": 0.4094416428701622, + "grad_norm": 0.24427703022956848, + "learning_rate": 6.431400626061565e-05, + "loss": 1.1486, + "step": 6739 + }, + { + "epoch": 0.40950239990278875, + "grad_norm": 5.299543857574463, + "learning_rate": 6.430483388447772e-05, + "loss": 1.0831, + "step": 6740 + }, + { + "epoch": 0.4095631569354153, + "grad_norm": 0.33214589953422546, + "learning_rate": 6.429566098399334e-05, + "loss": 1.0609, + "step": 6741 + }, + { + "epoch": 0.4096239139680418, + "grad_norm": 1.0026065111160278, + "learning_rate": 6.428648755949873e-05, + "loss": 1.1172, + "step": 6742 + }, + { + "epoch": 0.40968467100066835, + "grad_norm": 0.23603606224060059, + "learning_rate": 6.427731361133016e-05, + "loss": 1.1307, + "step": 6743 + }, + { + "epoch": 0.40974542803329483, + "grad_norm": 0.23268668353557587, + "learning_rate": 6.426813913982392e-05, + "loss": 1.2111, + "step": 6744 + }, + { + "epoch": 0.40980618506592137, + "grad_norm": 0.22991184890270233, + "learning_rate": 6.425896414531627e-05, + "loss": 1.1557, + "step": 6745 + }, + { + "epoch": 0.4098669420985479, + "grad_norm": 0.22204375267028809, + "learning_rate": 6.424978862814352e-05, + "loss": 1.1935, + "step": 6746 + }, + { + "epoch": 0.40992769913117444, + "grad_norm": 0.234690859913826, + "learning_rate": 6.424061258864203e-05, + "loss": 1.0437, + "step": 6747 + }, + { + "epoch": 0.409988456163801, + "grad_norm": 0.26384657621383667, + "learning_rate": 6.423143602714813e-05, + "loss": 1.0959, + "step": 6748 + }, + { + "epoch": 0.4100492131964275, + "grad_norm": 0.20148693025112152, + "learning_rate": 6.422225894399818e-05, + "loss": 1.0443, + "step": 6749 + }, + { + "epoch": 0.410109970229054, + "grad_norm": 0.2733219265937805, + "learning_rate": 6.421308133952858e-05, + "loss": 1.0501, + "step": 6750 + }, + { + "epoch": 0.4101707272616805, + "grad_norm": 0.24468663334846497, + "learning_rate": 6.420390321407575e-05, + "loss": 1.1939, + "step": 6751 + }, + { + "epoch": 0.41023148429430706, + "grad_norm": 0.20528742671012878, + "learning_rate": 6.41947245679761e-05, + "loss": 1.1118, + "step": 6752 + }, + { + "epoch": 0.4102922413269336, + "grad_norm": 0.4635936915874481, + "learning_rate": 6.418554540156607e-05, + "loss": 1.1397, + "step": 6753 + }, + { + "epoch": 0.41035299835956013, + "grad_norm": 0.1885901838541031, + "learning_rate": 6.417636571518213e-05, + "loss": 1.0626, + "step": 6754 + }, + { + "epoch": 0.41041375539218666, + "grad_norm": 0.36399173736572266, + "learning_rate": 6.416718550916078e-05, + "loss": 1.1929, + "step": 6755 + }, + { + "epoch": 0.4104745124248132, + "grad_norm": 0.23233751952648163, + "learning_rate": 6.415800478383849e-05, + "loss": 1.0845, + "step": 6756 + }, + { + "epoch": 0.4105352694574397, + "grad_norm": 0.38636812567710876, + "learning_rate": 6.41488235395518e-05, + "loss": 1.1785, + "step": 6757 + }, + { + "epoch": 0.4105960264900662, + "grad_norm": 0.1635327786207199, + "learning_rate": 6.413964177663726e-05, + "loss": 1.0562, + "step": 6758 + }, + { + "epoch": 0.41065678352269275, + "grad_norm": 0.2912883460521698, + "learning_rate": 6.41304594954314e-05, + "loss": 1.2798, + "step": 6759 + }, + { + "epoch": 0.4107175405553193, + "grad_norm": 0.29877224564552307, + "learning_rate": 6.412127669627083e-05, + "loss": 1.0287, + "step": 6760 + }, + { + "epoch": 0.4107782975879458, + "grad_norm": 0.34040167927742004, + "learning_rate": 6.411209337949214e-05, + "loss": 1.1679, + "step": 6761 + }, + { + "epoch": 0.41083905462057235, + "grad_norm": 0.16134081780910492, + "learning_rate": 6.410290954543191e-05, + "loss": 1.0179, + "step": 6762 + }, + { + "epoch": 0.41089981165319883, + "grad_norm": 1.497665524482727, + "learning_rate": 6.409372519442682e-05, + "loss": 1.2176, + "step": 6763 + }, + { + "epoch": 0.41096056868582537, + "grad_norm": 0.2162693440914154, + "learning_rate": 6.408454032681352e-05, + "loss": 1.081, + "step": 6764 + }, + { + "epoch": 0.4110213257184519, + "grad_norm": 0.26887598633766174, + "learning_rate": 6.407535494292867e-05, + "loss": 1.166, + "step": 6765 + }, + { + "epoch": 0.41108208275107844, + "grad_norm": 0.22757427394390106, + "learning_rate": 6.406616904310894e-05, + "loss": 1.126, + "step": 6766 + }, + { + "epoch": 0.41114283978370497, + "grad_norm": 0.9545022249221802, + "learning_rate": 6.405698262769109e-05, + "loss": 1.0854, + "step": 6767 + }, + { + "epoch": 0.4112035968163315, + "grad_norm": 0.3087695240974426, + "learning_rate": 6.404779569701182e-05, + "loss": 1.0157, + "step": 6768 + }, + { + "epoch": 0.41126435384895804, + "grad_norm": 0.18219462037086487, + "learning_rate": 6.403860825140789e-05, + "loss": 1.084, + "step": 6769 + }, + { + "epoch": 0.4113251108815845, + "grad_norm": 0.2968722879886627, + "learning_rate": 6.402942029121603e-05, + "loss": 1.2485, + "step": 6770 + }, + { + "epoch": 0.41138586791421106, + "grad_norm": 0.3325619697570801, + "learning_rate": 6.40202318167731e-05, + "loss": 1.1333, + "step": 6771 + }, + { + "epoch": 0.4114466249468376, + "grad_norm": 0.2572338283061981, + "learning_rate": 6.401104282841583e-05, + "loss": 1.2048, + "step": 6772 + }, + { + "epoch": 0.4115073819794641, + "grad_norm": 0.31122156977653503, + "learning_rate": 6.400185332648108e-05, + "loss": 1.1508, + "step": 6773 + }, + { + "epoch": 0.41156813901209066, + "grad_norm": 0.4123188853263855, + "learning_rate": 6.399266331130571e-05, + "loss": 1.1834, + "step": 6774 + }, + { + "epoch": 0.4116288960447172, + "grad_norm": 0.18454305827617645, + "learning_rate": 6.398347278322654e-05, + "loss": 1.1116, + "step": 6775 + }, + { + "epoch": 0.4116896530773437, + "grad_norm": 0.2674819827079773, + "learning_rate": 6.397428174258047e-05, + "loss": 1.1396, + "step": 6776 + }, + { + "epoch": 0.4117504101099702, + "grad_norm": 0.15941821038722992, + "learning_rate": 6.396509018970442e-05, + "loss": 1.0497, + "step": 6777 + }, + { + "epoch": 0.41181116714259675, + "grad_norm": 0.2549004554748535, + "learning_rate": 6.395589812493528e-05, + "loss": 1.1235, + "step": 6778 + }, + { + "epoch": 0.4118719241752233, + "grad_norm": 0.22597821056842804, + "learning_rate": 6.394670554861002e-05, + "loss": 1.0403, + "step": 6779 + }, + { + "epoch": 0.4119326812078498, + "grad_norm": 0.27535757422447205, + "learning_rate": 6.393751246106553e-05, + "loss": 1.1161, + "step": 6780 + }, + { + "epoch": 0.41199343824047635, + "grad_norm": 0.21634641289710999, + "learning_rate": 6.392831886263886e-05, + "loss": 1.079, + "step": 6781 + }, + { + "epoch": 0.4120541952731029, + "grad_norm": 0.49080634117126465, + "learning_rate": 6.391912475366695e-05, + "loss": 1.0574, + "step": 6782 + }, + { + "epoch": 0.41211495230572937, + "grad_norm": 0.2115047127008438, + "learning_rate": 6.390993013448683e-05, + "loss": 1.1185, + "step": 6783 + }, + { + "epoch": 0.4121757093383559, + "grad_norm": 0.38748422265052795, + "learning_rate": 6.390073500543556e-05, + "loss": 1.1152, + "step": 6784 + }, + { + "epoch": 0.41223646637098244, + "grad_norm": 0.2313445657491684, + "learning_rate": 6.389153936685012e-05, + "loss": 1.0813, + "step": 6785 + }, + { + "epoch": 0.41229722340360897, + "grad_norm": 0.15617817640304565, + "learning_rate": 6.388234321906765e-05, + "loss": 1.0625, + "step": 6786 + }, + { + "epoch": 0.4123579804362355, + "grad_norm": 0.28310829401016235, + "learning_rate": 6.38731465624252e-05, + "loss": 1.1617, + "step": 6787 + }, + { + "epoch": 0.41241873746886204, + "grad_norm": 4.891958236694336, + "learning_rate": 6.386394939725989e-05, + "loss": 1.0826, + "step": 6788 + }, + { + "epoch": 0.4124794945014886, + "grad_norm": 0.3007974922657013, + "learning_rate": 6.385475172390882e-05, + "loss": 1.087, + "step": 6789 + }, + { + "epoch": 0.41254025153411505, + "grad_norm": 0.21039758622646332, + "learning_rate": 6.384555354270913e-05, + "loss": 1.0918, + "step": 6790 + }, + { + "epoch": 0.4126010085667416, + "grad_norm": 2.276334285736084, + "learning_rate": 6.383635485399804e-05, + "loss": 1.0908, + "step": 6791 + }, + { + "epoch": 0.4126617655993681, + "grad_norm": 0.2619171738624573, + "learning_rate": 6.382715565811267e-05, + "loss": 1.0975, + "step": 6792 + }, + { + "epoch": 0.41272252263199466, + "grad_norm": 0.17960914969444275, + "learning_rate": 6.381795595539023e-05, + "loss": 1.062, + "step": 6793 + }, + { + "epoch": 0.4127832796646212, + "grad_norm": 0.1766301840543747, + "learning_rate": 6.380875574616795e-05, + "loss": 1.0367, + "step": 6794 + }, + { + "epoch": 0.41284403669724773, + "grad_norm": 0.18725599348545074, + "learning_rate": 6.379955503078306e-05, + "loss": 0.99, + "step": 6795 + }, + { + "epoch": 0.4129047937298742, + "grad_norm": 0.2478991448879242, + "learning_rate": 6.379035380957281e-05, + "loss": 1.0751, + "step": 6796 + }, + { + "epoch": 0.41296555076250074, + "grad_norm": 0.21442686021327972, + "learning_rate": 6.378115208287447e-05, + "loss": 1.096, + "step": 6797 + }, + { + "epoch": 0.4130263077951273, + "grad_norm": 0.3224743604660034, + "learning_rate": 6.377194985102534e-05, + "loss": 1.0786, + "step": 6798 + }, + { + "epoch": 0.4130870648277538, + "grad_norm": 0.3733947277069092, + "learning_rate": 6.376274711436274e-05, + "loss": 1.3206, + "step": 6799 + }, + { + "epoch": 0.41314782186038035, + "grad_norm": 0.32536762952804565, + "learning_rate": 6.375354387322397e-05, + "loss": 1.2171, + "step": 6800 + }, + { + "epoch": 0.4132085788930069, + "grad_norm": 0.2596377432346344, + "learning_rate": 6.37443401279464e-05, + "loss": 1.0716, + "step": 6801 + }, + { + "epoch": 0.4132693359256334, + "grad_norm": 0.2035432904958725, + "learning_rate": 6.373513587886737e-05, + "loss": 1.0573, + "step": 6802 + }, + { + "epoch": 0.4133300929582599, + "grad_norm": 0.45669201016426086, + "learning_rate": 6.37259311263243e-05, + "loss": 1.2967, + "step": 6803 + }, + { + "epoch": 0.41339084999088643, + "grad_norm": 0.2870173156261444, + "learning_rate": 6.371672587065456e-05, + "loss": 1.1081, + "step": 6804 + }, + { + "epoch": 0.41345160702351297, + "grad_norm": 0.22790570557117462, + "learning_rate": 6.370752011219559e-05, + "loss": 1.0425, + "step": 6805 + }, + { + "epoch": 0.4135123640561395, + "grad_norm": 0.3643816113471985, + "learning_rate": 6.369831385128479e-05, + "loss": 1.0715, + "step": 6806 + }, + { + "epoch": 0.41357312108876604, + "grad_norm": 0.3681482970714569, + "learning_rate": 6.368910708825968e-05, + "loss": 1.1834, + "step": 6807 + }, + { + "epoch": 0.4136338781213926, + "grad_norm": 0.19639968872070312, + "learning_rate": 6.36798998234577e-05, + "loss": 1.0867, + "step": 6808 + }, + { + "epoch": 0.41369463515401905, + "grad_norm": 0.5523821711540222, + "learning_rate": 6.367069205721634e-05, + "loss": 1.0426, + "step": 6809 + }, + { + "epoch": 0.4137553921866456, + "grad_norm": 0.16349327564239502, + "learning_rate": 6.366148378987312e-05, + "loss": 1.0534, + "step": 6810 + }, + { + "epoch": 0.4138161492192721, + "grad_norm": 0.3108346462249756, + "learning_rate": 6.365227502176557e-05, + "loss": 1.1469, + "step": 6811 + }, + { + "epoch": 0.41387690625189866, + "grad_norm": 0.3740333020687103, + "learning_rate": 6.364306575323124e-05, + "loss": 1.0315, + "step": 6812 + }, + { + "epoch": 0.4139376632845252, + "grad_norm": 0.1601964235305786, + "learning_rate": 6.363385598460772e-05, + "loss": 1.0722, + "step": 6813 + }, + { + "epoch": 0.4139984203171517, + "grad_norm": 0.34711453318595886, + "learning_rate": 6.362464571623254e-05, + "loss": 1.1837, + "step": 6814 + }, + { + "epoch": 0.41405917734977826, + "grad_norm": 0.43022093176841736, + "learning_rate": 6.361543494844337e-05, + "loss": 1.1429, + "step": 6815 + }, + { + "epoch": 0.41411993438240474, + "grad_norm": 0.23998253047466278, + "learning_rate": 6.360622368157776e-05, + "loss": 1.1566, + "step": 6816 + }, + { + "epoch": 0.4141806914150313, + "grad_norm": 0.3077058792114258, + "learning_rate": 6.359701191597342e-05, + "loss": 1.1365, + "step": 6817 + }, + { + "epoch": 0.4142414484476578, + "grad_norm": 0.6322752833366394, + "learning_rate": 6.358779965196799e-05, + "loss": 1.0272, + "step": 6818 + }, + { + "epoch": 0.41430220548028435, + "grad_norm": 0.20555078983306885, + "learning_rate": 6.35785868898991e-05, + "loss": 1.1345, + "step": 6819 + }, + { + "epoch": 0.4143629625129109, + "grad_norm": 0.5081337094306946, + "learning_rate": 6.356937363010452e-05, + "loss": 1.0639, + "step": 6820 + }, + { + "epoch": 0.4144237195455374, + "grad_norm": 0.6650211215019226, + "learning_rate": 6.35601598729219e-05, + "loss": 1.026, + "step": 6821 + }, + { + "epoch": 0.4144844765781639, + "grad_norm": 0.4151865541934967, + "learning_rate": 6.355094561868901e-05, + "loss": 1.1562, + "step": 6822 + }, + { + "epoch": 0.41454523361079043, + "grad_norm": 0.5195971727371216, + "learning_rate": 6.35417308677436e-05, + "loss": 1.051, + "step": 6823 + }, + { + "epoch": 0.41460599064341697, + "grad_norm": 0.23720337450504303, + "learning_rate": 6.353251562042342e-05, + "loss": 1.0551, + "step": 6824 + }, + { + "epoch": 0.4146667476760435, + "grad_norm": 1.4978221654891968, + "learning_rate": 6.352329987706625e-05, + "loss": 1.2318, + "step": 6825 + }, + { + "epoch": 0.41472750470867004, + "grad_norm": 0.473763644695282, + "learning_rate": 6.351408363800993e-05, + "loss": 1.052, + "step": 6826 + }, + { + "epoch": 0.41478826174129657, + "grad_norm": 0.263308048248291, + "learning_rate": 6.350486690359223e-05, + "loss": 1.0841, + "step": 6827 + }, + { + "epoch": 0.4148490187739231, + "grad_norm": 0.5219154953956604, + "learning_rate": 6.349564967415105e-05, + "loss": 1.1694, + "step": 6828 + }, + { + "epoch": 0.4149097758065496, + "grad_norm": 0.2892615497112274, + "learning_rate": 6.348643195002422e-05, + "loss": 1.235, + "step": 6829 + }, + { + "epoch": 0.4149705328391761, + "grad_norm": 0.24835431575775146, + "learning_rate": 6.34772137315496e-05, + "loss": 1.0966, + "step": 6830 + }, + { + "epoch": 0.41503128987180266, + "grad_norm": 0.1564081758260727, + "learning_rate": 6.346799501906513e-05, + "loss": 1.0097, + "step": 6831 + }, + { + "epoch": 0.4150920469044292, + "grad_norm": 0.3965553343296051, + "learning_rate": 6.34587758129087e-05, + "loss": 1.2222, + "step": 6832 + }, + { + "epoch": 0.4151528039370557, + "grad_norm": 0.2442299723625183, + "learning_rate": 6.344955611341822e-05, + "loss": 1.1695, + "step": 6833 + }, + { + "epoch": 0.41521356096968226, + "grad_norm": 0.19349533319473267, + "learning_rate": 6.344033592093169e-05, + "loss": 1.0874, + "step": 6834 + }, + { + "epoch": 0.41527431800230874, + "grad_norm": 0.3289538025856018, + "learning_rate": 6.343111523578703e-05, + "loss": 1.07, + "step": 6835 + }, + { + "epoch": 0.4153350750349353, + "grad_norm": 0.22725960612297058, + "learning_rate": 6.342189405832225e-05, + "loss": 1.0558, + "step": 6836 + }, + { + "epoch": 0.4153958320675618, + "grad_norm": 0.24844259023666382, + "learning_rate": 6.341267238887534e-05, + "loss": 1.0979, + "step": 6837 + }, + { + "epoch": 0.41545658910018834, + "grad_norm": 0.44897714257240295, + "learning_rate": 6.340345022778435e-05, + "loss": 1.0446, + "step": 6838 + }, + { + "epoch": 0.4155173461328149, + "grad_norm": 0.23755867779254913, + "learning_rate": 6.339422757538727e-05, + "loss": 1.1465, + "step": 6839 + }, + { + "epoch": 0.4155781031654414, + "grad_norm": 1.8848681449890137, + "learning_rate": 6.33850044320222e-05, + "loss": 1.0859, + "step": 6840 + }, + { + "epoch": 0.41563886019806795, + "grad_norm": 0.41928839683532715, + "learning_rate": 6.337578079802724e-05, + "loss": 1.0347, + "step": 6841 + }, + { + "epoch": 0.41569961723069443, + "grad_norm": 0.3091930150985718, + "learning_rate": 6.33665566737404e-05, + "loss": 1.2452, + "step": 6842 + }, + { + "epoch": 0.41576037426332096, + "grad_norm": 0.37989819049835205, + "learning_rate": 6.335733205949988e-05, + "loss": 1.0769, + "step": 6843 + }, + { + "epoch": 0.4158211312959475, + "grad_norm": 0.20921245217323303, + "learning_rate": 6.334810695564375e-05, + "loss": 1.1397, + "step": 6844 + }, + { + "epoch": 0.41588188832857403, + "grad_norm": 0.21210356056690216, + "learning_rate": 6.333888136251018e-05, + "loss": 1.0584, + "step": 6845 + }, + { + "epoch": 0.41594264536120057, + "grad_norm": 0.36930814385414124, + "learning_rate": 6.332965528043735e-05, + "loss": 1.1401, + "step": 6846 + }, + { + "epoch": 0.4160034023938271, + "grad_norm": 0.4475872218608856, + "learning_rate": 6.33204287097634e-05, + "loss": 1.0812, + "step": 6847 + }, + { + "epoch": 0.41606415942645364, + "grad_norm": 0.18987978994846344, + "learning_rate": 6.331120165082659e-05, + "loss": 1.0716, + "step": 6848 + }, + { + "epoch": 0.4161249164590801, + "grad_norm": 0.2781064808368683, + "learning_rate": 6.33019741039651e-05, + "loss": 1.1273, + "step": 6849 + }, + { + "epoch": 0.41618567349170665, + "grad_norm": 0.37764081358909607, + "learning_rate": 6.329274606951718e-05, + "loss": 1.0913, + "step": 6850 + }, + { + "epoch": 0.4162464305243332, + "grad_norm": 0.18990030884742737, + "learning_rate": 6.32835175478211e-05, + "loss": 1.055, + "step": 6851 + }, + { + "epoch": 0.4163071875569597, + "grad_norm": 0.3165608048439026, + "learning_rate": 6.327428853921508e-05, + "loss": 1.0534, + "step": 6852 + }, + { + "epoch": 0.41636794458958626, + "grad_norm": 0.480694055557251, + "learning_rate": 6.326505904403745e-05, + "loss": 1.2529, + "step": 6853 + }, + { + "epoch": 0.4164287016222128, + "grad_norm": 0.28392675518989563, + "learning_rate": 6.325582906262655e-05, + "loss": 0.9947, + "step": 6854 + }, + { + "epoch": 0.4164894586548393, + "grad_norm": 0.284906268119812, + "learning_rate": 6.324659859532065e-05, + "loss": 1.1589, + "step": 6855 + }, + { + "epoch": 0.4165502156874658, + "grad_norm": 1.1190378665924072, + "learning_rate": 6.323736764245813e-05, + "loss": 1.0782, + "step": 6856 + }, + { + "epoch": 0.41661097272009234, + "grad_norm": 0.28652533888816833, + "learning_rate": 6.322813620437732e-05, + "loss": 1.1048, + "step": 6857 + }, + { + "epoch": 0.4166717297527189, + "grad_norm": 0.49379172921180725, + "learning_rate": 6.321890428141665e-05, + "loss": 1.146, + "step": 6858 + }, + { + "epoch": 0.4167324867853454, + "grad_norm": 0.33676677942276, + "learning_rate": 6.320967187391447e-05, + "loss": 1.1649, + "step": 6859 + }, + { + "epoch": 0.41679324381797195, + "grad_norm": 0.2703011631965637, + "learning_rate": 6.32004389822092e-05, + "loss": 1.205, + "step": 6860 + }, + { + "epoch": 0.4168540008505985, + "grad_norm": 0.20038853585720062, + "learning_rate": 6.319120560663932e-05, + "loss": 1.1349, + "step": 6861 + }, + { + "epoch": 0.41691475788322496, + "grad_norm": 0.401863694190979, + "learning_rate": 6.318197174754325e-05, + "loss": 1.1057, + "step": 6862 + }, + { + "epoch": 0.4169755149158515, + "grad_norm": 0.19146637618541718, + "learning_rate": 6.317273740525942e-05, + "loss": 1.1561, + "step": 6863 + }, + { + "epoch": 0.41703627194847803, + "grad_norm": 0.36278507113456726, + "learning_rate": 6.316350258012638e-05, + "loss": 1.1183, + "step": 6864 + }, + { + "epoch": 0.41709702898110457, + "grad_norm": 0.21764323115348816, + "learning_rate": 6.31542672724826e-05, + "loss": 1.115, + "step": 6865 + }, + { + "epoch": 0.4171577860137311, + "grad_norm": 0.24897782504558563, + "learning_rate": 6.314503148266662e-05, + "loss": 1.1734, + "step": 6866 + }, + { + "epoch": 0.41721854304635764, + "grad_norm": 1.1859842538833618, + "learning_rate": 6.313579521101697e-05, + "loss": 1.1562, + "step": 6867 + }, + { + "epoch": 0.4172793000789841, + "grad_norm": 0.3171798586845398, + "learning_rate": 6.31265584578722e-05, + "loss": 1.0295, + "step": 6868 + }, + { + "epoch": 0.41734005711161065, + "grad_norm": 1.6364203691482544, + "learning_rate": 6.31173212235709e-05, + "loss": 1.0894, + "step": 6869 + }, + { + "epoch": 0.4174008141442372, + "grad_norm": 0.30226317048072815, + "learning_rate": 6.310808350845164e-05, + "loss": 1.0878, + "step": 6870 + }, + { + "epoch": 0.4174615711768637, + "grad_norm": 0.6053403615951538, + "learning_rate": 6.309884531285307e-05, + "loss": 1.2685, + "step": 6871 + }, + { + "epoch": 0.41752232820949026, + "grad_norm": 0.37546706199645996, + "learning_rate": 6.308960663711376e-05, + "loss": 1.0683, + "step": 6872 + }, + { + "epoch": 0.4175830852421168, + "grad_norm": 0.3015528619289398, + "learning_rate": 6.30803674815724e-05, + "loss": 1.0267, + "step": 6873 + }, + { + "epoch": 0.4176438422747433, + "grad_norm": 0.2958393692970276, + "learning_rate": 6.307112784656766e-05, + "loss": 1.0618, + "step": 6874 + }, + { + "epoch": 0.4177045993073698, + "grad_norm": 0.44382444024086, + "learning_rate": 6.306188773243819e-05, + "loss": 0.9976, + "step": 6875 + }, + { + "epoch": 0.41776535633999634, + "grad_norm": 1.6360039710998535, + "learning_rate": 6.305264713952269e-05, + "loss": 1.1885, + "step": 6876 + }, + { + "epoch": 0.4178261133726229, + "grad_norm": 0.32199013233184814, + "learning_rate": 6.30434060681599e-05, + "loss": 1.3084, + "step": 6877 + }, + { + "epoch": 0.4178868704052494, + "grad_norm": 0.49860745668411255, + "learning_rate": 6.303416451868853e-05, + "loss": 1.1388, + "step": 6878 + }, + { + "epoch": 0.41794762743787595, + "grad_norm": 0.18165679275989532, + "learning_rate": 6.302492249144736e-05, + "loss": 1.0454, + "step": 6879 + }, + { + "epoch": 0.4180083844705025, + "grad_norm": 0.39735913276672363, + "learning_rate": 6.301567998677512e-05, + "loss": 1.1953, + "step": 6880 + }, + { + "epoch": 0.41806914150312896, + "grad_norm": 0.48428261280059814, + "learning_rate": 6.300643700501061e-05, + "loss": 1.3017, + "step": 6881 + }, + { + "epoch": 0.4181298985357555, + "grad_norm": 0.3692059814929962, + "learning_rate": 6.299719354649266e-05, + "loss": 1.0707, + "step": 6882 + }, + { + "epoch": 0.41819065556838203, + "grad_norm": 0.45567071437835693, + "learning_rate": 6.298794961156004e-05, + "loss": 1.3264, + "step": 6883 + }, + { + "epoch": 0.41825141260100857, + "grad_norm": 0.4696560800075531, + "learning_rate": 6.297870520055163e-05, + "loss": 1.1312, + "step": 6884 + }, + { + "epoch": 0.4183121696336351, + "grad_norm": 0.27619731426239014, + "learning_rate": 6.296946031380628e-05, + "loss": 1.0484, + "step": 6885 + }, + { + "epoch": 0.41837292666626164, + "grad_norm": 0.20989221334457397, + "learning_rate": 6.296021495166284e-05, + "loss": 1.0487, + "step": 6886 + }, + { + "epoch": 0.41843368369888817, + "grad_norm": 0.24997614324092865, + "learning_rate": 6.295096911446023e-05, + "loss": 1.115, + "step": 6887 + }, + { + "epoch": 0.41849444073151465, + "grad_norm": 0.21510380506515503, + "learning_rate": 6.294172280253734e-05, + "loss": 1.091, + "step": 6888 + }, + { + "epoch": 0.4185551977641412, + "grad_norm": 0.22667115926742554, + "learning_rate": 6.29324760162331e-05, + "loss": 1.0943, + "step": 6889 + }, + { + "epoch": 0.4186159547967677, + "grad_norm": 0.28604239225387573, + "learning_rate": 6.292322875588645e-05, + "loss": 1.1824, + "step": 6890 + }, + { + "epoch": 0.41867671182939425, + "grad_norm": 0.23618191480636597, + "learning_rate": 6.291398102183634e-05, + "loss": 1.0638, + "step": 6891 + }, + { + "epoch": 0.4187374688620208, + "grad_norm": 0.3630085289478302, + "learning_rate": 6.290473281442177e-05, + "loss": 1.1466, + "step": 6892 + }, + { + "epoch": 0.4187982258946473, + "grad_norm": 0.19133654236793518, + "learning_rate": 6.289548413398172e-05, + "loss": 1.0326, + "step": 6893 + }, + { + "epoch": 0.41885898292727386, + "grad_norm": 0.3283313512802124, + "learning_rate": 6.288623498085522e-05, + "loss": 1.0748, + "step": 6894 + }, + { + "epoch": 0.41891973995990034, + "grad_norm": 5.2213358879089355, + "learning_rate": 6.287698535538129e-05, + "loss": 1.1173, + "step": 6895 + }, + { + "epoch": 0.4189804969925269, + "grad_norm": 0.49817875027656555, + "learning_rate": 6.286773525789894e-05, + "loss": 1.0804, + "step": 6896 + }, + { + "epoch": 0.4190412540251534, + "grad_norm": 0.39500582218170166, + "learning_rate": 6.285848468874728e-05, + "loss": 1.0316, + "step": 6897 + }, + { + "epoch": 0.41910201105777994, + "grad_norm": 0.4489462375640869, + "learning_rate": 6.28492336482654e-05, + "loss": 1.1196, + "step": 6898 + }, + { + "epoch": 0.4191627680904065, + "grad_norm": 0.7229900360107422, + "learning_rate": 6.283998213679236e-05, + "loss": 1.0607, + "step": 6899 + }, + { + "epoch": 0.419223525123033, + "grad_norm": 0.17782610654830933, + "learning_rate": 6.28307301546673e-05, + "loss": 1.0754, + "step": 6900 + }, + { + "epoch": 0.4192842821556595, + "grad_norm": 0.4473031461238861, + "learning_rate": 6.282147770222932e-05, + "loss": 1.085, + "step": 6901 + }, + { + "epoch": 0.41934503918828603, + "grad_norm": 0.20528604090213776, + "learning_rate": 6.281222477981762e-05, + "loss": 1.0526, + "step": 6902 + }, + { + "epoch": 0.41940579622091256, + "grad_norm": 0.351974755525589, + "learning_rate": 6.280297138777135e-05, + "loss": 1.0832, + "step": 6903 + }, + { + "epoch": 0.4194665532535391, + "grad_norm": 0.23716016113758087, + "learning_rate": 6.279371752642968e-05, + "loss": 1.0459, + "step": 6904 + }, + { + "epoch": 0.41952731028616563, + "grad_norm": 0.16316290199756622, + "learning_rate": 6.278446319613183e-05, + "loss": 1.0964, + "step": 6905 + }, + { + "epoch": 0.41958806731879217, + "grad_norm": 0.17871583998203278, + "learning_rate": 6.2775208397217e-05, + "loss": 1.0196, + "step": 6906 + }, + { + "epoch": 0.4196488243514187, + "grad_norm": 0.2344989776611328, + "learning_rate": 6.276595313002442e-05, + "loss": 1.054, + "step": 6907 + }, + { + "epoch": 0.4197095813840452, + "grad_norm": 0.18178525567054749, + "learning_rate": 6.275669739489339e-05, + "loss": 1.1282, + "step": 6908 + }, + { + "epoch": 0.4197703384166717, + "grad_norm": 0.2146780788898468, + "learning_rate": 6.274744119216314e-05, + "loss": 1.0458, + "step": 6909 + }, + { + "epoch": 0.41983109544929825, + "grad_norm": 0.18630409240722656, + "learning_rate": 6.273818452217295e-05, + "loss": 1.072, + "step": 6910 + }, + { + "epoch": 0.4198918524819248, + "grad_norm": 0.5675243735313416, + "learning_rate": 6.27289273852622e-05, + "loss": 1.1493, + "step": 6911 + }, + { + "epoch": 0.4199526095145513, + "grad_norm": 1.4813159704208374, + "learning_rate": 6.271966978177009e-05, + "loss": 1.0578, + "step": 6912 + }, + { + "epoch": 0.42001336654717786, + "grad_norm": 0.2947629392147064, + "learning_rate": 6.271041171203606e-05, + "loss": 1.043, + "step": 6913 + }, + { + "epoch": 0.42007412357980434, + "grad_norm": 0.253913015127182, + "learning_rate": 6.270115317639943e-05, + "loss": 1.1318, + "step": 6914 + }, + { + "epoch": 0.42013488061243087, + "grad_norm": 0.38708487153053284, + "learning_rate": 6.269189417519957e-05, + "loss": 1.1264, + "step": 6915 + }, + { + "epoch": 0.4201956376450574, + "grad_norm": 0.2900838553905487, + "learning_rate": 6.268263470877587e-05, + "loss": 1.2168, + "step": 6916 + }, + { + "epoch": 0.42025639467768394, + "grad_norm": 0.21828092634677887, + "learning_rate": 6.267337477746776e-05, + "loss": 1.117, + "step": 6917 + }, + { + "epoch": 0.4203171517103105, + "grad_norm": 0.19790278375148773, + "learning_rate": 6.266411438161465e-05, + "loss": 1.0845, + "step": 6918 + }, + { + "epoch": 0.420377908742937, + "grad_norm": 0.25607606768608093, + "learning_rate": 6.265485352155595e-05, + "loss": 1.1479, + "step": 6919 + }, + { + "epoch": 0.42043866577556355, + "grad_norm": 0.2072649747133255, + "learning_rate": 6.264559219763116e-05, + "loss": 1.0726, + "step": 6920 + }, + { + "epoch": 0.42049942280819, + "grad_norm": 1.4104026556015015, + "learning_rate": 6.263633041017975e-05, + "loss": 1.1149, + "step": 6921 + }, + { + "epoch": 0.42056017984081656, + "grad_norm": 0.17129090428352356, + "learning_rate": 6.262706815954121e-05, + "loss": 1.0587, + "step": 6922 + }, + { + "epoch": 0.4206209368734431, + "grad_norm": 0.1692197173833847, + "learning_rate": 6.261780544605504e-05, + "loss": 1.0804, + "step": 6923 + }, + { + "epoch": 0.42068169390606963, + "grad_norm": 0.5294912457466125, + "learning_rate": 6.260854227006079e-05, + "loss": 1.1233, + "step": 6924 + }, + { + "epoch": 0.42074245093869617, + "grad_norm": 0.2499631643295288, + "learning_rate": 6.259927863189796e-05, + "loss": 1.1111, + "step": 6925 + }, + { + "epoch": 0.4208032079713227, + "grad_norm": 0.2553824484348297, + "learning_rate": 6.259001453190615e-05, + "loss": 1.0785, + "step": 6926 + }, + { + "epoch": 0.4208639650039492, + "grad_norm": 0.24110537767410278, + "learning_rate": 6.258074997042493e-05, + "loss": 1.1807, + "step": 6927 + }, + { + "epoch": 0.4209247220365757, + "grad_norm": 0.32907015085220337, + "learning_rate": 6.257148494779387e-05, + "loss": 1.2637, + "step": 6928 + }, + { + "epoch": 0.42098547906920225, + "grad_norm": 0.2085009664297104, + "learning_rate": 6.256221946435262e-05, + "loss": 1.0907, + "step": 6929 + }, + { + "epoch": 0.4210462361018288, + "grad_norm": 0.48711133003234863, + "learning_rate": 6.255295352044075e-05, + "loss": 1.0228, + "step": 6930 + }, + { + "epoch": 0.4211069931344553, + "grad_norm": 0.23919853568077087, + "learning_rate": 6.254368711639797e-05, + "loss": 1.1675, + "step": 6931 + }, + { + "epoch": 0.42116775016708186, + "grad_norm": 0.2253812998533249, + "learning_rate": 6.253442025256392e-05, + "loss": 1.0854, + "step": 6932 + }, + { + "epoch": 0.4212285071997084, + "grad_norm": 3.282294750213623, + "learning_rate": 6.252515292927827e-05, + "loss": 1.1366, + "step": 6933 + }, + { + "epoch": 0.42128926423233487, + "grad_norm": 0.16123583912849426, + "learning_rate": 6.25158851468807e-05, + "loss": 1.0766, + "step": 6934 + }, + { + "epoch": 0.4213500212649614, + "grad_norm": 0.22628316283226013, + "learning_rate": 6.250661690571095e-05, + "loss": 1.0993, + "step": 6935 + }, + { + "epoch": 0.42141077829758794, + "grad_norm": 0.22598251700401306, + "learning_rate": 6.249734820610876e-05, + "loss": 1.0114, + "step": 6936 + }, + { + "epoch": 0.4214715353302145, + "grad_norm": 0.17020723223686218, + "learning_rate": 6.248807904841383e-05, + "loss": 1.0546, + "step": 6937 + }, + { + "epoch": 0.421532292362841, + "grad_norm": 1.1962331533432007, + "learning_rate": 6.247880943296596e-05, + "loss": 1.1146, + "step": 6938 + }, + { + "epoch": 0.42159304939546755, + "grad_norm": 0.21614405512809753, + "learning_rate": 6.24695393601049e-05, + "loss": 1.0311, + "step": 6939 + }, + { + "epoch": 0.421653806428094, + "grad_norm": 0.6102123260498047, + "learning_rate": 6.246026883017048e-05, + "loss": 1.1876, + "step": 6940 + }, + { + "epoch": 0.42171456346072056, + "grad_norm": 0.1711275428533554, + "learning_rate": 6.24509978435025e-05, + "loss": 1.0451, + "step": 6941 + }, + { + "epoch": 0.4217753204933471, + "grad_norm": 0.21674102544784546, + "learning_rate": 6.244172640044081e-05, + "loss": 1.1087, + "step": 6942 + }, + { + "epoch": 0.42183607752597363, + "grad_norm": 0.1944355070590973, + "learning_rate": 6.24324545013252e-05, + "loss": 1.0906, + "step": 6943 + }, + { + "epoch": 0.42189683455860016, + "grad_norm": 0.17115402221679688, + "learning_rate": 6.242318214649557e-05, + "loss": 1.0745, + "step": 6944 + }, + { + "epoch": 0.4219575915912267, + "grad_norm": 0.147663876414299, + "learning_rate": 6.24139093362918e-05, + "loss": 1.0403, + "step": 6945 + }, + { + "epoch": 0.42201834862385323, + "grad_norm": 0.2002735137939453, + "learning_rate": 6.240463607105377e-05, + "loss": 1.0105, + "step": 6946 + }, + { + "epoch": 0.4220791056564797, + "grad_norm": 1.7828419208526611, + "learning_rate": 6.239536235112144e-05, + "loss": 1.0311, + "step": 6947 + }, + { + "epoch": 0.42213986268910625, + "grad_norm": 0.2207784354686737, + "learning_rate": 6.238608817683468e-05, + "loss": 1.075, + "step": 6948 + }, + { + "epoch": 0.4222006197217328, + "grad_norm": 0.43635377287864685, + "learning_rate": 6.237681354853349e-05, + "loss": 1.1691, + "step": 6949 + }, + { + "epoch": 0.4222613767543593, + "grad_norm": 0.27593547105789185, + "learning_rate": 6.236753846655778e-05, + "loss": 1.1987, + "step": 6950 + }, + { + "epoch": 0.42232213378698585, + "grad_norm": 0.223277285695076, + "learning_rate": 6.235826293124756e-05, + "loss": 1.0322, + "step": 6951 + }, + { + "epoch": 0.4223828908196124, + "grad_norm": 0.2753872275352478, + "learning_rate": 6.234898694294284e-05, + "loss": 1.1791, + "step": 6952 + }, + { + "epoch": 0.4224436478522389, + "grad_norm": 0.3902237117290497, + "learning_rate": 6.233971050198361e-05, + "loss": 1.075, + "step": 6953 + }, + { + "epoch": 0.4225044048848654, + "grad_norm": 5.590801239013672, + "learning_rate": 6.23304336087099e-05, + "loss": 1.167, + "step": 6954 + }, + { + "epoch": 0.42256516191749194, + "grad_norm": 0.21328012645244598, + "learning_rate": 6.232115626346178e-05, + "loss": 1.1084, + "step": 6955 + }, + { + "epoch": 0.4226259189501185, + "grad_norm": 0.19147354364395142, + "learning_rate": 6.231187846657927e-05, + "loss": 1.043, + "step": 6956 + }, + { + "epoch": 0.422686675982745, + "grad_norm": 0.18278789520263672, + "learning_rate": 6.23026002184025e-05, + "loss": 1.1318, + "step": 6957 + }, + { + "epoch": 0.42274743301537154, + "grad_norm": 0.3119962215423584, + "learning_rate": 6.229332151927151e-05, + "loss": 1.3811, + "step": 6958 + }, + { + "epoch": 0.4228081900479981, + "grad_norm": 0.1596004217863083, + "learning_rate": 6.228404236952647e-05, + "loss": 1.0336, + "step": 6959 + }, + { + "epoch": 0.42286894708062456, + "grad_norm": 0.17637209594249725, + "learning_rate": 6.227476276950748e-05, + "loss": 1.0416, + "step": 6960 + }, + { + "epoch": 0.4229297041132511, + "grad_norm": 0.3152308166027069, + "learning_rate": 6.226548271955466e-05, + "loss": 1.0221, + "step": 6961 + }, + { + "epoch": 0.4229904611458776, + "grad_norm": 0.2226824015378952, + "learning_rate": 6.225620222000826e-05, + "loss": 1.1271, + "step": 6962 + }, + { + "epoch": 0.42305121817850416, + "grad_norm": 0.28845229744911194, + "learning_rate": 6.224692127120834e-05, + "loss": 1.0389, + "step": 6963 + }, + { + "epoch": 0.4231119752111307, + "grad_norm": 0.8248730897903442, + "learning_rate": 6.223763987349518e-05, + "loss": 1.1454, + "step": 6964 + }, + { + "epoch": 0.42317273224375723, + "grad_norm": 0.2716536819934845, + "learning_rate": 6.222835802720897e-05, + "loss": 1.0442, + "step": 6965 + }, + { + "epoch": 0.42323348927638377, + "grad_norm": 0.2945568859577179, + "learning_rate": 6.22190757326899e-05, + "loss": 1.1755, + "step": 6966 + }, + { + "epoch": 0.42329424630901025, + "grad_norm": 0.2167447805404663, + "learning_rate": 6.220979299027829e-05, + "loss": 1.1861, + "step": 6967 + }, + { + "epoch": 0.4233550033416368, + "grad_norm": 0.4990527927875519, + "learning_rate": 6.220050980031433e-05, + "loss": 1.1438, + "step": 6968 + }, + { + "epoch": 0.4234157603742633, + "grad_norm": 0.44489094614982605, + "learning_rate": 6.219122616313832e-05, + "loss": 1.2035, + "step": 6969 + }, + { + "epoch": 0.42347651740688985, + "grad_norm": 0.2766324579715729, + "learning_rate": 6.218194207909058e-05, + "loss": 1.178, + "step": 6970 + }, + { + "epoch": 0.4235372744395164, + "grad_norm": 0.4277685284614563, + "learning_rate": 6.21726575485114e-05, + "loss": 1.0901, + "step": 6971 + }, + { + "epoch": 0.4235980314721429, + "grad_norm": 0.1649603694677353, + "learning_rate": 6.216337257174109e-05, + "loss": 1.1034, + "step": 6972 + }, + { + "epoch": 0.4236587885047694, + "grad_norm": 0.22138148546218872, + "learning_rate": 6.215408714912001e-05, + "loss": 1.1494, + "step": 6973 + }, + { + "epoch": 0.42371954553739594, + "grad_norm": 0.1960325837135315, + "learning_rate": 6.214480128098851e-05, + "loss": 1.0427, + "step": 6974 + }, + { + "epoch": 0.42378030257002247, + "grad_norm": 0.21071135997772217, + "learning_rate": 6.2135514967687e-05, + "loss": 1.0538, + "step": 6975 + }, + { + "epoch": 0.423841059602649, + "grad_norm": 0.25464069843292236, + "learning_rate": 6.212622820955582e-05, + "loss": 1.1369, + "step": 6976 + }, + { + "epoch": 0.42390181663527554, + "grad_norm": 0.20464183390140533, + "learning_rate": 6.211694100693538e-05, + "loss": 1.113, + "step": 6977 + }, + { + "epoch": 0.4239625736679021, + "grad_norm": 0.24915523827075958, + "learning_rate": 6.210765336016616e-05, + "loss": 1.0567, + "step": 6978 + }, + { + "epoch": 0.4240233307005286, + "grad_norm": 0.4983226954936981, + "learning_rate": 6.209836526958854e-05, + "loss": 1.1191, + "step": 6979 + }, + { + "epoch": 0.4240840877331551, + "grad_norm": 0.18474292755126953, + "learning_rate": 6.208907673554303e-05, + "loss": 1.0233, + "step": 6980 + }, + { + "epoch": 0.4241448447657816, + "grad_norm": 0.29805389046669006, + "learning_rate": 6.207978775837005e-05, + "loss": 1.1866, + "step": 6981 + }, + { + "epoch": 0.42420560179840816, + "grad_norm": 1.2930165529251099, + "learning_rate": 6.207049833841014e-05, + "loss": 1.1336, + "step": 6982 + }, + { + "epoch": 0.4242663588310347, + "grad_norm": 0.21525900065898895, + "learning_rate": 6.206120847600377e-05, + "loss": 1.0741, + "step": 6983 + }, + { + "epoch": 0.42432711586366123, + "grad_norm": 2.705293893814087, + "learning_rate": 6.205191817149145e-05, + "loss": 1.0981, + "step": 6984 + }, + { + "epoch": 0.42438787289628777, + "grad_norm": 0.3428841233253479, + "learning_rate": 6.204262742521377e-05, + "loss": 1.0927, + "step": 6985 + }, + { + "epoch": 0.42444862992891425, + "grad_norm": 0.1760510951280594, + "learning_rate": 6.203333623751124e-05, + "loss": 1.0513, + "step": 6986 + }, + { + "epoch": 0.4245093869615408, + "grad_norm": 0.4959033727645874, + "learning_rate": 6.202404460872444e-05, + "loss": 1.1172, + "step": 6987 + }, + { + "epoch": 0.4245701439941673, + "grad_norm": 0.20579026639461517, + "learning_rate": 6.201475253919397e-05, + "loss": 1.041, + "step": 6988 + }, + { + "epoch": 0.42463090102679385, + "grad_norm": 0.1990087479352951, + "learning_rate": 6.200546002926043e-05, + "loss": 1.0961, + "step": 6989 + }, + { + "epoch": 0.4246916580594204, + "grad_norm": 0.29986920952796936, + "learning_rate": 6.199616707926441e-05, + "loss": 1.1378, + "step": 6990 + }, + { + "epoch": 0.4247524150920469, + "grad_norm": 0.2791280150413513, + "learning_rate": 6.198687368954659e-05, + "loss": 1.1365, + "step": 6991 + }, + { + "epoch": 0.42481317212467345, + "grad_norm": 2.504725933074951, + "learning_rate": 6.197757986044759e-05, + "loss": 1.1038, + "step": 6992 + }, + { + "epoch": 0.42487392915729993, + "grad_norm": 0.20729687809944153, + "learning_rate": 6.196828559230809e-05, + "loss": 1.0985, + "step": 6993 + }, + { + "epoch": 0.42493468618992647, + "grad_norm": 0.2646080553531647, + "learning_rate": 6.195899088546875e-05, + "loss": 1.1176, + "step": 6994 + }, + { + "epoch": 0.424995443222553, + "grad_norm": 0.16805411875247955, + "learning_rate": 6.19496957402703e-05, + "loss": 1.0522, + "step": 6995 + }, + { + "epoch": 0.42505620025517954, + "grad_norm": 3.4537909030914307, + "learning_rate": 6.194040015705344e-05, + "loss": 1.2543, + "step": 6996 + }, + { + "epoch": 0.4251169572878061, + "grad_norm": 0.2683354318141937, + "learning_rate": 6.19311041361589e-05, + "loss": 1.0537, + "step": 6997 + }, + { + "epoch": 0.4251777143204326, + "grad_norm": 0.17208538949489594, + "learning_rate": 6.192180767792743e-05, + "loss": 1.0631, + "step": 6998 + }, + { + "epoch": 0.42523847135305914, + "grad_norm": 0.24395471811294556, + "learning_rate": 6.191251078269981e-05, + "loss": 1.0476, + "step": 6999 + }, + { + "epoch": 0.4252992283856856, + "grad_norm": 0.23021046817302704, + "learning_rate": 6.19032134508168e-05, + "loss": 1.0905, + "step": 7000 + }, + { + "epoch": 0.42535998541831216, + "grad_norm": 0.238621786236763, + "learning_rate": 6.189391568261919e-05, + "loss": 1.1898, + "step": 7001 + }, + { + "epoch": 0.4254207424509387, + "grad_norm": 0.30596378445625305, + "learning_rate": 6.188461747844781e-05, + "loss": 1.0999, + "step": 7002 + }, + { + "epoch": 0.42548149948356523, + "grad_norm": 0.2080690562725067, + "learning_rate": 6.187531883864348e-05, + "loss": 1.14, + "step": 7003 + }, + { + "epoch": 0.42554225651619176, + "grad_norm": 0.18276236951351166, + "learning_rate": 6.186601976354705e-05, + "loss": 1.1003, + "step": 7004 + }, + { + "epoch": 0.4256030135488183, + "grad_norm": 0.2544565498828888, + "learning_rate": 6.185672025349935e-05, + "loss": 1.1126, + "step": 7005 + }, + { + "epoch": 0.4256637705814448, + "grad_norm": 0.24650682508945465, + "learning_rate": 6.18474203088413e-05, + "loss": 1.23, + "step": 7006 + }, + { + "epoch": 0.4257245276140713, + "grad_norm": 0.23842039704322815, + "learning_rate": 6.183811992991375e-05, + "loss": 1.0975, + "step": 7007 + }, + { + "epoch": 0.42578528464669785, + "grad_norm": 1.4173132181167603, + "learning_rate": 6.182881911705762e-05, + "loss": 1.2305, + "step": 7008 + }, + { + "epoch": 0.4258460416793244, + "grad_norm": 0.2906324565410614, + "learning_rate": 6.181951787061388e-05, + "loss": 1.1231, + "step": 7009 + }, + { + "epoch": 0.4259067987119509, + "grad_norm": 0.21424710750579834, + "learning_rate": 6.181021619092339e-05, + "loss": 1.1434, + "step": 7010 + }, + { + "epoch": 0.42596755574457745, + "grad_norm": 0.2213815301656723, + "learning_rate": 6.180091407832715e-05, + "loss": 1.017, + "step": 7011 + }, + { + "epoch": 0.426028312777204, + "grad_norm": 0.17848485708236694, + "learning_rate": 6.179161153316614e-05, + "loss": 1.0263, + "step": 7012 + }, + { + "epoch": 0.42608906980983047, + "grad_norm": 0.23258467018604279, + "learning_rate": 6.17823085557813e-05, + "loss": 1.185, + "step": 7013 + }, + { + "epoch": 0.426149826842457, + "grad_norm": 0.3531310260295868, + "learning_rate": 6.177300514651367e-05, + "loss": 1.2604, + "step": 7014 + }, + { + "epoch": 0.42621058387508354, + "grad_norm": 0.24880331754684448, + "learning_rate": 6.176370130570428e-05, + "loss": 1.1103, + "step": 7015 + }, + { + "epoch": 0.4262713409077101, + "grad_norm": 0.3091300129890442, + "learning_rate": 6.175439703369414e-05, + "loss": 1.1514, + "step": 7016 + }, + { + "epoch": 0.4263320979403366, + "grad_norm": 0.24292035400867462, + "learning_rate": 6.17450923308243e-05, + "loss": 1.0565, + "step": 7017 + }, + { + "epoch": 0.42639285497296314, + "grad_norm": 0.21694618463516235, + "learning_rate": 6.173578719743581e-05, + "loss": 1.0994, + "step": 7018 + }, + { + "epoch": 0.4264536120055896, + "grad_norm": 0.21258015930652618, + "learning_rate": 6.17264816338698e-05, + "loss": 1.0902, + "step": 7019 + }, + { + "epoch": 0.42651436903821616, + "grad_norm": 0.18397225439548492, + "learning_rate": 6.171717564046735e-05, + "loss": 1.0615, + "step": 7020 + }, + { + "epoch": 0.4265751260708427, + "grad_norm": 0.19073963165283203, + "learning_rate": 6.170786921756951e-05, + "loss": 1.0109, + "step": 7021 + }, + { + "epoch": 0.4266358831034692, + "grad_norm": 0.2373439520597458, + "learning_rate": 6.169856236551751e-05, + "loss": 1.1686, + "step": 7022 + }, + { + "epoch": 0.42669664013609576, + "grad_norm": 0.1631343811750412, + "learning_rate": 6.168925508465241e-05, + "loss": 1.0118, + "step": 7023 + }, + { + "epoch": 0.4267573971687223, + "grad_norm": 13.181735038757324, + "learning_rate": 6.167994737531543e-05, + "loss": 1.0682, + "step": 7024 + }, + { + "epoch": 0.42681815420134883, + "grad_norm": 0.21763722598552704, + "learning_rate": 6.167063923784771e-05, + "loss": 1.0181, + "step": 7025 + }, + { + "epoch": 0.4268789112339753, + "grad_norm": 0.23271769285202026, + "learning_rate": 6.166133067259047e-05, + "loss": 1.0706, + "step": 7026 + }, + { + "epoch": 0.42693966826660185, + "grad_norm": 0.2533799111843109, + "learning_rate": 6.165202167988488e-05, + "loss": 1.1143, + "step": 7027 + }, + { + "epoch": 0.4270004252992284, + "grad_norm": 0.22938448190689087, + "learning_rate": 6.16427122600722e-05, + "loss": 1.1542, + "step": 7028 + }, + { + "epoch": 0.4270611823318549, + "grad_norm": 0.8390161395072937, + "learning_rate": 6.163340241349365e-05, + "loss": 1.2637, + "step": 7029 + }, + { + "epoch": 0.42712193936448145, + "grad_norm": 0.256451815366745, + "learning_rate": 6.162409214049047e-05, + "loss": 1.0632, + "step": 7030 + }, + { + "epoch": 0.427182696397108, + "grad_norm": 0.200017049908638, + "learning_rate": 6.161478144140395e-05, + "loss": 1.122, + "step": 7031 + }, + { + "epoch": 0.42724345342973447, + "grad_norm": 0.34692782163619995, + "learning_rate": 6.160547031657538e-05, + "loss": 1.1935, + "step": 7032 + }, + { + "epoch": 0.427304210462361, + "grad_norm": 0.24514400959014893, + "learning_rate": 6.159615876634603e-05, + "loss": 1.1907, + "step": 7033 + }, + { + "epoch": 0.42736496749498754, + "grad_norm": 0.18958242237567902, + "learning_rate": 6.158684679105727e-05, + "loss": 1.0376, + "step": 7034 + }, + { + "epoch": 0.42742572452761407, + "grad_norm": 0.2087637484073639, + "learning_rate": 6.15775343910504e-05, + "loss": 1.0332, + "step": 7035 + }, + { + "epoch": 0.4274864815602406, + "grad_norm": 0.3054821789264679, + "learning_rate": 6.156822156666674e-05, + "loss": 1.1919, + "step": 7036 + }, + { + "epoch": 0.42754723859286714, + "grad_norm": 0.1871086210012436, + "learning_rate": 6.155890831824769e-05, + "loss": 1.0436, + "step": 7037 + }, + { + "epoch": 0.4276079956254937, + "grad_norm": 0.24374210834503174, + "learning_rate": 6.154959464613464e-05, + "loss": 0.9868, + "step": 7038 + }, + { + "epoch": 0.42766875265812015, + "grad_norm": 0.2566034495830536, + "learning_rate": 6.154028055066896e-05, + "loss": 1.0794, + "step": 7039 + }, + { + "epoch": 0.4277295096907467, + "grad_norm": 0.22213801741600037, + "learning_rate": 6.153096603219207e-05, + "loss": 1.1473, + "step": 7040 + }, + { + "epoch": 0.4277902667233732, + "grad_norm": 0.23980431258678436, + "learning_rate": 6.152165109104536e-05, + "loss": 1.1479, + "step": 7041 + }, + { + "epoch": 0.42785102375599976, + "grad_norm": 0.4484161138534546, + "learning_rate": 6.151233572757034e-05, + "loss": 1.1218, + "step": 7042 + }, + { + "epoch": 0.4279117807886263, + "grad_norm": 0.30772411823272705, + "learning_rate": 6.150301994210841e-05, + "loss": 1.2412, + "step": 7043 + }, + { + "epoch": 0.42797253782125283, + "grad_norm": 8.698909759521484, + "learning_rate": 6.149370373500105e-05, + "loss": 1.0452, + "step": 7044 + }, + { + "epoch": 0.4280332948538793, + "grad_norm": 0.23927760124206543, + "learning_rate": 6.14843871065898e-05, + "loss": 1.1999, + "step": 7045 + }, + { + "epoch": 0.42809405188650584, + "grad_norm": 0.3440180718898773, + "learning_rate": 6.147507005721607e-05, + "loss": 1.1653, + "step": 7046 + }, + { + "epoch": 0.4281548089191324, + "grad_norm": 0.3408317565917969, + "learning_rate": 6.146575258722146e-05, + "loss": 1.2005, + "step": 7047 + }, + { + "epoch": 0.4282155659517589, + "grad_norm": 0.26534155011177063, + "learning_rate": 6.145643469694747e-05, + "loss": 1.1075, + "step": 7048 + }, + { + "epoch": 0.42827632298438545, + "grad_norm": 0.2901115119457245, + "learning_rate": 6.144711638673564e-05, + "loss": 1.1558, + "step": 7049 + }, + { + "epoch": 0.428337080017012, + "grad_norm": 0.30954569578170776, + "learning_rate": 6.143779765692756e-05, + "loss": 1.0587, + "step": 7050 + }, + { + "epoch": 0.4283978370496385, + "grad_norm": 0.24421045184135437, + "learning_rate": 6.142847850786477e-05, + "loss": 1.2399, + "step": 7051 + }, + { + "epoch": 0.428458594082265, + "grad_norm": 0.45197248458862305, + "learning_rate": 6.141915893988892e-05, + "loss": 1.1269, + "step": 7052 + }, + { + "epoch": 0.42851935111489153, + "grad_norm": 0.2274293303489685, + "learning_rate": 6.140983895334158e-05, + "loss": 1.102, + "step": 7053 + }, + { + "epoch": 0.42858010814751807, + "grad_norm": 0.2515104115009308, + "learning_rate": 6.140051854856437e-05, + "loss": 1.2256, + "step": 7054 + }, + { + "epoch": 0.4286408651801446, + "grad_norm": 0.4127196669578552, + "learning_rate": 6.139119772589899e-05, + "loss": 1.1181, + "step": 7055 + }, + { + "epoch": 0.42870162221277114, + "grad_norm": 0.6216238737106323, + "learning_rate": 6.138187648568704e-05, + "loss": 1.2906, + "step": 7056 + }, + { + "epoch": 0.4287623792453977, + "grad_norm": 0.34634873270988464, + "learning_rate": 6.137255482827018e-05, + "loss": 1.1571, + "step": 7057 + }, + { + "epoch": 0.4288231362780242, + "grad_norm": 0.29207488894462585, + "learning_rate": 6.136323275399016e-05, + "loss": 1.0384, + "step": 7058 + }, + { + "epoch": 0.4288838933106507, + "grad_norm": 0.3194137215614319, + "learning_rate": 6.135391026318862e-05, + "loss": 1.4005, + "step": 7059 + }, + { + "epoch": 0.4289446503432772, + "grad_norm": 0.32703548669815063, + "learning_rate": 6.134458735620732e-05, + "loss": 1.0091, + "step": 7060 + }, + { + "epoch": 0.42900540737590376, + "grad_norm": 1.7010349035263062, + "learning_rate": 6.133526403338798e-05, + "loss": 1.0535, + "step": 7061 + }, + { + "epoch": 0.4290661644085303, + "grad_norm": 0.21688340604305267, + "learning_rate": 6.132594029507234e-05, + "loss": 1.0874, + "step": 7062 + }, + { + "epoch": 0.42912692144115683, + "grad_norm": 0.30177998542785645, + "learning_rate": 6.131661614160217e-05, + "loss": 1.1022, + "step": 7063 + }, + { + "epoch": 0.42918767847378336, + "grad_norm": 0.1850363314151764, + "learning_rate": 6.130729157331924e-05, + "loss": 1.1559, + "step": 7064 + }, + { + "epoch": 0.42924843550640984, + "grad_norm": 0.2068721354007721, + "learning_rate": 6.129796659056535e-05, + "loss": 1.147, + "step": 7065 + }, + { + "epoch": 0.4293091925390364, + "grad_norm": 0.2177405208349228, + "learning_rate": 6.128864119368234e-05, + "loss": 1.0812, + "step": 7066 + }, + { + "epoch": 0.4293699495716629, + "grad_norm": 0.32686927914619446, + "learning_rate": 6.127931538301198e-05, + "loss": 1.1009, + "step": 7067 + }, + { + "epoch": 0.42943070660428945, + "grad_norm": 0.21827568113803864, + "learning_rate": 6.126998915889612e-05, + "loss": 1.1053, + "step": 7068 + }, + { + "epoch": 0.429491463636916, + "grad_norm": 0.1896175742149353, + "learning_rate": 6.126066252167667e-05, + "loss": 1.0587, + "step": 7069 + }, + { + "epoch": 0.4295522206695425, + "grad_norm": 0.5136411786079407, + "learning_rate": 6.125133547169543e-05, + "loss": 1.2522, + "step": 7070 + }, + { + "epoch": 0.42961297770216905, + "grad_norm": 0.15417514741420746, + "learning_rate": 6.124200800929431e-05, + "loss": 1.0681, + "step": 7071 + }, + { + "epoch": 0.42967373473479553, + "grad_norm": 0.30215397477149963, + "learning_rate": 6.123268013481523e-05, + "loss": 1.0652, + "step": 7072 + }, + { + "epoch": 0.42973449176742207, + "grad_norm": 0.181355819106102, + "learning_rate": 6.122335184860008e-05, + "loss": 1.0866, + "step": 7073 + }, + { + "epoch": 0.4297952488000486, + "grad_norm": 0.21743904054164886, + "learning_rate": 6.121402315099081e-05, + "loss": 1.1242, + "step": 7074 + }, + { + "epoch": 0.42985600583267514, + "grad_norm": 0.25111785531044006, + "learning_rate": 6.120469404232932e-05, + "loss": 1.0344, + "step": 7075 + }, + { + "epoch": 0.42991676286530167, + "grad_norm": 0.1931886076927185, + "learning_rate": 6.119536452295765e-05, + "loss": 1.132, + "step": 7076 + }, + { + "epoch": 0.4299775198979282, + "grad_norm": 0.37709370255470276, + "learning_rate": 6.118603459321772e-05, + "loss": 1.1562, + "step": 7077 + }, + { + "epoch": 0.4300382769305547, + "grad_norm": 0.2053460329771042, + "learning_rate": 6.11767042534515e-05, + "loss": 1.0189, + "step": 7078 + }, + { + "epoch": 0.4300990339631812, + "grad_norm": 0.18062633275985718, + "learning_rate": 6.116737350400105e-05, + "loss": 1.038, + "step": 7079 + }, + { + "epoch": 0.43015979099580776, + "grad_norm": 0.2100735753774643, + "learning_rate": 6.115804234520834e-05, + "loss": 1.0255, + "step": 7080 + }, + { + "epoch": 0.4302205480284343, + "grad_norm": 0.15320998430252075, + "learning_rate": 6.114871077741546e-05, + "loss": 1.0471, + "step": 7081 + }, + { + "epoch": 0.4302813050610608, + "grad_norm": 0.17768436670303345, + "learning_rate": 6.113937880096442e-05, + "loss": 1.1092, + "step": 7082 + }, + { + "epoch": 0.43034206209368736, + "grad_norm": 0.18748368322849274, + "learning_rate": 6.11300464161973e-05, + "loss": 1.0546, + "step": 7083 + }, + { + "epoch": 0.4304028191263139, + "grad_norm": 0.319961279630661, + "learning_rate": 6.112071362345617e-05, + "loss": 1.1565, + "step": 7084 + }, + { + "epoch": 0.4304635761589404, + "grad_norm": 0.30719900131225586, + "learning_rate": 6.111138042308314e-05, + "loss": 1.0294, + "step": 7085 + }, + { + "epoch": 0.4305243331915669, + "grad_norm": 0.2661246657371521, + "learning_rate": 6.110204681542031e-05, + "loss": 1.0586, + "step": 7086 + }, + { + "epoch": 0.43058509022419345, + "grad_norm": 0.461480051279068, + "learning_rate": 6.109271280080981e-05, + "loss": 1.3065, + "step": 7087 + }, + { + "epoch": 0.43064584725682, + "grad_norm": 0.23279330134391785, + "learning_rate": 6.108337837959377e-05, + "loss": 1.117, + "step": 7088 + }, + { + "epoch": 0.4307066042894465, + "grad_norm": 0.1974182426929474, + "learning_rate": 6.107404355211437e-05, + "loss": 1.0772, + "step": 7089 + }, + { + "epoch": 0.43076736132207305, + "grad_norm": 0.1773347705602646, + "learning_rate": 6.106470831871374e-05, + "loss": 1.0802, + "step": 7090 + }, + { + "epoch": 0.43082811835469953, + "grad_norm": 0.26244738698005676, + "learning_rate": 6.10553726797341e-05, + "loss": 1.0751, + "step": 7091 + }, + { + "epoch": 0.43088887538732606, + "grad_norm": 0.3744717240333557, + "learning_rate": 6.104603663551764e-05, + "loss": 1.1258, + "step": 7092 + }, + { + "epoch": 0.4309496324199526, + "grad_norm": 0.16945628821849823, + "learning_rate": 6.103670018640657e-05, + "loss": 1.0804, + "step": 7093 + }, + { + "epoch": 0.43101038945257913, + "grad_norm": 0.22216671705245972, + "learning_rate": 6.102736333274313e-05, + "loss": 1.1903, + "step": 7094 + }, + { + "epoch": 0.43107114648520567, + "grad_norm": 0.21010053157806396, + "learning_rate": 6.1018026074869536e-05, + "loss": 1.1549, + "step": 7095 + }, + { + "epoch": 0.4311319035178322, + "grad_norm": 0.27313464879989624, + "learning_rate": 6.100868841312808e-05, + "loss": 1.0819, + "step": 7096 + }, + { + "epoch": 0.43119266055045874, + "grad_norm": 0.6617570519447327, + "learning_rate": 6.0999350347861016e-05, + "loss": 1.0489, + "step": 7097 + }, + { + "epoch": 0.4312534175830852, + "grad_norm": 0.18625076115131378, + "learning_rate": 6.099001187941063e-05, + "loss": 1.123, + "step": 7098 + }, + { + "epoch": 0.43131417461571175, + "grad_norm": 0.2985597252845764, + "learning_rate": 6.098067300811927e-05, + "loss": 1.1781, + "step": 7099 + }, + { + "epoch": 0.4313749316483383, + "grad_norm": 0.21250446140766144, + "learning_rate": 6.09713337343292e-05, + "loss": 1.1662, + "step": 7100 + }, + { + "epoch": 0.4314356886809648, + "grad_norm": 0.17271845042705536, + "learning_rate": 6.096199405838274e-05, + "loss": 1.0463, + "step": 7101 + }, + { + "epoch": 0.43149644571359136, + "grad_norm": 0.21951226890087128, + "learning_rate": 6.095265398062231e-05, + "loss": 1.0445, + "step": 7102 + }, + { + "epoch": 0.4315572027462179, + "grad_norm": 0.1777946949005127, + "learning_rate": 6.094331350139021e-05, + "loss": 1.0219, + "step": 7103 + }, + { + "epoch": 0.43161795977884443, + "grad_norm": 0.275790810585022, + "learning_rate": 6.093397262102884e-05, + "loss": 1.0747, + "step": 7104 + }, + { + "epoch": 0.4316787168114709, + "grad_norm": 0.44109880924224854, + "learning_rate": 6.09246313398806e-05, + "loss": 1.1365, + "step": 7105 + }, + { + "epoch": 0.43173947384409744, + "grad_norm": 0.2122659683227539, + "learning_rate": 6.0915289658287877e-05, + "loss": 1.0851, + "step": 7106 + }, + { + "epoch": 0.431800230876724, + "grad_norm": 1.0892996788024902, + "learning_rate": 6.090594757659311e-05, + "loss": 1.0409, + "step": 7107 + }, + { + "epoch": 0.4318609879093505, + "grad_norm": 0.27642083168029785, + "learning_rate": 6.089660509513872e-05, + "loss": 1.0163, + "step": 7108 + }, + { + "epoch": 0.43192174494197705, + "grad_norm": 0.19626495242118835, + "learning_rate": 6.0887262214267174e-05, + "loss": 1.0738, + "step": 7109 + }, + { + "epoch": 0.4319825019746036, + "grad_norm": 0.3603537678718567, + "learning_rate": 6.0877918934320924e-05, + "loss": 1.0888, + "step": 7110 + }, + { + "epoch": 0.43204325900723006, + "grad_norm": 0.38488778471946716, + "learning_rate": 6.086857525564244e-05, + "loss": 0.9978, + "step": 7111 + }, + { + "epoch": 0.4321040160398566, + "grad_norm": 0.31866276264190674, + "learning_rate": 6.085923117857424e-05, + "loss": 1.0383, + "step": 7112 + }, + { + "epoch": 0.43216477307248313, + "grad_norm": 0.2901649475097656, + "learning_rate": 6.0849886703458826e-05, + "loss": 1.2359, + "step": 7113 + }, + { + "epoch": 0.43222553010510967, + "grad_norm": 0.2501998245716095, + "learning_rate": 6.08405418306387e-05, + "loss": 1.0623, + "step": 7114 + }, + { + "epoch": 0.4322862871377362, + "grad_norm": 0.2938055694103241, + "learning_rate": 6.083119656045643e-05, + "loss": 1.1897, + "step": 7115 + }, + { + "epoch": 0.43234704417036274, + "grad_norm": 0.39185917377471924, + "learning_rate": 6.082185089325455e-05, + "loss": 1.1501, + "step": 7116 + }, + { + "epoch": 0.4324078012029893, + "grad_norm": 0.5710768699645996, + "learning_rate": 6.0812504829375636e-05, + "loss": 1.1061, + "step": 7117 + }, + { + "epoch": 0.43246855823561575, + "grad_norm": 0.2557678520679474, + "learning_rate": 6.080315836916228e-05, + "loss": 1.0722, + "step": 7118 + }, + { + "epoch": 0.4325293152682423, + "grad_norm": 0.2071082592010498, + "learning_rate": 6.079381151295706e-05, + "loss": 1.0622, + "step": 7119 + }, + { + "epoch": 0.4325900723008688, + "grad_norm": 0.30488649010658264, + "learning_rate": 6.0784464261102584e-05, + "loss": 1.0478, + "step": 7120 + }, + { + "epoch": 0.43265082933349536, + "grad_norm": 0.23083153367042542, + "learning_rate": 6.07751166139415e-05, + "loss": 1.1327, + "step": 7121 + }, + { + "epoch": 0.4327115863661219, + "grad_norm": 0.18727731704711914, + "learning_rate": 6.076576857181642e-05, + "loss": 1.0537, + "step": 7122 + }, + { + "epoch": 0.4327723433987484, + "grad_norm": 0.5068754553794861, + "learning_rate": 6.0756420135070016e-05, + "loss": 1.1071, + "step": 7123 + }, + { + "epoch": 0.4328331004313749, + "grad_norm": 0.1897944062948227, + "learning_rate": 6.0747071304044943e-05, + "loss": 1.0807, + "step": 7124 + }, + { + "epoch": 0.43289385746400144, + "grad_norm": 0.31676748394966125, + "learning_rate": 6.07377220790839e-05, + "loss": 1.1074, + "step": 7125 + }, + { + "epoch": 0.432954614496628, + "grad_norm": 0.2479688823223114, + "learning_rate": 6.0728372460529594e-05, + "loss": 1.1648, + "step": 7126 + }, + { + "epoch": 0.4330153715292545, + "grad_norm": 0.23466546833515167, + "learning_rate": 6.07190224487247e-05, + "loss": 1.1554, + "step": 7127 + }, + { + "epoch": 0.43307612856188105, + "grad_norm": 0.24507245421409607, + "learning_rate": 6.070967204401199e-05, + "loss": 1.1184, + "step": 7128 + }, + { + "epoch": 0.4331368855945076, + "grad_norm": 0.7541384100914001, + "learning_rate": 6.070032124673416e-05, + "loss": 1.1483, + "step": 7129 + }, + { + "epoch": 0.4331976426271341, + "grad_norm": 0.3497978448867798, + "learning_rate": 6.0690970057234e-05, + "loss": 1.0262, + "step": 7130 + }, + { + "epoch": 0.4332583996597606, + "grad_norm": 0.21554039418697357, + "learning_rate": 6.068161847585426e-05, + "loss": 1.0548, + "step": 7131 + }, + { + "epoch": 0.43331915669238713, + "grad_norm": 0.32862359285354614, + "learning_rate": 6.067226650293774e-05, + "loss": 1.1009, + "step": 7132 + }, + { + "epoch": 0.43337991372501367, + "grad_norm": 0.42050445079803467, + "learning_rate": 6.066291413882721e-05, + "loss": 1.0418, + "step": 7133 + }, + { + "epoch": 0.4334406707576402, + "grad_norm": 0.18236806988716125, + "learning_rate": 6.06535613838655e-05, + "loss": 1.1199, + "step": 7134 + }, + { + "epoch": 0.43350142779026674, + "grad_norm": 0.20139387249946594, + "learning_rate": 6.064420823839545e-05, + "loss": 1.0995, + "step": 7135 + }, + { + "epoch": 0.43356218482289327, + "grad_norm": 0.3556334674358368, + "learning_rate": 6.0634854702759904e-05, + "loss": 1.1444, + "step": 7136 + }, + { + "epoch": 0.43362294185551975, + "grad_norm": 0.23456314206123352, + "learning_rate": 6.062550077730167e-05, + "loss": 1.0669, + "step": 7137 + }, + { + "epoch": 0.4336836988881463, + "grad_norm": 0.24361416697502136, + "learning_rate": 6.0616146462363664e-05, + "loss": 1.0452, + "step": 7138 + }, + { + "epoch": 0.4337444559207728, + "grad_norm": 0.22430114448070526, + "learning_rate": 6.060679175828877e-05, + "loss": 1.0871, + "step": 7139 + }, + { + "epoch": 0.43380521295339936, + "grad_norm": 0.1967409998178482, + "learning_rate": 6.059743666541986e-05, + "loss": 1.1067, + "step": 7140 + }, + { + "epoch": 0.4338659699860259, + "grad_norm": 0.25552281737327576, + "learning_rate": 6.0588081184099865e-05, + "loss": 1.1189, + "step": 7141 + }, + { + "epoch": 0.4339267270186524, + "grad_norm": 0.19199378788471222, + "learning_rate": 6.05787253146717e-05, + "loss": 1.1522, + "step": 7142 + }, + { + "epoch": 0.43398748405127896, + "grad_norm": 0.21110743284225464, + "learning_rate": 6.056936905747834e-05, + "loss": 1.0959, + "step": 7143 + }, + { + "epoch": 0.43404824108390544, + "grad_norm": 0.1889297068119049, + "learning_rate": 6.05600124128627e-05, + "loss": 1.053, + "step": 7144 + }, + { + "epoch": 0.434108998116532, + "grad_norm": 0.22505521774291992, + "learning_rate": 6.055065538116774e-05, + "loss": 1.1633, + "step": 7145 + }, + { + "epoch": 0.4341697551491585, + "grad_norm": 0.5630159974098206, + "learning_rate": 6.054129796273651e-05, + "loss": 1.0809, + "step": 7146 + }, + { + "epoch": 0.43423051218178504, + "grad_norm": 0.24188190698623657, + "learning_rate": 6.0531940157911924e-05, + "loss": 1.128, + "step": 7147 + }, + { + "epoch": 0.4342912692144116, + "grad_norm": 0.20126736164093018, + "learning_rate": 6.052258196703706e-05, + "loss": 1.0624, + "step": 7148 + }, + { + "epoch": 0.4343520262470381, + "grad_norm": 0.5249847173690796, + "learning_rate": 6.051322339045493e-05, + "loss": 1.2799, + "step": 7149 + }, + { + "epoch": 0.4344127832796646, + "grad_norm": 0.21452663838863373, + "learning_rate": 6.050386442850854e-05, + "loss": 1.0795, + "step": 7150 + }, + { + "epoch": 0.43447354031229113, + "grad_norm": 0.36117103695869446, + "learning_rate": 6.049450508154098e-05, + "loss": 1.0423, + "step": 7151 + }, + { + "epoch": 0.43453429734491766, + "grad_norm": 0.2547288239002228, + "learning_rate": 6.04851453498953e-05, + "loss": 1.1726, + "step": 7152 + }, + { + "epoch": 0.4345950543775442, + "grad_norm": 0.2436836063861847, + "learning_rate": 6.047578523391459e-05, + "loss": 1.1537, + "step": 7153 + }, + { + "epoch": 0.43465581141017073, + "grad_norm": 0.25048500299453735, + "learning_rate": 6.046642473394196e-05, + "loss": 1.1834, + "step": 7154 + }, + { + "epoch": 0.43471656844279727, + "grad_norm": 0.25666344165802, + "learning_rate": 6.045706385032049e-05, + "loss": 1.2689, + "step": 7155 + }, + { + "epoch": 0.4347773254754238, + "grad_norm": 0.19280289113521576, + "learning_rate": 6.044770258339335e-05, + "loss": 1.0907, + "step": 7156 + }, + { + "epoch": 0.4348380825080503, + "grad_norm": 0.23388700187206268, + "learning_rate": 6.0438340933503644e-05, + "loss": 1.0544, + "step": 7157 + }, + { + "epoch": 0.4348988395406768, + "grad_norm": 0.14203333854675293, + "learning_rate": 6.0428978900994516e-05, + "loss": 1.0147, + "step": 7158 + }, + { + "epoch": 0.43495959657330335, + "grad_norm": 0.1574932336807251, + "learning_rate": 6.0419616486209174e-05, + "loss": 1.0008, + "step": 7159 + }, + { + "epoch": 0.4350203536059299, + "grad_norm": 0.20448267459869385, + "learning_rate": 6.041025368949076e-05, + "loss": 1.0016, + "step": 7160 + }, + { + "epoch": 0.4350811106385564, + "grad_norm": 0.1934046596288681, + "learning_rate": 6.0400890511182485e-05, + "loss": 1.1076, + "step": 7161 + }, + { + "epoch": 0.43514186767118296, + "grad_norm": 0.5857439041137695, + "learning_rate": 6.039152695162756e-05, + "loss": 1.076, + "step": 7162 + }, + { + "epoch": 0.4352026247038095, + "grad_norm": 0.3041735887527466, + "learning_rate": 6.038216301116921e-05, + "loss": 1.2234, + "step": 7163 + }, + { + "epoch": 0.435263381736436, + "grad_norm": 0.18242640793323517, + "learning_rate": 6.037279869015067e-05, + "loss": 1.0702, + "step": 7164 + }, + { + "epoch": 0.4353241387690625, + "grad_norm": 3.94091796875, + "learning_rate": 6.0363433988915176e-05, + "loss": 1.0258, + "step": 7165 + }, + { + "epoch": 0.43538489580168904, + "grad_norm": 0.5485088229179382, + "learning_rate": 6.0354068907806014e-05, + "loss": 1.0211, + "step": 7166 + }, + { + "epoch": 0.4354456528343156, + "grad_norm": 7.205066680908203, + "learning_rate": 6.034470344716646e-05, + "loss": 1.077, + "step": 7167 + }, + { + "epoch": 0.4355064098669421, + "grad_norm": 0.7411412596702576, + "learning_rate": 6.033533760733978e-05, + "loss": 1.0501, + "step": 7168 + }, + { + "epoch": 0.43556716689956865, + "grad_norm": 0.3527250289916992, + "learning_rate": 6.0325971388669334e-05, + "loss": 1.136, + "step": 7169 + }, + { + "epoch": 0.4356279239321951, + "grad_norm": 0.44140979647636414, + "learning_rate": 6.03166047914984e-05, + "loss": 1.0466, + "step": 7170 + }, + { + "epoch": 0.43568868096482166, + "grad_norm": 0.3918663561344147, + "learning_rate": 6.0307237816170304e-05, + "loss": 1.1146, + "step": 7171 + }, + { + "epoch": 0.4357494379974482, + "grad_norm": 0.9930682182312012, + "learning_rate": 6.029787046302843e-05, + "loss": 1.3, + "step": 7172 + }, + { + "epoch": 0.43581019503007473, + "grad_norm": 0.3567764461040497, + "learning_rate": 6.028850273241612e-05, + "loss": 1.0644, + "step": 7173 + }, + { + "epoch": 0.43587095206270127, + "grad_norm": 0.39049032330513, + "learning_rate": 6.027913462467676e-05, + "loss": 1.0485, + "step": 7174 + }, + { + "epoch": 0.4359317090953278, + "grad_norm": 0.18666154146194458, + "learning_rate": 6.026976614015373e-05, + "loss": 1.0616, + "step": 7175 + }, + { + "epoch": 0.43599246612795434, + "grad_norm": 0.7160224914550781, + "learning_rate": 6.0260397279190437e-05, + "loss": 1.1421, + "step": 7176 + }, + { + "epoch": 0.4360532231605808, + "grad_norm": 6.1905083656311035, + "learning_rate": 6.02510280421303e-05, + "loss": 1.2278, + "step": 7177 + }, + { + "epoch": 0.43611398019320735, + "grad_norm": 0.46130645275115967, + "learning_rate": 6.0241658429316747e-05, + "loss": 1.1489, + "step": 7178 + }, + { + "epoch": 0.4361747372258339, + "grad_norm": 0.2688387334346771, + "learning_rate": 6.0232288441093224e-05, + "loss": 1.1051, + "step": 7179 + }, + { + "epoch": 0.4362354942584604, + "grad_norm": 0.4547428786754608, + "learning_rate": 6.022291807780319e-05, + "loss": 1.0823, + "step": 7180 + }, + { + "epoch": 0.43629625129108696, + "grad_norm": 0.38336896896362305, + "learning_rate": 6.0213547339790097e-05, + "loss": 1.0559, + "step": 7181 + }, + { + "epoch": 0.4363570083237135, + "grad_norm": 0.1940966695547104, + "learning_rate": 6.020417622739748e-05, + "loss": 1.0875, + "step": 7182 + }, + { + "epoch": 0.43641776535633997, + "grad_norm": 0.309395968914032, + "learning_rate": 6.01948047409688e-05, + "loss": 1.1861, + "step": 7183 + }, + { + "epoch": 0.4364785223889665, + "grad_norm": 0.6491801142692566, + "learning_rate": 6.018543288084759e-05, + "loss": 1.2353, + "step": 7184 + }, + { + "epoch": 0.43653927942159304, + "grad_norm": 0.7617549300193787, + "learning_rate": 6.0176060647377367e-05, + "loss": 1.0836, + "step": 7185 + }, + { + "epoch": 0.4366000364542196, + "grad_norm": 0.9441420435905457, + "learning_rate": 6.016668804090168e-05, + "loss": 1.0852, + "step": 7186 + }, + { + "epoch": 0.4366607934868461, + "grad_norm": 0.41086944937705994, + "learning_rate": 6.0157315061764066e-05, + "loss": 1.2826, + "step": 7187 + }, + { + "epoch": 0.43672155051947265, + "grad_norm": 0.2185354083776474, + "learning_rate": 6.014794171030811e-05, + "loss": 1.1008, + "step": 7188 + }, + { + "epoch": 0.4367823075520992, + "grad_norm": 0.6336871385574341, + "learning_rate": 6.013856798687738e-05, + "loss": 1.0553, + "step": 7189 + }, + { + "epoch": 0.43684306458472566, + "grad_norm": 0.2452002763748169, + "learning_rate": 6.012919389181549e-05, + "loss": 1.0461, + "step": 7190 + }, + { + "epoch": 0.4369038216173522, + "grad_norm": 0.244609534740448, + "learning_rate": 6.011981942546603e-05, + "loss": 1.1824, + "step": 7191 + }, + { + "epoch": 0.43696457864997873, + "grad_norm": 0.18046297132968903, + "learning_rate": 6.0110444588172645e-05, + "loss": 1.0984, + "step": 7192 + }, + { + "epoch": 0.43702533568260526, + "grad_norm": 0.31075358390808105, + "learning_rate": 6.010106938027896e-05, + "loss": 1.0725, + "step": 7193 + }, + { + "epoch": 0.4370860927152318, + "grad_norm": 0.22597366571426392, + "learning_rate": 6.0091693802128614e-05, + "loss": 1.083, + "step": 7194 + }, + { + "epoch": 0.43714684974785833, + "grad_norm": 0.48745831847190857, + "learning_rate": 6.0082317854065285e-05, + "loss": 1.1966, + "step": 7195 + }, + { + "epoch": 0.4372076067804848, + "grad_norm": 0.300906240940094, + "learning_rate": 6.0072941536432644e-05, + "loss": 1.1221, + "step": 7196 + }, + { + "epoch": 0.43726836381311135, + "grad_norm": 0.9893830418586731, + "learning_rate": 6.0063564849574385e-05, + "loss": 1.1042, + "step": 7197 + }, + { + "epoch": 0.4373291208457379, + "grad_norm": 1.2854351997375488, + "learning_rate": 6.005418779383421e-05, + "loss": 1.0536, + "step": 7198 + }, + { + "epoch": 0.4373898778783644, + "grad_norm": 0.25542309880256653, + "learning_rate": 6.0044810369555846e-05, + "loss": 1.1166, + "step": 7199 + }, + { + "epoch": 0.43745063491099095, + "grad_norm": 0.16165263950824738, + "learning_rate": 6.003543257708302e-05, + "loss": 1.0141, + "step": 7200 + }, + { + "epoch": 0.4375113919436175, + "grad_norm": 0.17820800840854645, + "learning_rate": 6.002605441675946e-05, + "loss": 1.029, + "step": 7201 + }, + { + "epoch": 0.437572148976244, + "grad_norm": 0.30711546540260315, + "learning_rate": 6.001667588892893e-05, + "loss": 1.1343, + "step": 7202 + }, + { + "epoch": 0.4376329060088705, + "grad_norm": 0.15225905179977417, + "learning_rate": 6.000729699393523e-05, + "loss": 1.0488, + "step": 7203 + }, + { + "epoch": 0.43769366304149704, + "grad_norm": 0.20828142762184143, + "learning_rate": 5.999791773212211e-05, + "loss": 1.1039, + "step": 7204 + }, + { + "epoch": 0.4377544200741236, + "grad_norm": 0.16067703068256378, + "learning_rate": 5.998853810383339e-05, + "loss": 0.9796, + "step": 7205 + }, + { + "epoch": 0.4378151771067501, + "grad_norm": 0.8489102125167847, + "learning_rate": 5.9979158109412894e-05, + "loss": 1.1576, + "step": 7206 + }, + { + "epoch": 0.43787593413937664, + "grad_norm": 0.15138572454452515, + "learning_rate": 5.996977774920439e-05, + "loss": 1.0861, + "step": 7207 + }, + { + "epoch": 0.4379366911720032, + "grad_norm": 0.8432535529136658, + "learning_rate": 5.996039702355177e-05, + "loss": 1.1289, + "step": 7208 + }, + { + "epoch": 0.4379974482046297, + "grad_norm": 0.5914003252983093, + "learning_rate": 5.995101593279888e-05, + "loss": 1.0423, + "step": 7209 + }, + { + "epoch": 0.4380582052372562, + "grad_norm": 0.3116024136543274, + "learning_rate": 5.9941634477289576e-05, + "loss": 1.1056, + "step": 7210 + }, + { + "epoch": 0.43811896226988273, + "grad_norm": 0.1728050857782364, + "learning_rate": 5.993225265736774e-05, + "loss": 1.1182, + "step": 7211 + }, + { + "epoch": 0.43817971930250926, + "grad_norm": 0.4064423441886902, + "learning_rate": 5.9922870473377244e-05, + "loss": 1.0148, + "step": 7212 + }, + { + "epoch": 0.4382404763351358, + "grad_norm": 0.3010304272174835, + "learning_rate": 5.991348792566204e-05, + "loss": 1.028, + "step": 7213 + }, + { + "epoch": 0.43830123336776233, + "grad_norm": 0.413234680891037, + "learning_rate": 5.9904105014566e-05, + "loss": 1.1938, + "step": 7214 + }, + { + "epoch": 0.43836199040038887, + "grad_norm": 0.4640549123287201, + "learning_rate": 5.989472174043308e-05, + "loss": 1.0765, + "step": 7215 + }, + { + "epoch": 0.43842274743301535, + "grad_norm": 0.41762325167655945, + "learning_rate": 5.988533810360724e-05, + "loss": 1.053, + "step": 7216 + }, + { + "epoch": 0.4384835044656419, + "grad_norm": 0.544173538684845, + "learning_rate": 5.9875954104432394e-05, + "loss": 1.1147, + "step": 7217 + }, + { + "epoch": 0.4385442614982684, + "grad_norm": 0.3211129605770111, + "learning_rate": 5.9866569743252555e-05, + "loss": 1.1055, + "step": 7218 + }, + { + "epoch": 0.43860501853089495, + "grad_norm": 0.3134716749191284, + "learning_rate": 5.98571850204117e-05, + "loss": 1.0485, + "step": 7219 + }, + { + "epoch": 0.4386657755635215, + "grad_norm": 0.3417632281780243, + "learning_rate": 5.984779993625381e-05, + "loss": 1.0242, + "step": 7220 + }, + { + "epoch": 0.438726532596148, + "grad_norm": 0.3981088697910309, + "learning_rate": 5.983841449112292e-05, + "loss": 1.0642, + "step": 7221 + }, + { + "epoch": 0.43878728962877456, + "grad_norm": 0.1976952701807022, + "learning_rate": 5.9829028685363044e-05, + "loss": 1.0538, + "step": 7222 + }, + { + "epoch": 0.43884804666140104, + "grad_norm": 0.2523045837879181, + "learning_rate": 5.981964251931822e-05, + "loss": 1.0796, + "step": 7223 + }, + { + "epoch": 0.43890880369402757, + "grad_norm": 0.36842867732048035, + "learning_rate": 5.98102559933325e-05, + "loss": 1.133, + "step": 7224 + }, + { + "epoch": 0.4389695607266541, + "grad_norm": 0.3475496768951416, + "learning_rate": 5.980086910774994e-05, + "loss": 1.1207, + "step": 7225 + }, + { + "epoch": 0.43903031775928064, + "grad_norm": 0.26218128204345703, + "learning_rate": 5.9791481862914645e-05, + "loss": 1.0942, + "step": 7226 + }, + { + "epoch": 0.4390910747919072, + "grad_norm": 0.3836635649204254, + "learning_rate": 5.978209425917067e-05, + "loss": 1.2107, + "step": 7227 + }, + { + "epoch": 0.4391518318245337, + "grad_norm": 0.24452392756938934, + "learning_rate": 5.977270629686216e-05, + "loss": 1.0993, + "step": 7228 + }, + { + "epoch": 0.4392125888571602, + "grad_norm": 0.3940485715866089, + "learning_rate": 5.976331797633321e-05, + "loss": 1.0852, + "step": 7229 + }, + { + "epoch": 0.4392733458897867, + "grad_norm": 0.20935490727424622, + "learning_rate": 5.9753929297927945e-05, + "loss": 1.1085, + "step": 7230 + }, + { + "epoch": 0.43933410292241326, + "grad_norm": 0.6556986570358276, + "learning_rate": 5.974454026199051e-05, + "loss": 1.1047, + "step": 7231 + }, + { + "epoch": 0.4393948599550398, + "grad_norm": 0.24542610347270966, + "learning_rate": 5.9735150868865086e-05, + "loss": 1.1351, + "step": 7232 + }, + { + "epoch": 0.43945561698766633, + "grad_norm": 0.3383590579032898, + "learning_rate": 5.9725761118895816e-05, + "loss": 1.1119, + "step": 7233 + }, + { + "epoch": 0.43951637402029287, + "grad_norm": 0.19016574323177338, + "learning_rate": 5.9716371012426885e-05, + "loss": 1.0846, + "step": 7234 + }, + { + "epoch": 0.4395771310529194, + "grad_norm": 0.5306042432785034, + "learning_rate": 5.9706980549802494e-05, + "loss": 1.2283, + "step": 7235 + }, + { + "epoch": 0.4396378880855459, + "grad_norm": 0.2208300679922104, + "learning_rate": 5.969758973136688e-05, + "loss": 1.1858, + "step": 7236 + }, + { + "epoch": 0.4396986451181724, + "grad_norm": 0.16660301387310028, + "learning_rate": 5.968819855746423e-05, + "loss": 1.0528, + "step": 7237 + }, + { + "epoch": 0.43975940215079895, + "grad_norm": 0.27683889865875244, + "learning_rate": 5.9678807028438766e-05, + "loss": 1.1296, + "step": 7238 + }, + { + "epoch": 0.4398201591834255, + "grad_norm": 0.334265798330307, + "learning_rate": 5.9669415144634775e-05, + "loss": 1.0643, + "step": 7239 + }, + { + "epoch": 0.439880916216052, + "grad_norm": 0.18286532163619995, + "learning_rate": 5.966002290639651e-05, + "loss": 1.0524, + "step": 7240 + }, + { + "epoch": 0.43994167324867856, + "grad_norm": 0.24221351742744446, + "learning_rate": 5.9650630314068226e-05, + "loss": 1.0941, + "step": 7241 + }, + { + "epoch": 0.44000243028130503, + "grad_norm": 0.38189131021499634, + "learning_rate": 5.964123736799423e-05, + "loss": 1.0639, + "step": 7242 + }, + { + "epoch": 0.44006318731393157, + "grad_norm": 0.4260561466217041, + "learning_rate": 5.9631844068518813e-05, + "loss": 1.1941, + "step": 7243 + }, + { + "epoch": 0.4401239443465581, + "grad_norm": 0.4665431082248688, + "learning_rate": 5.962245041598629e-05, + "loss": 1.0958, + "step": 7244 + }, + { + "epoch": 0.44018470137918464, + "grad_norm": 0.29358017444610596, + "learning_rate": 5.961305641074099e-05, + "loss": 1.1509, + "step": 7245 + }, + { + "epoch": 0.4402454584118112, + "grad_norm": 0.2900998890399933, + "learning_rate": 5.960366205312725e-05, + "loss": 1.0892, + "step": 7246 + }, + { + "epoch": 0.4403062154444377, + "grad_norm": 0.33606261014938354, + "learning_rate": 5.959426734348942e-05, + "loss": 1.1816, + "step": 7247 + }, + { + "epoch": 0.44036697247706424, + "grad_norm": 0.4606006145477295, + "learning_rate": 5.9584872282171845e-05, + "loss": 1.212, + "step": 7248 + }, + { + "epoch": 0.4404277295096907, + "grad_norm": 0.1799398809671402, + "learning_rate": 5.957547686951894e-05, + "loss": 1.0738, + "step": 7249 + }, + { + "epoch": 0.44048848654231726, + "grad_norm": 0.19885380566120148, + "learning_rate": 5.9566081105875093e-05, + "loss": 1.0906, + "step": 7250 + }, + { + "epoch": 0.4405492435749438, + "grad_norm": 0.24253475666046143, + "learning_rate": 5.955668499158468e-05, + "loss": 1.0935, + "step": 7251 + }, + { + "epoch": 0.44061000060757033, + "grad_norm": 0.22813253104686737, + "learning_rate": 5.954728852699213e-05, + "loss": 1.093, + "step": 7252 + }, + { + "epoch": 0.44067075764019686, + "grad_norm": 0.23932334780693054, + "learning_rate": 5.9537891712441885e-05, + "loss": 1.1362, + "step": 7253 + }, + { + "epoch": 0.4407315146728234, + "grad_norm": 0.9550638198852539, + "learning_rate": 5.952849454827837e-05, + "loss": 1.4036, + "step": 7254 + }, + { + "epoch": 0.4407922717054499, + "grad_norm": 0.21288806200027466, + "learning_rate": 5.951909703484604e-05, + "loss": 1.0194, + "step": 7255 + }, + { + "epoch": 0.4408530287380764, + "grad_norm": 0.1944960653781891, + "learning_rate": 5.950969917248938e-05, + "loss": 1.0356, + "step": 7256 + }, + { + "epoch": 0.44091378577070295, + "grad_norm": 0.18893203139305115, + "learning_rate": 5.950030096155285e-05, + "loss": 1.1194, + "step": 7257 + }, + { + "epoch": 0.4409745428033295, + "grad_norm": 0.19390052556991577, + "learning_rate": 5.9490902402380944e-05, + "loss": 1.0992, + "step": 7258 + }, + { + "epoch": 0.441035299835956, + "grad_norm": 0.17596518993377686, + "learning_rate": 5.948150349531818e-05, + "loss": 1.0515, + "step": 7259 + }, + { + "epoch": 0.44109605686858255, + "grad_norm": 0.2870226502418518, + "learning_rate": 5.947210424070908e-05, + "loss": 1.051, + "step": 7260 + }, + { + "epoch": 0.4411568139012091, + "grad_norm": 0.1645200401544571, + "learning_rate": 5.946270463889816e-05, + "loss": 1.0505, + "step": 7261 + }, + { + "epoch": 0.44121757093383557, + "grad_norm": 0.2147611528635025, + "learning_rate": 5.945330469022997e-05, + "loss": 1.1132, + "step": 7262 + }, + { + "epoch": 0.4412783279664621, + "grad_norm": 0.19651545584201813, + "learning_rate": 5.94439043950491e-05, + "loss": 1.1823, + "step": 7263 + }, + { + "epoch": 0.44133908499908864, + "grad_norm": 0.5738714933395386, + "learning_rate": 5.943450375370006e-05, + "loss": 1.1729, + "step": 7264 + }, + { + "epoch": 0.4413998420317152, + "grad_norm": 0.9765027165412903, + "learning_rate": 5.942510276652746e-05, + "loss": 1.2881, + "step": 7265 + }, + { + "epoch": 0.4414605990643417, + "grad_norm": 0.704963743686676, + "learning_rate": 5.94157014338759e-05, + "loss": 1.2059, + "step": 7266 + }, + { + "epoch": 0.44152135609696824, + "grad_norm": 0.2541247308254242, + "learning_rate": 5.940629975608999e-05, + "loss": 1.097, + "step": 7267 + }, + { + "epoch": 0.4415821131295948, + "grad_norm": 0.4678594768047333, + "learning_rate": 5.939689773351434e-05, + "loss": 1.0629, + "step": 7268 + }, + { + "epoch": 0.44164287016222126, + "grad_norm": 0.18991893529891968, + "learning_rate": 5.938749536649357e-05, + "loss": 1.0508, + "step": 7269 + }, + { + "epoch": 0.4417036271948478, + "grad_norm": 0.2352619469165802, + "learning_rate": 5.937809265537239e-05, + "loss": 1.0585, + "step": 7270 + }, + { + "epoch": 0.4417643842274743, + "grad_norm": 0.22449132800102234, + "learning_rate": 5.936868960049538e-05, + "loss": 1.1263, + "step": 7271 + }, + { + "epoch": 0.44182514126010086, + "grad_norm": 0.43905478715896606, + "learning_rate": 5.935928620220723e-05, + "loss": 1.1324, + "step": 7272 + }, + { + "epoch": 0.4418858982927274, + "grad_norm": 0.338352233171463, + "learning_rate": 5.934988246085266e-05, + "loss": 1.0595, + "step": 7273 + }, + { + "epoch": 0.44194665532535393, + "grad_norm": 0.27160581946372986, + "learning_rate": 5.934047837677632e-05, + "loss": 1.0747, + "step": 7274 + }, + { + "epoch": 0.4420074123579804, + "grad_norm": 0.34117817878723145, + "learning_rate": 5.933107395032295e-05, + "loss": 1.1169, + "step": 7275 + }, + { + "epoch": 0.44206816939060695, + "grad_norm": 0.4565107524394989, + "learning_rate": 5.932166918183726e-05, + "loss": 1.0686, + "step": 7276 + }, + { + "epoch": 0.4421289264232335, + "grad_norm": 0.24625347554683685, + "learning_rate": 5.931226407166398e-05, + "loss": 1.128, + "step": 7277 + }, + { + "epoch": 0.44218968345586, + "grad_norm": 0.6617821455001831, + "learning_rate": 5.930285862014787e-05, + "loss": 1.0555, + "step": 7278 + }, + { + "epoch": 0.44225044048848655, + "grad_norm": 0.1967105120420456, + "learning_rate": 5.929345282763367e-05, + "loss": 1.0896, + "step": 7279 + }, + { + "epoch": 0.4423111975211131, + "grad_norm": 0.21876326203346252, + "learning_rate": 5.928404669446617e-05, + "loss": 1.0606, + "step": 7280 + }, + { + "epoch": 0.4423719545537396, + "grad_norm": 0.23728539049625397, + "learning_rate": 5.927464022099014e-05, + "loss": 1.0428, + "step": 7281 + }, + { + "epoch": 0.4424327115863661, + "grad_norm": 0.8700002431869507, + "learning_rate": 5.926523340755037e-05, + "loss": 1.1376, + "step": 7282 + }, + { + "epoch": 0.44249346861899264, + "grad_norm": 0.23871907591819763, + "learning_rate": 5.92558262544917e-05, + "loss": 1.0542, + "step": 7283 + }, + { + "epoch": 0.44255422565161917, + "grad_norm": 0.2090703397989273, + "learning_rate": 5.924641876215891e-05, + "loss": 1.0821, + "step": 7284 + }, + { + "epoch": 0.4426149826842457, + "grad_norm": 0.16961465775966644, + "learning_rate": 5.9237010930896856e-05, + "loss": 1.0593, + "step": 7285 + }, + { + "epoch": 0.44267573971687224, + "grad_norm": 1.667732834815979, + "learning_rate": 5.922760276105038e-05, + "loss": 1.1344, + "step": 7286 + }, + { + "epoch": 0.4427364967494988, + "grad_norm": 1.5297890901565552, + "learning_rate": 5.9218194252964355e-05, + "loss": 1.0974, + "step": 7287 + }, + { + "epoch": 0.44279725378212526, + "grad_norm": 0.5030049085617065, + "learning_rate": 5.9208785406983624e-05, + "loss": 1.0825, + "step": 7288 + }, + { + "epoch": 0.4428580108147518, + "grad_norm": 0.20052428543567657, + "learning_rate": 5.919937622345309e-05, + "loss": 1.1851, + "step": 7289 + }, + { + "epoch": 0.4429187678473783, + "grad_norm": 0.3303828537464142, + "learning_rate": 5.9189966702717636e-05, + "loss": 1.0509, + "step": 7290 + }, + { + "epoch": 0.44297952488000486, + "grad_norm": 0.32798075675964355, + "learning_rate": 5.918055684512218e-05, + "loss": 1.1315, + "step": 7291 + }, + { + "epoch": 0.4430402819126314, + "grad_norm": 0.22705096006393433, + "learning_rate": 5.917114665101163e-05, + "loss": 1.1541, + "step": 7292 + }, + { + "epoch": 0.44310103894525793, + "grad_norm": 0.3118288516998291, + "learning_rate": 5.916173612073095e-05, + "loss": 1.1059, + "step": 7293 + }, + { + "epoch": 0.44316179597788447, + "grad_norm": 0.2543984353542328, + "learning_rate": 5.915232525462505e-05, + "loss": 1.0738, + "step": 7294 + }, + { + "epoch": 0.44322255301051094, + "grad_norm": 0.1721610575914383, + "learning_rate": 5.914291405303889e-05, + "loss": 1.042, + "step": 7295 + }, + { + "epoch": 0.4432833100431375, + "grad_norm": 0.2705150842666626, + "learning_rate": 5.913350251631745e-05, + "loss": 1.1422, + "step": 7296 + }, + { + "epoch": 0.443344067075764, + "grad_norm": 1.6969562768936157, + "learning_rate": 5.9124090644805706e-05, + "loss": 1.0718, + "step": 7297 + }, + { + "epoch": 0.44340482410839055, + "grad_norm": 0.23043838143348694, + "learning_rate": 5.911467843884867e-05, + "loss": 1.1493, + "step": 7298 + }, + { + "epoch": 0.4434655811410171, + "grad_norm": 0.2982065677642822, + "learning_rate": 5.910526589879133e-05, + "loss": 1.1018, + "step": 7299 + }, + { + "epoch": 0.4435263381736436, + "grad_norm": 3.794044256210327, + "learning_rate": 5.9095853024978706e-05, + "loss": 1.087, + "step": 7300 + }, + { + "epoch": 0.4435870952062701, + "grad_norm": 0.23261511325836182, + "learning_rate": 5.9086439817755833e-05, + "loss": 1.1784, + "step": 7301 + }, + { + "epoch": 0.44364785223889663, + "grad_norm": 5.770605087280273, + "learning_rate": 5.907702627746775e-05, + "loss": 1.0421, + "step": 7302 + }, + { + "epoch": 0.44370860927152317, + "grad_norm": 0.24627025425434113, + "learning_rate": 5.9067612404459516e-05, + "loss": 1.0787, + "step": 7303 + }, + { + "epoch": 0.4437693663041497, + "grad_norm": 0.2217102199792862, + "learning_rate": 5.90581981990762e-05, + "loss": 1.0988, + "step": 7304 + }, + { + "epoch": 0.44383012333677624, + "grad_norm": 0.3626880943775177, + "learning_rate": 5.904878366166285e-05, + "loss": 1.2038, + "step": 7305 + }, + { + "epoch": 0.4438908803694028, + "grad_norm": 0.4255942404270172, + "learning_rate": 5.903936879256462e-05, + "loss": 1.1582, + "step": 7306 + }, + { + "epoch": 0.4439516374020293, + "grad_norm": 0.17142796516418457, + "learning_rate": 5.9029953592126575e-05, + "loss": 1.0468, + "step": 7307 + }, + { + "epoch": 0.4440123944346558, + "grad_norm": 0.22460462152957916, + "learning_rate": 5.902053806069382e-05, + "loss": 1.0342, + "step": 7308 + }, + { + "epoch": 0.4440731514672823, + "grad_norm": 0.32491299510002136, + "learning_rate": 5.9011122198611515e-05, + "loss": 1.1127, + "step": 7309 + }, + { + "epoch": 0.44413390849990886, + "grad_norm": 0.2077612727880478, + "learning_rate": 5.900170600622477e-05, + "loss": 1.1599, + "step": 7310 + }, + { + "epoch": 0.4441946655325354, + "grad_norm": 0.27063629031181335, + "learning_rate": 5.899228948387876e-05, + "loss": 1.156, + "step": 7311 + }, + { + "epoch": 0.44425542256516193, + "grad_norm": 0.16541171073913574, + "learning_rate": 5.898287263191864e-05, + "loss": 1.0417, + "step": 7312 + }, + { + "epoch": 0.44431617959778846, + "grad_norm": 0.18324244022369385, + "learning_rate": 5.8973455450689586e-05, + "loss": 1.1376, + "step": 7313 + }, + { + "epoch": 0.444376936630415, + "grad_norm": 0.20537932217121124, + "learning_rate": 5.896403794053679e-05, + "loss": 1.1703, + "step": 7314 + }, + { + "epoch": 0.4444376936630415, + "grad_norm": 0.27177220582962036, + "learning_rate": 5.895462010180545e-05, + "loss": 1.1099, + "step": 7315 + }, + { + "epoch": 0.444498450695668, + "grad_norm": 0.21656423807144165, + "learning_rate": 5.894520193484077e-05, + "loss": 1.0373, + "step": 7316 + }, + { + "epoch": 0.44455920772829455, + "grad_norm": 0.33748769760131836, + "learning_rate": 5.8935783439988004e-05, + "loss": 1.0538, + "step": 7317 + }, + { + "epoch": 0.4446199647609211, + "grad_norm": 0.20813708007335663, + "learning_rate": 5.892636461759234e-05, + "loss": 1.0488, + "step": 7318 + }, + { + "epoch": 0.4446807217935476, + "grad_norm": 0.27373820543289185, + "learning_rate": 5.891694546799909e-05, + "loss": 1.1244, + "step": 7319 + }, + { + "epoch": 0.44474147882617415, + "grad_norm": 0.3898060917854309, + "learning_rate": 5.8907525991553444e-05, + "loss": 1.2062, + "step": 7320 + }, + { + "epoch": 0.44480223585880063, + "grad_norm": 0.16159678995609283, + "learning_rate": 5.889810618860074e-05, + "loss": 1.0811, + "step": 7321 + }, + { + "epoch": 0.44486299289142717, + "grad_norm": 0.4525826871395111, + "learning_rate": 5.888868605948622e-05, + "loss": 1.0349, + "step": 7322 + }, + { + "epoch": 0.4449237499240537, + "grad_norm": 0.24378934502601624, + "learning_rate": 5.88792656045552e-05, + "loss": 1.0912, + "step": 7323 + }, + { + "epoch": 0.44498450695668024, + "grad_norm": 0.5948576331138611, + "learning_rate": 5.8869844824152985e-05, + "loss": 1.1911, + "step": 7324 + }, + { + "epoch": 0.44504526398930677, + "grad_norm": 0.3675840198993683, + "learning_rate": 5.886042371862489e-05, + "loss": 0.9903, + "step": 7325 + }, + { + "epoch": 0.4451060210219333, + "grad_norm": 0.18750004470348358, + "learning_rate": 5.8851002288316234e-05, + "loss": 1.1016, + "step": 7326 + }, + { + "epoch": 0.44516677805455984, + "grad_norm": 0.54721599817276, + "learning_rate": 5.884158053357241e-05, + "loss": 1.1382, + "step": 7327 + }, + { + "epoch": 0.4452275350871863, + "grad_norm": 0.3777073919773102, + "learning_rate": 5.883215845473872e-05, + "loss": 1.0551, + "step": 7328 + }, + { + "epoch": 0.44528829211981286, + "grad_norm": 0.529330849647522, + "learning_rate": 5.8822736052160553e-05, + "loss": 1.1569, + "step": 7329 + }, + { + "epoch": 0.4453490491524394, + "grad_norm": 0.8565409779548645, + "learning_rate": 5.881331332618332e-05, + "loss": 1.1089, + "step": 7330 + }, + { + "epoch": 0.4454098061850659, + "grad_norm": 1.0311236381530762, + "learning_rate": 5.880389027715235e-05, + "loss": 1.4039, + "step": 7331 + }, + { + "epoch": 0.44547056321769246, + "grad_norm": 0.293642520904541, + "learning_rate": 5.8794466905413095e-05, + "loss": 1.1009, + "step": 7332 + }, + { + "epoch": 0.445531320250319, + "grad_norm": 0.16232475638389587, + "learning_rate": 5.8785043211310954e-05, + "loss": 1.0388, + "step": 7333 + }, + { + "epoch": 0.4455920772829455, + "grad_norm": 0.20965005457401276, + "learning_rate": 5.8775619195191355e-05, + "loss": 1.1059, + "step": 7334 + }, + { + "epoch": 0.445652834315572, + "grad_norm": 0.18931862711906433, + "learning_rate": 5.876619485739974e-05, + "loss": 1.1358, + "step": 7335 + }, + { + "epoch": 0.44571359134819855, + "grad_norm": 0.16539393365383148, + "learning_rate": 5.875677019828154e-05, + "loss": 1.0615, + "step": 7336 + }, + { + "epoch": 0.4457743483808251, + "grad_norm": 0.22868390381336212, + "learning_rate": 5.874734521818227e-05, + "loss": 1.0826, + "step": 7337 + }, + { + "epoch": 0.4458351054134516, + "grad_norm": 0.16299359500408173, + "learning_rate": 5.873791991744735e-05, + "loss": 1.1033, + "step": 7338 + }, + { + "epoch": 0.44589586244607815, + "grad_norm": 0.1611664742231369, + "learning_rate": 5.872849429642228e-05, + "loss": 0.9899, + "step": 7339 + }, + { + "epoch": 0.4459566194787047, + "grad_norm": 0.17430822551250458, + "learning_rate": 5.871906835545259e-05, + "loss": 1.0631, + "step": 7340 + }, + { + "epoch": 0.44601737651133117, + "grad_norm": 0.1778005063533783, + "learning_rate": 5.8709642094883736e-05, + "loss": 1.0179, + "step": 7341 + }, + { + "epoch": 0.4460781335439577, + "grad_norm": 0.2959222197532654, + "learning_rate": 5.870021551506127e-05, + "loss": 1.0645, + "step": 7342 + }, + { + "epoch": 0.44613889057658423, + "grad_norm": 0.31365785002708435, + "learning_rate": 5.869078861633075e-05, + "loss": 1.1967, + "step": 7343 + }, + { + "epoch": 0.44619964760921077, + "grad_norm": 0.26272115111351013, + "learning_rate": 5.868136139903766e-05, + "loss": 1.1033, + "step": 7344 + }, + { + "epoch": 0.4462604046418373, + "grad_norm": 0.19225920736789703, + "learning_rate": 5.8671933863527606e-05, + "loss": 1.0533, + "step": 7345 + }, + { + "epoch": 0.44632116167446384, + "grad_norm": 0.2123839259147644, + "learning_rate": 5.8662506010146146e-05, + "loss": 1.0629, + "step": 7346 + }, + { + "epoch": 0.4463819187070903, + "grad_norm": 0.16645079851150513, + "learning_rate": 5.865307783923885e-05, + "loss": 1.0523, + "step": 7347 + }, + { + "epoch": 0.44644267573971685, + "grad_norm": 0.17835716903209686, + "learning_rate": 5.86436493511513e-05, + "loss": 1.0763, + "step": 7348 + }, + { + "epoch": 0.4465034327723434, + "grad_norm": 1.0383397340774536, + "learning_rate": 5.8634220546229115e-05, + "loss": 1.1288, + "step": 7349 + }, + { + "epoch": 0.4465641898049699, + "grad_norm": 0.2143787443637848, + "learning_rate": 5.8624791424817926e-05, + "loss": 1.0281, + "step": 7350 + }, + { + "epoch": 0.44662494683759646, + "grad_norm": 0.15249891579151154, + "learning_rate": 5.8615361987263325e-05, + "loss": 1.0411, + "step": 7351 + }, + { + "epoch": 0.446685703870223, + "grad_norm": 0.28516435623168945, + "learning_rate": 5.860593223391096e-05, + "loss": 1.0104, + "step": 7352 + }, + { + "epoch": 0.44674646090284953, + "grad_norm": 0.18722732365131378, + "learning_rate": 5.8596502165106494e-05, + "loss": 1.0085, + "step": 7353 + }, + { + "epoch": 0.446807217935476, + "grad_norm": 0.18831618130207062, + "learning_rate": 5.858707178119558e-05, + "loss": 1.0373, + "step": 7354 + }, + { + "epoch": 0.44686797496810254, + "grad_norm": 0.18261127173900604, + "learning_rate": 5.857764108252389e-05, + "loss": 1.0358, + "step": 7355 + }, + { + "epoch": 0.4469287320007291, + "grad_norm": 0.15932577848434448, + "learning_rate": 5.856821006943711e-05, + "loss": 0.9971, + "step": 7356 + }, + { + "epoch": 0.4469894890333556, + "grad_norm": 0.22685480117797852, + "learning_rate": 5.855877874228093e-05, + "loss": 1.0898, + "step": 7357 + }, + { + "epoch": 0.44705024606598215, + "grad_norm": 0.20138020813465118, + "learning_rate": 5.8549347101401066e-05, + "loss": 1.1175, + "step": 7358 + }, + { + "epoch": 0.4471110030986087, + "grad_norm": 4.473047733306885, + "learning_rate": 5.853991514714322e-05, + "loss": 1.0632, + "step": 7359 + }, + { + "epoch": 0.44717176013123516, + "grad_norm": 0.3470970690250397, + "learning_rate": 5.8530482879853145e-05, + "loss": 0.9923, + "step": 7360 + }, + { + "epoch": 0.4472325171638617, + "grad_norm": 3.6041197776794434, + "learning_rate": 5.8521050299876566e-05, + "loss": 1.1897, + "step": 7361 + }, + { + "epoch": 0.44729327419648823, + "grad_norm": 0.2523518204689026, + "learning_rate": 5.851161740755924e-05, + "loss": 1.0896, + "step": 7362 + }, + { + "epoch": 0.44735403122911477, + "grad_norm": 0.38514822721481323, + "learning_rate": 5.8502184203246924e-05, + "loss": 1.0995, + "step": 7363 + }, + { + "epoch": 0.4474147882617413, + "grad_norm": 0.2697890102863312, + "learning_rate": 5.849275068728543e-05, + "loss": 1.1077, + "step": 7364 + }, + { + "epoch": 0.44747554529436784, + "grad_norm": 0.33724379539489746, + "learning_rate": 5.848331686002049e-05, + "loss": 1.1073, + "step": 7365 + }, + { + "epoch": 0.4475363023269944, + "grad_norm": 0.20428842306137085, + "learning_rate": 5.847388272179795e-05, + "loss": 1.1274, + "step": 7366 + }, + { + "epoch": 0.44759705935962085, + "grad_norm": 0.17937445640563965, + "learning_rate": 5.846444827296359e-05, + "loss": 1.0689, + "step": 7367 + }, + { + "epoch": 0.4476578163922474, + "grad_norm": 0.42051345109939575, + "learning_rate": 5.845501351386324e-05, + "loss": 1.1235, + "step": 7368 + }, + { + "epoch": 0.4477185734248739, + "grad_norm": 0.2665199637413025, + "learning_rate": 5.844557844484274e-05, + "loss": 1.0015, + "step": 7369 + }, + { + "epoch": 0.44777933045750046, + "grad_norm": 0.1776839643716812, + "learning_rate": 5.8436143066247936e-05, + "loss": 1.087, + "step": 7370 + }, + { + "epoch": 0.447840087490127, + "grad_norm": 0.28367990255355835, + "learning_rate": 5.8426707378424675e-05, + "loss": 1.0041, + "step": 7371 + }, + { + "epoch": 0.4479008445227535, + "grad_norm": 0.2045825868844986, + "learning_rate": 5.841727138171882e-05, + "loss": 1.0636, + "step": 7372 + }, + { + "epoch": 0.44796160155538006, + "grad_norm": 0.25175440311431885, + "learning_rate": 5.840783507647625e-05, + "loss": 1.0998, + "step": 7373 + }, + { + "epoch": 0.44802235858800654, + "grad_norm": 0.15413321554660797, + "learning_rate": 5.839839846304287e-05, + "loss": 1.0882, + "step": 7374 + }, + { + "epoch": 0.4480831156206331, + "grad_norm": 8.23060131072998, + "learning_rate": 5.838896154176456e-05, + "loss": 1.0511, + "step": 7375 + }, + { + "epoch": 0.4481438726532596, + "grad_norm": 0.19782058894634247, + "learning_rate": 5.837952431298726e-05, + "loss": 1.0542, + "step": 7376 + }, + { + "epoch": 0.44820462968588615, + "grad_norm": 0.23401859402656555, + "learning_rate": 5.8370086777056866e-05, + "loss": 1.0465, + "step": 7377 + }, + { + "epoch": 0.4482653867185127, + "grad_norm": 0.17422102391719818, + "learning_rate": 5.836064893431933e-05, + "loss": 1.062, + "step": 7378 + }, + { + "epoch": 0.4483261437511392, + "grad_norm": 0.1864158660173416, + "learning_rate": 5.835121078512058e-05, + "loss": 1.1608, + "step": 7379 + }, + { + "epoch": 0.4483869007837657, + "grad_norm": 0.22185742855072021, + "learning_rate": 5.83417723298066e-05, + "loss": 1.1381, + "step": 7380 + }, + { + "epoch": 0.44844765781639223, + "grad_norm": 0.1733950972557068, + "learning_rate": 5.833233356872333e-05, + "loss": 1.0276, + "step": 7381 + }, + { + "epoch": 0.44850841484901877, + "grad_norm": 0.22213006019592285, + "learning_rate": 5.832289450221677e-05, + "loss": 1.0762, + "step": 7382 + }, + { + "epoch": 0.4485691718816453, + "grad_norm": 0.18503446877002716, + "learning_rate": 5.831345513063289e-05, + "loss": 1.0899, + "step": 7383 + }, + { + "epoch": 0.44862992891427184, + "grad_norm": 0.21543844044208527, + "learning_rate": 5.8304015454317726e-05, + "loss": 1.0582, + "step": 7384 + }, + { + "epoch": 0.44869068594689837, + "grad_norm": 0.45483461022377014, + "learning_rate": 5.8294575473617254e-05, + "loss": 1.2154, + "step": 7385 + }, + { + "epoch": 0.4487514429795249, + "grad_norm": 0.2606468200683594, + "learning_rate": 5.8285135188877517e-05, + "loss": 1.104, + "step": 7386 + }, + { + "epoch": 0.4488122000121514, + "grad_norm": 1.2074331045150757, + "learning_rate": 5.8275694600444565e-05, + "loss": 1.2674, + "step": 7387 + }, + { + "epoch": 0.4488729570447779, + "grad_norm": 0.2907407879829407, + "learning_rate": 5.8266253708664396e-05, + "loss": 1.1812, + "step": 7388 + }, + { + "epoch": 0.44893371407740446, + "grad_norm": 0.1780625879764557, + "learning_rate": 5.825681251388313e-05, + "loss": 1.0933, + "step": 7389 + }, + { + "epoch": 0.448994471110031, + "grad_norm": 5.302388668060303, + "learning_rate": 5.824737101644679e-05, + "loss": 1.1402, + "step": 7390 + }, + { + "epoch": 0.4490552281426575, + "grad_norm": 0.4203245937824249, + "learning_rate": 5.823792921670147e-05, + "loss": 1.1359, + "step": 7391 + }, + { + "epoch": 0.44911598517528406, + "grad_norm": 0.21080723404884338, + "learning_rate": 5.822848711499327e-05, + "loss": 1.0874, + "step": 7392 + }, + { + "epoch": 0.44917674220791054, + "grad_norm": 0.23848605155944824, + "learning_rate": 5.8219044711668266e-05, + "loss": 1.1006, + "step": 7393 + }, + { + "epoch": 0.4492374992405371, + "grad_norm": 0.15253865718841553, + "learning_rate": 5.8209602007072615e-05, + "loss": 1.0659, + "step": 7394 + }, + { + "epoch": 0.4492982562731636, + "grad_norm": 1.5301845073699951, + "learning_rate": 5.8200159001552415e-05, + "loss": 1.0608, + "step": 7395 + }, + { + "epoch": 0.44935901330579014, + "grad_norm": 0.27607110142707825, + "learning_rate": 5.819071569545378e-05, + "loss": 1.0852, + "step": 7396 + }, + { + "epoch": 0.4494197703384167, + "grad_norm": 0.15938974916934967, + "learning_rate": 5.81812720891229e-05, + "loss": 1.0374, + "step": 7397 + }, + { + "epoch": 0.4494805273710432, + "grad_norm": 0.368619829416275, + "learning_rate": 5.817182818290588e-05, + "loss": 1.0503, + "step": 7398 + }, + { + "epoch": 0.44954128440366975, + "grad_norm": 0.6582369208335876, + "learning_rate": 5.816238397714895e-05, + "loss": 1.0836, + "step": 7399 + }, + { + "epoch": 0.44960204143629623, + "grad_norm": 0.16880817711353302, + "learning_rate": 5.815293947219825e-05, + "loss": 1.0447, + "step": 7400 + }, + { + "epoch": 0.44966279846892276, + "grad_norm": 0.3757792115211487, + "learning_rate": 5.814349466839997e-05, + "loss": 1.1212, + "step": 7401 + }, + { + "epoch": 0.4497235555015493, + "grad_norm": 0.3731338083744049, + "learning_rate": 5.8134049566100325e-05, + "loss": 1.0621, + "step": 7402 + }, + { + "epoch": 0.44978431253417583, + "grad_norm": 0.17359742522239685, + "learning_rate": 5.812460416564551e-05, + "loss": 1.0372, + "step": 7403 + }, + { + "epoch": 0.44984506956680237, + "grad_norm": 0.21395398676395416, + "learning_rate": 5.811515846738177e-05, + "loss": 1.0529, + "step": 7404 + }, + { + "epoch": 0.4499058265994289, + "grad_norm": 0.20391100645065308, + "learning_rate": 5.810571247165533e-05, + "loss": 1.0265, + "step": 7405 + }, + { + "epoch": 0.4499665836320554, + "grad_norm": 0.26830974221229553, + "learning_rate": 5.809626617881241e-05, + "loss": 0.9902, + "step": 7406 + }, + { + "epoch": 0.4500273406646819, + "grad_norm": 0.3767789304256439, + "learning_rate": 5.808681958919932e-05, + "loss": 1.1464, + "step": 7407 + }, + { + "epoch": 0.45008809769730845, + "grad_norm": 0.21065771579742432, + "learning_rate": 5.807737270316228e-05, + "loss": 1.0878, + "step": 7408 + }, + { + "epoch": 0.450148854729935, + "grad_norm": 0.2770909070968628, + "learning_rate": 5.806792552104757e-05, + "loss": 1.0054, + "step": 7409 + }, + { + "epoch": 0.4502096117625615, + "grad_norm": 0.19188305735588074, + "learning_rate": 5.8058478043201526e-05, + "loss": 1.0504, + "step": 7410 + }, + { + "epoch": 0.45027036879518806, + "grad_norm": 0.3214500844478607, + "learning_rate": 5.8049030269970374e-05, + "loss": 1.0814, + "step": 7411 + }, + { + "epoch": 0.4503311258278146, + "grad_norm": 0.2563783824443817, + "learning_rate": 5.803958220170047e-05, + "loss": 1.16, + "step": 7412 + }, + { + "epoch": 0.4503918828604411, + "grad_norm": 0.24616369605064392, + "learning_rate": 5.8030133838738146e-05, + "loss": 1.1287, + "step": 7413 + }, + { + "epoch": 0.4504526398930676, + "grad_norm": 0.1581309288740158, + "learning_rate": 5.802068518142969e-05, + "loss": 1.0248, + "step": 7414 + }, + { + "epoch": 0.45051339692569414, + "grad_norm": 0.30103766918182373, + "learning_rate": 5.801123623012148e-05, + "loss": 1.0568, + "step": 7415 + }, + { + "epoch": 0.4505741539583207, + "grad_norm": 0.1456846445798874, + "learning_rate": 5.8001786985159856e-05, + "loss": 1.0329, + "step": 7416 + }, + { + "epoch": 0.4506349109909472, + "grad_norm": 0.47445839643478394, + "learning_rate": 5.7992337446891186e-05, + "loss": 1.1646, + "step": 7417 + }, + { + "epoch": 0.45069566802357375, + "grad_norm": 3.8807661533355713, + "learning_rate": 5.798288761566184e-05, + "loss": 1.241, + "step": 7418 + }, + { + "epoch": 0.4507564250562003, + "grad_norm": 0.3615172803401947, + "learning_rate": 5.797343749181819e-05, + "loss": 1.1357, + "step": 7419 + }, + { + "epoch": 0.45081718208882676, + "grad_norm": 0.3322477340698242, + "learning_rate": 5.796398707570668e-05, + "loss": 1.0446, + "step": 7420 + }, + { + "epoch": 0.4508779391214533, + "grad_norm": 0.2173192948102951, + "learning_rate": 5.795453636767366e-05, + "loss": 1.0559, + "step": 7421 + }, + { + "epoch": 0.45093869615407983, + "grad_norm": 0.18061821162700653, + "learning_rate": 5.794508536806558e-05, + "loss": 1.0454, + "step": 7422 + }, + { + "epoch": 0.45099945318670637, + "grad_norm": 0.1681375801563263, + "learning_rate": 5.793563407722886e-05, + "loss": 1.0294, + "step": 7423 + }, + { + "epoch": 0.4510602102193329, + "grad_norm": 0.20406576991081238, + "learning_rate": 5.792618249550994e-05, + "loss": 1.1102, + "step": 7424 + }, + { + "epoch": 0.45112096725195944, + "grad_norm": 0.29774633049964905, + "learning_rate": 5.791673062325527e-05, + "loss": 1.1459, + "step": 7425 + }, + { + "epoch": 0.4511817242845859, + "grad_norm": 0.5764881372451782, + "learning_rate": 5.79072784608113e-05, + "loss": 1.1653, + "step": 7426 + }, + { + "epoch": 0.45124248131721245, + "grad_norm": 1.2865904569625854, + "learning_rate": 5.789782600852452e-05, + "loss": 1.0483, + "step": 7427 + }, + { + "epoch": 0.451303238349839, + "grad_norm": 0.1547137349843979, + "learning_rate": 5.7888373266741394e-05, + "loss": 1.0537, + "step": 7428 + }, + { + "epoch": 0.4513639953824655, + "grad_norm": 0.18341271579265594, + "learning_rate": 5.787892023580841e-05, + "loss": 1.0131, + "step": 7429 + }, + { + "epoch": 0.45142475241509206, + "grad_norm": 0.29318100214004517, + "learning_rate": 5.786946691607209e-05, + "loss": 1.08, + "step": 7430 + }, + { + "epoch": 0.4514855094477186, + "grad_norm": 0.21907569468021393, + "learning_rate": 5.786001330787896e-05, + "loss": 1.085, + "step": 7431 + }, + { + "epoch": 0.4515462664803451, + "grad_norm": 0.1893540620803833, + "learning_rate": 5.78505594115755e-05, + "loss": 1.0948, + "step": 7432 + }, + { + "epoch": 0.4516070235129716, + "grad_norm": 0.6454317569732666, + "learning_rate": 5.784110522750828e-05, + "loss": 1.0561, + "step": 7433 + }, + { + "epoch": 0.45166778054559814, + "grad_norm": 0.3014373779296875, + "learning_rate": 5.783165075602381e-05, + "loss": 1.1023, + "step": 7434 + }, + { + "epoch": 0.4517285375782247, + "grad_norm": 0.4181009531021118, + "learning_rate": 5.78221959974687e-05, + "loss": 1.2981, + "step": 7435 + }, + { + "epoch": 0.4517892946108512, + "grad_norm": 0.19378197193145752, + "learning_rate": 5.781274095218946e-05, + "loss": 1.0489, + "step": 7436 + }, + { + "epoch": 0.45185005164347775, + "grad_norm": 0.21390679478645325, + "learning_rate": 5.78032856205327e-05, + "loss": 1.125, + "step": 7437 + }, + { + "epoch": 0.4519108086761043, + "grad_norm": 0.1993975192308426, + "learning_rate": 5.7793830002845e-05, + "loss": 1.0994, + "step": 7438 + }, + { + "epoch": 0.45197156570873076, + "grad_norm": 0.16427543759346008, + "learning_rate": 5.778437409947295e-05, + "loss": 1.0348, + "step": 7439 + }, + { + "epoch": 0.4520323227413573, + "grad_norm": 0.8129555583000183, + "learning_rate": 5.777491791076315e-05, + "loss": 1.2974, + "step": 7440 + }, + { + "epoch": 0.45209307977398383, + "grad_norm": 5.783533096313477, + "learning_rate": 5.776546143706225e-05, + "loss": 1.1173, + "step": 7441 + }, + { + "epoch": 0.45215383680661037, + "grad_norm": 0.5417641997337341, + "learning_rate": 5.7756004678716847e-05, + "loss": 1.0549, + "step": 7442 + }, + { + "epoch": 0.4522145938392369, + "grad_norm": 0.33411550521850586, + "learning_rate": 5.77465476360736e-05, + "loss": 1.0838, + "step": 7443 + }, + { + "epoch": 0.45227535087186344, + "grad_norm": 0.4922631084918976, + "learning_rate": 5.773709030947916e-05, + "loss": 1.031, + "step": 7444 + }, + { + "epoch": 0.45233610790448997, + "grad_norm": 0.34361937642097473, + "learning_rate": 5.7727632699280166e-05, + "loss": 1.1276, + "step": 7445 + }, + { + "epoch": 0.45239686493711645, + "grad_norm": 0.2769177258014679, + "learning_rate": 5.771817480582331e-05, + "loss": 1.111, + "step": 7446 + }, + { + "epoch": 0.452457621969743, + "grad_norm": 0.28126955032348633, + "learning_rate": 5.7708716629455264e-05, + "loss": 1.0044, + "step": 7447 + }, + { + "epoch": 0.4525183790023695, + "grad_norm": 0.2639360725879669, + "learning_rate": 5.7699258170522705e-05, + "loss": 1.0767, + "step": 7448 + }, + { + "epoch": 0.45257913603499605, + "grad_norm": 0.4880031943321228, + "learning_rate": 5.768979942937237e-05, + "loss": 1.0442, + "step": 7449 + }, + { + "epoch": 0.4526398930676226, + "grad_norm": 0.26361867785453796, + "learning_rate": 5.768034040635093e-05, + "loss": 1.1781, + "step": 7450 + }, + { + "epoch": 0.4527006501002491, + "grad_norm": 0.4553070366382599, + "learning_rate": 5.767088110180515e-05, + "loss": 1.1122, + "step": 7451 + }, + { + "epoch": 0.4527614071328756, + "grad_norm": 0.24578317999839783, + "learning_rate": 5.766142151608174e-05, + "loss": 1.0801, + "step": 7452 + }, + { + "epoch": 0.45282216416550214, + "grad_norm": 0.19122429192066193, + "learning_rate": 5.765196164952742e-05, + "loss": 1.0392, + "step": 7453 + }, + { + "epoch": 0.4528829211981287, + "grad_norm": 0.19522784650325775, + "learning_rate": 5.7642501502488986e-05, + "loss": 1.1267, + "step": 7454 + }, + { + "epoch": 0.4529436782307552, + "grad_norm": 0.47257351875305176, + "learning_rate": 5.7633041075313176e-05, + "loss": 1.029, + "step": 7455 + }, + { + "epoch": 0.45300443526338174, + "grad_norm": 0.2940736413002014, + "learning_rate": 5.7623580368346765e-05, + "loss": 1.2428, + "step": 7456 + }, + { + "epoch": 0.4530651922960083, + "grad_norm": 0.40287768840789795, + "learning_rate": 5.761411938193656e-05, + "loss": 1.0292, + "step": 7457 + }, + { + "epoch": 0.4531259493286348, + "grad_norm": 0.3594214618206024, + "learning_rate": 5.760465811642932e-05, + "loss": 1.1151, + "step": 7458 + }, + { + "epoch": 0.4531867063612613, + "grad_norm": 0.23864169418811798, + "learning_rate": 5.759519657217187e-05, + "loss": 1.0729, + "step": 7459 + }, + { + "epoch": 0.45324746339388783, + "grad_norm": 0.2789094150066376, + "learning_rate": 5.758573474951102e-05, + "loss": 1.064, + "step": 7460 + }, + { + "epoch": 0.45330822042651436, + "grad_norm": 0.25875720381736755, + "learning_rate": 5.75762726487936e-05, + "loss": 1.1066, + "step": 7461 + }, + { + "epoch": 0.4533689774591409, + "grad_norm": 0.44665399193763733, + "learning_rate": 5.7566810270366445e-05, + "loss": 1.0445, + "step": 7462 + }, + { + "epoch": 0.45342973449176743, + "grad_norm": 0.24276277422904968, + "learning_rate": 5.7557347614576384e-05, + "loss": 1.2205, + "step": 7463 + }, + { + "epoch": 0.45349049152439397, + "grad_norm": 0.1874784380197525, + "learning_rate": 5.75478846817703e-05, + "loss": 1.0794, + "step": 7464 + }, + { + "epoch": 0.45355124855702045, + "grad_norm": 0.23392121493816376, + "learning_rate": 5.753842147229504e-05, + "loss": 1.1179, + "step": 7465 + }, + { + "epoch": 0.453612005589647, + "grad_norm": 0.2731075882911682, + "learning_rate": 5.752895798649748e-05, + "loss": 1.0634, + "step": 7466 + }, + { + "epoch": 0.4536727626222735, + "grad_norm": 0.2631729543209076, + "learning_rate": 5.751949422472452e-05, + "loss": 1.1628, + "step": 7467 + }, + { + "epoch": 0.45373351965490005, + "grad_norm": 0.20215779542922974, + "learning_rate": 5.751003018732303e-05, + "loss": 1.0465, + "step": 7468 + }, + { + "epoch": 0.4537942766875266, + "grad_norm": 3.2414135932922363, + "learning_rate": 5.750056587463995e-05, + "loss": 1.0799, + "step": 7469 + }, + { + "epoch": 0.4538550337201531, + "grad_norm": 0.212851420044899, + "learning_rate": 5.7491101287022154e-05, + "loss": 1.0968, + "step": 7470 + }, + { + "epoch": 0.45391579075277966, + "grad_norm": 1.4891316890716553, + "learning_rate": 5.748163642481661e-05, + "loss": 1.0689, + "step": 7471 + }, + { + "epoch": 0.45397654778540614, + "grad_norm": 0.3126177191734314, + "learning_rate": 5.7472171288370235e-05, + "loss": 1.1238, + "step": 7472 + }, + { + "epoch": 0.45403730481803267, + "grad_norm": 0.5243487358093262, + "learning_rate": 5.746270587802998e-05, + "loss": 1.2805, + "step": 7473 + }, + { + "epoch": 0.4540980618506592, + "grad_norm": 0.24549609422683716, + "learning_rate": 5.745324019414279e-05, + "loss": 1.0696, + "step": 7474 + }, + { + "epoch": 0.45415881888328574, + "grad_norm": 0.27857351303100586, + "learning_rate": 5.7443774237055644e-05, + "loss": 1.0511, + "step": 7475 + }, + { + "epoch": 0.4542195759159123, + "grad_norm": 0.21311511099338531, + "learning_rate": 5.74343080071155e-05, + "loss": 1.1063, + "step": 7476 + }, + { + "epoch": 0.4542803329485388, + "grad_norm": 0.2704142928123474, + "learning_rate": 5.742484150466939e-05, + "loss": 1.1988, + "step": 7477 + }, + { + "epoch": 0.45434108998116535, + "grad_norm": 0.19638675451278687, + "learning_rate": 5.741537473006425e-05, + "loss": 1.108, + "step": 7478 + }, + { + "epoch": 0.4544018470137918, + "grad_norm": 0.1897803246974945, + "learning_rate": 5.7405907683647134e-05, + "loss": 1.108, + "step": 7479 + }, + { + "epoch": 0.45446260404641836, + "grad_norm": 0.1746230572462082, + "learning_rate": 5.739644036576502e-05, + "loss": 1.0038, + "step": 7480 + }, + { + "epoch": 0.4545233610790449, + "grad_norm": 0.3049311339855194, + "learning_rate": 5.738697277676498e-05, + "loss": 1.0239, + "step": 7481 + }, + { + "epoch": 0.45458411811167143, + "grad_norm": 0.2433568686246872, + "learning_rate": 5.737750491699402e-05, + "loss": 1.2201, + "step": 7482 + }, + { + "epoch": 0.45464487514429797, + "grad_norm": 0.2792401909828186, + "learning_rate": 5.736803678679918e-05, + "loss": 1.0529, + "step": 7483 + }, + { + "epoch": 0.4547056321769245, + "grad_norm": 0.3068558871746063, + "learning_rate": 5.735856838652754e-05, + "loss": 1.0503, + "step": 7484 + }, + { + "epoch": 0.454766389209551, + "grad_norm": 0.21400170028209686, + "learning_rate": 5.734909971652614e-05, + "loss": 1.0205, + "step": 7485 + }, + { + "epoch": 0.4548271462421775, + "grad_norm": 0.35003596544265747, + "learning_rate": 5.733963077714206e-05, + "loss": 1.2883, + "step": 7486 + }, + { + "epoch": 0.45488790327480405, + "grad_norm": 0.31050729751586914, + "learning_rate": 5.7330161568722426e-05, + "loss": 1.0363, + "step": 7487 + }, + { + "epoch": 0.4549486603074306, + "grad_norm": 0.43270885944366455, + "learning_rate": 5.732069209161429e-05, + "loss": 1.2469, + "step": 7488 + }, + { + "epoch": 0.4550094173400571, + "grad_norm": 1.9055674076080322, + "learning_rate": 5.731122234616476e-05, + "loss": 1.0529, + "step": 7489 + }, + { + "epoch": 0.45507017437268366, + "grad_norm": 0.6104212403297424, + "learning_rate": 5.730175233272097e-05, + "loss": 1.0598, + "step": 7490 + }, + { + "epoch": 0.4551309314053102, + "grad_norm": 0.25065895915031433, + "learning_rate": 5.729228205163003e-05, + "loss": 1.0421, + "step": 7491 + }, + { + "epoch": 0.45519168843793667, + "grad_norm": 0.32287582755088806, + "learning_rate": 5.7282811503239096e-05, + "loss": 1.1318, + "step": 7492 + }, + { + "epoch": 0.4552524454705632, + "grad_norm": 0.6019036173820496, + "learning_rate": 5.727334068789529e-05, + "loss": 1.1485, + "step": 7493 + }, + { + "epoch": 0.45531320250318974, + "grad_norm": 0.27581578493118286, + "learning_rate": 5.7263869605945775e-05, + "loss": 1.0438, + "step": 7494 + }, + { + "epoch": 0.4553739595358163, + "grad_norm": 2.6000022888183594, + "learning_rate": 5.725439825773772e-05, + "loss": 1.0887, + "step": 7495 + }, + { + "epoch": 0.4554347165684428, + "grad_norm": 0.6515898704528809, + "learning_rate": 5.724492664361829e-05, + "loss": 1.0506, + "step": 7496 + }, + { + "epoch": 0.45549547360106935, + "grad_norm": 0.3576430082321167, + "learning_rate": 5.7235454763934657e-05, + "loss": 1.078, + "step": 7497 + }, + { + "epoch": 0.4555562306336958, + "grad_norm": 0.21200697124004364, + "learning_rate": 5.722598261903407e-05, + "loss": 1.0509, + "step": 7498 + }, + { + "epoch": 0.45561698766632236, + "grad_norm": 0.33290961384773254, + "learning_rate": 5.721651020926366e-05, + "loss": 1.0664, + "step": 7499 + }, + { + "epoch": 0.4556777446989489, + "grad_norm": 0.38332200050354004, + "learning_rate": 5.7207037534970686e-05, + "loss": 1.1415, + "step": 7500 + }, + { + "epoch": 0.45573850173157543, + "grad_norm": 0.43867045640945435, + "learning_rate": 5.719756459650237e-05, + "loss": 1.0889, + "step": 7501 + }, + { + "epoch": 0.45579925876420196, + "grad_norm": 0.34436342120170593, + "learning_rate": 5.718809139420591e-05, + "loss": 1.2425, + "step": 7502 + }, + { + "epoch": 0.4558600157968285, + "grad_norm": 0.7236154079437256, + "learning_rate": 5.717861792842858e-05, + "loss": 1.3504, + "step": 7503 + }, + { + "epoch": 0.45592077282945503, + "grad_norm": 0.3888036906719208, + "learning_rate": 5.716914419951762e-05, + "loss": 1.0574, + "step": 7504 + }, + { + "epoch": 0.4559815298620815, + "grad_norm": 0.2139088362455368, + "learning_rate": 5.7159670207820305e-05, + "loss": 1.0357, + "step": 7505 + }, + { + "epoch": 0.45604228689470805, + "grad_norm": 0.22628968954086304, + "learning_rate": 5.715019595368388e-05, + "loss": 1.0762, + "step": 7506 + }, + { + "epoch": 0.4561030439273346, + "grad_norm": 0.1528075486421585, + "learning_rate": 5.714072143745564e-05, + "loss": 1.0314, + "step": 7507 + }, + { + "epoch": 0.4561638009599611, + "grad_norm": 0.24585914611816406, + "learning_rate": 5.71312466594829e-05, + "loss": 1.1637, + "step": 7508 + }, + { + "epoch": 0.45622455799258765, + "grad_norm": 0.1709943562746048, + "learning_rate": 5.712177162011291e-05, + "loss": 1.0149, + "step": 7509 + }, + { + "epoch": 0.4562853150252142, + "grad_norm": 0.17621967196464539, + "learning_rate": 5.7112296319693004e-05, + "loss": 1.0565, + "step": 7510 + }, + { + "epoch": 0.45634607205784067, + "grad_norm": 0.640906035900116, + "learning_rate": 5.710282075857052e-05, + "loss": 1.1682, + "step": 7511 + }, + { + "epoch": 0.4564068290904672, + "grad_norm": 0.2556917369365692, + "learning_rate": 5.709334493709274e-05, + "loss": 1.021, + "step": 7512 + }, + { + "epoch": 0.45646758612309374, + "grad_norm": 0.3286280035972595, + "learning_rate": 5.7083868855607045e-05, + "loss": 1.0597, + "step": 7513 + }, + { + "epoch": 0.4565283431557203, + "grad_norm": 0.2545030117034912, + "learning_rate": 5.707439251446077e-05, + "loss": 1.0774, + "step": 7514 + }, + { + "epoch": 0.4565891001883468, + "grad_norm": 0.24428969621658325, + "learning_rate": 5.706491591400128e-05, + "loss": 1.0625, + "step": 7515 + }, + { + "epoch": 0.45664985722097334, + "grad_norm": 0.32242196798324585, + "learning_rate": 5.705543905457592e-05, + "loss": 1.152, + "step": 7516 + }, + { + "epoch": 0.4567106142535999, + "grad_norm": 0.269846647977829, + "learning_rate": 5.704596193653208e-05, + "loss": 1.1542, + "step": 7517 + }, + { + "epoch": 0.45677137128622636, + "grad_norm": 7.235185623168945, + "learning_rate": 5.703648456021714e-05, + "loss": 1.0554, + "step": 7518 + }, + { + "epoch": 0.4568321283188529, + "grad_norm": 0.4652956426143646, + "learning_rate": 5.702700692597851e-05, + "loss": 1.1809, + "step": 7519 + }, + { + "epoch": 0.4568928853514794, + "grad_norm": 0.3133206069469452, + "learning_rate": 5.701752903416356e-05, + "loss": 1.1176, + "step": 7520 + }, + { + "epoch": 0.45695364238410596, + "grad_norm": 0.313586562871933, + "learning_rate": 5.7008050885119754e-05, + "loss": 1.0712, + "step": 7521 + }, + { + "epoch": 0.4570143994167325, + "grad_norm": 0.5457965135574341, + "learning_rate": 5.699857247919447e-05, + "loss": 1.1377, + "step": 7522 + }, + { + "epoch": 0.45707515644935903, + "grad_norm": 0.4500719904899597, + "learning_rate": 5.698909381673516e-05, + "loss": 1.0508, + "step": 7523 + }, + { + "epoch": 0.45713591348198557, + "grad_norm": 0.19196027517318726, + "learning_rate": 5.697961489808927e-05, + "loss": 1.1031, + "step": 7524 + }, + { + "epoch": 0.45719667051461205, + "grad_norm": 0.42987099289894104, + "learning_rate": 5.697013572360424e-05, + "loss": 1.0904, + "step": 7525 + }, + { + "epoch": 0.4572574275472386, + "grad_norm": 0.8578873872756958, + "learning_rate": 5.696065629362755e-05, + "loss": 1.2499, + "step": 7526 + }, + { + "epoch": 0.4573181845798651, + "grad_norm": 0.40459948778152466, + "learning_rate": 5.695117660850665e-05, + "loss": 1.0491, + "step": 7527 + }, + { + "epoch": 0.45737894161249165, + "grad_norm": 0.2900027632713318, + "learning_rate": 5.694169666858903e-05, + "loss": 1.0492, + "step": 7528 + }, + { + "epoch": 0.4574396986451182, + "grad_norm": 0.4772969186306, + "learning_rate": 5.693221647422218e-05, + "loss": 1.0519, + "step": 7529 + }, + { + "epoch": 0.4575004556777447, + "grad_norm": 0.19082361459732056, + "learning_rate": 5.692273602575358e-05, + "loss": 1.0898, + "step": 7530 + }, + { + "epoch": 0.4575612127103712, + "grad_norm": 0.2132323831319809, + "learning_rate": 5.691325532353078e-05, + "loss": 1.0265, + "step": 7531 + }, + { + "epoch": 0.45762196974299774, + "grad_norm": 0.2564988434314728, + "learning_rate": 5.690377436790125e-05, + "loss": 1.15, + "step": 7532 + }, + { + "epoch": 0.45768272677562427, + "grad_norm": 0.1844327747821808, + "learning_rate": 5.689429315921253e-05, + "loss": 1.0428, + "step": 7533 + }, + { + "epoch": 0.4577434838082508, + "grad_norm": 0.26687854528427124, + "learning_rate": 5.688481169781218e-05, + "loss": 1.1461, + "step": 7534 + }, + { + "epoch": 0.45780424084087734, + "grad_norm": 0.2788293957710266, + "learning_rate": 5.687532998404771e-05, + "loss": 1.1519, + "step": 7535 + }, + { + "epoch": 0.4578649978735039, + "grad_norm": 0.1986551731824875, + "learning_rate": 5.68658480182667e-05, + "loss": 1.0773, + "step": 7536 + }, + { + "epoch": 0.4579257549061304, + "grad_norm": 0.15250664949417114, + "learning_rate": 5.68563658008167e-05, + "loss": 1.0351, + "step": 7537 + }, + { + "epoch": 0.4579865119387569, + "grad_norm": 2.257981300354004, + "learning_rate": 5.6846883332045284e-05, + "loss": 1.1185, + "step": 7538 + }, + { + "epoch": 0.4580472689713834, + "grad_norm": 0.24484288692474365, + "learning_rate": 5.683740061230005e-05, + "loss": 1.0431, + "step": 7539 + }, + { + "epoch": 0.45810802600400996, + "grad_norm": 0.2593284547328949, + "learning_rate": 5.6827917641928566e-05, + "loss": 1.1365, + "step": 7540 + }, + { + "epoch": 0.4581687830366365, + "grad_norm": 0.2856282591819763, + "learning_rate": 5.681843442127844e-05, + "loss": 1.162, + "step": 7541 + }, + { + "epoch": 0.45822954006926303, + "grad_norm": 0.19559557735919952, + "learning_rate": 5.680895095069728e-05, + "loss": 1.0445, + "step": 7542 + }, + { + "epoch": 0.45829029710188957, + "grad_norm": 0.30117544531822205, + "learning_rate": 5.67994672305327e-05, + "loss": 1.2411, + "step": 7543 + }, + { + "epoch": 0.45835105413451604, + "grad_norm": 0.21897847950458527, + "learning_rate": 5.678998326113236e-05, + "loss": 1.0962, + "step": 7544 + }, + { + "epoch": 0.4584118111671426, + "grad_norm": 0.2691689431667328, + "learning_rate": 5.678049904284385e-05, + "loss": 1.0596, + "step": 7545 + }, + { + "epoch": 0.4584725681997691, + "grad_norm": 0.1827954351902008, + "learning_rate": 5.6771014576014836e-05, + "loss": 1.0687, + "step": 7546 + }, + { + "epoch": 0.45853332523239565, + "grad_norm": 0.5373795628547668, + "learning_rate": 5.676152986099298e-05, + "loss": 1.2414, + "step": 7547 + }, + { + "epoch": 0.4585940822650222, + "grad_norm": 0.25555118918418884, + "learning_rate": 5.675204489812594e-05, + "loss": 1.0606, + "step": 7548 + }, + { + "epoch": 0.4586548392976487, + "grad_norm": 3.226719379425049, + "learning_rate": 5.674255968776139e-05, + "loss": 1.1138, + "step": 7549 + }, + { + "epoch": 0.45871559633027525, + "grad_norm": 0.6467882394790649, + "learning_rate": 5.673307423024701e-05, + "loss": 1.0434, + "step": 7550 + }, + { + "epoch": 0.45877635336290173, + "grad_norm": 0.21346168220043182, + "learning_rate": 5.6723588525930496e-05, + "loss": 1.1489, + "step": 7551 + }, + { + "epoch": 0.45883711039552827, + "grad_norm": 0.3944863975048065, + "learning_rate": 5.671410257515955e-05, + "loss": 1.282, + "step": 7552 + }, + { + "epoch": 0.4588978674281548, + "grad_norm": 0.889251708984375, + "learning_rate": 5.670461637828186e-05, + "loss": 1.0842, + "step": 7553 + }, + { + "epoch": 0.45895862446078134, + "grad_norm": 0.20962730050086975, + "learning_rate": 5.669512993564516e-05, + "loss": 1.1435, + "step": 7554 + }, + { + "epoch": 0.4590193814934079, + "grad_norm": 0.28050366044044495, + "learning_rate": 5.66856432475972e-05, + "loss": 1.1962, + "step": 7555 + }, + { + "epoch": 0.4590801385260344, + "grad_norm": 0.3554040491580963, + "learning_rate": 5.6676156314485665e-05, + "loss": 1.0125, + "step": 7556 + }, + { + "epoch": 0.4591408955586609, + "grad_norm": 0.2129548341035843, + "learning_rate": 5.6666669136658343e-05, + "loss": 1.0412, + "step": 7557 + }, + { + "epoch": 0.4592016525912874, + "grad_norm": 0.3159272074699402, + "learning_rate": 5.665718171446299e-05, + "loss": 1.0819, + "step": 7558 + }, + { + "epoch": 0.45926240962391396, + "grad_norm": 0.23154796659946442, + "learning_rate": 5.6647694048247335e-05, + "loss": 1.0858, + "step": 7559 + }, + { + "epoch": 0.4593231666565405, + "grad_norm": 0.28524935245513916, + "learning_rate": 5.6638206138359185e-05, + "loss": 1.2442, + "step": 7560 + }, + { + "epoch": 0.45938392368916703, + "grad_norm": 0.2034040093421936, + "learning_rate": 5.662871798514629e-05, + "loss": 1.0717, + "step": 7561 + }, + { + "epoch": 0.45944468072179356, + "grad_norm": 0.6708499789237976, + "learning_rate": 5.661922958895647e-05, + "loss": 1.2046, + "step": 7562 + }, + { + "epoch": 0.4595054377544201, + "grad_norm": 0.20701876282691956, + "learning_rate": 5.66097409501375e-05, + "loss": 1.0446, + "step": 7563 + }, + { + "epoch": 0.4595661947870466, + "grad_norm": 0.22035159170627594, + "learning_rate": 5.6600252069037206e-05, + "loss": 1.0289, + "step": 7564 + }, + { + "epoch": 0.4596269518196731, + "grad_norm": 0.28180405497550964, + "learning_rate": 5.65907629460034e-05, + "loss": 1.0859, + "step": 7565 + }, + { + "epoch": 0.45968770885229965, + "grad_norm": 0.2643662393093109, + "learning_rate": 5.65812735813839e-05, + "loss": 1.0958, + "step": 7566 + }, + { + "epoch": 0.4597484658849262, + "grad_norm": 0.17394453287124634, + "learning_rate": 5.6571783975526523e-05, + "loss": 1.108, + "step": 7567 + }, + { + "epoch": 0.4598092229175527, + "grad_norm": 0.18648993968963623, + "learning_rate": 5.656229412877917e-05, + "loss": 1.0394, + "step": 7568 + }, + { + "epoch": 0.45986997995017925, + "grad_norm": 0.21176445484161377, + "learning_rate": 5.655280404148963e-05, + "loss": 1.0508, + "step": 7569 + }, + { + "epoch": 0.45993073698280573, + "grad_norm": 0.24990487098693848, + "learning_rate": 5.65433137140058e-05, + "loss": 1.1106, + "step": 7570 + }, + { + "epoch": 0.45999149401543227, + "grad_norm": 0.22738945484161377, + "learning_rate": 5.6533823146675546e-05, + "loss": 1.0538, + "step": 7571 + }, + { + "epoch": 0.4600522510480588, + "grad_norm": 0.20849516987800598, + "learning_rate": 5.652433233984673e-05, + "loss": 1.0828, + "step": 7572 + }, + { + "epoch": 0.46011300808068534, + "grad_norm": 0.15324100852012634, + "learning_rate": 5.651484129386726e-05, + "loss": 1.0432, + "step": 7573 + }, + { + "epoch": 0.4601737651133119, + "grad_norm": 0.2635593116283417, + "learning_rate": 5.6505350009085024e-05, + "loss": 1.0036, + "step": 7574 + }, + { + "epoch": 0.4602345221459384, + "grad_norm": 0.7034757733345032, + "learning_rate": 5.649585848584793e-05, + "loss": 1.1751, + "step": 7575 + }, + { + "epoch": 0.46029527917856494, + "grad_norm": 0.27834203839302063, + "learning_rate": 5.648636672450388e-05, + "loss": 1.0722, + "step": 7576 + }, + { + "epoch": 0.4603560362111914, + "grad_norm": 0.25042808055877686, + "learning_rate": 5.6476874725400795e-05, + "loss": 1.1264, + "step": 7577 + }, + { + "epoch": 0.46041679324381796, + "grad_norm": 0.21791227161884308, + "learning_rate": 5.646738248888663e-05, + "loss": 1.0775, + "step": 7578 + }, + { + "epoch": 0.4604775502764445, + "grad_norm": 0.30885154008865356, + "learning_rate": 5.64578900153093e-05, + "loss": 1.258, + "step": 7579 + }, + { + "epoch": 0.460538307309071, + "grad_norm": 0.32601040601730347, + "learning_rate": 5.6448397305016766e-05, + "loss": 1.0429, + "step": 7580 + }, + { + "epoch": 0.46059906434169756, + "grad_norm": 0.8380441069602966, + "learning_rate": 5.6438904358357004e-05, + "loss": 1.0496, + "step": 7581 + }, + { + "epoch": 0.4606598213743241, + "grad_norm": 0.2596938908100128, + "learning_rate": 5.642941117567793e-05, + "loss": 1.0875, + "step": 7582 + }, + { + "epoch": 0.46072057840695063, + "grad_norm": 0.3099798262119293, + "learning_rate": 5.6419917757327555e-05, + "loss": 1.0741, + "step": 7583 + }, + { + "epoch": 0.4607813354395771, + "grad_norm": 0.3536452353000641, + "learning_rate": 5.6410424103653844e-05, + "loss": 1.1275, + "step": 7584 + }, + { + "epoch": 0.46084209247220365, + "grad_norm": 0.18998849391937256, + "learning_rate": 5.640093021500481e-05, + "loss": 1.0513, + "step": 7585 + }, + { + "epoch": 0.4609028495048302, + "grad_norm": 0.2061043381690979, + "learning_rate": 5.6391436091728446e-05, + "loss": 1.1378, + "step": 7586 + }, + { + "epoch": 0.4609636065374567, + "grad_norm": 0.2220660299062729, + "learning_rate": 5.638194173417274e-05, + "loss": 1.0378, + "step": 7587 + }, + { + "epoch": 0.46102436357008325, + "grad_norm": 3.2256340980529785, + "learning_rate": 5.637244714268575e-05, + "loss": 1.1368, + "step": 7588 + }, + { + "epoch": 0.4610851206027098, + "grad_norm": 0.21574297547340393, + "learning_rate": 5.636295231761547e-05, + "loss": 1.097, + "step": 7589 + }, + { + "epoch": 0.46114587763533627, + "grad_norm": 0.18217594921588898, + "learning_rate": 5.6353457259309925e-05, + "loss": 1.0705, + "step": 7590 + }, + { + "epoch": 0.4612066346679628, + "grad_norm": 0.1580817848443985, + "learning_rate": 5.6343961968117196e-05, + "loss": 1.0574, + "step": 7591 + }, + { + "epoch": 0.46126739170058934, + "grad_norm": 0.25099265575408936, + "learning_rate": 5.63344664443853e-05, + "loss": 1.0504, + "step": 7592 + }, + { + "epoch": 0.46132814873321587, + "grad_norm": 0.20136821269989014, + "learning_rate": 5.632497068846232e-05, + "loss": 1.0695, + "step": 7593 + }, + { + "epoch": 0.4613889057658424, + "grad_norm": 0.2960567772388458, + "learning_rate": 5.631547470069631e-05, + "loss": 1.0178, + "step": 7594 + }, + { + "epoch": 0.46144966279846894, + "grad_norm": 1.6741943359375, + "learning_rate": 5.630597848143537e-05, + "loss": 1.0705, + "step": 7595 + }, + { + "epoch": 0.4615104198310955, + "grad_norm": 2.326185703277588, + "learning_rate": 5.6296482031027565e-05, + "loss": 1.1659, + "step": 7596 + }, + { + "epoch": 0.46157117686372195, + "grad_norm": 0.3455941677093506, + "learning_rate": 5.6286985349820986e-05, + "loss": 1.1375, + "step": 7597 + }, + { + "epoch": 0.4616319338963485, + "grad_norm": 0.2876083254814148, + "learning_rate": 5.627748843816375e-05, + "loss": 1.1847, + "step": 7598 + }, + { + "epoch": 0.461692690928975, + "grad_norm": 0.5531802773475647, + "learning_rate": 5.626799129640395e-05, + "loss": 1.2097, + "step": 7599 + }, + { + "epoch": 0.46175344796160156, + "grad_norm": 0.298249751329422, + "learning_rate": 5.625849392488972e-05, + "loss": 1.2368, + "step": 7600 + }, + { + "epoch": 0.4618142049942281, + "grad_norm": 0.2267497479915619, + "learning_rate": 5.62489963239692e-05, + "loss": 1.1745, + "step": 7601 + }, + { + "epoch": 0.46187496202685463, + "grad_norm": 0.3270699381828308, + "learning_rate": 5.623949849399052e-05, + "loss": 1.014, + "step": 7602 + }, + { + "epoch": 0.4619357190594811, + "grad_norm": 0.1773277223110199, + "learning_rate": 5.623000043530179e-05, + "loss": 1.0578, + "step": 7603 + }, + { + "epoch": 0.46199647609210764, + "grad_norm": 0.2467549443244934, + "learning_rate": 5.6220502148251197e-05, + "loss": 1.1229, + "step": 7604 + }, + { + "epoch": 0.4620572331247342, + "grad_norm": 0.2550158202648163, + "learning_rate": 5.62110036331869e-05, + "loss": 1.0562, + "step": 7605 + }, + { + "epoch": 0.4621179901573607, + "grad_norm": 0.24993577599525452, + "learning_rate": 5.620150489045706e-05, + "loss": 1.1294, + "step": 7606 + }, + { + "epoch": 0.46217874718998725, + "grad_norm": 0.6363841891288757, + "learning_rate": 5.619200592040986e-05, + "loss": 1.0342, + "step": 7607 + }, + { + "epoch": 0.4622395042226138, + "grad_norm": 0.2022799253463745, + "learning_rate": 5.618250672339349e-05, + "loss": 1.0597, + "step": 7608 + }, + { + "epoch": 0.4623002612552403, + "grad_norm": 0.2077684998512268, + "learning_rate": 5.617300729975614e-05, + "loss": 1.117, + "step": 7609 + }, + { + "epoch": 0.4623610182878668, + "grad_norm": 0.20101527869701385, + "learning_rate": 5.6163507649846026e-05, + "loss": 1.0662, + "step": 7610 + }, + { + "epoch": 0.46242177532049333, + "grad_norm": 0.24869295954704285, + "learning_rate": 5.615400777401133e-05, + "loss": 1.1274, + "step": 7611 + }, + { + "epoch": 0.46248253235311987, + "grad_norm": 0.2294936627149582, + "learning_rate": 5.61445076726003e-05, + "loss": 1.139, + "step": 7612 + }, + { + "epoch": 0.4625432893857464, + "grad_norm": 0.1827412098646164, + "learning_rate": 5.613500734596113e-05, + "loss": 1.083, + "step": 7613 + }, + { + "epoch": 0.46260404641837294, + "grad_norm": 0.21870110929012299, + "learning_rate": 5.612550679444211e-05, + "loss": 1.0259, + "step": 7614 + }, + { + "epoch": 0.4626648034509995, + "grad_norm": 0.25700849294662476, + "learning_rate": 5.611600601839144e-05, + "loss": 1.2374, + "step": 7615 + }, + { + "epoch": 0.46272556048362595, + "grad_norm": 0.3086835741996765, + "learning_rate": 5.610650501815739e-05, + "loss": 1.0732, + "step": 7616 + }, + { + "epoch": 0.4627863175162525, + "grad_norm": 0.19841092824935913, + "learning_rate": 5.609700379408822e-05, + "loss": 1.0698, + "step": 7617 + }, + { + "epoch": 0.462847074548879, + "grad_norm": 0.42023199796676636, + "learning_rate": 5.60875023465322e-05, + "loss": 1.2194, + "step": 7618 + }, + { + "epoch": 0.46290783158150556, + "grad_norm": 0.3178770840167999, + "learning_rate": 5.6078000675837594e-05, + "loss": 1.0463, + "step": 7619 + }, + { + "epoch": 0.4629685886141321, + "grad_norm": 0.18007050454616547, + "learning_rate": 5.6068498782352707e-05, + "loss": 1.0615, + "step": 7620 + }, + { + "epoch": 0.4630293456467586, + "grad_norm": 0.4302619695663452, + "learning_rate": 5.605899666642582e-05, + "loss": 1.0386, + "step": 7621 + }, + { + "epoch": 0.46309010267938516, + "grad_norm": 0.22086502611637115, + "learning_rate": 5.604949432840525e-05, + "loss": 1.0386, + "step": 7622 + }, + { + "epoch": 0.46315085971201164, + "grad_norm": 0.24988168478012085, + "learning_rate": 5.6039991768639276e-05, + "loss": 1.1299, + "step": 7623 + }, + { + "epoch": 0.4632116167446382, + "grad_norm": 0.2916835844516754, + "learning_rate": 5.603048898747625e-05, + "loss": 1.1352, + "step": 7624 + }, + { + "epoch": 0.4632723737772647, + "grad_norm": 0.22294054925441742, + "learning_rate": 5.6020985985264486e-05, + "loss": 1.1004, + "step": 7625 + }, + { + "epoch": 0.46333313080989125, + "grad_norm": 0.21702668070793152, + "learning_rate": 5.601148276235231e-05, + "loss": 1.1274, + "step": 7626 + }, + { + "epoch": 0.4633938878425178, + "grad_norm": 0.2935025990009308, + "learning_rate": 5.600197931908807e-05, + "loss": 1.1458, + "step": 7627 + }, + { + "epoch": 0.4634546448751443, + "grad_norm": 0.19048158824443817, + "learning_rate": 5.599247565582012e-05, + "loss": 1.1037, + "step": 7628 + }, + { + "epoch": 0.46351540190777085, + "grad_norm": 1.4627227783203125, + "learning_rate": 5.598297177289682e-05, + "loss": 1.179, + "step": 7629 + }, + { + "epoch": 0.46357615894039733, + "grad_norm": 0.1989573836326599, + "learning_rate": 5.597346767066654e-05, + "loss": 1.0247, + "step": 7630 + }, + { + "epoch": 0.46363691597302387, + "grad_norm": 0.4160252511501312, + "learning_rate": 5.5963963349477636e-05, + "loss": 1.0716, + "step": 7631 + }, + { + "epoch": 0.4636976730056504, + "grad_norm": 0.2152736485004425, + "learning_rate": 5.59544588096785e-05, + "loss": 1.0719, + "step": 7632 + }, + { + "epoch": 0.46375843003827694, + "grad_norm": 0.44221436977386475, + "learning_rate": 5.594495405161754e-05, + "loss": 1.2015, + "step": 7633 + }, + { + "epoch": 0.46381918707090347, + "grad_norm": 0.4225747287273407, + "learning_rate": 5.593544907564312e-05, + "loss": 1.0873, + "step": 7634 + }, + { + "epoch": 0.46387994410353, + "grad_norm": 1.4651278257369995, + "learning_rate": 5.592594388210369e-05, + "loss": 1.0768, + "step": 7635 + }, + { + "epoch": 0.4639407011361565, + "grad_norm": 0.33938834071159363, + "learning_rate": 5.591643847134762e-05, + "loss": 1.0559, + "step": 7636 + }, + { + "epoch": 0.464001458168783, + "grad_norm": 0.23585732281208038, + "learning_rate": 5.590693284372336e-05, + "loss": 1.0887, + "step": 7637 + }, + { + "epoch": 0.46406221520140956, + "grad_norm": 0.17854046821594238, + "learning_rate": 5.589742699957935e-05, + "loss": 1.0743, + "step": 7638 + }, + { + "epoch": 0.4641229722340361, + "grad_norm": 0.2324364334344864, + "learning_rate": 5.588792093926399e-05, + "loss": 1.1019, + "step": 7639 + }, + { + "epoch": 0.4641837292666626, + "grad_norm": 0.36458832025527954, + "learning_rate": 5.5878414663125765e-05, + "loss": 1.1308, + "step": 7640 + }, + { + "epoch": 0.46424448629928916, + "grad_norm": 0.25196847319602966, + "learning_rate": 5.586890817151311e-05, + "loss": 1.0397, + "step": 7641 + }, + { + "epoch": 0.4643052433319157, + "grad_norm": 0.24359384179115295, + "learning_rate": 5.58594014647745e-05, + "loss": 1.0884, + "step": 7642 + }, + { + "epoch": 0.4643660003645422, + "grad_norm": 0.4225873053073883, + "learning_rate": 5.5849894543258396e-05, + "loss": 1.1787, + "step": 7643 + }, + { + "epoch": 0.4644267573971687, + "grad_norm": 0.2095779925584793, + "learning_rate": 5.5840387407313254e-05, + "loss": 1.03, + "step": 7644 + }, + { + "epoch": 0.46448751442979525, + "grad_norm": 0.21092787384986877, + "learning_rate": 5.5830880057287616e-05, + "loss": 1.0801, + "step": 7645 + }, + { + "epoch": 0.4645482714624218, + "grad_norm": 0.44746696949005127, + "learning_rate": 5.5821372493529924e-05, + "loss": 1.0027, + "step": 7646 + }, + { + "epoch": 0.4646090284950483, + "grad_norm": 0.22135557234287262, + "learning_rate": 5.581186471638868e-05, + "loss": 1.074, + "step": 7647 + }, + { + "epoch": 0.46466978552767485, + "grad_norm": 0.36693277955055237, + "learning_rate": 5.580235672621246e-05, + "loss": 1.1372, + "step": 7648 + }, + { + "epoch": 0.46473054256030133, + "grad_norm": 0.3321906626224518, + "learning_rate": 5.5792848523349684e-05, + "loss": 1.1874, + "step": 7649 + }, + { + "epoch": 0.46479129959292786, + "grad_norm": 0.19388987123966217, + "learning_rate": 5.578334010814895e-05, + "loss": 1.0818, + "step": 7650 + }, + { + "epoch": 0.4648520566255544, + "grad_norm": 0.44777408242225647, + "learning_rate": 5.577383148095876e-05, + "loss": 1.0718, + "step": 7651 + }, + { + "epoch": 0.46491281365818093, + "grad_norm": 0.1984647512435913, + "learning_rate": 5.5764322642127665e-05, + "loss": 1.1116, + "step": 7652 + }, + { + "epoch": 0.46497357069080747, + "grad_norm": 0.5073148608207703, + "learning_rate": 5.575481359200421e-05, + "loss": 1.174, + "step": 7653 + }, + { + "epoch": 0.465034327723434, + "grad_norm": 0.35818105936050415, + "learning_rate": 5.5745304330936945e-05, + "loss": 1.3209, + "step": 7654 + }, + { + "epoch": 0.46509508475606054, + "grad_norm": 0.2559344172477722, + "learning_rate": 5.5735794859274436e-05, + "loss": 1.01, + "step": 7655 + }, + { + "epoch": 0.465155841788687, + "grad_norm": 0.20523402094841003, + "learning_rate": 5.5726285177365265e-05, + "loss": 1.2053, + "step": 7656 + }, + { + "epoch": 0.46521659882131355, + "grad_norm": 0.20450018346309662, + "learning_rate": 5.5716775285558e-05, + "loss": 1.0743, + "step": 7657 + }, + { + "epoch": 0.4652773558539401, + "grad_norm": 0.2277001589536667, + "learning_rate": 5.570726518420124e-05, + "loss": 1.0999, + "step": 7658 + }, + { + "epoch": 0.4653381128865666, + "grad_norm": 0.29164692759513855, + "learning_rate": 5.569775487364357e-05, + "loss": 1.1295, + "step": 7659 + }, + { + "epoch": 0.46539886991919316, + "grad_norm": 0.1650746762752533, + "learning_rate": 5.5688244354233577e-05, + "loss": 1.0826, + "step": 7660 + }, + { + "epoch": 0.4654596269518197, + "grad_norm": 0.24266475439071655, + "learning_rate": 5.56787336263199e-05, + "loss": 1.0939, + "step": 7661 + }, + { + "epoch": 0.4655203839844462, + "grad_norm": 0.18860836327075958, + "learning_rate": 5.5669222690251146e-05, + "loss": 1.0571, + "step": 7662 + }, + { + "epoch": 0.4655811410170727, + "grad_norm": 0.18324267864227295, + "learning_rate": 5.5659711546375935e-05, + "loss": 1.0665, + "step": 7663 + }, + { + "epoch": 0.46564189804969924, + "grad_norm": 0.29037347435951233, + "learning_rate": 5.5650200195042914e-05, + "loss": 1.1407, + "step": 7664 + }, + { + "epoch": 0.4657026550823258, + "grad_norm": 0.4183213710784912, + "learning_rate": 5.56406886366007e-05, + "loss": 1.1691, + "step": 7665 + }, + { + "epoch": 0.4657634121149523, + "grad_norm": 0.34661880135536194, + "learning_rate": 5.563117687139796e-05, + "loss": 1.1954, + "step": 7666 + }, + { + "epoch": 0.46582416914757885, + "grad_norm": 0.198090061545372, + "learning_rate": 5.5621664899783345e-05, + "loss": 1.1319, + "step": 7667 + }, + { + "epoch": 0.4658849261802054, + "grad_norm": 0.21464644372463226, + "learning_rate": 5.561215272210551e-05, + "loss": 1.1038, + "step": 7668 + }, + { + "epoch": 0.46594568321283186, + "grad_norm": 0.30723217129707336, + "learning_rate": 5.560264033871313e-05, + "loss": 1.1879, + "step": 7669 + }, + { + "epoch": 0.4660064402454584, + "grad_norm": 0.49183136224746704, + "learning_rate": 5.559312774995489e-05, + "loss": 1.1351, + "step": 7670 + }, + { + "epoch": 0.46606719727808493, + "grad_norm": 0.1799955517053604, + "learning_rate": 5.5583614956179464e-05, + "loss": 1.0627, + "step": 7671 + }, + { + "epoch": 0.46612795431071147, + "grad_norm": 4.573728561401367, + "learning_rate": 5.5574101957735556e-05, + "loss": 1.0747, + "step": 7672 + }, + { + "epoch": 0.466188711343338, + "grad_norm": 0.18264299631118774, + "learning_rate": 5.556458875497187e-05, + "loss": 1.0492, + "step": 7673 + }, + { + "epoch": 0.46624946837596454, + "grad_norm": 0.18759137392044067, + "learning_rate": 5.555507534823711e-05, + "loss": 1.0664, + "step": 7674 + }, + { + "epoch": 0.466310225408591, + "grad_norm": 0.18537776172161102, + "learning_rate": 5.554556173787998e-05, + "loss": 1.0686, + "step": 7675 + }, + { + "epoch": 0.46637098244121755, + "grad_norm": 0.3475114703178406, + "learning_rate": 5.553604792424922e-05, + "loss": 1.0717, + "step": 7676 + }, + { + "epoch": 0.4664317394738441, + "grad_norm": 1.095594048500061, + "learning_rate": 5.552653390769356e-05, + "loss": 1.0563, + "step": 7677 + }, + { + "epoch": 0.4664924965064706, + "grad_norm": 0.6307897567749023, + "learning_rate": 5.551701968856171e-05, + "loss": 1.0499, + "step": 7678 + }, + { + "epoch": 0.46655325353909716, + "grad_norm": 0.272800087928772, + "learning_rate": 5.5507505267202456e-05, + "loss": 1.1306, + "step": 7679 + }, + { + "epoch": 0.4666140105717237, + "grad_norm": 0.190135657787323, + "learning_rate": 5.5497990643964505e-05, + "loss": 1.1253, + "step": 7680 + }, + { + "epoch": 0.4666747676043502, + "grad_norm": 0.37058550119400024, + "learning_rate": 5.5488475819196674e-05, + "loss": 1.1407, + "step": 7681 + }, + { + "epoch": 0.4667355246369767, + "grad_norm": 0.17635153234004974, + "learning_rate": 5.547896079324769e-05, + "loss": 1.0384, + "step": 7682 + }, + { + "epoch": 0.46679628166960324, + "grad_norm": 0.37154144048690796, + "learning_rate": 5.546944556646634e-05, + "loss": 1.0488, + "step": 7683 + }, + { + "epoch": 0.4668570387022298, + "grad_norm": 0.16774867475032806, + "learning_rate": 5.54599301392014e-05, + "loss": 1.0499, + "step": 7684 + }, + { + "epoch": 0.4669177957348563, + "grad_norm": 0.9479286670684814, + "learning_rate": 5.545041451180167e-05, + "loss": 1.1242, + "step": 7685 + }, + { + "epoch": 0.46697855276748285, + "grad_norm": 0.30662405490875244, + "learning_rate": 5.544089868461594e-05, + "loss": 1.0707, + "step": 7686 + }, + { + "epoch": 0.4670393098001094, + "grad_norm": 0.22174064815044403, + "learning_rate": 5.5431382657993025e-05, + "loss": 1.1581, + "step": 7687 + }, + { + "epoch": 0.4671000668327359, + "grad_norm": 0.2018086016178131, + "learning_rate": 5.542186643228172e-05, + "loss": 1.061, + "step": 7688 + }, + { + "epoch": 0.4671608238653624, + "grad_norm": 0.22511062026023865, + "learning_rate": 5.541235000783086e-05, + "loss": 1.0867, + "step": 7689 + }, + { + "epoch": 0.46722158089798893, + "grad_norm": 0.2326614260673523, + "learning_rate": 5.540283338498926e-05, + "loss": 1.0807, + "step": 7690 + }, + { + "epoch": 0.46728233793061547, + "grad_norm": 0.23883451521396637, + "learning_rate": 5.539331656410575e-05, + "loss": 1.0459, + "step": 7691 + }, + { + "epoch": 0.467343094963242, + "grad_norm": 0.33507469296455383, + "learning_rate": 5.5383799545529205e-05, + "loss": 1.0789, + "step": 7692 + }, + { + "epoch": 0.46740385199586854, + "grad_norm": 0.20195689797401428, + "learning_rate": 5.5374282329608404e-05, + "loss": 1.0183, + "step": 7693 + }, + { + "epoch": 0.46746460902849507, + "grad_norm": 0.4699666202068329, + "learning_rate": 5.5364764916692277e-05, + "loss": 1.1584, + "step": 7694 + }, + { + "epoch": 0.46752536606112155, + "grad_norm": 0.2054796665906906, + "learning_rate": 5.5355247307129664e-05, + "loss": 1.0693, + "step": 7695 + }, + { + "epoch": 0.4675861230937481, + "grad_norm": 0.25129929184913635, + "learning_rate": 5.5345729501269396e-05, + "loss": 1.2148, + "step": 7696 + }, + { + "epoch": 0.4676468801263746, + "grad_norm": 0.3782368004322052, + "learning_rate": 5.53362114994604e-05, + "loss": 1.0663, + "step": 7697 + }, + { + "epoch": 0.46770763715900115, + "grad_norm": 0.22399115562438965, + "learning_rate": 5.5326693302051537e-05, + "loss": 1.0572, + "step": 7698 + }, + { + "epoch": 0.4677683941916277, + "grad_norm": 0.1674412488937378, + "learning_rate": 5.531717490939169e-05, + "loss": 1.0555, + "step": 7699 + }, + { + "epoch": 0.4678291512242542, + "grad_norm": 0.19757743179798126, + "learning_rate": 5.530765632182978e-05, + "loss": 1.1174, + "step": 7700 + }, + { + "epoch": 0.46788990825688076, + "grad_norm": 0.3858512341976166, + "learning_rate": 5.529813753971469e-05, + "loss": 1.1597, + "step": 7701 + }, + { + "epoch": 0.46795066528950724, + "grad_norm": 0.20808576047420502, + "learning_rate": 5.5288618563395366e-05, + "loss": 1.0432, + "step": 7702 + }, + { + "epoch": 0.4680114223221338, + "grad_norm": 0.20657944679260254, + "learning_rate": 5.527909939322069e-05, + "loss": 1.0488, + "step": 7703 + }, + { + "epoch": 0.4680721793547603, + "grad_norm": 0.29501286149024963, + "learning_rate": 5.5269580029539605e-05, + "loss": 1.1164, + "step": 7704 + }, + { + "epoch": 0.46813293638738684, + "grad_norm": 0.20441164076328278, + "learning_rate": 5.526006047270106e-05, + "loss": 1.0772, + "step": 7705 + }, + { + "epoch": 0.4681936934200134, + "grad_norm": 0.2311316728591919, + "learning_rate": 5.525054072305395e-05, + "loss": 1.0474, + "step": 7706 + }, + { + "epoch": 0.4682544504526399, + "grad_norm": 0.4041880667209625, + "learning_rate": 5.5241020780947284e-05, + "loss": 1.0077, + "step": 7707 + }, + { + "epoch": 0.4683152074852664, + "grad_norm": 0.31208887696266174, + "learning_rate": 5.523150064672999e-05, + "loss": 1.2131, + "step": 7708 + }, + { + "epoch": 0.46837596451789293, + "grad_norm": 0.20872093737125397, + "learning_rate": 5.522198032075101e-05, + "loss": 1.0719, + "step": 7709 + }, + { + "epoch": 0.46843672155051946, + "grad_norm": 0.33208340406417847, + "learning_rate": 5.521245980335934e-05, + "loss": 1.0578, + "step": 7710 + }, + { + "epoch": 0.468497478583146, + "grad_norm": 0.17058779299259186, + "learning_rate": 5.520293909490396e-05, + "loss": 1.061, + "step": 7711 + }, + { + "epoch": 0.46855823561577253, + "grad_norm": 0.20519152283668518, + "learning_rate": 5.5193418195733825e-05, + "loss": 1.0947, + "step": 7712 + }, + { + "epoch": 0.46861899264839907, + "grad_norm": 0.31395551562309265, + "learning_rate": 5.5183897106197945e-05, + "loss": 1.1705, + "step": 7713 + }, + { + "epoch": 0.4686797496810256, + "grad_norm": 0.456880122423172, + "learning_rate": 5.517437582664531e-05, + "loss": 1.0784, + "step": 7714 + }, + { + "epoch": 0.4687405067136521, + "grad_norm": 0.18932782113552094, + "learning_rate": 5.516485435742496e-05, + "loss": 1.1595, + "step": 7715 + }, + { + "epoch": 0.4688012637462786, + "grad_norm": 0.37593695521354675, + "learning_rate": 5.515533269888583e-05, + "loss": 1.0062, + "step": 7716 + }, + { + "epoch": 0.46886202077890515, + "grad_norm": 0.24021102488040924, + "learning_rate": 5.514581085137701e-05, + "loss": 1.0815, + "step": 7717 + }, + { + "epoch": 0.4689227778115317, + "grad_norm": 0.17040550708770752, + "learning_rate": 5.51362888152475e-05, + "loss": 1.0756, + "step": 7718 + }, + { + "epoch": 0.4689835348441582, + "grad_norm": 0.3467976450920105, + "learning_rate": 5.512676659084631e-05, + "loss": 1.0739, + "step": 7719 + }, + { + "epoch": 0.46904429187678476, + "grad_norm": 0.2088669389486313, + "learning_rate": 5.511724417852252e-05, + "loss": 1.1374, + "step": 7720 + }, + { + "epoch": 0.46910504890941124, + "grad_norm": 1.1059761047363281, + "learning_rate": 5.510772157862515e-05, + "loss": 1.0778, + "step": 7721 + }, + { + "epoch": 0.4691658059420378, + "grad_norm": 1.0863369703292847, + "learning_rate": 5.5098198791503265e-05, + "loss": 1.0299, + "step": 7722 + }, + { + "epoch": 0.4692265629746643, + "grad_norm": 1.8171181678771973, + "learning_rate": 5.5088675817505905e-05, + "loss": 1.1278, + "step": 7723 + }, + { + "epoch": 0.46928732000729084, + "grad_norm": 0.47837790846824646, + "learning_rate": 5.507915265698215e-05, + "loss": 1.0923, + "step": 7724 + }, + { + "epoch": 0.4693480770399174, + "grad_norm": 0.3408852517604828, + "learning_rate": 5.506962931028109e-05, + "loss": 1.0945, + "step": 7725 + }, + { + "epoch": 0.4694088340725439, + "grad_norm": 0.17354249954223633, + "learning_rate": 5.506010577775177e-05, + "loss": 1.0556, + "step": 7726 + }, + { + "epoch": 0.46946959110517045, + "grad_norm": 0.22707052528858185, + "learning_rate": 5.5050582059743296e-05, + "loss": 1.1182, + "step": 7727 + }, + { + "epoch": 0.4695303481377969, + "grad_norm": 0.18686246871948242, + "learning_rate": 5.5041058156604766e-05, + "loss": 1.0486, + "step": 7728 + }, + { + "epoch": 0.46959110517042346, + "grad_norm": 0.277308851480484, + "learning_rate": 5.503153406868527e-05, + "loss": 1.1801, + "step": 7729 + }, + { + "epoch": 0.46965186220305, + "grad_norm": 0.16434863209724426, + "learning_rate": 5.502200979633392e-05, + "loss": 1.0116, + "step": 7730 + }, + { + "epoch": 0.46971261923567653, + "grad_norm": 0.18023914098739624, + "learning_rate": 5.5012485339899835e-05, + "loss": 1.0992, + "step": 7731 + }, + { + "epoch": 0.46977337626830307, + "grad_norm": 0.23739442229270935, + "learning_rate": 5.5002960699732135e-05, + "loss": 1.1705, + "step": 7732 + }, + { + "epoch": 0.4698341333009296, + "grad_norm": 0.2114521563053131, + "learning_rate": 5.499343587617993e-05, + "loss": 1.0961, + "step": 7733 + }, + { + "epoch": 0.46989489033355614, + "grad_norm": 0.1890164464712143, + "learning_rate": 5.4983910869592384e-05, + "loss": 1.0898, + "step": 7734 + }, + { + "epoch": 0.4699556473661826, + "grad_norm": 0.17923152446746826, + "learning_rate": 5.497438568031861e-05, + "loss": 1.0141, + "step": 7735 + }, + { + "epoch": 0.47001640439880915, + "grad_norm": 0.42414745688438416, + "learning_rate": 5.496486030870777e-05, + "loss": 1.1367, + "step": 7736 + }, + { + "epoch": 0.4700771614314357, + "grad_norm": 0.19310948252677917, + "learning_rate": 5.4955334755109014e-05, + "loss": 1.0948, + "step": 7737 + }, + { + "epoch": 0.4701379184640622, + "grad_norm": 0.2536373734474182, + "learning_rate": 5.49458090198715e-05, + "loss": 1.0185, + "step": 7738 + }, + { + "epoch": 0.47019867549668876, + "grad_norm": 1.459326982498169, + "learning_rate": 5.493628310334442e-05, + "loss": 1.2247, + "step": 7739 + }, + { + "epoch": 0.4702594325293153, + "grad_norm": 0.24550150334835052, + "learning_rate": 5.4926757005876904e-05, + "loss": 1.1075, + "step": 7740 + }, + { + "epoch": 0.47032018956194177, + "grad_norm": 0.3844994604587555, + "learning_rate": 5.491723072781817e-05, + "loss": 1.136, + "step": 7741 + }, + { + "epoch": 0.4703809465945683, + "grad_norm": 0.24745212495326996, + "learning_rate": 5.4907704269517394e-05, + "loss": 1.1117, + "step": 7742 + }, + { + "epoch": 0.47044170362719484, + "grad_norm": 0.20928412675857544, + "learning_rate": 5.489817763132377e-05, + "loss": 1.108, + "step": 7743 + }, + { + "epoch": 0.4705024606598214, + "grad_norm": 0.2411159873008728, + "learning_rate": 5.4888650813586494e-05, + "loss": 1.1067, + "step": 7744 + }, + { + "epoch": 0.4705632176924479, + "grad_norm": 0.4837057292461395, + "learning_rate": 5.4879123816654786e-05, + "loss": 1.1648, + "step": 7745 + }, + { + "epoch": 0.47062397472507445, + "grad_norm": 0.3472537398338318, + "learning_rate": 5.486959664087784e-05, + "loss": 1.0803, + "step": 7746 + }, + { + "epoch": 0.470684731757701, + "grad_norm": 0.3306402862071991, + "learning_rate": 5.4860069286604896e-05, + "loss": 1.0154, + "step": 7747 + }, + { + "epoch": 0.47074548879032746, + "grad_norm": 0.2244849056005478, + "learning_rate": 5.485054175418516e-05, + "loss": 1.0233, + "step": 7748 + }, + { + "epoch": 0.470806245822954, + "grad_norm": 0.22034361958503723, + "learning_rate": 5.48410140439679e-05, + "loss": 1.0861, + "step": 7749 + }, + { + "epoch": 0.47086700285558053, + "grad_norm": 0.23178568482398987, + "learning_rate": 5.483148615630231e-05, + "loss": 1.1683, + "step": 7750 + }, + { + "epoch": 0.47092775988820706, + "grad_norm": 0.1639622002840042, + "learning_rate": 5.4821958091537675e-05, + "loss": 1.0644, + "step": 7751 + }, + { + "epoch": 0.4709885169208336, + "grad_norm": 0.2664664089679718, + "learning_rate": 5.481242985002325e-05, + "loss": 1.1444, + "step": 7752 + }, + { + "epoch": 0.47104927395346013, + "grad_norm": 0.23814383149147034, + "learning_rate": 5.480290143210825e-05, + "loss": 1.1395, + "step": 7753 + }, + { + "epoch": 0.4711100309860866, + "grad_norm": 0.3044374883174896, + "learning_rate": 5.4793372838141974e-05, + "loss": 1.0543, + "step": 7754 + }, + { + "epoch": 0.47117078801871315, + "grad_norm": 0.30997759103775024, + "learning_rate": 5.47838440684737e-05, + "loss": 1.193, + "step": 7755 + }, + { + "epoch": 0.4712315450513397, + "grad_norm": 0.2885420024394989, + "learning_rate": 5.4774315123452703e-05, + "loss": 1.1926, + "step": 7756 + }, + { + "epoch": 0.4712923020839662, + "grad_norm": 0.218947172164917, + "learning_rate": 5.476478600342825e-05, + "loss": 1.0548, + "step": 7757 + }, + { + "epoch": 0.47135305911659275, + "grad_norm": 0.29494941234588623, + "learning_rate": 5.4755256708749635e-05, + "loss": 1.0934, + "step": 7758 + }, + { + "epoch": 0.4714138161492193, + "grad_norm": 0.22374698519706726, + "learning_rate": 5.4745727239766195e-05, + "loss": 1.0637, + "step": 7759 + }, + { + "epoch": 0.4714745731818458, + "grad_norm": 0.16914936900138855, + "learning_rate": 5.4736197596827185e-05, + "loss": 1.054, + "step": 7760 + }, + { + "epoch": 0.4715353302144723, + "grad_norm": 0.2004675269126892, + "learning_rate": 5.4726667780281935e-05, + "loss": 1.1143, + "step": 7761 + }, + { + "epoch": 0.47159608724709884, + "grad_norm": 0.23474137485027313, + "learning_rate": 5.471713779047979e-05, + "loss": 1.0678, + "step": 7762 + }, + { + "epoch": 0.4716568442797254, + "grad_norm": 2.1227939128875732, + "learning_rate": 5.470760762777001e-05, + "loss": 1.112, + "step": 7763 + }, + { + "epoch": 0.4717176013123519, + "grad_norm": 0.2037692368030548, + "learning_rate": 5.469807729250198e-05, + "loss": 1.0661, + "step": 7764 + }, + { + "epoch": 0.47177835834497844, + "grad_norm": 0.2998400628566742, + "learning_rate": 5.468854678502503e-05, + "loss": 1.1151, + "step": 7765 + }, + { + "epoch": 0.471839115377605, + "grad_norm": 0.342376172542572, + "learning_rate": 5.4679016105688475e-05, + "loss": 1.0344, + "step": 7766 + }, + { + "epoch": 0.47189987241023146, + "grad_norm": 0.4901200532913208, + "learning_rate": 5.466948525484169e-05, + "loss": 1.053, + "step": 7767 + }, + { + "epoch": 0.471960629442858, + "grad_norm": 0.20413914322853088, + "learning_rate": 5.465995423283401e-05, + "loss": 1.1101, + "step": 7768 + }, + { + "epoch": 0.47202138647548453, + "grad_norm": 0.2600811719894409, + "learning_rate": 5.465042304001482e-05, + "loss": 1.1723, + "step": 7769 + }, + { + "epoch": 0.47208214350811106, + "grad_norm": 0.7267048358917236, + "learning_rate": 5.4640891676733466e-05, + "loss": 1.1537, + "step": 7770 + }, + { + "epoch": 0.4721429005407376, + "grad_norm": 0.2122507244348526, + "learning_rate": 5.463136014333933e-05, + "loss": 1.1439, + "step": 7771 + }, + { + "epoch": 0.47220365757336413, + "grad_norm": 0.2675290107727051, + "learning_rate": 5.46218284401818e-05, + "loss": 1.0424, + "step": 7772 + }, + { + "epoch": 0.47226441460599067, + "grad_norm": 0.46526435017585754, + "learning_rate": 5.4612296567610244e-05, + "loss": 1.1332, + "step": 7773 + }, + { + "epoch": 0.47232517163861715, + "grad_norm": 0.20805221796035767, + "learning_rate": 5.4602764525974074e-05, + "loss": 1.1375, + "step": 7774 + }, + { + "epoch": 0.4723859286712437, + "grad_norm": 0.21143968403339386, + "learning_rate": 5.459323231562269e-05, + "loss": 1.0874, + "step": 7775 + }, + { + "epoch": 0.4724466857038702, + "grad_norm": 0.23523233830928802, + "learning_rate": 5.458369993690546e-05, + "loss": 1.1074, + "step": 7776 + }, + { + "epoch": 0.47250744273649675, + "grad_norm": 0.2994513213634491, + "learning_rate": 5.4574167390171836e-05, + "loss": 1.1226, + "step": 7777 + }, + { + "epoch": 0.4725681997691233, + "grad_norm": 0.18978257477283478, + "learning_rate": 5.456463467577123e-05, + "loss": 1.0449, + "step": 7778 + }, + { + "epoch": 0.4726289568017498, + "grad_norm": 0.2601977586746216, + "learning_rate": 5.4555101794053054e-05, + "loss": 1.1276, + "step": 7779 + }, + { + "epoch": 0.4726897138343763, + "grad_norm": 0.19588837027549744, + "learning_rate": 5.4545568745366734e-05, + "loss": 1.0557, + "step": 7780 + }, + { + "epoch": 0.47275047086700284, + "grad_norm": 0.30455899238586426, + "learning_rate": 5.4536035530061716e-05, + "loss": 1.0827, + "step": 7781 + }, + { + "epoch": 0.47281122789962937, + "grad_norm": 0.16870777308940887, + "learning_rate": 5.452650214848746e-05, + "loss": 0.9718, + "step": 7782 + }, + { + "epoch": 0.4728719849322559, + "grad_norm": 1.0959722995758057, + "learning_rate": 5.451696860099337e-05, + "loss": 1.0684, + "step": 7783 + }, + { + "epoch": 0.47293274196488244, + "grad_norm": 0.19158869981765747, + "learning_rate": 5.450743488792892e-05, + "loss": 1.1778, + "step": 7784 + }, + { + "epoch": 0.472993498997509, + "grad_norm": 0.1884506791830063, + "learning_rate": 5.4497901009643594e-05, + "loss": 1.0506, + "step": 7785 + }, + { + "epoch": 0.4730542560301355, + "grad_norm": 0.24310314655303955, + "learning_rate": 5.4488366966486804e-05, + "loss": 1.2111, + "step": 7786 + }, + { + "epoch": 0.473115013062762, + "grad_norm": 0.15878768265247345, + "learning_rate": 5.447883275880809e-05, + "loss": 1.0577, + "step": 7787 + }, + { + "epoch": 0.4731757700953885, + "grad_norm": 0.2634182870388031, + "learning_rate": 5.446929838695687e-05, + "loss": 1.2852, + "step": 7788 + }, + { + "epoch": 0.47323652712801506, + "grad_norm": 0.17359024286270142, + "learning_rate": 5.445976385128266e-05, + "loss": 1.032, + "step": 7789 + }, + { + "epoch": 0.4732972841606416, + "grad_norm": 0.1711236834526062, + "learning_rate": 5.445022915213496e-05, + "loss": 1.1343, + "step": 7790 + }, + { + "epoch": 0.47335804119326813, + "grad_norm": 0.30514198541641235, + "learning_rate": 5.444069428986324e-05, + "loss": 1.0568, + "step": 7791 + }, + { + "epoch": 0.47341879822589467, + "grad_norm": 0.20935852825641632, + "learning_rate": 5.4431159264817e-05, + "loss": 1.0469, + "step": 7792 + }, + { + "epoch": 0.4734795552585212, + "grad_norm": 0.19876043498516083, + "learning_rate": 5.442162407734578e-05, + "loss": 1.0778, + "step": 7793 + }, + { + "epoch": 0.4735403122911477, + "grad_norm": 0.24242311716079712, + "learning_rate": 5.441208872779905e-05, + "loss": 1.1314, + "step": 7794 + }, + { + "epoch": 0.4736010693237742, + "grad_norm": 0.1591338813304901, + "learning_rate": 5.440255321652637e-05, + "loss": 1.0152, + "step": 7795 + }, + { + "epoch": 0.47366182635640075, + "grad_norm": 0.29040589928627014, + "learning_rate": 5.4393017543877266e-05, + "loss": 1.1594, + "step": 7796 + }, + { + "epoch": 0.4737225833890273, + "grad_norm": 0.213766947388649, + "learning_rate": 5.438348171020123e-05, + "loss": 1.0947, + "step": 7797 + }, + { + "epoch": 0.4737833404216538, + "grad_norm": 0.17584848403930664, + "learning_rate": 5.437394571584784e-05, + "loss": 1.0801, + "step": 7798 + }, + { + "epoch": 0.47384409745428036, + "grad_norm": 0.7073409557342529, + "learning_rate": 5.436440956116662e-05, + "loss": 1.3632, + "step": 7799 + }, + { + "epoch": 0.47390485448690683, + "grad_norm": 0.20954829454421997, + "learning_rate": 5.435487324650712e-05, + "loss": 1.0561, + "step": 7800 + }, + { + "epoch": 0.47396561151953337, + "grad_norm": 0.3228541612625122, + "learning_rate": 5.43453367722189e-05, + "loss": 1.0519, + "step": 7801 + }, + { + "epoch": 0.4740263685521599, + "grad_norm": 0.21747608482837677, + "learning_rate": 5.433580013865153e-05, + "loss": 1.0395, + "step": 7802 + }, + { + "epoch": 0.47408712558478644, + "grad_norm": 0.1869489550590515, + "learning_rate": 5.432626334615456e-05, + "loss": 1.0954, + "step": 7803 + }, + { + "epoch": 0.474147882617413, + "grad_norm": 0.29950642585754395, + "learning_rate": 5.431672639507758e-05, + "loss": 1.0835, + "step": 7804 + }, + { + "epoch": 0.4742086396500395, + "grad_norm": 0.15465764701366425, + "learning_rate": 5.430718928577013e-05, + "loss": 1.0591, + "step": 7805 + }, + { + "epoch": 0.47426939668266604, + "grad_norm": 0.19578099250793457, + "learning_rate": 5.429765201858186e-05, + "loss": 1.0376, + "step": 7806 + }, + { + "epoch": 0.4743301537152925, + "grad_norm": 0.19985279440879822, + "learning_rate": 5.4288114593862305e-05, + "loss": 1.1542, + "step": 7807 + }, + { + "epoch": 0.47439091074791906, + "grad_norm": 0.18950261175632477, + "learning_rate": 5.4278577011961084e-05, + "loss": 1.0616, + "step": 7808 + }, + { + "epoch": 0.4744516677805456, + "grad_norm": 0.20801934599876404, + "learning_rate": 5.4269039273227796e-05, + "loss": 1.1511, + "step": 7809 + }, + { + "epoch": 0.47451242481317213, + "grad_norm": 10.029825210571289, + "learning_rate": 5.4259501378012055e-05, + "loss": 1.0551, + "step": 7810 + }, + { + "epoch": 0.47457318184579866, + "grad_norm": 0.19913385808467865, + "learning_rate": 5.424996332666347e-05, + "loss": 1.1989, + "step": 7811 + }, + { + "epoch": 0.4746339388784252, + "grad_norm": 0.16805100440979004, + "learning_rate": 5.424042511953165e-05, + "loss": 1.0016, + "step": 7812 + }, + { + "epoch": 0.4746946959110517, + "grad_norm": 0.1698824167251587, + "learning_rate": 5.4230886756966224e-05, + "loss": 1.0809, + "step": 7813 + }, + { + "epoch": 0.4747554529436782, + "grad_norm": 0.1393348127603531, + "learning_rate": 5.4221348239316826e-05, + "loss": 1.014, + "step": 7814 + }, + { + "epoch": 0.47481620997630475, + "grad_norm": 0.1568973809480667, + "learning_rate": 5.421180956693309e-05, + "loss": 1.0708, + "step": 7815 + }, + { + "epoch": 0.4748769670089313, + "grad_norm": 0.25952744483947754, + "learning_rate": 5.4202270740164676e-05, + "loss": 1.0871, + "step": 7816 + }, + { + "epoch": 0.4749377240415578, + "grad_norm": 0.19380545616149902, + "learning_rate": 5.4192731759361194e-05, + "loss": 1.1282, + "step": 7817 + }, + { + "epoch": 0.47499848107418435, + "grad_norm": 0.2848016321659088, + "learning_rate": 5.418319262487234e-05, + "loss": 1.0987, + "step": 7818 + }, + { + "epoch": 0.4750592381068109, + "grad_norm": 0.1970958709716797, + "learning_rate": 5.417365333704775e-05, + "loss": 1.0458, + "step": 7819 + }, + { + "epoch": 0.47511999513943737, + "grad_norm": 0.3210126459598541, + "learning_rate": 5.4164113896237065e-05, + "loss": 1.213, + "step": 7820 + }, + { + "epoch": 0.4751807521720639, + "grad_norm": 0.19475558400154114, + "learning_rate": 5.4154574302790006e-05, + "loss": 1.0498, + "step": 7821 + }, + { + "epoch": 0.47524150920469044, + "grad_norm": 0.17350833117961884, + "learning_rate": 5.4145034557056216e-05, + "loss": 1.0129, + "step": 7822 + }, + { + "epoch": 0.475302266237317, + "grad_norm": 0.21677222847938538, + "learning_rate": 5.4135494659385375e-05, + "loss": 1.0869, + "step": 7823 + }, + { + "epoch": 0.4753630232699435, + "grad_norm": 0.22733552753925323, + "learning_rate": 5.4125954610127194e-05, + "loss": 1.0816, + "step": 7824 + }, + { + "epoch": 0.47542378030257004, + "grad_norm": 0.18232934176921844, + "learning_rate": 5.4116414409631334e-05, + "loss": 1.117, + "step": 7825 + }, + { + "epoch": 0.4754845373351965, + "grad_norm": 0.18153983354568481, + "learning_rate": 5.4106874058247526e-05, + "loss": 1.0698, + "step": 7826 + }, + { + "epoch": 0.47554529436782306, + "grad_norm": 0.397493451833725, + "learning_rate": 5.409733355632544e-05, + "loss": 1.1118, + "step": 7827 + }, + { + "epoch": 0.4756060514004496, + "grad_norm": 0.18194492161273956, + "learning_rate": 5.40877929042148e-05, + "loss": 1.0677, + "step": 7828 + }, + { + "epoch": 0.4756668084330761, + "grad_norm": 0.32024818658828735, + "learning_rate": 5.407825210226534e-05, + "loss": 1.0804, + "step": 7829 + }, + { + "epoch": 0.47572756546570266, + "grad_norm": 0.7912682890892029, + "learning_rate": 5.406871115082672e-05, + "loss": 1.1365, + "step": 7830 + }, + { + "epoch": 0.4757883224983292, + "grad_norm": 0.22342143952846527, + "learning_rate": 5.405917005024874e-05, + "loss": 1.1383, + "step": 7831 + }, + { + "epoch": 0.47584907953095573, + "grad_norm": 0.15908388793468475, + "learning_rate": 5.4049628800881105e-05, + "loss": 1.0193, + "step": 7832 + }, + { + "epoch": 0.4759098365635822, + "grad_norm": 0.23341946303844452, + "learning_rate": 5.404008740307351e-05, + "loss": 1.0452, + "step": 7833 + }, + { + "epoch": 0.47597059359620875, + "grad_norm": 0.21023912727832794, + "learning_rate": 5.4030545857175754e-05, + "loss": 1.0898, + "step": 7834 + }, + { + "epoch": 0.4760313506288353, + "grad_norm": 0.21621491014957428, + "learning_rate": 5.402100416353756e-05, + "loss": 1.1054, + "step": 7835 + }, + { + "epoch": 0.4760921076614618, + "grad_norm": 2.251018762588501, + "learning_rate": 5.4011462322508674e-05, + "loss": 1.0566, + "step": 7836 + }, + { + "epoch": 0.47615286469408835, + "grad_norm": 0.16898095607757568, + "learning_rate": 5.400192033443886e-05, + "loss": 1.0483, + "step": 7837 + }, + { + "epoch": 0.4762136217267149, + "grad_norm": 0.7878841161727905, + "learning_rate": 5.399237819967788e-05, + "loss": 1.0475, + "step": 7838 + }, + { + "epoch": 0.4762743787593414, + "grad_norm": 0.2765450179576874, + "learning_rate": 5.398283591857552e-05, + "loss": 1.0852, + "step": 7839 + }, + { + "epoch": 0.4763351357919679, + "grad_norm": 0.2913093566894531, + "learning_rate": 5.397329349148154e-05, + "loss": 1.1516, + "step": 7840 + }, + { + "epoch": 0.47639589282459444, + "grad_norm": 0.1654752641916275, + "learning_rate": 5.3963750918745694e-05, + "loss": 1.0111, + "step": 7841 + }, + { + "epoch": 0.47645664985722097, + "grad_norm": 0.28692102432250977, + "learning_rate": 5.3954208200717814e-05, + "loss": 1.2241, + "step": 7842 + }, + { + "epoch": 0.4765174068898475, + "grad_norm": 0.19560185074806213, + "learning_rate": 5.3944665337747655e-05, + "loss": 1.0942, + "step": 7843 + }, + { + "epoch": 0.47657816392247404, + "grad_norm": 0.24038812518119812, + "learning_rate": 5.393512233018504e-05, + "loss": 1.0271, + "step": 7844 + }, + { + "epoch": 0.4766389209551006, + "grad_norm": 0.8283798098564148, + "learning_rate": 5.3925579178379745e-05, + "loss": 1.1276, + "step": 7845 + }, + { + "epoch": 0.47669967798772706, + "grad_norm": 0.3438664674758911, + "learning_rate": 5.3916035882681595e-05, + "loss": 1.2006, + "step": 7846 + }, + { + "epoch": 0.4767604350203536, + "grad_norm": 0.1746523380279541, + "learning_rate": 5.3906492443440384e-05, + "loss": 1.115, + "step": 7847 + }, + { + "epoch": 0.4768211920529801, + "grad_norm": 0.19465163350105286, + "learning_rate": 5.3896948861005935e-05, + "loss": 1.0282, + "step": 7848 + }, + { + "epoch": 0.47688194908560666, + "grad_norm": 0.31141549348831177, + "learning_rate": 5.388740513572809e-05, + "loss": 1.0645, + "step": 7849 + }, + { + "epoch": 0.4769427061182332, + "grad_norm": 1.1318401098251343, + "learning_rate": 5.3877861267956645e-05, + "loss": 1.2812, + "step": 7850 + }, + { + "epoch": 0.47700346315085973, + "grad_norm": 0.36720946431159973, + "learning_rate": 5.386831725804143e-05, + "loss": 1.002, + "step": 7851 + }, + { + "epoch": 0.47706422018348627, + "grad_norm": 0.3570851981639862, + "learning_rate": 5.385877310633233e-05, + "loss": 1.2786, + "step": 7852 + }, + { + "epoch": 0.47712497721611274, + "grad_norm": 0.19337321817874908, + "learning_rate": 5.3849228813179144e-05, + "loss": 1.1081, + "step": 7853 + }, + { + "epoch": 0.4771857342487393, + "grad_norm": 0.39098045229911804, + "learning_rate": 5.3839684378931724e-05, + "loss": 1.0026, + "step": 7854 + }, + { + "epoch": 0.4772464912813658, + "grad_norm": 0.2751114070415497, + "learning_rate": 5.383013980393993e-05, + "loss": 1.0589, + "step": 7855 + }, + { + "epoch": 0.47730724831399235, + "grad_norm": 0.27905410528182983, + "learning_rate": 5.3820595088553626e-05, + "loss": 1.1522, + "step": 7856 + }, + { + "epoch": 0.4773680053466189, + "grad_norm": 0.2844397723674774, + "learning_rate": 5.381105023312267e-05, + "loss": 1.0732, + "step": 7857 + }, + { + "epoch": 0.4774287623792454, + "grad_norm": 0.4838232696056366, + "learning_rate": 5.380150523799693e-05, + "loss": 1.1858, + "step": 7858 + }, + { + "epoch": 0.4774895194118719, + "grad_norm": 4.603612899780273, + "learning_rate": 5.3791960103526284e-05, + "loss": 1.13, + "step": 7859 + }, + { + "epoch": 0.47755027644449843, + "grad_norm": 0.4172646701335907, + "learning_rate": 5.378241483006061e-05, + "loss": 1.3196, + "step": 7860 + }, + { + "epoch": 0.47761103347712497, + "grad_norm": 1.6208860874176025, + "learning_rate": 5.377286941794979e-05, + "loss": 1.053, + "step": 7861 + }, + { + "epoch": 0.4776717905097515, + "grad_norm": 0.27274438738822937, + "learning_rate": 5.376332386754369e-05, + "loss": 1.203, + "step": 7862 + }, + { + "epoch": 0.47773254754237804, + "grad_norm": 0.3362331986427307, + "learning_rate": 5.375377817919226e-05, + "loss": 1.0181, + "step": 7863 + }, + { + "epoch": 0.4777933045750046, + "grad_norm": 0.32621508836746216, + "learning_rate": 5.374423235324534e-05, + "loss": 1.1533, + "step": 7864 + }, + { + "epoch": 0.4778540616076311, + "grad_norm": 0.569141685962677, + "learning_rate": 5.3734686390052866e-05, + "loss": 1.1443, + "step": 7865 + }, + { + "epoch": 0.4779148186402576, + "grad_norm": 0.31646814942359924, + "learning_rate": 5.372514028996475e-05, + "loss": 1.0688, + "step": 7866 + }, + { + "epoch": 0.4779755756728841, + "grad_norm": 0.26794108748435974, + "learning_rate": 5.371559405333089e-05, + "loss": 1.0643, + "step": 7867 + }, + { + "epoch": 0.47803633270551066, + "grad_norm": 0.28026628494262695, + "learning_rate": 5.370604768050121e-05, + "loss": 1.1335, + "step": 7868 + }, + { + "epoch": 0.4780970897381372, + "grad_norm": 0.433650940656662, + "learning_rate": 5.369650117182564e-05, + "loss": 1.0646, + "step": 7869 + }, + { + "epoch": 0.47815784677076373, + "grad_norm": 0.2206176221370697, + "learning_rate": 5.3686954527654096e-05, + "loss": 1.1771, + "step": 7870 + }, + { + "epoch": 0.47821860380339026, + "grad_norm": 0.3216512203216553, + "learning_rate": 5.367740774833653e-05, + "loss": 1.1278, + "step": 7871 + }, + { + "epoch": 0.47827936083601674, + "grad_norm": 0.31839242577552795, + "learning_rate": 5.366786083422286e-05, + "loss": 1.1425, + "step": 7872 + }, + { + "epoch": 0.4783401178686433, + "grad_norm": 0.3995634615421295, + "learning_rate": 5.365831378566306e-05, + "loss": 1.3483, + "step": 7873 + }, + { + "epoch": 0.4784008749012698, + "grad_norm": 0.3371184468269348, + "learning_rate": 5.364876660300705e-05, + "loss": 1.053, + "step": 7874 + }, + { + "epoch": 0.47846163193389635, + "grad_norm": 0.17160677909851074, + "learning_rate": 5.363921928660479e-05, + "loss": 1.0064, + "step": 7875 + }, + { + "epoch": 0.4785223889665229, + "grad_norm": 0.29481077194213867, + "learning_rate": 5.362967183680627e-05, + "loss": 1.1155, + "step": 7876 + }, + { + "epoch": 0.4785831459991494, + "grad_norm": 0.2002861499786377, + "learning_rate": 5.36201242539614e-05, + "loss": 1.0319, + "step": 7877 + }, + { + "epoch": 0.47864390303177595, + "grad_norm": 0.25530293583869934, + "learning_rate": 5.3610576538420185e-05, + "loss": 1.1633, + "step": 7878 + }, + { + "epoch": 0.47870466006440243, + "grad_norm": 0.39925727248191833, + "learning_rate": 5.3601028690532586e-05, + "loss": 1.1845, + "step": 7879 + }, + { + "epoch": 0.47876541709702897, + "grad_norm": 0.1766461580991745, + "learning_rate": 5.359148071064859e-05, + "loss": 1.096, + "step": 7880 + }, + { + "epoch": 0.4788261741296555, + "grad_norm": 0.19777780771255493, + "learning_rate": 5.358193259911817e-05, + "loss": 1.1013, + "step": 7881 + }, + { + "epoch": 0.47888693116228204, + "grad_norm": 0.3180660903453827, + "learning_rate": 5.35723843562913e-05, + "loss": 1.0667, + "step": 7882 + }, + { + "epoch": 0.47894768819490857, + "grad_norm": 0.2779911160469055, + "learning_rate": 5.356283598251803e-05, + "loss": 1.2735, + "step": 7883 + }, + { + "epoch": 0.4790084452275351, + "grad_norm": 0.18989722430706024, + "learning_rate": 5.3553287478148295e-05, + "loss": 1.0091, + "step": 7884 + }, + { + "epoch": 0.4790692022601616, + "grad_norm": 0.230073943734169, + "learning_rate": 5.3543738843532124e-05, + "loss": 1.1439, + "step": 7885 + }, + { + "epoch": 0.4791299592927881, + "grad_norm": 0.32565414905548096, + "learning_rate": 5.353419007901953e-05, + "loss": 1.0704, + "step": 7886 + }, + { + "epoch": 0.47919071632541466, + "grad_norm": 0.33661413192749023, + "learning_rate": 5.352464118496049e-05, + "loss": 1.3265, + "step": 7887 + }, + { + "epoch": 0.4792514733580412, + "grad_norm": 0.21748971939086914, + "learning_rate": 5.351509216170507e-05, + "loss": 1.0546, + "step": 7888 + }, + { + "epoch": 0.4793122303906677, + "grad_norm": 6.381424427032471, + "learning_rate": 5.350554300960326e-05, + "loss": 1.1051, + "step": 7889 + }, + { + "epoch": 0.47937298742329426, + "grad_norm": 0.39335599541664124, + "learning_rate": 5.349599372900509e-05, + "loss": 1.1911, + "step": 7890 + }, + { + "epoch": 0.4794337444559208, + "grad_norm": 0.16682802140712738, + "learning_rate": 5.34864443202606e-05, + "loss": 1.0778, + "step": 7891 + }, + { + "epoch": 0.4794945014885473, + "grad_norm": 0.16996996104717255, + "learning_rate": 5.347689478371982e-05, + "loss": 1.0737, + "step": 7892 + }, + { + "epoch": 0.4795552585211738, + "grad_norm": 0.18690958619117737, + "learning_rate": 5.346734511973278e-05, + "loss": 1.1178, + "step": 7893 + }, + { + "epoch": 0.47961601555380035, + "grad_norm": 0.1597827821969986, + "learning_rate": 5.345779532864954e-05, + "loss": 1.0428, + "step": 7894 + }, + { + "epoch": 0.4796767725864269, + "grad_norm": 0.17395959794521332, + "learning_rate": 5.344824541082014e-05, + "loss": 1.0341, + "step": 7895 + }, + { + "epoch": 0.4797375296190534, + "grad_norm": 0.2134961187839508, + "learning_rate": 5.343869536659465e-05, + "loss": 1.0158, + "step": 7896 + }, + { + "epoch": 0.47979828665167995, + "grad_norm": 0.18007944524288177, + "learning_rate": 5.342914519632312e-05, + "loss": 1.0363, + "step": 7897 + }, + { + "epoch": 0.4798590436843065, + "grad_norm": 0.22623908519744873, + "learning_rate": 5.3419594900355585e-05, + "loss": 1.0509, + "step": 7898 + }, + { + "epoch": 0.47991980071693296, + "grad_norm": 0.21957845985889435, + "learning_rate": 5.3410044479042166e-05, + "loss": 1.0023, + "step": 7899 + }, + { + "epoch": 0.4799805577495595, + "grad_norm": 0.23291805386543274, + "learning_rate": 5.3400493932732873e-05, + "loss": 1.1619, + "step": 7900 + }, + { + "epoch": 0.48004131478218603, + "grad_norm": 0.19825111329555511, + "learning_rate": 5.339094326177785e-05, + "loss": 1.1127, + "step": 7901 + }, + { + "epoch": 0.48010207181481257, + "grad_norm": 0.2806437313556671, + "learning_rate": 5.3381392466527135e-05, + "loss": 1.0672, + "step": 7902 + }, + { + "epoch": 0.4801628288474391, + "grad_norm": 0.1874779462814331, + "learning_rate": 5.3371841547330825e-05, + "loss": 1.0857, + "step": 7903 + }, + { + "epoch": 0.48022358588006564, + "grad_norm": 0.2725803256034851, + "learning_rate": 5.3362290504539005e-05, + "loss": 1.0452, + "step": 7904 + }, + { + "epoch": 0.4802843429126921, + "grad_norm": 0.2927266061306, + "learning_rate": 5.335273933850179e-05, + "loss": 1.0522, + "step": 7905 + }, + { + "epoch": 0.48034509994531865, + "grad_norm": 0.6349446177482605, + "learning_rate": 5.334318804956926e-05, + "loss": 1.0302, + "step": 7906 + }, + { + "epoch": 0.4804058569779452, + "grad_norm": 0.40387478470802307, + "learning_rate": 5.333363663809152e-05, + "loss": 1.0709, + "step": 7907 + }, + { + "epoch": 0.4804666140105717, + "grad_norm": 0.9047887325286865, + "learning_rate": 5.3324085104418683e-05, + "loss": 1.0253, + "step": 7908 + }, + { + "epoch": 0.48052737104319826, + "grad_norm": 0.16983115673065186, + "learning_rate": 5.331453344890088e-05, + "loss": 1.0906, + "step": 7909 + }, + { + "epoch": 0.4805881280758248, + "grad_norm": 0.21618200838565826, + "learning_rate": 5.33049816718882e-05, + "loss": 1.2072, + "step": 7910 + }, + { + "epoch": 0.48064888510845133, + "grad_norm": 0.35862234234809875, + "learning_rate": 5.3295429773730774e-05, + "loss": 1.2408, + "step": 7911 + }, + { + "epoch": 0.4807096421410778, + "grad_norm": 0.2044491171836853, + "learning_rate": 5.328587775477875e-05, + "loss": 1.1155, + "step": 7912 + }, + { + "epoch": 0.48077039917370434, + "grad_norm": 1.7568200826644897, + "learning_rate": 5.327632561538223e-05, + "loss": 1.0571, + "step": 7913 + }, + { + "epoch": 0.4808311562063309, + "grad_norm": 0.1703987866640091, + "learning_rate": 5.3266773355891365e-05, + "loss": 1.0036, + "step": 7914 + }, + { + "epoch": 0.4808919132389574, + "grad_norm": 0.2591935396194458, + "learning_rate": 5.325722097665629e-05, + "loss": 1.0722, + "step": 7915 + }, + { + "epoch": 0.48095267027158395, + "grad_norm": 0.24154451489448547, + "learning_rate": 5.3247668478027144e-05, + "loss": 1.0436, + "step": 7916 + }, + { + "epoch": 0.4810134273042105, + "grad_norm": 0.16714993119239807, + "learning_rate": 5.32381158603541e-05, + "loss": 1.02, + "step": 7917 + }, + { + "epoch": 0.48107418433683696, + "grad_norm": 0.33737269043922424, + "learning_rate": 5.3228563123987266e-05, + "loss": 1.1601, + "step": 7918 + }, + { + "epoch": 0.4811349413694635, + "grad_norm": 0.395826131105423, + "learning_rate": 5.321901026927685e-05, + "loss": 1.0787, + "step": 7919 + }, + { + "epoch": 0.48119569840209003, + "grad_norm": 0.21179351210594177, + "learning_rate": 5.320945729657298e-05, + "loss": 1.1457, + "step": 7920 + }, + { + "epoch": 0.48125645543471657, + "grad_norm": 0.2807459533214569, + "learning_rate": 5.319990420622583e-05, + "loss": 1.0744, + "step": 7921 + }, + { + "epoch": 0.4813172124673431, + "grad_norm": 0.28016483783721924, + "learning_rate": 5.319035099858558e-05, + "loss": 1.2205, + "step": 7922 + }, + { + "epoch": 0.48137796949996964, + "grad_norm": 0.24991266429424286, + "learning_rate": 5.3180797674002394e-05, + "loss": 1.3637, + "step": 7923 + }, + { + "epoch": 0.4814387265325962, + "grad_norm": 0.2709290683269501, + "learning_rate": 5.317124423282646e-05, + "loss": 1.0746, + "step": 7924 + }, + { + "epoch": 0.48149948356522265, + "grad_norm": 0.1968274712562561, + "learning_rate": 5.316169067540795e-05, + "loss": 1.0709, + "step": 7925 + }, + { + "epoch": 0.4815602405978492, + "grad_norm": 0.22891372442245483, + "learning_rate": 5.315213700209706e-05, + "loss": 1.0884, + "step": 7926 + }, + { + "epoch": 0.4816209976304757, + "grad_norm": 0.31204307079315186, + "learning_rate": 5.314258321324398e-05, + "loss": 1.0492, + "step": 7927 + }, + { + "epoch": 0.48168175466310226, + "grad_norm": 0.6906654238700867, + "learning_rate": 5.31330293091989e-05, + "loss": 1.0805, + "step": 7928 + }, + { + "epoch": 0.4817425116957288, + "grad_norm": 0.26162949204444885, + "learning_rate": 5.312347529031203e-05, + "loss": 1.1814, + "step": 7929 + }, + { + "epoch": 0.4818032687283553, + "grad_norm": 0.32089963555336, + "learning_rate": 5.3113921156933566e-05, + "loss": 1.0508, + "step": 7930 + }, + { + "epoch": 0.4818640257609818, + "grad_norm": 0.27471739053726196, + "learning_rate": 5.3104366909413705e-05, + "loss": 1.2752, + "step": 7931 + }, + { + "epoch": 0.48192478279360834, + "grad_norm": 0.2196688950061798, + "learning_rate": 5.309481254810269e-05, + "loss": 1.0167, + "step": 7932 + }, + { + "epoch": 0.4819855398262349, + "grad_norm": 0.27987876534461975, + "learning_rate": 5.308525807335074e-05, + "loss": 1.0579, + "step": 7933 + }, + { + "epoch": 0.4820462968588614, + "grad_norm": 0.15253625810146332, + "learning_rate": 5.3075703485508036e-05, + "loss": 1.0422, + "step": 7934 + }, + { + "epoch": 0.48210705389148795, + "grad_norm": 0.16870753467082977, + "learning_rate": 5.306614878492483e-05, + "loss": 1.0783, + "step": 7935 + }, + { + "epoch": 0.4821678109241145, + "grad_norm": 0.17224101722240448, + "learning_rate": 5.305659397195135e-05, + "loss": 1.0266, + "step": 7936 + }, + { + "epoch": 0.482228567956741, + "grad_norm": 0.3229813575744629, + "learning_rate": 5.304703904693783e-05, + "loss": 1.002, + "step": 7937 + }, + { + "epoch": 0.4822893249893675, + "grad_norm": 0.27749088406562805, + "learning_rate": 5.303748401023451e-05, + "loss": 1.1536, + "step": 7938 + }, + { + "epoch": 0.48235008202199403, + "grad_norm": 0.3479826748371124, + "learning_rate": 5.30279288621916e-05, + "loss": 1.0564, + "step": 7939 + }, + { + "epoch": 0.48241083905462057, + "grad_norm": 0.5479521751403809, + "learning_rate": 5.30183736031594e-05, + "loss": 1.1711, + "step": 7940 + }, + { + "epoch": 0.4824715960872471, + "grad_norm": 0.29602673649787903, + "learning_rate": 5.3008818233488124e-05, + "loss": 1.0142, + "step": 7941 + }, + { + "epoch": 0.48253235311987364, + "grad_norm": 0.30238693952560425, + "learning_rate": 5.299926275352802e-05, + "loss": 1.0344, + "step": 7942 + }, + { + "epoch": 0.48259311015250017, + "grad_norm": 0.3397599756717682, + "learning_rate": 5.298970716362939e-05, + "loss": 1.0593, + "step": 7943 + }, + { + "epoch": 0.4826538671851267, + "grad_norm": 3.074270009994507, + "learning_rate": 5.2980151464142434e-05, + "loss": 1.1325, + "step": 7944 + }, + { + "epoch": 0.4827146242177532, + "grad_norm": 0.3386601507663727, + "learning_rate": 5.297059565541746e-05, + "loss": 1.1664, + "step": 7945 + }, + { + "epoch": 0.4827753812503797, + "grad_norm": 0.28568035364151, + "learning_rate": 5.296103973780475e-05, + "loss": 1.092, + "step": 7946 + }, + { + "epoch": 0.48283613828300626, + "grad_norm": 0.17234770953655243, + "learning_rate": 5.2951483711654534e-05, + "loss": 1.0452, + "step": 7947 + }, + { + "epoch": 0.4828968953156328, + "grad_norm": 0.27826181054115295, + "learning_rate": 5.294192757731712e-05, + "loss": 1.1433, + "step": 7948 + }, + { + "epoch": 0.4829576523482593, + "grad_norm": 0.3185092806816101, + "learning_rate": 5.2932371335142784e-05, + "loss": 1.0513, + "step": 7949 + }, + { + "epoch": 0.48301840938088586, + "grad_norm": 0.177689790725708, + "learning_rate": 5.29228149854818e-05, + "loss": 1.0752, + "step": 7950 + }, + { + "epoch": 0.48307916641351234, + "grad_norm": 0.3550652861595154, + "learning_rate": 5.2913258528684476e-05, + "loss": 1.2053, + "step": 7951 + }, + { + "epoch": 0.4831399234461389, + "grad_norm": 0.3355729579925537, + "learning_rate": 5.290370196510108e-05, + "loss": 1.2328, + "step": 7952 + }, + { + "epoch": 0.4832006804787654, + "grad_norm": 0.23810675740242004, + "learning_rate": 5.2894145295081954e-05, + "loss": 1.0849, + "step": 7953 + }, + { + "epoch": 0.48326143751139194, + "grad_norm": 0.6307013034820557, + "learning_rate": 5.2884588518977364e-05, + "loss": 1.1542, + "step": 7954 + }, + { + "epoch": 0.4833221945440185, + "grad_norm": 0.2562393248081207, + "learning_rate": 5.287503163713761e-05, + "loss": 1.1341, + "step": 7955 + }, + { + "epoch": 0.483382951576645, + "grad_norm": 0.15776176750659943, + "learning_rate": 5.286547464991304e-05, + "loss": 1.0376, + "step": 7956 + }, + { + "epoch": 0.48344370860927155, + "grad_norm": 0.22272121906280518, + "learning_rate": 5.285591755765391e-05, + "loss": 1.0644, + "step": 7957 + }, + { + "epoch": 0.48350446564189803, + "grad_norm": 0.16499771177768707, + "learning_rate": 5.28463603607106e-05, + "loss": 1.0957, + "step": 7958 + }, + { + "epoch": 0.48356522267452456, + "grad_norm": 6.767419338226318, + "learning_rate": 5.2836803059433385e-05, + "loss": 1.0489, + "step": 7959 + }, + { + "epoch": 0.4836259797071511, + "grad_norm": 0.3651256859302521, + "learning_rate": 5.282724565417261e-05, + "loss": 1.0886, + "step": 7960 + }, + { + "epoch": 0.48368673673977763, + "grad_norm": 0.2786197066307068, + "learning_rate": 5.28176881452786e-05, + "loss": 1.1265, + "step": 7961 + }, + { + "epoch": 0.48374749377240417, + "grad_norm": 0.2558015286922455, + "learning_rate": 5.280813053310168e-05, + "loss": 1.2717, + "step": 7962 + }, + { + "epoch": 0.4838082508050307, + "grad_norm": 0.2377810776233673, + "learning_rate": 5.279857281799221e-05, + "loss": 1.0478, + "step": 7963 + }, + { + "epoch": 0.4838690078376572, + "grad_norm": 0.3320736289024353, + "learning_rate": 5.27890150003005e-05, + "loss": 1.0051, + "step": 7964 + }, + { + "epoch": 0.4839297648702837, + "grad_norm": 2.577547550201416, + "learning_rate": 5.2779457080376895e-05, + "loss": 1.1186, + "step": 7965 + }, + { + "epoch": 0.48399052190291025, + "grad_norm": 0.20932815968990326, + "learning_rate": 5.276989905857178e-05, + "loss": 1.0776, + "step": 7966 + }, + { + "epoch": 0.4840512789355368, + "grad_norm": 0.26246631145477295, + "learning_rate": 5.276034093523545e-05, + "loss": 1.0351, + "step": 7967 + }, + { + "epoch": 0.4841120359681633, + "grad_norm": 1.684192180633545, + "learning_rate": 5.27507827107183e-05, + "loss": 1.1569, + "step": 7968 + }, + { + "epoch": 0.48417279300078986, + "grad_norm": 0.2570514678955078, + "learning_rate": 5.274122438537069e-05, + "loss": 1.1023, + "step": 7969 + }, + { + "epoch": 0.4842335500334164, + "grad_norm": 0.21962356567382812, + "learning_rate": 5.2731665959542964e-05, + "loss": 1.0941, + "step": 7970 + }, + { + "epoch": 0.4842943070660429, + "grad_norm": 0.2093268483877182, + "learning_rate": 5.272210743358549e-05, + "loss": 1.0872, + "step": 7971 + }, + { + "epoch": 0.4843550640986694, + "grad_norm": 0.20752331614494324, + "learning_rate": 5.271254880784865e-05, + "loss": 1.0518, + "step": 7972 + }, + { + "epoch": 0.48441582113129594, + "grad_norm": 0.24322804808616638, + "learning_rate": 5.27029900826828e-05, + "loss": 1.1143, + "step": 7973 + }, + { + "epoch": 0.4844765781639225, + "grad_norm": 0.21228475868701935, + "learning_rate": 5.269343125843833e-05, + "loss": 1.0554, + "step": 7974 + }, + { + "epoch": 0.484537335196549, + "grad_norm": 0.15959379076957703, + "learning_rate": 5.268387233546561e-05, + "loss": 1.0845, + "step": 7975 + }, + { + "epoch": 0.48459809222917555, + "grad_norm": 0.21946778893470764, + "learning_rate": 5.267431331411504e-05, + "loss": 1.0358, + "step": 7976 + }, + { + "epoch": 0.484658849261802, + "grad_norm": 0.23403900861740112, + "learning_rate": 5.2664754194737007e-05, + "loss": 1.0361, + "step": 7977 + }, + { + "epoch": 0.48471960629442856, + "grad_norm": 0.18266230821609497, + "learning_rate": 5.2655194977681864e-05, + "loss": 1.0328, + "step": 7978 + }, + { + "epoch": 0.4847803633270551, + "grad_norm": 0.46280965209007263, + "learning_rate": 5.264563566330006e-05, + "loss": 1.0103, + "step": 7979 + }, + { + "epoch": 0.48484112035968163, + "grad_norm": 0.938435971736908, + "learning_rate": 5.263607625194197e-05, + "loss": 1.0395, + "step": 7980 + }, + { + "epoch": 0.48490187739230817, + "grad_norm": 0.27795302867889404, + "learning_rate": 5.262651674395799e-05, + "loss": 1.1518, + "step": 7981 + }, + { + "epoch": 0.4849626344249347, + "grad_norm": 0.19984029233455658, + "learning_rate": 5.261695713969853e-05, + "loss": 1.0652, + "step": 7982 + }, + { + "epoch": 0.48502339145756124, + "grad_norm": 0.25176432728767395, + "learning_rate": 5.2607397439514e-05, + "loss": 1.1041, + "step": 7983 + }, + { + "epoch": 0.4850841484901877, + "grad_norm": 0.4109284281730652, + "learning_rate": 5.259783764375481e-05, + "loss": 1.227, + "step": 7984 + }, + { + "epoch": 0.48514490552281425, + "grad_norm": 0.2165309190750122, + "learning_rate": 5.2588277752771383e-05, + "loss": 1.2029, + "step": 7985 + }, + { + "epoch": 0.4852056625554408, + "grad_norm": 0.17905232310295105, + "learning_rate": 5.257871776691413e-05, + "loss": 1.0589, + "step": 7986 + }, + { + "epoch": 0.4852664195880673, + "grad_norm": 1.559589147567749, + "learning_rate": 5.2569157686533486e-05, + "loss": 1.1237, + "step": 7987 + }, + { + "epoch": 0.48532717662069386, + "grad_norm": 0.44913190603256226, + "learning_rate": 5.2559597511979865e-05, + "loss": 1.2577, + "step": 7988 + }, + { + "epoch": 0.4853879336533204, + "grad_norm": 0.19089102745056152, + "learning_rate": 5.25500372436037e-05, + "loss": 1.0281, + "step": 7989 + }, + { + "epoch": 0.48544869068594687, + "grad_norm": 0.20560552179813385, + "learning_rate": 5.254047688175544e-05, + "loss": 1.0759, + "step": 7990 + }, + { + "epoch": 0.4855094477185734, + "grad_norm": 0.22604411840438843, + "learning_rate": 5.253091642678549e-05, + "loss": 1.142, + "step": 7991 + }, + { + "epoch": 0.48557020475119994, + "grad_norm": 0.46328455209732056, + "learning_rate": 5.252135587904432e-05, + "loss": 1.2114, + "step": 7992 + }, + { + "epoch": 0.4856309617838265, + "grad_norm": 0.23439888656139374, + "learning_rate": 5.251179523888237e-05, + "loss": 1.1328, + "step": 7993 + }, + { + "epoch": 0.485691718816453, + "grad_norm": 0.26095741987228394, + "learning_rate": 5.2502234506650074e-05, + "loss": 1.0043, + "step": 7994 + }, + { + "epoch": 0.48575247584907955, + "grad_norm": 0.40629056096076965, + "learning_rate": 5.249267368269788e-05, + "loss": 1.1671, + "step": 7995 + }, + { + "epoch": 0.4858132328817061, + "grad_norm": 0.35910409688949585, + "learning_rate": 5.2483112767376267e-05, + "loss": 1.0387, + "step": 7996 + }, + { + "epoch": 0.48587398991433256, + "grad_norm": 0.46119558811187744, + "learning_rate": 5.247355176103567e-05, + "loss": 1.2785, + "step": 7997 + }, + { + "epoch": 0.4859347469469591, + "grad_norm": 0.20692366361618042, + "learning_rate": 5.246399066402654e-05, + "loss": 1.0704, + "step": 7998 + }, + { + "epoch": 0.48599550397958563, + "grad_norm": 0.2578948736190796, + "learning_rate": 5.2454429476699365e-05, + "loss": 1.2423, + "step": 7999 + }, + { + "epoch": 0.48605626101221217, + "grad_norm": 0.6680195331573486, + "learning_rate": 5.2444868199404616e-05, + "loss": 1.0593, + "step": 8000 + }, + { + "epoch": 0.4861170180448387, + "grad_norm": 0.29338738322257996, + "learning_rate": 5.2435306832492724e-05, + "loss": 1.1678, + "step": 8001 + }, + { + "epoch": 0.48617777507746524, + "grad_norm": 2.068232536315918, + "learning_rate": 5.2425745376314206e-05, + "loss": 1.0352, + "step": 8002 + }, + { + "epoch": 0.48623853211009177, + "grad_norm": 0.513767421245575, + "learning_rate": 5.2416183831219535e-05, + "loss": 1.0754, + "step": 8003 + }, + { + "epoch": 0.48629928914271825, + "grad_norm": 0.24635492265224457, + "learning_rate": 5.240662219755915e-05, + "loss": 1.1173, + "step": 8004 + }, + { + "epoch": 0.4863600461753448, + "grad_norm": 0.29896867275238037, + "learning_rate": 5.239706047568358e-05, + "loss": 1.1242, + "step": 8005 + }, + { + "epoch": 0.4864208032079713, + "grad_norm": 0.3185860812664032, + "learning_rate": 5.2387498665943294e-05, + "loss": 1.27, + "step": 8006 + }, + { + "epoch": 0.48648156024059785, + "grad_norm": 0.2312432825565338, + "learning_rate": 5.237793676868879e-05, + "loss": 1.1531, + "step": 8007 + }, + { + "epoch": 0.4865423172732244, + "grad_norm": 0.20754635334014893, + "learning_rate": 5.2368374784270544e-05, + "loss": 1.1506, + "step": 8008 + }, + { + "epoch": 0.4866030743058509, + "grad_norm": 0.2468145191669464, + "learning_rate": 5.2358812713039054e-05, + "loss": 1.1492, + "step": 8009 + }, + { + "epoch": 0.4866638313384774, + "grad_norm": 0.29405614733695984, + "learning_rate": 5.234925055534485e-05, + "loss": 1.0992, + "step": 8010 + }, + { + "epoch": 0.48672458837110394, + "grad_norm": 0.19453515112400055, + "learning_rate": 5.233968831153839e-05, + "loss": 1.0571, + "step": 8011 + }, + { + "epoch": 0.4867853454037305, + "grad_norm": 0.3593937158584595, + "learning_rate": 5.233012598197021e-05, + "loss": 1.092, + "step": 8012 + }, + { + "epoch": 0.486846102436357, + "grad_norm": 0.19887514412403107, + "learning_rate": 5.232056356699083e-05, + "loss": 1.0396, + "step": 8013 + }, + { + "epoch": 0.48690685946898354, + "grad_norm": 0.24907569587230682, + "learning_rate": 5.2311001066950715e-05, + "loss": 1.0815, + "step": 8014 + }, + { + "epoch": 0.4869676165016101, + "grad_norm": 0.2843758761882782, + "learning_rate": 5.2301438482200425e-05, + "loss": 1.0194, + "step": 8015 + }, + { + "epoch": 0.4870283735342366, + "grad_norm": 0.2814374268054962, + "learning_rate": 5.2291875813090465e-05, + "loss": 1.0022, + "step": 8016 + }, + { + "epoch": 0.4870891305668631, + "grad_norm": 0.33612093329429626, + "learning_rate": 5.228231305997136e-05, + "loss": 1.139, + "step": 8017 + }, + { + "epoch": 0.48714988759948963, + "grad_norm": 0.1629321128129959, + "learning_rate": 5.227275022319361e-05, + "loss": 1.0385, + "step": 8018 + }, + { + "epoch": 0.48721064463211616, + "grad_norm": 0.16132812201976776, + "learning_rate": 5.226318730310776e-05, + "loss": 1.0532, + "step": 8019 + }, + { + "epoch": 0.4872714016647427, + "grad_norm": 0.3240479826927185, + "learning_rate": 5.2253624300064365e-05, + "loss": 1.0874, + "step": 8020 + }, + { + "epoch": 0.48733215869736923, + "grad_norm": 0.3388752341270447, + "learning_rate": 5.224406121441392e-05, + "loss": 1.1151, + "step": 8021 + }, + { + "epoch": 0.48739291572999577, + "grad_norm": 0.34432846307754517, + "learning_rate": 5.223449804650697e-05, + "loss": 1.0012, + "step": 8022 + }, + { + "epoch": 0.48745367276262225, + "grad_norm": 0.33019089698791504, + "learning_rate": 5.222493479669409e-05, + "loss": 1.1519, + "step": 8023 + }, + { + "epoch": 0.4875144297952488, + "grad_norm": 0.27106785774230957, + "learning_rate": 5.221537146532577e-05, + "loss": 1.1537, + "step": 8024 + }, + { + "epoch": 0.4875751868278753, + "grad_norm": 0.2733124792575836, + "learning_rate": 5.220580805275258e-05, + "loss": 1.0676, + "step": 8025 + }, + { + "epoch": 0.48763594386050185, + "grad_norm": 1.0885729789733887, + "learning_rate": 5.219624455932508e-05, + "loss": 1.0906, + "step": 8026 + }, + { + "epoch": 0.4876967008931284, + "grad_norm": 0.3168673515319824, + "learning_rate": 5.218668098539381e-05, + "loss": 1.2171, + "step": 8027 + }, + { + "epoch": 0.4877574579257549, + "grad_norm": 0.24438971281051636, + "learning_rate": 5.217711733130932e-05, + "loss": 1.027, + "step": 8028 + }, + { + "epoch": 0.48781821495838146, + "grad_norm": 0.24599319696426392, + "learning_rate": 5.216755359742218e-05, + "loss": 1.0928, + "step": 8029 + }, + { + "epoch": 0.48787897199100794, + "grad_norm": 0.32027167081832886, + "learning_rate": 5.215798978408295e-05, + "loss": 1.0825, + "step": 8030 + }, + { + "epoch": 0.48793972902363447, + "grad_norm": 0.1972193866968155, + "learning_rate": 5.214842589164218e-05, + "loss": 1.0476, + "step": 8031 + }, + { + "epoch": 0.488000486056261, + "grad_norm": 0.35687872767448425, + "learning_rate": 5.213886192045043e-05, + "loss": 1.0338, + "step": 8032 + }, + { + "epoch": 0.48806124308888754, + "grad_norm": 0.2560797929763794, + "learning_rate": 5.2129297870858304e-05, + "loss": 1.0295, + "step": 8033 + }, + { + "epoch": 0.4881220001215141, + "grad_norm": 0.215506449341774, + "learning_rate": 5.2119733743216346e-05, + "loss": 1.1401, + "step": 8034 + }, + { + "epoch": 0.4881827571541406, + "grad_norm": 0.3178851008415222, + "learning_rate": 5.2110169537875134e-05, + "loss": 1.2365, + "step": 8035 + }, + { + "epoch": 0.4882435141867671, + "grad_norm": 0.997816801071167, + "learning_rate": 5.210060525518525e-05, + "loss": 1.0707, + "step": 8036 + }, + { + "epoch": 0.4883042712193936, + "grad_norm": 0.20935621857643127, + "learning_rate": 5.209104089549728e-05, + "loss": 1.1169, + "step": 8037 + }, + { + "epoch": 0.48836502825202016, + "grad_norm": 0.194577157497406, + "learning_rate": 5.2081476459161796e-05, + "loss": 1.0488, + "step": 8038 + }, + { + "epoch": 0.4884257852846467, + "grad_norm": 0.2573005259037018, + "learning_rate": 5.20719119465294e-05, + "loss": 1.1219, + "step": 8039 + }, + { + "epoch": 0.48848654231727323, + "grad_norm": 0.24838697910308838, + "learning_rate": 5.2062347357950655e-05, + "loss": 1.212, + "step": 8040 + }, + { + "epoch": 0.48854729934989977, + "grad_norm": 0.28555455803871155, + "learning_rate": 5.205278269377618e-05, + "loss": 1.0834, + "step": 8041 + }, + { + "epoch": 0.4886080563825263, + "grad_norm": 0.4251202642917633, + "learning_rate": 5.204321795435656e-05, + "loss": 1.1438, + "step": 8042 + }, + { + "epoch": 0.4886688134151528, + "grad_norm": 0.1663186103105545, + "learning_rate": 5.2033653140042384e-05, + "loss": 1.0903, + "step": 8043 + }, + { + "epoch": 0.4887295704477793, + "grad_norm": 0.2342243492603302, + "learning_rate": 5.202408825118427e-05, + "loss": 1.1687, + "step": 8044 + }, + { + "epoch": 0.48879032748040585, + "grad_norm": 0.5047648549079895, + "learning_rate": 5.201452328813279e-05, + "loss": 1.0408, + "step": 8045 + }, + { + "epoch": 0.4888510845130324, + "grad_norm": 1.8601551055908203, + "learning_rate": 5.200495825123858e-05, + "loss": 1.0697, + "step": 8046 + }, + { + "epoch": 0.4889118415456589, + "grad_norm": 0.19322028756141663, + "learning_rate": 5.199539314085224e-05, + "loss": 1.0023, + "step": 8047 + }, + { + "epoch": 0.48897259857828546, + "grad_norm": 0.19818828999996185, + "learning_rate": 5.1985827957324376e-05, + "loss": 1.1574, + "step": 8048 + }, + { + "epoch": 0.489033355610912, + "grad_norm": 0.19978925585746765, + "learning_rate": 5.197626270100561e-05, + "loss": 1.1195, + "step": 8049 + }, + { + "epoch": 0.48909411264353847, + "grad_norm": 0.25012704730033875, + "learning_rate": 5.196669737224654e-05, + "loss": 1.1455, + "step": 8050 + }, + { + "epoch": 0.489154869676165, + "grad_norm": 0.23060300946235657, + "learning_rate": 5.195713197139781e-05, + "loss": 1.1287, + "step": 8051 + }, + { + "epoch": 0.48921562670879154, + "grad_norm": 0.23804767429828644, + "learning_rate": 5.194756649881003e-05, + "loss": 1.1001, + "step": 8052 + }, + { + "epoch": 0.4892763837414181, + "grad_norm": 0.15199387073516846, + "learning_rate": 5.1938000954833824e-05, + "loss": 1.2227, + "step": 8053 + }, + { + "epoch": 0.4893371407740446, + "grad_norm": 0.20485831797122955, + "learning_rate": 5.192843533981981e-05, + "loss": 1.1068, + "step": 8054 + }, + { + "epoch": 0.48939789780667114, + "grad_norm": 0.3025892972946167, + "learning_rate": 5.191886965411863e-05, + "loss": 1.0623, + "step": 8055 + }, + { + "epoch": 0.4894586548392976, + "grad_norm": 0.5403998494148254, + "learning_rate": 5.190930389808091e-05, + "loss": 1.2265, + "step": 8056 + }, + { + "epoch": 0.48951941187192416, + "grad_norm": 0.36432701349258423, + "learning_rate": 5.1899738072057294e-05, + "loss": 1.0946, + "step": 8057 + }, + { + "epoch": 0.4895801689045507, + "grad_norm": 0.2495393455028534, + "learning_rate": 5.18901721763984e-05, + "loss": 1.0576, + "step": 8058 + }, + { + "epoch": 0.48964092593717723, + "grad_norm": 0.35778117179870605, + "learning_rate": 5.1880606211454886e-05, + "loss": 1.0789, + "step": 8059 + }, + { + "epoch": 0.48970168296980376, + "grad_norm": 0.1856382191181183, + "learning_rate": 5.187104017757738e-05, + "loss": 1.0719, + "step": 8060 + }, + { + "epoch": 0.4897624400024303, + "grad_norm": 0.27437591552734375, + "learning_rate": 5.186147407511654e-05, + "loss": 1.0364, + "step": 8061 + }, + { + "epoch": 0.48982319703505683, + "grad_norm": 0.16722312569618225, + "learning_rate": 5.185190790442301e-05, + "loss": 1.0424, + "step": 8062 + }, + { + "epoch": 0.4898839540676833, + "grad_norm": 1.6140021085739136, + "learning_rate": 5.184234166584745e-05, + "loss": 1.0874, + "step": 8063 + }, + { + "epoch": 0.48994471110030985, + "grad_norm": 0.24838882684707642, + "learning_rate": 5.1832775359740484e-05, + "loss": 1.043, + "step": 8064 + }, + { + "epoch": 0.4900054681329364, + "grad_norm": 0.2292795330286026, + "learning_rate": 5.1823208986452786e-05, + "loss": 0.9817, + "step": 8065 + }, + { + "epoch": 0.4900662251655629, + "grad_norm": 0.2428232729434967, + "learning_rate": 5.181364254633501e-05, + "loss": 1.0416, + "step": 8066 + }, + { + "epoch": 0.49012698219818945, + "grad_norm": 0.3024384379386902, + "learning_rate": 5.180407603973783e-05, + "loss": 1.089, + "step": 8067 + }, + { + "epoch": 0.490187739230816, + "grad_norm": 0.1872648000717163, + "learning_rate": 5.1794509467011864e-05, + "loss": 1.0833, + "step": 8068 + }, + { + "epoch": 0.49024849626344247, + "grad_norm": 0.2842201888561249, + "learning_rate": 5.178494282850782e-05, + "loss": 1.0954, + "step": 8069 + }, + { + "epoch": 0.490309253296069, + "grad_norm": 0.6653691530227661, + "learning_rate": 5.177537612457637e-05, + "loss": 1.2517, + "step": 8070 + }, + { + "epoch": 0.49037001032869554, + "grad_norm": 0.29714158177375793, + "learning_rate": 5.176580935556814e-05, + "loss": 1.0609, + "step": 8071 + }, + { + "epoch": 0.4904307673613221, + "grad_norm": 0.1815546751022339, + "learning_rate": 5.175624252183383e-05, + "loss": 1.0718, + "step": 8072 + }, + { + "epoch": 0.4904915243939486, + "grad_norm": 0.19997532665729523, + "learning_rate": 5.174667562372413e-05, + "loss": 1.0917, + "step": 8073 + }, + { + "epoch": 0.49055228142657514, + "grad_norm": 0.23305092751979828, + "learning_rate": 5.1737108661589685e-05, + "loss": 1.1018, + "step": 8074 + }, + { + "epoch": 0.4906130384592017, + "grad_norm": 1.61652410030365, + "learning_rate": 5.172754163578118e-05, + "loss": 1.0752, + "step": 8075 + }, + { + "epoch": 0.49067379549182816, + "grad_norm": 0.23336483538150787, + "learning_rate": 5.17179745466493e-05, + "loss": 1.1296, + "step": 8076 + }, + { + "epoch": 0.4907345525244547, + "grad_norm": 0.22476141154766083, + "learning_rate": 5.1708407394544755e-05, + "loss": 1.149, + "step": 8077 + }, + { + "epoch": 0.4907953095570812, + "grad_norm": 0.23487325012683868, + "learning_rate": 5.1698840179818186e-05, + "loss": 1.0902, + "step": 8078 + }, + { + "epoch": 0.49085606658970776, + "grad_norm": 0.21375839412212372, + "learning_rate": 5.1689272902820296e-05, + "loss": 1.0758, + "step": 8079 + }, + { + "epoch": 0.4909168236223343, + "grad_norm": 0.26866990327835083, + "learning_rate": 5.16797055639018e-05, + "loss": 1.2417, + "step": 8080 + }, + { + "epoch": 0.49097758065496083, + "grad_norm": 0.20398926734924316, + "learning_rate": 5.167013816341336e-05, + "loss": 1.0053, + "step": 8081 + }, + { + "epoch": 0.4910383376875873, + "grad_norm": 0.2140093594789505, + "learning_rate": 5.166057070170569e-05, + "loss": 1.0887, + "step": 8082 + }, + { + "epoch": 0.49109909472021385, + "grad_norm": 0.19469992816448212, + "learning_rate": 5.1651003179129476e-05, + "loss": 1.0483, + "step": 8083 + }, + { + "epoch": 0.4911598517528404, + "grad_norm": 0.1425102800130844, + "learning_rate": 5.164143559603543e-05, + "loss": 1.0549, + "step": 8084 + }, + { + "epoch": 0.4912206087854669, + "grad_norm": 0.17900612950325012, + "learning_rate": 5.163186795277423e-05, + "loss": 1.0992, + "step": 8085 + }, + { + "epoch": 0.49128136581809345, + "grad_norm": 0.19786173105239868, + "learning_rate": 5.162230024969661e-05, + "loss": 1.052, + "step": 8086 + }, + { + "epoch": 0.49134212285072, + "grad_norm": 0.2261119931936264, + "learning_rate": 5.161273248715325e-05, + "loss": 1.1011, + "step": 8087 + }, + { + "epoch": 0.4914028798833465, + "grad_norm": 4.01869010925293, + "learning_rate": 5.1603164665494877e-05, + "loss": 1.1361, + "step": 8088 + }, + { + "epoch": 0.491463636915973, + "grad_norm": 0.162197545170784, + "learning_rate": 5.159359678507217e-05, + "loss": 1.0234, + "step": 8089 + }, + { + "epoch": 0.49152439394859954, + "grad_norm": 0.6136916279792786, + "learning_rate": 5.15840288462359e-05, + "loss": 1.0305, + "step": 8090 + }, + { + "epoch": 0.49158515098122607, + "grad_norm": 0.19052204489707947, + "learning_rate": 5.1574460849336724e-05, + "loss": 1.0725, + "step": 8091 + }, + { + "epoch": 0.4916459080138526, + "grad_norm": 0.15237952768802643, + "learning_rate": 5.156489279472537e-05, + "loss": 1.0256, + "step": 8092 + }, + { + "epoch": 0.49170666504647914, + "grad_norm": 0.3050539195537567, + "learning_rate": 5.155532468275259e-05, + "loss": 1.1796, + "step": 8093 + }, + { + "epoch": 0.4917674220791057, + "grad_norm": 0.18298310041427612, + "learning_rate": 5.1545756513769084e-05, + "loss": 1.0597, + "step": 8094 + }, + { + "epoch": 0.49182817911173216, + "grad_norm": 0.15779797732830048, + "learning_rate": 5.1536188288125565e-05, + "loss": 1.0197, + "step": 8095 + }, + { + "epoch": 0.4918889361443587, + "grad_norm": 1.2724928855895996, + "learning_rate": 5.152662000617277e-05, + "loss": 1.1207, + "step": 8096 + }, + { + "epoch": 0.4919496931769852, + "grad_norm": 1.925052523612976, + "learning_rate": 5.151705166826143e-05, + "loss": 1.1157, + "step": 8097 + }, + { + "epoch": 0.49201045020961176, + "grad_norm": 0.1940930187702179, + "learning_rate": 5.1507483274742266e-05, + "loss": 1.0831, + "step": 8098 + }, + { + "epoch": 0.4920712072422383, + "grad_norm": 0.2818032503128052, + "learning_rate": 5.1497914825965996e-05, + "loss": 1.1162, + "step": 8099 + }, + { + "epoch": 0.49213196427486483, + "grad_norm": 0.22024548053741455, + "learning_rate": 5.1488346322283384e-05, + "loss": 1.1157, + "step": 8100 + }, + { + "epoch": 0.49219272130749137, + "grad_norm": 0.36588212847709656, + "learning_rate": 5.147877776404515e-05, + "loss": 1.0445, + "step": 8101 + }, + { + "epoch": 0.49225347834011784, + "grad_norm": 0.2352423220872879, + "learning_rate": 5.1469209151602014e-05, + "loss": 1.1273, + "step": 8102 + }, + { + "epoch": 0.4923142353727444, + "grad_norm": 0.15190699696540833, + "learning_rate": 5.145964048530475e-05, + "loss": 1.0346, + "step": 8103 + }, + { + "epoch": 0.4923749924053709, + "grad_norm": 0.16297903656959534, + "learning_rate": 5.145007176550408e-05, + "loss": 1.0727, + "step": 8104 + }, + { + "epoch": 0.49243574943799745, + "grad_norm": 0.26103314757347107, + "learning_rate": 5.144050299255074e-05, + "loss": 1.3343, + "step": 8105 + }, + { + "epoch": 0.492496506470624, + "grad_norm": 0.16598579287528992, + "learning_rate": 5.143093416679549e-05, + "loss": 1.048, + "step": 8106 + }, + { + "epoch": 0.4925572635032505, + "grad_norm": 0.6972333788871765, + "learning_rate": 5.142136528858907e-05, + "loss": 1.0946, + "step": 8107 + }, + { + "epoch": 0.49261802053587705, + "grad_norm": 0.2653838098049164, + "learning_rate": 5.1411796358282235e-05, + "loss": 1.1102, + "step": 8108 + }, + { + "epoch": 0.49267877756850353, + "grad_norm": 0.16588179767131805, + "learning_rate": 5.140222737622573e-05, + "loss": 1.0192, + "step": 8109 + }, + { + "epoch": 0.49273953460113007, + "grad_norm": 0.20079882442951202, + "learning_rate": 5.139265834277032e-05, + "loss": 1.1232, + "step": 8110 + }, + { + "epoch": 0.4928002916337566, + "grad_norm": 0.1658407300710678, + "learning_rate": 5.1383089258266725e-05, + "loss": 1.0255, + "step": 8111 + }, + { + "epoch": 0.49286104866638314, + "grad_norm": 0.1716679036617279, + "learning_rate": 5.1373520123065735e-05, + "loss": 1.0567, + "step": 8112 + }, + { + "epoch": 0.4929218056990097, + "grad_norm": 0.17032849788665771, + "learning_rate": 5.1363950937518104e-05, + "loss": 1.0506, + "step": 8113 + }, + { + "epoch": 0.4929825627316362, + "grad_norm": 0.1883547157049179, + "learning_rate": 5.1354381701974597e-05, + "loss": 1.0829, + "step": 8114 + }, + { + "epoch": 0.4930433197642627, + "grad_norm": 0.23675614595413208, + "learning_rate": 5.1344812416785946e-05, + "loss": 1.0787, + "step": 8115 + }, + { + "epoch": 0.4931040767968892, + "grad_norm": 2.5445520877838135, + "learning_rate": 5.133524308230295e-05, + "loss": 1.0633, + "step": 8116 + }, + { + "epoch": 0.49316483382951576, + "grad_norm": 0.22502104938030243, + "learning_rate": 5.1325673698876356e-05, + "loss": 1.1015, + "step": 8117 + }, + { + "epoch": 0.4932255908621423, + "grad_norm": 0.17098243534564972, + "learning_rate": 5.1316104266856936e-05, + "loss": 1.1028, + "step": 8118 + }, + { + "epoch": 0.49328634789476883, + "grad_norm": 0.15951326489448547, + "learning_rate": 5.130653478659546e-05, + "loss": 1.0924, + "step": 8119 + }, + { + "epoch": 0.49334710492739536, + "grad_norm": 0.4295470416545868, + "learning_rate": 5.1296965258442706e-05, + "loss": 1.0507, + "step": 8120 + }, + { + "epoch": 0.4934078619600219, + "grad_norm": 0.1790187805891037, + "learning_rate": 5.128739568274944e-05, + "loss": 1.1161, + "step": 8121 + }, + { + "epoch": 0.4934686189926484, + "grad_norm": 0.15050630271434784, + "learning_rate": 5.127782605986643e-05, + "loss": 1.0825, + "step": 8122 + }, + { + "epoch": 0.4935293760252749, + "grad_norm": 0.24988164007663727, + "learning_rate": 5.1268256390144444e-05, + "loss": 1.184, + "step": 8123 + }, + { + "epoch": 0.49359013305790145, + "grad_norm": 0.32614824175834656, + "learning_rate": 5.125868667393431e-05, + "loss": 1.1276, + "step": 8124 + }, + { + "epoch": 0.493650890090528, + "grad_norm": 0.23005157709121704, + "learning_rate": 5.124911691158674e-05, + "loss": 1.14, + "step": 8125 + }, + { + "epoch": 0.4937116471231545, + "grad_norm": 1.2748064994812012, + "learning_rate": 5.123954710345256e-05, + "loss": 1.0184, + "step": 8126 + }, + { + "epoch": 0.49377240415578105, + "grad_norm": 0.16370514035224915, + "learning_rate": 5.122997724988255e-05, + "loss": 1.0601, + "step": 8127 + }, + { + "epoch": 0.49383316118840753, + "grad_norm": 0.26546815037727356, + "learning_rate": 5.122040735122747e-05, + "loss": 1.2883, + "step": 8128 + }, + { + "epoch": 0.49389391822103407, + "grad_norm": 0.34020674228668213, + "learning_rate": 5.121083740783813e-05, + "loss": 1.2201, + "step": 8129 + }, + { + "epoch": 0.4939546752536606, + "grad_norm": 0.2647283673286438, + "learning_rate": 5.120126742006531e-05, + "loss": 1.1109, + "step": 8130 + }, + { + "epoch": 0.49401543228628714, + "grad_norm": 0.22371411323547363, + "learning_rate": 5.119169738825981e-05, + "loss": 1.1341, + "step": 8131 + }, + { + "epoch": 0.49407618931891367, + "grad_norm": 0.23701615631580353, + "learning_rate": 5.11821273127724e-05, + "loss": 1.1404, + "step": 8132 + }, + { + "epoch": 0.4941369463515402, + "grad_norm": 0.18504586815834045, + "learning_rate": 5.117255719395388e-05, + "loss": 1.0327, + "step": 8133 + }, + { + "epoch": 0.49419770338416674, + "grad_norm": 0.26946520805358887, + "learning_rate": 5.1162987032155074e-05, + "loss": 1.1587, + "step": 8134 + }, + { + "epoch": 0.4942584604167932, + "grad_norm": 0.16372235119342804, + "learning_rate": 5.115341682772674e-05, + "loss": 1.0186, + "step": 8135 + }, + { + "epoch": 0.49431921744941976, + "grad_norm": 0.29886800050735474, + "learning_rate": 5.114384658101967e-05, + "loss": 1.0712, + "step": 8136 + }, + { + "epoch": 0.4943799744820463, + "grad_norm": 0.21416710317134857, + "learning_rate": 5.113427629238472e-05, + "loss": 1.1035, + "step": 8137 + }, + { + "epoch": 0.4944407315146728, + "grad_norm": 0.24226157367229462, + "learning_rate": 5.112470596217261e-05, + "loss": 1.1255, + "step": 8138 + }, + { + "epoch": 0.49450148854729936, + "grad_norm": 1.6867119073867798, + "learning_rate": 5.111513559073422e-05, + "loss": 1.0376, + "step": 8139 + }, + { + "epoch": 0.4945622455799259, + "grad_norm": 0.18056273460388184, + "learning_rate": 5.110556517842031e-05, + "loss": 1.1054, + "step": 8140 + }, + { + "epoch": 0.4946230026125524, + "grad_norm": 0.22449879348278046, + "learning_rate": 5.109599472558169e-05, + "loss": 1.1894, + "step": 8141 + }, + { + "epoch": 0.4946837596451789, + "grad_norm": 0.2928978502750397, + "learning_rate": 5.108642423256918e-05, + "loss": 1.217, + "step": 8142 + }, + { + "epoch": 0.49474451667780545, + "grad_norm": 0.3114981949329376, + "learning_rate": 5.1076853699733576e-05, + "loss": 1.1359, + "step": 8143 + }, + { + "epoch": 0.494805273710432, + "grad_norm": 0.23422235250473022, + "learning_rate": 5.106728312742568e-05, + "loss": 1.1303, + "step": 8144 + }, + { + "epoch": 0.4948660307430585, + "grad_norm": 0.22376172244548798, + "learning_rate": 5.105771251599633e-05, + "loss": 1.1063, + "step": 8145 + }, + { + "epoch": 0.49492678777568505, + "grad_norm": 0.15352259576320648, + "learning_rate": 5.104814186579631e-05, + "loss": 1.0369, + "step": 8146 + }, + { + "epoch": 0.4949875448083116, + "grad_norm": 0.2079126089811325, + "learning_rate": 5.103857117717647e-05, + "loss": 1.0625, + "step": 8147 + }, + { + "epoch": 0.49504830184093807, + "grad_norm": 2.71842098236084, + "learning_rate": 5.102900045048758e-05, + "loss": 1.071, + "step": 8148 + }, + { + "epoch": 0.4951090588735646, + "grad_norm": 0.2165210247039795, + "learning_rate": 5.101942968608048e-05, + "loss": 1.0842, + "step": 8149 + }, + { + "epoch": 0.49516981590619114, + "grad_norm": 0.9022283554077148, + "learning_rate": 5.1009858884305995e-05, + "loss": 1.1223, + "step": 8150 + }, + { + "epoch": 0.49523057293881767, + "grad_norm": 0.17668123543262482, + "learning_rate": 5.100028804551493e-05, + "loss": 1.097, + "step": 8151 + }, + { + "epoch": 0.4952913299714442, + "grad_norm": 0.29282906651496887, + "learning_rate": 5.099071717005812e-05, + "loss": 1.146, + "step": 8152 + }, + { + "epoch": 0.49535208700407074, + "grad_norm": 0.2608746290206909, + "learning_rate": 5.0981146258286385e-05, + "loss": 1.0779, + "step": 8153 + }, + { + "epoch": 0.4954128440366973, + "grad_norm": 0.8095966577529907, + "learning_rate": 5.0971575310550546e-05, + "loss": 1.1914, + "step": 8154 + }, + { + "epoch": 0.49547360106932375, + "grad_norm": 0.21081599593162537, + "learning_rate": 5.0962004327201416e-05, + "loss": 1.0941, + "step": 8155 + }, + { + "epoch": 0.4955343581019503, + "grad_norm": 0.26327022910118103, + "learning_rate": 5.0952433308589844e-05, + "loss": 1.11, + "step": 8156 + }, + { + "epoch": 0.4955951151345768, + "grad_norm": 0.20215706527233124, + "learning_rate": 5.094286225506664e-05, + "loss": 1.0067, + "step": 8157 + }, + { + "epoch": 0.49565587216720336, + "grad_norm": 0.21349425613880157, + "learning_rate": 5.093329116698263e-05, + "loss": 1.0662, + "step": 8158 + }, + { + "epoch": 0.4957166291998299, + "grad_norm": 0.40824735164642334, + "learning_rate": 5.092372004468865e-05, + "loss": 1.0828, + "step": 8159 + }, + { + "epoch": 0.49577738623245643, + "grad_norm": 0.21477216482162476, + "learning_rate": 5.091414888853554e-05, + "loss": 1.0749, + "step": 8160 + }, + { + "epoch": 0.4958381432650829, + "grad_norm": 0.1776292771100998, + "learning_rate": 5.090457769887412e-05, + "loss": 1.0421, + "step": 8161 + }, + { + "epoch": 0.49589890029770944, + "grad_norm": 0.38771185278892517, + "learning_rate": 5.089500647605524e-05, + "loss": 1.1759, + "step": 8162 + }, + { + "epoch": 0.495959657330336, + "grad_norm": 0.1728784590959549, + "learning_rate": 5.088543522042972e-05, + "loss": 1.0329, + "step": 8163 + }, + { + "epoch": 0.4960204143629625, + "grad_norm": 0.18178987503051758, + "learning_rate": 5.08758639323484e-05, + "loss": 1.0763, + "step": 8164 + }, + { + "epoch": 0.49608117139558905, + "grad_norm": 0.263397753238678, + "learning_rate": 5.086629261216211e-05, + "loss": 1.0709, + "step": 8165 + }, + { + "epoch": 0.4961419284282156, + "grad_norm": 0.1619594395160675, + "learning_rate": 5.0856721260221716e-05, + "loss": 1.0626, + "step": 8166 + }, + { + "epoch": 0.4962026854608421, + "grad_norm": 0.1985131800174713, + "learning_rate": 5.084714987687803e-05, + "loss": 1.0852, + "step": 8167 + }, + { + "epoch": 0.4962634424934686, + "grad_norm": 0.25719335675239563, + "learning_rate": 5.08375784624819e-05, + "loss": 1.1421, + "step": 8168 + }, + { + "epoch": 0.49632419952609513, + "grad_norm": 0.1936257779598236, + "learning_rate": 5.082800701738416e-05, + "loss": 1.0928, + "step": 8169 + }, + { + "epoch": 0.49638495655872167, + "grad_norm": 0.7118650078773499, + "learning_rate": 5.0818435541935675e-05, + "loss": 1.1254, + "step": 8170 + }, + { + "epoch": 0.4964457135913482, + "grad_norm": 0.19936439394950867, + "learning_rate": 5.080886403648728e-05, + "loss": 1.1454, + "step": 8171 + }, + { + "epoch": 0.49650647062397474, + "grad_norm": 0.44367027282714844, + "learning_rate": 5.079929250138981e-05, + "loss": 1.0751, + "step": 8172 + }, + { + "epoch": 0.4965672276566013, + "grad_norm": 0.17225752770900726, + "learning_rate": 5.0789720936994115e-05, + "loss": 1.0732, + "step": 8173 + }, + { + "epoch": 0.49662798468922775, + "grad_norm": 0.23918890953063965, + "learning_rate": 5.078014934365105e-05, + "loss": 1.0264, + "step": 8174 + }, + { + "epoch": 0.4966887417218543, + "grad_norm": 0.20254014432430267, + "learning_rate": 5.077057772171146e-05, + "loss": 1.1512, + "step": 8175 + }, + { + "epoch": 0.4967494987544808, + "grad_norm": 0.19923125207424164, + "learning_rate": 5.076100607152621e-05, + "loss": 1.026, + "step": 8176 + }, + { + "epoch": 0.49681025578710736, + "grad_norm": 0.15291431546211243, + "learning_rate": 5.0751434393446116e-05, + "loss": 1.0038, + "step": 8177 + }, + { + "epoch": 0.4968710128197339, + "grad_norm": 0.15776586532592773, + "learning_rate": 5.074186268782205e-05, + "loss": 1.0309, + "step": 8178 + }, + { + "epoch": 0.4969317698523604, + "grad_norm": 0.26682406663894653, + "learning_rate": 5.073229095500488e-05, + "loss": 1.1215, + "step": 8179 + }, + { + "epoch": 0.49699252688498696, + "grad_norm": 0.2540592551231384, + "learning_rate": 5.0722719195345415e-05, + "loss": 1.118, + "step": 8180 + }, + { + "epoch": 0.49705328391761344, + "grad_norm": 0.26085564494132996, + "learning_rate": 5.071314740919457e-05, + "loss": 1.1761, + "step": 8181 + }, + { + "epoch": 0.49711404095024, + "grad_norm": 0.17339974641799927, + "learning_rate": 5.070357559690314e-05, + "loss": 1.0336, + "step": 8182 + }, + { + "epoch": 0.4971747979828665, + "grad_norm": 0.23360802233219147, + "learning_rate": 5.069400375882203e-05, + "loss": 1.126, + "step": 8183 + }, + { + "epoch": 0.49723555501549305, + "grad_norm": 0.1652679294347763, + "learning_rate": 5.0684431895302086e-05, + "loss": 1.0562, + "step": 8184 + }, + { + "epoch": 0.4972963120481196, + "grad_norm": 0.21372082829475403, + "learning_rate": 5.067486000669414e-05, + "loss": 1.1579, + "step": 8185 + }, + { + "epoch": 0.4973570690807461, + "grad_norm": 0.8117876052856445, + "learning_rate": 5.066528809334907e-05, + "loss": 1.1195, + "step": 8186 + }, + { + "epoch": 0.4974178261133726, + "grad_norm": 0.18187633156776428, + "learning_rate": 5.065571615561774e-05, + "loss": 1.0892, + "step": 8187 + }, + { + "epoch": 0.49747858314599913, + "grad_norm": 0.22736068069934845, + "learning_rate": 5.064614419385101e-05, + "loss": 1.2401, + "step": 8188 + }, + { + "epoch": 0.49753934017862567, + "grad_norm": 0.17300830781459808, + "learning_rate": 5.0636572208399743e-05, + "loss": 1.0463, + "step": 8189 + }, + { + "epoch": 0.4976000972112522, + "grad_norm": 0.3146394193172455, + "learning_rate": 5.062700019961478e-05, + "loss": 1.1547, + "step": 8190 + }, + { + "epoch": 0.49766085424387874, + "grad_norm": 0.25705644488334656, + "learning_rate": 5.061742816784704e-05, + "loss": 1.1139, + "step": 8191 + }, + { + "epoch": 0.49772161127650527, + "grad_norm": 0.20825406908988953, + "learning_rate": 5.0607856113447324e-05, + "loss": 1.1092, + "step": 8192 + }, + { + "epoch": 0.4977823683091318, + "grad_norm": 0.341325044631958, + "learning_rate": 5.059828403676652e-05, + "loss": 1.0897, + "step": 8193 + }, + { + "epoch": 0.4978431253417583, + "grad_norm": 5.687345027923584, + "learning_rate": 5.058871193815552e-05, + "loss": 1.0987, + "step": 8194 + }, + { + "epoch": 0.4979038823743848, + "grad_norm": 0.20438973605632782, + "learning_rate": 5.0579139817965146e-05, + "loss": 1.0228, + "step": 8195 + }, + { + "epoch": 0.49796463940701136, + "grad_norm": 0.19468270242214203, + "learning_rate": 5.0569567676546304e-05, + "loss": 1.0973, + "step": 8196 + }, + { + "epoch": 0.4980253964396379, + "grad_norm": 0.27129459381103516, + "learning_rate": 5.055999551424986e-05, + "loss": 1.0885, + "step": 8197 + }, + { + "epoch": 0.4980861534722644, + "grad_norm": 0.17092439532279968, + "learning_rate": 5.055042333142666e-05, + "loss": 1.0321, + "step": 8198 + }, + { + "epoch": 0.49814691050489096, + "grad_norm": 0.20761604607105255, + "learning_rate": 5.054085112842758e-05, + "loss": 1.0426, + "step": 8199 + }, + { + "epoch": 0.49820766753751744, + "grad_norm": 0.23158133029937744, + "learning_rate": 5.053127890560351e-05, + "loss": 1.1288, + "step": 8200 + }, + { + "epoch": 0.498268424570144, + "grad_norm": 0.22966934740543365, + "learning_rate": 5.05217066633053e-05, + "loss": 1.3035, + "step": 8201 + }, + { + "epoch": 0.4983291816027705, + "grad_norm": 0.2019810527563095, + "learning_rate": 5.051213440188383e-05, + "loss": 1.1071, + "step": 8202 + }, + { + "epoch": 0.49838993863539705, + "grad_norm": 0.8647695779800415, + "learning_rate": 5.050256212168997e-05, + "loss": 1.0596, + "step": 8203 + }, + { + "epoch": 0.4984506956680236, + "grad_norm": 0.3693030774593353, + "learning_rate": 5.049298982307462e-05, + "loss": 1.1393, + "step": 8204 + }, + { + "epoch": 0.4985114527006501, + "grad_norm": 0.25215888023376465, + "learning_rate": 5.048341750638862e-05, + "loss": 1.0697, + "step": 8205 + }, + { + "epoch": 0.49857220973327665, + "grad_norm": 0.20735566318035126, + "learning_rate": 5.0473845171982844e-05, + "loss": 1.0416, + "step": 8206 + }, + { + "epoch": 0.49863296676590313, + "grad_norm": 0.17883895337581635, + "learning_rate": 5.0464272820208214e-05, + "loss": 1.008, + "step": 8207 + }, + { + "epoch": 0.49869372379852966, + "grad_norm": 0.3284088373184204, + "learning_rate": 5.045470045141554e-05, + "loss": 1.1015, + "step": 8208 + }, + { + "epoch": 0.4987544808311562, + "grad_norm": 0.1435699462890625, + "learning_rate": 5.044512806595575e-05, + "loss": 1.0701, + "step": 8209 + }, + { + "epoch": 0.49881523786378273, + "grad_norm": 0.2953360974788666, + "learning_rate": 5.04355556641797e-05, + "loss": 1.1973, + "step": 8210 + }, + { + "epoch": 0.49887599489640927, + "grad_norm": 0.2505517303943634, + "learning_rate": 5.042598324643828e-05, + "loss": 1.0803, + "step": 8211 + }, + { + "epoch": 0.4989367519290358, + "grad_norm": 0.5078453421592712, + "learning_rate": 5.041641081308236e-05, + "loss": 1.0037, + "step": 8212 + }, + { + "epoch": 0.49899750896166234, + "grad_norm": 1.1556284427642822, + "learning_rate": 5.0406838364462804e-05, + "loss": 1.0326, + "step": 8213 + }, + { + "epoch": 0.4990582659942888, + "grad_norm": 0.2433190792798996, + "learning_rate": 5.039726590093053e-05, + "loss": 1.2376, + "step": 8214 + }, + { + "epoch": 0.49911902302691535, + "grad_norm": 0.18118420243263245, + "learning_rate": 5.03876934228364e-05, + "loss": 1.0542, + "step": 8215 + }, + { + "epoch": 0.4991797800595419, + "grad_norm": 10.443779945373535, + "learning_rate": 5.037812093053127e-05, + "loss": 1.0751, + "step": 8216 + }, + { + "epoch": 0.4992405370921684, + "grad_norm": 0.3541370928287506, + "learning_rate": 5.036854842436608e-05, + "loss": 1.1199, + "step": 8217 + }, + { + "epoch": 0.49930129412479496, + "grad_norm": 0.19256718456745148, + "learning_rate": 5.035897590469164e-05, + "loss": 1.1256, + "step": 8218 + }, + { + "epoch": 0.4993620511574215, + "grad_norm": 0.16446253657341003, + "learning_rate": 5.0349403371858886e-05, + "loss": 1.0089, + "step": 8219 + }, + { + "epoch": 0.499422808190048, + "grad_norm": 0.2226303666830063, + "learning_rate": 5.033983082621868e-05, + "loss": 1.1213, + "step": 8220 + }, + { + "epoch": 0.4994835652226745, + "grad_norm": 1.332863688468933, + "learning_rate": 5.033025826812191e-05, + "loss": 1.0561, + "step": 8221 + }, + { + "epoch": 0.49954432225530104, + "grad_norm": 0.24998003244400024, + "learning_rate": 5.0320685697919465e-05, + "loss": 1.1075, + "step": 8222 + }, + { + "epoch": 0.4996050792879276, + "grad_norm": 0.2404552400112152, + "learning_rate": 5.0311113115962215e-05, + "loss": 1.1034, + "step": 8223 + }, + { + "epoch": 0.4996658363205541, + "grad_norm": 0.32798367738723755, + "learning_rate": 5.030154052260106e-05, + "loss": 1.1327, + "step": 8224 + }, + { + "epoch": 0.49972659335318065, + "grad_norm": 0.2546255588531494, + "learning_rate": 5.0291967918186875e-05, + "loss": 1.0999, + "step": 8225 + }, + { + "epoch": 0.4997873503858072, + "grad_norm": 0.2445601522922516, + "learning_rate": 5.028239530307054e-05, + "loss": 1.0681, + "step": 8226 + }, + { + "epoch": 0.49984810741843366, + "grad_norm": 0.22328834235668182, + "learning_rate": 5.027282267760296e-05, + "loss": 0.9897, + "step": 8227 + }, + { + "epoch": 0.4999088644510602, + "grad_norm": 0.19838844239711761, + "learning_rate": 5.026325004213504e-05, + "loss": 1.0791, + "step": 8228 + }, + { + "epoch": 0.49996962148368673, + "grad_norm": 0.2523570656776428, + "learning_rate": 5.02536773970176e-05, + "loss": 1.0724, + "step": 8229 + }, + { + "epoch": 0.5000303785163133, + "grad_norm": 0.24826928973197937, + "learning_rate": 5.024410474260157e-05, + "loss": 1.0645, + "step": 8230 + }, + { + "epoch": 0.5000911355489398, + "grad_norm": 0.6889215111732483, + "learning_rate": 5.023453207923784e-05, + "loss": 1.4693, + "step": 8231 + }, + { + "epoch": 0.5001518925815663, + "grad_norm": 0.5733199119567871, + "learning_rate": 5.0224959407277296e-05, + "loss": 1.0365, + "step": 8232 + }, + { + "epoch": 0.5002126496141929, + "grad_norm": 0.20520754158496857, + "learning_rate": 5.0215386727070824e-05, + "loss": 1.0884, + "step": 8233 + }, + { + "epoch": 0.5002734066468194, + "grad_norm": 0.2768336236476898, + "learning_rate": 5.02058140389693e-05, + "loss": 1.0754, + "step": 8234 + }, + { + "epoch": 0.5003341636794459, + "grad_norm": 0.21394649147987366, + "learning_rate": 5.019624134332363e-05, + "loss": 1.0475, + "step": 8235 + }, + { + "epoch": 0.5003949207120725, + "grad_norm": 0.23658888041973114, + "learning_rate": 5.018666864048469e-05, + "loss": 1.1899, + "step": 8236 + }, + { + "epoch": 0.5004556777446989, + "grad_norm": 0.18030115962028503, + "learning_rate": 5.017709593080338e-05, + "loss": 1.0283, + "step": 8237 + }, + { + "epoch": 0.5005164347773254, + "grad_norm": 0.3575654625892639, + "learning_rate": 5.0167523214630586e-05, + "loss": 1.0891, + "step": 8238 + }, + { + "epoch": 0.500577191809952, + "grad_norm": 0.2031727284193039, + "learning_rate": 5.015795049231717e-05, + "loss": 1.0439, + "step": 8239 + }, + { + "epoch": 0.5006379488425785, + "grad_norm": 0.14056791365146637, + "learning_rate": 5.0148377764214084e-05, + "loss": 1.0618, + "step": 8240 + }, + { + "epoch": 0.500698705875205, + "grad_norm": 0.1523967683315277, + "learning_rate": 5.0138805030672176e-05, + "loss": 1.0517, + "step": 8241 + }, + { + "epoch": 0.5007594629078316, + "grad_norm": 0.28027790784835815, + "learning_rate": 5.0129232292042325e-05, + "loss": 1.0235, + "step": 8242 + }, + { + "epoch": 0.5008202199404581, + "grad_norm": 0.7287334203720093, + "learning_rate": 5.0119659548675446e-05, + "loss": 1.0748, + "step": 8243 + }, + { + "epoch": 0.5008809769730846, + "grad_norm": 0.27022311091423035, + "learning_rate": 5.011008680092243e-05, + "loss": 1.1026, + "step": 8244 + }, + { + "epoch": 0.5009417340057112, + "grad_norm": 0.2327490746974945, + "learning_rate": 5.010051404913416e-05, + "loss": 1.0762, + "step": 8245 + }, + { + "epoch": 0.5010024910383377, + "grad_norm": 0.20018459856510162, + "learning_rate": 5.009094129366153e-05, + "loss": 1.0982, + "step": 8246 + }, + { + "epoch": 0.5010632480709643, + "grad_norm": 0.4067973792552948, + "learning_rate": 5.008136853485541e-05, + "loss": 1.036, + "step": 8247 + }, + { + "epoch": 0.5011240051035908, + "grad_norm": 0.1380934715270996, + "learning_rate": 5.007179577306674e-05, + "loss": 1.1001, + "step": 8248 + }, + { + "epoch": 0.5011847621362173, + "grad_norm": 0.185125470161438, + "learning_rate": 5.006222300864637e-05, + "loss": 1.0734, + "step": 8249 + }, + { + "epoch": 0.5012455191688437, + "grad_norm": 0.27016109228134155, + "learning_rate": 5.005265024194519e-05, + "loss": 1.1485, + "step": 8250 + }, + { + "epoch": 0.5013062762014703, + "grad_norm": 0.7843911647796631, + "learning_rate": 5.004307747331413e-05, + "loss": 1.0634, + "step": 8251 + }, + { + "epoch": 0.5013670332340968, + "grad_norm": 0.4777998626232147, + "learning_rate": 5.003350470310403e-05, + "loss": 1.2149, + "step": 8252 + }, + { + "epoch": 0.5014277902667233, + "grad_norm": 0.4653027355670929, + "learning_rate": 5.002393193166582e-05, + "loss": 1.1084, + "step": 8253 + }, + { + "epoch": 0.5014885472993499, + "grad_norm": 0.40227171778678894, + "learning_rate": 5.0014359159350385e-05, + "loss": 1.0713, + "step": 8254 + }, + { + "epoch": 0.5015493043319764, + "grad_norm": 0.3518075942993164, + "learning_rate": 5.0004786386508616e-05, + "loss": 1.1206, + "step": 8255 + }, + { + "epoch": 0.501610061364603, + "grad_norm": 0.4867675006389618, + "learning_rate": 4.9995213613491396e-05, + "loss": 0.9991, + "step": 8256 + }, + { + "epoch": 0.5016708183972295, + "grad_norm": 0.360953152179718, + "learning_rate": 4.998564084064962e-05, + "loss": 1.0151, + "step": 8257 + }, + { + "epoch": 0.501731575429856, + "grad_norm": 0.40292972326278687, + "learning_rate": 4.9976068068334176e-05, + "loss": 1.045, + "step": 8258 + }, + { + "epoch": 0.5017923324624826, + "grad_norm": 0.32671424746513367, + "learning_rate": 4.9966495296895974e-05, + "loss": 1.136, + "step": 8259 + }, + { + "epoch": 0.5018530894951091, + "grad_norm": 0.2950899004936218, + "learning_rate": 4.995692252668589e-05, + "loss": 1.0622, + "step": 8260 + }, + { + "epoch": 0.5019138465277356, + "grad_norm": 5.081988334655762, + "learning_rate": 4.9947349758054816e-05, + "loss": 1.1071, + "step": 8261 + }, + { + "epoch": 0.5019746035603622, + "grad_norm": 0.5468086004257202, + "learning_rate": 4.993777699135364e-05, + "loss": 1.08, + "step": 8262 + }, + { + "epoch": 0.5020353605929886, + "grad_norm": 0.3869243562221527, + "learning_rate": 4.9928204226933264e-05, + "loss": 1.2303, + "step": 8263 + }, + { + "epoch": 0.5020961176256151, + "grad_norm": 0.1941031813621521, + "learning_rate": 4.991863146514459e-05, + "loss": 1.0434, + "step": 8264 + }, + { + "epoch": 0.5021568746582417, + "grad_norm": 0.40264517068862915, + "learning_rate": 4.990905870633848e-05, + "loss": 1.0696, + "step": 8265 + }, + { + "epoch": 0.5022176316908682, + "grad_norm": 0.501337468624115, + "learning_rate": 4.989948595086586e-05, + "loss": 1.2877, + "step": 8266 + }, + { + "epoch": 0.5022783887234947, + "grad_norm": 0.2265487015247345, + "learning_rate": 4.988991319907758e-05, + "loss": 1.0701, + "step": 8267 + }, + { + "epoch": 0.5023391457561213, + "grad_norm": 0.4228810966014862, + "learning_rate": 4.988034045132456e-05, + "loss": 1.0327, + "step": 8268 + }, + { + "epoch": 0.5023999027887478, + "grad_norm": 0.3455961048603058, + "learning_rate": 4.9870767707957686e-05, + "loss": 1.1779, + "step": 8269 + }, + { + "epoch": 0.5024606598213743, + "grad_norm": 0.39894402027130127, + "learning_rate": 4.9861194969327835e-05, + "loss": 1.0985, + "step": 8270 + }, + { + "epoch": 0.5025214168540009, + "grad_norm": 0.2011348307132721, + "learning_rate": 4.985162223578592e-05, + "loss": 1.0313, + "step": 8271 + }, + { + "epoch": 0.5025821738866274, + "grad_norm": 0.39950379729270935, + "learning_rate": 4.984204950768282e-05, + "loss": 1.1286, + "step": 8272 + }, + { + "epoch": 0.5026429309192539, + "grad_norm": 1.3239810466766357, + "learning_rate": 4.983247678536944e-05, + "loss": 1.1146, + "step": 8273 + }, + { + "epoch": 0.5027036879518805, + "grad_norm": 0.19271592795848846, + "learning_rate": 4.982290406919664e-05, + "loss": 1.0964, + "step": 8274 + }, + { + "epoch": 0.502764444984507, + "grad_norm": 0.26949405670166016, + "learning_rate": 4.981333135951532e-05, + "loss": 1.1392, + "step": 8275 + }, + { + "epoch": 0.5028252020171334, + "grad_norm": 0.4041360020637512, + "learning_rate": 4.980375865667638e-05, + "loss": 1.071, + "step": 8276 + }, + { + "epoch": 0.50288595904976, + "grad_norm": 0.17877624928951263, + "learning_rate": 4.979418596103071e-05, + "loss": 1.0653, + "step": 8277 + }, + { + "epoch": 0.5029467160823865, + "grad_norm": 0.15727831423282623, + "learning_rate": 4.978461327292918e-05, + "loss": 1.0633, + "step": 8278 + }, + { + "epoch": 0.503007473115013, + "grad_norm": 0.2671699523925781, + "learning_rate": 4.977504059272273e-05, + "loss": 1.1095, + "step": 8279 + }, + { + "epoch": 0.5030682301476396, + "grad_norm": 0.23585407435894012, + "learning_rate": 4.9765467920762164e-05, + "loss": 1.1794, + "step": 8280 + }, + { + "epoch": 0.5031289871802661, + "grad_norm": 0.3933916389942169, + "learning_rate": 4.975589525739844e-05, + "loss": 1.1089, + "step": 8281 + }, + { + "epoch": 0.5031897442128926, + "grad_norm": 0.1952313780784607, + "learning_rate": 4.974632260298242e-05, + "loss": 1.0315, + "step": 8282 + }, + { + "epoch": 0.5032505012455192, + "grad_norm": 0.2035033404827118, + "learning_rate": 4.973674995786498e-05, + "loss": 1.1676, + "step": 8283 + }, + { + "epoch": 0.5033112582781457, + "grad_norm": 0.26324570178985596, + "learning_rate": 4.9727177322397034e-05, + "loss": 1.0755, + "step": 8284 + }, + { + "epoch": 0.5033720153107722, + "grad_norm": 0.38725632429122925, + "learning_rate": 4.971760469692946e-05, + "loss": 1.1501, + "step": 8285 + }, + { + "epoch": 0.5034327723433988, + "grad_norm": 0.14480295777320862, + "learning_rate": 4.9708032081813144e-05, + "loss": 1.0281, + "step": 8286 + }, + { + "epoch": 0.5034935293760253, + "grad_norm": 0.2777232527732849, + "learning_rate": 4.969845947739896e-05, + "loss": 1.0953, + "step": 8287 + }, + { + "epoch": 0.5035542864086519, + "grad_norm": 0.2686401605606079, + "learning_rate": 4.96888868840378e-05, + "loss": 1.2153, + "step": 8288 + }, + { + "epoch": 0.5036150434412783, + "grad_norm": 0.3003457188606262, + "learning_rate": 4.967931430208055e-05, + "loss": 1.0593, + "step": 8289 + }, + { + "epoch": 0.5036758004739048, + "grad_norm": 0.22697874903678894, + "learning_rate": 4.96697417318781e-05, + "loss": 1.0035, + "step": 8290 + }, + { + "epoch": 0.5037365575065313, + "grad_norm": 0.28194892406463623, + "learning_rate": 4.9660169173781316e-05, + "loss": 1.2416, + "step": 8291 + }, + { + "epoch": 0.5037973145391579, + "grad_norm": 0.2653176188468933, + "learning_rate": 4.965059662814112e-05, + "loss": 1.0722, + "step": 8292 + }, + { + "epoch": 0.5038580715717844, + "grad_norm": 0.4104539155960083, + "learning_rate": 4.964102409530837e-05, + "loss": 0.9946, + "step": 8293 + }, + { + "epoch": 0.503918828604411, + "grad_norm": 0.15923184156417847, + "learning_rate": 4.963145157563395e-05, + "loss": 1.088, + "step": 8294 + }, + { + "epoch": 0.5039795856370375, + "grad_norm": 0.36037835478782654, + "learning_rate": 4.962187906946873e-05, + "loss": 1.0251, + "step": 8295 + }, + { + "epoch": 0.504040342669664, + "grad_norm": 2.212118625640869, + "learning_rate": 4.961230657716361e-05, + "loss": 1.06, + "step": 8296 + }, + { + "epoch": 0.5041010997022906, + "grad_norm": 4.713129997253418, + "learning_rate": 4.960273409906947e-05, + "loss": 1.0258, + "step": 8297 + }, + { + "epoch": 0.5041618567349171, + "grad_norm": 0.2231234312057495, + "learning_rate": 4.9593161635537194e-05, + "loss": 1.1255, + "step": 8298 + }, + { + "epoch": 0.5042226137675436, + "grad_norm": 0.24116811156272888, + "learning_rate": 4.958358918691764e-05, + "loss": 1.0771, + "step": 8299 + }, + { + "epoch": 0.5042833708001702, + "grad_norm": 0.2011670172214508, + "learning_rate": 4.9574016753561744e-05, + "loss": 1.0831, + "step": 8300 + }, + { + "epoch": 0.5043441278327967, + "grad_norm": 0.3340749442577362, + "learning_rate": 4.956444433582031e-05, + "loss": 1.0657, + "step": 8301 + }, + { + "epoch": 0.5044048848654231, + "grad_norm": 0.49979835748672485, + "learning_rate": 4.955487193404426e-05, + "loss": 1.0305, + "step": 8302 + }, + { + "epoch": 0.5044656418980497, + "grad_norm": 0.34475070238113403, + "learning_rate": 4.954529954858447e-05, + "loss": 1.1703, + "step": 8303 + }, + { + "epoch": 0.5045263989306762, + "grad_norm": 0.5290508270263672, + "learning_rate": 4.9535727179791804e-05, + "loss": 1.0386, + "step": 8304 + }, + { + "epoch": 0.5045871559633027, + "grad_norm": 0.29755330085754395, + "learning_rate": 4.9526154828017154e-05, + "loss": 1.224, + "step": 8305 + }, + { + "epoch": 0.5046479129959293, + "grad_norm": 0.17868642508983612, + "learning_rate": 4.95165824936114e-05, + "loss": 1.0599, + "step": 8306 + }, + { + "epoch": 0.5047086700285558, + "grad_norm": 1.4540058374404907, + "learning_rate": 4.950701017692541e-05, + "loss": 1.1147, + "step": 8307 + }, + { + "epoch": 0.5047694270611823, + "grad_norm": 0.35950079560279846, + "learning_rate": 4.9497437878310045e-05, + "loss": 1.0554, + "step": 8308 + }, + { + "epoch": 0.5048301840938089, + "grad_norm": 0.2150409072637558, + "learning_rate": 4.9487865598116177e-05, + "loss": 1.0517, + "step": 8309 + }, + { + "epoch": 0.5048909411264354, + "grad_norm": 0.15251877903938293, + "learning_rate": 4.947829333669471e-05, + "loss": 1.0753, + "step": 8310 + }, + { + "epoch": 0.5049516981590619, + "grad_norm": 0.1598837673664093, + "learning_rate": 4.946872109439651e-05, + "loss": 1.0195, + "step": 8311 + }, + { + "epoch": 0.5050124551916885, + "grad_norm": 0.25122588872909546, + "learning_rate": 4.945914887157242e-05, + "loss": 1.1703, + "step": 8312 + }, + { + "epoch": 0.505073212224315, + "grad_norm": 0.19530469179153442, + "learning_rate": 4.9449576668573374e-05, + "loss": 1.0009, + "step": 8313 + }, + { + "epoch": 0.5051339692569415, + "grad_norm": 0.15277330577373505, + "learning_rate": 4.9440004485750165e-05, + "loss": 1.0521, + "step": 8314 + }, + { + "epoch": 0.505194726289568, + "grad_norm": 0.1940031796693802, + "learning_rate": 4.943043232345371e-05, + "loss": 1.1934, + "step": 8315 + }, + { + "epoch": 0.5052554833221945, + "grad_norm": 0.6129154562950134, + "learning_rate": 4.942086018203486e-05, + "loss": 1.2076, + "step": 8316 + }, + { + "epoch": 0.505316240354821, + "grad_norm": 0.16677913069725037, + "learning_rate": 4.941128806184449e-05, + "loss": 1.0356, + "step": 8317 + }, + { + "epoch": 0.5053769973874476, + "grad_norm": 0.19937124848365784, + "learning_rate": 4.940171596323348e-05, + "loss": 1.088, + "step": 8318 + }, + { + "epoch": 0.5054377544200741, + "grad_norm": 0.25840383768081665, + "learning_rate": 4.939214388655269e-05, + "loss": 1.226, + "step": 8319 + }, + { + "epoch": 0.5054985114527006, + "grad_norm": 0.17602191865444183, + "learning_rate": 4.938257183215299e-05, + "loss": 1.0555, + "step": 8320 + }, + { + "epoch": 0.5055592684853272, + "grad_norm": 0.1881481111049652, + "learning_rate": 4.937299980038523e-05, + "loss": 1.1354, + "step": 8321 + }, + { + "epoch": 0.5056200255179537, + "grad_norm": 0.16850614547729492, + "learning_rate": 4.936342779160027e-05, + "loss": 1.1078, + "step": 8322 + }, + { + "epoch": 0.5056807825505802, + "grad_norm": 0.17757046222686768, + "learning_rate": 4.9353855806149003e-05, + "loss": 1.0198, + "step": 8323 + }, + { + "epoch": 0.5057415395832068, + "grad_norm": 0.2864724397659302, + "learning_rate": 4.934428384438227e-05, + "loss": 1.0673, + "step": 8324 + }, + { + "epoch": 0.5058022966158333, + "grad_norm": 0.25471657514572144, + "learning_rate": 4.933471190665093e-05, + "loss": 1.1365, + "step": 8325 + }, + { + "epoch": 0.5058630536484598, + "grad_norm": 0.22538802027702332, + "learning_rate": 4.9325139993305894e-05, + "loss": 1.1099, + "step": 8326 + }, + { + "epoch": 0.5059238106810864, + "grad_norm": 0.18097706139087677, + "learning_rate": 4.931556810469794e-05, + "loss": 1.0658, + "step": 8327 + }, + { + "epoch": 0.5059845677137128, + "grad_norm": 0.16827042400836945, + "learning_rate": 4.930599624117798e-05, + "loss": 1.0279, + "step": 8328 + }, + { + "epoch": 0.5060453247463393, + "grad_norm": 0.688311755657196, + "learning_rate": 4.929642440309687e-05, + "loss": 1.0539, + "step": 8329 + }, + { + "epoch": 0.5061060817789659, + "grad_norm": 0.22743169963359833, + "learning_rate": 4.9286852590805446e-05, + "loss": 1.0649, + "step": 8330 + }, + { + "epoch": 0.5061668388115924, + "grad_norm": 0.25786951184272766, + "learning_rate": 4.927728080465458e-05, + "loss": 1.1628, + "step": 8331 + }, + { + "epoch": 0.506227595844219, + "grad_norm": 0.1663835197687149, + "learning_rate": 4.9267709044995136e-05, + "loss": 1.0168, + "step": 8332 + }, + { + "epoch": 0.5062883528768455, + "grad_norm": 0.3063949644565582, + "learning_rate": 4.925813731217797e-05, + "loss": 1.2884, + "step": 8333 + }, + { + "epoch": 0.506349109909472, + "grad_norm": 0.9124975800514221, + "learning_rate": 4.92485656065539e-05, + "loss": 1.2994, + "step": 8334 + }, + { + "epoch": 0.5064098669420986, + "grad_norm": 0.24231532216072083, + "learning_rate": 4.923899392847381e-05, + "loss": 1.1115, + "step": 8335 + }, + { + "epoch": 0.5064706239747251, + "grad_norm": 0.3301926255226135, + "learning_rate": 4.922942227828855e-05, + "loss": 1.145, + "step": 8336 + }, + { + "epoch": 0.5065313810073516, + "grad_norm": 0.1696614921092987, + "learning_rate": 4.921985065634896e-05, + "loss": 1.0368, + "step": 8337 + }, + { + "epoch": 0.5065921380399782, + "grad_norm": 0.37339261174201965, + "learning_rate": 4.921027906300588e-05, + "loss": 1.0324, + "step": 8338 + }, + { + "epoch": 0.5066528950726047, + "grad_norm": 0.19580639898777008, + "learning_rate": 4.920070749861019e-05, + "loss": 1.0498, + "step": 8339 + }, + { + "epoch": 0.5067136521052312, + "grad_norm": 0.1741429567337036, + "learning_rate": 4.9191135963512744e-05, + "loss": 1.0373, + "step": 8340 + }, + { + "epoch": 0.5067744091378578, + "grad_norm": 0.17351596057415009, + "learning_rate": 4.918156445806434e-05, + "loss": 1.0253, + "step": 8341 + }, + { + "epoch": 0.5068351661704842, + "grad_norm": 0.21828864514827728, + "learning_rate": 4.917199298261585e-05, + "loss": 1.1723, + "step": 8342 + }, + { + "epoch": 0.5068959232031107, + "grad_norm": 0.21093842387199402, + "learning_rate": 4.916242153751811e-05, + "loss": 1.0689, + "step": 8343 + }, + { + "epoch": 0.5069566802357373, + "grad_norm": 0.21721325814723969, + "learning_rate": 4.9152850123121977e-05, + "loss": 1.034, + "step": 8344 + }, + { + "epoch": 0.5070174372683638, + "grad_norm": 0.20075462758541107, + "learning_rate": 4.9143278739778296e-05, + "loss": 1.1439, + "step": 8345 + }, + { + "epoch": 0.5070781943009903, + "grad_norm": 0.20121555030345917, + "learning_rate": 4.91337073878379e-05, + "loss": 1.0693, + "step": 8346 + }, + { + "epoch": 0.5071389513336169, + "grad_norm": 0.4003956913948059, + "learning_rate": 4.912413606765161e-05, + "loss": 1.2686, + "step": 8347 + }, + { + "epoch": 0.5071997083662434, + "grad_norm": 0.20956532657146454, + "learning_rate": 4.911456477957029e-05, + "loss": 1.0459, + "step": 8348 + }, + { + "epoch": 0.5072604653988699, + "grad_norm": 0.18146494030952454, + "learning_rate": 4.910499352394478e-05, + "loss": 1.0578, + "step": 8349 + }, + { + "epoch": 0.5073212224314965, + "grad_norm": 0.25708597898483276, + "learning_rate": 4.909542230112589e-05, + "loss": 1.1834, + "step": 8350 + }, + { + "epoch": 0.507381979464123, + "grad_norm": 0.19937890768051147, + "learning_rate": 4.9085851111464464e-05, + "loss": 1.029, + "step": 8351 + }, + { + "epoch": 0.5074427364967495, + "grad_norm": 0.19355802237987518, + "learning_rate": 4.9076279955311357e-05, + "loss": 1.0688, + "step": 8352 + }, + { + "epoch": 0.5075034935293761, + "grad_norm": 0.23764842748641968, + "learning_rate": 4.9066708833017396e-05, + "loss": 1.1637, + "step": 8353 + }, + { + "epoch": 0.5075642505620026, + "grad_norm": 0.20620958507061005, + "learning_rate": 4.905713774493339e-05, + "loss": 1.1997, + "step": 8354 + }, + { + "epoch": 0.507625007594629, + "grad_norm": 3.9620556831359863, + "learning_rate": 4.9047566691410174e-05, + "loss": 1.0827, + "step": 8355 + }, + { + "epoch": 0.5076857646272556, + "grad_norm": 7.365649223327637, + "learning_rate": 4.9037995672798595e-05, + "loss": 1.0573, + "step": 8356 + }, + { + "epoch": 0.5077465216598821, + "grad_norm": 0.20697364211082458, + "learning_rate": 4.9028424689449466e-05, + "loss": 1.0855, + "step": 8357 + }, + { + "epoch": 0.5078072786925086, + "grad_norm": 0.24241343140602112, + "learning_rate": 4.901885374171362e-05, + "loss": 1.0851, + "step": 8358 + }, + { + "epoch": 0.5078680357251352, + "grad_norm": 0.3855138123035431, + "learning_rate": 4.900928282994187e-05, + "loss": 1.064, + "step": 8359 + }, + { + "epoch": 0.5079287927577617, + "grad_norm": 0.2671750783920288, + "learning_rate": 4.8999711954485075e-05, + "loss": 1.1063, + "step": 8360 + }, + { + "epoch": 0.5079895497903882, + "grad_norm": 0.20583464205265045, + "learning_rate": 4.8990141115694024e-05, + "loss": 1.0292, + "step": 8361 + }, + { + "epoch": 0.5080503068230148, + "grad_norm": 0.22781920433044434, + "learning_rate": 4.898057031391953e-05, + "loss": 1.1234, + "step": 8362 + }, + { + "epoch": 0.5081110638556413, + "grad_norm": 0.17401471734046936, + "learning_rate": 4.897099954951243e-05, + "loss": 1.0441, + "step": 8363 + }, + { + "epoch": 0.5081718208882678, + "grad_norm": 0.21992945671081543, + "learning_rate": 4.896142882282355e-05, + "loss": 1.1126, + "step": 8364 + }, + { + "epoch": 0.5082325779208944, + "grad_norm": 0.23331819474697113, + "learning_rate": 4.895185813420369e-05, + "loss": 1.1378, + "step": 8365 + }, + { + "epoch": 0.5082933349535209, + "grad_norm": 0.30200889706611633, + "learning_rate": 4.8942287484003684e-05, + "loss": 1.1332, + "step": 8366 + }, + { + "epoch": 0.5083540919861474, + "grad_norm": 0.15863487124443054, + "learning_rate": 4.893271687257434e-05, + "loss": 1.0418, + "step": 8367 + }, + { + "epoch": 0.5084148490187739, + "grad_norm": 0.19647139310836792, + "learning_rate": 4.892314630026644e-05, + "loss": 1.1144, + "step": 8368 + }, + { + "epoch": 0.5084756060514004, + "grad_norm": 0.4112296402454376, + "learning_rate": 4.8913575767430834e-05, + "loss": 1.0632, + "step": 8369 + }, + { + "epoch": 0.5085363630840269, + "grad_norm": 0.15378917753696442, + "learning_rate": 4.890400527441832e-05, + "loss": 1.0217, + "step": 8370 + }, + { + "epoch": 0.5085971201166535, + "grad_norm": 1.3834155797958374, + "learning_rate": 4.889443482157969e-05, + "loss": 1.2294, + "step": 8371 + }, + { + "epoch": 0.50865787714928, + "grad_norm": 0.28758037090301514, + "learning_rate": 4.8884864409265786e-05, + "loss": 1.116, + "step": 8372 + }, + { + "epoch": 0.5087186341819065, + "grad_norm": 0.615566611289978, + "learning_rate": 4.887529403782739e-05, + "loss": 1.0497, + "step": 8373 + }, + { + "epoch": 0.5087793912145331, + "grad_norm": 0.2549811899662018, + "learning_rate": 4.886572370761531e-05, + "loss": 1.0458, + "step": 8374 + }, + { + "epoch": 0.5088401482471596, + "grad_norm": 0.22646912932395935, + "learning_rate": 4.885615341898034e-05, + "loss": 1.066, + "step": 8375 + }, + { + "epoch": 0.5089009052797862, + "grad_norm": 0.7892501354217529, + "learning_rate": 4.8846583172273275e-05, + "loss": 1.0698, + "step": 8376 + }, + { + "epoch": 0.5089616623124127, + "grad_norm": 0.17599248886108398, + "learning_rate": 4.8837012967844944e-05, + "loss": 1.0354, + "step": 8377 + }, + { + "epoch": 0.5090224193450392, + "grad_norm": 0.24540600180625916, + "learning_rate": 4.882744280604612e-05, + "loss": 1.0899, + "step": 8378 + }, + { + "epoch": 0.5090831763776658, + "grad_norm": 0.28214186429977417, + "learning_rate": 4.8817872687227594e-05, + "loss": 1.0887, + "step": 8379 + }, + { + "epoch": 0.5091439334102923, + "grad_norm": 0.15951783955097198, + "learning_rate": 4.8808302611740215e-05, + "loss": 1.0507, + "step": 8380 + }, + { + "epoch": 0.5092046904429187, + "grad_norm": 2.525709390640259, + "learning_rate": 4.87987325799347e-05, + "loss": 1.0659, + "step": 8381 + }, + { + "epoch": 0.5092654474755453, + "grad_norm": 0.2444886565208435, + "learning_rate": 4.878916259216188e-05, + "loss": 1.0763, + "step": 8382 + }, + { + "epoch": 0.5093262045081718, + "grad_norm": 0.2378772646188736, + "learning_rate": 4.877959264877254e-05, + "loss": 1.0252, + "step": 8383 + }, + { + "epoch": 0.5093869615407983, + "grad_norm": 0.2676268219947815, + "learning_rate": 4.877002275011746e-05, + "loss": 1.1438, + "step": 8384 + }, + { + "epoch": 0.5094477185734249, + "grad_norm": 0.27486440539360046, + "learning_rate": 4.876045289654744e-05, + "loss": 1.0593, + "step": 8385 + }, + { + "epoch": 0.5095084756060514, + "grad_norm": 0.14516188204288483, + "learning_rate": 4.875088308841326e-05, + "loss": 1.0092, + "step": 8386 + }, + { + "epoch": 0.5095692326386779, + "grad_norm": 0.21178965270519257, + "learning_rate": 4.8741313326065715e-05, + "loss": 1.036, + "step": 8387 + }, + { + "epoch": 0.5096299896713045, + "grad_norm": 0.22538858652114868, + "learning_rate": 4.873174360985557e-05, + "loss": 1.0922, + "step": 8388 + }, + { + "epoch": 0.509690746703931, + "grad_norm": 0.1463756412267685, + "learning_rate": 4.872217394013359e-05, + "loss": 1.0551, + "step": 8389 + }, + { + "epoch": 0.5097515037365575, + "grad_norm": 0.20578160881996155, + "learning_rate": 4.8712604317250576e-05, + "loss": 1.1027, + "step": 8390 + }, + { + "epoch": 0.5098122607691841, + "grad_norm": 0.2149629443883896, + "learning_rate": 4.87030347415573e-05, + "loss": 1.1288, + "step": 8391 + }, + { + "epoch": 0.5098730178018106, + "grad_norm": 0.1702302247285843, + "learning_rate": 4.869346521340453e-05, + "loss": 1.0004, + "step": 8392 + }, + { + "epoch": 0.5099337748344371, + "grad_norm": 0.18756654858589172, + "learning_rate": 4.868389573314308e-05, + "loss": 1.1217, + "step": 8393 + }, + { + "epoch": 0.5099945318670636, + "grad_norm": 0.34728094935417175, + "learning_rate": 4.867432630112365e-05, + "loss": 1.0238, + "step": 8394 + }, + { + "epoch": 0.5100552888996901, + "grad_norm": 0.240312397480011, + "learning_rate": 4.8664756917697066e-05, + "loss": 1.1023, + "step": 8395 + }, + { + "epoch": 0.5101160459323166, + "grad_norm": 0.6608803272247314, + "learning_rate": 4.8655187583214066e-05, + "loss": 1.024, + "step": 8396 + }, + { + "epoch": 0.5101768029649432, + "grad_norm": 0.22536662220954895, + "learning_rate": 4.8645618298025415e-05, + "loss": 1.0739, + "step": 8397 + }, + { + "epoch": 0.5102375599975697, + "grad_norm": 0.22818554937839508, + "learning_rate": 4.86360490624819e-05, + "loss": 1.1352, + "step": 8398 + }, + { + "epoch": 0.5102983170301962, + "grad_norm": 0.1539105772972107, + "learning_rate": 4.862647987693427e-05, + "loss": 0.9977, + "step": 8399 + }, + { + "epoch": 0.5103590740628228, + "grad_norm": 0.3032917082309723, + "learning_rate": 4.8616910741733294e-05, + "loss": 1.0902, + "step": 8400 + }, + { + "epoch": 0.5104198310954493, + "grad_norm": 0.19875693321228027, + "learning_rate": 4.860734165722971e-05, + "loss": 1.0483, + "step": 8401 + }, + { + "epoch": 0.5104805881280758, + "grad_norm": 0.14216336607933044, + "learning_rate": 4.859777262377428e-05, + "loss": 1.0217, + "step": 8402 + }, + { + "epoch": 0.5105413451607024, + "grad_norm": 0.20183049142360687, + "learning_rate": 4.858820364171777e-05, + "loss": 1.0636, + "step": 8403 + }, + { + "epoch": 0.5106021021933289, + "grad_norm": 0.29829126596450806, + "learning_rate": 4.8578634711410933e-05, + "loss": 1.0617, + "step": 8404 + }, + { + "epoch": 0.5106628592259554, + "grad_norm": 0.22386310994625092, + "learning_rate": 4.856906583320451e-05, + "loss": 1.0373, + "step": 8405 + }, + { + "epoch": 0.510723616258582, + "grad_norm": 0.148743137717247, + "learning_rate": 4.855949700744926e-05, + "loss": 1.0557, + "step": 8406 + }, + { + "epoch": 0.5107843732912084, + "grad_norm": 0.2074337750673294, + "learning_rate": 4.854992823449594e-05, + "loss": 1.1606, + "step": 8407 + }, + { + "epoch": 0.5108451303238349, + "grad_norm": 0.22323819994926453, + "learning_rate": 4.854035951469527e-05, + "loss": 1.0881, + "step": 8408 + }, + { + "epoch": 0.5109058873564615, + "grad_norm": 0.18411579728126526, + "learning_rate": 4.8530790848398e-05, + "loss": 1.1037, + "step": 8409 + }, + { + "epoch": 0.510966644389088, + "grad_norm": 0.19367371499538422, + "learning_rate": 4.8521222235954864e-05, + "loss": 1.1413, + "step": 8410 + }, + { + "epoch": 0.5110274014217145, + "grad_norm": 0.18465657532215118, + "learning_rate": 4.851165367771663e-05, + "loss": 1.0716, + "step": 8411 + }, + { + "epoch": 0.5110881584543411, + "grad_norm": 0.23191657662391663, + "learning_rate": 4.8502085174034015e-05, + "loss": 1.0233, + "step": 8412 + }, + { + "epoch": 0.5111489154869676, + "grad_norm": 0.1745026558637619, + "learning_rate": 4.8492516725257766e-05, + "loss": 1.0905, + "step": 8413 + }, + { + "epoch": 0.5112096725195941, + "grad_norm": 0.6412477493286133, + "learning_rate": 4.84829483317386e-05, + "loss": 1.0578, + "step": 8414 + }, + { + "epoch": 0.5112704295522207, + "grad_norm": 0.6531240940093994, + "learning_rate": 4.847337999382724e-05, + "loss": 1.1362, + "step": 8415 + }, + { + "epoch": 0.5113311865848472, + "grad_norm": 0.1458996832370758, + "learning_rate": 4.8463811711874446e-05, + "loss": 1.0177, + "step": 8416 + }, + { + "epoch": 0.5113919436174738, + "grad_norm": 0.1669774055480957, + "learning_rate": 4.8454243486230934e-05, + "loss": 1.1078, + "step": 8417 + }, + { + "epoch": 0.5114527006501003, + "grad_norm": 0.25376182794570923, + "learning_rate": 4.844467531724741e-05, + "loss": 1.2829, + "step": 8418 + }, + { + "epoch": 0.5115134576827268, + "grad_norm": 0.17448492348194122, + "learning_rate": 4.843510720527462e-05, + "loss": 1.0821, + "step": 8419 + }, + { + "epoch": 0.5115742147153532, + "grad_norm": 1.7603847980499268, + "learning_rate": 4.8425539150663294e-05, + "loss": 1.0836, + "step": 8420 + }, + { + "epoch": 0.5116349717479798, + "grad_norm": 0.3951775133609772, + "learning_rate": 4.8415971153764126e-05, + "loss": 1.0018, + "step": 8421 + }, + { + "epoch": 0.5116957287806063, + "grad_norm": 0.20730450749397278, + "learning_rate": 4.840640321492784e-05, + "loss": 1.1222, + "step": 8422 + }, + { + "epoch": 0.5117564858132329, + "grad_norm": 0.17273102700710297, + "learning_rate": 4.839683533450514e-05, + "loss": 1.051, + "step": 8423 + }, + { + "epoch": 0.5118172428458594, + "grad_norm": 0.5942500829696655, + "learning_rate": 4.8387267512846764e-05, + "loss": 1.0724, + "step": 8424 + }, + { + "epoch": 0.5118779998784859, + "grad_norm": 0.18783478438854218, + "learning_rate": 4.8377699750303404e-05, + "loss": 1.0904, + "step": 8425 + }, + { + "epoch": 0.5119387569111125, + "grad_norm": 0.3454197347164154, + "learning_rate": 4.836813204722577e-05, + "loss": 1.3102, + "step": 8426 + }, + { + "epoch": 0.511999513943739, + "grad_norm": 0.18253149092197418, + "learning_rate": 4.83585644039646e-05, + "loss": 1.0623, + "step": 8427 + }, + { + "epoch": 0.5120602709763655, + "grad_norm": 0.21053765714168549, + "learning_rate": 4.8348996820870535e-05, + "loss": 1.1343, + "step": 8428 + }, + { + "epoch": 0.5121210280089921, + "grad_norm": 0.1903291493654251, + "learning_rate": 4.833942929829433e-05, + "loss": 1.0293, + "step": 8429 + }, + { + "epoch": 0.5121817850416186, + "grad_norm": 0.15340793132781982, + "learning_rate": 4.8329861836586645e-05, + "loss": 1.0797, + "step": 8430 + }, + { + "epoch": 0.5122425420742451, + "grad_norm": 0.3817113935947418, + "learning_rate": 4.832029443609821e-05, + "loss": 1.0064, + "step": 8431 + }, + { + "epoch": 0.5123032991068717, + "grad_norm": 0.16961467266082764, + "learning_rate": 4.83107270971797e-05, + "loss": 1.0621, + "step": 8432 + }, + { + "epoch": 0.5123640561394981, + "grad_norm": 0.15949055552482605, + "learning_rate": 4.830115982018183e-05, + "loss": 1.0719, + "step": 8433 + }, + { + "epoch": 0.5124248131721246, + "grad_norm": 0.48601076006889343, + "learning_rate": 4.829159260545527e-05, + "loss": 1.1127, + "step": 8434 + }, + { + "epoch": 0.5124855702047512, + "grad_norm": 0.35952597856521606, + "learning_rate": 4.828202545335071e-05, + "loss": 1.1334, + "step": 8435 + }, + { + "epoch": 0.5125463272373777, + "grad_norm": 0.2469291239976883, + "learning_rate": 4.827245836421883e-05, + "loss": 1.0609, + "step": 8436 + }, + { + "epoch": 0.5126070842700042, + "grad_norm": 1.2532098293304443, + "learning_rate": 4.8262891338410334e-05, + "loss": 1.0957, + "step": 8437 + }, + { + "epoch": 0.5126678413026308, + "grad_norm": 0.3030371069908142, + "learning_rate": 4.825332437627589e-05, + "loss": 1.1575, + "step": 8438 + }, + { + "epoch": 0.5127285983352573, + "grad_norm": 0.1868620216846466, + "learning_rate": 4.824375747816616e-05, + "loss": 1.0613, + "step": 8439 + }, + { + "epoch": 0.5127893553678838, + "grad_norm": 0.2763935625553131, + "learning_rate": 4.8234190644431874e-05, + "loss": 1.1691, + "step": 8440 + }, + { + "epoch": 0.5128501124005104, + "grad_norm": 0.6206904649734497, + "learning_rate": 4.822462387542365e-05, + "loss": 1.0912, + "step": 8441 + }, + { + "epoch": 0.5129108694331369, + "grad_norm": 0.4281459450721741, + "learning_rate": 4.821505717149219e-05, + "loss": 1.1015, + "step": 8442 + }, + { + "epoch": 0.5129716264657634, + "grad_norm": 0.43776676058769226, + "learning_rate": 4.8205490532988154e-05, + "loss": 1.0941, + "step": 8443 + }, + { + "epoch": 0.51303238349839, + "grad_norm": 0.38347285985946655, + "learning_rate": 4.819592396026219e-05, + "loss": 1.1222, + "step": 8444 + }, + { + "epoch": 0.5130931405310165, + "grad_norm": 0.2629888951778412, + "learning_rate": 4.8186357453665e-05, + "loss": 1.0917, + "step": 8445 + }, + { + "epoch": 0.513153897563643, + "grad_norm": 0.21480029821395874, + "learning_rate": 4.8176791013547226e-05, + "loss": 1.078, + "step": 8446 + }, + { + "epoch": 0.5132146545962695, + "grad_norm": 0.18832212686538696, + "learning_rate": 4.816722464025954e-05, + "loss": 0.9976, + "step": 8447 + }, + { + "epoch": 0.513275411628896, + "grad_norm": 0.31812986731529236, + "learning_rate": 4.8157658334152566e-05, + "loss": 1.2205, + "step": 8448 + }, + { + "epoch": 0.5133361686615225, + "grad_norm": 0.24824932217597961, + "learning_rate": 4.8148092095576994e-05, + "loss": 1.1552, + "step": 8449 + }, + { + "epoch": 0.5133969256941491, + "grad_norm": 1.7643104791641235, + "learning_rate": 4.813852592488346e-05, + "loss": 1.3281, + "step": 8450 + }, + { + "epoch": 0.5134576827267756, + "grad_norm": 0.265901118516922, + "learning_rate": 4.812895982242262e-05, + "loss": 1.1347, + "step": 8451 + }, + { + "epoch": 0.5135184397594021, + "grad_norm": 0.2735331952571869, + "learning_rate": 4.811939378854512e-05, + "loss": 1.0572, + "step": 8452 + }, + { + "epoch": 0.5135791967920287, + "grad_norm": 0.17309542000293732, + "learning_rate": 4.81098278236016e-05, + "loss": 1.0864, + "step": 8453 + }, + { + "epoch": 0.5136399538246552, + "grad_norm": 3.8620998859405518, + "learning_rate": 4.810026192794273e-05, + "loss": 1.0211, + "step": 8454 + }, + { + "epoch": 0.5137007108572818, + "grad_norm": 0.2926393151283264, + "learning_rate": 4.809069610191911e-05, + "loss": 1.1101, + "step": 8455 + }, + { + "epoch": 0.5137614678899083, + "grad_norm": 1.1867011785507202, + "learning_rate": 4.808113034588138e-05, + "loss": 1.0184, + "step": 8456 + }, + { + "epoch": 0.5138222249225348, + "grad_norm": 0.1908213496208191, + "learning_rate": 4.80715646601802e-05, + "loss": 1.0821, + "step": 8457 + }, + { + "epoch": 0.5138829819551614, + "grad_norm": 0.18448446691036224, + "learning_rate": 4.806199904516619e-05, + "loss": 1.0473, + "step": 8458 + }, + { + "epoch": 0.5139437389877879, + "grad_norm": 0.35471311211586, + "learning_rate": 4.805243350118998e-05, + "loss": 1.0745, + "step": 8459 + }, + { + "epoch": 0.5140044960204143, + "grad_norm": 0.2805604636669159, + "learning_rate": 4.804286802860221e-05, + "loss": 1.2587, + "step": 8460 + }, + { + "epoch": 0.5140652530530408, + "grad_norm": 2.183051824569702, + "learning_rate": 4.803330262775346e-05, + "loss": 1.0809, + "step": 8461 + }, + { + "epoch": 0.5141260100856674, + "grad_norm": 0.5683111548423767, + "learning_rate": 4.8023737298994404e-05, + "loss": 1.1388, + "step": 8462 + }, + { + "epoch": 0.5141867671182939, + "grad_norm": 0.26664626598358154, + "learning_rate": 4.8014172042675636e-05, + "loss": 1.1324, + "step": 8463 + }, + { + "epoch": 0.5142475241509205, + "grad_norm": 0.6080256104469299, + "learning_rate": 4.800460685914776e-05, + "loss": 1.0653, + "step": 8464 + }, + { + "epoch": 0.514308281183547, + "grad_norm": 2.6850974559783936, + "learning_rate": 4.799504174876142e-05, + "loss": 1.0763, + "step": 8465 + }, + { + "epoch": 0.5143690382161735, + "grad_norm": 0.3416141867637634, + "learning_rate": 4.7985476711867215e-05, + "loss": 1.2903, + "step": 8466 + }, + { + "epoch": 0.5144297952488001, + "grad_norm": 0.35317638516426086, + "learning_rate": 4.797591174881576e-05, + "loss": 1.1088, + "step": 8467 + }, + { + "epoch": 0.5144905522814266, + "grad_norm": 0.31679823994636536, + "learning_rate": 4.796634685995764e-05, + "loss": 1.1256, + "step": 8468 + }, + { + "epoch": 0.5145513093140531, + "grad_norm": 0.18249650299549103, + "learning_rate": 4.795678204564345e-05, + "loss": 1.0715, + "step": 8469 + }, + { + "epoch": 0.5146120663466797, + "grad_norm": 0.19725148379802704, + "learning_rate": 4.794721730622383e-05, + "loss": 1.0493, + "step": 8470 + }, + { + "epoch": 0.5146728233793062, + "grad_norm": 1.3708916902542114, + "learning_rate": 4.793765264204935e-05, + "loss": 1.1309, + "step": 8471 + }, + { + "epoch": 0.5147335804119327, + "grad_norm": 0.2045522928237915, + "learning_rate": 4.79280880534706e-05, + "loss": 1.0973, + "step": 8472 + }, + { + "epoch": 0.5147943374445592, + "grad_norm": 0.17927329242229462, + "learning_rate": 4.791852354083821e-05, + "loss": 1.0256, + "step": 8473 + }, + { + "epoch": 0.5148550944771857, + "grad_norm": 0.2411138266324997, + "learning_rate": 4.790895910450273e-05, + "loss": 1.1083, + "step": 8474 + }, + { + "epoch": 0.5149158515098122, + "grad_norm": 0.22161926329135895, + "learning_rate": 4.789939474481476e-05, + "loss": 1.0955, + "step": 8475 + }, + { + "epoch": 0.5149766085424388, + "grad_norm": 0.19313666224479675, + "learning_rate": 4.788983046212488e-05, + "loss": 1.0753, + "step": 8476 + }, + { + "epoch": 0.5150373655750653, + "grad_norm": 1.1039561033248901, + "learning_rate": 4.7880266256783666e-05, + "loss": 1.077, + "step": 8477 + }, + { + "epoch": 0.5150981226076918, + "grad_norm": 0.20608150959014893, + "learning_rate": 4.78707021291417e-05, + "loss": 1.0071, + "step": 8478 + }, + { + "epoch": 0.5151588796403184, + "grad_norm": 0.27479755878448486, + "learning_rate": 4.7861138079549566e-05, + "loss": 1.0335, + "step": 8479 + }, + { + "epoch": 0.5152196366729449, + "grad_norm": 0.16679172217845917, + "learning_rate": 4.7851574108357845e-05, + "loss": 1.0778, + "step": 8480 + }, + { + "epoch": 0.5152803937055714, + "grad_norm": 0.2005607634782791, + "learning_rate": 4.784201021591707e-05, + "loss": 1.0776, + "step": 8481 + }, + { + "epoch": 0.515341150738198, + "grad_norm": 0.33597978949546814, + "learning_rate": 4.7832446402577825e-05, + "loss": 1.0479, + "step": 8482 + }, + { + "epoch": 0.5154019077708245, + "grad_norm": 0.24618178606033325, + "learning_rate": 4.7822882668690685e-05, + "loss": 1.1081, + "step": 8483 + }, + { + "epoch": 0.515462664803451, + "grad_norm": 0.19817900657653809, + "learning_rate": 4.78133190146062e-05, + "loss": 0.9781, + "step": 8484 + }, + { + "epoch": 0.5155234218360776, + "grad_norm": 0.2262955904006958, + "learning_rate": 4.7803755440674916e-05, + "loss": 1.1484, + "step": 8485 + }, + { + "epoch": 0.515584178868704, + "grad_norm": 0.17157413065433502, + "learning_rate": 4.7794191947247414e-05, + "loss": 1.0717, + "step": 8486 + }, + { + "epoch": 0.5156449359013305, + "grad_norm": 0.28720659017562866, + "learning_rate": 4.778462853467425e-05, + "loss": 1.1296, + "step": 8487 + }, + { + "epoch": 0.5157056929339571, + "grad_norm": 0.15859709680080414, + "learning_rate": 4.777506520330593e-05, + "loss": 0.997, + "step": 8488 + }, + { + "epoch": 0.5157664499665836, + "grad_norm": 0.6856150031089783, + "learning_rate": 4.7765501953493033e-05, + "loss": 1.1583, + "step": 8489 + }, + { + "epoch": 0.5158272069992101, + "grad_norm": 0.22150512039661407, + "learning_rate": 4.7755938785586086e-05, + "loss": 1.0802, + "step": 8490 + }, + { + "epoch": 0.5158879640318367, + "grad_norm": 0.18280759453773499, + "learning_rate": 4.7746375699935646e-05, + "loss": 1.0589, + "step": 8491 + }, + { + "epoch": 0.5159487210644632, + "grad_norm": 0.20087938010692596, + "learning_rate": 4.773681269689224e-05, + "loss": 1.062, + "step": 8492 + }, + { + "epoch": 0.5160094780970897, + "grad_norm": 0.46621841192245483, + "learning_rate": 4.772724977680639e-05, + "loss": 1.271, + "step": 8493 + }, + { + "epoch": 0.5160702351297163, + "grad_norm": 0.2155219316482544, + "learning_rate": 4.7717686940028675e-05, + "loss": 1.0454, + "step": 8494 + }, + { + "epoch": 0.5161309921623428, + "grad_norm": 0.27062565088272095, + "learning_rate": 4.770812418690954e-05, + "loss": 1.0471, + "step": 8495 + }, + { + "epoch": 0.5161917491949694, + "grad_norm": 0.5824044346809387, + "learning_rate": 4.769856151779959e-05, + "loss": 1.1867, + "step": 8496 + }, + { + "epoch": 0.5162525062275959, + "grad_norm": 0.3077709376811981, + "learning_rate": 4.768899893304929e-05, + "loss": 1.099, + "step": 8497 + }, + { + "epoch": 0.5163132632602224, + "grad_norm": 0.18326666951179504, + "learning_rate": 4.767943643300918e-05, + "loss": 1.0484, + "step": 8498 + }, + { + "epoch": 0.5163740202928488, + "grad_norm": 0.32287177443504333, + "learning_rate": 4.7669874018029785e-05, + "loss": 1.1542, + "step": 8499 + }, + { + "epoch": 0.5164347773254754, + "grad_norm": 0.22765077650547028, + "learning_rate": 4.766031168846162e-05, + "loss": 1.1085, + "step": 8500 + }, + { + "epoch": 0.5164955343581019, + "grad_norm": 0.2173711061477661, + "learning_rate": 4.765074944465517e-05, + "loss": 1.0717, + "step": 8501 + }, + { + "epoch": 0.5165562913907285, + "grad_norm": 0.1884503960609436, + "learning_rate": 4.764118728696096e-05, + "loss": 1.0948, + "step": 8502 + }, + { + "epoch": 0.516617048423355, + "grad_norm": 0.32299309968948364, + "learning_rate": 4.763162521572947e-05, + "loss": 0.9888, + "step": 8503 + }, + { + "epoch": 0.5166778054559815, + "grad_norm": 0.2179969847202301, + "learning_rate": 4.7622063231311225e-05, + "loss": 1.0916, + "step": 8504 + }, + { + "epoch": 0.516738562488608, + "grad_norm": 0.24838489294052124, + "learning_rate": 4.761250133405671e-05, + "loss": 1.0983, + "step": 8505 + }, + { + "epoch": 0.5167993195212346, + "grad_norm": 0.16738073527812958, + "learning_rate": 4.760293952431642e-05, + "loss": 1.0455, + "step": 8506 + }, + { + "epoch": 0.5168600765538611, + "grad_norm": 0.23851539194583893, + "learning_rate": 4.759337780244087e-05, + "loss": 1.0596, + "step": 8507 + }, + { + "epoch": 0.5169208335864877, + "grad_norm": 0.234355628490448, + "learning_rate": 4.758381616878049e-05, + "loss": 1.1127, + "step": 8508 + }, + { + "epoch": 0.5169815906191142, + "grad_norm": 0.292145699262619, + "learning_rate": 4.7574254623685806e-05, + "loss": 1.3777, + "step": 8509 + }, + { + "epoch": 0.5170423476517407, + "grad_norm": 0.20824772119522095, + "learning_rate": 4.756469316750729e-05, + "loss": 1.0486, + "step": 8510 + }, + { + "epoch": 0.5171031046843673, + "grad_norm": 0.24053503572940826, + "learning_rate": 4.7555131800595395e-05, + "loss": 1.0519, + "step": 8511 + }, + { + "epoch": 0.5171638617169937, + "grad_norm": 0.20052233338356018, + "learning_rate": 4.754557052330063e-05, + "loss": 1.1001, + "step": 8512 + }, + { + "epoch": 0.5172246187496202, + "grad_norm": 0.326852411031723, + "learning_rate": 4.753600933597346e-05, + "loss": 1.0734, + "step": 8513 + }, + { + "epoch": 0.5172853757822468, + "grad_norm": 0.20768706500530243, + "learning_rate": 4.752644823896436e-05, + "loss": 1.0818, + "step": 8514 + }, + { + "epoch": 0.5173461328148733, + "grad_norm": 0.17522835731506348, + "learning_rate": 4.751688723262376e-05, + "loss": 1.0557, + "step": 8515 + }, + { + "epoch": 0.5174068898474998, + "grad_norm": 0.24332275986671448, + "learning_rate": 4.750732631730212e-05, + "loss": 1.0524, + "step": 8516 + }, + { + "epoch": 0.5174676468801264, + "grad_norm": 0.22207848727703094, + "learning_rate": 4.749776549334994e-05, + "loss": 1.0975, + "step": 8517 + }, + { + "epoch": 0.5175284039127529, + "grad_norm": 0.1809588372707367, + "learning_rate": 4.7488204761117645e-05, + "loss": 1.0708, + "step": 8518 + }, + { + "epoch": 0.5175891609453794, + "grad_norm": 0.20144391059875488, + "learning_rate": 4.747864412095567e-05, + "loss": 1.0641, + "step": 8519 + }, + { + "epoch": 0.517649917978006, + "grad_norm": 0.16503779590129852, + "learning_rate": 4.7469083573214507e-05, + "loss": 1.0942, + "step": 8520 + }, + { + "epoch": 0.5177106750106325, + "grad_norm": 0.2766873240470886, + "learning_rate": 4.745952311824458e-05, + "loss": 1.2788, + "step": 8521 + }, + { + "epoch": 0.517771432043259, + "grad_norm": 0.22772829234600067, + "learning_rate": 4.744996275639631e-05, + "loss": 1.1454, + "step": 8522 + }, + { + "epoch": 0.5178321890758856, + "grad_norm": 0.32286950945854187, + "learning_rate": 4.744040248802015e-05, + "loss": 1.1332, + "step": 8523 + }, + { + "epoch": 0.5178929461085121, + "grad_norm": 0.32638970017433167, + "learning_rate": 4.7430842313466526e-05, + "loss": 1.1437, + "step": 8524 + }, + { + "epoch": 0.5179537031411385, + "grad_norm": 0.2093062400817871, + "learning_rate": 4.742128223308587e-05, + "loss": 1.0101, + "step": 8525 + }, + { + "epoch": 0.5180144601737651, + "grad_norm": 0.17655521631240845, + "learning_rate": 4.741172224722863e-05, + "loss": 1.0396, + "step": 8526 + }, + { + "epoch": 0.5180752172063916, + "grad_norm": 0.4212319552898407, + "learning_rate": 4.740216235624521e-05, + "loss": 1.1793, + "step": 8527 + }, + { + "epoch": 0.5181359742390181, + "grad_norm": 0.14244121313095093, + "learning_rate": 4.739260256048603e-05, + "loss": 1.0444, + "step": 8528 + }, + { + "epoch": 0.5181967312716447, + "grad_norm": 0.2686750292778015, + "learning_rate": 4.7383042860301486e-05, + "loss": 1.1137, + "step": 8529 + }, + { + "epoch": 0.5182574883042712, + "grad_norm": 0.1605362594127655, + "learning_rate": 4.737348325604203e-05, + "loss": 1.0196, + "step": 8530 + }, + { + "epoch": 0.5183182453368977, + "grad_norm": 0.2831780016422272, + "learning_rate": 4.736392374805805e-05, + "loss": 1.2475, + "step": 8531 + }, + { + "epoch": 0.5183790023695243, + "grad_norm": 0.17395345866680145, + "learning_rate": 4.7354364336699944e-05, + "loss": 1.066, + "step": 8532 + }, + { + "epoch": 0.5184397594021508, + "grad_norm": 0.293475478887558, + "learning_rate": 4.734480502231813e-05, + "loss": 1.2149, + "step": 8533 + }, + { + "epoch": 0.5185005164347773, + "grad_norm": 0.7362297773361206, + "learning_rate": 4.733524580526302e-05, + "loss": 1.3998, + "step": 8534 + }, + { + "epoch": 0.5185612734674039, + "grad_norm": 0.2028091847896576, + "learning_rate": 4.732568668588497e-05, + "loss": 1.0354, + "step": 8535 + }, + { + "epoch": 0.5186220305000304, + "grad_norm": 0.19668975472450256, + "learning_rate": 4.7316127664534406e-05, + "loss": 1.1224, + "step": 8536 + }, + { + "epoch": 0.518682787532657, + "grad_norm": 0.4082567095756531, + "learning_rate": 4.730656874156168e-05, + "loss": 1.0873, + "step": 8537 + }, + { + "epoch": 0.5187435445652834, + "grad_norm": 0.16568146646022797, + "learning_rate": 4.729700991731721e-05, + "loss": 0.9573, + "step": 8538 + }, + { + "epoch": 0.5188043015979099, + "grad_norm": 0.18939757347106934, + "learning_rate": 4.728745119215136e-05, + "loss": 1.0562, + "step": 8539 + }, + { + "epoch": 0.5188650586305364, + "grad_norm": 0.1677064299583435, + "learning_rate": 4.7277892566414505e-05, + "loss": 1.0295, + "step": 8540 + }, + { + "epoch": 0.518925815663163, + "grad_norm": 0.36986127495765686, + "learning_rate": 4.726833404045705e-05, + "loss": 1.1454, + "step": 8541 + }, + { + "epoch": 0.5189865726957895, + "grad_norm": 0.2969701290130615, + "learning_rate": 4.725877561462932e-05, + "loss": 1.1034, + "step": 8542 + }, + { + "epoch": 0.519047329728416, + "grad_norm": 0.22680231928825378, + "learning_rate": 4.72492172892817e-05, + "loss": 1.0683, + "step": 8543 + }, + { + "epoch": 0.5191080867610426, + "grad_norm": 0.22423718869686127, + "learning_rate": 4.7239659064764555e-05, + "loss": 1.1155, + "step": 8544 + }, + { + "epoch": 0.5191688437936691, + "grad_norm": 0.2142300009727478, + "learning_rate": 4.723010094142824e-05, + "loss": 1.0843, + "step": 8545 + }, + { + "epoch": 0.5192296008262957, + "grad_norm": 0.22921018302440643, + "learning_rate": 4.722054291962311e-05, + "loss": 1.1121, + "step": 8546 + }, + { + "epoch": 0.5192903578589222, + "grad_norm": 0.1932772696018219, + "learning_rate": 4.7210984999699525e-05, + "loss": 1.0972, + "step": 8547 + }, + { + "epoch": 0.5193511148915487, + "grad_norm": 0.3662955164909363, + "learning_rate": 4.720142718200781e-05, + "loss": 1.0357, + "step": 8548 + }, + { + "epoch": 0.5194118719241753, + "grad_norm": 0.24416397511959076, + "learning_rate": 4.719186946689832e-05, + "loss": 1.0857, + "step": 8549 + }, + { + "epoch": 0.5194726289568018, + "grad_norm": 0.23753342032432556, + "learning_rate": 4.7182311854721405e-05, + "loss": 1.1916, + "step": 8550 + }, + { + "epoch": 0.5195333859894283, + "grad_norm": 2.3685197830200195, + "learning_rate": 4.7172754345827395e-05, + "loss": 1.0556, + "step": 8551 + }, + { + "epoch": 0.5195941430220548, + "grad_norm": 0.39524737000465393, + "learning_rate": 4.716319694056662e-05, + "loss": 1.1472, + "step": 8552 + }, + { + "epoch": 0.5196549000546813, + "grad_norm": 0.21601912379264832, + "learning_rate": 4.7153639639289405e-05, + "loss": 1.0176, + "step": 8553 + }, + { + "epoch": 0.5197156570873078, + "grad_norm": 0.18951432406902313, + "learning_rate": 4.7144082442346096e-05, + "loss": 1.1086, + "step": 8554 + }, + { + "epoch": 0.5197764141199344, + "grad_norm": 0.17754366993904114, + "learning_rate": 4.713452535008698e-05, + "loss": 1.0628, + "step": 8555 + }, + { + "epoch": 0.5198371711525609, + "grad_norm": 2.1686081886291504, + "learning_rate": 4.71249683628624e-05, + "loss": 1.1413, + "step": 8556 + }, + { + "epoch": 0.5198979281851874, + "grad_norm": 0.16676998138427734, + "learning_rate": 4.711541148102265e-05, + "loss": 1.0087, + "step": 8557 + }, + { + "epoch": 0.519958685217814, + "grad_norm": 0.2818245589733124, + "learning_rate": 4.710585470491806e-05, + "loss": 1.1161, + "step": 8558 + }, + { + "epoch": 0.5200194422504405, + "grad_norm": 0.270172655582428, + "learning_rate": 4.7096298034898916e-05, + "loss": 1.0783, + "step": 8559 + }, + { + "epoch": 0.520080199283067, + "grad_norm": 0.29348498582839966, + "learning_rate": 4.7086741471315535e-05, + "loss": 1.0775, + "step": 8560 + }, + { + "epoch": 0.5201409563156936, + "grad_norm": 0.1927543431520462, + "learning_rate": 4.707718501451822e-05, + "loss": 1.1403, + "step": 8561 + }, + { + "epoch": 0.5202017133483201, + "grad_norm": 0.22708195447921753, + "learning_rate": 4.7067628664857234e-05, + "loss": 1.1406, + "step": 8562 + }, + { + "epoch": 0.5202624703809466, + "grad_norm": 0.19285310804843903, + "learning_rate": 4.705807242268289e-05, + "loss": 1.0484, + "step": 8563 + }, + { + "epoch": 0.5203232274135732, + "grad_norm": 0.17300379276275635, + "learning_rate": 4.704851628834548e-05, + "loss": 1.1145, + "step": 8564 + }, + { + "epoch": 0.5203839844461996, + "grad_norm": 0.13172629475593567, + "learning_rate": 4.703896026219525e-05, + "loss": 1.095, + "step": 8565 + }, + { + "epoch": 0.5204447414788261, + "grad_norm": 0.15263254940509796, + "learning_rate": 4.702940434458253e-05, + "loss": 1.0371, + "step": 8566 + }, + { + "epoch": 0.5205054985114527, + "grad_norm": 0.2412646859884262, + "learning_rate": 4.701984853585757e-05, + "loss": 1.1224, + "step": 8567 + }, + { + "epoch": 0.5205662555440792, + "grad_norm": 0.23671112954616547, + "learning_rate": 4.7010292836370636e-05, + "loss": 1.2462, + "step": 8568 + }, + { + "epoch": 0.5206270125767057, + "grad_norm": 0.34110501408576965, + "learning_rate": 4.7000737246471985e-05, + "loss": 1.0054, + "step": 8569 + }, + { + "epoch": 0.5206877696093323, + "grad_norm": 0.374319851398468, + "learning_rate": 4.699118176651188e-05, + "loss": 1.0959, + "step": 8570 + }, + { + "epoch": 0.5207485266419588, + "grad_norm": 0.3077484369277954, + "learning_rate": 4.69816263968406e-05, + "loss": 1.0298, + "step": 8571 + }, + { + "epoch": 0.5208092836745853, + "grad_norm": 0.3725254535675049, + "learning_rate": 4.6972071137808396e-05, + "loss": 1.0552, + "step": 8572 + }, + { + "epoch": 0.5208700407072119, + "grad_norm": 0.2622997760772705, + "learning_rate": 4.69625159897655e-05, + "loss": 1.1776, + "step": 8573 + }, + { + "epoch": 0.5209307977398384, + "grad_norm": 0.1841263473033905, + "learning_rate": 4.6952960953062194e-05, + "loss": 1.058, + "step": 8574 + }, + { + "epoch": 0.520991554772465, + "grad_norm": 0.2738814651966095, + "learning_rate": 4.694340602804866e-05, + "loss": 1.059, + "step": 8575 + }, + { + "epoch": 0.5210523118050915, + "grad_norm": 0.43523672223091125, + "learning_rate": 4.693385121507518e-05, + "loss": 1.1433, + "step": 8576 + }, + { + "epoch": 0.521113068837718, + "grad_norm": 0.22403864562511444, + "learning_rate": 4.6924296514491976e-05, + "loss": 1.1057, + "step": 8577 + }, + { + "epoch": 0.5211738258703444, + "grad_norm": 0.43765559792518616, + "learning_rate": 4.691474192664927e-05, + "loss": 1.1184, + "step": 8578 + }, + { + "epoch": 0.521234582902971, + "grad_norm": 1.787830114364624, + "learning_rate": 4.6905187451897306e-05, + "loss": 1.0691, + "step": 8579 + }, + { + "epoch": 0.5212953399355975, + "grad_norm": 2.979149103164673, + "learning_rate": 4.6895633090586287e-05, + "loss": 1.0278, + "step": 8580 + }, + { + "epoch": 0.521356096968224, + "grad_norm": 0.3662060499191284, + "learning_rate": 4.688607884306645e-05, + "loss": 1.1374, + "step": 8581 + }, + { + "epoch": 0.5214168540008506, + "grad_norm": 0.37562650442123413, + "learning_rate": 4.6876524709687995e-05, + "loss": 1.0339, + "step": 8582 + }, + { + "epoch": 0.5214776110334771, + "grad_norm": 0.23909929394721985, + "learning_rate": 4.686697069080111e-05, + "loss": 1.0005, + "step": 8583 + }, + { + "epoch": 0.5215383680661037, + "grad_norm": 0.31724080443382263, + "learning_rate": 4.6857416786756036e-05, + "loss": 1.1538, + "step": 8584 + }, + { + "epoch": 0.5215991250987302, + "grad_norm": 0.19796723127365112, + "learning_rate": 4.684786299790295e-05, + "loss": 1.0423, + "step": 8585 + }, + { + "epoch": 0.5216598821313567, + "grad_norm": 0.31002169847488403, + "learning_rate": 4.683830932459205e-05, + "loss": 1.1083, + "step": 8586 + }, + { + "epoch": 0.5217206391639833, + "grad_norm": 0.2076428383588791, + "learning_rate": 4.682875576717354e-05, + "loss": 1.1461, + "step": 8587 + }, + { + "epoch": 0.5217813961966098, + "grad_norm": 0.7217409014701843, + "learning_rate": 4.681920232599761e-05, + "loss": 1.0741, + "step": 8588 + }, + { + "epoch": 0.5218421532292363, + "grad_norm": 0.2279433310031891, + "learning_rate": 4.680964900141443e-05, + "loss": 1.0486, + "step": 8589 + }, + { + "epoch": 0.5219029102618629, + "grad_norm": 0.4954124689102173, + "learning_rate": 4.680009579377418e-05, + "loss": 1.0849, + "step": 8590 + }, + { + "epoch": 0.5219636672944893, + "grad_norm": 0.23280012607574463, + "learning_rate": 4.679054270342703e-05, + "loss": 1.0696, + "step": 8591 + }, + { + "epoch": 0.5220244243271158, + "grad_norm": 0.2656466066837311, + "learning_rate": 4.6780989730723155e-05, + "loss": 1.0537, + "step": 8592 + }, + { + "epoch": 0.5220851813597424, + "grad_norm": 0.48484793305397034, + "learning_rate": 4.677143687601274e-05, + "loss": 0.9976, + "step": 8593 + }, + { + "epoch": 0.5221459383923689, + "grad_norm": 0.33468562364578247, + "learning_rate": 4.676188413964593e-05, + "loss": 1.0073, + "step": 8594 + }, + { + "epoch": 0.5222066954249954, + "grad_norm": 3.006819486618042, + "learning_rate": 4.675233152197287e-05, + "loss": 1.0498, + "step": 8595 + }, + { + "epoch": 0.522267452457622, + "grad_norm": 0.39103060960769653, + "learning_rate": 4.6742779023343725e-05, + "loss": 1.0399, + "step": 8596 + }, + { + "epoch": 0.5223282094902485, + "grad_norm": 0.34185585379600525, + "learning_rate": 4.673322664410865e-05, + "loss": 1.0346, + "step": 8597 + }, + { + "epoch": 0.522388966522875, + "grad_norm": 0.41087085008621216, + "learning_rate": 4.672367438461778e-05, + "loss": 1.1316, + "step": 8598 + }, + { + "epoch": 0.5224497235555016, + "grad_norm": 0.41497451066970825, + "learning_rate": 4.671412224522126e-05, + "loss": 1.0857, + "step": 8599 + }, + { + "epoch": 0.5225104805881281, + "grad_norm": 0.3524736166000366, + "learning_rate": 4.670457022626922e-05, + "loss": 1.0511, + "step": 8600 + }, + { + "epoch": 0.5225712376207546, + "grad_norm": 0.18988481163978577, + "learning_rate": 4.6695018328111814e-05, + "loss": 1.1677, + "step": 8601 + }, + { + "epoch": 0.5226319946533812, + "grad_norm": 0.28644412755966187, + "learning_rate": 4.668546655109914e-05, + "loss": 1.1783, + "step": 8602 + }, + { + "epoch": 0.5226927516860077, + "grad_norm": 0.2920369505882263, + "learning_rate": 4.667591489558133e-05, + "loss": 1.0607, + "step": 8603 + }, + { + "epoch": 0.5227535087186341, + "grad_norm": 0.2984852194786072, + "learning_rate": 4.666636336190849e-05, + "loss": 1.0549, + "step": 8604 + }, + { + "epoch": 0.5228142657512607, + "grad_norm": 0.26738280057907104, + "learning_rate": 4.6656811950430754e-05, + "loss": 1.0391, + "step": 8605 + }, + { + "epoch": 0.5228750227838872, + "grad_norm": 0.22442780435085297, + "learning_rate": 4.664726066149822e-05, + "loss": 1.0435, + "step": 8606 + }, + { + "epoch": 0.5229357798165137, + "grad_norm": 0.3146284222602844, + "learning_rate": 4.663770949546099e-05, + "loss": 1.1761, + "step": 8607 + }, + { + "epoch": 0.5229965368491403, + "grad_norm": 0.318392813205719, + "learning_rate": 4.66281584526692e-05, + "loss": 1.302, + "step": 8608 + }, + { + "epoch": 0.5230572938817668, + "grad_norm": 0.2535770833492279, + "learning_rate": 4.661860753347288e-05, + "loss": 1.0515, + "step": 8609 + }, + { + "epoch": 0.5231180509143933, + "grad_norm": 0.18526317179203033, + "learning_rate": 4.660905673822217e-05, + "loss": 1.0235, + "step": 8610 + }, + { + "epoch": 0.5231788079470199, + "grad_norm": 0.27258166670799255, + "learning_rate": 4.659950606726713e-05, + "loss": 1.172, + "step": 8611 + }, + { + "epoch": 0.5232395649796464, + "grad_norm": 0.22199583053588867, + "learning_rate": 4.658995552095785e-05, + "loss": 1.0528, + "step": 8612 + }, + { + "epoch": 0.523300322012273, + "grad_norm": 0.23652096092700958, + "learning_rate": 4.658040509964441e-05, + "loss": 1.1067, + "step": 8613 + }, + { + "epoch": 0.5233610790448995, + "grad_norm": 0.3175472915172577, + "learning_rate": 4.657085480367691e-05, + "loss": 1.083, + "step": 8614 + }, + { + "epoch": 0.523421836077526, + "grad_norm": 0.4318819046020508, + "learning_rate": 4.6561304633405365e-05, + "loss": 1.0959, + "step": 8615 + }, + { + "epoch": 0.5234825931101525, + "grad_norm": 0.1710147261619568, + "learning_rate": 4.6551754589179867e-05, + "loss": 1.0022, + "step": 8616 + }, + { + "epoch": 0.523543350142779, + "grad_norm": 0.21420492231845856, + "learning_rate": 4.654220467135047e-05, + "loss": 1.1827, + "step": 8617 + }, + { + "epoch": 0.5236041071754055, + "grad_norm": 0.23011763393878937, + "learning_rate": 4.653265488026723e-05, + "loss": 1.2113, + "step": 8618 + }, + { + "epoch": 0.523664864208032, + "grad_norm": 0.4787297546863556, + "learning_rate": 4.6523105216280196e-05, + "loss": 1.1649, + "step": 8619 + }, + { + "epoch": 0.5237256212406586, + "grad_norm": 0.18796595931053162, + "learning_rate": 4.65135556797394e-05, + "loss": 1.0406, + "step": 8620 + }, + { + "epoch": 0.5237863782732851, + "grad_norm": 0.2257094532251358, + "learning_rate": 4.6504006270994934e-05, + "loss": 1.0218, + "step": 8621 + }, + { + "epoch": 0.5238471353059116, + "grad_norm": 2.5749943256378174, + "learning_rate": 4.649445699039675e-05, + "loss": 1.0489, + "step": 8622 + }, + { + "epoch": 0.5239078923385382, + "grad_norm": 0.739093542098999, + "learning_rate": 4.648490783829495e-05, + "loss": 1.0532, + "step": 8623 + }, + { + "epoch": 0.5239686493711647, + "grad_norm": 0.1745673418045044, + "learning_rate": 4.6475358815039514e-05, + "loss": 1.0673, + "step": 8624 + }, + { + "epoch": 0.5240294064037913, + "grad_norm": 0.3874518573284149, + "learning_rate": 4.6465809920980485e-05, + "loss": 1.1084, + "step": 8625 + }, + { + "epoch": 0.5240901634364178, + "grad_norm": 0.457267701625824, + "learning_rate": 4.645626115646788e-05, + "loss": 0.9837, + "step": 8626 + }, + { + "epoch": 0.5241509204690443, + "grad_norm": 0.2391200214624405, + "learning_rate": 4.644671252185171e-05, + "loss": 1.1521, + "step": 8627 + }, + { + "epoch": 0.5242116775016709, + "grad_norm": 0.32125139236450195, + "learning_rate": 4.643716401748199e-05, + "loss": 1.0638, + "step": 8628 + }, + { + "epoch": 0.5242724345342974, + "grad_norm": 0.25773096084594727, + "learning_rate": 4.64276156437087e-05, + "loss": 1.0863, + "step": 8629 + }, + { + "epoch": 0.5243331915669238, + "grad_norm": 0.15947604179382324, + "learning_rate": 4.641806740088184e-05, + "loss": 1.0637, + "step": 8630 + }, + { + "epoch": 0.5243939485995504, + "grad_norm": 0.18176445364952087, + "learning_rate": 4.640851928935142e-05, + "loss": 0.9997, + "step": 8631 + }, + { + "epoch": 0.5244547056321769, + "grad_norm": 0.2632804214954376, + "learning_rate": 4.6398971309467426e-05, + "loss": 1.0377, + "step": 8632 + }, + { + "epoch": 0.5245154626648034, + "grad_norm": 1.0730253458023071, + "learning_rate": 4.638942346157982e-05, + "loss": 1.0579, + "step": 8633 + }, + { + "epoch": 0.52457621969743, + "grad_norm": 1.8335888385772705, + "learning_rate": 4.6379875746038625e-05, + "loss": 1.3026, + "step": 8634 + }, + { + "epoch": 0.5246369767300565, + "grad_norm": 0.3631538152694702, + "learning_rate": 4.6370328163193754e-05, + "loss": 1.0883, + "step": 8635 + }, + { + "epoch": 0.524697733762683, + "grad_norm": 1.2544866800308228, + "learning_rate": 4.6360780713395214e-05, + "loss": 1.1761, + "step": 8636 + }, + { + "epoch": 0.5247584907953096, + "grad_norm": 0.2968771755695343, + "learning_rate": 4.6351233396992966e-05, + "loss": 1.17, + "step": 8637 + }, + { + "epoch": 0.5248192478279361, + "grad_norm": 0.4506833255290985, + "learning_rate": 4.634168621433694e-05, + "loss": 1.0658, + "step": 8638 + }, + { + "epoch": 0.5248800048605626, + "grad_norm": 0.33651822805404663, + "learning_rate": 4.633213916577714e-05, + "loss": 1.0747, + "step": 8639 + }, + { + "epoch": 0.5249407618931892, + "grad_norm": 0.2771355211734772, + "learning_rate": 4.632259225166348e-05, + "loss": 1.026, + "step": 8640 + }, + { + "epoch": 0.5250015189258157, + "grad_norm": 0.2644052803516388, + "learning_rate": 4.631304547234592e-05, + "loss": 1.1088, + "step": 8641 + }, + { + "epoch": 0.5250622759584422, + "grad_norm": 0.5024870038032532, + "learning_rate": 4.630349882817437e-05, + "loss": 1.1483, + "step": 8642 + }, + { + "epoch": 0.5251230329910687, + "grad_norm": 0.22417831420898438, + "learning_rate": 4.62939523194988e-05, + "loss": 1.153, + "step": 8643 + }, + { + "epoch": 0.5251837900236952, + "grad_norm": 0.4100082218647003, + "learning_rate": 4.628440594666912e-05, + "loss": 1.18, + "step": 8644 + }, + { + "epoch": 0.5252445470563217, + "grad_norm": 0.20493429899215698, + "learning_rate": 4.627485971003527e-05, + "loss": 0.99, + "step": 8645 + }, + { + "epoch": 0.5253053040889483, + "grad_norm": 0.17279714345932007, + "learning_rate": 4.626531360994713e-05, + "loss": 1.0466, + "step": 8646 + }, + { + "epoch": 0.5253660611215748, + "grad_norm": 0.3003229796886444, + "learning_rate": 4.625576764675466e-05, + "loss": 1.137, + "step": 8647 + }, + { + "epoch": 0.5254268181542013, + "grad_norm": 0.19564509391784668, + "learning_rate": 4.624622182080776e-05, + "loss": 1.0694, + "step": 8648 + }, + { + "epoch": 0.5254875751868279, + "grad_norm": 0.1911187320947647, + "learning_rate": 4.623667613245632e-05, + "loss": 1.0992, + "step": 8649 + }, + { + "epoch": 0.5255483322194544, + "grad_norm": 0.17246036231517792, + "learning_rate": 4.622713058205023e-05, + "loss": 1.0619, + "step": 8650 + }, + { + "epoch": 0.5256090892520809, + "grad_norm": 0.22989396750926971, + "learning_rate": 4.62175851699394e-05, + "loss": 1.0405, + "step": 8651 + }, + { + "epoch": 0.5256698462847075, + "grad_norm": 0.1909831464290619, + "learning_rate": 4.620803989647373e-05, + "loss": 1.1153, + "step": 8652 + }, + { + "epoch": 0.525730603317334, + "grad_norm": 0.4238590896129608, + "learning_rate": 4.6198494762003076e-05, + "loss": 1.0564, + "step": 8653 + }, + { + "epoch": 0.5257913603499605, + "grad_norm": 0.45940205454826355, + "learning_rate": 4.618894976687733e-05, + "loss": 1.1193, + "step": 8654 + }, + { + "epoch": 0.5258521173825871, + "grad_norm": 0.24307286739349365, + "learning_rate": 4.6179404911446386e-05, + "loss": 1.1415, + "step": 8655 + }, + { + "epoch": 0.5259128744152136, + "grad_norm": 0.269954115152359, + "learning_rate": 4.6169860196060084e-05, + "loss": 1.1389, + "step": 8656 + }, + { + "epoch": 0.52597363144784, + "grad_norm": 0.39620450139045715, + "learning_rate": 4.6160315621068294e-05, + "loss": 1.0138, + "step": 8657 + }, + { + "epoch": 0.5260343884804666, + "grad_norm": 0.7389344573020935, + "learning_rate": 4.615077118682087e-05, + "loss": 1.0754, + "step": 8658 + }, + { + "epoch": 0.5260951455130931, + "grad_norm": 0.43096667528152466, + "learning_rate": 4.6141226893667684e-05, + "loss": 1.0342, + "step": 8659 + }, + { + "epoch": 0.5261559025457196, + "grad_norm": 0.20412592589855194, + "learning_rate": 4.6131682741958566e-05, + "loss": 1.0926, + "step": 8660 + }, + { + "epoch": 0.5262166595783462, + "grad_norm": 0.6854451894760132, + "learning_rate": 4.612213873204338e-05, + "loss": 1.1594, + "step": 8661 + }, + { + "epoch": 0.5262774166109727, + "grad_norm": 0.3041723370552063, + "learning_rate": 4.611259486427194e-05, + "loss": 1.1043, + "step": 8662 + }, + { + "epoch": 0.5263381736435992, + "grad_norm": 0.21460001170635223, + "learning_rate": 4.610305113899407e-05, + "loss": 1.0738, + "step": 8663 + }, + { + "epoch": 0.5263989306762258, + "grad_norm": 0.16254273056983948, + "learning_rate": 4.609350755655963e-05, + "loss": 1.0889, + "step": 8664 + }, + { + "epoch": 0.5264596877088523, + "grad_norm": 1.3207886219024658, + "learning_rate": 4.608396411731842e-05, + "loss": 1.039, + "step": 8665 + }, + { + "epoch": 0.5265204447414789, + "grad_norm": 3.1131958961486816, + "learning_rate": 4.607442082162026e-05, + "loss": 1.1992, + "step": 8666 + }, + { + "epoch": 0.5265812017741054, + "grad_norm": 0.2376040816307068, + "learning_rate": 4.606487766981496e-05, + "loss": 1.0394, + "step": 8667 + }, + { + "epoch": 0.5266419588067319, + "grad_norm": 0.374767005443573, + "learning_rate": 4.605533466225235e-05, + "loss": 1.2485, + "step": 8668 + }, + { + "epoch": 0.5267027158393585, + "grad_norm": 0.20279432833194733, + "learning_rate": 4.60457917992822e-05, + "loss": 1.1138, + "step": 8669 + }, + { + "epoch": 0.5267634728719849, + "grad_norm": 0.2781737744808197, + "learning_rate": 4.603624908125432e-05, + "loss": 1.1938, + "step": 8670 + }, + { + "epoch": 0.5268242299046114, + "grad_norm": 0.8089534640312195, + "learning_rate": 4.6026706508518474e-05, + "loss": 1.1955, + "step": 8671 + }, + { + "epoch": 0.526884986937238, + "grad_norm": 0.20195825397968292, + "learning_rate": 4.601716408142448e-05, + "loss": 1.0505, + "step": 8672 + }, + { + "epoch": 0.5269457439698645, + "grad_norm": 0.2275310754776001, + "learning_rate": 4.600762180032212e-05, + "loss": 1.0567, + "step": 8673 + }, + { + "epoch": 0.527006501002491, + "grad_norm": 0.19209061563014984, + "learning_rate": 4.599807966556113e-05, + "loss": 1.0714, + "step": 8674 + }, + { + "epoch": 0.5270672580351176, + "grad_norm": 0.16299524903297424, + "learning_rate": 4.598853767749135e-05, + "loss": 1.0481, + "step": 8675 + }, + { + "epoch": 0.5271280150677441, + "grad_norm": 0.1494925320148468, + "learning_rate": 4.597899583646245e-05, + "loss": 1.0091, + "step": 8676 + }, + { + "epoch": 0.5271887721003706, + "grad_norm": 0.29467105865478516, + "learning_rate": 4.596945414282425e-05, + "loss": 1.1704, + "step": 8677 + }, + { + "epoch": 0.5272495291329972, + "grad_norm": 0.30434074997901917, + "learning_rate": 4.59599125969265e-05, + "loss": 1.2978, + "step": 8678 + }, + { + "epoch": 0.5273102861656237, + "grad_norm": 0.49404481053352356, + "learning_rate": 4.5950371199118906e-05, + "loss": 1.0825, + "step": 8679 + }, + { + "epoch": 0.5273710431982502, + "grad_norm": 0.24953167140483856, + "learning_rate": 4.5940829949751255e-05, + "loss": 0.9784, + "step": 8680 + }, + { + "epoch": 0.5274318002308768, + "grad_norm": 4.299622535705566, + "learning_rate": 4.593128884917328e-05, + "loss": 1.0917, + "step": 8681 + }, + { + "epoch": 0.5274925572635033, + "grad_norm": 0.19633279740810394, + "learning_rate": 4.592174789773469e-05, + "loss": 1.0837, + "step": 8682 + }, + { + "epoch": 0.5275533142961297, + "grad_norm": 0.2561893165111542, + "learning_rate": 4.591220709578521e-05, + "loss": 1.0819, + "step": 8683 + }, + { + "epoch": 0.5276140713287563, + "grad_norm": 0.16112151741981506, + "learning_rate": 4.590266644367457e-05, + "loss": 1.0473, + "step": 8684 + }, + { + "epoch": 0.5276748283613828, + "grad_norm": 0.23485927283763885, + "learning_rate": 4.589312594175249e-05, + "loss": 1.1847, + "step": 8685 + }, + { + "epoch": 0.5277355853940093, + "grad_norm": 0.6381456255912781, + "learning_rate": 4.588358559036867e-05, + "loss": 1.2066, + "step": 8686 + }, + { + "epoch": 0.5277963424266359, + "grad_norm": 0.5587844252586365, + "learning_rate": 4.5874045389872804e-05, + "loss": 1.2188, + "step": 8687 + }, + { + "epoch": 0.5278570994592624, + "grad_norm": 0.23315288126468658, + "learning_rate": 4.5864505340614636e-05, + "loss": 1.3463, + "step": 8688 + }, + { + "epoch": 0.5279178564918889, + "grad_norm": 0.23430277407169342, + "learning_rate": 4.5854965442943796e-05, + "loss": 1.0829, + "step": 8689 + }, + { + "epoch": 0.5279786135245155, + "grad_norm": 0.2447742223739624, + "learning_rate": 4.5845425697210005e-05, + "loss": 1.0595, + "step": 8690 + }, + { + "epoch": 0.528039370557142, + "grad_norm": 0.6450576782226562, + "learning_rate": 4.583588610376294e-05, + "loss": 1.2013, + "step": 8691 + }, + { + "epoch": 0.5281001275897685, + "grad_norm": 0.21756316721439362, + "learning_rate": 4.582634666295226e-05, + "loss": 1.1237, + "step": 8692 + }, + { + "epoch": 0.5281608846223951, + "grad_norm": 0.33508235216140747, + "learning_rate": 4.581680737512767e-05, + "loss": 1.0485, + "step": 8693 + }, + { + "epoch": 0.5282216416550216, + "grad_norm": 0.24591495096683502, + "learning_rate": 4.58072682406388e-05, + "loss": 1.0134, + "step": 8694 + }, + { + "epoch": 0.5282823986876481, + "grad_norm": 0.23450876772403717, + "learning_rate": 4.579772925983535e-05, + "loss": 1.1129, + "step": 8695 + }, + { + "epoch": 0.5283431557202746, + "grad_norm": 0.2687796950340271, + "learning_rate": 4.578819043306692e-05, + "loss": 1.1422, + "step": 8696 + }, + { + "epoch": 0.5284039127529011, + "grad_norm": 0.5016398429870605, + "learning_rate": 4.5778651760683185e-05, + "loss": 1.03, + "step": 8697 + }, + { + "epoch": 0.5284646697855276, + "grad_norm": 0.5293537974357605, + "learning_rate": 4.576911324303379e-05, + "loss": 1.2491, + "step": 8698 + }, + { + "epoch": 0.5285254268181542, + "grad_norm": 0.47100186347961426, + "learning_rate": 4.575957488046837e-05, + "loss": 1.1795, + "step": 8699 + }, + { + "epoch": 0.5285861838507807, + "grad_norm": 0.20102228224277496, + "learning_rate": 4.5750036673336535e-05, + "loss": 1.0363, + "step": 8700 + }, + { + "epoch": 0.5286469408834072, + "grad_norm": 0.25759556889533997, + "learning_rate": 4.574049862198797e-05, + "loss": 1.0936, + "step": 8701 + }, + { + "epoch": 0.5287076979160338, + "grad_norm": 0.5296412110328674, + "learning_rate": 4.573096072677221e-05, + "loss": 1.0939, + "step": 8702 + }, + { + "epoch": 0.5287684549486603, + "grad_norm": 0.33371326327323914, + "learning_rate": 4.572142298803893e-05, + "loss": 1.0119, + "step": 8703 + }, + { + "epoch": 0.5288292119812869, + "grad_norm": 0.45812341570854187, + "learning_rate": 4.571188540613771e-05, + "loss": 1.1332, + "step": 8704 + }, + { + "epoch": 0.5288899690139134, + "grad_norm": 0.22400186955928802, + "learning_rate": 4.5702347981418146e-05, + "loss": 1.1201, + "step": 8705 + }, + { + "epoch": 0.5289507260465399, + "grad_norm": 0.29025670886039734, + "learning_rate": 4.569281071422987e-05, + "loss": 1.093, + "step": 8706 + }, + { + "epoch": 0.5290114830791665, + "grad_norm": 0.297935426235199, + "learning_rate": 4.568327360492244e-05, + "loss": 1.1372, + "step": 8707 + }, + { + "epoch": 0.529072240111793, + "grad_norm": 2.126323699951172, + "learning_rate": 4.567373665384546e-05, + "loss": 1.0434, + "step": 8708 + }, + { + "epoch": 0.5291329971444194, + "grad_norm": 0.22692044079303741, + "learning_rate": 4.5664199861348495e-05, + "loss": 1.054, + "step": 8709 + }, + { + "epoch": 0.529193754177046, + "grad_norm": 0.8416318297386169, + "learning_rate": 4.5654663227781106e-05, + "loss": 1.0643, + "step": 8710 + }, + { + "epoch": 0.5292545112096725, + "grad_norm": 0.2404051274061203, + "learning_rate": 4.564512675349289e-05, + "loss": 1.0837, + "step": 8711 + }, + { + "epoch": 0.529315268242299, + "grad_norm": 0.27754124999046326, + "learning_rate": 4.563559043883339e-05, + "loss": 1.1657, + "step": 8712 + }, + { + "epoch": 0.5293760252749256, + "grad_norm": 0.3691059350967407, + "learning_rate": 4.5626054284152165e-05, + "loss": 1.1949, + "step": 8713 + }, + { + "epoch": 0.5294367823075521, + "grad_norm": 0.21631082892417908, + "learning_rate": 4.561651828979877e-05, + "loss": 1.1633, + "step": 8714 + }, + { + "epoch": 0.5294975393401786, + "grad_norm": 0.3909175395965576, + "learning_rate": 4.560698245612275e-05, + "loss": 1.0459, + "step": 8715 + }, + { + "epoch": 0.5295582963728052, + "grad_norm": 0.1706586331129074, + "learning_rate": 4.559744678347363e-05, + "loss": 1.0365, + "step": 8716 + }, + { + "epoch": 0.5296190534054317, + "grad_norm": 0.2772104740142822, + "learning_rate": 4.558791127220096e-05, + "loss": 1.0044, + "step": 8717 + }, + { + "epoch": 0.5296798104380582, + "grad_norm": 0.20732614398002625, + "learning_rate": 4.5578375922654235e-05, + "loss": 1.1125, + "step": 8718 + }, + { + "epoch": 0.5297405674706848, + "grad_norm": 0.26610267162323, + "learning_rate": 4.5568840735183005e-05, + "loss": 1.0649, + "step": 8719 + }, + { + "epoch": 0.5298013245033113, + "grad_norm": 0.1740850955247879, + "learning_rate": 4.555930571013678e-05, + "loss": 1.0994, + "step": 8720 + }, + { + "epoch": 0.5298620815359378, + "grad_norm": 0.4105677902698517, + "learning_rate": 4.554977084786507e-05, + "loss": 1.0887, + "step": 8721 + }, + { + "epoch": 0.5299228385685643, + "grad_norm": 0.2515830993652344, + "learning_rate": 4.5540236148717356e-05, + "loss": 1.0865, + "step": 8722 + }, + { + "epoch": 0.5299835956011908, + "grad_norm": 0.15293537080287933, + "learning_rate": 4.553070161304314e-05, + "loss": 1.0087, + "step": 8723 + }, + { + "epoch": 0.5300443526338173, + "grad_norm": 0.4082357883453369, + "learning_rate": 4.552116724119193e-05, + "loss": 1.0852, + "step": 8724 + }, + { + "epoch": 0.5301051096664439, + "grad_norm": 0.28447583317756653, + "learning_rate": 4.55116330335132e-05, + "loss": 1.0412, + "step": 8725 + }, + { + "epoch": 0.5301658666990704, + "grad_norm": 0.19071820378303528, + "learning_rate": 4.5502098990356425e-05, + "loss": 1.0509, + "step": 8726 + }, + { + "epoch": 0.5302266237316969, + "grad_norm": 0.23470738530158997, + "learning_rate": 4.549256511207108e-05, + "loss": 1.0599, + "step": 8727 + }, + { + "epoch": 0.5302873807643235, + "grad_norm": 0.31913673877716064, + "learning_rate": 4.548303139900666e-05, + "loss": 1.0002, + "step": 8728 + }, + { + "epoch": 0.53034813779695, + "grad_norm": 0.2100980132818222, + "learning_rate": 4.547349785151257e-05, + "loss": 1.0213, + "step": 8729 + }, + { + "epoch": 0.5304088948295765, + "grad_norm": 0.6650189161300659, + "learning_rate": 4.546396446993829e-05, + "loss": 1.029, + "step": 8730 + }, + { + "epoch": 0.5304696518622031, + "grad_norm": 0.20567159354686737, + "learning_rate": 4.545443125463327e-05, + "loss": 1.0826, + "step": 8731 + }, + { + "epoch": 0.5305304088948296, + "grad_norm": 10.988689422607422, + "learning_rate": 4.544489820594696e-05, + "loss": 1.1196, + "step": 8732 + }, + { + "epoch": 0.5305911659274561, + "grad_norm": 0.2841789126396179, + "learning_rate": 4.543536532422878e-05, + "loss": 1.1644, + "step": 8733 + }, + { + "epoch": 0.5306519229600827, + "grad_norm": 0.2592669427394867, + "learning_rate": 4.5425832609828156e-05, + "loss": 1.1513, + "step": 8734 + }, + { + "epoch": 0.5307126799927091, + "grad_norm": 0.19854146242141724, + "learning_rate": 4.541630006309456e-05, + "loss": 0.9931, + "step": 8735 + }, + { + "epoch": 0.5307734370253356, + "grad_norm": 0.17326615750789642, + "learning_rate": 4.5406767684377336e-05, + "loss": 1.054, + "step": 8736 + }, + { + "epoch": 0.5308341940579622, + "grad_norm": 0.14202813804149628, + "learning_rate": 4.539723547402594e-05, + "loss": 0.9811, + "step": 8737 + }, + { + "epoch": 0.5308949510905887, + "grad_norm": 0.8649278283119202, + "learning_rate": 4.5387703432389775e-05, + "loss": 1.1635, + "step": 8738 + }, + { + "epoch": 0.5309557081232152, + "grad_norm": 0.23742815852165222, + "learning_rate": 4.537817155981821e-05, + "loss": 1.1623, + "step": 8739 + }, + { + "epoch": 0.5310164651558418, + "grad_norm": 6.8257012367248535, + "learning_rate": 4.5368639856660674e-05, + "loss": 1.2524, + "step": 8740 + }, + { + "epoch": 0.5310772221884683, + "grad_norm": 0.19616559147834778, + "learning_rate": 4.535910832326654e-05, + "loss": 1.105, + "step": 8741 + }, + { + "epoch": 0.5311379792210948, + "grad_norm": 0.22356723248958588, + "learning_rate": 4.53495769599852e-05, + "loss": 1.0543, + "step": 8742 + }, + { + "epoch": 0.5311987362537214, + "grad_norm": 0.2927730083465576, + "learning_rate": 4.534004576716599e-05, + "loss": 1.1248, + "step": 8743 + }, + { + "epoch": 0.5312594932863479, + "grad_norm": 0.35202234983444214, + "learning_rate": 4.5330514745158324e-05, + "loss": 1.1401, + "step": 8744 + }, + { + "epoch": 0.5313202503189745, + "grad_norm": 0.20556995272636414, + "learning_rate": 4.532098389431154e-05, + "loss": 1.1198, + "step": 8745 + }, + { + "epoch": 0.531381007351601, + "grad_norm": 0.3086276948451996, + "learning_rate": 4.531145321497498e-05, + "loss": 1.0645, + "step": 8746 + }, + { + "epoch": 0.5314417643842275, + "grad_norm": 0.2044759839773178, + "learning_rate": 4.530192270749801e-05, + "loss": 1.0891, + "step": 8747 + }, + { + "epoch": 0.531502521416854, + "grad_norm": 0.15689334273338318, + "learning_rate": 4.5292392372229996e-05, + "loss": 1.0349, + "step": 8748 + }, + { + "epoch": 0.5315632784494805, + "grad_norm": 0.8245697617530823, + "learning_rate": 4.5282862209520236e-05, + "loss": 1.0564, + "step": 8749 + }, + { + "epoch": 0.531624035482107, + "grad_norm": 0.4029900133609772, + "learning_rate": 4.5273332219718076e-05, + "loss": 1.0237, + "step": 8750 + }, + { + "epoch": 0.5316847925147336, + "grad_norm": 0.19824060797691345, + "learning_rate": 4.526380240317282e-05, + "loss": 1.0965, + "step": 8751 + }, + { + "epoch": 0.5317455495473601, + "grad_norm": 0.6450697183609009, + "learning_rate": 4.525427276023382e-05, + "loss": 1.1045, + "step": 8752 + }, + { + "epoch": 0.5318063065799866, + "grad_norm": 0.19679449498653412, + "learning_rate": 4.524474329125036e-05, + "loss": 1.0302, + "step": 8753 + }, + { + "epoch": 0.5318670636126132, + "grad_norm": 0.31215226650238037, + "learning_rate": 4.523521399657176e-05, + "loss": 1.0626, + "step": 8754 + }, + { + "epoch": 0.5319278206452397, + "grad_norm": 1.2342747449874878, + "learning_rate": 4.522568487654733e-05, + "loss": 1.2307, + "step": 8755 + }, + { + "epoch": 0.5319885776778662, + "grad_norm": 0.17231105268001556, + "learning_rate": 4.5216155931526306e-05, + "loss": 1.0354, + "step": 8756 + }, + { + "epoch": 0.5320493347104928, + "grad_norm": 0.20127499103546143, + "learning_rate": 4.520662716185803e-05, + "loss": 1.0726, + "step": 8757 + }, + { + "epoch": 0.5321100917431193, + "grad_norm": 0.6923670172691345, + "learning_rate": 4.519709856789176e-05, + "loss": 1.1466, + "step": 8758 + }, + { + "epoch": 0.5321708487757458, + "grad_norm": 0.24884703755378723, + "learning_rate": 4.5187570149976764e-05, + "loss": 1.1436, + "step": 8759 + }, + { + "epoch": 0.5322316058083724, + "grad_norm": 0.3187161982059479, + "learning_rate": 4.517804190846232e-05, + "loss": 1.0801, + "step": 8760 + }, + { + "epoch": 0.5322923628409989, + "grad_norm": 0.14839410781860352, + "learning_rate": 4.5168513843697686e-05, + "loss": 1.051, + "step": 8761 + }, + { + "epoch": 0.5323531198736253, + "grad_norm": 0.17094096541404724, + "learning_rate": 4.5158985956032127e-05, + "loss": 1.0038, + "step": 8762 + }, + { + "epoch": 0.5324138769062519, + "grad_norm": 0.21987712383270264, + "learning_rate": 4.5149458245814854e-05, + "loss": 1.1476, + "step": 8763 + }, + { + "epoch": 0.5324746339388784, + "grad_norm": 0.19184069335460663, + "learning_rate": 4.513993071339511e-05, + "loss": 1.1131, + "step": 8764 + }, + { + "epoch": 0.5325353909715049, + "grad_norm": 0.18839313089847565, + "learning_rate": 4.513040335912217e-05, + "loss": 1.0253, + "step": 8765 + }, + { + "epoch": 0.5325961480041315, + "grad_norm": 0.1808563619852066, + "learning_rate": 4.5120876183345226e-05, + "loss": 1.0414, + "step": 8766 + }, + { + "epoch": 0.532656905036758, + "grad_norm": 0.3202536702156067, + "learning_rate": 4.5111349186413504e-05, + "loss": 1.1599, + "step": 8767 + }, + { + "epoch": 0.5327176620693845, + "grad_norm": 0.2105739265680313, + "learning_rate": 4.510182236867625e-05, + "loss": 1.0966, + "step": 8768 + }, + { + "epoch": 0.5327784191020111, + "grad_norm": 0.2985372245311737, + "learning_rate": 4.509229573048261e-05, + "loss": 1.0367, + "step": 8769 + }, + { + "epoch": 0.5328391761346376, + "grad_norm": 0.18381811678409576, + "learning_rate": 4.5082769272181844e-05, + "loss": 1.0668, + "step": 8770 + }, + { + "epoch": 0.5328999331672641, + "grad_norm": 0.22565197944641113, + "learning_rate": 4.507324299412311e-05, + "loss": 1.0274, + "step": 8771 + }, + { + "epoch": 0.5329606901998907, + "grad_norm": 0.20653429627418518, + "learning_rate": 4.50637168966556e-05, + "loss": 1.0666, + "step": 8772 + }, + { + "epoch": 0.5330214472325172, + "grad_norm": 1.1182798147201538, + "learning_rate": 4.5054190980128494e-05, + "loss": 1.1064, + "step": 8773 + }, + { + "epoch": 0.5330822042651437, + "grad_norm": 0.2298218458890915, + "learning_rate": 4.504466524489099e-05, + "loss": 1.0217, + "step": 8774 + }, + { + "epoch": 0.5331429612977702, + "grad_norm": 0.24565666913986206, + "learning_rate": 4.503513969129225e-05, + "loss": 1.1378, + "step": 8775 + }, + { + "epoch": 0.5332037183303967, + "grad_norm": 0.2195633053779602, + "learning_rate": 4.502561431968141e-05, + "loss": 1.0342, + "step": 8776 + }, + { + "epoch": 0.5332644753630232, + "grad_norm": 0.22323140501976013, + "learning_rate": 4.501608913040763e-05, + "loss": 1.0384, + "step": 8777 + }, + { + "epoch": 0.5333252323956498, + "grad_norm": 0.15649636089801788, + "learning_rate": 4.5006564123820074e-05, + "loss": 1.048, + "step": 8778 + }, + { + "epoch": 0.5333859894282763, + "grad_norm": 0.1592605710029602, + "learning_rate": 4.499703930026788e-05, + "loss": 1.0397, + "step": 8779 + }, + { + "epoch": 0.5334467464609028, + "grad_norm": 0.22037093341350555, + "learning_rate": 4.498751466010017e-05, + "loss": 1.0359, + "step": 8780 + }, + { + "epoch": 0.5335075034935294, + "grad_norm": 0.23057422041893005, + "learning_rate": 4.497799020366608e-05, + "loss": 1.1141, + "step": 8781 + }, + { + "epoch": 0.5335682605261559, + "grad_norm": 0.3333434462547302, + "learning_rate": 4.496846593131474e-05, + "loss": 1.0964, + "step": 8782 + }, + { + "epoch": 0.5336290175587824, + "grad_norm": 0.7253627777099609, + "learning_rate": 4.495894184339525e-05, + "loss": 1.0309, + "step": 8783 + }, + { + "epoch": 0.533689774591409, + "grad_norm": 0.2005721479654312, + "learning_rate": 4.494941794025672e-05, + "loss": 1.0123, + "step": 8784 + }, + { + "epoch": 0.5337505316240355, + "grad_norm": 0.13499990105628967, + "learning_rate": 4.493989422224824e-05, + "loss": 1.031, + "step": 8785 + }, + { + "epoch": 0.533811288656662, + "grad_norm": 4.074336051940918, + "learning_rate": 4.493037068971892e-05, + "loss": 1.0348, + "step": 8786 + }, + { + "epoch": 0.5338720456892886, + "grad_norm": 0.24117591977119446, + "learning_rate": 4.492084734301785e-05, + "loss": 1.1298, + "step": 8787 + }, + { + "epoch": 0.533932802721915, + "grad_norm": 0.1795424520969391, + "learning_rate": 4.491132418249411e-05, + "loss": 1.0086, + "step": 8788 + }, + { + "epoch": 0.5339935597545415, + "grad_norm": 0.4908912777900696, + "learning_rate": 4.490180120849676e-05, + "loss": 1.1793, + "step": 8789 + }, + { + "epoch": 0.5340543167871681, + "grad_norm": 0.17141738533973694, + "learning_rate": 4.489227842137485e-05, + "loss": 1.1165, + "step": 8790 + }, + { + "epoch": 0.5341150738197946, + "grad_norm": 0.15687523782253265, + "learning_rate": 4.488275582147749e-05, + "loss": 1.0843, + "step": 8791 + }, + { + "epoch": 0.5341758308524212, + "grad_norm": 0.3246571123600006, + "learning_rate": 4.4873233409153695e-05, + "loss": 1.0905, + "step": 8792 + }, + { + "epoch": 0.5342365878850477, + "grad_norm": 0.3337225317955017, + "learning_rate": 4.486371118475251e-05, + "loss": 1.026, + "step": 8793 + }, + { + "epoch": 0.5342973449176742, + "grad_norm": 0.16708073019981384, + "learning_rate": 4.4854189148622994e-05, + "loss": 1.0406, + "step": 8794 + }, + { + "epoch": 0.5343581019503008, + "grad_norm": 0.34432846307754517, + "learning_rate": 4.484466730111418e-05, + "loss": 1.1127, + "step": 8795 + }, + { + "epoch": 0.5344188589829273, + "grad_norm": 0.16187772154808044, + "learning_rate": 4.4835145642575074e-05, + "loss": 1.0453, + "step": 8796 + }, + { + "epoch": 0.5344796160155538, + "grad_norm": 0.19312690198421478, + "learning_rate": 4.48256241733547e-05, + "loss": 1.0812, + "step": 8797 + }, + { + "epoch": 0.5345403730481804, + "grad_norm": 0.31319379806518555, + "learning_rate": 4.481610289380206e-05, + "loss": 1.1497, + "step": 8798 + }, + { + "epoch": 0.5346011300808069, + "grad_norm": 0.1938648372888565, + "learning_rate": 4.4806581804266186e-05, + "loss": 1.1734, + "step": 8799 + }, + { + "epoch": 0.5346618871134334, + "grad_norm": 0.22219060361385345, + "learning_rate": 4.4797060905096055e-05, + "loss": 1.131, + "step": 8800 + }, + { + "epoch": 0.5347226441460599, + "grad_norm": 0.524145245552063, + "learning_rate": 4.478754019664065e-05, + "loss": 1.1164, + "step": 8801 + }, + { + "epoch": 0.5347834011786864, + "grad_norm": 0.15736441314220428, + "learning_rate": 4.4778019679249004e-05, + "loss": 1.0997, + "step": 8802 + }, + { + "epoch": 0.5348441582113129, + "grad_norm": 0.410472571849823, + "learning_rate": 4.4768499353270026e-05, + "loss": 1.116, + "step": 8803 + }, + { + "epoch": 0.5349049152439395, + "grad_norm": 0.2566424012184143, + "learning_rate": 4.475897921905272e-05, + "loss": 1.0658, + "step": 8804 + }, + { + "epoch": 0.534965672276566, + "grad_norm": 0.1699775755405426, + "learning_rate": 4.474945927694605e-05, + "loss": 1.0691, + "step": 8805 + }, + { + "epoch": 0.5350264293091925, + "grad_norm": 0.2633498013019562, + "learning_rate": 4.473993952729895e-05, + "loss": 1.0602, + "step": 8806 + }, + { + "epoch": 0.5350871863418191, + "grad_norm": 0.2832537889480591, + "learning_rate": 4.47304199704604e-05, + "loss": 1.029, + "step": 8807 + }, + { + "epoch": 0.5351479433744456, + "grad_norm": 0.28935495018959045, + "learning_rate": 4.472090060677932e-05, + "loss": 1.0639, + "step": 8808 + }, + { + "epoch": 0.5352087004070721, + "grad_norm": 0.16809530556201935, + "learning_rate": 4.471138143660465e-05, + "loss": 1.0595, + "step": 8809 + }, + { + "epoch": 0.5352694574396987, + "grad_norm": 0.2816694378852844, + "learning_rate": 4.470186246028532e-05, + "loss": 1.0022, + "step": 8810 + }, + { + "epoch": 0.5353302144723252, + "grad_norm": 0.24123434722423553, + "learning_rate": 4.469234367817023e-05, + "loss": 1.1071, + "step": 8811 + }, + { + "epoch": 0.5353909715049517, + "grad_norm": 0.23041997849941254, + "learning_rate": 4.4682825090608314e-05, + "loss": 1.1522, + "step": 8812 + }, + { + "epoch": 0.5354517285375783, + "grad_norm": 0.29295098781585693, + "learning_rate": 4.467330669794848e-05, + "loss": 1.034, + "step": 8813 + }, + { + "epoch": 0.5355124855702047, + "grad_norm": 0.2621358036994934, + "learning_rate": 4.4663788500539603e-05, + "loss": 1.102, + "step": 8814 + }, + { + "epoch": 0.5355732426028312, + "grad_norm": 0.21559928357601166, + "learning_rate": 4.465427049873062e-05, + "loss": 1.0333, + "step": 8815 + }, + { + "epoch": 0.5356339996354578, + "grad_norm": 0.21570293605327606, + "learning_rate": 4.464475269287036e-05, + "loss": 1.0633, + "step": 8816 + }, + { + "epoch": 0.5356947566680843, + "grad_norm": 0.20846299827098846, + "learning_rate": 4.463523508330773e-05, + "loss": 1.0649, + "step": 8817 + }, + { + "epoch": 0.5357555137007108, + "grad_norm": 0.28448206186294556, + "learning_rate": 4.46257176703916e-05, + "loss": 1.0711, + "step": 8818 + }, + { + "epoch": 0.5358162707333374, + "grad_norm": 0.2049059420824051, + "learning_rate": 4.461620045447081e-05, + "loss": 1.1068, + "step": 8819 + }, + { + "epoch": 0.5358770277659639, + "grad_norm": 0.3479083180427551, + "learning_rate": 4.4606683435894255e-05, + "loss": 1.1367, + "step": 8820 + }, + { + "epoch": 0.5359377847985904, + "grad_norm": 0.36638060212135315, + "learning_rate": 4.459716661501075e-05, + "loss": 1.061, + "step": 8821 + }, + { + "epoch": 0.535998541831217, + "grad_norm": 0.25744181871414185, + "learning_rate": 4.458764999216916e-05, + "loss": 1.1474, + "step": 8822 + }, + { + "epoch": 0.5360592988638435, + "grad_norm": 0.40011727809906006, + "learning_rate": 4.45781335677183e-05, + "loss": 1.1719, + "step": 8823 + }, + { + "epoch": 0.53612005589647, + "grad_norm": 0.3097867965698242, + "learning_rate": 4.4568617342006986e-05, + "loss": 1.1098, + "step": 8824 + }, + { + "epoch": 0.5361808129290966, + "grad_norm": 0.18512561917304993, + "learning_rate": 4.4559101315384066e-05, + "loss": 1.0888, + "step": 8825 + }, + { + "epoch": 0.5362415699617231, + "grad_norm": 1.5040671825408936, + "learning_rate": 4.454958548819834e-05, + "loss": 1.1379, + "step": 8826 + }, + { + "epoch": 0.5363023269943495, + "grad_norm": 0.4778504967689514, + "learning_rate": 4.45400698607986e-05, + "loss": 1.3045, + "step": 8827 + }, + { + "epoch": 0.5363630840269761, + "grad_norm": 0.20005851984024048, + "learning_rate": 4.4530554433533666e-05, + "loss": 1.1166, + "step": 8828 + }, + { + "epoch": 0.5364238410596026, + "grad_norm": 0.35838696360588074, + "learning_rate": 4.452103920675232e-05, + "loss": 1.1669, + "step": 8829 + }, + { + "epoch": 0.5364845980922291, + "grad_norm": 0.29099833965301514, + "learning_rate": 4.4511524180803345e-05, + "loss": 1.1901, + "step": 8830 + }, + { + "epoch": 0.5365453551248557, + "grad_norm": 0.18893519043922424, + "learning_rate": 4.45020093560355e-05, + "loss": 1.1108, + "step": 8831 + }, + { + "epoch": 0.5366061121574822, + "grad_norm": 0.23872722685337067, + "learning_rate": 4.4492494732797556e-05, + "loss": 1.0239, + "step": 8832 + }, + { + "epoch": 0.5366668691901088, + "grad_norm": 0.2653682231903076, + "learning_rate": 4.44829803114383e-05, + "loss": 1.1755, + "step": 8833 + }, + { + "epoch": 0.5367276262227353, + "grad_norm": 0.226792111992836, + "learning_rate": 4.447346609230646e-05, + "loss": 1.0812, + "step": 8834 + }, + { + "epoch": 0.5367883832553618, + "grad_norm": 0.2463216334581375, + "learning_rate": 4.44639520757508e-05, + "loss": 1.0662, + "step": 8835 + }, + { + "epoch": 0.5368491402879884, + "grad_norm": 0.28490152955055237, + "learning_rate": 4.445443826212003e-05, + "loss": 1.1865, + "step": 8836 + }, + { + "epoch": 0.5369098973206149, + "grad_norm": 0.34240463376045227, + "learning_rate": 4.44449246517629e-05, + "loss": 1.1926, + "step": 8837 + }, + { + "epoch": 0.5369706543532414, + "grad_norm": 0.15790100395679474, + "learning_rate": 4.4435411245028134e-05, + "loss": 1.0582, + "step": 8838 + }, + { + "epoch": 0.537031411385868, + "grad_norm": 1.1619457006454468, + "learning_rate": 4.442589804226445e-05, + "loss": 1.0583, + "step": 8839 + }, + { + "epoch": 0.5370921684184944, + "grad_norm": 0.18283167481422424, + "learning_rate": 4.4416385043820534e-05, + "loss": 1.0415, + "step": 8840 + }, + { + "epoch": 0.5371529254511209, + "grad_norm": 0.1915820688009262, + "learning_rate": 4.4406872250045115e-05, + "loss": 1.0834, + "step": 8841 + }, + { + "epoch": 0.5372136824837475, + "grad_norm": 0.203061044216156, + "learning_rate": 4.4397359661286886e-05, + "loss": 1.1244, + "step": 8842 + }, + { + "epoch": 0.537274439516374, + "grad_norm": 0.23137392103672028, + "learning_rate": 4.438784727789451e-05, + "loss": 1.1124, + "step": 8843 + }, + { + "epoch": 0.5373351965490005, + "grad_norm": 0.25040385127067566, + "learning_rate": 4.437833510021667e-05, + "loss": 1.0968, + "step": 8844 + }, + { + "epoch": 0.5373959535816271, + "grad_norm": 0.16145800054073334, + "learning_rate": 4.436882312860205e-05, + "loss": 1.0865, + "step": 8845 + }, + { + "epoch": 0.5374567106142536, + "grad_norm": 0.23443464934825897, + "learning_rate": 4.435931136339931e-05, + "loss": 1.1157, + "step": 8846 + }, + { + "epoch": 0.5375174676468801, + "grad_norm": 0.1523744761943817, + "learning_rate": 4.43497998049571e-05, + "loss": 1.0633, + "step": 8847 + }, + { + "epoch": 0.5375782246795067, + "grad_norm": 0.21774600446224213, + "learning_rate": 4.4340288453624057e-05, + "loss": 1.1414, + "step": 8848 + }, + { + "epoch": 0.5376389817121332, + "grad_norm": 4.096322536468506, + "learning_rate": 4.433077730974886e-05, + "loss": 1.1068, + "step": 8849 + }, + { + "epoch": 0.5376997387447597, + "grad_norm": 0.2267347127199173, + "learning_rate": 4.432126637368011e-05, + "loss": 1.0561, + "step": 8850 + }, + { + "epoch": 0.5377604957773863, + "grad_norm": 0.1721230000257492, + "learning_rate": 4.4311755645766435e-05, + "loss": 1.0411, + "step": 8851 + }, + { + "epoch": 0.5378212528100128, + "grad_norm": 0.1902976781129837, + "learning_rate": 4.430224512635644e-05, + "loss": 1.0677, + "step": 8852 + }, + { + "epoch": 0.5378820098426392, + "grad_norm": 0.25539112091064453, + "learning_rate": 4.4292734815798767e-05, + "loss": 1.1289, + "step": 8853 + }, + { + "epoch": 0.5379427668752658, + "grad_norm": 0.1767245978116989, + "learning_rate": 4.4283224714442e-05, + "loss": 1.0714, + "step": 8854 + }, + { + "epoch": 0.5380035239078923, + "grad_norm": 0.22714431583881378, + "learning_rate": 4.4273714822634754e-05, + "loss": 1.1573, + "step": 8855 + }, + { + "epoch": 0.5380642809405188, + "grad_norm": 0.24854111671447754, + "learning_rate": 4.426420514072558e-05, + "loss": 1.0807, + "step": 8856 + }, + { + "epoch": 0.5381250379731454, + "grad_norm": 0.26030534505844116, + "learning_rate": 4.4254695669063066e-05, + "loss": 1.1098, + "step": 8857 + }, + { + "epoch": 0.5381857950057719, + "grad_norm": 0.210786372423172, + "learning_rate": 4.4245186407995804e-05, + "loss": 1.1537, + "step": 8858 + }, + { + "epoch": 0.5382465520383984, + "grad_norm": 0.2915708124637604, + "learning_rate": 4.4235677357872346e-05, + "loss": 1.1065, + "step": 8859 + }, + { + "epoch": 0.538307309071025, + "grad_norm": 0.14643916487693787, + "learning_rate": 4.422616851904124e-05, + "loss": 1.0373, + "step": 8860 + }, + { + "epoch": 0.5383680661036515, + "grad_norm": 0.2378307282924652, + "learning_rate": 4.4216659891851054e-05, + "loss": 1.0801, + "step": 8861 + }, + { + "epoch": 0.538428823136278, + "grad_norm": 0.2153945118188858, + "learning_rate": 4.420715147665033e-05, + "loss": 1.0393, + "step": 8862 + }, + { + "epoch": 0.5384895801689046, + "grad_norm": 0.2054414600133896, + "learning_rate": 4.419764327378757e-05, + "loss": 1.1661, + "step": 8863 + }, + { + "epoch": 0.5385503372015311, + "grad_norm": 0.16983453929424286, + "learning_rate": 4.4188135283611324e-05, + "loss": 1.0472, + "step": 8864 + }, + { + "epoch": 0.5386110942341577, + "grad_norm": 0.1611793041229248, + "learning_rate": 4.417862750647009e-05, + "loss": 1.0509, + "step": 8865 + }, + { + "epoch": 0.5386718512667842, + "grad_norm": 0.32169076800346375, + "learning_rate": 4.41691199427124e-05, + "loss": 1.0394, + "step": 8866 + }, + { + "epoch": 0.5387326082994106, + "grad_norm": 0.28186050057411194, + "learning_rate": 4.4159612592686744e-05, + "loss": 1.1758, + "step": 8867 + }, + { + "epoch": 0.5387933653320371, + "grad_norm": 0.1641460508108139, + "learning_rate": 4.415010545674161e-05, + "loss": 1.0238, + "step": 8868 + }, + { + "epoch": 0.5388541223646637, + "grad_norm": 0.21418187022209167, + "learning_rate": 4.414059853522552e-05, + "loss": 1.1519, + "step": 8869 + }, + { + "epoch": 0.5389148793972902, + "grad_norm": 0.29164063930511475, + "learning_rate": 4.4131091828486894e-05, + "loss": 1.0809, + "step": 8870 + }, + { + "epoch": 0.5389756364299167, + "grad_norm": 1.893638014793396, + "learning_rate": 4.412158533687424e-05, + "loss": 1.2398, + "step": 8871 + }, + { + "epoch": 0.5390363934625433, + "grad_norm": 0.24341514706611633, + "learning_rate": 4.411207906073601e-05, + "loss": 1.1126, + "step": 8872 + }, + { + "epoch": 0.5390971504951698, + "grad_norm": 0.2361065000295639, + "learning_rate": 4.4102573000420657e-05, + "loss": 1.0796, + "step": 8873 + }, + { + "epoch": 0.5391579075277964, + "grad_norm": 0.21174132823944092, + "learning_rate": 4.409306715627663e-05, + "loss": 1.0345, + "step": 8874 + }, + { + "epoch": 0.5392186645604229, + "grad_norm": 0.17392396926879883, + "learning_rate": 4.4083561528652384e-05, + "loss": 1.0732, + "step": 8875 + }, + { + "epoch": 0.5392794215930494, + "grad_norm": 0.22369810938835144, + "learning_rate": 4.407405611789633e-05, + "loss": 1.0752, + "step": 8876 + }, + { + "epoch": 0.539340178625676, + "grad_norm": 0.2319834679365158, + "learning_rate": 4.406455092435689e-05, + "loss": 1.075, + "step": 8877 + }, + { + "epoch": 0.5394009356583025, + "grad_norm": 0.20145393908023834, + "learning_rate": 4.405504594838247e-05, + "loss": 1.1029, + "step": 8878 + }, + { + "epoch": 0.539461692690929, + "grad_norm": 0.26017045974731445, + "learning_rate": 4.40455411903215e-05, + "loss": 1.1013, + "step": 8879 + }, + { + "epoch": 0.5395224497235555, + "grad_norm": 0.14146822690963745, + "learning_rate": 4.4036036650522376e-05, + "loss": 1.0661, + "step": 8880 + }, + { + "epoch": 0.539583206756182, + "grad_norm": 0.16516050696372986, + "learning_rate": 4.402653232933346e-05, + "loss": 1.1134, + "step": 8881 + }, + { + "epoch": 0.5396439637888085, + "grad_norm": 2.6282753944396973, + "learning_rate": 4.40170282271032e-05, + "loss": 1.1107, + "step": 8882 + }, + { + "epoch": 0.5397047208214351, + "grad_norm": 0.17772163450717926, + "learning_rate": 4.400752434417989e-05, + "loss": 1.0468, + "step": 8883 + }, + { + "epoch": 0.5397654778540616, + "grad_norm": 0.16070041060447693, + "learning_rate": 4.399802068091195e-05, + "loss": 1.0466, + "step": 8884 + }, + { + "epoch": 0.5398262348866881, + "grad_norm": 0.37689968943595886, + "learning_rate": 4.398851723764771e-05, + "loss": 1.3833, + "step": 8885 + }, + { + "epoch": 0.5398869919193147, + "grad_norm": 0.17482414841651917, + "learning_rate": 4.397901401473552e-05, + "loss": 1.04, + "step": 8886 + }, + { + "epoch": 0.5399477489519412, + "grad_norm": 0.1751963049173355, + "learning_rate": 4.396951101252376e-05, + "loss": 1.0447, + "step": 8887 + }, + { + "epoch": 0.5400085059845677, + "grad_norm": 0.24033649265766144, + "learning_rate": 4.396000823136073e-05, + "loss": 1.1257, + "step": 8888 + }, + { + "epoch": 0.5400692630171943, + "grad_norm": 0.1542900651693344, + "learning_rate": 4.395050567159478e-05, + "loss": 1.0309, + "step": 8889 + }, + { + "epoch": 0.5401300200498208, + "grad_norm": 0.2711865305900574, + "learning_rate": 4.39410033335742e-05, + "loss": 1.073, + "step": 8890 + }, + { + "epoch": 0.5401907770824473, + "grad_norm": 1.1377654075622559, + "learning_rate": 4.3931501217647305e-05, + "loss": 1.1022, + "step": 8891 + }, + { + "epoch": 0.5402515341150739, + "grad_norm": 0.21627171337604523, + "learning_rate": 4.392199932416242e-05, + "loss": 1.0551, + "step": 8892 + }, + { + "epoch": 0.5403122911477003, + "grad_norm": 0.2038409411907196, + "learning_rate": 4.3912497653467815e-05, + "loss": 1.0661, + "step": 8893 + }, + { + "epoch": 0.5403730481803268, + "grad_norm": 0.2486761510372162, + "learning_rate": 4.3902996205911784e-05, + "loss": 1.0997, + "step": 8894 + }, + { + "epoch": 0.5404338052129534, + "grad_norm": 0.17991070449352264, + "learning_rate": 4.389349498184261e-05, + "loss": 1.1051, + "step": 8895 + }, + { + "epoch": 0.5404945622455799, + "grad_norm": 0.24209366738796234, + "learning_rate": 4.3883993981608576e-05, + "loss": 1.0426, + "step": 8896 + }, + { + "epoch": 0.5405553192782064, + "grad_norm": 0.3186430335044861, + "learning_rate": 4.387449320555791e-05, + "loss": 1.0707, + "step": 8897 + }, + { + "epoch": 0.540616076310833, + "grad_norm": 0.20038948953151703, + "learning_rate": 4.3864992654038875e-05, + "loss": 1.0347, + "step": 8898 + }, + { + "epoch": 0.5406768333434595, + "grad_norm": 1.3247672319412231, + "learning_rate": 4.3855492327399713e-05, + "loss": 1.1095, + "step": 8899 + }, + { + "epoch": 0.540737590376086, + "grad_norm": 0.27208730578422546, + "learning_rate": 4.384599222598868e-05, + "loss": 1.1142, + "step": 8900 + }, + { + "epoch": 0.5407983474087126, + "grad_norm": 0.2105126827955246, + "learning_rate": 4.383649235015399e-05, + "loss": 1.1003, + "step": 8901 + }, + { + "epoch": 0.5408591044413391, + "grad_norm": 0.6108821034431458, + "learning_rate": 4.3826992700243876e-05, + "loss": 1.1047, + "step": 8902 + }, + { + "epoch": 0.5409198614739656, + "grad_norm": 0.18587887287139893, + "learning_rate": 4.381749327660652e-05, + "loss": 1.041, + "step": 8903 + }, + { + "epoch": 0.5409806185065922, + "grad_norm": 0.20236864686012268, + "learning_rate": 4.3807994079590145e-05, + "loss": 1.0108, + "step": 8904 + }, + { + "epoch": 0.5410413755392187, + "grad_norm": 0.16722428798675537, + "learning_rate": 4.3798495109542946e-05, + "loss": 1.0372, + "step": 8905 + }, + { + "epoch": 0.5411021325718451, + "grad_norm": 0.2907964289188385, + "learning_rate": 4.378899636681311e-05, + "loss": 1.1795, + "step": 8906 + }, + { + "epoch": 0.5411628896044717, + "grad_norm": 0.154231458902359, + "learning_rate": 4.37794978517488e-05, + "loss": 1.1313, + "step": 8907 + }, + { + "epoch": 0.5412236466370982, + "grad_norm": 0.187501922249794, + "learning_rate": 4.376999956469822e-05, + "loss": 1.077, + "step": 8908 + }, + { + "epoch": 0.5412844036697247, + "grad_norm": 0.6247522234916687, + "learning_rate": 4.376050150600951e-05, + "loss": 1.0745, + "step": 8909 + }, + { + "epoch": 0.5413451607023513, + "grad_norm": 0.20852908492088318, + "learning_rate": 4.375100367603081e-05, + "loss": 1.1112, + "step": 8910 + }, + { + "epoch": 0.5414059177349778, + "grad_norm": 0.19609388709068298, + "learning_rate": 4.3741506075110285e-05, + "loss": 1.1564, + "step": 8911 + }, + { + "epoch": 0.5414666747676044, + "grad_norm": 0.1402510106563568, + "learning_rate": 4.373200870359605e-05, + "loss": 1.0503, + "step": 8912 + }, + { + "epoch": 0.5415274318002309, + "grad_norm": 0.2151438295841217, + "learning_rate": 4.372251156183626e-05, + "loss": 1.1271, + "step": 8913 + }, + { + "epoch": 0.5415881888328574, + "grad_norm": 0.519421398639679, + "learning_rate": 4.3713014650179026e-05, + "loss": 1.0579, + "step": 8914 + }, + { + "epoch": 0.541648945865484, + "grad_norm": 0.17627203464508057, + "learning_rate": 4.370351796897244e-05, + "loss": 1.0705, + "step": 8915 + }, + { + "epoch": 0.5417097028981105, + "grad_norm": 0.41577693819999695, + "learning_rate": 4.369402151856465e-05, + "loss": 1.0792, + "step": 8916 + }, + { + "epoch": 0.541770459930737, + "grad_norm": 2.274890661239624, + "learning_rate": 4.368452529930369e-05, + "loss": 1.2179, + "step": 8917 + }, + { + "epoch": 0.5418312169633636, + "grad_norm": 0.2140251249074936, + "learning_rate": 4.367502931153769e-05, + "loss": 1.0894, + "step": 8918 + }, + { + "epoch": 0.54189197399599, + "grad_norm": 0.21850328147411346, + "learning_rate": 4.366553355561471e-05, + "loss": 1.0797, + "step": 8919 + }, + { + "epoch": 0.5419527310286165, + "grad_norm": 0.23538626730442047, + "learning_rate": 4.3656038031882816e-05, + "loss": 1.1845, + "step": 8920 + }, + { + "epoch": 0.542013488061243, + "grad_norm": 0.15424662828445435, + "learning_rate": 4.364654274069008e-05, + "loss": 1.0535, + "step": 8921 + }, + { + "epoch": 0.5420742450938696, + "grad_norm": 0.22279484570026398, + "learning_rate": 4.363704768238456e-05, + "loss": 1.1117, + "step": 8922 + }, + { + "epoch": 0.5421350021264961, + "grad_norm": 0.15371371805667877, + "learning_rate": 4.362755285731427e-05, + "loss": 0.9905, + "step": 8923 + }, + { + "epoch": 0.5421957591591227, + "grad_norm": 0.242278054356575, + "learning_rate": 4.3618058265827264e-05, + "loss": 1.1103, + "step": 8924 + }, + { + "epoch": 0.5422565161917492, + "grad_norm": 0.19425413012504578, + "learning_rate": 4.360856390827156e-05, + "loss": 1.1123, + "step": 8925 + }, + { + "epoch": 0.5423172732243757, + "grad_norm": 0.1810752898454666, + "learning_rate": 4.359906978499519e-05, + "loss": 1.0152, + "step": 8926 + }, + { + "epoch": 0.5423780302570023, + "grad_norm": 0.20560535788536072, + "learning_rate": 4.358957589634616e-05, + "loss": 1.0752, + "step": 8927 + }, + { + "epoch": 0.5424387872896288, + "grad_norm": 0.8879467248916626, + "learning_rate": 4.358008224267245e-05, + "loss": 1.2007, + "step": 8928 + }, + { + "epoch": 0.5424995443222553, + "grad_norm": 0.2015133947134018, + "learning_rate": 4.3570588824322096e-05, + "loss": 1.0494, + "step": 8929 + }, + { + "epoch": 0.5425603013548819, + "grad_norm": 0.1606956571340561, + "learning_rate": 4.356109564164302e-05, + "loss": 1.0867, + "step": 8930 + }, + { + "epoch": 0.5426210583875084, + "grad_norm": 0.15887141227722168, + "learning_rate": 4.355160269498324e-05, + "loss": 1.0422, + "step": 8931 + }, + { + "epoch": 0.5426818154201348, + "grad_norm": 0.8587523698806763, + "learning_rate": 4.354210998469071e-05, + "loss": 1.1711, + "step": 8932 + }, + { + "epoch": 0.5427425724527614, + "grad_norm": 0.18437370657920837, + "learning_rate": 4.353261751111337e-05, + "loss": 1.0402, + "step": 8933 + }, + { + "epoch": 0.5428033294853879, + "grad_norm": 0.1429387331008911, + "learning_rate": 4.352312527459921e-05, + "loss": 1.0218, + "step": 8934 + }, + { + "epoch": 0.5428640865180144, + "grad_norm": 0.4876413643360138, + "learning_rate": 4.351363327549613e-05, + "loss": 1.3728, + "step": 8935 + }, + { + "epoch": 0.542924843550641, + "grad_norm": 0.2826386094093323, + "learning_rate": 4.35041415141521e-05, + "loss": 1.2288, + "step": 8936 + }, + { + "epoch": 0.5429856005832675, + "grad_norm": 0.28653237223625183, + "learning_rate": 4.349464999091498e-05, + "loss": 1.1639, + "step": 8937 + }, + { + "epoch": 0.543046357615894, + "grad_norm": 0.16596131026744843, + "learning_rate": 4.3485158706132744e-05, + "loss": 1.0659, + "step": 8938 + }, + { + "epoch": 0.5431071146485206, + "grad_norm": 0.23432305455207825, + "learning_rate": 4.347566766015327e-05, + "loss": 1.1101, + "step": 8939 + }, + { + "epoch": 0.5431678716811471, + "grad_norm": 0.1488552987575531, + "learning_rate": 4.346617685332447e-05, + "loss": 1.0448, + "step": 8940 + }, + { + "epoch": 0.5432286287137736, + "grad_norm": 0.2940818965435028, + "learning_rate": 4.3456686285994205e-05, + "loss": 0.9933, + "step": 8941 + }, + { + "epoch": 0.5432893857464002, + "grad_norm": 0.3219488561153412, + "learning_rate": 4.3447195958510384e-05, + "loss": 0.998, + "step": 8942 + }, + { + "epoch": 0.5433501427790267, + "grad_norm": 0.22157925367355347, + "learning_rate": 4.343770587122086e-05, + "loss": 1.194, + "step": 8943 + }, + { + "epoch": 0.5434108998116532, + "grad_norm": 0.17392872273921967, + "learning_rate": 4.342821602447348e-05, + "loss": 1.016, + "step": 8944 + }, + { + "epoch": 0.5434716568442797, + "grad_norm": 0.2347022145986557, + "learning_rate": 4.341872641861612e-05, + "loss": 1.1059, + "step": 8945 + }, + { + "epoch": 0.5435324138769062, + "grad_norm": 0.4344325363636017, + "learning_rate": 4.3409237053996615e-05, + "loss": 1.0744, + "step": 8946 + }, + { + "epoch": 0.5435931709095327, + "grad_norm": 0.1492263674736023, + "learning_rate": 4.3399747930962805e-05, + "loss": 1.0134, + "step": 8947 + }, + { + "epoch": 0.5436539279421593, + "grad_norm": 0.16743223369121552, + "learning_rate": 4.3390259049862503e-05, + "loss": 1.0641, + "step": 8948 + }, + { + "epoch": 0.5437146849747858, + "grad_norm": 0.3208848237991333, + "learning_rate": 4.338077041104355e-05, + "loss": 1.0777, + "step": 8949 + }, + { + "epoch": 0.5437754420074123, + "grad_norm": 0.19929815828800201, + "learning_rate": 4.337128201485371e-05, + "loss": 1.1567, + "step": 8950 + }, + { + "epoch": 0.5438361990400389, + "grad_norm": 0.26201045513153076, + "learning_rate": 4.336179386164083e-05, + "loss": 1.2322, + "step": 8951 + }, + { + "epoch": 0.5438969560726654, + "grad_norm": 0.28423893451690674, + "learning_rate": 4.335230595175267e-05, + "loss": 1.1492, + "step": 8952 + }, + { + "epoch": 0.543957713105292, + "grad_norm": 0.33184707164764404, + "learning_rate": 4.334281828553701e-05, + "loss": 1.1445, + "step": 8953 + }, + { + "epoch": 0.5440184701379185, + "grad_norm": 0.21927012503147125, + "learning_rate": 4.333333086334165e-05, + "loss": 1.1078, + "step": 8954 + }, + { + "epoch": 0.544079227170545, + "grad_norm": 0.6765453815460205, + "learning_rate": 4.332384368551433e-05, + "loss": 1.0798, + "step": 8955 + }, + { + "epoch": 0.5441399842031716, + "grad_norm": 0.3112942576408386, + "learning_rate": 4.3314356752402826e-05, + "loss": 1.1382, + "step": 8956 + }, + { + "epoch": 0.5442007412357981, + "grad_norm": 0.3520110249519348, + "learning_rate": 4.330487006435485e-05, + "loss": 1.0355, + "step": 8957 + }, + { + "epoch": 0.5442614982684245, + "grad_norm": 0.317983478307724, + "learning_rate": 4.329538362171815e-05, + "loss": 1.0519, + "step": 8958 + }, + { + "epoch": 0.544322255301051, + "grad_norm": 0.34609484672546387, + "learning_rate": 4.328589742484047e-05, + "loss": 1.2751, + "step": 8959 + }, + { + "epoch": 0.5443830123336776, + "grad_norm": 0.2747074067592621, + "learning_rate": 4.3276411474069515e-05, + "loss": 1.1372, + "step": 8960 + }, + { + "epoch": 0.5444437693663041, + "grad_norm": 0.25901108980178833, + "learning_rate": 4.326692576975299e-05, + "loss": 1.0382, + "step": 8961 + }, + { + "epoch": 0.5445045263989307, + "grad_norm": 6.6069746017456055, + "learning_rate": 4.325744031223861e-05, + "loss": 1.1756, + "step": 8962 + }, + { + "epoch": 0.5445652834315572, + "grad_norm": 0.3036087453365326, + "learning_rate": 4.324795510187407e-05, + "loss": 1.1476, + "step": 8963 + }, + { + "epoch": 0.5446260404641837, + "grad_norm": 0.16914160549640656, + "learning_rate": 4.323847013900703e-05, + "loss": 1.0447, + "step": 8964 + }, + { + "epoch": 0.5446867974968103, + "grad_norm": 0.20247714221477509, + "learning_rate": 4.3228985423985176e-05, + "loss": 1.0606, + "step": 8965 + }, + { + "epoch": 0.5447475545294368, + "grad_norm": 0.21544943749904633, + "learning_rate": 4.3219500957156157e-05, + "loss": 1.0985, + "step": 8966 + }, + { + "epoch": 0.5448083115620633, + "grad_norm": 0.21818697452545166, + "learning_rate": 4.321001673886765e-05, + "loss": 1.1421, + "step": 8967 + }, + { + "epoch": 0.5448690685946899, + "grad_norm": 0.17124179005622864, + "learning_rate": 4.32005327694673e-05, + "loss": 0.9977, + "step": 8968 + }, + { + "epoch": 0.5449298256273164, + "grad_norm": 0.25139614939689636, + "learning_rate": 4.319104904930274e-05, + "loss": 1.1718, + "step": 8969 + }, + { + "epoch": 0.5449905826599429, + "grad_norm": 0.2423376590013504, + "learning_rate": 4.3181565578721584e-05, + "loss": 1.1447, + "step": 8970 + }, + { + "epoch": 0.5450513396925695, + "grad_norm": 1.4261865615844727, + "learning_rate": 4.3172082358071446e-05, + "loss": 1.2622, + "step": 8971 + }, + { + "epoch": 0.5451120967251959, + "grad_norm": 0.22845768928527832, + "learning_rate": 4.3162599387699966e-05, + "loss": 1.1317, + "step": 8972 + }, + { + "epoch": 0.5451728537578224, + "grad_norm": 0.16972555220127106, + "learning_rate": 4.315311666795472e-05, + "loss": 1.0174, + "step": 8973 + }, + { + "epoch": 0.545233610790449, + "grad_norm": 0.20607098937034607, + "learning_rate": 4.31436341991833e-05, + "loss": 1.0501, + "step": 8974 + }, + { + "epoch": 0.5452943678230755, + "grad_norm": 0.19274453818798065, + "learning_rate": 4.31341519817333e-05, + "loss": 1.0576, + "step": 8975 + }, + { + "epoch": 0.545355124855702, + "grad_norm": 0.1549033373594284, + "learning_rate": 4.31246700159523e-05, + "loss": 1.0316, + "step": 8976 + }, + { + "epoch": 0.5454158818883286, + "grad_norm": 0.24633774161338806, + "learning_rate": 4.3115188302187844e-05, + "loss": 1.0469, + "step": 8977 + }, + { + "epoch": 0.5454766389209551, + "grad_norm": 0.30001553893089294, + "learning_rate": 4.310570684078749e-05, + "loss": 1.1914, + "step": 8978 + }, + { + "epoch": 0.5455373959535816, + "grad_norm": 0.18207748234272003, + "learning_rate": 4.309622563209876e-05, + "loss": 1.0831, + "step": 8979 + }, + { + "epoch": 0.5455981529862082, + "grad_norm": 0.2777933180332184, + "learning_rate": 4.308674467646924e-05, + "loss": 1.0864, + "step": 8980 + }, + { + "epoch": 0.5456589100188347, + "grad_norm": 0.2313176840543747, + "learning_rate": 4.307726397424642e-05, + "loss": 1.0389, + "step": 8981 + }, + { + "epoch": 0.5457196670514612, + "grad_norm": 3.7525827884674072, + "learning_rate": 4.306778352577783e-05, + "loss": 1.0571, + "step": 8982 + }, + { + "epoch": 0.5457804240840878, + "grad_norm": 2.3151063919067383, + "learning_rate": 4.3058303331410994e-05, + "loss": 1.0816, + "step": 8983 + }, + { + "epoch": 0.5458411811167143, + "grad_norm": 0.23251238465309143, + "learning_rate": 4.3048823391493366e-05, + "loss": 1.0648, + "step": 8984 + }, + { + "epoch": 0.5459019381493407, + "grad_norm": 0.2674936056137085, + "learning_rate": 4.3039343706372465e-05, + "loss": 1.0787, + "step": 8985 + }, + { + "epoch": 0.5459626951819673, + "grad_norm": 0.22900545597076416, + "learning_rate": 4.302986427639577e-05, + "loss": 1.0995, + "step": 8986 + }, + { + "epoch": 0.5460234522145938, + "grad_norm": 0.1827414184808731, + "learning_rate": 4.302038510191073e-05, + "loss": 1.0983, + "step": 8987 + }, + { + "epoch": 0.5460842092472203, + "grad_norm": 0.17570970952510834, + "learning_rate": 4.301090618326484e-05, + "loss": 0.9943, + "step": 8988 + }, + { + "epoch": 0.5461449662798469, + "grad_norm": 0.24477963149547577, + "learning_rate": 4.300142752080555e-05, + "loss": 1.1172, + "step": 8989 + }, + { + "epoch": 0.5462057233124734, + "grad_norm": 0.20596130192279816, + "learning_rate": 4.299194911488027e-05, + "loss": 1.0818, + "step": 8990 + }, + { + "epoch": 0.5462664803451, + "grad_norm": 0.2426786869764328, + "learning_rate": 4.2982470965836455e-05, + "loss": 1.0147, + "step": 8991 + }, + { + "epoch": 0.5463272373777265, + "grad_norm": 0.4715384244918823, + "learning_rate": 4.297299307402151e-05, + "loss": 1.156, + "step": 8992 + }, + { + "epoch": 0.546387994410353, + "grad_norm": 1.037922739982605, + "learning_rate": 4.2963515439782864e-05, + "loss": 1.0595, + "step": 8993 + }, + { + "epoch": 0.5464487514429796, + "grad_norm": 0.2757258713245392, + "learning_rate": 4.295403806346793e-05, + "loss": 1.1039, + "step": 8994 + }, + { + "epoch": 0.5465095084756061, + "grad_norm": 0.24645380675792694, + "learning_rate": 4.294456094542408e-05, + "loss": 1.062, + "step": 8995 + }, + { + "epoch": 0.5465702655082326, + "grad_norm": 0.2649162709712982, + "learning_rate": 4.2935084085998745e-05, + "loss": 1.1658, + "step": 8996 + }, + { + "epoch": 0.5466310225408592, + "grad_norm": 0.22574616968631744, + "learning_rate": 4.2925607485539235e-05, + "loss": 1.0502, + "step": 8997 + }, + { + "epoch": 0.5466917795734856, + "grad_norm": 6.891936779022217, + "learning_rate": 4.291613114439296e-05, + "loss": 1.1343, + "step": 8998 + }, + { + "epoch": 0.5467525366061121, + "grad_norm": 0.38878217339515686, + "learning_rate": 4.290665506290727e-05, + "loss": 1.1139, + "step": 8999 + }, + { + "epoch": 0.5468132936387387, + "grad_norm": 0.24717079102993011, + "learning_rate": 4.28971792414295e-05, + "loss": 1.0018, + "step": 9000 + }, + { + "epoch": 0.5468740506713652, + "grad_norm": 0.28075629472732544, + "learning_rate": 4.2887703680307e-05, + "loss": 1.0864, + "step": 9001 + }, + { + "epoch": 0.5469348077039917, + "grad_norm": 0.33770015835762024, + "learning_rate": 4.28782283798871e-05, + "loss": 1.1541, + "step": 9002 + }, + { + "epoch": 0.5469955647366183, + "grad_norm": 0.25801584124565125, + "learning_rate": 4.2868753340517135e-05, + "loss": 1.1056, + "step": 9003 + }, + { + "epoch": 0.5470563217692448, + "grad_norm": 0.1953677535057068, + "learning_rate": 4.2859278562544373e-05, + "loss": 1.0498, + "step": 9004 + }, + { + "epoch": 0.5471170788018713, + "grad_norm": 0.2809269428253174, + "learning_rate": 4.284980404631613e-05, + "loss": 1.0482, + "step": 9005 + }, + { + "epoch": 0.5471778358344979, + "grad_norm": 0.24413439631462097, + "learning_rate": 4.2840329792179713e-05, + "loss": 1.1493, + "step": 9006 + }, + { + "epoch": 0.5472385928671244, + "grad_norm": 0.220168337225914, + "learning_rate": 4.283085580048239e-05, + "loss": 1.2008, + "step": 9007 + }, + { + "epoch": 0.5472993498997509, + "grad_norm": 0.21722915768623352, + "learning_rate": 4.282138207157142e-05, + "loss": 1.1166, + "step": 9008 + }, + { + "epoch": 0.5473601069323775, + "grad_norm": 0.5612629652023315, + "learning_rate": 4.281190860579411e-05, + "loss": 1.1371, + "step": 9009 + }, + { + "epoch": 0.547420863965004, + "grad_norm": 0.3778144121170044, + "learning_rate": 4.2802435403497654e-05, + "loss": 1.0481, + "step": 9010 + }, + { + "epoch": 0.5474816209976304, + "grad_norm": 0.2898939847946167, + "learning_rate": 4.2792962465029326e-05, + "loss": 1.0428, + "step": 9011 + }, + { + "epoch": 0.547542378030257, + "grad_norm": 0.25594231486320496, + "learning_rate": 4.278348979073635e-05, + "loss": 1.0493, + "step": 9012 + }, + { + "epoch": 0.5476031350628835, + "grad_norm": 0.1432192325592041, + "learning_rate": 4.2774017380965944e-05, + "loss": 1.027, + "step": 9013 + }, + { + "epoch": 0.54766389209551, + "grad_norm": 0.16895335912704468, + "learning_rate": 4.2764545236065335e-05, + "loss": 1.0947, + "step": 9014 + }, + { + "epoch": 0.5477246491281366, + "grad_norm": 0.5842527747154236, + "learning_rate": 4.275507335638172e-05, + "loss": 1.1227, + "step": 9015 + }, + { + "epoch": 0.5477854061607631, + "grad_norm": 0.28798651695251465, + "learning_rate": 4.2745601742262306e-05, + "loss": 1.1446, + "step": 9016 + }, + { + "epoch": 0.5478461631933896, + "grad_norm": 0.2142031341791153, + "learning_rate": 4.273613039405425e-05, + "loss": 1.1682, + "step": 9017 + }, + { + "epoch": 0.5479069202260162, + "grad_norm": 0.24835893511772156, + "learning_rate": 4.272665931210472e-05, + "loss": 1.1082, + "step": 9018 + }, + { + "epoch": 0.5479676772586427, + "grad_norm": 0.22329942882061005, + "learning_rate": 4.2717188496760916e-05, + "loss": 1.0958, + "step": 9019 + }, + { + "epoch": 0.5480284342912692, + "grad_norm": 0.1584528088569641, + "learning_rate": 4.2707717948369974e-05, + "loss": 1.0691, + "step": 9020 + }, + { + "epoch": 0.5480891913238958, + "grad_norm": 0.16641826927661896, + "learning_rate": 4.2698247667279036e-05, + "loss": 1.0122, + "step": 9021 + }, + { + "epoch": 0.5481499483565223, + "grad_norm": 0.1589505523443222, + "learning_rate": 4.268877765383524e-05, + "loss": 1.0183, + "step": 9022 + }, + { + "epoch": 0.5482107053891488, + "grad_norm": 0.2100931853055954, + "learning_rate": 4.267930790838573e-05, + "loss": 1.0983, + "step": 9023 + }, + { + "epoch": 0.5482714624217753, + "grad_norm": 0.15332570672035217, + "learning_rate": 4.266983843127759e-05, + "loss": 1.0303, + "step": 9024 + }, + { + "epoch": 0.5483322194544018, + "grad_norm": 0.14308324456214905, + "learning_rate": 4.2660369222857944e-05, + "loss": 1.0282, + "step": 9025 + }, + { + "epoch": 0.5483929764870283, + "grad_norm": 0.15541036427021027, + "learning_rate": 4.265090028347387e-05, + "loss": 1.0489, + "step": 9026 + }, + { + "epoch": 0.5484537335196549, + "grad_norm": 0.2705892324447632, + "learning_rate": 4.2641431613472475e-05, + "loss": 1.0498, + "step": 9027 + }, + { + "epoch": 0.5485144905522814, + "grad_norm": 0.14915095269680023, + "learning_rate": 4.263196321320083e-05, + "loss": 1.0351, + "step": 9028 + }, + { + "epoch": 0.5485752475849079, + "grad_norm": 0.20254473388195038, + "learning_rate": 4.262249508300598e-05, + "loss": 1.0629, + "step": 9029 + }, + { + "epoch": 0.5486360046175345, + "grad_norm": 0.21875502169132233, + "learning_rate": 4.261302722323504e-05, + "loss": 1.12, + "step": 9030 + }, + { + "epoch": 0.548696761650161, + "grad_norm": 0.23802493512630463, + "learning_rate": 4.260355963423498e-05, + "loss": 1.0504, + "step": 9031 + }, + { + "epoch": 0.5487575186827875, + "grad_norm": 0.337667316198349, + "learning_rate": 4.2594092316352885e-05, + "loss": 1.105, + "step": 9032 + }, + { + "epoch": 0.5488182757154141, + "grad_norm": 0.19323278963565826, + "learning_rate": 4.258462526993576e-05, + "loss": 1.0416, + "step": 9033 + }, + { + "epoch": 0.5488790327480406, + "grad_norm": 0.20298446714878082, + "learning_rate": 4.257515849533063e-05, + "loss": 1.0754, + "step": 9034 + }, + { + "epoch": 0.5489397897806672, + "grad_norm": 0.3311987519264221, + "learning_rate": 4.25656919928845e-05, + "loss": 1.0656, + "step": 9035 + }, + { + "epoch": 0.5490005468132937, + "grad_norm": 0.2473273128271103, + "learning_rate": 4.255622576294438e-05, + "loss": 1.1436, + "step": 9036 + }, + { + "epoch": 0.5490613038459201, + "grad_norm": 0.29233354330062866, + "learning_rate": 4.254675980585723e-05, + "loss": 1.1455, + "step": 9037 + }, + { + "epoch": 0.5491220608785466, + "grad_norm": 0.2671107053756714, + "learning_rate": 4.253729412197004e-05, + "loss": 1.1262, + "step": 9038 + }, + { + "epoch": 0.5491828179111732, + "grad_norm": 0.32056140899658203, + "learning_rate": 4.2527828711629784e-05, + "loss": 1.0675, + "step": 9039 + }, + { + "epoch": 0.5492435749437997, + "grad_norm": 2.0505712032318115, + "learning_rate": 4.25183635751834e-05, + "loss": 1.07, + "step": 9040 + }, + { + "epoch": 0.5493043319764263, + "grad_norm": 0.39371100068092346, + "learning_rate": 4.250889871297785e-05, + "loss": 1.127, + "step": 9041 + }, + { + "epoch": 0.5493650890090528, + "grad_norm": 0.2312161922454834, + "learning_rate": 4.249943412536006e-05, + "loss": 1.0766, + "step": 9042 + }, + { + "epoch": 0.5494258460416793, + "grad_norm": 0.17141279578208923, + "learning_rate": 4.248996981267698e-05, + "loss": 1.0366, + "step": 9043 + }, + { + "epoch": 0.5494866030743059, + "grad_norm": 0.2223292887210846, + "learning_rate": 4.24805057752755e-05, + "loss": 1.0165, + "step": 9044 + }, + { + "epoch": 0.5495473601069324, + "grad_norm": 2.325972080230713, + "learning_rate": 4.2471042013502534e-05, + "loss": 1.0174, + "step": 9045 + }, + { + "epoch": 0.5496081171395589, + "grad_norm": 2.8682708740234375, + "learning_rate": 4.246157852770496e-05, + "loss": 1.2493, + "step": 9046 + }, + { + "epoch": 0.5496688741721855, + "grad_norm": 0.30333712697029114, + "learning_rate": 4.245211531822971e-05, + "loss": 1.1715, + "step": 9047 + }, + { + "epoch": 0.549729631204812, + "grad_norm": 0.24614906311035156, + "learning_rate": 4.244265238542362e-05, + "loss": 1.116, + "step": 9048 + }, + { + "epoch": 0.5497903882374385, + "grad_norm": 3.7144737243652344, + "learning_rate": 4.2433189729633566e-05, + "loss": 1.0046, + "step": 9049 + }, + { + "epoch": 0.549851145270065, + "grad_norm": 0.24935682117938995, + "learning_rate": 4.242372735120642e-05, + "loss": 1.071, + "step": 9050 + }, + { + "epoch": 0.5499119023026915, + "grad_norm": 0.17469541728496552, + "learning_rate": 4.241426525048899e-05, + "loss": 1.028, + "step": 9051 + }, + { + "epoch": 0.549972659335318, + "grad_norm": 0.22592346370220184, + "learning_rate": 4.240480342782814e-05, + "loss": 1.1127, + "step": 9052 + }, + { + "epoch": 0.5500334163679446, + "grad_norm": 0.15003247559070587, + "learning_rate": 4.239534188357069e-05, + "loss": 1.0556, + "step": 9053 + }, + { + "epoch": 0.5500941734005711, + "grad_norm": 0.21591393649578094, + "learning_rate": 4.238588061806345e-05, + "loss": 1.0109, + "step": 9054 + }, + { + "epoch": 0.5501549304331976, + "grad_norm": 0.18142741918563843, + "learning_rate": 4.237641963165323e-05, + "loss": 1.0372, + "step": 9055 + }, + { + "epoch": 0.5502156874658242, + "grad_norm": 0.18496496975421906, + "learning_rate": 4.236695892468684e-05, + "loss": 1.0226, + "step": 9056 + }, + { + "epoch": 0.5502764444984507, + "grad_norm": 0.24200059473514557, + "learning_rate": 4.2357498497511026e-05, + "loss": 1.2184, + "step": 9057 + }, + { + "epoch": 0.5503372015310772, + "grad_norm": 0.491447776556015, + "learning_rate": 4.234803835047259e-05, + "loss": 1.1226, + "step": 9058 + }, + { + "epoch": 0.5503979585637038, + "grad_norm": 0.21173228323459625, + "learning_rate": 4.233857848391828e-05, + "loss": 1.0305, + "step": 9059 + }, + { + "epoch": 0.5504587155963303, + "grad_norm": 0.26113247871398926, + "learning_rate": 4.232911889819486e-05, + "loss": 1.1014, + "step": 9060 + }, + { + "epoch": 0.5505194726289568, + "grad_norm": 0.2716074585914612, + "learning_rate": 4.231965959364907e-05, + "loss": 1.0685, + "step": 9061 + }, + { + "epoch": 0.5505802296615834, + "grad_norm": 0.29457083344459534, + "learning_rate": 4.231020057062763e-05, + "loss": 1.0172, + "step": 9062 + }, + { + "epoch": 0.5506409866942098, + "grad_norm": 0.2366848886013031, + "learning_rate": 4.230074182947731e-05, + "loss": 1.1461, + "step": 9063 + }, + { + "epoch": 0.5507017437268363, + "grad_norm": 0.17740757763385773, + "learning_rate": 4.2291283370544754e-05, + "loss": 1.0657, + "step": 9064 + }, + { + "epoch": 0.5507625007594629, + "grad_norm": 0.21471595764160156, + "learning_rate": 4.228182519417671e-05, + "loss": 1.0848, + "step": 9065 + }, + { + "epoch": 0.5508232577920894, + "grad_norm": 0.2537383735179901, + "learning_rate": 4.2272367300719846e-05, + "loss": 1.1122, + "step": 9066 + }, + { + "epoch": 0.5508840148247159, + "grad_norm": 0.19461524486541748, + "learning_rate": 4.2262909690520845e-05, + "loss": 1.0697, + "step": 9067 + }, + { + "epoch": 0.5509447718573425, + "grad_norm": 0.3284063935279846, + "learning_rate": 4.22534523639264e-05, + "loss": 1.0488, + "step": 9068 + }, + { + "epoch": 0.551005528889969, + "grad_norm": 0.17272064089775085, + "learning_rate": 4.224399532128315e-05, + "loss": 1.0829, + "step": 9069 + }, + { + "epoch": 0.5510662859225955, + "grad_norm": 0.15949051082134247, + "learning_rate": 4.223453856293777e-05, + "loss": 1.0453, + "step": 9070 + }, + { + "epoch": 0.5511270429552221, + "grad_norm": 0.1752612590789795, + "learning_rate": 4.222508208923686e-05, + "loss": 1.0201, + "step": 9071 + }, + { + "epoch": 0.5511877999878486, + "grad_norm": 0.24758635461330414, + "learning_rate": 4.221562590052707e-05, + "loss": 1.1434, + "step": 9072 + }, + { + "epoch": 0.5512485570204751, + "grad_norm": 11.714828491210938, + "learning_rate": 4.220616999715501e-05, + "loss": 1.0928, + "step": 9073 + }, + { + "epoch": 0.5513093140531017, + "grad_norm": 0.19036731123924255, + "learning_rate": 4.219671437946731e-05, + "loss": 1.0486, + "step": 9074 + }, + { + "epoch": 0.5513700710857282, + "grad_norm": 0.23644955456256866, + "learning_rate": 4.218725904781054e-05, + "loss": 1.1566, + "step": 9075 + }, + { + "epoch": 0.5514308281183548, + "grad_norm": 0.2116638720035553, + "learning_rate": 4.2177804002531326e-05, + "loss": 1.1388, + "step": 9076 + }, + { + "epoch": 0.5514915851509812, + "grad_norm": 0.4794771671295166, + "learning_rate": 4.2168349243976194e-05, + "loss": 1.1787, + "step": 9077 + }, + { + "epoch": 0.5515523421836077, + "grad_norm": 0.18746381998062134, + "learning_rate": 4.215889477249174e-05, + "loss": 1.1297, + "step": 9078 + }, + { + "epoch": 0.5516130992162342, + "grad_norm": 0.14614573121070862, + "learning_rate": 4.2149440588424515e-05, + "loss": 1.0127, + "step": 9079 + }, + { + "epoch": 0.5516738562488608, + "grad_norm": 0.25298500061035156, + "learning_rate": 4.2139986692121045e-05, + "loss": 1.1663, + "step": 9080 + }, + { + "epoch": 0.5517346132814873, + "grad_norm": 0.15330007672309875, + "learning_rate": 4.21305330839279e-05, + "loss": 1.0376, + "step": 9081 + }, + { + "epoch": 0.5517953703141139, + "grad_norm": 0.14236119389533997, + "learning_rate": 4.212107976419159e-05, + "loss": 1.0504, + "step": 9082 + }, + { + "epoch": 0.5518561273467404, + "grad_norm": 0.2261180579662323, + "learning_rate": 4.211162673325863e-05, + "loss": 1.1196, + "step": 9083 + }, + { + "epoch": 0.5519168843793669, + "grad_norm": 0.2575688064098358, + "learning_rate": 4.210217399147551e-05, + "loss": 1.1672, + "step": 9084 + }, + { + "epoch": 0.5519776414119935, + "grad_norm": 0.40841105580329895, + "learning_rate": 4.209272153918871e-05, + "loss": 1.0573, + "step": 9085 + }, + { + "epoch": 0.55203839844462, + "grad_norm": 0.34509265422821045, + "learning_rate": 4.208326937674475e-05, + "loss": 1.1387, + "step": 9086 + }, + { + "epoch": 0.5520991554772465, + "grad_norm": 0.9564976692199707, + "learning_rate": 4.2073817504490076e-05, + "loss": 1.1565, + "step": 9087 + }, + { + "epoch": 0.5521599125098731, + "grad_norm": 0.4409131109714508, + "learning_rate": 4.2064365922771145e-05, + "loss": 1.0207, + "step": 9088 + }, + { + "epoch": 0.5522206695424996, + "grad_norm": 0.7595028281211853, + "learning_rate": 4.205491463193442e-05, + "loss": 1.084, + "step": 9089 + }, + { + "epoch": 0.552281426575126, + "grad_norm": 0.250903844833374, + "learning_rate": 4.204546363232635e-05, + "loss": 1.0377, + "step": 9090 + }, + { + "epoch": 0.5523421836077526, + "grad_norm": 0.393002986907959, + "learning_rate": 4.203601292429334e-05, + "loss": 1.0006, + "step": 9091 + }, + { + "epoch": 0.5524029406403791, + "grad_norm": 6.953628063201904, + "learning_rate": 4.202656250818181e-05, + "loss": 1.0976, + "step": 9092 + }, + { + "epoch": 0.5524636976730056, + "grad_norm": 0.2244192510843277, + "learning_rate": 4.201711238433817e-05, + "loss": 1.0562, + "step": 9093 + }, + { + "epoch": 0.5525244547056322, + "grad_norm": 0.24274368584156036, + "learning_rate": 4.2007662553108825e-05, + "loss": 1.0227, + "step": 9094 + }, + { + "epoch": 0.5525852117382587, + "grad_norm": 0.20965056121349335, + "learning_rate": 4.199821301484015e-05, + "loss": 1.0045, + "step": 9095 + }, + { + "epoch": 0.5526459687708852, + "grad_norm": 0.21032750606536865, + "learning_rate": 4.198876376987851e-05, + "loss": 1.0779, + "step": 9096 + }, + { + "epoch": 0.5527067258035118, + "grad_norm": 0.19821666181087494, + "learning_rate": 4.197931481857033e-05, + "loss": 1.0882, + "step": 9097 + }, + { + "epoch": 0.5527674828361383, + "grad_norm": 0.23598548769950867, + "learning_rate": 4.196986616126187e-05, + "loss": 1.0428, + "step": 9098 + }, + { + "epoch": 0.5528282398687648, + "grad_norm": 0.30615752935409546, + "learning_rate": 4.1960417798299536e-05, + "loss": 1.0756, + "step": 9099 + }, + { + "epoch": 0.5528889969013914, + "grad_norm": 0.364027202129364, + "learning_rate": 4.195096973002964e-05, + "loss": 1.0719, + "step": 9100 + }, + { + "epoch": 0.5529497539340179, + "grad_norm": 0.21617257595062256, + "learning_rate": 4.194152195679849e-05, + "loss": 1.1, + "step": 9101 + }, + { + "epoch": 0.5530105109666444, + "grad_norm": 0.1933189481496811, + "learning_rate": 4.193207447895243e-05, + "loss": 1.0792, + "step": 9102 + }, + { + "epoch": 0.5530712679992709, + "grad_norm": 0.29247280955314636, + "learning_rate": 4.192262729683774e-05, + "loss": 1.119, + "step": 9103 + }, + { + "epoch": 0.5531320250318974, + "grad_norm": 0.5027613043785095, + "learning_rate": 4.19131804108007e-05, + "loss": 1.1913, + "step": 9104 + }, + { + "epoch": 0.5531927820645239, + "grad_norm": 0.30796974897384644, + "learning_rate": 4.1903733821187594e-05, + "loss": 1.0852, + "step": 9105 + }, + { + "epoch": 0.5532535390971505, + "grad_norm": 0.7791163325309753, + "learning_rate": 4.189428752834468e-05, + "loss": 1.1095, + "step": 9106 + }, + { + "epoch": 0.553314296129777, + "grad_norm": 0.8154774308204651, + "learning_rate": 4.188484153261824e-05, + "loss": 1.4459, + "step": 9107 + }, + { + "epoch": 0.5533750531624035, + "grad_norm": 0.879037618637085, + "learning_rate": 4.18753958343545e-05, + "loss": 1.0436, + "step": 9108 + }, + { + "epoch": 0.5534358101950301, + "grad_norm": 0.22916442155838013, + "learning_rate": 4.186595043389968e-05, + "loss": 1.0558, + "step": 9109 + }, + { + "epoch": 0.5534965672276566, + "grad_norm": 0.3045390844345093, + "learning_rate": 4.185650533160005e-05, + "loss": 1.1019, + "step": 9110 + }, + { + "epoch": 0.5535573242602831, + "grad_norm": 0.3844398856163025, + "learning_rate": 4.1847060527801766e-05, + "loss": 1.0448, + "step": 9111 + }, + { + "epoch": 0.5536180812929097, + "grad_norm": 0.3754139840602875, + "learning_rate": 4.183761602285106e-05, + "loss": 1.0212, + "step": 9112 + }, + { + "epoch": 0.5536788383255362, + "grad_norm": 0.22499346733093262, + "learning_rate": 4.182817181709412e-05, + "loss": 1.0144, + "step": 9113 + }, + { + "epoch": 0.5537395953581628, + "grad_norm": 0.4877437949180603, + "learning_rate": 4.181872791087711e-05, + "loss": 1.085, + "step": 9114 + }, + { + "epoch": 0.5538003523907893, + "grad_norm": 0.3879311680793762, + "learning_rate": 4.180928430454622e-05, + "loss": 1.0082, + "step": 9115 + }, + { + "epoch": 0.5538611094234157, + "grad_norm": 0.2792123258113861, + "learning_rate": 4.17998409984476e-05, + "loss": 1.0465, + "step": 9116 + }, + { + "epoch": 0.5539218664560422, + "grad_norm": 0.35094037652015686, + "learning_rate": 4.17903979929274e-05, + "loss": 1.0636, + "step": 9117 + }, + { + "epoch": 0.5539826234886688, + "grad_norm": 0.24818752706050873, + "learning_rate": 4.178095528833174e-05, + "loss": 1.0511, + "step": 9118 + }, + { + "epoch": 0.5540433805212953, + "grad_norm": 0.1677802950143814, + "learning_rate": 4.177151288500674e-05, + "loss": 1.0231, + "step": 9119 + }, + { + "epoch": 0.5541041375539218, + "grad_norm": 0.28627899289131165, + "learning_rate": 4.176207078329853e-05, + "loss": 1.0662, + "step": 9120 + }, + { + "epoch": 0.5541648945865484, + "grad_norm": 0.16133445501327515, + "learning_rate": 4.175262898355322e-05, + "loss": 1.0533, + "step": 9121 + }, + { + "epoch": 0.5542256516191749, + "grad_norm": 0.3476414382457733, + "learning_rate": 4.174318748611687e-05, + "loss": 1.122, + "step": 9122 + }, + { + "epoch": 0.5542864086518015, + "grad_norm": 1.2353898286819458, + "learning_rate": 4.1733746291335616e-05, + "loss": 1.0186, + "step": 9123 + }, + { + "epoch": 0.554347165684428, + "grad_norm": 0.23779815435409546, + "learning_rate": 4.172430539955546e-05, + "loss": 1.0497, + "step": 9124 + }, + { + "epoch": 0.5544079227170545, + "grad_norm": 1.0510014295578003, + "learning_rate": 4.1714864811122495e-05, + "loss": 1.2105, + "step": 9125 + }, + { + "epoch": 0.5544686797496811, + "grad_norm": 0.2436598837375641, + "learning_rate": 4.1705424526382765e-05, + "loss": 1.0261, + "step": 9126 + }, + { + "epoch": 0.5545294367823076, + "grad_norm": 0.270253986120224, + "learning_rate": 4.1695984545682285e-05, + "loss": 1.0321, + "step": 9127 + }, + { + "epoch": 0.5545901938149341, + "grad_norm": 0.3611792325973511, + "learning_rate": 4.1686544869367114e-05, + "loss": 1.0108, + "step": 9128 + }, + { + "epoch": 0.5546509508475606, + "grad_norm": 0.17206333577632904, + "learning_rate": 4.167710549778324e-05, + "loss": 1.0491, + "step": 9129 + }, + { + "epoch": 0.5547117078801871, + "grad_norm": 0.18194781243801117, + "learning_rate": 4.166766643127669e-05, + "loss": 1.0523, + "step": 9130 + }, + { + "epoch": 0.5547724649128136, + "grad_norm": 0.2747882008552551, + "learning_rate": 4.165822767019343e-05, + "loss": 1.0981, + "step": 9131 + }, + { + "epoch": 0.5548332219454402, + "grad_norm": 0.3847016990184784, + "learning_rate": 4.164878921487943e-05, + "loss": 1.0825, + "step": 9132 + }, + { + "epoch": 0.5548939789780667, + "grad_norm": 0.19339650869369507, + "learning_rate": 4.163935106568069e-05, + "loss": 1.1253, + "step": 9133 + }, + { + "epoch": 0.5549547360106932, + "grad_norm": 0.3405653238296509, + "learning_rate": 4.1629913222943146e-05, + "loss": 1.129, + "step": 9134 + }, + { + "epoch": 0.5550154930433198, + "grad_norm": 1.524662733078003, + "learning_rate": 4.162047568701275e-05, + "loss": 1.1189, + "step": 9135 + }, + { + "epoch": 0.5550762500759463, + "grad_norm": 0.2920164167881012, + "learning_rate": 4.1611038458235435e-05, + "loss": 1.2205, + "step": 9136 + }, + { + "epoch": 0.5551370071085728, + "grad_norm": 0.17263273894786835, + "learning_rate": 4.1601601536957144e-05, + "loss": 1.0617, + "step": 9137 + }, + { + "epoch": 0.5551977641411994, + "grad_norm": 0.2837333083152771, + "learning_rate": 4.159216492352377e-05, + "loss": 1.0372, + "step": 9138 + }, + { + "epoch": 0.5552585211738259, + "grad_norm": 0.23039424419403076, + "learning_rate": 4.15827286182812e-05, + "loss": 1.0977, + "step": 9139 + }, + { + "epoch": 0.5553192782064524, + "grad_norm": 0.15666313469409943, + "learning_rate": 4.1573292621575344e-05, + "loss": 1.0651, + "step": 9140 + }, + { + "epoch": 0.555380035239079, + "grad_norm": 0.289902001619339, + "learning_rate": 4.1563856933752075e-05, + "loss": 1.0638, + "step": 9141 + }, + { + "epoch": 0.5554407922717054, + "grad_norm": 0.7130061984062195, + "learning_rate": 4.1554421555157266e-05, + "loss": 1.0386, + "step": 9142 + }, + { + "epoch": 0.5555015493043319, + "grad_norm": 0.23457007110118866, + "learning_rate": 4.1544986486136784e-05, + "loss": 1.0534, + "step": 9143 + }, + { + "epoch": 0.5555623063369585, + "grad_norm": 0.2509375810623169, + "learning_rate": 4.153555172703643e-05, + "loss": 1.0661, + "step": 9144 + }, + { + "epoch": 0.555623063369585, + "grad_norm": 0.513267993927002, + "learning_rate": 4.152611727820207e-05, + "loss": 1.2331, + "step": 9145 + }, + { + "epoch": 0.5556838204022115, + "grad_norm": 0.2004573941230774, + "learning_rate": 4.151668313997952e-05, + "loss": 1.0924, + "step": 9146 + }, + { + "epoch": 0.5557445774348381, + "grad_norm": 0.182392418384552, + "learning_rate": 4.150724931271458e-05, + "loss": 1.0412, + "step": 9147 + }, + { + "epoch": 0.5558053344674646, + "grad_norm": 3.0133233070373535, + "learning_rate": 4.149781579675307e-05, + "loss": 1.1814, + "step": 9148 + }, + { + "epoch": 0.5558660915000911, + "grad_norm": 0.36485525965690613, + "learning_rate": 4.148838259244077e-05, + "loss": 0.9993, + "step": 9149 + }, + { + "epoch": 0.5559268485327177, + "grad_norm": 0.26010411977767944, + "learning_rate": 4.147894970012345e-05, + "loss": 1.0486, + "step": 9150 + }, + { + "epoch": 0.5559876055653442, + "grad_norm": 0.15971556305885315, + "learning_rate": 4.146951712014688e-05, + "loss": 1.0504, + "step": 9151 + }, + { + "epoch": 0.5560483625979707, + "grad_norm": 0.31349319219589233, + "learning_rate": 4.146008485285678e-05, + "loss": 1.0768, + "step": 9152 + }, + { + "epoch": 0.5561091196305973, + "grad_norm": 0.22753791511058807, + "learning_rate": 4.145065289859895e-05, + "loss": 1.1757, + "step": 9153 + }, + { + "epoch": 0.5561698766632238, + "grad_norm": 0.17676420509815216, + "learning_rate": 4.144122125771908e-05, + "loss": 1.0032, + "step": 9154 + }, + { + "epoch": 0.5562306336958502, + "grad_norm": 0.2902083396911621, + "learning_rate": 4.143178993056289e-05, + "loss": 1.0564, + "step": 9155 + }, + { + "epoch": 0.5562913907284768, + "grad_norm": 0.31940898299217224, + "learning_rate": 4.1422358917476114e-05, + "loss": 1.1449, + "step": 9156 + }, + { + "epoch": 0.5563521477611033, + "grad_norm": 0.1790735125541687, + "learning_rate": 4.1412928218804434e-05, + "loss": 1.0625, + "step": 9157 + }, + { + "epoch": 0.5564129047937298, + "grad_norm": 0.2167191207408905, + "learning_rate": 4.140349783489352e-05, + "loss": 1.1103, + "step": 9158 + }, + { + "epoch": 0.5564736618263564, + "grad_norm": 0.2603096067905426, + "learning_rate": 4.139406776608905e-05, + "loss": 1.1368, + "step": 9159 + }, + { + "epoch": 0.5565344188589829, + "grad_norm": 0.20530657470226288, + "learning_rate": 4.138463801273668e-05, + "loss": 1.0757, + "step": 9160 + }, + { + "epoch": 0.5565951758916095, + "grad_norm": 0.2085428088903427, + "learning_rate": 4.1375208575182085e-05, + "loss": 1.1452, + "step": 9161 + }, + { + "epoch": 0.556655932924236, + "grad_norm": 0.14733819663524628, + "learning_rate": 4.136577945377088e-05, + "loss": 1.0276, + "step": 9162 + }, + { + "epoch": 0.5567166899568625, + "grad_norm": 0.2207883894443512, + "learning_rate": 4.13563506488487e-05, + "loss": 1.175, + "step": 9163 + }, + { + "epoch": 0.556777446989489, + "grad_norm": 0.23376551270484924, + "learning_rate": 4.1346922160761184e-05, + "loss": 1.1682, + "step": 9164 + }, + { + "epoch": 0.5568382040221156, + "grad_norm": 0.19631215929985046, + "learning_rate": 4.133749398985387e-05, + "loss": 1.1163, + "step": 9165 + }, + { + "epoch": 0.5568989610547421, + "grad_norm": 0.22082148492336273, + "learning_rate": 4.1328066136472405e-05, + "loss": 1.2073, + "step": 9166 + }, + { + "epoch": 0.5569597180873687, + "grad_norm": 0.19571058452129364, + "learning_rate": 4.1318638600962346e-05, + "loss": 1.1216, + "step": 9167 + }, + { + "epoch": 0.5570204751199951, + "grad_norm": 1.0029270648956299, + "learning_rate": 4.130921138366926e-05, + "loss": 1.0816, + "step": 9168 + }, + { + "epoch": 0.5570812321526216, + "grad_norm": 0.17740383744239807, + "learning_rate": 4.1299784484938726e-05, + "loss": 1.0439, + "step": 9169 + }, + { + "epoch": 0.5571419891852482, + "grad_norm": 0.22257710993289948, + "learning_rate": 4.1290357905116276e-05, + "loss": 1.0191, + "step": 9170 + }, + { + "epoch": 0.5572027462178747, + "grad_norm": 0.1655077040195465, + "learning_rate": 4.1280931644547435e-05, + "loss": 1.135, + "step": 9171 + }, + { + "epoch": 0.5572635032505012, + "grad_norm": 0.541623055934906, + "learning_rate": 4.1271505703577726e-05, + "loss": 1.0594, + "step": 9172 + }, + { + "epoch": 0.5573242602831278, + "grad_norm": 0.17888161540031433, + "learning_rate": 4.1262080082552656e-05, + "loss": 1.0207, + "step": 9173 + }, + { + "epoch": 0.5573850173157543, + "grad_norm": 0.3191627860069275, + "learning_rate": 4.1252654781817746e-05, + "loss": 1.2455, + "step": 9174 + }, + { + "epoch": 0.5574457743483808, + "grad_norm": 0.24886097013950348, + "learning_rate": 4.124322980171845e-05, + "loss": 1.1333, + "step": 9175 + }, + { + "epoch": 0.5575065313810074, + "grad_norm": 0.18209686875343323, + "learning_rate": 4.1233805142600265e-05, + "loss": 1.0285, + "step": 9176 + }, + { + "epoch": 0.5575672884136339, + "grad_norm": 0.2062995731830597, + "learning_rate": 4.122438080480867e-05, + "loss": 1.0932, + "step": 9177 + }, + { + "epoch": 0.5576280454462604, + "grad_norm": 0.2716463506221771, + "learning_rate": 4.121495678868906e-05, + "loss": 1.089, + "step": 9178 + }, + { + "epoch": 0.557688802478887, + "grad_norm": 0.2622074782848358, + "learning_rate": 4.1205533094586916e-05, + "loss": 1.121, + "step": 9179 + }, + { + "epoch": 0.5577495595115135, + "grad_norm": 0.2517070174217224, + "learning_rate": 4.1196109722847666e-05, + "loss": 1.15, + "step": 9180 + }, + { + "epoch": 0.55781031654414, + "grad_norm": 0.1623789221048355, + "learning_rate": 4.118668667381669e-05, + "loss": 1.0665, + "step": 9181 + }, + { + "epoch": 0.5578710735767665, + "grad_norm": 0.23418322205543518, + "learning_rate": 4.117726394783944e-05, + "loss": 1.0651, + "step": 9182 + }, + { + "epoch": 0.557931830609393, + "grad_norm": 0.3779861330986023, + "learning_rate": 4.116784154526128e-05, + "loss": 1.1092, + "step": 9183 + }, + { + "epoch": 0.5579925876420195, + "grad_norm": 0.15771810710430145, + "learning_rate": 4.115841946642761e-05, + "loss": 1.0342, + "step": 9184 + }, + { + "epoch": 0.5580533446746461, + "grad_norm": 0.2844100296497345, + "learning_rate": 4.114899771168377e-05, + "loss": 1.1593, + "step": 9185 + }, + { + "epoch": 0.5581141017072726, + "grad_norm": 0.21968477964401245, + "learning_rate": 4.1139576281375125e-05, + "loss": 1.1067, + "step": 9186 + }, + { + "epoch": 0.5581748587398991, + "grad_norm": 0.22269387543201447, + "learning_rate": 4.1130155175847026e-05, + "loss": 1.0721, + "step": 9187 + }, + { + "epoch": 0.5582356157725257, + "grad_norm": 0.16716499626636505, + "learning_rate": 4.112073439544481e-05, + "loss": 1.0418, + "step": 9188 + }, + { + "epoch": 0.5582963728051522, + "grad_norm": 0.30056872963905334, + "learning_rate": 4.1111313940513776e-05, + "loss": 1.0304, + "step": 9189 + }, + { + "epoch": 0.5583571298377787, + "grad_norm": 0.2642410397529602, + "learning_rate": 4.110189381139928e-05, + "loss": 1.1716, + "step": 9190 + }, + { + "epoch": 0.5584178868704053, + "grad_norm": 0.21299555897712708, + "learning_rate": 4.109247400844656e-05, + "loss": 1.0674, + "step": 9191 + }, + { + "epoch": 0.5584786439030318, + "grad_norm": 0.21428631246089935, + "learning_rate": 4.108305453200093e-05, + "loss": 1.1411, + "step": 9192 + }, + { + "epoch": 0.5585394009356583, + "grad_norm": 0.2423153817653656, + "learning_rate": 4.107363538240767e-05, + "loss": 1.1577, + "step": 9193 + }, + { + "epoch": 0.5586001579682849, + "grad_norm": 0.3740766644477844, + "learning_rate": 4.106421656001201e-05, + "loss": 0.9827, + "step": 9194 + }, + { + "epoch": 0.5586609150009113, + "grad_norm": 0.1541246622800827, + "learning_rate": 4.105479806515923e-05, + "loss": 1.0468, + "step": 9195 + }, + { + "epoch": 0.5587216720335378, + "grad_norm": 0.42135143280029297, + "learning_rate": 4.104537989819456e-05, + "loss": 1.1536, + "step": 9196 + }, + { + "epoch": 0.5587824290661644, + "grad_norm": 0.21386271715164185, + "learning_rate": 4.103596205946323e-05, + "loss": 1.0511, + "step": 9197 + }, + { + "epoch": 0.5588431860987909, + "grad_norm": 0.20215994119644165, + "learning_rate": 4.102654454931043e-05, + "loss": 1.1233, + "step": 9198 + }, + { + "epoch": 0.5589039431314174, + "grad_norm": 0.27129048109054565, + "learning_rate": 4.101712736808137e-05, + "loss": 1.0069, + "step": 9199 + }, + { + "epoch": 0.558964700164044, + "grad_norm": 1.121099829673767, + "learning_rate": 4.100771051612125e-05, + "loss": 1.1117, + "step": 9200 + }, + { + "epoch": 0.5590254571966705, + "grad_norm": 0.20939838886260986, + "learning_rate": 4.0998293993775237e-05, + "loss": 1.1182, + "step": 9201 + }, + { + "epoch": 0.559086214229297, + "grad_norm": 0.3378422260284424, + "learning_rate": 4.098887780138849e-05, + "loss": 1.0446, + "step": 9202 + }, + { + "epoch": 0.5591469712619236, + "grad_norm": 0.27399373054504395, + "learning_rate": 4.097946193930618e-05, + "loss": 1.1747, + "step": 9203 + }, + { + "epoch": 0.5592077282945501, + "grad_norm": 0.2535637319087982, + "learning_rate": 4.0970046407873444e-05, + "loss": 1.2042, + "step": 9204 + }, + { + "epoch": 0.5592684853271767, + "grad_norm": 0.46783891320228577, + "learning_rate": 4.0960631207435394e-05, + "loss": 1.063, + "step": 9205 + }, + { + "epoch": 0.5593292423598032, + "grad_norm": 0.19766885042190552, + "learning_rate": 4.095121633833715e-05, + "loss": 1.0977, + "step": 9206 + }, + { + "epoch": 0.5593899993924297, + "grad_norm": 0.24793611466884613, + "learning_rate": 4.094180180092382e-05, + "loss": 1.0713, + "step": 9207 + }, + { + "epoch": 0.5594507564250562, + "grad_norm": 0.20317505300045013, + "learning_rate": 4.0932387595540496e-05, + "loss": 1.2405, + "step": 9208 + }, + { + "epoch": 0.5595115134576827, + "grad_norm": 0.2307678759098053, + "learning_rate": 4.0922973722532264e-05, + "loss": 1.1095, + "step": 9209 + }, + { + "epoch": 0.5595722704903092, + "grad_norm": 0.1522945910692215, + "learning_rate": 4.091356018224419e-05, + "loss": 1.0499, + "step": 9210 + }, + { + "epoch": 0.5596330275229358, + "grad_norm": 0.26950347423553467, + "learning_rate": 4.090414697502132e-05, + "loss": 1.1424, + "step": 9211 + }, + { + "epoch": 0.5596937845555623, + "grad_norm": 0.48247432708740234, + "learning_rate": 4.089473410120868e-05, + "loss": 1.0304, + "step": 9212 + }, + { + "epoch": 0.5597545415881888, + "grad_norm": 0.2127780318260193, + "learning_rate": 4.088532156115134e-05, + "loss": 1.0695, + "step": 9213 + }, + { + "epoch": 0.5598152986208154, + "grad_norm": 0.15464693307876587, + "learning_rate": 4.08759093551943e-05, + "loss": 1.0845, + "step": 9214 + }, + { + "epoch": 0.5598760556534419, + "grad_norm": 1.6254702806472778, + "learning_rate": 4.0866497483682556e-05, + "loss": 1.0868, + "step": 9215 + }, + { + "epoch": 0.5599368126860684, + "grad_norm": 1.9411534070968628, + "learning_rate": 4.0857085946961115e-05, + "loss": 1.2948, + "step": 9216 + }, + { + "epoch": 0.559997569718695, + "grad_norm": 0.15538617968559265, + "learning_rate": 4.0847674745374974e-05, + "loss": 0.9943, + "step": 9217 + }, + { + "epoch": 0.5600583267513215, + "grad_norm": 0.1765478551387787, + "learning_rate": 4.0838263879269073e-05, + "loss": 1.0839, + "step": 9218 + }, + { + "epoch": 0.560119083783948, + "grad_norm": 0.2513713836669922, + "learning_rate": 4.082885334898838e-05, + "loss": 1.0492, + "step": 9219 + }, + { + "epoch": 0.5601798408165746, + "grad_norm": 0.19734176993370056, + "learning_rate": 4.081944315487783e-05, + "loss": 1.1153, + "step": 9220 + }, + { + "epoch": 0.560240597849201, + "grad_norm": 0.17919893562793732, + "learning_rate": 4.0810033297282375e-05, + "loss": 1.0496, + "step": 9221 + }, + { + "epoch": 0.5603013548818275, + "grad_norm": 0.1951284408569336, + "learning_rate": 4.080062377654692e-05, + "loss": 1.1022, + "step": 9222 + }, + { + "epoch": 0.5603621119144541, + "grad_norm": 0.3104482889175415, + "learning_rate": 4.079121459301638e-05, + "loss": 1.1249, + "step": 9223 + }, + { + "epoch": 0.5604228689470806, + "grad_norm": 5.347562789916992, + "learning_rate": 4.078180574703567e-05, + "loss": 1.0733, + "step": 9224 + }, + { + "epoch": 0.5604836259797071, + "grad_norm": 0.1700235903263092, + "learning_rate": 4.077239723894963e-05, + "loss": 1.0706, + "step": 9225 + }, + { + "epoch": 0.5605443830123337, + "grad_norm": 0.458423376083374, + "learning_rate": 4.0762989069103156e-05, + "loss": 1.0901, + "step": 9226 + }, + { + "epoch": 0.5606051400449602, + "grad_norm": 0.22139477729797363, + "learning_rate": 4.075358123784111e-05, + "loss": 1.1026, + "step": 9227 + }, + { + "epoch": 0.5606658970775867, + "grad_norm": 0.29300954937934875, + "learning_rate": 4.074417374550831e-05, + "loss": 1.1606, + "step": 9228 + }, + { + "epoch": 0.5607266541102133, + "grad_norm": 0.27637046575546265, + "learning_rate": 4.073476659244963e-05, + "loss": 1.0302, + "step": 9229 + }, + { + "epoch": 0.5607874111428398, + "grad_norm": 0.29441776871681213, + "learning_rate": 4.072535977900988e-05, + "loss": 1.079, + "step": 9230 + }, + { + "epoch": 0.5608481681754663, + "grad_norm": 0.23698009550571442, + "learning_rate": 4.071595330553385e-05, + "loss": 1.1124, + "step": 9231 + }, + { + "epoch": 0.5609089252080929, + "grad_norm": 0.23744140565395355, + "learning_rate": 4.070654717236634e-05, + "loss": 1.0733, + "step": 9232 + }, + { + "epoch": 0.5609696822407194, + "grad_norm": 0.19375908374786377, + "learning_rate": 4.069714137985214e-05, + "loss": 1.0348, + "step": 9233 + }, + { + "epoch": 0.5610304392733458, + "grad_norm": 0.5568618774414062, + "learning_rate": 4.0687735928336024e-05, + "loss": 1.1706, + "step": 9234 + }, + { + "epoch": 0.5610911963059724, + "grad_norm": 0.19145552814006805, + "learning_rate": 4.067833081816275e-05, + "loss": 0.9916, + "step": 9235 + }, + { + "epoch": 0.5611519533385989, + "grad_norm": 0.30136987566947937, + "learning_rate": 4.066892604967705e-05, + "loss": 1.1305, + "step": 9236 + }, + { + "epoch": 0.5612127103712254, + "grad_norm": 0.22299502789974213, + "learning_rate": 4.06595216232237e-05, + "loss": 1.0803, + "step": 9237 + }, + { + "epoch": 0.561273467403852, + "grad_norm": 0.5463804602622986, + "learning_rate": 4.0650117539147366e-05, + "loss": 1.1334, + "step": 9238 + }, + { + "epoch": 0.5613342244364785, + "grad_norm": 0.6465639472007751, + "learning_rate": 4.064071379779278e-05, + "loss": 1.2016, + "step": 9239 + }, + { + "epoch": 0.561394981469105, + "grad_norm": 0.4862769544124603, + "learning_rate": 4.0631310399504634e-05, + "loss": 1.23, + "step": 9240 + }, + { + "epoch": 0.5614557385017316, + "grad_norm": 0.2430248260498047, + "learning_rate": 4.062190734462763e-05, + "loss": 1.1409, + "step": 9241 + }, + { + "epoch": 0.5615164955343581, + "grad_norm": 0.2663247585296631, + "learning_rate": 4.061250463350642e-05, + "loss": 1.2411, + "step": 9242 + }, + { + "epoch": 0.5615772525669847, + "grad_norm": 0.2994415760040283, + "learning_rate": 4.060310226648567e-05, + "loss": 1.0373, + "step": 9243 + }, + { + "epoch": 0.5616380095996112, + "grad_norm": 0.21191053092479706, + "learning_rate": 4.059370024391003e-05, + "loss": 1.103, + "step": 9244 + }, + { + "epoch": 0.5616987666322377, + "grad_norm": 0.2837742269039154, + "learning_rate": 4.058429856612411e-05, + "loss": 1.1077, + "step": 9245 + }, + { + "epoch": 0.5617595236648643, + "grad_norm": 0.24049492180347443, + "learning_rate": 4.057489723347255e-05, + "loss": 1.1256, + "step": 9246 + }, + { + "epoch": 0.5618202806974907, + "grad_norm": 0.21877828240394592, + "learning_rate": 4.056549624629996e-05, + "loss": 1.0158, + "step": 9247 + }, + { + "epoch": 0.5618810377301172, + "grad_norm": 0.18928594887256622, + "learning_rate": 4.055609560495092e-05, + "loss": 1.0608, + "step": 9248 + }, + { + "epoch": 0.5619417947627438, + "grad_norm": 0.27199506759643555, + "learning_rate": 4.054669530977002e-05, + "loss": 1.0955, + "step": 9249 + }, + { + "epoch": 0.5620025517953703, + "grad_norm": 0.262079119682312, + "learning_rate": 4.053729536110184e-05, + "loss": 1.0315, + "step": 9250 + }, + { + "epoch": 0.5620633088279968, + "grad_norm": 0.3923673927783966, + "learning_rate": 4.052789575929094e-05, + "loss": 1.0735, + "step": 9251 + }, + { + "epoch": 0.5621240658606234, + "grad_norm": 0.5480532050132751, + "learning_rate": 4.051849650468184e-05, + "loss": 1.0259, + "step": 9252 + }, + { + "epoch": 0.5621848228932499, + "grad_norm": 0.2838411331176758, + "learning_rate": 4.050909759761907e-05, + "loss": 1.0276, + "step": 9253 + }, + { + "epoch": 0.5622455799258764, + "grad_norm": 0.28498899936676025, + "learning_rate": 4.049969903844717e-05, + "loss": 1.1387, + "step": 9254 + }, + { + "epoch": 0.562306336958503, + "grad_norm": 0.338323712348938, + "learning_rate": 4.049030082751064e-05, + "loss": 1.0716, + "step": 9255 + }, + { + "epoch": 0.5623670939911295, + "grad_norm": 0.23632659018039703, + "learning_rate": 4.048090296515396e-05, + "loss": 1.1868, + "step": 9256 + }, + { + "epoch": 0.562427851023756, + "grad_norm": 0.2008480429649353, + "learning_rate": 4.047150545172166e-05, + "loss": 1.0984, + "step": 9257 + }, + { + "epoch": 0.5624886080563826, + "grad_norm": 0.326617568731308, + "learning_rate": 4.046210828755813e-05, + "loss": 1.1338, + "step": 9258 + }, + { + "epoch": 0.5625493650890091, + "grad_norm": 0.5674911141395569, + "learning_rate": 4.0452711473007873e-05, + "loss": 1.012, + "step": 9259 + }, + { + "epoch": 0.5626101221216355, + "grad_norm": 0.20757721364498138, + "learning_rate": 4.0443315008415334e-05, + "loss": 1.1024, + "step": 9260 + }, + { + "epoch": 0.5626708791542621, + "grad_norm": 0.21453210711479187, + "learning_rate": 4.043391889412491e-05, + "loss": 1.0424, + "step": 9261 + }, + { + "epoch": 0.5627316361868886, + "grad_norm": 0.36534321308135986, + "learning_rate": 4.042452313048105e-05, + "loss": 1.0274, + "step": 9262 + }, + { + "epoch": 0.5627923932195151, + "grad_norm": 0.2168557345867157, + "learning_rate": 4.041512771782815e-05, + "loss": 1.2108, + "step": 9263 + }, + { + "epoch": 0.5628531502521417, + "grad_norm": 0.2509775459766388, + "learning_rate": 4.040573265651061e-05, + "loss": 1.1154, + "step": 9264 + }, + { + "epoch": 0.5629139072847682, + "grad_norm": 0.4371791183948517, + "learning_rate": 4.0396337946872774e-05, + "loss": 1.1281, + "step": 9265 + }, + { + "epoch": 0.5629746643173947, + "grad_norm": 0.16411370038986206, + "learning_rate": 4.038694358925902e-05, + "loss": 1.0667, + "step": 9266 + }, + { + "epoch": 0.5630354213500213, + "grad_norm": 0.18401654064655304, + "learning_rate": 4.037754958401372e-05, + "loss": 1.0627, + "step": 9267 + }, + { + "epoch": 0.5630961783826478, + "grad_norm": 0.375328004360199, + "learning_rate": 4.036815593148119e-05, + "loss": 1.1676, + "step": 9268 + }, + { + "epoch": 0.5631569354152743, + "grad_norm": 0.16237080097198486, + "learning_rate": 4.035876263200577e-05, + "loss": 1.0365, + "step": 9269 + }, + { + "epoch": 0.5632176924479009, + "grad_norm": 0.16548191010951996, + "learning_rate": 4.034936968593177e-05, + "loss": 1.0932, + "step": 9270 + }, + { + "epoch": 0.5632784494805274, + "grad_norm": 0.21353679895401, + "learning_rate": 4.033997709360351e-05, + "loss": 1.0411, + "step": 9271 + }, + { + "epoch": 0.563339206513154, + "grad_norm": 0.3227792978286743, + "learning_rate": 4.033058485536524e-05, + "loss": 1.1475, + "step": 9272 + }, + { + "epoch": 0.5633999635457804, + "grad_norm": 0.21658164262771606, + "learning_rate": 4.0321192971561245e-05, + "loss": 1.0957, + "step": 9273 + }, + { + "epoch": 0.5634607205784069, + "grad_norm": 0.18437203764915466, + "learning_rate": 4.031180144253579e-05, + "loss": 1.0844, + "step": 9274 + }, + { + "epoch": 0.5635214776110334, + "grad_norm": 0.29198044538497925, + "learning_rate": 4.0302410268633134e-05, + "loss": 1.0693, + "step": 9275 + }, + { + "epoch": 0.56358223464366, + "grad_norm": 0.23750093579292297, + "learning_rate": 4.02930194501975e-05, + "loss": 1.1418, + "step": 9276 + }, + { + "epoch": 0.5636429916762865, + "grad_norm": 2.5402143001556396, + "learning_rate": 4.028362898757313e-05, + "loss": 1.0166, + "step": 9277 + }, + { + "epoch": 0.563703748708913, + "grad_norm": 0.23183394968509674, + "learning_rate": 4.027423888110421e-05, + "loss": 1.0511, + "step": 9278 + }, + { + "epoch": 0.5637645057415396, + "grad_norm": 0.3369738459587097, + "learning_rate": 4.026484913113493e-05, + "loss": 1.0823, + "step": 9279 + }, + { + "epoch": 0.5638252627741661, + "grad_norm": 0.3425249755382538, + "learning_rate": 4.0255459738009496e-05, + "loss": 1.1256, + "step": 9280 + }, + { + "epoch": 0.5638860198067926, + "grad_norm": 0.2142961472272873, + "learning_rate": 4.0246070702072066e-05, + "loss": 1.0371, + "step": 9281 + }, + { + "epoch": 0.5639467768394192, + "grad_norm": 0.4849246144294739, + "learning_rate": 4.0236682023666796e-05, + "loss": 1.0476, + "step": 9282 + }, + { + "epoch": 0.5640075338720457, + "grad_norm": 0.6679051518440247, + "learning_rate": 4.022729370313784e-05, + "loss": 1.1632, + "step": 9283 + }, + { + "epoch": 0.5640682909046723, + "grad_norm": 0.4357801675796509, + "learning_rate": 4.0217905740829334e-05, + "loss": 0.9835, + "step": 9284 + }, + { + "epoch": 0.5641290479372988, + "grad_norm": 0.14981092512607574, + "learning_rate": 4.020851813708537e-05, + "loss": 1.0204, + "step": 9285 + }, + { + "epoch": 0.5641898049699253, + "grad_norm": 0.1999199390411377, + "learning_rate": 4.0199130892250075e-05, + "loss": 1.157, + "step": 9286 + }, + { + "epoch": 0.5642505620025517, + "grad_norm": 0.21363025903701782, + "learning_rate": 4.018974400666751e-05, + "loss": 1.0474, + "step": 9287 + }, + { + "epoch": 0.5643113190351783, + "grad_norm": 0.22807073593139648, + "learning_rate": 4.0180357480681794e-05, + "loss": 1.1208, + "step": 9288 + }, + { + "epoch": 0.5643720760678048, + "grad_norm": 0.24565954506397247, + "learning_rate": 4.0170971314636974e-05, + "loss": 1.2094, + "step": 9289 + }, + { + "epoch": 0.5644328331004314, + "grad_norm": 0.2292650192975998, + "learning_rate": 4.0161585508877084e-05, + "loss": 1.1735, + "step": 9290 + }, + { + "epoch": 0.5644935901330579, + "grad_norm": 2.8376340866088867, + "learning_rate": 4.015220006374621e-05, + "loss": 1.0706, + "step": 9291 + }, + { + "epoch": 0.5645543471656844, + "grad_norm": 0.197785884141922, + "learning_rate": 4.0142814979588314e-05, + "loss": 1.0115, + "step": 9292 + }, + { + "epoch": 0.564615104198311, + "grad_norm": 0.2838975191116333, + "learning_rate": 4.0133430256747456e-05, + "loss": 1.126, + "step": 9293 + }, + { + "epoch": 0.5646758612309375, + "grad_norm": 0.22421903908252716, + "learning_rate": 4.012404589556762e-05, + "loss": 1.08, + "step": 9294 + }, + { + "epoch": 0.564736618263564, + "grad_norm": 0.3535689413547516, + "learning_rate": 4.011466189639277e-05, + "loss": 1.0079, + "step": 9295 + }, + { + "epoch": 0.5647973752961906, + "grad_norm": 0.19548100233078003, + "learning_rate": 4.010527825956692e-05, + "loss": 1.1313, + "step": 9296 + }, + { + "epoch": 0.5648581323288171, + "grad_norm": 0.3644217550754547, + "learning_rate": 4.009589498543401e-05, + "loss": 1.051, + "step": 9297 + }, + { + "epoch": 0.5649188893614436, + "grad_norm": 0.37840789556503296, + "learning_rate": 4.008651207433798e-05, + "loss": 1.1368, + "step": 9298 + }, + { + "epoch": 0.5649796463940702, + "grad_norm": 0.5139578580856323, + "learning_rate": 4.007712952662277e-05, + "loss": 1.1492, + "step": 9299 + }, + { + "epoch": 0.5650404034266966, + "grad_norm": 0.2759154438972473, + "learning_rate": 4.006774734263227e-05, + "loss": 1.1309, + "step": 9300 + }, + { + "epoch": 0.5651011604593231, + "grad_norm": 0.42458635568618774, + "learning_rate": 4.0058365522710435e-05, + "loss": 1.0908, + "step": 9301 + }, + { + "epoch": 0.5651619174919497, + "grad_norm": 0.2653268873691559, + "learning_rate": 4.0048984067201126e-05, + "loss": 1.3125, + "step": 9302 + }, + { + "epoch": 0.5652226745245762, + "grad_norm": 0.2967388927936554, + "learning_rate": 4.003960297644822e-05, + "loss": 1.0312, + "step": 9303 + }, + { + "epoch": 0.5652834315572027, + "grad_norm": 0.3151795566082001, + "learning_rate": 4.003022225079563e-05, + "loss": 1.1429, + "step": 9304 + }, + { + "epoch": 0.5653441885898293, + "grad_norm": 0.1982724368572235, + "learning_rate": 4.002084189058714e-05, + "loss": 1.0873, + "step": 9305 + }, + { + "epoch": 0.5654049456224558, + "grad_norm": 0.27187997102737427, + "learning_rate": 4.001146189616662e-05, + "loss": 0.9922, + "step": 9306 + }, + { + "epoch": 0.5654657026550823, + "grad_norm": 0.20016787946224213, + "learning_rate": 4.00020822678779e-05, + "loss": 1.0652, + "step": 9307 + }, + { + "epoch": 0.5655264596877089, + "grad_norm": 0.20849180221557617, + "learning_rate": 3.999270300606477e-05, + "loss": 1.1806, + "step": 9308 + }, + { + "epoch": 0.5655872167203354, + "grad_norm": 0.23631073534488678, + "learning_rate": 3.9983324111071064e-05, + "loss": 1.0401, + "step": 9309 + }, + { + "epoch": 0.5656479737529619, + "grad_norm": 0.3045595586299896, + "learning_rate": 3.9973945583240545e-05, + "loss": 1.0407, + "step": 9310 + }, + { + "epoch": 0.5657087307855885, + "grad_norm": 2.1587653160095215, + "learning_rate": 3.9964567422917e-05, + "loss": 1.1456, + "step": 9311 + }, + { + "epoch": 0.565769487818215, + "grad_norm": 0.8412651419639587, + "learning_rate": 3.995518963044417e-05, + "loss": 1.1793, + "step": 9312 + }, + { + "epoch": 0.5658302448508414, + "grad_norm": 0.20178738236427307, + "learning_rate": 3.994581220616579e-05, + "loss": 1.0541, + "step": 9313 + }, + { + "epoch": 0.565891001883468, + "grad_norm": 0.24081389605998993, + "learning_rate": 3.993643515042562e-05, + "loss": 1.0799, + "step": 9314 + }, + { + "epoch": 0.5659517589160945, + "grad_norm": 0.282135009765625, + "learning_rate": 3.992705846356737e-05, + "loss": 1.0364, + "step": 9315 + }, + { + "epoch": 0.566012515948721, + "grad_norm": 0.16417331993579865, + "learning_rate": 3.991768214593472e-05, + "loss": 1.0249, + "step": 9316 + }, + { + "epoch": 0.5660732729813476, + "grad_norm": 0.21000291407108307, + "learning_rate": 3.990830619787139e-05, + "loss": 1.0863, + "step": 9317 + }, + { + "epoch": 0.5661340300139741, + "grad_norm": 0.20507200062274933, + "learning_rate": 3.9898930619721065e-05, + "loss": 1.0697, + "step": 9318 + }, + { + "epoch": 0.5661947870466006, + "grad_norm": 0.20977991819381714, + "learning_rate": 3.9889555411827374e-05, + "loss": 1.0912, + "step": 9319 + }, + { + "epoch": 0.5662555440792272, + "grad_norm": 0.19044329226016998, + "learning_rate": 3.988018057453398e-05, + "loss": 1.0379, + "step": 9320 + }, + { + "epoch": 0.5663163011118537, + "grad_norm": 0.16950348019599915, + "learning_rate": 3.987080610818452e-05, + "loss": 1.0149, + "step": 9321 + }, + { + "epoch": 0.5663770581444802, + "grad_norm": 0.22774747014045715, + "learning_rate": 3.986143201312263e-05, + "loss": 1.0372, + "step": 9322 + }, + { + "epoch": 0.5664378151771068, + "grad_norm": 0.17480848729610443, + "learning_rate": 3.9852058289691905e-05, + "loss": 1.0536, + "step": 9323 + }, + { + "epoch": 0.5664985722097333, + "grad_norm": 0.2882568836212158, + "learning_rate": 3.9842684938235966e-05, + "loss": 1.0079, + "step": 9324 + }, + { + "epoch": 0.5665593292423599, + "grad_norm": 0.2890188992023468, + "learning_rate": 3.9833311959098354e-05, + "loss": 1.038, + "step": 9325 + }, + { + "epoch": 0.5666200862749863, + "grad_norm": 0.44669854640960693, + "learning_rate": 3.982393935262264e-05, + "loss": 1.2551, + "step": 9326 + }, + { + "epoch": 0.5666808433076128, + "grad_norm": 0.31870654225349426, + "learning_rate": 3.9814567119152416e-05, + "loss": 1.0635, + "step": 9327 + }, + { + "epoch": 0.5667416003402393, + "grad_norm": 0.28913289308547974, + "learning_rate": 3.98051952590312e-05, + "loss": 1.1112, + "step": 9328 + }, + { + "epoch": 0.5668023573728659, + "grad_norm": 0.28194379806518555, + "learning_rate": 3.979582377260252e-05, + "loss": 1.0897, + "step": 9329 + }, + { + "epoch": 0.5668631144054924, + "grad_norm": 0.37816929817199707, + "learning_rate": 3.9786452660209895e-05, + "loss": 1.0969, + "step": 9330 + }, + { + "epoch": 0.566923871438119, + "grad_norm": 0.20288652181625366, + "learning_rate": 3.977708192219683e-05, + "loss": 1.1511, + "step": 9331 + }, + { + "epoch": 0.5669846284707455, + "grad_norm": 0.190296933054924, + "learning_rate": 3.97677115589068e-05, + "loss": 1.0346, + "step": 9332 + }, + { + "epoch": 0.567045385503372, + "grad_norm": 0.17884548008441925, + "learning_rate": 3.9758341570683265e-05, + "loss": 1.0433, + "step": 9333 + }, + { + "epoch": 0.5671061425359986, + "grad_norm": 0.3744193911552429, + "learning_rate": 3.974897195786971e-05, + "loss": 1.0165, + "step": 9334 + }, + { + "epoch": 0.5671668995686251, + "grad_norm": 0.25330138206481934, + "learning_rate": 3.9739602720809575e-05, + "loss": 1.1322, + "step": 9335 + }, + { + "epoch": 0.5672276566012516, + "grad_norm": 0.1535419076681137, + "learning_rate": 3.9730233859846284e-05, + "loss": 1.0399, + "step": 9336 + }, + { + "epoch": 0.5672884136338782, + "grad_norm": 0.21788422763347626, + "learning_rate": 3.972086537532324e-05, + "loss": 1.0763, + "step": 9337 + }, + { + "epoch": 0.5673491706665047, + "grad_norm": 0.8888079524040222, + "learning_rate": 3.971149726758389e-05, + "loss": 1.1777, + "step": 9338 + }, + { + "epoch": 0.5674099276991311, + "grad_norm": 0.210330069065094, + "learning_rate": 3.9702129536971577e-05, + "loss": 1.1768, + "step": 9339 + }, + { + "epoch": 0.5674706847317577, + "grad_norm": 0.18061070144176483, + "learning_rate": 3.96927621838297e-05, + "loss": 1.0613, + "step": 9340 + }, + { + "epoch": 0.5675314417643842, + "grad_norm": 0.24510422348976135, + "learning_rate": 3.968339520850161e-05, + "loss": 1.1348, + "step": 9341 + }, + { + "epoch": 0.5675921987970107, + "grad_norm": 0.2316739559173584, + "learning_rate": 3.967402861133068e-05, + "loss": 1.0322, + "step": 9342 + }, + { + "epoch": 0.5676529558296373, + "grad_norm": 0.17277434468269348, + "learning_rate": 3.966466239266021e-05, + "loss": 1.1074, + "step": 9343 + }, + { + "epoch": 0.5677137128622638, + "grad_norm": 0.23352639377117157, + "learning_rate": 3.9655296552833565e-05, + "loss": 1.0078, + "step": 9344 + }, + { + "epoch": 0.5677744698948903, + "grad_norm": 0.21717698872089386, + "learning_rate": 3.9645931092194004e-05, + "loss": 1.0523, + "step": 9345 + }, + { + "epoch": 0.5678352269275169, + "grad_norm": 0.19367381930351257, + "learning_rate": 3.963656601108483e-05, + "loss": 1.0815, + "step": 9346 + }, + { + "epoch": 0.5678959839601434, + "grad_norm": 0.1775786578655243, + "learning_rate": 3.962720130984935e-05, + "loss": 1.0768, + "step": 9347 + }, + { + "epoch": 0.5679567409927699, + "grad_norm": 0.4386703073978424, + "learning_rate": 3.96178369888308e-05, + "loss": 1.1971, + "step": 9348 + }, + { + "epoch": 0.5680174980253965, + "grad_norm": 0.3666427731513977, + "learning_rate": 3.9608473048372444e-05, + "loss": 1.1057, + "step": 9349 + }, + { + "epoch": 0.568078255058023, + "grad_norm": 0.34049487113952637, + "learning_rate": 3.959910948881751e-05, + "loss": 1.0742, + "step": 9350 + }, + { + "epoch": 0.5681390120906495, + "grad_norm": 0.2525404691696167, + "learning_rate": 3.958974631050926e-05, + "loss": 1.0669, + "step": 9351 + }, + { + "epoch": 0.568199769123276, + "grad_norm": 0.19319120049476624, + "learning_rate": 3.9580383513790845e-05, + "loss": 1.056, + "step": 9352 + }, + { + "epoch": 0.5682605261559025, + "grad_norm": 0.1798635572195053, + "learning_rate": 3.9571021099005495e-05, + "loss": 1.0372, + "step": 9353 + }, + { + "epoch": 0.568321283188529, + "grad_norm": 0.18697211146354675, + "learning_rate": 3.956165906649637e-05, + "loss": 1.0928, + "step": 9354 + }, + { + "epoch": 0.5683820402211556, + "grad_norm": 0.46909356117248535, + "learning_rate": 3.955229741660665e-05, + "loss": 1.1336, + "step": 9355 + }, + { + "epoch": 0.5684427972537821, + "grad_norm": 0.3708920478820801, + "learning_rate": 3.95429361496795e-05, + "loss": 1.2131, + "step": 9356 + }, + { + "epoch": 0.5685035542864086, + "grad_norm": 0.17374636232852936, + "learning_rate": 3.953357526605803e-05, + "loss": 1.0214, + "step": 9357 + }, + { + "epoch": 0.5685643113190352, + "grad_norm": 0.19382081925868988, + "learning_rate": 3.952421476608542e-05, + "loss": 1.0515, + "step": 9358 + }, + { + "epoch": 0.5686250683516617, + "grad_norm": 0.229023739695549, + "learning_rate": 3.95148546501047e-05, + "loss": 1.1185, + "step": 9359 + }, + { + "epoch": 0.5686858253842882, + "grad_norm": 0.19368411600589752, + "learning_rate": 3.9505494918459034e-05, + "loss": 1.0514, + "step": 9360 + }, + { + "epoch": 0.5687465824169148, + "grad_norm": 0.17007040977478027, + "learning_rate": 3.9496135571491475e-05, + "loss": 1.0437, + "step": 9361 + }, + { + "epoch": 0.5688073394495413, + "grad_norm": 0.26762890815734863, + "learning_rate": 3.9486776609545086e-05, + "loss": 1.1159, + "step": 9362 + }, + { + "epoch": 0.5688680964821679, + "grad_norm": 0.2000059187412262, + "learning_rate": 3.947741803296294e-05, + "loss": 1.155, + "step": 9363 + }, + { + "epoch": 0.5689288535147944, + "grad_norm": 0.16094501316547394, + "learning_rate": 3.946805984208808e-05, + "loss": 1.0421, + "step": 9364 + }, + { + "epoch": 0.5689896105474208, + "grad_norm": 0.16497910022735596, + "learning_rate": 3.9458702037263515e-05, + "loss": 1.0166, + "step": 9365 + }, + { + "epoch": 0.5690503675800473, + "grad_norm": 0.16493213176727295, + "learning_rate": 3.9449344618832265e-05, + "loss": 1.0958, + "step": 9366 + }, + { + "epoch": 0.5691111246126739, + "grad_norm": 0.15834973752498627, + "learning_rate": 3.943998758713732e-05, + "loss": 1.0563, + "step": 9367 + }, + { + "epoch": 0.5691718816453004, + "grad_norm": 0.26308971643447876, + "learning_rate": 3.943063094252167e-05, + "loss": 1.2464, + "step": 9368 + }, + { + "epoch": 0.569232638677927, + "grad_norm": 0.18746700882911682, + "learning_rate": 3.94212746853283e-05, + "loss": 1.0543, + "step": 9369 + }, + { + "epoch": 0.5692933957105535, + "grad_norm": 1.9807474613189697, + "learning_rate": 3.9411918815900126e-05, + "loss": 1.088, + "step": 9370 + }, + { + "epoch": 0.56935415274318, + "grad_norm": 0.2528713345527649, + "learning_rate": 3.9402563334580155e-05, + "loss": 1.2444, + "step": 9371 + }, + { + "epoch": 0.5694149097758066, + "grad_norm": 0.1890864372253418, + "learning_rate": 3.939320824171124e-05, + "loss": 1.1033, + "step": 9372 + }, + { + "epoch": 0.5694756668084331, + "grad_norm": 0.35816794633865356, + "learning_rate": 3.938385353763634e-05, + "loss": 1.0452, + "step": 9373 + }, + { + "epoch": 0.5695364238410596, + "grad_norm": 0.1656225025653839, + "learning_rate": 3.937449922269834e-05, + "loss": 1.0127, + "step": 9374 + }, + { + "epoch": 0.5695971808736862, + "grad_norm": 0.5509020090103149, + "learning_rate": 3.9365145297240114e-05, + "loss": 1.1004, + "step": 9375 + }, + { + "epoch": 0.5696579379063127, + "grad_norm": 0.1663479208946228, + "learning_rate": 3.935579176160455e-05, + "loss": 1.1087, + "step": 9376 + }, + { + "epoch": 0.5697186949389392, + "grad_norm": 0.42133328318595886, + "learning_rate": 3.93464386161345e-05, + "loss": 1.1801, + "step": 9377 + }, + { + "epoch": 0.5697794519715657, + "grad_norm": 0.21763677895069122, + "learning_rate": 3.933708586117281e-05, + "loss": 0.9929, + "step": 9378 + }, + { + "epoch": 0.5698402090041922, + "grad_norm": 0.20823898911476135, + "learning_rate": 3.9327733497062294e-05, + "loss": 1.0443, + "step": 9379 + }, + { + "epoch": 0.5699009660368187, + "grad_norm": 0.2682357728481293, + "learning_rate": 3.931838152414575e-05, + "loss": 1.0882, + "step": 9380 + }, + { + "epoch": 0.5699617230694453, + "grad_norm": 4.591124534606934, + "learning_rate": 3.930902994276601e-05, + "loss": 1.0792, + "step": 9381 + }, + { + "epoch": 0.5700224801020718, + "grad_norm": 0.1722022145986557, + "learning_rate": 3.929967875326584e-05, + "loss": 1.0543, + "step": 9382 + }, + { + "epoch": 0.5700832371346983, + "grad_norm": 0.2967587411403656, + "learning_rate": 3.9290327955988014e-05, + "loss": 1.14, + "step": 9383 + }, + { + "epoch": 0.5701439941673249, + "grad_norm": 0.37682220339775085, + "learning_rate": 3.9280977551275296e-05, + "loss": 1.1673, + "step": 9384 + }, + { + "epoch": 0.5702047511999514, + "grad_norm": 0.24176017940044403, + "learning_rate": 3.9271627539470425e-05, + "loss": 1.0987, + "step": 9385 + }, + { + "epoch": 0.5702655082325779, + "grad_norm": 0.22621138393878937, + "learning_rate": 3.926227792091611e-05, + "loss": 1.0917, + "step": 9386 + }, + { + "epoch": 0.5703262652652045, + "grad_norm": 0.2588675320148468, + "learning_rate": 3.925292869595507e-05, + "loss": 1.0427, + "step": 9387 + }, + { + "epoch": 0.570387022297831, + "grad_norm": 0.19658410549163818, + "learning_rate": 3.9243579864929995e-05, + "loss": 1.0454, + "step": 9388 + }, + { + "epoch": 0.5704477793304575, + "grad_norm": 0.19282487034797668, + "learning_rate": 3.9234231428183594e-05, + "loss": 1.0758, + "step": 9389 + }, + { + "epoch": 0.5705085363630841, + "grad_norm": 0.2337033748626709, + "learning_rate": 3.922488338605852e-05, + "loss": 1.0466, + "step": 9390 + }, + { + "epoch": 0.5705692933957106, + "grad_norm": 0.1970319002866745, + "learning_rate": 3.9215535738897435e-05, + "loss": 1.0981, + "step": 9391 + }, + { + "epoch": 0.570630050428337, + "grad_norm": 0.282839298248291, + "learning_rate": 3.920618848704297e-05, + "loss": 1.1488, + "step": 9392 + }, + { + "epoch": 0.5706908074609636, + "grad_norm": 0.23851880431175232, + "learning_rate": 3.919684163083773e-05, + "loss": 1.1462, + "step": 9393 + }, + { + "epoch": 0.5707515644935901, + "grad_norm": 0.4173027575016022, + "learning_rate": 3.918749517062437e-05, + "loss": 1.2083, + "step": 9394 + }, + { + "epoch": 0.5708123215262166, + "grad_norm": 0.7515547275543213, + "learning_rate": 3.9178149106745454e-05, + "loss": 1.0593, + "step": 9395 + }, + { + "epoch": 0.5708730785588432, + "grad_norm": 0.24345393478870392, + "learning_rate": 3.9168803439543575e-05, + "loss": 1.044, + "step": 9396 + }, + { + "epoch": 0.5709338355914697, + "grad_norm": 0.20905762910842896, + "learning_rate": 3.91594581693613e-05, + "loss": 1.1368, + "step": 9397 + }, + { + "epoch": 0.5709945926240962, + "grad_norm": 0.25183889269828796, + "learning_rate": 3.91501132965412e-05, + "loss": 1.0476, + "step": 9398 + }, + { + "epoch": 0.5710553496567228, + "grad_norm": 0.28223717212677, + "learning_rate": 3.9140768821425775e-05, + "loss": 1.055, + "step": 9399 + }, + { + "epoch": 0.5711161066893493, + "grad_norm": 0.15855133533477783, + "learning_rate": 3.9131424744357574e-05, + "loss": 1.02, + "step": 9400 + }, + { + "epoch": 0.5711768637219758, + "grad_norm": 0.29612383246421814, + "learning_rate": 3.912208106567909e-05, + "loss": 1.1124, + "step": 9401 + }, + { + "epoch": 0.5712376207546024, + "grad_norm": 0.28322261571884155, + "learning_rate": 3.911273778573284e-05, + "loss": 1.017, + "step": 9402 + }, + { + "epoch": 0.5712983777872289, + "grad_norm": 0.24889986217021942, + "learning_rate": 3.910339490486128e-05, + "loss": 1.1211, + "step": 9403 + }, + { + "epoch": 0.5713591348198555, + "grad_norm": 0.2263394296169281, + "learning_rate": 3.9094052423406886e-05, + "loss": 1.0572, + "step": 9404 + }, + { + "epoch": 0.5714198918524819, + "grad_norm": 0.2520793676376343, + "learning_rate": 3.908471034171214e-05, + "loss": 1.0536, + "step": 9405 + }, + { + "epoch": 0.5714806488851084, + "grad_norm": 0.19945545494556427, + "learning_rate": 3.907536866011941e-05, + "loss": 1.099, + "step": 9406 + }, + { + "epoch": 0.571541405917735, + "grad_norm": 0.2662945091724396, + "learning_rate": 3.906602737897116e-05, + "loss": 1.0982, + "step": 9407 + }, + { + "epoch": 0.5716021629503615, + "grad_norm": 3.2412965297698975, + "learning_rate": 3.9056686498609803e-05, + "loss": 1.1985, + "step": 9408 + }, + { + "epoch": 0.571662919982988, + "grad_norm": 0.2522185742855072, + "learning_rate": 3.9047346019377695e-05, + "loss": 1.0721, + "step": 9409 + }, + { + "epoch": 0.5717236770156146, + "grad_norm": 0.21286186575889587, + "learning_rate": 3.903800594161725e-05, + "loss": 1.1048, + "step": 9410 + }, + { + "epoch": 0.5717844340482411, + "grad_norm": 0.19051653146743774, + "learning_rate": 3.9028666265670835e-05, + "loss": 1.0259, + "step": 9411 + }, + { + "epoch": 0.5718451910808676, + "grad_norm": 0.2275470495223999, + "learning_rate": 3.901932699188076e-05, + "loss": 1.0162, + "step": 9412 + }, + { + "epoch": 0.5719059481134942, + "grad_norm": 0.20308515429496765, + "learning_rate": 3.900998812058937e-05, + "loss": 1.004, + "step": 9413 + }, + { + "epoch": 0.5719667051461207, + "grad_norm": 0.1726303994655609, + "learning_rate": 3.900064965213899e-05, + "loss": 1.0506, + "step": 9414 + }, + { + "epoch": 0.5720274621787472, + "grad_norm": 0.17614808678627014, + "learning_rate": 3.8991311586871925e-05, + "loss": 1.1018, + "step": 9415 + }, + { + "epoch": 0.5720882192113738, + "grad_norm": 0.22601203620433807, + "learning_rate": 3.898197392513047e-05, + "loss": 1.0421, + "step": 9416 + }, + { + "epoch": 0.5721489762440003, + "grad_norm": 0.2628367841243744, + "learning_rate": 3.897263666725688e-05, + "loss": 1.0004, + "step": 9417 + }, + { + "epoch": 0.5722097332766267, + "grad_norm": 0.22797812521457672, + "learning_rate": 3.896329981359345e-05, + "loss": 1.0211, + "step": 9418 + }, + { + "epoch": 0.5722704903092533, + "grad_norm": 0.19651399552822113, + "learning_rate": 3.895396336448237e-05, + "loss": 1.0638, + "step": 9419 + }, + { + "epoch": 0.5723312473418798, + "grad_norm": 0.66497403383255, + "learning_rate": 3.894462732026591e-05, + "loss": 1.2284, + "step": 9420 + }, + { + "epoch": 0.5723920043745063, + "grad_norm": 0.2026027888059616, + "learning_rate": 3.893529168128627e-05, + "loss": 1.1192, + "step": 9421 + }, + { + "epoch": 0.5724527614071329, + "grad_norm": 0.491390585899353, + "learning_rate": 3.8925956447885645e-05, + "loss": 1.1345, + "step": 9422 + }, + { + "epoch": 0.5725135184397594, + "grad_norm": 0.21738183498382568, + "learning_rate": 3.8916621620406236e-05, + "loss": 1.0695, + "step": 9423 + }, + { + "epoch": 0.5725742754723859, + "grad_norm": 0.19323194026947021, + "learning_rate": 3.89072871991902e-05, + "loss": 1.0085, + "step": 9424 + }, + { + "epoch": 0.5726350325050125, + "grad_norm": 0.28661972284317017, + "learning_rate": 3.889795318457971e-05, + "loss": 1.2444, + "step": 9425 + }, + { + "epoch": 0.572695789537639, + "grad_norm": 0.2325149029493332, + "learning_rate": 3.888861957691688e-05, + "loss": 1.0553, + "step": 9426 + }, + { + "epoch": 0.5727565465702655, + "grad_norm": 0.22492030262947083, + "learning_rate": 3.8879286376543835e-05, + "loss": 1.1735, + "step": 9427 + }, + { + "epoch": 0.5728173036028921, + "grad_norm": 0.370980829000473, + "learning_rate": 3.886995358380272e-05, + "loss": 1.0696, + "step": 9428 + }, + { + "epoch": 0.5728780606355186, + "grad_norm": 0.2441726177930832, + "learning_rate": 3.886062119903559e-05, + "loss": 1.0709, + "step": 9429 + }, + { + "epoch": 0.5729388176681451, + "grad_norm": 0.4298327565193176, + "learning_rate": 3.885128922258454e-05, + "loss": 1.0172, + "step": 9430 + }, + { + "epoch": 0.5729995747007716, + "grad_norm": 0.33700448274612427, + "learning_rate": 3.884195765479166e-05, + "loss": 1.0773, + "step": 9431 + }, + { + "epoch": 0.5730603317333981, + "grad_norm": 0.8910164833068848, + "learning_rate": 3.8832626495998976e-05, + "loss": 1.0383, + "step": 9432 + }, + { + "epoch": 0.5731210887660246, + "grad_norm": 0.37752148509025574, + "learning_rate": 3.882329574654852e-05, + "loss": 1.0333, + "step": 9433 + }, + { + "epoch": 0.5731818457986512, + "grad_norm": 0.32926270365715027, + "learning_rate": 3.88139654067823e-05, + "loss": 1.1355, + "step": 9434 + }, + { + "epoch": 0.5732426028312777, + "grad_norm": 0.2624306082725525, + "learning_rate": 3.880463547704236e-05, + "loss": 1.1375, + "step": 9435 + }, + { + "epoch": 0.5733033598639042, + "grad_norm": 4.224170207977295, + "learning_rate": 3.879530595767067e-05, + "loss": 1.084, + "step": 9436 + }, + { + "epoch": 0.5733641168965308, + "grad_norm": 0.338049054145813, + "learning_rate": 3.878597684900921e-05, + "loss": 1.0815, + "step": 9437 + }, + { + "epoch": 0.5734248739291573, + "grad_norm": 0.19629283249378204, + "learning_rate": 3.877664815139994e-05, + "loss": 1.0886, + "step": 9438 + }, + { + "epoch": 0.5734856309617838, + "grad_norm": 1.5553627014160156, + "learning_rate": 3.8767319865184785e-05, + "loss": 1.0422, + "step": 9439 + }, + { + "epoch": 0.5735463879944104, + "grad_norm": 0.20345164835453033, + "learning_rate": 3.875799199070569e-05, + "loss": 1.0618, + "step": 9440 + }, + { + "epoch": 0.5736071450270369, + "grad_norm": 0.28878533840179443, + "learning_rate": 3.8748664528304585e-05, + "loss": 1.104, + "step": 9441 + }, + { + "epoch": 0.5736679020596634, + "grad_norm": 0.5350784659385681, + "learning_rate": 3.8739337478323346e-05, + "loss": 1.3909, + "step": 9442 + }, + { + "epoch": 0.57372865909229, + "grad_norm": 0.19057606160640717, + "learning_rate": 3.873001084110387e-05, + "loss": 1.0467, + "step": 9443 + }, + { + "epoch": 0.5737894161249164, + "grad_norm": 0.38166698813438416, + "learning_rate": 3.8720684616988035e-05, + "loss": 1.0939, + "step": 9444 + }, + { + "epoch": 0.5738501731575429, + "grad_norm": 0.2792218327522278, + "learning_rate": 3.871135880631769e-05, + "loss": 1.0951, + "step": 9445 + }, + { + "epoch": 0.5739109301901695, + "grad_norm": 0.467853844165802, + "learning_rate": 3.870203340943466e-05, + "loss": 1.0464, + "step": 9446 + }, + { + "epoch": 0.573971687222796, + "grad_norm": 0.29702049493789673, + "learning_rate": 3.869270842668077e-05, + "loss": 1.0243, + "step": 9447 + }, + { + "epoch": 0.5740324442554225, + "grad_norm": 0.23274128139019012, + "learning_rate": 3.868338385839785e-05, + "loss": 0.9763, + "step": 9448 + }, + { + "epoch": 0.5740932012880491, + "grad_norm": 0.14223021268844604, + "learning_rate": 3.867405970492767e-05, + "loss": 1.0302, + "step": 9449 + }, + { + "epoch": 0.5741539583206756, + "grad_norm": 0.17669309675693512, + "learning_rate": 3.866473596661202e-05, + "loss": 1.0613, + "step": 9450 + }, + { + "epoch": 0.5742147153533022, + "grad_norm": 0.2563076615333557, + "learning_rate": 3.865541264379268e-05, + "loss": 1.0056, + "step": 9451 + }, + { + "epoch": 0.5742754723859287, + "grad_norm": 0.20854590833187103, + "learning_rate": 3.864608973681139e-05, + "loss": 1.0919, + "step": 9452 + }, + { + "epoch": 0.5743362294185552, + "grad_norm": 0.1522424966096878, + "learning_rate": 3.863676724600986e-05, + "loss": 1.0682, + "step": 9453 + }, + { + "epoch": 0.5743969864511818, + "grad_norm": 0.22214928269386292, + "learning_rate": 3.862744517172982e-05, + "loss": 1.1354, + "step": 9454 + }, + { + "epoch": 0.5744577434838083, + "grad_norm": 0.271312952041626, + "learning_rate": 3.8618123514312976e-05, + "loss": 1.0331, + "step": 9455 + }, + { + "epoch": 0.5745185005164348, + "grad_norm": 0.16513663530349731, + "learning_rate": 3.860880227410102e-05, + "loss": 1.0453, + "step": 9456 + }, + { + "epoch": 0.5745792575490613, + "grad_norm": 0.19204393029212952, + "learning_rate": 3.859948145143562e-05, + "loss": 1.1045, + "step": 9457 + }, + { + "epoch": 0.5746400145816878, + "grad_norm": 0.15907074511051178, + "learning_rate": 3.859016104665844e-05, + "loss": 1.0467, + "step": 9458 + }, + { + "epoch": 0.5747007716143143, + "grad_norm": 0.1579115241765976, + "learning_rate": 3.85808410601111e-05, + "loss": 1.0647, + "step": 9459 + }, + { + "epoch": 0.5747615286469409, + "grad_norm": 0.23713088035583496, + "learning_rate": 3.8571521492135235e-05, + "loss": 1.0093, + "step": 9460 + }, + { + "epoch": 0.5748222856795674, + "grad_norm": 0.14569102227687836, + "learning_rate": 3.856220234307246e-05, + "loss": 0.9872, + "step": 9461 + }, + { + "epoch": 0.5748830427121939, + "grad_norm": 0.2717152237892151, + "learning_rate": 3.855288361326437e-05, + "loss": 1.0943, + "step": 9462 + }, + { + "epoch": 0.5749437997448205, + "grad_norm": 0.25329530239105225, + "learning_rate": 3.8543565303052545e-05, + "loss": 1.0627, + "step": 9463 + }, + { + "epoch": 0.575004556777447, + "grad_norm": 0.3086485266685486, + "learning_rate": 3.8534247412778545e-05, + "loss": 1.0727, + "step": 9464 + }, + { + "epoch": 0.5750653138100735, + "grad_norm": 0.4388085901737213, + "learning_rate": 3.852492994278394e-05, + "loss": 1.1258, + "step": 9465 + }, + { + "epoch": 0.5751260708427001, + "grad_norm": 0.4548232853412628, + "learning_rate": 3.851561289341023e-05, + "loss": 1.0766, + "step": 9466 + }, + { + "epoch": 0.5751868278753266, + "grad_norm": 0.17956282198429108, + "learning_rate": 3.850629626499895e-05, + "loss": 1.1115, + "step": 9467 + }, + { + "epoch": 0.5752475849079531, + "grad_norm": 0.23760104179382324, + "learning_rate": 3.84969800578916e-05, + "loss": 1.0182, + "step": 9468 + }, + { + "epoch": 0.5753083419405797, + "grad_norm": 0.21398918330669403, + "learning_rate": 3.848766427242967e-05, + "loss": 1.0857, + "step": 9469 + }, + { + "epoch": 0.5753690989732061, + "grad_norm": 0.35212039947509766, + "learning_rate": 3.847834890895464e-05, + "loss": 1.1686, + "step": 9470 + }, + { + "epoch": 0.5754298560058326, + "grad_norm": 0.18270991742610931, + "learning_rate": 3.8469033967807944e-05, + "loss": 1.1068, + "step": 9471 + }, + { + "epoch": 0.5754906130384592, + "grad_norm": 0.35226553678512573, + "learning_rate": 3.845971944933107e-05, + "loss": 1.1118, + "step": 9472 + }, + { + "epoch": 0.5755513700710857, + "grad_norm": 0.26989322900772095, + "learning_rate": 3.845040535386537e-05, + "loss": 1.1408, + "step": 9473 + }, + { + "epoch": 0.5756121271037122, + "grad_norm": 0.23112593591213226, + "learning_rate": 3.844109168175231e-05, + "loss": 1.0888, + "step": 9474 + }, + { + "epoch": 0.5756728841363388, + "grad_norm": 0.2957271933555603, + "learning_rate": 3.8431778433333274e-05, + "loss": 1.0571, + "step": 9475 + }, + { + "epoch": 0.5757336411689653, + "grad_norm": 0.35709816217422485, + "learning_rate": 3.842246560894962e-05, + "loss": 1.4099, + "step": 9476 + }, + { + "epoch": 0.5757943982015918, + "grad_norm": 0.2607353925704956, + "learning_rate": 3.841315320894274e-05, + "loss": 1.1942, + "step": 9477 + }, + { + "epoch": 0.5758551552342184, + "grad_norm": 0.25393474102020264, + "learning_rate": 3.840384123365397e-05, + "loss": 1.059, + "step": 9478 + }, + { + "epoch": 0.5759159122668449, + "grad_norm": 0.15938307344913483, + "learning_rate": 3.839452968342464e-05, + "loss": 1.0592, + "step": 9479 + }, + { + "epoch": 0.5759766692994714, + "grad_norm": 0.2824380397796631, + "learning_rate": 3.838521855859607e-05, + "loss": 1.131, + "step": 9480 + }, + { + "epoch": 0.576037426332098, + "grad_norm": 0.37770652770996094, + "learning_rate": 3.837590785950954e-05, + "loss": 1.2646, + "step": 9481 + }, + { + "epoch": 0.5760981833647245, + "grad_norm": 0.29892227053642273, + "learning_rate": 3.836659758650637e-05, + "loss": 1.0912, + "step": 9482 + }, + { + "epoch": 0.5761589403973509, + "grad_norm": 0.23888474702835083, + "learning_rate": 3.835728773992781e-05, + "loss": 1.1591, + "step": 9483 + }, + { + "epoch": 0.5762196974299775, + "grad_norm": 0.1620764285326004, + "learning_rate": 3.8347978320115116e-05, + "loss": 1.027, + "step": 9484 + }, + { + "epoch": 0.576280454462604, + "grad_norm": 0.18116234242916107, + "learning_rate": 3.8338669327409555e-05, + "loss": 1.0589, + "step": 9485 + }, + { + "epoch": 0.5763412114952305, + "grad_norm": 0.16242754459381104, + "learning_rate": 3.832936076215229e-05, + "loss": 1.0604, + "step": 9486 + }, + { + "epoch": 0.5764019685278571, + "grad_norm": 0.4309985637664795, + "learning_rate": 3.832005262468458e-05, + "loss": 1.0426, + "step": 9487 + }, + { + "epoch": 0.5764627255604836, + "grad_norm": 0.20062144100666046, + "learning_rate": 3.831074491534759e-05, + "loss": 1.071, + "step": 9488 + }, + { + "epoch": 0.5765234825931101, + "grad_norm": 0.19168207049369812, + "learning_rate": 3.83014376344825e-05, + "loss": 1.007, + "step": 9489 + }, + { + "epoch": 0.5765842396257367, + "grad_norm": 0.3013843894004822, + "learning_rate": 3.829213078243048e-05, + "loss": 1.184, + "step": 9490 + }, + { + "epoch": 0.5766449966583632, + "grad_norm": 0.1622961014509201, + "learning_rate": 3.828282435953267e-05, + "loss": 1.0542, + "step": 9491 + }, + { + "epoch": 0.5767057536909898, + "grad_norm": 0.1932172179222107, + "learning_rate": 3.827351836613021e-05, + "loss": 1.033, + "step": 9492 + }, + { + "epoch": 0.5767665107236163, + "grad_norm": 0.20858457684516907, + "learning_rate": 3.82642128025642e-05, + "loss": 1.0879, + "step": 9493 + }, + { + "epoch": 0.5768272677562428, + "grad_norm": 0.22788582742214203, + "learning_rate": 3.825490766917572e-05, + "loss": 1.0527, + "step": 9494 + }, + { + "epoch": 0.5768880247888694, + "grad_norm": 0.18508651852607727, + "learning_rate": 3.8245602966305876e-05, + "loss": 1.091, + "step": 9495 + }, + { + "epoch": 0.5769487818214959, + "grad_norm": 0.15247516334056854, + "learning_rate": 3.823629869429573e-05, + "loss": 1.0322, + "step": 9496 + }, + { + "epoch": 0.5770095388541223, + "grad_norm": 0.2967797815799713, + "learning_rate": 3.822699485348632e-05, + "loss": 1.0965, + "step": 9497 + }, + { + "epoch": 0.5770702958867489, + "grad_norm": 0.24553482234477997, + "learning_rate": 3.8217691444218725e-05, + "loss": 1.0676, + "step": 9498 + }, + { + "epoch": 0.5771310529193754, + "grad_norm": 0.25540921092033386, + "learning_rate": 3.8208388466833886e-05, + "loss": 1.1159, + "step": 9499 + }, + { + "epoch": 0.5771918099520019, + "grad_norm": 0.13839077949523926, + "learning_rate": 3.8199085921672864e-05, + "loss": 1.0276, + "step": 9500 + }, + { + "epoch": 0.5772525669846285, + "grad_norm": 0.17022272944450378, + "learning_rate": 3.818978380907663e-05, + "loss": 1.028, + "step": 9501 + }, + { + "epoch": 0.577313324017255, + "grad_norm": 0.23737899959087372, + "learning_rate": 3.818048212938614e-05, + "loss": 1.1206, + "step": 9502 + }, + { + "epoch": 0.5773740810498815, + "grad_norm": 0.1579214483499527, + "learning_rate": 3.817118088294237e-05, + "loss": 1.0553, + "step": 9503 + }, + { + "epoch": 0.5774348380825081, + "grad_norm": 0.1810448169708252, + "learning_rate": 3.8161880070086254e-05, + "loss": 1.0095, + "step": 9504 + }, + { + "epoch": 0.5774955951151346, + "grad_norm": 3.2540814876556396, + "learning_rate": 3.8152579691158726e-05, + "loss": 1.1686, + "step": 9505 + }, + { + "epoch": 0.5775563521477611, + "grad_norm": 0.1668989211320877, + "learning_rate": 3.814327974650067e-05, + "loss": 1.0186, + "step": 9506 + }, + { + "epoch": 0.5776171091803877, + "grad_norm": 0.3189834952354431, + "learning_rate": 3.813398023645297e-05, + "loss": 1.0309, + "step": 9507 + }, + { + "epoch": 0.5776778662130142, + "grad_norm": 0.15398141741752625, + "learning_rate": 3.812468116135653e-05, + "loss": 1.0804, + "step": 9508 + }, + { + "epoch": 0.5777386232456407, + "grad_norm": 0.19893692433834076, + "learning_rate": 3.81153825215522e-05, + "loss": 1.073, + "step": 9509 + }, + { + "epoch": 0.5777993802782672, + "grad_norm": 0.2759493291378021, + "learning_rate": 3.8106084317380806e-05, + "loss": 1.1294, + "step": 9510 + }, + { + "epoch": 0.5778601373108937, + "grad_norm": 0.2890907824039459, + "learning_rate": 3.80967865491832e-05, + "loss": 1.2188, + "step": 9511 + }, + { + "epoch": 0.5779208943435202, + "grad_norm": 0.13869395852088928, + "learning_rate": 3.8087489217300205e-05, + "loss": 1.0326, + "step": 9512 + }, + { + "epoch": 0.5779816513761468, + "grad_norm": 0.20128227770328522, + "learning_rate": 3.807819232207258e-05, + "loss": 1.0832, + "step": 9513 + }, + { + "epoch": 0.5780424084087733, + "grad_norm": 0.2938046455383301, + "learning_rate": 3.8068895863841114e-05, + "loss": 1.047, + "step": 9514 + }, + { + "epoch": 0.5781031654413998, + "grad_norm": 0.2997506558895111, + "learning_rate": 3.805959984294657e-05, + "loss": 1.0389, + "step": 9515 + }, + { + "epoch": 0.5781639224740264, + "grad_norm": 0.16949620842933655, + "learning_rate": 3.8050304259729705e-05, + "loss": 1.0718, + "step": 9516 + }, + { + "epoch": 0.5782246795066529, + "grad_norm": 0.36870646476745605, + "learning_rate": 3.804100911453126e-05, + "loss": 1.0549, + "step": 9517 + }, + { + "epoch": 0.5782854365392794, + "grad_norm": 4.605151176452637, + "learning_rate": 3.803171440769194e-05, + "loss": 1.0808, + "step": 9518 + }, + { + "epoch": 0.578346193571906, + "grad_norm": 0.19292248785495758, + "learning_rate": 3.802242013955243e-05, + "loss": 1.0715, + "step": 9519 + }, + { + "epoch": 0.5784069506045325, + "grad_norm": 0.16746164858341217, + "learning_rate": 3.8013126310453426e-05, + "loss": 1.0322, + "step": 9520 + }, + { + "epoch": 0.578467707637159, + "grad_norm": 0.29235774278640747, + "learning_rate": 3.800383292073559e-05, + "loss": 1.1147, + "step": 9521 + }, + { + "epoch": 0.5785284646697856, + "grad_norm": 0.19507966935634613, + "learning_rate": 3.799453997073959e-05, + "loss": 1.0958, + "step": 9522 + }, + { + "epoch": 0.578589221702412, + "grad_norm": 0.18736010789871216, + "learning_rate": 3.798524746080603e-05, + "loss": 1.0693, + "step": 9523 + }, + { + "epoch": 0.5786499787350385, + "grad_norm": 0.21031160652637482, + "learning_rate": 3.797595539127556e-05, + "loss": 1.1385, + "step": 9524 + }, + { + "epoch": 0.5787107357676651, + "grad_norm": 0.23980671167373657, + "learning_rate": 3.7966663762488775e-05, + "loss": 1.0462, + "step": 9525 + }, + { + "epoch": 0.5787714928002916, + "grad_norm": 0.23109614849090576, + "learning_rate": 3.795737257478625e-05, + "loss": 1.0792, + "step": 9526 + }, + { + "epoch": 0.5788322498329181, + "grad_norm": 0.15472590923309326, + "learning_rate": 3.794808182850856e-05, + "loss": 1.0852, + "step": 9527 + }, + { + "epoch": 0.5788930068655447, + "grad_norm": 0.14314022660255432, + "learning_rate": 3.793879152399625e-05, + "loss": 1.0433, + "step": 9528 + }, + { + "epoch": 0.5789537638981712, + "grad_norm": 0.44770875573158264, + "learning_rate": 3.7929501661589875e-05, + "loss": 1.0787, + "step": 9529 + }, + { + "epoch": 0.5790145209307977, + "grad_norm": 0.19374573230743408, + "learning_rate": 3.7920212241629954e-05, + "loss": 1.1091, + "step": 9530 + }, + { + "epoch": 0.5790752779634243, + "grad_norm": 0.15946152806282043, + "learning_rate": 3.7910923264456974e-05, + "loss": 1.0539, + "step": 9531 + }, + { + "epoch": 0.5791360349960508, + "grad_norm": 0.19574026763439178, + "learning_rate": 3.790163473041146e-05, + "loss": 0.9973, + "step": 9532 + }, + { + "epoch": 0.5791967920286774, + "grad_norm": 0.24165688455104828, + "learning_rate": 3.789234663983386e-05, + "loss": 1.078, + "step": 9533 + }, + { + "epoch": 0.5792575490613039, + "grad_norm": 0.5230681300163269, + "learning_rate": 3.788305899306463e-05, + "loss": 1.1782, + "step": 9534 + }, + { + "epoch": 0.5793183060939304, + "grad_norm": 0.35579901933670044, + "learning_rate": 3.7873771790444205e-05, + "loss": 1.044, + "step": 9535 + }, + { + "epoch": 0.5793790631265568, + "grad_norm": 0.25119584798812866, + "learning_rate": 3.786448503231302e-05, + "loss": 0.9992, + "step": 9536 + }, + { + "epoch": 0.5794398201591834, + "grad_norm": 4.793880462646484, + "learning_rate": 3.785519871901149e-05, + "loss": 1.12, + "step": 9537 + }, + { + "epoch": 0.5795005771918099, + "grad_norm": 0.13223712146282196, + "learning_rate": 3.784591285088e-05, + "loss": 1.0093, + "step": 9538 + }, + { + "epoch": 0.5795613342244365, + "grad_norm": 2.458625078201294, + "learning_rate": 3.783662742825893e-05, + "loss": 1.147, + "step": 9539 + }, + { + "epoch": 0.579622091257063, + "grad_norm": 0.1867521107196808, + "learning_rate": 3.7827342451488616e-05, + "loss": 1.0556, + "step": 9540 + }, + { + "epoch": 0.5796828482896895, + "grad_norm": 0.590216338634491, + "learning_rate": 3.7818057920909425e-05, + "loss": 1.0424, + "step": 9541 + }, + { + "epoch": 0.5797436053223161, + "grad_norm": 0.24295377731323242, + "learning_rate": 3.7808773836861676e-05, + "loss": 1.0783, + "step": 9542 + }, + { + "epoch": 0.5798043623549426, + "grad_norm": 1.3160758018493652, + "learning_rate": 3.7799490199685675e-05, + "loss": 1.0888, + "step": 9543 + }, + { + "epoch": 0.5798651193875691, + "grad_norm": 0.3098611533641815, + "learning_rate": 3.7790207009721724e-05, + "loss": 1.1051, + "step": 9544 + }, + { + "epoch": 0.5799258764201957, + "grad_norm": 0.1902393102645874, + "learning_rate": 3.77809242673101e-05, + "loss": 1.1022, + "step": 9545 + }, + { + "epoch": 0.5799866334528222, + "grad_norm": 0.20462064445018768, + "learning_rate": 3.7771641972791056e-05, + "loss": 1.0041, + "step": 9546 + }, + { + "epoch": 0.5800473904854487, + "grad_norm": 0.362136572599411, + "learning_rate": 3.776236012650484e-05, + "loss": 1.0463, + "step": 9547 + }, + { + "epoch": 0.5801081475180753, + "grad_norm": 0.5012286901473999, + "learning_rate": 3.7753078728791663e-05, + "loss": 1.0341, + "step": 9548 + }, + { + "epoch": 0.5801689045507017, + "grad_norm": 0.20635907351970673, + "learning_rate": 3.774379777999176e-05, + "loss": 1.1612, + "step": 9549 + }, + { + "epoch": 0.5802296615833282, + "grad_norm": 0.22217212617397308, + "learning_rate": 3.773451728044533e-05, + "loss": 1.0307, + "step": 9550 + }, + { + "epoch": 0.5802904186159548, + "grad_norm": 0.24988342821598053, + "learning_rate": 3.772523723049252e-05, + "loss": 1.0837, + "step": 9551 + }, + { + "epoch": 0.5803511756485813, + "grad_norm": 0.20271922647953033, + "learning_rate": 3.771595763047355e-05, + "loss": 1.1391, + "step": 9552 + }, + { + "epoch": 0.5804119326812078, + "grad_norm": 0.2066178172826767, + "learning_rate": 3.770667848072848e-05, + "loss": 1.0958, + "step": 9553 + }, + { + "epoch": 0.5804726897138344, + "grad_norm": 0.2927109897136688, + "learning_rate": 3.769739978159752e-05, + "loss": 1.1539, + "step": 9554 + }, + { + "epoch": 0.5805334467464609, + "grad_norm": 5.786839962005615, + "learning_rate": 3.768812153342074e-05, + "loss": 1.1534, + "step": 9555 + }, + { + "epoch": 0.5805942037790874, + "grad_norm": 0.2404061108827591, + "learning_rate": 3.7678843736538224e-05, + "loss": 1.0716, + "step": 9556 + }, + { + "epoch": 0.580654960811714, + "grad_norm": 0.3260616958141327, + "learning_rate": 3.76695663912901e-05, + "loss": 1.2228, + "step": 9557 + }, + { + "epoch": 0.5807157178443405, + "grad_norm": 0.6057564616203308, + "learning_rate": 3.7660289498016395e-05, + "loss": 1.0478, + "step": 9558 + }, + { + "epoch": 0.580776474876967, + "grad_norm": 0.18696148693561554, + "learning_rate": 3.7651013057057174e-05, + "loss": 1.1294, + "step": 9559 + }, + { + "epoch": 0.5808372319095936, + "grad_norm": 0.5084629654884338, + "learning_rate": 3.764173706875245e-05, + "loss": 1.2705, + "step": 9560 + }, + { + "epoch": 0.5808979889422201, + "grad_norm": 0.2301284372806549, + "learning_rate": 3.7632461533442223e-05, + "loss": 1.1544, + "step": 9561 + }, + { + "epoch": 0.5809587459748465, + "grad_norm": 0.3175599277019501, + "learning_rate": 3.7623186451466526e-05, + "loss": 1.1584, + "step": 9562 + }, + { + "epoch": 0.5810195030074731, + "grad_norm": 0.15435776114463806, + "learning_rate": 3.761391182316532e-05, + "loss": 1.0077, + "step": 9563 + }, + { + "epoch": 0.5810802600400996, + "grad_norm": 0.1965581327676773, + "learning_rate": 3.760463764887856e-05, + "loss": 1.0872, + "step": 9564 + }, + { + "epoch": 0.5811410170727261, + "grad_norm": 0.1429084837436676, + "learning_rate": 3.759536392894624e-05, + "loss": 1.0256, + "step": 9565 + }, + { + "epoch": 0.5812017741053527, + "grad_norm": 0.21005098521709442, + "learning_rate": 3.758609066370821e-05, + "loss": 1.1478, + "step": 9566 + }, + { + "epoch": 0.5812625311379792, + "grad_norm": 1.4452909231185913, + "learning_rate": 3.757681785350445e-05, + "loss": 1.1451, + "step": 9567 + }, + { + "epoch": 0.5813232881706057, + "grad_norm": 0.5599542260169983, + "learning_rate": 3.7567545498674824e-05, + "loss": 1.0341, + "step": 9568 + }, + { + "epoch": 0.5813840452032323, + "grad_norm": 0.42192867398262024, + "learning_rate": 3.7558273599559216e-05, + "loss": 1.1532, + "step": 9569 + }, + { + "epoch": 0.5814448022358588, + "grad_norm": 0.17911770939826965, + "learning_rate": 3.75490021564975e-05, + "loss": 1.0492, + "step": 9570 + }, + { + "epoch": 0.5815055592684854, + "grad_norm": 0.26514947414398193, + "learning_rate": 3.753973116982952e-05, + "loss": 1.0216, + "step": 9571 + }, + { + "epoch": 0.5815663163011119, + "grad_norm": 0.3300642967224121, + "learning_rate": 3.7530460639895106e-05, + "loss": 1.113, + "step": 9572 + }, + { + "epoch": 0.5816270733337384, + "grad_norm": 0.17882554233074188, + "learning_rate": 3.7521190567034065e-05, + "loss": 0.9791, + "step": 9573 + }, + { + "epoch": 0.581687830366365, + "grad_norm": 0.2927786111831665, + "learning_rate": 3.7511920951586185e-05, + "loss": 1.2508, + "step": 9574 + }, + { + "epoch": 0.5817485873989914, + "grad_norm": 1.1133081912994385, + "learning_rate": 3.750265179389126e-05, + "loss": 1.2559, + "step": 9575 + }, + { + "epoch": 0.5818093444316179, + "grad_norm": 1.1582667827606201, + "learning_rate": 3.7493383094289054e-05, + "loss": 1.0161, + "step": 9576 + }, + { + "epoch": 0.5818701014642444, + "grad_norm": 0.25841259956359863, + "learning_rate": 3.74841148531193e-05, + "loss": 1.0362, + "step": 9577 + }, + { + "epoch": 0.581930858496871, + "grad_norm": 0.22965793311595917, + "learning_rate": 3.747484707072174e-05, + "loss": 1.0114, + "step": 9578 + }, + { + "epoch": 0.5819916155294975, + "grad_norm": 0.31669101119041443, + "learning_rate": 3.7465579747436094e-05, + "loss": 1.2051, + "step": 9579 + }, + { + "epoch": 0.582052372562124, + "grad_norm": 0.3147105872631073, + "learning_rate": 3.745631288360204e-05, + "loss": 1.068, + "step": 9580 + }, + { + "epoch": 0.5821131295947506, + "grad_norm": 0.21404792368412018, + "learning_rate": 3.744704647955926e-05, + "loss": 1.0699, + "step": 9581 + }, + { + "epoch": 0.5821738866273771, + "grad_norm": 0.1704101413488388, + "learning_rate": 3.74377805356474e-05, + "loss": 1.05, + "step": 9582 + }, + { + "epoch": 0.5822346436600037, + "grad_norm": 1.9669069051742554, + "learning_rate": 3.7428515052206134e-05, + "loss": 1.1154, + "step": 9583 + }, + { + "epoch": 0.5822954006926302, + "grad_norm": 0.17890682816505432, + "learning_rate": 3.741925002957509e-05, + "loss": 1.0121, + "step": 9584 + }, + { + "epoch": 0.5823561577252567, + "grad_norm": 0.27350977063179016, + "learning_rate": 3.740998546809387e-05, + "loss": 1.0443, + "step": 9585 + }, + { + "epoch": 0.5824169147578833, + "grad_norm": 0.19773802161216736, + "learning_rate": 3.7400721368102054e-05, + "loss": 1.0614, + "step": 9586 + }, + { + "epoch": 0.5824776717905098, + "grad_norm": 0.2138923555612564, + "learning_rate": 3.7391457729939226e-05, + "loss": 1.0826, + "step": 9587 + }, + { + "epoch": 0.5825384288231362, + "grad_norm": 0.24452747404575348, + "learning_rate": 3.738219455394496e-05, + "loss": 1.0296, + "step": 9588 + }, + { + "epoch": 0.5825991858557628, + "grad_norm": 0.4696047008037567, + "learning_rate": 3.737293184045879e-05, + "loss": 1.1074, + "step": 9589 + }, + { + "epoch": 0.5826599428883893, + "grad_norm": 0.25147488713264465, + "learning_rate": 3.736366958982024e-05, + "loss": 1.1036, + "step": 9590 + }, + { + "epoch": 0.5827206999210158, + "grad_norm": 0.20051923394203186, + "learning_rate": 3.735440780236883e-05, + "loss": 1.1045, + "step": 9591 + }, + { + "epoch": 0.5827814569536424, + "grad_norm": 0.2920598089694977, + "learning_rate": 3.734514647844406e-05, + "loss": 1.0118, + "step": 9592 + }, + { + "epoch": 0.5828422139862689, + "grad_norm": 0.20527487993240356, + "learning_rate": 3.733588561838538e-05, + "loss": 1.1199, + "step": 9593 + }, + { + "epoch": 0.5829029710188954, + "grad_norm": 0.6568442583084106, + "learning_rate": 3.7326625222532256e-05, + "loss": 1.2707, + "step": 9594 + }, + { + "epoch": 0.582963728051522, + "grad_norm": 0.29720866680145264, + "learning_rate": 3.7317365291224124e-05, + "loss": 1.0931, + "step": 9595 + }, + { + "epoch": 0.5830244850841485, + "grad_norm": 0.1876746565103531, + "learning_rate": 3.7308105824800436e-05, + "loss": 1.0213, + "step": 9596 + }, + { + "epoch": 0.583085242116775, + "grad_norm": 0.20846034586429596, + "learning_rate": 3.729884682360058e-05, + "loss": 1.0613, + "step": 9597 + }, + { + "epoch": 0.5831459991494016, + "grad_norm": 0.20632073283195496, + "learning_rate": 3.728958828796394e-05, + "loss": 1.0926, + "step": 9598 + }, + { + "epoch": 0.5832067561820281, + "grad_norm": 0.24521151185035706, + "learning_rate": 3.728033021822992e-05, + "loss": 1.1114, + "step": 9599 + }, + { + "epoch": 0.5832675132146546, + "grad_norm": 0.435982346534729, + "learning_rate": 3.7271072614737834e-05, + "loss": 1.1257, + "step": 9600 + }, + { + "epoch": 0.5833282702472812, + "grad_norm": 0.24102437496185303, + "learning_rate": 3.7261815477827044e-05, + "loss": 1.0586, + "step": 9601 + }, + { + "epoch": 0.5833890272799076, + "grad_norm": 0.15575465559959412, + "learning_rate": 3.725255880783688e-05, + "loss": 1.0165, + "step": 9602 + }, + { + "epoch": 0.5834497843125341, + "grad_norm": 0.23034732043743134, + "learning_rate": 3.724330260510662e-05, + "loss": 1.1198, + "step": 9603 + }, + { + "epoch": 0.5835105413451607, + "grad_norm": 0.4199292063713074, + "learning_rate": 3.7234046869975575e-05, + "loss": 1.0487, + "step": 9604 + }, + { + "epoch": 0.5835712983777872, + "grad_norm": 0.24424058198928833, + "learning_rate": 3.722479160278301e-05, + "loss": 1.1255, + "step": 9605 + }, + { + "epoch": 0.5836320554104137, + "grad_norm": 0.41847699880599976, + "learning_rate": 3.7215536803868204e-05, + "loss": 1.0286, + "step": 9606 + }, + { + "epoch": 0.5836928124430403, + "grad_norm": 0.22996090352535248, + "learning_rate": 3.720628247357034e-05, + "loss": 1.0488, + "step": 9607 + }, + { + "epoch": 0.5837535694756668, + "grad_norm": 0.2643592953681946, + "learning_rate": 3.719702861222866e-05, + "loss": 0.9995, + "step": 9608 + }, + { + "epoch": 0.5838143265082933, + "grad_norm": 0.20129527151584625, + "learning_rate": 3.718777522018238e-05, + "loss": 1.0365, + "step": 9609 + }, + { + "epoch": 0.5838750835409199, + "grad_norm": 0.26840588450431824, + "learning_rate": 3.717852229777068e-05, + "loss": 1.065, + "step": 9610 + }, + { + "epoch": 0.5839358405735464, + "grad_norm": 0.3326706886291504, + "learning_rate": 3.7169269845332714e-05, + "loss": 1.0158, + "step": 9611 + }, + { + "epoch": 0.583996597606173, + "grad_norm": 2.5722506046295166, + "learning_rate": 3.716001786320766e-05, + "loss": 1.0495, + "step": 9612 + }, + { + "epoch": 0.5840573546387995, + "grad_norm": 0.2644788324832916, + "learning_rate": 3.7150766351734616e-05, + "loss": 1.0566, + "step": 9613 + }, + { + "epoch": 0.584118111671426, + "grad_norm": 0.3582479655742645, + "learning_rate": 3.7141515311252724e-05, + "loss": 1.0764, + "step": 9614 + }, + { + "epoch": 0.5841788687040524, + "grad_norm": 0.17954690754413605, + "learning_rate": 3.713226474210107e-05, + "loss": 1.065, + "step": 9615 + }, + { + "epoch": 0.584239625736679, + "grad_norm": 0.1953449696302414, + "learning_rate": 3.7123014644618726e-05, + "loss": 1.023, + "step": 9616 + }, + { + "epoch": 0.5843003827693055, + "grad_norm": 0.36411499977111816, + "learning_rate": 3.711376501914479e-05, + "loss": 1.1221, + "step": 9617 + }, + { + "epoch": 0.584361139801932, + "grad_norm": 0.1908355951309204, + "learning_rate": 3.7104515866018274e-05, + "loss": 1.1056, + "step": 9618 + }, + { + "epoch": 0.5844218968345586, + "grad_norm": 0.22664564847946167, + "learning_rate": 3.7095267185578244e-05, + "loss": 1.1395, + "step": 9619 + }, + { + "epoch": 0.5844826538671851, + "grad_norm": 0.4767666459083557, + "learning_rate": 3.708601897816367e-05, + "loss": 1.0731, + "step": 9620 + }, + { + "epoch": 0.5845434108998117, + "grad_norm": 0.2070866972208023, + "learning_rate": 3.7076771244113565e-05, + "loss": 1.1885, + "step": 9621 + }, + { + "epoch": 0.5846041679324382, + "grad_norm": 0.1383165568113327, + "learning_rate": 3.7067523983766915e-05, + "loss": 1.0633, + "step": 9622 + }, + { + "epoch": 0.5846649249650647, + "grad_norm": 0.2194654941558838, + "learning_rate": 3.705827719746267e-05, + "loss": 1.0532, + "step": 9623 + }, + { + "epoch": 0.5847256819976913, + "grad_norm": 1.0365620851516724, + "learning_rate": 3.704903088553977e-05, + "loss": 1.173, + "step": 9624 + }, + { + "epoch": 0.5847864390303178, + "grad_norm": 0.36042773723602295, + "learning_rate": 3.703978504833716e-05, + "loss": 1.2617, + "step": 9625 + }, + { + "epoch": 0.5848471960629443, + "grad_norm": 0.18460413813591003, + "learning_rate": 3.7030539686193734e-05, + "loss": 1.055, + "step": 9626 + }, + { + "epoch": 0.5849079530955709, + "grad_norm": 0.27213168144226074, + "learning_rate": 3.702129479944838e-05, + "loss": 1.0339, + "step": 9627 + }, + { + "epoch": 0.5849687101281973, + "grad_norm": 0.22336578369140625, + "learning_rate": 3.7012050388439966e-05, + "loss": 1.3019, + "step": 9628 + }, + { + "epoch": 0.5850294671608238, + "grad_norm": 0.16395431756973267, + "learning_rate": 3.7002806453507356e-05, + "loss": 1.0724, + "step": 9629 + }, + { + "epoch": 0.5850902241934504, + "grad_norm": 0.3815798759460449, + "learning_rate": 3.69935629949894e-05, + "loss": 1.0391, + "step": 9630 + }, + { + "epoch": 0.5851509812260769, + "grad_norm": 0.20346233248710632, + "learning_rate": 3.6984320013224894e-05, + "loss": 1.0193, + "step": 9631 + }, + { + "epoch": 0.5852117382587034, + "grad_norm": 0.19071389734745026, + "learning_rate": 3.697507750855267e-05, + "loss": 1.0532, + "step": 9632 + }, + { + "epoch": 0.58527249529133, + "grad_norm": 0.2737258970737457, + "learning_rate": 3.696583548131147e-05, + "loss": 1.2188, + "step": 9633 + }, + { + "epoch": 0.5853332523239565, + "grad_norm": 0.21068082749843597, + "learning_rate": 3.6956593931840114e-05, + "loss": 1.0273, + "step": 9634 + }, + { + "epoch": 0.585394009356583, + "grad_norm": 0.185922771692276, + "learning_rate": 3.694735286047731e-05, + "loss": 1.0466, + "step": 9635 + }, + { + "epoch": 0.5854547663892096, + "grad_norm": 0.1365990936756134, + "learning_rate": 3.693811226756182e-05, + "loss": 1.078, + "step": 9636 + }, + { + "epoch": 0.5855155234218361, + "grad_norm": 4.549467086791992, + "learning_rate": 3.692887215343235e-05, + "loss": 1.1415, + "step": 9637 + }, + { + "epoch": 0.5855762804544626, + "grad_norm": 0.20873530209064484, + "learning_rate": 3.691963251842759e-05, + "loss": 1.0117, + "step": 9638 + }, + { + "epoch": 0.5856370374870892, + "grad_norm": 0.22083909809589386, + "learning_rate": 3.6910393362886254e-05, + "loss": 1.0331, + "step": 9639 + }, + { + "epoch": 0.5856977945197157, + "grad_norm": 0.16727912425994873, + "learning_rate": 3.690115468714697e-05, + "loss": 1.0481, + "step": 9640 + }, + { + "epoch": 0.5857585515523421, + "grad_norm": 0.2531777620315552, + "learning_rate": 3.6891916491548366e-05, + "loss": 1.1882, + "step": 9641 + }, + { + "epoch": 0.5858193085849687, + "grad_norm": 0.47085076570510864, + "learning_rate": 3.688267877642912e-05, + "loss": 1.2304, + "step": 9642 + }, + { + "epoch": 0.5858800656175952, + "grad_norm": 0.23334349691867828, + "learning_rate": 3.687344154212781e-05, + "loss": 1.1255, + "step": 9643 + }, + { + "epoch": 0.5859408226502217, + "grad_norm": 0.2757597863674164, + "learning_rate": 3.686420478898304e-05, + "loss": 1.0738, + "step": 9644 + }, + { + "epoch": 0.5860015796828483, + "grad_norm": 0.20217609405517578, + "learning_rate": 3.685496851733338e-05, + "loss": 1.0769, + "step": 9645 + }, + { + "epoch": 0.5860623367154748, + "grad_norm": 0.3908464312553406, + "learning_rate": 3.6845732727517404e-05, + "loss": 1.0587, + "step": 9646 + }, + { + "epoch": 0.5861230937481013, + "grad_norm": 0.17710594832897186, + "learning_rate": 3.6836497419873626e-05, + "loss": 1.079, + "step": 9647 + }, + { + "epoch": 0.5861838507807279, + "grad_norm": 0.21035638451576233, + "learning_rate": 3.682726259474059e-05, + "loss": 1.0927, + "step": 9648 + }, + { + "epoch": 0.5862446078133544, + "grad_norm": 0.15505968034267426, + "learning_rate": 3.681802825245677e-05, + "loss": 1.0442, + "step": 9649 + }, + { + "epoch": 0.586305364845981, + "grad_norm": 0.3732808232307434, + "learning_rate": 3.680879439336069e-05, + "loss": 1.3353, + "step": 9650 + }, + { + "epoch": 0.5863661218786075, + "grad_norm": 2.652304172515869, + "learning_rate": 3.679956101779079e-05, + "loss": 1.2425, + "step": 9651 + }, + { + "epoch": 0.586426878911234, + "grad_norm": 0.21164259314537048, + "learning_rate": 3.679032812608555e-05, + "loss": 1.0172, + "step": 9652 + }, + { + "epoch": 0.5864876359438606, + "grad_norm": 0.1677429974079132, + "learning_rate": 3.678109571858337e-05, + "loss": 1.0891, + "step": 9653 + }, + { + "epoch": 0.586548392976487, + "grad_norm": 1.0468740463256836, + "learning_rate": 3.6771863795622674e-05, + "loss": 1.0293, + "step": 9654 + }, + { + "epoch": 0.5866091500091135, + "grad_norm": 9.567384719848633, + "learning_rate": 3.6762632357541884e-05, + "loss": 1.0653, + "step": 9655 + }, + { + "epoch": 0.58666990704174, + "grad_norm": 0.1603352278470993, + "learning_rate": 3.675340140467936e-05, + "loss": 1.032, + "step": 9656 + }, + { + "epoch": 0.5867306640743666, + "grad_norm": 1.9250242710113525, + "learning_rate": 3.6744170937373454e-05, + "loss": 1.0796, + "step": 9657 + }, + { + "epoch": 0.5867914211069931, + "grad_norm": 0.24214816093444824, + "learning_rate": 3.673494095596254e-05, + "loss": 1.1173, + "step": 9658 + }, + { + "epoch": 0.5868521781396197, + "grad_norm": 0.25979599356651306, + "learning_rate": 3.672571146078494e-05, + "loss": 1.1002, + "step": 9659 + }, + { + "epoch": 0.5869129351722462, + "grad_norm": 0.4308815896511078, + "learning_rate": 3.6716482452178934e-05, + "loss": 1.1845, + "step": 9660 + }, + { + "epoch": 0.5869736922048727, + "grad_norm": 0.470042884349823, + "learning_rate": 3.6707253930482835e-05, + "loss": 1.1311, + "step": 9661 + }, + { + "epoch": 0.5870344492374993, + "grad_norm": 0.19202539324760437, + "learning_rate": 3.66980258960349e-05, + "loss": 1.0794, + "step": 9662 + }, + { + "epoch": 0.5870952062701258, + "grad_norm": 0.1771136373281479, + "learning_rate": 3.6688798349173416e-05, + "loss": 1.04, + "step": 9663 + }, + { + "epoch": 0.5871559633027523, + "grad_norm": 0.2136196345090866, + "learning_rate": 3.66795712902366e-05, + "loss": 1.0826, + "step": 9664 + }, + { + "epoch": 0.5872167203353789, + "grad_norm": 0.25034767389297485, + "learning_rate": 3.6670344719562655e-05, + "loss": 1.2034, + "step": 9665 + }, + { + "epoch": 0.5872774773680054, + "grad_norm": 0.16104166209697723, + "learning_rate": 3.666111863748984e-05, + "loss": 1.0058, + "step": 9666 + }, + { + "epoch": 0.5873382344006318, + "grad_norm": 0.2443307489156723, + "learning_rate": 3.6651893044356264e-05, + "loss": 1.135, + "step": 9667 + }, + { + "epoch": 0.5873989914332584, + "grad_norm": 0.34711527824401855, + "learning_rate": 3.664266794050014e-05, + "loss": 1.0312, + "step": 9668 + }, + { + "epoch": 0.5874597484658849, + "grad_norm": 0.20483854413032532, + "learning_rate": 3.66334433262596e-05, + "loss": 1.0968, + "step": 9669 + }, + { + "epoch": 0.5875205054985114, + "grad_norm": 0.17550280690193176, + "learning_rate": 3.6624219201972776e-05, + "loss": 1.1139, + "step": 9670 + }, + { + "epoch": 0.587581262531138, + "grad_norm": 0.17304740846157074, + "learning_rate": 3.661499556797779e-05, + "loss": 1.0997, + "step": 9671 + }, + { + "epoch": 0.5876420195637645, + "grad_norm": 0.20152659714221954, + "learning_rate": 3.660577242461273e-05, + "loss": 1.0631, + "step": 9672 + }, + { + "epoch": 0.587702776596391, + "grad_norm": 0.22784605622291565, + "learning_rate": 3.6596549772215674e-05, + "loss": 1.0913, + "step": 9673 + }, + { + "epoch": 0.5877635336290176, + "grad_norm": 0.24763886630535126, + "learning_rate": 3.658732761112468e-05, + "loss": 1.1345, + "step": 9674 + }, + { + "epoch": 0.5878242906616441, + "grad_norm": 0.2244214564561844, + "learning_rate": 3.657810594167777e-05, + "loss": 1.0527, + "step": 9675 + }, + { + "epoch": 0.5878850476942706, + "grad_norm": 0.15864059329032898, + "learning_rate": 3.656888476421298e-05, + "loss": 1.0469, + "step": 9676 + }, + { + "epoch": 0.5879458047268972, + "grad_norm": 0.33790087699890137, + "learning_rate": 3.655966407906832e-05, + "loss": 1.0731, + "step": 9677 + }, + { + "epoch": 0.5880065617595237, + "grad_norm": 0.18550872802734375, + "learning_rate": 3.655044388658176e-05, + "loss": 1.0866, + "step": 9678 + }, + { + "epoch": 0.5880673187921502, + "grad_norm": 0.2145218551158905, + "learning_rate": 3.6541224187091324e-05, + "loss": 1.1261, + "step": 9679 + }, + { + "epoch": 0.5881280758247767, + "grad_norm": 0.22317452728748322, + "learning_rate": 3.6532004980934876e-05, + "loss": 1.0891, + "step": 9680 + }, + { + "epoch": 0.5881888328574032, + "grad_norm": 0.16283102333545685, + "learning_rate": 3.65227862684504e-05, + "loss": 1.0473, + "step": 9681 + }, + { + "epoch": 0.5882495898900297, + "grad_norm": 0.4891977906227112, + "learning_rate": 3.65135680499758e-05, + "loss": 1.0656, + "step": 9682 + }, + { + "epoch": 0.5883103469226563, + "grad_norm": 0.24216073751449585, + "learning_rate": 3.650435032584895e-05, + "loss": 1.1311, + "step": 9683 + }, + { + "epoch": 0.5883711039552828, + "grad_norm": 0.22788946330547333, + "learning_rate": 3.649513309640777e-05, + "loss": 1.1275, + "step": 9684 + }, + { + "epoch": 0.5884318609879093, + "grad_norm": 0.21033041179180145, + "learning_rate": 3.6485916361990094e-05, + "loss": 1.0822, + "step": 9685 + }, + { + "epoch": 0.5884926180205359, + "grad_norm": 0.2601049840450287, + "learning_rate": 3.647670012293377e-05, + "loss": 1.2399, + "step": 9686 + }, + { + "epoch": 0.5885533750531624, + "grad_norm": 0.19047017395496368, + "learning_rate": 3.646748437957661e-05, + "loss": 1.0391, + "step": 9687 + }, + { + "epoch": 0.5886141320857889, + "grad_norm": 0.1761654168367386, + "learning_rate": 3.645826913225642e-05, + "loss": 1.0459, + "step": 9688 + }, + { + "epoch": 0.5886748891184155, + "grad_norm": 0.9378427267074585, + "learning_rate": 3.6449054381311e-05, + "loss": 1.1059, + "step": 9689 + }, + { + "epoch": 0.588735646151042, + "grad_norm": 0.25633475184440613, + "learning_rate": 3.643984012707811e-05, + "loss": 1.133, + "step": 9690 + }, + { + "epoch": 0.5887964031836685, + "grad_norm": 5.662492275238037, + "learning_rate": 3.643062636989549e-05, + "loss": 1.0904, + "step": 9691 + }, + { + "epoch": 0.5888571602162951, + "grad_norm": 0.32295215129852295, + "learning_rate": 3.642141311010089e-05, + "loss": 1.0608, + "step": 9692 + }, + { + "epoch": 0.5889179172489215, + "grad_norm": 0.19396977126598358, + "learning_rate": 3.641220034803203e-05, + "loss": 1.0727, + "step": 9693 + }, + { + "epoch": 0.588978674281548, + "grad_norm": 0.18688930571079254, + "learning_rate": 3.640298808402659e-05, + "loss": 1.0382, + "step": 9694 + }, + { + "epoch": 0.5890394313141746, + "grad_norm": 0.18331369757652283, + "learning_rate": 3.639377631842225e-05, + "loss": 1.0034, + "step": 9695 + }, + { + "epoch": 0.5891001883468011, + "grad_norm": 0.20976726710796356, + "learning_rate": 3.6384565051556654e-05, + "loss": 1.0615, + "step": 9696 + }, + { + "epoch": 0.5891609453794276, + "grad_norm": 0.22687575221061707, + "learning_rate": 3.637535428376746e-05, + "loss": 1.1502, + "step": 9697 + }, + { + "epoch": 0.5892217024120542, + "grad_norm": 0.2188509851694107, + "learning_rate": 3.63661440153923e-05, + "loss": 1.0799, + "step": 9698 + }, + { + "epoch": 0.5892824594446807, + "grad_norm": 0.18039798736572266, + "learning_rate": 3.635693424676877e-05, + "loss": 1.0644, + "step": 9699 + }, + { + "epoch": 0.5893432164773073, + "grad_norm": 1.2981153726577759, + "learning_rate": 3.6347724978234444e-05, + "loss": 1.212, + "step": 9700 + }, + { + "epoch": 0.5894039735099338, + "grad_norm": 0.2115543782711029, + "learning_rate": 3.633851621012689e-05, + "loss": 1.0257, + "step": 9701 + }, + { + "epoch": 0.5894647305425603, + "grad_norm": 0.2282436639070511, + "learning_rate": 3.6329307942783676e-05, + "loss": 1.0515, + "step": 9702 + }, + { + "epoch": 0.5895254875751869, + "grad_norm": 0.26386338472366333, + "learning_rate": 3.632010017654232e-05, + "loss": 1.1066, + "step": 9703 + }, + { + "epoch": 0.5895862446078134, + "grad_norm": 0.214959517121315, + "learning_rate": 3.6310892911740325e-05, + "loss": 1.1442, + "step": 9704 + }, + { + "epoch": 0.5896470016404399, + "grad_norm": 0.1808650940656662, + "learning_rate": 3.63016861487152e-05, + "loss": 1.0621, + "step": 9705 + }, + { + "epoch": 0.5897077586730665, + "grad_norm": 0.20109710097312927, + "learning_rate": 3.629247988780444e-05, + "loss": 1.0266, + "step": 9706 + }, + { + "epoch": 0.5897685157056929, + "grad_norm": 0.1887139081954956, + "learning_rate": 3.628327412934546e-05, + "loss": 1.096, + "step": 9707 + }, + { + "epoch": 0.5898292727383194, + "grad_norm": 0.18158984184265137, + "learning_rate": 3.627406887367572e-05, + "loss": 1.0714, + "step": 9708 + }, + { + "epoch": 0.589890029770946, + "grad_norm": 0.2741205394268036, + "learning_rate": 3.626486412113264e-05, + "loss": 1.1329, + "step": 9709 + }, + { + "epoch": 0.5899507868035725, + "grad_norm": 0.1785549372434616, + "learning_rate": 3.6255659872053616e-05, + "loss": 1.0206, + "step": 9710 + }, + { + "epoch": 0.590011543836199, + "grad_norm": 0.18007971346378326, + "learning_rate": 3.6246456126776044e-05, + "loss": 1.2206, + "step": 9711 + }, + { + "epoch": 0.5900723008688256, + "grad_norm": 0.5703466534614563, + "learning_rate": 3.623725288563726e-05, + "loss": 1.159, + "step": 9712 + }, + { + "epoch": 0.5901330579014521, + "grad_norm": 0.1816219538450241, + "learning_rate": 3.6228050148974674e-05, + "loss": 1.0683, + "step": 9713 + }, + { + "epoch": 0.5901938149340786, + "grad_norm": 0.17640164494514465, + "learning_rate": 3.6218847917125535e-05, + "loss": 1.0357, + "step": 9714 + }, + { + "epoch": 0.5902545719667052, + "grad_norm": 0.2541182041168213, + "learning_rate": 3.62096461904272e-05, + "loss": 1.1345, + "step": 9715 + }, + { + "epoch": 0.5903153289993317, + "grad_norm": 0.29778921604156494, + "learning_rate": 3.620044496921695e-05, + "loss": 1.2558, + "step": 9716 + }, + { + "epoch": 0.5903760860319582, + "grad_norm": 0.8852308988571167, + "learning_rate": 3.619124425383206e-05, + "loss": 1.0309, + "step": 9717 + }, + { + "epoch": 0.5904368430645848, + "grad_norm": 0.17884086072444916, + "learning_rate": 3.618204404460978e-05, + "loss": 1.1617, + "step": 9718 + }, + { + "epoch": 0.5904976000972113, + "grad_norm": 0.24508096277713776, + "learning_rate": 3.617284434188735e-05, + "loss": 1.1106, + "step": 9719 + }, + { + "epoch": 0.5905583571298377, + "grad_norm": 4.022109031677246, + "learning_rate": 3.616364514600198e-05, + "loss": 1.0549, + "step": 9720 + }, + { + "epoch": 0.5906191141624643, + "grad_norm": 0.24402497708797455, + "learning_rate": 3.615444645729087e-05, + "loss": 1.1215, + "step": 9721 + }, + { + "epoch": 0.5906798711950908, + "grad_norm": 0.1783686876296997, + "learning_rate": 3.6145248276091194e-05, + "loss": 1.0745, + "step": 9722 + }, + { + "epoch": 0.5907406282277173, + "grad_norm": 0.20226113498210907, + "learning_rate": 3.613605060274013e-05, + "loss": 1.0445, + "step": 9723 + }, + { + "epoch": 0.5908013852603439, + "grad_norm": 0.22913703322410583, + "learning_rate": 3.612685343757481e-05, + "loss": 1.0653, + "step": 9724 + }, + { + "epoch": 0.5908621422929704, + "grad_norm": 1.4038991928100586, + "learning_rate": 3.611765678093235e-05, + "loss": 1.0415, + "step": 9725 + }, + { + "epoch": 0.5909228993255969, + "grad_norm": 0.32284995913505554, + "learning_rate": 3.610846063314988e-05, + "loss": 1.1034, + "step": 9726 + }, + { + "epoch": 0.5909836563582235, + "grad_norm": 0.33259737491607666, + "learning_rate": 3.609926499456446e-05, + "loss": 1.0633, + "step": 9727 + }, + { + "epoch": 0.59104441339085, + "grad_norm": 0.1786070168018341, + "learning_rate": 3.609006986551318e-05, + "loss": 1.0221, + "step": 9728 + }, + { + "epoch": 0.5911051704234765, + "grad_norm": 0.14675813913345337, + "learning_rate": 3.6080875246333065e-05, + "loss": 1.0233, + "step": 9729 + }, + { + "epoch": 0.5911659274561031, + "grad_norm": 1.6613141298294067, + "learning_rate": 3.6071681137361155e-05, + "loss": 1.0786, + "step": 9730 + }, + { + "epoch": 0.5912266844887296, + "grad_norm": 0.2805497646331787, + "learning_rate": 3.606248753893448e-05, + "loss": 1.0505, + "step": 9731 + }, + { + "epoch": 0.5912874415213561, + "grad_norm": 0.2032736837863922, + "learning_rate": 3.605329445139001e-05, + "loss": 1.0424, + "step": 9732 + }, + { + "epoch": 0.5913481985539826, + "grad_norm": 0.19396735727787018, + "learning_rate": 3.604410187506474e-05, + "loss": 1.0479, + "step": 9733 + }, + { + "epoch": 0.5914089555866091, + "grad_norm": 0.18111427128314972, + "learning_rate": 3.603490981029559e-05, + "loss": 1.0643, + "step": 9734 + }, + { + "epoch": 0.5914697126192356, + "grad_norm": 0.33942630887031555, + "learning_rate": 3.602571825741953e-05, + "loss": 1.0349, + "step": 9735 + }, + { + "epoch": 0.5915304696518622, + "grad_norm": 0.223747119307518, + "learning_rate": 3.6016527216773475e-05, + "loss": 1.0524, + "step": 9736 + }, + { + "epoch": 0.5915912266844887, + "grad_norm": 0.14011640846729279, + "learning_rate": 3.60073366886943e-05, + "loss": 1.0021, + "step": 9737 + }, + { + "epoch": 0.5916519837171152, + "grad_norm": 0.6361732482910156, + "learning_rate": 3.5998146673518915e-05, + "loss": 1.085, + "step": 9738 + }, + { + "epoch": 0.5917127407497418, + "grad_norm": 0.23272116482257843, + "learning_rate": 3.598895717158418e-05, + "loss": 1.0848, + "step": 9739 + }, + { + "epoch": 0.5917734977823683, + "grad_norm": 0.24441948533058167, + "learning_rate": 3.597976818322694e-05, + "loss": 1.1118, + "step": 9740 + }, + { + "epoch": 0.5918342548149949, + "grad_norm": 0.15202000737190247, + "learning_rate": 3.597057970878398e-05, + "loss": 1.0348, + "step": 9741 + }, + { + "epoch": 0.5918950118476214, + "grad_norm": 0.2850262224674225, + "learning_rate": 3.596139174859213e-05, + "loss": 1.0439, + "step": 9742 + }, + { + "epoch": 0.5919557688802479, + "grad_norm": 0.2803071439266205, + "learning_rate": 3.5952204302988194e-05, + "loss": 1.0507, + "step": 9743 + }, + { + "epoch": 0.5920165259128745, + "grad_norm": 0.2175857424736023, + "learning_rate": 3.594301737230892e-05, + "loss": 1.0637, + "step": 9744 + }, + { + "epoch": 0.592077282945501, + "grad_norm": 0.2298722267150879, + "learning_rate": 3.5933830956891054e-05, + "loss": 1.0501, + "step": 9745 + }, + { + "epoch": 0.5921380399781274, + "grad_norm": 0.5449656844139099, + "learning_rate": 3.592464505707136e-05, + "loss": 1.0485, + "step": 9746 + }, + { + "epoch": 0.592198797010754, + "grad_norm": 1.9638288021087646, + "learning_rate": 3.591545967318649e-05, + "loss": 1.0878, + "step": 9747 + }, + { + "epoch": 0.5922595540433805, + "grad_norm": 0.19098076224327087, + "learning_rate": 3.590627480557318e-05, + "loss": 1.0059, + "step": 9748 + }, + { + "epoch": 0.592320311076007, + "grad_norm": 0.15434515476226807, + "learning_rate": 3.5897090454568097e-05, + "loss": 1.0258, + "step": 9749 + }, + { + "epoch": 0.5923810681086336, + "grad_norm": 0.23777075111865997, + "learning_rate": 3.588790662050788e-05, + "loss": 1.0652, + "step": 9750 + }, + { + "epoch": 0.5924418251412601, + "grad_norm": 0.1457141786813736, + "learning_rate": 3.587872330372918e-05, + "loss": 1.0238, + "step": 9751 + }, + { + "epoch": 0.5925025821738866, + "grad_norm": 0.13772332668304443, + "learning_rate": 3.58695405045686e-05, + "loss": 0.9943, + "step": 9752 + }, + { + "epoch": 0.5925633392065132, + "grad_norm": 0.2583719491958618, + "learning_rate": 3.586035822336277e-05, + "loss": 1.0347, + "step": 9753 + }, + { + "epoch": 0.5926240962391397, + "grad_norm": 0.1442100554704666, + "learning_rate": 3.5851176460448216e-05, + "loss": 1.0294, + "step": 9754 + }, + { + "epoch": 0.5926848532717662, + "grad_norm": 0.15986347198486328, + "learning_rate": 3.5841995216161514e-05, + "loss": 1.1018, + "step": 9755 + }, + { + "epoch": 0.5927456103043928, + "grad_norm": 0.15791216492652893, + "learning_rate": 3.583281449083924e-05, + "loss": 1.0214, + "step": 9756 + }, + { + "epoch": 0.5928063673370193, + "grad_norm": 0.862037718296051, + "learning_rate": 3.582363428481788e-05, + "loss": 1.0694, + "step": 9757 + }, + { + "epoch": 0.5928671243696458, + "grad_norm": 0.7882353067398071, + "learning_rate": 3.5814454598433936e-05, + "loss": 1.2797, + "step": 9758 + }, + { + "epoch": 0.5929278814022723, + "grad_norm": 0.3240058124065399, + "learning_rate": 3.58052754320239e-05, + "loss": 1.0054, + "step": 9759 + }, + { + "epoch": 0.5929886384348988, + "grad_norm": 0.17563392221927643, + "learning_rate": 3.579609678592426e-05, + "loss": 1.0269, + "step": 9760 + }, + { + "epoch": 0.5930493954675253, + "grad_norm": 0.16642968356609344, + "learning_rate": 3.5786918660471424e-05, + "loss": 1.0761, + "step": 9761 + }, + { + "epoch": 0.5931101525001519, + "grad_norm": 1.1524062156677246, + "learning_rate": 3.5777741056001834e-05, + "loss": 1.0254, + "step": 9762 + }, + { + "epoch": 0.5931709095327784, + "grad_norm": 0.14050987362861633, + "learning_rate": 3.5768563972851886e-05, + "loss": 1.0458, + "step": 9763 + }, + { + "epoch": 0.5932316665654049, + "grad_norm": 0.23115402460098267, + "learning_rate": 3.575938741135798e-05, + "loss": 1.0947, + "step": 9764 + }, + { + "epoch": 0.5932924235980315, + "grad_norm": 0.3403030037879944, + "learning_rate": 3.575021137185649e-05, + "loss": 1.0796, + "step": 9765 + }, + { + "epoch": 0.593353180630658, + "grad_norm": 0.25544601678848267, + "learning_rate": 3.574103585468376e-05, + "loss": 1.1254, + "step": 9766 + }, + { + "epoch": 0.5934139376632845, + "grad_norm": 0.24200955033302307, + "learning_rate": 3.57318608601761e-05, + "loss": 1.1257, + "step": 9767 + }, + { + "epoch": 0.5934746946959111, + "grad_norm": 0.23298683762550354, + "learning_rate": 3.572268638866985e-05, + "loss": 1.0588, + "step": 9768 + }, + { + "epoch": 0.5935354517285376, + "grad_norm": 0.3837737739086151, + "learning_rate": 3.5713512440501276e-05, + "loss": 1.1171, + "step": 9769 + }, + { + "epoch": 0.5935962087611641, + "grad_norm": 0.16318723559379578, + "learning_rate": 3.570433901600668e-05, + "loss": 1.0501, + "step": 9770 + }, + { + "epoch": 0.5936569657937907, + "grad_norm": 0.4686637818813324, + "learning_rate": 3.569516611552228e-05, + "loss": 1.041, + "step": 9771 + }, + { + "epoch": 0.5937177228264171, + "grad_norm": 0.14293159544467926, + "learning_rate": 3.568599373938435e-05, + "loss": 1.0094, + "step": 9772 + }, + { + "epoch": 0.5937784798590436, + "grad_norm": 0.15207961201667786, + "learning_rate": 3.5676821887929095e-05, + "loss": 1.0415, + "step": 9773 + }, + { + "epoch": 0.5938392368916702, + "grad_norm": 0.25316423177719116, + "learning_rate": 3.566765056149269e-05, + "loss": 1.2249, + "step": 9774 + }, + { + "epoch": 0.5938999939242967, + "grad_norm": 0.2173895388841629, + "learning_rate": 3.5658479760411326e-05, + "loss": 1.0824, + "step": 9775 + }, + { + "epoch": 0.5939607509569232, + "grad_norm": 0.7083467841148376, + "learning_rate": 3.564930948502115e-05, + "loss": 1.1928, + "step": 9776 + }, + { + "epoch": 0.5940215079895498, + "grad_norm": 1.5885332822799683, + "learning_rate": 3.564013973565832e-05, + "loss": 1.0859, + "step": 9777 + }, + { + "epoch": 0.5940822650221763, + "grad_norm": 0.44661155343055725, + "learning_rate": 3.5630970512658935e-05, + "loss": 1.0547, + "step": 9778 + }, + { + "epoch": 0.5941430220548028, + "grad_norm": 0.2305700182914734, + "learning_rate": 3.56218018163591e-05, + "loss": 1.0233, + "step": 9779 + }, + { + "epoch": 0.5942037790874294, + "grad_norm": 0.28299281001091003, + "learning_rate": 3.561263364709493e-05, + "loss": 1.1789, + "step": 9780 + }, + { + "epoch": 0.5942645361200559, + "grad_norm": 0.3785856366157532, + "learning_rate": 3.560346600520242e-05, + "loss": 1.1161, + "step": 9781 + }, + { + "epoch": 0.5943252931526825, + "grad_norm": 0.3673478364944458, + "learning_rate": 3.5594298891017654e-05, + "loss": 1.0165, + "step": 9782 + }, + { + "epoch": 0.594386050185309, + "grad_norm": 0.16438870131969452, + "learning_rate": 3.5585132304876655e-05, + "loss": 1.0701, + "step": 9783 + }, + { + "epoch": 0.5944468072179355, + "grad_norm": 0.13779112696647644, + "learning_rate": 3.55759662471154e-05, + "loss": 1.0367, + "step": 9784 + }, + { + "epoch": 0.594507564250562, + "grad_norm": 0.7678408622741699, + "learning_rate": 3.5566800718069904e-05, + "loss": 1.11, + "step": 9785 + }, + { + "epoch": 0.5945683212831885, + "grad_norm": 0.16460973024368286, + "learning_rate": 3.555763571807613e-05, + "loss": 1.0419, + "step": 9786 + }, + { + "epoch": 0.594629078315815, + "grad_norm": 0.2706615924835205, + "learning_rate": 3.5548471247469994e-05, + "loss": 1.0761, + "step": 9787 + }, + { + "epoch": 0.5946898353484416, + "grad_norm": 0.4967660903930664, + "learning_rate": 3.5539307306587444e-05, + "loss": 1.0316, + "step": 9788 + }, + { + "epoch": 0.5947505923810681, + "grad_norm": 0.13874489068984985, + "learning_rate": 3.553014389576437e-05, + "loss": 1.0172, + "step": 9789 + }, + { + "epoch": 0.5948113494136946, + "grad_norm": 0.26009342074394226, + "learning_rate": 3.552098101533667e-05, + "loss": 1.0196, + "step": 9790 + }, + { + "epoch": 0.5948721064463212, + "grad_norm": 0.16447220742702484, + "learning_rate": 3.5511818665640215e-05, + "loss": 1.0118, + "step": 9791 + }, + { + "epoch": 0.5949328634789477, + "grad_norm": 0.16123321652412415, + "learning_rate": 3.550265684701084e-05, + "loss": 1.0091, + "step": 9792 + }, + { + "epoch": 0.5949936205115742, + "grad_norm": 0.1560126543045044, + "learning_rate": 3.5493495559784405e-05, + "loss": 0.9721, + "step": 9793 + }, + { + "epoch": 0.5950543775442008, + "grad_norm": 1.0804160833358765, + "learning_rate": 3.5484334804296674e-05, + "loss": 1.074, + "step": 9794 + }, + { + "epoch": 0.5951151345768273, + "grad_norm": 0.15108312666416168, + "learning_rate": 3.5475174580883456e-05, + "loss": 1.0149, + "step": 9795 + }, + { + "epoch": 0.5951758916094538, + "grad_norm": 0.2284315675497055, + "learning_rate": 3.546601488988053e-05, + "loss": 1.1437, + "step": 9796 + }, + { + "epoch": 0.5952366486420804, + "grad_norm": 0.7955669164657593, + "learning_rate": 3.545685573162362e-05, + "loss": 1.0113, + "step": 9797 + }, + { + "epoch": 0.5952974056747068, + "grad_norm": 0.19802354276180267, + "learning_rate": 3.544769710644849e-05, + "loss": 1.0149, + "step": 9798 + }, + { + "epoch": 0.5953581627073333, + "grad_norm": 0.2099938988685608, + "learning_rate": 3.543853901469083e-05, + "loss": 1.0366, + "step": 9799 + }, + { + "epoch": 0.5954189197399599, + "grad_norm": 0.18992555141448975, + "learning_rate": 3.542938145668635e-05, + "loss": 1.1239, + "step": 9800 + }, + { + "epoch": 0.5954796767725864, + "grad_norm": 0.24753396213054657, + "learning_rate": 3.54202244327707e-05, + "loss": 1.0365, + "step": 9801 + }, + { + "epoch": 0.5955404338052129, + "grad_norm": 0.24218875169754028, + "learning_rate": 3.541106794327953e-05, + "loss": 1.0225, + "step": 9802 + }, + { + "epoch": 0.5956011908378395, + "grad_norm": 0.14680911600589752, + "learning_rate": 3.54019119885485e-05, + "loss": 1.0519, + "step": 9803 + }, + { + "epoch": 0.595661947870466, + "grad_norm": 0.2127353847026825, + "learning_rate": 3.53927565689132e-05, + "loss": 1.0753, + "step": 9804 + }, + { + "epoch": 0.5957227049030925, + "grad_norm": 0.3327980041503906, + "learning_rate": 3.538360168470922e-05, + "loss": 1.0752, + "step": 9805 + }, + { + "epoch": 0.5957834619357191, + "grad_norm": 0.23177003860473633, + "learning_rate": 3.537444733627219e-05, + "loss": 1.0024, + "step": 9806 + }, + { + "epoch": 0.5958442189683456, + "grad_norm": 0.23101577162742615, + "learning_rate": 3.536529352393757e-05, + "loss": 1.0623, + "step": 9807 + }, + { + "epoch": 0.5959049760009721, + "grad_norm": 0.2258395403623581, + "learning_rate": 3.535614024804097e-05, + "loss": 1.0328, + "step": 9808 + }, + { + "epoch": 0.5959657330335987, + "grad_norm": 0.19233787059783936, + "learning_rate": 3.5346987508917876e-05, + "loss": 1.0831, + "step": 9809 + }, + { + "epoch": 0.5960264900662252, + "grad_norm": 0.4917679727077484, + "learning_rate": 3.533783530690377e-05, + "loss": 1.2587, + "step": 9810 + }, + { + "epoch": 0.5960872470988517, + "grad_norm": 0.20104597508907318, + "learning_rate": 3.532868364233416e-05, + "loss": 1.0997, + "step": 9811 + }, + { + "epoch": 0.5961480041314782, + "grad_norm": 0.7530564069747925, + "learning_rate": 3.531953251554449e-05, + "loss": 1.2812, + "step": 9812 + }, + { + "epoch": 0.5962087611641047, + "grad_norm": 0.21984761953353882, + "learning_rate": 3.531038192687021e-05, + "loss": 1.1273, + "step": 9813 + }, + { + "epoch": 0.5962695181967312, + "grad_norm": 0.17666193842887878, + "learning_rate": 3.530123187664669e-05, + "loss": 1.0712, + "step": 9814 + }, + { + "epoch": 0.5963302752293578, + "grad_norm": 0.2476043850183487, + "learning_rate": 3.529208236520936e-05, + "loss": 1.0713, + "step": 9815 + }, + { + "epoch": 0.5963910322619843, + "grad_norm": 0.22253242135047913, + "learning_rate": 3.52829333928936e-05, + "loss": 0.9759, + "step": 9816 + }, + { + "epoch": 0.5964517892946108, + "grad_norm": 1.2658627033233643, + "learning_rate": 3.527378496003476e-05, + "loss": 1.1197, + "step": 9817 + }, + { + "epoch": 0.5965125463272374, + "grad_norm": 0.27710920572280884, + "learning_rate": 3.5264637066968175e-05, + "loss": 0.992, + "step": 9818 + }, + { + "epoch": 0.5965733033598639, + "grad_norm": 0.2740899622440338, + "learning_rate": 3.5255489714029176e-05, + "loss": 1.0489, + "step": 9819 + }, + { + "epoch": 0.5966340603924905, + "grad_norm": 0.34065383672714233, + "learning_rate": 3.524634290155306e-05, + "loss": 1.1858, + "step": 9820 + }, + { + "epoch": 0.596694817425117, + "grad_norm": 0.26940885186195374, + "learning_rate": 3.523719662987508e-05, + "loss": 1.0623, + "step": 9821 + }, + { + "epoch": 0.5967555744577435, + "grad_norm": 0.2562985420227051, + "learning_rate": 3.522805089933052e-05, + "loss": 1.1766, + "step": 9822 + }, + { + "epoch": 0.5968163314903701, + "grad_norm": 0.23675651848316193, + "learning_rate": 3.52189057102546e-05, + "loss": 1.1555, + "step": 9823 + }, + { + "epoch": 0.5968770885229966, + "grad_norm": 0.4015543460845947, + "learning_rate": 3.520976106298255e-05, + "loss": 1.1338, + "step": 9824 + }, + { + "epoch": 0.596937845555623, + "grad_norm": 0.17750370502471924, + "learning_rate": 3.520061695784958e-05, + "loss": 1.0064, + "step": 9825 + }, + { + "epoch": 0.5969986025882495, + "grad_norm": 0.19568850100040436, + "learning_rate": 3.5191473395190834e-05, + "loss": 1.0665, + "step": 9826 + }, + { + "epoch": 0.5970593596208761, + "grad_norm": 0.299297958612442, + "learning_rate": 3.518233037534151e-05, + "loss": 1.0706, + "step": 9827 + }, + { + "epoch": 0.5971201166535026, + "grad_norm": 0.15247681736946106, + "learning_rate": 3.517318789863673e-05, + "loss": 1.0333, + "step": 9828 + }, + { + "epoch": 0.5971808736861292, + "grad_norm": 0.24903614819049835, + "learning_rate": 3.516404596541161e-05, + "loss": 1.0649, + "step": 9829 + }, + { + "epoch": 0.5972416307187557, + "grad_norm": 0.14784032106399536, + "learning_rate": 3.515490457600126e-05, + "loss": 1.0239, + "step": 9830 + }, + { + "epoch": 0.5973023877513822, + "grad_norm": 0.14974428713321686, + "learning_rate": 3.5145763730740735e-05, + "loss": 1.0498, + "step": 9831 + }, + { + "epoch": 0.5973631447840088, + "grad_norm": 0.1778467297554016, + "learning_rate": 3.513662342996512e-05, + "loss": 0.9952, + "step": 9832 + }, + { + "epoch": 0.5974239018166353, + "grad_norm": 0.2092350721359253, + "learning_rate": 3.512748367400945e-05, + "loss": 1.015, + "step": 9833 + }, + { + "epoch": 0.5974846588492618, + "grad_norm": 0.38251379132270813, + "learning_rate": 3.511834446320875e-05, + "loss": 1.1142, + "step": 9834 + }, + { + "epoch": 0.5975454158818884, + "grad_norm": 0.19206781685352325, + "learning_rate": 3.510920579789799e-05, + "loss": 1.0411, + "step": 9835 + }, + { + "epoch": 0.5976061729145149, + "grad_norm": 0.2836723327636719, + "learning_rate": 3.510006767841218e-05, + "loss": 1.1415, + "step": 9836 + }, + { + "epoch": 0.5976669299471414, + "grad_norm": 0.15572905540466309, + "learning_rate": 3.5090930105086264e-05, + "loss": 1.0818, + "step": 9837 + }, + { + "epoch": 0.5977276869797679, + "grad_norm": 0.19883112609386444, + "learning_rate": 3.508179307825518e-05, + "loss": 1.1327, + "step": 9838 + }, + { + "epoch": 0.5977884440123944, + "grad_norm": 0.21243979036808014, + "learning_rate": 3.5072656598253875e-05, + "loss": 1.0592, + "step": 9839 + }, + { + "epoch": 0.5978492010450209, + "grad_norm": 0.19978618621826172, + "learning_rate": 3.506352066541723e-05, + "loss": 1.1037, + "step": 9840 + }, + { + "epoch": 0.5979099580776475, + "grad_norm": 4.999427795410156, + "learning_rate": 3.5054385280080104e-05, + "loss": 1.1338, + "step": 9841 + }, + { + "epoch": 0.597970715110274, + "grad_norm": 0.1780209243297577, + "learning_rate": 3.504525044257738e-05, + "loss": 1.1042, + "step": 9842 + }, + { + "epoch": 0.5980314721429005, + "grad_norm": 0.18891137838363647, + "learning_rate": 3.503611615324388e-05, + "loss": 1.0875, + "step": 9843 + }, + { + "epoch": 0.5980922291755271, + "grad_norm": 0.25424355268478394, + "learning_rate": 3.502698241241445e-05, + "loss": 1.1019, + "step": 9844 + }, + { + "epoch": 0.5981529862081536, + "grad_norm": 0.17623987793922424, + "learning_rate": 3.501784922042387e-05, + "loss": 1.0878, + "step": 9845 + }, + { + "epoch": 0.5982137432407801, + "grad_norm": 0.18165570497512817, + "learning_rate": 3.500871657760692e-05, + "loss": 1.0253, + "step": 9846 + }, + { + "epoch": 0.5982745002734067, + "grad_norm": 0.13430428504943848, + "learning_rate": 3.4999584484298385e-05, + "loss": 1.0045, + "step": 9847 + }, + { + "epoch": 0.5983352573060332, + "grad_norm": 0.2260645031929016, + "learning_rate": 3.4990452940832954e-05, + "loss": 1.1109, + "step": 9848 + }, + { + "epoch": 0.5983960143386597, + "grad_norm": 0.174525648355484, + "learning_rate": 3.4981321947545374e-05, + "loss": 1.0606, + "step": 9849 + }, + { + "epoch": 0.5984567713712863, + "grad_norm": 0.1836102306842804, + "learning_rate": 3.497219150477035e-05, + "loss": 1.0905, + "step": 9850 + }, + { + "epoch": 0.5985175284039127, + "grad_norm": 0.2666669189929962, + "learning_rate": 3.496306161284254e-05, + "loss": 1.1221, + "step": 9851 + }, + { + "epoch": 0.5985782854365392, + "grad_norm": 0.17689351737499237, + "learning_rate": 3.495393227209662e-05, + "loss": 1.0923, + "step": 9852 + }, + { + "epoch": 0.5986390424691658, + "grad_norm": 0.1892380267381668, + "learning_rate": 3.4944803482867225e-05, + "loss": 1.0622, + "step": 9853 + }, + { + "epoch": 0.5986997995017923, + "grad_norm": 0.19593334197998047, + "learning_rate": 3.4935675245488976e-05, + "loss": 1.1085, + "step": 9854 + }, + { + "epoch": 0.5987605565344188, + "grad_norm": 0.26064756512641907, + "learning_rate": 3.4926547560296445e-05, + "loss": 1.0458, + "step": 9855 + }, + { + "epoch": 0.5988213135670454, + "grad_norm": 0.2056310772895813, + "learning_rate": 3.4917420427624224e-05, + "loss": 1.0525, + "step": 9856 + }, + { + "epoch": 0.5988820705996719, + "grad_norm": 0.15796516835689545, + "learning_rate": 3.4908293847806885e-05, + "loss": 1.0866, + "step": 9857 + }, + { + "epoch": 0.5989428276322984, + "grad_norm": 0.2432084083557129, + "learning_rate": 3.489916782117895e-05, + "loss": 1.0935, + "step": 9858 + }, + { + "epoch": 0.599003584664925, + "grad_norm": 0.2152310013771057, + "learning_rate": 3.489004234807492e-05, + "loss": 1.0321, + "step": 9859 + }, + { + "epoch": 0.5990643416975515, + "grad_norm": 0.14516909420490265, + "learning_rate": 3.488091742882935e-05, + "loss": 1.0565, + "step": 9860 + }, + { + "epoch": 0.599125098730178, + "grad_norm": 0.11877552419900894, + "learning_rate": 3.4871793063776636e-05, + "loss": 1.0373, + "step": 9861 + }, + { + "epoch": 0.5991858557628046, + "grad_norm": 0.2581102252006531, + "learning_rate": 3.486266925325129e-05, + "loss": 1.1465, + "step": 9862 + }, + { + "epoch": 0.5992466127954311, + "grad_norm": 0.18569573760032654, + "learning_rate": 3.4853545997587725e-05, + "loss": 1.0972, + "step": 9863 + }, + { + "epoch": 0.5993073698280575, + "grad_norm": 0.22401855885982513, + "learning_rate": 3.4844423297120345e-05, + "loss": 1.1157, + "step": 9864 + }, + { + "epoch": 0.5993681268606841, + "grad_norm": 0.23354563117027283, + "learning_rate": 3.4835301152183575e-05, + "loss": 1.1255, + "step": 9865 + }, + { + "epoch": 0.5994288838933106, + "grad_norm": 0.1931324452161789, + "learning_rate": 3.4826179563111764e-05, + "loss": 1.0688, + "step": 9866 + }, + { + "epoch": 0.5994896409259372, + "grad_norm": 0.1954026073217392, + "learning_rate": 3.4817058530239285e-05, + "loss": 1.0936, + "step": 9867 + }, + { + "epoch": 0.5995503979585637, + "grad_norm": 0.1532920002937317, + "learning_rate": 3.480793805390046e-05, + "loss": 1.0291, + "step": 9868 + }, + { + "epoch": 0.5996111549911902, + "grad_norm": 0.2932119071483612, + "learning_rate": 3.479881813442959e-05, + "loss": 1.0571, + "step": 9869 + }, + { + "epoch": 0.5996719120238168, + "grad_norm": 0.24775323271751404, + "learning_rate": 3.4789698772160986e-05, + "loss": 0.9745, + "step": 9870 + }, + { + "epoch": 0.5997326690564433, + "grad_norm": 0.19105170667171478, + "learning_rate": 3.478057996742891e-05, + "loss": 1.0311, + "step": 9871 + }, + { + "epoch": 0.5997934260890698, + "grad_norm": 0.28020021319389343, + "learning_rate": 3.477146172056761e-05, + "loss": 1.0701, + "step": 9872 + }, + { + "epoch": 0.5998541831216964, + "grad_norm": 0.8257371783256531, + "learning_rate": 3.4762344031911354e-05, + "loss": 1.2584, + "step": 9873 + }, + { + "epoch": 0.5999149401543229, + "grad_norm": 0.2680833339691162, + "learning_rate": 3.4753226901794294e-05, + "loss": 1.2408, + "step": 9874 + }, + { + "epoch": 0.5999756971869494, + "grad_norm": 0.3145543038845062, + "learning_rate": 3.474411033055066e-05, + "loss": 1.1397, + "step": 9875 + }, + { + "epoch": 0.600036454219576, + "grad_norm": 0.41395053267478943, + "learning_rate": 3.473499431851461e-05, + "loss": 1.1226, + "step": 9876 + }, + { + "epoch": 0.6000972112522024, + "grad_norm": 3.0896875858306885, + "learning_rate": 3.472587886602028e-05, + "loss": 1.0819, + "step": 9877 + }, + { + "epoch": 0.6001579682848289, + "grad_norm": 0.2806435823440552, + "learning_rate": 3.471676397340183e-05, + "loss": 1.2051, + "step": 9878 + }, + { + "epoch": 0.6002187253174555, + "grad_norm": 0.1837780773639679, + "learning_rate": 3.470764964099334e-05, + "loss": 1.0852, + "step": 9879 + }, + { + "epoch": 0.600279482350082, + "grad_norm": 0.2748752534389496, + "learning_rate": 3.4698535869128936e-05, + "loss": 1.0903, + "step": 9880 + }, + { + "epoch": 0.6003402393827085, + "grad_norm": 0.22958989441394806, + "learning_rate": 3.4689422658142635e-05, + "loss": 1.0263, + "step": 9881 + }, + { + "epoch": 0.6004009964153351, + "grad_norm": 0.16941148042678833, + "learning_rate": 3.46803100083685e-05, + "loss": 1.0737, + "step": 9882 + }, + { + "epoch": 0.6004617534479616, + "grad_norm": 0.16698552668094635, + "learning_rate": 3.4671197920140564e-05, + "loss": 1.0093, + "step": 9883 + }, + { + "epoch": 0.6005225104805881, + "grad_norm": 0.21274149417877197, + "learning_rate": 3.466208639379284e-05, + "loss": 1.1013, + "step": 9884 + }, + { + "epoch": 0.6005832675132147, + "grad_norm": 0.20625872910022736, + "learning_rate": 3.465297542965929e-05, + "loss": 1.0705, + "step": 9885 + }, + { + "epoch": 0.6006440245458412, + "grad_norm": 0.15981997549533844, + "learning_rate": 3.464386502807391e-05, + "loss": 1.04, + "step": 9886 + }, + { + "epoch": 0.6007047815784677, + "grad_norm": 0.18398982286453247, + "learning_rate": 3.463475518937063e-05, + "loss": 1.0179, + "step": 9887 + }, + { + "epoch": 0.6007655386110943, + "grad_norm": 0.19266727566719055, + "learning_rate": 3.4625645913883355e-05, + "loss": 1.099, + "step": 9888 + }, + { + "epoch": 0.6008262956437208, + "grad_norm": 0.26012739539146423, + "learning_rate": 3.4616537201946e-05, + "loss": 1.1973, + "step": 9889 + }, + { + "epoch": 0.6008870526763472, + "grad_norm": 0.18100190162658691, + "learning_rate": 3.460742905389244e-05, + "loss": 1.0731, + "step": 9890 + }, + { + "epoch": 0.6009478097089738, + "grad_norm": 0.16277164220809937, + "learning_rate": 3.459832147005656e-05, + "loss": 1.026, + "step": 9891 + }, + { + "epoch": 0.6010085667416003, + "grad_norm": 0.19103237986564636, + "learning_rate": 3.458921445077218e-05, + "loss": 1.0281, + "step": 9892 + }, + { + "epoch": 0.6010693237742268, + "grad_norm": 0.16430512070655823, + "learning_rate": 3.4580107996373106e-05, + "loss": 1.0663, + "step": 9893 + }, + { + "epoch": 0.6011300808068534, + "grad_norm": 0.14718285202980042, + "learning_rate": 3.4571002107193195e-05, + "loss": 1.0118, + "step": 9894 + }, + { + "epoch": 0.6011908378394799, + "grad_norm": 0.24228739738464355, + "learning_rate": 3.4561896783566145e-05, + "loss": 1.1433, + "step": 9895 + }, + { + "epoch": 0.6012515948721064, + "grad_norm": 0.2201833873987198, + "learning_rate": 3.455279202582577e-05, + "loss": 1.135, + "step": 9896 + }, + { + "epoch": 0.601312351904733, + "grad_norm": 0.2937956750392914, + "learning_rate": 3.454368783430578e-05, + "loss": 1.2391, + "step": 9897 + }, + { + "epoch": 0.6013731089373595, + "grad_norm": 0.2742892801761627, + "learning_rate": 3.453458420933989e-05, + "loss": 1.0692, + "step": 9898 + }, + { + "epoch": 0.601433865969986, + "grad_norm": 0.35242658853530884, + "learning_rate": 3.452548115126183e-05, + "loss": 1.0139, + "step": 9899 + }, + { + "epoch": 0.6014946230026126, + "grad_norm": 0.20755353569984436, + "learning_rate": 3.451637866040524e-05, + "loss": 1.226, + "step": 9900 + }, + { + "epoch": 0.6015553800352391, + "grad_norm": 0.2526429295539856, + "learning_rate": 3.450727673710379e-05, + "loss": 1.1713, + "step": 9901 + }, + { + "epoch": 0.6016161370678657, + "grad_norm": 0.2739250063896179, + "learning_rate": 3.44981753816911e-05, + "loss": 1.0398, + "step": 9902 + }, + { + "epoch": 0.6016768941004921, + "grad_norm": 0.23383377492427826, + "learning_rate": 3.448907459450078e-05, + "loss": 1.148, + "step": 9903 + }, + { + "epoch": 0.6017376511331186, + "grad_norm": 0.21326719224452972, + "learning_rate": 3.447997437586643e-05, + "loss": 1.0056, + "step": 9904 + }, + { + "epoch": 0.6017984081657451, + "grad_norm": 0.32057198882102966, + "learning_rate": 3.447087472612163e-05, + "loss": 1.0241, + "step": 9905 + }, + { + "epoch": 0.6018591651983717, + "grad_norm": 0.14795731008052826, + "learning_rate": 3.44617756455999e-05, + "loss": 1.0749, + "step": 9906 + }, + { + "epoch": 0.6019199222309982, + "grad_norm": 0.19698049128055573, + "learning_rate": 3.445267713463482e-05, + "loss": 1.1842, + "step": 9907 + }, + { + "epoch": 0.6019806792636248, + "grad_norm": 0.2921554148197174, + "learning_rate": 3.444357919355984e-05, + "loss": 1.0999, + "step": 9908 + }, + { + "epoch": 0.6020414362962513, + "grad_norm": 0.3522489666938782, + "learning_rate": 3.443448182270848e-05, + "loss": 1.1065, + "step": 9909 + }, + { + "epoch": 0.6021021933288778, + "grad_norm": 0.17259444296360016, + "learning_rate": 3.442538502241419e-05, + "loss": 1.0567, + "step": 9910 + }, + { + "epoch": 0.6021629503615044, + "grad_norm": 0.31218039989471436, + "learning_rate": 3.441628879301042e-05, + "loss": 1.0535, + "step": 9911 + }, + { + "epoch": 0.6022237073941309, + "grad_norm": 0.42425158619880676, + "learning_rate": 3.44071931348306e-05, + "loss": 1.0004, + "step": 9912 + }, + { + "epoch": 0.6022844644267574, + "grad_norm": 0.2720471918582916, + "learning_rate": 3.439809804820814e-05, + "loss": 1.007, + "step": 9913 + }, + { + "epoch": 0.602345221459384, + "grad_norm": 0.1511344164609909, + "learning_rate": 3.438900353347641e-05, + "loss": 1.0107, + "step": 9914 + }, + { + "epoch": 0.6024059784920105, + "grad_norm": 0.13928796350955963, + "learning_rate": 3.437990959096877e-05, + "loss": 1.0183, + "step": 9915 + }, + { + "epoch": 0.602466735524637, + "grad_norm": 0.17372334003448486, + "learning_rate": 3.437081622101855e-05, + "loss": 1.0903, + "step": 9916 + }, + { + "epoch": 0.6025274925572635, + "grad_norm": 0.2963823676109314, + "learning_rate": 3.436172342395909e-05, + "loss": 1.0785, + "step": 9917 + }, + { + "epoch": 0.60258824958989, + "grad_norm": 0.23828846216201782, + "learning_rate": 3.435263120012369e-05, + "loss": 1.0775, + "step": 9918 + }, + { + "epoch": 0.6026490066225165, + "grad_norm": 0.17726165056228638, + "learning_rate": 3.4343539549845604e-05, + "loss": 1.0718, + "step": 9919 + }, + { + "epoch": 0.6027097636551431, + "grad_norm": 0.22877749800682068, + "learning_rate": 3.433444847345812e-05, + "loss": 1.1014, + "step": 9920 + }, + { + "epoch": 0.6027705206877696, + "grad_norm": 0.26189354062080383, + "learning_rate": 3.4325357971294447e-05, + "loss": 1.0412, + "step": 9921 + }, + { + "epoch": 0.6028312777203961, + "grad_norm": 0.3123627305030823, + "learning_rate": 3.4316268043687806e-05, + "loss": 1.0906, + "step": 9922 + }, + { + "epoch": 0.6028920347530227, + "grad_norm": 0.17752701044082642, + "learning_rate": 3.430717869097139e-05, + "loss": 1.0538, + "step": 9923 + }, + { + "epoch": 0.6029527917856492, + "grad_norm": 0.27796122431755066, + "learning_rate": 3.4298089913478365e-05, + "loss": 1.0322, + "step": 9924 + }, + { + "epoch": 0.6030135488182757, + "grad_norm": 0.14859558641910553, + "learning_rate": 3.42890017115419e-05, + "loss": 1.0261, + "step": 9925 + }, + { + "epoch": 0.6030743058509023, + "grad_norm": 0.3564405143260956, + "learning_rate": 3.427991408549511e-05, + "loss": 1.0366, + "step": 9926 + }, + { + "epoch": 0.6031350628835288, + "grad_norm": 0.2351512312889099, + "learning_rate": 3.427082703567113e-05, + "loss": 1.1517, + "step": 9927 + }, + { + "epoch": 0.6031958199161553, + "grad_norm": 0.16360944509506226, + "learning_rate": 3.426174056240299e-05, + "loss": 1.0955, + "step": 9928 + }, + { + "epoch": 0.6032565769487819, + "grad_norm": 0.2725212275981903, + "learning_rate": 3.425265466602381e-05, + "loss": 1.0085, + "step": 9929 + }, + { + "epoch": 0.6033173339814083, + "grad_norm": 0.37230294942855835, + "learning_rate": 3.424356934686661e-05, + "loss": 1.1356, + "step": 9930 + }, + { + "epoch": 0.6033780910140348, + "grad_norm": 0.23116841912269592, + "learning_rate": 3.4234484605264416e-05, + "loss": 1.1199, + "step": 9931 + }, + { + "epoch": 0.6034388480466614, + "grad_norm": 0.1707472801208496, + "learning_rate": 3.422540044155025e-05, + "loss": 1.0362, + "step": 9932 + }, + { + "epoch": 0.6034996050792879, + "grad_norm": 0.27363428473472595, + "learning_rate": 3.421631685605707e-05, + "loss": 1.0311, + "step": 9933 + }, + { + "epoch": 0.6035603621119144, + "grad_norm": 0.2761203646659851, + "learning_rate": 3.420723384911787e-05, + "loss": 1.089, + "step": 9934 + }, + { + "epoch": 0.603621119144541, + "grad_norm": 0.2761535942554474, + "learning_rate": 3.4198151421065546e-05, + "loss": 1.1277, + "step": 9935 + }, + { + "epoch": 0.6036818761771675, + "grad_norm": 0.20255540311336517, + "learning_rate": 3.418906957223303e-05, + "loss": 1.0311, + "step": 9936 + }, + { + "epoch": 0.603742633209794, + "grad_norm": 0.26905450224876404, + "learning_rate": 3.4179988302953235e-05, + "loss": 1.0735, + "step": 9937 + }, + { + "epoch": 0.6038033902424206, + "grad_norm": 0.49789196252822876, + "learning_rate": 3.417090761355902e-05, + "loss": 1.0563, + "step": 9938 + }, + { + "epoch": 0.6038641472750471, + "grad_norm": 0.2046404927968979, + "learning_rate": 3.416182750438325e-05, + "loss": 1.0557, + "step": 9939 + }, + { + "epoch": 0.6039249043076736, + "grad_norm": 0.2748302221298218, + "learning_rate": 3.415274797575878e-05, + "loss": 1.1093, + "step": 9940 + }, + { + "epoch": 0.6039856613403002, + "grad_norm": 0.18984586000442505, + "learning_rate": 3.414366902801836e-05, + "loss": 1.01, + "step": 9941 + }, + { + "epoch": 0.6040464183729267, + "grad_norm": 0.16108585894107819, + "learning_rate": 3.413459066149483e-05, + "loss": 1.0759, + "step": 9942 + }, + { + "epoch": 0.6041071754055531, + "grad_norm": 0.17295798659324646, + "learning_rate": 3.412551287652094e-05, + "loss": 1.0281, + "step": 9943 + }, + { + "epoch": 0.6041679324381797, + "grad_norm": 0.16011109948158264, + "learning_rate": 3.411643567342944e-05, + "loss": 1.0717, + "step": 9944 + }, + { + "epoch": 0.6042286894708062, + "grad_norm": 4.958183765411377, + "learning_rate": 3.410735905255307e-05, + "loss": 1.1984, + "step": 9945 + }, + { + "epoch": 0.6042894465034327, + "grad_norm": 0.19258776307106018, + "learning_rate": 3.409828301422453e-05, + "loss": 1.1049, + "step": 9946 + }, + { + "epoch": 0.6043502035360593, + "grad_norm": 0.15082234144210815, + "learning_rate": 3.408920755877651e-05, + "loss": 1.0331, + "step": 9947 + }, + { + "epoch": 0.6044109605686858, + "grad_norm": 0.1778273731470108, + "learning_rate": 3.4080132686541644e-05, + "loss": 1.1144, + "step": 9948 + }, + { + "epoch": 0.6044717176013124, + "grad_norm": 0.19462087750434875, + "learning_rate": 3.4071058397852586e-05, + "loss": 1.1473, + "step": 9949 + }, + { + "epoch": 0.6045324746339389, + "grad_norm": 0.22963158786296844, + "learning_rate": 3.406198469304197e-05, + "loss": 1.0567, + "step": 9950 + }, + { + "epoch": 0.6045932316665654, + "grad_norm": 0.24596284329891205, + "learning_rate": 3.405291157244238e-05, + "loss": 1.1347, + "step": 9951 + }, + { + "epoch": 0.604653988699192, + "grad_norm": 0.17323586344718933, + "learning_rate": 3.40438390363864e-05, + "loss": 1.0281, + "step": 9952 + }, + { + "epoch": 0.6047147457318185, + "grad_norm": 0.4308653771877289, + "learning_rate": 3.4034767085206586e-05, + "loss": 1.2055, + "step": 9953 + }, + { + "epoch": 0.604775502764445, + "grad_norm": 0.15391525626182556, + "learning_rate": 3.4025695719235484e-05, + "loss": 0.991, + "step": 9954 + }, + { + "epoch": 0.6048362597970716, + "grad_norm": 0.16045191884040833, + "learning_rate": 3.401662493880559e-05, + "loss": 1.1035, + "step": 9955 + }, + { + "epoch": 0.604897016829698, + "grad_norm": 0.28901949524879456, + "learning_rate": 3.400755474424939e-05, + "loss": 1.135, + "step": 9956 + }, + { + "epoch": 0.6049577738623245, + "grad_norm": 1.2806038856506348, + "learning_rate": 3.399848513589935e-05, + "loss": 1.0231, + "step": 9957 + }, + { + "epoch": 0.6050185308949511, + "grad_norm": 1.5975866317749023, + "learning_rate": 3.3989416114087946e-05, + "loss": 1.0401, + "step": 9958 + }, + { + "epoch": 0.6050792879275776, + "grad_norm": 0.20212270319461823, + "learning_rate": 3.398034767914759e-05, + "loss": 0.9994, + "step": 9959 + }, + { + "epoch": 0.6051400449602041, + "grad_norm": 0.22192856669425964, + "learning_rate": 3.397127983141067e-05, + "loss": 1.0946, + "step": 9960 + }, + { + "epoch": 0.6052008019928307, + "grad_norm": 0.233612060546875, + "learning_rate": 3.396221257120962e-05, + "loss": 1.0374, + "step": 9961 + }, + { + "epoch": 0.6052615590254572, + "grad_norm": 0.13939522206783295, + "learning_rate": 3.3953145898876746e-05, + "loss": 1.0563, + "step": 9962 + }, + { + "epoch": 0.6053223160580837, + "grad_norm": 0.2529875934123993, + "learning_rate": 3.394407981474441e-05, + "loss": 1.0903, + "step": 9963 + }, + { + "epoch": 0.6053830730907103, + "grad_norm": 0.2044186145067215, + "learning_rate": 3.393501431914493e-05, + "loss": 1.0597, + "step": 9964 + }, + { + "epoch": 0.6054438301233368, + "grad_norm": 0.24036280810832977, + "learning_rate": 3.3925949412410605e-05, + "loss": 1.0417, + "step": 9965 + }, + { + "epoch": 0.6055045871559633, + "grad_norm": 0.15199844539165497, + "learning_rate": 3.391688509487372e-05, + "loss": 1.0517, + "step": 9966 + }, + { + "epoch": 0.6055653441885899, + "grad_norm": 0.27164608240127563, + "learning_rate": 3.390782136686652e-05, + "loss": 1.1532, + "step": 9967 + }, + { + "epoch": 0.6056261012212164, + "grad_norm": 0.2821643054485321, + "learning_rate": 3.389875822872124e-05, + "loss": 1.0903, + "step": 9968 + }, + { + "epoch": 0.6056868582538428, + "grad_norm": 0.17748329043388367, + "learning_rate": 3.3889695680770085e-05, + "loss": 1.0308, + "step": 9969 + }, + { + "epoch": 0.6057476152864694, + "grad_norm": 0.15312828123569489, + "learning_rate": 3.388063372334524e-05, + "loss": 1.0406, + "step": 9970 + }, + { + "epoch": 0.6058083723190959, + "grad_norm": 0.28643226623535156, + "learning_rate": 3.387157235677889e-05, + "loss": 1.0783, + "step": 9971 + }, + { + "epoch": 0.6058691293517224, + "grad_norm": 0.13088832795619965, + "learning_rate": 3.386251158140317e-05, + "loss": 1.0474, + "step": 9972 + }, + { + "epoch": 0.605929886384349, + "grad_norm": 0.25981202721595764, + "learning_rate": 3.385345139755021e-05, + "loss": 1.1581, + "step": 9973 + }, + { + "epoch": 0.6059906434169755, + "grad_norm": 0.27505216002464294, + "learning_rate": 3.3844391805552124e-05, + "loss": 1.12, + "step": 9974 + }, + { + "epoch": 0.606051400449602, + "grad_norm": 0.22454072535037994, + "learning_rate": 3.383533280574096e-05, + "loss": 1.1202, + "step": 9975 + }, + { + "epoch": 0.6061121574822286, + "grad_norm": 0.14103436470031738, + "learning_rate": 3.38262743984488e-05, + "loss": 1.0262, + "step": 9976 + }, + { + "epoch": 0.6061729145148551, + "grad_norm": 0.25919297337532043, + "learning_rate": 3.3817216584007685e-05, + "loss": 1.1609, + "step": 9977 + }, + { + "epoch": 0.6062336715474816, + "grad_norm": 0.1673460453748703, + "learning_rate": 3.3808159362749605e-05, + "loss": 1.0426, + "step": 9978 + }, + { + "epoch": 0.6062944285801082, + "grad_norm": 0.1934138834476471, + "learning_rate": 3.379910273500659e-05, + "loss": 1.1272, + "step": 9979 + }, + { + "epoch": 0.6063551856127347, + "grad_norm": 0.19239842891693115, + "learning_rate": 3.3790046701110596e-05, + "loss": 1.119, + "step": 9980 + }, + { + "epoch": 0.6064159426453613, + "grad_norm": 0.3604520857334137, + "learning_rate": 3.378099126139358e-05, + "loss": 1.0276, + "step": 9981 + }, + { + "epoch": 0.6064766996779877, + "grad_norm": 0.2085612565279007, + "learning_rate": 3.3771936416187456e-05, + "loss": 1.1088, + "step": 9982 + }, + { + "epoch": 0.6065374567106142, + "grad_norm": 0.1883450150489807, + "learning_rate": 3.376288216582413e-05, + "loss": 1.0743, + "step": 9983 + }, + { + "epoch": 0.6065982137432407, + "grad_norm": 0.3065178394317627, + "learning_rate": 3.375382851063551e-05, + "loss": 1.0458, + "step": 9984 + }, + { + "epoch": 0.6066589707758673, + "grad_norm": 0.42738983035087585, + "learning_rate": 3.374477545095344e-05, + "loss": 1.0105, + "step": 9985 + }, + { + "epoch": 0.6067197278084938, + "grad_norm": 0.49046090245246887, + "learning_rate": 3.3735722987109764e-05, + "loss": 1.0706, + "step": 9986 + }, + { + "epoch": 0.6067804848411203, + "grad_norm": 0.1829921156167984, + "learning_rate": 3.3726671119436325e-05, + "loss": 1.074, + "step": 9987 + }, + { + "epoch": 0.6068412418737469, + "grad_norm": 0.6046299934387207, + "learning_rate": 3.371761984826488e-05, + "loss": 1.0743, + "step": 9988 + }, + { + "epoch": 0.6069019989063734, + "grad_norm": 0.6445074081420898, + "learning_rate": 3.370856917392724e-05, + "loss": 1.1915, + "step": 9989 + }, + { + "epoch": 0.606962755939, + "grad_norm": 0.1471179872751236, + "learning_rate": 3.369951909675514e-05, + "loss": 1.0141, + "step": 9990 + }, + { + "epoch": 0.6070235129716265, + "grad_norm": 0.1867695152759552, + "learning_rate": 3.3690469617080315e-05, + "loss": 1.0777, + "step": 9991 + }, + { + "epoch": 0.607084270004253, + "grad_norm": 0.7203973531723022, + "learning_rate": 3.368142073523448e-05, + "loss": 1.0851, + "step": 9992 + }, + { + "epoch": 0.6071450270368796, + "grad_norm": 3.4259936809539795, + "learning_rate": 3.367237245154933e-05, + "loss": 1.0796, + "step": 9993 + }, + { + "epoch": 0.6072057840695061, + "grad_norm": 0.2626950442790985, + "learning_rate": 3.366332476635653e-05, + "loss": 1.0772, + "step": 9994 + }, + { + "epoch": 0.6072665411021325, + "grad_norm": 0.1615108698606491, + "learning_rate": 3.36542776799877e-05, + "loss": 1.0147, + "step": 9995 + }, + { + "epoch": 0.607327298134759, + "grad_norm": 0.623496413230896, + "learning_rate": 3.3645231192774486e-05, + "loss": 1.075, + "step": 9996 + }, + { + "epoch": 0.6073880551673856, + "grad_norm": 0.24349495768547058, + "learning_rate": 3.363618530504848e-05, + "loss": 1.0404, + "step": 9997 + }, + { + "epoch": 0.6074488122000121, + "grad_norm": 0.44003167748451233, + "learning_rate": 3.362714001714128e-05, + "loss": 1.1066, + "step": 9998 + }, + { + "epoch": 0.6075095692326387, + "grad_norm": 0.22849029302597046, + "learning_rate": 3.361809532938441e-05, + "loss": 1.0509, + "step": 9999 + }, + { + "epoch": 0.6075703262652652, + "grad_norm": 0.29770034551620483, + "learning_rate": 3.360905124210943e-05, + "loss": 1.0423, + "step": 10000 + }, + { + "epoch": 0.6076310832978917, + "grad_norm": 0.18990741670131683, + "learning_rate": 3.360000775564785e-05, + "loss": 0.9872, + "step": 10001 + }, + { + "epoch": 0.6076918403305183, + "grad_norm": 0.1775653213262558, + "learning_rate": 3.3590964870331157e-05, + "loss": 1.0289, + "step": 10002 + }, + { + "epoch": 0.6077525973631448, + "grad_norm": 0.14696259796619415, + "learning_rate": 3.3581922586490815e-05, + "loss": 1.0463, + "step": 10003 + }, + { + "epoch": 0.6078133543957713, + "grad_norm": 0.3024024963378906, + "learning_rate": 3.357288090445827e-05, + "loss": 1.153, + "step": 10004 + }, + { + "epoch": 0.6078741114283979, + "grad_norm": 0.3817656636238098, + "learning_rate": 3.356383982456495e-05, + "loss": 1.0385, + "step": 10005 + }, + { + "epoch": 0.6079348684610244, + "grad_norm": 0.1715405136346817, + "learning_rate": 3.355479934714226e-05, + "loss": 1.0504, + "step": 10006 + }, + { + "epoch": 0.6079956254936509, + "grad_norm": 0.20394110679626465, + "learning_rate": 3.354575947252159e-05, + "loss": 1.0274, + "step": 10007 + }, + { + "epoch": 0.6080563825262774, + "grad_norm": 0.18502674996852875, + "learning_rate": 3.353672020103428e-05, + "loss": 1.0618, + "step": 10008 + }, + { + "epoch": 0.6081171395589039, + "grad_norm": 0.23505271971225739, + "learning_rate": 3.352768153301166e-05, + "loss": 1.1551, + "step": 10009 + }, + { + "epoch": 0.6081778965915304, + "grad_norm": 0.26736024022102356, + "learning_rate": 3.351864346878507e-05, + "loss": 1.1147, + "step": 10010 + }, + { + "epoch": 0.608238653624157, + "grad_norm": 0.19149713218212128, + "learning_rate": 3.350960600868579e-05, + "loss": 1.0708, + "step": 10011 + }, + { + "epoch": 0.6082994106567835, + "grad_norm": 0.2358231097459793, + "learning_rate": 3.350056915304508e-05, + "loss": 1.0199, + "step": 10012 + }, + { + "epoch": 0.60836016768941, + "grad_norm": 0.189131960272789, + "learning_rate": 3.34915329021942e-05, + "loss": 1.0564, + "step": 10013 + }, + { + "epoch": 0.6084209247220366, + "grad_norm": 0.22099080681800842, + "learning_rate": 3.348249725646439e-05, + "loss": 0.9979, + "step": 10014 + }, + { + "epoch": 0.6084816817546631, + "grad_norm": 0.27197638154029846, + "learning_rate": 3.347346221618682e-05, + "loss": 1.1879, + "step": 10015 + }, + { + "epoch": 0.6085424387872896, + "grad_norm": 0.14551647007465363, + "learning_rate": 3.3464427781692695e-05, + "loss": 1.0468, + "step": 10016 + }, + { + "epoch": 0.6086031958199162, + "grad_norm": 0.5579959154129028, + "learning_rate": 3.345539395331315e-05, + "loss": 1.1695, + "step": 10017 + }, + { + "epoch": 0.6086639528525427, + "grad_norm": 0.13692618906497955, + "learning_rate": 3.344636073137934e-05, + "loss": 1.057, + "step": 10018 + }, + { + "epoch": 0.6087247098851692, + "grad_norm": 0.17220866680145264, + "learning_rate": 3.343732811622239e-05, + "loss": 1.0682, + "step": 10019 + }, + { + "epoch": 0.6087854669177958, + "grad_norm": 0.18368585407733917, + "learning_rate": 3.342829610817337e-05, + "loss": 1.0565, + "step": 10020 + }, + { + "epoch": 0.6088462239504223, + "grad_norm": 0.2398783415555954, + "learning_rate": 3.341926470756337e-05, + "loss": 1.1139, + "step": 10021 + }, + { + "epoch": 0.6089069809830487, + "grad_norm": 0.23918280005455017, + "learning_rate": 3.341023391472342e-05, + "loss": 1.0632, + "step": 10022 + }, + { + "epoch": 0.6089677380156753, + "grad_norm": 0.2737674415111542, + "learning_rate": 3.340120372998455e-05, + "loss": 1.0539, + "step": 10023 + }, + { + "epoch": 0.6090284950483018, + "grad_norm": 0.14114613831043243, + "learning_rate": 3.339217415367778e-05, + "loss": 1.0543, + "step": 10024 + }, + { + "epoch": 0.6090892520809283, + "grad_norm": 0.24521104991436005, + "learning_rate": 3.338314518613406e-05, + "loss": 1.2214, + "step": 10025 + }, + { + "epoch": 0.6091500091135549, + "grad_norm": 0.1915372610092163, + "learning_rate": 3.337411682768436e-05, + "loss": 1.0355, + "step": 10026 + }, + { + "epoch": 0.6092107661461814, + "grad_norm": 0.1637675166130066, + "learning_rate": 3.336508907865964e-05, + "loss": 1.035, + "step": 10027 + }, + { + "epoch": 0.609271523178808, + "grad_norm": 0.16703706979751587, + "learning_rate": 3.3356061939390795e-05, + "loss": 1.0529, + "step": 10028 + }, + { + "epoch": 0.6093322802114345, + "grad_norm": 0.3652595579624176, + "learning_rate": 3.3347035410208705e-05, + "loss": 1.2051, + "step": 10029 + }, + { + "epoch": 0.609393037244061, + "grad_norm": 0.25046488642692566, + "learning_rate": 3.333800949144426e-05, + "loss": 1.0073, + "step": 10030 + }, + { + "epoch": 0.6094537942766876, + "grad_norm": 0.18478882312774658, + "learning_rate": 3.332898418342829e-05, + "loss": 1.0532, + "step": 10031 + }, + { + "epoch": 0.6095145513093141, + "grad_norm": 0.27501046657562256, + "learning_rate": 3.331995948649161e-05, + "loss": 1.0086, + "step": 10032 + }, + { + "epoch": 0.6095753083419406, + "grad_norm": 0.2536748945713043, + "learning_rate": 3.331093540096507e-05, + "loss": 1.0779, + "step": 10033 + }, + { + "epoch": 0.6096360653745672, + "grad_norm": 0.416840136051178, + "learning_rate": 3.330191192717942e-05, + "loss": 1.2009, + "step": 10034 + }, + { + "epoch": 0.6096968224071936, + "grad_norm": 0.36155620217323303, + "learning_rate": 3.3292889065465405e-05, + "loss": 1.3119, + "step": 10035 + }, + { + "epoch": 0.6097575794398201, + "grad_norm": 0.1383623480796814, + "learning_rate": 3.328386681615377e-05, + "loss": 1.051, + "step": 10036 + }, + { + "epoch": 0.6098183364724467, + "grad_norm": 0.17462247610092163, + "learning_rate": 3.327484517957522e-05, + "loss": 1.0308, + "step": 10037 + }, + { + "epoch": 0.6098790935050732, + "grad_norm": 0.2941528558731079, + "learning_rate": 3.326582415606046e-05, + "loss": 1.024, + "step": 10038 + }, + { + "epoch": 0.6099398505376997, + "grad_norm": 6.779785633087158, + "learning_rate": 3.325680374594015e-05, + "loss": 1.2247, + "step": 10039 + }, + { + "epoch": 0.6100006075703263, + "grad_norm": 0.24607540667057037, + "learning_rate": 3.324778394954493e-05, + "loss": 1.0587, + "step": 10040 + }, + { + "epoch": 0.6100613646029528, + "grad_norm": 0.22240161895751953, + "learning_rate": 3.323876476720546e-05, + "loss": 1.1059, + "step": 10041 + }, + { + "epoch": 0.6101221216355793, + "grad_norm": 0.15586572885513306, + "learning_rate": 3.322974619925227e-05, + "loss": 1.0291, + "step": 10042 + }, + { + "epoch": 0.6101828786682059, + "grad_norm": 0.1533995121717453, + "learning_rate": 3.322072824601598e-05, + "loss": 1.0315, + "step": 10043 + }, + { + "epoch": 0.6102436357008324, + "grad_norm": 0.34493890404701233, + "learning_rate": 3.321171090782715e-05, + "loss": 1.2354, + "step": 10044 + }, + { + "epoch": 0.6103043927334589, + "grad_norm": 0.6337910890579224, + "learning_rate": 3.320269418501628e-05, + "loss": 1.0521, + "step": 10045 + }, + { + "epoch": 0.6103651497660855, + "grad_norm": 0.24548110365867615, + "learning_rate": 3.3193678077913926e-05, + "loss": 1.0847, + "step": 10046 + }, + { + "epoch": 0.610425906798712, + "grad_norm": 0.24315710365772247, + "learning_rate": 3.3184662586850545e-05, + "loss": 1.0599, + "step": 10047 + }, + { + "epoch": 0.6104866638313384, + "grad_norm": 0.21195003390312195, + "learning_rate": 3.317564771215661e-05, + "loss": 1.0291, + "step": 10048 + }, + { + "epoch": 0.610547420863965, + "grad_norm": 0.22682352364063263, + "learning_rate": 3.316663345416256e-05, + "loss": 1.1126, + "step": 10049 + }, + { + "epoch": 0.6106081778965915, + "grad_norm": 0.22534573078155518, + "learning_rate": 3.315761981319881e-05, + "loss": 1.123, + "step": 10050 + }, + { + "epoch": 0.610668934929218, + "grad_norm": 0.4557771384716034, + "learning_rate": 3.314860678959575e-05, + "loss": 1.0316, + "step": 10051 + }, + { + "epoch": 0.6107296919618446, + "grad_norm": 0.1513921022415161, + "learning_rate": 3.313959438368378e-05, + "loss": 1.013, + "step": 10052 + }, + { + "epoch": 0.6107904489944711, + "grad_norm": 0.2849206328392029, + "learning_rate": 3.3130582595793226e-05, + "loss": 1.1463, + "step": 10053 + }, + { + "epoch": 0.6108512060270976, + "grad_norm": 0.3992820978164673, + "learning_rate": 3.3121571426254456e-05, + "loss": 1.0656, + "step": 10054 + }, + { + "epoch": 0.6109119630597242, + "grad_norm": 5.082508563995361, + "learning_rate": 3.311256087539771e-05, + "loss": 1.1011, + "step": 10055 + }, + { + "epoch": 0.6109727200923507, + "grad_norm": 0.35456353425979614, + "learning_rate": 3.310355094355333e-05, + "loss": 1.1687, + "step": 10056 + }, + { + "epoch": 0.6110334771249772, + "grad_norm": 1.137507677078247, + "learning_rate": 3.309454163105156e-05, + "loss": 1.0212, + "step": 10057 + }, + { + "epoch": 0.6110942341576038, + "grad_norm": 0.5151012539863586, + "learning_rate": 3.308553293822262e-05, + "loss": 1.1776, + "step": 10058 + }, + { + "epoch": 0.6111549911902303, + "grad_norm": 0.20152853429317474, + "learning_rate": 3.307652486539674e-05, + "loss": 1.091, + "step": 10059 + }, + { + "epoch": 0.6112157482228568, + "grad_norm": 0.24182768166065216, + "learning_rate": 3.306751741290413e-05, + "loss": 1.1749, + "step": 10060 + }, + { + "epoch": 0.6112765052554833, + "grad_norm": 0.30117037892341614, + "learning_rate": 3.305851058107494e-05, + "loss": 1.1618, + "step": 10061 + }, + { + "epoch": 0.6113372622881098, + "grad_norm": 2.719102144241333, + "learning_rate": 3.30495043702393e-05, + "loss": 1.0605, + "step": 10062 + }, + { + "epoch": 0.6113980193207363, + "grad_norm": 0.20633748173713684, + "learning_rate": 3.3040498780727355e-05, + "loss": 1.004, + "step": 10063 + }, + { + "epoch": 0.6114587763533629, + "grad_norm": 0.17750100791454315, + "learning_rate": 3.303149381286922e-05, + "loss": 1.088, + "step": 10064 + }, + { + "epoch": 0.6115195333859894, + "grad_norm": 0.17178645730018616, + "learning_rate": 3.3022489466994954e-05, + "loss": 1.0471, + "step": 10065 + }, + { + "epoch": 0.611580290418616, + "grad_norm": 0.3692171573638916, + "learning_rate": 3.3013485743434615e-05, + "loss": 1.2249, + "step": 10066 + }, + { + "epoch": 0.6116410474512425, + "grad_norm": 0.2835707366466522, + "learning_rate": 3.300448264251824e-05, + "loss": 1.0679, + "step": 10067 + }, + { + "epoch": 0.611701804483869, + "grad_norm": 0.19177448749542236, + "learning_rate": 3.299548016457586e-05, + "loss": 1.036, + "step": 10068 + }, + { + "epoch": 0.6117625615164956, + "grad_norm": 0.21391868591308594, + "learning_rate": 3.298647830993742e-05, + "loss": 0.9773, + "step": 10069 + }, + { + "epoch": 0.6118233185491221, + "grad_norm": 0.2547816336154938, + "learning_rate": 3.2977477078932904e-05, + "loss": 1.0381, + "step": 10070 + }, + { + "epoch": 0.6118840755817486, + "grad_norm": 0.16034641861915588, + "learning_rate": 3.296847647189225e-05, + "loss": 1.0685, + "step": 10071 + }, + { + "epoch": 0.6119448326143752, + "grad_norm": 0.1585560142993927, + "learning_rate": 3.29594764891454e-05, + "loss": 0.9959, + "step": 10072 + }, + { + "epoch": 0.6120055896470017, + "grad_norm": 0.19166015088558197, + "learning_rate": 3.295047713102223e-05, + "loss": 1.0392, + "step": 10073 + }, + { + "epoch": 0.6120663466796281, + "grad_norm": 0.1464860737323761, + "learning_rate": 3.294147839785262e-05, + "loss": 1.0339, + "step": 10074 + }, + { + "epoch": 0.6121271037122546, + "grad_norm": 0.23034080862998962, + "learning_rate": 3.293248028996642e-05, + "loss": 1.0842, + "step": 10075 + }, + { + "epoch": 0.6121878607448812, + "grad_norm": 0.17896944284439087, + "learning_rate": 3.292348280769343e-05, + "loss": 0.9821, + "step": 10076 + }, + { + "epoch": 0.6122486177775077, + "grad_norm": 0.18529874086380005, + "learning_rate": 3.29144859513635e-05, + "loss": 1.0268, + "step": 10077 + }, + { + "epoch": 0.6123093748101343, + "grad_norm": 0.1673300564289093, + "learning_rate": 3.2905489721306385e-05, + "loss": 1.0663, + "step": 10078 + }, + { + "epoch": 0.6123701318427608, + "grad_norm": 0.2462995946407318, + "learning_rate": 3.289649411785183e-05, + "loss": 1.1534, + "step": 10079 + }, + { + "epoch": 0.6124308888753873, + "grad_norm": 0.19418473541736603, + "learning_rate": 3.2887499141329606e-05, + "loss": 1.0584, + "step": 10080 + }, + { + "epoch": 0.6124916459080139, + "grad_norm": 0.2694394886493683, + "learning_rate": 3.287850479206942e-05, + "loss": 1.043, + "step": 10081 + }, + { + "epoch": 0.6125524029406404, + "grad_norm": 0.21942435204982758, + "learning_rate": 3.286951107040094e-05, + "loss": 1.046, + "step": 10082 + }, + { + "epoch": 0.6126131599732669, + "grad_norm": 0.20871299505233765, + "learning_rate": 3.2860517976653845e-05, + "loss": 0.9946, + "step": 10083 + }, + { + "epoch": 0.6126739170058935, + "grad_norm": 0.3086872100830078, + "learning_rate": 3.2851525511157764e-05, + "loss": 1.0421, + "step": 10084 + }, + { + "epoch": 0.61273467403852, + "grad_norm": 0.22375884652137756, + "learning_rate": 3.284253367424234e-05, + "loss": 1.015, + "step": 10085 + }, + { + "epoch": 0.6127954310711465, + "grad_norm": 0.44149845838546753, + "learning_rate": 3.283354246623716e-05, + "loss": 1.0307, + "step": 10086 + }, + { + "epoch": 0.612856188103773, + "grad_norm": 0.23019739985466003, + "learning_rate": 3.282455188747179e-05, + "loss": 1.0719, + "step": 10087 + }, + { + "epoch": 0.6129169451363995, + "grad_norm": 0.31628966331481934, + "learning_rate": 3.2815561938275815e-05, + "loss": 1.0126, + "step": 10088 + }, + { + "epoch": 0.612977702169026, + "grad_norm": 0.24273350834846497, + "learning_rate": 3.280657261897872e-05, + "loss": 1.1768, + "step": 10089 + }, + { + "epoch": 0.6130384592016526, + "grad_norm": 2.0942304134368896, + "learning_rate": 3.279758392991004e-05, + "loss": 1.0591, + "step": 10090 + }, + { + "epoch": 0.6130992162342791, + "grad_norm": 0.3770155906677246, + "learning_rate": 3.278859587139923e-05, + "loss": 1.0132, + "step": 10091 + }, + { + "epoch": 0.6131599732669056, + "grad_norm": 1.4221439361572266, + "learning_rate": 3.277960844377577e-05, + "loss": 1.2447, + "step": 10092 + }, + { + "epoch": 0.6132207302995322, + "grad_norm": 0.13615866005420685, + "learning_rate": 3.277062164736909e-05, + "loss": 1.0225, + "step": 10093 + }, + { + "epoch": 0.6132814873321587, + "grad_norm": 0.1682521104812622, + "learning_rate": 3.276163548250862e-05, + "loss": 1.0212, + "step": 10094 + }, + { + "epoch": 0.6133422443647852, + "grad_norm": 0.6234630346298218, + "learning_rate": 3.275264994952373e-05, + "loss": 1.0978, + "step": 10095 + }, + { + "epoch": 0.6134030013974118, + "grad_norm": 0.22052881121635437, + "learning_rate": 3.2743665048743785e-05, + "loss": 1.0972, + "step": 10096 + }, + { + "epoch": 0.6134637584300383, + "grad_norm": 0.1781950443983078, + "learning_rate": 3.273468078049812e-05, + "loss": 1.0566, + "step": 10097 + }, + { + "epoch": 0.6135245154626648, + "grad_norm": 0.18953752517700195, + "learning_rate": 3.272569714511607e-05, + "loss": 1.0312, + "step": 10098 + }, + { + "epoch": 0.6135852724952914, + "grad_norm": 0.207096666097641, + "learning_rate": 3.2716714142926953e-05, + "loss": 1.004, + "step": 10099 + }, + { + "epoch": 0.6136460295279178, + "grad_norm": 0.3253735601902008, + "learning_rate": 3.270773177425999e-05, + "loss": 1.1431, + "step": 10100 + }, + { + "epoch": 0.6137067865605443, + "grad_norm": 0.15784801542758942, + "learning_rate": 3.2698750039444504e-05, + "loss": 1.0496, + "step": 10101 + }, + { + "epoch": 0.6137675435931709, + "grad_norm": 0.18263408541679382, + "learning_rate": 3.268976893880964e-05, + "loss": 1.0797, + "step": 10102 + }, + { + "epoch": 0.6138283006257974, + "grad_norm": 0.30145519971847534, + "learning_rate": 3.268078847268465e-05, + "loss": 1.1939, + "step": 10103 + }, + { + "epoch": 0.6138890576584239, + "grad_norm": 0.19124795496463776, + "learning_rate": 3.2671808641398714e-05, + "loss": 1.0378, + "step": 10104 + }, + { + "epoch": 0.6139498146910505, + "grad_norm": 0.15886706113815308, + "learning_rate": 3.266282944528097e-05, + "loss": 1.0284, + "step": 10105 + }, + { + "epoch": 0.614010571723677, + "grad_norm": 0.15414468944072723, + "learning_rate": 3.2653850884660576e-05, + "loss": 1.0758, + "step": 10106 + }, + { + "epoch": 0.6140713287563035, + "grad_norm": 0.44625595211982727, + "learning_rate": 3.2644872959866626e-05, + "loss": 1.2355, + "step": 10107 + }, + { + "epoch": 0.6141320857889301, + "grad_norm": 0.14319609105587006, + "learning_rate": 3.263589567122823e-05, + "loss": 1.0299, + "step": 10108 + }, + { + "epoch": 0.6141928428215566, + "grad_norm": 0.25840696692466736, + "learning_rate": 3.262691901907442e-05, + "loss": 1.2118, + "step": 10109 + }, + { + "epoch": 0.6142535998541832, + "grad_norm": 0.20369909703731537, + "learning_rate": 3.261794300373424e-05, + "loss": 1.0718, + "step": 10110 + }, + { + "epoch": 0.6143143568868097, + "grad_norm": 0.17461955547332764, + "learning_rate": 3.260896762553673e-05, + "loss": 1.1146, + "step": 10111 + }, + { + "epoch": 0.6143751139194362, + "grad_norm": 0.4261322021484375, + "learning_rate": 3.259999288481087e-05, + "loss": 1.228, + "step": 10112 + }, + { + "epoch": 0.6144358709520626, + "grad_norm": 0.19938696920871735, + "learning_rate": 3.259101878188563e-05, + "loss": 1.0336, + "step": 10113 + }, + { + "epoch": 0.6144966279846892, + "grad_norm": 0.1384919434785843, + "learning_rate": 3.2582045317089966e-05, + "loss": 1.0045, + "step": 10114 + }, + { + "epoch": 0.6145573850173157, + "grad_norm": 0.30236369371414185, + "learning_rate": 3.257307249075281e-05, + "loss": 1.2236, + "step": 10115 + }, + { + "epoch": 0.6146181420499423, + "grad_norm": 0.1642615795135498, + "learning_rate": 3.2564100303203035e-05, + "loss": 1.0748, + "step": 10116 + }, + { + "epoch": 0.6146788990825688, + "grad_norm": 0.27784842252731323, + "learning_rate": 3.2555128754769544e-05, + "loss": 1.0977, + "step": 10117 + }, + { + "epoch": 0.6147396561151953, + "grad_norm": 0.16382871568202972, + "learning_rate": 3.254615784578117e-05, + "loss": 1.0394, + "step": 10118 + }, + { + "epoch": 0.6148004131478219, + "grad_norm": 0.17480334639549255, + "learning_rate": 3.253718757656675e-05, + "loss": 1.0365, + "step": 10119 + }, + { + "epoch": 0.6148611701804484, + "grad_norm": 0.20663578808307648, + "learning_rate": 3.252821794745511e-05, + "loss": 1.0959, + "step": 10120 + }, + { + "epoch": 0.6149219272130749, + "grad_norm": 0.23179388046264648, + "learning_rate": 3.251924895877502e-05, + "loss": 1.0659, + "step": 10121 + }, + { + "epoch": 0.6149826842457015, + "grad_norm": 0.21536658704280853, + "learning_rate": 3.251028061085522e-05, + "loss": 1.1325, + "step": 10122 + }, + { + "epoch": 0.615043441278328, + "grad_norm": 0.18155747652053833, + "learning_rate": 3.250131290402448e-05, + "loss": 1.0185, + "step": 10123 + }, + { + "epoch": 0.6151041983109545, + "grad_norm": 0.17900489270687103, + "learning_rate": 3.249234583861149e-05, + "loss": 1.0363, + "step": 10124 + }, + { + "epoch": 0.6151649553435811, + "grad_norm": 0.2936435043811798, + "learning_rate": 3.248337941494496e-05, + "loss": 1.0695, + "step": 10125 + }, + { + "epoch": 0.6152257123762076, + "grad_norm": 0.1760258972644806, + "learning_rate": 3.2474413633353515e-05, + "loss": 1.0364, + "step": 10126 + }, + { + "epoch": 0.615286469408834, + "grad_norm": 0.21129821240901947, + "learning_rate": 3.246544849416585e-05, + "loss": 1.0617, + "step": 10127 + }, + { + "epoch": 0.6153472264414606, + "grad_norm": 0.15322187542915344, + "learning_rate": 3.245648399771057e-05, + "loss": 1.0695, + "step": 10128 + }, + { + "epoch": 0.6154079834740871, + "grad_norm": 0.16864202916622162, + "learning_rate": 3.2447520144316246e-05, + "loss": 1.103, + "step": 10129 + }, + { + "epoch": 0.6154687405067136, + "grad_norm": 0.1438341736793518, + "learning_rate": 3.243855693431146e-05, + "loss": 1.0315, + "step": 10130 + }, + { + "epoch": 0.6155294975393402, + "grad_norm": 0.1905677318572998, + "learning_rate": 3.2429594368024766e-05, + "loss": 1.1002, + "step": 10131 + }, + { + "epoch": 0.6155902545719667, + "grad_norm": 0.19476403295993805, + "learning_rate": 3.2420632445784685e-05, + "loss": 1.0096, + "step": 10132 + }, + { + "epoch": 0.6156510116045932, + "grad_norm": 0.21540936827659607, + "learning_rate": 3.241167116791971e-05, + "loss": 1.1881, + "step": 10133 + }, + { + "epoch": 0.6157117686372198, + "grad_norm": 0.17864125967025757, + "learning_rate": 3.240271053475834e-05, + "loss": 1.0248, + "step": 10134 + }, + { + "epoch": 0.6157725256698463, + "grad_norm": 0.1908658891916275, + "learning_rate": 3.239375054662903e-05, + "loss": 1.0532, + "step": 10135 + }, + { + "epoch": 0.6158332827024728, + "grad_norm": 0.20724207162857056, + "learning_rate": 3.238479120386019e-05, + "loss": 1.0477, + "step": 10136 + }, + { + "epoch": 0.6158940397350994, + "grad_norm": 0.1920655518770218, + "learning_rate": 3.237583250678022e-05, + "loss": 1.0304, + "step": 10137 + }, + { + "epoch": 0.6159547967677259, + "grad_norm": 0.16035227477550507, + "learning_rate": 3.236687445571752e-05, + "loss": 1.0243, + "step": 10138 + }, + { + "epoch": 0.6160155538003524, + "grad_norm": 0.24631831049919128, + "learning_rate": 3.2357917051000454e-05, + "loss": 1.335, + "step": 10139 + }, + { + "epoch": 0.6160763108329789, + "grad_norm": 0.4307802617549896, + "learning_rate": 3.2348960292957345e-05, + "loss": 1.1465, + "step": 10140 + }, + { + "epoch": 0.6161370678656054, + "grad_norm": 0.17299385368824005, + "learning_rate": 3.234000418191652e-05, + "loss": 1.0124, + "step": 10141 + }, + { + "epoch": 0.6161978248982319, + "grad_norm": 0.5028324127197266, + "learning_rate": 3.233104871820623e-05, + "loss": 1.2705, + "step": 10142 + }, + { + "epoch": 0.6162585819308585, + "grad_norm": 0.17996293306350708, + "learning_rate": 3.2322093902154776e-05, + "loss": 1.0546, + "step": 10143 + }, + { + "epoch": 0.616319338963485, + "grad_norm": 0.25919288396835327, + "learning_rate": 3.2313139734090395e-05, + "loss": 1.0141, + "step": 10144 + }, + { + "epoch": 0.6163800959961115, + "grad_norm": 0.5761468410491943, + "learning_rate": 3.2304186214341284e-05, + "loss": 1.2037, + "step": 10145 + }, + { + "epoch": 0.6164408530287381, + "grad_norm": 0.25784218311309814, + "learning_rate": 3.2295233343235645e-05, + "loss": 1.1031, + "step": 10146 + }, + { + "epoch": 0.6165016100613646, + "grad_norm": 0.18163453042507172, + "learning_rate": 3.228628112110166e-05, + "loss": 1.0009, + "step": 10147 + }, + { + "epoch": 0.6165623670939911, + "grad_norm": 0.26517972350120544, + "learning_rate": 3.227732954826747e-05, + "loss": 1.0857, + "step": 10148 + }, + { + "epoch": 0.6166231241266177, + "grad_norm": 0.2820921242237091, + "learning_rate": 3.2268378625061185e-05, + "loss": 1.1452, + "step": 10149 + }, + { + "epoch": 0.6166838811592442, + "grad_norm": 0.3755902349948883, + "learning_rate": 3.225942835181091e-05, + "loss": 1.2728, + "step": 10150 + }, + { + "epoch": 0.6167446381918708, + "grad_norm": 0.29077252745628357, + "learning_rate": 3.2250478728844704e-05, + "loss": 1.0967, + "step": 10151 + }, + { + "epoch": 0.6168053952244973, + "grad_norm": 0.22165298461914062, + "learning_rate": 3.224152975649064e-05, + "loss": 1.0133, + "step": 10152 + }, + { + "epoch": 0.6168661522571237, + "grad_norm": 0.1433059126138687, + "learning_rate": 3.223258143507674e-05, + "loss": 1.0462, + "step": 10153 + }, + { + "epoch": 0.6169269092897502, + "grad_norm": 0.24733296036720276, + "learning_rate": 3.2223633764931e-05, + "loss": 1.1458, + "step": 10154 + }, + { + "epoch": 0.6169876663223768, + "grad_norm": 0.17426623404026031, + "learning_rate": 3.2214686746381415e-05, + "loss": 1.0322, + "step": 10155 + }, + { + "epoch": 0.6170484233550033, + "grad_norm": 0.44509753584861755, + "learning_rate": 3.22057403797559e-05, + "loss": 1.1648, + "step": 10156 + }, + { + "epoch": 0.6171091803876299, + "grad_norm": 0.17009349167346954, + "learning_rate": 3.219679466538242e-05, + "loss": 1.0442, + "step": 10157 + }, + { + "epoch": 0.6171699374202564, + "grad_norm": 1.046763300895691, + "learning_rate": 3.218784960358888e-05, + "loss": 1.0137, + "step": 10158 + }, + { + "epoch": 0.6172306944528829, + "grad_norm": 0.20213112235069275, + "learning_rate": 3.217890519470313e-05, + "loss": 1.0051, + "step": 10159 + }, + { + "epoch": 0.6172914514855095, + "grad_norm": 0.43099287152290344, + "learning_rate": 3.216996143905308e-05, + "loss": 1.073, + "step": 10160 + }, + { + "epoch": 0.617352208518136, + "grad_norm": 0.1688334345817566, + "learning_rate": 3.216101833696655e-05, + "loss": 1.0683, + "step": 10161 + }, + { + "epoch": 0.6174129655507625, + "grad_norm": 0.34064480662345886, + "learning_rate": 3.215207588877133e-05, + "loss": 1.0363, + "step": 10162 + }, + { + "epoch": 0.6174737225833891, + "grad_norm": 0.25082653760910034, + "learning_rate": 3.214313409479521e-05, + "loss": 1.0643, + "step": 10163 + }, + { + "epoch": 0.6175344796160156, + "grad_norm": 0.20345745980739594, + "learning_rate": 3.2134192955365954e-05, + "loss": 1.0658, + "step": 10164 + }, + { + "epoch": 0.6175952366486421, + "grad_norm": 0.16762571036815643, + "learning_rate": 3.2125252470811315e-05, + "loss": 1.0177, + "step": 10165 + }, + { + "epoch": 0.6176559936812686, + "grad_norm": 0.1598019301891327, + "learning_rate": 3.211631264145901e-05, + "loss": 1.0282, + "step": 10166 + }, + { + "epoch": 0.6177167507138951, + "grad_norm": 1.0749990940093994, + "learning_rate": 3.2107373467636715e-05, + "loss": 1.0397, + "step": 10167 + }, + { + "epoch": 0.6177775077465216, + "grad_norm": 0.15978261828422546, + "learning_rate": 3.209843494967213e-05, + "loss": 1.0698, + "step": 10168 + }, + { + "epoch": 0.6178382647791482, + "grad_norm": 0.1726628839969635, + "learning_rate": 3.208949708789284e-05, + "loss": 1.02, + "step": 10169 + }, + { + "epoch": 0.6178990218117747, + "grad_norm": 0.17053233087062836, + "learning_rate": 3.2080559882626505e-05, + "loss": 1.0256, + "step": 10170 + }, + { + "epoch": 0.6179597788444012, + "grad_norm": 0.15688998997211456, + "learning_rate": 3.2071623334200714e-05, + "loss": 1.0144, + "step": 10171 + }, + { + "epoch": 0.6180205358770278, + "grad_norm": 0.17563804984092712, + "learning_rate": 3.2062687442943023e-05, + "loss": 1.0854, + "step": 10172 + }, + { + "epoch": 0.6180812929096543, + "grad_norm": 0.20266608893871307, + "learning_rate": 3.2053752209181004e-05, + "loss": 1.1281, + "step": 10173 + }, + { + "epoch": 0.6181420499422808, + "grad_norm": 0.22245417535305023, + "learning_rate": 3.2044817633242165e-05, + "loss": 1.1485, + "step": 10174 + }, + { + "epoch": 0.6182028069749074, + "grad_norm": 0.18357734382152557, + "learning_rate": 3.203588371545402e-05, + "loss": 0.9876, + "step": 10175 + }, + { + "epoch": 0.6182635640075339, + "grad_norm": 0.16036096215248108, + "learning_rate": 3.2026950456144013e-05, + "loss": 1.0487, + "step": 10176 + }, + { + "epoch": 0.6183243210401604, + "grad_norm": 0.16604144871234894, + "learning_rate": 3.2018017855639606e-05, + "loss": 1.0172, + "step": 10177 + }, + { + "epoch": 0.618385078072787, + "grad_norm": 0.1468852162361145, + "learning_rate": 3.200908591426823e-05, + "loss": 1.054, + "step": 10178 + }, + { + "epoch": 0.6184458351054134, + "grad_norm": 0.13226790726184845, + "learning_rate": 3.2000154632357305e-05, + "loss": 1.0423, + "step": 10179 + }, + { + "epoch": 0.6185065921380399, + "grad_norm": 0.25710687041282654, + "learning_rate": 3.199122401023417e-05, + "loss": 1.1567, + "step": 10180 + }, + { + "epoch": 0.6185673491706665, + "grad_norm": 0.20025576651096344, + "learning_rate": 3.198229404822621e-05, + "loss": 1.0509, + "step": 10181 + }, + { + "epoch": 0.618628106203293, + "grad_norm": 0.20072899758815765, + "learning_rate": 3.197336474666076e-05, + "loss": 1.1505, + "step": 10182 + }, + { + "epoch": 0.6186888632359195, + "grad_norm": 0.18351779878139496, + "learning_rate": 3.196443610586509e-05, + "loss": 1.156, + "step": 10183 + }, + { + "epoch": 0.6187496202685461, + "grad_norm": 0.5188824534416199, + "learning_rate": 3.195550812616651e-05, + "loss": 1.1634, + "step": 10184 + }, + { + "epoch": 0.6188103773011726, + "grad_norm": 0.6952223777770996, + "learning_rate": 3.1946580807892255e-05, + "loss": 1.0882, + "step": 10185 + }, + { + "epoch": 0.6188711343337991, + "grad_norm": 0.1457202136516571, + "learning_rate": 3.1937654151369584e-05, + "loss": 1.0466, + "step": 10186 + }, + { + "epoch": 0.6189318913664257, + "grad_norm": 0.16795623302459717, + "learning_rate": 3.192872815692569e-05, + "loss": 1.0858, + "step": 10187 + }, + { + "epoch": 0.6189926483990522, + "grad_norm": 0.16598843038082123, + "learning_rate": 3.191980282488776e-05, + "loss": 1.053, + "step": 10188 + }, + { + "epoch": 0.6190534054316787, + "grad_norm": 0.2908492386341095, + "learning_rate": 3.191087815558295e-05, + "loss": 1.0529, + "step": 10189 + }, + { + "epoch": 0.6191141624643053, + "grad_norm": 0.24747945368289948, + "learning_rate": 3.190195414933839e-05, + "loss": 1.1561, + "step": 10190 + }, + { + "epoch": 0.6191749194969318, + "grad_norm": 0.21363358199596405, + "learning_rate": 3.18930308064812e-05, + "loss": 1.1603, + "step": 10191 + }, + { + "epoch": 0.6192356765295582, + "grad_norm": 0.13437344133853912, + "learning_rate": 3.188410812733847e-05, + "loss": 1.0521, + "step": 10192 + }, + { + "epoch": 0.6192964335621848, + "grad_norm": 0.32211780548095703, + "learning_rate": 3.187518611223725e-05, + "loss": 1.0275, + "step": 10193 + }, + { + "epoch": 0.6193571905948113, + "grad_norm": 0.28306442499160767, + "learning_rate": 3.186626476150459e-05, + "loss": 1.0678, + "step": 10194 + }, + { + "epoch": 0.6194179476274378, + "grad_norm": 0.20646969974040985, + "learning_rate": 3.185734407546752e-05, + "loss": 1.0628, + "step": 10195 + }, + { + "epoch": 0.6194787046600644, + "grad_norm": 0.14269468188285828, + "learning_rate": 3.1848424054452994e-05, + "loss": 1.0502, + "step": 10196 + }, + { + "epoch": 0.6195394616926909, + "grad_norm": 0.28430935740470886, + "learning_rate": 3.1839504698787995e-05, + "loss": 1.0202, + "step": 10197 + }, + { + "epoch": 0.6196002187253175, + "grad_norm": 0.2105807065963745, + "learning_rate": 3.1830586008799454e-05, + "loss": 1.0476, + "step": 10198 + }, + { + "epoch": 0.619660975757944, + "grad_norm": 0.2991226613521576, + "learning_rate": 3.1821667984814305e-05, + "loss": 1.0679, + "step": 10199 + }, + { + "epoch": 0.6197217327905705, + "grad_norm": 0.26441940665245056, + "learning_rate": 3.181275062715943e-05, + "loss": 1.002, + "step": 10200 + }, + { + "epoch": 0.6197824898231971, + "grad_norm": 0.28742945194244385, + "learning_rate": 3.1803833936161676e-05, + "loss": 1.1182, + "step": 10201 + }, + { + "epoch": 0.6198432468558236, + "grad_norm": 0.21710854768753052, + "learning_rate": 3.179491791214795e-05, + "loss": 1.0291, + "step": 10202 + }, + { + "epoch": 0.6199040038884501, + "grad_norm": 0.16838732361793518, + "learning_rate": 3.178600255544498e-05, + "loss": 1.0186, + "step": 10203 + }, + { + "epoch": 0.6199647609210767, + "grad_norm": 0.38471871614456177, + "learning_rate": 3.177708786637963e-05, + "loss": 1.1346, + "step": 10204 + }, + { + "epoch": 0.6200255179537031, + "grad_norm": 0.3194098174571991, + "learning_rate": 3.1768173845278645e-05, + "loss": 1.1237, + "step": 10205 + }, + { + "epoch": 0.6200862749863296, + "grad_norm": 0.16880454123020172, + "learning_rate": 3.1759260492468754e-05, + "loss": 1.0552, + "step": 10206 + }, + { + "epoch": 0.6201470320189562, + "grad_norm": 0.260284423828125, + "learning_rate": 3.175034780827672e-05, + "loss": 0.9924, + "step": 10207 + }, + { + "epoch": 0.6202077890515827, + "grad_norm": 0.27796441316604614, + "learning_rate": 3.174143579302922e-05, + "loss": 1.0398, + "step": 10208 + }, + { + "epoch": 0.6202685460842092, + "grad_norm": 0.21115843951702118, + "learning_rate": 3.1732524447052904e-05, + "loss": 1.0761, + "step": 10209 + }, + { + "epoch": 0.6203293031168358, + "grad_norm": 0.2716456949710846, + "learning_rate": 3.1723613770674436e-05, + "loss": 1.1493, + "step": 10210 + }, + { + "epoch": 0.6203900601494623, + "grad_norm": 0.1650032252073288, + "learning_rate": 3.171470376422043e-05, + "loss": 1.0183, + "step": 10211 + }, + { + "epoch": 0.6204508171820888, + "grad_norm": 0.21368509531021118, + "learning_rate": 3.1705794428017506e-05, + "loss": 1.0168, + "step": 10212 + }, + { + "epoch": 0.6205115742147154, + "grad_norm": 0.24319179356098175, + "learning_rate": 3.169688576239222e-05, + "loss": 1.0352, + "step": 10213 + }, + { + "epoch": 0.6205723312473419, + "grad_norm": 0.2903781235218048, + "learning_rate": 3.168797776767112e-05, + "loss": 1.1074, + "step": 10214 + }, + { + "epoch": 0.6206330882799684, + "grad_norm": 0.16955311596393585, + "learning_rate": 3.167907044418074e-05, + "loss": 0.9948, + "step": 10215 + }, + { + "epoch": 0.620693845312595, + "grad_norm": 0.28391364216804504, + "learning_rate": 3.1670163792247574e-05, + "loss": 1.0954, + "step": 10216 + }, + { + "epoch": 0.6207546023452215, + "grad_norm": 0.5145750641822815, + "learning_rate": 3.16612578121981e-05, + "loss": 1.0596, + "step": 10217 + }, + { + "epoch": 0.6208153593778479, + "grad_norm": 0.23124797642230988, + "learning_rate": 3.165235250435876e-05, + "loss": 1.1413, + "step": 10218 + }, + { + "epoch": 0.6208761164104745, + "grad_norm": 0.4238141179084778, + "learning_rate": 3.164344786905598e-05, + "loss": 1.0319, + "step": 10219 + }, + { + "epoch": 0.620936873443101, + "grad_norm": 0.23900336027145386, + "learning_rate": 3.163454390661617e-05, + "loss": 1.068, + "step": 10220 + }, + { + "epoch": 0.6209976304757275, + "grad_norm": 0.21207095682621002, + "learning_rate": 3.162564061736571e-05, + "loss": 1.0632, + "step": 10221 + }, + { + "epoch": 0.6210583875083541, + "grad_norm": 0.26243436336517334, + "learning_rate": 3.1616738001630954e-05, + "loss": 1.1335, + "step": 10222 + }, + { + "epoch": 0.6211191445409806, + "grad_norm": 0.4101652503013611, + "learning_rate": 3.1607836059738196e-05, + "loss": 1.09, + "step": 10223 + }, + { + "epoch": 0.6211799015736071, + "grad_norm": 0.24113506078720093, + "learning_rate": 3.159893479201377e-05, + "loss": 1.1806, + "step": 10224 + }, + { + "epoch": 0.6212406586062337, + "grad_norm": 0.1972876489162445, + "learning_rate": 3.1590034198783956e-05, + "loss": 1.03, + "step": 10225 + }, + { + "epoch": 0.6213014156388602, + "grad_norm": 0.26333487033843994, + "learning_rate": 3.158113428037499e-05, + "loss": 1.12, + "step": 10226 + }, + { + "epoch": 0.6213621726714867, + "grad_norm": 0.2510044276714325, + "learning_rate": 3.15722350371131e-05, + "loss": 1.0337, + "step": 10227 + }, + { + "epoch": 0.6214229297041133, + "grad_norm": 0.24830971658229828, + "learning_rate": 3.156333646932451e-05, + "loss": 1.0788, + "step": 10228 + }, + { + "epoch": 0.6214836867367398, + "grad_norm": 0.16549190878868103, + "learning_rate": 3.155443857733539e-05, + "loss": 1.0938, + "step": 10229 + }, + { + "epoch": 0.6215444437693664, + "grad_norm": 0.21437163650989532, + "learning_rate": 3.1545541361471876e-05, + "loss": 1.0935, + "step": 10230 + }, + { + "epoch": 0.6216052008019929, + "grad_norm": 0.30399852991104126, + "learning_rate": 3.1536644822060105e-05, + "loss": 1.0723, + "step": 10231 + }, + { + "epoch": 0.6216659578346193, + "grad_norm": 0.2795153856277466, + "learning_rate": 3.1527748959426194e-05, + "loss": 1.1576, + "step": 10232 + }, + { + "epoch": 0.6217267148672458, + "grad_norm": 0.1609867960214615, + "learning_rate": 3.151885377389622e-05, + "loss": 1.0341, + "step": 10233 + }, + { + "epoch": 0.6217874718998724, + "grad_norm": 0.24298638105392456, + "learning_rate": 3.150995926579622e-05, + "loss": 1.0688, + "step": 10234 + }, + { + "epoch": 0.6218482289324989, + "grad_norm": 0.36414822936058044, + "learning_rate": 3.150106543545227e-05, + "loss": 1.1314, + "step": 10235 + }, + { + "epoch": 0.6219089859651254, + "grad_norm": 0.5812063217163086, + "learning_rate": 3.1492172283190314e-05, + "loss": 1.0452, + "step": 10236 + }, + { + "epoch": 0.621969742997752, + "grad_norm": 0.2066909521818161, + "learning_rate": 3.148327980933636e-05, + "loss": 1.06, + "step": 10237 + }, + { + "epoch": 0.6220305000303785, + "grad_norm": 0.3819483518600464, + "learning_rate": 3.1474388014216376e-05, + "loss": 1.1038, + "step": 10238 + }, + { + "epoch": 0.622091257063005, + "grad_norm": 0.16834676265716553, + "learning_rate": 3.1465496898156266e-05, + "loss": 1.0358, + "step": 10239 + }, + { + "epoch": 0.6221520140956316, + "grad_norm": 0.207370787858963, + "learning_rate": 3.145660646148196e-05, + "loss": 1.0374, + "step": 10240 + }, + { + "epoch": 0.6222127711282581, + "grad_norm": 0.4460877478122711, + "learning_rate": 3.1447716704519326e-05, + "loss": 1.2075, + "step": 10241 + }, + { + "epoch": 0.6222735281608847, + "grad_norm": 0.4133269786834717, + "learning_rate": 3.143882762759424e-05, + "loss": 1.2827, + "step": 10242 + }, + { + "epoch": 0.6223342851935112, + "grad_norm": 0.29919084906578064, + "learning_rate": 3.1429939231032494e-05, + "loss": 1.1151, + "step": 10243 + }, + { + "epoch": 0.6223950422261377, + "grad_norm": 0.2154017686843872, + "learning_rate": 3.142105151515991e-05, + "loss": 1.124, + "step": 10244 + }, + { + "epoch": 0.6224557992587642, + "grad_norm": 0.2455597072839737, + "learning_rate": 3.1412164480302286e-05, + "loss": 1.1155, + "step": 10245 + }, + { + "epoch": 0.6225165562913907, + "grad_norm": 0.2005384862422943, + "learning_rate": 3.140327812678536e-05, + "loss": 0.9988, + "step": 10246 + }, + { + "epoch": 0.6225773133240172, + "grad_norm": 0.1392596811056137, + "learning_rate": 3.139439245493486e-05, + "loss": 1.0388, + "step": 10247 + }, + { + "epoch": 0.6226380703566438, + "grad_norm": 0.1672644466161728, + "learning_rate": 3.138550746507652e-05, + "loss": 1.0235, + "step": 10248 + }, + { + "epoch": 0.6226988273892703, + "grad_norm": 1.942883014678955, + "learning_rate": 3.137662315753601e-05, + "loss": 1.0025, + "step": 10249 + }, + { + "epoch": 0.6227595844218968, + "grad_norm": 0.24306084215641022, + "learning_rate": 3.1367739532638964e-05, + "loss": 1.1036, + "step": 10250 + }, + { + "epoch": 0.6228203414545234, + "grad_norm": 0.3663457930088043, + "learning_rate": 3.135885659071102e-05, + "loss": 1.2677, + "step": 10251 + }, + { + "epoch": 0.6228810984871499, + "grad_norm": 0.1878475695848465, + "learning_rate": 3.134997433207779e-05, + "loss": 1.0698, + "step": 10252 + }, + { + "epoch": 0.6229418555197764, + "grad_norm": 0.2427508533000946, + "learning_rate": 3.1341092757064864e-05, + "loss": 1.1229, + "step": 10253 + }, + { + "epoch": 0.623002612552403, + "grad_norm": 0.15525515377521515, + "learning_rate": 3.1332211865997786e-05, + "loss": 0.9977, + "step": 10254 + }, + { + "epoch": 0.6230633695850295, + "grad_norm": 0.16334670782089233, + "learning_rate": 3.132333165920211e-05, + "loss": 1.0496, + "step": 10255 + }, + { + "epoch": 0.623124126617656, + "grad_norm": 0.22797450423240662, + "learning_rate": 3.1314452137003305e-05, + "loss": 1.0938, + "step": 10256 + }, + { + "epoch": 0.6231848836502826, + "grad_norm": 0.2155742645263672, + "learning_rate": 3.130557329972685e-05, + "loss": 1.1173, + "step": 10257 + }, + { + "epoch": 0.623245640682909, + "grad_norm": 1.5453203916549683, + "learning_rate": 3.129669514769825e-05, + "loss": 1.0055, + "step": 10258 + }, + { + "epoch": 0.6233063977155355, + "grad_norm": 0.31893816590309143, + "learning_rate": 3.128781768124289e-05, + "loss": 1.1359, + "step": 10259 + }, + { + "epoch": 0.6233671547481621, + "grad_norm": 0.19334197044372559, + "learning_rate": 3.127894090068618e-05, + "loss": 1.1145, + "step": 10260 + }, + { + "epoch": 0.6234279117807886, + "grad_norm": 0.1989111751317978, + "learning_rate": 3.1270064806353525e-05, + "loss": 1.0848, + "step": 10261 + }, + { + "epoch": 0.6234886688134151, + "grad_norm": 0.20146800577640533, + "learning_rate": 3.126118939857027e-05, + "loss": 1.0897, + "step": 10262 + }, + { + "epoch": 0.6235494258460417, + "grad_norm": 0.16453489661216736, + "learning_rate": 3.1252314677661734e-05, + "loss": 1.1312, + "step": 10263 + }, + { + "epoch": 0.6236101828786682, + "grad_norm": 0.2294052094221115, + "learning_rate": 3.124344064395322e-05, + "loss": 1.0412, + "step": 10264 + }, + { + "epoch": 0.6236709399112947, + "grad_norm": 0.537531316280365, + "learning_rate": 3.1234567297770025e-05, + "loss": 1.0698, + "step": 10265 + }, + { + "epoch": 0.6237316969439213, + "grad_norm": 0.18664078414440155, + "learning_rate": 3.122569463943739e-05, + "loss": 1.1387, + "step": 10266 + }, + { + "epoch": 0.6237924539765478, + "grad_norm": 0.20761434733867645, + "learning_rate": 3.1216822669280545e-05, + "loss": 1.0998, + "step": 10267 + }, + { + "epoch": 0.6238532110091743, + "grad_norm": 3.007406234741211, + "learning_rate": 3.120795138762468e-05, + "loss": 1.1117, + "step": 10268 + }, + { + "epoch": 0.6239139680418009, + "grad_norm": 0.20627567172050476, + "learning_rate": 3.119908079479503e-05, + "loss": 1.1595, + "step": 10269 + }, + { + "epoch": 0.6239747250744274, + "grad_norm": 0.2101801037788391, + "learning_rate": 3.1190210891116674e-05, + "loss": 1.0403, + "step": 10270 + }, + { + "epoch": 0.6240354821070538, + "grad_norm": 0.1659596562385559, + "learning_rate": 3.118134167691478e-05, + "loss": 1.0183, + "step": 10271 + }, + { + "epoch": 0.6240962391396804, + "grad_norm": 1.942302942276001, + "learning_rate": 3.117247315251445e-05, + "loss": 1.0013, + "step": 10272 + }, + { + "epoch": 0.6241569961723069, + "grad_norm": 0.1856427937746048, + "learning_rate": 3.116360531824074e-05, + "loss": 1.1701, + "step": 10273 + }, + { + "epoch": 0.6242177532049334, + "grad_norm": 0.395821213722229, + "learning_rate": 3.115473817441873e-05, + "loss": 1.082, + "step": 10274 + }, + { + "epoch": 0.62427851023756, + "grad_norm": 0.20507104694843292, + "learning_rate": 3.114587172137345e-05, + "loss": 1.0906, + "step": 10275 + }, + { + "epoch": 0.6243392672701865, + "grad_norm": 0.2472638338804245, + "learning_rate": 3.113700595942988e-05, + "loss": 1.1301, + "step": 10276 + }, + { + "epoch": 0.624400024302813, + "grad_norm": 0.7463474869728088, + "learning_rate": 3.1128140888912994e-05, + "loss": 1.3888, + "step": 10277 + }, + { + "epoch": 0.6244607813354396, + "grad_norm": 0.29054322838783264, + "learning_rate": 3.1119276510147744e-05, + "loss": 1.0355, + "step": 10278 + }, + { + "epoch": 0.6245215383680661, + "grad_norm": 0.40375226736068726, + "learning_rate": 3.111041282345907e-05, + "loss": 1.2001, + "step": 10279 + }, + { + "epoch": 0.6245822954006927, + "grad_norm": 0.14367203414440155, + "learning_rate": 3.110154982917187e-05, + "loss": 1.0334, + "step": 10280 + }, + { + "epoch": 0.6246430524333192, + "grad_norm": 0.2817092537879944, + "learning_rate": 3.1092687527610994e-05, + "loss": 0.9946, + "step": 10281 + }, + { + "epoch": 0.6247038094659457, + "grad_norm": 0.2110549658536911, + "learning_rate": 3.1083825919101354e-05, + "loss": 1.0181, + "step": 10282 + }, + { + "epoch": 0.6247645664985723, + "grad_norm": 0.1490904688835144, + "learning_rate": 3.107496500396769e-05, + "loss": 1.0308, + "step": 10283 + }, + { + "epoch": 0.6248253235311987, + "grad_norm": 1.7914942502975464, + "learning_rate": 3.1066104782534866e-05, + "loss": 1.056, + "step": 10284 + }, + { + "epoch": 0.6248860805638252, + "grad_norm": 0.24690064787864685, + "learning_rate": 3.105724525512762e-05, + "loss": 1.0339, + "step": 10285 + }, + { + "epoch": 0.6249468375964518, + "grad_norm": 0.1767684519290924, + "learning_rate": 3.1048386422070704e-05, + "loss": 1.0612, + "step": 10286 + }, + { + "epoch": 0.6250075946290783, + "grad_norm": 0.171746164560318, + "learning_rate": 3.1039528283688854e-05, + "loss": 1.0702, + "step": 10287 + }, + { + "epoch": 0.6250683516617048, + "grad_norm": 0.2601718306541443, + "learning_rate": 3.1030670840306766e-05, + "loss": 1.1602, + "step": 10288 + }, + { + "epoch": 0.6251291086943314, + "grad_norm": 0.17959505319595337, + "learning_rate": 3.102181409224911e-05, + "loss": 1.0642, + "step": 10289 + }, + { + "epoch": 0.6251898657269579, + "grad_norm": 0.17178158462047577, + "learning_rate": 3.101295803984052e-05, + "loss": 1.0658, + "step": 10290 + }, + { + "epoch": 0.6252506227595844, + "grad_norm": 0.21338163316249847, + "learning_rate": 3.100410268340561e-05, + "loss": 1.0217, + "step": 10291 + }, + { + "epoch": 0.625311379792211, + "grad_norm": 0.18336667120456696, + "learning_rate": 3.0995248023269e-05, + "loss": 1.0482, + "step": 10292 + }, + { + "epoch": 0.6253721368248375, + "grad_norm": 0.13656415045261383, + "learning_rate": 3.098639405975525e-05, + "loss": 1.0289, + "step": 10293 + }, + { + "epoch": 0.625432893857464, + "grad_norm": 0.26850664615631104, + "learning_rate": 3.097754079318889e-05, + "loss": 1.0094, + "step": 10294 + }, + { + "epoch": 0.6254936508900906, + "grad_norm": 1.1956884860992432, + "learning_rate": 3.096868822389448e-05, + "loss": 1.0807, + "step": 10295 + }, + { + "epoch": 0.6255544079227171, + "grad_norm": 0.16048064827919006, + "learning_rate": 3.095983635219645e-05, + "loss": 1.0184, + "step": 10296 + }, + { + "epoch": 0.6256151649553435, + "grad_norm": 0.1715308129787445, + "learning_rate": 3.09509851784193e-05, + "loss": 1.0635, + "step": 10297 + }, + { + "epoch": 0.6256759219879701, + "grad_norm": 0.15712113678455353, + "learning_rate": 3.0942134702887485e-05, + "loss": 1.0528, + "step": 10298 + }, + { + "epoch": 0.6257366790205966, + "grad_norm": 0.18228782713413239, + "learning_rate": 3.0933284925925386e-05, + "loss": 1.1484, + "step": 10299 + }, + { + "epoch": 0.6257974360532231, + "grad_norm": 0.3532848060131073, + "learning_rate": 3.092443584785742e-05, + "loss": 1.0452, + "step": 10300 + }, + { + "epoch": 0.6258581930858497, + "grad_norm": 0.25136813521385193, + "learning_rate": 3.0915587469007954e-05, + "loss": 1.1038, + "step": 10301 + }, + { + "epoch": 0.6259189501184762, + "grad_norm": 0.15983234345912933, + "learning_rate": 3.090673978970133e-05, + "loss": 1.1001, + "step": 10302 + }, + { + "epoch": 0.6259797071511027, + "grad_norm": 0.18707209825515747, + "learning_rate": 3.0897892810261844e-05, + "loss": 1.079, + "step": 10303 + }, + { + "epoch": 0.6260404641837293, + "grad_norm": 0.2426978349685669, + "learning_rate": 3.0889046531013765e-05, + "loss": 1.1763, + "step": 10304 + }, + { + "epoch": 0.6261012212163558, + "grad_norm": 0.7528966069221497, + "learning_rate": 3.0880200952281395e-05, + "loss": 1.0481, + "step": 10305 + }, + { + "epoch": 0.6261619782489823, + "grad_norm": 0.25707924365997314, + "learning_rate": 3.087135607438896e-05, + "loss": 1.0276, + "step": 10306 + }, + { + "epoch": 0.6262227352816089, + "grad_norm": 0.2600954473018646, + "learning_rate": 3.086251189766065e-05, + "loss": 1.0219, + "step": 10307 + }, + { + "epoch": 0.6262834923142354, + "grad_norm": 0.29236140847206116, + "learning_rate": 3.0853668422420675e-05, + "loss": 1.0593, + "step": 10308 + }, + { + "epoch": 0.626344249346862, + "grad_norm": 1.7043170928955078, + "learning_rate": 3.08448256489932e-05, + "loss": 1.0903, + "step": 10309 + }, + { + "epoch": 0.6264050063794884, + "grad_norm": 0.21469131112098694, + "learning_rate": 3.083598357770233e-05, + "loss": 1.0929, + "step": 10310 + }, + { + "epoch": 0.6264657634121149, + "grad_norm": 0.5576730370521545, + "learning_rate": 3.082714220887218e-05, + "loss": 1.0203, + "step": 10311 + }, + { + "epoch": 0.6265265204447414, + "grad_norm": 0.16076548397541046, + "learning_rate": 3.081830154282684e-05, + "loss": 1.0392, + "step": 10312 + }, + { + "epoch": 0.626587277477368, + "grad_norm": 0.2302761673927307, + "learning_rate": 3.080946157989037e-05, + "loss": 1.1515, + "step": 10313 + }, + { + "epoch": 0.6266480345099945, + "grad_norm": 0.3249976336956024, + "learning_rate": 3.0800622320386795e-05, + "loss": 1.0438, + "step": 10314 + }, + { + "epoch": 0.626708791542621, + "grad_norm": 0.2937881350517273, + "learning_rate": 3.0791783764640134e-05, + "loss": 1.0637, + "step": 10315 + }, + { + "epoch": 0.6267695485752476, + "grad_norm": 0.1553046852350235, + "learning_rate": 3.078294591297432e-05, + "loss": 1.0231, + "step": 10316 + }, + { + "epoch": 0.6268303056078741, + "grad_norm": 0.17441143095493317, + "learning_rate": 3.077410876571336e-05, + "loss": 1.0538, + "step": 10317 + }, + { + "epoch": 0.6268910626405007, + "grad_norm": 0.24811109900474548, + "learning_rate": 3.076527232318116e-05, + "loss": 1.1122, + "step": 10318 + }, + { + "epoch": 0.6269518196731272, + "grad_norm": 0.29914605617523193, + "learning_rate": 3.075643658570161e-05, + "loss": 1.1012, + "step": 10319 + }, + { + "epoch": 0.6270125767057537, + "grad_norm": 0.16781513392925262, + "learning_rate": 3.0747601553598596e-05, + "loss": 1.0693, + "step": 10320 + }, + { + "epoch": 0.6270733337383803, + "grad_norm": 0.2300262451171875, + "learning_rate": 3.0738767227195976e-05, + "loss": 1.1262, + "step": 10321 + }, + { + "epoch": 0.6271340907710068, + "grad_norm": 0.22071987390518188, + "learning_rate": 3.072993360681757e-05, + "loss": 0.9952, + "step": 10322 + }, + { + "epoch": 0.6271948478036332, + "grad_norm": 0.22672142088413239, + "learning_rate": 3.0721100692787177e-05, + "loss": 1.0912, + "step": 10323 + }, + { + "epoch": 0.6272556048362598, + "grad_norm": 0.18394985795021057, + "learning_rate": 3.0712268485428544e-05, + "loss": 1.1525, + "step": 10324 + }, + { + "epoch": 0.6273163618688863, + "grad_norm": 0.358905166387558, + "learning_rate": 3.070343698506545e-05, + "loss": 1.1071, + "step": 10325 + }, + { + "epoch": 0.6273771189015128, + "grad_norm": 0.6166578531265259, + "learning_rate": 3.06946061920216e-05, + "loss": 1.2225, + "step": 10326 + }, + { + "epoch": 0.6274378759341394, + "grad_norm": 0.25921401381492615, + "learning_rate": 3.0685776106620704e-05, + "loss": 1.2086, + "step": 10327 + }, + { + "epoch": 0.6274986329667659, + "grad_norm": 0.18612873554229736, + "learning_rate": 3.0676946729186395e-05, + "loss": 1.0232, + "step": 10328 + }, + { + "epoch": 0.6275593899993924, + "grad_norm": 0.1875188648700714, + "learning_rate": 3.066811806004236e-05, + "loss": 1.0063, + "step": 10329 + }, + { + "epoch": 0.627620147032019, + "grad_norm": 0.1939394474029541, + "learning_rate": 3.065929009951219e-05, + "loss": 1.0834, + "step": 10330 + }, + { + "epoch": 0.6276809040646455, + "grad_norm": 0.15116910636425018, + "learning_rate": 3.0650462847919476e-05, + "loss": 1.013, + "step": 10331 + }, + { + "epoch": 0.627741661097272, + "grad_norm": 0.21394391357898712, + "learning_rate": 3.0641636305587776e-05, + "loss": 1.0688, + "step": 10332 + }, + { + "epoch": 0.6278024181298986, + "grad_norm": 0.3425639569759369, + "learning_rate": 3.063281047284065e-05, + "loss": 1.3683, + "step": 10333 + }, + { + "epoch": 0.6278631751625251, + "grad_norm": 0.14759787917137146, + "learning_rate": 3.0623985350001606e-05, + "loss": 1.0161, + "step": 10334 + }, + { + "epoch": 0.6279239321951516, + "grad_norm": 0.3689730763435364, + "learning_rate": 3.0615160937394105e-05, + "loss": 1.2082, + "step": 10335 + }, + { + "epoch": 0.6279846892277782, + "grad_norm": 0.2506732642650604, + "learning_rate": 3.060633723534167e-05, + "loss": 1.1904, + "step": 10336 + }, + { + "epoch": 0.6280454462604046, + "grad_norm": 0.28354790806770325, + "learning_rate": 3.0597514244167647e-05, + "loss": 1.0584, + "step": 10337 + }, + { + "epoch": 0.6281062032930311, + "grad_norm": 0.20032499730587006, + "learning_rate": 3.05886919641955e-05, + "loss": 1.0362, + "step": 10338 + }, + { + "epoch": 0.6281669603256577, + "grad_norm": 0.22291158139705658, + "learning_rate": 3.057987039574861e-05, + "loss": 1.0914, + "step": 10339 + }, + { + "epoch": 0.6282277173582842, + "grad_norm": 0.1998777687549591, + "learning_rate": 3.057104953915031e-05, + "loss": 1.158, + "step": 10340 + }, + { + "epoch": 0.6282884743909107, + "grad_norm": 0.16948993504047394, + "learning_rate": 3.0562229394723956e-05, + "loss": 1.029, + "step": 10341 + }, + { + "epoch": 0.6283492314235373, + "grad_norm": 2.924109697341919, + "learning_rate": 3.055340996279285e-05, + "loss": 1.0376, + "step": 10342 + }, + { + "epoch": 0.6284099884561638, + "grad_norm": 0.14696475863456726, + "learning_rate": 3.054459124368025e-05, + "loss": 1.0518, + "step": 10343 + }, + { + "epoch": 0.6284707454887903, + "grad_norm": 0.1799822598695755, + "learning_rate": 3.053577323770942e-05, + "loss": 1.0174, + "step": 10344 + }, + { + "epoch": 0.6285315025214169, + "grad_norm": 0.2542928457260132, + "learning_rate": 3.0526955945203576e-05, + "loss": 1.0794, + "step": 10345 + }, + { + "epoch": 0.6285922595540434, + "grad_norm": 0.2205812782049179, + "learning_rate": 3.051813936648593e-05, + "loss": 1.1783, + "step": 10346 + }, + { + "epoch": 0.6286530165866699, + "grad_norm": 0.16196352243423462, + "learning_rate": 3.050932350187966e-05, + "loss": 1.0555, + "step": 10347 + }, + { + "epoch": 0.6287137736192965, + "grad_norm": 0.21102426946163177, + "learning_rate": 3.050050835170789e-05, + "loss": 1.0418, + "step": 10348 + }, + { + "epoch": 0.628774530651923, + "grad_norm": 0.5248153805732727, + "learning_rate": 3.0491693916293783e-05, + "loss": 1.1487, + "step": 10349 + }, + { + "epoch": 0.6288352876845494, + "grad_norm": 0.20198465883731842, + "learning_rate": 3.0482880195960382e-05, + "loss": 1.0532, + "step": 10350 + }, + { + "epoch": 0.628896044717176, + "grad_norm": 0.17303092777729034, + "learning_rate": 3.0474067191030785e-05, + "loss": 1.0712, + "step": 10351 + }, + { + "epoch": 0.6289568017498025, + "grad_norm": 0.2339252382516861, + "learning_rate": 3.0465254901828032e-05, + "loss": 1.0711, + "step": 10352 + }, + { + "epoch": 0.629017558782429, + "grad_norm": 0.21205520629882812, + "learning_rate": 3.0456443328675133e-05, + "loss": 1.0668, + "step": 10353 + }, + { + "epoch": 0.6290783158150556, + "grad_norm": 0.24284236133098602, + "learning_rate": 3.0447632471895083e-05, + "loss": 1.1664, + "step": 10354 + }, + { + "epoch": 0.6291390728476821, + "grad_norm": 0.1455504298210144, + "learning_rate": 3.0438822331810846e-05, + "loss": 1.0268, + "step": 10355 + }, + { + "epoch": 0.6291998298803086, + "grad_norm": 0.24306775629520416, + "learning_rate": 3.0430012908745363e-05, + "loss": 1.0104, + "step": 10356 + }, + { + "epoch": 0.6292605869129352, + "grad_norm": 0.14489811658859253, + "learning_rate": 3.0421204203021536e-05, + "loss": 0.9991, + "step": 10357 + }, + { + "epoch": 0.6293213439455617, + "grad_norm": 0.23985731601715088, + "learning_rate": 3.041239621496224e-05, + "loss": 1.3103, + "step": 10358 + }, + { + "epoch": 0.6293821009781883, + "grad_norm": 0.19825688004493713, + "learning_rate": 3.040358894489036e-05, + "loss": 1.0579, + "step": 10359 + }, + { + "epoch": 0.6294428580108148, + "grad_norm": 0.15460100769996643, + "learning_rate": 3.0394782393128712e-05, + "loss": 1.0688, + "step": 10360 + }, + { + "epoch": 0.6295036150434413, + "grad_norm": 0.3532441556453705, + "learning_rate": 3.0385976560000096e-05, + "loss": 1.118, + "step": 10361 + }, + { + "epoch": 0.6295643720760679, + "grad_norm": 0.23219802975654602, + "learning_rate": 3.037717144582733e-05, + "loss": 1.0071, + "step": 10362 + }, + { + "epoch": 0.6296251291086943, + "grad_norm": 1.3610126972198486, + "learning_rate": 3.036836705093311e-05, + "loss": 1.0447, + "step": 10363 + }, + { + "epoch": 0.6296858861413208, + "grad_norm": 0.22628170251846313, + "learning_rate": 3.0359563375640198e-05, + "loss": 1.0029, + "step": 10364 + }, + { + "epoch": 0.6297466431739474, + "grad_norm": 0.2650076150894165, + "learning_rate": 3.0350760420271285e-05, + "loss": 1.1269, + "step": 10365 + }, + { + "epoch": 0.6298074002065739, + "grad_norm": 0.19506101310253143, + "learning_rate": 3.0341958185149043e-05, + "loss": 1.1458, + "step": 10366 + }, + { + "epoch": 0.6298681572392004, + "grad_norm": 0.18995454907417297, + "learning_rate": 3.033315667059613e-05, + "loss": 1.0704, + "step": 10367 + }, + { + "epoch": 0.629928914271827, + "grad_norm": 0.17256519198417664, + "learning_rate": 3.0324355876935163e-05, + "loss": 1.0462, + "step": 10368 + }, + { + "epoch": 0.6299896713044535, + "grad_norm": 0.13936391472816467, + "learning_rate": 3.0315555804488737e-05, + "loss": 1.0395, + "step": 10369 + }, + { + "epoch": 0.63005042833708, + "grad_norm": 0.12892702221870422, + "learning_rate": 3.0306756453579417e-05, + "loss": 0.9678, + "step": 10370 + }, + { + "epoch": 0.6301111853697066, + "grad_norm": 0.2197466641664505, + "learning_rate": 3.029795782452973e-05, + "loss": 1.1068, + "step": 10371 + }, + { + "epoch": 0.6301719424023331, + "grad_norm": 0.1719978302717209, + "learning_rate": 3.0289159917662223e-05, + "loss": 1.042, + "step": 10372 + }, + { + "epoch": 0.6302326994349596, + "grad_norm": 0.24360531568527222, + "learning_rate": 3.0280362733299362e-05, + "loss": 1.1427, + "step": 10373 + }, + { + "epoch": 0.6302934564675862, + "grad_norm": 3.230790376663208, + "learning_rate": 3.0271566271763613e-05, + "loss": 1.1461, + "step": 10374 + }, + { + "epoch": 0.6303542135002127, + "grad_norm": 0.26681554317474365, + "learning_rate": 3.026277053337742e-05, + "loss": 1.1925, + "step": 10375 + }, + { + "epoch": 0.6304149705328391, + "grad_norm": 0.16658267378807068, + "learning_rate": 3.0253975518463206e-05, + "loss": 1.0063, + "step": 10376 + }, + { + "epoch": 0.6304757275654657, + "grad_norm": 0.21557575464248657, + "learning_rate": 3.0245181227343322e-05, + "loss": 1.0731, + "step": 10377 + }, + { + "epoch": 0.6305364845980922, + "grad_norm": 0.5277007818222046, + "learning_rate": 3.023638766034014e-05, + "loss": 1.1235, + "step": 10378 + }, + { + "epoch": 0.6305972416307187, + "grad_norm": 0.16555750370025635, + "learning_rate": 3.0227594817775974e-05, + "loss": 1.0407, + "step": 10379 + }, + { + "epoch": 0.6306579986633453, + "grad_norm": 3.706537961959839, + "learning_rate": 3.0218802699973152e-05, + "loss": 1.0587, + "step": 10380 + }, + { + "epoch": 0.6307187556959718, + "grad_norm": 1.3219647407531738, + "learning_rate": 3.0210011307253945e-05, + "loss": 1.051, + "step": 10381 + }, + { + "epoch": 0.6307795127285983, + "grad_norm": 0.34503835439682007, + "learning_rate": 3.0201220639940608e-05, + "loss": 1.1472, + "step": 10382 + }, + { + "epoch": 0.6308402697612249, + "grad_norm": 0.1647673398256302, + "learning_rate": 3.0192430698355346e-05, + "loss": 1.0496, + "step": 10383 + }, + { + "epoch": 0.6309010267938514, + "grad_norm": 3.149202585220337, + "learning_rate": 3.0183641482820353e-05, + "loss": 1.1211, + "step": 10384 + }, + { + "epoch": 0.6309617838264779, + "grad_norm": 0.25413912534713745, + "learning_rate": 3.017485299365782e-05, + "loss": 1.1744, + "step": 10385 + }, + { + "epoch": 0.6310225408591045, + "grad_norm": 0.18522965908050537, + "learning_rate": 3.0166065231189888e-05, + "loss": 1.0472, + "step": 10386 + }, + { + "epoch": 0.631083297891731, + "grad_norm": 0.2938874661922455, + "learning_rate": 3.015727819573865e-05, + "loss": 1.1607, + "step": 10387 + }, + { + "epoch": 0.6311440549243575, + "grad_norm": 0.16385316848754883, + "learning_rate": 3.0148491887626235e-05, + "loss": 0.9929, + "step": 10388 + }, + { + "epoch": 0.631204811956984, + "grad_norm": 0.3231884241104126, + "learning_rate": 3.0139706307174698e-05, + "loss": 1.1225, + "step": 10389 + }, + { + "epoch": 0.6312655689896105, + "grad_norm": 0.28571072220802307, + "learning_rate": 3.013092145470605e-05, + "loss": 1.0748, + "step": 10390 + }, + { + "epoch": 0.631326326022237, + "grad_norm": 0.17621327936649323, + "learning_rate": 3.0122137330542318e-05, + "loss": 1.0589, + "step": 10391 + }, + { + "epoch": 0.6313870830548636, + "grad_norm": 0.19008980691432953, + "learning_rate": 3.0113353935005474e-05, + "loss": 1.0006, + "step": 10392 + }, + { + "epoch": 0.6314478400874901, + "grad_norm": 0.2760668098926544, + "learning_rate": 3.010457126841749e-05, + "loss": 1.1455, + "step": 10393 + }, + { + "epoch": 0.6315085971201166, + "grad_norm": 0.19350466132164001, + "learning_rate": 3.00957893311003e-05, + "loss": 1.0957, + "step": 10394 + }, + { + "epoch": 0.6315693541527432, + "grad_norm": 0.1403573602437973, + "learning_rate": 3.0087008123375787e-05, + "loss": 1.0295, + "step": 10395 + }, + { + "epoch": 0.6316301111853697, + "grad_norm": 0.20007528364658356, + "learning_rate": 3.007822764556587e-05, + "loss": 0.9887, + "step": 10396 + }, + { + "epoch": 0.6316908682179962, + "grad_norm": 0.22924551367759705, + "learning_rate": 3.0069447897992342e-05, + "loss": 1.1165, + "step": 10397 + }, + { + "epoch": 0.6317516252506228, + "grad_norm": 0.2109244465827942, + "learning_rate": 3.006066888097706e-05, + "loss": 1.0242, + "step": 10398 + }, + { + "epoch": 0.6318123822832493, + "grad_norm": 0.2021513134241104, + "learning_rate": 3.0051890594841815e-05, + "loss": 1.0501, + "step": 10399 + }, + { + "epoch": 0.6318731393158759, + "grad_norm": 0.5345461964607239, + "learning_rate": 3.0043113039908367e-05, + "loss": 1.2014, + "step": 10400 + }, + { + "epoch": 0.6319338963485024, + "grad_norm": 0.1507575511932373, + "learning_rate": 3.0034336216498477e-05, + "loss": 1.0561, + "step": 10401 + }, + { + "epoch": 0.6319946533811288, + "grad_norm": 0.21786066889762878, + "learning_rate": 3.0025560124933856e-05, + "loss": 1.0132, + "step": 10402 + }, + { + "epoch": 0.6320554104137553, + "grad_norm": 0.19458146393299103, + "learning_rate": 3.0016784765536194e-05, + "loss": 1.1476, + "step": 10403 + }, + { + "epoch": 0.6321161674463819, + "grad_norm": 0.25485432147979736, + "learning_rate": 3.0008010138627142e-05, + "loss": 1.174, + "step": 10404 + }, + { + "epoch": 0.6321769244790084, + "grad_norm": 0.2786526381969452, + "learning_rate": 2.9999236244528333e-05, + "loss": 0.9925, + "step": 10405 + }, + { + "epoch": 0.632237681511635, + "grad_norm": 0.6503855586051941, + "learning_rate": 2.999046308356139e-05, + "loss": 1.1727, + "step": 10406 + }, + { + "epoch": 0.6322984385442615, + "grad_norm": 0.2684127688407898, + "learning_rate": 2.99816906560479e-05, + "loss": 1.03, + "step": 10407 + }, + { + "epoch": 0.632359195576888, + "grad_norm": 0.31423836946487427, + "learning_rate": 2.9972918962309393e-05, + "loss": 1.0175, + "step": 10408 + }, + { + "epoch": 0.6324199526095146, + "grad_norm": 0.20827743411064148, + "learning_rate": 2.9964148002667437e-05, + "loss": 1.0332, + "step": 10409 + }, + { + "epoch": 0.6324807096421411, + "grad_norm": 0.19096548855304718, + "learning_rate": 2.9955377777443493e-05, + "loss": 1.1087, + "step": 10410 + }, + { + "epoch": 0.6325414666747676, + "grad_norm": 0.31815269589424133, + "learning_rate": 2.9946608286959056e-05, + "loss": 1.096, + "step": 10411 + }, + { + "epoch": 0.6326022237073942, + "grad_norm": 0.3105093240737915, + "learning_rate": 2.993783953153556e-05, + "loss": 1.0344, + "step": 10412 + }, + { + "epoch": 0.6326629807400207, + "grad_norm": 0.13965001702308655, + "learning_rate": 2.9929071511494433e-05, + "loss": 1.0471, + "step": 10413 + }, + { + "epoch": 0.6327237377726472, + "grad_norm": 0.22606025636196136, + "learning_rate": 2.9920304227157082e-05, + "loss": 1.0132, + "step": 10414 + }, + { + "epoch": 0.6327844948052737, + "grad_norm": 0.37282949686050415, + "learning_rate": 2.9911537678844846e-05, + "loss": 1.1137, + "step": 10415 + }, + { + "epoch": 0.6328452518379002, + "grad_norm": 0.16844916343688965, + "learning_rate": 2.9902771866879104e-05, + "loss": 1.1162, + "step": 10416 + }, + { + "epoch": 0.6329060088705267, + "grad_norm": 0.1481255143880844, + "learning_rate": 2.989400679158112e-05, + "loss": 1.0738, + "step": 10417 + }, + { + "epoch": 0.6329667659031533, + "grad_norm": 0.18795540928840637, + "learning_rate": 2.988524245327221e-05, + "loss": 1.073, + "step": 10418 + }, + { + "epoch": 0.6330275229357798, + "grad_norm": 0.36018675565719604, + "learning_rate": 2.9876478852273626e-05, + "loss": 1.2306, + "step": 10419 + }, + { + "epoch": 0.6330882799684063, + "grad_norm": 0.22278249263763428, + "learning_rate": 2.9867715988906602e-05, + "loss": 1.1534, + "step": 10420 + }, + { + "epoch": 0.6331490370010329, + "grad_norm": 0.2623753249645233, + "learning_rate": 2.9858953863492334e-05, + "loss": 1.1199, + "step": 10421 + }, + { + "epoch": 0.6332097940336594, + "grad_norm": 0.21146170794963837, + "learning_rate": 2.9850192476352006e-05, + "loss": 1.0331, + "step": 10422 + }, + { + "epoch": 0.6332705510662859, + "grad_norm": 0.24176397919654846, + "learning_rate": 2.9841431827806787e-05, + "loss": 1.0571, + "step": 10423 + }, + { + "epoch": 0.6333313080989125, + "grad_norm": 0.15834620594978333, + "learning_rate": 2.9832671918177762e-05, + "loss": 1.0438, + "step": 10424 + }, + { + "epoch": 0.633392065131539, + "grad_norm": 0.23575207591056824, + "learning_rate": 2.9823912747786043e-05, + "loss": 1.1623, + "step": 10425 + }, + { + "epoch": 0.6334528221641655, + "grad_norm": 0.36830687522888184, + "learning_rate": 2.9815154316952708e-05, + "loss": 1.0274, + "step": 10426 + }, + { + "epoch": 0.6335135791967921, + "grad_norm": 0.23382653295993805, + "learning_rate": 2.9806396625998794e-05, + "loss": 1.1273, + "step": 10427 + }, + { + "epoch": 0.6335743362294185, + "grad_norm": 0.20687885582447052, + "learning_rate": 2.9797639675245305e-05, + "loss": 1.0482, + "step": 10428 + }, + { + "epoch": 0.633635093262045, + "grad_norm": 0.3687138259410858, + "learning_rate": 2.9788883465013274e-05, + "loss": 1.0696, + "step": 10429 + }, + { + "epoch": 0.6336958502946716, + "grad_norm": 0.34979742765426636, + "learning_rate": 2.9780127995623587e-05, + "loss": 1.0403, + "step": 10430 + }, + { + "epoch": 0.6337566073272981, + "grad_norm": 0.1880851536989212, + "learning_rate": 2.977137326739723e-05, + "loss": 1.054, + "step": 10431 + }, + { + "epoch": 0.6338173643599246, + "grad_norm": 0.27019190788269043, + "learning_rate": 2.97626192806551e-05, + "loss": 1.0941, + "step": 10432 + }, + { + "epoch": 0.6338781213925512, + "grad_norm": 0.5246180891990662, + "learning_rate": 2.9753866035718058e-05, + "loss": 1.2413, + "step": 10433 + }, + { + "epoch": 0.6339388784251777, + "grad_norm": 0.4369150698184967, + "learning_rate": 2.974511353290698e-05, + "loss": 1.18, + "step": 10434 + }, + { + "epoch": 0.6339996354578042, + "grad_norm": 0.22921566665172577, + "learning_rate": 2.9736361772542687e-05, + "loss": 1.0522, + "step": 10435 + }, + { + "epoch": 0.6340603924904308, + "grad_norm": 0.2939261198043823, + "learning_rate": 2.9727610754945972e-05, + "loss": 1.0832, + "step": 10436 + }, + { + "epoch": 0.6341211495230573, + "grad_norm": 0.2623286843299866, + "learning_rate": 2.9718860480437594e-05, + "loss": 1.0758, + "step": 10437 + }, + { + "epoch": 0.6341819065556838, + "grad_norm": 0.24846018850803375, + "learning_rate": 2.9710110949338298e-05, + "loss": 0.9839, + "step": 10438 + }, + { + "epoch": 0.6342426635883104, + "grad_norm": 0.18253391981124878, + "learning_rate": 2.970136216196882e-05, + "loss": 1.0987, + "step": 10439 + }, + { + "epoch": 0.6343034206209369, + "grad_norm": 0.20280775427818298, + "learning_rate": 2.969261411864983e-05, + "loss": 1.0605, + "step": 10440 + }, + { + "epoch": 0.6343641776535635, + "grad_norm": 0.24208495020866394, + "learning_rate": 2.968386681970199e-05, + "loss": 1.0568, + "step": 10441 + }, + { + "epoch": 0.6344249346861899, + "grad_norm": 0.15895819664001465, + "learning_rate": 2.9675120265445944e-05, + "loss": 1.0663, + "step": 10442 + }, + { + "epoch": 0.6344856917188164, + "grad_norm": 0.14222745597362518, + "learning_rate": 2.9666374456202306e-05, + "loss": 1.0806, + "step": 10443 + }, + { + "epoch": 0.634546448751443, + "grad_norm": 0.20936381816864014, + "learning_rate": 2.9657629392291632e-05, + "loss": 1.0944, + "step": 10444 + }, + { + "epoch": 0.6346072057840695, + "grad_norm": 0.21725332736968994, + "learning_rate": 2.9648885074034483e-05, + "loss": 1.1289, + "step": 10445 + }, + { + "epoch": 0.634667962816696, + "grad_norm": 0.3673194944858551, + "learning_rate": 2.9640141501751384e-05, + "loss": 1.331, + "step": 10446 + }, + { + "epoch": 0.6347287198493226, + "grad_norm": 0.16266264021396637, + "learning_rate": 2.9631398675762835e-05, + "loss": 1.0118, + "step": 10447 + }, + { + "epoch": 0.6347894768819491, + "grad_norm": 0.19139760732650757, + "learning_rate": 2.9622656596389308e-05, + "loss": 1.0736, + "step": 10448 + }, + { + "epoch": 0.6348502339145756, + "grad_norm": 0.274005264043808, + "learning_rate": 2.9613915263951247e-05, + "loss": 1.0101, + "step": 10449 + }, + { + "epoch": 0.6349109909472022, + "grad_norm": 0.24060678482055664, + "learning_rate": 2.9605174678769054e-05, + "loss": 1.1097, + "step": 10450 + }, + { + "epoch": 0.6349717479798287, + "grad_norm": 0.30272573232650757, + "learning_rate": 2.9596434841163124e-05, + "loss": 1.1192, + "step": 10451 + }, + { + "epoch": 0.6350325050124552, + "grad_norm": 0.18629448115825653, + "learning_rate": 2.958769575145382e-05, + "loss": 1.1211, + "step": 10452 + }, + { + "epoch": 0.6350932620450818, + "grad_norm": 0.1965755969285965, + "learning_rate": 2.9578957409961478e-05, + "loss": 1.087, + "step": 10453 + }, + { + "epoch": 0.6351540190777083, + "grad_norm": 0.18011681735515594, + "learning_rate": 2.9570219817006384e-05, + "loss": 1.2226, + "step": 10454 + }, + { + "epoch": 0.6352147761103347, + "grad_norm": 0.20786981284618378, + "learning_rate": 2.956148297290885e-05, + "loss": 1.1677, + "step": 10455 + }, + { + "epoch": 0.6352755331429613, + "grad_norm": 0.5168510675430298, + "learning_rate": 2.9552746877989113e-05, + "loss": 1.2309, + "step": 10456 + }, + { + "epoch": 0.6353362901755878, + "grad_norm": 0.31981950998306274, + "learning_rate": 2.9544011532567385e-05, + "loss": 1.1919, + "step": 10457 + }, + { + "epoch": 0.6353970472082143, + "grad_norm": 0.19786275923252106, + "learning_rate": 2.9535276936963864e-05, + "loss": 1.1059, + "step": 10458 + }, + { + "epoch": 0.6354578042408409, + "grad_norm": 0.24380019307136536, + "learning_rate": 2.9526543091498715e-05, + "loss": 1.1379, + "step": 10459 + }, + { + "epoch": 0.6355185612734674, + "grad_norm": 0.22043776512145996, + "learning_rate": 2.9517809996492096e-05, + "loss": 1.1056, + "step": 10460 + }, + { + "epoch": 0.6355793183060939, + "grad_norm": 0.16715116798877716, + "learning_rate": 2.9509077652264106e-05, + "loss": 0.9982, + "step": 10461 + }, + { + "epoch": 0.6356400753387205, + "grad_norm": 0.3566102087497711, + "learning_rate": 2.950034605913483e-05, + "loss": 1.1912, + "step": 10462 + }, + { + "epoch": 0.635700832371347, + "grad_norm": 0.181471586227417, + "learning_rate": 2.949161521742436e-05, + "loss": 1.0679, + "step": 10463 + }, + { + "epoch": 0.6357615894039735, + "grad_norm": 0.2609163224697113, + "learning_rate": 2.9482885127452676e-05, + "loss": 1.0026, + "step": 10464 + }, + { + "epoch": 0.6358223464366001, + "grad_norm": 0.21590977907180786, + "learning_rate": 2.9474155789539803e-05, + "loss": 1.0331, + "step": 10465 + }, + { + "epoch": 0.6358831034692266, + "grad_norm": 0.27672919631004333, + "learning_rate": 2.9465427204005725e-05, + "loss": 1.2331, + "step": 10466 + }, + { + "epoch": 0.6359438605018531, + "grad_norm": 0.22714777290821075, + "learning_rate": 2.9456699371170372e-05, + "loss": 1.0911, + "step": 10467 + }, + { + "epoch": 0.6360046175344796, + "grad_norm": 0.22333866357803345, + "learning_rate": 2.9447972291353676e-05, + "loss": 1.0183, + "step": 10468 + }, + { + "epoch": 0.6360653745671061, + "grad_norm": 0.17011399567127228, + "learning_rate": 2.943924596487554e-05, + "loss": 1.1153, + "step": 10469 + }, + { + "epoch": 0.6361261315997326, + "grad_norm": 0.175033301115036, + "learning_rate": 2.9430520392055817e-05, + "loss": 1.0429, + "step": 10470 + }, + { + "epoch": 0.6361868886323592, + "grad_norm": 0.20927269756793976, + "learning_rate": 2.9421795573214337e-05, + "loss": 1.138, + "step": 10471 + }, + { + "epoch": 0.6362476456649857, + "grad_norm": 0.23329657316207886, + "learning_rate": 2.9413071508670907e-05, + "loss": 1.1094, + "step": 10472 + }, + { + "epoch": 0.6363084026976122, + "grad_norm": 0.25079572200775146, + "learning_rate": 2.9404348198745334e-05, + "loss": 1.0295, + "step": 10473 + }, + { + "epoch": 0.6363691597302388, + "grad_norm": 0.16970273852348328, + "learning_rate": 2.9395625643757365e-05, + "loss": 1.0696, + "step": 10474 + }, + { + "epoch": 0.6364299167628653, + "grad_norm": 0.22155103087425232, + "learning_rate": 2.93869038440267e-05, + "loss": 1.1434, + "step": 10475 + }, + { + "epoch": 0.6364906737954918, + "grad_norm": 0.19520479440689087, + "learning_rate": 2.9378182799873094e-05, + "loss": 1.1095, + "step": 10476 + }, + { + "epoch": 0.6365514308281184, + "grad_norm": 0.34213370084762573, + "learning_rate": 2.9369462511616154e-05, + "loss": 1.1218, + "step": 10477 + }, + { + "epoch": 0.6366121878607449, + "grad_norm": 0.18259897828102112, + "learning_rate": 2.936074297957556e-05, + "loss": 1.1138, + "step": 10478 + }, + { + "epoch": 0.6366729448933715, + "grad_norm": 0.29943394660949707, + "learning_rate": 2.9352024204070927e-05, + "loss": 1.0347, + "step": 10479 + }, + { + "epoch": 0.636733701925998, + "grad_norm": 0.27895477414131165, + "learning_rate": 2.934330618542182e-05, + "loss": 1.208, + "step": 10480 + }, + { + "epoch": 0.6367944589586244, + "grad_norm": 0.24743030965328217, + "learning_rate": 2.9334588923947838e-05, + "loss": 1.0078, + "step": 10481 + }, + { + "epoch": 0.636855215991251, + "grad_norm": 0.27887317538261414, + "learning_rate": 2.932587241996848e-05, + "loss": 1.0658, + "step": 10482 + }, + { + "epoch": 0.6369159730238775, + "grad_norm": 11.776100158691406, + "learning_rate": 2.9317156673803282e-05, + "loss": 1.0528, + "step": 10483 + }, + { + "epoch": 0.636976730056504, + "grad_norm": 0.4576953947544098, + "learning_rate": 2.9308441685771692e-05, + "loss": 1.0197, + "step": 10484 + }, + { + "epoch": 0.6370374870891305, + "grad_norm": 0.2606697976589203, + "learning_rate": 2.9299727456193172e-05, + "loss": 1.1538, + "step": 10485 + }, + { + "epoch": 0.6370982441217571, + "grad_norm": 0.21457089483737946, + "learning_rate": 2.929101398538714e-05, + "loss": 1.0659, + "step": 10486 + }, + { + "epoch": 0.6371590011543836, + "grad_norm": 0.1925864815711975, + "learning_rate": 2.9282301273672996e-05, + "loss": 1.0166, + "step": 10487 + }, + { + "epoch": 0.6372197581870102, + "grad_norm": 0.26542410254478455, + "learning_rate": 2.9273589321370098e-05, + "loss": 1.0924, + "step": 10488 + }, + { + "epoch": 0.6372805152196367, + "grad_norm": 0.19813016057014465, + "learning_rate": 2.9264878128797802e-05, + "loss": 1.0827, + "step": 10489 + }, + { + "epoch": 0.6373412722522632, + "grad_norm": 0.27986443042755127, + "learning_rate": 2.925616769627541e-05, + "loss": 1.0324, + "step": 10490 + }, + { + "epoch": 0.6374020292848898, + "grad_norm": 0.15699289739131927, + "learning_rate": 2.9247458024122193e-05, + "loss": 1.0217, + "step": 10491 + }, + { + "epoch": 0.6374627863175163, + "grad_norm": 0.21668890118598938, + "learning_rate": 2.9238749112657414e-05, + "loss": 1.0848, + "step": 10492 + }, + { + "epoch": 0.6375235433501428, + "grad_norm": 0.17873866856098175, + "learning_rate": 2.923004096220029e-05, + "loss": 1.0169, + "step": 10493 + }, + { + "epoch": 0.6375843003827693, + "grad_norm": 0.1731361746788025, + "learning_rate": 2.922133357307003e-05, + "loss": 1.0905, + "step": 10494 + }, + { + "epoch": 0.6376450574153958, + "grad_norm": 0.21488367021083832, + "learning_rate": 2.921262694558582e-05, + "loss": 1.0656, + "step": 10495 + }, + { + "epoch": 0.6377058144480223, + "grad_norm": 0.16641485691070557, + "learning_rate": 2.9203921080066786e-05, + "loss": 0.9826, + "step": 10496 + }, + { + "epoch": 0.6377665714806489, + "grad_norm": 0.2193039059638977, + "learning_rate": 2.9195215976832034e-05, + "loss": 1.1042, + "step": 10497 + }, + { + "epoch": 0.6378273285132754, + "grad_norm": 0.14269620180130005, + "learning_rate": 2.9186511636200666e-05, + "loss": 1.0595, + "step": 10498 + }, + { + "epoch": 0.6378880855459019, + "grad_norm": 0.2796608507633209, + "learning_rate": 2.917780805849173e-05, + "loss": 1.1199, + "step": 10499 + }, + { + "epoch": 0.6379488425785285, + "grad_norm": 0.24705006182193756, + "learning_rate": 2.9169105244024275e-05, + "loss": 1.122, + "step": 10500 + }, + { + "epoch": 0.638009599611155, + "grad_norm": 0.18371637165546417, + "learning_rate": 2.9160403193117292e-05, + "loss": 1.1006, + "step": 10501 + }, + { + "epoch": 0.6380703566437815, + "grad_norm": 0.545051097869873, + "learning_rate": 2.9151701906089735e-05, + "loss": 1.2048, + "step": 10502 + }, + { + "epoch": 0.6381311136764081, + "grad_norm": 0.17087338864803314, + "learning_rate": 2.9143001383260614e-05, + "loss": 1.1946, + "step": 10503 + }, + { + "epoch": 0.6381918707090346, + "grad_norm": 0.17378097772598267, + "learning_rate": 2.913430162494879e-05, + "loss": 1.0514, + "step": 10504 + }, + { + "epoch": 0.6382526277416611, + "grad_norm": 0.18431510031223297, + "learning_rate": 2.9125602631473176e-05, + "loss": 1.0299, + "step": 10505 + }, + { + "epoch": 0.6383133847742877, + "grad_norm": 0.2168143391609192, + "learning_rate": 2.911690440315263e-05, + "loss": 1.1374, + "step": 10506 + }, + { + "epoch": 0.6383741418069141, + "grad_norm": 0.17916178703308105, + "learning_rate": 2.9108206940305972e-05, + "loss": 1.0721, + "step": 10507 + }, + { + "epoch": 0.6384348988395406, + "grad_norm": 0.1370878517627716, + "learning_rate": 2.9099510243252042e-05, + "loss": 1.0238, + "step": 10508 + }, + { + "epoch": 0.6384956558721672, + "grad_norm": 0.14136061072349548, + "learning_rate": 2.909081431230961e-05, + "loss": 1.0493, + "step": 10509 + }, + { + "epoch": 0.6385564129047937, + "grad_norm": 0.1441960334777832, + "learning_rate": 2.9082119147797437e-05, + "loss": 1.0298, + "step": 10510 + }, + { + "epoch": 0.6386171699374202, + "grad_norm": 0.13976144790649414, + "learning_rate": 2.907342475003421e-05, + "loss": 1.0534, + "step": 10511 + }, + { + "epoch": 0.6386779269700468, + "grad_norm": 0.15226629376411438, + "learning_rate": 2.906473111933863e-05, + "loss": 1.2284, + "step": 10512 + }, + { + "epoch": 0.6387386840026733, + "grad_norm": 3.459918975830078, + "learning_rate": 2.9056038256029388e-05, + "loss": 1.1127, + "step": 10513 + }, + { + "epoch": 0.6387994410352998, + "grad_norm": 0.1456443816423416, + "learning_rate": 2.9047346160425114e-05, + "loss": 1.0495, + "step": 10514 + }, + { + "epoch": 0.6388601980679264, + "grad_norm": 0.28783121705055237, + "learning_rate": 2.9038654832844425e-05, + "loss": 1.0764, + "step": 10515 + }, + { + "epoch": 0.6389209551005529, + "grad_norm": 0.1500846892595291, + "learning_rate": 2.9029964273605902e-05, + "loss": 1.0399, + "step": 10516 + }, + { + "epoch": 0.6389817121331794, + "grad_norm": 0.2088969200849533, + "learning_rate": 2.902127448302806e-05, + "loss": 1.0634, + "step": 10517 + }, + { + "epoch": 0.639042469165806, + "grad_norm": 0.20981422066688538, + "learning_rate": 2.9012585461429475e-05, + "loss": 1.0278, + "step": 10518 + }, + { + "epoch": 0.6391032261984325, + "grad_norm": 0.14799395203590393, + "learning_rate": 2.9003897209128627e-05, + "loss": 1.0114, + "step": 10519 + }, + { + "epoch": 0.6391639832310589, + "grad_norm": 0.12527261674404144, + "learning_rate": 2.8995209726443988e-05, + "loss": 1.0022, + "step": 10520 + }, + { + "epoch": 0.6392247402636855, + "grad_norm": 0.6986045241355896, + "learning_rate": 2.8986523013694e-05, + "loss": 1.0523, + "step": 10521 + }, + { + "epoch": 0.639285497296312, + "grad_norm": 2.4540939331054688, + "learning_rate": 2.8977837071197067e-05, + "loss": 1.1382, + "step": 10522 + }, + { + "epoch": 0.6393462543289385, + "grad_norm": 0.17050419747829437, + "learning_rate": 2.8969151899271573e-05, + "loss": 1.0568, + "step": 10523 + }, + { + "epoch": 0.6394070113615651, + "grad_norm": 0.21459123492240906, + "learning_rate": 2.8960467498235887e-05, + "loss": 1.0887, + "step": 10524 + }, + { + "epoch": 0.6394677683941916, + "grad_norm": 0.16448813676834106, + "learning_rate": 2.895178386840834e-05, + "loss": 1.0222, + "step": 10525 + }, + { + "epoch": 0.6395285254268182, + "grad_norm": 0.2933340072631836, + "learning_rate": 2.8943101010107214e-05, + "loss": 1.1288, + "step": 10526 + }, + { + "epoch": 0.6395892824594447, + "grad_norm": 0.2306685596704483, + "learning_rate": 2.8934418923650795e-05, + "loss": 1.0219, + "step": 10527 + }, + { + "epoch": 0.6396500394920712, + "grad_norm": 0.2205544412136078, + "learning_rate": 2.8925737609357296e-05, + "loss": 1.0167, + "step": 10528 + }, + { + "epoch": 0.6397107965246978, + "grad_norm": 2.6001367568969727, + "learning_rate": 2.8917057067544983e-05, + "loss": 1.1208, + "step": 10529 + }, + { + "epoch": 0.6397715535573243, + "grad_norm": 0.267107218503952, + "learning_rate": 2.8908377298532042e-05, + "loss": 1.0691, + "step": 10530 + }, + { + "epoch": 0.6398323105899508, + "grad_norm": 0.26263532042503357, + "learning_rate": 2.8899698302636586e-05, + "loss": 1.1598, + "step": 10531 + }, + { + "epoch": 0.6398930676225774, + "grad_norm": 0.19372712075710297, + "learning_rate": 2.8891020080176767e-05, + "loss": 1.0785, + "step": 10532 + }, + { + "epoch": 0.6399538246552038, + "grad_norm": 0.21181721985340118, + "learning_rate": 2.888234263147066e-05, + "loss": 1.0545, + "step": 10533 + }, + { + "epoch": 0.6400145816878303, + "grad_norm": 0.20640496909618378, + "learning_rate": 2.887366595683639e-05, + "loss": 1.0757, + "step": 10534 + }, + { + "epoch": 0.6400753387204569, + "grad_norm": 0.27535438537597656, + "learning_rate": 2.8864990056591984e-05, + "loss": 1.187, + "step": 10535 + }, + { + "epoch": 0.6401360957530834, + "grad_norm": 3.1241238117218018, + "learning_rate": 2.885631493105545e-05, + "loss": 1.0769, + "step": 10536 + }, + { + "epoch": 0.6401968527857099, + "grad_norm": 0.31507977843284607, + "learning_rate": 2.88476405805448e-05, + "loss": 1.2688, + "step": 10537 + }, + { + "epoch": 0.6402576098183365, + "grad_norm": 0.18061679601669312, + "learning_rate": 2.8838967005377927e-05, + "loss": 1.1098, + "step": 10538 + }, + { + "epoch": 0.640318366850963, + "grad_norm": 0.24573034048080444, + "learning_rate": 2.8830294205872833e-05, + "loss": 1.1624, + "step": 10539 + }, + { + "epoch": 0.6403791238835895, + "grad_norm": 0.3537411391735077, + "learning_rate": 2.8821622182347395e-05, + "loss": 1.1502, + "step": 10540 + }, + { + "epoch": 0.6404398809162161, + "grad_norm": 0.20023702085018158, + "learning_rate": 2.8812950935119498e-05, + "loss": 1.0958, + "step": 10541 + }, + { + "epoch": 0.6405006379488426, + "grad_norm": 0.23242689669132233, + "learning_rate": 2.8804280464506973e-05, + "loss": 1.1428, + "step": 10542 + }, + { + "epoch": 0.6405613949814691, + "grad_norm": 0.1741657853126526, + "learning_rate": 2.8795610770827643e-05, + "loss": 1.0519, + "step": 10543 + }, + { + "epoch": 0.6406221520140957, + "grad_norm": 0.1436629444360733, + "learning_rate": 2.8786941854399297e-05, + "loss": 1.0456, + "step": 10544 + }, + { + "epoch": 0.6406829090467222, + "grad_norm": 0.1663539558649063, + "learning_rate": 2.8778273715539706e-05, + "loss": 1.0853, + "step": 10545 + }, + { + "epoch": 0.6407436660793487, + "grad_norm": 0.13799241185188293, + "learning_rate": 2.8769606354566585e-05, + "loss": 1.0829, + "step": 10546 + }, + { + "epoch": 0.6408044231119752, + "grad_norm": 0.19414234161376953, + "learning_rate": 2.8760939771797656e-05, + "loss": 1.0779, + "step": 10547 + }, + { + "epoch": 0.6408651801446017, + "grad_norm": 0.19542182981967926, + "learning_rate": 2.875227396755058e-05, + "loss": 1.077, + "step": 10548 + }, + { + "epoch": 0.6409259371772282, + "grad_norm": 0.2462919056415558, + "learning_rate": 2.8743608942142986e-05, + "loss": 1.1231, + "step": 10549 + }, + { + "epoch": 0.6409866942098548, + "grad_norm": 0.2779509723186493, + "learning_rate": 2.8734944695892564e-05, + "loss": 1.1776, + "step": 10550 + }, + { + "epoch": 0.6410474512424813, + "grad_norm": 2.1580617427825928, + "learning_rate": 2.8726281229116826e-05, + "loss": 1.0829, + "step": 10551 + }, + { + "epoch": 0.6411082082751078, + "grad_norm": 0.20197483897209167, + "learning_rate": 2.8717618542133362e-05, + "loss": 1.0887, + "step": 10552 + }, + { + "epoch": 0.6411689653077344, + "grad_norm": 0.26988357305526733, + "learning_rate": 2.870895663525971e-05, + "loss": 1.0227, + "step": 10553 + }, + { + "epoch": 0.6412297223403609, + "grad_norm": 1.6526679992675781, + "learning_rate": 2.8700295508813342e-05, + "loss": 1.2466, + "step": 10554 + }, + { + "epoch": 0.6412904793729874, + "grad_norm": 0.22585134208202362, + "learning_rate": 2.869163516311178e-05, + "loss": 1.1099, + "step": 10555 + }, + { + "epoch": 0.641351236405614, + "grad_norm": 0.6732856035232544, + "learning_rate": 2.868297559847245e-05, + "loss": 1.0901, + "step": 10556 + }, + { + "epoch": 0.6414119934382405, + "grad_norm": 0.20253092050552368, + "learning_rate": 2.8674316815212797e-05, + "loss": 1.0721, + "step": 10557 + }, + { + "epoch": 0.641472750470867, + "grad_norm": 0.2992306649684906, + "learning_rate": 2.8665658813650153e-05, + "loss": 1.3024, + "step": 10558 + }, + { + "epoch": 0.6415335075034936, + "grad_norm": 0.1389251947402954, + "learning_rate": 2.8657001594101895e-05, + "loss": 1.0472, + "step": 10559 + }, + { + "epoch": 0.64159426453612, + "grad_norm": 0.26110389828681946, + "learning_rate": 2.864834515688539e-05, + "loss": 1.0948, + "step": 10560 + }, + { + "epoch": 0.6416550215687465, + "grad_norm": 0.16514131426811218, + "learning_rate": 2.8639689502317917e-05, + "loss": 1.0646, + "step": 10561 + }, + { + "epoch": 0.6417157786013731, + "grad_norm": 0.3476346433162689, + "learning_rate": 2.863103463071676e-05, + "loss": 1.0125, + "step": 10562 + }, + { + "epoch": 0.6417765356339996, + "grad_norm": 0.13853426277637482, + "learning_rate": 2.862238054239915e-05, + "loss": 1.0696, + "step": 10563 + }, + { + "epoch": 0.6418372926666261, + "grad_norm": 0.15854158997535706, + "learning_rate": 2.861372723768232e-05, + "loss": 1.039, + "step": 10564 + }, + { + "epoch": 0.6418980496992527, + "grad_norm": 0.13335920870304108, + "learning_rate": 2.8605074716883452e-05, + "loss": 1.0486, + "step": 10565 + }, + { + "epoch": 0.6419588067318792, + "grad_norm": 0.16531626880168915, + "learning_rate": 2.8596422980319704e-05, + "loss": 1.0436, + "step": 10566 + }, + { + "epoch": 0.6420195637645058, + "grad_norm": 0.15582625567913055, + "learning_rate": 2.8587772028308208e-05, + "loss": 1.0141, + "step": 10567 + }, + { + "epoch": 0.6420803207971323, + "grad_norm": 2.886087417602539, + "learning_rate": 2.857912186116607e-05, + "loss": 1.069, + "step": 10568 + }, + { + "epoch": 0.6421410778297588, + "grad_norm": 0.27575862407684326, + "learning_rate": 2.8570472479210354e-05, + "loss": 0.989, + "step": 10569 + }, + { + "epoch": 0.6422018348623854, + "grad_norm": 0.31458139419555664, + "learning_rate": 2.856182388275812e-05, + "loss": 0.9971, + "step": 10570 + }, + { + "epoch": 0.6422625918950119, + "grad_norm": 0.16712796688079834, + "learning_rate": 2.855317607212637e-05, + "loss": 1.0364, + "step": 10571 + }, + { + "epoch": 0.6423233489276384, + "grad_norm": 0.19279715418815613, + "learning_rate": 2.8544529047632095e-05, + "loss": 1.1397, + "step": 10572 + }, + { + "epoch": 0.6423841059602649, + "grad_norm": 0.2562878131866455, + "learning_rate": 2.853588280959226e-05, + "loss": 0.9937, + "step": 10573 + }, + { + "epoch": 0.6424448629928914, + "grad_norm": 0.17854396998882294, + "learning_rate": 2.8527237358323768e-05, + "loss": 1.0986, + "step": 10574 + }, + { + "epoch": 0.6425056200255179, + "grad_norm": 2.2411932945251465, + "learning_rate": 2.8518592694143564e-05, + "loss": 1.0866, + "step": 10575 + }, + { + "epoch": 0.6425663770581445, + "grad_norm": 0.1470174789428711, + "learning_rate": 2.8509948817368492e-05, + "loss": 1.0742, + "step": 10576 + }, + { + "epoch": 0.642627134090771, + "grad_norm": 0.15624035894870758, + "learning_rate": 2.850130572831542e-05, + "loss": 1.0394, + "step": 10577 + }, + { + "epoch": 0.6426878911233975, + "grad_norm": 0.20398494601249695, + "learning_rate": 2.8492663427301127e-05, + "loss": 1.0353, + "step": 10578 + }, + { + "epoch": 0.6427486481560241, + "grad_norm": 0.24344800412654877, + "learning_rate": 2.848402191464239e-05, + "loss": 1.134, + "step": 10579 + }, + { + "epoch": 0.6428094051886506, + "grad_norm": 0.9971254467964172, + "learning_rate": 2.8475381190656013e-05, + "loss": 1.0954, + "step": 10580 + }, + { + "epoch": 0.6428701622212771, + "grad_norm": 0.16742849349975586, + "learning_rate": 2.8466741255658693e-05, + "loss": 1.1294, + "step": 10581 + }, + { + "epoch": 0.6429309192539037, + "grad_norm": 0.283002644777298, + "learning_rate": 2.8458102109967133e-05, + "loss": 1.2108, + "step": 10582 + }, + { + "epoch": 0.6429916762865302, + "grad_norm": 0.22541570663452148, + "learning_rate": 2.844946375389803e-05, + "loss": 1.1209, + "step": 10583 + }, + { + "epoch": 0.6430524333191567, + "grad_norm": 2.935861587524414, + "learning_rate": 2.844082618776796e-05, + "loss": 1.0855, + "step": 10584 + }, + { + "epoch": 0.6431131903517833, + "grad_norm": 0.19058136641979218, + "learning_rate": 2.843218941189359e-05, + "loss": 1.081, + "step": 10585 + }, + { + "epoch": 0.6431739473844097, + "grad_norm": 0.275488018989563, + "learning_rate": 2.8423553426591486e-05, + "loss": 1.0881, + "step": 10586 + }, + { + "epoch": 0.6432347044170362, + "grad_norm": 0.13175101578235626, + "learning_rate": 2.84149182321782e-05, + "loss": 0.9927, + "step": 10587 + }, + { + "epoch": 0.6432954614496628, + "grad_norm": 0.20485453307628632, + "learning_rate": 2.840628382897026e-05, + "loss": 1.072, + "step": 10588 + }, + { + "epoch": 0.6433562184822893, + "grad_norm": 0.2235424965620041, + "learning_rate": 2.8397650217284165e-05, + "loss": 1.1719, + "step": 10589 + }, + { + "epoch": 0.6434169755149158, + "grad_norm": 0.19105499982833862, + "learning_rate": 2.838901739743637e-05, + "loss": 0.997, + "step": 10590 + }, + { + "epoch": 0.6434777325475424, + "grad_norm": 0.21691671013832092, + "learning_rate": 2.8380385369743325e-05, + "loss": 1.1267, + "step": 10591 + }, + { + "epoch": 0.6435384895801689, + "grad_norm": 0.14863766729831696, + "learning_rate": 2.837175413452143e-05, + "loss": 1.0837, + "step": 10592 + }, + { + "epoch": 0.6435992466127954, + "grad_norm": 0.24956850707530975, + "learning_rate": 2.836312369208707e-05, + "loss": 1.0515, + "step": 10593 + }, + { + "epoch": 0.643660003645422, + "grad_norm": 0.14244811236858368, + "learning_rate": 2.835449404275659e-05, + "loss": 1.0155, + "step": 10594 + }, + { + "epoch": 0.6437207606780485, + "grad_norm": 0.14746737480163574, + "learning_rate": 2.83458651868463e-05, + "loss": 1.0048, + "step": 10595 + }, + { + "epoch": 0.643781517710675, + "grad_norm": 0.2184237390756607, + "learning_rate": 2.833723712467252e-05, + "loss": 1.1635, + "step": 10596 + }, + { + "epoch": 0.6438422747433016, + "grad_norm": 0.19648896157741547, + "learning_rate": 2.8328609856551524e-05, + "loss": 0.9968, + "step": 10597 + }, + { + "epoch": 0.6439030317759281, + "grad_norm": 0.18149517476558685, + "learning_rate": 2.83199833827995e-05, + "loss": 1.0613, + "step": 10598 + }, + { + "epoch": 0.6439637888085545, + "grad_norm": 0.20802715420722961, + "learning_rate": 2.8311357703732683e-05, + "loss": 1.0998, + "step": 10599 + }, + { + "epoch": 0.6440245458411811, + "grad_norm": 0.5641850829124451, + "learning_rate": 2.830273281966722e-05, + "loss": 1.0677, + "step": 10600 + }, + { + "epoch": 0.6440853028738076, + "grad_norm": 0.13975229859352112, + "learning_rate": 2.8294108730919295e-05, + "loss": 1.0522, + "step": 10601 + }, + { + "epoch": 0.6441460599064341, + "grad_norm": 2.315631151199341, + "learning_rate": 2.8285485437805016e-05, + "loss": 1.1565, + "step": 10602 + }, + { + "epoch": 0.6442068169390607, + "grad_norm": 0.16507162153720856, + "learning_rate": 2.827686294064048e-05, + "loss": 1.0438, + "step": 10603 + }, + { + "epoch": 0.6442675739716872, + "grad_norm": 0.7756314873695374, + "learning_rate": 2.8268241239741712e-05, + "loss": 1.1539, + "step": 10604 + }, + { + "epoch": 0.6443283310043137, + "grad_norm": 0.15813900530338287, + "learning_rate": 2.825962033542474e-05, + "loss": 1.086, + "step": 10605 + }, + { + "epoch": 0.6443890880369403, + "grad_norm": 0.24553470313549042, + "learning_rate": 2.8251000228005614e-05, + "loss": 1.099, + "step": 10606 + }, + { + "epoch": 0.6444498450695668, + "grad_norm": 0.14088580012321472, + "learning_rate": 2.8242380917800278e-05, + "loss": 1.0038, + "step": 10607 + }, + { + "epoch": 0.6445106021021934, + "grad_norm": 0.1802026480436325, + "learning_rate": 2.823376240512467e-05, + "loss": 1.0699, + "step": 10608 + }, + { + "epoch": 0.6445713591348199, + "grad_norm": 0.21146097779273987, + "learning_rate": 2.8225144690294714e-05, + "loss": 1.071, + "step": 10609 + }, + { + "epoch": 0.6446321161674464, + "grad_norm": 0.43120068311691284, + "learning_rate": 2.8216527773626277e-05, + "loss": 1.3179, + "step": 10610 + }, + { + "epoch": 0.644692873200073, + "grad_norm": 0.18573731184005737, + "learning_rate": 2.8207911655435238e-05, + "loss": 1.1409, + "step": 10611 + }, + { + "epoch": 0.6447536302326994, + "grad_norm": 0.198276549577713, + "learning_rate": 2.8199296336037406e-05, + "loss": 1.0698, + "step": 10612 + }, + { + "epoch": 0.6448143872653259, + "grad_norm": 0.2512544095516205, + "learning_rate": 2.8190681815748577e-05, + "loss": 0.9989, + "step": 10613 + }, + { + "epoch": 0.6448751442979525, + "grad_norm": 0.174990713596344, + "learning_rate": 2.8182068094884527e-05, + "loss": 1.1116, + "step": 10614 + }, + { + "epoch": 0.644935901330579, + "grad_norm": 0.21055661141872406, + "learning_rate": 2.8173455173760987e-05, + "loss": 1.1876, + "step": 10615 + }, + { + "epoch": 0.6449966583632055, + "grad_norm": 0.3719698488712311, + "learning_rate": 2.8164843052693645e-05, + "loss": 0.999, + "step": 10616 + }, + { + "epoch": 0.6450574153958321, + "grad_norm": 0.6674947738647461, + "learning_rate": 2.815623173199824e-05, + "loss": 1.1726, + "step": 10617 + }, + { + "epoch": 0.6451181724284586, + "grad_norm": 0.1900763362646103, + "learning_rate": 2.814762121199036e-05, + "loss": 1.0677, + "step": 10618 + }, + { + "epoch": 0.6451789294610851, + "grad_norm": 0.3120037913322449, + "learning_rate": 2.8139011492985657e-05, + "loss": 0.9995, + "step": 10619 + }, + { + "epoch": 0.6452396864937117, + "grad_norm": 0.19436316192150116, + "learning_rate": 2.813040257529971e-05, + "loss": 1.1655, + "step": 10620 + }, + { + "epoch": 0.6453004435263382, + "grad_norm": 0.22700625658035278, + "learning_rate": 2.8121794459248063e-05, + "loss": 1.1123, + "step": 10621 + }, + { + "epoch": 0.6453612005589647, + "grad_norm": 0.18110990524291992, + "learning_rate": 2.811318714514629e-05, + "loss": 1.0206, + "step": 10622 + }, + { + "epoch": 0.6454219575915913, + "grad_norm": 3.4883077144622803, + "learning_rate": 2.8104580633309872e-05, + "loss": 1.0241, + "step": 10623 + }, + { + "epoch": 0.6454827146242178, + "grad_norm": 0.16194044053554535, + "learning_rate": 2.8095974924054303e-05, + "loss": 1.0837, + "step": 10624 + }, + { + "epoch": 0.6455434716568442, + "grad_norm": 0.14323647320270538, + "learning_rate": 2.8087370017694992e-05, + "loss": 1.0427, + "step": 10625 + }, + { + "epoch": 0.6456042286894708, + "grad_norm": 0.1778782308101654, + "learning_rate": 2.8078765914547345e-05, + "loss": 1.0857, + "step": 10626 + }, + { + "epoch": 0.6456649857220973, + "grad_norm": 0.22986264526844025, + "learning_rate": 2.8070162614926787e-05, + "loss": 1.145, + "step": 10627 + }, + { + "epoch": 0.6457257427547238, + "grad_norm": 0.20973241329193115, + "learning_rate": 2.8061560119148662e-05, + "loss": 1.0565, + "step": 10628 + }, + { + "epoch": 0.6457864997873504, + "grad_norm": 0.13414113223552704, + "learning_rate": 2.8052958427528285e-05, + "loss": 0.9844, + "step": 10629 + }, + { + "epoch": 0.6458472568199769, + "grad_norm": 0.18374726176261902, + "learning_rate": 2.8044357540380984e-05, + "loss": 1.044, + "step": 10630 + }, + { + "epoch": 0.6459080138526034, + "grad_norm": 0.18679040670394897, + "learning_rate": 2.803575745802196e-05, + "loss": 1.0756, + "step": 10631 + }, + { + "epoch": 0.64596877088523, + "grad_norm": 0.19506610929965973, + "learning_rate": 2.8027158180766518e-05, + "loss": 1.0639, + "step": 10632 + }, + { + "epoch": 0.6460295279178565, + "grad_norm": 0.3637720048427582, + "learning_rate": 2.8018559708929837e-05, + "loss": 1.0504, + "step": 10633 + }, + { + "epoch": 0.646090284950483, + "grad_norm": 0.1827133595943451, + "learning_rate": 2.8009962042827098e-05, + "loss": 1.0662, + "step": 10634 + }, + { + "epoch": 0.6461510419831096, + "grad_norm": 0.8316200971603394, + "learning_rate": 2.8001365182773447e-05, + "loss": 1.1236, + "step": 10635 + }, + { + "epoch": 0.6462117990157361, + "grad_norm": 0.25573089718818665, + "learning_rate": 2.7992769129084005e-05, + "loss": 1.1083, + "step": 10636 + }, + { + "epoch": 0.6462725560483626, + "grad_norm": 0.24525128304958344, + "learning_rate": 2.7984173882073872e-05, + "loss": 1.0189, + "step": 10637 + }, + { + "epoch": 0.6463333130809891, + "grad_norm": 0.31146132946014404, + "learning_rate": 2.7975579442058102e-05, + "loss": 1.0097, + "step": 10638 + }, + { + "epoch": 0.6463940701136156, + "grad_norm": 0.25546589493751526, + "learning_rate": 2.796698580935172e-05, + "loss": 1.0849, + "step": 10639 + }, + { + "epoch": 0.6464548271462421, + "grad_norm": 0.894803524017334, + "learning_rate": 2.795839298426973e-05, + "loss": 1.1175, + "step": 10640 + }, + { + "epoch": 0.6465155841788687, + "grad_norm": 0.2752804458141327, + "learning_rate": 2.7949800967127116e-05, + "loss": 1.1628, + "step": 10641 + }, + { + "epoch": 0.6465763412114952, + "grad_norm": 0.6775246262550354, + "learning_rate": 2.7941209758238786e-05, + "loss": 1.234, + "step": 10642 + }, + { + "epoch": 0.6466370982441217, + "grad_norm": 0.21429786086082458, + "learning_rate": 2.7932619357919687e-05, + "loss": 1.133, + "step": 10643 + }, + { + "epoch": 0.6466978552767483, + "grad_norm": 0.14917413890361786, + "learning_rate": 2.7924029766484717e-05, + "loss": 1.0198, + "step": 10644 + }, + { + "epoch": 0.6467586123093748, + "grad_norm": 0.21316313743591309, + "learning_rate": 2.7915440984248686e-05, + "loss": 1.0283, + "step": 10645 + }, + { + "epoch": 0.6468193693420013, + "grad_norm": 0.21296539902687073, + "learning_rate": 2.7906853011526435e-05, + "loss": 1.0907, + "step": 10646 + }, + { + "epoch": 0.6468801263746279, + "grad_norm": 0.16633765399456024, + "learning_rate": 2.7898265848632733e-05, + "loss": 1.0838, + "step": 10647 + }, + { + "epoch": 0.6469408834072544, + "grad_norm": 0.6495085954666138, + "learning_rate": 2.7889679495882385e-05, + "loss": 1.0541, + "step": 10648 + }, + { + "epoch": 0.647001640439881, + "grad_norm": 0.2259526252746582, + "learning_rate": 2.788109395359012e-05, + "loss": 1.0873, + "step": 10649 + }, + { + "epoch": 0.6470623974725075, + "grad_norm": 0.17805831134319305, + "learning_rate": 2.7872509222070642e-05, + "loss": 1.0759, + "step": 10650 + }, + { + "epoch": 0.647123154505134, + "grad_norm": 0.31398478150367737, + "learning_rate": 2.7863925301638598e-05, + "loss": 1.0644, + "step": 10651 + }, + { + "epoch": 0.6471839115377604, + "grad_norm": 0.20052233338356018, + "learning_rate": 2.7855342192608637e-05, + "loss": 1.165, + "step": 10652 + }, + { + "epoch": 0.647244668570387, + "grad_norm": 0.14741869270801544, + "learning_rate": 2.7846759895295397e-05, + "loss": 1.0258, + "step": 10653 + }, + { + "epoch": 0.6473054256030135, + "grad_norm": 0.15443217754364014, + "learning_rate": 2.783817841001346e-05, + "loss": 1.0823, + "step": 10654 + }, + { + "epoch": 0.64736618263564, + "grad_norm": 0.14515122771263123, + "learning_rate": 2.7829597737077385e-05, + "loss": 1.0846, + "step": 10655 + }, + { + "epoch": 0.6474269396682666, + "grad_norm": 0.174307718873024, + "learning_rate": 2.782101787680168e-05, + "loss": 1.0551, + "step": 10656 + }, + { + "epoch": 0.6474876967008931, + "grad_norm": 0.19094575941562653, + "learning_rate": 2.781243882950085e-05, + "loss": 1.0566, + "step": 10657 + }, + { + "epoch": 0.6475484537335197, + "grad_norm": 0.15695898234844208, + "learning_rate": 2.7803860595489373e-05, + "loss": 1.1067, + "step": 10658 + }, + { + "epoch": 0.6476092107661462, + "grad_norm": 0.1791105419397354, + "learning_rate": 2.7795283175081678e-05, + "loss": 1.1023, + "step": 10659 + }, + { + "epoch": 0.6476699677987727, + "grad_norm": 0.17872925102710724, + "learning_rate": 2.7786706568592163e-05, + "loss": 1.1592, + "step": 10660 + }, + { + "epoch": 0.6477307248313993, + "grad_norm": 0.23921625316143036, + "learning_rate": 2.7778130776335217e-05, + "loss": 1.0509, + "step": 10661 + }, + { + "epoch": 0.6477914818640258, + "grad_norm": 0.280554860830307, + "learning_rate": 2.776955579862518e-05, + "loss": 1.0445, + "step": 10662 + }, + { + "epoch": 0.6478522388966523, + "grad_norm": 0.23960524797439575, + "learning_rate": 2.776098163577635e-05, + "loss": 1.1746, + "step": 10663 + }, + { + "epoch": 0.6479129959292789, + "grad_norm": 0.31280699372291565, + "learning_rate": 2.7752408288103082e-05, + "loss": 1.0956, + "step": 10664 + }, + { + "epoch": 0.6479737529619053, + "grad_norm": 0.42136716842651367, + "learning_rate": 2.774383575591956e-05, + "loss": 1.0866, + "step": 10665 + }, + { + "epoch": 0.6480345099945318, + "grad_norm": 0.20635396242141724, + "learning_rate": 2.773526403954005e-05, + "loss": 1.0903, + "step": 10666 + }, + { + "epoch": 0.6480952670271584, + "grad_norm": 0.22826144099235535, + "learning_rate": 2.7726693139278716e-05, + "loss": 1.0864, + "step": 10667 + }, + { + "epoch": 0.6481560240597849, + "grad_norm": 1.1304939985275269, + "learning_rate": 2.7718123055449773e-05, + "loss": 1.0331, + "step": 10668 + }, + { + "epoch": 0.6482167810924114, + "grad_norm": 0.19250664114952087, + "learning_rate": 2.770955378836733e-05, + "loss": 1.1023, + "step": 10669 + }, + { + "epoch": 0.648277538125038, + "grad_norm": 0.16241875290870667, + "learning_rate": 2.7700985338345525e-05, + "loss": 1.1281, + "step": 10670 + }, + { + "epoch": 0.6483382951576645, + "grad_norm": 0.1643594652414322, + "learning_rate": 2.769241770569839e-05, + "loss": 1.0134, + "step": 10671 + }, + { + "epoch": 0.648399052190291, + "grad_norm": 0.15217629075050354, + "learning_rate": 2.768385089073998e-05, + "loss": 1.0827, + "step": 10672 + }, + { + "epoch": 0.6484598092229176, + "grad_norm": 0.14066019654273987, + "learning_rate": 2.7675284893784353e-05, + "loss": 1.02, + "step": 10673 + }, + { + "epoch": 0.6485205662555441, + "grad_norm": 0.15381215512752533, + "learning_rate": 2.766671971514547e-05, + "loss": 1.0581, + "step": 10674 + }, + { + "epoch": 0.6485813232881706, + "grad_norm": 0.22001467645168304, + "learning_rate": 2.7658155355137295e-05, + "loss": 1.1725, + "step": 10675 + }, + { + "epoch": 0.6486420803207972, + "grad_norm": 0.15447553992271423, + "learning_rate": 2.764959181407375e-05, + "loss": 1.0787, + "step": 10676 + }, + { + "epoch": 0.6487028373534237, + "grad_norm": 0.23059430718421936, + "learning_rate": 2.7641029092268733e-05, + "loss": 1.1331, + "step": 10677 + }, + { + "epoch": 0.6487635943860501, + "grad_norm": 0.25465190410614014, + "learning_rate": 2.7632467190036128e-05, + "loss": 1.1865, + "step": 10678 + }, + { + "epoch": 0.6488243514186767, + "grad_norm": 0.22285905480384827, + "learning_rate": 2.762390610768975e-05, + "loss": 1.1388, + "step": 10679 + }, + { + "epoch": 0.6488851084513032, + "grad_norm": 0.1846393346786499, + "learning_rate": 2.761534584554343e-05, + "loss": 1.049, + "step": 10680 + }, + { + "epoch": 0.6489458654839297, + "grad_norm": 0.18521639704704285, + "learning_rate": 2.7606786403910923e-05, + "loss": 1.1037, + "step": 10681 + }, + { + "epoch": 0.6490066225165563, + "grad_norm": 0.5514048337936401, + "learning_rate": 2.7598227783105995e-05, + "loss": 1.0481, + "step": 10682 + }, + { + "epoch": 0.6490673795491828, + "grad_norm": 0.2728809416294098, + "learning_rate": 2.758966998344234e-05, + "loss": 1.2928, + "step": 10683 + }, + { + "epoch": 0.6491281365818093, + "grad_norm": 0.13598595559597015, + "learning_rate": 2.75811130052337e-05, + "loss": 0.9835, + "step": 10684 + }, + { + "epoch": 0.6491888936144359, + "grad_norm": 0.6162171959877014, + "learning_rate": 2.757255684879368e-05, + "loss": 1.0309, + "step": 10685 + }, + { + "epoch": 0.6492496506470624, + "grad_norm": 0.1570122390985489, + "learning_rate": 2.756400151443592e-05, + "loss": 1.2754, + "step": 10686 + }, + { + "epoch": 0.649310407679689, + "grad_norm": 0.13723613321781158, + "learning_rate": 2.7555447002474023e-05, + "loss": 1.019, + "step": 10687 + }, + { + "epoch": 0.6493711647123155, + "grad_norm": 0.5135619044303894, + "learning_rate": 2.7546893313221535e-05, + "loss": 1.0437, + "step": 10688 + }, + { + "epoch": 0.649431921744942, + "grad_norm": 0.22871555387973785, + "learning_rate": 2.753834044699203e-05, + "loss": 1.2032, + "step": 10689 + }, + { + "epoch": 0.6494926787775686, + "grad_norm": 0.15492059290409088, + "learning_rate": 2.7529788404098998e-05, + "loss": 1.0435, + "step": 10690 + }, + { + "epoch": 0.649553435810195, + "grad_norm": 0.14886534214019775, + "learning_rate": 2.752123718485594e-05, + "loss": 1.0079, + "step": 10691 + }, + { + "epoch": 0.6496141928428215, + "grad_norm": 0.24716533720493317, + "learning_rate": 2.7512686789576253e-05, + "loss": 1.0739, + "step": 10692 + }, + { + "epoch": 0.649674949875448, + "grad_norm": 0.20316872000694275, + "learning_rate": 2.750413721857336e-05, + "loss": 1.0939, + "step": 10693 + }, + { + "epoch": 0.6497357069080746, + "grad_norm": 0.1298951506614685, + "learning_rate": 2.7495588472160687e-05, + "loss": 1.047, + "step": 10694 + }, + { + "epoch": 0.6497964639407011, + "grad_norm": 0.14157339930534363, + "learning_rate": 2.7487040550651564e-05, + "loss": 1.0443, + "step": 10695 + }, + { + "epoch": 0.6498572209733277, + "grad_norm": 0.219308003783226, + "learning_rate": 2.7478493454359334e-05, + "loss": 1.1908, + "step": 10696 + }, + { + "epoch": 0.6499179780059542, + "grad_norm": 0.1654965728521347, + "learning_rate": 2.746994718359729e-05, + "loss": 1.0678, + "step": 10697 + }, + { + "epoch": 0.6499787350385807, + "grad_norm": 0.1433989554643631, + "learning_rate": 2.7461401738678643e-05, + "loss": 1.0704, + "step": 10698 + }, + { + "epoch": 0.6500394920712073, + "grad_norm": 0.7129765748977661, + "learning_rate": 2.7452857119916696e-05, + "loss": 1.3095, + "step": 10699 + }, + { + "epoch": 0.6501002491038338, + "grad_norm": 0.4937276244163513, + "learning_rate": 2.7444313327624632e-05, + "loss": 1.2971, + "step": 10700 + }, + { + "epoch": 0.6501610061364603, + "grad_norm": 0.19503894448280334, + "learning_rate": 2.743577036211562e-05, + "loss": 1.1052, + "step": 10701 + }, + { + "epoch": 0.6502217631690869, + "grad_norm": 0.1849939376115799, + "learning_rate": 2.7427228223702805e-05, + "loss": 1.0471, + "step": 10702 + }, + { + "epoch": 0.6502825202017134, + "grad_norm": 0.18988533318042755, + "learning_rate": 2.741868691269931e-05, + "loss": 1.0588, + "step": 10703 + }, + { + "epoch": 0.6503432772343398, + "grad_norm": 0.1430758237838745, + "learning_rate": 2.7410146429418206e-05, + "loss": 1.0454, + "step": 10704 + }, + { + "epoch": 0.6504040342669664, + "grad_norm": 0.1616276055574417, + "learning_rate": 2.740160677417255e-05, + "loss": 1.0295, + "step": 10705 + }, + { + "epoch": 0.6504647912995929, + "grad_norm": 0.6280221343040466, + "learning_rate": 2.739306794727537e-05, + "loss": 1.2155, + "step": 10706 + }, + { + "epoch": 0.6505255483322194, + "grad_norm": 0.1893477737903595, + "learning_rate": 2.738452994903965e-05, + "loss": 1.1385, + "step": 10707 + }, + { + "epoch": 0.650586305364846, + "grad_norm": 0.1733779013156891, + "learning_rate": 2.737599277977836e-05, + "loss": 1.018, + "step": 10708 + }, + { + "epoch": 0.6506470623974725, + "grad_norm": 0.27373600006103516, + "learning_rate": 2.7367456439804407e-05, + "loss": 1.0807, + "step": 10709 + }, + { + "epoch": 0.650707819430099, + "grad_norm": 0.23534506559371948, + "learning_rate": 2.735892092943073e-05, + "loss": 1.0407, + "step": 10710 + }, + { + "epoch": 0.6507685764627256, + "grad_norm": 0.2373695969581604, + "learning_rate": 2.7350386248970207e-05, + "loss": 1.1795, + "step": 10711 + }, + { + "epoch": 0.6508293334953521, + "grad_norm": 0.15411032736301422, + "learning_rate": 2.734185239873563e-05, + "loss": 1.028, + "step": 10712 + }, + { + "epoch": 0.6508900905279786, + "grad_norm": 0.20434895157814026, + "learning_rate": 2.7333319379039834e-05, + "loss": 1.1669, + "step": 10713 + }, + { + "epoch": 0.6509508475606052, + "grad_norm": 0.21243466436862946, + "learning_rate": 2.7324787190195577e-05, + "loss": 1.1053, + "step": 10714 + }, + { + "epoch": 0.6510116045932317, + "grad_norm": 0.1453329175710678, + "learning_rate": 2.731625583251565e-05, + "loss": 1.0459, + "step": 10715 + }, + { + "epoch": 0.6510723616258582, + "grad_norm": 0.2651583254337311, + "learning_rate": 2.730772530631275e-05, + "loss": 1.0279, + "step": 10716 + }, + { + "epoch": 0.6511331186584847, + "grad_norm": 0.1741637885570526, + "learning_rate": 2.7299195611899587e-05, + "loss": 1.0585, + "step": 10717 + }, + { + "epoch": 0.6511938756911112, + "grad_norm": 0.20416787266731262, + "learning_rate": 2.729066674958878e-05, + "loss": 1.1147, + "step": 10718 + }, + { + "epoch": 0.6512546327237377, + "grad_norm": 0.20872555673122406, + "learning_rate": 2.728213871969294e-05, + "loss": 1.093, + "step": 10719 + }, + { + "epoch": 0.6513153897563643, + "grad_norm": 0.1848808079957962, + "learning_rate": 2.7273611522524728e-05, + "loss": 1.0659, + "step": 10720 + }, + { + "epoch": 0.6513761467889908, + "grad_norm": 0.19201789796352386, + "learning_rate": 2.7265085158396675e-05, + "loss": 1.0122, + "step": 10721 + }, + { + "epoch": 0.6514369038216173, + "grad_norm": 0.16538840532302856, + "learning_rate": 2.725655962762132e-05, + "loss": 1.0691, + "step": 10722 + }, + { + "epoch": 0.6514976608542439, + "grad_norm": 1.0865978002548218, + "learning_rate": 2.724803493051117e-05, + "loss": 1.0639, + "step": 10723 + }, + { + "epoch": 0.6515584178868704, + "grad_norm": 0.14194871485233307, + "learning_rate": 2.7239511067378692e-05, + "loss": 1.0336, + "step": 10724 + }, + { + "epoch": 0.651619174919497, + "grad_norm": 0.1483648270368576, + "learning_rate": 2.723098803853634e-05, + "loss": 1.0395, + "step": 10725 + }, + { + "epoch": 0.6516799319521235, + "grad_norm": 0.18544408679008484, + "learning_rate": 2.722246584429652e-05, + "loss": 1.1637, + "step": 10726 + }, + { + "epoch": 0.65174068898475, + "grad_norm": 0.5413992404937744, + "learning_rate": 2.7213944484971622e-05, + "loss": 1.076, + "step": 10727 + }, + { + "epoch": 0.6518014460173766, + "grad_norm": 0.20047779381275177, + "learning_rate": 2.7205423960873987e-05, + "loss": 1.0786, + "step": 10728 + }, + { + "epoch": 0.6518622030500031, + "grad_norm": 0.6364901065826416, + "learning_rate": 2.7196904272315947e-05, + "loss": 1.1365, + "step": 10729 + }, + { + "epoch": 0.6519229600826295, + "grad_norm": 0.17390680313110352, + "learning_rate": 2.718838541960977e-05, + "loss": 1.1025, + "step": 10730 + }, + { + "epoch": 0.651983717115256, + "grad_norm": 0.17652803659439087, + "learning_rate": 2.717986740306777e-05, + "loss": 1.1007, + "step": 10731 + }, + { + "epoch": 0.6520444741478826, + "grad_norm": 0.47258687019348145, + "learning_rate": 2.7171350223002123e-05, + "loss": 1.1786, + "step": 10732 + }, + { + "epoch": 0.6521052311805091, + "grad_norm": 0.2359837144613266, + "learning_rate": 2.716283387972505e-05, + "loss": 1.0408, + "step": 10733 + }, + { + "epoch": 0.6521659882131357, + "grad_norm": 0.18744009733200073, + "learning_rate": 2.715431837354871e-05, + "loss": 1.0424, + "step": 10734 + }, + { + "epoch": 0.6522267452457622, + "grad_norm": 0.3179978132247925, + "learning_rate": 2.7145803704785234e-05, + "loss": 1.1588, + "step": 10735 + }, + { + "epoch": 0.6522875022783887, + "grad_norm": 1.2414882183074951, + "learning_rate": 2.7137289873746752e-05, + "loss": 1.1484, + "step": 10736 + }, + { + "epoch": 0.6523482593110153, + "grad_norm": 0.14354966580867767, + "learning_rate": 2.7128776880745354e-05, + "loss": 1.0491, + "step": 10737 + }, + { + "epoch": 0.6524090163436418, + "grad_norm": 0.17942111194133759, + "learning_rate": 2.7120264726093038e-05, + "loss": 1.0957, + "step": 10738 + }, + { + "epoch": 0.6524697733762683, + "grad_norm": 0.24916771054267883, + "learning_rate": 2.7111753410101838e-05, + "loss": 1.1401, + "step": 10739 + }, + { + "epoch": 0.6525305304088949, + "grad_norm": 0.3501843214035034, + "learning_rate": 2.710324293308373e-05, + "loss": 1.0164, + "step": 10740 + }, + { + "epoch": 0.6525912874415214, + "grad_norm": 2.5333755016326904, + "learning_rate": 2.709473329535069e-05, + "loss": 1.072, + "step": 10741 + }, + { + "epoch": 0.6526520444741479, + "grad_norm": 0.1803043633699417, + "learning_rate": 2.708622449721463e-05, + "loss": 1.0425, + "step": 10742 + }, + { + "epoch": 0.6527128015067744, + "grad_norm": 0.3447245955467224, + "learning_rate": 2.707771653898744e-05, + "loss": 1.0177, + "step": 10743 + }, + { + "epoch": 0.6527735585394009, + "grad_norm": 0.23754844069480896, + "learning_rate": 2.7069209420981002e-05, + "loss": 1.1551, + "step": 10744 + }, + { + "epoch": 0.6528343155720274, + "grad_norm": 0.18579238653182983, + "learning_rate": 2.7060703143507088e-05, + "loss": 1.0542, + "step": 10745 + }, + { + "epoch": 0.652895072604654, + "grad_norm": 0.17283645272254944, + "learning_rate": 2.7052197706877553e-05, + "loss": 1.0634, + "step": 10746 + }, + { + "epoch": 0.6529558296372805, + "grad_norm": 0.35922971367836, + "learning_rate": 2.7043693111404146e-05, + "loss": 1.0449, + "step": 10747 + }, + { + "epoch": 0.653016586669907, + "grad_norm": 1.4799396991729736, + "learning_rate": 2.7035189357398605e-05, + "loss": 1.0094, + "step": 10748 + }, + { + "epoch": 0.6530773437025336, + "grad_norm": 0.19621512293815613, + "learning_rate": 2.7026686445172634e-05, + "loss": 1.0893, + "step": 10749 + }, + { + "epoch": 0.6531381007351601, + "grad_norm": 0.30941030383110046, + "learning_rate": 2.7018184375037913e-05, + "loss": 1.1085, + "step": 10750 + }, + { + "epoch": 0.6531988577677866, + "grad_norm": 0.2522110342979431, + "learning_rate": 2.700968314730609e-05, + "loss": 1.12, + "step": 10751 + }, + { + "epoch": 0.6532596148004132, + "grad_norm": 0.25645413994789124, + "learning_rate": 2.7001182762288774e-05, + "loss": 1.0734, + "step": 10752 + }, + { + "epoch": 0.6533203718330397, + "grad_norm": 0.27849042415618896, + "learning_rate": 2.699268322029755e-05, + "loss": 1.0293, + "step": 10753 + }, + { + "epoch": 0.6533811288656662, + "grad_norm": 0.49275463819503784, + "learning_rate": 2.6984184521643968e-05, + "loss": 1.4073, + "step": 10754 + }, + { + "epoch": 0.6534418858982928, + "grad_norm": 0.16096359491348267, + "learning_rate": 2.6975686666639556e-05, + "loss": 1.0263, + "step": 10755 + }, + { + "epoch": 0.6535026429309193, + "grad_norm": 0.2383432686328888, + "learning_rate": 2.696718965559577e-05, + "loss": 1.0986, + "step": 10756 + }, + { + "epoch": 0.6535633999635457, + "grad_norm": 1.6519004106521606, + "learning_rate": 2.6958693488824127e-05, + "loss": 1.0993, + "step": 10757 + }, + { + "epoch": 0.6536241569961723, + "grad_norm": 0.17147605121135712, + "learning_rate": 2.6950198166636042e-05, + "loss": 1.0164, + "step": 10758 + }, + { + "epoch": 0.6536849140287988, + "grad_norm": 0.29691648483276367, + "learning_rate": 2.6941703689342885e-05, + "loss": 1.1414, + "step": 10759 + }, + { + "epoch": 0.6537456710614253, + "grad_norm": 0.16076034307479858, + "learning_rate": 2.6933210057256014e-05, + "loss": 0.9836, + "step": 10760 + }, + { + "epoch": 0.6538064280940519, + "grad_norm": 0.345527708530426, + "learning_rate": 2.6924717270686807e-05, + "loss": 1.1093, + "step": 10761 + }, + { + "epoch": 0.6538671851266784, + "grad_norm": 0.15248987078666687, + "learning_rate": 2.691622532994655e-05, + "loss": 1.0257, + "step": 10762 + }, + { + "epoch": 0.6539279421593049, + "grad_norm": 0.20922008156776428, + "learning_rate": 2.6907734235346515e-05, + "loss": 1.1448, + "step": 10763 + }, + { + "epoch": 0.6539886991919315, + "grad_norm": 0.14480626583099365, + "learning_rate": 2.6899243987197963e-05, + "loss": 1.0029, + "step": 10764 + }, + { + "epoch": 0.654049456224558, + "grad_norm": 0.5519189834594727, + "learning_rate": 2.689075458581205e-05, + "loss": 1.2369, + "step": 10765 + }, + { + "epoch": 0.6541102132571845, + "grad_norm": 0.3156692087650299, + "learning_rate": 2.6882266031500016e-05, + "loss": 1.0895, + "step": 10766 + }, + { + "epoch": 0.6541709702898111, + "grad_norm": 0.25416848063468933, + "learning_rate": 2.6873778324572986e-05, + "loss": 1.0951, + "step": 10767 + }, + { + "epoch": 0.6542317273224376, + "grad_norm": 0.20531107485294342, + "learning_rate": 2.686529146534208e-05, + "loss": 1.0627, + "step": 10768 + }, + { + "epoch": 0.6542924843550642, + "grad_norm": 0.21995213627815247, + "learning_rate": 2.685680545411839e-05, + "loss": 1.0852, + "step": 10769 + }, + { + "epoch": 0.6543532413876906, + "grad_norm": 0.1890331357717514, + "learning_rate": 2.684832029121297e-05, + "loss": 1.0976, + "step": 10770 + }, + { + "epoch": 0.6544139984203171, + "grad_norm": 0.23430286347866058, + "learning_rate": 2.6839835976936843e-05, + "loss": 1.0412, + "step": 10771 + }, + { + "epoch": 0.6544747554529436, + "grad_norm": 0.22663167119026184, + "learning_rate": 2.6831352511601006e-05, + "loss": 1.1589, + "step": 10772 + }, + { + "epoch": 0.6545355124855702, + "grad_norm": 0.15309225022792816, + "learning_rate": 2.6822869895516424e-05, + "loss": 1.0205, + "step": 10773 + }, + { + "epoch": 0.6545962695181967, + "grad_norm": 0.20319217443466187, + "learning_rate": 2.6814388128994023e-05, + "loss": 1.0851, + "step": 10774 + }, + { + "epoch": 0.6546570265508233, + "grad_norm": 0.3437619209289551, + "learning_rate": 2.6805907212344705e-05, + "loss": 1.1001, + "step": 10775 + }, + { + "epoch": 0.6547177835834498, + "grad_norm": 0.1484859585762024, + "learning_rate": 2.6797427145879327e-05, + "loss": 1.0208, + "step": 10776 + }, + { + "epoch": 0.6547785406160763, + "grad_norm": 3.4825735092163086, + "learning_rate": 2.6788947929908757e-05, + "loss": 1.0692, + "step": 10777 + }, + { + "epoch": 0.6548392976487029, + "grad_norm": 0.2094874233007431, + "learning_rate": 2.678046956474381e-05, + "loss": 1.0845, + "step": 10778 + }, + { + "epoch": 0.6549000546813294, + "grad_norm": 0.15171664953231812, + "learning_rate": 2.6771992050695215e-05, + "loss": 1.0524, + "step": 10779 + }, + { + "epoch": 0.6549608117139559, + "grad_norm": 0.1377207338809967, + "learning_rate": 2.6763515388073745e-05, + "loss": 1.0545, + "step": 10780 + }, + { + "epoch": 0.6550215687465825, + "grad_norm": 0.21109136939048767, + "learning_rate": 2.6755039577190087e-05, + "loss": 1.1314, + "step": 10781 + }, + { + "epoch": 0.655082325779209, + "grad_norm": 0.3014861047267914, + "learning_rate": 2.6746564618354974e-05, + "loss": 1.0761, + "step": 10782 + }, + { + "epoch": 0.6551430828118354, + "grad_norm": 0.17874892055988312, + "learning_rate": 2.6738090511879022e-05, + "loss": 1.0352, + "step": 10783 + }, + { + "epoch": 0.655203839844462, + "grad_norm": 0.19415998458862305, + "learning_rate": 2.6729617258072883e-05, + "loss": 1.0716, + "step": 10784 + }, + { + "epoch": 0.6552645968770885, + "grad_norm": 0.1818680465221405, + "learning_rate": 2.67211448572471e-05, + "loss": 1.1019, + "step": 10785 + }, + { + "epoch": 0.655325353909715, + "grad_norm": 0.2367117553949356, + "learning_rate": 2.6712673309712232e-05, + "loss": 1.1145, + "step": 10786 + }, + { + "epoch": 0.6553861109423416, + "grad_norm": 0.25574055314064026, + "learning_rate": 2.6704202615778844e-05, + "loss": 1.1256, + "step": 10787 + }, + { + "epoch": 0.6554468679749681, + "grad_norm": 3.9128382205963135, + "learning_rate": 2.669573277575741e-05, + "loss": 1.1848, + "step": 10788 + }, + { + "epoch": 0.6555076250075946, + "grad_norm": 0.26186618208885193, + "learning_rate": 2.6687263789958407e-05, + "loss": 1.0211, + "step": 10789 + }, + { + "epoch": 0.6555683820402212, + "grad_norm": 0.13495513796806335, + "learning_rate": 2.6678795658692247e-05, + "loss": 1.0714, + "step": 10790 + }, + { + "epoch": 0.6556291390728477, + "grad_norm": 0.39009809494018555, + "learning_rate": 2.6670328382269338e-05, + "loss": 0.9886, + "step": 10791 + }, + { + "epoch": 0.6556898961054742, + "grad_norm": 0.20703744888305664, + "learning_rate": 2.6661861961000058e-05, + "loss": 1.1396, + "step": 10792 + }, + { + "epoch": 0.6557506531381008, + "grad_norm": 0.21005354821681976, + "learning_rate": 2.6653396395194728e-05, + "loss": 1.1716, + "step": 10793 + }, + { + "epoch": 0.6558114101707273, + "grad_norm": 0.19532352685928345, + "learning_rate": 2.6644931685163678e-05, + "loss": 1.0761, + "step": 10794 + }, + { + "epoch": 0.6558721672033538, + "grad_norm": 0.1387837827205658, + "learning_rate": 2.663646783121716e-05, + "loss": 1.0652, + "step": 10795 + }, + { + "epoch": 0.6559329242359803, + "grad_norm": 0.1897311806678772, + "learning_rate": 2.6628004833665443e-05, + "loss": 1.1005, + "step": 10796 + }, + { + "epoch": 0.6559936812686068, + "grad_norm": 13.215658187866211, + "learning_rate": 2.66195426928187e-05, + "loss": 1.0429, + "step": 10797 + }, + { + "epoch": 0.6560544383012333, + "grad_norm": 0.2883339822292328, + "learning_rate": 2.6611081408987177e-05, + "loss": 1.2024, + "step": 10798 + }, + { + "epoch": 0.6561151953338599, + "grad_norm": 0.4316248595714569, + "learning_rate": 2.660262098248097e-05, + "loss": 1.0301, + "step": 10799 + }, + { + "epoch": 0.6561759523664864, + "grad_norm": 0.1754373461008072, + "learning_rate": 2.6594161413610207e-05, + "loss": 1.0359, + "step": 10800 + }, + { + "epoch": 0.6562367093991129, + "grad_norm": 0.23220029473304749, + "learning_rate": 2.6585702702684988e-05, + "loss": 1.0612, + "step": 10801 + }, + { + "epoch": 0.6562974664317395, + "grad_norm": 0.32575371861457825, + "learning_rate": 2.6577244850015337e-05, + "loss": 1.0015, + "step": 10802 + }, + { + "epoch": 0.656358223464366, + "grad_norm": 0.2115354835987091, + "learning_rate": 2.656878785591133e-05, + "loss": 1.0791, + "step": 10803 + }, + { + "epoch": 0.6564189804969925, + "grad_norm": 0.24314740300178528, + "learning_rate": 2.6560331720682952e-05, + "loss": 1.1457, + "step": 10804 + }, + { + "epoch": 0.6564797375296191, + "grad_norm": 0.17754894495010376, + "learning_rate": 2.6551876444640123e-05, + "loss": 1.0808, + "step": 10805 + }, + { + "epoch": 0.6565404945622456, + "grad_norm": 0.16858181357383728, + "learning_rate": 2.6543422028092797e-05, + "loss": 0.9977, + "step": 10806 + }, + { + "epoch": 0.6566012515948721, + "grad_norm": 0.18263766169548035, + "learning_rate": 2.6534968471350852e-05, + "loss": 1.0857, + "step": 10807 + }, + { + "epoch": 0.6566620086274987, + "grad_norm": 0.17301493883132935, + "learning_rate": 2.6526515774724193e-05, + "loss": 1.0439, + "step": 10808 + }, + { + "epoch": 0.6567227656601251, + "grad_norm": 0.2595754861831665, + "learning_rate": 2.651806393852263e-05, + "loss": 1.0691, + "step": 10809 + }, + { + "epoch": 0.6567835226927516, + "grad_norm": 0.25913873314857483, + "learning_rate": 2.6509612963055974e-05, + "loss": 1.0058, + "step": 10810 + }, + { + "epoch": 0.6568442797253782, + "grad_norm": 0.21605786681175232, + "learning_rate": 2.650116284863402e-05, + "loss": 1.0932, + "step": 10811 + }, + { + "epoch": 0.6569050367580047, + "grad_norm": 0.17078648507595062, + "learning_rate": 2.6492713595566443e-05, + "loss": 1.0492, + "step": 10812 + }, + { + "epoch": 0.6569657937906312, + "grad_norm": 0.20615684986114502, + "learning_rate": 2.6484265204163005e-05, + "loss": 1.075, + "step": 10813 + }, + { + "epoch": 0.6570265508232578, + "grad_norm": 1.6909211874008179, + "learning_rate": 2.6475817674733383e-05, + "loss": 1.0755, + "step": 10814 + }, + { + "epoch": 0.6570873078558843, + "grad_norm": 0.2131003588438034, + "learning_rate": 2.6467371007587203e-05, + "loss": 1.0454, + "step": 10815 + }, + { + "epoch": 0.6571480648885109, + "grad_norm": 0.15673495829105377, + "learning_rate": 2.645892520303409e-05, + "loss": 1.0793, + "step": 10816 + }, + { + "epoch": 0.6572088219211374, + "grad_norm": 1.3247718811035156, + "learning_rate": 2.6450480261383626e-05, + "loss": 1.2371, + "step": 10817 + }, + { + "epoch": 0.6572695789537639, + "grad_norm": 0.21553361415863037, + "learning_rate": 2.644203618294536e-05, + "loss": 1.1809, + "step": 10818 + }, + { + "epoch": 0.6573303359863905, + "grad_norm": 0.18700170516967773, + "learning_rate": 2.6433592968028807e-05, + "loss": 1.0949, + "step": 10819 + }, + { + "epoch": 0.657391093019017, + "grad_norm": 0.17333029210567474, + "learning_rate": 2.6425150616943463e-05, + "loss": 1.0644, + "step": 10820 + }, + { + "epoch": 0.6574518500516435, + "grad_norm": 0.22418756783008575, + "learning_rate": 2.6416709129998783e-05, + "loss": 1.1202, + "step": 10821 + }, + { + "epoch": 0.65751260708427, + "grad_norm": 0.33728528022766113, + "learning_rate": 2.6408268507504186e-05, + "loss": 1.1299, + "step": 10822 + }, + { + "epoch": 0.6575733641168965, + "grad_norm": 0.16564756631851196, + "learning_rate": 2.639982874976904e-05, + "loss": 1.0783, + "step": 10823 + }, + { + "epoch": 0.657634121149523, + "grad_norm": 0.3284679353237152, + "learning_rate": 2.6391389857102762e-05, + "loss": 1.0486, + "step": 10824 + }, + { + "epoch": 0.6576948781821496, + "grad_norm": 0.2077442854642868, + "learning_rate": 2.638295182981467e-05, + "loss": 1.0909, + "step": 10825 + }, + { + "epoch": 0.6577556352147761, + "grad_norm": 0.19988960027694702, + "learning_rate": 2.6374514668214024e-05, + "loss": 1.0815, + "step": 10826 + }, + { + "epoch": 0.6578163922474026, + "grad_norm": 0.22184252738952637, + "learning_rate": 2.6366078372610103e-05, + "loss": 1.0007, + "step": 10827 + }, + { + "epoch": 0.6578771492800292, + "grad_norm": 0.25345343351364136, + "learning_rate": 2.6357642943312134e-05, + "loss": 1.0859, + "step": 10828 + }, + { + "epoch": 0.6579379063126557, + "grad_norm": 0.29638853669166565, + "learning_rate": 2.6349208380629353e-05, + "loss": 1.184, + "step": 10829 + }, + { + "epoch": 0.6579986633452822, + "grad_norm": 0.23870056867599487, + "learning_rate": 2.6340774684870904e-05, + "loss": 1.1198, + "step": 10830 + }, + { + "epoch": 0.6580594203779088, + "grad_norm": 0.46421536803245544, + "learning_rate": 2.633234185634596e-05, + "loss": 1.0948, + "step": 10831 + }, + { + "epoch": 0.6581201774105353, + "grad_norm": 0.25881460309028625, + "learning_rate": 2.6323909895363575e-05, + "loss": 1.0702, + "step": 10832 + }, + { + "epoch": 0.6581809344431618, + "grad_norm": 0.9096970558166504, + "learning_rate": 2.6315478802232835e-05, + "loss": 1.1899, + "step": 10833 + }, + { + "epoch": 0.6582416914757884, + "grad_norm": 0.20603151619434357, + "learning_rate": 2.6307048577262815e-05, + "loss": 1.0596, + "step": 10834 + }, + { + "epoch": 0.6583024485084148, + "grad_norm": 0.17520055174827576, + "learning_rate": 2.629861922076251e-05, + "loss": 1.0425, + "step": 10835 + }, + { + "epoch": 0.6583632055410413, + "grad_norm": 0.2357771396636963, + "learning_rate": 2.6290190733040897e-05, + "loss": 1.0215, + "step": 10836 + }, + { + "epoch": 0.6584239625736679, + "grad_norm": 0.22373662889003754, + "learning_rate": 2.6281763114406923e-05, + "loss": 1.1218, + "step": 10837 + }, + { + "epoch": 0.6584847196062944, + "grad_norm": 0.21005676686763763, + "learning_rate": 2.627333636516951e-05, + "loss": 1.0554, + "step": 10838 + }, + { + "epoch": 0.6585454766389209, + "grad_norm": 0.21170301735401154, + "learning_rate": 2.626491048563754e-05, + "loss": 1.0992, + "step": 10839 + }, + { + "epoch": 0.6586062336715475, + "grad_norm": 0.15383166074752808, + "learning_rate": 2.6256485476119863e-05, + "loss": 1.0243, + "step": 10840 + }, + { + "epoch": 0.658666990704174, + "grad_norm": 0.22549141943454742, + "learning_rate": 2.6248061336925302e-05, + "loss": 1.1731, + "step": 10841 + }, + { + "epoch": 0.6587277477368005, + "grad_norm": 0.3167867660522461, + "learning_rate": 2.6239638068362636e-05, + "loss": 1.0387, + "step": 10842 + }, + { + "epoch": 0.6587885047694271, + "grad_norm": 0.27193403244018555, + "learning_rate": 2.623121567074064e-05, + "loss": 0.9966, + "step": 10843 + }, + { + "epoch": 0.6588492618020536, + "grad_norm": 0.369173526763916, + "learning_rate": 2.6222794144368e-05, + "loss": 1.1154, + "step": 10844 + }, + { + "epoch": 0.6589100188346801, + "grad_norm": 0.28341004252433777, + "learning_rate": 2.621437348955348e-05, + "loss": 1.0608, + "step": 10845 + }, + { + "epoch": 0.6589707758673067, + "grad_norm": 0.21559014916419983, + "learning_rate": 2.620595370660568e-05, + "loss": 1.0494, + "step": 10846 + }, + { + "epoch": 0.6590315328999332, + "grad_norm": 0.15215207636356354, + "learning_rate": 2.619753479583324e-05, + "loss": 1.0374, + "step": 10847 + }, + { + "epoch": 0.6590922899325596, + "grad_norm": 1.1964733600616455, + "learning_rate": 2.6189116757544764e-05, + "loss": 1.0635, + "step": 10848 + }, + { + "epoch": 0.6591530469651862, + "grad_norm": 0.23914112150669098, + "learning_rate": 2.6180699592048795e-05, + "loss": 1.0778, + "step": 10849 + }, + { + "epoch": 0.6592138039978127, + "grad_norm": 0.253040075302124, + "learning_rate": 2.617228329965391e-05, + "loss": 1.112, + "step": 10850 + }, + { + "epoch": 0.6592745610304392, + "grad_norm": 0.1336577832698822, + "learning_rate": 2.6163867880668614e-05, + "loss": 1.0101, + "step": 10851 + }, + { + "epoch": 0.6593353180630658, + "grad_norm": 0.2121431678533554, + "learning_rate": 2.6155453335401315e-05, + "loss": 1.1612, + "step": 10852 + }, + { + "epoch": 0.6593960750956923, + "grad_norm": 0.16817377507686615, + "learning_rate": 2.6147039664160488e-05, + "loss": 1.095, + "step": 10853 + }, + { + "epoch": 0.6594568321283188, + "grad_norm": 0.2734683156013489, + "learning_rate": 2.6138626867254512e-05, + "loss": 1.0377, + "step": 10854 + }, + { + "epoch": 0.6595175891609454, + "grad_norm": 0.15286396443843842, + "learning_rate": 2.6130214944991804e-05, + "loss": 1.0487, + "step": 10855 + }, + { + "epoch": 0.6595783461935719, + "grad_norm": 0.2807813882827759, + "learning_rate": 2.612180389768068e-05, + "loss": 1.0491, + "step": 10856 + }, + { + "epoch": 0.6596391032261985, + "grad_norm": 0.2670130729675293, + "learning_rate": 2.6113393725629447e-05, + "loss": 1.077, + "step": 10857 + }, + { + "epoch": 0.659699860258825, + "grad_norm": 0.21060161292552948, + "learning_rate": 2.6104984429146384e-05, + "loss": 1.1224, + "step": 10858 + }, + { + "epoch": 0.6597606172914515, + "grad_norm": 0.22572708129882812, + "learning_rate": 2.609657600853974e-05, + "loss": 1.1151, + "step": 10859 + }, + { + "epoch": 0.6598213743240781, + "grad_norm": 0.21359983086585999, + "learning_rate": 2.608816846411772e-05, + "loss": 1.1513, + "step": 10860 + }, + { + "epoch": 0.6598821313567046, + "grad_norm": 0.14728495478630066, + "learning_rate": 2.6079761796188508e-05, + "loss": 1.0513, + "step": 10861 + }, + { + "epoch": 0.659942888389331, + "grad_norm": 0.5978706479072571, + "learning_rate": 2.6071356005060248e-05, + "loss": 1.1289, + "step": 10862 + }, + { + "epoch": 0.6600036454219576, + "grad_norm": 0.20156753063201904, + "learning_rate": 2.606295109104106e-05, + "loss": 1.0216, + "step": 10863 + }, + { + "epoch": 0.6600644024545841, + "grad_norm": 0.17050205171108246, + "learning_rate": 2.6054547054439026e-05, + "loss": 1.0221, + "step": 10864 + }, + { + "epoch": 0.6601251594872106, + "grad_norm": 0.15981563925743103, + "learning_rate": 2.60461438955622e-05, + "loss": 1.0068, + "step": 10865 + }, + { + "epoch": 0.6601859165198372, + "grad_norm": 0.16660384833812714, + "learning_rate": 2.6037741614718602e-05, + "loss": 1.0684, + "step": 10866 + }, + { + "epoch": 0.6602466735524637, + "grad_norm": 0.2430313378572464, + "learning_rate": 2.602934021221621e-05, + "loss": 1.1309, + "step": 10867 + }, + { + "epoch": 0.6603074305850902, + "grad_norm": 0.276727557182312, + "learning_rate": 2.6020939688362995e-05, + "loss": 1.1355, + "step": 10868 + }, + { + "epoch": 0.6603681876177168, + "grad_norm": 0.5443997383117676, + "learning_rate": 2.601254004346685e-05, + "loss": 1.4188, + "step": 10869 + }, + { + "epoch": 0.6604289446503433, + "grad_norm": 0.44094929099082947, + "learning_rate": 2.6004141277835703e-05, + "loss": 1.162, + "step": 10870 + }, + { + "epoch": 0.6604897016829698, + "grad_norm": 0.3927478790283203, + "learning_rate": 2.599574339177741e-05, + "loss": 1.0397, + "step": 10871 + }, + { + "epoch": 0.6605504587155964, + "grad_norm": 0.33909982442855835, + "learning_rate": 2.598734638559977e-05, + "loss": 1.0335, + "step": 10872 + }, + { + "epoch": 0.6606112157482229, + "grad_norm": 1.763257384300232, + "learning_rate": 2.597895025961059e-05, + "loss": 1.1261, + "step": 10873 + }, + { + "epoch": 0.6606719727808494, + "grad_norm": 0.3060975968837738, + "learning_rate": 2.5970555014117605e-05, + "loss": 1.2755, + "step": 10874 + }, + { + "epoch": 0.6607327298134759, + "grad_norm": 0.483625203371048, + "learning_rate": 2.59621606494286e-05, + "loss": 1.09, + "step": 10875 + }, + { + "epoch": 0.6607934868461024, + "grad_norm": 0.1726621687412262, + "learning_rate": 2.5953767165851238e-05, + "loss": 1.0384, + "step": 10876 + }, + { + "epoch": 0.6608542438787289, + "grad_norm": 0.49828317761421204, + "learning_rate": 2.5945374563693185e-05, + "loss": 1.0328, + "step": 10877 + }, + { + "epoch": 0.6609150009113555, + "grad_norm": 0.38602131605148315, + "learning_rate": 2.5936982843262102e-05, + "loss": 1.038, + "step": 10878 + }, + { + "epoch": 0.660975757943982, + "grad_norm": 0.16009782254695892, + "learning_rate": 2.5928592004865526e-05, + "loss": 1.0214, + "step": 10879 + }, + { + "epoch": 0.6610365149766085, + "grad_norm": 0.14416159689426422, + "learning_rate": 2.5920202048811075e-05, + "loss": 1.0658, + "step": 10880 + }, + { + "epoch": 0.6610972720092351, + "grad_norm": 0.2827083170413971, + "learning_rate": 2.591181297540628e-05, + "loss": 1.1918, + "step": 10881 + }, + { + "epoch": 0.6611580290418616, + "grad_norm": 0.21076542139053345, + "learning_rate": 2.590342478495863e-05, + "loss": 1.0488, + "step": 10882 + }, + { + "epoch": 0.6612187860744881, + "grad_norm": 0.18713364005088806, + "learning_rate": 2.589503747777561e-05, + "loss": 1.0734, + "step": 10883 + }, + { + "epoch": 0.6612795431071147, + "grad_norm": 0.17940637469291687, + "learning_rate": 2.5886651054164647e-05, + "loss": 1.007, + "step": 10884 + }, + { + "epoch": 0.6613403001397412, + "grad_norm": 0.2840965986251831, + "learning_rate": 2.5878265514433148e-05, + "loss": 1.1014, + "step": 10885 + }, + { + "epoch": 0.6614010571723677, + "grad_norm": 0.2022271603345871, + "learning_rate": 2.586988085888849e-05, + "loss": 1.1154, + "step": 10886 + }, + { + "epoch": 0.6614618142049943, + "grad_norm": 0.3066585063934326, + "learning_rate": 2.5861497087838016e-05, + "loss": 0.9902, + "step": 10887 + }, + { + "epoch": 0.6615225712376207, + "grad_norm": 1.5410419702529907, + "learning_rate": 2.585311420158903e-05, + "loss": 1.0858, + "step": 10888 + }, + { + "epoch": 0.6615833282702472, + "grad_norm": 0.30357712507247925, + "learning_rate": 2.5844732200448813e-05, + "loss": 1.4205, + "step": 10889 + }, + { + "epoch": 0.6616440853028738, + "grad_norm": 5.377013206481934, + "learning_rate": 2.5836351084724587e-05, + "loss": 1.053, + "step": 10890 + }, + { + "epoch": 0.6617048423355003, + "grad_norm": 0.3312297463417053, + "learning_rate": 2.5827970854723615e-05, + "loss": 1.2416, + "step": 10891 + }, + { + "epoch": 0.6617655993681268, + "grad_norm": 0.343467116355896, + "learning_rate": 2.581959151075303e-05, + "loss": 1.119, + "step": 10892 + }, + { + "epoch": 0.6618263564007534, + "grad_norm": 0.14207501709461212, + "learning_rate": 2.581121305311999e-05, + "loss": 1.0723, + "step": 10893 + }, + { + "epoch": 0.6618871134333799, + "grad_norm": 0.18890921771526337, + "learning_rate": 2.5802835482131616e-05, + "loss": 1.0369, + "step": 10894 + }, + { + "epoch": 0.6619478704660064, + "grad_norm": 0.16411785781383514, + "learning_rate": 2.5794458798094955e-05, + "loss": 1.0683, + "step": 10895 + }, + { + "epoch": 0.662008627498633, + "grad_norm": 0.7556265592575073, + "learning_rate": 2.578608300131712e-05, + "loss": 1.012, + "step": 10896 + }, + { + "epoch": 0.6620693845312595, + "grad_norm": 0.2138395756483078, + "learning_rate": 2.577770809210508e-05, + "loss": 1.105, + "step": 10897 + }, + { + "epoch": 0.662130141563886, + "grad_norm": 0.5602432489395142, + "learning_rate": 2.5769334070765856e-05, + "loss": 1.0097, + "step": 10898 + }, + { + "epoch": 0.6621908985965126, + "grad_norm": 0.182882159948349, + "learning_rate": 2.5760960937606353e-05, + "loss": 1.059, + "step": 10899 + }, + { + "epoch": 0.6622516556291391, + "grad_norm": 0.22839391231536865, + "learning_rate": 2.57525886929335e-05, + "loss": 1.2029, + "step": 10900 + }, + { + "epoch": 0.6623124126617655, + "grad_norm": 0.2634769678115845, + "learning_rate": 2.574421733705421e-05, + "loss": 1.0108, + "step": 10901 + }, + { + "epoch": 0.6623731696943921, + "grad_norm": 0.16778959333896637, + "learning_rate": 2.5735846870275322e-05, + "loss": 0.9681, + "step": 10902 + }, + { + "epoch": 0.6624339267270186, + "grad_norm": 0.22366110980510712, + "learning_rate": 2.5727477292903655e-05, + "loss": 1.0987, + "step": 10903 + }, + { + "epoch": 0.6624946837596452, + "grad_norm": 0.2736739218235016, + "learning_rate": 2.5719108605246e-05, + "loss": 1.0678, + "step": 10904 + }, + { + "epoch": 0.6625554407922717, + "grad_norm": 0.1888788938522339, + "learning_rate": 2.5710740807609117e-05, + "loss": 1.164, + "step": 10905 + }, + { + "epoch": 0.6626161978248982, + "grad_norm": 0.8267762064933777, + "learning_rate": 2.5702373900299725e-05, + "loss": 1.3509, + "step": 10906 + }, + { + "epoch": 0.6626769548575248, + "grad_norm": 0.2201201170682907, + "learning_rate": 2.569400788362451e-05, + "loss": 1.0303, + "step": 10907 + }, + { + "epoch": 0.6627377118901513, + "grad_norm": 0.38446947932243347, + "learning_rate": 2.5685642757890138e-05, + "loss": 0.9632, + "step": 10908 + }, + { + "epoch": 0.6627984689227778, + "grad_norm": 0.2634279131889343, + "learning_rate": 2.5677278523403226e-05, + "loss": 1.1132, + "step": 10909 + }, + { + "epoch": 0.6628592259554044, + "grad_norm": 0.6921120285987854, + "learning_rate": 2.5668915180470376e-05, + "loss": 1.1708, + "step": 10910 + }, + { + "epoch": 0.6629199829880309, + "grad_norm": 0.21103468537330627, + "learning_rate": 2.5660552729398118e-05, + "loss": 1.1614, + "step": 10911 + }, + { + "epoch": 0.6629807400206574, + "grad_norm": 0.30797332525253296, + "learning_rate": 2.5652191170493045e-05, + "loss": 1.1117, + "step": 10912 + }, + { + "epoch": 0.663041497053284, + "grad_norm": 0.17501062154769897, + "learning_rate": 2.5643830504061594e-05, + "loss": 1.0945, + "step": 10913 + }, + { + "epoch": 0.6631022540859104, + "grad_norm": 0.20496191084384918, + "learning_rate": 2.5635470730410243e-05, + "loss": 1.0451, + "step": 10914 + }, + { + "epoch": 0.6631630111185369, + "grad_norm": 0.1966884583234787, + "learning_rate": 2.562711184984542e-05, + "loss": 1.0168, + "step": 10915 + }, + { + "epoch": 0.6632237681511635, + "grad_norm": 0.23123915493488312, + "learning_rate": 2.5618753862673506e-05, + "loss": 1.1167, + "step": 10916 + }, + { + "epoch": 0.66328452518379, + "grad_norm": 0.275493860244751, + "learning_rate": 2.5610396769200906e-05, + "loss": 1.0708, + "step": 10917 + }, + { + "epoch": 0.6633452822164165, + "grad_norm": 0.19200129806995392, + "learning_rate": 2.5602040569733936e-05, + "loss": 1.1544, + "step": 10918 + }, + { + "epoch": 0.6634060392490431, + "grad_norm": 0.25271710753440857, + "learning_rate": 2.559368526457887e-05, + "loss": 1.0096, + "step": 10919 + }, + { + "epoch": 0.6634667962816696, + "grad_norm": 0.1910032480955124, + "learning_rate": 2.5585330854041988e-05, + "loss": 1.073, + "step": 10920 + }, + { + "epoch": 0.6635275533142961, + "grad_norm": 0.3941117823123932, + "learning_rate": 2.557697733842951e-05, + "loss": 1.1439, + "step": 10921 + }, + { + "epoch": 0.6635883103469227, + "grad_norm": 0.1746150255203247, + "learning_rate": 2.556862471804766e-05, + "loss": 1.0191, + "step": 10922 + }, + { + "epoch": 0.6636490673795492, + "grad_norm": 0.22310568392276764, + "learning_rate": 2.5560272993202594e-05, + "loss": 1.0487, + "step": 10923 + }, + { + "epoch": 0.6637098244121757, + "grad_norm": 0.5468060970306396, + "learning_rate": 2.555192216420045e-05, + "loss": 1.2641, + "step": 10924 + }, + { + "epoch": 0.6637705814448023, + "grad_norm": 0.21269801259040833, + "learning_rate": 2.5543572231347345e-05, + "loss": 1.0662, + "step": 10925 + }, + { + "epoch": 0.6638313384774288, + "grad_norm": 0.13296100497245789, + "learning_rate": 2.553522319494928e-05, + "loss": 1.0158, + "step": 10926 + }, + { + "epoch": 0.6638920955100552, + "grad_norm": 0.17591160535812378, + "learning_rate": 2.552687505531236e-05, + "loss": 1.0383, + "step": 10927 + }, + { + "epoch": 0.6639528525426818, + "grad_norm": 0.30921345949172974, + "learning_rate": 2.551852781274257e-05, + "loss": 1.0309, + "step": 10928 + }, + { + "epoch": 0.6640136095753083, + "grad_norm": 0.1526738554239273, + "learning_rate": 2.5510181467545873e-05, + "loss": 1.0468, + "step": 10929 + }, + { + "epoch": 0.6640743666079348, + "grad_norm": 0.14515064656734467, + "learning_rate": 2.5501836020028208e-05, + "loss": 1.0715, + "step": 10930 + }, + { + "epoch": 0.6641351236405614, + "grad_norm": 0.28222352266311646, + "learning_rate": 2.549349147049548e-05, + "loss": 1.0249, + "step": 10931 + }, + { + "epoch": 0.6641958806731879, + "grad_norm": 0.1975465565919876, + "learning_rate": 2.5485147819253556e-05, + "loss": 1.0677, + "step": 10932 + }, + { + "epoch": 0.6642566377058144, + "grad_norm": 0.22246982157230377, + "learning_rate": 2.547680506660828e-05, + "loss": 1.0531, + "step": 10933 + }, + { + "epoch": 0.664317394738441, + "grad_norm": 0.2736908197402954, + "learning_rate": 2.5468463212865455e-05, + "loss": 1.2134, + "step": 10934 + }, + { + "epoch": 0.6643781517710675, + "grad_norm": 0.12339150905609131, + "learning_rate": 2.5460122258330853e-05, + "loss": 1.0419, + "step": 10935 + }, + { + "epoch": 0.664438908803694, + "grad_norm": 0.18034951388835907, + "learning_rate": 2.5451782203310208e-05, + "loss": 1.1107, + "step": 10936 + }, + { + "epoch": 0.6644996658363206, + "grad_norm": 0.22330309450626373, + "learning_rate": 2.5443443048109217e-05, + "loss": 1.0017, + "step": 10937 + }, + { + "epoch": 0.6645604228689471, + "grad_norm": 0.152618870139122, + "learning_rate": 2.5435104793033605e-05, + "loss": 1.0673, + "step": 10938 + }, + { + "epoch": 0.6646211799015737, + "grad_norm": 0.21899329125881195, + "learning_rate": 2.5426767438388954e-05, + "loss": 1.131, + "step": 10939 + }, + { + "epoch": 0.6646819369342001, + "grad_norm": 0.3989003896713257, + "learning_rate": 2.54184309844809e-05, + "loss": 1.1454, + "step": 10940 + }, + { + "epoch": 0.6647426939668266, + "grad_norm": 0.5937917828559875, + "learning_rate": 2.5410095431615e-05, + "loss": 1.1011, + "step": 10941 + }, + { + "epoch": 0.6648034509994531, + "grad_norm": 0.2024754136800766, + "learning_rate": 2.5401760780096794e-05, + "loss": 1.0301, + "step": 10942 + }, + { + "epoch": 0.6648642080320797, + "grad_norm": 0.22539468109607697, + "learning_rate": 2.5393427030231816e-05, + "loss": 1.1722, + "step": 10943 + }, + { + "epoch": 0.6649249650647062, + "grad_norm": 0.18778689205646515, + "learning_rate": 2.5385094182325525e-05, + "loss": 1.1002, + "step": 10944 + }, + { + "epoch": 0.6649857220973328, + "grad_norm": 0.2289135456085205, + "learning_rate": 2.537676223668338e-05, + "loss": 1.048, + "step": 10945 + }, + { + "epoch": 0.6650464791299593, + "grad_norm": 0.12880998849868774, + "learning_rate": 2.5368431193610763e-05, + "loss": 1.0391, + "step": 10946 + }, + { + "epoch": 0.6651072361625858, + "grad_norm": 0.19280768930912018, + "learning_rate": 2.5360101053413045e-05, + "loss": 1.0774, + "step": 10947 + }, + { + "epoch": 0.6651679931952124, + "grad_norm": 0.15065492689609528, + "learning_rate": 2.5351771816395597e-05, + "loss": 1.0655, + "step": 10948 + }, + { + "epoch": 0.6652287502278389, + "grad_norm": 0.1503314971923828, + "learning_rate": 2.5343443482863728e-05, + "loss": 1.0409, + "step": 10949 + }, + { + "epoch": 0.6652895072604654, + "grad_norm": 0.23627124726772308, + "learning_rate": 2.53351160531227e-05, + "loss": 0.9834, + "step": 10950 + }, + { + "epoch": 0.665350264293092, + "grad_norm": 0.49488502740859985, + "learning_rate": 2.5326789527477756e-05, + "loss": 1.1526, + "step": 10951 + }, + { + "epoch": 0.6654110213257185, + "grad_norm": 0.1916038691997528, + "learning_rate": 2.5318463906234113e-05, + "loss": 1.1377, + "step": 10952 + }, + { + "epoch": 0.6654717783583449, + "grad_norm": 0.519318699836731, + "learning_rate": 2.5310139189696945e-05, + "loss": 1.2742, + "step": 10953 + }, + { + "epoch": 0.6655325353909715, + "grad_norm": 0.20849274098873138, + "learning_rate": 2.53018153781714e-05, + "loss": 0.9833, + "step": 10954 + }, + { + "epoch": 0.665593292423598, + "grad_norm": 0.24163778126239777, + "learning_rate": 2.5293492471962587e-05, + "loss": 1.1532, + "step": 10955 + }, + { + "epoch": 0.6656540494562245, + "grad_norm": 0.2965753376483917, + "learning_rate": 2.528517047137558e-05, + "loss": 1.1128, + "step": 10956 + }, + { + "epoch": 0.6657148064888511, + "grad_norm": 0.25049108266830444, + "learning_rate": 2.527684937671543e-05, + "loss": 1.2467, + "step": 10957 + }, + { + "epoch": 0.6657755635214776, + "grad_norm": 0.12429855763912201, + "learning_rate": 2.5268529188287137e-05, + "loss": 1.04, + "step": 10958 + }, + { + "epoch": 0.6658363205541041, + "grad_norm": 0.1712392419576645, + "learning_rate": 2.52602099063957e-05, + "loss": 1.072, + "step": 10959 + }, + { + "epoch": 0.6658970775867307, + "grad_norm": 0.24250173568725586, + "learning_rate": 2.5251891531346038e-05, + "loss": 1.084, + "step": 10960 + }, + { + "epoch": 0.6659578346193572, + "grad_norm": 0.8053963780403137, + "learning_rate": 2.524357406344309e-05, + "loss": 0.9998, + "step": 10961 + }, + { + "epoch": 0.6660185916519837, + "grad_norm": 0.19810481369495392, + "learning_rate": 2.523525750299169e-05, + "loss": 1.1692, + "step": 10962 + }, + { + "epoch": 0.6660793486846103, + "grad_norm": 0.3961082398891449, + "learning_rate": 2.5226941850296737e-05, + "loss": 1.1178, + "step": 10963 + }, + { + "epoch": 0.6661401057172368, + "grad_norm": 0.1585647463798523, + "learning_rate": 2.5218627105663017e-05, + "loss": 1.0099, + "step": 10964 + }, + { + "epoch": 0.6662008627498633, + "grad_norm": 0.25369367003440857, + "learning_rate": 2.5210313269395326e-05, + "loss": 1.075, + "step": 10965 + }, + { + "epoch": 0.6662616197824899, + "grad_norm": 2.3266091346740723, + "learning_rate": 2.520200034179838e-05, + "loss": 1.0934, + "step": 10966 + }, + { + "epoch": 0.6663223768151163, + "grad_norm": 0.9236332178115845, + "learning_rate": 2.5193688323176882e-05, + "loss": 1.0706, + "step": 10967 + }, + { + "epoch": 0.6663831338477428, + "grad_norm": 0.15891312062740326, + "learning_rate": 2.518537721383556e-05, + "loss": 1.0429, + "step": 10968 + }, + { + "epoch": 0.6664438908803694, + "grad_norm": 0.13613361120224, + "learning_rate": 2.5177067014079026e-05, + "loss": 1.0159, + "step": 10969 + }, + { + "epoch": 0.6665046479129959, + "grad_norm": 0.18868862092494965, + "learning_rate": 2.51687577242119e-05, + "loss": 1.0157, + "step": 10970 + }, + { + "epoch": 0.6665654049456224, + "grad_norm": 0.15194952487945557, + "learning_rate": 2.516044934453876e-05, + "loss": 1.0796, + "step": 10971 + }, + { + "epoch": 0.666626161978249, + "grad_norm": 3.5493323802948, + "learning_rate": 2.5152141875364148e-05, + "loss": 1.0211, + "step": 10972 + }, + { + "epoch": 0.6666869190108755, + "grad_norm": 0.16274473071098328, + "learning_rate": 2.514383531699258e-05, + "loss": 1.0061, + "step": 10973 + }, + { + "epoch": 0.666747676043502, + "grad_norm": 0.20875021815299988, + "learning_rate": 2.5135529669728537e-05, + "loss": 1.1112, + "step": 10974 + }, + { + "epoch": 0.6668084330761286, + "grad_norm": 0.2209598422050476, + "learning_rate": 2.512722493387646e-05, + "loss": 1.0591, + "step": 10975 + }, + { + "epoch": 0.6668691901087551, + "grad_norm": 0.1865786463022232, + "learning_rate": 2.5118921109740757e-05, + "loss": 1.078, + "step": 10976 + }, + { + "epoch": 0.6669299471413817, + "grad_norm": 0.25866395235061646, + "learning_rate": 2.5110618197625806e-05, + "loss": 1.0105, + "step": 10977 + }, + { + "epoch": 0.6669907041740082, + "grad_norm": 0.21549391746520996, + "learning_rate": 2.510231619783594e-05, + "loss": 1.0858, + "step": 10978 + }, + { + "epoch": 0.6670514612066347, + "grad_norm": 0.1742464005947113, + "learning_rate": 2.509401511067553e-05, + "loss": 1.0673, + "step": 10979 + }, + { + "epoch": 0.6671122182392611, + "grad_norm": 0.40994831919670105, + "learning_rate": 2.508571493644879e-05, + "loss": 1.0436, + "step": 10980 + }, + { + "epoch": 0.6671729752718877, + "grad_norm": 0.8475384712219238, + "learning_rate": 2.5077415675459982e-05, + "loss": 1.022, + "step": 10981 + }, + { + "epoch": 0.6672337323045142, + "grad_norm": 0.2136261761188507, + "learning_rate": 2.506911732801332e-05, + "loss": 1.1737, + "step": 10982 + }, + { + "epoch": 0.6672944893371408, + "grad_norm": 0.3071012496948242, + "learning_rate": 2.5060819894412963e-05, + "loss": 1.1155, + "step": 10983 + }, + { + "epoch": 0.6673552463697673, + "grad_norm": 0.28201860189437866, + "learning_rate": 2.505252337496309e-05, + "loss": 1.0918, + "step": 10984 + }, + { + "epoch": 0.6674160034023938, + "grad_norm": 0.1462487280368805, + "learning_rate": 2.504422776996782e-05, + "loss": 1.0448, + "step": 10985 + }, + { + "epoch": 0.6674767604350204, + "grad_norm": 0.22179818153381348, + "learning_rate": 2.5035933079731176e-05, + "loss": 1.0473, + "step": 10986 + }, + { + "epoch": 0.6675375174676469, + "grad_norm": 0.20765475928783417, + "learning_rate": 2.5027639304557236e-05, + "loss": 1.1328, + "step": 10987 + }, + { + "epoch": 0.6675982745002734, + "grad_norm": 0.17914514243602753, + "learning_rate": 2.5019346444749992e-05, + "loss": 1.0287, + "step": 10988 + }, + { + "epoch": 0.6676590315329, + "grad_norm": 0.14450326561927795, + "learning_rate": 2.5011054500613446e-05, + "loss": 1.0057, + "step": 10989 + }, + { + "epoch": 0.6677197885655265, + "grad_norm": 0.155588299036026, + "learning_rate": 2.5002763472451528e-05, + "loss": 1.0331, + "step": 10990 + }, + { + "epoch": 0.667780545598153, + "grad_norm": 0.21720454096794128, + "learning_rate": 2.4994473360568145e-05, + "loss": 1.036, + "step": 10991 + }, + { + "epoch": 0.6678413026307796, + "grad_norm": 0.19388063251972198, + "learning_rate": 2.4986184165267195e-05, + "loss": 1.0527, + "step": 10992 + }, + { + "epoch": 0.667902059663406, + "grad_norm": 0.13869836926460266, + "learning_rate": 2.497789588685246e-05, + "loss": 1.04, + "step": 10993 + }, + { + "epoch": 0.6679628166960325, + "grad_norm": 0.14747163653373718, + "learning_rate": 2.4969608525627807e-05, + "loss": 1.0516, + "step": 10994 + }, + { + "epoch": 0.6680235737286591, + "grad_norm": 0.5105185508728027, + "learning_rate": 2.4961322081896994e-05, + "loss": 1.0632, + "step": 10995 + }, + { + "epoch": 0.6680843307612856, + "grad_norm": 0.21015068888664246, + "learning_rate": 2.4953036555963754e-05, + "loss": 1.0111, + "step": 10996 + }, + { + "epoch": 0.6681450877939121, + "grad_norm": 0.181083083152771, + "learning_rate": 2.4944751948131806e-05, + "loss": 1.0751, + "step": 10997 + }, + { + "epoch": 0.6682058448265387, + "grad_norm": 0.20577938854694366, + "learning_rate": 2.493646825870482e-05, + "loss": 1.0901, + "step": 10998 + }, + { + "epoch": 0.6682666018591652, + "grad_norm": 0.1656748205423355, + "learning_rate": 2.4928185487986433e-05, + "loss": 1.0071, + "step": 10999 + }, + { + "epoch": 0.6683273588917917, + "grad_norm": 0.14758124947547913, + "learning_rate": 2.491990363628025e-05, + "loss": 1.1109, + "step": 11000 + }, + { + "epoch": 0.6683881159244183, + "grad_norm": 0.6431008577346802, + "learning_rate": 2.4911622703889848e-05, + "loss": 1.1959, + "step": 11001 + }, + { + "epoch": 0.6684488729570448, + "grad_norm": 0.2164120376110077, + "learning_rate": 2.490334269111877e-05, + "loss": 1.1697, + "step": 11002 + }, + { + "epoch": 0.6685096299896713, + "grad_norm": 1.476911187171936, + "learning_rate": 2.489506359827051e-05, + "loss": 1.2094, + "step": 11003 + }, + { + "epoch": 0.6685703870222979, + "grad_norm": 0.19703175127506256, + "learning_rate": 2.4886785425648533e-05, + "loss": 1.0112, + "step": 11004 + }, + { + "epoch": 0.6686311440549244, + "grad_norm": 0.15433557331562042, + "learning_rate": 2.4878508173556327e-05, + "loss": 1.0065, + "step": 11005 + }, + { + "epoch": 0.6686919010875508, + "grad_norm": 0.45005103945732117, + "learning_rate": 2.4870231842297236e-05, + "loss": 1.1688, + "step": 11006 + }, + { + "epoch": 0.6687526581201774, + "grad_norm": 0.16318713128566742, + "learning_rate": 2.486195643217466e-05, + "loss": 1.0986, + "step": 11007 + }, + { + "epoch": 0.6688134151528039, + "grad_norm": 0.49916303157806396, + "learning_rate": 2.485368194349193e-05, + "loss": 1.0651, + "step": 11008 + }, + { + "epoch": 0.6688741721854304, + "grad_norm": 0.28305336833000183, + "learning_rate": 2.4845408376552325e-05, + "loss": 1.0533, + "step": 11009 + }, + { + "epoch": 0.668934929218057, + "grad_norm": 0.2912352383136749, + "learning_rate": 2.4837135731659168e-05, + "loss": 1.2482, + "step": 11010 + }, + { + "epoch": 0.6689956862506835, + "grad_norm": 0.16589808464050293, + "learning_rate": 2.4828864009115654e-05, + "loss": 1.0576, + "step": 11011 + }, + { + "epoch": 0.66905644328331, + "grad_norm": 0.2287047952413559, + "learning_rate": 2.4820593209225017e-05, + "loss": 1.1325, + "step": 11012 + }, + { + "epoch": 0.6691172003159366, + "grad_norm": 0.1572786271572113, + "learning_rate": 2.4812323332290384e-05, + "loss": 1.0992, + "step": 11013 + }, + { + "epoch": 0.6691779573485631, + "grad_norm": 0.18290910124778748, + "learning_rate": 2.4804054378614887e-05, + "loss": 1.0295, + "step": 11014 + }, + { + "epoch": 0.6692387143811896, + "grad_norm": 0.1316484957933426, + "learning_rate": 2.479578634850167e-05, + "loss": 1.0121, + "step": 11015 + }, + { + "epoch": 0.6692994714138162, + "grad_norm": 0.17698898911476135, + "learning_rate": 2.4787519242253776e-05, + "loss": 1.0377, + "step": 11016 + }, + { + "epoch": 0.6693602284464427, + "grad_norm": 0.15125074982643127, + "learning_rate": 2.4779253060174233e-05, + "loss": 1.0643, + "step": 11017 + }, + { + "epoch": 0.6694209854790693, + "grad_norm": 0.2028488665819168, + "learning_rate": 2.4770987802566054e-05, + "loss": 1.0597, + "step": 11018 + }, + { + "epoch": 0.6694817425116957, + "grad_norm": 0.1695825606584549, + "learning_rate": 2.4762723469732186e-05, + "loss": 1.0118, + "step": 11019 + }, + { + "epoch": 0.6695424995443222, + "grad_norm": 0.19617390632629395, + "learning_rate": 2.475446006197557e-05, + "loss": 1.0303, + "step": 11020 + }, + { + "epoch": 0.6696032565769487, + "grad_norm": 0.22058182954788208, + "learning_rate": 2.4746197579599095e-05, + "loss": 1.1491, + "step": 11021 + }, + { + "epoch": 0.6696640136095753, + "grad_norm": 0.18038059771060944, + "learning_rate": 2.4737936022905632e-05, + "loss": 1.1096, + "step": 11022 + }, + { + "epoch": 0.6697247706422018, + "grad_norm": 0.19032666087150574, + "learning_rate": 2.472967539219801e-05, + "loss": 1.0679, + "step": 11023 + }, + { + "epoch": 0.6697855276748284, + "grad_norm": 0.14399196207523346, + "learning_rate": 2.472141568777902e-05, + "loss": 1.0467, + "step": 11024 + }, + { + "epoch": 0.6698462847074549, + "grad_norm": 0.17795169353485107, + "learning_rate": 2.4713156909951424e-05, + "loss": 1.0071, + "step": 11025 + }, + { + "epoch": 0.6699070417400814, + "grad_norm": 0.2153054177761078, + "learning_rate": 2.470489905901795e-05, + "loss": 1.1091, + "step": 11026 + }, + { + "epoch": 0.669967798772708, + "grad_norm": 0.1929326355457306, + "learning_rate": 2.469664213528129e-05, + "loss": 1.0465, + "step": 11027 + }, + { + "epoch": 0.6700285558053345, + "grad_norm": 0.18597277998924255, + "learning_rate": 2.46883861390441e-05, + "loss": 1.085, + "step": 11028 + }, + { + "epoch": 0.670089312837961, + "grad_norm": 0.18936829268932343, + "learning_rate": 2.4680131070609015e-05, + "loss": 1.0632, + "step": 11029 + }, + { + "epoch": 0.6701500698705876, + "grad_norm": 0.2460111528635025, + "learning_rate": 2.4671876930278598e-05, + "loss": 1.1121, + "step": 11030 + }, + { + "epoch": 0.6702108269032141, + "grad_norm": 0.15778033435344696, + "learning_rate": 2.4663623718355444e-05, + "loss": 1.1002, + "step": 11031 + }, + { + "epoch": 0.6702715839358405, + "grad_norm": 0.22407162189483643, + "learning_rate": 2.4655371435142084e-05, + "loss": 1.0951, + "step": 11032 + }, + { + "epoch": 0.670332340968467, + "grad_norm": 0.17315182089805603, + "learning_rate": 2.4647120080940962e-05, + "loss": 1.0453, + "step": 11033 + }, + { + "epoch": 0.6703930980010936, + "grad_norm": 0.2125493437051773, + "learning_rate": 2.4638869656054554e-05, + "loss": 1.0994, + "step": 11034 + }, + { + "epoch": 0.6704538550337201, + "grad_norm": 3.5579452514648438, + "learning_rate": 2.463062016078526e-05, + "loss": 1.1151, + "step": 11035 + }, + { + "epoch": 0.6705146120663467, + "grad_norm": 0.13283728063106537, + "learning_rate": 2.4622371595435506e-05, + "loss": 1.0074, + "step": 11036 + }, + { + "epoch": 0.6705753690989732, + "grad_norm": 0.20148830115795135, + "learning_rate": 2.4614123960307634e-05, + "loss": 1.1444, + "step": 11037 + }, + { + "epoch": 0.6706361261315997, + "grad_norm": 0.20042972266674042, + "learning_rate": 2.4605877255703947e-05, + "loss": 1.0531, + "step": 11038 + }, + { + "epoch": 0.6706968831642263, + "grad_norm": 0.20785197615623474, + "learning_rate": 2.459763148192676e-05, + "loss": 1.0253, + "step": 11039 + }, + { + "epoch": 0.6707576401968528, + "grad_norm": 5.0876688957214355, + "learning_rate": 2.458938663927826e-05, + "loss": 0.9735, + "step": 11040 + }, + { + "epoch": 0.6708183972294793, + "grad_norm": 0.30966678261756897, + "learning_rate": 2.458114272806073e-05, + "loss": 1.0284, + "step": 11041 + }, + { + "epoch": 0.6708791542621059, + "grad_norm": 0.18019120395183563, + "learning_rate": 2.457289974857632e-05, + "loss": 1.0814, + "step": 11042 + }, + { + "epoch": 0.6709399112947324, + "grad_norm": 0.13959722220897675, + "learning_rate": 2.4564657701127193e-05, + "loss": 1.0011, + "step": 11043 + }, + { + "epoch": 0.6710006683273589, + "grad_norm": 0.2098534107208252, + "learning_rate": 2.4556416586015446e-05, + "loss": 1.0953, + "step": 11044 + }, + { + "epoch": 0.6710614253599854, + "grad_norm": 0.23567137122154236, + "learning_rate": 2.454817640354317e-05, + "loss": 1.0994, + "step": 11045 + }, + { + "epoch": 0.6711221823926119, + "grad_norm": 0.25719496607780457, + "learning_rate": 2.453993715401241e-05, + "loss": 1.1038, + "step": 11046 + }, + { + "epoch": 0.6711829394252384, + "grad_norm": 0.19635805487632751, + "learning_rate": 2.4531698837725175e-05, + "loss": 1.1122, + "step": 11047 + }, + { + "epoch": 0.671243696457865, + "grad_norm": 0.14776401221752167, + "learning_rate": 2.4523461454983436e-05, + "loss": 1.0596, + "step": 11048 + }, + { + "epoch": 0.6713044534904915, + "grad_norm": 0.1352836936712265, + "learning_rate": 2.4515225006089147e-05, + "loss": 1.0225, + "step": 11049 + }, + { + "epoch": 0.671365210523118, + "grad_norm": 0.26755475997924805, + "learning_rate": 2.4506989491344213e-05, + "loss": 1.1653, + "step": 11050 + }, + { + "epoch": 0.6714259675557446, + "grad_norm": 6.276671409606934, + "learning_rate": 2.4498754911050485e-05, + "loss": 0.9789, + "step": 11051 + }, + { + "epoch": 0.6714867245883711, + "grad_norm": 0.20639406144618988, + "learning_rate": 2.4490521265509863e-05, + "loss": 1.125, + "step": 11052 + }, + { + "epoch": 0.6715474816209976, + "grad_norm": 0.2370862364768982, + "learning_rate": 2.4482288555024097e-05, + "loss": 1.029, + "step": 11053 + }, + { + "epoch": 0.6716082386536242, + "grad_norm": 0.23639339208602905, + "learning_rate": 2.4474056779894978e-05, + "loss": 1.1837, + "step": 11054 + }, + { + "epoch": 0.6716689956862507, + "grad_norm": 0.19829902052879333, + "learning_rate": 2.4465825940424236e-05, + "loss": 1.1225, + "step": 11055 + }, + { + "epoch": 0.6717297527188772, + "grad_norm": 0.19954749941825867, + "learning_rate": 2.4457596036913567e-05, + "loss": 1.0775, + "step": 11056 + }, + { + "epoch": 0.6717905097515038, + "grad_norm": 0.19838108122348785, + "learning_rate": 2.444936706966467e-05, + "loss": 1.0712, + "step": 11057 + }, + { + "epoch": 0.6718512667841302, + "grad_norm": 0.14401325583457947, + "learning_rate": 2.444113903897916e-05, + "loss": 1.0334, + "step": 11058 + }, + { + "epoch": 0.6719120238167567, + "grad_norm": 0.145975723862648, + "learning_rate": 2.4432911945158653e-05, + "loss": 1.0019, + "step": 11059 + }, + { + "epoch": 0.6719727808493833, + "grad_norm": 0.3535212576389313, + "learning_rate": 2.442468578850467e-05, + "loss": 1.1775, + "step": 11060 + }, + { + "epoch": 0.6720335378820098, + "grad_norm": 0.206547349691391, + "learning_rate": 2.4416460569318796e-05, + "loss": 1.0877, + "step": 11061 + }, + { + "epoch": 0.6720942949146363, + "grad_norm": 0.2937861382961273, + "learning_rate": 2.4408236287902497e-05, + "loss": 1.0951, + "step": 11062 + }, + { + "epoch": 0.6721550519472629, + "grad_norm": 0.3986402153968811, + "learning_rate": 2.4400012944557244e-05, + "loss": 1.1216, + "step": 11063 + }, + { + "epoch": 0.6722158089798894, + "grad_norm": 0.17387425899505615, + "learning_rate": 2.439179053958447e-05, + "loss": 1.0512, + "step": 11064 + }, + { + "epoch": 0.672276566012516, + "grad_norm": 0.5248031616210938, + "learning_rate": 2.438356907328556e-05, + "loss": 1.2145, + "step": 11065 + }, + { + "epoch": 0.6723373230451425, + "grad_norm": 0.3510620594024658, + "learning_rate": 2.4375348545961878e-05, + "loss": 1.1028, + "step": 11066 + }, + { + "epoch": 0.672398080077769, + "grad_norm": 0.14842994511127472, + "learning_rate": 2.4367128957914746e-05, + "loss": 1.0475, + "step": 11067 + }, + { + "epoch": 0.6724588371103956, + "grad_norm": 0.3787466585636139, + "learning_rate": 2.435891030944546e-05, + "loss": 1.1714, + "step": 11068 + }, + { + "epoch": 0.6725195941430221, + "grad_norm": 0.20507948100566864, + "learning_rate": 2.4350692600855274e-05, + "loss": 0.9776, + "step": 11069 + }, + { + "epoch": 0.6725803511756486, + "grad_norm": 1.5550682544708252, + "learning_rate": 2.4342475832445405e-05, + "loss": 1.0736, + "step": 11070 + }, + { + "epoch": 0.6726411082082752, + "grad_norm": 0.9204466342926025, + "learning_rate": 2.4334260004517022e-05, + "loss": 1.253, + "step": 11071 + }, + { + "epoch": 0.6727018652409016, + "grad_norm": 0.1978798657655716, + "learning_rate": 2.4326045117371343e-05, + "loss": 1.1098, + "step": 11072 + }, + { + "epoch": 0.6727626222735281, + "grad_norm": 0.8262424468994141, + "learning_rate": 2.4317831171309423e-05, + "loss": 1.0754, + "step": 11073 + }, + { + "epoch": 0.6728233793061547, + "grad_norm": 0.20507487654685974, + "learning_rate": 2.430961816663237e-05, + "loss": 1.0422, + "step": 11074 + }, + { + "epoch": 0.6728841363387812, + "grad_norm": 0.18610550463199615, + "learning_rate": 2.430140610364122e-05, + "loss": 1.1004, + "step": 11075 + }, + { + "epoch": 0.6729448933714077, + "grad_norm": 0.2774583399295807, + "learning_rate": 2.4293194982636986e-05, + "loss": 1.11, + "step": 11076 + }, + { + "epoch": 0.6730056504040343, + "grad_norm": 0.2201559841632843, + "learning_rate": 2.4284984803920673e-05, + "loss": 1.0738, + "step": 11077 + }, + { + "epoch": 0.6730664074366608, + "grad_norm": 0.16596855223178864, + "learning_rate": 2.4276775567793214e-05, + "loss": 0.9997, + "step": 11078 + }, + { + "epoch": 0.6731271644692873, + "grad_norm": 0.16794554889202118, + "learning_rate": 2.4268567274555538e-05, + "loss": 1.0283, + "step": 11079 + }, + { + "epoch": 0.6731879215019139, + "grad_norm": 0.15709200501441956, + "learning_rate": 2.426035992450848e-05, + "loss": 0.9906, + "step": 11080 + }, + { + "epoch": 0.6732486785345404, + "grad_norm": 0.3478946089744568, + "learning_rate": 2.4252153517952884e-05, + "loss": 1.1353, + "step": 11081 + }, + { + "epoch": 0.6733094355671669, + "grad_norm": 0.3211744427680969, + "learning_rate": 2.4243948055189602e-05, + "loss": 1.1944, + "step": 11082 + }, + { + "epoch": 0.6733701925997935, + "grad_norm": 0.6750484704971313, + "learning_rate": 2.4235743536519383e-05, + "loss": 1.1815, + "step": 11083 + }, + { + "epoch": 0.67343094963242, + "grad_norm": 0.2308436781167984, + "learning_rate": 2.422753996224296e-05, + "loss": 1.0419, + "step": 11084 + }, + { + "epoch": 0.6734917066650464, + "grad_norm": 3.1786699295043945, + "learning_rate": 2.4219337332661045e-05, + "loss": 1.1096, + "step": 11085 + }, + { + "epoch": 0.673552463697673, + "grad_norm": 0.2450891137123108, + "learning_rate": 2.421113564807431e-05, + "loss": 1.1498, + "step": 11086 + }, + { + "epoch": 0.6736132207302995, + "grad_norm": 0.2958107590675354, + "learning_rate": 2.420293490878337e-05, + "loss": 1.1212, + "step": 11087 + }, + { + "epoch": 0.673673977762926, + "grad_norm": 0.3859425187110901, + "learning_rate": 2.4194735115088846e-05, + "loss": 1.2029, + "step": 11088 + }, + { + "epoch": 0.6737347347955526, + "grad_norm": 0.15924671292304993, + "learning_rate": 2.4186536267291287e-05, + "loss": 1.0857, + "step": 11089 + }, + { + "epoch": 0.6737954918281791, + "grad_norm": 0.23284924030303955, + "learning_rate": 2.4178338365691238e-05, + "loss": 1.077, + "step": 11090 + }, + { + "epoch": 0.6738562488608056, + "grad_norm": 0.23263482749462128, + "learning_rate": 2.4170141410589185e-05, + "loss": 1.2115, + "step": 11091 + }, + { + "epoch": 0.6739170058934322, + "grad_norm": 0.21997298300266266, + "learning_rate": 2.416194540228559e-05, + "loss": 1.0304, + "step": 11092 + }, + { + "epoch": 0.6739777629260587, + "grad_norm": 0.2990148961544037, + "learning_rate": 2.415375034108088e-05, + "loss": 1.3009, + "step": 11093 + }, + { + "epoch": 0.6740385199586852, + "grad_norm": 0.17576488852500916, + "learning_rate": 2.414555622727545e-05, + "loss": 1.0791, + "step": 11094 + }, + { + "epoch": 0.6740992769913118, + "grad_norm": 0.19239294528961182, + "learning_rate": 2.4137363061169656e-05, + "loss": 1.1216, + "step": 11095 + }, + { + "epoch": 0.6741600340239383, + "grad_norm": 0.40333837270736694, + "learning_rate": 2.4129170843063816e-05, + "loss": 1.0542, + "step": 11096 + }, + { + "epoch": 0.6742207910565648, + "grad_norm": 0.3484377861022949, + "learning_rate": 2.4120979573258195e-05, + "loss": 1.1068, + "step": 11097 + }, + { + "epoch": 0.6742815480891913, + "grad_norm": 0.8450833559036255, + "learning_rate": 2.4112789252053092e-05, + "loss": 1.1145, + "step": 11098 + }, + { + "epoch": 0.6743423051218178, + "grad_norm": 0.2068619281053543, + "learning_rate": 2.4104599879748724e-05, + "loss": 1.103, + "step": 11099 + }, + { + "epoch": 0.6744030621544443, + "grad_norm": 0.3146471083164215, + "learning_rate": 2.4096411456645235e-05, + "loss": 1.0489, + "step": 11100 + }, + { + "epoch": 0.6744638191870709, + "grad_norm": 0.1908671259880066, + "learning_rate": 2.408822398304279e-05, + "loss": 1.0369, + "step": 11101 + }, + { + "epoch": 0.6745245762196974, + "grad_norm": 0.19635352492332458, + "learning_rate": 2.4080037459241482e-05, + "loss": 1.0954, + "step": 11102 + }, + { + "epoch": 0.674585333252324, + "grad_norm": 0.15185540914535522, + "learning_rate": 2.4071851885541435e-05, + "loss": 1.0769, + "step": 11103 + }, + { + "epoch": 0.6746460902849505, + "grad_norm": 0.19959725439548492, + "learning_rate": 2.4063667262242662e-05, + "loss": 1.094, + "step": 11104 + }, + { + "epoch": 0.674706847317577, + "grad_norm": 0.2039928287267685, + "learning_rate": 2.4055483589645185e-05, + "loss": 1.0868, + "step": 11105 + }, + { + "epoch": 0.6747676043502036, + "grad_norm": 0.18697185814380646, + "learning_rate": 2.404730086804899e-05, + "loss": 1.0626, + "step": 11106 + }, + { + "epoch": 0.6748283613828301, + "grad_norm": 0.22379055619239807, + "learning_rate": 2.403911909775396e-05, + "loss": 1.004, + "step": 11107 + }, + { + "epoch": 0.6748891184154566, + "grad_norm": 0.5437240600585938, + "learning_rate": 2.4030938279060054e-05, + "loss": 1.1416, + "step": 11108 + }, + { + "epoch": 0.6749498754480832, + "grad_norm": 0.16238808631896973, + "learning_rate": 2.4022758412267126e-05, + "loss": 1.0336, + "step": 11109 + }, + { + "epoch": 0.6750106324807097, + "grad_norm": 0.3537892997264862, + "learning_rate": 2.401457949767501e-05, + "loss": 1.0863, + "step": 11110 + }, + { + "epoch": 0.6750713895133361, + "grad_norm": 0.3461466431617737, + "learning_rate": 2.4006401535583507e-05, + "loss": 0.9815, + "step": 11111 + }, + { + "epoch": 0.6751321465459627, + "grad_norm": 0.26014795899391174, + "learning_rate": 2.3998224526292377e-05, + "loss": 1.1775, + "step": 11112 + }, + { + "epoch": 0.6751929035785892, + "grad_norm": 0.14943625032901764, + "learning_rate": 2.3990048470101355e-05, + "loss": 1.0948, + "step": 11113 + }, + { + "epoch": 0.6752536606112157, + "grad_norm": 0.24261227250099182, + "learning_rate": 2.3981873367310136e-05, + "loss": 1.1228, + "step": 11114 + }, + { + "epoch": 0.6753144176438423, + "grad_norm": 0.3985610604286194, + "learning_rate": 2.3973699218218375e-05, + "loss": 1.1104, + "step": 11115 + }, + { + "epoch": 0.6753751746764688, + "grad_norm": 0.14404666423797607, + "learning_rate": 2.39655260231257e-05, + "loss": 1.0451, + "step": 11116 + }, + { + "epoch": 0.6754359317090953, + "grad_norm": 0.22327010333538055, + "learning_rate": 2.3957353782331706e-05, + "loss": 1.0971, + "step": 11117 + }, + { + "epoch": 0.6754966887417219, + "grad_norm": 0.23971788585186005, + "learning_rate": 2.3949182496135913e-05, + "loss": 1.0056, + "step": 11118 + }, + { + "epoch": 0.6755574457743484, + "grad_norm": 0.21099717915058136, + "learning_rate": 2.3941012164837917e-05, + "loss": 1.0424, + "step": 11119 + }, + { + "epoch": 0.6756182028069749, + "grad_norm": 0.2086886614561081, + "learning_rate": 2.3932842788737135e-05, + "loss": 1.1086, + "step": 11120 + }, + { + "epoch": 0.6756789598396015, + "grad_norm": 0.25138169527053833, + "learning_rate": 2.3924674368133042e-05, + "loss": 1.1164, + "step": 11121 + }, + { + "epoch": 0.675739716872228, + "grad_norm": 0.28817299008369446, + "learning_rate": 2.3916506903325053e-05, + "loss": 1.0486, + "step": 11122 + }, + { + "epoch": 0.6758004739048545, + "grad_norm": 0.2897702753543854, + "learning_rate": 2.390834039461252e-05, + "loss": 1.1302, + "step": 11123 + }, + { + "epoch": 0.675861230937481, + "grad_norm": 0.8575665354728699, + "learning_rate": 2.3900174842294836e-05, + "loss": 1.074, + "step": 11124 + }, + { + "epoch": 0.6759219879701075, + "grad_norm": 0.24161235988140106, + "learning_rate": 2.3892010246671287e-05, + "loss": 1.0705, + "step": 11125 + }, + { + "epoch": 0.675982745002734, + "grad_norm": 1.0593019723892212, + "learning_rate": 2.388384660804117e-05, + "loss": 1.1032, + "step": 11126 + }, + { + "epoch": 0.6760435020353606, + "grad_norm": 0.19444672763347626, + "learning_rate": 2.3875683926703683e-05, + "loss": 1.0939, + "step": 11127 + }, + { + "epoch": 0.6761042590679871, + "grad_norm": 0.18553577363491058, + "learning_rate": 2.386752220295803e-05, + "loss": 1.0718, + "step": 11128 + }, + { + "epoch": 0.6761650161006136, + "grad_norm": 0.33524391055107117, + "learning_rate": 2.3859361437103428e-05, + "loss": 1.2949, + "step": 11129 + }, + { + "epoch": 0.6762257731332402, + "grad_norm": 0.14581957459449768, + "learning_rate": 2.385120162943898e-05, + "loss": 1.0275, + "step": 11130 + }, + { + "epoch": 0.6762865301658667, + "grad_norm": 0.18833357095718384, + "learning_rate": 2.384304278026379e-05, + "loss": 1.1095, + "step": 11131 + }, + { + "epoch": 0.6763472871984932, + "grad_norm": 0.17623968422412872, + "learning_rate": 2.3834884889876925e-05, + "loss": 1.081, + "step": 11132 + }, + { + "epoch": 0.6764080442311198, + "grad_norm": 0.15596522390842438, + "learning_rate": 2.3826727958577412e-05, + "loss": 1.0518, + "step": 11133 + }, + { + "epoch": 0.6764688012637463, + "grad_norm": 0.20820166170597076, + "learning_rate": 2.381857198666424e-05, + "loss": 1.0181, + "step": 11134 + }, + { + "epoch": 0.6765295582963728, + "grad_norm": 0.219509556889534, + "learning_rate": 2.3810416974436378e-05, + "loss": 1.1095, + "step": 11135 + }, + { + "epoch": 0.6765903153289994, + "grad_norm": 0.13036322593688965, + "learning_rate": 2.3802262922192742e-05, + "loss": 1.0446, + "step": 11136 + }, + { + "epoch": 0.6766510723616258, + "grad_norm": 0.23557572066783905, + "learning_rate": 2.3794109830232226e-05, + "loss": 1.1137, + "step": 11137 + }, + { + "epoch": 0.6767118293942523, + "grad_norm": 0.3936331272125244, + "learning_rate": 2.378595769885368e-05, + "loss": 1.1259, + "step": 11138 + }, + { + "epoch": 0.6767725864268789, + "grad_norm": 2.1370038986206055, + "learning_rate": 2.3777806528355918e-05, + "loss": 1.0503, + "step": 11139 + }, + { + "epoch": 0.6768333434595054, + "grad_norm": 0.19872157275676727, + "learning_rate": 2.3769656319037724e-05, + "loss": 1.0054, + "step": 11140 + }, + { + "epoch": 0.676894100492132, + "grad_norm": 1.2535362243652344, + "learning_rate": 2.3761507071197854e-05, + "loss": 1.0494, + "step": 11141 + }, + { + "epoch": 0.6769548575247585, + "grad_norm": 0.5356118083000183, + "learning_rate": 2.3753358785135005e-05, + "loss": 1.0562, + "step": 11142 + }, + { + "epoch": 0.677015614557385, + "grad_norm": 0.3028128743171692, + "learning_rate": 2.374521146114787e-05, + "loss": 1.1766, + "step": 11143 + }, + { + "epoch": 0.6770763715900115, + "grad_norm": 0.18396839499473572, + "learning_rate": 2.3737065099535066e-05, + "loss": 1.0601, + "step": 11144 + }, + { + "epoch": 0.6771371286226381, + "grad_norm": 0.2422734498977661, + "learning_rate": 2.3728919700595233e-05, + "loss": 1.1758, + "step": 11145 + }, + { + "epoch": 0.6771978856552646, + "grad_norm": 0.2132338434457779, + "learning_rate": 2.372077526462695e-05, + "loss": 1.0577, + "step": 11146 + }, + { + "epoch": 0.6772586426878912, + "grad_norm": 0.23936808109283447, + "learning_rate": 2.3712631791928703e-05, + "loss": 1.1006, + "step": 11147 + }, + { + "epoch": 0.6773193997205177, + "grad_norm": 0.18331767618656158, + "learning_rate": 2.3704489282799025e-05, + "loss": 1.0921, + "step": 11148 + }, + { + "epoch": 0.6773801567531442, + "grad_norm": 0.16875343024730682, + "learning_rate": 2.3696347737536352e-05, + "loss": 1.0681, + "step": 11149 + }, + { + "epoch": 0.6774409137857706, + "grad_norm": 0.21655164659023285, + "learning_rate": 2.3688207156439156e-05, + "loss": 1.1579, + "step": 11150 + }, + { + "epoch": 0.6775016708183972, + "grad_norm": 0.2149423360824585, + "learning_rate": 2.3680067539805817e-05, + "loss": 1.0427, + "step": 11151 + }, + { + "epoch": 0.6775624278510237, + "grad_norm": 0.3484596610069275, + "learning_rate": 2.3671928887934687e-05, + "loss": 0.9838, + "step": 11152 + }, + { + "epoch": 0.6776231848836503, + "grad_norm": 0.26690927147865295, + "learning_rate": 2.366379120112409e-05, + "loss": 1.0969, + "step": 11153 + }, + { + "epoch": 0.6776839419162768, + "grad_norm": 0.3210144639015198, + "learning_rate": 2.3655654479672324e-05, + "loss": 1.1417, + "step": 11154 + }, + { + "epoch": 0.6777446989489033, + "grad_norm": 0.22170284390449524, + "learning_rate": 2.3647518723877632e-05, + "loss": 1.1104, + "step": 11155 + }, + { + "epoch": 0.6778054559815299, + "grad_norm": 0.2140231728553772, + "learning_rate": 2.3639383934038233e-05, + "loss": 1.062, + "step": 11156 + }, + { + "epoch": 0.6778662130141564, + "grad_norm": 0.2692912220954895, + "learning_rate": 2.3631250110452312e-05, + "loss": 1.0472, + "step": 11157 + }, + { + "epoch": 0.6779269700467829, + "grad_norm": 0.17037907242774963, + "learning_rate": 2.362311725341802e-05, + "loss": 1.0894, + "step": 11158 + }, + { + "epoch": 0.6779877270794095, + "grad_norm": 0.24714340269565582, + "learning_rate": 2.361498536323346e-05, + "loss": 1.0768, + "step": 11159 + }, + { + "epoch": 0.678048484112036, + "grad_norm": 0.19160813093185425, + "learning_rate": 2.3606854440196714e-05, + "loss": 1.0708, + "step": 11160 + }, + { + "epoch": 0.6781092411446625, + "grad_norm": 0.1635282039642334, + "learning_rate": 2.3598724484605815e-05, + "loss": 1.0502, + "step": 11161 + }, + { + "epoch": 0.6781699981772891, + "grad_norm": 0.24812854826450348, + "learning_rate": 2.3590595496758776e-05, + "loss": 1.0369, + "step": 11162 + }, + { + "epoch": 0.6782307552099155, + "grad_norm": 0.2336500883102417, + "learning_rate": 2.3582467476953563e-05, + "loss": 1.0164, + "step": 11163 + }, + { + "epoch": 0.678291512242542, + "grad_norm": 0.38903817534446716, + "learning_rate": 2.357434042548809e-05, + "loss": 0.9906, + "step": 11164 + }, + { + "epoch": 0.6783522692751686, + "grad_norm": 0.2568333148956299, + "learning_rate": 2.3566214342660304e-05, + "loss": 1.0233, + "step": 11165 + }, + { + "epoch": 0.6784130263077951, + "grad_norm": 0.3248317539691925, + "learning_rate": 2.3558089228768048e-05, + "loss": 1.2193, + "step": 11166 + }, + { + "epoch": 0.6784737833404216, + "grad_norm": 0.27131161093711853, + "learning_rate": 2.3549965084109132e-05, + "loss": 1.0929, + "step": 11167 + }, + { + "epoch": 0.6785345403730482, + "grad_norm": 0.2610524296760559, + "learning_rate": 2.3541841908981353e-05, + "loss": 1.1519, + "step": 11168 + }, + { + "epoch": 0.6785952974056747, + "grad_norm": 0.17496512830257416, + "learning_rate": 2.353371970368246e-05, + "loss": 1.0924, + "step": 11169 + }, + { + "epoch": 0.6786560544383012, + "grad_norm": 0.16868022084236145, + "learning_rate": 2.3525598468510208e-05, + "loss": 1.0545, + "step": 11170 + }, + { + "epoch": 0.6787168114709278, + "grad_norm": 0.18174971640110016, + "learning_rate": 2.3517478203762255e-05, + "loss": 1.0063, + "step": 11171 + }, + { + "epoch": 0.6787775685035543, + "grad_norm": 0.18934668600559235, + "learning_rate": 2.3509358909736256e-05, + "loss": 1.0338, + "step": 11172 + }, + { + "epoch": 0.6788383255361808, + "grad_norm": 0.1509687304496765, + "learning_rate": 2.350124058672985e-05, + "loss": 1.0443, + "step": 11173 + }, + { + "epoch": 0.6788990825688074, + "grad_norm": 1.0682342052459717, + "learning_rate": 2.3493123235040554e-05, + "loss": 1.0717, + "step": 11174 + }, + { + "epoch": 0.6789598396014339, + "grad_norm": 0.18538779020309448, + "learning_rate": 2.348500685496597e-05, + "loss": 1.0428, + "step": 11175 + }, + { + "epoch": 0.6790205966340604, + "grad_norm": 0.32143542170524597, + "learning_rate": 2.347689144680358e-05, + "loss": 1.2803, + "step": 11176 + }, + { + "epoch": 0.6790813536666869, + "grad_norm": 0.16064034402370453, + "learning_rate": 2.3468777010850863e-05, + "loss": 1.0804, + "step": 11177 + }, + { + "epoch": 0.6791421106993134, + "grad_norm": 0.1633293181657791, + "learning_rate": 2.3460663547405253e-05, + "loss": 1.0098, + "step": 11178 + }, + { + "epoch": 0.6792028677319399, + "grad_norm": 0.20480169355869293, + "learning_rate": 2.3452551056764156e-05, + "loss": 1.144, + "step": 11179 + }, + { + "epoch": 0.6792636247645665, + "grad_norm": 0.24426697194576263, + "learning_rate": 2.344443953922492e-05, + "loss": 1.1013, + "step": 11180 + }, + { + "epoch": 0.679324381797193, + "grad_norm": 0.170455664396286, + "learning_rate": 2.3436328995084895e-05, + "loss": 1.0054, + "step": 11181 + }, + { + "epoch": 0.6793851388298195, + "grad_norm": 0.23188428580760956, + "learning_rate": 2.3428219424641363e-05, + "loss": 1.1733, + "step": 11182 + }, + { + "epoch": 0.6794458958624461, + "grad_norm": 0.19851355254650116, + "learning_rate": 2.3420110828191584e-05, + "loss": 1.0342, + "step": 11183 + }, + { + "epoch": 0.6795066528950726, + "grad_norm": 0.5072343945503235, + "learning_rate": 2.3412003206032786e-05, + "loss": 1.1174, + "step": 11184 + }, + { + "epoch": 0.6795674099276992, + "grad_norm": 0.2877500653266907, + "learning_rate": 2.3403896558462118e-05, + "loss": 1.1002, + "step": 11185 + }, + { + "epoch": 0.6796281669603257, + "grad_norm": 0.47695472836494446, + "learning_rate": 2.3395790885776807e-05, + "loss": 1.1226, + "step": 11186 + }, + { + "epoch": 0.6796889239929522, + "grad_norm": 0.1909618228673935, + "learning_rate": 2.338768618827391e-05, + "loss": 1.0526, + "step": 11187 + }, + { + "epoch": 0.6797496810255788, + "grad_norm": 3.508739709854126, + "learning_rate": 2.3379582466250515e-05, + "loss": 1.0633, + "step": 11188 + }, + { + "epoch": 0.6798104380582053, + "grad_norm": 0.41409316658973694, + "learning_rate": 2.337147972000367e-05, + "loss": 1.0333, + "step": 11189 + }, + { + "epoch": 0.6798711950908317, + "grad_norm": 0.19435419142246246, + "learning_rate": 2.3363377949830368e-05, + "loss": 1.0445, + "step": 11190 + }, + { + "epoch": 0.6799319521234582, + "grad_norm": 0.19912934303283691, + "learning_rate": 2.335527715602761e-05, + "loss": 1.083, + "step": 11191 + }, + { + "epoch": 0.6799927091560848, + "grad_norm": 0.4310612678527832, + "learning_rate": 2.3347177338892318e-05, + "loss": 1.1046, + "step": 11192 + }, + { + "epoch": 0.6800534661887113, + "grad_norm": 0.6860024333000183, + "learning_rate": 2.3339078498721413e-05, + "loss": 1.1939, + "step": 11193 + }, + { + "epoch": 0.6801142232213379, + "grad_norm": 0.20367735624313354, + "learning_rate": 2.3330980635811722e-05, + "loss": 1.0724, + "step": 11194 + }, + { + "epoch": 0.6801749802539644, + "grad_norm": 0.7026233077049255, + "learning_rate": 2.3322883750460068e-05, + "loss": 1.0815, + "step": 11195 + }, + { + "epoch": 0.6802357372865909, + "grad_norm": 0.23127532005310059, + "learning_rate": 2.3314787842963286e-05, + "loss": 1.1596, + "step": 11196 + }, + { + "epoch": 0.6802964943192175, + "grad_norm": 0.13683272898197174, + "learning_rate": 2.3306692913618116e-05, + "loss": 1.0108, + "step": 11197 + }, + { + "epoch": 0.680357251351844, + "grad_norm": 0.1855653077363968, + "learning_rate": 2.329859896272128e-05, + "loss": 0.9997, + "step": 11198 + }, + { + "epoch": 0.6804180083844705, + "grad_norm": 0.24435703456401825, + "learning_rate": 2.3290505990569457e-05, + "loss": 1.1188, + "step": 11199 + }, + { + "epoch": 0.6804787654170971, + "grad_norm": 0.1810896396636963, + "learning_rate": 2.3282413997459295e-05, + "loss": 1.0671, + "step": 11200 + }, + { + "epoch": 0.6805395224497236, + "grad_norm": 0.18404601514339447, + "learning_rate": 2.3274322983687417e-05, + "loss": 1.0386, + "step": 11201 + }, + { + "epoch": 0.6806002794823501, + "grad_norm": 0.17581501603126526, + "learning_rate": 2.3266232949550397e-05, + "loss": 1.0398, + "step": 11202 + }, + { + "epoch": 0.6806610365149766, + "grad_norm": 0.17367181181907654, + "learning_rate": 2.3258143895344774e-05, + "loss": 1.0648, + "step": 11203 + }, + { + "epoch": 0.6807217935476031, + "grad_norm": 0.19805051386356354, + "learning_rate": 2.3250055821367056e-05, + "loss": 1.1179, + "step": 11204 + }, + { + "epoch": 0.6807825505802296, + "grad_norm": 3.0617306232452393, + "learning_rate": 2.3241968727913716e-05, + "loss": 1.0029, + "step": 11205 + }, + { + "epoch": 0.6808433076128562, + "grad_norm": 0.17073345184326172, + "learning_rate": 2.3233882615281184e-05, + "loss": 1.039, + "step": 11206 + }, + { + "epoch": 0.6809040646454827, + "grad_norm": 0.20036056637763977, + "learning_rate": 2.3225797483765855e-05, + "loss": 1.0363, + "step": 11207 + }, + { + "epoch": 0.6809648216781092, + "grad_norm": 0.2302485853433609, + "learning_rate": 2.321771333366409e-05, + "loss": 1.1018, + "step": 11208 + }, + { + "epoch": 0.6810255787107358, + "grad_norm": 0.1728275567293167, + "learning_rate": 2.320963016527223e-05, + "loss": 1.074, + "step": 11209 + }, + { + "epoch": 0.6810863357433623, + "grad_norm": 0.9067436456680298, + "learning_rate": 2.3201547978886544e-05, + "loss": 1.1704, + "step": 11210 + }, + { + "epoch": 0.6811470927759888, + "grad_norm": 0.34167492389678955, + "learning_rate": 2.3193466774803284e-05, + "loss": 1.0944, + "step": 11211 + }, + { + "epoch": 0.6812078498086154, + "grad_norm": 0.1406320482492447, + "learning_rate": 2.31853865533187e-05, + "loss": 1.0559, + "step": 11212 + }, + { + "epoch": 0.6812686068412419, + "grad_norm": 0.15037281811237335, + "learning_rate": 2.3177307314728974e-05, + "loss": 1.0542, + "step": 11213 + }, + { + "epoch": 0.6813293638738684, + "grad_norm": 0.24401754140853882, + "learning_rate": 2.3169229059330217e-05, + "loss": 1.0588, + "step": 11214 + }, + { + "epoch": 0.681390120906495, + "grad_norm": 0.21447701752185822, + "learning_rate": 2.316115178741855e-05, + "loss": 1.1251, + "step": 11215 + }, + { + "epoch": 0.6814508779391214, + "grad_norm": 0.22562271356582642, + "learning_rate": 2.3153075499290034e-05, + "loss": 1.0384, + "step": 11216 + }, + { + "epoch": 0.6815116349717479, + "grad_norm": 0.14040498435497284, + "learning_rate": 2.3145000195240745e-05, + "loss": 1.0174, + "step": 11217 + }, + { + "epoch": 0.6815723920043745, + "grad_norm": 0.29808682203292847, + "learning_rate": 2.313692587556667e-05, + "loss": 1.3053, + "step": 11218 + }, + { + "epoch": 0.681633149037001, + "grad_norm": 0.17840616405010223, + "learning_rate": 2.3128852540563762e-05, + "loss": 1.0553, + "step": 11219 + }, + { + "epoch": 0.6816939060696275, + "grad_norm": 0.21508195996284485, + "learning_rate": 2.3120780190527975e-05, + "loss": 1.0915, + "step": 11220 + }, + { + "epoch": 0.6817546631022541, + "grad_norm": 0.9147571921348572, + "learning_rate": 2.3112708825755153e-05, + "loss": 1.0457, + "step": 11221 + }, + { + "epoch": 0.6818154201348806, + "grad_norm": 0.5478297472000122, + "learning_rate": 2.3104638446541204e-05, + "loss": 1.1704, + "step": 11222 + }, + { + "epoch": 0.6818761771675071, + "grad_norm": 0.24100801348686218, + "learning_rate": 2.3096569053181933e-05, + "loss": 1.0067, + "step": 11223 + }, + { + "epoch": 0.6819369342001337, + "grad_norm": 0.2985419034957886, + "learning_rate": 2.308850064597312e-05, + "loss": 1.0274, + "step": 11224 + }, + { + "epoch": 0.6819976912327602, + "grad_norm": 1.4251482486724854, + "learning_rate": 2.3080433225210523e-05, + "loss": 1.0094, + "step": 11225 + }, + { + "epoch": 0.6820584482653868, + "grad_norm": 0.31346243619918823, + "learning_rate": 2.3072366791189847e-05, + "loss": 1.0816, + "step": 11226 + }, + { + "epoch": 0.6821192052980133, + "grad_norm": 0.15792682766914368, + "learning_rate": 2.3064301344206768e-05, + "loss": 1.04, + "step": 11227 + }, + { + "epoch": 0.6821799623306398, + "grad_norm": 0.19466426968574524, + "learning_rate": 2.3056236884556937e-05, + "loss": 1.0924, + "step": 11228 + }, + { + "epoch": 0.6822407193632662, + "grad_norm": 0.1803024709224701, + "learning_rate": 2.3048173412535945e-05, + "loss": 1.0707, + "step": 11229 + }, + { + "epoch": 0.6823014763958928, + "grad_norm": 0.33263081312179565, + "learning_rate": 2.3040110928439367e-05, + "loss": 1.0686, + "step": 11230 + }, + { + "epoch": 0.6823622334285193, + "grad_norm": 0.3582042157649994, + "learning_rate": 2.3032049432562734e-05, + "loss": 1.06, + "step": 11231 + }, + { + "epoch": 0.6824229904611459, + "grad_norm": 0.15211142599582672, + "learning_rate": 2.3023988925201527e-05, + "loss": 1.0452, + "step": 11232 + }, + { + "epoch": 0.6824837474937724, + "grad_norm": 0.27796515822410583, + "learning_rate": 2.3015929406651255e-05, + "loss": 1.1062, + "step": 11233 + }, + { + "epoch": 0.6825445045263989, + "grad_norm": 0.22183015942573547, + "learning_rate": 2.3007870877207288e-05, + "loss": 1.0068, + "step": 11234 + }, + { + "epoch": 0.6826052615590255, + "grad_norm": 0.26027265191078186, + "learning_rate": 2.2999813337165034e-05, + "loss": 1.1568, + "step": 11235 + }, + { + "epoch": 0.682666018591652, + "grad_norm": 0.18528512120246887, + "learning_rate": 2.2991756786819833e-05, + "loss": 1.0801, + "step": 11236 + }, + { + "epoch": 0.6827267756242785, + "grad_norm": 0.18161265552043915, + "learning_rate": 2.2983701226467e-05, + "loss": 0.9999, + "step": 11237 + }, + { + "epoch": 0.6827875326569051, + "grad_norm": 0.17594307661056519, + "learning_rate": 2.297564665640183e-05, + "loss": 1.0642, + "step": 11238 + }, + { + "epoch": 0.6828482896895316, + "grad_norm": 6.580755710601807, + "learning_rate": 2.2967593076919558e-05, + "loss": 1.1316, + "step": 11239 + }, + { + "epoch": 0.6829090467221581, + "grad_norm": 0.22040574252605438, + "learning_rate": 2.2959540488315408e-05, + "loss": 1.0124, + "step": 11240 + }, + { + "epoch": 0.6829698037547847, + "grad_norm": 0.20321033895015717, + "learning_rate": 2.295148889088451e-05, + "loss": 1.0961, + "step": 11241 + }, + { + "epoch": 0.6830305607874111, + "grad_norm": 0.18429984152317047, + "learning_rate": 2.2943438284921994e-05, + "loss": 1.1044, + "step": 11242 + }, + { + "epoch": 0.6830913178200376, + "grad_norm": 0.4786635637283325, + "learning_rate": 2.2935388670722996e-05, + "loss": 1.2073, + "step": 11243 + }, + { + "epoch": 0.6831520748526642, + "grad_norm": 0.15585094690322876, + "learning_rate": 2.2927340048582563e-05, + "loss": 1.0529, + "step": 11244 + }, + { + "epoch": 0.6832128318852907, + "grad_norm": 0.2617243230342865, + "learning_rate": 2.2919292418795706e-05, + "loss": 1.1023, + "step": 11245 + }, + { + "epoch": 0.6832735889179172, + "grad_norm": 0.13891160488128662, + "learning_rate": 2.2911245781657425e-05, + "loss": 1.0123, + "step": 11246 + }, + { + "epoch": 0.6833343459505438, + "grad_norm": 0.17713722586631775, + "learning_rate": 2.2903200137462668e-05, + "loss": 1.1189, + "step": 11247 + }, + { + "epoch": 0.6833951029831703, + "grad_norm": 0.3277253806591034, + "learning_rate": 2.2895155486506337e-05, + "loss": 0.9648, + "step": 11248 + }, + { + "epoch": 0.6834558600157968, + "grad_norm": 0.4767155349254608, + "learning_rate": 2.288711182908333e-05, + "loss": 1.0776, + "step": 11249 + }, + { + "epoch": 0.6835166170484234, + "grad_norm": 0.28766533732414246, + "learning_rate": 2.2879069165488477e-05, + "loss": 1.1746, + "step": 11250 + }, + { + "epoch": 0.6835773740810499, + "grad_norm": 0.19111867249011993, + "learning_rate": 2.287102749601659e-05, + "loss": 1.0675, + "step": 11251 + }, + { + "epoch": 0.6836381311136764, + "grad_norm": 0.35480576753616333, + "learning_rate": 2.2862986820962428e-05, + "loss": 1.0191, + "step": 11252 + }, + { + "epoch": 0.683698888146303, + "grad_norm": 0.26815661787986755, + "learning_rate": 2.285494714062073e-05, + "loss": 1.0483, + "step": 11253 + }, + { + "epoch": 0.6837596451789295, + "grad_norm": 0.19890573620796204, + "learning_rate": 2.2846908455286198e-05, + "loss": 1.1322, + "step": 11254 + }, + { + "epoch": 0.6838204022115559, + "grad_norm": 0.17780862748622894, + "learning_rate": 2.283887076525348e-05, + "loss": 1.0966, + "step": 11255 + }, + { + "epoch": 0.6838811592441825, + "grad_norm": 0.2777501344680786, + "learning_rate": 2.2830834070817204e-05, + "loss": 1.0406, + "step": 11256 + }, + { + "epoch": 0.683941916276809, + "grad_norm": 0.19387133419513702, + "learning_rate": 2.282279837227194e-05, + "loss": 1.0479, + "step": 11257 + }, + { + "epoch": 0.6840026733094355, + "grad_norm": 0.21947185695171356, + "learning_rate": 2.2814763669912276e-05, + "loss": 1.1643, + "step": 11258 + }, + { + "epoch": 0.6840634303420621, + "grad_norm": 0.19117461144924164, + "learning_rate": 2.2806729964032696e-05, + "loss": 1.0234, + "step": 11259 + }, + { + "epoch": 0.6841241873746886, + "grad_norm": 0.17242500185966492, + "learning_rate": 2.2798697254927716e-05, + "loss": 1.0603, + "step": 11260 + }, + { + "epoch": 0.6841849444073151, + "grad_norm": 0.16445142030715942, + "learning_rate": 2.2790665542891716e-05, + "loss": 1.0385, + "step": 11261 + }, + { + "epoch": 0.6842457014399417, + "grad_norm": 0.25113439559936523, + "learning_rate": 2.278263482821912e-05, + "loss": 1.1353, + "step": 11262 + }, + { + "epoch": 0.6843064584725682, + "grad_norm": 0.16451561450958252, + "learning_rate": 2.2774605111204317e-05, + "loss": 1.0438, + "step": 11263 + }, + { + "epoch": 0.6843672155051947, + "grad_norm": 0.3460811674594879, + "learning_rate": 2.2766576392141632e-05, + "loss": 1.1477, + "step": 11264 + }, + { + "epoch": 0.6844279725378213, + "grad_norm": 0.18698544800281525, + "learning_rate": 2.2758548671325348e-05, + "loss": 1.0236, + "step": 11265 + }, + { + "epoch": 0.6844887295704478, + "grad_norm": 0.3839779496192932, + "learning_rate": 2.275052194904973e-05, + "loss": 1.1806, + "step": 11266 + }, + { + "epoch": 0.6845494866030744, + "grad_norm": 0.23553353548049927, + "learning_rate": 2.2742496225609e-05, + "loss": 1.0373, + "step": 11267 + }, + { + "epoch": 0.6846102436357008, + "grad_norm": 0.37944063544273376, + "learning_rate": 2.273447150129734e-05, + "loss": 1.1049, + "step": 11268 + }, + { + "epoch": 0.6846710006683273, + "grad_norm": 0.1851978600025177, + "learning_rate": 2.2726447776408894e-05, + "loss": 1.0685, + "step": 11269 + }, + { + "epoch": 0.6847317577009538, + "grad_norm": 0.19793568551540375, + "learning_rate": 2.271842505123778e-05, + "loss": 1.0422, + "step": 11270 + }, + { + "epoch": 0.6847925147335804, + "grad_norm": 0.212863028049469, + "learning_rate": 2.2710403326078066e-05, + "loss": 1.0474, + "step": 11271 + }, + { + "epoch": 0.6848532717662069, + "grad_norm": 0.21480274200439453, + "learning_rate": 2.270238260122379e-05, + "loss": 1.0763, + "step": 11272 + }, + { + "epoch": 0.6849140287988335, + "grad_norm": 0.1596066653728485, + "learning_rate": 2.2694362876968965e-05, + "loss": 1.0465, + "step": 11273 + }, + { + "epoch": 0.68497478583146, + "grad_norm": 0.18858887255191803, + "learning_rate": 2.268634415360754e-05, + "loss": 1.0094, + "step": 11274 + }, + { + "epoch": 0.6850355428640865, + "grad_norm": 0.2269919067621231, + "learning_rate": 2.2678326431433454e-05, + "loss": 1.0913, + "step": 11275 + }, + { + "epoch": 0.6850962998967131, + "grad_norm": 0.7135042548179626, + "learning_rate": 2.267030971074059e-05, + "loss": 1.2034, + "step": 11276 + }, + { + "epoch": 0.6851570569293396, + "grad_norm": 0.17940901219844818, + "learning_rate": 2.2662293991822807e-05, + "loss": 1.0175, + "step": 11277 + }, + { + "epoch": 0.6852178139619661, + "grad_norm": 0.21202413737773895, + "learning_rate": 2.2654279274973905e-05, + "loss": 1.035, + "step": 11278 + }, + { + "epoch": 0.6852785709945927, + "grad_norm": 0.2816675007343292, + "learning_rate": 2.26462655604877e-05, + "loss": 1.1469, + "step": 11279 + }, + { + "epoch": 0.6853393280272192, + "grad_norm": 0.2332669049501419, + "learning_rate": 2.2638252848657938e-05, + "loss": 1.0268, + "step": 11280 + }, + { + "epoch": 0.6854000850598457, + "grad_norm": 0.23355479538440704, + "learning_rate": 2.2630241139778286e-05, + "loss": 1.0487, + "step": 11281 + }, + { + "epoch": 0.6854608420924722, + "grad_norm": 0.21977652609348297, + "learning_rate": 2.2622230434142438e-05, + "loss": 1.1666, + "step": 11282 + }, + { + "epoch": 0.6855215991250987, + "grad_norm": 0.2945498824119568, + "learning_rate": 2.2614220732044006e-05, + "loss": 1.1174, + "step": 11283 + }, + { + "epoch": 0.6855823561577252, + "grad_norm": 1.6675939559936523, + "learning_rate": 2.2606212033776625e-05, + "loss": 1.1673, + "step": 11284 + }, + { + "epoch": 0.6856431131903518, + "grad_norm": 0.3116815686225891, + "learning_rate": 2.2598204339633833e-05, + "loss": 1.1359, + "step": 11285 + }, + { + "epoch": 0.6857038702229783, + "grad_norm": 0.19051456451416016, + "learning_rate": 2.259019764990916e-05, + "loss": 1.0671, + "step": 11286 + }, + { + "epoch": 0.6857646272556048, + "grad_norm": 0.13889628648757935, + "learning_rate": 2.258219196489611e-05, + "loss": 1.0345, + "step": 11287 + }, + { + "epoch": 0.6858253842882314, + "grad_norm": 0.2179017812013626, + "learning_rate": 2.2574187284888076e-05, + "loss": 1.0352, + "step": 11288 + }, + { + "epoch": 0.6858861413208579, + "grad_norm": 0.22188392281532288, + "learning_rate": 2.2566183610178525e-05, + "loss": 1.0148, + "step": 11289 + }, + { + "epoch": 0.6859468983534844, + "grad_norm": 0.2314874678850174, + "learning_rate": 2.2558180941060823e-05, + "loss": 1.1924, + "step": 11290 + }, + { + "epoch": 0.686007655386111, + "grad_norm": 0.19055108726024628, + "learning_rate": 2.25501792778283e-05, + "loss": 1.0713, + "step": 11291 + }, + { + "epoch": 0.6860684124187375, + "grad_norm": 0.19924232363700867, + "learning_rate": 2.2542178620774257e-05, + "loss": 1.014, + "step": 11292 + }, + { + "epoch": 0.686129169451364, + "grad_norm": 0.15228872001171112, + "learning_rate": 2.2534178970191966e-05, + "loss": 0.9752, + "step": 11293 + }, + { + "epoch": 0.6861899264839906, + "grad_norm": 0.13316605985164642, + "learning_rate": 2.2526180326374656e-05, + "loss": 1.0204, + "step": 11294 + }, + { + "epoch": 0.686250683516617, + "grad_norm": 0.14829912781715393, + "learning_rate": 2.2518182689615518e-05, + "loss": 1.0148, + "step": 11295 + }, + { + "epoch": 0.6863114405492435, + "grad_norm": 0.14959383010864258, + "learning_rate": 2.2510186060207707e-05, + "loss": 1.0272, + "step": 11296 + }, + { + "epoch": 0.6863721975818701, + "grad_norm": 0.14643090963363647, + "learning_rate": 2.250219043844433e-05, + "loss": 1.0315, + "step": 11297 + }, + { + "epoch": 0.6864329546144966, + "grad_norm": 0.16283442080020905, + "learning_rate": 2.2494195824618485e-05, + "loss": 1.0372, + "step": 11298 + }, + { + "epoch": 0.6864937116471231, + "grad_norm": 0.16768315434455872, + "learning_rate": 2.2486202219023188e-05, + "loss": 1.0662, + "step": 11299 + }, + { + "epoch": 0.6865544686797497, + "grad_norm": 0.12032515555620193, + "learning_rate": 2.2478209621951508e-05, + "loss": 1.0163, + "step": 11300 + }, + { + "epoch": 0.6866152257123762, + "grad_norm": 0.22903306782245636, + "learning_rate": 2.247021803369635e-05, + "loss": 1.1028, + "step": 11301 + }, + { + "epoch": 0.6866759827450027, + "grad_norm": 0.14379461109638214, + "learning_rate": 2.2462227454550677e-05, + "loss": 1.1773, + "step": 11302 + }, + { + "epoch": 0.6867367397776293, + "grad_norm": 0.17039012908935547, + "learning_rate": 2.245423788480737e-05, + "loss": 1.1126, + "step": 11303 + }, + { + "epoch": 0.6867974968102558, + "grad_norm": 0.19883230328559875, + "learning_rate": 2.244624932475929e-05, + "loss": 1.0568, + "step": 11304 + }, + { + "epoch": 0.6868582538428823, + "grad_norm": 0.13795390725135803, + "learning_rate": 2.2438261774699283e-05, + "loss": 1.004, + "step": 11305 + }, + { + "epoch": 0.6869190108755089, + "grad_norm": 0.20832762122154236, + "learning_rate": 2.2430275234920118e-05, + "loss": 1.1307, + "step": 11306 + }, + { + "epoch": 0.6869797679081354, + "grad_norm": 0.16005542874336243, + "learning_rate": 2.242228970571456e-05, + "loss": 1.0764, + "step": 11307 + }, + { + "epoch": 0.6870405249407618, + "grad_norm": 0.1926472932100296, + "learning_rate": 2.2414305187375286e-05, + "loss": 1.1899, + "step": 11308 + }, + { + "epoch": 0.6871012819733884, + "grad_norm": 0.17942726612091064, + "learning_rate": 2.2406321680194965e-05, + "loss": 1.122, + "step": 11309 + }, + { + "epoch": 0.6871620390060149, + "grad_norm": 0.12999144196510315, + "learning_rate": 2.2398339184466278e-05, + "loss": 1.0033, + "step": 11310 + }, + { + "epoch": 0.6872227960386414, + "grad_norm": 0.20175376534461975, + "learning_rate": 2.2390357700481808e-05, + "loss": 1.1422, + "step": 11311 + }, + { + "epoch": 0.687283553071268, + "grad_norm": 0.18946079909801483, + "learning_rate": 2.238237722853411e-05, + "loss": 1.1267, + "step": 11312 + }, + { + "epoch": 0.6873443101038945, + "grad_norm": 0.14222770929336548, + "learning_rate": 2.237439776891573e-05, + "loss": 1.0206, + "step": 11313 + }, + { + "epoch": 0.687405067136521, + "grad_norm": 0.20201309025287628, + "learning_rate": 2.2366419321919102e-05, + "loss": 1.1295, + "step": 11314 + }, + { + "epoch": 0.6874658241691476, + "grad_norm": 0.36559391021728516, + "learning_rate": 2.2358441887836735e-05, + "loss": 1.3033, + "step": 11315 + }, + { + "epoch": 0.6875265812017741, + "grad_norm": 0.29848337173461914, + "learning_rate": 2.2350465466961024e-05, + "loss": 1.1858, + "step": 11316 + }, + { + "epoch": 0.6875873382344007, + "grad_norm": 0.132566437125206, + "learning_rate": 2.2342490059584347e-05, + "loss": 1.0598, + "step": 11317 + }, + { + "epoch": 0.6876480952670272, + "grad_norm": 0.18899880349636078, + "learning_rate": 2.233451566599904e-05, + "loss": 1.0682, + "step": 11318 + }, + { + "epoch": 0.6877088522996537, + "grad_norm": 0.171379953622818, + "learning_rate": 2.232654228649741e-05, + "loss": 1.0827, + "step": 11319 + }, + { + "epoch": 0.6877696093322803, + "grad_norm": 0.14324839413166046, + "learning_rate": 2.2318569921371725e-05, + "loss": 1.1639, + "step": 11320 + }, + { + "epoch": 0.6878303663649067, + "grad_norm": 0.13648715615272522, + "learning_rate": 2.2310598570914204e-05, + "loss": 1.0107, + "step": 11321 + }, + { + "epoch": 0.6878911233975332, + "grad_norm": 0.23263072967529297, + "learning_rate": 2.230262823541705e-05, + "loss": 1.166, + "step": 11322 + }, + { + "epoch": 0.6879518804301598, + "grad_norm": 0.19581371545791626, + "learning_rate": 2.229465891517241e-05, + "loss": 1.1124, + "step": 11323 + }, + { + "epoch": 0.6880126374627863, + "grad_norm": 0.2769557237625122, + "learning_rate": 2.2286690610472404e-05, + "loss": 1.2821, + "step": 11324 + }, + { + "epoch": 0.6880733944954128, + "grad_norm": 0.16085684299468994, + "learning_rate": 2.227872332160909e-05, + "loss": 1.0957, + "step": 11325 + }, + { + "epoch": 0.6881341515280394, + "grad_norm": 0.24761411547660828, + "learning_rate": 2.2270757048874553e-05, + "loss": 1.1874, + "step": 11326 + }, + { + "epoch": 0.6881949085606659, + "grad_norm": 0.16628389060497284, + "learning_rate": 2.226279179256079e-05, + "loss": 0.9982, + "step": 11327 + }, + { + "epoch": 0.6882556655932924, + "grad_norm": 0.1527760922908783, + "learning_rate": 2.225482755295974e-05, + "loss": 1.0424, + "step": 11328 + }, + { + "epoch": 0.688316422625919, + "grad_norm": 0.14394831657409668, + "learning_rate": 2.2246864330363348e-05, + "loss": 1.0152, + "step": 11329 + }, + { + "epoch": 0.6883771796585455, + "grad_norm": 0.1702050119638443, + "learning_rate": 2.2238902125063487e-05, + "loss": 1.0403, + "step": 11330 + }, + { + "epoch": 0.688437936691172, + "grad_norm": 0.13897405564785004, + "learning_rate": 2.2230940937352057e-05, + "loss": 1.0261, + "step": 11331 + }, + { + "epoch": 0.6884986937237986, + "grad_norm": 0.140434131026268, + "learning_rate": 2.2222980767520856e-05, + "loss": 1.0222, + "step": 11332 + }, + { + "epoch": 0.6885594507564251, + "grad_norm": 0.15677984058856964, + "learning_rate": 2.2215021615861658e-05, + "loss": 1.0072, + "step": 11333 + }, + { + "epoch": 0.6886202077890515, + "grad_norm": 0.1727035641670227, + "learning_rate": 2.220706348266624e-05, + "loss": 1.2287, + "step": 11334 + }, + { + "epoch": 0.6886809648216781, + "grad_norm": 2.200191020965576, + "learning_rate": 2.219910636822624e-05, + "loss": 1.0754, + "step": 11335 + }, + { + "epoch": 0.6887417218543046, + "grad_norm": 0.4675142467021942, + "learning_rate": 2.219115027283339e-05, + "loss": 1.2032, + "step": 11336 + }, + { + "epoch": 0.6888024788869311, + "grad_norm": 0.19623936712741852, + "learning_rate": 2.21831951967793e-05, + "loss": 1.0519, + "step": 11337 + }, + { + "epoch": 0.6888632359195577, + "grad_norm": 0.22839213907718658, + "learning_rate": 2.2175241140355568e-05, + "loss": 1.0324, + "step": 11338 + }, + { + "epoch": 0.6889239929521842, + "grad_norm": 0.2076042890548706, + "learning_rate": 2.216728810385375e-05, + "loss": 1.099, + "step": 11339 + }, + { + "epoch": 0.6889847499848107, + "grad_norm": 0.1393222063779831, + "learning_rate": 2.215933608756537e-05, + "loss": 1.0584, + "step": 11340 + }, + { + "epoch": 0.6890455070174373, + "grad_norm": 0.1925438791513443, + "learning_rate": 2.2151385091781907e-05, + "loss": 1.0081, + "step": 11341 + }, + { + "epoch": 0.6891062640500638, + "grad_norm": 2.038776397705078, + "learning_rate": 2.2143435116794803e-05, + "loss": 1.0655, + "step": 11342 + }, + { + "epoch": 0.6891670210826903, + "grad_norm": 0.28692448139190674, + "learning_rate": 2.2135486162895468e-05, + "loss": 1.0754, + "step": 11343 + }, + { + "epoch": 0.6892277781153169, + "grad_norm": 0.19050706923007965, + "learning_rate": 2.2127538230375277e-05, + "loss": 1.0829, + "step": 11344 + }, + { + "epoch": 0.6892885351479434, + "grad_norm": 8.978743553161621, + "learning_rate": 2.2119591319525557e-05, + "loss": 1.0621, + "step": 11345 + }, + { + "epoch": 0.68934929218057, + "grad_norm": 0.27118831872940063, + "learning_rate": 2.211164543063759e-05, + "loss": 1.0741, + "step": 11346 + }, + { + "epoch": 0.6894100492131964, + "grad_norm": 0.2255370318889618, + "learning_rate": 2.2103700564002688e-05, + "loss": 1.074, + "step": 11347 + }, + { + "epoch": 0.6894708062458229, + "grad_norm": 0.24943530559539795, + "learning_rate": 2.2095756719912014e-05, + "loss": 1.0757, + "step": 11348 + }, + { + "epoch": 0.6895315632784494, + "grad_norm": 0.16472876071929932, + "learning_rate": 2.2087813898656774e-05, + "loss": 1.0199, + "step": 11349 + }, + { + "epoch": 0.689592320311076, + "grad_norm": 0.14655499160289764, + "learning_rate": 2.2079872100528108e-05, + "loss": 1.0738, + "step": 11350 + }, + { + "epoch": 0.6896530773437025, + "grad_norm": 0.13503171503543854, + "learning_rate": 2.2071931325817114e-05, + "loss": 1.0126, + "step": 11351 + }, + { + "epoch": 0.689713834376329, + "grad_norm": 0.13849663734436035, + "learning_rate": 2.2063991574814884e-05, + "loss": 1.0029, + "step": 11352 + }, + { + "epoch": 0.6897745914089556, + "grad_norm": 0.1910209208726883, + "learning_rate": 2.2056052847812454e-05, + "loss": 1.0502, + "step": 11353 + }, + { + "epoch": 0.6898353484415821, + "grad_norm": 0.2907213866710663, + "learning_rate": 2.2048115145100817e-05, + "loss": 1.127, + "step": 11354 + }, + { + "epoch": 0.6898961054742087, + "grad_norm": 4.602910995483398, + "learning_rate": 2.204017846697089e-05, + "loss": 1.1551, + "step": 11355 + }, + { + "epoch": 0.6899568625068352, + "grad_norm": 0.4151572287082672, + "learning_rate": 2.2032242813713643e-05, + "loss": 1.0565, + "step": 11356 + }, + { + "epoch": 0.6900176195394617, + "grad_norm": 0.17964577674865723, + "learning_rate": 2.2024308185619946e-05, + "loss": 1.0526, + "step": 11357 + }, + { + "epoch": 0.6900783765720883, + "grad_norm": 0.5173928141593933, + "learning_rate": 2.2016374582980638e-05, + "loss": 1.2322, + "step": 11358 + }, + { + "epoch": 0.6901391336047148, + "grad_norm": 1.9041574001312256, + "learning_rate": 2.200844200608652e-05, + "loss": 1.0078, + "step": 11359 + }, + { + "epoch": 0.6901998906373412, + "grad_norm": 0.16573524475097656, + "learning_rate": 2.200051045522838e-05, + "loss": 1.0667, + "step": 11360 + }, + { + "epoch": 0.6902606476699678, + "grad_norm": 0.21009685099124908, + "learning_rate": 2.1992579930696945e-05, + "loss": 1.0307, + "step": 11361 + }, + { + "epoch": 0.6903214047025943, + "grad_norm": 0.17714422941207886, + "learning_rate": 2.19846504327829e-05, + "loss": 1.0811, + "step": 11362 + }, + { + "epoch": 0.6903821617352208, + "grad_norm": 0.2438088059425354, + "learning_rate": 2.197672196177691e-05, + "loss": 1.1472, + "step": 11363 + }, + { + "epoch": 0.6904429187678474, + "grad_norm": 0.15369157493114471, + "learning_rate": 2.1968794517969588e-05, + "loss": 1.0944, + "step": 11364 + }, + { + "epoch": 0.6905036758004739, + "grad_norm": 0.17585086822509766, + "learning_rate": 2.1960868101651527e-05, + "loss": 1.0238, + "step": 11365 + }, + { + "epoch": 0.6905644328331004, + "grad_norm": 0.33924809098243713, + "learning_rate": 2.1952942713113246e-05, + "loss": 1.0834, + "step": 11366 + }, + { + "epoch": 0.690625189865727, + "grad_norm": 0.13669896125793457, + "learning_rate": 2.1945018352645308e-05, + "loss": 1.0282, + "step": 11367 + }, + { + "epoch": 0.6906859468983535, + "grad_norm": 0.22358156740665436, + "learning_rate": 2.193709502053813e-05, + "loss": 1.1051, + "step": 11368 + }, + { + "epoch": 0.69074670393098, + "grad_norm": 0.2107873558998108, + "learning_rate": 2.1929172717082152e-05, + "loss": 1.1208, + "step": 11369 + }, + { + "epoch": 0.6908074609636066, + "grad_norm": 0.21886497735977173, + "learning_rate": 2.1921251442567782e-05, + "loss": 1.1293, + "step": 11370 + }, + { + "epoch": 0.6908682179962331, + "grad_norm": 0.2542925775051117, + "learning_rate": 2.191333119728534e-05, + "loss": 1.0513, + "step": 11371 + }, + { + "epoch": 0.6909289750288596, + "grad_norm": 0.1539592742919922, + "learning_rate": 2.1905411981525197e-05, + "loss": 1.0377, + "step": 11372 + }, + { + "epoch": 0.6909897320614861, + "grad_norm": 0.19022776186466217, + "learning_rate": 2.1897493795577606e-05, + "loss": 1.0494, + "step": 11373 + }, + { + "epoch": 0.6910504890941126, + "grad_norm": 0.5520213842391968, + "learning_rate": 2.1889576639732834e-05, + "loss": 1.1947, + "step": 11374 + }, + { + "epoch": 0.6911112461267391, + "grad_norm": 0.2908802032470703, + "learning_rate": 2.1881660514281043e-05, + "loss": 1.2382, + "step": 11375 + }, + { + "epoch": 0.6911720031593657, + "grad_norm": 0.24713121354579926, + "learning_rate": 2.1873745419512403e-05, + "loss": 1.1282, + "step": 11376 + }, + { + "epoch": 0.6912327601919922, + "grad_norm": 0.3395889699459076, + "learning_rate": 2.1865831355717075e-05, + "loss": 1.0455, + "step": 11377 + }, + { + "epoch": 0.6912935172246187, + "grad_norm": 0.2502438724040985, + "learning_rate": 2.1857918323185145e-05, + "loss": 1.1153, + "step": 11378 + }, + { + "epoch": 0.6913542742572453, + "grad_norm": 0.1530296355485916, + "learning_rate": 2.1850006322206656e-05, + "loss": 1.0229, + "step": 11379 + }, + { + "epoch": 0.6914150312898718, + "grad_norm": 1.3855068683624268, + "learning_rate": 2.184209535307164e-05, + "loss": 1.3426, + "step": 11380 + }, + { + "epoch": 0.6914757883224983, + "grad_norm": 0.18322540819644928, + "learning_rate": 2.183418541607003e-05, + "loss": 1.0446, + "step": 11381 + }, + { + "epoch": 0.6915365453551249, + "grad_norm": 0.2326284945011139, + "learning_rate": 2.1826276511491815e-05, + "loss": 1.1355, + "step": 11382 + }, + { + "epoch": 0.6915973023877514, + "grad_norm": 0.2206798493862152, + "learning_rate": 2.181836863962688e-05, + "loss": 1.069, + "step": 11383 + }, + { + "epoch": 0.691658059420378, + "grad_norm": 0.7081030607223511, + "learning_rate": 2.181046180076509e-05, + "loss": 1.2429, + "step": 11384 + }, + { + "epoch": 0.6917188164530045, + "grad_norm": 0.2798891067504883, + "learning_rate": 2.1802555995196273e-05, + "loss": 0.9498, + "step": 11385 + }, + { + "epoch": 0.691779573485631, + "grad_norm": 0.22451475262641907, + "learning_rate": 2.179465122321021e-05, + "loss": 1.1414, + "step": 11386 + }, + { + "epoch": 0.6918403305182574, + "grad_norm": 0.2164708375930786, + "learning_rate": 2.1786747485096665e-05, + "loss": 1.0037, + "step": 11387 + }, + { + "epoch": 0.691901087550884, + "grad_norm": 0.2843039333820343, + "learning_rate": 2.1778844781145347e-05, + "loss": 1.0184, + "step": 11388 + }, + { + "epoch": 0.6919618445835105, + "grad_norm": 0.19108203053474426, + "learning_rate": 2.1770943111645918e-05, + "loss": 1.02, + "step": 11389 + }, + { + "epoch": 0.692022601616137, + "grad_norm": 0.1766122728586197, + "learning_rate": 2.1763042476888036e-05, + "loss": 1.0444, + "step": 11390 + }, + { + "epoch": 0.6920833586487636, + "grad_norm": 0.19636991620063782, + "learning_rate": 2.1755142877161282e-05, + "loss": 1.024, + "step": 11391 + }, + { + "epoch": 0.6921441156813901, + "grad_norm": 0.14932505786418915, + "learning_rate": 2.1747244312755218e-05, + "loss": 1.0952, + "step": 11392 + }, + { + "epoch": 0.6922048727140167, + "grad_norm": 0.18734018504619598, + "learning_rate": 2.1739346783959387e-05, + "loss": 1.1221, + "step": 11393 + }, + { + "epoch": 0.6922656297466432, + "grad_norm": 0.20437929034233093, + "learning_rate": 2.1731450291063275e-05, + "loss": 1.1044, + "step": 11394 + }, + { + "epoch": 0.6923263867792697, + "grad_norm": 0.2038939893245697, + "learning_rate": 2.1723554834356307e-05, + "loss": 1.1164, + "step": 11395 + }, + { + "epoch": 0.6923871438118963, + "grad_norm": 0.2606448531150818, + "learning_rate": 2.1715660414127896e-05, + "loss": 1.0993, + "step": 11396 + }, + { + "epoch": 0.6924479008445228, + "grad_norm": 0.17546270787715912, + "learning_rate": 2.1707767030667403e-05, + "loss": 1.1104, + "step": 11397 + }, + { + "epoch": 0.6925086578771493, + "grad_norm": 0.13139212131500244, + "learning_rate": 2.16998746842642e-05, + "loss": 1.0268, + "step": 11398 + }, + { + "epoch": 0.6925694149097759, + "grad_norm": 0.14681825041770935, + "learning_rate": 2.1691983375207557e-05, + "loss": 1.0511, + "step": 11399 + }, + { + "epoch": 0.6926301719424023, + "grad_norm": 0.12257730960845947, + "learning_rate": 2.168409310378674e-05, + "loss": 1.0093, + "step": 11400 + }, + { + "epoch": 0.6926909289750288, + "grad_norm": 0.13903465867042542, + "learning_rate": 2.167620387029098e-05, + "loss": 1.0362, + "step": 11401 + }, + { + "epoch": 0.6927516860076554, + "grad_norm": 0.1359190195798874, + "learning_rate": 2.1668315675009404e-05, + "loss": 1.029, + "step": 11402 + }, + { + "epoch": 0.6928124430402819, + "grad_norm": 0.12885840237140656, + "learning_rate": 2.1660428518231207e-05, + "loss": 1.0043, + "step": 11403 + }, + { + "epoch": 0.6928732000729084, + "grad_norm": 0.1406404674053192, + "learning_rate": 2.165254240024549e-05, + "loss": 1.0413, + "step": 11404 + }, + { + "epoch": 0.692933957105535, + "grad_norm": 0.44082289934158325, + "learning_rate": 2.1644657321341304e-05, + "loss": 1.215, + "step": 11405 + }, + { + "epoch": 0.6929947141381615, + "grad_norm": 0.6462645530700684, + "learning_rate": 2.1636773281807687e-05, + "loss": 1.2577, + "step": 11406 + }, + { + "epoch": 0.693055471170788, + "grad_norm": 1.1320308446884155, + "learning_rate": 2.1628890281933627e-05, + "loss": 1.0412, + "step": 11407 + }, + { + "epoch": 0.6931162282034146, + "grad_norm": 0.16635125875473022, + "learning_rate": 2.162100832200808e-05, + "loss": 1.0922, + "step": 11408 + }, + { + "epoch": 0.6931769852360411, + "grad_norm": 0.2206936925649643, + "learning_rate": 2.1613127402319954e-05, + "loss": 1.056, + "step": 11409 + }, + { + "epoch": 0.6932377422686676, + "grad_norm": 0.25520187616348267, + "learning_rate": 2.1605247523158133e-05, + "loss": 1.1161, + "step": 11410 + }, + { + "epoch": 0.6932984993012942, + "grad_norm": 0.6448429822921753, + "learning_rate": 2.1597368684811454e-05, + "loss": 1.1879, + "step": 11411 + }, + { + "epoch": 0.6933592563339207, + "grad_norm": 0.18067897856235504, + "learning_rate": 2.1589490887568715e-05, + "loss": 1.0699, + "step": 11412 + }, + { + "epoch": 0.6934200133665471, + "grad_norm": 0.5042697191238403, + "learning_rate": 2.1581614131718663e-05, + "loss": 1.0833, + "step": 11413 + }, + { + "epoch": 0.6934807703991737, + "grad_norm": 0.30842918157577515, + "learning_rate": 2.1573738417550072e-05, + "loss": 1.1131, + "step": 11414 + }, + { + "epoch": 0.6935415274318002, + "grad_norm": 0.1606072634458542, + "learning_rate": 2.1565863745351577e-05, + "loss": 1.1041, + "step": 11415 + }, + { + "epoch": 0.6936022844644267, + "grad_norm": 0.4934183955192566, + "learning_rate": 2.1557990115411843e-05, + "loss": 1.1857, + "step": 11416 + }, + { + "epoch": 0.6936630414970533, + "grad_norm": 0.2131747156381607, + "learning_rate": 2.1550117528019474e-05, + "loss": 1.012, + "step": 11417 + }, + { + "epoch": 0.6937237985296798, + "grad_norm": 0.18851107358932495, + "learning_rate": 2.1542245983463035e-05, + "loss": 1.0448, + "step": 11418 + }, + { + "epoch": 0.6937845555623063, + "grad_norm": 0.3615238666534424, + "learning_rate": 2.153437548203108e-05, + "loss": 1.0659, + "step": 11419 + }, + { + "epoch": 0.6938453125949329, + "grad_norm": 0.1720396727323532, + "learning_rate": 2.15265060240121e-05, + "loss": 1.0425, + "step": 11420 + }, + { + "epoch": 0.6939060696275594, + "grad_norm": 0.18849465250968933, + "learning_rate": 2.1518637609694565e-05, + "loss": 1.0363, + "step": 11421 + }, + { + "epoch": 0.6939668266601859, + "grad_norm": 0.562273383140564, + "learning_rate": 2.151077023936685e-05, + "loss": 1.1877, + "step": 11422 + }, + { + "epoch": 0.6940275836928125, + "grad_norm": 0.21640536189079285, + "learning_rate": 2.1502903913317346e-05, + "loss": 1.0472, + "step": 11423 + }, + { + "epoch": 0.694088340725439, + "grad_norm": 0.2354108691215515, + "learning_rate": 2.1495038631834426e-05, + "loss": 1.079, + "step": 11424 + }, + { + "epoch": 0.6941490977580655, + "grad_norm": 0.16371072828769684, + "learning_rate": 2.148717439520637e-05, + "loss": 0.9997, + "step": 11425 + }, + { + "epoch": 0.694209854790692, + "grad_norm": 0.23535358905792236, + "learning_rate": 2.1479311203721452e-05, + "loss": 0.9784, + "step": 11426 + }, + { + "epoch": 0.6942706118233185, + "grad_norm": 0.15247629582881927, + "learning_rate": 2.1471449057667907e-05, + "loss": 1.0305, + "step": 11427 + }, + { + "epoch": 0.694331368855945, + "grad_norm": 0.2790956199169159, + "learning_rate": 2.146358795733388e-05, + "loss": 1.1881, + "step": 11428 + }, + { + "epoch": 0.6943921258885716, + "grad_norm": 0.22870954871177673, + "learning_rate": 2.145572790300757e-05, + "loss": 1.0897, + "step": 11429 + }, + { + "epoch": 0.6944528829211981, + "grad_norm": 0.39218559861183167, + "learning_rate": 2.1447868894977073e-05, + "loss": 1.2788, + "step": 11430 + }, + { + "epoch": 0.6945136399538246, + "grad_norm": 0.19512523710727692, + "learning_rate": 2.1440010933530456e-05, + "loss": 1.0724, + "step": 11431 + }, + { + "epoch": 0.6945743969864512, + "grad_norm": 1.016347050666809, + "learning_rate": 2.1432154018955763e-05, + "loss": 1.0872, + "step": 11432 + }, + { + "epoch": 0.6946351540190777, + "grad_norm": 0.5006350874900818, + "learning_rate": 2.1424298151540983e-05, + "loss": 1.1219, + "step": 11433 + }, + { + "epoch": 0.6946959110517043, + "grad_norm": 0.24114114046096802, + "learning_rate": 2.141644333157408e-05, + "loss": 1.0459, + "step": 11434 + }, + { + "epoch": 0.6947566680843308, + "grad_norm": 0.24277754127979279, + "learning_rate": 2.1408589559342974e-05, + "loss": 1.1165, + "step": 11435 + }, + { + "epoch": 0.6948174251169573, + "grad_norm": 0.25387048721313477, + "learning_rate": 2.1400736835135545e-05, + "loss": 1.0543, + "step": 11436 + }, + { + "epoch": 0.6948781821495839, + "grad_norm": 0.20717036724090576, + "learning_rate": 2.1392885159239628e-05, + "loss": 1.0502, + "step": 11437 + }, + { + "epoch": 0.6949389391822104, + "grad_norm": 0.1640082150697708, + "learning_rate": 2.138503453194304e-05, + "loss": 1.0485, + "step": 11438 + }, + { + "epoch": 0.6949996962148368, + "grad_norm": 0.18056482076644897, + "learning_rate": 2.1377184953533526e-05, + "loss": 1.0591, + "step": 11439 + }, + { + "epoch": 0.6950604532474634, + "grad_norm": 0.16109666228294373, + "learning_rate": 2.136933642429884e-05, + "loss": 1.0164, + "step": 11440 + }, + { + "epoch": 0.6951212102800899, + "grad_norm": 0.15336517989635468, + "learning_rate": 2.1361488944526686e-05, + "loss": 1.0069, + "step": 11441 + }, + { + "epoch": 0.6951819673127164, + "grad_norm": 4.9177703857421875, + "learning_rate": 2.1353642514504675e-05, + "loss": 1.0231, + "step": 11442 + }, + { + "epoch": 0.695242724345343, + "grad_norm": 0.16507165133953094, + "learning_rate": 2.134579713452043e-05, + "loss": 0.9966, + "step": 11443 + }, + { + "epoch": 0.6953034813779695, + "grad_norm": 5.878389835357666, + "learning_rate": 2.1337952804861515e-05, + "loss": 1.2295, + "step": 11444 + }, + { + "epoch": 0.695364238410596, + "grad_norm": 0.19645503163337708, + "learning_rate": 2.1330109525815494e-05, + "loss": 1.1185, + "step": 11445 + }, + { + "epoch": 0.6954249954432226, + "grad_norm": 0.2985987663269043, + "learning_rate": 2.132226729766985e-05, + "loss": 1.155, + "step": 11446 + }, + { + "epoch": 0.6954857524758491, + "grad_norm": 0.18776501715183258, + "learning_rate": 2.1314426120712066e-05, + "loss": 1.1019, + "step": 11447 + }, + { + "epoch": 0.6955465095084756, + "grad_norm": 0.3698202073574066, + "learning_rate": 2.13065859952295e-05, + "loss": 1.0404, + "step": 11448 + }, + { + "epoch": 0.6956072665411022, + "grad_norm": 0.19307704269886017, + "learning_rate": 2.1298746921509582e-05, + "loss": 1.1082, + "step": 11449 + }, + { + "epoch": 0.6956680235737287, + "grad_norm": 0.13047917187213898, + "learning_rate": 2.1290908899839653e-05, + "loss": 0.9696, + "step": 11450 + }, + { + "epoch": 0.6957287806063552, + "grad_norm": 0.19268940389156342, + "learning_rate": 2.1283071930506998e-05, + "loss": 1.0331, + "step": 11451 + }, + { + "epoch": 0.6957895376389817, + "grad_norm": 0.38352927565574646, + "learning_rate": 2.12752360137989e-05, + "loss": 1.141, + "step": 11452 + }, + { + "epoch": 0.6958502946716082, + "grad_norm": 0.22159363329410553, + "learning_rate": 2.126740115000257e-05, + "loss": 1.057, + "step": 11453 + }, + { + "epoch": 0.6959110517042347, + "grad_norm": 0.1856004297733307, + "learning_rate": 2.1259567339405213e-05, + "loss": 1.0248, + "step": 11454 + }, + { + "epoch": 0.6959718087368613, + "grad_norm": 0.15835565328598022, + "learning_rate": 2.1251734582293963e-05, + "loss": 1.0641, + "step": 11455 + }, + { + "epoch": 0.6960325657694878, + "grad_norm": 0.21616755425930023, + "learning_rate": 2.1243902878955946e-05, + "loss": 1.0739, + "step": 11456 + }, + { + "epoch": 0.6960933228021143, + "grad_norm": 0.14011913537979126, + "learning_rate": 2.1236072229678223e-05, + "loss": 1.0519, + "step": 11457 + }, + { + "epoch": 0.6961540798347409, + "grad_norm": 0.3400423526763916, + "learning_rate": 2.1228242634747835e-05, + "loss": 1.0477, + "step": 11458 + }, + { + "epoch": 0.6962148368673674, + "grad_norm": 0.2795734405517578, + "learning_rate": 2.1220414094451752e-05, + "loss": 1.0435, + "step": 11459 + }, + { + "epoch": 0.6962755938999939, + "grad_norm": 0.3440396189689636, + "learning_rate": 2.121258660907698e-05, + "loss": 1.1067, + "step": 11460 + }, + { + "epoch": 0.6963363509326205, + "grad_norm": 0.237813800573349, + "learning_rate": 2.1204760178910422e-05, + "loss": 1.0229, + "step": 11461 + }, + { + "epoch": 0.696397107965247, + "grad_norm": 0.2600637376308441, + "learning_rate": 2.119693480423893e-05, + "loss": 1.074, + "step": 11462 + }, + { + "epoch": 0.6964578649978735, + "grad_norm": 0.2820562422275543, + "learning_rate": 2.1189110485349354e-05, + "loss": 1.1358, + "step": 11463 + }, + { + "epoch": 0.6965186220305001, + "grad_norm": 0.17977745831012726, + "learning_rate": 2.118128722252849e-05, + "loss": 1.0119, + "step": 11464 + }, + { + "epoch": 0.6965793790631265, + "grad_norm": 0.2673608362674713, + "learning_rate": 2.1173465016063134e-05, + "loss": 1.1141, + "step": 11465 + }, + { + "epoch": 0.696640136095753, + "grad_norm": 0.2825843095779419, + "learning_rate": 2.1165643866239982e-05, + "loss": 1.0833, + "step": 11466 + }, + { + "epoch": 0.6967008931283796, + "grad_norm": 0.20057903230190277, + "learning_rate": 2.1157823773345752e-05, + "loss": 0.9864, + "step": 11467 + }, + { + "epoch": 0.6967616501610061, + "grad_norm": 0.20720891654491425, + "learning_rate": 2.1150004737667045e-05, + "loss": 1.0858, + "step": 11468 + }, + { + "epoch": 0.6968224071936326, + "grad_norm": 0.17461656033992767, + "learning_rate": 2.1142186759490473e-05, + "loss": 1.0345, + "step": 11469 + }, + { + "epoch": 0.6968831642262592, + "grad_norm": 0.13701646029949188, + "learning_rate": 2.1134369839102646e-05, + "loss": 1.0402, + "step": 11470 + }, + { + "epoch": 0.6969439212588857, + "grad_norm": 0.7899938225746155, + "learning_rate": 2.112655397679007e-05, + "loss": 1.127, + "step": 11471 + }, + { + "epoch": 0.6970046782915122, + "grad_norm": 0.24129366874694824, + "learning_rate": 2.1118739172839236e-05, + "loss": 1.1692, + "step": 11472 + }, + { + "epoch": 0.6970654353241388, + "grad_norm": 0.20598256587982178, + "learning_rate": 2.1110925427536604e-05, + "loss": 1.134, + "step": 11473 + }, + { + "epoch": 0.6971261923567653, + "grad_norm": 0.14116506278514862, + "learning_rate": 2.1103112741168578e-05, + "loss": 1.0173, + "step": 11474 + }, + { + "epoch": 0.6971869493893919, + "grad_norm": 0.1788707673549652, + "learning_rate": 2.109530111402155e-05, + "loss": 1.0844, + "step": 11475 + }, + { + "epoch": 0.6972477064220184, + "grad_norm": 0.1461603343486786, + "learning_rate": 2.108749054638184e-05, + "loss": 1.0381, + "step": 11476 + }, + { + "epoch": 0.6973084634546449, + "grad_norm": 0.4907006323337555, + "learning_rate": 2.1079681038535753e-05, + "loss": 1.2609, + "step": 11477 + }, + { + "epoch": 0.6973692204872713, + "grad_norm": 0.1815682202577591, + "learning_rate": 2.1071872590769554e-05, + "loss": 1.0539, + "step": 11478 + }, + { + "epoch": 0.6974299775198979, + "grad_norm": 0.24304334819316864, + "learning_rate": 2.1064065203369448e-05, + "loss": 1.1511, + "step": 11479 + }, + { + "epoch": 0.6974907345525244, + "grad_norm": 0.19885475933551788, + "learning_rate": 2.1056258876621616e-05, + "loss": 1.1197, + "step": 11480 + }, + { + "epoch": 0.697551491585151, + "grad_norm": 0.23619434237480164, + "learning_rate": 2.104845361081224e-05, + "loss": 1.0279, + "step": 11481 + }, + { + "epoch": 0.6976122486177775, + "grad_norm": 0.2090185582637787, + "learning_rate": 2.1040649406227376e-05, + "loss": 1.1036, + "step": 11482 + }, + { + "epoch": 0.697673005650404, + "grad_norm": 0.31826069951057434, + "learning_rate": 2.1032846263153105e-05, + "loss": 1.1052, + "step": 11483 + }, + { + "epoch": 0.6977337626830306, + "grad_norm": 0.1840842217206955, + "learning_rate": 2.1025044181875453e-05, + "loss": 1.0686, + "step": 11484 + }, + { + "epoch": 0.6977945197156571, + "grad_norm": 0.1592065393924713, + "learning_rate": 2.101724316268039e-05, + "loss": 1.0198, + "step": 11485 + }, + { + "epoch": 0.6978552767482836, + "grad_norm": 0.20020297169685364, + "learning_rate": 2.1009443205853902e-05, + "loss": 1.1155, + "step": 11486 + }, + { + "epoch": 0.6979160337809102, + "grad_norm": 0.15480680763721466, + "learning_rate": 2.1001644311681873e-05, + "loss": 1.0017, + "step": 11487 + }, + { + "epoch": 0.6979767908135367, + "grad_norm": 0.13564394414424896, + "learning_rate": 2.0993846480450198e-05, + "loss": 1.0452, + "step": 11488 + }, + { + "epoch": 0.6980375478461632, + "grad_norm": 0.18872347474098206, + "learning_rate": 2.0986049712444666e-05, + "loss": 1.0095, + "step": 11489 + }, + { + "epoch": 0.6980983048787898, + "grad_norm": 0.2226688712835312, + "learning_rate": 2.0978254007951077e-05, + "loss": 1.0175, + "step": 11490 + }, + { + "epoch": 0.6981590619114163, + "grad_norm": 0.15198230743408203, + "learning_rate": 2.0970459367255207e-05, + "loss": 1.1107, + "step": 11491 + }, + { + "epoch": 0.6982198189440427, + "grad_norm": 0.22816002368927002, + "learning_rate": 2.0962665790642765e-05, + "loss": 1.0099, + "step": 11492 + }, + { + "epoch": 0.6982805759766693, + "grad_norm": 0.15984629094600677, + "learning_rate": 2.0954873278399422e-05, + "loss": 1.0706, + "step": 11493 + }, + { + "epoch": 0.6983413330092958, + "grad_norm": 0.6445143222808838, + "learning_rate": 2.0947081830810833e-05, + "loss": 1.1796, + "step": 11494 + }, + { + "epoch": 0.6984020900419223, + "grad_norm": 1.1893699169158936, + "learning_rate": 2.0939291448162534e-05, + "loss": 1.1002, + "step": 11495 + }, + { + "epoch": 0.6984628470745489, + "grad_norm": 0.21028423309326172, + "learning_rate": 2.093150213074015e-05, + "loss": 1.0816, + "step": 11496 + }, + { + "epoch": 0.6985236041071754, + "grad_norm": 0.17226791381835938, + "learning_rate": 2.0923713878829165e-05, + "loss": 1.0336, + "step": 11497 + }, + { + "epoch": 0.6985843611398019, + "grad_norm": 0.23297980427742004, + "learning_rate": 2.0915926692715073e-05, + "loss": 1.1557, + "step": 11498 + }, + { + "epoch": 0.6986451181724285, + "grad_norm": 0.1710096001625061, + "learning_rate": 2.090814057268331e-05, + "loss": 1.065, + "step": 11499 + }, + { + "epoch": 0.698705875205055, + "grad_norm": 0.16053786873817444, + "learning_rate": 2.0900355519019277e-05, + "loss": 1.0168, + "step": 11500 + }, + { + "epoch": 0.6987666322376815, + "grad_norm": 0.20687149465084076, + "learning_rate": 2.0892571532008338e-05, + "loss": 1.0098, + "step": 11501 + }, + { + "epoch": 0.6988273892703081, + "grad_norm": 0.21350055932998657, + "learning_rate": 2.0884788611935813e-05, + "loss": 1.1792, + "step": 11502 + }, + { + "epoch": 0.6988881463029346, + "grad_norm": 0.14661294221878052, + "learning_rate": 2.0877006759086993e-05, + "loss": 1.0112, + "step": 11503 + }, + { + "epoch": 0.6989489033355611, + "grad_norm": 0.1975366324186325, + "learning_rate": 2.086922597374712e-05, + "loss": 1.0469, + "step": 11504 + }, + { + "epoch": 0.6990096603681876, + "grad_norm": 0.15287348628044128, + "learning_rate": 2.0861446256201394e-05, + "loss": 1.0124, + "step": 11505 + }, + { + "epoch": 0.6990704174008141, + "grad_norm": 0.3244244158267975, + "learning_rate": 2.085366760673497e-05, + "loss": 1.1327, + "step": 11506 + }, + { + "epoch": 0.6991311744334406, + "grad_norm": 0.16270513832569122, + "learning_rate": 2.084589002563302e-05, + "loss": 1.0321, + "step": 11507 + }, + { + "epoch": 0.6991919314660672, + "grad_norm": 0.13466598093509674, + "learning_rate": 2.083811351318062e-05, + "loss": 1.0005, + "step": 11508 + }, + { + "epoch": 0.6992526884986937, + "grad_norm": 0.16684868931770325, + "learning_rate": 2.0830338069662785e-05, + "loss": 1.1199, + "step": 11509 + }, + { + "epoch": 0.6993134455313202, + "grad_norm": 0.1840648353099823, + "learning_rate": 2.082256369536455e-05, + "loss": 1.0308, + "step": 11510 + }, + { + "epoch": 0.6993742025639468, + "grad_norm": 0.2114047259092331, + "learning_rate": 2.0814790390570865e-05, + "loss": 1.0889, + "step": 11511 + }, + { + "epoch": 0.6994349595965733, + "grad_norm": 0.29512572288513184, + "learning_rate": 2.08070181555667e-05, + "loss": 1.0743, + "step": 11512 + }, + { + "epoch": 0.6994957166291998, + "grad_norm": 0.16722717881202698, + "learning_rate": 2.0799246990636922e-05, + "loss": 1.0672, + "step": 11513 + }, + { + "epoch": 0.6995564736618264, + "grad_norm": 0.14865122735500336, + "learning_rate": 2.0791476896066407e-05, + "loss": 1.0655, + "step": 11514 + }, + { + "epoch": 0.6996172306944529, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.078370787213994e-05, + "loss": 1.035, + "step": 11515 + }, + { + "epoch": 0.6996779877270795, + "grad_norm": 0.18893763422966003, + "learning_rate": 2.0775939919142284e-05, + "loss": 1.1232, + "step": 11516 + }, + { + "epoch": 0.699738744759706, + "grad_norm": 0.2534438669681549, + "learning_rate": 2.0768173037358224e-05, + "loss": 1.0249, + "step": 11517 + }, + { + "epoch": 0.6997995017923324, + "grad_norm": 0.2629739046096802, + "learning_rate": 2.0760407227072424e-05, + "loss": 1.155, + "step": 11518 + }, + { + "epoch": 0.699860258824959, + "grad_norm": 0.12234020233154297, + "learning_rate": 2.0752642488569556e-05, + "loss": 1.0074, + "step": 11519 + }, + { + "epoch": 0.6999210158575855, + "grad_norm": 0.18215914070606232, + "learning_rate": 2.074487882213423e-05, + "loss": 1.0417, + "step": 11520 + }, + { + "epoch": 0.699981772890212, + "grad_norm": 0.1524747610092163, + "learning_rate": 2.0737116228051023e-05, + "loss": 1.0601, + "step": 11521 + }, + { + "epoch": 0.7000425299228386, + "grad_norm": 0.2565464377403259, + "learning_rate": 2.0729354706604476e-05, + "loss": 1.031, + "step": 11522 + }, + { + "epoch": 0.7001032869554651, + "grad_norm": 0.18529529869556427, + "learning_rate": 2.0721594258079093e-05, + "loss": 1.0457, + "step": 11523 + }, + { + "epoch": 0.7001640439880916, + "grad_norm": 0.1715671718120575, + "learning_rate": 2.071383488275933e-05, + "loss": 1.0733, + "step": 11524 + }, + { + "epoch": 0.7002248010207182, + "grad_norm": 0.2327313870191574, + "learning_rate": 2.070607658092961e-05, + "loss": 1.0151, + "step": 11525 + }, + { + "epoch": 0.7002855580533447, + "grad_norm": 0.1822744905948639, + "learning_rate": 2.069831935287432e-05, + "loss": 1.08, + "step": 11526 + }, + { + "epoch": 0.7003463150859712, + "grad_norm": 0.19206973910331726, + "learning_rate": 2.0690563198877776e-05, + "loss": 1.074, + "step": 11527 + }, + { + "epoch": 0.7004070721185978, + "grad_norm": 0.13440202176570892, + "learning_rate": 2.068280811922434e-05, + "loss": 1.0193, + "step": 11528 + }, + { + "epoch": 0.7004678291512243, + "grad_norm": 0.3235919177532196, + "learning_rate": 2.067505411419822e-05, + "loss": 1.0554, + "step": 11529 + }, + { + "epoch": 0.7005285861838508, + "grad_norm": 0.18741081655025482, + "learning_rate": 2.066730118408366e-05, + "loss": 1.0765, + "step": 11530 + }, + { + "epoch": 0.7005893432164773, + "grad_norm": 0.18201807141304016, + "learning_rate": 2.0659549329164845e-05, + "loss": 1.0565, + "step": 11531 + }, + { + "epoch": 0.7006501002491038, + "grad_norm": 0.15730296075344086, + "learning_rate": 2.0651798549725903e-05, + "loss": 1.027, + "step": 11532 + }, + { + "epoch": 0.7007108572817303, + "grad_norm": 0.14078876376152039, + "learning_rate": 2.0644048846050977e-05, + "loss": 1.074, + "step": 11533 + }, + { + "epoch": 0.7007716143143569, + "grad_norm": 0.2253931760787964, + "learning_rate": 2.0636300218424136e-05, + "loss": 1.146, + "step": 11534 + }, + { + "epoch": 0.7008323713469834, + "grad_norm": 0.1523474007844925, + "learning_rate": 2.0628552667129363e-05, + "loss": 1.0598, + "step": 11535 + }, + { + "epoch": 0.7008931283796099, + "grad_norm": 0.21230202913284302, + "learning_rate": 2.062080619245067e-05, + "loss": 1.0682, + "step": 11536 + }, + { + "epoch": 0.7009538854122365, + "grad_norm": 0.33155712485313416, + "learning_rate": 2.061306079467199e-05, + "loss": 1.2552, + "step": 11537 + }, + { + "epoch": 0.701014642444863, + "grad_norm": 0.13536348938941956, + "learning_rate": 2.0605316474077264e-05, + "loss": 1.0903, + "step": 11538 + }, + { + "epoch": 0.7010753994774895, + "grad_norm": 0.1689533293247223, + "learning_rate": 2.0597573230950347e-05, + "loss": 1.0782, + "step": 11539 + }, + { + "epoch": 0.7011361565101161, + "grad_norm": 0.20579645037651062, + "learning_rate": 2.058983106557506e-05, + "loss": 1.0724, + "step": 11540 + }, + { + "epoch": 0.7011969135427426, + "grad_norm": 0.17070773243904114, + "learning_rate": 2.0582089978235204e-05, + "loss": 1.032, + "step": 11541 + }, + { + "epoch": 0.7012576705753691, + "grad_norm": 0.23072439432144165, + "learning_rate": 2.0574349969214524e-05, + "loss": 1.0272, + "step": 11542 + }, + { + "epoch": 0.7013184276079957, + "grad_norm": 0.22161272168159485, + "learning_rate": 2.0566611038796735e-05, + "loss": 1.0641, + "step": 11543 + }, + { + "epoch": 0.7013791846406221, + "grad_norm": 0.1966264247894287, + "learning_rate": 2.055887318726551e-05, + "loss": 1.0354, + "step": 11544 + }, + { + "epoch": 0.7014399416732486, + "grad_norm": 0.1992717981338501, + "learning_rate": 2.0551136414904475e-05, + "loss": 0.9886, + "step": 11545 + }, + { + "epoch": 0.7015006987058752, + "grad_norm": 0.12861256301403046, + "learning_rate": 2.054340072199723e-05, + "loss": 1.0183, + "step": 11546 + }, + { + "epoch": 0.7015614557385017, + "grad_norm": 0.3239456117153168, + "learning_rate": 2.0535666108827323e-05, + "loss": 1.0586, + "step": 11547 + }, + { + "epoch": 0.7016222127711282, + "grad_norm": 0.2051747739315033, + "learning_rate": 2.052793257567827e-05, + "loss": 1.1515, + "step": 11548 + }, + { + "epoch": 0.7016829698037548, + "grad_norm": 0.14052969217300415, + "learning_rate": 2.052020012283355e-05, + "loss": 0.9945, + "step": 11549 + }, + { + "epoch": 0.7017437268363813, + "grad_norm": 0.18644526600837708, + "learning_rate": 2.0512468750576592e-05, + "loss": 1.054, + "step": 11550 + }, + { + "epoch": 0.7018044838690078, + "grad_norm": 0.19870607554912567, + "learning_rate": 2.0504738459190785e-05, + "loss": 1.1244, + "step": 11551 + }, + { + "epoch": 0.7018652409016344, + "grad_norm": 0.20699620246887207, + "learning_rate": 2.04970092489595e-05, + "loss": 0.9984, + "step": 11552 + }, + { + "epoch": 0.7019259979342609, + "grad_norm": 0.33848607540130615, + "learning_rate": 2.048928112016602e-05, + "loss": 1.1457, + "step": 11553 + }, + { + "epoch": 0.7019867549668874, + "grad_norm": 0.1388871669769287, + "learning_rate": 2.0481554073093666e-05, + "loss": 1.0411, + "step": 11554 + }, + { + "epoch": 0.702047511999514, + "grad_norm": 0.3523934781551361, + "learning_rate": 2.0473828108025666e-05, + "loss": 1.1226, + "step": 11555 + }, + { + "epoch": 0.7021082690321405, + "grad_norm": 0.18747149407863617, + "learning_rate": 2.0466103225245192e-05, + "loss": 1.0583, + "step": 11556 + }, + { + "epoch": 0.7021690260647669, + "grad_norm": 0.2482525110244751, + "learning_rate": 2.045837942503539e-05, + "loss": 1.0638, + "step": 11557 + }, + { + "epoch": 0.7022297830973935, + "grad_norm": 1.4105141162872314, + "learning_rate": 2.0450656707679422e-05, + "loss": 1.2988, + "step": 11558 + }, + { + "epoch": 0.70229054013002, + "grad_norm": 0.1828053891658783, + "learning_rate": 2.0442935073460345e-05, + "loss": 1.0224, + "step": 11559 + }, + { + "epoch": 0.7023512971626465, + "grad_norm": 0.27935758233070374, + "learning_rate": 2.0435214522661194e-05, + "loss": 1.0704, + "step": 11560 + }, + { + "epoch": 0.7024120541952731, + "grad_norm": 0.19481541216373444, + "learning_rate": 2.042749505556499e-05, + "loss": 1.055, + "step": 11561 + }, + { + "epoch": 0.7024728112278996, + "grad_norm": 0.14184653759002686, + "learning_rate": 2.0419776672454634e-05, + "loss": 1.0025, + "step": 11562 + }, + { + "epoch": 0.7025335682605262, + "grad_norm": 0.2467447966337204, + "learning_rate": 2.0412059373613096e-05, + "loss": 1.097, + "step": 11563 + }, + { + "epoch": 0.7025943252931527, + "grad_norm": 0.2457072138786316, + "learning_rate": 2.0404343159323243e-05, + "loss": 1.0427, + "step": 11564 + }, + { + "epoch": 0.7026550823257792, + "grad_norm": 0.16685888171195984, + "learning_rate": 2.039662802986791e-05, + "loss": 1.0648, + "step": 11565 + }, + { + "epoch": 0.7027158393584058, + "grad_norm": 0.23430736362934113, + "learning_rate": 2.03889139855299e-05, + "loss": 1.177, + "step": 11566 + }, + { + "epoch": 0.7027765963910323, + "grad_norm": 0.18079927563667297, + "learning_rate": 2.038120102659198e-05, + "loss": 1.1157, + "step": 11567 + }, + { + "epoch": 0.7028373534236588, + "grad_norm": 0.16697730123996735, + "learning_rate": 2.037348915333685e-05, + "loss": 1.1179, + "step": 11568 + }, + { + "epoch": 0.7028981104562854, + "grad_norm": 1.8018940687179565, + "learning_rate": 2.0365778366047206e-05, + "loss": 1.0403, + "step": 11569 + }, + { + "epoch": 0.7029588674889118, + "grad_norm": 0.3022644817829132, + "learning_rate": 2.0358068665005685e-05, + "loss": 1.066, + "step": 11570 + }, + { + "epoch": 0.7030196245215383, + "grad_norm": 0.1648973673582077, + "learning_rate": 2.0350360050494888e-05, + "loss": 1.0182, + "step": 11571 + }, + { + "epoch": 0.7030803815541649, + "grad_norm": 0.18266484141349792, + "learning_rate": 2.0342652522797374e-05, + "loss": 1.0565, + "step": 11572 + }, + { + "epoch": 0.7031411385867914, + "grad_norm": 0.31234896183013916, + "learning_rate": 2.0334946082195643e-05, + "loss": 1.0213, + "step": 11573 + }, + { + "epoch": 0.7032018956194179, + "grad_norm": 0.6090006232261658, + "learning_rate": 2.0327240728972215e-05, + "loss": 1.113, + "step": 11574 + }, + { + "epoch": 0.7032626526520445, + "grad_norm": 0.33582568168640137, + "learning_rate": 2.0319536463409533e-05, + "loss": 1.0339, + "step": 11575 + }, + { + "epoch": 0.703323409684671, + "grad_norm": 0.13417084515094757, + "learning_rate": 2.0311833285789968e-05, + "loss": 0.9871, + "step": 11576 + }, + { + "epoch": 0.7033841667172975, + "grad_norm": 0.9197388291358948, + "learning_rate": 2.030413119639588e-05, + "loss": 1.0314, + "step": 11577 + }, + { + "epoch": 0.7034449237499241, + "grad_norm": 0.2015281617641449, + "learning_rate": 2.0296430195509597e-05, + "loss": 1.0767, + "step": 11578 + }, + { + "epoch": 0.7035056807825506, + "grad_norm": 0.4248289167881012, + "learning_rate": 2.0288730283413416e-05, + "loss": 1.1484, + "step": 11579 + }, + { + "epoch": 0.7035664378151771, + "grad_norm": 0.28621014952659607, + "learning_rate": 2.0281031460389576e-05, + "loss": 0.9784, + "step": 11580 + }, + { + "epoch": 0.7036271948478037, + "grad_norm": 0.14086662232875824, + "learning_rate": 2.0273333726720284e-05, + "loss": 1.0617, + "step": 11581 + }, + { + "epoch": 0.7036879518804302, + "grad_norm": 0.19791261851787567, + "learning_rate": 2.0265637082687677e-05, + "loss": 1.0685, + "step": 11582 + }, + { + "epoch": 0.7037487089130566, + "grad_norm": 0.2872837483882904, + "learning_rate": 2.0257941528573872e-05, + "loss": 1.2324, + "step": 11583 + }, + { + "epoch": 0.7038094659456832, + "grad_norm": 0.2727854251861572, + "learning_rate": 2.0250247064660987e-05, + "loss": 1.1419, + "step": 11584 + }, + { + "epoch": 0.7038702229783097, + "grad_norm": 0.27872875332832336, + "learning_rate": 2.024255369123104e-05, + "loss": 1.135, + "step": 11585 + }, + { + "epoch": 0.7039309800109362, + "grad_norm": 0.20736072957515717, + "learning_rate": 2.0234861408566052e-05, + "loss": 1.1277, + "step": 11586 + }, + { + "epoch": 0.7039917370435628, + "grad_norm": 0.18182599544525146, + "learning_rate": 2.0227170216947967e-05, + "loss": 1.1208, + "step": 11587 + }, + { + "epoch": 0.7040524940761893, + "grad_norm": 0.24280962347984314, + "learning_rate": 2.0219480116658707e-05, + "loss": 1.0713, + "step": 11588 + }, + { + "epoch": 0.7041132511088158, + "grad_norm": 0.13895347714424133, + "learning_rate": 2.0211791107980167e-05, + "loss": 0.9686, + "step": 11589 + }, + { + "epoch": 0.7041740081414424, + "grad_norm": 0.22623944282531738, + "learning_rate": 2.0204103191194178e-05, + "loss": 1.2075, + "step": 11590 + }, + { + "epoch": 0.7042347651740689, + "grad_norm": 0.14495201408863068, + "learning_rate": 2.0196416366582543e-05, + "loss": 1.0394, + "step": 11591 + }, + { + "epoch": 0.7042955222066954, + "grad_norm": 0.19287265837192535, + "learning_rate": 2.0188730634427035e-05, + "loss": 1.0501, + "step": 11592 + }, + { + "epoch": 0.704356279239322, + "grad_norm": 0.17191262543201447, + "learning_rate": 2.018104599500936e-05, + "loss": 1.0337, + "step": 11593 + }, + { + "epoch": 0.7044170362719485, + "grad_norm": 0.1741471290588379, + "learning_rate": 2.0173362448611195e-05, + "loss": 1.0893, + "step": 11594 + }, + { + "epoch": 0.704477793304575, + "grad_norm": 0.3732975125312805, + "learning_rate": 2.016567999551423e-05, + "loss": 1.2385, + "step": 11595 + }, + { + "epoch": 0.7045385503372016, + "grad_norm": 0.14746242761611938, + "learning_rate": 2.0157998636000015e-05, + "loss": 1.0061, + "step": 11596 + }, + { + "epoch": 0.704599307369828, + "grad_norm": 0.18906036019325256, + "learning_rate": 2.0150318370350125e-05, + "loss": 1.1081, + "step": 11597 + }, + { + "epoch": 0.7046600644024545, + "grad_norm": 0.14103391766548157, + "learning_rate": 2.0142639198846086e-05, + "loss": 1.0384, + "step": 11598 + }, + { + "epoch": 0.7047208214350811, + "grad_norm": 0.16637089848518372, + "learning_rate": 2.0134961121769364e-05, + "loss": 1.0588, + "step": 11599 + }, + { + "epoch": 0.7047815784677076, + "grad_norm": 0.1269828975200653, + "learning_rate": 2.0127284139401426e-05, + "loss": 0.9952, + "step": 11600 + }, + { + "epoch": 0.7048423355003341, + "grad_norm": 0.15160314738750458, + "learning_rate": 2.0119608252023688e-05, + "loss": 1.0417, + "step": 11601 + }, + { + "epoch": 0.7049030925329607, + "grad_norm": 0.16122202575206757, + "learning_rate": 2.011193345991746e-05, + "loss": 1.037, + "step": 11602 + }, + { + "epoch": 0.7049638495655872, + "grad_norm": 0.1910794973373413, + "learning_rate": 2.010425976336409e-05, + "loss": 1.0773, + "step": 11603 + }, + { + "epoch": 0.7050246065982138, + "grad_norm": 0.15524131059646606, + "learning_rate": 2.0096587162644842e-05, + "loss": 1.0579, + "step": 11604 + }, + { + "epoch": 0.7050853636308403, + "grad_norm": 0.14267157018184662, + "learning_rate": 2.0088915658040992e-05, + "loss": 1.0312, + "step": 11605 + }, + { + "epoch": 0.7051461206634668, + "grad_norm": 0.16957224905490875, + "learning_rate": 2.0081245249833713e-05, + "loss": 1.0294, + "step": 11606 + }, + { + "epoch": 0.7052068776960934, + "grad_norm": 0.17651763558387756, + "learning_rate": 2.0073575938304178e-05, + "loss": 1.0667, + "step": 11607 + }, + { + "epoch": 0.7052676347287199, + "grad_norm": 0.14146997034549713, + "learning_rate": 2.006590772373352e-05, + "loss": 1.056, + "step": 11608 + }, + { + "epoch": 0.7053283917613464, + "grad_norm": 0.6753054857254028, + "learning_rate": 2.0058240606402755e-05, + "loss": 1.0144, + "step": 11609 + }, + { + "epoch": 0.7053891487939729, + "grad_norm": 0.13592104613780975, + "learning_rate": 2.0050574586592996e-05, + "loss": 1.0133, + "step": 11610 + }, + { + "epoch": 0.7054499058265994, + "grad_norm": 0.25048956274986267, + "learning_rate": 2.0042909664585214e-05, + "loss": 1.1687, + "step": 11611 + }, + { + "epoch": 0.7055106628592259, + "grad_norm": 0.18366578221321106, + "learning_rate": 2.0035245840660367e-05, + "loss": 1.0869, + "step": 11612 + }, + { + "epoch": 0.7055714198918525, + "grad_norm": 0.1804017573595047, + "learning_rate": 2.0027583115099375e-05, + "loss": 1.1131, + "step": 11613 + }, + { + "epoch": 0.705632176924479, + "grad_norm": 0.1677662432193756, + "learning_rate": 2.0019921488183118e-05, + "loss": 1.0499, + "step": 11614 + }, + { + "epoch": 0.7056929339571055, + "grad_norm": 0.16356715559959412, + "learning_rate": 2.001226096019243e-05, + "loss": 1.0601, + "step": 11615 + }, + { + "epoch": 0.7057536909897321, + "grad_norm": 0.1592373549938202, + "learning_rate": 2.000460153140812e-05, + "loss": 1.0627, + "step": 11616 + }, + { + "epoch": 0.7058144480223586, + "grad_norm": 0.12925437092781067, + "learning_rate": 1.9996943202110934e-05, + "loss": 1.0097, + "step": 11617 + }, + { + "epoch": 0.7058752050549851, + "grad_norm": 0.19854521751403809, + "learning_rate": 1.9989285972581595e-05, + "loss": 1.1422, + "step": 11618 + }, + { + "epoch": 0.7059359620876117, + "grad_norm": 0.2509424090385437, + "learning_rate": 1.9981629843100784e-05, + "loss": 1.1168, + "step": 11619 + }, + { + "epoch": 0.7059967191202382, + "grad_norm": 0.16597990691661835, + "learning_rate": 1.9973974813949115e-05, + "loss": 1.0073, + "step": 11620 + }, + { + "epoch": 0.7060574761528647, + "grad_norm": 0.15888646245002747, + "learning_rate": 1.9966320885407213e-05, + "loss": 1.0793, + "step": 11621 + }, + { + "epoch": 0.7061182331854913, + "grad_norm": 0.13420553505420685, + "learning_rate": 1.995866805775565e-05, + "loss": 0.9821, + "step": 11622 + }, + { + "epoch": 0.7061789902181177, + "grad_norm": 0.21412307024002075, + "learning_rate": 1.9951016331274897e-05, + "loss": 1.06, + "step": 11623 + }, + { + "epoch": 0.7062397472507442, + "grad_norm": 0.18910609185695648, + "learning_rate": 1.9943365706245447e-05, + "loss": 1.0242, + "step": 11624 + }, + { + "epoch": 0.7063005042833708, + "grad_norm": 0.16253310441970825, + "learning_rate": 1.9935716182947716e-05, + "loss": 1.1194, + "step": 11625 + }, + { + "epoch": 0.7063612613159973, + "grad_norm": 0.19903503358364105, + "learning_rate": 1.9928067761662146e-05, + "loss": 1.0616, + "step": 11626 + }, + { + "epoch": 0.7064220183486238, + "grad_norm": 0.46064460277557373, + "learning_rate": 1.992042044266906e-05, + "loss": 1.173, + "step": 11627 + }, + { + "epoch": 0.7064827753812504, + "grad_norm": 0.13893476128578186, + "learning_rate": 1.991277422624879e-05, + "loss": 1.0529, + "step": 11628 + }, + { + "epoch": 0.7065435324138769, + "grad_norm": 0.1632026880979538, + "learning_rate": 1.9905129112681585e-05, + "loss": 1.0646, + "step": 11629 + }, + { + "epoch": 0.7066042894465034, + "grad_norm": 0.20791426301002502, + "learning_rate": 1.989748510224767e-05, + "loss": 1.1313, + "step": 11630 + }, + { + "epoch": 0.70666504647913, + "grad_norm": 0.1572713851928711, + "learning_rate": 1.9889842195227276e-05, + "loss": 0.9989, + "step": 11631 + }, + { + "epoch": 0.7067258035117565, + "grad_norm": 0.21711012721061707, + "learning_rate": 1.9882200391900536e-05, + "loss": 1.1575, + "step": 11632 + }, + { + "epoch": 0.706786560544383, + "grad_norm": 0.5225393176078796, + "learning_rate": 1.987455969254756e-05, + "loss": 1.0546, + "step": 11633 + }, + { + "epoch": 0.7068473175770096, + "grad_norm": 0.15523073077201843, + "learning_rate": 1.9866920097448426e-05, + "loss": 1.0006, + "step": 11634 + }, + { + "epoch": 0.7069080746096361, + "grad_norm": 0.21019184589385986, + "learning_rate": 1.9859281606883158e-05, + "loss": 1.1159, + "step": 11635 + }, + { + "epoch": 0.7069688316422625, + "grad_norm": 0.2340015470981598, + "learning_rate": 1.985164422113175e-05, + "loss": 1.0891, + "step": 11636 + }, + { + "epoch": 0.7070295886748891, + "grad_norm": 0.143396258354187, + "learning_rate": 1.9844007940474153e-05, + "loss": 1.0368, + "step": 11637 + }, + { + "epoch": 0.7070903457075156, + "grad_norm": 0.1841418445110321, + "learning_rate": 1.983637276519027e-05, + "loss": 1.1125, + "step": 11638 + }, + { + "epoch": 0.7071511027401421, + "grad_norm": 0.23582510650157928, + "learning_rate": 1.9828738695559972e-05, + "loss": 1.0792, + "step": 11639 + }, + { + "epoch": 0.7072118597727687, + "grad_norm": 0.1381693333387375, + "learning_rate": 1.9821105731863093e-05, + "loss": 1.0311, + "step": 11640 + }, + { + "epoch": 0.7072726168053952, + "grad_norm": 0.2427506446838379, + "learning_rate": 1.9813473874379395e-05, + "loss": 1.0883, + "step": 11641 + }, + { + "epoch": 0.7073333738380218, + "grad_norm": 0.888260006904602, + "learning_rate": 1.9805843123388686e-05, + "loss": 1.0858, + "step": 11642 + }, + { + "epoch": 0.7073941308706483, + "grad_norm": 0.14071987569332123, + "learning_rate": 1.9798213479170613e-05, + "loss": 1.0104, + "step": 11643 + }, + { + "epoch": 0.7074548879032748, + "grad_norm": 0.14098162949085236, + "learning_rate": 1.9790584942004862e-05, + "loss": 1.0373, + "step": 11644 + }, + { + "epoch": 0.7075156449359014, + "grad_norm": 0.13122422993183136, + "learning_rate": 1.9782957512171068e-05, + "loss": 1.0371, + "step": 11645 + }, + { + "epoch": 0.7075764019685279, + "grad_norm": 0.1470564305782318, + "learning_rate": 1.977533118994878e-05, + "loss": 1.0184, + "step": 11646 + }, + { + "epoch": 0.7076371590011544, + "grad_norm": 0.18352192640304565, + "learning_rate": 1.976770597561759e-05, + "loss": 1.0112, + "step": 11647 + }, + { + "epoch": 0.707697916033781, + "grad_norm": 0.12152980268001556, + "learning_rate": 1.9760081869457005e-05, + "loss": 1.0554, + "step": 11648 + }, + { + "epoch": 0.7077586730664074, + "grad_norm": 0.19066199660301208, + "learning_rate": 1.9752458871746437e-05, + "loss": 1.1092, + "step": 11649 + }, + { + "epoch": 0.7078194300990339, + "grad_norm": 0.2313549816608429, + "learning_rate": 1.9744836982765325e-05, + "loss": 1.1101, + "step": 11650 + }, + { + "epoch": 0.7078801871316605, + "grad_norm": 0.14385811984539032, + "learning_rate": 1.9737216202793073e-05, + "loss": 0.9768, + "step": 11651 + }, + { + "epoch": 0.707940944164287, + "grad_norm": 0.13443192839622498, + "learning_rate": 1.972959653210902e-05, + "loss": 1.0053, + "step": 11652 + }, + { + "epoch": 0.7080017011969135, + "grad_norm": 1.8586245775222778, + "learning_rate": 1.972197797099245e-05, + "loss": 1.1716, + "step": 11653 + }, + { + "epoch": 0.7080624582295401, + "grad_norm": 0.14931176602840424, + "learning_rate": 1.971436051972263e-05, + "loss": 1.0111, + "step": 11654 + }, + { + "epoch": 0.7081232152621666, + "grad_norm": 1.2358441352844238, + "learning_rate": 1.9706744178578783e-05, + "loss": 1.0333, + "step": 11655 + }, + { + "epoch": 0.7081839722947931, + "grad_norm": 0.21095271408557892, + "learning_rate": 1.969912894784008e-05, + "loss": 1.0178, + "step": 11656 + }, + { + "epoch": 0.7082447293274197, + "grad_norm": 0.2366623878479004, + "learning_rate": 1.9691514827785663e-05, + "loss": 1.1005, + "step": 11657 + }, + { + "epoch": 0.7083054863600462, + "grad_norm": 0.18599346280097961, + "learning_rate": 1.9683901818694632e-05, + "loss": 1.0193, + "step": 11658 + }, + { + "epoch": 0.7083662433926727, + "grad_norm": 0.17395125329494476, + "learning_rate": 1.9676289920846038e-05, + "loss": 1.0926, + "step": 11659 + }, + { + "epoch": 0.7084270004252993, + "grad_norm": 5.945157051086426, + "learning_rate": 1.9668679134518898e-05, + "loss": 1.0542, + "step": 11660 + }, + { + "epoch": 0.7084877574579258, + "grad_norm": 0.24029621481895447, + "learning_rate": 1.966106945999217e-05, + "loss": 1.1849, + "step": 11661 + }, + { + "epoch": 0.7085485144905522, + "grad_norm": 0.21729281544685364, + "learning_rate": 1.965346089754484e-05, + "loss": 1.071, + "step": 11662 + }, + { + "epoch": 0.7086092715231788, + "grad_norm": 0.20860406756401062, + "learning_rate": 1.9645853447455748e-05, + "loss": 1.1751, + "step": 11663 + }, + { + "epoch": 0.7086700285558053, + "grad_norm": 0.19256676733493805, + "learning_rate": 1.9638247110003765e-05, + "loss": 1.1393, + "step": 11664 + }, + { + "epoch": 0.7087307855884318, + "grad_norm": 0.1320914328098297, + "learning_rate": 1.9630641885467704e-05, + "loss": 0.9998, + "step": 11665 + }, + { + "epoch": 0.7087915426210584, + "grad_norm": 0.17847011983394623, + "learning_rate": 1.9623037774126317e-05, + "loss": 1.0191, + "step": 11666 + }, + { + "epoch": 0.7088522996536849, + "grad_norm": 0.21056631207466125, + "learning_rate": 1.9615434776258372e-05, + "loss": 1.008, + "step": 11667 + }, + { + "epoch": 0.7089130566863114, + "grad_norm": 0.17908596992492676, + "learning_rate": 1.9607832892142554e-05, + "loss": 1.1247, + "step": 11668 + }, + { + "epoch": 0.708973813718938, + "grad_norm": 0.18597280979156494, + "learning_rate": 1.9600232122057478e-05, + "loss": 1.1093, + "step": 11669 + }, + { + "epoch": 0.7090345707515645, + "grad_norm": 0.20764264464378357, + "learning_rate": 1.9592632466281775e-05, + "loss": 1.0129, + "step": 11670 + }, + { + "epoch": 0.709095327784191, + "grad_norm": 0.19079600274562836, + "learning_rate": 1.9585033925093982e-05, + "loss": 1.0978, + "step": 11671 + }, + { + "epoch": 0.7091560848168176, + "grad_norm": 0.34723880887031555, + "learning_rate": 1.9577436498772666e-05, + "loss": 1.0818, + "step": 11672 + }, + { + "epoch": 0.7092168418494441, + "grad_norm": 0.24341143667697906, + "learning_rate": 1.95698401875963e-05, + "loss": 1.0249, + "step": 11673 + }, + { + "epoch": 0.7092775988820706, + "grad_norm": 0.23948602378368378, + "learning_rate": 1.9562244991843327e-05, + "loss": 1.0803, + "step": 11674 + }, + { + "epoch": 0.7093383559146971, + "grad_norm": 0.35194844007492065, + "learning_rate": 1.955465091179216e-05, + "loss": 1.205, + "step": 11675 + }, + { + "epoch": 0.7093991129473236, + "grad_norm": 0.5696149468421936, + "learning_rate": 1.9547057947721116e-05, + "loss": 1.0481, + "step": 11676 + }, + { + "epoch": 0.7094598699799501, + "grad_norm": 0.16976945102214813, + "learning_rate": 1.9539466099908565e-05, + "loss": 1.0362, + "step": 11677 + }, + { + "epoch": 0.7095206270125767, + "grad_norm": 0.1537586748600006, + "learning_rate": 1.9531875368632774e-05, + "loss": 1.0579, + "step": 11678 + }, + { + "epoch": 0.7095813840452032, + "grad_norm": 0.19641618430614471, + "learning_rate": 1.9524285754171978e-05, + "loss": 1.0663, + "step": 11679 + }, + { + "epoch": 0.7096421410778297, + "grad_norm": 0.19217193126678467, + "learning_rate": 1.9516697256804378e-05, + "loss": 1.0334, + "step": 11680 + }, + { + "epoch": 0.7097028981104563, + "grad_norm": 0.1387718766927719, + "learning_rate": 1.9509109876808135e-05, + "loss": 1.0125, + "step": 11681 + }, + { + "epoch": 0.7097636551430828, + "grad_norm": 0.1341647356748581, + "learning_rate": 1.9501523614461365e-05, + "loss": 1.0509, + "step": 11682 + }, + { + "epoch": 0.7098244121757094, + "grad_norm": 0.19160161912441254, + "learning_rate": 1.949393847004214e-05, + "loss": 1.0956, + "step": 11683 + }, + { + "epoch": 0.7098851692083359, + "grad_norm": 0.2551026940345764, + "learning_rate": 1.9486354443828498e-05, + "loss": 1.0854, + "step": 11684 + }, + { + "epoch": 0.7099459262409624, + "grad_norm": 0.14049676060676575, + "learning_rate": 1.947877153609843e-05, + "loss": 0.9942, + "step": 11685 + }, + { + "epoch": 0.710006683273589, + "grad_norm": 0.15755221247673035, + "learning_rate": 1.947118974712989e-05, + "loss": 1.0336, + "step": 11686 + }, + { + "epoch": 0.7100674403062155, + "grad_norm": 0.17419813573360443, + "learning_rate": 1.9463609077200774e-05, + "loss": 1.1191, + "step": 11687 + }, + { + "epoch": 0.7101281973388419, + "grad_norm": 0.278252512216568, + "learning_rate": 1.945602952658901e-05, + "loss": 0.9771, + "step": 11688 + }, + { + "epoch": 0.7101889543714685, + "grad_norm": 0.24212829768657684, + "learning_rate": 1.9448451095572362e-05, + "loss": 1.1141, + "step": 11689 + }, + { + "epoch": 0.710249711404095, + "grad_norm": 0.20440900325775146, + "learning_rate": 1.944087378442865e-05, + "loss": 1.0639, + "step": 11690 + }, + { + "epoch": 0.7103104684367215, + "grad_norm": 0.17397455871105194, + "learning_rate": 1.9433297593435617e-05, + "loss": 1.0332, + "step": 11691 + }, + { + "epoch": 0.710371225469348, + "grad_norm": 0.49350184202194214, + "learning_rate": 1.9425722522870948e-05, + "loss": 1.053, + "step": 11692 + }, + { + "epoch": 0.7104319825019746, + "grad_norm": 0.20767243206501007, + "learning_rate": 1.9418148573012352e-05, + "loss": 1.1248, + "step": 11693 + }, + { + "epoch": 0.7104927395346011, + "grad_norm": 0.19705867767333984, + "learning_rate": 1.9410575744137426e-05, + "loss": 1.0385, + "step": 11694 + }, + { + "epoch": 0.7105534965672277, + "grad_norm": 0.1586126685142517, + "learning_rate": 1.940300403652378e-05, + "loss": 1.0438, + "step": 11695 + }, + { + "epoch": 0.7106142535998542, + "grad_norm": 0.5229145288467407, + "learning_rate": 1.9395433450448914e-05, + "loss": 1.224, + "step": 11696 + }, + { + "epoch": 0.7106750106324807, + "grad_norm": 0.1643553078174591, + "learning_rate": 1.9387863986190337e-05, + "loss": 1.073, + "step": 11697 + }, + { + "epoch": 0.7107357676651073, + "grad_norm": 0.14517661929130554, + "learning_rate": 1.9380295644025538e-05, + "loss": 1.0384, + "step": 11698 + }, + { + "epoch": 0.7107965246977338, + "grad_norm": 0.2176668345928192, + "learning_rate": 1.9372728424231923e-05, + "loss": 1.0376, + "step": 11699 + }, + { + "epoch": 0.7108572817303603, + "grad_norm": 0.20165039598941803, + "learning_rate": 1.9365162327086867e-05, + "loss": 1.0196, + "step": 11700 + }, + { + "epoch": 0.7109180387629869, + "grad_norm": 0.15427738428115845, + "learning_rate": 1.9357597352867706e-05, + "loss": 1.0007, + "step": 11701 + }, + { + "epoch": 0.7109787957956133, + "grad_norm": 0.4921993613243103, + "learning_rate": 1.9350033501851737e-05, + "loss": 1.0594, + "step": 11702 + }, + { + "epoch": 0.7110395528282398, + "grad_norm": 0.1364239752292633, + "learning_rate": 1.9342470774316217e-05, + "loss": 1.0332, + "step": 11703 + }, + { + "epoch": 0.7111003098608664, + "grad_norm": 0.15141117572784424, + "learning_rate": 1.9334909170538356e-05, + "loss": 0.9993, + "step": 11704 + }, + { + "epoch": 0.7111610668934929, + "grad_norm": 0.2824283540248871, + "learning_rate": 1.9327348690795326e-05, + "loss": 1.1113, + "step": 11705 + }, + { + "epoch": 0.7112218239261194, + "grad_norm": 0.16449201107025146, + "learning_rate": 1.9319789335364258e-05, + "loss": 1.0395, + "step": 11706 + }, + { + "epoch": 0.711282580958746, + "grad_norm": 0.30580344796180725, + "learning_rate": 1.9312231104522242e-05, + "loss": 1.0159, + "step": 11707 + }, + { + "epoch": 0.7113433379913725, + "grad_norm": 0.14460936188697815, + "learning_rate": 1.9304673998546312e-05, + "loss": 1.0078, + "step": 11708 + }, + { + "epoch": 0.711404095023999, + "grad_norm": 0.3564567565917969, + "learning_rate": 1.9297118017713524e-05, + "loss": 1.2239, + "step": 11709 + }, + { + "epoch": 0.7114648520566256, + "grad_norm": 0.14387594163417816, + "learning_rate": 1.9289563162300787e-05, + "loss": 1.0771, + "step": 11710 + }, + { + "epoch": 0.7115256090892521, + "grad_norm": 0.5036328434944153, + "learning_rate": 1.9282009432585056e-05, + "loss": 1.162, + "step": 11711 + }, + { + "epoch": 0.7115863661218786, + "grad_norm": 0.1609337478876114, + "learning_rate": 1.9274456828843206e-05, + "loss": 1.0489, + "step": 11712 + }, + { + "epoch": 0.7116471231545052, + "grad_norm": 0.2098923623561859, + "learning_rate": 1.926690535135206e-05, + "loss": 0.974, + "step": 11713 + }, + { + "epoch": 0.7117078801871317, + "grad_norm": 0.17633172869682312, + "learning_rate": 1.9259355000388457e-05, + "loss": 1.1127, + "step": 11714 + }, + { + "epoch": 0.7117686372197581, + "grad_norm": 0.2741782069206238, + "learning_rate": 1.9251805776229153e-05, + "loss": 1.0958, + "step": 11715 + }, + { + "epoch": 0.7118293942523847, + "grad_norm": 0.19436228275299072, + "learning_rate": 1.924425767915084e-05, + "loss": 0.9963, + "step": 11716 + }, + { + "epoch": 0.7118901512850112, + "grad_norm": 0.20139910280704498, + "learning_rate": 1.92367107094302e-05, + "loss": 1.0399, + "step": 11717 + }, + { + "epoch": 0.7119509083176377, + "grad_norm": 0.24296055734157562, + "learning_rate": 1.9229164867343863e-05, + "loss": 1.0312, + "step": 11718 + }, + { + "epoch": 0.7120116653502643, + "grad_norm": 0.15330387651920319, + "learning_rate": 1.922162015316845e-05, + "loss": 1.0703, + "step": 11719 + }, + { + "epoch": 0.7120724223828908, + "grad_norm": 0.1494821161031723, + "learning_rate": 1.9214076567180506e-05, + "loss": 0.9594, + "step": 11720 + }, + { + "epoch": 0.7121331794155173, + "grad_norm": 0.16394075751304626, + "learning_rate": 1.9206534109656532e-05, + "loss": 0.9887, + "step": 11721 + }, + { + "epoch": 0.7121939364481439, + "grad_norm": 0.21139316260814667, + "learning_rate": 1.9198992780873015e-05, + "loss": 1.0868, + "step": 11722 + }, + { + "epoch": 0.7122546934807704, + "grad_norm": 0.19013117253780365, + "learning_rate": 1.9191452581106346e-05, + "loss": 1.0259, + "step": 11723 + }, + { + "epoch": 0.712315450513397, + "grad_norm": 0.1768871545791626, + "learning_rate": 1.9183913510632945e-05, + "loss": 1.0805, + "step": 11724 + }, + { + "epoch": 0.7123762075460235, + "grad_norm": 0.22870860993862152, + "learning_rate": 1.917637556972916e-05, + "loss": 1.01, + "step": 11725 + }, + { + "epoch": 0.71243696457865, + "grad_norm": 6.653565406799316, + "learning_rate": 1.9168838758671282e-05, + "loss": 1.095, + "step": 11726 + }, + { + "epoch": 0.7124977216112766, + "grad_norm": 0.3509865701198578, + "learning_rate": 1.9161303077735575e-05, + "loss": 1.1927, + "step": 11727 + }, + { + "epoch": 0.712558478643903, + "grad_norm": 0.17832794785499573, + "learning_rate": 1.9153768527198272e-05, + "loss": 1.0448, + "step": 11728 + }, + { + "epoch": 0.7126192356765295, + "grad_norm": 0.15734076499938965, + "learning_rate": 1.9146235107335543e-05, + "loss": 1.029, + "step": 11729 + }, + { + "epoch": 0.712679992709156, + "grad_norm": 0.3242066204547882, + "learning_rate": 1.913870281842353e-05, + "loss": 1.1418, + "step": 11730 + }, + { + "epoch": 0.7127407497417826, + "grad_norm": 0.2586510479450226, + "learning_rate": 1.9131171660738335e-05, + "loss": 1.0509, + "step": 11731 + }, + { + "epoch": 0.7128015067744091, + "grad_norm": 0.2569168508052826, + "learning_rate": 1.9123641634556e-05, + "loss": 1.1629, + "step": 11732 + }, + { + "epoch": 0.7128622638070357, + "grad_norm": 0.3447055518627167, + "learning_rate": 1.911611274015256e-05, + "loss": 1.1809, + "step": 11733 + }, + { + "epoch": 0.7129230208396622, + "grad_norm": 0.17897450923919678, + "learning_rate": 1.910858497780395e-05, + "loss": 1.09, + "step": 11734 + }, + { + "epoch": 0.7129837778722887, + "grad_norm": 0.41136646270751953, + "learning_rate": 1.9101058347786173e-05, + "loss": 1.0028, + "step": 11735 + }, + { + "epoch": 0.7130445349049153, + "grad_norm": 0.19103768467903137, + "learning_rate": 1.909353285037505e-05, + "loss": 1.061, + "step": 11736 + }, + { + "epoch": 0.7131052919375418, + "grad_norm": 0.2931089997291565, + "learning_rate": 1.908600848584646e-05, + "loss": 1.0472, + "step": 11737 + }, + { + "epoch": 0.7131660489701683, + "grad_norm": 0.4355412721633911, + "learning_rate": 1.9078485254476198e-05, + "loss": 1.0361, + "step": 11738 + }, + { + "epoch": 0.7132268060027949, + "grad_norm": 0.31621620059013367, + "learning_rate": 1.9070963156540023e-05, + "loss": 1.062, + "step": 11739 + }, + { + "epoch": 0.7132875630354214, + "grad_norm": 0.4727327227592468, + "learning_rate": 1.9063442192313687e-05, + "loss": 1.0561, + "step": 11740 + }, + { + "epoch": 0.7133483200680478, + "grad_norm": 0.22823044657707214, + "learning_rate": 1.9055922362072858e-05, + "loss": 1.1216, + "step": 11741 + }, + { + "epoch": 0.7134090771006744, + "grad_norm": 0.45678579807281494, + "learning_rate": 1.9048403666093195e-05, + "loss": 1.126, + "step": 11742 + }, + { + "epoch": 0.7134698341333009, + "grad_norm": 0.9400303959846497, + "learning_rate": 1.904088610465024e-05, + "loss": 1.0602, + "step": 11743 + }, + { + "epoch": 0.7135305911659274, + "grad_norm": 0.21936200559139252, + "learning_rate": 1.9033369678019613e-05, + "loss": 1.0769, + "step": 11744 + }, + { + "epoch": 0.713591348198554, + "grad_norm": 0.2604951560497284, + "learning_rate": 1.9025854386476798e-05, + "loss": 1.1154, + "step": 11745 + }, + { + "epoch": 0.7136521052311805, + "grad_norm": 0.2755037844181061, + "learning_rate": 1.9018340230297288e-05, + "loss": 1.1143, + "step": 11746 + }, + { + "epoch": 0.713712862263807, + "grad_norm": 0.21337789297103882, + "learning_rate": 1.90108272097565e-05, + "loss": 1.0665, + "step": 11747 + }, + { + "epoch": 0.7137736192964336, + "grad_norm": 0.2862176299095154, + "learning_rate": 1.900331532512983e-05, + "loss": 1.1884, + "step": 11748 + }, + { + "epoch": 0.7138343763290601, + "grad_norm": 0.15735752880573273, + "learning_rate": 1.8995804576692627e-05, + "loss": 1.0585, + "step": 11749 + }, + { + "epoch": 0.7138951333616866, + "grad_norm": 0.22634190320968628, + "learning_rate": 1.89882949647202e-05, + "loss": 1.0264, + "step": 11750 + }, + { + "epoch": 0.7139558903943132, + "grad_norm": 0.18327176570892334, + "learning_rate": 1.898078648948782e-05, + "loss": 1.0409, + "step": 11751 + }, + { + "epoch": 0.7140166474269397, + "grad_norm": 0.1615905463695526, + "learning_rate": 1.8973279151270702e-05, + "loss": 0.9897, + "step": 11752 + }, + { + "epoch": 0.7140774044595662, + "grad_norm": 0.17042972147464752, + "learning_rate": 1.8965772950344042e-05, + "loss": 1.0432, + "step": 11753 + }, + { + "epoch": 0.7141381614921927, + "grad_norm": 0.15339641273021698, + "learning_rate": 1.8958267886982945e-05, + "loss": 1.0214, + "step": 11754 + }, + { + "epoch": 0.7141989185248192, + "grad_norm": 0.29294249415397644, + "learning_rate": 1.8950763961462582e-05, + "loss": 1.2104, + "step": 11755 + }, + { + "epoch": 0.7142596755574457, + "grad_norm": 2.8276686668395996, + "learning_rate": 1.8943261174057948e-05, + "loss": 1.0371, + "step": 11756 + }, + { + "epoch": 0.7143204325900723, + "grad_norm": 0.1949310153722763, + "learning_rate": 1.893575952504408e-05, + "loss": 1.0399, + "step": 11757 + }, + { + "epoch": 0.7143811896226988, + "grad_norm": 0.24249492585659027, + "learning_rate": 1.8928259014695948e-05, + "loss": 1.0053, + "step": 11758 + }, + { + "epoch": 0.7144419466553253, + "grad_norm": 0.17023661732673645, + "learning_rate": 1.892075964328847e-05, + "loss": 1.1327, + "step": 11759 + }, + { + "epoch": 0.7145027036879519, + "grad_norm": 0.222820445895195, + "learning_rate": 1.8913261411096572e-05, + "loss": 1.0563, + "step": 11760 + }, + { + "epoch": 0.7145634607205784, + "grad_norm": 0.16460531949996948, + "learning_rate": 1.8905764318395093e-05, + "loss": 1.012, + "step": 11761 + }, + { + "epoch": 0.714624217753205, + "grad_norm": 0.22096559405326843, + "learning_rate": 1.8898268365458844e-05, + "loss": 1.3159, + "step": 11762 + }, + { + "epoch": 0.7146849747858315, + "grad_norm": 0.12816064059734344, + "learning_rate": 1.889077355256256e-05, + "loss": 1.0263, + "step": 11763 + }, + { + "epoch": 0.714745731818458, + "grad_norm": 0.13777105510234833, + "learning_rate": 1.8883279879980976e-05, + "loss": 1.0277, + "step": 11764 + }, + { + "epoch": 0.7148064888510846, + "grad_norm": 0.21803885698318481, + "learning_rate": 1.8875787347988793e-05, + "loss": 1.0466, + "step": 11765 + }, + { + "epoch": 0.7148672458837111, + "grad_norm": 0.31762614846229553, + "learning_rate": 1.8868295956860642e-05, + "loss": 1.1985, + "step": 11766 + }, + { + "epoch": 0.7149280029163375, + "grad_norm": 0.18634694814682007, + "learning_rate": 1.886080570687112e-05, + "loss": 1.0743, + "step": 11767 + }, + { + "epoch": 0.714988759948964, + "grad_norm": 0.29539746046066284, + "learning_rate": 1.885331659829479e-05, + "loss": 1.1563, + "step": 11768 + }, + { + "epoch": 0.7150495169815906, + "grad_norm": 0.1802942156791687, + "learning_rate": 1.8845828631406153e-05, + "loss": 1.061, + "step": 11769 + }, + { + "epoch": 0.7151102740142171, + "grad_norm": 0.17608396708965302, + "learning_rate": 1.883834180647969e-05, + "loss": 1.089, + "step": 11770 + }, + { + "epoch": 0.7151710310468437, + "grad_norm": 0.1908123940229416, + "learning_rate": 1.8830856123789836e-05, + "loss": 1.0438, + "step": 11771 + }, + { + "epoch": 0.7152317880794702, + "grad_norm": 0.18337857723236084, + "learning_rate": 1.8823371583610977e-05, + "loss": 1.0756, + "step": 11772 + }, + { + "epoch": 0.7152925451120967, + "grad_norm": 0.14490655064582825, + "learning_rate": 1.8815888186217457e-05, + "loss": 0.9957, + "step": 11773 + }, + { + "epoch": 0.7153533021447233, + "grad_norm": 0.1863926202058792, + "learning_rate": 1.8808405931883583e-05, + "loss": 1.0232, + "step": 11774 + }, + { + "epoch": 0.7154140591773498, + "grad_norm": 0.19431519508361816, + "learning_rate": 1.8800924820883597e-05, + "loss": 1.1579, + "step": 11775 + }, + { + "epoch": 0.7154748162099763, + "grad_norm": 0.13856180012226105, + "learning_rate": 1.8793444853491783e-05, + "loss": 1.0823, + "step": 11776 + }, + { + "epoch": 0.7155355732426029, + "grad_norm": 0.20456789433956146, + "learning_rate": 1.878596602998226e-05, + "loss": 1.0933, + "step": 11777 + }, + { + "epoch": 0.7155963302752294, + "grad_norm": 0.22810259461402893, + "learning_rate": 1.8778488350629187e-05, + "loss": 1.049, + "step": 11778 + }, + { + "epoch": 0.7156570873078559, + "grad_norm": 0.3905121088027954, + "learning_rate": 1.8771011815706657e-05, + "loss": 1.0871, + "step": 11779 + }, + { + "epoch": 0.7157178443404824, + "grad_norm": 0.22155234217643738, + "learning_rate": 1.8763536425488704e-05, + "loss": 1.0919, + "step": 11780 + }, + { + "epoch": 0.7157786013731089, + "grad_norm": 0.21493957936763763, + "learning_rate": 1.8756062180249385e-05, + "loss": 1.1387, + "step": 11781 + }, + { + "epoch": 0.7158393584057354, + "grad_norm": 0.34415602684020996, + "learning_rate": 1.8748589080262662e-05, + "loss": 1.0701, + "step": 11782 + }, + { + "epoch": 0.715900115438362, + "grad_norm": 0.22448748350143433, + "learning_rate": 1.8741117125802427e-05, + "loss": 1.1398, + "step": 11783 + }, + { + "epoch": 0.7159608724709885, + "grad_norm": 0.19061557948589325, + "learning_rate": 1.873364631714259e-05, + "loss": 1.0931, + "step": 11784 + }, + { + "epoch": 0.716021629503615, + "grad_norm": 0.19973041117191315, + "learning_rate": 1.872617665455697e-05, + "loss": 1.0462, + "step": 11785 + }, + { + "epoch": 0.7160823865362416, + "grad_norm": 0.25780004262924194, + "learning_rate": 1.871870813831941e-05, + "loss": 1.0319, + "step": 11786 + }, + { + "epoch": 0.7161431435688681, + "grad_norm": 0.2405567318201065, + "learning_rate": 1.8711240768703647e-05, + "loss": 1.1448, + "step": 11787 + }, + { + "epoch": 0.7162039006014946, + "grad_norm": 0.7255765795707703, + "learning_rate": 1.8703774545983405e-05, + "loss": 1.0277, + "step": 11788 + }, + { + "epoch": 0.7162646576341212, + "grad_norm": 0.16757984459400177, + "learning_rate": 1.8696309470432372e-05, + "loss": 1.019, + "step": 11789 + }, + { + "epoch": 0.7163254146667477, + "grad_norm": 0.22032125294208527, + "learning_rate": 1.868884554232413e-05, + "loss": 1.1366, + "step": 11790 + }, + { + "epoch": 0.7163861716993742, + "grad_norm": 0.15231327712535858, + "learning_rate": 1.8681382761932324e-05, + "loss": 1.0492, + "step": 11791 + }, + { + "epoch": 0.7164469287320008, + "grad_norm": 0.20620961487293243, + "learning_rate": 1.8673921129530492e-05, + "loss": 1.0697, + "step": 11792 + }, + { + "epoch": 0.7165076857646272, + "grad_norm": 0.19030196964740753, + "learning_rate": 1.866646064539213e-05, + "loss": 1.0771, + "step": 11793 + }, + { + "epoch": 0.7165684427972537, + "grad_norm": 0.20335501432418823, + "learning_rate": 1.8659001309790715e-05, + "loss": 1.1924, + "step": 11794 + }, + { + "epoch": 0.7166291998298803, + "grad_norm": 0.17489959299564362, + "learning_rate": 1.865154312299966e-05, + "loss": 1.014, + "step": 11795 + }, + { + "epoch": 0.7166899568625068, + "grad_norm": 0.21956895291805267, + "learning_rate": 1.8644086085292357e-05, + "loss": 1.1071, + "step": 11796 + }, + { + "epoch": 0.7167507138951333, + "grad_norm": 0.18535366654396057, + "learning_rate": 1.8636630196942136e-05, + "loss": 1.0875, + "step": 11797 + }, + { + "epoch": 0.7168114709277599, + "grad_norm": 0.2833510637283325, + "learning_rate": 1.8629175458222293e-05, + "loss": 1.0917, + "step": 11798 + }, + { + "epoch": 0.7168722279603864, + "grad_norm": 0.31574761867523193, + "learning_rate": 1.8621721869406095e-05, + "loss": 1.1497, + "step": 11799 + }, + { + "epoch": 0.716932984993013, + "grad_norm": 0.13919039070606232, + "learning_rate": 1.861426943076674e-05, + "loss": 1.0434, + "step": 11800 + }, + { + "epoch": 0.7169937420256395, + "grad_norm": 0.26820647716522217, + "learning_rate": 1.8606818142577392e-05, + "loss": 1.1887, + "step": 11801 + }, + { + "epoch": 0.717054499058266, + "grad_norm": 0.29692527651786804, + "learning_rate": 1.8599368005111233e-05, + "loss": 1.2781, + "step": 11802 + }, + { + "epoch": 0.7171152560908925, + "grad_norm": 1.0516659021377563, + "learning_rate": 1.8591919018641284e-05, + "loss": 1.2364, + "step": 11803 + }, + { + "epoch": 0.7171760131235191, + "grad_norm": 0.31023380160331726, + "learning_rate": 1.8584471183440623e-05, + "loss": 1.1054, + "step": 11804 + }, + { + "epoch": 0.7172367701561456, + "grad_norm": 0.2953868806362152, + "learning_rate": 1.8577024499782237e-05, + "loss": 1.04, + "step": 11805 + }, + { + "epoch": 0.7172975271887722, + "grad_norm": 0.5045631527900696, + "learning_rate": 1.8569578967939073e-05, + "loss": 1.0502, + "step": 11806 + }, + { + "epoch": 0.7173582842213986, + "grad_norm": 0.2013852745294571, + "learning_rate": 1.856213458818409e-05, + "loss": 1.1774, + "step": 11807 + }, + { + "epoch": 0.7174190412540251, + "grad_norm": 0.17298844456672668, + "learning_rate": 1.855469136079014e-05, + "loss": 1.1149, + "step": 11808 + }, + { + "epoch": 0.7174797982866516, + "grad_norm": 0.20909567177295685, + "learning_rate": 1.8547249286030065e-05, + "loss": 1.0664, + "step": 11809 + }, + { + "epoch": 0.7175405553192782, + "grad_norm": 0.12829434871673584, + "learning_rate": 1.8539808364176635e-05, + "loss": 1.0788, + "step": 11810 + }, + { + "epoch": 0.7176013123519047, + "grad_norm": 0.16210830211639404, + "learning_rate": 1.85323685955026e-05, + "loss": 1.0178, + "step": 11811 + }, + { + "epoch": 0.7176620693845313, + "grad_norm": 0.16762219369411469, + "learning_rate": 1.8524929980280685e-05, + "loss": 1.0274, + "step": 11812 + }, + { + "epoch": 0.7177228264171578, + "grad_norm": 0.20096519589424133, + "learning_rate": 1.851749251878355e-05, + "loss": 1.1316, + "step": 11813 + }, + { + "epoch": 0.7177835834497843, + "grad_norm": 0.27306151390075684, + "learning_rate": 1.851005621128381e-05, + "loss": 1.114, + "step": 11814 + }, + { + "epoch": 0.7178443404824109, + "grad_norm": 0.25230035185813904, + "learning_rate": 1.8502621058054048e-05, + "loss": 1.0558, + "step": 11815 + }, + { + "epoch": 0.7179050975150374, + "grad_norm": 0.13505718111991882, + "learning_rate": 1.8495187059366804e-05, + "loss": 1.0261, + "step": 11816 + }, + { + "epoch": 0.7179658545476639, + "grad_norm": 0.14621128141880035, + "learning_rate": 1.8487754215494564e-05, + "loss": 1.0273, + "step": 11817 + }, + { + "epoch": 0.7180266115802905, + "grad_norm": 0.17648938298225403, + "learning_rate": 1.8480322526709783e-05, + "loss": 0.9785, + "step": 11818 + }, + { + "epoch": 0.718087368612917, + "grad_norm": 0.14451909065246582, + "learning_rate": 1.8472891993284867e-05, + "loss": 0.9983, + "step": 11819 + }, + { + "epoch": 0.7181481256455434, + "grad_norm": 0.3215439021587372, + "learning_rate": 1.8465462615492196e-05, + "loss": 1.191, + "step": 11820 + }, + { + "epoch": 0.71820888267817, + "grad_norm": 0.15757663547992706, + "learning_rate": 1.845803439360409e-05, + "loss": 1.0338, + "step": 11821 + }, + { + "epoch": 0.7182696397107965, + "grad_norm": 0.20301571488380432, + "learning_rate": 1.845060732789283e-05, + "loss": 1.0343, + "step": 11822 + }, + { + "epoch": 0.718330396743423, + "grad_norm": 0.13986554741859436, + "learning_rate": 1.844318141863065e-05, + "loss": 1.0079, + "step": 11823 + }, + { + "epoch": 0.7183911537760496, + "grad_norm": 0.21361786127090454, + "learning_rate": 1.843575666608976e-05, + "loss": 1.0589, + "step": 11824 + }, + { + "epoch": 0.7184519108086761, + "grad_norm": 0.23442141711711884, + "learning_rate": 1.8428333070542308e-05, + "loss": 0.9961, + "step": 11825 + }, + { + "epoch": 0.7185126678413026, + "grad_norm": 0.13107125461101532, + "learning_rate": 1.8420910632260414e-05, + "loss": 0.9765, + "step": 11826 + }, + { + "epoch": 0.7185734248739292, + "grad_norm": 0.19029468297958374, + "learning_rate": 1.8413489351516128e-05, + "loss": 1.0595, + "step": 11827 + }, + { + "epoch": 0.7186341819065557, + "grad_norm": 0.1447376012802124, + "learning_rate": 1.840606922858151e-05, + "loss": 1.0331, + "step": 11828 + }, + { + "epoch": 0.7186949389391822, + "grad_norm": 0.18577484786510468, + "learning_rate": 1.8398650263728547e-05, + "loss": 1.0898, + "step": 11829 + }, + { + "epoch": 0.7187556959718088, + "grad_norm": 0.15740127861499786, + "learning_rate": 1.8391232457229153e-05, + "loss": 1.0561, + "step": 11830 + }, + { + "epoch": 0.7188164530044353, + "grad_norm": 0.12481498718261719, + "learning_rate": 1.8383815809355237e-05, + "loss": 0.9825, + "step": 11831 + }, + { + "epoch": 0.7188772100370618, + "grad_norm": 0.18631814420223236, + "learning_rate": 1.8376400320378654e-05, + "loss": 1.1024, + "step": 11832 + }, + { + "epoch": 0.7189379670696883, + "grad_norm": 0.20976340770721436, + "learning_rate": 1.836898599057124e-05, + "loss": 1.1144, + "step": 11833 + }, + { + "epoch": 0.7189987241023148, + "grad_norm": 0.16727115213871002, + "learning_rate": 1.8361572820204768e-05, + "loss": 0.9636, + "step": 11834 + }, + { + "epoch": 0.7190594811349413, + "grad_norm": 0.22351495921611786, + "learning_rate": 1.8354160809550952e-05, + "loss": 1.0531, + "step": 11835 + }, + { + "epoch": 0.7191202381675679, + "grad_norm": 5.896280765533447, + "learning_rate": 1.8346749958881492e-05, + "loss": 1.0582, + "step": 11836 + }, + { + "epoch": 0.7191809952001944, + "grad_norm": 0.44322556257247925, + "learning_rate": 1.833934026846803e-05, + "loss": 1.0771, + "step": 11837 + }, + { + "epoch": 0.7192417522328209, + "grad_norm": 0.9600245952606201, + "learning_rate": 1.833193173858217e-05, + "loss": 1.0224, + "step": 11838 + }, + { + "epoch": 0.7193025092654475, + "grad_norm": 0.17735330760478973, + "learning_rate": 1.8324524369495476e-05, + "loss": 1.084, + "step": 11839 + }, + { + "epoch": 0.719363266298074, + "grad_norm": 0.17591097950935364, + "learning_rate": 1.8317118161479463e-05, + "loss": 1.0838, + "step": 11840 + }, + { + "epoch": 0.7194240233307005, + "grad_norm": 0.2737712562084198, + "learning_rate": 1.8309713114805603e-05, + "loss": 1.1082, + "step": 11841 + }, + { + "epoch": 0.7194847803633271, + "grad_norm": 0.1341107189655304, + "learning_rate": 1.830230922974534e-05, + "loss": 0.9896, + "step": 11842 + }, + { + "epoch": 0.7195455373959536, + "grad_norm": 0.23046131432056427, + "learning_rate": 1.829490650657006e-05, + "loss": 1.2145, + "step": 11843 + }, + { + "epoch": 0.7196062944285802, + "grad_norm": 0.192766934633255, + "learning_rate": 1.8287504945551104e-05, + "loss": 1.0303, + "step": 11844 + }, + { + "epoch": 0.7196670514612067, + "grad_norm": 0.2848644554615021, + "learning_rate": 1.828010454695979e-05, + "loss": 1.0097, + "step": 11845 + }, + { + "epoch": 0.7197278084938331, + "grad_norm": 0.15069590508937836, + "learning_rate": 1.8272705311067374e-05, + "loss": 1.0247, + "step": 11846 + }, + { + "epoch": 0.7197885655264596, + "grad_norm": 0.1978844702243805, + "learning_rate": 1.8265307238145074e-05, + "loss": 1.0733, + "step": 11847 + }, + { + "epoch": 0.7198493225590862, + "grad_norm": 0.32169559597969055, + "learning_rate": 1.8257910328464057e-05, + "loss": 1.1357, + "step": 11848 + }, + { + "epoch": 0.7199100795917127, + "grad_norm": 0.15754975378513336, + "learning_rate": 1.825051458229551e-05, + "loss": 1.0107, + "step": 11849 + }, + { + "epoch": 0.7199708366243392, + "grad_norm": 0.20106159150600433, + "learning_rate": 1.824311999991046e-05, + "loss": 1.0666, + "step": 11850 + }, + { + "epoch": 0.7200315936569658, + "grad_norm": 0.13575264811515808, + "learning_rate": 1.8235726581579998e-05, + "loss": 0.9896, + "step": 11851 + }, + { + "epoch": 0.7200923506895923, + "grad_norm": 0.13829223811626434, + "learning_rate": 1.8228334327575097e-05, + "loss": 0.9818, + "step": 11852 + }, + { + "epoch": 0.7201531077222189, + "grad_norm": 0.2209959179162979, + "learning_rate": 1.8220943238166754e-05, + "loss": 1.0888, + "step": 11853 + }, + { + "epoch": 0.7202138647548454, + "grad_norm": 0.14765897393226624, + "learning_rate": 1.8213553313625888e-05, + "loss": 1.0681, + "step": 11854 + }, + { + "epoch": 0.7202746217874719, + "grad_norm": 0.20563434064388275, + "learning_rate": 1.8206164554223358e-05, + "loss": 1.1194, + "step": 11855 + }, + { + "epoch": 0.7203353788200985, + "grad_norm": 0.13147789239883423, + "learning_rate": 1.819877696023004e-05, + "loss": 1.0303, + "step": 11856 + }, + { + "epoch": 0.720396135852725, + "grad_norm": 0.1730482280254364, + "learning_rate": 1.8191390531916663e-05, + "loss": 1.0658, + "step": 11857 + }, + { + "epoch": 0.7204568928853515, + "grad_norm": 0.2509239912033081, + "learning_rate": 1.8184005269554027e-05, + "loss": 1.1152, + "step": 11858 + }, + { + "epoch": 0.720517649917978, + "grad_norm": 0.23618800938129425, + "learning_rate": 1.8176621173412834e-05, + "loss": 1.1606, + "step": 11859 + }, + { + "epoch": 0.7205784069506045, + "grad_norm": 0.8432917594909668, + "learning_rate": 1.8169238243763735e-05, + "loss": 1.0307, + "step": 11860 + }, + { + "epoch": 0.720639163983231, + "grad_norm": 0.1562713235616684, + "learning_rate": 1.8161856480877365e-05, + "loss": 1.0446, + "step": 11861 + }, + { + "epoch": 0.7206999210158576, + "grad_norm": 0.20497797429561615, + "learning_rate": 1.81544758850243e-05, + "loss": 1.0846, + "step": 11862 + }, + { + "epoch": 0.7207606780484841, + "grad_norm": 0.22672149538993835, + "learning_rate": 1.814709645647508e-05, + "loss": 1.1928, + "step": 11863 + }, + { + "epoch": 0.7208214350811106, + "grad_norm": 0.19137275218963623, + "learning_rate": 1.8139718195500193e-05, + "loss": 1.128, + "step": 11864 + }, + { + "epoch": 0.7208821921137372, + "grad_norm": 0.12515410780906677, + "learning_rate": 1.81323411023701e-05, + "loss": 1.0248, + "step": 11865 + }, + { + "epoch": 0.7209429491463637, + "grad_norm": 0.1429346650838852, + "learning_rate": 1.8124965177355195e-05, + "loss": 1.0492, + "step": 11866 + }, + { + "epoch": 0.7210037061789902, + "grad_norm": 4.356032848358154, + "learning_rate": 1.8117590420725856e-05, + "loss": 1.0648, + "step": 11867 + }, + { + "epoch": 0.7210644632116168, + "grad_norm": 0.19401760399341583, + "learning_rate": 1.8110216832752387e-05, + "loss": 0.9963, + "step": 11868 + }, + { + "epoch": 0.7211252202442433, + "grad_norm": 0.22318823635578156, + "learning_rate": 1.8102844413705117e-05, + "loss": 1.0719, + "step": 11869 + }, + { + "epoch": 0.7211859772768698, + "grad_norm": 0.17227022349834442, + "learning_rate": 1.8095473163854237e-05, + "loss": 1.0587, + "step": 11870 + }, + { + "epoch": 0.7212467343094964, + "grad_norm": 0.19656383991241455, + "learning_rate": 1.808810308346995e-05, + "loss": 1.0387, + "step": 11871 + }, + { + "epoch": 0.7213074913421228, + "grad_norm": 0.7389742732048035, + "learning_rate": 1.8080734172822417e-05, + "loss": 1.0626, + "step": 11872 + }, + { + "epoch": 0.7213682483747493, + "grad_norm": 0.846412718296051, + "learning_rate": 1.807336643218172e-05, + "loss": 1.0224, + "step": 11873 + }, + { + "epoch": 0.7214290054073759, + "grad_norm": 0.14130057394504547, + "learning_rate": 1.8065999861817972e-05, + "loss": 1.0384, + "step": 11874 + }, + { + "epoch": 0.7214897624400024, + "grad_norm": 0.11677338182926178, + "learning_rate": 1.8058634462001168e-05, + "loss": 1.0223, + "step": 11875 + }, + { + "epoch": 0.7215505194726289, + "grad_norm": 0.14386677742004395, + "learning_rate": 1.805127023300131e-05, + "loss": 1.0859, + "step": 11876 + }, + { + "epoch": 0.7216112765052555, + "grad_norm": 0.13483205437660217, + "learning_rate": 1.8043907175088298e-05, + "loss": 1.0045, + "step": 11877 + }, + { + "epoch": 0.721672033537882, + "grad_norm": 0.35182690620422363, + "learning_rate": 1.803654528853203e-05, + "loss": 1.2449, + "step": 11878 + }, + { + "epoch": 0.7217327905705085, + "grad_norm": 0.14638377726078033, + "learning_rate": 1.8029184573602394e-05, + "loss": 1.0598, + "step": 11879 + }, + { + "epoch": 0.7217935476031351, + "grad_norm": 0.15669922530651093, + "learning_rate": 1.8021825030569168e-05, + "loss": 0.9741, + "step": 11880 + }, + { + "epoch": 0.7218543046357616, + "grad_norm": 0.2722494602203369, + "learning_rate": 1.8014466659702133e-05, + "loss": 1.107, + "step": 11881 + }, + { + "epoch": 0.7219150616683881, + "grad_norm": 0.3362126648426056, + "learning_rate": 1.8007109461271006e-05, + "loss": 1.0359, + "step": 11882 + }, + { + "epoch": 0.7219758187010147, + "grad_norm": 0.5372222065925598, + "learning_rate": 1.7999753435545464e-05, + "loss": 1.2054, + "step": 11883 + }, + { + "epoch": 0.7220365757336412, + "grad_norm": 0.1586301028728485, + "learning_rate": 1.7992398582795146e-05, + "loss": 1.0304, + "step": 11884 + }, + { + "epoch": 0.7220973327662676, + "grad_norm": 0.14964400231838226, + "learning_rate": 1.7985044903289645e-05, + "loss": 1.0148, + "step": 11885 + }, + { + "epoch": 0.7221580897988942, + "grad_norm": 0.2265174686908722, + "learning_rate": 1.7977692397298507e-05, + "loss": 0.9982, + "step": 11886 + }, + { + "epoch": 0.7222188468315207, + "grad_norm": 0.156300351023674, + "learning_rate": 1.7970341065091245e-05, + "loss": 1.0446, + "step": 11887 + }, + { + "epoch": 0.7222796038641472, + "grad_norm": 0.8448696732521057, + "learning_rate": 1.7962990906937326e-05, + "loss": 1.1855, + "step": 11888 + }, + { + "epoch": 0.7223403608967738, + "grad_norm": 0.17689655721187592, + "learning_rate": 1.795564192310617e-05, + "loss": 1.0396, + "step": 11889 + }, + { + "epoch": 0.7224011179294003, + "grad_norm": 0.12835215032100677, + "learning_rate": 1.7948294113867146e-05, + "loss": 1.0317, + "step": 11890 + }, + { + "epoch": 0.7224618749620269, + "grad_norm": 0.22271111607551575, + "learning_rate": 1.7940947479489605e-05, + "loss": 1.1451, + "step": 11891 + }, + { + "epoch": 0.7225226319946534, + "grad_norm": 0.15610936284065247, + "learning_rate": 1.7933602020242823e-05, + "loss": 1.008, + "step": 11892 + }, + { + "epoch": 0.7225833890272799, + "grad_norm": 0.26250407099723816, + "learning_rate": 1.7926257736396062e-05, + "loss": 1.1122, + "step": 11893 + }, + { + "epoch": 0.7226441460599065, + "grad_norm": 0.17547118663787842, + "learning_rate": 1.791891462821851e-05, + "loss": 1.1361, + "step": 11894 + }, + { + "epoch": 0.722704903092533, + "grad_norm": 0.16468536853790283, + "learning_rate": 1.791157269597935e-05, + "loss": 1.0869, + "step": 11895 + }, + { + "epoch": 0.7227656601251595, + "grad_norm": 0.1483166366815567, + "learning_rate": 1.7904231939947724e-05, + "loss": 1.052, + "step": 11896 + }, + { + "epoch": 0.7228264171577861, + "grad_norm": 0.1373625099658966, + "learning_rate": 1.7896892360392654e-05, + "loss": 1.0255, + "step": 11897 + }, + { + "epoch": 0.7228871741904125, + "grad_norm": 0.1702195256948471, + "learning_rate": 1.7889553957583205e-05, + "loss": 1.0096, + "step": 11898 + }, + { + "epoch": 0.722947931223039, + "grad_norm": 0.18581053614616394, + "learning_rate": 1.788221673178834e-05, + "loss": 1.0439, + "step": 11899 + }, + { + "epoch": 0.7230086882556656, + "grad_norm": 0.23636052012443542, + "learning_rate": 1.7874880683277053e-05, + "loss": 1.1015, + "step": 11900 + }, + { + "epoch": 0.7230694452882921, + "grad_norm": 0.3609596788883209, + "learning_rate": 1.786754581231822e-05, + "loss": 1.1433, + "step": 11901 + }, + { + "epoch": 0.7231302023209186, + "grad_norm": 0.2429107427597046, + "learning_rate": 1.786021211918071e-05, + "loss": 1.0333, + "step": 11902 + }, + { + "epoch": 0.7231909593535452, + "grad_norm": 0.18088723719120026, + "learning_rate": 1.785287960413335e-05, + "loss": 1.2514, + "step": 11903 + }, + { + "epoch": 0.7232517163861717, + "grad_norm": 0.5686538219451904, + "learning_rate": 1.784554826744487e-05, + "loss": 1.0384, + "step": 11904 + }, + { + "epoch": 0.7233124734187982, + "grad_norm": 0.29504433274269104, + "learning_rate": 1.7838218109384053e-05, + "loss": 1.0303, + "step": 11905 + }, + { + "epoch": 0.7233732304514248, + "grad_norm": 0.26908689737319946, + "learning_rate": 1.7830889130219564e-05, + "loss": 1.0332, + "step": 11906 + }, + { + "epoch": 0.7234339874840513, + "grad_norm": 0.16701938211917877, + "learning_rate": 1.782356133022005e-05, + "loss": 1.0346, + "step": 11907 + }, + { + "epoch": 0.7234947445166778, + "grad_norm": 0.1376688927412033, + "learning_rate": 1.7816234709654124e-05, + "loss": 1.031, + "step": 11908 + }, + { + "epoch": 0.7235555015493044, + "grad_norm": 0.41214826703071594, + "learning_rate": 1.7808909268790323e-05, + "loss": 1.0156, + "step": 11909 + }, + { + "epoch": 0.7236162585819309, + "grad_norm": 0.24938194453716278, + "learning_rate": 1.7801585007897188e-05, + "loss": 1.2372, + "step": 11910 + }, + { + "epoch": 0.7236770156145574, + "grad_norm": 0.17977459728717804, + "learning_rate": 1.7794261927243173e-05, + "loss": 1.0269, + "step": 11911 + }, + { + "epoch": 0.7237377726471839, + "grad_norm": 0.17882023751735687, + "learning_rate": 1.778694002709671e-05, + "loss": 1.0469, + "step": 11912 + }, + { + "epoch": 0.7237985296798104, + "grad_norm": 0.21102429926395416, + "learning_rate": 1.7779619307726186e-05, + "loss": 1.0472, + "step": 11913 + }, + { + "epoch": 0.7238592867124369, + "grad_norm": 2.8092527389526367, + "learning_rate": 1.7772299769399948e-05, + "loss": 1.1479, + "step": 11914 + }, + { + "epoch": 0.7239200437450635, + "grad_norm": 0.19103248417377472, + "learning_rate": 1.7764981412386273e-05, + "loss": 1.0604, + "step": 11915 + }, + { + "epoch": 0.72398080077769, + "grad_norm": 0.2117866575717926, + "learning_rate": 1.7757664236953468e-05, + "loss": 1.0757, + "step": 11916 + }, + { + "epoch": 0.7240415578103165, + "grad_norm": 0.24587732553482056, + "learning_rate": 1.7750348243369703e-05, + "loss": 1.0895, + "step": 11917 + }, + { + "epoch": 0.7241023148429431, + "grad_norm": 0.21032288670539856, + "learning_rate": 1.774303343190315e-05, + "loss": 1.169, + "step": 11918 + }, + { + "epoch": 0.7241630718755696, + "grad_norm": 0.16315431892871857, + "learning_rate": 1.7735719802821943e-05, + "loss": 1.0285, + "step": 11919 + }, + { + "epoch": 0.7242238289081961, + "grad_norm": 0.17456279695034027, + "learning_rate": 1.7728407356394144e-05, + "loss": 1.0103, + "step": 11920 + }, + { + "epoch": 0.7242845859408227, + "grad_norm": 0.31635865569114685, + "learning_rate": 1.7721096092887833e-05, + "loss": 1.1727, + "step": 11921 + }, + { + "epoch": 0.7243453429734492, + "grad_norm": 0.24123826622962952, + "learning_rate": 1.7713786012570983e-05, + "loss": 1.0896, + "step": 11922 + }, + { + "epoch": 0.7244061000060757, + "grad_norm": 0.14461125433444977, + "learning_rate": 1.7706477115711563e-05, + "loss": 1.0288, + "step": 11923 + }, + { + "epoch": 0.7244668570387023, + "grad_norm": 0.17050157487392426, + "learning_rate": 1.7699169402577447e-05, + "loss": 1.0833, + "step": 11924 + }, + { + "epoch": 0.7245276140713287, + "grad_norm": 0.24234454333782196, + "learning_rate": 1.7691862873436504e-05, + "loss": 1.0838, + "step": 11925 + }, + { + "epoch": 0.7245883711039552, + "grad_norm": 0.1897338330745697, + "learning_rate": 1.7684557528556595e-05, + "loss": 1.0483, + "step": 11926 + }, + { + "epoch": 0.7246491281365818, + "grad_norm": 0.15417616069316864, + "learning_rate": 1.7677253368205472e-05, + "loss": 1.0307, + "step": 11927 + }, + { + "epoch": 0.7247098851692083, + "grad_norm": 0.2668818235397339, + "learning_rate": 1.7669950392650876e-05, + "loss": 1.2201, + "step": 11928 + }, + { + "epoch": 0.7247706422018348, + "grad_norm": 0.19664618372917175, + "learning_rate": 1.7662648602160493e-05, + "loss": 1.1476, + "step": 11929 + }, + { + "epoch": 0.7248313992344614, + "grad_norm": 0.2859170734882355, + "learning_rate": 1.765534799700198e-05, + "loss": 1.2732, + "step": 11930 + }, + { + "epoch": 0.7248921562670879, + "grad_norm": 4.838929653167725, + "learning_rate": 1.764804857744294e-05, + "loss": 1.1169, + "step": 11931 + }, + { + "epoch": 0.7249529132997145, + "grad_norm": 3.48994517326355, + "learning_rate": 1.764075034375093e-05, + "loss": 0.9972, + "step": 11932 + }, + { + "epoch": 0.725013670332341, + "grad_norm": 0.22503824532032013, + "learning_rate": 1.763345329619347e-05, + "loss": 1.0182, + "step": 11933 + }, + { + "epoch": 0.7250744273649675, + "grad_norm": 0.269570916891098, + "learning_rate": 1.762615743503804e-05, + "loss": 1.1426, + "step": 11934 + }, + { + "epoch": 0.7251351843975941, + "grad_norm": 0.23704588413238525, + "learning_rate": 1.7618862760552063e-05, + "loss": 1.1345, + "step": 11935 + }, + { + "epoch": 0.7251959414302206, + "grad_norm": 0.28229567408561707, + "learning_rate": 1.761156927300293e-05, + "loss": 1.1202, + "step": 11936 + }, + { + "epoch": 0.7252566984628471, + "grad_norm": 0.1913503110408783, + "learning_rate": 1.760427697265799e-05, + "loss": 1.0663, + "step": 11937 + }, + { + "epoch": 0.7253174554954736, + "grad_norm": 0.13558129966259003, + "learning_rate": 1.7596985859784537e-05, + "loss": 1.041, + "step": 11938 + }, + { + "epoch": 0.7253782125281001, + "grad_norm": 0.1980980634689331, + "learning_rate": 1.7589695934649837e-05, + "loss": 1.0458, + "step": 11939 + }, + { + "epoch": 0.7254389695607266, + "grad_norm": 0.22802449762821198, + "learning_rate": 1.758240719752109e-05, + "loss": 1.0628, + "step": 11940 + }, + { + "epoch": 0.7254997265933532, + "grad_norm": 0.18528719246387482, + "learning_rate": 1.7575119648665457e-05, + "loss": 1.0695, + "step": 11941 + }, + { + "epoch": 0.7255604836259797, + "grad_norm": 0.1925671100616455, + "learning_rate": 1.75678332883501e-05, + "loss": 1.0335, + "step": 11942 + }, + { + "epoch": 0.7256212406586062, + "grad_norm": 1.4555299282073975, + "learning_rate": 1.7560548116842097e-05, + "loss": 1.075, + "step": 11943 + }, + { + "epoch": 0.7256819976912328, + "grad_norm": 0.26899704337120056, + "learning_rate": 1.755326413440846e-05, + "loss": 1.1103, + "step": 11944 + }, + { + "epoch": 0.7257427547238593, + "grad_norm": 0.2331940084695816, + "learning_rate": 1.7545981341316174e-05, + "loss": 1.1571, + "step": 11945 + }, + { + "epoch": 0.7258035117564858, + "grad_norm": 0.21935653686523438, + "learning_rate": 1.7538699737832236e-05, + "loss": 1.0396, + "step": 11946 + }, + { + "epoch": 0.7258642687891124, + "grad_norm": 0.197550430893898, + "learning_rate": 1.7531419324223536e-05, + "loss": 1.0458, + "step": 11947 + }, + { + "epoch": 0.7259250258217389, + "grad_norm": 0.2114322930574417, + "learning_rate": 1.7524140100756937e-05, + "loss": 1.0094, + "step": 11948 + }, + { + "epoch": 0.7259857828543654, + "grad_norm": 0.8108474612236023, + "learning_rate": 1.751686206769926e-05, + "loss": 1.2194, + "step": 11949 + }, + { + "epoch": 0.726046539886992, + "grad_norm": 0.24939508736133575, + "learning_rate": 1.750958522531728e-05, + "loss": 1.1657, + "step": 11950 + }, + { + "epoch": 0.7261072969196184, + "grad_norm": 0.20924407243728638, + "learning_rate": 1.7502309573877735e-05, + "loss": 0.9986, + "step": 11951 + }, + { + "epoch": 0.7261680539522449, + "grad_norm": 0.18297210335731506, + "learning_rate": 1.7495035113647312e-05, + "loss": 0.9844, + "step": 11952 + }, + { + "epoch": 0.7262288109848715, + "grad_norm": 0.218725323677063, + "learning_rate": 1.7487761844892663e-05, + "loss": 1.0683, + "step": 11953 + }, + { + "epoch": 0.726289568017498, + "grad_norm": 0.22949229180812836, + "learning_rate": 1.7480489767880388e-05, + "loss": 1.1788, + "step": 11954 + }, + { + "epoch": 0.7263503250501245, + "grad_norm": 0.23814409971237183, + "learning_rate": 1.7473218882877047e-05, + "loss": 1.228, + "step": 11955 + }, + { + "epoch": 0.7264110820827511, + "grad_norm": 1.6760926246643066, + "learning_rate": 1.7465949190149155e-05, + "loss": 1.0772, + "step": 11956 + }, + { + "epoch": 0.7264718391153776, + "grad_norm": 0.15535205602645874, + "learning_rate": 1.745868068996318e-05, + "loss": 1.0198, + "step": 11957 + }, + { + "epoch": 0.7265325961480041, + "grad_norm": 0.18121644854545593, + "learning_rate": 1.7451413382585557e-05, + "loss": 1.0179, + "step": 11958 + }, + { + "epoch": 0.7265933531806307, + "grad_norm": 1.4574313163757324, + "learning_rate": 1.7444147268282673e-05, + "loss": 1.0778, + "step": 11959 + }, + { + "epoch": 0.7266541102132572, + "grad_norm": 0.18895384669303894, + "learning_rate": 1.7436882347320856e-05, + "loss": 1.0517, + "step": 11960 + }, + { + "epoch": 0.7267148672458837, + "grad_norm": 0.1729179173707962, + "learning_rate": 1.74296186199664e-05, + "loss": 1.051, + "step": 11961 + }, + { + "epoch": 0.7267756242785103, + "grad_norm": 0.13464312255382538, + "learning_rate": 1.742235608648558e-05, + "loss": 1.0244, + "step": 11962 + }, + { + "epoch": 0.7268363813111368, + "grad_norm": 0.19090884923934937, + "learning_rate": 1.7415094747144617e-05, + "loss": 1.1041, + "step": 11963 + }, + { + "epoch": 0.7268971383437632, + "grad_norm": 0.2165987491607666, + "learning_rate": 1.7407834602209633e-05, + "loss": 1.0307, + "step": 11964 + }, + { + "epoch": 0.7269578953763898, + "grad_norm": 0.20466408133506775, + "learning_rate": 1.7400575651946778e-05, + "loss": 1.1328, + "step": 11965 + }, + { + "epoch": 0.7270186524090163, + "grad_norm": 0.22854872047901154, + "learning_rate": 1.7393317896622103e-05, + "loss": 1.1265, + "step": 11966 + }, + { + "epoch": 0.7270794094416428, + "grad_norm": 0.13764043152332306, + "learning_rate": 1.738606133650168e-05, + "loss": 1.0287, + "step": 11967 + }, + { + "epoch": 0.7271401664742694, + "grad_norm": 0.14283035695552826, + "learning_rate": 1.737880597185148e-05, + "loss": 1.0139, + "step": 11968 + }, + { + "epoch": 0.7272009235068959, + "grad_norm": 0.24339868128299713, + "learning_rate": 1.7371551802937456e-05, + "loss": 1.1074, + "step": 11969 + }, + { + "epoch": 0.7272616805395224, + "grad_norm": 0.1635340303182602, + "learning_rate": 1.736429883002553e-05, + "loss": 1.097, + "step": 11970 + }, + { + "epoch": 0.727322437572149, + "grad_norm": 0.1373094618320465, + "learning_rate": 1.7357047053381492e-05, + "loss": 1.0461, + "step": 11971 + }, + { + "epoch": 0.7273831946047755, + "grad_norm": 0.1404392421245575, + "learning_rate": 1.734979647327123e-05, + "loss": 1.0539, + "step": 11972 + }, + { + "epoch": 0.727443951637402, + "grad_norm": 0.1432620733976364, + "learning_rate": 1.7342547089960487e-05, + "loss": 1.041, + "step": 11973 + }, + { + "epoch": 0.7275047086700286, + "grad_norm": 0.16150271892547607, + "learning_rate": 1.733529890371499e-05, + "loss": 1.1117, + "step": 11974 + }, + { + "epoch": 0.7275654657026551, + "grad_norm": 0.1858302354812622, + "learning_rate": 1.7328051914800432e-05, + "loss": 1.077, + "step": 11975 + }, + { + "epoch": 0.7276262227352817, + "grad_norm": 0.14762087166309357, + "learning_rate": 1.7320806123482447e-05, + "loss": 1.0158, + "step": 11976 + }, + { + "epoch": 0.7276869797679081, + "grad_norm": 0.232815682888031, + "learning_rate": 1.7313561530026624e-05, + "loss": 1.1021, + "step": 11977 + }, + { + "epoch": 0.7277477368005346, + "grad_norm": 0.24023056030273438, + "learning_rate": 1.7306318134698523e-05, + "loss": 1.0765, + "step": 11978 + }, + { + "epoch": 0.7278084938331612, + "grad_norm": 0.23971092700958252, + "learning_rate": 1.7299075937763654e-05, + "loss": 1.0813, + "step": 11979 + }, + { + "epoch": 0.7278692508657877, + "grad_norm": 0.16451777517795563, + "learning_rate": 1.7291834939487473e-05, + "loss": 1.1001, + "step": 11980 + }, + { + "epoch": 0.7279300078984142, + "grad_norm": 0.18622159957885742, + "learning_rate": 1.7284595140135408e-05, + "loss": 1.0351, + "step": 11981 + }, + { + "epoch": 0.7279907649310408, + "grad_norm": 0.19943372905254364, + "learning_rate": 1.7277356539972812e-05, + "loss": 1.0809, + "step": 11982 + }, + { + "epoch": 0.7280515219636673, + "grad_norm": 0.22772513329982758, + "learning_rate": 1.7270119139265072e-05, + "loss": 1.0629, + "step": 11983 + }, + { + "epoch": 0.7281122789962938, + "grad_norm": 0.2238074392080307, + "learning_rate": 1.7262882938277425e-05, + "loss": 1.0609, + "step": 11984 + }, + { + "epoch": 0.7281730360289204, + "grad_norm": 0.14449524879455566, + "learning_rate": 1.7255647937275133e-05, + "loss": 0.9983, + "step": 11985 + }, + { + "epoch": 0.7282337930615469, + "grad_norm": 0.18908697366714478, + "learning_rate": 1.7248414136523396e-05, + "loss": 1.0114, + "step": 11986 + }, + { + "epoch": 0.7282945500941734, + "grad_norm": 0.44009923934936523, + "learning_rate": 1.724118153628735e-05, + "loss": 1.067, + "step": 11987 + }, + { + "epoch": 0.7283553071268, + "grad_norm": 0.14333678781986237, + "learning_rate": 1.7233950136832143e-05, + "loss": 1.0583, + "step": 11988 + }, + { + "epoch": 0.7284160641594265, + "grad_norm": 0.26585403084754944, + "learning_rate": 1.722671993842283e-05, + "loss": 1.0155, + "step": 11989 + }, + { + "epoch": 0.7284768211920529, + "grad_norm": 0.25955429673194885, + "learning_rate": 1.7219490941324444e-05, + "loss": 1.092, + "step": 11990 + }, + { + "epoch": 0.7285375782246795, + "grad_norm": 0.3298370838165283, + "learning_rate": 1.7212263145801943e-05, + "loss": 1.028, + "step": 11991 + }, + { + "epoch": 0.728598335257306, + "grad_norm": 0.16183407604694366, + "learning_rate": 1.7205036552120257e-05, + "loss": 1.0536, + "step": 11992 + }, + { + "epoch": 0.7286590922899325, + "grad_norm": 0.35626351833343506, + "learning_rate": 1.7197811160544307e-05, + "loss": 1.1647, + "step": 11993 + }, + { + "epoch": 0.7287198493225591, + "grad_norm": 0.19814515113830566, + "learning_rate": 1.7190586971338935e-05, + "loss": 1.0597, + "step": 11994 + }, + { + "epoch": 0.7287806063551856, + "grad_norm": 0.2423521727323532, + "learning_rate": 1.7183363984768935e-05, + "loss": 1.1947, + "step": 11995 + }, + { + "epoch": 0.7288413633878121, + "grad_norm": 0.16210925579071045, + "learning_rate": 1.7176142201099077e-05, + "loss": 1.0147, + "step": 11996 + }, + { + "epoch": 0.7289021204204387, + "grad_norm": 0.15628249943256378, + "learning_rate": 1.716892162059407e-05, + "loss": 1.0709, + "step": 11997 + }, + { + "epoch": 0.7289628774530652, + "grad_norm": 0.11739705502986908, + "learning_rate": 1.7161702243518585e-05, + "loss": 1.0196, + "step": 11998 + }, + { + "epoch": 0.7290236344856917, + "grad_norm": 2.2876222133636475, + "learning_rate": 1.715448407013725e-05, + "loss": 1.057, + "step": 11999 + }, + { + "epoch": 0.7290843915183183, + "grad_norm": 0.1299145370721817, + "learning_rate": 1.7147267100714654e-05, + "loss": 1.0623, + "step": 12000 + }, + { + "epoch": 0.7291451485509448, + "grad_norm": 0.25023528933525085, + "learning_rate": 1.714005133551533e-05, + "loss": 1.1499, + "step": 12001 + }, + { + "epoch": 0.7292059055835713, + "grad_norm": 0.16520905494689941, + "learning_rate": 1.713283677480378e-05, + "loss": 1.0335, + "step": 12002 + }, + { + "epoch": 0.7292666626161978, + "grad_norm": 0.11456679552793503, + "learning_rate": 1.7125623418844444e-05, + "loss": 1.0105, + "step": 12003 + }, + { + "epoch": 0.7293274196488243, + "grad_norm": 0.1832854002714157, + "learning_rate": 1.711841126790174e-05, + "loss": 1.0955, + "step": 12004 + }, + { + "epoch": 0.7293881766814508, + "grad_norm": 0.191285640001297, + "learning_rate": 1.7111200322240027e-05, + "loss": 1.1003, + "step": 12005 + }, + { + "epoch": 0.7294489337140774, + "grad_norm": 0.5203683376312256, + "learning_rate": 1.7103990582123624e-05, + "loss": 1.0494, + "step": 12006 + }, + { + "epoch": 0.7295096907467039, + "grad_norm": 0.18868228793144226, + "learning_rate": 1.7096782047816805e-05, + "loss": 1.1349, + "step": 12007 + }, + { + "epoch": 0.7295704477793304, + "grad_norm": 0.16170978546142578, + "learning_rate": 1.7089574719583785e-05, + "loss": 1.072, + "step": 12008 + }, + { + "epoch": 0.729631204811957, + "grad_norm": 0.15769550204277039, + "learning_rate": 1.708236859768878e-05, + "loss": 1.0047, + "step": 12009 + }, + { + "epoch": 0.7296919618445835, + "grad_norm": 0.23780043423175812, + "learning_rate": 1.7075163682395933e-05, + "loss": 1.0906, + "step": 12010 + }, + { + "epoch": 0.72975271887721, + "grad_norm": 0.1336865872144699, + "learning_rate": 1.706795997396931e-05, + "loss": 1.0535, + "step": 12011 + }, + { + "epoch": 0.7298134759098366, + "grad_norm": 0.21559228003025055, + "learning_rate": 1.7060757472672977e-05, + "loss": 1.1183, + "step": 12012 + }, + { + "epoch": 0.7298742329424631, + "grad_norm": 0.2091226726770401, + "learning_rate": 1.705355617877093e-05, + "loss": 1.0263, + "step": 12013 + }, + { + "epoch": 0.7299349899750897, + "grad_norm": 0.1745561957359314, + "learning_rate": 1.704635609252717e-05, + "loss": 1.0521, + "step": 12014 + }, + { + "epoch": 0.7299957470077162, + "grad_norm": 0.28505438566207886, + "learning_rate": 1.7039157214205593e-05, + "loss": 1.0815, + "step": 12015 + }, + { + "epoch": 0.7300565040403427, + "grad_norm": 0.1525629460811615, + "learning_rate": 1.7031959544070082e-05, + "loss": 1.0378, + "step": 12016 + }, + { + "epoch": 0.7301172610729691, + "grad_norm": 0.3699992001056671, + "learning_rate": 1.7024763082384483e-05, + "loss": 1.162, + "step": 12017 + }, + { + "epoch": 0.7301780181055957, + "grad_norm": 0.33950337767601013, + "learning_rate": 1.7017567829412535e-05, + "loss": 1.0941, + "step": 12018 + }, + { + "epoch": 0.7302387751382222, + "grad_norm": 0.14099258184432983, + "learning_rate": 1.7010373785418028e-05, + "loss": 1.0709, + "step": 12019 + }, + { + "epoch": 0.7302995321708488, + "grad_norm": 0.18766345083713531, + "learning_rate": 1.7003180950664642e-05, + "loss": 1.0713, + "step": 12020 + }, + { + "epoch": 0.7303602892034753, + "grad_norm": 0.24312561750411987, + "learning_rate": 1.6995989325416046e-05, + "loss": 0.9712, + "step": 12021 + }, + { + "epoch": 0.7304210462361018, + "grad_norm": 0.12350136041641235, + "learning_rate": 1.6988798909935833e-05, + "loss": 1.0382, + "step": 12022 + }, + { + "epoch": 0.7304818032687284, + "grad_norm": 0.47753164172172546, + "learning_rate": 1.6981609704487583e-05, + "loss": 1.0398, + "step": 12023 + }, + { + "epoch": 0.7305425603013549, + "grad_norm": 0.1921849548816681, + "learning_rate": 1.6974421709334808e-05, + "loss": 1.0841, + "step": 12024 + }, + { + "epoch": 0.7306033173339814, + "grad_norm": 5.323007583618164, + "learning_rate": 1.696723492474099e-05, + "loss": 1.0734, + "step": 12025 + }, + { + "epoch": 0.730664074366608, + "grad_norm": 0.4686852693557739, + "learning_rate": 1.696004935096956e-05, + "loss": 1.2119, + "step": 12026 + }, + { + "epoch": 0.7307248313992345, + "grad_norm": 0.17589835822582245, + "learning_rate": 1.6952864988283907e-05, + "loss": 1.1086, + "step": 12027 + }, + { + "epoch": 0.730785588431861, + "grad_norm": 2.8391520977020264, + "learning_rate": 1.6945681836947374e-05, + "loss": 1.0221, + "step": 12028 + }, + { + "epoch": 0.7308463454644876, + "grad_norm": 0.19321005046367645, + "learning_rate": 1.6938499897223252e-05, + "loss": 0.9946, + "step": 12029 + }, + { + "epoch": 0.730907102497114, + "grad_norm": 0.40528377890586853, + "learning_rate": 1.6931319169374836e-05, + "loss": 1.0487, + "step": 12030 + }, + { + "epoch": 0.7309678595297405, + "grad_norm": 0.15739288926124573, + "learning_rate": 1.6924139653665293e-05, + "loss": 1.0249, + "step": 12031 + }, + { + "epoch": 0.7310286165623671, + "grad_norm": 0.1646651029586792, + "learning_rate": 1.6916961350357802e-05, + "loss": 0.9869, + "step": 12032 + }, + { + "epoch": 0.7310893735949936, + "grad_norm": 0.19569137692451477, + "learning_rate": 1.690978425971549e-05, + "loss": 1.0127, + "step": 12033 + }, + { + "epoch": 0.7311501306276201, + "grad_norm": 0.26992595195770264, + "learning_rate": 1.6902608382001415e-05, + "loss": 1.028, + "step": 12034 + }, + { + "epoch": 0.7312108876602467, + "grad_norm": 0.5256550312042236, + "learning_rate": 1.6895433717478643e-05, + "loss": 1.1255, + "step": 12035 + }, + { + "epoch": 0.7312716446928732, + "grad_norm": 0.1441774070262909, + "learning_rate": 1.688826026641015e-05, + "loss": 1.0263, + "step": 12036 + }, + { + "epoch": 0.7313324017254997, + "grad_norm": 0.13416828215122223, + "learning_rate": 1.688108802905889e-05, + "loss": 1.0203, + "step": 12037 + }, + { + "epoch": 0.7313931587581263, + "grad_norm": 0.21125821769237518, + "learning_rate": 1.687391700568772e-05, + "loss": 1.1447, + "step": 12038 + }, + { + "epoch": 0.7314539157907528, + "grad_norm": 0.15147802233695984, + "learning_rate": 1.686674719655954e-05, + "loss": 1.09, + "step": 12039 + }, + { + "epoch": 0.7315146728233793, + "grad_norm": 0.22955982387065887, + "learning_rate": 1.6859578601937144e-05, + "loss": 1.1211, + "step": 12040 + }, + { + "epoch": 0.7315754298560059, + "grad_norm": 0.1889542192220688, + "learning_rate": 1.6852411222083302e-05, + "loss": 1.0976, + "step": 12041 + }, + { + "epoch": 0.7316361868886324, + "grad_norm": 0.13393111526966095, + "learning_rate": 1.6845245057260734e-05, + "loss": 0.9946, + "step": 12042 + }, + { + "epoch": 0.7316969439212588, + "grad_norm": 0.18932290375232697, + "learning_rate": 1.6838080107732117e-05, + "loss": 1.0776, + "step": 12043 + }, + { + "epoch": 0.7317577009538854, + "grad_norm": 0.9971408247947693, + "learning_rate": 1.6830916373760087e-05, + "loss": 1.1474, + "step": 12044 + }, + { + "epoch": 0.7318184579865119, + "grad_norm": 0.23503944277763367, + "learning_rate": 1.6823753855607222e-05, + "loss": 1.2034, + "step": 12045 + }, + { + "epoch": 0.7318792150191384, + "grad_norm": 0.22641731798648834, + "learning_rate": 1.6816592553536077e-05, + "loss": 1.0473, + "step": 12046 + }, + { + "epoch": 0.731939972051765, + "grad_norm": 0.2162696272134781, + "learning_rate": 1.6809432467809144e-05, + "loss": 1.013, + "step": 12047 + }, + { + "epoch": 0.7320007290843915, + "grad_norm": 0.15569138526916504, + "learning_rate": 1.6802273598688884e-05, + "loss": 1.016, + "step": 12048 + }, + { + "epoch": 0.732061486117018, + "grad_norm": 0.24236425757408142, + "learning_rate": 1.67951159464377e-05, + "loss": 1.1059, + "step": 12049 + }, + { + "epoch": 0.7321222431496446, + "grad_norm": 0.21469469368457794, + "learning_rate": 1.6787959511317958e-05, + "loss": 1.0494, + "step": 12050 + }, + { + "epoch": 0.7321830001822711, + "grad_norm": 0.250492662191391, + "learning_rate": 1.678080429359198e-05, + "loss": 1.0414, + "step": 12051 + }, + { + "epoch": 0.7322437572148977, + "grad_norm": 0.1607767790555954, + "learning_rate": 1.677365029352204e-05, + "loss": 1.0617, + "step": 12052 + }, + { + "epoch": 0.7323045142475242, + "grad_norm": 0.3698523938655853, + "learning_rate": 1.6766497511370376e-05, + "loss": 1.12, + "step": 12053 + }, + { + "epoch": 0.7323652712801507, + "grad_norm": 0.24637648463249207, + "learning_rate": 1.6759345947399145e-05, + "loss": 1.0573, + "step": 12054 + }, + { + "epoch": 0.7324260283127773, + "grad_norm": 0.17111331224441528, + "learning_rate": 1.675219560187054e-05, + "loss": 1.0794, + "step": 12055 + }, + { + "epoch": 0.7324867853454037, + "grad_norm": 0.15086501836776733, + "learning_rate": 1.6745046475046627e-05, + "loss": 1.0361, + "step": 12056 + }, + { + "epoch": 0.7325475423780302, + "grad_norm": 0.24123866856098175, + "learning_rate": 1.673789856718948e-05, + "loss": 1.0922, + "step": 12057 + }, + { + "epoch": 0.7326082994106567, + "grad_norm": 0.14296065270900726, + "learning_rate": 1.6730751878561074e-05, + "loss": 1.0501, + "step": 12058 + }, + { + "epoch": 0.7326690564432833, + "grad_norm": 0.1941286325454712, + "learning_rate": 1.672360640942337e-05, + "loss": 1.0964, + "step": 12059 + }, + { + "epoch": 0.7327298134759098, + "grad_norm": 0.20296406745910645, + "learning_rate": 1.6716462160038325e-05, + "loss": 1.1387, + "step": 12060 + }, + { + "epoch": 0.7327905705085364, + "grad_norm": 0.21449318528175354, + "learning_rate": 1.6709319130667793e-05, + "loss": 1.0423, + "step": 12061 + }, + { + "epoch": 0.7328513275411629, + "grad_norm": 0.12751971185207367, + "learning_rate": 1.6702177321573603e-05, + "loss": 1.0225, + "step": 12062 + }, + { + "epoch": 0.7329120845737894, + "grad_norm": 0.4117542505264282, + "learning_rate": 1.669503673301754e-05, + "loss": 1.0386, + "step": 12063 + }, + { + "epoch": 0.732972841606416, + "grad_norm": 0.18628422915935516, + "learning_rate": 1.668789736526134e-05, + "loss": 1.1576, + "step": 12064 + }, + { + "epoch": 0.7330335986390425, + "grad_norm": 0.1504087597131729, + "learning_rate": 1.6680759218566704e-05, + "loss": 1.0233, + "step": 12065 + }, + { + "epoch": 0.733094355671669, + "grad_norm": 1.1308015584945679, + "learning_rate": 1.667362229319528e-05, + "loss": 1.0369, + "step": 12066 + }, + { + "epoch": 0.7331551127042956, + "grad_norm": 0.13854114711284637, + "learning_rate": 1.666648658940867e-05, + "loss": 1.0185, + "step": 12067 + }, + { + "epoch": 0.7332158697369221, + "grad_norm": 0.3720371425151825, + "learning_rate": 1.665935210746844e-05, + "loss": 1.1112, + "step": 12068 + }, + { + "epoch": 0.7332766267695485, + "grad_norm": 0.1353306919336319, + "learning_rate": 1.6652218847636098e-05, + "loss": 0.9797, + "step": 12069 + }, + { + "epoch": 0.7333373838021751, + "grad_norm": 0.13254907727241516, + "learning_rate": 1.664508681017313e-05, + "loss": 1.0369, + "step": 12070 + }, + { + "epoch": 0.7333981408348016, + "grad_norm": 0.16061079502105713, + "learning_rate": 1.663795599534094e-05, + "loss": 1.0829, + "step": 12071 + }, + { + "epoch": 0.7334588978674281, + "grad_norm": 0.1492386758327484, + "learning_rate": 1.663082640340093e-05, + "loss": 1.063, + "step": 12072 + }, + { + "epoch": 0.7335196549000547, + "grad_norm": 1.5714997053146362, + "learning_rate": 1.6623698034614428e-05, + "loss": 1.1545, + "step": 12073 + }, + { + "epoch": 0.7335804119326812, + "grad_norm": 0.21003714203834534, + "learning_rate": 1.661657088924272e-05, + "loss": 1.0533, + "step": 12074 + }, + { + "epoch": 0.7336411689653077, + "grad_norm": 0.15426969528198242, + "learning_rate": 1.660944496754705e-05, + "loss": 1.0477, + "step": 12075 + }, + { + "epoch": 0.7337019259979343, + "grad_norm": 0.21045248210430145, + "learning_rate": 1.660232026978864e-05, + "loss": 1.0337, + "step": 12076 + }, + { + "epoch": 0.7337626830305608, + "grad_norm": 0.25851884484291077, + "learning_rate": 1.6595196796228655e-05, + "loss": 1.0252, + "step": 12077 + }, + { + "epoch": 0.7338234400631873, + "grad_norm": 0.14234012365341187, + "learning_rate": 1.6588074547128173e-05, + "loss": 0.981, + "step": 12078 + }, + { + "epoch": 0.7338841970958139, + "grad_norm": 0.18745067715644836, + "learning_rate": 1.658095352274827e-05, + "loss": 1.0726, + "step": 12079 + }, + { + "epoch": 0.7339449541284404, + "grad_norm": 0.22129912674427032, + "learning_rate": 1.657383372334997e-05, + "loss": 1.1589, + "step": 12080 + }, + { + "epoch": 0.7340057111610669, + "grad_norm": 0.23530568182468414, + "learning_rate": 1.656671514919427e-05, + "loss": 1.0865, + "step": 12081 + }, + { + "epoch": 0.7340664681936934, + "grad_norm": 0.29802191257476807, + "learning_rate": 1.6559597800542082e-05, + "loss": 1.1477, + "step": 12082 + }, + { + "epoch": 0.7341272252263199, + "grad_norm": 0.44481226801872253, + "learning_rate": 1.6552481677654302e-05, + "loss": 1.2021, + "step": 12083 + }, + { + "epoch": 0.7341879822589464, + "grad_norm": 0.2596678137779236, + "learning_rate": 1.6545366780791792e-05, + "loss": 1.1738, + "step": 12084 + }, + { + "epoch": 0.734248739291573, + "grad_norm": 0.30902406573295593, + "learning_rate": 1.6538253110215292e-05, + "loss": 1.2012, + "step": 12085 + }, + { + "epoch": 0.7343094963241995, + "grad_norm": 0.2586876451969147, + "learning_rate": 1.6531140666185614e-05, + "loss": 1.0681, + "step": 12086 + }, + { + "epoch": 0.734370253356826, + "grad_norm": 0.237026646733284, + "learning_rate": 1.652402944896344e-05, + "loss": 1.0179, + "step": 12087 + }, + { + "epoch": 0.7344310103894526, + "grad_norm": 0.2303038388490677, + "learning_rate": 1.651691945880944e-05, + "loss": 1.077, + "step": 12088 + }, + { + "epoch": 0.7344917674220791, + "grad_norm": 0.24662743508815765, + "learning_rate": 1.6509810695984225e-05, + "loss": 0.9797, + "step": 12089 + }, + { + "epoch": 0.7345525244547056, + "grad_norm": 0.20820677280426025, + "learning_rate": 1.650270316074838e-05, + "loss": 1.0595, + "step": 12090 + }, + { + "epoch": 0.7346132814873322, + "grad_norm": 0.21978025138378143, + "learning_rate": 1.6495596853362415e-05, + "loss": 1.1082, + "step": 12091 + }, + { + "epoch": 0.7346740385199587, + "grad_norm": 0.17106331884860992, + "learning_rate": 1.6488491774086834e-05, + "loss": 1.0377, + "step": 12092 + }, + { + "epoch": 0.7347347955525853, + "grad_norm": 0.2536803185939789, + "learning_rate": 1.6481387923182055e-05, + "loss": 1.0939, + "step": 12093 + }, + { + "epoch": 0.7347955525852118, + "grad_norm": 0.19816751778125763, + "learning_rate": 1.6474285300908486e-05, + "loss": 1.0509, + "step": 12094 + }, + { + "epoch": 0.7348563096178382, + "grad_norm": 0.26066505908966064, + "learning_rate": 1.6467183907526472e-05, + "loss": 1.027, + "step": 12095 + }, + { + "epoch": 0.7349170666504647, + "grad_norm": 0.25249990820884705, + "learning_rate": 1.6460083743296296e-05, + "loss": 0.9966, + "step": 12096 + }, + { + "epoch": 0.7349778236830913, + "grad_norm": 0.24441476166248322, + "learning_rate": 1.6452984808478267e-05, + "loss": 1.154, + "step": 12097 + }, + { + "epoch": 0.7350385807157178, + "grad_norm": 0.4724961519241333, + "learning_rate": 1.644588710333254e-05, + "loss": 1.0172, + "step": 12098 + }, + { + "epoch": 0.7350993377483444, + "grad_norm": 0.2478560358285904, + "learning_rate": 1.6438790628119317e-05, + "loss": 1.179, + "step": 12099 + }, + { + "epoch": 0.7351600947809709, + "grad_norm": 0.19919981062412262, + "learning_rate": 1.6431695383098704e-05, + "loss": 1.088, + "step": 12100 + }, + { + "epoch": 0.7352208518135974, + "grad_norm": 0.25076404213905334, + "learning_rate": 1.6424601368530772e-05, + "loss": 1.1931, + "step": 12101 + }, + { + "epoch": 0.735281608846224, + "grad_norm": 0.22720111906528473, + "learning_rate": 1.6417508584675583e-05, + "loss": 1.1454, + "step": 12102 + }, + { + "epoch": 0.7353423658788505, + "grad_norm": 0.21579253673553467, + "learning_rate": 1.6410417031793112e-05, + "loss": 1.0267, + "step": 12103 + }, + { + "epoch": 0.735403122911477, + "grad_norm": 0.19889214634895325, + "learning_rate": 1.6403326710143306e-05, + "loss": 1.0709, + "step": 12104 + }, + { + "epoch": 0.7354638799441036, + "grad_norm": 0.16239091753959656, + "learning_rate": 1.6396237619986042e-05, + "loss": 1.0589, + "step": 12105 + }, + { + "epoch": 0.7355246369767301, + "grad_norm": 0.17132940888404846, + "learning_rate": 1.638914976158117e-05, + "loss": 1.0116, + "step": 12106 + }, + { + "epoch": 0.7355853940093566, + "grad_norm": 0.1622474044561386, + "learning_rate": 1.6382063135188517e-05, + "loss": 1.0393, + "step": 12107 + }, + { + "epoch": 0.735646151041983, + "grad_norm": 0.15770988166332245, + "learning_rate": 1.6374977741067847e-05, + "loss": 1.0716, + "step": 12108 + }, + { + "epoch": 0.7357069080746096, + "grad_norm": 0.13953624665737152, + "learning_rate": 1.6367893579478865e-05, + "loss": 1.0459, + "step": 12109 + }, + { + "epoch": 0.7357676651072361, + "grad_norm": 0.18816706538200378, + "learning_rate": 1.6360810650681262e-05, + "loss": 1.0948, + "step": 12110 + }, + { + "epoch": 0.7358284221398627, + "grad_norm": 0.46075230836868286, + "learning_rate": 1.635372895493461e-05, + "loss": 1.0202, + "step": 12111 + }, + { + "epoch": 0.7358891791724892, + "grad_norm": 0.16444611549377441, + "learning_rate": 1.634664849249855e-05, + "loss": 1.1129, + "step": 12112 + }, + { + "epoch": 0.7359499362051157, + "grad_norm": 0.14240719377994537, + "learning_rate": 1.6339569263632588e-05, + "loss": 1.0604, + "step": 12113 + }, + { + "epoch": 0.7360106932377423, + "grad_norm": 0.12777021527290344, + "learning_rate": 1.6332491268596223e-05, + "loss": 1.0154, + "step": 12114 + }, + { + "epoch": 0.7360714502703688, + "grad_norm": 1.069507122039795, + "learning_rate": 1.63254145076489e-05, + "loss": 1.1258, + "step": 12115 + }, + { + "epoch": 0.7361322073029953, + "grad_norm": 0.18157348036766052, + "learning_rate": 1.631833898105002e-05, + "loss": 1.0709, + "step": 12116 + }, + { + "epoch": 0.7361929643356219, + "grad_norm": 0.13767734169960022, + "learning_rate": 1.631126468905893e-05, + "loss": 1.0901, + "step": 12117 + }, + { + "epoch": 0.7362537213682484, + "grad_norm": 0.18010273575782776, + "learning_rate": 1.6304191631934945e-05, + "loss": 1.0138, + "step": 12118 + }, + { + "epoch": 0.7363144784008749, + "grad_norm": 0.2551179528236389, + "learning_rate": 1.6297119809937328e-05, + "loss": 1.1061, + "step": 12119 + }, + { + "epoch": 0.7363752354335015, + "grad_norm": 0.13636304438114166, + "learning_rate": 1.6290049223325304e-05, + "loss": 1.066, + "step": 12120 + }, + { + "epoch": 0.736435992466128, + "grad_norm": 0.19603414833545685, + "learning_rate": 1.628297987235804e-05, + "loss": 1.0329, + "step": 12121 + }, + { + "epoch": 0.7364967494987544, + "grad_norm": 0.2311643362045288, + "learning_rate": 1.627591175729465e-05, + "loss": 1.1393, + "step": 12122 + }, + { + "epoch": 0.736557506531381, + "grad_norm": 0.12488242983818054, + "learning_rate": 1.6268844878394247e-05, + "loss": 0.9805, + "step": 12123 + }, + { + "epoch": 0.7366182635640075, + "grad_norm": 0.2144162505865097, + "learning_rate": 1.626177923591587e-05, + "loss": 1.0972, + "step": 12124 + }, + { + "epoch": 0.736679020596634, + "grad_norm": 0.15694460272789001, + "learning_rate": 1.6254714830118482e-05, + "loss": 1.0641, + "step": 12125 + }, + { + "epoch": 0.7367397776292606, + "grad_norm": 0.33337998390197754, + "learning_rate": 1.6247651661261047e-05, + "loss": 1.1084, + "step": 12126 + }, + { + "epoch": 0.7368005346618871, + "grad_norm": 0.1301247924566269, + "learning_rate": 1.6240589729602446e-05, + "loss": 1.0549, + "step": 12127 + }, + { + "epoch": 0.7368612916945136, + "grad_norm": 0.28256696462631226, + "learning_rate": 1.6233529035401575e-05, + "loss": 1.1697, + "step": 12128 + }, + { + "epoch": 0.7369220487271402, + "grad_norm": 0.24133507907390594, + "learning_rate": 1.622646957891722e-05, + "loss": 1.0532, + "step": 12129 + }, + { + "epoch": 0.7369828057597667, + "grad_norm": 0.14497233927249908, + "learning_rate": 1.6219411360408148e-05, + "loss": 1.0777, + "step": 12130 + }, + { + "epoch": 0.7370435627923932, + "grad_norm": 0.1488426774740219, + "learning_rate": 1.6212354380133088e-05, + "loss": 1.0438, + "step": 12131 + }, + { + "epoch": 0.7371043198250198, + "grad_norm": 0.2765292823314667, + "learning_rate": 1.6205298638350704e-05, + "loss": 1.0146, + "step": 12132 + }, + { + "epoch": 0.7371650768576463, + "grad_norm": 0.25340497493743896, + "learning_rate": 1.6198244135319636e-05, + "loss": 1.1333, + "step": 12133 + }, + { + "epoch": 0.7372258338902729, + "grad_norm": 0.23027236759662628, + "learning_rate": 1.6191190871298457e-05, + "loss": 1.0926, + "step": 12134 + }, + { + "epoch": 0.7372865909228993, + "grad_norm": 0.21186301112174988, + "learning_rate": 1.6184138846545714e-05, + "loss": 1.0404, + "step": 12135 + }, + { + "epoch": 0.7373473479555258, + "grad_norm": 0.20262014865875244, + "learning_rate": 1.6177088061319902e-05, + "loss": 0.9998, + "step": 12136 + }, + { + "epoch": 0.7374081049881523, + "grad_norm": 0.14068935811519623, + "learning_rate": 1.617003851587946e-05, + "loss": 1.0396, + "step": 12137 + }, + { + "epoch": 0.7374688620207789, + "grad_norm": 0.1755463182926178, + "learning_rate": 1.61629902104828e-05, + "loss": 1.0763, + "step": 12138 + }, + { + "epoch": 0.7375296190534054, + "grad_norm": 0.2780778706073761, + "learning_rate": 1.615594314538827e-05, + "loss": 1.018, + "step": 12139 + }, + { + "epoch": 0.737590376086032, + "grad_norm": 7.305284023284912, + "learning_rate": 1.614889732085419e-05, + "loss": 1.1851, + "step": 12140 + }, + { + "epoch": 0.7376511331186585, + "grad_norm": 0.15769775211811066, + "learning_rate": 1.614185273713882e-05, + "loss": 0.9965, + "step": 12141 + }, + { + "epoch": 0.737711890151285, + "grad_norm": 0.1451137512922287, + "learning_rate": 1.613480939450038e-05, + "loss": 1.0706, + "step": 12142 + }, + { + "epoch": 0.7377726471839116, + "grad_norm": 0.15863953530788422, + "learning_rate": 1.6127767293197034e-05, + "loss": 1.0746, + "step": 12143 + }, + { + "epoch": 0.7378334042165381, + "grad_norm": 0.1679084748029709, + "learning_rate": 1.6120726433486953e-05, + "loss": 0.9851, + "step": 12144 + }, + { + "epoch": 0.7378941612491646, + "grad_norm": 0.19867177307605743, + "learning_rate": 1.6113686815628187e-05, + "loss": 1.0316, + "step": 12145 + }, + { + "epoch": 0.7379549182817912, + "grad_norm": 0.1528138667345047, + "learning_rate": 1.6106648439878776e-05, + "loss": 1.1034, + "step": 12146 + }, + { + "epoch": 0.7380156753144177, + "grad_norm": 0.1515488177537918, + "learning_rate": 1.6099611306496698e-05, + "loss": 1.0712, + "step": 12147 + }, + { + "epoch": 0.7380764323470441, + "grad_norm": 0.16822655498981476, + "learning_rate": 1.6092575415739943e-05, + "loss": 1.0301, + "step": 12148 + }, + { + "epoch": 0.7381371893796707, + "grad_norm": 0.14853422343730927, + "learning_rate": 1.608554076786638e-05, + "loss": 1.0065, + "step": 12149 + }, + { + "epoch": 0.7381979464122972, + "grad_norm": 2.5558056831359863, + "learning_rate": 1.6078507363133888e-05, + "loss": 1.1037, + "step": 12150 + }, + { + "epoch": 0.7382587034449237, + "grad_norm": 0.1307159662246704, + "learning_rate": 1.6071475201800272e-05, + "loss": 1.0103, + "step": 12151 + }, + { + "epoch": 0.7383194604775503, + "grad_norm": 0.5363821983337402, + "learning_rate": 1.6064444284123265e-05, + "loss": 1.0812, + "step": 12152 + }, + { + "epoch": 0.7383802175101768, + "grad_norm": 0.18014459311962128, + "learning_rate": 1.6057414610360627e-05, + "loss": 1.0229, + "step": 12153 + }, + { + "epoch": 0.7384409745428033, + "grad_norm": 0.1398610770702362, + "learning_rate": 1.6050386180770017e-05, + "loss": 1.0451, + "step": 12154 + }, + { + "epoch": 0.7385017315754299, + "grad_norm": 0.36795589327812195, + "learning_rate": 1.604335899560906e-05, + "loss": 1.0674, + "step": 12155 + }, + { + "epoch": 0.7385624886080564, + "grad_norm": 0.18371464312076569, + "learning_rate": 1.6036333055135344e-05, + "loss": 1.0292, + "step": 12156 + }, + { + "epoch": 0.7386232456406829, + "grad_norm": 0.15505965054035187, + "learning_rate": 1.602930835960641e-05, + "loss": 0.9417, + "step": 12157 + }, + { + "epoch": 0.7386840026733095, + "grad_norm": 0.1787785142660141, + "learning_rate": 1.602228490927974e-05, + "loss": 1.0618, + "step": 12158 + }, + { + "epoch": 0.738744759705936, + "grad_norm": 0.19735237956047058, + "learning_rate": 1.6015262704412787e-05, + "loss": 1.1298, + "step": 12159 + }, + { + "epoch": 0.7388055167385625, + "grad_norm": 0.17755354940891266, + "learning_rate": 1.6008241745262947e-05, + "loss": 1.154, + "step": 12160 + }, + { + "epoch": 0.738866273771189, + "grad_norm": 0.1529655009508133, + "learning_rate": 1.6001222032087575e-05, + "loss": 1.0649, + "step": 12161 + }, + { + "epoch": 0.7389270308038155, + "grad_norm": 0.25719979405403137, + "learning_rate": 1.5994203565143982e-05, + "loss": 1.0902, + "step": 12162 + }, + { + "epoch": 0.738987787836442, + "grad_norm": 0.13515958189964294, + "learning_rate": 1.598718634468942e-05, + "loss": 1.0549, + "step": 12163 + }, + { + "epoch": 0.7390485448690686, + "grad_norm": 0.16471487283706665, + "learning_rate": 1.598017037098114e-05, + "loss": 1.0648, + "step": 12164 + }, + { + "epoch": 0.7391093019016951, + "grad_norm": 0.14137525856494904, + "learning_rate": 1.5973155644276284e-05, + "loss": 1.0381, + "step": 12165 + }, + { + "epoch": 0.7391700589343216, + "grad_norm": 0.17968054115772247, + "learning_rate": 1.5966142164831976e-05, + "loss": 1.0825, + "step": 12166 + }, + { + "epoch": 0.7392308159669482, + "grad_norm": 0.19986937940120697, + "learning_rate": 1.5959129932905314e-05, + "loss": 1.1131, + "step": 12167 + }, + { + "epoch": 0.7392915729995747, + "grad_norm": 0.2529922127723694, + "learning_rate": 1.5952118948753302e-05, + "loss": 1.0874, + "step": 12168 + }, + { + "epoch": 0.7393523300322012, + "grad_norm": 0.16164061427116394, + "learning_rate": 1.594510921263297e-05, + "loss": 1.0671, + "step": 12169 + }, + { + "epoch": 0.7394130870648278, + "grad_norm": 0.19416692852973938, + "learning_rate": 1.593810072480124e-05, + "loss": 1.0499, + "step": 12170 + }, + { + "epoch": 0.7394738440974543, + "grad_norm": 3.7689201831817627, + "learning_rate": 1.593109348551503e-05, + "loss": 1.0785, + "step": 12171 + }, + { + "epoch": 0.7395346011300808, + "grad_norm": 0.27557870745658875, + "learning_rate": 1.592408749503116e-05, + "loss": 1.011, + "step": 12172 + }, + { + "epoch": 0.7395953581627074, + "grad_norm": 0.21448230743408203, + "learning_rate": 1.5917082753606434e-05, + "loss": 1.0125, + "step": 12173 + }, + { + "epoch": 0.7396561151953338, + "grad_norm": 0.19736918807029724, + "learning_rate": 1.591007926149764e-05, + "loss": 1.0277, + "step": 12174 + }, + { + "epoch": 0.7397168722279603, + "grad_norm": 0.15478961169719696, + "learning_rate": 1.5903077018961477e-05, + "loss": 1.014, + "step": 12175 + }, + { + "epoch": 0.7397776292605869, + "grad_norm": 0.17717431485652924, + "learning_rate": 1.5896076026254626e-05, + "loss": 0.9961, + "step": 12176 + }, + { + "epoch": 0.7398383862932134, + "grad_norm": 0.5351703763008118, + "learning_rate": 1.5889076283633707e-05, + "loss": 1.1098, + "step": 12177 + }, + { + "epoch": 0.73989914332584, + "grad_norm": 0.30535200238227844, + "learning_rate": 1.5882077791355265e-05, + "loss": 1.1526, + "step": 12178 + }, + { + "epoch": 0.7399599003584665, + "grad_norm": 0.23752780258655548, + "learning_rate": 1.587508054967587e-05, + "loss": 1.0547, + "step": 12179 + }, + { + "epoch": 0.740020657391093, + "grad_norm": 0.4709758460521698, + "learning_rate": 1.5868084558851994e-05, + "loss": 1.1721, + "step": 12180 + }, + { + "epoch": 0.7400814144237196, + "grad_norm": 0.19341643154621124, + "learning_rate": 1.5861089819140074e-05, + "loss": 1.0268, + "step": 12181 + }, + { + "epoch": 0.7401421714563461, + "grad_norm": 0.2940843105316162, + "learning_rate": 1.5854096330796507e-05, + "loss": 1.084, + "step": 12182 + }, + { + "epoch": 0.7402029284889726, + "grad_norm": 0.23170220851898193, + "learning_rate": 1.5847104094077636e-05, + "loss": 1.0548, + "step": 12183 + }, + { + "epoch": 0.7402636855215992, + "grad_norm": 0.30767157673835754, + "learning_rate": 1.5840113109239772e-05, + "loss": 1.1224, + "step": 12184 + }, + { + "epoch": 0.7403244425542257, + "grad_norm": 0.13287043571472168, + "learning_rate": 1.583312337653916e-05, + "loss": 0.9962, + "step": 12185 + }, + { + "epoch": 0.7403851995868522, + "grad_norm": 0.24579693377017975, + "learning_rate": 1.582613489623202e-05, + "loss": 1.0275, + "step": 12186 + }, + { + "epoch": 0.7404459566194787, + "grad_norm": 0.2078135758638382, + "learning_rate": 1.5819147668574507e-05, + "loss": 1.1267, + "step": 12187 + }, + { + "epoch": 0.7405067136521052, + "grad_norm": 0.16390381753444672, + "learning_rate": 1.5812161693822747e-05, + "loss": 1.0798, + "step": 12188 + }, + { + "epoch": 0.7405674706847317, + "grad_norm": 0.31575071811676025, + "learning_rate": 1.580517697223279e-05, + "loss": 1.1279, + "step": 12189 + }, + { + "epoch": 0.7406282277173583, + "grad_norm": 5.849077224731445, + "learning_rate": 1.5798193504060692e-05, + "loss": 1.0477, + "step": 12190 + }, + { + "epoch": 0.7406889847499848, + "grad_norm": 0.15946334600448608, + "learning_rate": 1.5791211289562447e-05, + "loss": 1.0247, + "step": 12191 + }, + { + "epoch": 0.7407497417826113, + "grad_norm": 0.5073317885398865, + "learning_rate": 1.5784230328993942e-05, + "loss": 1.1454, + "step": 12192 + }, + { + "epoch": 0.7408104988152379, + "grad_norm": 0.17353080213069916, + "learning_rate": 1.5777250622611096e-05, + "loss": 1.0467, + "step": 12193 + }, + { + "epoch": 0.7408712558478644, + "grad_norm": 0.2100648432970047, + "learning_rate": 1.577027217066972e-05, + "loss": 1.1676, + "step": 12194 + }, + { + "epoch": 0.7409320128804909, + "grad_norm": 0.2722139060497284, + "learning_rate": 1.576329497342565e-05, + "loss": 1.0308, + "step": 12195 + }, + { + "epoch": 0.7409927699131175, + "grad_norm": 0.6022172570228577, + "learning_rate": 1.5756319031134626e-05, + "loss": 1.0453, + "step": 12196 + }, + { + "epoch": 0.741053526945744, + "grad_norm": 0.14008285105228424, + "learning_rate": 1.574934434405234e-05, + "loss": 0.9995, + "step": 12197 + }, + { + "epoch": 0.7411142839783705, + "grad_norm": 0.4958098232746124, + "learning_rate": 1.574237091243448e-05, + "loss": 1.2204, + "step": 12198 + }, + { + "epoch": 0.7411750410109971, + "grad_norm": 0.1818794459104538, + "learning_rate": 1.57353987365366e-05, + "loss": 1.0686, + "step": 12199 + }, + { + "epoch": 0.7412357980436235, + "grad_norm": 10.184050559997559, + "learning_rate": 1.572842781661432e-05, + "loss": 1.1654, + "step": 12200 + }, + { + "epoch": 0.74129655507625, + "grad_norm": 0.2594745457172394, + "learning_rate": 1.5721458152923142e-05, + "loss": 1.0455, + "step": 12201 + }, + { + "epoch": 0.7413573121088766, + "grad_norm": 0.18343879282474518, + "learning_rate": 1.571448974571854e-05, + "loss": 1.0065, + "step": 12202 + }, + { + "epoch": 0.7414180691415031, + "grad_norm": 0.16876137256622314, + "learning_rate": 1.5707522595255953e-05, + "loss": 1.091, + "step": 12203 + }, + { + "epoch": 0.7414788261741296, + "grad_norm": 0.21117544174194336, + "learning_rate": 1.570055670179074e-05, + "loss": 1.1016, + "step": 12204 + }, + { + "epoch": 0.7415395832067562, + "grad_norm": 0.21855662763118744, + "learning_rate": 1.569359206557826e-05, + "loss": 1.1726, + "step": 12205 + }, + { + "epoch": 0.7416003402393827, + "grad_norm": 0.23321466147899628, + "learning_rate": 1.5686628686873796e-05, + "loss": 1.067, + "step": 12206 + }, + { + "epoch": 0.7416610972720092, + "grad_norm": 0.19129067659378052, + "learning_rate": 1.5679666565932584e-05, + "loss": 0.9956, + "step": 12207 + }, + { + "epoch": 0.7417218543046358, + "grad_norm": 0.19235219061374664, + "learning_rate": 1.5672705703009828e-05, + "loss": 1.0223, + "step": 12208 + }, + { + "epoch": 0.7417826113372623, + "grad_norm": 0.2211027294397354, + "learning_rate": 1.5665746098360685e-05, + "loss": 1.1023, + "step": 12209 + }, + { + "epoch": 0.7418433683698888, + "grad_norm": 0.32120731472969055, + "learning_rate": 1.5658787752240234e-05, + "loss": 1.1182, + "step": 12210 + }, + { + "epoch": 0.7419041254025154, + "grad_norm": 0.4269195795059204, + "learning_rate": 1.5651830664903588e-05, + "loss": 1.0232, + "step": 12211 + }, + { + "epoch": 0.7419648824351419, + "grad_norm": 0.21186895668506622, + "learning_rate": 1.5644874836605716e-05, + "loss": 1.0773, + "step": 12212 + }, + { + "epoch": 0.7420256394677683, + "grad_norm": 0.29658156633377075, + "learning_rate": 1.5637920267601597e-05, + "loss": 1.1244, + "step": 12213 + }, + { + "epoch": 0.7420863965003949, + "grad_norm": 0.137900710105896, + "learning_rate": 1.5630966958146154e-05, + "loss": 1.0204, + "step": 12214 + }, + { + "epoch": 0.7421471535330214, + "grad_norm": 0.1505892127752304, + "learning_rate": 1.562401490849424e-05, + "loss": 1.0496, + "step": 12215 + }, + { + "epoch": 0.7422079105656479, + "grad_norm": 0.20965510606765747, + "learning_rate": 1.561706411890072e-05, + "loss": 1.1094, + "step": 12216 + }, + { + "epoch": 0.7422686675982745, + "grad_norm": 0.1287902593612671, + "learning_rate": 1.5610114589620366e-05, + "loss": 0.9976, + "step": 12217 + }, + { + "epoch": 0.742329424630901, + "grad_norm": 0.34176474809646606, + "learning_rate": 1.560316632090792e-05, + "loss": 1.2779, + "step": 12218 + }, + { + "epoch": 0.7423901816635275, + "grad_norm": 0.8237595558166504, + "learning_rate": 1.559621931301805e-05, + "loss": 1.2582, + "step": 12219 + }, + { + "epoch": 0.7424509386961541, + "grad_norm": 0.2588569223880768, + "learning_rate": 1.5589273566205394e-05, + "loss": 1.1221, + "step": 12220 + }, + { + "epoch": 0.7425116957287806, + "grad_norm": 0.24077297747135162, + "learning_rate": 1.558232908072458e-05, + "loss": 1.0018, + "step": 12221 + }, + { + "epoch": 0.7425724527614072, + "grad_norm": 0.17504651844501495, + "learning_rate": 1.557538585683015e-05, + "loss": 1.0586, + "step": 12222 + }, + { + "epoch": 0.7426332097940337, + "grad_norm": 0.12876471877098083, + "learning_rate": 1.5568443894776602e-05, + "loss": 1.0215, + "step": 12223 + }, + { + "epoch": 0.7426939668266602, + "grad_norm": 0.12596672773361206, + "learning_rate": 1.55615031948184e-05, + "loss": 1.0659, + "step": 12224 + }, + { + "epoch": 0.7427547238592868, + "grad_norm": 0.14167657494544983, + "learning_rate": 1.5554563757209956e-05, + "loss": 1.0561, + "step": 12225 + }, + { + "epoch": 0.7428154808919133, + "grad_norm": 0.1953197419643402, + "learning_rate": 1.554762558220563e-05, + "loss": 1.0653, + "step": 12226 + }, + { + "epoch": 0.7428762379245397, + "grad_norm": 0.1634196639060974, + "learning_rate": 1.5540688670059754e-05, + "loss": 1.0507, + "step": 12227 + }, + { + "epoch": 0.7429369949571663, + "grad_norm": 0.14514052867889404, + "learning_rate": 1.55337530210266e-05, + "loss": 1.077, + "step": 12228 + }, + { + "epoch": 0.7429977519897928, + "grad_norm": 0.20578783750534058, + "learning_rate": 1.552681863536038e-05, + "loss": 1.0528, + "step": 12229 + }, + { + "epoch": 0.7430585090224193, + "grad_norm": 0.2471075803041458, + "learning_rate": 1.551988551331529e-05, + "loss": 1.0799, + "step": 12230 + }, + { + "epoch": 0.7431192660550459, + "grad_norm": 0.19242486357688904, + "learning_rate": 1.5512953655145467e-05, + "loss": 1.0414, + "step": 12231 + }, + { + "epoch": 0.7431800230876724, + "grad_norm": 0.202849343419075, + "learning_rate": 1.550602306110499e-05, + "loss": 1.0479, + "step": 12232 + }, + { + "epoch": 0.7432407801202989, + "grad_norm": 0.1721251904964447, + "learning_rate": 1.5499093731447905e-05, + "loss": 1.0092, + "step": 12233 + }, + { + "epoch": 0.7433015371529255, + "grad_norm": 0.2746470272541046, + "learning_rate": 1.549216566642821e-05, + "loss": 1.0006, + "step": 12234 + }, + { + "epoch": 0.743362294185552, + "grad_norm": 0.2515316605567932, + "learning_rate": 1.548523886629985e-05, + "loss": 1.124, + "step": 12235 + }, + { + "epoch": 0.7434230512181785, + "grad_norm": 0.2106676548719406, + "learning_rate": 1.547831333131672e-05, + "loss": 1.0863, + "step": 12236 + }, + { + "epoch": 0.7434838082508051, + "grad_norm": 0.5567908883094788, + "learning_rate": 1.54713890617327e-05, + "loss": 1.0123, + "step": 12237 + }, + { + "epoch": 0.7435445652834316, + "grad_norm": 0.2612905204296112, + "learning_rate": 1.54644660578016e-05, + "loss": 1.0803, + "step": 12238 + }, + { + "epoch": 0.7436053223160581, + "grad_norm": 0.235815167427063, + "learning_rate": 1.5457544319777163e-05, + "loss": 1.1488, + "step": 12239 + }, + { + "epoch": 0.7436660793486846, + "grad_norm": 0.15918777883052826, + "learning_rate": 1.54506238479131e-05, + "loss": 1.0201, + "step": 12240 + }, + { + "epoch": 0.7437268363813111, + "grad_norm": 0.1546841263771057, + "learning_rate": 1.544370464246311e-05, + "loss": 1.0125, + "step": 12241 + }, + { + "epoch": 0.7437875934139376, + "grad_norm": 0.16376914083957672, + "learning_rate": 1.5436786703680806e-05, + "loss": 1.0385, + "step": 12242 + }, + { + "epoch": 0.7438483504465642, + "grad_norm": 1.0175727605819702, + "learning_rate": 1.5429870031819766e-05, + "loss": 1.097, + "step": 12243 + }, + { + "epoch": 0.7439091074791907, + "grad_norm": 0.20679105818271637, + "learning_rate": 1.542295462713354e-05, + "loss": 1.0734, + "step": 12244 + }, + { + "epoch": 0.7439698645118172, + "grad_norm": 0.34706827998161316, + "learning_rate": 1.5416040489875565e-05, + "loss": 1.0282, + "step": 12245 + }, + { + "epoch": 0.7440306215444438, + "grad_norm": 0.13362546265125275, + "learning_rate": 1.5409127620299326e-05, + "loss": 0.9661, + "step": 12246 + }, + { + "epoch": 0.7440913785770703, + "grad_norm": 0.2383270561695099, + "learning_rate": 1.54022160186582e-05, + "loss": 1.0415, + "step": 12247 + }, + { + "epoch": 0.7441521356096968, + "grad_norm": 0.24554871022701263, + "learning_rate": 1.539530568520553e-05, + "loss": 1.0332, + "step": 12248 + }, + { + "epoch": 0.7442128926423234, + "grad_norm": 0.19278277456760406, + "learning_rate": 1.5388396620194623e-05, + "loss": 1.1077, + "step": 12249 + }, + { + "epoch": 0.7442736496749499, + "grad_norm": 0.19051454961299896, + "learning_rate": 1.5381488823878725e-05, + "loss": 1.0476, + "step": 12250 + }, + { + "epoch": 0.7443344067075764, + "grad_norm": 0.9674873352050781, + "learning_rate": 1.5374582296511053e-05, + "loss": 1.1084, + "step": 12251 + }, + { + "epoch": 0.744395163740203, + "grad_norm": 0.15226489305496216, + "learning_rate": 1.5367677038344754e-05, + "loss": 1.0067, + "step": 12252 + }, + { + "epoch": 0.7444559207728294, + "grad_norm": 0.1480061411857605, + "learning_rate": 1.536077304963295e-05, + "loss": 1.0296, + "step": 12253 + }, + { + "epoch": 0.7445166778054559, + "grad_norm": 0.14703676104545593, + "learning_rate": 1.53538703306287e-05, + "loss": 0.9824, + "step": 12254 + }, + { + "epoch": 0.7445774348380825, + "grad_norm": 0.5437590479850769, + "learning_rate": 1.5346968881585034e-05, + "loss": 1.2147, + "step": 12255 + }, + { + "epoch": 0.744638191870709, + "grad_norm": 0.20612148940563202, + "learning_rate": 1.5340068702754906e-05, + "loss": 1.0621, + "step": 12256 + }, + { + "epoch": 0.7446989489033355, + "grad_norm": 0.25368958711624146, + "learning_rate": 1.533316979439128e-05, + "loss": 1.0408, + "step": 12257 + }, + { + "epoch": 0.7447597059359621, + "grad_norm": 0.15697847306728363, + "learning_rate": 1.5326272156747024e-05, + "loss": 1.0156, + "step": 12258 + }, + { + "epoch": 0.7448204629685886, + "grad_norm": 0.12871995568275452, + "learning_rate": 1.531937579007495e-05, + "loss": 1.0318, + "step": 12259 + }, + { + "epoch": 0.7448812200012151, + "grad_norm": 0.13674461841583252, + "learning_rate": 1.5312480694627862e-05, + "loss": 1.0183, + "step": 12260 + }, + { + "epoch": 0.7449419770338417, + "grad_norm": 0.1678137332201004, + "learning_rate": 1.5305586870658485e-05, + "loss": 1.0912, + "step": 12261 + }, + { + "epoch": 0.7450027340664682, + "grad_norm": 0.15710866451263428, + "learning_rate": 1.5298694318419536e-05, + "loss": 1.0398, + "step": 12262 + }, + { + "epoch": 0.7450634910990948, + "grad_norm": 0.16869790852069855, + "learning_rate": 1.5291803038163665e-05, + "loss": 1.035, + "step": 12263 + }, + { + "epoch": 0.7451242481317213, + "grad_norm": 0.13926680386066437, + "learning_rate": 1.5284913030143472e-05, + "loss": 1.0175, + "step": 12264 + }, + { + "epoch": 0.7451850051643478, + "grad_norm": 0.38392913341522217, + "learning_rate": 1.5278024294611487e-05, + "loss": 1.0453, + "step": 12265 + }, + { + "epoch": 0.7452457621969742, + "grad_norm": 0.1388578712940216, + "learning_rate": 1.5271136831820214e-05, + "loss": 1.0305, + "step": 12266 + }, + { + "epoch": 0.7453065192296008, + "grad_norm": 0.22907735407352448, + "learning_rate": 1.5264250642022153e-05, + "loss": 1.0016, + "step": 12267 + }, + { + "epoch": 0.7453672762622273, + "grad_norm": 0.21486346423625946, + "learning_rate": 1.5257365725469703e-05, + "loss": 1.1617, + "step": 12268 + }, + { + "epoch": 0.7454280332948539, + "grad_norm": 0.25346431136131287, + "learning_rate": 1.5250482082415223e-05, + "loss": 1.2133, + "step": 12269 + }, + { + "epoch": 0.7454887903274804, + "grad_norm": 0.9438855648040771, + "learning_rate": 1.524359971311104e-05, + "loss": 1.1637, + "step": 12270 + }, + { + "epoch": 0.7455495473601069, + "grad_norm": 0.20238009095191956, + "learning_rate": 1.5236718617809425e-05, + "loss": 1.1451, + "step": 12271 + }, + { + "epoch": 0.7456103043927335, + "grad_norm": 5.505398273468018, + "learning_rate": 1.5229838796762613e-05, + "loss": 1.1153, + "step": 12272 + }, + { + "epoch": 0.74567106142536, + "grad_norm": 0.14844673871994019, + "learning_rate": 1.522296025022278e-05, + "loss": 1.0177, + "step": 12273 + }, + { + "epoch": 0.7457318184579865, + "grad_norm": 0.2235460728406906, + "learning_rate": 1.521608297844206e-05, + "loss": 1.0879, + "step": 12274 + }, + { + "epoch": 0.7457925754906131, + "grad_norm": 0.13267450034618378, + "learning_rate": 1.520920698167254e-05, + "loss": 1.0715, + "step": 12275 + }, + { + "epoch": 0.7458533325232396, + "grad_norm": 0.18545234203338623, + "learning_rate": 1.5202332260166269e-05, + "loss": 1.1111, + "step": 12276 + }, + { + "epoch": 0.7459140895558661, + "grad_norm": 0.16761964559555054, + "learning_rate": 1.5195458814175217e-05, + "loss": 1.0835, + "step": 12277 + }, + { + "epoch": 0.7459748465884927, + "grad_norm": 0.26743146777153015, + "learning_rate": 1.518858664395138e-05, + "loss": 1.0647, + "step": 12278 + }, + { + "epoch": 0.7460356036211191, + "grad_norm": 0.13091523945331573, + "learning_rate": 1.518171574974661e-05, + "loss": 1.0053, + "step": 12279 + }, + { + "epoch": 0.7460963606537456, + "grad_norm": 0.2550339102745056, + "learning_rate": 1.5174846131812787e-05, + "loss": 1.0744, + "step": 12280 + }, + { + "epoch": 0.7461571176863722, + "grad_norm": 0.19582781195640564, + "learning_rate": 1.5167977790401705e-05, + "loss": 0.9705, + "step": 12281 + }, + { + "epoch": 0.7462178747189987, + "grad_norm": 3.913116693496704, + "learning_rate": 1.5161110725765116e-05, + "loss": 1.0084, + "step": 12282 + }, + { + "epoch": 0.7462786317516252, + "grad_norm": 0.25027304887771606, + "learning_rate": 1.515424493815476e-05, + "loss": 0.9917, + "step": 12283 + }, + { + "epoch": 0.7463393887842518, + "grad_norm": 0.16813236474990845, + "learning_rate": 1.5147380427822295e-05, + "loss": 1.0695, + "step": 12284 + }, + { + "epoch": 0.7464001458168783, + "grad_norm": 0.1373838186264038, + "learning_rate": 1.514051719501935e-05, + "loss": 1.0583, + "step": 12285 + }, + { + "epoch": 0.7464609028495048, + "grad_norm": 0.255839467048645, + "learning_rate": 1.5133655239997474e-05, + "loss": 1.2117, + "step": 12286 + }, + { + "epoch": 0.7465216598821314, + "grad_norm": 0.18116353452205658, + "learning_rate": 1.5126794563008184e-05, + "loss": 1.1223, + "step": 12287 + }, + { + "epoch": 0.7465824169147579, + "grad_norm": 0.1951504349708557, + "learning_rate": 1.5119935164303e-05, + "loss": 1.0484, + "step": 12288 + }, + { + "epoch": 0.7466431739473844, + "grad_norm": 0.12804356217384338, + "learning_rate": 1.5113077044133334e-05, + "loss": 1.0293, + "step": 12289 + }, + { + "epoch": 0.746703930980011, + "grad_norm": 0.2731664478778839, + "learning_rate": 1.510622020275057e-05, + "loss": 1.1645, + "step": 12290 + }, + { + "epoch": 0.7467646880126375, + "grad_norm": 0.18990570306777954, + "learning_rate": 1.5099364640406066e-05, + "loss": 1.0534, + "step": 12291 + }, + { + "epoch": 0.7468254450452639, + "grad_norm": 0.38706693053245544, + "learning_rate": 1.509251035735107e-05, + "loss": 1.2173, + "step": 12292 + }, + { + "epoch": 0.7468862020778905, + "grad_norm": 0.1403743177652359, + "learning_rate": 1.5085657353836863e-05, + "loss": 1.0624, + "step": 12293 + }, + { + "epoch": 0.746946959110517, + "grad_norm": 0.28369081020355225, + "learning_rate": 1.5078805630114639e-05, + "loss": 1.0966, + "step": 12294 + }, + { + "epoch": 0.7470077161431435, + "grad_norm": 0.5037837028503418, + "learning_rate": 1.5071955186435544e-05, + "loss": 1.0562, + "step": 12295 + }, + { + "epoch": 0.7470684731757701, + "grad_norm": 19.89148712158203, + "learning_rate": 1.506510602305068e-05, + "loss": 1.0537, + "step": 12296 + }, + { + "epoch": 0.7471292302083966, + "grad_norm": 2.0159053802490234, + "learning_rate": 1.5058258140211112e-05, + "loss": 1.116, + "step": 12297 + }, + { + "epoch": 0.7471899872410231, + "grad_norm": 0.2304600030183792, + "learning_rate": 1.5051411538167843e-05, + "loss": 1.1843, + "step": 12298 + }, + { + "epoch": 0.7472507442736497, + "grad_norm": 0.17824512720108032, + "learning_rate": 1.504456621717184e-05, + "loss": 0.9964, + "step": 12299 + }, + { + "epoch": 0.7473115013062762, + "grad_norm": 2.6903367042541504, + "learning_rate": 1.5037722177474023e-05, + "loss": 1.0271, + "step": 12300 + }, + { + "epoch": 0.7473722583389028, + "grad_norm": 0.22106531262397766, + "learning_rate": 1.503087941932525e-05, + "loss": 1.0254, + "step": 12301 + }, + { + "epoch": 0.7474330153715293, + "grad_norm": 0.2453022301197052, + "learning_rate": 1.502403794297636e-05, + "loss": 1.0016, + "step": 12302 + }, + { + "epoch": 0.7474937724041558, + "grad_norm": 0.24402932822704315, + "learning_rate": 1.5017197748678096e-05, + "loss": 1.0559, + "step": 12303 + }, + { + "epoch": 0.7475545294367824, + "grad_norm": 0.23583471775054932, + "learning_rate": 1.5010358836681232e-05, + "loss": 1.0738, + "step": 12304 + }, + { + "epoch": 0.7476152864694088, + "grad_norm": 0.18598178029060364, + "learning_rate": 1.5003521207236443e-05, + "loss": 1.0477, + "step": 12305 + }, + { + "epoch": 0.7476760435020353, + "grad_norm": 0.3215706944465637, + "learning_rate": 1.4996684860594329e-05, + "loss": 1.0716, + "step": 12306 + }, + { + "epoch": 0.7477368005346618, + "grad_norm": 0.2177494764328003, + "learning_rate": 1.49898497970055e-05, + "loss": 1.0777, + "step": 12307 + }, + { + "epoch": 0.7477975575672884, + "grad_norm": 0.9029622673988342, + "learning_rate": 1.498301601672048e-05, + "loss": 1.0692, + "step": 12308 + }, + { + "epoch": 0.7478583145999149, + "grad_norm": 0.2609107494354248, + "learning_rate": 1.497618351998979e-05, + "loss": 1.0517, + "step": 12309 + }, + { + "epoch": 0.7479190716325415, + "grad_norm": 0.4564414620399475, + "learning_rate": 1.4969352307063866e-05, + "loss": 1.0879, + "step": 12310 + }, + { + "epoch": 0.747979828665168, + "grad_norm": 0.5797598958015442, + "learning_rate": 1.4962522378193118e-05, + "loss": 1.132, + "step": 12311 + }, + { + "epoch": 0.7480405856977945, + "grad_norm": 0.1990519016981125, + "learning_rate": 1.4955693733627867e-05, + "loss": 1.0022, + "step": 12312 + }, + { + "epoch": 0.7481013427304211, + "grad_norm": 0.13997657597064972, + "learning_rate": 1.4948866373618426e-05, + "loss": 1.0538, + "step": 12313 + }, + { + "epoch": 0.7481620997630476, + "grad_norm": 0.15543818473815918, + "learning_rate": 1.4942040298415072e-05, + "loss": 0.9834, + "step": 12314 + }, + { + "epoch": 0.7482228567956741, + "grad_norm": 0.3180289566516876, + "learning_rate": 1.4935215508268014e-05, + "loss": 1.2051, + "step": 12315 + }, + { + "epoch": 0.7482836138283007, + "grad_norm": 0.24254123866558075, + "learning_rate": 1.4928392003427404e-05, + "loss": 1.0292, + "step": 12316 + }, + { + "epoch": 0.7483443708609272, + "grad_norm": 0.28811800479888916, + "learning_rate": 1.4921569784143368e-05, + "loss": 1.0298, + "step": 12317 + }, + { + "epoch": 0.7484051278935536, + "grad_norm": 0.15601962804794312, + "learning_rate": 1.491474885066597e-05, + "loss": 1.0638, + "step": 12318 + }, + { + "epoch": 0.7484658849261802, + "grad_norm": 0.21351198852062225, + "learning_rate": 1.4907929203245235e-05, + "loss": 1.0937, + "step": 12319 + }, + { + "epoch": 0.7485266419588067, + "grad_norm": 1.4613151550292969, + "learning_rate": 1.4901110842131138e-05, + "loss": 1.0156, + "step": 12320 + }, + { + "epoch": 0.7485873989914332, + "grad_norm": 0.3001377284526825, + "learning_rate": 1.4894293767573613e-05, + "loss": 1.0785, + "step": 12321 + }, + { + "epoch": 0.7486481560240598, + "grad_norm": 0.14506889879703522, + "learning_rate": 1.488747797982253e-05, + "loss": 1.088, + "step": 12322 + }, + { + "epoch": 0.7487089130566863, + "grad_norm": 0.13142932951450348, + "learning_rate": 1.4880663479127727e-05, + "loss": 1.0269, + "step": 12323 + }, + { + "epoch": 0.7487696700893128, + "grad_norm": 0.23391607403755188, + "learning_rate": 1.487385026573898e-05, + "loss": 1.1165, + "step": 12324 + }, + { + "epoch": 0.7488304271219394, + "grad_norm": 0.29040634632110596, + "learning_rate": 1.4867038339906076e-05, + "loss": 1.0916, + "step": 12325 + }, + { + "epoch": 0.7488911841545659, + "grad_norm": 0.1516198068857193, + "learning_rate": 1.4860227701878654e-05, + "loss": 1.0486, + "step": 12326 + }, + { + "epoch": 0.7489519411871924, + "grad_norm": 0.18298299610614777, + "learning_rate": 1.4853418351906374e-05, + "loss": 1.0405, + "step": 12327 + }, + { + "epoch": 0.749012698219819, + "grad_norm": 0.1442565619945526, + "learning_rate": 1.4846610290238843e-05, + "loss": 1.0387, + "step": 12328 + }, + { + "epoch": 0.7490734552524455, + "grad_norm": 6.812145709991455, + "learning_rate": 1.483980351712559e-05, + "loss": 1.069, + "step": 12329 + }, + { + "epoch": 0.749134212285072, + "grad_norm": 0.24135959148406982, + "learning_rate": 1.4832998032816148e-05, + "loss": 1.1652, + "step": 12330 + }, + { + "epoch": 0.7491949693176986, + "grad_norm": 0.28880661725997925, + "learning_rate": 1.4826193837559981e-05, + "loss": 1.2853, + "step": 12331 + }, + { + "epoch": 0.749255726350325, + "grad_norm": 0.24991200864315033, + "learning_rate": 1.4819390931606458e-05, + "loss": 1.051, + "step": 12332 + }, + { + "epoch": 0.7493164833829515, + "grad_norm": 0.27602991461753845, + "learning_rate": 1.481258931520495e-05, + "loss": 1.1029, + "step": 12333 + }, + { + "epoch": 0.7493772404155781, + "grad_norm": 0.14343565702438354, + "learning_rate": 1.4805788988604791e-05, + "loss": 1.053, + "step": 12334 + }, + { + "epoch": 0.7494379974482046, + "grad_norm": 0.1919344961643219, + "learning_rate": 1.4798989952055248e-05, + "loss": 1.0802, + "step": 12335 + }, + { + "epoch": 0.7494987544808311, + "grad_norm": 0.17783115804195404, + "learning_rate": 1.4792192205805527e-05, + "loss": 0.9741, + "step": 12336 + }, + { + "epoch": 0.7495595115134577, + "grad_norm": 0.1460333913564682, + "learning_rate": 1.4785395750104808e-05, + "loss": 0.9945, + "step": 12337 + }, + { + "epoch": 0.7496202685460842, + "grad_norm": 0.17845812439918518, + "learning_rate": 1.4778600585202213e-05, + "loss": 1.0083, + "step": 12338 + }, + { + "epoch": 0.7496810255787107, + "grad_norm": 0.1490906924009323, + "learning_rate": 1.477180671134682e-05, + "loss": 1.0631, + "step": 12339 + }, + { + "epoch": 0.7497417826113373, + "grad_norm": 5.508548736572266, + "learning_rate": 1.4765014128787663e-05, + "loss": 1.1918, + "step": 12340 + }, + { + "epoch": 0.7498025396439638, + "grad_norm": 0.2161664217710495, + "learning_rate": 1.4758222837773727e-05, + "loss": 1.0659, + "step": 12341 + }, + { + "epoch": 0.7498632966765904, + "grad_norm": 0.5580757260322571, + "learning_rate": 1.475143283855394e-05, + "loss": 1.1433, + "step": 12342 + }, + { + "epoch": 0.7499240537092169, + "grad_norm": 0.15779034793376923, + "learning_rate": 1.47446441313772e-05, + "loss": 1.0161, + "step": 12343 + }, + { + "epoch": 0.7499848107418434, + "grad_norm": 0.3191524147987366, + "learning_rate": 1.4737856716492344e-05, + "loss": 1.0842, + "step": 12344 + }, + { + "epoch": 0.7500455677744698, + "grad_norm": 0.16471239924430847, + "learning_rate": 1.4731070594148166e-05, + "loss": 1.0808, + "step": 12345 + }, + { + "epoch": 0.7501063248070964, + "grad_norm": 0.24525564908981323, + "learning_rate": 1.472428576459341e-05, + "loss": 1.1787, + "step": 12346 + }, + { + "epoch": 0.7501670818397229, + "grad_norm": 0.26087355613708496, + "learning_rate": 1.4717502228076779e-05, + "loss": 1.1006, + "step": 12347 + }, + { + "epoch": 0.7502278388723495, + "grad_norm": 0.19485615193843842, + "learning_rate": 1.4710719984846927e-05, + "loss": 1.1066, + "step": 12348 + }, + { + "epoch": 0.750288595904976, + "grad_norm": 1.0824217796325684, + "learning_rate": 1.4703939035152436e-05, + "loss": 1.0456, + "step": 12349 + }, + { + "epoch": 0.7503493529376025, + "grad_norm": 0.15943478047847748, + "learning_rate": 1.4697159379241899e-05, + "loss": 1.0436, + "step": 12350 + }, + { + "epoch": 0.7504101099702291, + "grad_norm": 0.17657478153705597, + "learning_rate": 1.4690381017363808e-05, + "loss": 1.0812, + "step": 12351 + }, + { + "epoch": 0.7504708670028556, + "grad_norm": 0.20391564071178436, + "learning_rate": 1.4683603949766639e-05, + "loss": 1.0561, + "step": 12352 + }, + { + "epoch": 0.7505316240354821, + "grad_norm": 4.867756366729736, + "learning_rate": 1.4676828176698782e-05, + "loss": 1.0601, + "step": 12353 + }, + { + "epoch": 0.7505923810681087, + "grad_norm": 0.21434496343135834, + "learning_rate": 1.4670053698408598e-05, + "loss": 1.0839, + "step": 12354 + }, + { + "epoch": 0.7506531381007352, + "grad_norm": 0.23476846516132355, + "learning_rate": 1.4663280515144445e-05, + "loss": 1.2399, + "step": 12355 + }, + { + "epoch": 0.7507138951333617, + "grad_norm": 0.1305151879787445, + "learning_rate": 1.465650862715457e-05, + "loss": 1.0177, + "step": 12356 + }, + { + "epoch": 0.7507746521659883, + "grad_norm": 0.1966802030801773, + "learning_rate": 1.4649738034687206e-05, + "loss": 1.1407, + "step": 12357 + }, + { + "epoch": 0.7508354091986147, + "grad_norm": 3.1399667263031006, + "learning_rate": 1.4642968737990543e-05, + "loss": 1.1104, + "step": 12358 + }, + { + "epoch": 0.7508961662312412, + "grad_norm": 0.20816075801849365, + "learning_rate": 1.463620073731266e-05, + "loss": 1.0138, + "step": 12359 + }, + { + "epoch": 0.7509569232638678, + "grad_norm": 0.20617693662643433, + "learning_rate": 1.4629434032901696e-05, + "loss": 1.1315, + "step": 12360 + }, + { + "epoch": 0.7510176802964943, + "grad_norm": 0.18313680589199066, + "learning_rate": 1.462266862500566e-05, + "loss": 0.9971, + "step": 12361 + }, + { + "epoch": 0.7510784373291208, + "grad_norm": 0.1659594625234604, + "learning_rate": 1.4615904513872547e-05, + "loss": 1.1037, + "step": 12362 + }, + { + "epoch": 0.7511391943617474, + "grad_norm": 0.16755665838718414, + "learning_rate": 1.4609141699750295e-05, + "loss": 1.0385, + "step": 12363 + }, + { + "epoch": 0.7511999513943739, + "grad_norm": 0.14337384700775146, + "learning_rate": 1.4602380182886788e-05, + "loss": 1.0341, + "step": 12364 + }, + { + "epoch": 0.7512607084270004, + "grad_norm": 0.43202534317970276, + "learning_rate": 1.4595619963529883e-05, + "loss": 1.0164, + "step": 12365 + }, + { + "epoch": 0.751321465459627, + "grad_norm": 0.20758479833602905, + "learning_rate": 1.458886104192737e-05, + "loss": 1.0871, + "step": 12366 + }, + { + "epoch": 0.7513822224922535, + "grad_norm": 0.19898706674575806, + "learning_rate": 1.4582103418327003e-05, + "loss": 1.0324, + "step": 12367 + }, + { + "epoch": 0.75144297952488, + "grad_norm": 1.4586150646209717, + "learning_rate": 1.4575347092976477e-05, + "loss": 1.044, + "step": 12368 + }, + { + "epoch": 0.7515037365575066, + "grad_norm": 0.3867838382720947, + "learning_rate": 1.4568592066123454e-05, + "loss": 1.2002, + "step": 12369 + }, + { + "epoch": 0.7515644935901331, + "grad_norm": 0.4569261968135834, + "learning_rate": 1.456183833801552e-05, + "loss": 1.0396, + "step": 12370 + }, + { + "epoch": 0.7516252506227595, + "grad_norm": 2.977109432220459, + "learning_rate": 1.4555085908900267e-05, + "loss": 1.1348, + "step": 12371 + }, + { + "epoch": 0.7516860076553861, + "grad_norm": 0.2584514915943146, + "learning_rate": 1.4548334779025213e-05, + "loss": 1.0938, + "step": 12372 + }, + { + "epoch": 0.7517467646880126, + "grad_norm": 0.1292036771774292, + "learning_rate": 1.4541584948637776e-05, + "loss": 1.0045, + "step": 12373 + }, + { + "epoch": 0.7518075217206391, + "grad_norm": 1.24665367603302, + "learning_rate": 1.4534836417985403e-05, + "loss": 1.0347, + "step": 12374 + }, + { + "epoch": 0.7518682787532657, + "grad_norm": 0.23743687570095062, + "learning_rate": 1.4528089187315436e-05, + "loss": 1.098, + "step": 12375 + }, + { + "epoch": 0.7519290357858922, + "grad_norm": 0.19594256579875946, + "learning_rate": 1.4521343256875235e-05, + "loss": 1.0775, + "step": 12376 + }, + { + "epoch": 0.7519897928185187, + "grad_norm": 0.21354098618030548, + "learning_rate": 1.4514598626912057e-05, + "loss": 1.1437, + "step": 12377 + }, + { + "epoch": 0.7520505498511453, + "grad_norm": 0.30547693371772766, + "learning_rate": 1.450785529767314e-05, + "loss": 1.1916, + "step": 12378 + }, + { + "epoch": 0.7521113068837718, + "grad_norm": 0.14954106509685516, + "learning_rate": 1.450111326940563e-05, + "loss": 1.0604, + "step": 12379 + }, + { + "epoch": 0.7521720639163983, + "grad_norm": 0.23873847723007202, + "learning_rate": 1.449437254235666e-05, + "loss": 1.0063, + "step": 12380 + }, + { + "epoch": 0.7522328209490249, + "grad_norm": 0.22796984016895294, + "learning_rate": 1.4487633116773341e-05, + "loss": 1.0628, + "step": 12381 + }, + { + "epoch": 0.7522935779816514, + "grad_norm": 0.41442883014678955, + "learning_rate": 1.4480894992902694e-05, + "loss": 1.2479, + "step": 12382 + }, + { + "epoch": 0.752354335014278, + "grad_norm": 0.22826454043388367, + "learning_rate": 1.4474158170991708e-05, + "loss": 0.9943, + "step": 12383 + }, + { + "epoch": 0.7524150920469044, + "grad_norm": 0.3993418216705322, + "learning_rate": 1.4467422651287326e-05, + "loss": 1.1464, + "step": 12384 + }, + { + "epoch": 0.7524758490795309, + "grad_norm": 0.21205493807792664, + "learning_rate": 1.4460688434036429e-05, + "loss": 0.986, + "step": 12385 + }, + { + "epoch": 0.7525366061121574, + "grad_norm": 0.18927305936813354, + "learning_rate": 1.4453955519485868e-05, + "loss": 1.0789, + "step": 12386 + }, + { + "epoch": 0.752597363144784, + "grad_norm": 0.13985902070999146, + "learning_rate": 1.444722390788244e-05, + "loss": 1.034, + "step": 12387 + }, + { + "epoch": 0.7526581201774105, + "grad_norm": 0.18720990419387817, + "learning_rate": 1.4440493599472894e-05, + "loss": 1.0848, + "step": 12388 + }, + { + "epoch": 0.752718877210037, + "grad_norm": 0.2454121857881546, + "learning_rate": 1.4433764594503923e-05, + "loss": 1.1308, + "step": 12389 + }, + { + "epoch": 0.7527796342426636, + "grad_norm": 0.9478039145469666, + "learning_rate": 1.442703689322219e-05, + "loss": 1.3211, + "step": 12390 + }, + { + "epoch": 0.7528403912752901, + "grad_norm": 0.2874387204647064, + "learning_rate": 1.4420310495874279e-05, + "loss": 1.1813, + "step": 12391 + }, + { + "epoch": 0.7529011483079167, + "grad_norm": 0.19087065756320953, + "learning_rate": 1.4413585402706798e-05, + "loss": 1.0579, + "step": 12392 + }, + { + "epoch": 0.7529619053405432, + "grad_norm": 0.16360221803188324, + "learning_rate": 1.4406861613966205e-05, + "loss": 1.0133, + "step": 12393 + }, + { + "epoch": 0.7530226623731697, + "grad_norm": 0.31029802560806274, + "learning_rate": 1.4400139129898977e-05, + "loss": 1.1558, + "step": 12394 + }, + { + "epoch": 0.7530834194057963, + "grad_norm": 0.1791743040084839, + "learning_rate": 1.439341795075153e-05, + "loss": 1.036, + "step": 12395 + }, + { + "epoch": 0.7531441764384228, + "grad_norm": 0.23604726791381836, + "learning_rate": 1.4386698076770216e-05, + "loss": 1.0967, + "step": 12396 + }, + { + "epoch": 0.7532049334710492, + "grad_norm": 0.23559719324111938, + "learning_rate": 1.4379979508201385e-05, + "loss": 1.1214, + "step": 12397 + }, + { + "epoch": 0.7532656905036758, + "grad_norm": 0.36540910601615906, + "learning_rate": 1.4373262245291308e-05, + "loss": 1.1712, + "step": 12398 + }, + { + "epoch": 0.7533264475363023, + "grad_norm": 0.20903797447681427, + "learning_rate": 1.4366546288286176e-05, + "loss": 0.9842, + "step": 12399 + }, + { + "epoch": 0.7533872045689288, + "grad_norm": 0.20424382388591766, + "learning_rate": 1.435983163743217e-05, + "loss": 1.1409, + "step": 12400 + }, + { + "epoch": 0.7534479616015554, + "grad_norm": 0.20547153055667877, + "learning_rate": 1.4353118292975414e-05, + "loss": 1.1166, + "step": 12401 + }, + { + "epoch": 0.7535087186341819, + "grad_norm": 2.086050271987915, + "learning_rate": 1.434640625516201e-05, + "loss": 1.0576, + "step": 12402 + }, + { + "epoch": 0.7535694756668084, + "grad_norm": 0.2608780860900879, + "learning_rate": 1.4339695524237979e-05, + "loss": 1.1662, + "step": 12403 + }, + { + "epoch": 0.753630232699435, + "grad_norm": 0.25754937529563904, + "learning_rate": 1.4332986100449303e-05, + "loss": 1.0922, + "step": 12404 + }, + { + "epoch": 0.7536909897320615, + "grad_norm": 0.2551557719707489, + "learning_rate": 1.4326277984041925e-05, + "loss": 1.0635, + "step": 12405 + }, + { + "epoch": 0.753751746764688, + "grad_norm": 0.19687290489673615, + "learning_rate": 1.4319571175261698e-05, + "loss": 1.092, + "step": 12406 + }, + { + "epoch": 0.7538125037973146, + "grad_norm": 0.17205794155597687, + "learning_rate": 1.4312865674354497e-05, + "loss": 1.0242, + "step": 12407 + }, + { + "epoch": 0.7538732608299411, + "grad_norm": 0.23441699147224426, + "learning_rate": 1.4306161481566105e-05, + "loss": 1.0849, + "step": 12408 + }, + { + "epoch": 0.7539340178625676, + "grad_norm": 0.679172694683075, + "learning_rate": 1.4299458597142267e-05, + "loss": 1.0321, + "step": 12409 + }, + { + "epoch": 0.7539947748951941, + "grad_norm": 0.2127111405134201, + "learning_rate": 1.4292757021328673e-05, + "loss": 1.0576, + "step": 12410 + }, + { + "epoch": 0.7540555319278206, + "grad_norm": 0.278123676776886, + "learning_rate": 1.4286056754370968e-05, + "loss": 1.0795, + "step": 12411 + }, + { + "epoch": 0.7541162889604471, + "grad_norm": 0.19436343014240265, + "learning_rate": 1.4279357796514759e-05, + "loss": 1.0664, + "step": 12412 + }, + { + "epoch": 0.7541770459930737, + "grad_norm": 0.158972829580307, + "learning_rate": 1.4272660148005596e-05, + "loss": 1.0398, + "step": 12413 + }, + { + "epoch": 0.7542378030257002, + "grad_norm": 0.15032760798931122, + "learning_rate": 1.4265963809088977e-05, + "loss": 1.0151, + "step": 12414 + }, + { + "epoch": 0.7542985600583267, + "grad_norm": 0.20882120728492737, + "learning_rate": 1.4259268780010371e-05, + "loss": 1.0736, + "step": 12415 + }, + { + "epoch": 0.7543593170909533, + "grad_norm": 0.1399017870426178, + "learning_rate": 1.4252575061015166e-05, + "loss": 1.0058, + "step": 12416 + }, + { + "epoch": 0.7544200741235798, + "grad_norm": 0.23553623259067535, + "learning_rate": 1.4245882652348724e-05, + "loss": 1.1341, + "step": 12417 + }, + { + "epoch": 0.7544808311562063, + "grad_norm": 0.19232937693595886, + "learning_rate": 1.4239191554256376e-05, + "loss": 1.0599, + "step": 12418 + }, + { + "epoch": 0.7545415881888329, + "grad_norm": 0.4818721115589142, + "learning_rate": 1.4232501766983391e-05, + "loss": 1.0685, + "step": 12419 + }, + { + "epoch": 0.7546023452214594, + "grad_norm": 0.14236430823802948, + "learning_rate": 1.4225813290774948e-05, + "loss": 1.0347, + "step": 12420 + }, + { + "epoch": 0.754663102254086, + "grad_norm": 0.2666613757610321, + "learning_rate": 1.4219126125876237e-05, + "loss": 1.1159, + "step": 12421 + }, + { + "epoch": 0.7547238592867125, + "grad_norm": 9.05506420135498, + "learning_rate": 1.4212440272532362e-05, + "loss": 0.9965, + "step": 12422 + }, + { + "epoch": 0.754784616319339, + "grad_norm": 0.27398842573165894, + "learning_rate": 1.4205755730988418e-05, + "loss": 1.0423, + "step": 12423 + }, + { + "epoch": 0.7548453733519654, + "grad_norm": 0.2804865539073944, + "learning_rate": 1.4199072501489418e-05, + "loss": 1.1814, + "step": 12424 + }, + { + "epoch": 0.754906130384592, + "grad_norm": 0.28536710143089294, + "learning_rate": 1.4192390584280346e-05, + "loss": 1.0509, + "step": 12425 + }, + { + "epoch": 0.7549668874172185, + "grad_norm": 0.2250620573759079, + "learning_rate": 1.4185709979606094e-05, + "loss": 0.9514, + "step": 12426 + }, + { + "epoch": 0.755027644449845, + "grad_norm": 0.179813414812088, + "learning_rate": 1.417903068771158e-05, + "loss": 1.0782, + "step": 12427 + }, + { + "epoch": 0.7550884014824716, + "grad_norm": 0.48168784379959106, + "learning_rate": 1.4172352708841618e-05, + "loss": 1.1688, + "step": 12428 + }, + { + "epoch": 0.7551491585150981, + "grad_norm": 0.14817216992378235, + "learning_rate": 1.4165676043240995e-05, + "loss": 1.0376, + "step": 12429 + }, + { + "epoch": 0.7552099155477247, + "grad_norm": 0.1814858764410019, + "learning_rate": 1.4159000691154439e-05, + "loss": 1.0662, + "step": 12430 + }, + { + "epoch": 0.7552706725803512, + "grad_norm": 0.1313580572605133, + "learning_rate": 1.4152326652826647e-05, + "loss": 1.0179, + "step": 12431 + }, + { + "epoch": 0.7553314296129777, + "grad_norm": 0.2124365270137787, + "learning_rate": 1.4145653928502245e-05, + "loss": 1.0225, + "step": 12432 + }, + { + "epoch": 0.7553921866456043, + "grad_norm": 0.13885805010795593, + "learning_rate": 1.4138982518425831e-05, + "loss": 1.0018, + "step": 12433 + }, + { + "epoch": 0.7554529436782308, + "grad_norm": 0.20052242279052734, + "learning_rate": 1.4132312422841947e-05, + "loss": 1.1465, + "step": 12434 + }, + { + "epoch": 0.7555137007108573, + "grad_norm": 0.16675755381584167, + "learning_rate": 1.4125643641995078e-05, + "loss": 1.1143, + "step": 12435 + }, + { + "epoch": 0.7555744577434839, + "grad_norm": 0.13732722401618958, + "learning_rate": 1.4118976176129684e-05, + "loss": 1.0273, + "step": 12436 + }, + { + "epoch": 0.7556352147761103, + "grad_norm": 0.16032195091247559, + "learning_rate": 1.411231002549015e-05, + "loss": 1.0242, + "step": 12437 + }, + { + "epoch": 0.7556959718087368, + "grad_norm": 0.15149761736392975, + "learning_rate": 1.410564519032081e-05, + "loss": 1.0795, + "step": 12438 + }, + { + "epoch": 0.7557567288413634, + "grad_norm": 0.5057505369186401, + "learning_rate": 1.4098981670866018e-05, + "loss": 1.0265, + "step": 12439 + }, + { + "epoch": 0.7558174858739899, + "grad_norm": 0.21293695271015167, + "learning_rate": 1.4092319467369974e-05, + "loss": 1.1148, + "step": 12440 + }, + { + "epoch": 0.7558782429066164, + "grad_norm": 0.2395654171705246, + "learning_rate": 1.4085658580076905e-05, + "loss": 1.1404, + "step": 12441 + }, + { + "epoch": 0.755938999939243, + "grad_norm": 0.15418641269207, + "learning_rate": 1.4078999009230947e-05, + "loss": 1.0306, + "step": 12442 + }, + { + "epoch": 0.7559997569718695, + "grad_norm": 1.8570897579193115, + "learning_rate": 1.4072340755076235e-05, + "loss": 1.0551, + "step": 12443 + }, + { + "epoch": 0.756060514004496, + "grad_norm": 0.2591049373149872, + "learning_rate": 1.4065683817856823e-05, + "loss": 1.153, + "step": 12444 + }, + { + "epoch": 0.7561212710371226, + "grad_norm": 0.23178306221961975, + "learning_rate": 1.4059028197816725e-05, + "loss": 0.9838, + "step": 12445 + }, + { + "epoch": 0.7561820280697491, + "grad_norm": 0.42688867449760437, + "learning_rate": 1.4052373895199888e-05, + "loss": 1.2381, + "step": 12446 + }, + { + "epoch": 0.7562427851023756, + "grad_norm": 0.16974328458309174, + "learning_rate": 1.4045720910250216e-05, + "loss": 1.0711, + "step": 12447 + }, + { + "epoch": 0.7563035421350022, + "grad_norm": 0.14781701564788818, + "learning_rate": 1.4039069243211605e-05, + "loss": 1.0224, + "step": 12448 + }, + { + "epoch": 0.7563642991676287, + "grad_norm": 0.4157654941082001, + "learning_rate": 1.403241889432787e-05, + "loss": 1.1431, + "step": 12449 + }, + { + "epoch": 0.7564250562002551, + "grad_norm": 0.2512097954750061, + "learning_rate": 1.4025769863842764e-05, + "loss": 1.0131, + "step": 12450 + }, + { + "epoch": 0.7564858132328817, + "grad_norm": 0.1476621925830841, + "learning_rate": 1.4019122152000025e-05, + "loss": 1.0308, + "step": 12451 + }, + { + "epoch": 0.7565465702655082, + "grad_norm": 0.20870807766914368, + "learning_rate": 1.4012475759043314e-05, + "loss": 1.1327, + "step": 12452 + }, + { + "epoch": 0.7566073272981347, + "grad_norm": 0.1429072618484497, + "learning_rate": 1.4005830685216259e-05, + "loss": 1.0273, + "step": 12453 + }, + { + "epoch": 0.7566680843307613, + "grad_norm": 0.19838672876358032, + "learning_rate": 1.3999186930762442e-05, + "loss": 1.0912, + "step": 12454 + }, + { + "epoch": 0.7567288413633878, + "grad_norm": 0.4041643440723419, + "learning_rate": 1.3992544495925386e-05, + "loss": 1.0361, + "step": 12455 + }, + { + "epoch": 0.7567895983960143, + "grad_norm": 0.16092495620250702, + "learning_rate": 1.3985903380948573e-05, + "loss": 1.0438, + "step": 12456 + }, + { + "epoch": 0.7568503554286409, + "grad_norm": 0.12724977731704712, + "learning_rate": 1.3979263586075426e-05, + "loss": 1.0578, + "step": 12457 + }, + { + "epoch": 0.7569111124612674, + "grad_norm": 0.13582748174667358, + "learning_rate": 1.3972625111549325e-05, + "loss": 0.9963, + "step": 12458 + }, + { + "epoch": 0.756971869493894, + "grad_norm": 0.14830635488033295, + "learning_rate": 1.3965987957613647e-05, + "loss": 1.0649, + "step": 12459 + }, + { + "epoch": 0.7570326265265205, + "grad_norm": 0.21304182708263397, + "learning_rate": 1.3959352124511627e-05, + "loss": 1.0879, + "step": 12460 + }, + { + "epoch": 0.757093383559147, + "grad_norm": 0.14700187742710114, + "learning_rate": 1.3952717612486522e-05, + "loss": 1.0418, + "step": 12461 + }, + { + "epoch": 0.7571541405917736, + "grad_norm": 0.21427591145038605, + "learning_rate": 1.394608442178152e-05, + "loss": 1.0994, + "step": 12462 + }, + { + "epoch": 0.7572148976244, + "grad_norm": 0.14718963205814362, + "learning_rate": 1.3939452552639748e-05, + "loss": 1.0917, + "step": 12463 + }, + { + "epoch": 0.7572756546570265, + "grad_norm": 0.2005254328250885, + "learning_rate": 1.3932822005304325e-05, + "loss": 1.0448, + "step": 12464 + }, + { + "epoch": 0.757336411689653, + "grad_norm": 0.1698300540447235, + "learning_rate": 1.3926192780018305e-05, + "loss": 1.1302, + "step": 12465 + }, + { + "epoch": 0.7573971687222796, + "grad_norm": 0.3536456227302551, + "learning_rate": 1.3919564877024644e-05, + "loss": 1.0642, + "step": 12466 + }, + { + "epoch": 0.7574579257549061, + "grad_norm": 0.1791263222694397, + "learning_rate": 1.3912938296566302e-05, + "loss": 1.0609, + "step": 12467 + }, + { + "epoch": 0.7575186827875326, + "grad_norm": 0.20895130932331085, + "learning_rate": 1.3906313038886171e-05, + "loss": 1.1871, + "step": 12468 + }, + { + "epoch": 0.7575794398201592, + "grad_norm": 0.1597427874803543, + "learning_rate": 1.3899689104227126e-05, + "loss": 1.0243, + "step": 12469 + }, + { + "epoch": 0.7576401968527857, + "grad_norm": 0.14181551337242126, + "learning_rate": 1.3893066492831957e-05, + "loss": 1.0252, + "step": 12470 + }, + { + "epoch": 0.7577009538854123, + "grad_norm": 0.12455352395772934, + "learning_rate": 1.3886445204943416e-05, + "loss": 1.0062, + "step": 12471 + }, + { + "epoch": 0.7577617109180388, + "grad_norm": 0.1531362533569336, + "learning_rate": 1.3879825240804217e-05, + "loss": 0.9953, + "step": 12472 + }, + { + "epoch": 0.7578224679506653, + "grad_norm": 0.1934642344713211, + "learning_rate": 1.387320660065698e-05, + "loss": 1.0349, + "step": 12473 + }, + { + "epoch": 0.7578832249832919, + "grad_norm": 1.9845266342163086, + "learning_rate": 1.3866589284744352e-05, + "loss": 1.0351, + "step": 12474 + }, + { + "epoch": 0.7579439820159184, + "grad_norm": 0.1374569833278656, + "learning_rate": 1.3859973293308876e-05, + "loss": 1.0516, + "step": 12475 + }, + { + "epoch": 0.7580047390485448, + "grad_norm": 0.13528355956077576, + "learning_rate": 1.3853358626593065e-05, + "loss": 1.0393, + "step": 12476 + }, + { + "epoch": 0.7580654960811714, + "grad_norm": 0.16686031222343445, + "learning_rate": 1.3846745284839385e-05, + "loss": 1.075, + "step": 12477 + }, + { + "epoch": 0.7581262531137979, + "grad_norm": 0.13151110708713531, + "learning_rate": 1.3840133268290239e-05, + "loss": 0.9814, + "step": 12478 + }, + { + "epoch": 0.7581870101464244, + "grad_norm": 0.2679908573627472, + "learning_rate": 1.3833522577188002e-05, + "loss": 1.2532, + "step": 12479 + }, + { + "epoch": 0.758247767179051, + "grad_norm": 0.8641195297241211, + "learning_rate": 1.3826913211774983e-05, + "loss": 1.1201, + "step": 12480 + }, + { + "epoch": 0.7583085242116775, + "grad_norm": 0.20159924030303955, + "learning_rate": 1.3820305172293451e-05, + "loss": 1.0265, + "step": 12481 + }, + { + "epoch": 0.758369281244304, + "grad_norm": 0.1631007194519043, + "learning_rate": 1.381369845898563e-05, + "loss": 1.0716, + "step": 12482 + }, + { + "epoch": 0.7584300382769306, + "grad_norm": 0.1293521672487259, + "learning_rate": 1.3807093072093686e-05, + "loss": 1.0377, + "step": 12483 + }, + { + "epoch": 0.7584907953095571, + "grad_norm": 0.22282932698726654, + "learning_rate": 1.3800489011859724e-05, + "loss": 1.1226, + "step": 12484 + }, + { + "epoch": 0.7585515523421836, + "grad_norm": 0.23399317264556885, + "learning_rate": 1.3793886278525853e-05, + "loss": 1.088, + "step": 12485 + }, + { + "epoch": 0.7586123093748102, + "grad_norm": 0.20826606452465057, + "learning_rate": 1.3787284872334095e-05, + "loss": 1.0888, + "step": 12486 + }, + { + "epoch": 0.7586730664074367, + "grad_norm": 0.20823538303375244, + "learning_rate": 1.3780684793526389e-05, + "loss": 1.1186, + "step": 12487 + }, + { + "epoch": 0.7587338234400632, + "grad_norm": 0.1876756250858307, + "learning_rate": 1.3774086042344686e-05, + "loss": 1.1105, + "step": 12488 + }, + { + "epoch": 0.7587945804726897, + "grad_norm": 0.15832959115505219, + "learning_rate": 1.3767488619030843e-05, + "loss": 1.0007, + "step": 12489 + }, + { + "epoch": 0.7588553375053162, + "grad_norm": 0.28893133997917175, + "learning_rate": 1.3760892523826729e-05, + "loss": 1.2394, + "step": 12490 + }, + { + "epoch": 0.7589160945379427, + "grad_norm": 0.17421387135982513, + "learning_rate": 1.3754297756974099e-05, + "loss": 1.1036, + "step": 12491 + }, + { + "epoch": 0.7589768515705693, + "grad_norm": 0.20006486773490906, + "learning_rate": 1.374770431871471e-05, + "loss": 1.1163, + "step": 12492 + }, + { + "epoch": 0.7590376086031958, + "grad_norm": 0.23953378200531006, + "learning_rate": 1.3741112209290214e-05, + "loss": 1.0147, + "step": 12493 + }, + { + "epoch": 0.7590983656358223, + "grad_norm": 0.5787145495414734, + "learning_rate": 1.3734521428942238e-05, + "loss": 1.0768, + "step": 12494 + }, + { + "epoch": 0.7591591226684489, + "grad_norm": 0.1401488482952118, + "learning_rate": 1.3727931977912407e-05, + "loss": 1.0687, + "step": 12495 + }, + { + "epoch": 0.7592198797010754, + "grad_norm": 0.2855076193809509, + "learning_rate": 1.3721343856442238e-05, + "loss": 1.0808, + "step": 12496 + }, + { + "epoch": 0.7592806367337019, + "grad_norm": 0.19481967389583588, + "learning_rate": 1.3714757064773226e-05, + "loss": 1.0567, + "step": 12497 + }, + { + "epoch": 0.7593413937663285, + "grad_norm": 0.17138898372650146, + "learning_rate": 1.3708171603146803e-05, + "loss": 1.1023, + "step": 12498 + }, + { + "epoch": 0.759402150798955, + "grad_norm": 0.15591728687286377, + "learning_rate": 1.3701587471804367e-05, + "loss": 0.983, + "step": 12499 + }, + { + "epoch": 0.7594629078315815, + "grad_norm": 3.773162364959717, + "learning_rate": 1.369500467098726e-05, + "loss": 1.091, + "step": 12500 + }, + { + "epoch": 0.7595236648642081, + "grad_norm": 0.431063175201416, + "learning_rate": 1.3688423200936773e-05, + "loss": 1.136, + "step": 12501 + }, + { + "epoch": 0.7595844218968345, + "grad_norm": 0.23997826874256134, + "learning_rate": 1.368184306189415e-05, + "loss": 1.0783, + "step": 12502 + }, + { + "epoch": 0.759645178929461, + "grad_norm": 0.2836208939552307, + "learning_rate": 1.3675264254100595e-05, + "loss": 1.114, + "step": 12503 + }, + { + "epoch": 0.7597059359620876, + "grad_norm": 0.16494353115558624, + "learning_rate": 1.3668686777797247e-05, + "loss": 1.029, + "step": 12504 + }, + { + "epoch": 0.7597666929947141, + "grad_norm": 0.12692925333976746, + "learning_rate": 1.3662110633225196e-05, + "loss": 1.016, + "step": 12505 + }, + { + "epoch": 0.7598274500273406, + "grad_norm": 0.2563782334327698, + "learning_rate": 1.3655535820625532e-05, + "loss": 1.0231, + "step": 12506 + }, + { + "epoch": 0.7598882070599672, + "grad_norm": 0.19143162667751312, + "learning_rate": 1.3648962340239208e-05, + "loss": 1.0772, + "step": 12507 + }, + { + "epoch": 0.7599489640925937, + "grad_norm": 0.19717441499233246, + "learning_rate": 1.3642390192307204e-05, + "loss": 1.1513, + "step": 12508 + }, + { + "epoch": 0.7600097211252203, + "grad_norm": 0.5800662636756897, + "learning_rate": 1.3635819377070407e-05, + "loss": 1.0033, + "step": 12509 + }, + { + "epoch": 0.7600704781578468, + "grad_norm": 0.17145879566669464, + "learning_rate": 1.3629249894769668e-05, + "loss": 1.0768, + "step": 12510 + }, + { + "epoch": 0.7601312351904733, + "grad_norm": 0.5142497420310974, + "learning_rate": 1.3622681745645821e-05, + "loss": 1.0568, + "step": 12511 + }, + { + "epoch": 0.7601919922230999, + "grad_norm": 0.2719340920448303, + "learning_rate": 1.3616114929939622e-05, + "loss": 1.0749, + "step": 12512 + }, + { + "epoch": 0.7602527492557264, + "grad_norm": 0.26659440994262695, + "learning_rate": 1.3609549447891745e-05, + "loss": 1.1094, + "step": 12513 + }, + { + "epoch": 0.7603135062883529, + "grad_norm": 0.20936378836631775, + "learning_rate": 1.360298529974287e-05, + "loss": 1.0626, + "step": 12514 + }, + { + "epoch": 0.7603742633209793, + "grad_norm": 0.22962205111980438, + "learning_rate": 1.3596422485733589e-05, + "loss": 1.0994, + "step": 12515 + }, + { + "epoch": 0.7604350203536059, + "grad_norm": 0.15905433893203735, + "learning_rate": 1.3589861006104494e-05, + "loss": 1.0488, + "step": 12516 + }, + { + "epoch": 0.7604957773862324, + "grad_norm": 0.12805649638175964, + "learning_rate": 1.3583300861096077e-05, + "loss": 1.001, + "step": 12517 + }, + { + "epoch": 0.760556534418859, + "grad_norm": 0.1723327338695526, + "learning_rate": 1.357674205094881e-05, + "loss": 1.1031, + "step": 12518 + }, + { + "epoch": 0.7606172914514855, + "grad_norm": 0.21429628133773804, + "learning_rate": 1.3570184575903105e-05, + "loss": 1.0603, + "step": 12519 + }, + { + "epoch": 0.760678048484112, + "grad_norm": 0.23544301092624664, + "learning_rate": 1.3563628436199322e-05, + "loss": 1.1064, + "step": 12520 + }, + { + "epoch": 0.7607388055167386, + "grad_norm": 0.20912890136241913, + "learning_rate": 1.3557073632077783e-05, + "loss": 1.0331, + "step": 12521 + }, + { + "epoch": 0.7607995625493651, + "grad_norm": 0.16583947837352753, + "learning_rate": 1.3550520163778757e-05, + "loss": 0.9913, + "step": 12522 + }, + { + "epoch": 0.7608603195819916, + "grad_norm": 1.9085358381271362, + "learning_rate": 1.3543968031542459e-05, + "loss": 1.0307, + "step": 12523 + }, + { + "epoch": 0.7609210766146182, + "grad_norm": 0.12597978115081787, + "learning_rate": 1.3537417235609057e-05, + "loss": 1.0041, + "step": 12524 + }, + { + "epoch": 0.7609818336472447, + "grad_norm": 0.1700996607542038, + "learning_rate": 1.3530867776218675e-05, + "loss": 1.0224, + "step": 12525 + }, + { + "epoch": 0.7610425906798712, + "grad_norm": 0.19211909174919128, + "learning_rate": 1.3524319653611384e-05, + "loss": 1.0858, + "step": 12526 + }, + { + "epoch": 0.7611033477124978, + "grad_norm": 0.14007562398910522, + "learning_rate": 1.3517772868027206e-05, + "loss": 1.0763, + "step": 12527 + }, + { + "epoch": 0.7611641047451243, + "grad_norm": 0.1527559608221054, + "learning_rate": 1.3511227419706119e-05, + "loss": 0.9902, + "step": 12528 + }, + { + "epoch": 0.7612248617777507, + "grad_norm": 0.2170749455690384, + "learning_rate": 1.3504683308888043e-05, + "loss": 1.0738, + "step": 12529 + }, + { + "epoch": 0.7612856188103773, + "grad_norm": 0.38003069162368774, + "learning_rate": 1.3498140535812852e-05, + "loss": 1.052, + "step": 12530 + }, + { + "epoch": 0.7613463758430038, + "grad_norm": 0.2978039085865021, + "learning_rate": 1.3491599100720365e-05, + "loss": 1.1331, + "step": 12531 + }, + { + "epoch": 0.7614071328756303, + "grad_norm": 0.15652498602867126, + "learning_rate": 1.3485059003850392e-05, + "loss": 1.0519, + "step": 12532 + }, + { + "epoch": 0.7614678899082569, + "grad_norm": 1.7316577434539795, + "learning_rate": 1.3478520245442633e-05, + "loss": 1.0909, + "step": 12533 + }, + { + "epoch": 0.7615286469408834, + "grad_norm": 0.12219748646020889, + "learning_rate": 1.3471982825736767e-05, + "loss": 1.0017, + "step": 12534 + }, + { + "epoch": 0.7615894039735099, + "grad_norm": 0.20619073510169983, + "learning_rate": 1.346544674497242e-05, + "loss": 1.125, + "step": 12535 + }, + { + "epoch": 0.7616501610061365, + "grad_norm": 0.22032634913921356, + "learning_rate": 1.3458912003389201e-05, + "loss": 1.0848, + "step": 12536 + }, + { + "epoch": 0.761710918038763, + "grad_norm": 0.5504510998725891, + "learning_rate": 1.3452378601226624e-05, + "loss": 1.1116, + "step": 12537 + }, + { + "epoch": 0.7617716750713895, + "grad_norm": 0.24087081849575043, + "learning_rate": 1.3445846538724171e-05, + "loss": 1.1509, + "step": 12538 + }, + { + "epoch": 0.7618324321040161, + "grad_norm": 0.13236486911773682, + "learning_rate": 1.3439315816121294e-05, + "loss": 1.0175, + "step": 12539 + }, + { + "epoch": 0.7618931891366426, + "grad_norm": 0.24486103653907776, + "learning_rate": 1.3432786433657335e-05, + "loss": 1.2894, + "step": 12540 + }, + { + "epoch": 0.7619539461692691, + "grad_norm": 0.2049788534641266, + "learning_rate": 1.3426258391571677e-05, + "loss": 1.0707, + "step": 12541 + }, + { + "epoch": 0.7620147032018956, + "grad_norm": 0.16394475102424622, + "learning_rate": 1.3419731690103581e-05, + "loss": 1.0375, + "step": 12542 + }, + { + "epoch": 0.7620754602345221, + "grad_norm": 0.19463828206062317, + "learning_rate": 1.3413206329492295e-05, + "loss": 1.141, + "step": 12543 + }, + { + "epoch": 0.7621362172671486, + "grad_norm": 0.48082804679870605, + "learning_rate": 1.3406682309976998e-05, + "loss": 1.0763, + "step": 12544 + }, + { + "epoch": 0.7621969742997752, + "grad_norm": 0.17180514335632324, + "learning_rate": 1.3400159631796838e-05, + "loss": 1.0973, + "step": 12545 + }, + { + "epoch": 0.7622577313324017, + "grad_norm": 0.16259151697158813, + "learning_rate": 1.3393638295190896e-05, + "loss": 1.0541, + "step": 12546 + }, + { + "epoch": 0.7623184883650282, + "grad_norm": 0.23652586340904236, + "learning_rate": 1.3387118300398227e-05, + "loss": 1.1115, + "step": 12547 + }, + { + "epoch": 0.7623792453976548, + "grad_norm": 0.15315882861614227, + "learning_rate": 1.338059964765781e-05, + "loss": 1.079, + "step": 12548 + }, + { + "epoch": 0.7624400024302813, + "grad_norm": 0.21819977462291718, + "learning_rate": 1.3374082337208587e-05, + "loss": 0.9945, + "step": 12549 + }, + { + "epoch": 0.7625007594629079, + "grad_norm": 0.19960103929042816, + "learning_rate": 1.3367566369289457e-05, + "loss": 1.184, + "step": 12550 + }, + { + "epoch": 0.7625615164955344, + "grad_norm": 2.7370994091033936, + "learning_rate": 1.3361051744139253e-05, + "loss": 0.9905, + "step": 12551 + }, + { + "epoch": 0.7626222735281609, + "grad_norm": 0.20495986938476562, + "learning_rate": 1.3354538461996807e-05, + "loss": 1.1323, + "step": 12552 + }, + { + "epoch": 0.7626830305607875, + "grad_norm": 0.15847209095954895, + "learning_rate": 1.334802652310082e-05, + "loss": 1.0239, + "step": 12553 + }, + { + "epoch": 0.762743787593414, + "grad_norm": 0.17811959981918335, + "learning_rate": 1.3341515927690007e-05, + "loss": 1.0818, + "step": 12554 + }, + { + "epoch": 0.7628045446260404, + "grad_norm": 0.22100414335727692, + "learning_rate": 1.3335006676003014e-05, + "loss": 1.1106, + "step": 12555 + }, + { + "epoch": 0.762865301658667, + "grad_norm": 0.19699764251708984, + "learning_rate": 1.332849876827842e-05, + "loss": 1.0875, + "step": 12556 + }, + { + "epoch": 0.7629260586912935, + "grad_norm": 0.8191133141517639, + "learning_rate": 1.3321992204754813e-05, + "loss": 1.1787, + "step": 12557 + }, + { + "epoch": 0.76298681572392, + "grad_norm": 0.1856895536184311, + "learning_rate": 1.3315486985670667e-05, + "loss": 1.0054, + "step": 12558 + }, + { + "epoch": 0.7630475727565466, + "grad_norm": 0.1923539638519287, + "learning_rate": 1.330898311126445e-05, + "loss": 1.1287, + "step": 12559 + }, + { + "epoch": 0.7631083297891731, + "grad_norm": 0.15246796607971191, + "learning_rate": 1.330248058177454e-05, + "loss": 1.0166, + "step": 12560 + }, + { + "epoch": 0.7631690868217996, + "grad_norm": 0.16366790235042572, + "learning_rate": 1.3295979397439279e-05, + "loss": 1.0139, + "step": 12561 + }, + { + "epoch": 0.7632298438544262, + "grad_norm": 0.1890559047460556, + "learning_rate": 1.3289479558497003e-05, + "loss": 1.0079, + "step": 12562 + }, + { + "epoch": 0.7632906008870527, + "grad_norm": 0.2307078093290329, + "learning_rate": 1.328298106518595e-05, + "loss": 1.0396, + "step": 12563 + }, + { + "epoch": 0.7633513579196792, + "grad_norm": 0.17899379134178162, + "learning_rate": 1.3276483917744326e-05, + "loss": 1.0502, + "step": 12564 + }, + { + "epoch": 0.7634121149523058, + "grad_norm": 0.18211714923381805, + "learning_rate": 1.3269988116410276e-05, + "loss": 1.0564, + "step": 12565 + }, + { + "epoch": 0.7634728719849323, + "grad_norm": 0.15848837792873383, + "learning_rate": 1.3263493661421916e-05, + "loss": 1.0691, + "step": 12566 + }, + { + "epoch": 0.7635336290175588, + "grad_norm": 0.1611543744802475, + "learning_rate": 1.3257000553017296e-05, + "loss": 1.0334, + "step": 12567 + }, + { + "epoch": 0.7635943860501853, + "grad_norm": 0.16787272691726685, + "learning_rate": 1.3250508791434424e-05, + "loss": 0.9899, + "step": 12568 + }, + { + "epoch": 0.7636551430828118, + "grad_norm": 0.13397838175296783, + "learning_rate": 1.3244018376911255e-05, + "loss": 1.0435, + "step": 12569 + }, + { + "epoch": 0.7637159001154383, + "grad_norm": 1.822006344795227, + "learning_rate": 1.3237529309685697e-05, + "loss": 0.9972, + "step": 12570 + }, + { + "epoch": 0.7637766571480649, + "grad_norm": 0.19742923974990845, + "learning_rate": 1.3231041589995607e-05, + "loss": 1.0007, + "step": 12571 + }, + { + "epoch": 0.7638374141806914, + "grad_norm": 6.730983257293701, + "learning_rate": 1.3224555218078782e-05, + "loss": 1.0355, + "step": 12572 + }, + { + "epoch": 0.7638981712133179, + "grad_norm": 0.17334315180778503, + "learning_rate": 1.3218070194173027e-05, + "loss": 1.0476, + "step": 12573 + }, + { + "epoch": 0.7639589282459445, + "grad_norm": 0.20793648064136505, + "learning_rate": 1.3211586518516e-05, + "loss": 1.1287, + "step": 12574 + }, + { + "epoch": 0.764019685278571, + "grad_norm": 0.20518960058689117, + "learning_rate": 1.320510419134538e-05, + "loss": 1.1102, + "step": 12575 + }, + { + "epoch": 0.7640804423111975, + "grad_norm": 0.1878655105829239, + "learning_rate": 1.3198623212898776e-05, + "loss": 1.0169, + "step": 12576 + }, + { + "epoch": 0.7641411993438241, + "grad_norm": 0.4014975130558014, + "learning_rate": 1.319214358341374e-05, + "loss": 1.1533, + "step": 12577 + }, + { + "epoch": 0.7642019563764506, + "grad_norm": 0.12942343950271606, + "learning_rate": 1.3185665303127809e-05, + "loss": 0.9908, + "step": 12578 + }, + { + "epoch": 0.7642627134090771, + "grad_norm": 0.23763976991176605, + "learning_rate": 1.317918837227845e-05, + "loss": 1.1463, + "step": 12579 + }, + { + "epoch": 0.7643234704417037, + "grad_norm": 0.22604866325855255, + "learning_rate": 1.3172712791103043e-05, + "loss": 1.1382, + "step": 12580 + }, + { + "epoch": 0.7643842274743301, + "grad_norm": 0.12796372175216675, + "learning_rate": 1.316623855983896e-05, + "loss": 1.0069, + "step": 12581 + }, + { + "epoch": 0.7644449845069566, + "grad_norm": 0.12333228439092636, + "learning_rate": 1.3159765678723518e-05, + "loss": 1.0297, + "step": 12582 + }, + { + "epoch": 0.7645057415395832, + "grad_norm": 0.2215987592935562, + "learning_rate": 1.3153294147993995e-05, + "loss": 1.171, + "step": 12583 + }, + { + "epoch": 0.7645664985722097, + "grad_norm": 0.15680423378944397, + "learning_rate": 1.31468239678876e-05, + "loss": 1.1158, + "step": 12584 + }, + { + "epoch": 0.7646272556048362, + "grad_norm": 0.156818687915802, + "learning_rate": 1.3140355138641496e-05, + "loss": 1.1016, + "step": 12585 + }, + { + "epoch": 0.7646880126374628, + "grad_norm": 0.23134243488311768, + "learning_rate": 1.313388766049281e-05, + "loss": 1.2143, + "step": 12586 + }, + { + "epoch": 0.7647487696700893, + "grad_norm": 0.37965860962867737, + "learning_rate": 1.3127421533678575e-05, + "loss": 1.2922, + "step": 12587 + }, + { + "epoch": 0.7648095267027158, + "grad_norm": 0.12810705602169037, + "learning_rate": 1.3120956758435837e-05, + "loss": 1.0707, + "step": 12588 + }, + { + "epoch": 0.7648702837353424, + "grad_norm": 0.13619191944599152, + "learning_rate": 1.3114493335001565e-05, + "loss": 1.0116, + "step": 12589 + }, + { + "epoch": 0.7649310407679689, + "grad_norm": 0.19306838512420654, + "learning_rate": 1.310803126361267e-05, + "loss": 1.075, + "step": 12590 + }, + { + "epoch": 0.7649917978005955, + "grad_norm": 0.8585348129272461, + "learning_rate": 1.3101570544506014e-05, + "loss": 1.1009, + "step": 12591 + }, + { + "epoch": 0.765052554833222, + "grad_norm": 0.14882126450538635, + "learning_rate": 1.3095111177918423e-05, + "loss": 1.0363, + "step": 12592 + }, + { + "epoch": 0.7651133118658485, + "grad_norm": 0.21391494572162628, + "learning_rate": 1.3088653164086667e-05, + "loss": 1.1112, + "step": 12593 + }, + { + "epoch": 0.765174068898475, + "grad_norm": 0.26765546202659607, + "learning_rate": 1.3082196503247469e-05, + "loss": 1.0611, + "step": 12594 + }, + { + "epoch": 0.7652348259311015, + "grad_norm": 0.15519912540912628, + "learning_rate": 1.3075741195637487e-05, + "loss": 1.0595, + "step": 12595 + }, + { + "epoch": 0.765295582963728, + "grad_norm": 0.21657052636146545, + "learning_rate": 1.306928724149335e-05, + "loss": 1.1862, + "step": 12596 + }, + { + "epoch": 0.7653563399963546, + "grad_norm": 0.46756860613822937, + "learning_rate": 1.306283464105163e-05, + "loss": 1.1897, + "step": 12597 + }, + { + "epoch": 0.7654170970289811, + "grad_norm": 0.22280685603618622, + "learning_rate": 1.3056383394548832e-05, + "loss": 1.0864, + "step": 12598 + }, + { + "epoch": 0.7654778540616076, + "grad_norm": 0.25368282198905945, + "learning_rate": 1.3049933502221468e-05, + "loss": 1.099, + "step": 12599 + }, + { + "epoch": 0.7655386110942342, + "grad_norm": 0.22470787167549133, + "learning_rate": 1.3043484964305924e-05, + "loss": 1.0122, + "step": 12600 + }, + { + "epoch": 0.7655993681268607, + "grad_norm": 0.13428638875484467, + "learning_rate": 1.303703778103858e-05, + "loss": 1.0225, + "step": 12601 + }, + { + "epoch": 0.7656601251594872, + "grad_norm": 0.2146756500005722, + "learning_rate": 1.3030591952655757e-05, + "loss": 1.0652, + "step": 12602 + }, + { + "epoch": 0.7657208821921138, + "grad_norm": 0.17361494898796082, + "learning_rate": 1.3024147479393722e-05, + "loss": 1.0342, + "step": 12603 + }, + { + "epoch": 0.7657816392247403, + "grad_norm": 0.20369866490364075, + "learning_rate": 1.301770436148872e-05, + "loss": 1.033, + "step": 12604 + }, + { + "epoch": 0.7658423962573668, + "grad_norm": 0.19120484590530396, + "learning_rate": 1.3011262599176915e-05, + "loss": 1.0603, + "step": 12605 + }, + { + "epoch": 0.7659031532899934, + "grad_norm": 0.26522183418273926, + "learning_rate": 1.3004822192694444e-05, + "loss": 1.2301, + "step": 12606 + }, + { + "epoch": 0.7659639103226198, + "grad_norm": 0.1325247436761856, + "learning_rate": 1.2998383142277348e-05, + "loss": 0.9969, + "step": 12607 + }, + { + "epoch": 0.7660246673552463, + "grad_norm": 0.3074871003627777, + "learning_rate": 1.2991945448161658e-05, + "loss": 1.1574, + "step": 12608 + }, + { + "epoch": 0.7660854243878729, + "grad_norm": 3.3955633640289307, + "learning_rate": 1.2985509110583378e-05, + "loss": 1.1354, + "step": 12609 + }, + { + "epoch": 0.7661461814204994, + "grad_norm": 0.1966390162706375, + "learning_rate": 1.2979074129778419e-05, + "loss": 1.1341, + "step": 12610 + }, + { + "epoch": 0.7662069384531259, + "grad_norm": 0.19091841578483582, + "learning_rate": 1.2972640505982647e-05, + "loss": 1.1589, + "step": 12611 + }, + { + "epoch": 0.7662676954857525, + "grad_norm": 0.20663481950759888, + "learning_rate": 1.2966208239431899e-05, + "loss": 0.959, + "step": 12612 + }, + { + "epoch": 0.766328452518379, + "grad_norm": 0.30960577726364136, + "learning_rate": 1.2959777330361938e-05, + "loss": 1.0448, + "step": 12613 + }, + { + "epoch": 0.7663892095510055, + "grad_norm": 0.16737186908721924, + "learning_rate": 1.2953347779008508e-05, + "loss": 1.054, + "step": 12614 + }, + { + "epoch": 0.7664499665836321, + "grad_norm": 0.12703509628772736, + "learning_rate": 1.2946919585607275e-05, + "loss": 1.0037, + "step": 12615 + }, + { + "epoch": 0.7665107236162586, + "grad_norm": 0.2665083706378937, + "learning_rate": 1.2940492750393863e-05, + "loss": 1.1709, + "step": 12616 + }, + { + "epoch": 0.7665714806488851, + "grad_norm": 0.16834868490695953, + "learning_rate": 1.2934067273603856e-05, + "loss": 1.0448, + "step": 12617 + }, + { + "epoch": 0.7666322376815117, + "grad_norm": 11.661815643310547, + "learning_rate": 1.2927643155472774e-05, + "loss": 1.0463, + "step": 12618 + }, + { + "epoch": 0.7666929947141382, + "grad_norm": 0.24716760218143463, + "learning_rate": 1.29212203962361e-05, + "loss": 1.2316, + "step": 12619 + }, + { + "epoch": 0.7667537517467646, + "grad_norm": 0.1429780274629593, + "learning_rate": 1.2914798996129262e-05, + "loss": 1.0169, + "step": 12620 + }, + { + "epoch": 0.7668145087793912, + "grad_norm": 0.23687613010406494, + "learning_rate": 1.2908378955387634e-05, + "loss": 1.1476, + "step": 12621 + }, + { + "epoch": 0.7668752658120177, + "grad_norm": 0.1551547795534134, + "learning_rate": 1.2901960274246544e-05, + "loss": 0.9943, + "step": 12622 + }, + { + "epoch": 0.7669360228446442, + "grad_norm": 0.2121308147907257, + "learning_rate": 1.289554295294127e-05, + "loss": 1.1035, + "step": 12623 + }, + { + "epoch": 0.7669967798772708, + "grad_norm": 0.14433489739894867, + "learning_rate": 1.288912699170703e-05, + "loss": 1.0564, + "step": 12624 + }, + { + "epoch": 0.7670575369098973, + "grad_norm": 0.16125920414924622, + "learning_rate": 1.2882712390779022e-05, + "loss": 1.0376, + "step": 12625 + }, + { + "epoch": 0.7671182939425238, + "grad_norm": 0.1961602121591568, + "learning_rate": 1.287629915039239e-05, + "loss": 1.0698, + "step": 12626 + }, + { + "epoch": 0.7671790509751504, + "grad_norm": 1.2741436958312988, + "learning_rate": 1.2869887270782165e-05, + "loss": 1.1429, + "step": 12627 + }, + { + "epoch": 0.7672398080077769, + "grad_norm": 0.15076661109924316, + "learning_rate": 1.286347675218339e-05, + "loss": 1.0434, + "step": 12628 + }, + { + "epoch": 0.7673005650404034, + "grad_norm": 0.16556218266487122, + "learning_rate": 1.2857067594831063e-05, + "loss": 0.9676, + "step": 12629 + }, + { + "epoch": 0.76736132207303, + "grad_norm": 0.31363898515701294, + "learning_rate": 1.2850659798960107e-05, + "loss": 1.1758, + "step": 12630 + }, + { + "epoch": 0.7674220791056565, + "grad_norm": 0.176557257771492, + "learning_rate": 1.2844253364805398e-05, + "loss": 1.0464, + "step": 12631 + }, + { + "epoch": 0.767482836138283, + "grad_norm": 0.4316345751285553, + "learning_rate": 1.2837848292601755e-05, + "loss": 1.0975, + "step": 12632 + }, + { + "epoch": 0.7675435931709096, + "grad_norm": 0.23363591730594635, + "learning_rate": 1.2831444582583973e-05, + "loss": 1.0303, + "step": 12633 + }, + { + "epoch": 0.767604350203536, + "grad_norm": 0.1661541908979416, + "learning_rate": 1.2825042234986779e-05, + "loss": 1.0608, + "step": 12634 + }, + { + "epoch": 0.7676651072361625, + "grad_norm": 0.8694517612457275, + "learning_rate": 1.2818641250044838e-05, + "loss": 1.1047, + "step": 12635 + }, + { + "epoch": 0.7677258642687891, + "grad_norm": 0.21763907372951508, + "learning_rate": 1.2812241627992793e-05, + "loss": 1.1164, + "step": 12636 + }, + { + "epoch": 0.7677866213014156, + "grad_norm": 0.19925454258918762, + "learning_rate": 1.2805843369065223e-05, + "loss": 1.0242, + "step": 12637 + }, + { + "epoch": 0.7678473783340422, + "grad_norm": 0.21951137483119965, + "learning_rate": 1.2799446473496651e-05, + "loss": 1.0786, + "step": 12638 + }, + { + "epoch": 0.7679081353666687, + "grad_norm": 0.7524231672286987, + "learning_rate": 1.2793050941521556e-05, + "loss": 1.1314, + "step": 12639 + }, + { + "epoch": 0.7679688923992952, + "grad_norm": 0.2013801634311676, + "learning_rate": 1.2786656773374373e-05, + "loss": 1.2932, + "step": 12640 + }, + { + "epoch": 0.7680296494319218, + "grad_norm": 0.18967460095882416, + "learning_rate": 1.278026396928948e-05, + "loss": 1.0744, + "step": 12641 + }, + { + "epoch": 0.7680904064645483, + "grad_norm": 0.20179706811904907, + "learning_rate": 1.2773872529501207e-05, + "loss": 1.0947, + "step": 12642 + }, + { + "epoch": 0.7681511634971748, + "grad_norm": 0.17518985271453857, + "learning_rate": 1.2767482454243829e-05, + "loss": 1.0282, + "step": 12643 + }, + { + "epoch": 0.7682119205298014, + "grad_norm": 0.12890887260437012, + "learning_rate": 1.2761093743751568e-05, + "loss": 1.0017, + "step": 12644 + }, + { + "epoch": 0.7682726775624279, + "grad_norm": 0.19834722578525543, + "learning_rate": 1.2754706398258625e-05, + "loss": 1.1114, + "step": 12645 + }, + { + "epoch": 0.7683334345950544, + "grad_norm": 0.1358412206172943, + "learning_rate": 1.2748320417999132e-05, + "loss": 1.0017, + "step": 12646 + }, + { + "epoch": 0.7683941916276809, + "grad_norm": 0.25472673773765564, + "learning_rate": 1.2741935803207144e-05, + "loss": 1.1055, + "step": 12647 + }, + { + "epoch": 0.7684549486603074, + "grad_norm": 0.19049274921417236, + "learning_rate": 1.2735552554116698e-05, + "loss": 0.9992, + "step": 12648 + }, + { + "epoch": 0.7685157056929339, + "grad_norm": 0.28876855969429016, + "learning_rate": 1.272917067096176e-05, + "loss": 1.153, + "step": 12649 + }, + { + "epoch": 0.7685764627255605, + "grad_norm": 0.13519832491874695, + "learning_rate": 1.2722790153976294e-05, + "loss": 1.0455, + "step": 12650 + }, + { + "epoch": 0.768637219758187, + "grad_norm": 0.22434791922569275, + "learning_rate": 1.2716411003394157e-05, + "loss": 1.147, + "step": 12651 + }, + { + "epoch": 0.7686979767908135, + "grad_norm": 0.20125633478164673, + "learning_rate": 1.2710033219449185e-05, + "loss": 1.0073, + "step": 12652 + }, + { + "epoch": 0.7687587338234401, + "grad_norm": 0.16753990948200226, + "learning_rate": 1.2703656802375164e-05, + "loss": 1.0718, + "step": 12653 + }, + { + "epoch": 0.7688194908560666, + "grad_norm": 0.21424466371536255, + "learning_rate": 1.2697281752405787e-05, + "loss": 1.035, + "step": 12654 + }, + { + "epoch": 0.7688802478886931, + "grad_norm": 0.29243433475494385, + "learning_rate": 1.269090806977477e-05, + "loss": 1.1827, + "step": 12655 + }, + { + "epoch": 0.7689410049213197, + "grad_norm": 0.16744136810302734, + "learning_rate": 1.2684535754715732e-05, + "loss": 0.9542, + "step": 12656 + }, + { + "epoch": 0.7690017619539462, + "grad_norm": 0.14315524697303772, + "learning_rate": 1.2678164807462244e-05, + "loss": 1.0151, + "step": 12657 + }, + { + "epoch": 0.7690625189865727, + "grad_norm": 0.20562465488910675, + "learning_rate": 1.267179522824784e-05, + "loss": 1.0601, + "step": 12658 + }, + { + "epoch": 0.7691232760191993, + "grad_norm": 0.14764274656772614, + "learning_rate": 1.2665427017306002e-05, + "loss": 1.0825, + "step": 12659 + }, + { + "epoch": 0.7691840330518257, + "grad_norm": 0.22509382665157318, + "learning_rate": 1.2659060174870146e-05, + "loss": 1.194, + "step": 12660 + }, + { + "epoch": 0.7692447900844522, + "grad_norm": 0.24258023500442505, + "learning_rate": 1.2652694701173662e-05, + "loss": 1.1811, + "step": 12661 + }, + { + "epoch": 0.7693055471170788, + "grad_norm": 0.7514196038246155, + "learning_rate": 1.2646330596449873e-05, + "loss": 1.0712, + "step": 12662 + }, + { + "epoch": 0.7693663041497053, + "grad_norm": 0.16079604625701904, + "learning_rate": 1.2639967860932056e-05, + "loss": 1.0188, + "step": 12663 + }, + { + "epoch": 0.7694270611823318, + "grad_norm": 0.21602831780910492, + "learning_rate": 1.263360649485344e-05, + "loss": 1.1112, + "step": 12664 + }, + { + "epoch": 0.7694878182149584, + "grad_norm": 0.24385912716388702, + "learning_rate": 1.2627246498447182e-05, + "loss": 1.1428, + "step": 12665 + }, + { + "epoch": 0.7695485752475849, + "grad_norm": 0.1655386984348297, + "learning_rate": 1.2620887871946462e-05, + "loss": 1.0865, + "step": 12666 + }, + { + "epoch": 0.7696093322802114, + "grad_norm": 0.19480988383293152, + "learning_rate": 1.2614530615584303e-05, + "loss": 1.0863, + "step": 12667 + }, + { + "epoch": 0.769670089312838, + "grad_norm": 0.18022222816944122, + "learning_rate": 1.2608174729593753e-05, + "loss": 1.0657, + "step": 12668 + }, + { + "epoch": 0.7697308463454645, + "grad_norm": 0.2053578794002533, + "learning_rate": 1.2601820214207783e-05, + "loss": 1.1192, + "step": 12669 + }, + { + "epoch": 0.769791603378091, + "grad_norm": 0.13606128096580505, + "learning_rate": 1.2595467069659312e-05, + "loss": 1.0384, + "step": 12670 + }, + { + "epoch": 0.7698523604107176, + "grad_norm": 2.9864773750305176, + "learning_rate": 1.2589115296181237e-05, + "loss": 1.1974, + "step": 12671 + }, + { + "epoch": 0.7699131174433441, + "grad_norm": 0.23922699689865112, + "learning_rate": 1.2582764894006371e-05, + "loss": 1.1817, + "step": 12672 + }, + { + "epoch": 0.7699738744759705, + "grad_norm": 0.1395544707775116, + "learning_rate": 1.2576415863367508e-05, + "loss": 1.0142, + "step": 12673 + }, + { + "epoch": 0.7700346315085971, + "grad_norm": 0.13987411558628082, + "learning_rate": 1.2570068204497343e-05, + "loss": 1.0488, + "step": 12674 + }, + { + "epoch": 0.7700953885412236, + "grad_norm": 0.13964740931987762, + "learning_rate": 1.2563721917628546e-05, + "loss": 1.0365, + "step": 12675 + }, + { + "epoch": 0.7701561455738501, + "grad_norm": 0.1822194755077362, + "learning_rate": 1.255737700299377e-05, + "loss": 1.0578, + "step": 12676 + }, + { + "epoch": 0.7702169026064767, + "grad_norm": 0.8207707405090332, + "learning_rate": 1.2551033460825579e-05, + "loss": 1.0061, + "step": 12677 + }, + { + "epoch": 0.7702776596391032, + "grad_norm": 0.15862637758255005, + "learning_rate": 1.2544691291356497e-05, + "loss": 1.04, + "step": 12678 + }, + { + "epoch": 0.7703384166717298, + "grad_norm": 0.2402767688035965, + "learning_rate": 1.2538350494818996e-05, + "loss": 1.092, + "step": 12679 + }, + { + "epoch": 0.7703991737043563, + "grad_norm": 0.2024051696062088, + "learning_rate": 1.25320110714455e-05, + "loss": 1.0859, + "step": 12680 + }, + { + "epoch": 0.7704599307369828, + "grad_norm": 0.12398771941661835, + "learning_rate": 1.2525673021468376e-05, + "loss": 0.9886, + "step": 12681 + }, + { + "epoch": 0.7705206877696094, + "grad_norm": 0.6527822613716125, + "learning_rate": 1.2519336345119953e-05, + "loss": 1.0609, + "step": 12682 + }, + { + "epoch": 0.7705814448022359, + "grad_norm": 1.5776901245117188, + "learning_rate": 1.25130010426325e-05, + "loss": 1.0187, + "step": 12683 + }, + { + "epoch": 0.7706422018348624, + "grad_norm": 0.2468462437391281, + "learning_rate": 1.250666711423824e-05, + "loss": 1.193, + "step": 12684 + }, + { + "epoch": 0.770702958867489, + "grad_norm": 0.18418419361114502, + "learning_rate": 1.2500334560169347e-05, + "loss": 1.0906, + "step": 12685 + }, + { + "epoch": 0.7707637159001154, + "grad_norm": 1.829376220703125, + "learning_rate": 1.2494003380657937e-05, + "loss": 1.086, + "step": 12686 + }, + { + "epoch": 0.7708244729327419, + "grad_norm": 7.173383712768555, + "learning_rate": 1.2487673575936082e-05, + "loss": 1.0388, + "step": 12687 + }, + { + "epoch": 0.7708852299653685, + "grad_norm": 0.3399854004383087, + "learning_rate": 1.2481345146235807e-05, + "loss": 1.3064, + "step": 12688 + }, + { + "epoch": 0.770945986997995, + "grad_norm": 0.20304113626480103, + "learning_rate": 1.2475018091789075e-05, + "loss": 1.0986, + "step": 12689 + }, + { + "epoch": 0.7710067440306215, + "grad_norm": 0.1646611988544464, + "learning_rate": 1.246869241282781e-05, + "loss": 1.0619, + "step": 12690 + }, + { + "epoch": 0.7710675010632481, + "grad_norm": 0.5155800580978394, + "learning_rate": 1.2462368109583861e-05, + "loss": 1.1667, + "step": 12691 + }, + { + "epoch": 0.7711282580958746, + "grad_norm": 0.209584042429924, + "learning_rate": 1.2456045182289083e-05, + "loss": 1.1234, + "step": 12692 + }, + { + "epoch": 0.7711890151285011, + "grad_norm": 3.7057366371154785, + "learning_rate": 1.2449723631175237e-05, + "loss": 1.228, + "step": 12693 + }, + { + "epoch": 0.7712497721611277, + "grad_norm": 0.23990826308727264, + "learning_rate": 1.2443403456474017e-05, + "loss": 1.0294, + "step": 12694 + }, + { + "epoch": 0.7713105291937542, + "grad_norm": 0.13924945890903473, + "learning_rate": 1.2437084658417104e-05, + "loss": 1.0082, + "step": 12695 + }, + { + "epoch": 0.7713712862263807, + "grad_norm": 0.4962828755378723, + "learning_rate": 1.2430767237236096e-05, + "loss": 1.0764, + "step": 12696 + }, + { + "epoch": 0.7714320432590073, + "grad_norm": 0.26206403970718384, + "learning_rate": 1.2424451193162594e-05, + "loss": 1.0254, + "step": 12697 + }, + { + "epoch": 0.7714928002916338, + "grad_norm": 0.2997289299964905, + "learning_rate": 1.2418136526428098e-05, + "loss": 1.1666, + "step": 12698 + }, + { + "epoch": 0.7715535573242602, + "grad_norm": 0.1936078518629074, + "learning_rate": 1.2411823237264069e-05, + "loss": 1.116, + "step": 12699 + }, + { + "epoch": 0.7716143143568868, + "grad_norm": 0.21960070729255676, + "learning_rate": 1.2405511325901942e-05, + "loss": 1.259, + "step": 12700 + }, + { + "epoch": 0.7716750713895133, + "grad_norm": 0.14493779838085175, + "learning_rate": 1.239920079257303e-05, + "loss": 0.9811, + "step": 12701 + }, + { + "epoch": 0.7717358284221398, + "grad_norm": 0.3576783835887909, + "learning_rate": 1.2392891637508702e-05, + "loss": 1.1068, + "step": 12702 + }, + { + "epoch": 0.7717965854547664, + "grad_norm": 0.17253521084785461, + "learning_rate": 1.2386583860940198e-05, + "loss": 1.0417, + "step": 12703 + }, + { + "epoch": 0.7718573424873929, + "grad_norm": 0.19465260207653046, + "learning_rate": 1.2380277463098733e-05, + "loss": 1.0805, + "step": 12704 + }, + { + "epoch": 0.7719180995200194, + "grad_norm": 0.21865147352218628, + "learning_rate": 1.2373972444215464e-05, + "loss": 1.1365, + "step": 12705 + }, + { + "epoch": 0.771978856552646, + "grad_norm": 1.6978946924209595, + "learning_rate": 1.2367668804521515e-05, + "loss": 1.2326, + "step": 12706 + }, + { + "epoch": 0.7720396135852725, + "grad_norm": 0.9822807312011719, + "learning_rate": 1.2361366544247937e-05, + "loss": 1.153, + "step": 12707 + }, + { + "epoch": 0.772100370617899, + "grad_norm": 0.24755197763442993, + "learning_rate": 1.2355065663625747e-05, + "loss": 1.065, + "step": 12708 + }, + { + "epoch": 0.7721611276505256, + "grad_norm": 0.2357681393623352, + "learning_rate": 1.23487661628859e-05, + "loss": 1.0823, + "step": 12709 + }, + { + "epoch": 0.7722218846831521, + "grad_norm": 0.24175310134887695, + "learning_rate": 1.2342468042259309e-05, + "loss": 1.1038, + "step": 12710 + }, + { + "epoch": 0.7722826417157787, + "grad_norm": 0.139458566904068, + "learning_rate": 1.2336171301976824e-05, + "loss": 1.0314, + "step": 12711 + }, + { + "epoch": 0.7723433987484051, + "grad_norm": 0.2423827350139618, + "learning_rate": 1.2329875942269253e-05, + "loss": 1.08, + "step": 12712 + }, + { + "epoch": 0.7724041557810316, + "grad_norm": 0.14969079196453094, + "learning_rate": 1.2323581963367386e-05, + "loss": 1.0476, + "step": 12713 + }, + { + "epoch": 0.7724649128136581, + "grad_norm": 0.19709183275699615, + "learning_rate": 1.2317289365501894e-05, + "loss": 1.0757, + "step": 12714 + }, + { + "epoch": 0.7725256698462847, + "grad_norm": 1.8492051362991333, + "learning_rate": 1.2310998148903441e-05, + "loss": 1.08, + "step": 12715 + }, + { + "epoch": 0.7725864268789112, + "grad_norm": 0.19617119431495667, + "learning_rate": 1.2304708313802632e-05, + "loss": 1.044, + "step": 12716 + }, + { + "epoch": 0.7726471839115377, + "grad_norm": 2.5987331867218018, + "learning_rate": 1.2298419860430016e-05, + "loss": 0.9942, + "step": 12717 + }, + { + "epoch": 0.7727079409441643, + "grad_norm": 0.13710376620292664, + "learning_rate": 1.2292132789016125e-05, + "loss": 1.019, + "step": 12718 + }, + { + "epoch": 0.7727686979767908, + "grad_norm": 0.17622964084148407, + "learning_rate": 1.2285847099791391e-05, + "loss": 1.1331, + "step": 12719 + }, + { + "epoch": 0.7728294550094174, + "grad_norm": 0.2079898864030838, + "learning_rate": 1.2279562792986238e-05, + "loss": 1.0009, + "step": 12720 + }, + { + "epoch": 0.7728902120420439, + "grad_norm": 0.578728437423706, + "learning_rate": 1.2273279868830973e-05, + "loss": 1.1084, + "step": 12721 + }, + { + "epoch": 0.7729509690746704, + "grad_norm": 0.2308521568775177, + "learning_rate": 1.2266998327555939e-05, + "loss": 1.036, + "step": 12722 + }, + { + "epoch": 0.773011726107297, + "grad_norm": 0.8228200078010559, + "learning_rate": 1.2260718169391377e-05, + "loss": 1.1543, + "step": 12723 + }, + { + "epoch": 0.7730724831399235, + "grad_norm": 0.23016905784606934, + "learning_rate": 1.225443939456749e-05, + "loss": 1.26, + "step": 12724 + }, + { + "epoch": 0.7731332401725499, + "grad_norm": 0.1707516461610794, + "learning_rate": 1.2248162003314418e-05, + "loss": 1.0304, + "step": 12725 + }, + { + "epoch": 0.7731939972051765, + "grad_norm": 0.2643856108188629, + "learning_rate": 1.2241885995862262e-05, + "loss": 1.1139, + "step": 12726 + }, + { + "epoch": 0.773254754237803, + "grad_norm": 0.23546276986598969, + "learning_rate": 1.223561137244108e-05, + "loss": 1.1348, + "step": 12727 + }, + { + "epoch": 0.7733155112704295, + "grad_norm": 0.1173090934753418, + "learning_rate": 1.222933813328086e-05, + "loss": 0.9697, + "step": 12728 + }, + { + "epoch": 0.7733762683030561, + "grad_norm": 0.1456197202205658, + "learning_rate": 1.222306627861155e-05, + "loss": 1.0885, + "step": 12729 + }, + { + "epoch": 0.7734370253356826, + "grad_norm": 0.13828636705875397, + "learning_rate": 1.221679580866305e-05, + "loss": 1.0179, + "step": 12730 + }, + { + "epoch": 0.7734977823683091, + "grad_norm": 0.16994591057300568, + "learning_rate": 1.22105267236652e-05, + "loss": 1.0966, + "step": 12731 + }, + { + "epoch": 0.7735585394009357, + "grad_norm": 0.1895648092031479, + "learning_rate": 1.2204259023847796e-05, + "loss": 1.0971, + "step": 12732 + }, + { + "epoch": 0.7736192964335622, + "grad_norm": 0.1948980838060379, + "learning_rate": 1.2197992709440592e-05, + "loss": 1.1203, + "step": 12733 + }, + { + "epoch": 0.7736800534661887, + "grad_norm": 0.2512350380420685, + "learning_rate": 1.2191727780673262e-05, + "loss": 1.1205, + "step": 12734 + }, + { + "epoch": 0.7737408104988153, + "grad_norm": 0.12344150245189667, + "learning_rate": 1.2185464237775467e-05, + "loss": 1.0426, + "step": 12735 + }, + { + "epoch": 0.7738015675314418, + "grad_norm": 0.18382316827774048, + "learning_rate": 1.217920208097678e-05, + "loss": 1.0423, + "step": 12736 + }, + { + "epoch": 0.7738623245640683, + "grad_norm": 0.2553408741950989, + "learning_rate": 1.2172941310506748e-05, + "loss": 1.0807, + "step": 12737 + }, + { + "epoch": 0.7739230815966949, + "grad_norm": 0.2251313030719757, + "learning_rate": 1.216668192659487e-05, + "loss": 1.0173, + "step": 12738 + }, + { + "epoch": 0.7739838386293213, + "grad_norm": 0.17356465756893158, + "learning_rate": 1.2160423929470582e-05, + "loss": 1.0715, + "step": 12739 + }, + { + "epoch": 0.7740445956619478, + "grad_norm": 0.1568404883146286, + "learning_rate": 1.2154167319363286e-05, + "loss": 0.9737, + "step": 12740 + }, + { + "epoch": 0.7741053526945744, + "grad_norm": 0.5185936689376831, + "learning_rate": 1.2147912096502285e-05, + "loss": 1.0369, + "step": 12741 + }, + { + "epoch": 0.7741661097272009, + "grad_norm": 1.6018164157867432, + "learning_rate": 1.2141658261116867e-05, + "loss": 1.0532, + "step": 12742 + }, + { + "epoch": 0.7742268667598274, + "grad_norm": 0.1557290107011795, + "learning_rate": 1.2135405813436301e-05, + "loss": 1.0627, + "step": 12743 + }, + { + "epoch": 0.774287623792454, + "grad_norm": 0.14853310585021973, + "learning_rate": 1.2129154753689754e-05, + "loss": 1.1059, + "step": 12744 + }, + { + "epoch": 0.7743483808250805, + "grad_norm": 0.14633236825466156, + "learning_rate": 1.2122905082106357e-05, + "loss": 1.0171, + "step": 12745 + }, + { + "epoch": 0.774409137857707, + "grad_norm": 0.20377807319164276, + "learning_rate": 1.21166567989152e-05, + "loss": 1.2629, + "step": 12746 + }, + { + "epoch": 0.7744698948903336, + "grad_norm": 0.18756474554538727, + "learning_rate": 1.2110409904345305e-05, + "loss": 1.1194, + "step": 12747 + }, + { + "epoch": 0.7745306519229601, + "grad_norm": 0.19438649713993073, + "learning_rate": 1.2104164398625662e-05, + "loss": 1.127, + "step": 12748 + }, + { + "epoch": 0.7745914089555866, + "grad_norm": 0.1491466760635376, + "learning_rate": 1.2097920281985198e-05, + "loss": 1.0527, + "step": 12749 + }, + { + "epoch": 0.7746521659882132, + "grad_norm": 0.15273001790046692, + "learning_rate": 1.2091677554652792e-05, + "loss": 1.1116, + "step": 12750 + }, + { + "epoch": 0.7747129230208397, + "grad_norm": 0.12652656435966492, + "learning_rate": 1.2085436216857277e-05, + "loss": 0.9958, + "step": 12751 + }, + { + "epoch": 0.7747736800534661, + "grad_norm": 0.13640187680721283, + "learning_rate": 1.2079196268827419e-05, + "loss": 1.0352, + "step": 12752 + }, + { + "epoch": 0.7748344370860927, + "grad_norm": 0.18746574223041534, + "learning_rate": 1.2072957710791955e-05, + "loss": 1.0042, + "step": 12753 + }, + { + "epoch": 0.7748951941187192, + "grad_norm": 0.2179529219865799, + "learning_rate": 1.206672054297956e-05, + "loss": 1.1643, + "step": 12754 + }, + { + "epoch": 0.7749559511513457, + "grad_norm": 0.3338319659233093, + "learning_rate": 1.2060484765618851e-05, + "loss": 1.2326, + "step": 12755 + }, + { + "epoch": 0.7750167081839723, + "grad_norm": 0.181010439991951, + "learning_rate": 1.205425037893841e-05, + "loss": 1.0376, + "step": 12756 + }, + { + "epoch": 0.7750774652165988, + "grad_norm": 0.14557655155658722, + "learning_rate": 1.2048017383166753e-05, + "loss": 0.9751, + "step": 12757 + }, + { + "epoch": 0.7751382222492254, + "grad_norm": 0.13821597397327423, + "learning_rate": 1.2041785778532339e-05, + "loss": 1.0189, + "step": 12758 + }, + { + "epoch": 0.7751989792818519, + "grad_norm": 0.18181969225406647, + "learning_rate": 1.2035555565263623e-05, + "loss": 1.0965, + "step": 12759 + }, + { + "epoch": 0.7752597363144784, + "grad_norm": 0.14281710982322693, + "learning_rate": 1.202932674358897e-05, + "loss": 1.0465, + "step": 12760 + }, + { + "epoch": 0.775320493347105, + "grad_norm": 0.18732373416423798, + "learning_rate": 1.202309931373667e-05, + "loss": 1.0257, + "step": 12761 + }, + { + "epoch": 0.7753812503797315, + "grad_norm": 0.12003648281097412, + "learning_rate": 1.2016873275935004e-05, + "loss": 1.0153, + "step": 12762 + }, + { + "epoch": 0.775442007412358, + "grad_norm": 0.1572399139404297, + "learning_rate": 1.2010648630412175e-05, + "loss": 1.0505, + "step": 12763 + }, + { + "epoch": 0.7755027644449846, + "grad_norm": 0.2403363138437271, + "learning_rate": 1.2004425377396383e-05, + "loss": 1.1641, + "step": 12764 + }, + { + "epoch": 0.775563521477611, + "grad_norm": 0.17444270849227905, + "learning_rate": 1.199820351711572e-05, + "loss": 1.0718, + "step": 12765 + }, + { + "epoch": 0.7756242785102375, + "grad_norm": 0.1672670841217041, + "learning_rate": 1.199198304979825e-05, + "loss": 1.0727, + "step": 12766 + }, + { + "epoch": 0.775685035542864, + "grad_norm": 2.492030620574951, + "learning_rate": 1.1985763975672004e-05, + "loss": 0.9946, + "step": 12767 + }, + { + "epoch": 0.7757457925754906, + "grad_norm": 0.29291072487831116, + "learning_rate": 1.1979546294964899e-05, + "loss": 1.2147, + "step": 12768 + }, + { + "epoch": 0.7758065496081171, + "grad_norm": 0.19696073234081268, + "learning_rate": 1.1973330007904892e-05, + "loss": 1.1234, + "step": 12769 + }, + { + "epoch": 0.7758673066407437, + "grad_norm": 0.18862523138523102, + "learning_rate": 1.1967115114719818e-05, + "loss": 0.9919, + "step": 12770 + }, + { + "epoch": 0.7759280636733702, + "grad_norm": 0.34294137358665466, + "learning_rate": 1.1960901615637494e-05, + "loss": 1.1353, + "step": 12771 + }, + { + "epoch": 0.7759888207059967, + "grad_norm": 0.17390456795692444, + "learning_rate": 1.1954689510885675e-05, + "loss": 1.0821, + "step": 12772 + }, + { + "epoch": 0.7760495777386233, + "grad_norm": 0.197773277759552, + "learning_rate": 1.1948478800692065e-05, + "loss": 1.1015, + "step": 12773 + }, + { + "epoch": 0.7761103347712498, + "grad_norm": 0.14509107172489166, + "learning_rate": 1.194226948528432e-05, + "loss": 1.0324, + "step": 12774 + }, + { + "epoch": 0.7761710918038763, + "grad_norm": 0.15422256290912628, + "learning_rate": 1.1936061564890044e-05, + "loss": 1.0256, + "step": 12775 + }, + { + "epoch": 0.7762318488365029, + "grad_norm": 0.15667399764060974, + "learning_rate": 1.1929855039736786e-05, + "loss": 1.0422, + "step": 12776 + }, + { + "epoch": 0.7762926058691294, + "grad_norm": 0.2082609385251999, + "learning_rate": 1.1923649910052054e-05, + "loss": 1.0968, + "step": 12777 + }, + { + "epoch": 0.7763533629017558, + "grad_norm": 0.20585180819034576, + "learning_rate": 1.191744617606329e-05, + "loss": 1.0893, + "step": 12778 + }, + { + "epoch": 0.7764141199343824, + "grad_norm": 0.19863329827785492, + "learning_rate": 1.1911243837997887e-05, + "loss": 1.088, + "step": 12779 + }, + { + "epoch": 0.7764748769670089, + "grad_norm": 0.15365231037139893, + "learning_rate": 1.1905042896083235e-05, + "loss": 1.081, + "step": 12780 + }, + { + "epoch": 0.7765356339996354, + "grad_norm": 0.222613126039505, + "learning_rate": 1.1898843350546579e-05, + "loss": 1.1622, + "step": 12781 + }, + { + "epoch": 0.776596391032262, + "grad_norm": 0.17033858597278595, + "learning_rate": 1.1892645201615188e-05, + "loss": 1.0851, + "step": 12782 + }, + { + "epoch": 0.7766571480648885, + "grad_norm": 0.1271793097257614, + "learning_rate": 1.1886448449516251e-05, + "loss": 1.0049, + "step": 12783 + }, + { + "epoch": 0.776717905097515, + "grad_norm": 0.2466173619031906, + "learning_rate": 1.1880253094476901e-05, + "loss": 1.0342, + "step": 12784 + }, + { + "epoch": 0.7767786621301416, + "grad_norm": 0.20461808145046234, + "learning_rate": 1.1874059136724253e-05, + "loss": 1.0653, + "step": 12785 + }, + { + "epoch": 0.7768394191627681, + "grad_norm": 0.3232489824295044, + "learning_rate": 1.1867866576485342e-05, + "loss": 1.2115, + "step": 12786 + }, + { + "epoch": 0.7769001761953946, + "grad_norm": 0.1660570502281189, + "learning_rate": 1.1861675413987162e-05, + "loss": 1.0677, + "step": 12787 + }, + { + "epoch": 0.7769609332280212, + "grad_norm": 0.18697388470172882, + "learning_rate": 1.1855485649456633e-05, + "loss": 1.1331, + "step": 12788 + }, + { + "epoch": 0.7770216902606477, + "grad_norm": 0.19730299711227417, + "learning_rate": 1.184929728312063e-05, + "loss": 1.0843, + "step": 12789 + }, + { + "epoch": 0.7770824472932742, + "grad_norm": 0.1668861359357834, + "learning_rate": 1.184311031520603e-05, + "loss": 1.0098, + "step": 12790 + }, + { + "epoch": 0.7771432043259007, + "grad_norm": 0.23333723843097687, + "learning_rate": 1.1836924745939593e-05, + "loss": 1.1329, + "step": 12791 + }, + { + "epoch": 0.7772039613585272, + "grad_norm": 0.15302376449108124, + "learning_rate": 1.1830740575548055e-05, + "loss": 1.0918, + "step": 12792 + }, + { + "epoch": 0.7772647183911537, + "grad_norm": 0.22767987847328186, + "learning_rate": 1.1824557804258101e-05, + "loss": 1.1105, + "step": 12793 + }, + { + "epoch": 0.7773254754237803, + "grad_norm": 0.207039937376976, + "learning_rate": 1.1818376432296358e-05, + "loss": 1.0976, + "step": 12794 + }, + { + "epoch": 0.7773862324564068, + "grad_norm": 0.2451179027557373, + "learning_rate": 1.1812196459889413e-05, + "loss": 1.2209, + "step": 12795 + }, + { + "epoch": 0.7774469894890333, + "grad_norm": 4.473360061645508, + "learning_rate": 1.1806017887263782e-05, + "loss": 1.0589, + "step": 12796 + }, + { + "epoch": 0.7775077465216599, + "grad_norm": 0.28177523612976074, + "learning_rate": 1.179984071464595e-05, + "loss": 1.0592, + "step": 12797 + }, + { + "epoch": 0.7775685035542864, + "grad_norm": 0.1926187127828598, + "learning_rate": 1.179366494226234e-05, + "loss": 1.0853, + "step": 12798 + }, + { + "epoch": 0.777629260586913, + "grad_norm": 0.2756271958351135, + "learning_rate": 1.1787490570339326e-05, + "loss": 1.1968, + "step": 12799 + }, + { + "epoch": 0.7776900176195395, + "grad_norm": 0.1432493031024933, + "learning_rate": 1.1781317599103236e-05, + "loss": 1.0136, + "step": 12800 + }, + { + "epoch": 0.777750774652166, + "grad_norm": 0.18142318725585938, + "learning_rate": 1.1775146028780332e-05, + "loss": 1.1109, + "step": 12801 + }, + { + "epoch": 0.7778115316847926, + "grad_norm": 0.14041787385940552, + "learning_rate": 1.1768975859596843e-05, + "loss": 1.0336, + "step": 12802 + }, + { + "epoch": 0.7778722887174191, + "grad_norm": 0.21915173530578613, + "learning_rate": 1.1762807091778933e-05, + "loss": 1.0639, + "step": 12803 + }, + { + "epoch": 0.7779330457500455, + "grad_norm": 0.1835729479789734, + "learning_rate": 1.175663972555272e-05, + "loss": 1.096, + "step": 12804 + }, + { + "epoch": 0.777993802782672, + "grad_norm": 0.164793461561203, + "learning_rate": 1.1750473761144254e-05, + "loss": 1.0392, + "step": 12805 + }, + { + "epoch": 0.7780545598152986, + "grad_norm": 1.603026032447815, + "learning_rate": 1.174430919877958e-05, + "loss": 1.006, + "step": 12806 + }, + { + "epoch": 0.7781153168479251, + "grad_norm": 0.41648948192596436, + "learning_rate": 1.1738146038684662e-05, + "loss": 1.2888, + "step": 12807 + }, + { + "epoch": 0.7781760738805517, + "grad_norm": 0.16222132742404938, + "learning_rate": 1.1731984281085385e-05, + "loss": 1.0412, + "step": 12808 + }, + { + "epoch": 0.7782368309131782, + "grad_norm": 0.15327848494052887, + "learning_rate": 1.1725823926207612e-05, + "loss": 1.0065, + "step": 12809 + }, + { + "epoch": 0.7782975879458047, + "grad_norm": 0.1405220776796341, + "learning_rate": 1.1719664974277156e-05, + "loss": 1.0269, + "step": 12810 + }, + { + "epoch": 0.7783583449784313, + "grad_norm": 0.18382728099822998, + "learning_rate": 1.1713507425519793e-05, + "loss": 1.1189, + "step": 12811 + }, + { + "epoch": 0.7784191020110578, + "grad_norm": 0.210740864276886, + "learning_rate": 1.1707351280161216e-05, + "loss": 1.1991, + "step": 12812 + }, + { + "epoch": 0.7784798590436843, + "grad_norm": 0.23782692849636078, + "learning_rate": 1.1701196538427079e-05, + "loss": 1.1494, + "step": 12813 + }, + { + "epoch": 0.7785406160763109, + "grad_norm": 0.13225312530994415, + "learning_rate": 1.1695043200542982e-05, + "loss": 1.0072, + "step": 12814 + }, + { + "epoch": 0.7786013731089374, + "grad_norm": 0.12108033895492554, + "learning_rate": 1.1688891266734486e-05, + "loss": 1.0378, + "step": 12815 + }, + { + "epoch": 0.7786621301415639, + "grad_norm": 0.4010210335254669, + "learning_rate": 1.1682740737227083e-05, + "loss": 1.1048, + "step": 12816 + }, + { + "epoch": 0.7787228871741904, + "grad_norm": 0.2249818593263626, + "learning_rate": 1.1676591612246224e-05, + "loss": 1.0446, + "step": 12817 + }, + { + "epoch": 0.7787836442068169, + "grad_norm": 0.22185011208057404, + "learning_rate": 1.1670443892017314e-05, + "loss": 1.1528, + "step": 12818 + }, + { + "epoch": 0.7788444012394434, + "grad_norm": 0.24274806678295135, + "learning_rate": 1.1664297576765687e-05, + "loss": 1.1416, + "step": 12819 + }, + { + "epoch": 0.77890515827207, + "grad_norm": 0.1520518958568573, + "learning_rate": 1.1658152666716638e-05, + "loss": 1.0134, + "step": 12820 + }, + { + "epoch": 0.7789659153046965, + "grad_norm": 0.6281973719596863, + "learning_rate": 1.1652009162095424e-05, + "loss": 1.1456, + "step": 12821 + }, + { + "epoch": 0.779026672337323, + "grad_norm": 0.13250429928302765, + "learning_rate": 1.1645867063127219e-05, + "loss": 1.0163, + "step": 12822 + }, + { + "epoch": 0.7790874293699496, + "grad_norm": 0.2489403486251831, + "learning_rate": 1.1639726370037174e-05, + "loss": 1.016, + "step": 12823 + }, + { + "epoch": 0.7791481864025761, + "grad_norm": 0.2153848558664322, + "learning_rate": 1.1633587083050374e-05, + "loss": 1.1401, + "step": 12824 + }, + { + "epoch": 0.7792089434352026, + "grad_norm": 0.20696254074573517, + "learning_rate": 1.1627449202391849e-05, + "loss": 1.0268, + "step": 12825 + }, + { + "epoch": 0.7792697004678292, + "grad_norm": 0.1527950018644333, + "learning_rate": 1.1621312728286581e-05, + "loss": 1.0376, + "step": 12826 + }, + { + "epoch": 0.7793304575004557, + "grad_norm": 0.30250343680381775, + "learning_rate": 1.1615177660959547e-05, + "loss": 1.103, + "step": 12827 + }, + { + "epoch": 0.7793912145330822, + "grad_norm": 0.15964163839817047, + "learning_rate": 1.1609044000635577e-05, + "loss": 1.0302, + "step": 12828 + }, + { + "epoch": 0.7794519715657088, + "grad_norm": 0.12987543642520905, + "learning_rate": 1.160291174753952e-05, + "loss": 1.0481, + "step": 12829 + }, + { + "epoch": 0.7795127285983352, + "grad_norm": 0.21817538142204285, + "learning_rate": 1.1596780901896138e-05, + "loss": 1.0825, + "step": 12830 + }, + { + "epoch": 0.7795734856309617, + "grad_norm": 0.22308793663978577, + "learning_rate": 1.1590651463930191e-05, + "loss": 1.0925, + "step": 12831 + }, + { + "epoch": 0.7796342426635883, + "grad_norm": 0.2033763974905014, + "learning_rate": 1.158452343386634e-05, + "loss": 0.9828, + "step": 12832 + }, + { + "epoch": 0.7796949996962148, + "grad_norm": 0.7395886182785034, + "learning_rate": 1.157839681192921e-05, + "loss": 1.2438, + "step": 12833 + }, + { + "epoch": 0.7797557567288413, + "grad_norm": 0.17386341094970703, + "learning_rate": 1.1572271598343381e-05, + "loss": 1.0586, + "step": 12834 + }, + { + "epoch": 0.7798165137614679, + "grad_norm": 0.18324099481105804, + "learning_rate": 1.1566147793333343e-05, + "loss": 1.0765, + "step": 12835 + }, + { + "epoch": 0.7798772707940944, + "grad_norm": 0.20710918307304382, + "learning_rate": 1.1560025397123592e-05, + "loss": 1.1381, + "step": 12836 + }, + { + "epoch": 0.779938027826721, + "grad_norm": 0.3197503685951233, + "learning_rate": 1.1553904409938548e-05, + "loss": 1.063, + "step": 12837 + }, + { + "epoch": 0.7799987848593475, + "grad_norm": 0.17713426053524017, + "learning_rate": 1.1547784832002567e-05, + "loss": 1.0525, + "step": 12838 + }, + { + "epoch": 0.780059541891974, + "grad_norm": 0.15182489156723022, + "learning_rate": 1.1541666663539963e-05, + "loss": 0.9802, + "step": 12839 + }, + { + "epoch": 0.7801202989246006, + "grad_norm": 0.17221994698047638, + "learning_rate": 1.1535549904775006e-05, + "loss": 1.1004, + "step": 12840 + }, + { + "epoch": 0.7801810559572271, + "grad_norm": 0.2795870006084442, + "learning_rate": 1.15294345559319e-05, + "loss": 1.0667, + "step": 12841 + }, + { + "epoch": 0.7802418129898536, + "grad_norm": 0.4730730950832367, + "learning_rate": 1.1523320617234806e-05, + "loss": 1.1184, + "step": 12842 + }, + { + "epoch": 0.7803025700224802, + "grad_norm": 0.15566584467887878, + "learning_rate": 1.1517208088907833e-05, + "loss": 1.0317, + "step": 12843 + }, + { + "epoch": 0.7803633270551066, + "grad_norm": 0.14927321672439575, + "learning_rate": 1.1511096971175034e-05, + "loss": 1.0545, + "step": 12844 + }, + { + "epoch": 0.7804240840877331, + "grad_norm": 0.13444799184799194, + "learning_rate": 1.1504987264260414e-05, + "loss": 1.0496, + "step": 12845 + }, + { + "epoch": 0.7804848411203597, + "grad_norm": 0.1637629270553589, + "learning_rate": 1.1498878968387916e-05, + "loss": 1.0804, + "step": 12846 + }, + { + "epoch": 0.7805455981529862, + "grad_norm": 0.24465370178222656, + "learning_rate": 1.1492772083781478e-05, + "loss": 1.01, + "step": 12847 + }, + { + "epoch": 0.7806063551856127, + "grad_norm": 0.5342153906822205, + "learning_rate": 1.1486666610664908e-05, + "loss": 1.135, + "step": 12848 + }, + { + "epoch": 0.7806671122182393, + "grad_norm": 0.15922164916992188, + "learning_rate": 1.1480562549262019e-05, + "loss": 1.0144, + "step": 12849 + }, + { + "epoch": 0.7807278692508658, + "grad_norm": 0.18404781818389893, + "learning_rate": 1.1474459899796559e-05, + "loss": 1.0046, + "step": 12850 + }, + { + "epoch": 0.7807886262834923, + "grad_norm": 0.20605361461639404, + "learning_rate": 1.1468358662492202e-05, + "loss": 1.0144, + "step": 12851 + }, + { + "epoch": 0.7808493833161189, + "grad_norm": 0.3070259690284729, + "learning_rate": 1.1462258837572615e-05, + "loss": 1.0438, + "step": 12852 + }, + { + "epoch": 0.7809101403487454, + "grad_norm": 0.13349731266498566, + "learning_rate": 1.1456160425261386e-05, + "loss": 1.0165, + "step": 12853 + }, + { + "epoch": 0.7809708973813719, + "grad_norm": 0.16883695125579834, + "learning_rate": 1.145006342578206e-05, + "loss": 1.0347, + "step": 12854 + }, + { + "epoch": 0.7810316544139985, + "grad_norm": 0.26615482568740845, + "learning_rate": 1.1443967839358094e-05, + "loss": 1.1229, + "step": 12855 + }, + { + "epoch": 0.781092411446625, + "grad_norm": 0.46201959252357483, + "learning_rate": 1.1437873666212928e-05, + "loss": 1.0904, + "step": 12856 + }, + { + "epoch": 0.7811531684792514, + "grad_norm": 0.15440763533115387, + "learning_rate": 1.1431780906569966e-05, + "loss": 1.0631, + "step": 12857 + }, + { + "epoch": 0.781213925511878, + "grad_norm": 0.18126122653484344, + "learning_rate": 1.1425689560652536e-05, + "loss": 1.0768, + "step": 12858 + }, + { + "epoch": 0.7812746825445045, + "grad_norm": 0.15048764646053314, + "learning_rate": 1.1419599628683909e-05, + "loss": 1.0787, + "step": 12859 + }, + { + "epoch": 0.781335439577131, + "grad_norm": 0.2822747230529785, + "learning_rate": 1.1413511110887314e-05, + "loss": 1.0675, + "step": 12860 + }, + { + "epoch": 0.7813961966097576, + "grad_norm": 0.13867425918579102, + "learning_rate": 1.1407424007485929e-05, + "loss": 0.9799, + "step": 12861 + }, + { + "epoch": 0.7814569536423841, + "grad_norm": 0.22998060286045074, + "learning_rate": 1.1401338318702876e-05, + "loss": 1.0121, + "step": 12862 + }, + { + "epoch": 0.7815177106750106, + "grad_norm": 0.17861399054527283, + "learning_rate": 1.1395254044761227e-05, + "loss": 1.1013, + "step": 12863 + }, + { + "epoch": 0.7815784677076372, + "grad_norm": 0.17674842476844788, + "learning_rate": 1.1389171185884006e-05, + "loss": 1.1025, + "step": 12864 + }, + { + "epoch": 0.7816392247402637, + "grad_norm": 0.2055133879184723, + "learning_rate": 1.1383089742294179e-05, + "loss": 1.1331, + "step": 12865 + }, + { + "epoch": 0.7816999817728902, + "grad_norm": 0.32318824529647827, + "learning_rate": 1.1377009714214658e-05, + "loss": 1.1074, + "step": 12866 + }, + { + "epoch": 0.7817607388055168, + "grad_norm": 0.9395774006843567, + "learning_rate": 1.1370931101868315e-05, + "loss": 1.0715, + "step": 12867 + }, + { + "epoch": 0.7818214958381433, + "grad_norm": 0.11810673028230667, + "learning_rate": 1.1364853905477956e-05, + "loss": 1.0127, + "step": 12868 + }, + { + "epoch": 0.7818822528707698, + "grad_norm": 0.19320668280124664, + "learning_rate": 1.1358778125266344e-05, + "loss": 1.1191, + "step": 12869 + }, + { + "epoch": 0.7819430099033963, + "grad_norm": 0.554922878742218, + "learning_rate": 1.1352703761456196e-05, + "loss": 1.1517, + "step": 12870 + }, + { + "epoch": 0.7820037669360228, + "grad_norm": 0.19171680510044098, + "learning_rate": 1.1346630814270154e-05, + "loss": 1.0756, + "step": 12871 + }, + { + "epoch": 0.7820645239686493, + "grad_norm": 0.14509311318397522, + "learning_rate": 1.1340559283930823e-05, + "loss": 1.0557, + "step": 12872 + }, + { + "epoch": 0.7821252810012759, + "grad_norm": 0.16991806030273438, + "learning_rate": 1.1334489170660779e-05, + "loss": 1.0556, + "step": 12873 + }, + { + "epoch": 0.7821860380339024, + "grad_norm": 0.30027997493743896, + "learning_rate": 1.1328420474682516e-05, + "loss": 1.1818, + "step": 12874 + }, + { + "epoch": 0.7822467950665289, + "grad_norm": 0.197781503200531, + "learning_rate": 1.1322353196218465e-05, + "loss": 1.0407, + "step": 12875 + }, + { + "epoch": 0.7823075520991555, + "grad_norm": 0.3027670681476593, + "learning_rate": 1.1316287335491032e-05, + "loss": 1.082, + "step": 12876 + }, + { + "epoch": 0.782368309131782, + "grad_norm": 0.1760532259941101, + "learning_rate": 1.1310222892722556e-05, + "loss": 1.1299, + "step": 12877 + }, + { + "epoch": 0.7824290661644085, + "grad_norm": 0.2045224905014038, + "learning_rate": 1.1304159868135344e-05, + "loss": 1.0964, + "step": 12878 + }, + { + "epoch": 0.7824898231970351, + "grad_norm": 0.228510320186615, + "learning_rate": 1.129809826195164e-05, + "loss": 1.1317, + "step": 12879 + }, + { + "epoch": 0.7825505802296616, + "grad_norm": 0.26512598991394043, + "learning_rate": 1.1292038074393618e-05, + "loss": 1.2146, + "step": 12880 + }, + { + "epoch": 0.7826113372622882, + "grad_norm": 0.24852553009986877, + "learning_rate": 1.1285979305683441e-05, + "loss": 1.0834, + "step": 12881 + }, + { + "epoch": 0.7826720942949147, + "grad_norm": 0.2732764482498169, + "learning_rate": 1.1279921956043144e-05, + "loss": 1.0748, + "step": 12882 + }, + { + "epoch": 0.7827328513275411, + "grad_norm": 0.1470334529876709, + "learning_rate": 1.1273866025694813e-05, + "loss": 1.061, + "step": 12883 + }, + { + "epoch": 0.7827936083601676, + "grad_norm": 0.22543005645275116, + "learning_rate": 1.1267811514860399e-05, + "loss": 1.1029, + "step": 12884 + }, + { + "epoch": 0.7828543653927942, + "grad_norm": 0.25249505043029785, + "learning_rate": 1.1261758423761848e-05, + "loss": 1.1558, + "step": 12885 + }, + { + "epoch": 0.7829151224254207, + "grad_norm": 0.23568686842918396, + "learning_rate": 1.1255706752621026e-05, + "loss": 1.081, + "step": 12886 + }, + { + "epoch": 0.7829758794580473, + "grad_norm": 0.1936972439289093, + "learning_rate": 1.1249656501659767e-05, + "loss": 1.084, + "step": 12887 + }, + { + "epoch": 0.7830366364906738, + "grad_norm": 0.1833094209432602, + "learning_rate": 1.1243607671099831e-05, + "loss": 1.0052, + "step": 12888 + }, + { + "epoch": 0.7830973935233003, + "grad_norm": 0.13908442854881287, + "learning_rate": 1.1237560261162954e-05, + "loss": 1.0437, + "step": 12889 + }, + { + "epoch": 0.7831581505559269, + "grad_norm": 0.16442590951919556, + "learning_rate": 1.1231514272070798e-05, + "loss": 1.0116, + "step": 12890 + }, + { + "epoch": 0.7832189075885534, + "grad_norm": 0.1866680085659027, + "learning_rate": 1.1225469704044978e-05, + "loss": 1.072, + "step": 12891 + }, + { + "epoch": 0.7832796646211799, + "grad_norm": 0.21712692081928253, + "learning_rate": 1.1219426557307061e-05, + "loss": 1.0955, + "step": 12892 + }, + { + "epoch": 0.7833404216538065, + "grad_norm": 0.13343209028244019, + "learning_rate": 1.1213384832078545e-05, + "loss": 1.0252, + "step": 12893 + }, + { + "epoch": 0.783401178686433, + "grad_norm": 0.16088207066059113, + "learning_rate": 1.1207344528580937e-05, + "loss": 1.2694, + "step": 12894 + }, + { + "epoch": 0.7834619357190595, + "grad_norm": 0.12686337530612946, + "learning_rate": 1.1201305647035598e-05, + "loss": 1.0013, + "step": 12895 + }, + { + "epoch": 0.783522692751686, + "grad_norm": 0.21064142882823944, + "learning_rate": 1.1195268187663899e-05, + "loss": 1.0148, + "step": 12896 + }, + { + "epoch": 0.7835834497843125, + "grad_norm": 0.17827895283699036, + "learning_rate": 1.1189232150687145e-05, + "loss": 1.0413, + "step": 12897 + }, + { + "epoch": 0.783644206816939, + "grad_norm": 0.6548519730567932, + "learning_rate": 1.1183197536326573e-05, + "loss": 1.1437, + "step": 12898 + }, + { + "epoch": 0.7837049638495656, + "grad_norm": 0.12864194810390472, + "learning_rate": 1.1177164344803415e-05, + "loss": 1.0469, + "step": 12899 + }, + { + "epoch": 0.7837657208821921, + "grad_norm": 5.710908889770508, + "learning_rate": 1.11711325763388e-05, + "loss": 1.0727, + "step": 12900 + }, + { + "epoch": 0.7838264779148186, + "grad_norm": 0.4368632435798645, + "learning_rate": 1.1165102231153846e-05, + "loss": 1.078, + "step": 12901 + }, + { + "epoch": 0.7838872349474452, + "grad_norm": 0.3430633544921875, + "learning_rate": 1.1159073309469553e-05, + "loss": 1.1596, + "step": 12902 + }, + { + "epoch": 0.7839479919800717, + "grad_norm": 0.20942199230194092, + "learning_rate": 1.115304581150693e-05, + "loss": 1.1527, + "step": 12903 + }, + { + "epoch": 0.7840087490126982, + "grad_norm": 0.19040735065937042, + "learning_rate": 1.114701973748693e-05, + "loss": 1.0313, + "step": 12904 + }, + { + "epoch": 0.7840695060453248, + "grad_norm": 0.20312823355197906, + "learning_rate": 1.1140995087630434e-05, + "loss": 1.086, + "step": 12905 + }, + { + "epoch": 0.7841302630779513, + "grad_norm": 0.24687902629375458, + "learning_rate": 1.1134971862158277e-05, + "loss": 1.0639, + "step": 12906 + }, + { + "epoch": 0.7841910201105778, + "grad_norm": 0.2628224492073059, + "learning_rate": 1.112895006129125e-05, + "loss": 1.0989, + "step": 12907 + }, + { + "epoch": 0.7842517771432044, + "grad_norm": 0.18097111582756042, + "learning_rate": 1.1122929685250045e-05, + "loss": 1.0878, + "step": 12908 + }, + { + "epoch": 0.7843125341758308, + "grad_norm": 0.22957685589790344, + "learning_rate": 1.1116910734255375e-05, + "loss": 1.12, + "step": 12909 + }, + { + "epoch": 0.7843732912084573, + "grad_norm": 0.14877353608608246, + "learning_rate": 1.1110893208527866e-05, + "loss": 1.0706, + "step": 12910 + }, + { + "epoch": 0.7844340482410839, + "grad_norm": 0.18764512240886688, + "learning_rate": 1.1104877108288082e-05, + "loss": 1.0614, + "step": 12911 + }, + { + "epoch": 0.7844948052737104, + "grad_norm": 0.17056480050086975, + "learning_rate": 1.1098862433756545e-05, + "loss": 1.0641, + "step": 12912 + }, + { + "epoch": 0.7845555623063369, + "grad_norm": 0.2187301367521286, + "learning_rate": 1.1092849185153725e-05, + "loss": 1.0192, + "step": 12913 + }, + { + "epoch": 0.7846163193389635, + "grad_norm": 0.23513098061084747, + "learning_rate": 1.1086837362700036e-05, + "loss": 1.0261, + "step": 12914 + }, + { + "epoch": 0.78467707637159, + "grad_norm": 0.4457797706127167, + "learning_rate": 1.1080826966615848e-05, + "loss": 1.0269, + "step": 12915 + }, + { + "epoch": 0.7847378334042165, + "grad_norm": 0.15601439774036407, + "learning_rate": 1.1074817997121472e-05, + "loss": 1.032, + "step": 12916 + }, + { + "epoch": 0.7847985904368431, + "grad_norm": 0.19874560832977295, + "learning_rate": 1.1068810454437161e-05, + "loss": 1.0979, + "step": 12917 + }, + { + "epoch": 0.7848593474694696, + "grad_norm": 0.24426984786987305, + "learning_rate": 1.1062804338783134e-05, + "loss": 1.1915, + "step": 12918 + }, + { + "epoch": 0.7849201045020961, + "grad_norm": 0.19095976650714874, + "learning_rate": 1.1056799650379523e-05, + "loss": 1.0519, + "step": 12919 + }, + { + "epoch": 0.7849808615347227, + "grad_norm": 0.16295483708381653, + "learning_rate": 1.1050796389446466e-05, + "loss": 1.0388, + "step": 12920 + }, + { + "epoch": 0.7850416185673492, + "grad_norm": 0.4360133409500122, + "learning_rate": 1.1044794556204009e-05, + "loss": 1.1429, + "step": 12921 + }, + { + "epoch": 0.7851023755999756, + "grad_norm": 0.2733905613422394, + "learning_rate": 1.1038794150872116e-05, + "loss": 1.1178, + "step": 12922 + }, + { + "epoch": 0.7851631326326022, + "grad_norm": 3.4097211360931396, + "learning_rate": 1.1032795173670747e-05, + "loss": 1.0824, + "step": 12923 + }, + { + "epoch": 0.7852238896652287, + "grad_norm": 0.45448949933052063, + "learning_rate": 1.1026797624819818e-05, + "loss": 1.1621, + "step": 12924 + }, + { + "epoch": 0.7852846466978552, + "grad_norm": 0.2282106876373291, + "learning_rate": 1.1020801504539152e-05, + "loss": 1.0797, + "step": 12925 + }, + { + "epoch": 0.7853454037304818, + "grad_norm": 0.2467745542526245, + "learning_rate": 1.1014806813048545e-05, + "loss": 1.0698, + "step": 12926 + }, + { + "epoch": 0.7854061607631083, + "grad_norm": 0.20632979273796082, + "learning_rate": 1.1008813550567725e-05, + "loss": 0.9979, + "step": 12927 + }, + { + "epoch": 0.7854669177957349, + "grad_norm": 0.16501988470554352, + "learning_rate": 1.1002821717316387e-05, + "loss": 1.0401, + "step": 12928 + }, + { + "epoch": 0.7855276748283614, + "grad_norm": 0.20816127955913544, + "learning_rate": 1.099683131351415e-05, + "loss": 1.0578, + "step": 12929 + }, + { + "epoch": 0.7855884318609879, + "grad_norm": 1.7719972133636475, + "learning_rate": 1.0990842339380608e-05, + "loss": 1.1191, + "step": 12930 + }, + { + "epoch": 0.7856491888936145, + "grad_norm": 0.3286794126033783, + "learning_rate": 1.0984854795135274e-05, + "loss": 1.0649, + "step": 12931 + }, + { + "epoch": 0.785709945926241, + "grad_norm": 0.32665374875068665, + "learning_rate": 1.0978868680997634e-05, + "loss": 1.1057, + "step": 12932 + }, + { + "epoch": 0.7857707029588675, + "grad_norm": 0.14533042907714844, + "learning_rate": 1.09728839971871e-05, + "loss": 1.0373, + "step": 12933 + }, + { + "epoch": 0.7858314599914941, + "grad_norm": 0.1998651772737503, + "learning_rate": 1.0966900743923052e-05, + "loss": 1.0682, + "step": 12934 + }, + { + "epoch": 0.7858922170241205, + "grad_norm": 1.7451353073120117, + "learning_rate": 1.0960918921424801e-05, + "loss": 1.0055, + "step": 12935 + }, + { + "epoch": 0.785952974056747, + "grad_norm": 0.29143303632736206, + "learning_rate": 1.0954938529911608e-05, + "loss": 1.0614, + "step": 12936 + }, + { + "epoch": 0.7860137310893736, + "grad_norm": 0.16676843166351318, + "learning_rate": 1.0948959569602695e-05, + "loss": 1.0732, + "step": 12937 + }, + { + "epoch": 0.7860744881220001, + "grad_norm": 0.24016249179840088, + "learning_rate": 1.094298204071722e-05, + "loss": 1.117, + "step": 12938 + }, + { + "epoch": 0.7861352451546266, + "grad_norm": 0.19974175095558167, + "learning_rate": 1.0937005943474266e-05, + "loss": 0.9847, + "step": 12939 + }, + { + "epoch": 0.7861960021872532, + "grad_norm": 0.1295679211616516, + "learning_rate": 1.093103127809293e-05, + "loss": 1.0085, + "step": 12940 + }, + { + "epoch": 0.7862567592198797, + "grad_norm": 0.17401383817195892, + "learning_rate": 1.0925058044792208e-05, + "loss": 1.0566, + "step": 12941 + }, + { + "epoch": 0.7863175162525062, + "grad_norm": 0.2954272925853729, + "learning_rate": 1.091908624379102e-05, + "loss": 1.1135, + "step": 12942 + }, + { + "epoch": 0.7863782732851328, + "grad_norm": 0.22138236463069916, + "learning_rate": 1.0913115875308283e-05, + "loss": 1.1429, + "step": 12943 + }, + { + "epoch": 0.7864390303177593, + "grad_norm": 1.0461632013320923, + "learning_rate": 1.0907146939562818e-05, + "loss": 0.9982, + "step": 12944 + }, + { + "epoch": 0.7864997873503858, + "grad_norm": 0.1303156614303589, + "learning_rate": 1.0901179436773462e-05, + "loss": 1.0225, + "step": 12945 + }, + { + "epoch": 0.7865605443830124, + "grad_norm": 0.15005025267601013, + "learning_rate": 1.0895213367158924e-05, + "loss": 1.0625, + "step": 12946 + }, + { + "epoch": 0.7866213014156389, + "grad_norm": 0.2538699209690094, + "learning_rate": 1.08892487309379e-05, + "loss": 1.0657, + "step": 12947 + }, + { + "epoch": 0.7866820584482654, + "grad_norm": 0.1738605946302414, + "learning_rate": 1.088328552832904e-05, + "loss": 1.025, + "step": 12948 + }, + { + "epoch": 0.7867428154808919, + "grad_norm": 0.24402832984924316, + "learning_rate": 1.0877323759550884e-05, + "loss": 1.0226, + "step": 12949 + }, + { + "epoch": 0.7868035725135184, + "grad_norm": 0.15884552896022797, + "learning_rate": 1.0871363424821995e-05, + "loss": 0.9838, + "step": 12950 + }, + { + "epoch": 0.7868643295461449, + "grad_norm": 0.14920371770858765, + "learning_rate": 1.086540452436085e-05, + "loss": 1.0197, + "step": 12951 + }, + { + "epoch": 0.7869250865787715, + "grad_norm": 0.41768917441368103, + "learning_rate": 1.0859447058385862e-05, + "loss": 1.1925, + "step": 12952 + }, + { + "epoch": 0.786985843611398, + "grad_norm": 0.19942624866962433, + "learning_rate": 1.085349102711541e-05, + "loss": 1.0505, + "step": 12953 + }, + { + "epoch": 0.7870466006440245, + "grad_norm": 0.11639998108148575, + "learning_rate": 1.0847536430767813e-05, + "loss": 0.9723, + "step": 12954 + }, + { + "epoch": 0.7871073576766511, + "grad_norm": 0.22840236127376556, + "learning_rate": 1.0841583269561334e-05, + "loss": 1.0261, + "step": 12955 + }, + { + "epoch": 0.7871681147092776, + "grad_norm": 0.1629054993391037, + "learning_rate": 1.0835631543714186e-05, + "loss": 0.9821, + "step": 12956 + }, + { + "epoch": 0.7872288717419041, + "grad_norm": 0.16970194876194, + "learning_rate": 1.082968125344454e-05, + "loss": 1.0355, + "step": 12957 + }, + { + "epoch": 0.7872896287745307, + "grad_norm": 0.1864146739244461, + "learning_rate": 1.0823732398970492e-05, + "loss": 1.0718, + "step": 12958 + }, + { + "epoch": 0.7873503858071572, + "grad_norm": 0.12201405316591263, + "learning_rate": 1.0817784980510115e-05, + "loss": 0.9739, + "step": 12959 + }, + { + "epoch": 0.7874111428397838, + "grad_norm": 0.1581563800573349, + "learning_rate": 1.0811838998281382e-05, + "loss": 1.0263, + "step": 12960 + }, + { + "epoch": 0.7874718998724103, + "grad_norm": 0.3651217222213745, + "learning_rate": 1.0805894452502296e-05, + "loss": 1.2811, + "step": 12961 + }, + { + "epoch": 0.7875326569050367, + "grad_norm": 0.13549727201461792, + "learning_rate": 1.0799951343390714e-05, + "loss": 1.063, + "step": 12962 + }, + { + "epoch": 0.7875934139376632, + "grad_norm": 1.701201319694519, + "learning_rate": 1.0794009671164484e-05, + "loss": 1.1026, + "step": 12963 + }, + { + "epoch": 0.7876541709702898, + "grad_norm": 1.9763509035110474, + "learning_rate": 1.0788069436041415e-05, + "loss": 1.0382, + "step": 12964 + }, + { + "epoch": 0.7877149280029163, + "grad_norm": 0.15107344090938568, + "learning_rate": 1.0782130638239224e-05, + "loss": 0.996, + "step": 12965 + }, + { + "epoch": 0.7877756850355428, + "grad_norm": 0.7377314567565918, + "learning_rate": 1.0776193277975633e-05, + "loss": 1.0722, + "step": 12966 + }, + { + "epoch": 0.7878364420681694, + "grad_norm": 0.2262158840894699, + "learning_rate": 1.0770257355468256e-05, + "loss": 1.019, + "step": 12967 + }, + { + "epoch": 0.7878971991007959, + "grad_norm": 0.21163472533226013, + "learning_rate": 1.076432287093469e-05, + "loss": 1.1732, + "step": 12968 + }, + { + "epoch": 0.7879579561334225, + "grad_norm": 0.5530480742454529, + "learning_rate": 1.075838982459244e-05, + "loss": 1.1754, + "step": 12969 + }, + { + "epoch": 0.788018713166049, + "grad_norm": 0.17665937542915344, + "learning_rate": 1.0752458216658984e-05, + "loss": 1.0306, + "step": 12970 + }, + { + "epoch": 0.7880794701986755, + "grad_norm": 0.3275659680366516, + "learning_rate": 1.0746528047351772e-05, + "loss": 0.9853, + "step": 12971 + }, + { + "epoch": 0.7881402272313021, + "grad_norm": 0.24079535901546478, + "learning_rate": 1.0740599316888167e-05, + "loss": 1.0657, + "step": 12972 + }, + { + "epoch": 0.7882009842639286, + "grad_norm": 0.620383620262146, + "learning_rate": 1.0734672025485476e-05, + "loss": 1.087, + "step": 12973 + }, + { + "epoch": 0.7882617412965551, + "grad_norm": 0.20066192746162415, + "learning_rate": 1.0728746173360987e-05, + "loss": 1.1134, + "step": 12974 + }, + { + "epoch": 0.7883224983291816, + "grad_norm": 0.18508143723011017, + "learning_rate": 1.0722821760731871e-05, + "loss": 1.0356, + "step": 12975 + }, + { + "epoch": 0.7883832553618081, + "grad_norm": 0.15734495222568512, + "learning_rate": 1.0716898787815327e-05, + "loss": 1.0556, + "step": 12976 + }, + { + "epoch": 0.7884440123944346, + "grad_norm": 0.16295857727527618, + "learning_rate": 1.0710977254828458e-05, + "loss": 1.0724, + "step": 12977 + }, + { + "epoch": 0.7885047694270612, + "grad_norm": 0.8983916640281677, + "learning_rate": 1.0705057161988308e-05, + "loss": 1.0387, + "step": 12978 + }, + { + "epoch": 0.7885655264596877, + "grad_norm": 0.12436520308256149, + "learning_rate": 1.069913850951188e-05, + "loss": 1.0569, + "step": 12979 + }, + { + "epoch": 0.7886262834923142, + "grad_norm": 0.46720337867736816, + "learning_rate": 1.069322129761613e-05, + "loss": 1.1187, + "step": 12980 + }, + { + "epoch": 0.7886870405249408, + "grad_norm": 0.29703789949417114, + "learning_rate": 1.0687305526517949e-05, + "loss": 1.1326, + "step": 12981 + }, + { + "epoch": 0.7887477975575673, + "grad_norm": 0.24593999981880188, + "learning_rate": 1.0681391196434187e-05, + "loss": 1.1158, + "step": 12982 + }, + { + "epoch": 0.7888085545901938, + "grad_norm": 0.2105180323123932, + "learning_rate": 1.0675478307581627e-05, + "loss": 1.0047, + "step": 12983 + }, + { + "epoch": 0.7888693116228204, + "grad_norm": 0.1270182579755783, + "learning_rate": 1.066956686017701e-05, + "loss": 0.9825, + "step": 12984 + }, + { + "epoch": 0.7889300686554469, + "grad_norm": 0.22556433081626892, + "learning_rate": 1.0663656854437027e-05, + "loss": 1.0982, + "step": 12985 + }, + { + "epoch": 0.7889908256880734, + "grad_norm": 0.12535783648490906, + "learning_rate": 1.065774829057829e-05, + "loss": 1.0132, + "step": 12986 + }, + { + "epoch": 0.7890515827207, + "grad_norm": 0.1302294284105301, + "learning_rate": 1.0651841168817406e-05, + "loss": 1.0178, + "step": 12987 + }, + { + "epoch": 0.7891123397533264, + "grad_norm": 0.28794702887535095, + "learning_rate": 1.06459354893709e-05, + "loss": 1.1472, + "step": 12988 + }, + { + "epoch": 0.7891730967859529, + "grad_norm": 0.17235715687274933, + "learning_rate": 1.0640031252455229e-05, + "loss": 1.0678, + "step": 12989 + }, + { + "epoch": 0.7892338538185795, + "grad_norm": 0.13959282636642456, + "learning_rate": 1.0634128458286819e-05, + "loss": 0.9866, + "step": 12990 + }, + { + "epoch": 0.789294610851206, + "grad_norm": 0.25495338439941406, + "learning_rate": 1.0628227107082022e-05, + "loss": 1.1001, + "step": 12991 + }, + { + "epoch": 0.7893553678838325, + "grad_norm": 0.20715157687664032, + "learning_rate": 1.0622327199057192e-05, + "loss": 1.0743, + "step": 12992 + }, + { + "epoch": 0.7894161249164591, + "grad_norm": 0.1756540983915329, + "learning_rate": 1.0616428734428562e-05, + "loss": 1.0502, + "step": 12993 + }, + { + "epoch": 0.7894768819490856, + "grad_norm": 0.14392903447151184, + "learning_rate": 1.0610531713412352e-05, + "loss": 1.035, + "step": 12994 + }, + { + "epoch": 0.7895376389817121, + "grad_norm": 0.21065950393676758, + "learning_rate": 1.0604636136224733e-05, + "loss": 1.0361, + "step": 12995 + }, + { + "epoch": 0.7895983960143387, + "grad_norm": 0.20337913930416107, + "learning_rate": 1.0598742003081762e-05, + "loss": 1.1143, + "step": 12996 + }, + { + "epoch": 0.7896591530469652, + "grad_norm": 0.19963736832141876, + "learning_rate": 1.0592849314199532e-05, + "loss": 1.0542, + "step": 12997 + }, + { + "epoch": 0.7897199100795917, + "grad_norm": 0.17754365503787994, + "learning_rate": 1.0586958069794028e-05, + "loss": 1.0107, + "step": 12998 + }, + { + "epoch": 0.7897806671122183, + "grad_norm": 0.345092236995697, + "learning_rate": 1.058106827008119e-05, + "loss": 1.0364, + "step": 12999 + }, + { + "epoch": 0.7898414241448448, + "grad_norm": 0.27234238386154175, + "learning_rate": 1.0575179915276917e-05, + "loss": 1.0535, + "step": 13000 + }, + { + "epoch": 0.7899021811774712, + "grad_norm": 0.17322777211666107, + "learning_rate": 1.0569293005597048e-05, + "loss": 1.1007, + "step": 13001 + }, + { + "epoch": 0.7899629382100978, + "grad_norm": 0.2300783395767212, + "learning_rate": 1.0563407541257364e-05, + "loss": 1.1109, + "step": 13002 + }, + { + "epoch": 0.7900236952427243, + "grad_norm": 0.17129576206207275, + "learning_rate": 1.05575235224736e-05, + "loss": 1.064, + "step": 13003 + }, + { + "epoch": 0.7900844522753508, + "grad_norm": 1.7680379152297974, + "learning_rate": 1.0551640949461433e-05, + "loss": 1.0624, + "step": 13004 + }, + { + "epoch": 0.7901452093079774, + "grad_norm": 0.2216646522283554, + "learning_rate": 1.0545759822436496e-05, + "loss": 1.1581, + "step": 13005 + }, + { + "epoch": 0.7902059663406039, + "grad_norm": 0.2410593181848526, + "learning_rate": 1.0539880141614355e-05, + "loss": 1.0774, + "step": 13006 + }, + { + "epoch": 0.7902667233732305, + "grad_norm": 0.16837775707244873, + "learning_rate": 1.0534001907210522e-05, + "loss": 1.0974, + "step": 13007 + }, + { + "epoch": 0.790327480405857, + "grad_norm": 0.14947804808616638, + "learning_rate": 1.0528125119440507e-05, + "loss": 1.053, + "step": 13008 + }, + { + "epoch": 0.7903882374384835, + "grad_norm": 0.19796188175678253, + "learning_rate": 1.0522249778519677e-05, + "loss": 1.0847, + "step": 13009 + }, + { + "epoch": 0.7904489944711101, + "grad_norm": 0.1273566484451294, + "learning_rate": 1.051637588466342e-05, + "loss": 0.9977, + "step": 13010 + }, + { + "epoch": 0.7905097515037366, + "grad_norm": 0.18867747485637665, + "learning_rate": 1.0510503438087033e-05, + "loss": 1.0252, + "step": 13011 + }, + { + "epoch": 0.7905705085363631, + "grad_norm": 0.19973605871200562, + "learning_rate": 1.0504632439005762e-05, + "loss": 1.0415, + "step": 13012 + }, + { + "epoch": 0.7906312655689897, + "grad_norm": 2.0492427349090576, + "learning_rate": 1.0498762887634832e-05, + "loss": 1.1655, + "step": 13013 + }, + { + "epoch": 0.7906920226016161, + "grad_norm": 0.19302739202976227, + "learning_rate": 1.0492894784189389e-05, + "loss": 0.9903, + "step": 13014 + }, + { + "epoch": 0.7907527796342426, + "grad_norm": 2.463564395904541, + "learning_rate": 1.0487028128884541e-05, + "loss": 1.0511, + "step": 13015 + }, + { + "epoch": 0.7908135366668692, + "grad_norm": 0.29714342951774597, + "learning_rate": 1.0481162921935283e-05, + "loss": 1.0726, + "step": 13016 + }, + { + "epoch": 0.7908742936994957, + "grad_norm": 0.15738515555858612, + "learning_rate": 1.0475299163556651e-05, + "loss": 1.0875, + "step": 13017 + }, + { + "epoch": 0.7909350507321222, + "grad_norm": 0.16154788434505463, + "learning_rate": 1.0469436853963566e-05, + "loss": 1.0843, + "step": 13018 + }, + { + "epoch": 0.7909958077647488, + "grad_norm": 0.7143619060516357, + "learning_rate": 1.0463575993370917e-05, + "loss": 1.1183, + "step": 13019 + }, + { + "epoch": 0.7910565647973753, + "grad_norm": 0.19619981944561005, + "learning_rate": 1.0457716581993531e-05, + "loss": 1.1223, + "step": 13020 + }, + { + "epoch": 0.7911173218300018, + "grad_norm": 0.5528907775878906, + "learning_rate": 1.045185862004619e-05, + "loss": 1.0527, + "step": 13021 + }, + { + "epoch": 0.7911780788626284, + "grad_norm": 0.23793160915374756, + "learning_rate": 1.0446002107743612e-05, + "loss": 1.0527, + "step": 13022 + }, + { + "epoch": 0.7912388358952549, + "grad_norm": 0.15528060495853424, + "learning_rate": 1.0440147045300474e-05, + "loss": 1.0466, + "step": 13023 + }, + { + "epoch": 0.7912995929278814, + "grad_norm": 0.12719902396202087, + "learning_rate": 1.0434293432931398e-05, + "loss": 1.006, + "step": 13024 + }, + { + "epoch": 0.791360349960508, + "grad_norm": 0.2412288635969162, + "learning_rate": 1.0428441270850935e-05, + "loss": 1.1342, + "step": 13025 + }, + { + "epoch": 0.7914211069931345, + "grad_norm": 0.19873397052288055, + "learning_rate": 1.0422590559273609e-05, + "loss": 1.0307, + "step": 13026 + }, + { + "epoch": 0.7914818640257609, + "grad_norm": 2.8691272735595703, + "learning_rate": 1.0416741298413878e-05, + "loss": 1.1884, + "step": 13027 + }, + { + "epoch": 0.7915426210583875, + "grad_norm": 0.2128138393163681, + "learning_rate": 1.0410893488486145e-05, + "loss": 1.1226, + "step": 13028 + }, + { + "epoch": 0.791603378091014, + "grad_norm": 0.296351820230484, + "learning_rate": 1.0405047129704764e-05, + "loss": 1.2245, + "step": 13029 + }, + { + "epoch": 0.7916641351236405, + "grad_norm": 0.21697619557380676, + "learning_rate": 1.0399202222284033e-05, + "loss": 1.0821, + "step": 13030 + }, + { + "epoch": 0.7917248921562671, + "grad_norm": 0.20011933147907257, + "learning_rate": 1.0393358766438199e-05, + "loss": 1.0277, + "step": 13031 + }, + { + "epoch": 0.7917856491888936, + "grad_norm": 0.24124833941459656, + "learning_rate": 1.0387516762381444e-05, + "loss": 1.0422, + "step": 13032 + }, + { + "epoch": 0.7918464062215201, + "grad_norm": 0.19516292214393616, + "learning_rate": 1.0381676210327934e-05, + "loss": 1.0102, + "step": 13033 + }, + { + "epoch": 0.7919071632541467, + "grad_norm": 1.7926177978515625, + "learning_rate": 1.0375837110491737e-05, + "loss": 1.0311, + "step": 13034 + }, + { + "epoch": 0.7919679202867732, + "grad_norm": 0.2014506757259369, + "learning_rate": 1.0369999463086905e-05, + "loss": 1.0919, + "step": 13035 + }, + { + "epoch": 0.7920286773193997, + "grad_norm": 0.3101705312728882, + "learning_rate": 1.0364163268327387e-05, + "loss": 1.104, + "step": 13036 + }, + { + "epoch": 0.7920894343520263, + "grad_norm": 0.5237692594528198, + "learning_rate": 1.0358328526427113e-05, + "loss": 1.0874, + "step": 13037 + }, + { + "epoch": 0.7921501913846528, + "grad_norm": 0.4921267628669739, + "learning_rate": 1.0352495237599985e-05, + "loss": 1.0428, + "step": 13038 + }, + { + "epoch": 0.7922109484172793, + "grad_norm": 0.5928593277931213, + "learning_rate": 1.0346663402059809e-05, + "loss": 1.2923, + "step": 13039 + }, + { + "epoch": 0.7922717054499058, + "grad_norm": 0.4457022249698639, + "learning_rate": 1.0340833020020347e-05, + "loss": 1.0714, + "step": 13040 + }, + { + "epoch": 0.7923324624825323, + "grad_norm": 0.19737984240055084, + "learning_rate": 1.0335004091695333e-05, + "loss": 1.0852, + "step": 13041 + }, + { + "epoch": 0.7923932195151588, + "grad_norm": 0.16164802014827728, + "learning_rate": 1.0329176617298381e-05, + "loss": 1.0001, + "step": 13042 + }, + { + "epoch": 0.7924539765477854, + "grad_norm": 0.29158398509025574, + "learning_rate": 1.0323350597043141e-05, + "loss": 1.138, + "step": 13043 + }, + { + "epoch": 0.7925147335804119, + "grad_norm": 0.45492804050445557, + "learning_rate": 1.031752603114316e-05, + "loss": 1.0492, + "step": 13044 + }, + { + "epoch": 0.7925754906130384, + "grad_norm": 0.31973597407341003, + "learning_rate": 1.0311702919811928e-05, + "loss": 1.1239, + "step": 13045 + }, + { + "epoch": 0.792636247645665, + "grad_norm": 0.1498488187789917, + "learning_rate": 1.03058812632629e-05, + "loss": 1.0653, + "step": 13046 + }, + { + "epoch": 0.7926970046782915, + "grad_norm": 0.26506003737449646, + "learning_rate": 1.0300061061709465e-05, + "loss": 1.1142, + "step": 13047 + }, + { + "epoch": 0.792757761710918, + "grad_norm": 0.1572459638118744, + "learning_rate": 1.0294242315364966e-05, + "loss": 1.0269, + "step": 13048 + }, + { + "epoch": 0.7928185187435446, + "grad_norm": 0.2039637267589569, + "learning_rate": 1.0288425024442689e-05, + "loss": 0.9826, + "step": 13049 + }, + { + "epoch": 0.7928792757761711, + "grad_norm": 0.14731352031230927, + "learning_rate": 1.0282609189155867e-05, + "loss": 1.003, + "step": 13050 + }, + { + "epoch": 0.7929400328087977, + "grad_norm": 0.788032054901123, + "learning_rate": 1.0276794809717688e-05, + "loss": 1.172, + "step": 13051 + }, + { + "epoch": 0.7930007898414242, + "grad_norm": 1.0155340433120728, + "learning_rate": 1.0270981886341268e-05, + "loss": 1.0171, + "step": 13052 + }, + { + "epoch": 0.7930615468740507, + "grad_norm": 1.05502188205719, + "learning_rate": 1.0265170419239678e-05, + "loss": 1.0715, + "step": 13053 + }, + { + "epoch": 0.7931223039066772, + "grad_norm": 0.14731916785240173, + "learning_rate": 1.0259360408625957e-05, + "loss": 1.0065, + "step": 13054 + }, + { + "epoch": 0.7931830609393037, + "grad_norm": 0.17426668107509613, + "learning_rate": 1.0253551854713079e-05, + "loss": 1.0427, + "step": 13055 + }, + { + "epoch": 0.7932438179719302, + "grad_norm": 0.14183762669563293, + "learning_rate": 1.0247744757713923e-05, + "loss": 1.0788, + "step": 13056 + }, + { + "epoch": 0.7933045750045568, + "grad_norm": 0.2188291698694229, + "learning_rate": 1.0241939117841371e-05, + "loss": 1.0301, + "step": 13057 + }, + { + "epoch": 0.7933653320371833, + "grad_norm": 0.13384585082530975, + "learning_rate": 1.0236134935308205e-05, + "loss": 1.0326, + "step": 13058 + }, + { + "epoch": 0.7934260890698098, + "grad_norm": 0.16608786582946777, + "learning_rate": 1.023033221032722e-05, + "loss": 1.1128, + "step": 13059 + }, + { + "epoch": 0.7934868461024364, + "grad_norm": 0.16568394005298615, + "learning_rate": 1.0224530943111094e-05, + "loss": 1.0671, + "step": 13060 + }, + { + "epoch": 0.7935476031350629, + "grad_norm": 0.1522781401872635, + "learning_rate": 1.0218731133872484e-05, + "loss": 1.0354, + "step": 13061 + }, + { + "epoch": 0.7936083601676894, + "grad_norm": 0.1276731640100479, + "learning_rate": 1.0212932782823965e-05, + "loss": 1.0515, + "step": 13062 + }, + { + "epoch": 0.793669117200316, + "grad_norm": 0.17167283594608307, + "learning_rate": 1.0207135890178066e-05, + "loss": 1.0336, + "step": 13063 + }, + { + "epoch": 0.7937298742329425, + "grad_norm": 1.3844966888427734, + "learning_rate": 1.0201340456147313e-05, + "loss": 1.0621, + "step": 13064 + }, + { + "epoch": 0.793790631265569, + "grad_norm": 0.1329297423362732, + "learning_rate": 1.0195546480944113e-05, + "loss": 1.0241, + "step": 13065 + }, + { + "epoch": 0.7938513882981956, + "grad_norm": 0.6698585748672485, + "learning_rate": 1.0189753964780857e-05, + "loss": 1.0558, + "step": 13066 + }, + { + "epoch": 0.793912145330822, + "grad_norm": 0.1619594246149063, + "learning_rate": 1.018396290786986e-05, + "loss": 1.0411, + "step": 13067 + }, + { + "epoch": 0.7939729023634485, + "grad_norm": 0.13525299727916718, + "learning_rate": 1.01781733104234e-05, + "loss": 1.0191, + "step": 13068 + }, + { + "epoch": 0.7940336593960751, + "grad_norm": 0.15498274564743042, + "learning_rate": 1.0172385172653692e-05, + "loss": 1.0232, + "step": 13069 + }, + { + "epoch": 0.7940944164287016, + "grad_norm": 0.2080819457769394, + "learning_rate": 1.0166598494772905e-05, + "loss": 1.0382, + "step": 13070 + }, + { + "epoch": 0.7941551734613281, + "grad_norm": 0.15991561114788055, + "learning_rate": 1.0160813276993153e-05, + "loss": 1.0354, + "step": 13071 + }, + { + "epoch": 0.7942159304939547, + "grad_norm": 0.9621621966362, + "learning_rate": 1.0155029519526487e-05, + "loss": 1.1395, + "step": 13072 + }, + { + "epoch": 0.7942766875265812, + "grad_norm": 0.18404975533485413, + "learning_rate": 1.0149247222584923e-05, + "loss": 1.0195, + "step": 13073 + }, + { + "epoch": 0.7943374445592077, + "grad_norm": 0.21162302792072296, + "learning_rate": 1.0143466386380385e-05, + "loss": 1.0573, + "step": 13074 + }, + { + "epoch": 0.7943982015918343, + "grad_norm": 0.12591677904129028, + "learning_rate": 1.013768701112482e-05, + "loss": 1.0265, + "step": 13075 + }, + { + "epoch": 0.7944589586244608, + "grad_norm": 0.12545669078826904, + "learning_rate": 1.013190909703003e-05, + "loss": 1.0017, + "step": 13076 + }, + { + "epoch": 0.7945197156570873, + "grad_norm": 0.11995256692171097, + "learning_rate": 1.0126132644307817e-05, + "loss": 1.0046, + "step": 13077 + }, + { + "epoch": 0.7945804726897139, + "grad_norm": 0.1493687480688095, + "learning_rate": 1.0120357653169921e-05, + "loss": 1.0086, + "step": 13078 + }, + { + "epoch": 0.7946412297223404, + "grad_norm": 0.18389169871807098, + "learning_rate": 1.0114584123828008e-05, + "loss": 1.0157, + "step": 13079 + }, + { + "epoch": 0.7947019867549668, + "grad_norm": 0.21025362610816956, + "learning_rate": 1.0108812056493738e-05, + "loss": 1.0666, + "step": 13080 + }, + { + "epoch": 0.7947627437875934, + "grad_norm": 0.19835101068019867, + "learning_rate": 1.010304145137867e-05, + "loss": 1.1159, + "step": 13081 + }, + { + "epoch": 0.7948235008202199, + "grad_norm": 0.23146812617778778, + "learning_rate": 1.0097272308694345e-05, + "loss": 1.2661, + "step": 13082 + }, + { + "epoch": 0.7948842578528464, + "grad_norm": 0.126046821475029, + "learning_rate": 1.0091504628652204e-05, + "loss": 1.0327, + "step": 13083 + }, + { + "epoch": 0.794945014885473, + "grad_norm": 0.13559676706790924, + "learning_rate": 1.0085738411463664e-05, + "loss": 0.9878, + "step": 13084 + }, + { + "epoch": 0.7950057719180995, + "grad_norm": 0.14273463189601898, + "learning_rate": 1.0079973657340108e-05, + "loss": 1.0633, + "step": 13085 + }, + { + "epoch": 0.795066528950726, + "grad_norm": 0.3947140872478485, + "learning_rate": 1.0074210366492831e-05, + "loss": 1.1929, + "step": 13086 + }, + { + "epoch": 0.7951272859833526, + "grad_norm": 0.14574673771858215, + "learning_rate": 1.0068448539133096e-05, + "loss": 1.0569, + "step": 13087 + }, + { + "epoch": 0.7951880430159791, + "grad_norm": 0.4020739495754242, + "learning_rate": 1.0062688175472113e-05, + "loss": 1.1214, + "step": 13088 + }, + { + "epoch": 0.7952488000486057, + "grad_norm": 0.1905318796634674, + "learning_rate": 1.0056929275720982e-05, + "loss": 1.0107, + "step": 13089 + }, + { + "epoch": 0.7953095570812322, + "grad_norm": 0.19335868954658508, + "learning_rate": 1.0051171840090845e-05, + "loss": 1.0595, + "step": 13090 + }, + { + "epoch": 0.7953703141138587, + "grad_norm": 0.14467576146125793, + "learning_rate": 1.0045415868792729e-05, + "loss": 1.0468, + "step": 13091 + }, + { + "epoch": 0.7954310711464853, + "grad_norm": 0.14078576862812042, + "learning_rate": 1.003966136203761e-05, + "loss": 1.0305, + "step": 13092 + }, + { + "epoch": 0.7954918281791117, + "grad_norm": 0.2030366212129593, + "learning_rate": 1.0033908320036433e-05, + "loss": 1.0547, + "step": 13093 + }, + { + "epoch": 0.7955525852117382, + "grad_norm": 0.13857269287109375, + "learning_rate": 1.0028156743000072e-05, + "loss": 1.0541, + "step": 13094 + }, + { + "epoch": 0.7956133422443648, + "grad_norm": 0.1501915603876114, + "learning_rate": 1.0022406631139352e-05, + "loss": 1.0309, + "step": 13095 + }, + { + "epoch": 0.7956740992769913, + "grad_norm": 0.16208404302597046, + "learning_rate": 1.0016657984665045e-05, + "loss": 1.1082, + "step": 13096 + }, + { + "epoch": 0.7957348563096178, + "grad_norm": 0.24816077947616577, + "learning_rate": 1.0010910803787865e-05, + "loss": 1.159, + "step": 13097 + }, + { + "epoch": 0.7957956133422444, + "grad_norm": 1.9412277936935425, + "learning_rate": 1.0005165088718483e-05, + "loss": 1.0615, + "step": 13098 + }, + { + "epoch": 0.7958563703748709, + "grad_norm": 0.14652501046657562, + "learning_rate": 9.999420839667505e-06, + "loss": 1.042, + "step": 13099 + }, + { + "epoch": 0.7959171274074974, + "grad_norm": 0.21301037073135376, + "learning_rate": 9.993678056845474e-06, + "loss": 1.1881, + "step": 13100 + }, + { + "epoch": 0.795977884440124, + "grad_norm": 0.15191751718521118, + "learning_rate": 9.987936740462923e-06, + "loss": 1.0566, + "step": 13101 + }, + { + "epoch": 0.7960386414727505, + "grad_norm": 0.21193943917751312, + "learning_rate": 9.982196890730294e-06, + "loss": 1.1001, + "step": 13102 + }, + { + "epoch": 0.796099398505377, + "grad_norm": 0.15799377858638763, + "learning_rate": 9.97645850785796e-06, + "loss": 1.0175, + "step": 13103 + }, + { + "epoch": 0.7961601555380036, + "grad_norm": 0.5373066067695618, + "learning_rate": 9.970721592056276e-06, + "loss": 1.2123, + "step": 13104 + }, + { + "epoch": 0.7962209125706301, + "grad_norm": 0.3757920265197754, + "learning_rate": 9.964986143535515e-06, + "loss": 1.0519, + "step": 13105 + }, + { + "epoch": 0.7962816696032565, + "grad_norm": 0.170795738697052, + "learning_rate": 9.959252162505944e-06, + "loss": 1.0996, + "step": 13106 + }, + { + "epoch": 0.7963424266358831, + "grad_norm": 0.21175409853458405, + "learning_rate": 9.953519649177717e-06, + "loss": 1.0812, + "step": 13107 + }, + { + "epoch": 0.7964031836685096, + "grad_norm": 0.1776556372642517, + "learning_rate": 9.947788603760987e-06, + "loss": 1.0642, + "step": 13108 + }, + { + "epoch": 0.7964639407011361, + "grad_norm": 0.30178311467170715, + "learning_rate": 9.942059026465794e-06, + "loss": 1.0165, + "step": 13109 + }, + { + "epoch": 0.7965246977337627, + "grad_norm": 0.16325952112674713, + "learning_rate": 9.936330917502157e-06, + "loss": 1.0038, + "step": 13110 + }, + { + "epoch": 0.7965854547663892, + "grad_norm": 0.19428667426109314, + "learning_rate": 9.93060427708007e-06, + "loss": 1.0483, + "step": 13111 + }, + { + "epoch": 0.7966462117990157, + "grad_norm": 0.31289345026016235, + "learning_rate": 9.924879105409423e-06, + "loss": 1.0879, + "step": 13112 + }, + { + "epoch": 0.7967069688316423, + "grad_norm": 0.1870843768119812, + "learning_rate": 9.919155402700082e-06, + "loss": 1.0821, + "step": 13113 + }, + { + "epoch": 0.7967677258642688, + "grad_norm": 0.11789702624082565, + "learning_rate": 9.913433169161845e-06, + "loss": 1.0126, + "step": 13114 + }, + { + "epoch": 0.7968284828968953, + "grad_norm": 0.2346077263355255, + "learning_rate": 9.907712405004466e-06, + "loss": 1.1872, + "step": 13115 + }, + { + "epoch": 0.7968892399295219, + "grad_norm": 0.18692351877689362, + "learning_rate": 9.901993110437635e-06, + "loss": 1.0626, + "step": 13116 + }, + { + "epoch": 0.7969499969621484, + "grad_norm": 0.2516316771507263, + "learning_rate": 9.896275285671003e-06, + "loss": 1.0388, + "step": 13117 + }, + { + "epoch": 0.797010753994775, + "grad_norm": 0.20555947721004486, + "learning_rate": 9.890558930914146e-06, + "loss": 1.0558, + "step": 13118 + }, + { + "epoch": 0.7970715110274014, + "grad_norm": 1.6272326707839966, + "learning_rate": 9.884844046376607e-06, + "loss": 1.0304, + "step": 13119 + }, + { + "epoch": 0.7971322680600279, + "grad_norm": 0.1337209790945053, + "learning_rate": 9.879130632267864e-06, + "loss": 1.0167, + "step": 13120 + }, + { + "epoch": 0.7971930250926544, + "grad_norm": 0.18764923512935638, + "learning_rate": 9.873418688797332e-06, + "loss": 1.0621, + "step": 13121 + }, + { + "epoch": 0.797253782125281, + "grad_norm": 0.30124372243881226, + "learning_rate": 9.86770821617442e-06, + "loss": 1.1652, + "step": 13122 + }, + { + "epoch": 0.7973145391579075, + "grad_norm": 0.20750819146633148, + "learning_rate": 9.861999214608403e-06, + "loss": 1.0759, + "step": 13123 + }, + { + "epoch": 0.797375296190534, + "grad_norm": 0.16530326008796692, + "learning_rate": 9.856291684308566e-06, + "loss": 1.0144, + "step": 13124 + }, + { + "epoch": 0.7974360532231606, + "grad_norm": 0.24845075607299805, + "learning_rate": 9.850585625484105e-06, + "loss": 1.1352, + "step": 13125 + }, + { + "epoch": 0.7974968102557871, + "grad_norm": 0.13926587998867035, + "learning_rate": 9.8448810383442e-06, + "loss": 1.0449, + "step": 13126 + }, + { + "epoch": 0.7975575672884136, + "grad_norm": 0.3441337049007416, + "learning_rate": 9.839177923097942e-06, + "loss": 1.0993, + "step": 13127 + }, + { + "epoch": 0.7976183243210402, + "grad_norm": 0.1459483802318573, + "learning_rate": 9.833476279954395e-06, + "loss": 1.0686, + "step": 13128 + }, + { + "epoch": 0.7976790813536667, + "grad_norm": 0.2620517313480377, + "learning_rate": 9.827776109122526e-06, + "loss": 1.1243, + "step": 13129 + }, + { + "epoch": 0.7977398383862933, + "grad_norm": 0.12931400537490845, + "learning_rate": 9.822077410811275e-06, + "loss": 1.0111, + "step": 13130 + }, + { + "epoch": 0.7978005954189198, + "grad_norm": 0.3266181945800781, + "learning_rate": 9.816380185229556e-06, + "loss": 1.232, + "step": 13131 + }, + { + "epoch": 0.7978613524515462, + "grad_norm": 1.6434526443481445, + "learning_rate": 9.810684432586188e-06, + "loss": 1.107, + "step": 13132 + }, + { + "epoch": 0.7979221094841727, + "grad_norm": 0.20164422690868378, + "learning_rate": 9.80499015308995e-06, + "loss": 1.0428, + "step": 13133 + }, + { + "epoch": 0.7979828665167993, + "grad_norm": 0.1739140897989273, + "learning_rate": 9.79929734694957e-06, + "loss": 1.1125, + "step": 13134 + }, + { + "epoch": 0.7980436235494258, + "grad_norm": 0.16407619416713715, + "learning_rate": 9.793606014373712e-06, + "loss": 1.0394, + "step": 13135 + }, + { + "epoch": 0.7981043805820524, + "grad_norm": 0.1528838574886322, + "learning_rate": 9.787916155571003e-06, + "loss": 1.02, + "step": 13136 + }, + { + "epoch": 0.7981651376146789, + "grad_norm": 0.5488142967224121, + "learning_rate": 9.782227770749997e-06, + "loss": 1.2317, + "step": 13137 + }, + { + "epoch": 0.7982258946473054, + "grad_norm": 0.5633705258369446, + "learning_rate": 9.776540860119205e-06, + "loss": 1.053, + "step": 13138 + }, + { + "epoch": 0.798286651679932, + "grad_norm": 0.180685892701149, + "learning_rate": 9.770855423887088e-06, + "loss": 1.0372, + "step": 13139 + }, + { + "epoch": 0.7983474087125585, + "grad_norm": 0.14401106536388397, + "learning_rate": 9.765171462262036e-06, + "loss": 1.0261, + "step": 13140 + }, + { + "epoch": 0.798408165745185, + "grad_norm": 0.13949011266231537, + "learning_rate": 9.759488975452386e-06, + "loss": 1.0573, + "step": 13141 + }, + { + "epoch": 0.7984689227778116, + "grad_norm": 0.2961820960044861, + "learning_rate": 9.753807963666478e-06, + "loss": 1.0412, + "step": 13142 + }, + { + "epoch": 0.7985296798104381, + "grad_norm": 0.17980928719043732, + "learning_rate": 9.748128427112502e-06, + "loss": 1.0616, + "step": 13143 + }, + { + "epoch": 0.7985904368430646, + "grad_norm": 0.15300007164478302, + "learning_rate": 9.742450365998663e-06, + "loss": 1.0218, + "step": 13144 + }, + { + "epoch": 0.7986511938756911, + "grad_norm": 0.2574880123138428, + "learning_rate": 9.736773780533081e-06, + "loss": 1.1172, + "step": 13145 + }, + { + "epoch": 0.7987119509083176, + "grad_norm": 0.20608893036842346, + "learning_rate": 9.731098670923832e-06, + "loss": 1.14, + "step": 13146 + }, + { + "epoch": 0.7987727079409441, + "grad_norm": 0.21915572881698608, + "learning_rate": 9.725425037378954e-06, + "loss": 1.0076, + "step": 13147 + }, + { + "epoch": 0.7988334649735707, + "grad_norm": 0.23594032227993011, + "learning_rate": 9.71975288010641e-06, + "loss": 1.092, + "step": 13148 + }, + { + "epoch": 0.7988942220061972, + "grad_norm": 0.2264941781759262, + "learning_rate": 9.714082199314118e-06, + "loss": 1.0556, + "step": 13149 + }, + { + "epoch": 0.7989549790388237, + "grad_norm": 0.1644909530878067, + "learning_rate": 9.708412995209921e-06, + "loss": 1.0417, + "step": 13150 + }, + { + "epoch": 0.7990157360714503, + "grad_norm": 0.7126345634460449, + "learning_rate": 9.702745268001617e-06, + "loss": 1.2163, + "step": 13151 + }, + { + "epoch": 0.7990764931040768, + "grad_norm": 0.19588220119476318, + "learning_rate": 9.697079017896993e-06, + "loss": 1.002, + "step": 13152 + }, + { + "epoch": 0.7991372501367033, + "grad_norm": 0.22844715416431427, + "learning_rate": 9.691414245103725e-06, + "loss": 1.0962, + "step": 13153 + }, + { + "epoch": 0.7991980071693299, + "grad_norm": 0.23295533657073975, + "learning_rate": 9.685750949829458e-06, + "loss": 1.1175, + "step": 13154 + }, + { + "epoch": 0.7992587642019564, + "grad_norm": 0.14564889669418335, + "learning_rate": 9.6800891322818e-06, + "loss": 1.0597, + "step": 13155 + }, + { + "epoch": 0.7993195212345829, + "grad_norm": 0.14614324271678925, + "learning_rate": 9.674428792668238e-06, + "loss": 0.9953, + "step": 13156 + }, + { + "epoch": 0.7993802782672095, + "grad_norm": 0.1915741264820099, + "learning_rate": 9.668769931196297e-06, + "loss": 1.0321, + "step": 13157 + }, + { + "epoch": 0.799441035299836, + "grad_norm": 0.18455077707767487, + "learning_rate": 9.663112548073394e-06, + "loss": 1.0814, + "step": 13158 + }, + { + "epoch": 0.7995017923324624, + "grad_norm": 0.177991583943367, + "learning_rate": 9.65745664350689e-06, + "loss": 1.0617, + "step": 13159 + }, + { + "epoch": 0.799562549365089, + "grad_norm": 0.20916886627674103, + "learning_rate": 9.65180221770412e-06, + "loss": 1.1006, + "step": 13160 + }, + { + "epoch": 0.7996233063977155, + "grad_norm": 0.7610392570495605, + "learning_rate": 9.646149270872334e-06, + "loss": 1.0518, + "step": 13161 + }, + { + "epoch": 0.799684063430342, + "grad_norm": 0.2485802322626114, + "learning_rate": 9.640497803218745e-06, + "loss": 1.0896, + "step": 13162 + }, + { + "epoch": 0.7997448204629686, + "grad_norm": 0.22942118346691132, + "learning_rate": 9.634847814950509e-06, + "loss": 1.1499, + "step": 13163 + }, + { + "epoch": 0.7998055774955951, + "grad_norm": 0.1731702983379364, + "learning_rate": 9.629199306274733e-06, + "loss": 1.0092, + "step": 13164 + }, + { + "epoch": 0.7998663345282216, + "grad_norm": 0.19127246737480164, + "learning_rate": 9.623552277398461e-06, + "loss": 1.0703, + "step": 13165 + }, + { + "epoch": 0.7999270915608482, + "grad_norm": 0.12116823345422745, + "learning_rate": 9.61790672852868e-06, + "loss": 1.0012, + "step": 13166 + }, + { + "epoch": 0.7999878485934747, + "grad_norm": 0.16096466779708862, + "learning_rate": 9.612262659872322e-06, + "loss": 1.0205, + "step": 13167 + }, + { + "epoch": 0.8000486056261013, + "grad_norm": 0.14403480291366577, + "learning_rate": 9.606620071636302e-06, + "loss": 1.0564, + "step": 13168 + }, + { + "epoch": 0.8001093626587278, + "grad_norm": 0.13184787333011627, + "learning_rate": 9.600978964027435e-06, + "loss": 0.9757, + "step": 13169 + }, + { + "epoch": 0.8001701196913543, + "grad_norm": 0.22259201109409332, + "learning_rate": 9.595339337252485e-06, + "loss": 1.0527, + "step": 13170 + }, + { + "epoch": 0.8002308767239809, + "grad_norm": 0.7236748337745667, + "learning_rate": 9.589701191518186e-06, + "loss": 1.4077, + "step": 13171 + }, + { + "epoch": 0.8002916337566073, + "grad_norm": 0.240531787276268, + "learning_rate": 9.584064527031178e-06, + "loss": 1.0313, + "step": 13172 + }, + { + "epoch": 0.8003523907892338, + "grad_norm": 0.7037509083747864, + "learning_rate": 9.578429343998119e-06, + "loss": 1.4326, + "step": 13173 + }, + { + "epoch": 0.8004131478218603, + "grad_norm": 0.16630302369594574, + "learning_rate": 9.572795642625537e-06, + "loss": 1.0412, + "step": 13174 + }, + { + "epoch": 0.8004739048544869, + "grad_norm": 0.46538224816322327, + "learning_rate": 9.567163423119968e-06, + "loss": 1.1428, + "step": 13175 + }, + { + "epoch": 0.8005346618871134, + "grad_norm": 0.13478706777095795, + "learning_rate": 9.561532685687818e-06, + "loss": 1.0105, + "step": 13176 + }, + { + "epoch": 0.80059541891974, + "grad_norm": 0.17174896597862244, + "learning_rate": 9.555903430535496e-06, + "loss": 1.1286, + "step": 13177 + }, + { + "epoch": 0.8006561759523665, + "grad_norm": 0.151593416929245, + "learning_rate": 9.550275657869362e-06, + "loss": 1.0432, + "step": 13178 + }, + { + "epoch": 0.800716932984993, + "grad_norm": 0.14841178059577942, + "learning_rate": 9.544649367895692e-06, + "loss": 1.0529, + "step": 13179 + }, + { + "epoch": 0.8007776900176196, + "grad_norm": 0.24656477570533752, + "learning_rate": 9.53902456082072e-06, + "loss": 1.0105, + "step": 13180 + }, + { + "epoch": 0.8008384470502461, + "grad_norm": 0.1308651566505432, + "learning_rate": 9.533401236850626e-06, + "loss": 0.9877, + "step": 13181 + }, + { + "epoch": 0.8008992040828726, + "grad_norm": 0.12034396082162857, + "learning_rate": 9.527779396191528e-06, + "loss": 0.9737, + "step": 13182 + }, + { + "epoch": 0.8009599611154992, + "grad_norm": 0.21325190365314484, + "learning_rate": 9.5221590390495e-06, + "loss": 1.1495, + "step": 13183 + }, + { + "epoch": 0.8010207181481257, + "grad_norm": 0.16935792565345764, + "learning_rate": 9.516540165630556e-06, + "loss": 1.0421, + "step": 13184 + }, + { + "epoch": 0.8010814751807521, + "grad_norm": 0.14506414532661438, + "learning_rate": 9.510922776140663e-06, + "loss": 1.0673, + "step": 13185 + }, + { + "epoch": 0.8011422322133787, + "grad_norm": 0.21398234367370605, + "learning_rate": 9.505306870785713e-06, + "loss": 1.1746, + "step": 13186 + }, + { + "epoch": 0.8012029892460052, + "grad_norm": 0.17942854762077332, + "learning_rate": 9.499692449771574e-06, + "loss": 1.0208, + "step": 13187 + }, + { + "epoch": 0.8012637462786317, + "grad_norm": 0.5118858218193054, + "learning_rate": 9.494079513304021e-06, + "loss": 1.1892, + "step": 13188 + }, + { + "epoch": 0.8013245033112583, + "grad_norm": 0.1561407893896103, + "learning_rate": 9.48846806158884e-06, + "loss": 1.0323, + "step": 13189 + }, + { + "epoch": 0.8013852603438848, + "grad_norm": 0.24394792318344116, + "learning_rate": 9.482858094831681e-06, + "loss": 1.0518, + "step": 13190 + }, + { + "epoch": 0.8014460173765113, + "grad_norm": 2.0651354789733887, + "learning_rate": 9.477249613238186e-06, + "loss": 1.099, + "step": 13191 + }, + { + "epoch": 0.8015067744091379, + "grad_norm": 0.5809557437896729, + "learning_rate": 9.471642617013948e-06, + "loss": 1.2911, + "step": 13192 + }, + { + "epoch": 0.8015675314417644, + "grad_norm": 0.2300063669681549, + "learning_rate": 9.466037106364462e-06, + "loss": 1.128, + "step": 13193 + }, + { + "epoch": 0.8016282884743909, + "grad_norm": 0.17587332427501678, + "learning_rate": 9.46043308149524e-06, + "loss": 1.0409, + "step": 13194 + }, + { + "epoch": 0.8016890455070175, + "grad_norm": 0.19758258759975433, + "learning_rate": 9.454830542611686e-06, + "loss": 1.0353, + "step": 13195 + }, + { + "epoch": 0.801749802539644, + "grad_norm": 0.24176762998104095, + "learning_rate": 9.449229489919147e-06, + "loss": 1.0895, + "step": 13196 + }, + { + "epoch": 0.8018105595722705, + "grad_norm": 0.19048945605754852, + "learning_rate": 9.443629923622937e-06, + "loss": 1.0456, + "step": 13197 + }, + { + "epoch": 0.801871316604897, + "grad_norm": 0.29022201895713806, + "learning_rate": 9.438031843928302e-06, + "loss": 1.2181, + "step": 13198 + }, + { + "epoch": 0.8019320736375235, + "grad_norm": 0.13227444887161255, + "learning_rate": 9.43243525104046e-06, + "loss": 1.0348, + "step": 13199 + }, + { + "epoch": 0.80199283067015, + "grad_norm": 0.22261390089988708, + "learning_rate": 9.426840145164551e-06, + "loss": 1.1417, + "step": 13200 + }, + { + "epoch": 0.8020535877027766, + "grad_norm": 0.23207907378673553, + "learning_rate": 9.421246526505656e-06, + "loss": 1.068, + "step": 13201 + }, + { + "epoch": 0.8021143447354031, + "grad_norm": 0.14394530653953552, + "learning_rate": 9.415654395268826e-06, + "loss": 1.0118, + "step": 13202 + }, + { + "epoch": 0.8021751017680296, + "grad_norm": 1.2191189527511597, + "learning_rate": 9.410063751658998e-06, + "loss": 1.0538, + "step": 13203 + }, + { + "epoch": 0.8022358588006562, + "grad_norm": 0.24817383289337158, + "learning_rate": 9.404474595881146e-06, + "loss": 1.0722, + "step": 13204 + }, + { + "epoch": 0.8022966158332827, + "grad_norm": 0.14038868248462677, + "learning_rate": 9.398886928140127e-06, + "loss": 1.0173, + "step": 13205 + }, + { + "epoch": 0.8023573728659092, + "grad_norm": 0.13583721220493317, + "learning_rate": 9.393300748640755e-06, + "loss": 1.0751, + "step": 13206 + }, + { + "epoch": 0.8024181298985358, + "grad_norm": 0.282446026802063, + "learning_rate": 9.387716057587792e-06, + "loss": 1.2265, + "step": 13207 + }, + { + "epoch": 0.8024788869311623, + "grad_norm": 0.2882765829563141, + "learning_rate": 9.382132855185943e-06, + "loss": 1.2356, + "step": 13208 + }, + { + "epoch": 0.8025396439637889, + "grad_norm": 0.17146949470043182, + "learning_rate": 9.376551141639872e-06, + "loss": 1.0283, + "step": 13209 + }, + { + "epoch": 0.8026004009964154, + "grad_norm": 0.14216981828212738, + "learning_rate": 9.370970917154165e-06, + "loss": 1.046, + "step": 13210 + }, + { + "epoch": 0.8026611580290418, + "grad_norm": 0.1327461451292038, + "learning_rate": 9.36539218193338e-06, + "loss": 1.0389, + "step": 13211 + }, + { + "epoch": 0.8027219150616683, + "grad_norm": 0.2947297692298889, + "learning_rate": 9.359814936181992e-06, + "loss": 1.2184, + "step": 13212 + }, + { + "epoch": 0.8027826720942949, + "grad_norm": 0.4561639428138733, + "learning_rate": 9.354239180104445e-06, + "loss": 1.0143, + "step": 13213 + }, + { + "epoch": 0.8028434291269214, + "grad_norm": 0.13544520735740662, + "learning_rate": 9.348664913905108e-06, + "loss": 1.0249, + "step": 13214 + }, + { + "epoch": 0.802904186159548, + "grad_norm": 0.13862724602222443, + "learning_rate": 9.343092137788329e-06, + "loss": 1.0759, + "step": 13215 + }, + { + "epoch": 0.8029649431921745, + "grad_norm": 0.15352535247802734, + "learning_rate": 9.337520851958376e-06, + "loss": 1.0244, + "step": 13216 + }, + { + "epoch": 0.803025700224801, + "grad_norm": 0.33616775274276733, + "learning_rate": 9.331951056619442e-06, + "loss": 1.1414, + "step": 13217 + }, + { + "epoch": 0.8030864572574276, + "grad_norm": 0.20559456944465637, + "learning_rate": 9.326382751975693e-06, + "loss": 1.0697, + "step": 13218 + }, + { + "epoch": 0.8031472142900541, + "grad_norm": 0.15643100440502167, + "learning_rate": 9.320815938231258e-06, + "loss": 1.0681, + "step": 13219 + }, + { + "epoch": 0.8032079713226806, + "grad_norm": 0.1923975646495819, + "learning_rate": 9.315250615590182e-06, + "loss": 1.1125, + "step": 13220 + }, + { + "epoch": 0.8032687283553072, + "grad_norm": 0.14880885183811188, + "learning_rate": 9.309686784256456e-06, + "loss": 1.0386, + "step": 13221 + }, + { + "epoch": 0.8033294853879337, + "grad_norm": 0.1341559886932373, + "learning_rate": 9.304124444434037e-06, + "loss": 0.9996, + "step": 13222 + }, + { + "epoch": 0.8033902424205602, + "grad_norm": 0.1369275003671646, + "learning_rate": 9.298563596326777e-06, + "loss": 1.0421, + "step": 13223 + }, + { + "epoch": 0.8034509994531867, + "grad_norm": 0.12403661012649536, + "learning_rate": 9.293004240138548e-06, + "loss": 1.0752, + "step": 13224 + }, + { + "epoch": 0.8035117564858132, + "grad_norm": 0.18919095396995544, + "learning_rate": 9.287446376073123e-06, + "loss": 1.1214, + "step": 13225 + }, + { + "epoch": 0.8035725135184397, + "grad_norm": 0.6603615283966064, + "learning_rate": 9.281890004334214e-06, + "loss": 1.1746, + "step": 13226 + }, + { + "epoch": 0.8036332705510663, + "grad_norm": 0.18802323937416077, + "learning_rate": 9.276335125125501e-06, + "loss": 1.1217, + "step": 13227 + }, + { + "epoch": 0.8036940275836928, + "grad_norm": 0.1377108097076416, + "learning_rate": 9.270781738650597e-06, + "loss": 1.0236, + "step": 13228 + }, + { + "epoch": 0.8037547846163193, + "grad_norm": 0.12592169642448425, + "learning_rate": 9.26522984511306e-06, + "loss": 1.0401, + "step": 13229 + }, + { + "epoch": 0.8038155416489459, + "grad_norm": 0.13563883304595947, + "learning_rate": 9.259679444716397e-06, + "loss": 0.9896, + "step": 13230 + }, + { + "epoch": 0.8038762986815724, + "grad_norm": 0.129482701420784, + "learning_rate": 9.254130537664057e-06, + "loss": 1.0187, + "step": 13231 + }, + { + "epoch": 0.8039370557141989, + "grad_norm": 0.19505463540554047, + "learning_rate": 9.248583124159438e-06, + "loss": 1.0927, + "step": 13232 + }, + { + "epoch": 0.8039978127468255, + "grad_norm": 0.13852141797542572, + "learning_rate": 9.24303720440588e-06, + "loss": 1.0773, + "step": 13233 + }, + { + "epoch": 0.804058569779452, + "grad_norm": 0.1287098079919815, + "learning_rate": 9.237492778606665e-06, + "loss": 0.9954, + "step": 13234 + }, + { + "epoch": 0.8041193268120785, + "grad_norm": 0.1506652981042862, + "learning_rate": 9.231949846965038e-06, + "loss": 1.01, + "step": 13235 + }, + { + "epoch": 0.8041800838447051, + "grad_norm": 0.14575420320034027, + "learning_rate": 9.226408409684189e-06, + "loss": 0.9891, + "step": 13236 + }, + { + "epoch": 0.8042408408773315, + "grad_norm": 0.16629280149936676, + "learning_rate": 9.220868466967203e-06, + "loss": 1.113, + "step": 13237 + }, + { + "epoch": 0.804301597909958, + "grad_norm": 0.37589314579963684, + "learning_rate": 9.215330019017165e-06, + "loss": 1.0194, + "step": 13238 + }, + { + "epoch": 0.8043623549425846, + "grad_norm": 0.6745169758796692, + "learning_rate": 9.209793066037081e-06, + "loss": 1.2085, + "step": 13239 + }, + { + "epoch": 0.8044231119752111, + "grad_norm": 0.18978334963321686, + "learning_rate": 9.204257608229928e-06, + "loss": 1.092, + "step": 13240 + }, + { + "epoch": 0.8044838690078376, + "grad_norm": 0.13404926657676697, + "learning_rate": 9.198723645798595e-06, + "loss": 1.0441, + "step": 13241 + }, + { + "epoch": 0.8045446260404642, + "grad_norm": 0.13275112211704254, + "learning_rate": 9.193191178945948e-06, + "loss": 0.9891, + "step": 13242 + }, + { + "epoch": 0.8046053830730907, + "grad_norm": 0.2029099315404892, + "learning_rate": 9.187660207874754e-06, + "loss": 1.1108, + "step": 13243 + }, + { + "epoch": 0.8046661401057172, + "grad_norm": 0.1622231900691986, + "learning_rate": 9.182130732787747e-06, + "loss": 0.9912, + "step": 13244 + }, + { + "epoch": 0.8047268971383438, + "grad_norm": 0.14454412460327148, + "learning_rate": 9.176602753887648e-06, + "loss": 1.0711, + "step": 13245 + }, + { + "epoch": 0.8047876541709703, + "grad_norm": 0.3318987190723419, + "learning_rate": 9.171076271377061e-06, + "loss": 1.1898, + "step": 13246 + }, + { + "epoch": 0.8048484112035968, + "grad_norm": 2.674285650253296, + "learning_rate": 9.165551285458562e-06, + "loss": 1.0837, + "step": 13247 + }, + { + "epoch": 0.8049091682362234, + "grad_norm": 0.29047971963882446, + "learning_rate": 9.16002779633468e-06, + "loss": 1.0366, + "step": 13248 + }, + { + "epoch": 0.8049699252688499, + "grad_norm": 0.14001703262329102, + "learning_rate": 9.154505804207875e-06, + "loss": 1.0174, + "step": 13249 + }, + { + "epoch": 0.8050306823014763, + "grad_norm": 0.16648462414741516, + "learning_rate": 9.14898530928055e-06, + "loss": 1.0345, + "step": 13250 + }, + { + "epoch": 0.8050914393341029, + "grad_norm": 0.12677448987960815, + "learning_rate": 9.143466311755067e-06, + "loss": 0.9752, + "step": 13251 + }, + { + "epoch": 0.8051521963667294, + "grad_norm": 0.21500088274478912, + "learning_rate": 9.137948811833718e-06, + "loss": 1.1437, + "step": 13252 + }, + { + "epoch": 0.805212953399356, + "grad_norm": 0.1623937338590622, + "learning_rate": 9.13243280971876e-06, + "loss": 1.0463, + "step": 13253 + }, + { + "epoch": 0.8052737104319825, + "grad_norm": 0.19963528215885162, + "learning_rate": 9.126918305612376e-06, + "loss": 1.024, + "step": 13254 + }, + { + "epoch": 0.805334467464609, + "grad_norm": 0.26031923294067383, + "learning_rate": 9.12140529971669e-06, + "loss": 1.0991, + "step": 13255 + }, + { + "epoch": 0.8053952244972356, + "grad_norm": 1.8737348318099976, + "learning_rate": 9.11589379223382e-06, + "loss": 1.0637, + "step": 13256 + }, + { + "epoch": 0.8054559815298621, + "grad_norm": 0.13388176262378693, + "learning_rate": 9.11038378336575e-06, + "loss": 1.0089, + "step": 13257 + }, + { + "epoch": 0.8055167385624886, + "grad_norm": 0.16195866465568542, + "learning_rate": 9.104875273314472e-06, + "loss": 1.0736, + "step": 13258 + }, + { + "epoch": 0.8055774955951152, + "grad_norm": 0.39653831720352173, + "learning_rate": 9.099368262281887e-06, + "loss": 1.0378, + "step": 13259 + }, + { + "epoch": 0.8056382526277417, + "grad_norm": 0.20170144736766815, + "learning_rate": 9.093862750469862e-06, + "loss": 1.0889, + "step": 13260 + }, + { + "epoch": 0.8056990096603682, + "grad_norm": 0.15302474796772003, + "learning_rate": 9.08835873808021e-06, + "loss": 1.045, + "step": 13261 + }, + { + "epoch": 0.8057597666929948, + "grad_norm": 0.11412181705236435, + "learning_rate": 9.08285622531469e-06, + "loss": 0.9985, + "step": 13262 + }, + { + "epoch": 0.8058205237256213, + "grad_norm": 0.18597553670406342, + "learning_rate": 9.077355212374972e-06, + "loss": 1.126, + "step": 13263 + }, + { + "epoch": 0.8058812807582477, + "grad_norm": 0.17911003530025482, + "learning_rate": 9.071855699462712e-06, + "loss": 1.0652, + "step": 13264 + }, + { + "epoch": 0.8059420377908743, + "grad_norm": 0.20484846830368042, + "learning_rate": 9.066357686779475e-06, + "loss": 1.1135, + "step": 13265 + }, + { + "epoch": 0.8060027948235008, + "grad_norm": 0.1649646908044815, + "learning_rate": 9.060861174526824e-06, + "loss": 1.0293, + "step": 13266 + }, + { + "epoch": 0.8060635518561273, + "grad_norm": 0.22354108095169067, + "learning_rate": 9.05536616290622e-06, + "loss": 1.1475, + "step": 13267 + }, + { + "epoch": 0.8061243088887539, + "grad_norm": 0.1758941411972046, + "learning_rate": 9.049872652119085e-06, + "loss": 1.019, + "step": 13268 + }, + { + "epoch": 0.8061850659213804, + "grad_norm": 0.23655955493450165, + "learning_rate": 9.044380642366795e-06, + "loss": 1.1358, + "step": 13269 + }, + { + "epoch": 0.8062458229540069, + "grad_norm": 0.6504842042922974, + "learning_rate": 9.038890133850625e-06, + "loss": 1.0315, + "step": 13270 + }, + { + "epoch": 0.8063065799866335, + "grad_norm": 0.26177307963371277, + "learning_rate": 9.033401126771867e-06, + "loss": 1.0649, + "step": 13271 + }, + { + "epoch": 0.80636733701926, + "grad_norm": 0.17449449002742767, + "learning_rate": 9.027913621331707e-06, + "loss": 1.1005, + "step": 13272 + }, + { + "epoch": 0.8064280940518865, + "grad_norm": 5.714333534240723, + "learning_rate": 9.022427617731293e-06, + "loss": 1.074, + "step": 13273 + }, + { + "epoch": 0.8064888510845131, + "grad_norm": 0.1787169724702835, + "learning_rate": 9.016943116171722e-06, + "loss": 1.0287, + "step": 13274 + }, + { + "epoch": 0.8065496081171396, + "grad_norm": 0.2203814834356308, + "learning_rate": 9.011460116854015e-06, + "loss": 1.0245, + "step": 13275 + }, + { + "epoch": 0.8066103651497661, + "grad_norm": 0.2701554000377655, + "learning_rate": 9.005978619979167e-06, + "loss": 0.9926, + "step": 13276 + }, + { + "epoch": 0.8066711221823926, + "grad_norm": 0.23274683952331543, + "learning_rate": 9.000498625748088e-06, + "loss": 1.0639, + "step": 13277 + }, + { + "epoch": 0.8067318792150191, + "grad_norm": 0.22466088831424713, + "learning_rate": 8.995020134361664e-06, + "loss": 1.1279, + "step": 13278 + }, + { + "epoch": 0.8067926362476456, + "grad_norm": 0.21790164709091187, + "learning_rate": 8.989543146020701e-06, + "loss": 1.0845, + "step": 13279 + }, + { + "epoch": 0.8068533932802722, + "grad_norm": 0.4302048683166504, + "learning_rate": 8.98406766092596e-06, + "loss": 1.2063, + "step": 13280 + }, + { + "epoch": 0.8069141503128987, + "grad_norm": 0.13729354739189148, + "learning_rate": 8.978593679278136e-06, + "loss": 1.0238, + "step": 13281 + }, + { + "epoch": 0.8069749073455252, + "grad_norm": 0.21958285570144653, + "learning_rate": 8.973121201277901e-06, + "loss": 1.1869, + "step": 13282 + }, + { + "epoch": 0.8070356643781518, + "grad_norm": 0.11550614982843399, + "learning_rate": 8.967650227125856e-06, + "loss": 1.0032, + "step": 13283 + }, + { + "epoch": 0.8070964214107783, + "grad_norm": 0.44592058658599854, + "learning_rate": 8.962180757022504e-06, + "loss": 1.1494, + "step": 13284 + }, + { + "epoch": 0.8071571784434048, + "grad_norm": 0.15201056003570557, + "learning_rate": 8.956712791168354e-06, + "loss": 1.0633, + "step": 13285 + }, + { + "epoch": 0.8072179354760314, + "grad_norm": 0.19275164604187012, + "learning_rate": 8.951246329763813e-06, + "loss": 1.0531, + "step": 13286 + }, + { + "epoch": 0.8072786925086579, + "grad_norm": 0.14944177865982056, + "learning_rate": 8.945781373009288e-06, + "loss": 1.0413, + "step": 13287 + }, + { + "epoch": 0.8073394495412844, + "grad_norm": 0.18710312247276306, + "learning_rate": 8.940317921105085e-06, + "loss": 1.0978, + "step": 13288 + }, + { + "epoch": 0.807400206573911, + "grad_norm": 0.23309001326560974, + "learning_rate": 8.934855974251472e-06, + "loss": 1.0829, + "step": 13289 + }, + { + "epoch": 0.8074609636065374, + "grad_norm": 0.1827009916305542, + "learning_rate": 8.929395532648643e-06, + "loss": 1.101, + "step": 13290 + }, + { + "epoch": 0.8075217206391639, + "grad_norm": 0.18320326507091522, + "learning_rate": 8.923936596496745e-06, + "loss": 1.0421, + "step": 13291 + }, + { + "epoch": 0.8075824776717905, + "grad_norm": 0.20420828461647034, + "learning_rate": 8.918479165995901e-06, + "loss": 1.043, + "step": 13292 + }, + { + "epoch": 0.807643234704417, + "grad_norm": 0.2002255916595459, + "learning_rate": 8.913023241346147e-06, + "loss": 1.1486, + "step": 13293 + }, + { + "epoch": 0.8077039917370435, + "grad_norm": 0.2503628730773926, + "learning_rate": 8.907568822747469e-06, + "loss": 1.0609, + "step": 13294 + }, + { + "epoch": 0.8077647487696701, + "grad_norm": 0.1662638932466507, + "learning_rate": 8.902115910399794e-06, + "loss": 1.1105, + "step": 13295 + }, + { + "epoch": 0.8078255058022966, + "grad_norm": 0.20438627898693085, + "learning_rate": 8.89666450450301e-06, + "loss": 1.0793, + "step": 13296 + }, + { + "epoch": 0.8078862628349232, + "grad_norm": 0.21877317130565643, + "learning_rate": 8.89121460525693e-06, + "loss": 1.0885, + "step": 13297 + }, + { + "epoch": 0.8079470198675497, + "grad_norm": 0.20138075947761536, + "learning_rate": 8.885766212861329e-06, + "loss": 1.0834, + "step": 13298 + }, + { + "epoch": 0.8080077769001762, + "grad_norm": 0.15688665211200714, + "learning_rate": 8.880319327515913e-06, + "loss": 1.005, + "step": 13299 + }, + { + "epoch": 0.8080685339328028, + "grad_norm": 0.14453080296516418, + "learning_rate": 8.874873949420336e-06, + "loss": 1.0609, + "step": 13300 + }, + { + "epoch": 0.8081292909654293, + "grad_norm": 0.2393285185098648, + "learning_rate": 8.869430078774205e-06, + "loss": 1.1016, + "step": 13301 + }, + { + "epoch": 0.8081900479980558, + "grad_norm": 0.23933076858520508, + "learning_rate": 8.863987715777055e-06, + "loss": 1.0316, + "step": 13302 + }, + { + "epoch": 0.8082508050306823, + "grad_norm": 0.1592375636100769, + "learning_rate": 8.858546860628415e-06, + "loss": 1.077, + "step": 13303 + }, + { + "epoch": 0.8083115620633088, + "grad_norm": 0.18349389731884003, + "learning_rate": 8.853107513527676e-06, + "loss": 1.0886, + "step": 13304 + }, + { + "epoch": 0.8083723190959353, + "grad_norm": 0.17248278856277466, + "learning_rate": 8.84766967467423e-06, + "loss": 1.0643, + "step": 13305 + }, + { + "epoch": 0.8084330761285619, + "grad_norm": 0.1565869003534317, + "learning_rate": 8.842233344267414e-06, + "loss": 1.0845, + "step": 13306 + }, + { + "epoch": 0.8084938331611884, + "grad_norm": 0.5369102358818054, + "learning_rate": 8.836798522506473e-06, + "loss": 1.025, + "step": 13307 + }, + { + "epoch": 0.8085545901938149, + "grad_norm": 2.527047634124756, + "learning_rate": 8.831365209590653e-06, + "loss": 1.01, + "step": 13308 + }, + { + "epoch": 0.8086153472264415, + "grad_norm": 0.22747299075126648, + "learning_rate": 8.825933405719112e-06, + "loss": 1.0744, + "step": 13309 + }, + { + "epoch": 0.808676104259068, + "grad_norm": 0.226390078663826, + "learning_rate": 8.820503111090927e-06, + "loss": 1.0483, + "step": 13310 + }, + { + "epoch": 0.8087368612916945, + "grad_norm": 0.1468891054391861, + "learning_rate": 8.815074325905147e-06, + "loss": 1.0239, + "step": 13311 + }, + { + "epoch": 0.8087976183243211, + "grad_norm": 0.12797042727470398, + "learning_rate": 8.809647050360793e-06, + "loss": 0.9901, + "step": 13312 + }, + { + "epoch": 0.8088583753569476, + "grad_norm": 0.37454310059547424, + "learning_rate": 8.804221284656788e-06, + "loss": 1.0737, + "step": 13313 + }, + { + "epoch": 0.8089191323895741, + "grad_norm": 0.16410471498966217, + "learning_rate": 8.798797028992018e-06, + "loss": 1.0239, + "step": 13314 + }, + { + "epoch": 0.8089798894222007, + "grad_norm": 0.16789379715919495, + "learning_rate": 8.793374283565303e-06, + "loss": 1.0111, + "step": 13315 + }, + { + "epoch": 0.8090406464548271, + "grad_norm": 0.20927859842777252, + "learning_rate": 8.787953048575414e-06, + "loss": 1.0971, + "step": 13316 + }, + { + "epoch": 0.8091014034874536, + "grad_norm": 0.22642621397972107, + "learning_rate": 8.782533324221081e-06, + "loss": 1.2081, + "step": 13317 + }, + { + "epoch": 0.8091621605200802, + "grad_norm": 0.25023552775382996, + "learning_rate": 8.77711511070095e-06, + "loss": 1.1731, + "step": 13318 + }, + { + "epoch": 0.8092229175527067, + "grad_norm": 0.15797598659992218, + "learning_rate": 8.771698408213635e-06, + "loss": 1.0491, + "step": 13319 + }, + { + "epoch": 0.8092836745853332, + "grad_norm": 0.16422073543071747, + "learning_rate": 8.766283216957689e-06, + "loss": 1.0507, + "step": 13320 + }, + { + "epoch": 0.8093444316179598, + "grad_norm": 0.23154456913471222, + "learning_rate": 8.760869537131594e-06, + "loss": 1.0266, + "step": 13321 + }, + { + "epoch": 0.8094051886505863, + "grad_norm": 0.2109970897436142, + "learning_rate": 8.755457368933801e-06, + "loss": 1.1519, + "step": 13322 + }, + { + "epoch": 0.8094659456832128, + "grad_norm": 0.20074287056922913, + "learning_rate": 8.750046712562693e-06, + "loss": 1.1237, + "step": 13323 + }, + { + "epoch": 0.8095267027158394, + "grad_norm": 0.1959058940410614, + "learning_rate": 8.74463756821659e-06, + "loss": 0.9941, + "step": 13324 + }, + { + "epoch": 0.8095874597484659, + "grad_norm": 1.0562632083892822, + "learning_rate": 8.73922993609378e-06, + "loss": 1.0978, + "step": 13325 + }, + { + "epoch": 0.8096482167810924, + "grad_norm": 0.16476699709892273, + "learning_rate": 8.733823816392466e-06, + "loss": 1.0027, + "step": 13326 + }, + { + "epoch": 0.809708973813719, + "grad_norm": 0.9567370414733887, + "learning_rate": 8.7284192093108e-06, + "loss": 1.0147, + "step": 13327 + }, + { + "epoch": 0.8097697308463455, + "grad_norm": 0.1337011605501175, + "learning_rate": 8.723016115046928e-06, + "loss": 1.018, + "step": 13328 + }, + { + "epoch": 0.8098304878789719, + "grad_norm": 0.16990529000759125, + "learning_rate": 8.717614533798885e-06, + "loss": 1.1262, + "step": 13329 + }, + { + "epoch": 0.8098912449115985, + "grad_norm": 0.1297598034143448, + "learning_rate": 8.71221446576465e-06, + "loss": 1.0592, + "step": 13330 + }, + { + "epoch": 0.809952001944225, + "grad_norm": 0.317297101020813, + "learning_rate": 8.706815911142175e-06, + "loss": 1.0905, + "step": 13331 + }, + { + "epoch": 0.8100127589768515, + "grad_norm": 0.16625577211380005, + "learning_rate": 8.70141887012933e-06, + "loss": 0.9879, + "step": 13332 + }, + { + "epoch": 0.8100735160094781, + "grad_norm": 0.1807892918586731, + "learning_rate": 8.696023342923976e-06, + "loss": 0.9726, + "step": 13333 + }, + { + "epoch": 0.8101342730421046, + "grad_norm": 0.1418323516845703, + "learning_rate": 8.690629329723865e-06, + "loss": 1.0661, + "step": 13334 + }, + { + "epoch": 0.8101950300747311, + "grad_norm": 0.17416544258594513, + "learning_rate": 8.685236830726729e-06, + "loss": 1.1009, + "step": 13335 + }, + { + "epoch": 0.8102557871073577, + "grad_norm": 0.18850530683994293, + "learning_rate": 8.679845846130225e-06, + "loss": 1.0819, + "step": 13336 + }, + { + "epoch": 0.8103165441399842, + "grad_norm": 0.21120040118694305, + "learning_rate": 8.67445637613194e-06, + "loss": 1.0953, + "step": 13337 + }, + { + "epoch": 0.8103773011726108, + "grad_norm": 0.17070598900318146, + "learning_rate": 8.669068420929454e-06, + "loss": 1.0797, + "step": 13338 + }, + { + "epoch": 0.8104380582052373, + "grad_norm": 0.1434313803911209, + "learning_rate": 8.663681980720257e-06, + "loss": 1.0059, + "step": 13339 + }, + { + "epoch": 0.8104988152378638, + "grad_norm": 0.13803617656230927, + "learning_rate": 8.658297055701786e-06, + "loss": 1.0015, + "step": 13340 + }, + { + "epoch": 0.8105595722704904, + "grad_norm": 0.23951227962970734, + "learning_rate": 8.652913646071425e-06, + "loss": 1.0013, + "step": 13341 + }, + { + "epoch": 0.8106203293031168, + "grad_norm": 0.20786088705062866, + "learning_rate": 8.647531752026511e-06, + "loss": 1.1219, + "step": 13342 + }, + { + "epoch": 0.8106810863357433, + "grad_norm": 0.18287093937397003, + "learning_rate": 8.642151373764312e-06, + "loss": 1.1411, + "step": 13343 + }, + { + "epoch": 0.8107418433683699, + "grad_norm": 0.15448804199695587, + "learning_rate": 8.63677251148205e-06, + "loss": 1.0559, + "step": 13344 + }, + { + "epoch": 0.8108026004009964, + "grad_norm": 0.31872814893722534, + "learning_rate": 8.631395165376888e-06, + "loss": 1.127, + "step": 13345 + }, + { + "epoch": 0.8108633574336229, + "grad_norm": 0.26476767659187317, + "learning_rate": 8.62601933564593e-06, + "loss": 1.2587, + "step": 13346 + }, + { + "epoch": 0.8109241144662495, + "grad_norm": 0.19465158879756927, + "learning_rate": 8.62064502248623e-06, + "loss": 1.1149, + "step": 13347 + }, + { + "epoch": 0.810984871498876, + "grad_norm": 0.18735066056251526, + "learning_rate": 8.615272226094773e-06, + "loss": 1.0135, + "step": 13348 + }, + { + "epoch": 0.8110456285315025, + "grad_norm": 0.228982612490654, + "learning_rate": 8.609900946668537e-06, + "loss": 1.1133, + "step": 13349 + }, + { + "epoch": 0.8111063855641291, + "grad_norm": 0.12470275163650513, + "learning_rate": 8.604531184404368e-06, + "loss": 0.9894, + "step": 13350 + }, + { + "epoch": 0.8111671425967556, + "grad_norm": 0.13641098141670227, + "learning_rate": 8.599162939499111e-06, + "loss": 0.9884, + "step": 13351 + }, + { + "epoch": 0.8112278996293821, + "grad_norm": 0.12195431441068649, + "learning_rate": 8.59379621214954e-06, + "loss": 1.0245, + "step": 13352 + }, + { + "epoch": 0.8112886566620087, + "grad_norm": 0.36675962805747986, + "learning_rate": 8.588431002552361e-06, + "loss": 1.0472, + "step": 13353 + }, + { + "epoch": 0.8113494136946352, + "grad_norm": 0.21596698462963104, + "learning_rate": 8.583067310904264e-06, + "loss": 1.1098, + "step": 13354 + }, + { + "epoch": 0.8114101707272616, + "grad_norm": 0.16716881096363068, + "learning_rate": 8.577705137401837e-06, + "loss": 1.034, + "step": 13355 + }, + { + "epoch": 0.8114709277598882, + "grad_norm": 0.19974572956562042, + "learning_rate": 8.572344482241645e-06, + "loss": 1.0945, + "step": 13356 + }, + { + "epoch": 0.8115316847925147, + "grad_norm": 0.17221243679523468, + "learning_rate": 8.56698534562016e-06, + "loss": 1.0267, + "step": 13357 + }, + { + "epoch": 0.8115924418251412, + "grad_norm": 0.23640581965446472, + "learning_rate": 8.561627727733828e-06, + "loss": 1.132, + "step": 13358 + }, + { + "epoch": 0.8116531988577678, + "grad_norm": 0.24394047260284424, + "learning_rate": 8.556271628779056e-06, + "loss": 1.0503, + "step": 13359 + }, + { + "epoch": 0.8117139558903943, + "grad_norm": 2.147752523422241, + "learning_rate": 8.550917048952151e-06, + "loss": 1.2038, + "step": 13360 + }, + { + "epoch": 0.8117747129230208, + "grad_norm": 0.21752582490444183, + "learning_rate": 8.545563988449401e-06, + "loss": 1.1066, + "step": 13361 + }, + { + "epoch": 0.8118354699556474, + "grad_norm": 0.15506352484226227, + "learning_rate": 8.54021244746701e-06, + "loss": 1.2239, + "step": 13362 + }, + { + "epoch": 0.8118962269882739, + "grad_norm": 0.14957040548324585, + "learning_rate": 8.534862426201151e-06, + "loss": 1.035, + "step": 13363 + }, + { + "epoch": 0.8119569840209004, + "grad_norm": 0.20051157474517822, + "learning_rate": 8.52951392484792e-06, + "loss": 1.0851, + "step": 13364 + }, + { + "epoch": 0.812017741053527, + "grad_norm": 0.13006332516670227, + "learning_rate": 8.524166943603373e-06, + "loss": 1.0381, + "step": 13365 + }, + { + "epoch": 0.8120784980861535, + "grad_norm": 0.2170485556125641, + "learning_rate": 8.518821482663508e-06, + "loss": 1.0914, + "step": 13366 + }, + { + "epoch": 0.81213925511878, + "grad_norm": 0.14561296999454498, + "learning_rate": 8.513477542224258e-06, + "loss": 0.9718, + "step": 13367 + }, + { + "epoch": 0.8122000121514066, + "grad_norm": 0.23806257545948029, + "learning_rate": 8.508135122481503e-06, + "loss": 1.0701, + "step": 13368 + }, + { + "epoch": 0.812260769184033, + "grad_norm": 0.23510535061359406, + "learning_rate": 8.502794223631067e-06, + "loss": 1.0456, + "step": 13369 + }, + { + "epoch": 0.8123215262166595, + "grad_norm": 0.1686597764492035, + "learning_rate": 8.497454845868752e-06, + "loss": 1.0212, + "step": 13370 + }, + { + "epoch": 0.8123822832492861, + "grad_norm": 0.20919838547706604, + "learning_rate": 8.492116989390243e-06, + "loss": 1.0345, + "step": 13371 + }, + { + "epoch": 0.8124430402819126, + "grad_norm": 0.19253161549568176, + "learning_rate": 8.486780654391207e-06, + "loss": 1.0338, + "step": 13372 + }, + { + "epoch": 0.8125037973145391, + "grad_norm": 0.14951792359352112, + "learning_rate": 8.481445841067253e-06, + "loss": 1.0479, + "step": 13373 + }, + { + "epoch": 0.8125645543471657, + "grad_norm": 0.20469342172145844, + "learning_rate": 8.476112549613912e-06, + "loss": 1.0689, + "step": 13374 + }, + { + "epoch": 0.8126253113797922, + "grad_norm": 0.13289767503738403, + "learning_rate": 8.470780780226705e-06, + "loss": 1.0413, + "step": 13375 + }, + { + "epoch": 0.8126860684124187, + "grad_norm": 0.12832863628864288, + "learning_rate": 8.465450533101072e-06, + "loss": 1.0966, + "step": 13376 + }, + { + "epoch": 0.8127468254450453, + "grad_norm": 0.13569071888923645, + "learning_rate": 8.460121808432358e-06, + "loss": 1.0286, + "step": 13377 + }, + { + "epoch": 0.8128075824776718, + "grad_norm": 0.32198190689086914, + "learning_rate": 8.454794606415911e-06, + "loss": 1.0816, + "step": 13378 + }, + { + "epoch": 0.8128683395102984, + "grad_norm": 0.1469375342130661, + "learning_rate": 8.44946892724699e-06, + "loss": 1.0317, + "step": 13379 + }, + { + "epoch": 0.8129290965429249, + "grad_norm": 0.2232426553964615, + "learning_rate": 8.444144771120826e-06, + "loss": 1.1807, + "step": 13380 + }, + { + "epoch": 0.8129898535755514, + "grad_norm": 0.27788764238357544, + "learning_rate": 8.438822138232572e-06, + "loss": 1.109, + "step": 13381 + }, + { + "epoch": 0.8130506106081778, + "grad_norm": 0.12954086065292358, + "learning_rate": 8.433501028777318e-06, + "loss": 1.0132, + "step": 13382 + }, + { + "epoch": 0.8131113676408044, + "grad_norm": 0.14362236857414246, + "learning_rate": 8.42818144295014e-06, + "loss": 1.0398, + "step": 13383 + }, + { + "epoch": 0.8131721246734309, + "grad_norm": 0.14357590675354004, + "learning_rate": 8.422863380945977e-06, + "loss": 0.9966, + "step": 13384 + }, + { + "epoch": 0.8132328817060575, + "grad_norm": 0.4781111180782318, + "learning_rate": 8.417546842959806e-06, + "loss": 1.109, + "step": 13385 + }, + { + "epoch": 0.813293638738684, + "grad_norm": 0.23624849319458008, + "learning_rate": 8.412231829186495e-06, + "loss": 1.1504, + "step": 13386 + }, + { + "epoch": 0.8133543957713105, + "grad_norm": 0.1303945928812027, + "learning_rate": 8.406918339820864e-06, + "loss": 1.0027, + "step": 13387 + }, + { + "epoch": 0.8134151528039371, + "grad_norm": 0.18421292304992676, + "learning_rate": 8.40160637505768e-06, + "loss": 1.0976, + "step": 13388 + }, + { + "epoch": 0.8134759098365636, + "grad_norm": 0.15239883959293365, + "learning_rate": 8.396295935091659e-06, + "loss": 0.9529, + "step": 13389 + }, + { + "epoch": 0.8135366668691901, + "grad_norm": 1.5858125686645508, + "learning_rate": 8.390987020117447e-06, + "loss": 1.0573, + "step": 13390 + }, + { + "epoch": 0.8135974239018167, + "grad_norm": 0.1807781159877777, + "learning_rate": 8.385679630329652e-06, + "loss": 0.9892, + "step": 13391 + }, + { + "epoch": 0.8136581809344432, + "grad_norm": 0.13312286138534546, + "learning_rate": 8.380373765922812e-06, + "loss": 1.0448, + "step": 13392 + }, + { + "epoch": 0.8137189379670697, + "grad_norm": 0.13373954594135284, + "learning_rate": 8.37506942709142e-06, + "loss": 1.011, + "step": 13393 + }, + { + "epoch": 0.8137796949996963, + "grad_norm": 0.47429755330085754, + "learning_rate": 8.3697666140299e-06, + "loss": 1.1608, + "step": 13394 + }, + { + "epoch": 0.8138404520323227, + "grad_norm": 0.15928952395915985, + "learning_rate": 8.364465326932619e-06, + "loss": 1.0253, + "step": 13395 + }, + { + "epoch": 0.8139012090649492, + "grad_norm": 0.13044984638690948, + "learning_rate": 8.359165565993938e-06, + "loss": 1.0391, + "step": 13396 + }, + { + "epoch": 0.8139619660975758, + "grad_norm": 0.48919421434402466, + "learning_rate": 8.353867331408078e-06, + "loss": 1.0439, + "step": 13397 + }, + { + "epoch": 0.8140227231302023, + "grad_norm": 0.30776306986808777, + "learning_rate": 8.348570623369262e-06, + "loss": 1.1021, + "step": 13398 + }, + { + "epoch": 0.8140834801628288, + "grad_norm": 0.2298887073993683, + "learning_rate": 8.34327544207164e-06, + "loss": 1.0667, + "step": 13399 + }, + { + "epoch": 0.8141442371954554, + "grad_norm": 0.16253195703029633, + "learning_rate": 8.337981787709297e-06, + "loss": 1.0565, + "step": 13400 + }, + { + "epoch": 0.8142049942280819, + "grad_norm": 0.6355108618736267, + "learning_rate": 8.332689660476294e-06, + "loss": 1.0651, + "step": 13401 + }, + { + "epoch": 0.8142657512607084, + "grad_norm": 0.21033558249473572, + "learning_rate": 8.32739906056661e-06, + "loss": 1.1385, + "step": 13402 + }, + { + "epoch": 0.814326508293335, + "grad_norm": 0.13816000521183014, + "learning_rate": 8.322109988174177e-06, + "loss": 1.0152, + "step": 13403 + }, + { + "epoch": 0.8143872653259615, + "grad_norm": 0.15417973697185516, + "learning_rate": 8.31682244349285e-06, + "loss": 0.9743, + "step": 13404 + }, + { + "epoch": 0.814448022358588, + "grad_norm": 0.22606703639030457, + "learning_rate": 8.311536426716437e-06, + "loss": 1.0289, + "step": 13405 + }, + { + "epoch": 0.8145087793912146, + "grad_norm": 0.1775621920824051, + "learning_rate": 8.306251938038728e-06, + "loss": 1.1048, + "step": 13406 + }, + { + "epoch": 0.8145695364238411, + "grad_norm": 0.23691579699516296, + "learning_rate": 8.300968977653417e-06, + "loss": 1.0154, + "step": 13407 + }, + { + "epoch": 0.8146302934564675, + "grad_norm": 0.11363023519515991, + "learning_rate": 8.295687545754143e-06, + "loss": 1.0269, + "step": 13408 + }, + { + "epoch": 0.8146910504890941, + "grad_norm": 0.19208082556724548, + "learning_rate": 8.29040764253451e-06, + "loss": 0.9949, + "step": 13409 + }, + { + "epoch": 0.8147518075217206, + "grad_norm": 0.2103991061449051, + "learning_rate": 8.285129268188042e-06, + "loss": 1.1024, + "step": 13410 + }, + { + "epoch": 0.8148125645543471, + "grad_norm": 0.27126067876815796, + "learning_rate": 8.279852422908225e-06, + "loss": 1.0477, + "step": 13411 + }, + { + "epoch": 0.8148733215869737, + "grad_norm": 6.383580207824707, + "learning_rate": 8.274577106888481e-06, + "loss": 1.2579, + "step": 13412 + }, + { + "epoch": 0.8149340786196002, + "grad_norm": 0.14521083235740662, + "learning_rate": 8.26930332032218e-06, + "loss": 1.0221, + "step": 13413 + }, + { + "epoch": 0.8149948356522267, + "grad_norm": 0.1918259859085083, + "learning_rate": 8.264031063402627e-06, + "loss": 1.0048, + "step": 13414 + }, + { + "epoch": 0.8150555926848533, + "grad_norm": 6.2732038497924805, + "learning_rate": 8.25876033632309e-06, + "loss": 1.0932, + "step": 13415 + }, + { + "epoch": 0.8151163497174798, + "grad_norm": 0.19042551517486572, + "learning_rate": 8.253491139276759e-06, + "loss": 1.0493, + "step": 13416 + }, + { + "epoch": 0.8151771067501064, + "grad_norm": 1.6205908060073853, + "learning_rate": 8.248223472456778e-06, + "loss": 1.0339, + "step": 13417 + }, + { + "epoch": 0.8152378637827329, + "grad_norm": 0.5766290426254272, + "learning_rate": 8.24295733605624e-06, + "loss": 1.2525, + "step": 13418 + }, + { + "epoch": 0.8152986208153594, + "grad_norm": 0.2209654599428177, + "learning_rate": 8.23769273026817e-06, + "loss": 1.2182, + "step": 13419 + }, + { + "epoch": 0.815359377847986, + "grad_norm": 0.11979575455188751, + "learning_rate": 8.232429655285529e-06, + "loss": 1.0034, + "step": 13420 + }, + { + "epoch": 0.8154201348806124, + "grad_norm": 0.13452030718326569, + "learning_rate": 8.22716811130127e-06, + "loss": 1.0501, + "step": 13421 + }, + { + "epoch": 0.8154808919132389, + "grad_norm": 0.1485597938299179, + "learning_rate": 8.221908098508235e-06, + "loss": 1.0809, + "step": 13422 + }, + { + "epoch": 0.8155416489458654, + "grad_norm": 0.14308927953243256, + "learning_rate": 8.216649617099248e-06, + "loss": 1.0443, + "step": 13423 + }, + { + "epoch": 0.815602405978492, + "grad_norm": 0.14076745510101318, + "learning_rate": 8.211392667267037e-06, + "loss": 1.0371, + "step": 13424 + }, + { + "epoch": 0.8156631630111185, + "grad_norm": 0.14007572829723358, + "learning_rate": 8.206137249204288e-06, + "loss": 1.0568, + "step": 13425 + }, + { + "epoch": 0.815723920043745, + "grad_norm": 0.20115649700164795, + "learning_rate": 8.200883363103673e-06, + "loss": 1.0858, + "step": 13426 + }, + { + "epoch": 0.8157846770763716, + "grad_norm": 0.13836358487606049, + "learning_rate": 8.195631009157756e-06, + "loss": 1.0292, + "step": 13427 + }, + { + "epoch": 0.8158454341089981, + "grad_norm": 0.15522350370883942, + "learning_rate": 8.190380187559065e-06, + "loss": 1.0545, + "step": 13428 + }, + { + "epoch": 0.8159061911416247, + "grad_norm": 0.13885009288787842, + "learning_rate": 8.185130898500071e-06, + "loss": 1.0354, + "step": 13429 + }, + { + "epoch": 0.8159669481742512, + "grad_norm": 0.20412012934684753, + "learning_rate": 8.179883142173184e-06, + "loss": 1.2068, + "step": 13430 + }, + { + "epoch": 0.8160277052068777, + "grad_norm": 0.1694025844335556, + "learning_rate": 8.174636918770768e-06, + "loss": 0.9893, + "step": 13431 + }, + { + "epoch": 0.8160884622395043, + "grad_norm": 0.28125208616256714, + "learning_rate": 8.169392228485111e-06, + "loss": 1.2806, + "step": 13432 + }, + { + "epoch": 0.8161492192721308, + "grad_norm": 0.20255568623542786, + "learning_rate": 8.164149071508476e-06, + "loss": 1.1326, + "step": 13433 + }, + { + "epoch": 0.8162099763047572, + "grad_norm": 0.23800845444202423, + "learning_rate": 8.158907448033043e-06, + "loss": 1.2508, + "step": 13434 + }, + { + "epoch": 0.8162707333373838, + "grad_norm": 0.20871183276176453, + "learning_rate": 8.153667358250944e-06, + "loss": 1.0753, + "step": 13435 + }, + { + "epoch": 0.8163314903700103, + "grad_norm": 0.15837250649929047, + "learning_rate": 8.148428802354246e-06, + "loss": 1.006, + "step": 13436 + }, + { + "epoch": 0.8163922474026368, + "grad_norm": 0.14213843643665314, + "learning_rate": 8.143191780535003e-06, + "loss": 1.0264, + "step": 13437 + }, + { + "epoch": 0.8164530044352634, + "grad_norm": 0.18430973589420319, + "learning_rate": 8.137956292985144e-06, + "loss": 1.0485, + "step": 13438 + }, + { + "epoch": 0.8165137614678899, + "grad_norm": 0.2031368911266327, + "learning_rate": 8.132722339896588e-06, + "loss": 1.0085, + "step": 13439 + }, + { + "epoch": 0.8165745185005164, + "grad_norm": 1.0678596496582031, + "learning_rate": 8.127489921461185e-06, + "loss": 1.0639, + "step": 13440 + }, + { + "epoch": 0.816635275533143, + "grad_norm": 0.22343748807907104, + "learning_rate": 8.122259037870728e-06, + "loss": 1.1113, + "step": 13441 + }, + { + "epoch": 0.8166960325657695, + "grad_norm": 0.1397341936826706, + "learning_rate": 8.117029689316962e-06, + "loss": 1.0353, + "step": 13442 + }, + { + "epoch": 0.816756789598396, + "grad_norm": 0.1943097561597824, + "learning_rate": 8.111801875991593e-06, + "loss": 1.0561, + "step": 13443 + }, + { + "epoch": 0.8168175466310226, + "grad_norm": 0.19787752628326416, + "learning_rate": 8.106575598086203e-06, + "loss": 1.0429, + "step": 13444 + }, + { + "epoch": 0.8168783036636491, + "grad_norm": 0.27540239691734314, + "learning_rate": 8.101350855792383e-06, + "loss": 1.1675, + "step": 13445 + }, + { + "epoch": 0.8169390606962756, + "grad_norm": 0.24833445250988007, + "learning_rate": 8.096127649301638e-06, + "loss": 1.1424, + "step": 13446 + }, + { + "epoch": 0.8169998177289021, + "grad_norm": 0.12406907975673676, + "learning_rate": 8.09090597880544e-06, + "loss": 1.002, + "step": 13447 + }, + { + "epoch": 0.8170605747615286, + "grad_norm": 0.21743278205394745, + "learning_rate": 8.08568584449519e-06, + "loss": 1.1478, + "step": 13448 + }, + { + "epoch": 0.8171213317941551, + "grad_norm": 0.16476434469223022, + "learning_rate": 8.080467246562224e-06, + "loss": 1.0933, + "step": 13449 + }, + { + "epoch": 0.8171820888267817, + "grad_norm": 0.3470216989517212, + "learning_rate": 8.075250185197847e-06, + "loss": 1.0226, + "step": 13450 + }, + { + "epoch": 0.8172428458594082, + "grad_norm": 0.15369781851768494, + "learning_rate": 8.07003466059325e-06, + "loss": 1.0332, + "step": 13451 + }, + { + "epoch": 0.8173036028920347, + "grad_norm": 2.800419807434082, + "learning_rate": 8.064820672939654e-06, + "loss": 1.1167, + "step": 13452 + }, + { + "epoch": 0.8173643599246613, + "grad_norm": 0.2149336189031601, + "learning_rate": 8.059608222428162e-06, + "loss": 1.1375, + "step": 13453 + }, + { + "epoch": 0.8174251169572878, + "grad_norm": 0.23010069131851196, + "learning_rate": 8.054397309249839e-06, + "loss": 1.1677, + "step": 13454 + }, + { + "epoch": 0.8174858739899143, + "grad_norm": 0.1422623246908188, + "learning_rate": 8.049187933595686e-06, + "loss": 1.037, + "step": 13455 + }, + { + "epoch": 0.8175466310225409, + "grad_norm": 2.643388032913208, + "learning_rate": 8.043980095656662e-06, + "loss": 1.1552, + "step": 13456 + }, + { + "epoch": 0.8176073880551674, + "grad_norm": 0.22394973039627075, + "learning_rate": 8.038773795623656e-06, + "loss": 1.0905, + "step": 13457 + }, + { + "epoch": 0.817668145087794, + "grad_norm": 0.2807651162147522, + "learning_rate": 8.03356903368751e-06, + "loss": 1.1082, + "step": 13458 + }, + { + "epoch": 0.8177289021204205, + "grad_norm": 0.17401205003261566, + "learning_rate": 8.028365810039002e-06, + "loss": 1.0303, + "step": 13459 + }, + { + "epoch": 0.8177896591530469, + "grad_norm": 0.18030008673667908, + "learning_rate": 8.023164124868855e-06, + "loss": 1.0942, + "step": 13460 + }, + { + "epoch": 0.8178504161856734, + "grad_norm": 0.13947100937366486, + "learning_rate": 8.017963978367742e-06, + "loss": 1.0704, + "step": 13461 + }, + { + "epoch": 0.8179111732183, + "grad_norm": 0.11821171641349792, + "learning_rate": 8.012765370726261e-06, + "loss": 0.9945, + "step": 13462 + }, + { + "epoch": 0.8179719302509265, + "grad_norm": 0.7368493676185608, + "learning_rate": 8.007568302135011e-06, + "loss": 1.0885, + "step": 13463 + }, + { + "epoch": 0.818032687283553, + "grad_norm": 0.1966831386089325, + "learning_rate": 8.002372772784445e-06, + "loss": 1.0283, + "step": 13464 + }, + { + "epoch": 0.8180934443161796, + "grad_norm": 1.002602219581604, + "learning_rate": 7.997178782865023e-06, + "loss": 1.155, + "step": 13465 + }, + { + "epoch": 0.8181542013488061, + "grad_norm": 0.16526658833026886, + "learning_rate": 7.99198633256713e-06, + "loss": 1.0993, + "step": 13466 + }, + { + "epoch": 0.8182149583814327, + "grad_norm": 0.23592707514762878, + "learning_rate": 7.986795422081089e-06, + "loss": 1.1252, + "step": 13467 + }, + { + "epoch": 0.8182757154140592, + "grad_norm": 0.13643614947795868, + "learning_rate": 7.981606051597196e-06, + "loss": 0.9872, + "step": 13468 + }, + { + "epoch": 0.8183364724466857, + "grad_norm": 1.0766068696975708, + "learning_rate": 7.976418221305653e-06, + "loss": 1.1702, + "step": 13469 + }, + { + "epoch": 0.8183972294793123, + "grad_norm": 0.14028963446617126, + "learning_rate": 7.971231931396638e-06, + "loss": 1.0182, + "step": 13470 + }, + { + "epoch": 0.8184579865119388, + "grad_norm": 0.1792718768119812, + "learning_rate": 7.966047182060226e-06, + "loss": 1.0827, + "step": 13471 + }, + { + "epoch": 0.8185187435445653, + "grad_norm": 0.14816521108150482, + "learning_rate": 7.960863973486466e-06, + "loss": 1.0449, + "step": 13472 + }, + { + "epoch": 0.8185795005771919, + "grad_norm": 0.14968383312225342, + "learning_rate": 7.955682305865381e-06, + "loss": 1.0476, + "step": 13473 + }, + { + "epoch": 0.8186402576098183, + "grad_norm": 0.2761284410953522, + "learning_rate": 7.950502179386882e-06, + "loss": 1.1738, + "step": 13474 + }, + { + "epoch": 0.8187010146424448, + "grad_norm": 0.23043610155582428, + "learning_rate": 7.945323594240856e-06, + "loss": 1.0094, + "step": 13475 + }, + { + "epoch": 0.8187617716750714, + "grad_norm": 0.14452695846557617, + "learning_rate": 7.940146550617123e-06, + "loss": 1.0546, + "step": 13476 + }, + { + "epoch": 0.8188225287076979, + "grad_norm": 0.12471524626016617, + "learning_rate": 7.934971048705447e-06, + "loss": 1.0416, + "step": 13477 + }, + { + "epoch": 0.8188832857403244, + "grad_norm": 0.1552859991788864, + "learning_rate": 7.929797088695539e-06, + "loss": 1.0015, + "step": 13478 + }, + { + "epoch": 0.818944042772951, + "grad_norm": 0.14485594630241394, + "learning_rate": 7.924624670777047e-06, + "loss": 1.0263, + "step": 13479 + }, + { + "epoch": 0.8190047998055775, + "grad_norm": 0.1277191936969757, + "learning_rate": 7.919453795139576e-06, + "loss": 1.0427, + "step": 13480 + }, + { + "epoch": 0.819065556838204, + "grad_norm": 0.12876556813716888, + "learning_rate": 7.91428446197266e-06, + "loss": 0.9842, + "step": 13481 + }, + { + "epoch": 0.8191263138708306, + "grad_norm": 0.18227015435695648, + "learning_rate": 7.909116671465778e-06, + "loss": 1.0851, + "step": 13482 + }, + { + "epoch": 0.8191870709034571, + "grad_norm": 0.25828421115875244, + "learning_rate": 7.903950423808365e-06, + "loss": 1.1356, + "step": 13483 + }, + { + "epoch": 0.8192478279360836, + "grad_norm": 0.12338236719369888, + "learning_rate": 7.898785719189784e-06, + "loss": 1.038, + "step": 13484 + }, + { + "epoch": 0.8193085849687102, + "grad_norm": 0.13461928069591522, + "learning_rate": 7.893622557799346e-06, + "loss": 1.0035, + "step": 13485 + }, + { + "epoch": 0.8193693420013367, + "grad_norm": 3.3639883995056152, + "learning_rate": 7.88846093982632e-06, + "loss": 1.0893, + "step": 13486 + }, + { + "epoch": 0.8194300990339631, + "grad_norm": 0.23191241919994354, + "learning_rate": 7.883300865459891e-06, + "loss": 1.2116, + "step": 13487 + }, + { + "epoch": 0.8194908560665897, + "grad_norm": 0.19044698774814606, + "learning_rate": 7.878142334889205e-06, + "loss": 1.1618, + "step": 13488 + }, + { + "epoch": 0.8195516130992162, + "grad_norm": 0.20200209319591522, + "learning_rate": 7.872985348303358e-06, + "loss": 1.0979, + "step": 13489 + }, + { + "epoch": 0.8196123701318427, + "grad_norm": 0.132410928606987, + "learning_rate": 7.867829905891394e-06, + "loss": 0.9947, + "step": 13490 + }, + { + "epoch": 0.8196731271644693, + "grad_norm": 0.2555529475212097, + "learning_rate": 7.862676007842257e-06, + "loss": 1.0396, + "step": 13491 + }, + { + "epoch": 0.8197338841970958, + "grad_norm": 3.7266504764556885, + "learning_rate": 7.857523654344873e-06, + "loss": 1.0859, + "step": 13492 + }, + { + "epoch": 0.8197946412297223, + "grad_norm": 0.207106351852417, + "learning_rate": 7.85237284558809e-06, + "loss": 1.0741, + "step": 13493 + }, + { + "epoch": 0.8198553982623489, + "grad_norm": 0.17141056060791016, + "learning_rate": 7.847223581760744e-06, + "loss": 1.0287, + "step": 13494 + }, + { + "epoch": 0.8199161552949754, + "grad_norm": 0.15411804616451263, + "learning_rate": 7.842075863051567e-06, + "loss": 1.0283, + "step": 13495 + }, + { + "epoch": 0.819976912327602, + "grad_norm": 0.20126120746135712, + "learning_rate": 7.836929689649247e-06, + "loss": 1.1078, + "step": 13496 + }, + { + "epoch": 0.8200376693602285, + "grad_norm": 1.6396018266677856, + "learning_rate": 7.831785061742431e-06, + "loss": 1.0025, + "step": 13497 + }, + { + "epoch": 0.820098426392855, + "grad_norm": 0.16763454675674438, + "learning_rate": 7.826641979519655e-06, + "loss": 1.0789, + "step": 13498 + }, + { + "epoch": 0.8201591834254816, + "grad_norm": 0.5953937768936157, + "learning_rate": 7.821500443169487e-06, + "loss": 0.9732, + "step": 13499 + }, + { + "epoch": 0.820219940458108, + "grad_norm": 0.23744195699691772, + "learning_rate": 7.816360452880373e-06, + "loss": 1.0602, + "step": 13500 + }, + { + "epoch": 0.8202806974907345, + "grad_norm": 0.40224963426589966, + "learning_rate": 7.81122200884072e-06, + "loss": 1.1092, + "step": 13501 + }, + { + "epoch": 0.820341454523361, + "grad_norm": 0.5988378524780273, + "learning_rate": 7.80608511123888e-06, + "loss": 1.2588, + "step": 13502 + }, + { + "epoch": 0.8204022115559876, + "grad_norm": 0.18556249141693115, + "learning_rate": 7.800949760263144e-06, + "loss": 1.2559, + "step": 13503 + }, + { + "epoch": 0.8204629685886141, + "grad_norm": 0.2159474939107895, + "learning_rate": 7.795815956101755e-06, + "loss": 1.1485, + "step": 13504 + }, + { + "epoch": 0.8205237256212407, + "grad_norm": 0.17476050555706024, + "learning_rate": 7.790683698942885e-06, + "loss": 0.9955, + "step": 13505 + }, + { + "epoch": 0.8205844826538672, + "grad_norm": 0.21291862428188324, + "learning_rate": 7.785552988974665e-06, + "loss": 1.0953, + "step": 13506 + }, + { + "epoch": 0.8206452396864937, + "grad_norm": 0.9978319406509399, + "learning_rate": 7.780423826385158e-06, + "loss": 1.0195, + "step": 13507 + }, + { + "epoch": 0.8207059967191203, + "grad_norm": 0.18824568390846252, + "learning_rate": 7.775296211362382e-06, + "loss": 1.0633, + "step": 13508 + }, + { + "epoch": 0.8207667537517468, + "grad_norm": 0.1406991183757782, + "learning_rate": 7.770170144094269e-06, + "loss": 1.0321, + "step": 13509 + }, + { + "epoch": 0.8208275107843733, + "grad_norm": 0.20987100899219513, + "learning_rate": 7.765045624768758e-06, + "loss": 1.09, + "step": 13510 + }, + { + "epoch": 0.8208882678169999, + "grad_norm": 0.21866030991077423, + "learning_rate": 7.759922653573648e-06, + "loss": 1.0174, + "step": 13511 + }, + { + "epoch": 0.8209490248496264, + "grad_norm": 0.13357022404670715, + "learning_rate": 7.754801230696745e-06, + "loss": 1.0476, + "step": 13512 + }, + { + "epoch": 0.8210097818822528, + "grad_norm": 0.2136799395084381, + "learning_rate": 7.74968135632575e-06, + "loss": 1.1027, + "step": 13513 + }, + { + "epoch": 0.8210705389148794, + "grad_norm": 0.29462161660194397, + "learning_rate": 7.744563030648367e-06, + "loss": 1.2095, + "step": 13514 + }, + { + "epoch": 0.8211312959475059, + "grad_norm": 0.20674394071102142, + "learning_rate": 7.739446253852195e-06, + "loss": 1.0423, + "step": 13515 + }, + { + "epoch": 0.8211920529801324, + "grad_norm": 0.29262977838516235, + "learning_rate": 7.734331026124786e-06, + "loss": 1.0538, + "step": 13516 + }, + { + "epoch": 0.821252810012759, + "grad_norm": 0.160648912191391, + "learning_rate": 7.729217347653655e-06, + "loss": 0.9869, + "step": 13517 + }, + { + "epoch": 0.8213135670453855, + "grad_norm": 0.12834885716438293, + "learning_rate": 7.72410521862621e-06, + "loss": 1.0299, + "step": 13518 + }, + { + "epoch": 0.821374324078012, + "grad_norm": 0.1340564638376236, + "learning_rate": 7.71899463922987e-06, + "loss": 1.0415, + "step": 13519 + }, + { + "epoch": 0.8214350811106386, + "grad_norm": 0.4204949140548706, + "learning_rate": 7.71388560965196e-06, + "loss": 0.968, + "step": 13520 + }, + { + "epoch": 0.8214958381432651, + "grad_norm": 0.1897193342447281, + "learning_rate": 7.708778130079742e-06, + "loss": 1.0952, + "step": 13521 + }, + { + "epoch": 0.8215565951758916, + "grad_norm": 0.13421377539634705, + "learning_rate": 7.70367220070044e-06, + "loss": 1.036, + "step": 13522 + }, + { + "epoch": 0.8216173522085182, + "grad_norm": 0.21721796691417694, + "learning_rate": 7.698567821701203e-06, + "loss": 1.108, + "step": 13523 + }, + { + "epoch": 0.8216781092411447, + "grad_norm": 0.1946820169687271, + "learning_rate": 7.693464993269145e-06, + "loss": 1.0858, + "step": 13524 + }, + { + "epoch": 0.8217388662737712, + "grad_norm": 0.14116361737251282, + "learning_rate": 7.6883637155913e-06, + "loss": 1.0207, + "step": 13525 + }, + { + "epoch": 0.8217996233063977, + "grad_norm": 0.16092908382415771, + "learning_rate": 7.683263988854667e-06, + "loss": 1.0551, + "step": 13526 + }, + { + "epoch": 0.8218603803390242, + "grad_norm": 0.14869102835655212, + "learning_rate": 7.678165813246168e-06, + "loss": 1.025, + "step": 13527 + }, + { + "epoch": 0.8219211373716507, + "grad_norm": 0.15381605923175812, + "learning_rate": 7.673069188952686e-06, + "loss": 1.0699, + "step": 13528 + }, + { + "epoch": 0.8219818944042773, + "grad_norm": 0.19870859384536743, + "learning_rate": 7.66797411616102e-06, + "loss": 1.0613, + "step": 13529 + }, + { + "epoch": 0.8220426514369038, + "grad_norm": 0.18173258006572723, + "learning_rate": 7.66288059505797e-06, + "loss": 1.0101, + "step": 13530 + }, + { + "epoch": 0.8221034084695303, + "grad_norm": 0.17821475863456726, + "learning_rate": 7.657788625830204e-06, + "loss": 1.0864, + "step": 13531 + }, + { + "epoch": 0.8221641655021569, + "grad_norm": 0.1462254375219345, + "learning_rate": 7.652698208664378e-06, + "loss": 1.039, + "step": 13532 + }, + { + "epoch": 0.8222249225347834, + "grad_norm": 0.19708211719989777, + "learning_rate": 7.647609343747086e-06, + "loss": 1.1654, + "step": 13533 + }, + { + "epoch": 0.8222856795674099, + "grad_norm": 0.1642867624759674, + "learning_rate": 7.642522031264849e-06, + "loss": 1.0333, + "step": 13534 + }, + { + "epoch": 0.8223464366000365, + "grad_norm": 0.18503437936306, + "learning_rate": 7.637436271404168e-06, + "loss": 1.0646, + "step": 13535 + }, + { + "epoch": 0.822407193632663, + "grad_norm": 0.15961810946464539, + "learning_rate": 7.632352064351445e-06, + "loss": 1.0406, + "step": 13536 + }, + { + "epoch": 0.8224679506652895, + "grad_norm": 25.939237594604492, + "learning_rate": 7.627269410293059e-06, + "loss": 1.043, + "step": 13537 + }, + { + "epoch": 0.8225287076979161, + "grad_norm": 0.2772379517555237, + "learning_rate": 7.6221883094152955e-06, + "loss": 1.2218, + "step": 13538 + }, + { + "epoch": 0.8225894647305425, + "grad_norm": 0.41167664527893066, + "learning_rate": 7.617108761904396e-06, + "loss": 1.2303, + "step": 13539 + }, + { + "epoch": 0.822650221763169, + "grad_norm": 0.13770104944705963, + "learning_rate": 7.612030767946577e-06, + "loss": 1.0702, + "step": 13540 + }, + { + "epoch": 0.8227109787957956, + "grad_norm": 0.32722580432891846, + "learning_rate": 7.606954327727967e-06, + "loss": 1.0745, + "step": 13541 + }, + { + "epoch": 0.8227717358284221, + "grad_norm": 0.13500149548053741, + "learning_rate": 7.601879441434639e-06, + "loss": 1.0024, + "step": 13542 + }, + { + "epoch": 0.8228324928610486, + "grad_norm": 0.2498636692762375, + "learning_rate": 7.596806109252613e-06, + "loss": 1.0766, + "step": 13543 + }, + { + "epoch": 0.8228932498936752, + "grad_norm": 0.12592114508152008, + "learning_rate": 7.591734331367861e-06, + "loss": 1.2139, + "step": 13544 + }, + { + "epoch": 0.8229540069263017, + "grad_norm": 5.45838737487793, + "learning_rate": 7.586664107966279e-06, + "loss": 1.1088, + "step": 13545 + }, + { + "epoch": 0.8230147639589283, + "grad_norm": 0.2097436785697937, + "learning_rate": 7.58159543923373e-06, + "loss": 1.1148, + "step": 13546 + }, + { + "epoch": 0.8230755209915548, + "grad_norm": 0.18764664232730865, + "learning_rate": 7.576528325355991e-06, + "loss": 1.0027, + "step": 13547 + }, + { + "epoch": 0.8231362780241813, + "grad_norm": 0.26681292057037354, + "learning_rate": 7.571462766518811e-06, + "loss": 1.0629, + "step": 13548 + }, + { + "epoch": 0.8231970350568079, + "grad_norm": 0.1848364770412445, + "learning_rate": 7.566398762907861e-06, + "loss": 1.1097, + "step": 13549 + }, + { + "epoch": 0.8232577920894344, + "grad_norm": 0.16315023601055145, + "learning_rate": 7.561336314708773e-06, + "loss": 1.0309, + "step": 13550 + }, + { + "epoch": 0.8233185491220609, + "grad_norm": 0.12438501417636871, + "learning_rate": 7.5562754221071005e-06, + "loss": 1.0313, + "step": 13551 + }, + { + "epoch": 0.8233793061546874, + "grad_norm": 0.29777318239212036, + "learning_rate": 7.551216085288359e-06, + "loss": 1.1438, + "step": 13552 + }, + { + "epoch": 0.8234400631873139, + "grad_norm": 1.0374915599822998, + "learning_rate": 7.546158304437995e-06, + "loss": 1.1855, + "step": 13553 + }, + { + "epoch": 0.8235008202199404, + "grad_norm": 0.15251724421977997, + "learning_rate": 7.541102079741402e-06, + "loss": 0.9665, + "step": 13554 + }, + { + "epoch": 0.823561577252567, + "grad_norm": 0.11463949084281921, + "learning_rate": 7.536047411383912e-06, + "loss": 0.9762, + "step": 13555 + }, + { + "epoch": 0.8236223342851935, + "grad_norm": 0.13686974346637726, + "learning_rate": 7.530994299550826e-06, + "loss": 1.0809, + "step": 13556 + }, + { + "epoch": 0.82368309131782, + "grad_norm": 0.7147294878959656, + "learning_rate": 7.5259427444273614e-06, + "loss": 1.0601, + "step": 13557 + }, + { + "epoch": 0.8237438483504466, + "grad_norm": 0.1433509737253189, + "learning_rate": 7.520892746198666e-06, + "loss": 0.9901, + "step": 13558 + }, + { + "epoch": 0.8238046053830731, + "grad_norm": 0.16496989130973816, + "learning_rate": 7.5158443050498515e-06, + "loss": 1.0513, + "step": 13559 + }, + { + "epoch": 0.8238653624156996, + "grad_norm": 0.16630680859088898, + "learning_rate": 7.510797421165972e-06, + "loss": 1.2163, + "step": 13560 + }, + { + "epoch": 0.8239261194483262, + "grad_norm": 0.13437055051326752, + "learning_rate": 7.505752094732032e-06, + "loss": 1.0487, + "step": 13561 + }, + { + "epoch": 0.8239868764809527, + "grad_norm": 0.2328210473060608, + "learning_rate": 7.500708325932965e-06, + "loss": 1.1409, + "step": 13562 + }, + { + "epoch": 0.8240476335135792, + "grad_norm": 0.25564566254615784, + "learning_rate": 7.495666114953653e-06, + "loss": 1.1301, + "step": 13563 + }, + { + "epoch": 0.8241083905462058, + "grad_norm": 0.16897766292095184, + "learning_rate": 7.490625461978923e-06, + "loss": 1.0047, + "step": 13564 + }, + { + "epoch": 0.8241691475788322, + "grad_norm": 0.162090003490448, + "learning_rate": 7.485586367193509e-06, + "loss": 1.0891, + "step": 13565 + }, + { + "epoch": 0.8242299046114587, + "grad_norm": 0.2319522500038147, + "learning_rate": 7.480548830782164e-06, + "loss": 1.1554, + "step": 13566 + }, + { + "epoch": 0.8242906616440853, + "grad_norm": 0.195261612534523, + "learning_rate": 7.475512852929511e-06, + "loss": 1.0087, + "step": 13567 + }, + { + "epoch": 0.8243514186767118, + "grad_norm": 0.13499966263771057, + "learning_rate": 7.470478433820161e-06, + "loss": 1.0306, + "step": 13568 + }, + { + "epoch": 0.8244121757093383, + "grad_norm": 0.277130126953125, + "learning_rate": 7.4654455736386406e-06, + "loss": 1.1239, + "step": 13569 + }, + { + "epoch": 0.8244729327419649, + "grad_norm": 0.1491478979587555, + "learning_rate": 7.460414272569433e-06, + "loss": 1.0278, + "step": 13570 + }, + { + "epoch": 0.8245336897745914, + "grad_norm": 0.16289982199668884, + "learning_rate": 7.455384530796966e-06, + "loss": 1.0036, + "step": 13571 + }, + { + "epoch": 0.8245944468072179, + "grad_norm": 0.26611143350601196, + "learning_rate": 7.450356348505605e-06, + "loss": 1.0428, + "step": 13572 + }, + { + "epoch": 0.8246552038398445, + "grad_norm": 0.24199308454990387, + "learning_rate": 7.445329725879652e-06, + "loss": 1.1788, + "step": 13573 + }, + { + "epoch": 0.824715960872471, + "grad_norm": 0.1384393274784088, + "learning_rate": 7.4403046631033645e-06, + "loss": 1.0093, + "step": 13574 + }, + { + "epoch": 0.8247767179050975, + "grad_norm": 0.13248220086097717, + "learning_rate": 7.435281160360941e-06, + "loss": 1.0103, + "step": 13575 + }, + { + "epoch": 0.8248374749377241, + "grad_norm": 0.1331065595149994, + "learning_rate": 7.4302592178365e-06, + "loss": 1.0322, + "step": 13576 + }, + { + "epoch": 0.8248982319703506, + "grad_norm": 0.23227442800998688, + "learning_rate": 7.425238835714155e-06, + "loss": 1.0336, + "step": 13577 + }, + { + "epoch": 0.8249589890029771, + "grad_norm": 0.14967182278633118, + "learning_rate": 7.420220014177903e-06, + "loss": 1.0268, + "step": 13578 + }, + { + "epoch": 0.8250197460356036, + "grad_norm": 0.22763758897781372, + "learning_rate": 7.4152027534117155e-06, + "loss": 1.1197, + "step": 13579 + }, + { + "epoch": 0.8250805030682301, + "grad_norm": 0.4875481128692627, + "learning_rate": 7.410187053599499e-06, + "loss": 1.1967, + "step": 13580 + }, + { + "epoch": 0.8251412601008566, + "grad_norm": 0.185671865940094, + "learning_rate": 7.405172914925101e-06, + "loss": 1.0602, + "step": 13581 + }, + { + "epoch": 0.8252020171334832, + "grad_norm": 0.18877701461315155, + "learning_rate": 7.400160337572337e-06, + "loss": 0.9954, + "step": 13582 + }, + { + "epoch": 0.8252627741661097, + "grad_norm": 0.40660834312438965, + "learning_rate": 7.395149321724926e-06, + "loss": 1.1492, + "step": 13583 + }, + { + "epoch": 0.8253235311987362, + "grad_norm": 0.15399469435214996, + "learning_rate": 7.390139867566565e-06, + "loss": 1.0199, + "step": 13584 + }, + { + "epoch": 0.8253842882313628, + "grad_norm": 0.13096365332603455, + "learning_rate": 7.385131975280851e-06, + "loss": 0.997, + "step": 13585 + }, + { + "epoch": 0.8254450452639893, + "grad_norm": 1.1106717586517334, + "learning_rate": 7.38012564505135e-06, + "loss": 1.0361, + "step": 13586 + }, + { + "epoch": 0.8255058022966159, + "grad_norm": 0.16291289031505585, + "learning_rate": 7.375120877061592e-06, + "loss": 1.0508, + "step": 13587 + }, + { + "epoch": 0.8255665593292424, + "grad_norm": 0.2050158679485321, + "learning_rate": 7.370117671495019e-06, + "loss": 1.1252, + "step": 13588 + }, + { + "epoch": 0.8256273163618689, + "grad_norm": 0.13175736367702484, + "learning_rate": 7.365116028535024e-06, + "loss": 1.061, + "step": 13589 + }, + { + "epoch": 0.8256880733944955, + "grad_norm": 0.15284636616706848, + "learning_rate": 7.360115948364943e-06, + "loss": 1.0291, + "step": 13590 + }, + { + "epoch": 0.825748830427122, + "grad_norm": 0.13270540535449982, + "learning_rate": 7.355117431168052e-06, + "loss": 1.0443, + "step": 13591 + }, + { + "epoch": 0.8258095874597484, + "grad_norm": 0.1326657235622406, + "learning_rate": 7.350120477127576e-06, + "loss": 1.0206, + "step": 13592 + }, + { + "epoch": 0.825870344492375, + "grad_norm": 0.13718290627002716, + "learning_rate": 7.345125086426674e-06, + "loss": 1.0343, + "step": 13593 + }, + { + "epoch": 0.8259311015250015, + "grad_norm": 0.21384204924106598, + "learning_rate": 7.340131259248461e-06, + "loss": 1.0366, + "step": 13594 + }, + { + "epoch": 0.825991858557628, + "grad_norm": 0.6526817083358765, + "learning_rate": 7.335138995775981e-06, + "loss": 1.091, + "step": 13595 + }, + { + "epoch": 0.8260526155902546, + "grad_norm": 0.14492887258529663, + "learning_rate": 7.330148296192229e-06, + "loss": 0.9942, + "step": 13596 + }, + { + "epoch": 0.8261133726228811, + "grad_norm": 0.13426008820533752, + "learning_rate": 7.325159160680139e-06, + "loss": 1.0464, + "step": 13597 + }, + { + "epoch": 0.8261741296555076, + "grad_norm": 0.1337309181690216, + "learning_rate": 7.320171589422592e-06, + "loss": 1.0493, + "step": 13598 + }, + { + "epoch": 0.8262348866881342, + "grad_norm": 0.19757132232189178, + "learning_rate": 7.315185582602402e-06, + "loss": 1.152, + "step": 13599 + }, + { + "epoch": 0.8262956437207607, + "grad_norm": 0.20940552651882172, + "learning_rate": 7.310201140402334e-06, + "loss": 1.1391, + "step": 13600 + }, + { + "epoch": 0.8263564007533872, + "grad_norm": 0.2194860726594925, + "learning_rate": 7.305218263005098e-06, + "loss": 1.076, + "step": 13601 + }, + { + "epoch": 0.8264171577860138, + "grad_norm": 0.12998005747795105, + "learning_rate": 7.300236950593326e-06, + "loss": 1.0104, + "step": 13602 + }, + { + "epoch": 0.8264779148186403, + "grad_norm": 0.41599705815315247, + "learning_rate": 7.295257203349632e-06, + "loss": 1.084, + "step": 13603 + }, + { + "epoch": 0.8265386718512668, + "grad_norm": 0.18248318135738373, + "learning_rate": 7.290279021456553e-06, + "loss": 1.0422, + "step": 13604 + }, + { + "epoch": 0.8265994288838933, + "grad_norm": 0.1891697198152542, + "learning_rate": 7.285302405096539e-06, + "loss": 1.1301, + "step": 13605 + }, + { + "epoch": 0.8266601859165198, + "grad_norm": 0.24270778894424438, + "learning_rate": 7.280327354452027e-06, + "loss": 1.0417, + "step": 13606 + }, + { + "epoch": 0.8267209429491463, + "grad_norm": 0.5267266631126404, + "learning_rate": 7.275353869705353e-06, + "loss": 1.1932, + "step": 13607 + }, + { + "epoch": 0.8267816999817729, + "grad_norm": 3.1256840229034424, + "learning_rate": 7.270381951038863e-06, + "loss": 1.0305, + "step": 13608 + }, + { + "epoch": 0.8268424570143994, + "grad_norm": 0.23995907604694366, + "learning_rate": 7.265411598634775e-06, + "loss": 1.1144, + "step": 13609 + }, + { + "epoch": 0.8269032140470259, + "grad_norm": 0.26615986227989197, + "learning_rate": 7.260442812675288e-06, + "loss": 1.048, + "step": 13610 + }, + { + "epoch": 0.8269639710796525, + "grad_norm": 4.326955795288086, + "learning_rate": 7.255475593342531e-06, + "loss": 1.2705, + "step": 13611 + }, + { + "epoch": 0.827024728112279, + "grad_norm": 0.23952879011631012, + "learning_rate": 7.250509940818578e-06, + "loss": 1.0595, + "step": 13612 + }, + { + "epoch": 0.8270854851449055, + "grad_norm": 0.2289150357246399, + "learning_rate": 7.245545855285446e-06, + "loss": 1.0818, + "step": 13613 + }, + { + "epoch": 0.8271462421775321, + "grad_norm": 0.2617868483066559, + "learning_rate": 7.240583336925094e-06, + "loss": 1.0226, + "step": 13614 + }, + { + "epoch": 0.8272069992101586, + "grad_norm": 0.1575682908296585, + "learning_rate": 7.235622385919427e-06, + "loss": 1.0071, + "step": 13615 + }, + { + "epoch": 0.8272677562427851, + "grad_norm": 1.208633303642273, + "learning_rate": 7.230663002450289e-06, + "loss": 1.0768, + "step": 13616 + }, + { + "epoch": 0.8273285132754117, + "grad_norm": 0.19189471006393433, + "learning_rate": 7.22570518669946e-06, + "loss": 1.0235, + "step": 13617 + }, + { + "epoch": 0.8273892703080381, + "grad_norm": 0.15289440751075745, + "learning_rate": 7.220748938848682e-06, + "loss": 1.0619, + "step": 13618 + }, + { + "epoch": 0.8274500273406646, + "grad_norm": 0.3222452998161316, + "learning_rate": 7.215794259079617e-06, + "loss": 1.1296, + "step": 13619 + }, + { + "epoch": 0.8275107843732912, + "grad_norm": 0.21725346148014069, + "learning_rate": 7.2108411475738825e-06, + "loss": 1.0771, + "step": 13620 + }, + { + "epoch": 0.8275715414059177, + "grad_norm": 0.2506987154483795, + "learning_rate": 7.2058896045130384e-06, + "loss": 1.0167, + "step": 13621 + }, + { + "epoch": 0.8276322984385442, + "grad_norm": 0.15580271184444427, + "learning_rate": 7.200939630078573e-06, + "loss": 1.0068, + "step": 13622 + }, + { + "epoch": 0.8276930554711708, + "grad_norm": 0.18063943088054657, + "learning_rate": 7.195991224451948e-06, + "loss": 1.067, + "step": 13623 + }, + { + "epoch": 0.8277538125037973, + "grad_norm": 0.2540534436702728, + "learning_rate": 7.19104438781455e-06, + "loss": 1.0121, + "step": 13624 + }, + { + "epoch": 0.8278145695364238, + "grad_norm": 0.13147860765457153, + "learning_rate": 7.186099120347683e-06, + "loss": 1.0344, + "step": 13625 + }, + { + "epoch": 0.8278753265690504, + "grad_norm": 2.40822172164917, + "learning_rate": 7.181155422232633e-06, + "loss": 1.206, + "step": 13626 + }, + { + "epoch": 0.8279360836016769, + "grad_norm": 0.7841127514839172, + "learning_rate": 7.1762132936505914e-06, + "loss": 1.0385, + "step": 13627 + }, + { + "epoch": 0.8279968406343035, + "grad_norm": 0.26612499356269836, + "learning_rate": 7.171272734782741e-06, + "loss": 1.0558, + "step": 13628 + }, + { + "epoch": 0.82805759766693, + "grad_norm": 0.1856686770915985, + "learning_rate": 7.166333745810166e-06, + "loss": 1.0111, + "step": 13629 + }, + { + "epoch": 0.8281183546995565, + "grad_norm": 0.3251533806324005, + "learning_rate": 7.161396326913911e-06, + "loss": 1.1044, + "step": 13630 + }, + { + "epoch": 0.828179111732183, + "grad_norm": 0.3313359022140503, + "learning_rate": 7.156460478274962e-06, + "loss": 1.0695, + "step": 13631 + }, + { + "epoch": 0.8282398687648095, + "grad_norm": 0.28419503569602966, + "learning_rate": 7.1515262000742156e-06, + "loss": 1.1433, + "step": 13632 + }, + { + "epoch": 0.828300625797436, + "grad_norm": 0.3156418800354004, + "learning_rate": 7.146593492492565e-06, + "loss": 0.9382, + "step": 13633 + }, + { + "epoch": 0.8283613828300626, + "grad_norm": 0.2312171757221222, + "learning_rate": 7.141662355710821e-06, + "loss": 1.0374, + "step": 13634 + }, + { + "epoch": 0.8284221398626891, + "grad_norm": 0.32699596881866455, + "learning_rate": 7.136732789909723e-06, + "loss": 1.1878, + "step": 13635 + }, + { + "epoch": 0.8284828968953156, + "grad_norm": 0.16955196857452393, + "learning_rate": 7.1318047952699695e-06, + "loss": 1.0789, + "step": 13636 + }, + { + "epoch": 0.8285436539279422, + "grad_norm": 0.19326665997505188, + "learning_rate": 7.1268783719721955e-06, + "loss": 1.1487, + "step": 13637 + }, + { + "epoch": 0.8286044109605687, + "grad_norm": 0.7505744099617004, + "learning_rate": 7.121953520196989e-06, + "loss": 1.309, + "step": 13638 + }, + { + "epoch": 0.8286651679931952, + "grad_norm": 0.1524544656276703, + "learning_rate": 7.117030240124855e-06, + "loss": 1.2107, + "step": 13639 + }, + { + "epoch": 0.8287259250258218, + "grad_norm": 0.13313785195350647, + "learning_rate": 7.1121085319362736e-06, + "loss": 1.0041, + "step": 13640 + }, + { + "epoch": 0.8287866820584483, + "grad_norm": 0.5564085245132446, + "learning_rate": 7.107188395811643e-06, + "loss": 1.0974, + "step": 13641 + }, + { + "epoch": 0.8288474390910748, + "grad_norm": 0.2295825481414795, + "learning_rate": 7.102269831931313e-06, + "loss": 1.1039, + "step": 13642 + }, + { + "epoch": 0.8289081961237014, + "grad_norm": 0.8705514669418335, + "learning_rate": 7.097352840475563e-06, + "loss": 1.0188, + "step": 13643 + }, + { + "epoch": 0.8289689531563278, + "grad_norm": 0.5862904191017151, + "learning_rate": 7.092437421624659e-06, + "loss": 1.1064, + "step": 13644 + }, + { + "epoch": 0.8290297101889543, + "grad_norm": 0.1298293024301529, + "learning_rate": 7.087523575558741e-06, + "loss": 1.0666, + "step": 13645 + }, + { + "epoch": 0.8290904672215809, + "grad_norm": 0.1199067160487175, + "learning_rate": 7.082611302457948e-06, + "loss": 0.9803, + "step": 13646 + }, + { + "epoch": 0.8291512242542074, + "grad_norm": 0.1895984709262848, + "learning_rate": 7.07770060250233e-06, + "loss": 1.0977, + "step": 13647 + }, + { + "epoch": 0.8292119812868339, + "grad_norm": 0.25693055987358093, + "learning_rate": 7.0727914758718815e-06, + "loss": 1.1339, + "step": 13648 + }, + { + "epoch": 0.8292727383194605, + "grad_norm": 0.2380073517560959, + "learning_rate": 7.067883922746565e-06, + "loss": 1.1664, + "step": 13649 + }, + { + "epoch": 0.829333495352087, + "grad_norm": 0.2887047529220581, + "learning_rate": 7.06297794330627e-06, + "loss": 1.2974, + "step": 13650 + }, + { + "epoch": 0.8293942523847135, + "grad_norm": 0.1883433610200882, + "learning_rate": 7.0580735377308245e-06, + "loss": 1.0473, + "step": 13651 + }, + { + "epoch": 0.8294550094173401, + "grad_norm": 0.12130742520093918, + "learning_rate": 7.053170706199985e-06, + "loss": 1.0256, + "step": 13652 + }, + { + "epoch": 0.8295157664499666, + "grad_norm": 0.17242196202278137, + "learning_rate": 7.048269448893463e-06, + "loss": 1.0905, + "step": 13653 + }, + { + "epoch": 0.8295765234825931, + "grad_norm": 0.4169430434703827, + "learning_rate": 7.043369765990943e-06, + "loss": 1.1057, + "step": 13654 + }, + { + "epoch": 0.8296372805152197, + "grad_norm": 0.11508682370185852, + "learning_rate": 7.0384716576720025e-06, + "loss": 1.0484, + "step": 13655 + }, + { + "epoch": 0.8296980375478462, + "grad_norm": 0.22506853938102722, + "learning_rate": 7.033575124116187e-06, + "loss": 1.0783, + "step": 13656 + }, + { + "epoch": 0.8297587945804726, + "grad_norm": 0.36801278591156006, + "learning_rate": 7.028680165502982e-06, + "loss": 1.0785, + "step": 13657 + }, + { + "epoch": 0.8298195516130992, + "grad_norm": 0.3148743808269501, + "learning_rate": 7.0237867820118094e-06, + "loss": 1.2083, + "step": 13658 + }, + { + "epoch": 0.8298803086457257, + "grad_norm": 0.13508982956409454, + "learning_rate": 7.018894973822043e-06, + "loss": 1.0512, + "step": 13659 + }, + { + "epoch": 0.8299410656783522, + "grad_norm": 0.2319728285074234, + "learning_rate": 7.014004741112984e-06, + "loss": 1.0743, + "step": 13660 + }, + { + "epoch": 0.8300018227109788, + "grad_norm": 0.25792884826660156, + "learning_rate": 7.009116084063893e-06, + "loss": 1.0833, + "step": 13661 + }, + { + "epoch": 0.8300625797436053, + "grad_norm": 0.13387320935726166, + "learning_rate": 7.004229002853962e-06, + "loss": 1.0661, + "step": 13662 + }, + { + "epoch": 0.8301233367762318, + "grad_norm": 0.14482754468917847, + "learning_rate": 6.99934349766233e-06, + "loss": 1.0455, + "step": 13663 + }, + { + "epoch": 0.8301840938088584, + "grad_norm": 0.26328244805336, + "learning_rate": 6.994459568668071e-06, + "loss": 1.0599, + "step": 13664 + }, + { + "epoch": 0.8302448508414849, + "grad_norm": 0.18739193677902222, + "learning_rate": 6.989577216050214e-06, + "loss": 1.0692, + "step": 13665 + }, + { + "epoch": 0.8303056078741115, + "grad_norm": 0.21241840720176697, + "learning_rate": 6.984696439987715e-06, + "loss": 1.1271, + "step": 13666 + }, + { + "epoch": 0.830366364906738, + "grad_norm": 0.14011313021183014, + "learning_rate": 6.979817240659481e-06, + "loss": 1.0458, + "step": 13667 + }, + { + "epoch": 0.8304271219393645, + "grad_norm": 0.16087493300437927, + "learning_rate": 6.974939618244364e-06, + "loss": 1.0534, + "step": 13668 + }, + { + "epoch": 0.8304878789719911, + "grad_norm": 0.1285860389471054, + "learning_rate": 6.970063572921137e-06, + "loss": 1.0649, + "step": 13669 + }, + { + "epoch": 0.8305486360046175, + "grad_norm": 0.14683525264263153, + "learning_rate": 6.965189104868564e-06, + "loss": 1.0023, + "step": 13670 + }, + { + "epoch": 0.830609393037244, + "grad_norm": 0.1716129332780838, + "learning_rate": 6.960316214265311e-06, + "loss": 1.0433, + "step": 13671 + }, + { + "epoch": 0.8306701500698705, + "grad_norm": 0.8468506336212158, + "learning_rate": 6.9554449012899755e-06, + "loss": 1.1581, + "step": 13672 + }, + { + "epoch": 0.8307309071024971, + "grad_norm": 0.21197691559791565, + "learning_rate": 6.950575166121126e-06, + "loss": 1.1747, + "step": 13673 + }, + { + "epoch": 0.8307916641351236, + "grad_norm": 0.12325751036405563, + "learning_rate": 6.9457070089372575e-06, + "loss": 0.9929, + "step": 13674 + }, + { + "epoch": 0.8308524211677502, + "grad_norm": 0.12780198454856873, + "learning_rate": 6.940840429916828e-06, + "loss": 0.9879, + "step": 13675 + }, + { + "epoch": 0.8309131782003767, + "grad_norm": 0.3526577353477478, + "learning_rate": 6.935975429238218e-06, + "loss": 1.1289, + "step": 13676 + }, + { + "epoch": 0.8309739352330032, + "grad_norm": 0.15938465297222137, + "learning_rate": 6.931112007079749e-06, + "loss": 1.0694, + "step": 13677 + }, + { + "epoch": 0.8310346922656298, + "grad_norm": 0.6281047463417053, + "learning_rate": 6.926250163619713e-06, + "loss": 1.0654, + "step": 13678 + }, + { + "epoch": 0.8310954492982563, + "grad_norm": 0.14500994980335236, + "learning_rate": 6.9213898990362736e-06, + "loss": 1.0335, + "step": 13679 + }, + { + "epoch": 0.8311562063308828, + "grad_norm": 0.19841228425502777, + "learning_rate": 6.916531213507626e-06, + "loss": 1.1144, + "step": 13680 + }, + { + "epoch": 0.8312169633635094, + "grad_norm": 0.1488967388868332, + "learning_rate": 6.911674107211858e-06, + "loss": 1.0424, + "step": 13681 + }, + { + "epoch": 0.8312777203961359, + "grad_norm": 0.1816893070936203, + "learning_rate": 6.906818580327007e-06, + "loss": 1.0304, + "step": 13682 + }, + { + "epoch": 0.8313384774287624, + "grad_norm": 0.19307270646095276, + "learning_rate": 6.901964633031044e-06, + "loss": 1.0077, + "step": 13683 + }, + { + "epoch": 0.8313992344613889, + "grad_norm": 0.31373128294944763, + "learning_rate": 6.897112265501904e-06, + "loss": 1.0542, + "step": 13684 + }, + { + "epoch": 0.8314599914940154, + "grad_norm": 0.6235456466674805, + "learning_rate": 6.892261477917444e-06, + "loss": 1.2125, + "step": 13685 + }, + { + "epoch": 0.8315207485266419, + "grad_norm": 0.2876046299934387, + "learning_rate": 6.8874122704554676e-06, + "loss": 1.1041, + "step": 13686 + }, + { + "epoch": 0.8315815055592685, + "grad_norm": 0.11450030654668808, + "learning_rate": 6.882564643293732e-06, + "loss": 1.0987, + "step": 13687 + }, + { + "epoch": 0.831642262591895, + "grad_norm": 0.20706981420516968, + "learning_rate": 6.8777185966099225e-06, + "loss": 1.0484, + "step": 13688 + }, + { + "epoch": 0.8317030196245215, + "grad_norm": 0.1720835566520691, + "learning_rate": 6.872874130581674e-06, + "loss": 1.0961, + "step": 13689 + }, + { + "epoch": 0.8317637766571481, + "grad_norm": 0.2587965726852417, + "learning_rate": 6.868031245386547e-06, + "loss": 1.0861, + "step": 13690 + }, + { + "epoch": 0.8318245336897746, + "grad_norm": 0.5010504126548767, + "learning_rate": 6.8631899412020974e-06, + "loss": 1.0387, + "step": 13691 + }, + { + "epoch": 0.8318852907224011, + "grad_norm": 0.13519664108753204, + "learning_rate": 6.858350218205745e-06, + "loss": 1.054, + "step": 13692 + }, + { + "epoch": 0.8319460477550277, + "grad_norm": 0.23431579768657684, + "learning_rate": 6.853512076574908e-06, + "loss": 1.0873, + "step": 13693 + }, + { + "epoch": 0.8320068047876542, + "grad_norm": 0.24693140387535095, + "learning_rate": 6.848675516486924e-06, + "loss": 1.1179, + "step": 13694 + }, + { + "epoch": 0.8320675618202807, + "grad_norm": 0.2099170833826065, + "learning_rate": 6.843840538119067e-06, + "loss": 1.064, + "step": 13695 + }, + { + "epoch": 0.8321283188529073, + "grad_norm": 0.1371120810508728, + "learning_rate": 6.8390071416485905e-06, + "loss": 1.0414, + "step": 13696 + }, + { + "epoch": 0.8321890758855337, + "grad_norm": 2.413619041442871, + "learning_rate": 6.834175327252651e-06, + "loss": 1.071, + "step": 13697 + }, + { + "epoch": 0.8322498329181602, + "grad_norm": 0.17920811474323273, + "learning_rate": 6.829345095108364e-06, + "loss": 1.0515, + "step": 13698 + }, + { + "epoch": 0.8323105899507868, + "grad_norm": 0.3232126235961914, + "learning_rate": 6.82451644539277e-06, + "loss": 1.1795, + "step": 13699 + }, + { + "epoch": 0.8323713469834133, + "grad_norm": 0.3443066477775574, + "learning_rate": 6.819689378282862e-06, + "loss": 1.0738, + "step": 13700 + }, + { + "epoch": 0.8324321040160398, + "grad_norm": 0.1932527869939804, + "learning_rate": 6.814863893955598e-06, + "loss": 1.0778, + "step": 13701 + }, + { + "epoch": 0.8324928610486664, + "grad_norm": 1.3005937337875366, + "learning_rate": 6.810039992587841e-06, + "loss": 1.0394, + "step": 13702 + }, + { + "epoch": 0.8325536180812929, + "grad_norm": 0.19256550073623657, + "learning_rate": 6.805217674356418e-06, + "loss": 1.1068, + "step": 13703 + }, + { + "epoch": 0.8326143751139194, + "grad_norm": 0.15049053728580475, + "learning_rate": 6.8003969394380915e-06, + "loss": 0.9825, + "step": 13704 + }, + { + "epoch": 0.832675132146546, + "grad_norm": 0.13678212463855743, + "learning_rate": 6.7955777880095665e-06, + "loss": 1.0287, + "step": 13705 + }, + { + "epoch": 0.8327358891791725, + "grad_norm": 4.726916790008545, + "learning_rate": 6.790760220247488e-06, + "loss": 1.0599, + "step": 13706 + }, + { + "epoch": 0.832796646211799, + "grad_norm": 0.2263072282075882, + "learning_rate": 6.785944236328445e-06, + "loss": 1.1021, + "step": 13707 + }, + { + "epoch": 0.8328574032444256, + "grad_norm": 0.1345345377922058, + "learning_rate": 6.7811298364289675e-06, + "loss": 1.0164, + "step": 13708 + }, + { + "epoch": 0.8329181602770521, + "grad_norm": 0.19376248121261597, + "learning_rate": 6.776317020725531e-06, + "loss": 1.011, + "step": 13709 + }, + { + "epoch": 0.8329789173096785, + "grad_norm": 0.14592976868152618, + "learning_rate": 6.77150578939455e-06, + "loss": 1.02, + "step": 13710 + }, + { + "epoch": 0.8330396743423051, + "grad_norm": 0.13018591701984406, + "learning_rate": 6.766696142612383e-06, + "loss": 1.045, + "step": 13711 + }, + { + "epoch": 0.8331004313749316, + "grad_norm": 0.1391652375459671, + "learning_rate": 6.761888080555323e-06, + "loss": 1.0168, + "step": 13712 + }, + { + "epoch": 0.8331611884075582, + "grad_norm": 0.16378462314605713, + "learning_rate": 6.757081603399612e-06, + "loss": 1.0732, + "step": 13713 + }, + { + "epoch": 0.8332219454401847, + "grad_norm": 0.2526335120201111, + "learning_rate": 6.752276711321432e-06, + "loss": 1.2086, + "step": 13714 + }, + { + "epoch": 0.8332827024728112, + "grad_norm": 0.3432379364967346, + "learning_rate": 6.7474734044969016e-06, + "loss": 1.1103, + "step": 13715 + }, + { + "epoch": 0.8333434595054378, + "grad_norm": 0.20861634612083435, + "learning_rate": 6.742671683102103e-06, + "loss": 1.1704, + "step": 13716 + }, + { + "epoch": 0.8334042165380643, + "grad_norm": 0.1992982178926468, + "learning_rate": 6.737871547313035e-06, + "loss": 1.0913, + "step": 13717 + }, + { + "epoch": 0.8334649735706908, + "grad_norm": 0.1345367282629013, + "learning_rate": 6.733072997305661e-06, + "loss": 0.9768, + "step": 13718 + }, + { + "epoch": 0.8335257306033174, + "grad_norm": 0.10961595177650452, + "learning_rate": 6.728276033255848e-06, + "loss": 1.016, + "step": 13719 + }, + { + "epoch": 0.8335864876359439, + "grad_norm": 2.4814505577087402, + "learning_rate": 6.723480655339431e-06, + "loss": 1.0083, + "step": 13720 + }, + { + "epoch": 0.8336472446685704, + "grad_norm": 0.1933918446302414, + "learning_rate": 6.718686863732204e-06, + "loss": 1.1174, + "step": 13721 + }, + { + "epoch": 0.833708001701197, + "grad_norm": 0.18959343433380127, + "learning_rate": 6.713894658609876e-06, + "loss": 1.0764, + "step": 13722 + }, + { + "epoch": 0.8337687587338234, + "grad_norm": 0.21588563919067383, + "learning_rate": 6.709104040148106e-06, + "loss": 0.9542, + "step": 13723 + }, + { + "epoch": 0.8338295157664499, + "grad_norm": 0.311763197183609, + "learning_rate": 6.7043150085225e-06, + "loss": 1.1803, + "step": 13724 + }, + { + "epoch": 0.8338902727990765, + "grad_norm": 0.2754959166049957, + "learning_rate": 6.699527563908592e-06, + "loss": 1.2635, + "step": 13725 + }, + { + "epoch": 0.833951029831703, + "grad_norm": 0.1393309086561203, + "learning_rate": 6.694741706481872e-06, + "loss": 0.9915, + "step": 13726 + }, + { + "epoch": 0.8340117868643295, + "grad_norm": 0.11859650909900665, + "learning_rate": 6.689957436417766e-06, + "loss": 1.0168, + "step": 13727 + }, + { + "epoch": 0.8340725438969561, + "grad_norm": 0.1829681545495987, + "learning_rate": 6.685174753891643e-06, + "loss": 0.9885, + "step": 13728 + }, + { + "epoch": 0.8341333009295826, + "grad_norm": 0.23484736680984497, + "learning_rate": 6.68039365907881e-06, + "loss": 1.0932, + "step": 13729 + }, + { + "epoch": 0.8341940579622091, + "grad_norm": 0.2841300964355469, + "learning_rate": 6.675614152154519e-06, + "loss": 1.0995, + "step": 13730 + }, + { + "epoch": 0.8342548149948357, + "grad_norm": 0.13907818496227264, + "learning_rate": 6.67083623329397e-06, + "loss": 1.0795, + "step": 13731 + }, + { + "epoch": 0.8343155720274622, + "grad_norm": 0.1260407418012619, + "learning_rate": 6.666059902672295e-06, + "loss": 0.999, + "step": 13732 + }, + { + "epoch": 0.8343763290600887, + "grad_norm": 0.24311505258083344, + "learning_rate": 6.661285160464564e-06, + "loss": 1.0782, + "step": 13733 + }, + { + "epoch": 0.8344370860927153, + "grad_norm": 0.12827198207378387, + "learning_rate": 6.656512006845805e-06, + "loss": 1.0209, + "step": 13734 + }, + { + "epoch": 0.8344978431253418, + "grad_norm": 6.3856282234191895, + "learning_rate": 6.6517404419909835e-06, + "loss": 1.1047, + "step": 13735 + }, + { + "epoch": 0.8345586001579682, + "grad_norm": 0.13263681530952454, + "learning_rate": 6.646970466074975e-06, + "loss": 0.9896, + "step": 13736 + }, + { + "epoch": 0.8346193571905948, + "grad_norm": 0.25202369689941406, + "learning_rate": 6.642202079272658e-06, + "loss": 1.0475, + "step": 13737 + }, + { + "epoch": 0.8346801142232213, + "grad_norm": 3.9343008995056152, + "learning_rate": 6.637435281758819e-06, + "loss": 1.1194, + "step": 13738 + }, + { + "epoch": 0.8347408712558478, + "grad_norm": 0.3148392140865326, + "learning_rate": 6.632670073708158e-06, + "loss": 1.0564, + "step": 13739 + }, + { + "epoch": 0.8348016282884744, + "grad_norm": 0.19868198037147522, + "learning_rate": 6.627906455295358e-06, + "loss": 1.0087, + "step": 13740 + }, + { + "epoch": 0.8348623853211009, + "grad_norm": 0.11533943563699722, + "learning_rate": 6.623144426695016e-06, + "loss": 0.9914, + "step": 13741 + }, + { + "epoch": 0.8349231423537274, + "grad_norm": 0.12492358684539795, + "learning_rate": 6.618383988081717e-06, + "loss": 1.0144, + "step": 13742 + }, + { + "epoch": 0.834983899386354, + "grad_norm": 0.1906735748052597, + "learning_rate": 6.613625139629936e-06, + "loss": 1.0829, + "step": 13743 + }, + { + "epoch": 0.8350446564189805, + "grad_norm": 4.392884254455566, + "learning_rate": 6.608867881514108e-06, + "loss": 1.0937, + "step": 13744 + }, + { + "epoch": 0.835105413451607, + "grad_norm": 0.12836307287216187, + "learning_rate": 6.604112213908631e-06, + "loss": 1.0141, + "step": 13745 + }, + { + "epoch": 0.8351661704842336, + "grad_norm": 0.14812135696411133, + "learning_rate": 6.5993581369877874e-06, + "loss": 0.9844, + "step": 13746 + }, + { + "epoch": 0.8352269275168601, + "grad_norm": 0.23489461839199066, + "learning_rate": 6.594605650925867e-06, + "loss": 0.9894, + "step": 13747 + }, + { + "epoch": 0.8352876845494867, + "grad_norm": 0.14451515674591064, + "learning_rate": 6.589854755897068e-06, + "loss": 1.0561, + "step": 13748 + }, + { + "epoch": 0.8353484415821131, + "grad_norm": 0.12397667020559311, + "learning_rate": 6.585105452075535e-06, + "loss": 0.9948, + "step": 13749 + }, + { + "epoch": 0.8354091986147396, + "grad_norm": 0.29107969999313354, + "learning_rate": 6.580357739635357e-06, + "loss": 1.3138, + "step": 13750 + }, + { + "epoch": 0.8354699556473661, + "grad_norm": 0.34698495268821716, + "learning_rate": 6.575611618750555e-06, + "loss": 1.0767, + "step": 13751 + }, + { + "epoch": 0.8355307126799927, + "grad_norm": 0.13014842569828033, + "learning_rate": 6.5708670895951e-06, + "loss": 1.0865, + "step": 13752 + }, + { + "epoch": 0.8355914697126192, + "grad_norm": 0.17388007044792175, + "learning_rate": 6.566124152342912e-06, + "loss": 1.0445, + "step": 13753 + }, + { + "epoch": 0.8356522267452458, + "grad_norm": 0.1674811840057373, + "learning_rate": 6.5613828071678374e-06, + "loss": 1.042, + "step": 13754 + }, + { + "epoch": 0.8357129837778723, + "grad_norm": 0.12836028635501862, + "learning_rate": 6.556643054243672e-06, + "loss": 1.0259, + "step": 13755 + }, + { + "epoch": 0.8357737408104988, + "grad_norm": 0.12324405461549759, + "learning_rate": 6.551904893744154e-06, + "loss": 1.0335, + "step": 13756 + }, + { + "epoch": 0.8358344978431254, + "grad_norm": 0.14895343780517578, + "learning_rate": 6.547168325842945e-06, + "loss": 1.0451, + "step": 13757 + }, + { + "epoch": 0.8358952548757519, + "grad_norm": 0.1360788494348526, + "learning_rate": 6.542433350713706e-06, + "loss": 1.0168, + "step": 13758 + }, + { + "epoch": 0.8359560119083784, + "grad_norm": 0.14909063279628754, + "learning_rate": 6.537699968529965e-06, + "loss": 1.0466, + "step": 13759 + }, + { + "epoch": 0.836016768941005, + "grad_norm": 0.19881941378116608, + "learning_rate": 6.532968179465227e-06, + "loss": 1.1461, + "step": 13760 + }, + { + "epoch": 0.8360775259736315, + "grad_norm": 0.13095605373382568, + "learning_rate": 6.528237983692942e-06, + "loss": 1.0258, + "step": 13761 + }, + { + "epoch": 0.8361382830062579, + "grad_norm": 0.20683294534683228, + "learning_rate": 6.523509381386489e-06, + "loss": 1.0635, + "step": 13762 + }, + { + "epoch": 0.8361990400388845, + "grad_norm": 1.3160098791122437, + "learning_rate": 6.5187823727192125e-06, + "loss": 0.9818, + "step": 13763 + }, + { + "epoch": 0.836259797071511, + "grad_norm": 0.2944711446762085, + "learning_rate": 6.514056957864373e-06, + "loss": 1.3098, + "step": 13764 + }, + { + "epoch": 0.8363205541041375, + "grad_norm": 0.18848475813865662, + "learning_rate": 6.509333136995194e-06, + "loss": 0.983, + "step": 13765 + }, + { + "epoch": 0.8363813111367641, + "grad_norm": 0.17944465577602386, + "learning_rate": 6.504610910284803e-06, + "loss": 1.0251, + "step": 13766 + }, + { + "epoch": 0.8364420681693906, + "grad_norm": 0.12990576028823853, + "learning_rate": 6.499890277906295e-06, + "loss": 1.0546, + "step": 13767 + }, + { + "epoch": 0.8365028252020171, + "grad_norm": 0.24442777037620544, + "learning_rate": 6.495171240032733e-06, + "loss": 1.0422, + "step": 13768 + }, + { + "epoch": 0.8365635822346437, + "grad_norm": 0.22109390795230865, + "learning_rate": 6.490453796837071e-06, + "loss": 1.0062, + "step": 13769 + }, + { + "epoch": 0.8366243392672702, + "grad_norm": 0.18738365173339844, + "learning_rate": 6.4857379484922375e-06, + "loss": 1.1189, + "step": 13770 + }, + { + "epoch": 0.8366850962998967, + "grad_norm": 0.13478830456733704, + "learning_rate": 6.481023695171107e-06, + "loss": 1.0496, + "step": 13771 + }, + { + "epoch": 0.8367458533325233, + "grad_norm": 0.12917295098304749, + "learning_rate": 6.47631103704644e-06, + "loss": 1.0032, + "step": 13772 + }, + { + "epoch": 0.8368066103651498, + "grad_norm": 0.12523147463798523, + "learning_rate": 6.471599974291009e-06, + "loss": 1.0065, + "step": 13773 + }, + { + "epoch": 0.8368673673977763, + "grad_norm": 0.24927231669425964, + "learning_rate": 6.4668905070775055e-06, + "loss": 1.119, + "step": 13774 + }, + { + "epoch": 0.8369281244304028, + "grad_norm": 0.6039045453071594, + "learning_rate": 6.46218263557854e-06, + "loss": 1.014, + "step": 13775 + }, + { + "epoch": 0.8369888814630293, + "grad_norm": 0.6941308379173279, + "learning_rate": 6.4574763599666856e-06, + "loss": 1.0267, + "step": 13776 + }, + { + "epoch": 0.8370496384956558, + "grad_norm": 0.19110102951526642, + "learning_rate": 6.452771680414449e-06, + "loss": 1.1443, + "step": 13777 + }, + { + "epoch": 0.8371103955282824, + "grad_norm": 0.22842685878276825, + "learning_rate": 6.448068597094292e-06, + "loss": 1.1298, + "step": 13778 + }, + { + "epoch": 0.8371711525609089, + "grad_norm": 0.165660098195076, + "learning_rate": 6.443367110178594e-06, + "loss": 1.0517, + "step": 13779 + }, + { + "epoch": 0.8372319095935354, + "grad_norm": 0.19814038276672363, + "learning_rate": 6.438667219839689e-06, + "loss": 1.0658, + "step": 13780 + }, + { + "epoch": 0.837292666626162, + "grad_norm": 0.1399838626384735, + "learning_rate": 6.433968926249867e-06, + "loss": 1.0104, + "step": 13781 + }, + { + "epoch": 0.8373534236587885, + "grad_norm": 0.19463369250297546, + "learning_rate": 6.429272229581329e-06, + "loss": 1.086, + "step": 13782 + }, + { + "epoch": 0.837414180691415, + "grad_norm": 0.18467876315116882, + "learning_rate": 6.424577130006232e-06, + "loss": 0.9947, + "step": 13783 + }, + { + "epoch": 0.8374749377240416, + "grad_norm": 0.2543829679489136, + "learning_rate": 6.419883627696688e-06, + "loss": 1.1177, + "step": 13784 + }, + { + "epoch": 0.8375356947566681, + "grad_norm": 0.13764002919197083, + "learning_rate": 6.415191722824749e-06, + "loss": 1.0182, + "step": 13785 + }, + { + "epoch": 0.8375964517892946, + "grad_norm": 0.15422707796096802, + "learning_rate": 6.41050141556237e-06, + "loss": 1.0679, + "step": 13786 + }, + { + "epoch": 0.8376572088219212, + "grad_norm": 0.16771340370178223, + "learning_rate": 6.405812706081488e-06, + "loss": 1.0784, + "step": 13787 + }, + { + "epoch": 0.8377179658545477, + "grad_norm": 0.15040408074855804, + "learning_rate": 6.401125594553959e-06, + "loss": 0.9982, + "step": 13788 + }, + { + "epoch": 0.8377787228871741, + "grad_norm": 0.36628666520118713, + "learning_rate": 6.396440081151606e-06, + "loss": 1.1578, + "step": 13789 + }, + { + "epoch": 0.8378394799198007, + "grad_norm": 0.13845334947109222, + "learning_rate": 6.391756166046176e-06, + "loss": 1.036, + "step": 13790 + }, + { + "epoch": 0.8379002369524272, + "grad_norm": 0.17863404750823975, + "learning_rate": 6.387073849409347e-06, + "loss": 1.0475, + "step": 13791 + }, + { + "epoch": 0.8379609939850537, + "grad_norm": 0.2732945382595062, + "learning_rate": 6.3823931314127675e-06, + "loss": 1.0574, + "step": 13792 + }, + { + "epoch": 0.8380217510176803, + "grad_norm": 0.18042978644371033, + "learning_rate": 6.377714012227981e-06, + "loss": 1.0728, + "step": 13793 + }, + { + "epoch": 0.8380825080503068, + "grad_norm": 0.2728457748889923, + "learning_rate": 6.3730364920265295e-06, + "loss": 1.1764, + "step": 13794 + }, + { + "epoch": 0.8381432650829334, + "grad_norm": 0.13862240314483643, + "learning_rate": 6.3683605709798534e-06, + "loss": 1.0499, + "step": 13795 + }, + { + "epoch": 0.8382040221155599, + "grad_norm": 0.1333470195531845, + "learning_rate": 6.36368624925936e-06, + "loss": 1.0509, + "step": 13796 + }, + { + "epoch": 0.8382647791481864, + "grad_norm": 0.1823793351650238, + "learning_rate": 6.3590135270363784e-06, + "loss": 1.0646, + "step": 13797 + }, + { + "epoch": 0.838325536180813, + "grad_norm": 0.203656405210495, + "learning_rate": 6.354342404482194e-06, + "loss": 1.0316, + "step": 13798 + }, + { + "epoch": 0.8383862932134395, + "grad_norm": 0.917542576789856, + "learning_rate": 6.349672881768021e-06, + "loss": 1.1778, + "step": 13799 + }, + { + "epoch": 0.838447050246066, + "grad_norm": 0.15671344101428986, + "learning_rate": 6.345004959065032e-06, + "loss": 1.0712, + "step": 13800 + }, + { + "epoch": 0.8385078072786926, + "grad_norm": 0.12935946881771088, + "learning_rate": 6.340338636544319e-06, + "loss": 0.9947, + "step": 13801 + }, + { + "epoch": 0.838568564311319, + "grad_norm": 0.21335920691490173, + "learning_rate": 6.335673914376938e-06, + "loss": 1.0109, + "step": 13802 + }, + { + "epoch": 0.8386293213439455, + "grad_norm": 0.2272973209619522, + "learning_rate": 6.331010792733866e-06, + "loss": 1.1548, + "step": 13803 + }, + { + "epoch": 0.8386900783765721, + "grad_norm": 0.16329869627952576, + "learning_rate": 6.326349271786025e-06, + "loss": 1.021, + "step": 13804 + }, + { + "epoch": 0.8387508354091986, + "grad_norm": 0.2354801446199417, + "learning_rate": 6.321689351704313e-06, + "loss": 1.0048, + "step": 13805 + }, + { + "epoch": 0.8388115924418251, + "grad_norm": 0.21269145607948303, + "learning_rate": 6.31703103265951e-06, + "loss": 1.097, + "step": 13806 + }, + { + "epoch": 0.8388723494744517, + "grad_norm": 0.5345966815948486, + "learning_rate": 6.312374314822378e-06, + "loss": 1.2048, + "step": 13807 + }, + { + "epoch": 0.8389331065070782, + "grad_norm": 0.12914706766605377, + "learning_rate": 6.307719198363615e-06, + "loss": 1.0319, + "step": 13808 + }, + { + "epoch": 0.8389938635397047, + "grad_norm": 0.21124130487442017, + "learning_rate": 6.303065683453835e-06, + "loss": 1.0432, + "step": 13809 + }, + { + "epoch": 0.8390546205723313, + "grad_norm": 0.14039495587348938, + "learning_rate": 6.298413770263639e-06, + "loss": 1.0103, + "step": 13810 + }, + { + "epoch": 0.8391153776049578, + "grad_norm": 0.6047753095626831, + "learning_rate": 6.293763458963536e-06, + "loss": 1.0906, + "step": 13811 + }, + { + "epoch": 0.8391761346375843, + "grad_norm": 0.655311107635498, + "learning_rate": 6.28911474972399e-06, + "loss": 1.0563, + "step": 13812 + }, + { + "epoch": 0.8392368916702109, + "grad_norm": 0.13012945652008057, + "learning_rate": 6.284467642715375e-06, + "loss": 1.034, + "step": 13813 + }, + { + "epoch": 0.8392976487028374, + "grad_norm": 0.15770499408245087, + "learning_rate": 6.279822138108054e-06, + "loss": 1.0792, + "step": 13814 + }, + { + "epoch": 0.8393584057354638, + "grad_norm": 0.19351716339588165, + "learning_rate": 6.27517823607231e-06, + "loss": 1.0979, + "step": 13815 + }, + { + "epoch": 0.8394191627680904, + "grad_norm": 0.18629764020442963, + "learning_rate": 6.270535936778354e-06, + "loss": 1.1077, + "step": 13816 + }, + { + "epoch": 0.8394799198007169, + "grad_norm": 0.2389356642961502, + "learning_rate": 6.265895240396358e-06, + "loss": 1.1243, + "step": 13817 + }, + { + "epoch": 0.8395406768333434, + "grad_norm": 0.17585481703281403, + "learning_rate": 6.261256147096423e-06, + "loss": 1.0751, + "step": 13818 + }, + { + "epoch": 0.83960143386597, + "grad_norm": 0.1972997784614563, + "learning_rate": 6.256618657048602e-06, + "loss": 1.0476, + "step": 13819 + }, + { + "epoch": 0.8396621908985965, + "grad_norm": 0.20351679623126984, + "learning_rate": 6.2519827704228765e-06, + "loss": 1.0597, + "step": 13820 + }, + { + "epoch": 0.839722947931223, + "grad_norm": 0.20039457082748413, + "learning_rate": 6.247348487389182e-06, + "loss": 1.0685, + "step": 13821 + }, + { + "epoch": 0.8397837049638496, + "grad_norm": 0.22340111434459686, + "learning_rate": 6.242715808117383e-06, + "loss": 1.1812, + "step": 13822 + }, + { + "epoch": 0.8398444619964761, + "grad_norm": 0.234945148229599, + "learning_rate": 6.238084732777294e-06, + "loss": 1.1025, + "step": 13823 + }, + { + "epoch": 0.8399052190291026, + "grad_norm": 0.21438494324684143, + "learning_rate": 6.233455261538662e-06, + "loss": 1.1602, + "step": 13824 + }, + { + "epoch": 0.8399659760617292, + "grad_norm": 0.12407946586608887, + "learning_rate": 6.228827394571201e-06, + "loss": 0.9918, + "step": 13825 + }, + { + "epoch": 0.8400267330943557, + "grad_norm": 0.14407049119472504, + "learning_rate": 6.224201132044527e-06, + "loss": 0.9945, + "step": 13826 + }, + { + "epoch": 0.8400874901269823, + "grad_norm": 0.24188655614852905, + "learning_rate": 6.219576474128224e-06, + "loss": 1.0002, + "step": 13827 + }, + { + "epoch": 0.8401482471596087, + "grad_norm": 0.11767315864562988, + "learning_rate": 6.214953420991803e-06, + "loss": 1.064, + "step": 13828 + }, + { + "epoch": 0.8402090041922352, + "grad_norm": 0.15885762870311737, + "learning_rate": 6.210331972804723e-06, + "loss": 1.1193, + "step": 13829 + }, + { + "epoch": 0.8402697612248617, + "grad_norm": 0.302364319562912, + "learning_rate": 6.205712129736396e-06, + "loss": 1.3147, + "step": 13830 + }, + { + "epoch": 0.8403305182574883, + "grad_norm": 0.13661916553974152, + "learning_rate": 6.201093891956156e-06, + "loss": 1.0037, + "step": 13831 + }, + { + "epoch": 0.8403912752901148, + "grad_norm": 0.23532426357269287, + "learning_rate": 6.196477259633299e-06, + "loss": 1.0461, + "step": 13832 + }, + { + "epoch": 0.8404520323227413, + "grad_norm": 0.26025667786598206, + "learning_rate": 6.191862232937018e-06, + "loss": 1.1697, + "step": 13833 + }, + { + "epoch": 0.8405127893553679, + "grad_norm": 0.4833151698112488, + "learning_rate": 6.187248812036489e-06, + "loss": 1.1514, + "step": 13834 + }, + { + "epoch": 0.8405735463879944, + "grad_norm": 1.5074602365493774, + "learning_rate": 6.182636997100833e-06, + "loss": 0.9879, + "step": 13835 + }, + { + "epoch": 0.840634303420621, + "grad_norm": 0.2625046968460083, + "learning_rate": 6.178026788299091e-06, + "loss": 1.0867, + "step": 13836 + }, + { + "epoch": 0.8406950604532475, + "grad_norm": 0.3009073734283447, + "learning_rate": 6.17341818580024e-06, + "loss": 1.0169, + "step": 13837 + }, + { + "epoch": 0.840755817485874, + "grad_norm": 0.14638790488243103, + "learning_rate": 6.168811189773233e-06, + "loss": 1.0263, + "step": 13838 + }, + { + "epoch": 0.8408165745185006, + "grad_norm": 0.14669868350028992, + "learning_rate": 6.164205800386902e-06, + "loss": 1.047, + "step": 13839 + }, + { + "epoch": 0.8408773315511271, + "grad_norm": 0.16907386481761932, + "learning_rate": 6.15960201781009e-06, + "loss": 1.0771, + "step": 13840 + }, + { + "epoch": 0.8409380885837535, + "grad_norm": 0.1544034630060196, + "learning_rate": 6.154999842211545e-06, + "loss": 0.9448, + "step": 13841 + }, + { + "epoch": 0.84099884561638, + "grad_norm": 0.49631333351135254, + "learning_rate": 6.150399273759949e-06, + "loss": 1.0643, + "step": 13842 + }, + { + "epoch": 0.8410596026490066, + "grad_norm": 0.18336763978004456, + "learning_rate": 6.145800312623951e-06, + "loss": 1.0603, + "step": 13843 + }, + { + "epoch": 0.8411203596816331, + "grad_norm": 0.1715751737356186, + "learning_rate": 6.141202958972114e-06, + "loss": 0.9868, + "step": 13844 + }, + { + "epoch": 0.8411811167142597, + "grad_norm": 0.20169852674007416, + "learning_rate": 6.1366072129729615e-06, + "loss": 0.9886, + "step": 13845 + }, + { + "epoch": 0.8412418737468862, + "grad_norm": 0.254020094871521, + "learning_rate": 6.132013074794951e-06, + "loss": 1.2564, + "step": 13846 + }, + { + "epoch": 0.8413026307795127, + "grad_norm": 0.33783087134361267, + "learning_rate": 6.127420544606477e-06, + "loss": 1.1018, + "step": 13847 + }, + { + "epoch": 0.8413633878121393, + "grad_norm": 0.27870529890060425, + "learning_rate": 6.122829622575888e-06, + "loss": 1.1691, + "step": 13848 + }, + { + "epoch": 0.8414241448447658, + "grad_norm": 0.21987856924533844, + "learning_rate": 6.118240308871459e-06, + "loss": 0.9871, + "step": 13849 + }, + { + "epoch": 0.8414849018773923, + "grad_norm": 0.18808060884475708, + "learning_rate": 6.113652603661402e-06, + "loss": 1.0245, + "step": 13850 + }, + { + "epoch": 0.8415456589100189, + "grad_norm": 0.1884629875421524, + "learning_rate": 6.109066507113903e-06, + "loss": 1.1147, + "step": 13851 + }, + { + "epoch": 0.8416064159426454, + "grad_norm": 0.17707109451293945, + "learning_rate": 6.104482019397068e-06, + "loss": 1.1211, + "step": 13852 + }, + { + "epoch": 0.8416671729752719, + "grad_norm": 0.5027336478233337, + "learning_rate": 6.099899140678916e-06, + "loss": 1.2022, + "step": 13853 + }, + { + "epoch": 0.8417279300078984, + "grad_norm": 0.19862870872020721, + "learning_rate": 6.095317871127448e-06, + "loss": 1.1613, + "step": 13854 + }, + { + "epoch": 0.8417886870405249, + "grad_norm": 0.17862677574157715, + "learning_rate": 6.090738210910579e-06, + "loss": 1.0476, + "step": 13855 + }, + { + "epoch": 0.8418494440731514, + "grad_norm": 0.26751166582107544, + "learning_rate": 6.086160160196197e-06, + "loss": 1.0431, + "step": 13856 + }, + { + "epoch": 0.841910201105778, + "grad_norm": 0.24427662789821625, + "learning_rate": 6.081583719152106e-06, + "loss": 1.1561, + "step": 13857 + }, + { + "epoch": 0.8419709581384045, + "grad_norm": 0.827124297618866, + "learning_rate": 6.07700888794605e-06, + "loss": 1.3772, + "step": 13858 + }, + { + "epoch": 0.842031715171031, + "grad_norm": 0.24690276384353638, + "learning_rate": 6.072435666745735e-06, + "loss": 1.1618, + "step": 13859 + }, + { + "epoch": 0.8420924722036576, + "grad_norm": 0.12360241264104843, + "learning_rate": 6.067864055718764e-06, + "loss": 1.0097, + "step": 13860 + }, + { + "epoch": 0.8421532292362841, + "grad_norm": 0.1359236091375351, + "learning_rate": 6.063294055032731e-06, + "loss": 1.0476, + "step": 13861 + }, + { + "epoch": 0.8422139862689106, + "grad_norm": 0.2117934226989746, + "learning_rate": 6.058725664855153e-06, + "loss": 1.1616, + "step": 13862 + }, + { + "epoch": 0.8422747433015372, + "grad_norm": 0.3695821166038513, + "learning_rate": 6.054158885353478e-06, + "loss": 1.0294, + "step": 13863 + }, + { + "epoch": 0.8423355003341637, + "grad_norm": 0.2611178457736969, + "learning_rate": 6.049593716695107e-06, + "loss": 1.0921, + "step": 13864 + }, + { + "epoch": 0.8423962573667902, + "grad_norm": 0.12105994671583176, + "learning_rate": 6.045030159047366e-06, + "loss": 1.0092, + "step": 13865 + }, + { + "epoch": 0.8424570143994168, + "grad_norm": 0.2713248133659363, + "learning_rate": 6.040468212577549e-06, + "loss": 1.2138, + "step": 13866 + }, + { + "epoch": 0.8425177714320432, + "grad_norm": 0.21293891966342926, + "learning_rate": 6.0359078774528625e-06, + "loss": 1.0499, + "step": 13867 + }, + { + "epoch": 0.8425785284646697, + "grad_norm": 0.11982978135347366, + "learning_rate": 6.031349153840471e-06, + "loss": 1.0175, + "step": 13868 + }, + { + "epoch": 0.8426392854972963, + "grad_norm": 0.15622039139270782, + "learning_rate": 6.026792041907475e-06, + "loss": 0.9886, + "step": 13869 + }, + { + "epoch": 0.8427000425299228, + "grad_norm": 0.20125989615917206, + "learning_rate": 6.022236541820919e-06, + "loss": 1.0541, + "step": 13870 + }, + { + "epoch": 0.8427607995625493, + "grad_norm": 0.24577215313911438, + "learning_rate": 6.01768265374777e-06, + "loss": 1.0348, + "step": 13871 + }, + { + "epoch": 0.8428215565951759, + "grad_norm": 0.17107908427715302, + "learning_rate": 6.013130377854986e-06, + "loss": 1.0315, + "step": 13872 + }, + { + "epoch": 0.8428823136278024, + "grad_norm": 1.0838747024536133, + "learning_rate": 6.008579714309398e-06, + "loss": 1.1279, + "step": 13873 + }, + { + "epoch": 0.842943070660429, + "grad_norm": 0.6052243709564209, + "learning_rate": 6.004030663277832e-06, + "loss": 1.4386, + "step": 13874 + }, + { + "epoch": 0.8430038276930555, + "grad_norm": 0.12610965967178345, + "learning_rate": 5.999483224927016e-06, + "loss": 1.0349, + "step": 13875 + }, + { + "epoch": 0.843064584725682, + "grad_norm": 0.19866062700748444, + "learning_rate": 5.994937399423645e-06, + "loss": 1.04, + "step": 13876 + }, + { + "epoch": 0.8431253417583086, + "grad_norm": 0.15551242232322693, + "learning_rate": 5.990393186934357e-06, + "loss": 1.0299, + "step": 13877 + }, + { + "epoch": 0.8431860987909351, + "grad_norm": 0.150228351354599, + "learning_rate": 5.9858505876257136e-06, + "loss": 1.0849, + "step": 13878 + }, + { + "epoch": 0.8432468558235616, + "grad_norm": 0.21315334737300873, + "learning_rate": 5.981309601664237e-06, + "loss": 1.1063, + "step": 13879 + }, + { + "epoch": 0.843307612856188, + "grad_norm": 1.5416508913040161, + "learning_rate": 5.97677022921635e-06, + "loss": 1.0845, + "step": 13880 + }, + { + "epoch": 0.8433683698888146, + "grad_norm": 0.16873133182525635, + "learning_rate": 5.972232470448452e-06, + "loss": 0.961, + "step": 13881 + }, + { + "epoch": 0.8434291269214411, + "grad_norm": 0.2896052896976471, + "learning_rate": 5.967696325526895e-06, + "loss": 0.9863, + "step": 13882 + }, + { + "epoch": 0.8434898839540677, + "grad_norm": 0.1699301153421402, + "learning_rate": 5.9631617946179385e-06, + "loss": 1.0149, + "step": 13883 + }, + { + "epoch": 0.8435506409866942, + "grad_norm": 0.16436615586280823, + "learning_rate": 5.958628877887801e-06, + "loss": 1.0384, + "step": 13884 + }, + { + "epoch": 0.8436113980193207, + "grad_norm": 0.21357348561286926, + "learning_rate": 5.9540975755026426e-06, + "loss": 1.1106, + "step": 13885 + }, + { + "epoch": 0.8436721550519473, + "grad_norm": 0.1315164864063263, + "learning_rate": 5.949567887628532e-06, + "loss": 1.0085, + "step": 13886 + }, + { + "epoch": 0.8437329120845738, + "grad_norm": 0.2072315365076065, + "learning_rate": 5.945039814431541e-06, + "loss": 1.1103, + "step": 13887 + }, + { + "epoch": 0.8437936691172003, + "grad_norm": 0.36309561133384705, + "learning_rate": 5.940513356077626e-06, + "loss": 1.089, + "step": 13888 + }, + { + "epoch": 0.8438544261498269, + "grad_norm": 0.25347307324409485, + "learning_rate": 5.93598851273271e-06, + "loss": 1.0797, + "step": 13889 + }, + { + "epoch": 0.8439151831824534, + "grad_norm": 0.22249317169189453, + "learning_rate": 5.931465284562648e-06, + "loss": 1.1483, + "step": 13890 + }, + { + "epoch": 0.8439759402150799, + "grad_norm": 0.1306060403585434, + "learning_rate": 5.926943671733254e-06, + "loss": 1.0348, + "step": 13891 + }, + { + "epoch": 0.8440366972477065, + "grad_norm": 0.1565602421760559, + "learning_rate": 5.922423674410255e-06, + "loss": 1.0326, + "step": 13892 + }, + { + "epoch": 0.844097454280333, + "grad_norm": 0.16398748755455017, + "learning_rate": 5.917905292759334e-06, + "loss": 1.0373, + "step": 13893 + }, + { + "epoch": 0.8441582113129594, + "grad_norm": 0.1848883330821991, + "learning_rate": 5.913388526946117e-06, + "loss": 1.0739, + "step": 13894 + }, + { + "epoch": 0.844218968345586, + "grad_norm": 0.21457497775554657, + "learning_rate": 5.9088733771361645e-06, + "loss": 1.1354, + "step": 13895 + }, + { + "epoch": 0.8442797253782125, + "grad_norm": 0.19480466842651367, + "learning_rate": 5.904359843494978e-06, + "loss": 1.0938, + "step": 13896 + }, + { + "epoch": 0.844340482410839, + "grad_norm": 0.136036217212677, + "learning_rate": 5.899847926188001e-06, + "loss": 1.0341, + "step": 13897 + }, + { + "epoch": 0.8444012394434656, + "grad_norm": 0.129334956407547, + "learning_rate": 5.895337625380632e-06, + "loss": 0.9657, + "step": 13898 + }, + { + "epoch": 0.8444619964760921, + "grad_norm": 0.13704350590705872, + "learning_rate": 5.890828941238191e-06, + "loss": 1.063, + "step": 13899 + }, + { + "epoch": 0.8445227535087186, + "grad_norm": 0.20501579344272614, + "learning_rate": 5.886321873925937e-06, + "loss": 1.0644, + "step": 13900 + }, + { + "epoch": 0.8445835105413452, + "grad_norm": 0.2669939696788788, + "learning_rate": 5.881816423609082e-06, + "loss": 1.1002, + "step": 13901 + }, + { + "epoch": 0.8446442675739717, + "grad_norm": 0.2255709022283554, + "learning_rate": 5.87731259045276e-06, + "loss": 1.0099, + "step": 13902 + }, + { + "epoch": 0.8447050246065982, + "grad_norm": 0.14091111719608307, + "learning_rate": 5.8728103746220894e-06, + "loss": 1.0953, + "step": 13903 + }, + { + "epoch": 0.8447657816392248, + "grad_norm": 0.24488244950771332, + "learning_rate": 5.868309776282077e-06, + "loss": 1.0359, + "step": 13904 + }, + { + "epoch": 0.8448265386718513, + "grad_norm": 0.21729111671447754, + "learning_rate": 5.863810795597718e-06, + "loss": 1.1613, + "step": 13905 + }, + { + "epoch": 0.8448872957044778, + "grad_norm": 0.14834541082382202, + "learning_rate": 5.859313432733882e-06, + "loss": 1.0533, + "step": 13906 + }, + { + "epoch": 0.8449480527371043, + "grad_norm": 0.1557435542345047, + "learning_rate": 5.854817687855451e-06, + "loss": 1.0507, + "step": 13907 + }, + { + "epoch": 0.8450088097697308, + "grad_norm": 0.1901821345090866, + "learning_rate": 5.850323561127213e-06, + "loss": 1.1568, + "step": 13908 + }, + { + "epoch": 0.8450695668023573, + "grad_norm": 0.19953668117523193, + "learning_rate": 5.845831052713902e-06, + "loss": 1.0728, + "step": 13909 + }, + { + "epoch": 0.8451303238349839, + "grad_norm": 0.11421173065900803, + "learning_rate": 5.841340162780184e-06, + "loss": 1.0284, + "step": 13910 + }, + { + "epoch": 0.8451910808676104, + "grad_norm": 0.13216617703437805, + "learning_rate": 5.836850891490681e-06, + "loss": 0.9435, + "step": 13911 + }, + { + "epoch": 0.845251837900237, + "grad_norm": 0.1485094130039215, + "learning_rate": 5.832363239009947e-06, + "loss": 1.0187, + "step": 13912 + }, + { + "epoch": 0.8453125949328635, + "grad_norm": 0.14996857941150665, + "learning_rate": 5.827877205502469e-06, + "loss": 1.0129, + "step": 13913 + }, + { + "epoch": 0.84537335196549, + "grad_norm": 0.2526437044143677, + "learning_rate": 5.823392791132698e-06, + "loss": 1.1466, + "step": 13914 + }, + { + "epoch": 0.8454341089981166, + "grad_norm": 0.148070827126503, + "learning_rate": 5.818909996064998e-06, + "loss": 1.0118, + "step": 13915 + }, + { + "epoch": 0.8454948660307431, + "grad_norm": 0.23735779523849487, + "learning_rate": 5.814428820463691e-06, + "loss": 1.1202, + "step": 13916 + }, + { + "epoch": 0.8455556230633696, + "grad_norm": 0.12583379447460175, + "learning_rate": 5.809949264493031e-06, + "loss": 1.0018, + "step": 13917 + }, + { + "epoch": 0.8456163800959962, + "grad_norm": 5.075977325439453, + "learning_rate": 5.80547132831723e-06, + "loss": 1.0313, + "step": 13918 + }, + { + "epoch": 0.8456771371286227, + "grad_norm": 0.2426215410232544, + "learning_rate": 5.800995012100435e-06, + "loss": 1.1597, + "step": 13919 + }, + { + "epoch": 0.8457378941612491, + "grad_norm": 0.24412548542022705, + "learning_rate": 5.796520316006693e-06, + "loss": 1.1213, + "step": 13920 + }, + { + "epoch": 0.8457986511938757, + "grad_norm": 0.2506866753101349, + "learning_rate": 5.792047240200049e-06, + "loss": 1.0678, + "step": 13921 + }, + { + "epoch": 0.8458594082265022, + "grad_norm": 0.1865059733390808, + "learning_rate": 5.787575784844446e-06, + "loss": 1.0239, + "step": 13922 + }, + { + "epoch": 0.8459201652591287, + "grad_norm": 3.5601491928100586, + "learning_rate": 5.783105950103812e-06, + "loss": 1.0292, + "step": 13923 + }, + { + "epoch": 0.8459809222917553, + "grad_norm": 0.4360800087451935, + "learning_rate": 5.778637736141973e-06, + "loss": 0.9731, + "step": 13924 + }, + { + "epoch": 0.8460416793243818, + "grad_norm": 0.5295563340187073, + "learning_rate": 5.7741711431227255e-06, + "loss": 1.1438, + "step": 13925 + }, + { + "epoch": 0.8461024363570083, + "grad_norm": 0.40207743644714355, + "learning_rate": 5.769706171209777e-06, + "loss": 1.183, + "step": 13926 + }, + { + "epoch": 0.8461631933896349, + "grad_norm": 1.8269869089126587, + "learning_rate": 5.7652428205667855e-06, + "loss": 1.0609, + "step": 13927 + }, + { + "epoch": 0.8462239504222614, + "grad_norm": 0.14578919112682343, + "learning_rate": 5.760781091357376e-06, + "loss": 1.054, + "step": 13928 + }, + { + "epoch": 0.8462847074548879, + "grad_norm": 0.15031149983406067, + "learning_rate": 5.756320983745089e-06, + "loss": 1.0525, + "step": 13929 + }, + { + "epoch": 0.8463454644875145, + "grad_norm": 0.5003486275672913, + "learning_rate": 5.751862497893412e-06, + "loss": 1.1771, + "step": 13930 + }, + { + "epoch": 0.846406221520141, + "grad_norm": 0.3653428554534912, + "learning_rate": 5.747405633965763e-06, + "loss": 1.0193, + "step": 13931 + }, + { + "epoch": 0.8464669785527675, + "grad_norm": 0.26680123805999756, + "learning_rate": 5.742950392125518e-06, + "loss": 0.9612, + "step": 13932 + }, + { + "epoch": 0.846527735585394, + "grad_norm": 0.37356022000312805, + "learning_rate": 5.738496772535984e-06, + "loss": 1.2484, + "step": 13933 + }, + { + "epoch": 0.8465884926180205, + "grad_norm": 0.1461615413427353, + "learning_rate": 5.734044775360397e-06, + "loss": 1.0749, + "step": 13934 + }, + { + "epoch": 0.846649249650647, + "grad_norm": 0.16213323175907135, + "learning_rate": 5.729594400761961e-06, + "loss": 1.0527, + "step": 13935 + }, + { + "epoch": 0.8467100066832736, + "grad_norm": 0.27307310700416565, + "learning_rate": 5.725145648903801e-06, + "loss": 1.0225, + "step": 13936 + }, + { + "epoch": 0.8467707637159001, + "grad_norm": 4.1030473709106445, + "learning_rate": 5.720698519948986e-06, + "loss": 1.0553, + "step": 13937 + }, + { + "epoch": 0.8468315207485266, + "grad_norm": 0.2991310656070709, + "learning_rate": 5.716253014060513e-06, + "loss": 1.1273, + "step": 13938 + }, + { + "epoch": 0.8468922777811532, + "grad_norm": 0.3756154775619507, + "learning_rate": 5.7118091314013635e-06, + "loss": 1.0546, + "step": 13939 + }, + { + "epoch": 0.8469530348137797, + "grad_norm": 0.15142783522605896, + "learning_rate": 5.707366872134401e-06, + "loss": 1.0242, + "step": 13940 + }, + { + "epoch": 0.8470137918464062, + "grad_norm": 0.12368426471948624, + "learning_rate": 5.702926236422474e-06, + "loss": 0.9876, + "step": 13941 + }, + { + "epoch": 0.8470745488790328, + "grad_norm": 0.15336962044239044, + "learning_rate": 5.6984872244283395e-06, + "loss": 1.096, + "step": 13942 + }, + { + "epoch": 0.8471353059116593, + "grad_norm": 0.22228476405143738, + "learning_rate": 5.694049836314713e-06, + "loss": 1.1972, + "step": 13943 + }, + { + "epoch": 0.8471960629442858, + "grad_norm": 0.14296407997608185, + "learning_rate": 5.689614072244265e-06, + "loss": 1.0425, + "step": 13944 + }, + { + "epoch": 0.8472568199769124, + "grad_norm": 0.1372978687286377, + "learning_rate": 5.685179932379581e-06, + "loss": 1.0568, + "step": 13945 + }, + { + "epoch": 0.8473175770095388, + "grad_norm": 0.18370682001113892, + "learning_rate": 5.680747416883198e-06, + "loss": 1.0554, + "step": 13946 + }, + { + "epoch": 0.8473783340421653, + "grad_norm": 0.23924973607063293, + "learning_rate": 5.676316525917575e-06, + "loss": 1.1282, + "step": 13947 + }, + { + "epoch": 0.8474390910747919, + "grad_norm": 0.20873990654945374, + "learning_rate": 5.671887259645131e-06, + "loss": 0.9911, + "step": 13948 + }, + { + "epoch": 0.8474998481074184, + "grad_norm": 0.2274579554796219, + "learning_rate": 5.667459618228232e-06, + "loss": 1.2004, + "step": 13949 + }, + { + "epoch": 0.8475606051400449, + "grad_norm": 0.16203105449676514, + "learning_rate": 5.663033601829177e-06, + "loss": 1.0391, + "step": 13950 + }, + { + "epoch": 0.8476213621726715, + "grad_norm": 0.4387458860874176, + "learning_rate": 5.658609210610189e-06, + "loss": 1.2099, + "step": 13951 + }, + { + "epoch": 0.847682119205298, + "grad_norm": 0.2041759043931961, + "learning_rate": 5.6541864447334665e-06, + "loss": 1.1523, + "step": 13952 + }, + { + "epoch": 0.8477428762379245, + "grad_norm": 0.13228391110897064, + "learning_rate": 5.649765304361088e-06, + "loss": 1.0032, + "step": 13953 + }, + { + "epoch": 0.8478036332705511, + "grad_norm": 0.2011391818523407, + "learning_rate": 5.645345789655149e-06, + "loss": 1.0834, + "step": 13954 + }, + { + "epoch": 0.8478643903031776, + "grad_norm": 0.6580233573913574, + "learning_rate": 5.640927900777626e-06, + "loss": 1.1905, + "step": 13955 + }, + { + "epoch": 0.8479251473358042, + "grad_norm": 0.160218745470047, + "learning_rate": 5.636511637890474e-06, + "loss": 1.0322, + "step": 13956 + }, + { + "epoch": 0.8479859043684307, + "grad_norm": 0.23742710053920746, + "learning_rate": 5.632097001155556e-06, + "loss": 1.1126, + "step": 13957 + }, + { + "epoch": 0.8480466614010572, + "grad_norm": 0.19071397185325623, + "learning_rate": 5.6276839907347054e-06, + "loss": 1.1022, + "step": 13958 + }, + { + "epoch": 0.8481074184336836, + "grad_norm": 0.2212444245815277, + "learning_rate": 5.62327260678967e-06, + "loss": 1.106, + "step": 13959 + }, + { + "epoch": 0.8481681754663102, + "grad_norm": 0.16661453247070312, + "learning_rate": 5.618862849482159e-06, + "loss": 1.0882, + "step": 13960 + }, + { + "epoch": 0.8482289324989367, + "grad_norm": 0.12633618712425232, + "learning_rate": 5.6144547189738055e-06, + "loss": 1.0087, + "step": 13961 + }, + { + "epoch": 0.8482896895315633, + "grad_norm": 0.19354312121868134, + "learning_rate": 5.6100482154261945e-06, + "loss": 1.0743, + "step": 13962 + }, + { + "epoch": 0.8483504465641898, + "grad_norm": 0.18417000770568848, + "learning_rate": 5.605643339000849e-06, + "loss": 1.1809, + "step": 13963 + }, + { + "epoch": 0.8484112035968163, + "grad_norm": 0.12628979980945587, + "learning_rate": 5.6012400898592166e-06, + "loss": 1.0497, + "step": 13964 + }, + { + "epoch": 0.8484719606294429, + "grad_norm": 0.16008128225803375, + "learning_rate": 5.596838468162718e-06, + "loss": 1.0393, + "step": 13965 + }, + { + "epoch": 0.8485327176620694, + "grad_norm": 0.3026387095451355, + "learning_rate": 5.592438474072703e-06, + "loss": 1.1106, + "step": 13966 + }, + { + "epoch": 0.8485934746946959, + "grad_norm": 1.7673916816711426, + "learning_rate": 5.588040107750425e-06, + "loss": 1.0522, + "step": 13967 + }, + { + "epoch": 0.8486542317273225, + "grad_norm": 0.1445816457271576, + "learning_rate": 5.583643369357122e-06, + "loss": 1.0676, + "step": 13968 + }, + { + "epoch": 0.848714988759949, + "grad_norm": 0.1240842416882515, + "learning_rate": 5.579248259053949e-06, + "loss": 1.0183, + "step": 13969 + }, + { + "epoch": 0.8487757457925755, + "grad_norm": 0.14075827598571777, + "learning_rate": 5.574854777002026e-06, + "loss": 1.0699, + "step": 13970 + }, + { + "epoch": 0.8488365028252021, + "grad_norm": 0.21545542776584625, + "learning_rate": 5.570462923362391e-06, + "loss": 1.1221, + "step": 13971 + }, + { + "epoch": 0.8488972598578285, + "grad_norm": 0.21896840631961823, + "learning_rate": 5.566072698296032e-06, + "loss": 1.1925, + "step": 13972 + }, + { + "epoch": 0.848958016890455, + "grad_norm": 0.12367508560419083, + "learning_rate": 5.56168410196386e-06, + "loss": 1.0485, + "step": 13973 + }, + { + "epoch": 0.8490187739230816, + "grad_norm": 0.1447552740573883, + "learning_rate": 5.557297134526734e-06, + "loss": 1.0494, + "step": 13974 + }, + { + "epoch": 0.8490795309557081, + "grad_norm": 0.2258988469839096, + "learning_rate": 5.552911796145488e-06, + "loss": 1.1315, + "step": 13975 + }, + { + "epoch": 0.8491402879883346, + "grad_norm": 0.125199556350708, + "learning_rate": 5.548528086980848e-06, + "loss": 1.0325, + "step": 13976 + }, + { + "epoch": 0.8492010450209612, + "grad_norm": 0.2205362766981125, + "learning_rate": 5.544146007193501e-06, + "loss": 1.2265, + "step": 13977 + }, + { + "epoch": 0.8492618020535877, + "grad_norm": 0.1909557729959488, + "learning_rate": 5.5397655569440765e-06, + "loss": 1.0804, + "step": 13978 + }, + { + "epoch": 0.8493225590862142, + "grad_norm": 0.7384828329086304, + "learning_rate": 5.535386736393139e-06, + "loss": 1.1588, + "step": 13979 + }, + { + "epoch": 0.8493833161188408, + "grad_norm": 0.4256336987018585, + "learning_rate": 5.531009545701199e-06, + "loss": 1.0376, + "step": 13980 + }, + { + "epoch": 0.8494440731514673, + "grad_norm": 0.1923728585243225, + "learning_rate": 5.5266339850286955e-06, + "loss": 1.1178, + "step": 13981 + }, + { + "epoch": 0.8495048301840938, + "grad_norm": 0.2045133113861084, + "learning_rate": 5.522260054536022e-06, + "loss": 1.1339, + "step": 13982 + }, + { + "epoch": 0.8495655872167204, + "grad_norm": 0.1611107736825943, + "learning_rate": 5.517887754383505e-06, + "loss": 1.0379, + "step": 13983 + }, + { + "epoch": 0.8496263442493469, + "grad_norm": 0.21232765913009644, + "learning_rate": 5.5135170847314065e-06, + "loss": 0.9961, + "step": 13984 + }, + { + "epoch": 0.8496871012819733, + "grad_norm": 0.16088193655014038, + "learning_rate": 5.509148045739932e-06, + "loss": 1.0593, + "step": 13985 + }, + { + "epoch": 0.8497478583145999, + "grad_norm": 0.14637213945388794, + "learning_rate": 5.504780637569251e-06, + "loss": 1.0583, + "step": 13986 + }, + { + "epoch": 0.8498086153472264, + "grad_norm": 0.19225862622261047, + "learning_rate": 5.500414860379433e-06, + "loss": 1.0354, + "step": 13987 + }, + { + "epoch": 0.8498693723798529, + "grad_norm": 0.16727130115032196, + "learning_rate": 5.496050714330508e-06, + "loss": 1.0803, + "step": 13988 + }, + { + "epoch": 0.8499301294124795, + "grad_norm": 0.131273090839386, + "learning_rate": 5.491688199582446e-06, + "loss": 1.028, + "step": 13989 + }, + { + "epoch": 0.849990886445106, + "grad_norm": 0.2390243411064148, + "learning_rate": 5.487327316295143e-06, + "loss": 1.1606, + "step": 13990 + }, + { + "epoch": 0.8500516434777325, + "grad_norm": 0.16159091889858246, + "learning_rate": 5.482968064628474e-06, + "loss": 1.0243, + "step": 13991 + }, + { + "epoch": 0.8501124005103591, + "grad_norm": 0.1414962261915207, + "learning_rate": 5.478610444742227e-06, + "loss": 0.9823, + "step": 13992 + }, + { + "epoch": 0.8501731575429856, + "grad_norm": 0.1282268613576889, + "learning_rate": 5.474254456796107e-06, + "loss": 1.0432, + "step": 13993 + }, + { + "epoch": 0.8502339145756121, + "grad_norm": 0.10933861136436462, + "learning_rate": 5.469900100949798e-06, + "loss": 0.9937, + "step": 13994 + }, + { + "epoch": 0.8502946716082387, + "grad_norm": 0.17555667459964752, + "learning_rate": 5.465547377362895e-06, + "loss": 1.3069, + "step": 13995 + }, + { + "epoch": 0.8503554286408652, + "grad_norm": 0.13130657374858856, + "learning_rate": 5.461196286194975e-06, + "loss": 1.0471, + "step": 13996 + }, + { + "epoch": 0.8504161856734918, + "grad_norm": 0.22381973266601562, + "learning_rate": 5.4568468276055165e-06, + "loss": 1.1214, + "step": 13997 + }, + { + "epoch": 0.8504769427061183, + "grad_norm": 0.12606075406074524, + "learning_rate": 5.4524990017539415e-06, + "loss": 1.0661, + "step": 13998 + }, + { + "epoch": 0.8505376997387447, + "grad_norm": 0.2301390916109085, + "learning_rate": 5.4481528087996284e-06, + "loss": 1.1939, + "step": 13999 + }, + { + "epoch": 0.8505984567713712, + "grad_norm": 0.12403030693531036, + "learning_rate": 5.443808248901888e-06, + "loss": 1.0198, + "step": 14000 + }, + { + "epoch": 0.8506592138039978, + "grad_norm": 0.16402116417884827, + "learning_rate": 5.439465322219966e-06, + "loss": 1.0241, + "step": 14001 + }, + { + "epoch": 0.8507199708366243, + "grad_norm": 0.17502784729003906, + "learning_rate": 5.435124028913058e-06, + "loss": 0.9899, + "step": 14002 + }, + { + "epoch": 0.8507807278692509, + "grad_norm": 0.4833260774612427, + "learning_rate": 5.430784369140296e-06, + "loss": 1.1972, + "step": 14003 + }, + { + "epoch": 0.8508414849018774, + "grad_norm": 0.12640702724456787, + "learning_rate": 5.426446343060743e-06, + "loss": 1.0348, + "step": 14004 + }, + { + "epoch": 0.8509022419345039, + "grad_norm": 0.14349639415740967, + "learning_rate": 5.422109950833415e-06, + "loss": 1.0032, + "step": 14005 + }, + { + "epoch": 0.8509629989671305, + "grad_norm": 0.1341230273246765, + "learning_rate": 5.417775192617269e-06, + "loss": 0.9873, + "step": 14006 + }, + { + "epoch": 0.851023755999757, + "grad_norm": 0.17745395004749298, + "learning_rate": 5.413442068571184e-06, + "loss": 1.0676, + "step": 14007 + }, + { + "epoch": 0.8510845130323835, + "grad_norm": 0.2062326818704605, + "learning_rate": 5.4091105788540045e-06, + "loss": 1.0797, + "step": 14008 + }, + { + "epoch": 0.8511452700650101, + "grad_norm": 0.16171902418136597, + "learning_rate": 5.404780723624492e-06, + "loss": 1.0463, + "step": 14009 + }, + { + "epoch": 0.8512060270976366, + "grad_norm": 0.14392207562923431, + "learning_rate": 5.400452503041353e-06, + "loss": 1.0725, + "step": 14010 + }, + { + "epoch": 0.8512667841302631, + "grad_norm": 0.143628790974617, + "learning_rate": 5.396125917263256e-06, + "loss": 1.0204, + "step": 14011 + }, + { + "epoch": 0.8513275411628896, + "grad_norm": 0.21836933493614197, + "learning_rate": 5.39180096644879e-06, + "loss": 1.0219, + "step": 14012 + }, + { + "epoch": 0.8513882981955161, + "grad_norm": 0.17805074155330658, + "learning_rate": 5.387477650756489e-06, + "loss": 1.0509, + "step": 14013 + }, + { + "epoch": 0.8514490552281426, + "grad_norm": 0.11848554015159607, + "learning_rate": 5.383155970344806e-06, + "loss": 1.0262, + "step": 14014 + }, + { + "epoch": 0.8515098122607692, + "grad_norm": 0.12909625470638275, + "learning_rate": 5.378835925372161e-06, + "loss": 1.0362, + "step": 14015 + }, + { + "epoch": 0.8515705692933957, + "grad_norm": 0.1285289227962494, + "learning_rate": 5.374517515996919e-06, + "loss": 1.0072, + "step": 14016 + }, + { + "epoch": 0.8516313263260222, + "grad_norm": 0.15662997961044312, + "learning_rate": 5.370200742377357e-06, + "loss": 1.0216, + "step": 14017 + }, + { + "epoch": 0.8516920833586488, + "grad_norm": 0.13419532775878906, + "learning_rate": 5.365885604671722e-06, + "loss": 1.0597, + "step": 14018 + }, + { + "epoch": 0.8517528403912753, + "grad_norm": 0.934135913848877, + "learning_rate": 5.3615721030381815e-06, + "loss": 1.0648, + "step": 14019 + }, + { + "epoch": 0.8518135974239018, + "grad_norm": 0.18177230656147003, + "learning_rate": 5.357260237634826e-06, + "loss": 1.0845, + "step": 14020 + }, + { + "epoch": 0.8518743544565284, + "grad_norm": 0.19884257018566132, + "learning_rate": 5.3529500086197394e-06, + "loss": 1.0633, + "step": 14021 + }, + { + "epoch": 0.8519351114891549, + "grad_norm": 0.37935149669647217, + "learning_rate": 5.348641416150896e-06, + "loss": 1.0931, + "step": 14022 + }, + { + "epoch": 0.8519958685217814, + "grad_norm": 0.5824965238571167, + "learning_rate": 5.34433446038623e-06, + "loss": 1.0499, + "step": 14023 + }, + { + "epoch": 0.852056625554408, + "grad_norm": 0.21819645166397095, + "learning_rate": 5.340029141483621e-06, + "loss": 1.016, + "step": 14024 + }, + { + "epoch": 0.8521173825870344, + "grad_norm": 0.40328043699264526, + "learning_rate": 5.335725459600876e-06, + "loss": 1.1647, + "step": 14025 + }, + { + "epoch": 0.8521781396196609, + "grad_norm": 0.2532118260860443, + "learning_rate": 5.331423414895747e-06, + "loss": 1.12, + "step": 14026 + }, + { + "epoch": 0.8522388966522875, + "grad_norm": 0.6410199999809265, + "learning_rate": 5.327123007525925e-06, + "loss": 1.1991, + "step": 14027 + }, + { + "epoch": 0.852299653684914, + "grad_norm": 0.1411563754081726, + "learning_rate": 5.322824237649043e-06, + "loss": 1.0796, + "step": 14028 + }, + { + "epoch": 0.8523604107175405, + "grad_norm": 0.14017334580421448, + "learning_rate": 5.3185271054226825e-06, + "loss": 1.0271, + "step": 14029 + }, + { + "epoch": 0.8524211677501671, + "grad_norm": 0.5482425689697266, + "learning_rate": 5.314231611004339e-06, + "loss": 1.0074, + "step": 14030 + }, + { + "epoch": 0.8524819247827936, + "grad_norm": 0.2405431717634201, + "learning_rate": 5.309937754551469e-06, + "loss": 1.078, + "step": 14031 + }, + { + "epoch": 0.8525426818154201, + "grad_norm": 0.2137085497379303, + "learning_rate": 5.305645536221476e-06, + "loss": 1.1653, + "step": 14032 + }, + { + "epoch": 0.8526034388480467, + "grad_norm": 0.3224814534187317, + "learning_rate": 5.301354956171695e-06, + "loss": 1.1249, + "step": 14033 + }, + { + "epoch": 0.8526641958806732, + "grad_norm": 0.24360981583595276, + "learning_rate": 5.297066014559382e-06, + "loss": 0.9925, + "step": 14034 + }, + { + "epoch": 0.8527249529132997, + "grad_norm": 0.295287549495697, + "learning_rate": 5.292778711541751e-06, + "loss": 1.0517, + "step": 14035 + }, + { + "epoch": 0.8527857099459263, + "grad_norm": 0.19577530026435852, + "learning_rate": 5.28849304727595e-06, + "loss": 0.994, + "step": 14036 + }, + { + "epoch": 0.8528464669785528, + "grad_norm": 0.18572598695755005, + "learning_rate": 5.284209021919085e-06, + "loss": 1.0817, + "step": 14037 + }, + { + "epoch": 0.8529072240111792, + "grad_norm": 0.16208083927631378, + "learning_rate": 5.279926635628185e-06, + "loss": 0.9953, + "step": 14038 + }, + { + "epoch": 0.8529679810438058, + "grad_norm": 0.16196975111961365, + "learning_rate": 5.275645888560232e-06, + "loss": 1.1051, + "step": 14039 + }, + { + "epoch": 0.8530287380764323, + "grad_norm": 0.15718965232372284, + "learning_rate": 5.271366780872105e-06, + "loss": 1.0527, + "step": 14040 + }, + { + "epoch": 0.8530894951090588, + "grad_norm": 0.1637703776359558, + "learning_rate": 5.267089312720674e-06, + "loss": 1.0116, + "step": 14041 + }, + { + "epoch": 0.8531502521416854, + "grad_norm": 0.275718629360199, + "learning_rate": 5.262813484262735e-06, + "loss": 1.1544, + "step": 14042 + }, + { + "epoch": 0.8532110091743119, + "grad_norm": 0.22239762544631958, + "learning_rate": 5.258539295655019e-06, + "loss": 1.068, + "step": 14043 + }, + { + "epoch": 0.8532717662069385, + "grad_norm": 0.173714742064476, + "learning_rate": 5.254266747054188e-06, + "loss": 1.0666, + "step": 14044 + }, + { + "epoch": 0.853332523239565, + "grad_norm": 0.14588402211666107, + "learning_rate": 5.249995838616861e-06, + "loss": 1.0145, + "step": 14045 + }, + { + "epoch": 0.8533932802721915, + "grad_norm": 0.18313844501972198, + "learning_rate": 5.245726570499587e-06, + "loss": 1.0527, + "step": 14046 + }, + { + "epoch": 0.8534540373048181, + "grad_norm": 0.14554178714752197, + "learning_rate": 5.241458942858857e-06, + "loss": 1.0074, + "step": 14047 + }, + { + "epoch": 0.8535147943374446, + "grad_norm": 0.17254644632339478, + "learning_rate": 5.2371929558511e-06, + "loss": 1.0172, + "step": 14048 + }, + { + "epoch": 0.8535755513700711, + "grad_norm": 0.2271064966917038, + "learning_rate": 5.232928609632692e-06, + "loss": 1.147, + "step": 14049 + }, + { + "epoch": 0.8536363084026977, + "grad_norm": 0.21232888102531433, + "learning_rate": 5.228665904359936e-06, + "loss": 1.1114, + "step": 14050 + }, + { + "epoch": 0.8536970654353241, + "grad_norm": 0.16760732233524323, + "learning_rate": 5.224404840189084e-06, + "loss": 1.0413, + "step": 14051 + }, + { + "epoch": 0.8537578224679506, + "grad_norm": 0.15842761099338531, + "learning_rate": 5.2201454172763175e-06, + "loss": 1.0261, + "step": 14052 + }, + { + "epoch": 0.8538185795005772, + "grad_norm": 0.21494831144809723, + "learning_rate": 5.2158876357777995e-06, + "loss": 1.0875, + "step": 14053 + }, + { + "epoch": 0.8538793365332037, + "grad_norm": 0.19609951972961426, + "learning_rate": 5.211631495849562e-06, + "loss": 1.1049, + "step": 14054 + }, + { + "epoch": 0.8539400935658302, + "grad_norm": 0.21049846708774567, + "learning_rate": 5.2073769976476335e-06, + "loss": 1.0369, + "step": 14055 + }, + { + "epoch": 0.8540008505984568, + "grad_norm": 0.5333290696144104, + "learning_rate": 5.203124141327959e-06, + "loss": 1.1711, + "step": 14056 + }, + { + "epoch": 0.8540616076310833, + "grad_norm": 0.2019663155078888, + "learning_rate": 5.198872927046417e-06, + "loss": 1.0636, + "step": 14057 + }, + { + "epoch": 0.8541223646637098, + "grad_norm": 0.16061629354953766, + "learning_rate": 5.194623354958855e-06, + "loss": 1.0511, + "step": 14058 + }, + { + "epoch": 0.8541831216963364, + "grad_norm": 0.38564613461494446, + "learning_rate": 5.19037542522105e-06, + "loss": 1.1551, + "step": 14059 + }, + { + "epoch": 0.8542438787289629, + "grad_norm": 0.1669529527425766, + "learning_rate": 5.186129137988683e-06, + "loss": 1.0879, + "step": 14060 + }, + { + "epoch": 0.8543046357615894, + "grad_norm": 0.1854977011680603, + "learning_rate": 5.181884493417416e-06, + "loss": 1.0372, + "step": 14061 + }, + { + "epoch": 0.854365392794216, + "grad_norm": 0.5255594253540039, + "learning_rate": 5.177641491662821e-06, + "loss": 1.1982, + "step": 14062 + }, + { + "epoch": 0.8544261498268425, + "grad_norm": 0.1541188508272171, + "learning_rate": 5.173400132880457e-06, + "loss": 1.0644, + "step": 14063 + }, + { + "epoch": 0.8544869068594689, + "grad_norm": 0.14958515763282776, + "learning_rate": 5.169160417225771e-06, + "loss": 1.0409, + "step": 14064 + }, + { + "epoch": 0.8545476638920955, + "grad_norm": 0.14443592727184296, + "learning_rate": 5.1649223448541786e-06, + "loss": 1.0456, + "step": 14065 + }, + { + "epoch": 0.854608420924722, + "grad_norm": 0.3377833664417267, + "learning_rate": 5.160685915921032e-06, + "loss": 1.2471, + "step": 14066 + }, + { + "epoch": 0.8546691779573485, + "grad_norm": 0.1844770461320877, + "learning_rate": 5.156451130581597e-06, + "loss": 1.0585, + "step": 14067 + }, + { + "epoch": 0.8547299349899751, + "grad_norm": 0.14057545363903046, + "learning_rate": 5.152217988991115e-06, + "loss": 1.069, + "step": 14068 + }, + { + "epoch": 0.8547906920226016, + "grad_norm": 0.2722685933113098, + "learning_rate": 5.147986491304757e-06, + "loss": 1.0512, + "step": 14069 + }, + { + "epoch": 0.8548514490552281, + "grad_norm": 0.1453302502632141, + "learning_rate": 5.143756637677621e-06, + "loss": 1.0622, + "step": 14070 + }, + { + "epoch": 0.8549122060878547, + "grad_norm": 0.1505041867494583, + "learning_rate": 5.139528428264756e-06, + "loss": 1.0371, + "step": 14071 + }, + { + "epoch": 0.8549729631204812, + "grad_norm": 0.8098244071006775, + "learning_rate": 5.135301863221154e-06, + "loss": 0.9965, + "step": 14072 + }, + { + "epoch": 0.8550337201531077, + "grad_norm": 0.3148677349090576, + "learning_rate": 5.131076942701729e-06, + "loss": 1.0698, + "step": 14073 + }, + { + "epoch": 0.8550944771857343, + "grad_norm": 0.2173866182565689, + "learning_rate": 5.126853666861347e-06, + "loss": 1.0003, + "step": 14074 + }, + { + "epoch": 0.8551552342183608, + "grad_norm": 0.2378932386636734, + "learning_rate": 5.122632035854824e-06, + "loss": 1.0706, + "step": 14075 + }, + { + "epoch": 0.8552159912509874, + "grad_norm": 0.1560126394033432, + "learning_rate": 5.118412049836896e-06, + "loss": 1.0851, + "step": 14076 + }, + { + "epoch": 0.8552767482836138, + "grad_norm": 0.16882063448429108, + "learning_rate": 5.11419370896225e-06, + "loss": 1.0201, + "step": 14077 + }, + { + "epoch": 0.8553375053162403, + "grad_norm": 0.17609776556491852, + "learning_rate": 5.109977013385503e-06, + "loss": 1.1082, + "step": 14078 + }, + { + "epoch": 0.8553982623488668, + "grad_norm": 0.16709665954113007, + "learning_rate": 5.105761963261236e-06, + "loss": 1.0527, + "step": 14079 + }, + { + "epoch": 0.8554590193814934, + "grad_norm": 0.42474666237831116, + "learning_rate": 5.101548558743946e-06, + "loss": 1.0782, + "step": 14080 + }, + { + "epoch": 0.8555197764141199, + "grad_norm": 0.3871685862541199, + "learning_rate": 5.097336799988067e-06, + "loss": 1.0362, + "step": 14081 + }, + { + "epoch": 0.8555805334467464, + "grad_norm": 0.15563392639160156, + "learning_rate": 5.093126687147986e-06, + "loss": 1.0721, + "step": 14082 + }, + { + "epoch": 0.855641290479373, + "grad_norm": 0.1709534078836441, + "learning_rate": 5.088918220378014e-06, + "loss": 1.0267, + "step": 14083 + }, + { + "epoch": 0.8557020475119995, + "grad_norm": 0.15836656093597412, + "learning_rate": 5.084711399832437e-06, + "loss": 0.9817, + "step": 14084 + }, + { + "epoch": 0.855762804544626, + "grad_norm": 0.17920178174972534, + "learning_rate": 5.080506225665449e-06, + "loss": 1.0876, + "step": 14085 + }, + { + "epoch": 0.8558235615772526, + "grad_norm": 0.14687220752239227, + "learning_rate": 5.076302698031193e-06, + "loss": 1.0237, + "step": 14086 + }, + { + "epoch": 0.8558843186098791, + "grad_norm": 0.1988934874534607, + "learning_rate": 5.072100817083741e-06, + "loss": 1.1069, + "step": 14087 + }, + { + "epoch": 0.8559450756425057, + "grad_norm": 0.15474659204483032, + "learning_rate": 5.067900582977103e-06, + "loss": 1.0824, + "step": 14088 + }, + { + "epoch": 0.8560058326751322, + "grad_norm": 0.15512509644031525, + "learning_rate": 5.063701995865266e-06, + "loss": 1.0413, + "step": 14089 + }, + { + "epoch": 0.8560665897077586, + "grad_norm": 0.3030308783054352, + "learning_rate": 5.0595050559021196e-06, + "loss": 1.0625, + "step": 14090 + }, + { + "epoch": 0.8561273467403852, + "grad_norm": 0.17232871055603027, + "learning_rate": 5.055309763241501e-06, + "loss": 0.9881, + "step": 14091 + }, + { + "epoch": 0.8561881037730117, + "grad_norm": 0.11595418304204941, + "learning_rate": 5.0511161180371916e-06, + "loss": 1.0179, + "step": 14092 + }, + { + "epoch": 0.8562488608056382, + "grad_norm": 1.2898151874542236, + "learning_rate": 5.046924120442903e-06, + "loss": 1.118, + "step": 14093 + }, + { + "epoch": 0.8563096178382648, + "grad_norm": 0.18346980214118958, + "learning_rate": 5.042733770612307e-06, + "loss": 1.1157, + "step": 14094 + }, + { + "epoch": 0.8563703748708913, + "grad_norm": 0.1763455718755722, + "learning_rate": 5.038545068698991e-06, + "loss": 1.0263, + "step": 14095 + }, + { + "epoch": 0.8564311319035178, + "grad_norm": 0.18220920860767365, + "learning_rate": 5.034358014856499e-06, + "loss": 1.1212, + "step": 14096 + }, + { + "epoch": 0.8564918889361444, + "grad_norm": 0.45680078864097595, + "learning_rate": 5.030172609238304e-06, + "loss": 1.1165, + "step": 14097 + }, + { + "epoch": 0.8565526459687709, + "grad_norm": 1.0246318578720093, + "learning_rate": 5.025988851997826e-06, + "loss": 1.0172, + "step": 14098 + }, + { + "epoch": 0.8566134030013974, + "grad_norm": 0.1421239674091339, + "learning_rate": 5.02180674328841e-06, + "loss": 1.0446, + "step": 14099 + }, + { + "epoch": 0.856674160034024, + "grad_norm": 0.21484291553497314, + "learning_rate": 5.017626283263382e-06, + "loss": 1.1125, + "step": 14100 + }, + { + "epoch": 0.8567349170666505, + "grad_norm": 0.13472294807434082, + "learning_rate": 5.01344747207595e-06, + "loss": 1.08, + "step": 14101 + }, + { + "epoch": 0.856795674099277, + "grad_norm": 0.19898401200771332, + "learning_rate": 5.009270309879288e-06, + "loss": 1.014, + "step": 14102 + }, + { + "epoch": 0.8568564311319036, + "grad_norm": 0.1757878214120865, + "learning_rate": 5.005094796826526e-06, + "loss": 1.0828, + "step": 14103 + }, + { + "epoch": 0.85691718816453, + "grad_norm": 0.2288830727338791, + "learning_rate": 5.000920933070702e-06, + "loss": 1.0875, + "step": 14104 + }, + { + "epoch": 0.8569779451971565, + "grad_norm": 0.3268856406211853, + "learning_rate": 4.9967487187648266e-06, + "loss": 1.2686, + "step": 14105 + }, + { + "epoch": 0.8570387022297831, + "grad_norm": 0.3482927083969116, + "learning_rate": 4.992578154061839e-06, + "loss": 0.9995, + "step": 14106 + }, + { + "epoch": 0.8570994592624096, + "grad_norm": 0.19687116146087646, + "learning_rate": 4.988409239114583e-06, + "loss": 1.0215, + "step": 14107 + }, + { + "epoch": 0.8571602162950361, + "grad_norm": 0.2368345856666565, + "learning_rate": 4.984241974075882e-06, + "loss": 1.1544, + "step": 14108 + }, + { + "epoch": 0.8572209733276627, + "grad_norm": 0.17905186116695404, + "learning_rate": 4.980076359098501e-06, + "loss": 1.0494, + "step": 14109 + }, + { + "epoch": 0.8572817303602892, + "grad_norm": 0.1750558316707611, + "learning_rate": 4.975912394335125e-06, + "loss": 1.0804, + "step": 14110 + }, + { + "epoch": 0.8573424873929157, + "grad_norm": 0.2234957367181778, + "learning_rate": 4.971750079938381e-06, + "loss": 1.2517, + "step": 14111 + }, + { + "epoch": 0.8574032444255423, + "grad_norm": 0.17120343446731567, + "learning_rate": 4.967589416060842e-06, + "loss": 1.0317, + "step": 14112 + }, + { + "epoch": 0.8574640014581688, + "grad_norm": 0.11625444889068604, + "learning_rate": 4.9634304028550185e-06, + "loss": 1.0193, + "step": 14113 + }, + { + "epoch": 0.8575247584907953, + "grad_norm": 0.21900813281536102, + "learning_rate": 4.959273040473356e-06, + "loss": 1.1061, + "step": 14114 + }, + { + "epoch": 0.8575855155234219, + "grad_norm": 0.20168660581111908, + "learning_rate": 4.955117329068249e-06, + "loss": 1.1325, + "step": 14115 + }, + { + "epoch": 0.8576462725560484, + "grad_norm": 0.16049495339393616, + "learning_rate": 4.950963268792025e-06, + "loss": 1.127, + "step": 14116 + }, + { + "epoch": 0.8577070295886748, + "grad_norm": 0.16064921021461487, + "learning_rate": 4.9468108597969465e-06, + "loss": 1.0862, + "step": 14117 + }, + { + "epoch": 0.8577677866213014, + "grad_norm": 0.17822140455245972, + "learning_rate": 4.942660102235225e-06, + "loss": 1.0985, + "step": 14118 + }, + { + "epoch": 0.8578285436539279, + "grad_norm": 0.2689385712146759, + "learning_rate": 4.938510996258999e-06, + "loss": 1.081, + "step": 14119 + }, + { + "epoch": 0.8578893006865544, + "grad_norm": 0.13704459369182587, + "learning_rate": 4.934363542020381e-06, + "loss": 0.9976, + "step": 14120 + }, + { + "epoch": 0.857950057719181, + "grad_norm": 0.19898618757724762, + "learning_rate": 4.930217739671367e-06, + "loss": 1.1488, + "step": 14121 + }, + { + "epoch": 0.8580108147518075, + "grad_norm": 0.1647258996963501, + "learning_rate": 4.926073589363939e-06, + "loss": 1.0624, + "step": 14122 + }, + { + "epoch": 0.858071571784434, + "grad_norm": 0.6852223873138428, + "learning_rate": 4.921931091249993e-06, + "loss": 1.1155, + "step": 14123 + }, + { + "epoch": 0.8581323288170606, + "grad_norm": 0.1873166412115097, + "learning_rate": 4.917790245481363e-06, + "loss": 1.0263, + "step": 14124 + }, + { + "epoch": 0.8581930858496871, + "grad_norm": 0.1626402884721756, + "learning_rate": 4.913651052209861e-06, + "loss": 1.1047, + "step": 14125 + }, + { + "epoch": 0.8582538428823137, + "grad_norm": 0.10770536959171295, + "learning_rate": 4.909513511587199e-06, + "loss": 1.0017, + "step": 14126 + }, + { + "epoch": 0.8583145999149402, + "grad_norm": 0.19265878200531006, + "learning_rate": 4.905377623765029e-06, + "loss": 1.1107, + "step": 14127 + }, + { + "epoch": 0.8583753569475667, + "grad_norm": 0.4001806378364563, + "learning_rate": 4.901243388894961e-06, + "loss": 1.0973, + "step": 14128 + }, + { + "epoch": 0.8584361139801933, + "grad_norm": 0.19757820665836334, + "learning_rate": 4.897110807128519e-06, + "loss": 1.0669, + "step": 14129 + }, + { + "epoch": 0.8584968710128197, + "grad_norm": 0.24368375539779663, + "learning_rate": 4.89297987861721e-06, + "loss": 1.0638, + "step": 14130 + }, + { + "epoch": 0.8585576280454462, + "grad_norm": 0.169841930270195, + "learning_rate": 4.888850603512446e-06, + "loss": 1.0614, + "step": 14131 + }, + { + "epoch": 0.8586183850780728, + "grad_norm": 0.5794231295585632, + "learning_rate": 4.884722981965578e-06, + "loss": 1.0608, + "step": 14132 + }, + { + "epoch": 0.8586791421106993, + "grad_norm": 0.1510799378156662, + "learning_rate": 4.880597014127925e-06, + "loss": 1.0304, + "step": 14133 + }, + { + "epoch": 0.8587398991433258, + "grad_norm": 0.1357523649930954, + "learning_rate": 4.876472700150686e-06, + "loss": 1.0372, + "step": 14134 + }, + { + "epoch": 0.8588006561759524, + "grad_norm": 19.653453826904297, + "learning_rate": 4.87235004018507e-06, + "loss": 1.0823, + "step": 14135 + }, + { + "epoch": 0.8588614132085789, + "grad_norm": 3.774996757507324, + "learning_rate": 4.8682290343821875e-06, + "loss": 1.2738, + "step": 14136 + }, + { + "epoch": 0.8589221702412054, + "grad_norm": 0.12442640215158463, + "learning_rate": 4.864109682893098e-06, + "loss": 1.0246, + "step": 14137 + }, + { + "epoch": 0.858982927273832, + "grad_norm": 0.25545626878738403, + "learning_rate": 4.859991985868789e-06, + "loss": 1.186, + "step": 14138 + }, + { + "epoch": 0.8590436843064585, + "grad_norm": 0.30023759603500366, + "learning_rate": 4.855875943460203e-06, + "loss": 1.1279, + "step": 14139 + }, + { + "epoch": 0.859104441339085, + "grad_norm": 0.19071625173091888, + "learning_rate": 4.851761555818207e-06, + "loss": 1.1191, + "step": 14140 + }, + { + "epoch": 0.8591651983717116, + "grad_norm": 0.33429837226867676, + "learning_rate": 4.847648823093615e-06, + "loss": 1.1226, + "step": 14141 + }, + { + "epoch": 0.8592259554043381, + "grad_norm": 0.11961117386817932, + "learning_rate": 4.843537745437188e-06, + "loss": 0.9773, + "step": 14142 + }, + { + "epoch": 0.8592867124369645, + "grad_norm": 0.2013082653284073, + "learning_rate": 4.839428322999612e-06, + "loss": 1.1242, + "step": 14143 + }, + { + "epoch": 0.8593474694695911, + "grad_norm": 0.19582994282245636, + "learning_rate": 4.835320555931522e-06, + "loss": 1.0343, + "step": 14144 + }, + { + "epoch": 0.8594082265022176, + "grad_norm": 0.24116584658622742, + "learning_rate": 4.831214444383475e-06, + "loss": 1.0211, + "step": 14145 + }, + { + "epoch": 0.8594689835348441, + "grad_norm": 0.20090162754058838, + "learning_rate": 4.8271099885060125e-06, + "loss": 1.0891, + "step": 14146 + }, + { + "epoch": 0.8595297405674707, + "grad_norm": 0.13687479496002197, + "learning_rate": 4.823007188449557e-06, + "loss": 1.0029, + "step": 14147 + }, + { + "epoch": 0.8595904976000972, + "grad_norm": 0.6746511459350586, + "learning_rate": 4.818906044364507e-06, + "loss": 1.2412, + "step": 14148 + }, + { + "epoch": 0.8596512546327237, + "grad_norm": 0.2103484869003296, + "learning_rate": 4.814806556401186e-06, + "loss": 1.1112, + "step": 14149 + }, + { + "epoch": 0.8597120116653503, + "grad_norm": 0.12649881839752197, + "learning_rate": 4.810708724709856e-06, + "loss": 1.0218, + "step": 14150 + }, + { + "epoch": 0.8597727686979768, + "grad_norm": 0.11957068741321564, + "learning_rate": 4.806612549440742e-06, + "loss": 1.0046, + "step": 14151 + }, + { + "epoch": 0.8598335257306033, + "grad_norm": 1.204906940460205, + "learning_rate": 4.802518030743985e-06, + "loss": 1.0459, + "step": 14152 + }, + { + "epoch": 0.8598942827632299, + "grad_norm": 16.409915924072266, + "learning_rate": 4.798425168769671e-06, + "loss": 1.0573, + "step": 14153 + }, + { + "epoch": 0.8599550397958564, + "grad_norm": 0.24177242815494537, + "learning_rate": 4.794333963667813e-06, + "loss": 1.1715, + "step": 14154 + }, + { + "epoch": 0.860015796828483, + "grad_norm": 0.20045921206474304, + "learning_rate": 4.790244415588369e-06, + "loss": 1.0445, + "step": 14155 + }, + { + "epoch": 0.8600765538611094, + "grad_norm": 0.2165038138628006, + "learning_rate": 4.786156524681268e-06, + "loss": 1.2048, + "step": 14156 + }, + { + "epoch": 0.8601373108937359, + "grad_norm": 0.17372946441173553, + "learning_rate": 4.782070291096336e-06, + "loss": 1.0828, + "step": 14157 + }, + { + "epoch": 0.8601980679263624, + "grad_norm": 0.12380470335483551, + "learning_rate": 4.777985714983357e-06, + "loss": 1.021, + "step": 14158 + }, + { + "epoch": 0.860258824958989, + "grad_norm": 0.1486104428768158, + "learning_rate": 4.773902796492052e-06, + "loss": 1.0458, + "step": 14159 + }, + { + "epoch": 0.8603195819916155, + "grad_norm": 0.2402789145708084, + "learning_rate": 4.7698215357720895e-06, + "loss": 1.0655, + "step": 14160 + }, + { + "epoch": 0.860380339024242, + "grad_norm": 0.33339035511016846, + "learning_rate": 4.765741932973055e-06, + "loss": 1.1307, + "step": 14161 + }, + { + "epoch": 0.8604410960568686, + "grad_norm": 0.12551140785217285, + "learning_rate": 4.761663988244497e-06, + "loss": 1.0077, + "step": 14162 + }, + { + "epoch": 0.8605018530894951, + "grad_norm": 0.1715989112854004, + "learning_rate": 4.757587701735883e-06, + "loss": 1.0519, + "step": 14163 + }, + { + "epoch": 0.8605626101221217, + "grad_norm": 0.32183539867401123, + "learning_rate": 4.753513073596644e-06, + "loss": 1.0927, + "step": 14164 + }, + { + "epoch": 0.8606233671547482, + "grad_norm": 0.14816135168075562, + "learning_rate": 4.749440103976127e-06, + "loss": 1.0483, + "step": 14165 + }, + { + "epoch": 0.8606841241873747, + "grad_norm": 0.5146399140357971, + "learning_rate": 4.745368793023619e-06, + "loss": 1.2164, + "step": 14166 + }, + { + "epoch": 0.8607448812200013, + "grad_norm": 0.12622416019439697, + "learning_rate": 4.741299140888389e-06, + "loss": 1.0152, + "step": 14167 + }, + { + "epoch": 0.8608056382526278, + "grad_norm": 0.1136050745844841, + "learning_rate": 4.737231147719573e-06, + "loss": 1.0141, + "step": 14168 + }, + { + "epoch": 0.8608663952852542, + "grad_norm": 0.152277871966362, + "learning_rate": 4.733164813666302e-06, + "loss": 1.0436, + "step": 14169 + }, + { + "epoch": 0.8609271523178808, + "grad_norm": 0.1453281044960022, + "learning_rate": 4.729100138877623e-06, + "loss": 1.0096, + "step": 14170 + }, + { + "epoch": 0.8609879093505073, + "grad_norm": 11.08531379699707, + "learning_rate": 4.725037123502518e-06, + "loss": 1.0651, + "step": 14171 + }, + { + "epoch": 0.8610486663831338, + "grad_norm": 0.13586102426052094, + "learning_rate": 4.720975767689939e-06, + "loss": 1.0427, + "step": 14172 + }, + { + "epoch": 0.8611094234157604, + "grad_norm": 0.16124646365642548, + "learning_rate": 4.7169160715887566e-06, + "loss": 1.0724, + "step": 14173 + }, + { + "epoch": 0.8611701804483869, + "grad_norm": 0.22187605500221252, + "learning_rate": 4.712858035347756e-06, + "loss": 1.0902, + "step": 14174 + }, + { + "epoch": 0.8612309374810134, + "grad_norm": 0.22663149237632751, + "learning_rate": 4.708801659115702e-06, + "loss": 0.9782, + "step": 14175 + }, + { + "epoch": 0.86129169451364, + "grad_norm": 0.1696464568376541, + "learning_rate": 4.704746943041266e-06, + "loss": 1.0525, + "step": 14176 + }, + { + "epoch": 0.8613524515462665, + "grad_norm": 0.12996548414230347, + "learning_rate": 4.7006938872730945e-06, + "loss": 1.0753, + "step": 14177 + }, + { + "epoch": 0.861413208578893, + "grad_norm": 0.12015622109174728, + "learning_rate": 4.696642491959746e-06, + "loss": 1.0135, + "step": 14178 + }, + { + "epoch": 0.8614739656115196, + "grad_norm": 0.20397773385047913, + "learning_rate": 4.692592757249725e-06, + "loss": 1.0669, + "step": 14179 + }, + { + "epoch": 0.8615347226441461, + "grad_norm": 0.17050763964653015, + "learning_rate": 4.688544683291479e-06, + "loss": 1.1081, + "step": 14180 + }, + { + "epoch": 0.8615954796767726, + "grad_norm": 0.23892568051815033, + "learning_rate": 4.6844982702333725e-06, + "loss": 1.0249, + "step": 14181 + }, + { + "epoch": 0.8616562367093991, + "grad_norm": 0.14141437411308289, + "learning_rate": 4.6804535182237485e-06, + "loss": 1.0048, + "step": 14182 + }, + { + "epoch": 0.8617169937420256, + "grad_norm": 0.21362362802028656, + "learning_rate": 4.67641042741086e-06, + "loss": 1.1164, + "step": 14183 + }, + { + "epoch": 0.8617777507746521, + "grad_norm": 1.8475956916809082, + "learning_rate": 4.672368997942911e-06, + "loss": 1.1871, + "step": 14184 + }, + { + "epoch": 0.8618385078072787, + "grad_norm": 0.2093481421470642, + "learning_rate": 4.668329229968038e-06, + "loss": 1.0338, + "step": 14185 + }, + { + "epoch": 0.8618992648399052, + "grad_norm": 0.17458310723304749, + "learning_rate": 4.664291123634318e-06, + "loss": 1.0979, + "step": 14186 + }, + { + "epoch": 0.8619600218725317, + "grad_norm": 0.1326577514410019, + "learning_rate": 4.660254679089771e-06, + "loss": 1.044, + "step": 14187 + }, + { + "epoch": 0.8620207789051583, + "grad_norm": 0.12502983212471008, + "learning_rate": 4.65621989648235e-06, + "loss": 1.0099, + "step": 14188 + }, + { + "epoch": 0.8620815359377848, + "grad_norm": 0.1846161037683487, + "learning_rate": 4.652186775959954e-06, + "loss": 1.1069, + "step": 14189 + }, + { + "epoch": 0.8621422929704113, + "grad_norm": 1.5092058181762695, + "learning_rate": 4.648155317670422e-06, + "loss": 1.1222, + "step": 14190 + }, + { + "epoch": 0.8622030500030379, + "grad_norm": 0.18340329825878143, + "learning_rate": 4.644125521761516e-06, + "loss": 1.0259, + "step": 14191 + }, + { + "epoch": 0.8622638070356644, + "grad_norm": 0.13462935388088226, + "learning_rate": 4.640097388380954e-06, + "loss": 0.9688, + "step": 14192 + }, + { + "epoch": 0.8623245640682909, + "grad_norm": 0.1226174607872963, + "learning_rate": 4.636070917676405e-06, + "loss": 1.0447, + "step": 14193 + }, + { + "epoch": 0.8623853211009175, + "grad_norm": 0.12799270451068878, + "learning_rate": 4.632046109795435e-06, + "loss": 1.0039, + "step": 14194 + }, + { + "epoch": 0.8624460781335439, + "grad_norm": 0.2120084911584854, + "learning_rate": 4.628022964885587e-06, + "loss": 1.1573, + "step": 14195 + }, + { + "epoch": 0.8625068351661704, + "grad_norm": 0.22381483018398285, + "learning_rate": 4.62400148309432e-06, + "loss": 1.1666, + "step": 14196 + }, + { + "epoch": 0.862567592198797, + "grad_norm": 0.3212534487247467, + "learning_rate": 4.6199816645690505e-06, + "loss": 1.0364, + "step": 14197 + }, + { + "epoch": 0.8626283492314235, + "grad_norm": 2.2303309440612793, + "learning_rate": 4.615963509457128e-06, + "loss": 1.1159, + "step": 14198 + }, + { + "epoch": 0.86268910626405, + "grad_norm": 0.1360185444355011, + "learning_rate": 4.611947017905838e-06, + "loss": 1.0145, + "step": 14199 + }, + { + "epoch": 0.8627498632966766, + "grad_norm": 0.3206065595149994, + "learning_rate": 4.607932190062408e-06, + "loss": 1.1749, + "step": 14200 + }, + { + "epoch": 0.8628106203293031, + "grad_norm": 0.33713075518608093, + "learning_rate": 4.603919026073989e-06, + "loss": 1.096, + "step": 14201 + }, + { + "epoch": 0.8628713773619296, + "grad_norm": 0.15048936009407043, + "learning_rate": 4.5999075260876935e-06, + "loss": 1.0698, + "step": 14202 + }, + { + "epoch": 0.8629321343945562, + "grad_norm": 0.16683512926101685, + "learning_rate": 4.595897690250567e-06, + "loss": 1.0136, + "step": 14203 + }, + { + "epoch": 0.8629928914271827, + "grad_norm": 0.21444061398506165, + "learning_rate": 4.591889518709585e-06, + "loss": 1.2271, + "step": 14204 + }, + { + "epoch": 0.8630536484598093, + "grad_norm": 0.14498701691627502, + "learning_rate": 4.587883011611671e-06, + "loss": 1.0022, + "step": 14205 + }, + { + "epoch": 0.8631144054924358, + "grad_norm": 0.7216893434524536, + "learning_rate": 4.583878169103684e-06, + "loss": 1.0553, + "step": 14206 + }, + { + "epoch": 0.8631751625250623, + "grad_norm": 0.23811201751232147, + "learning_rate": 4.579874991332423e-06, + "loss": 1.1133, + "step": 14207 + }, + { + "epoch": 0.8632359195576889, + "grad_norm": 0.1916704922914505, + "learning_rate": 4.57587347844462e-06, + "loss": 1.03, + "step": 14208 + }, + { + "epoch": 0.8632966765903153, + "grad_norm": 0.22513344883918762, + "learning_rate": 4.571873630586954e-06, + "loss": 1.176, + "step": 14209 + }, + { + "epoch": 0.8633574336229418, + "grad_norm": 0.2513387203216553, + "learning_rate": 4.567875447906045e-06, + "loss": 1.0341, + "step": 14210 + }, + { + "epoch": 0.8634181906555684, + "grad_norm": 0.5784952640533447, + "learning_rate": 4.563878930548443e-06, + "loss": 1.095, + "step": 14211 + }, + { + "epoch": 0.8634789476881949, + "grad_norm": 3.994819402694702, + "learning_rate": 4.559884078660626e-06, + "loss": 1.1403, + "step": 14212 + }, + { + "epoch": 0.8635397047208214, + "grad_norm": 0.11445577442646027, + "learning_rate": 4.55589089238907e-06, + "loss": 0.9924, + "step": 14213 + }, + { + "epoch": 0.863600461753448, + "grad_norm": 0.11830813437700272, + "learning_rate": 4.551899371880103e-06, + "loss": 1.0261, + "step": 14214 + }, + { + "epoch": 0.8636612187860745, + "grad_norm": 0.12039349228143692, + "learning_rate": 4.547909517280047e-06, + "loss": 1.0398, + "step": 14215 + }, + { + "epoch": 0.863721975818701, + "grad_norm": 0.1437261402606964, + "learning_rate": 4.543921328735157e-06, + "loss": 1.0458, + "step": 14216 + }, + { + "epoch": 0.8637827328513276, + "grad_norm": 0.13213832676410675, + "learning_rate": 4.539934806391605e-06, + "loss": 0.9966, + "step": 14217 + }, + { + "epoch": 0.8638434898839541, + "grad_norm": 0.12144818902015686, + "learning_rate": 4.535949950395541e-06, + "loss": 1.044, + "step": 14218 + }, + { + "epoch": 0.8639042469165806, + "grad_norm": 0.12828093767166138, + "learning_rate": 4.53196676089302e-06, + "loss": 1.0408, + "step": 14219 + }, + { + "epoch": 0.8639650039492072, + "grad_norm": 0.18242663145065308, + "learning_rate": 4.5279852380300524e-06, + "loss": 1.0503, + "step": 14220 + }, + { + "epoch": 0.8640257609818337, + "grad_norm": 0.2777208387851715, + "learning_rate": 4.5240053819525726e-06, + "loss": 1.0532, + "step": 14221 + }, + { + "epoch": 0.8640865180144601, + "grad_norm": 0.14470389485359192, + "learning_rate": 4.520027192806453e-06, + "loss": 1.1006, + "step": 14222 + }, + { + "epoch": 0.8641472750470867, + "grad_norm": 0.18927332758903503, + "learning_rate": 4.516050670737537e-06, + "loss": 1.0486, + "step": 14223 + }, + { + "epoch": 0.8642080320797132, + "grad_norm": 0.18742652237415314, + "learning_rate": 4.5120758158915755e-06, + "loss": 1.0685, + "step": 14224 + }, + { + "epoch": 0.8642687891123397, + "grad_norm": 0.2889087498188019, + "learning_rate": 4.5081026284142625e-06, + "loss": 1.0519, + "step": 14225 + }, + { + "epoch": 0.8643295461449663, + "grad_norm": 0.13507655262947083, + "learning_rate": 4.504131108451248e-06, + "loss": 1.0357, + "step": 14226 + }, + { + "epoch": 0.8643903031775928, + "grad_norm": 0.20007358491420746, + "learning_rate": 4.5001612561481e-06, + "loss": 1.0209, + "step": 14227 + }, + { + "epoch": 0.8644510602102193, + "grad_norm": 0.17806778848171234, + "learning_rate": 4.496193071650334e-06, + "loss": 1.073, + "step": 14228 + }, + { + "epoch": 0.8645118172428459, + "grad_norm": 0.548748254776001, + "learning_rate": 4.492226555103407e-06, + "loss": 1.0572, + "step": 14229 + }, + { + "epoch": 0.8645725742754724, + "grad_norm": 0.153289332985878, + "learning_rate": 4.488261706652713e-06, + "loss": 1.0327, + "step": 14230 + }, + { + "epoch": 0.8646333313080989, + "grad_norm": 0.22419624030590057, + "learning_rate": 4.4842985264435864e-06, + "loss": 1.1498, + "step": 14231 + }, + { + "epoch": 0.8646940883407255, + "grad_norm": 0.2473427802324295, + "learning_rate": 4.480337014621294e-06, + "loss": 1.1119, + "step": 14232 + }, + { + "epoch": 0.864754845373352, + "grad_norm": 0.14901793003082275, + "learning_rate": 4.476377171331037e-06, + "loss": 1.0498, + "step": 14233 + }, + { + "epoch": 0.8648156024059785, + "grad_norm": 0.19615314900875092, + "learning_rate": 4.472418996717992e-06, + "loss": 1.0767, + "step": 14234 + }, + { + "epoch": 0.864876359438605, + "grad_norm": 0.18956589698791504, + "learning_rate": 4.468462490927216e-06, + "loss": 1.0912, + "step": 14235 + }, + { + "epoch": 0.8649371164712315, + "grad_norm": 0.16698569059371948, + "learning_rate": 4.464507654103756e-06, + "loss": 1.0223, + "step": 14236 + }, + { + "epoch": 0.864997873503858, + "grad_norm": 0.12168578803539276, + "learning_rate": 4.4605544863925654e-06, + "loss": 0.998, + "step": 14237 + }, + { + "epoch": 0.8650586305364846, + "grad_norm": 0.31648388504981995, + "learning_rate": 4.456602987938546e-06, + "loss": 1.1642, + "step": 14238 + }, + { + "epoch": 0.8651193875691111, + "grad_norm": 0.14555762708187103, + "learning_rate": 4.452653158886555e-06, + "loss": 1.0751, + "step": 14239 + }, + { + "epoch": 0.8651801446017376, + "grad_norm": 0.1772354245185852, + "learning_rate": 4.448704999381376e-06, + "loss": 1.0287, + "step": 14240 + }, + { + "epoch": 0.8652409016343642, + "grad_norm": 0.13597247004508972, + "learning_rate": 4.444758509567709e-06, + "loss": 1.102, + "step": 14241 + }, + { + "epoch": 0.8653016586669907, + "grad_norm": 0.25285211205482483, + "learning_rate": 4.4408136895902286e-06, + "loss": 1.0817, + "step": 14242 + }, + { + "epoch": 0.8653624156996172, + "grad_norm": 0.13240931928157806, + "learning_rate": 4.436870539593513e-06, + "loss": 1.0507, + "step": 14243 + }, + { + "epoch": 0.8654231727322438, + "grad_norm": 2.168606996536255, + "learning_rate": 4.4329290597221295e-06, + "loss": 1.1227, + "step": 14244 + }, + { + "epoch": 0.8654839297648703, + "grad_norm": 0.2772831320762634, + "learning_rate": 4.428989250120541e-06, + "loss": 1.2449, + "step": 14245 + }, + { + "epoch": 0.8655446867974969, + "grad_norm": 0.1646389663219452, + "learning_rate": 4.4250511109331534e-06, + "loss": 1.0684, + "step": 14246 + }, + { + "epoch": 0.8656054438301234, + "grad_norm": 0.3402803838253021, + "learning_rate": 4.421114642304342e-06, + "loss": 1.0697, + "step": 14247 + }, + { + "epoch": 0.8656662008627498, + "grad_norm": 0.20507340133190155, + "learning_rate": 4.417179844378361e-06, + "loss": 1.1473, + "step": 14248 + }, + { + "epoch": 0.8657269578953763, + "grad_norm": 0.15027055144309998, + "learning_rate": 4.413246717299474e-06, + "loss": 1.0046, + "step": 14249 + }, + { + "epoch": 0.8657877149280029, + "grad_norm": 0.13029612600803375, + "learning_rate": 4.4093152612118434e-06, + "loss": 1.0106, + "step": 14250 + }, + { + "epoch": 0.8658484719606294, + "grad_norm": 0.21464803814888, + "learning_rate": 4.405385476259571e-06, + "loss": 1.0109, + "step": 14251 + }, + { + "epoch": 0.865909228993256, + "grad_norm": 0.11926063150167465, + "learning_rate": 4.401457362586709e-06, + "loss": 1.0437, + "step": 14252 + }, + { + "epoch": 0.8659699860258825, + "grad_norm": 0.23871032893657684, + "learning_rate": 4.3975309203372396e-06, + "loss": 1.2118, + "step": 14253 + }, + { + "epoch": 0.866030743058509, + "grad_norm": 0.22356146574020386, + "learning_rate": 4.393606149655094e-06, + "loss": 0.9886, + "step": 14254 + }, + { + "epoch": 0.8660915000911356, + "grad_norm": 0.4629821479320526, + "learning_rate": 4.389683050684124e-06, + "loss": 1.0735, + "step": 14255 + }, + { + "epoch": 0.8661522571237621, + "grad_norm": 0.19734832644462585, + "learning_rate": 4.385761623568141e-06, + "loss": 1.2451, + "step": 14256 + }, + { + "epoch": 0.8662130141563886, + "grad_norm": 0.27320554852485657, + "learning_rate": 4.381841868450881e-06, + "loss": 1.0252, + "step": 14257 + }, + { + "epoch": 0.8662737711890152, + "grad_norm": 0.15424412488937378, + "learning_rate": 4.377923785476029e-06, + "loss": 1.0407, + "step": 14258 + }, + { + "epoch": 0.8663345282216417, + "grad_norm": 3.654623031616211, + "learning_rate": 4.3740073747871865e-06, + "loss": 1.0679, + "step": 14259 + }, + { + "epoch": 0.8663952852542682, + "grad_norm": 0.26374194025993347, + "learning_rate": 4.3700926365279445e-06, + "loss": 1.1504, + "step": 14260 + }, + { + "epoch": 0.8664560422868947, + "grad_norm": 0.13081814348697662, + "learning_rate": 4.366179570841766e-06, + "loss": 0.9815, + "step": 14261 + }, + { + "epoch": 0.8665167993195212, + "grad_norm": 0.173008531332016, + "learning_rate": 4.362268177872097e-06, + "loss": 1.0153, + "step": 14262 + }, + { + "epoch": 0.8665775563521477, + "grad_norm": 0.19696827232837677, + "learning_rate": 4.358358457762307e-06, + "loss": 0.992, + "step": 14263 + }, + { + "epoch": 0.8666383133847743, + "grad_norm": 9.755534172058105, + "learning_rate": 4.3544504106557025e-06, + "loss": 1.1945, + "step": 14264 + }, + { + "epoch": 0.8666990704174008, + "grad_norm": 0.697206437587738, + "learning_rate": 4.350544036695553e-06, + "loss": 1.1848, + "step": 14265 + }, + { + "epoch": 0.8667598274500273, + "grad_norm": 0.18340778350830078, + "learning_rate": 4.346639336025033e-06, + "loss": 1.0514, + "step": 14266 + }, + { + "epoch": 0.8668205844826539, + "grad_norm": 0.31481602787971497, + "learning_rate": 4.3427363087872815e-06, + "loss": 1.1426, + "step": 14267 + }, + { + "epoch": 0.8668813415152804, + "grad_norm": 0.1680772751569748, + "learning_rate": 4.338834955125343e-06, + "loss": 1.0525, + "step": 14268 + }, + { + "epoch": 0.8669420985479069, + "grad_norm": 0.17089292407035828, + "learning_rate": 4.334935275182234e-06, + "loss": 1.0594, + "step": 14269 + }, + { + "epoch": 0.8670028555805335, + "grad_norm": 0.18876045942306519, + "learning_rate": 4.331037269100901e-06, + "loss": 1.2044, + "step": 14270 + }, + { + "epoch": 0.86706361261316, + "grad_norm": 9.181546211242676, + "learning_rate": 4.327140937024232e-06, + "loss": 1.0792, + "step": 14271 + }, + { + "epoch": 0.8671243696457865, + "grad_norm": 0.18766233325004578, + "learning_rate": 4.323246279095033e-06, + "loss": 1.0715, + "step": 14272 + }, + { + "epoch": 0.8671851266784131, + "grad_norm": 0.27753812074661255, + "learning_rate": 4.319353295456074e-06, + "loss": 1.1098, + "step": 14273 + }, + { + "epoch": 0.8672458837110395, + "grad_norm": 0.18600651621818542, + "learning_rate": 4.315461986250052e-06, + "loss": 1.0502, + "step": 14274 + }, + { + "epoch": 0.867306640743666, + "grad_norm": 0.2704014182090759, + "learning_rate": 4.311572351619603e-06, + "loss": 1.048, + "step": 14275 + }, + { + "epoch": 0.8673673977762926, + "grad_norm": 0.2773779034614563, + "learning_rate": 4.3076843917072955e-06, + "loss": 1.1946, + "step": 14276 + }, + { + "epoch": 0.8674281548089191, + "grad_norm": 0.2364697903394699, + "learning_rate": 4.303798106655649e-06, + "loss": 1.0898, + "step": 14277 + }, + { + "epoch": 0.8674889118415456, + "grad_norm": 0.13476276397705078, + "learning_rate": 4.299913496607122e-06, + "loss": 1.0178, + "step": 14278 + }, + { + "epoch": 0.8675496688741722, + "grad_norm": 0.25110137462615967, + "learning_rate": 4.2960305617040955e-06, + "loss": 1.1914, + "step": 14279 + }, + { + "epoch": 0.8676104259067987, + "grad_norm": 0.260160893201828, + "learning_rate": 4.292149302088899e-06, + "loss": 1.0265, + "step": 14280 + }, + { + "epoch": 0.8676711829394252, + "grad_norm": 0.15583425760269165, + "learning_rate": 4.288269717903809e-06, + "loss": 0.9965, + "step": 14281 + }, + { + "epoch": 0.8677319399720518, + "grad_norm": 0.1341230720281601, + "learning_rate": 4.284391809291027e-06, + "loss": 1.0543, + "step": 14282 + }, + { + "epoch": 0.8677926970046783, + "grad_norm": 0.1910301297903061, + "learning_rate": 4.280515576392702e-06, + "loss": 1.0647, + "step": 14283 + }, + { + "epoch": 0.8678534540373049, + "grad_norm": 0.7497166395187378, + "learning_rate": 4.2766410193509136e-06, + "loss": 1.1618, + "step": 14284 + }, + { + "epoch": 0.8679142110699314, + "grad_norm": 0.14036765694618225, + "learning_rate": 4.2727681383076816e-06, + "loss": 1.059, + "step": 14285 + }, + { + "epoch": 0.8679749681025579, + "grad_norm": 0.30828654766082764, + "learning_rate": 4.268896933404981e-06, + "loss": 1.0976, + "step": 14286 + }, + { + "epoch": 0.8680357251351843, + "grad_norm": 0.2218739539384842, + "learning_rate": 4.265027404784705e-06, + "loss": 1.0008, + "step": 14287 + }, + { + "epoch": 0.8680964821678109, + "grad_norm": 0.1218310073018074, + "learning_rate": 4.261159552588689e-06, + "loss": 1.0811, + "step": 14288 + }, + { + "epoch": 0.8681572392004374, + "grad_norm": 0.17570051550865173, + "learning_rate": 4.2572933769587045e-06, + "loss": 1.0334, + "step": 14289 + }, + { + "epoch": 0.868217996233064, + "grad_norm": 0.42319509387016296, + "learning_rate": 4.25342887803647e-06, + "loss": 1.0024, + "step": 14290 + }, + { + "epoch": 0.8682787532656905, + "grad_norm": 0.1260683387517929, + "learning_rate": 4.24956605596365e-06, + "loss": 1.0412, + "step": 14291 + }, + { + "epoch": 0.868339510298317, + "grad_norm": 0.24070203304290771, + "learning_rate": 4.245704910881826e-06, + "loss": 1.0567, + "step": 14292 + }, + { + "epoch": 0.8684002673309436, + "grad_norm": 0.4098961055278778, + "learning_rate": 4.241845442932535e-06, + "loss": 1.1695, + "step": 14293 + }, + { + "epoch": 0.8684610243635701, + "grad_norm": 0.19528326392173767, + "learning_rate": 4.237987652257242e-06, + "loss": 1.0968, + "step": 14294 + }, + { + "epoch": 0.8685217813961966, + "grad_norm": 0.1666220724582672, + "learning_rate": 4.23413153899736e-06, + "loss": 1.0677, + "step": 14295 + }, + { + "epoch": 0.8685825384288232, + "grad_norm": 0.12665048241615295, + "learning_rate": 4.2302771032942334e-06, + "loss": 0.9958, + "step": 14296 + }, + { + "epoch": 0.8686432954614497, + "grad_norm": 0.12987719476222992, + "learning_rate": 4.226424345289143e-06, + "loss": 1.0701, + "step": 14297 + }, + { + "epoch": 0.8687040524940762, + "grad_norm": 0.45124661922454834, + "learning_rate": 4.2225732651233195e-06, + "loss": 1.03, + "step": 14298 + }, + { + "epoch": 0.8687648095267028, + "grad_norm": 0.2224099040031433, + "learning_rate": 4.218723862937917e-06, + "loss": 1.1409, + "step": 14299 + }, + { + "epoch": 0.8688255665593292, + "grad_norm": 0.3506018817424774, + "learning_rate": 4.214876138874046e-06, + "loss": 1.1634, + "step": 14300 + }, + { + "epoch": 0.8688863235919557, + "grad_norm": 0.14502625167369843, + "learning_rate": 4.211030093072738e-06, + "loss": 1.0211, + "step": 14301 + }, + { + "epoch": 0.8689470806245823, + "grad_norm": 0.17851543426513672, + "learning_rate": 4.2071857256749725e-06, + "loss": 1.0427, + "step": 14302 + }, + { + "epoch": 0.8690078376572088, + "grad_norm": 0.1145797073841095, + "learning_rate": 4.203343036821667e-06, + "loss": 1.0029, + "step": 14303 + }, + { + "epoch": 0.8690685946898353, + "grad_norm": 0.18751749396324158, + "learning_rate": 4.199502026653674e-06, + "loss": 1.1411, + "step": 14304 + }, + { + "epoch": 0.8691293517224619, + "grad_norm": 0.571515679359436, + "learning_rate": 4.195662695311786e-06, + "loss": 1.2234, + "step": 14305 + }, + { + "epoch": 0.8691901087550884, + "grad_norm": 0.21342016756534576, + "learning_rate": 4.1918250429367255e-06, + "loss": 1.1042, + "step": 14306 + }, + { + "epoch": 0.8692508657877149, + "grad_norm": 0.12836679816246033, + "learning_rate": 4.187989069669196e-06, + "loss": 1.0566, + "step": 14307 + }, + { + "epoch": 0.8693116228203415, + "grad_norm": 0.5403417944908142, + "learning_rate": 4.184154775649768e-06, + "loss": 1.093, + "step": 14308 + }, + { + "epoch": 0.869372379852968, + "grad_norm": 0.25868064165115356, + "learning_rate": 4.180322161019007e-06, + "loss": 1.0661, + "step": 14309 + }, + { + "epoch": 0.8694331368855945, + "grad_norm": 0.17786923050880432, + "learning_rate": 4.176491225917384e-06, + "loss": 1.077, + "step": 14310 + }, + { + "epoch": 0.8694938939182211, + "grad_norm": 0.1542423963546753, + "learning_rate": 4.172661970485336e-06, + "loss": 1.0135, + "step": 14311 + }, + { + "epoch": 0.8695546509508476, + "grad_norm": 0.2713870108127594, + "learning_rate": 4.168834394863231e-06, + "loss": 1.0388, + "step": 14312 + }, + { + "epoch": 0.8696154079834741, + "grad_norm": 0.18632525205612183, + "learning_rate": 4.165008499191353e-06, + "loss": 1.1053, + "step": 14313 + }, + { + "epoch": 0.8696761650161006, + "grad_norm": 0.2230849713087082, + "learning_rate": 4.161184283609959e-06, + "loss": 1.0935, + "step": 14314 + }, + { + "epoch": 0.8697369220487271, + "grad_norm": 0.9498096108436584, + "learning_rate": 4.157361748259198e-06, + "loss": 1.0418, + "step": 14315 + }, + { + "epoch": 0.8697976790813536, + "grad_norm": 0.16682825982570648, + "learning_rate": 4.153540893279212e-06, + "loss": 1.0434, + "step": 14316 + }, + { + "epoch": 0.8698584361139802, + "grad_norm": 0.16825038194656372, + "learning_rate": 4.149721718810051e-06, + "loss": 1.0156, + "step": 14317 + }, + { + "epoch": 0.8699191931466067, + "grad_norm": 0.21659637987613678, + "learning_rate": 4.145904224991698e-06, + "loss": 1.088, + "step": 14318 + }, + { + "epoch": 0.8699799501792332, + "grad_norm": 8.292357444763184, + "learning_rate": 4.1420884119640905e-06, + "loss": 1.0894, + "step": 14319 + }, + { + "epoch": 0.8700407072118598, + "grad_norm": 0.12114522606134415, + "learning_rate": 4.1382742798670935e-06, + "loss": 1.033, + "step": 14320 + }, + { + "epoch": 0.8701014642444863, + "grad_norm": 0.2814321219921112, + "learning_rate": 4.134461828840519e-06, + "loss": 1.2188, + "step": 14321 + }, + { + "epoch": 0.8701622212771128, + "grad_norm": 0.1923433393239975, + "learning_rate": 4.130651059024116e-06, + "loss": 1.0694, + "step": 14322 + }, + { + "epoch": 0.8702229783097394, + "grad_norm": 0.12352757900953293, + "learning_rate": 4.12684197055756e-06, + "loss": 1.0042, + "step": 14323 + }, + { + "epoch": 0.8702837353423659, + "grad_norm": 0.24481001496315002, + "learning_rate": 4.1230345635804804e-06, + "loss": 1.1462, + "step": 14324 + }, + { + "epoch": 0.8703444923749925, + "grad_norm": 0.20350763201713562, + "learning_rate": 4.119228838232436e-06, + "loss": 1.0968, + "step": 14325 + }, + { + "epoch": 0.870405249407619, + "grad_norm": 0.13247951865196228, + "learning_rate": 4.115424794652917e-06, + "loss": 1.0732, + "step": 14326 + }, + { + "epoch": 0.8704660064402454, + "grad_norm": 0.15297284722328186, + "learning_rate": 4.111622432981393e-06, + "loss": 1.0434, + "step": 14327 + }, + { + "epoch": 0.870526763472872, + "grad_norm": 0.17366892099380493, + "learning_rate": 4.10782175335721e-06, + "loss": 1.1129, + "step": 14328 + }, + { + "epoch": 0.8705875205054985, + "grad_norm": 0.14597034454345703, + "learning_rate": 4.104022755919689e-06, + "loss": 0.9715, + "step": 14329 + }, + { + "epoch": 0.870648277538125, + "grad_norm": 0.1815367043018341, + "learning_rate": 4.1002254408080865e-06, + "loss": 1.0741, + "step": 14330 + }, + { + "epoch": 0.8707090345707515, + "grad_norm": 0.14075976610183716, + "learning_rate": 4.096429808161578e-06, + "loss": 1.0304, + "step": 14331 + }, + { + "epoch": 0.8707697916033781, + "grad_norm": 0.13794153928756714, + "learning_rate": 4.092635858119326e-06, + "loss": 1.06, + "step": 14332 + }, + { + "epoch": 0.8708305486360046, + "grad_norm": 0.18431952595710754, + "learning_rate": 4.088843590820374e-06, + "loss": 1.0922, + "step": 14333 + }, + { + "epoch": 0.8708913056686312, + "grad_norm": 0.19584570825099945, + "learning_rate": 4.085053006403744e-06, + "loss": 1.027, + "step": 14334 + }, + { + "epoch": 0.8709520627012577, + "grad_norm": 0.17634980380535126, + "learning_rate": 4.081264105008364e-06, + "loss": 1.0536, + "step": 14335 + }, + { + "epoch": 0.8710128197338842, + "grad_norm": 0.15625038743019104, + "learning_rate": 4.077476886773118e-06, + "loss": 0.9958, + "step": 14336 + }, + { + "epoch": 0.8710735767665108, + "grad_norm": 0.18549366295337677, + "learning_rate": 4.073691351836839e-06, + "loss": 1.0813, + "step": 14337 + }, + { + "epoch": 0.8711343337991373, + "grad_norm": 0.31516486406326294, + "learning_rate": 4.069907500338277e-06, + "loss": 1.123, + "step": 14338 + }, + { + "epoch": 0.8711950908317638, + "grad_norm": 0.23729470372200012, + "learning_rate": 4.066125332416138e-06, + "loss": 1.182, + "step": 14339 + }, + { + "epoch": 0.8712558478643903, + "grad_norm": 0.176693394780159, + "learning_rate": 4.062344848209055e-06, + "loss": 1.0698, + "step": 14340 + }, + { + "epoch": 0.8713166048970168, + "grad_norm": 0.4220251739025116, + "learning_rate": 4.0585660478556e-06, + "loss": 1.016, + "step": 14341 + }, + { + "epoch": 0.8713773619296433, + "grad_norm": 0.1388482302427292, + "learning_rate": 4.0547889314942855e-06, + "loss": 0.9769, + "step": 14342 + }, + { + "epoch": 0.8714381189622699, + "grad_norm": 0.19124607741832733, + "learning_rate": 4.051013499263568e-06, + "loss": 1.1129, + "step": 14343 + }, + { + "epoch": 0.8714988759948964, + "grad_norm": 0.14204874634742737, + "learning_rate": 4.047239751301829e-06, + "loss": 1.031, + "step": 14344 + }, + { + "epoch": 0.8715596330275229, + "grad_norm": 0.27552685141563416, + "learning_rate": 4.043467687747399e-06, + "loss": 1.1357, + "step": 14345 + }, + { + "epoch": 0.8716203900601495, + "grad_norm": 0.14975929260253906, + "learning_rate": 4.039697308738544e-06, + "loss": 1.0463, + "step": 14346 + }, + { + "epoch": 0.871681147092776, + "grad_norm": 0.21219268441200256, + "learning_rate": 4.03592861441347e-06, + "loss": 1.1219, + "step": 14347 + }, + { + "epoch": 0.8717419041254025, + "grad_norm": 0.17066437005996704, + "learning_rate": 4.0321616049103116e-06, + "loss": 1.0276, + "step": 14348 + }, + { + "epoch": 0.8718026611580291, + "grad_norm": 0.1870967149734497, + "learning_rate": 4.028396280367164e-06, + "loss": 1.1364, + "step": 14349 + }, + { + "epoch": 0.8718634181906556, + "grad_norm": 0.3549751043319702, + "learning_rate": 4.024632640922027e-06, + "loss": 1.0748, + "step": 14350 + }, + { + "epoch": 0.8719241752232821, + "grad_norm": 0.22137728333473206, + "learning_rate": 4.0208706867128744e-06, + "loss": 1.0751, + "step": 14351 + }, + { + "epoch": 0.8719849322559087, + "grad_norm": 0.13092724978923798, + "learning_rate": 4.017110417877585e-06, + "loss": 1.0318, + "step": 14352 + }, + { + "epoch": 0.8720456892885351, + "grad_norm": 0.12395705282688141, + "learning_rate": 4.013351834554008e-06, + "loss": 1.034, + "step": 14353 + }, + { + "epoch": 0.8721064463211616, + "grad_norm": 0.15049493312835693, + "learning_rate": 4.009594936879918e-06, + "loss": 1.0209, + "step": 14354 + }, + { + "epoch": 0.8721672033537882, + "grad_norm": 0.12627385556697845, + "learning_rate": 4.00583972499301e-06, + "loss": 1.0481, + "step": 14355 + }, + { + "epoch": 0.8722279603864147, + "grad_norm": 0.13765235245227814, + "learning_rate": 4.00208619903093e-06, + "loss": 1.0115, + "step": 14356 + }, + { + "epoch": 0.8722887174190412, + "grad_norm": 0.13533249497413635, + "learning_rate": 3.998334359131267e-06, + "loss": 1.0523, + "step": 14357 + }, + { + "epoch": 0.8723494744516678, + "grad_norm": 0.11498196423053741, + "learning_rate": 3.994584205431562e-06, + "loss": 1.0033, + "step": 14358 + }, + { + "epoch": 0.8724102314842943, + "grad_norm": 0.6258643269538879, + "learning_rate": 3.99083573806926e-06, + "loss": 1.2826, + "step": 14359 + }, + { + "epoch": 0.8724709885169208, + "grad_norm": 0.1489703506231308, + "learning_rate": 3.987088957181772e-06, + "loss": 1.0403, + "step": 14360 + }, + { + "epoch": 0.8725317455495474, + "grad_norm": 0.3279097080230713, + "learning_rate": 3.983343862906436e-06, + "loss": 1.009, + "step": 14361 + }, + { + "epoch": 0.8725925025821739, + "grad_norm": 0.1710648536682129, + "learning_rate": 3.979600455380511e-06, + "loss": 1.0541, + "step": 14362 + }, + { + "epoch": 0.8726532596148004, + "grad_norm": 0.19468390941619873, + "learning_rate": 3.975858734741239e-06, + "loss": 1.0223, + "step": 14363 + }, + { + "epoch": 0.872714016647427, + "grad_norm": 0.19586096704006195, + "learning_rate": 3.972118701125754e-06, + "loss": 1.1144, + "step": 14364 + }, + { + "epoch": 0.8727747736800535, + "grad_norm": 0.20174065232276917, + "learning_rate": 3.968380354671164e-06, + "loss": 1.0855, + "step": 14365 + }, + { + "epoch": 0.8728355307126799, + "grad_norm": 0.12432872503995895, + "learning_rate": 3.964643695514481e-06, + "loss": 1.0079, + "step": 14366 + }, + { + "epoch": 0.8728962877453065, + "grad_norm": 0.12494869530200958, + "learning_rate": 3.960908723792689e-06, + "loss": 0.9607, + "step": 14367 + }, + { + "epoch": 0.872957044777933, + "grad_norm": 0.1332417130470276, + "learning_rate": 3.95717543964268e-06, + "loss": 1.0616, + "step": 14368 + }, + { + "epoch": 0.8730178018105595, + "grad_norm": 0.13644619286060333, + "learning_rate": 3.95344384320131e-06, + "loss": 0.98, + "step": 14369 + }, + { + "epoch": 0.8730785588431861, + "grad_norm": 0.34435564279556274, + "learning_rate": 3.949713934605354e-06, + "loss": 1.0503, + "step": 14370 + }, + { + "epoch": 0.8731393158758126, + "grad_norm": 0.23179131746292114, + "learning_rate": 3.945985713991534e-06, + "loss": 0.9915, + "step": 14371 + }, + { + "epoch": 0.8732000729084392, + "grad_norm": 0.3271268606185913, + "learning_rate": 3.942259181496516e-06, + "loss": 1.1319, + "step": 14372 + }, + { + "epoch": 0.8732608299410657, + "grad_norm": 0.12314756214618683, + "learning_rate": 3.938534337256877e-06, + "loss": 1.0205, + "step": 14373 + }, + { + "epoch": 0.8733215869736922, + "grad_norm": 0.21814750134944916, + "learning_rate": 3.934811181409181e-06, + "loss": 1.1315, + "step": 14374 + }, + { + "epoch": 0.8733823440063188, + "grad_norm": 0.19943195581436157, + "learning_rate": 3.9310897140898805e-06, + "loss": 1.1178, + "step": 14375 + }, + { + "epoch": 0.8734431010389453, + "grad_norm": 0.3602393567562103, + "learning_rate": 3.927369935435387e-06, + "loss": 1.1275, + "step": 14376 + }, + { + "epoch": 0.8735038580715718, + "grad_norm": 0.18363171815872192, + "learning_rate": 3.923651845582061e-06, + "loss": 1.1024, + "step": 14377 + }, + { + "epoch": 0.8735646151041984, + "grad_norm": 0.1145695298910141, + "learning_rate": 3.9199354446661675e-06, + "loss": 1.0115, + "step": 14378 + }, + { + "epoch": 0.8736253721368248, + "grad_norm": 0.23513160645961761, + "learning_rate": 3.916220732823961e-06, + "loss": 1.0716, + "step": 14379 + }, + { + "epoch": 0.8736861291694513, + "grad_norm": 0.12501150369644165, + "learning_rate": 3.9125077101915875e-06, + "loss": 1.0297, + "step": 14380 + }, + { + "epoch": 0.8737468862020779, + "grad_norm": 0.15812218189239502, + "learning_rate": 3.908796376905166e-06, + "loss": 1.0376, + "step": 14381 + }, + { + "epoch": 0.8738076432347044, + "grad_norm": 0.2314951866865158, + "learning_rate": 3.90508673310071e-06, + "loss": 1.1502, + "step": 14382 + }, + { + "epoch": 0.8738684002673309, + "grad_norm": 0.21201108396053314, + "learning_rate": 3.9013787789142e-06, + "loss": 1.0002, + "step": 14383 + }, + { + "epoch": 0.8739291572999575, + "grad_norm": 0.583795428276062, + "learning_rate": 3.89767251448157e-06, + "loss": 1.0366, + "step": 14384 + }, + { + "epoch": 0.873989914332584, + "grad_norm": 0.12199315428733826, + "learning_rate": 3.893967939938669e-06, + "loss": 1.083, + "step": 14385 + }, + { + "epoch": 0.8740506713652105, + "grad_norm": 0.21857741475105286, + "learning_rate": 3.890265055421283e-06, + "loss": 1.063, + "step": 14386 + }, + { + "epoch": 0.8741114283978371, + "grad_norm": 0.22139400243759155, + "learning_rate": 3.886563861065146e-06, + "loss": 1.1744, + "step": 14387 + }, + { + "epoch": 0.8741721854304636, + "grad_norm": 5.177894592285156, + "learning_rate": 3.882864357005922e-06, + "loss": 1.0381, + "step": 14388 + }, + { + "epoch": 0.8742329424630901, + "grad_norm": 0.18632982671260834, + "learning_rate": 3.87916654337922e-06, + "loss": 0.9924, + "step": 14389 + }, + { + "epoch": 0.8742936994957167, + "grad_norm": 0.16451622545719147, + "learning_rate": 3.875470420320582e-06, + "loss": 1.1061, + "step": 14390 + }, + { + "epoch": 0.8743544565283432, + "grad_norm": 0.3266952633857727, + "learning_rate": 3.871775987965498e-06, + "loss": 1.1054, + "step": 14391 + }, + { + "epoch": 0.8744152135609696, + "grad_norm": 0.2052331417798996, + "learning_rate": 3.868083246449372e-06, + "loss": 1.1249, + "step": 14392 + }, + { + "epoch": 0.8744759705935962, + "grad_norm": 0.20391325652599335, + "learning_rate": 3.864392195907579e-06, + "loss": 1.0508, + "step": 14393 + }, + { + "epoch": 0.8745367276262227, + "grad_norm": 0.19546261429786682, + "learning_rate": 3.860702836475405e-06, + "loss": 1.0907, + "step": 14394 + }, + { + "epoch": 0.8745974846588492, + "grad_norm": 0.11917006224393845, + "learning_rate": 3.8570151682880865e-06, + "loss": 1.0612, + "step": 14395 + }, + { + "epoch": 0.8746582416914758, + "grad_norm": 1.2734040021896362, + "learning_rate": 3.853329191480798e-06, + "loss": 1.1691, + "step": 14396 + }, + { + "epoch": 0.8747189987241023, + "grad_norm": 0.13734720647335052, + "learning_rate": 3.849644906188643e-06, + "loss": 0.961, + "step": 14397 + }, + { + "epoch": 0.8747797557567288, + "grad_norm": 0.34725672006607056, + "learning_rate": 3.845962312546681e-06, + "loss": 1.2299, + "step": 14398 + }, + { + "epoch": 0.8748405127893554, + "grad_norm": 0.17266348004341125, + "learning_rate": 3.842281410689879e-06, + "loss": 1.0801, + "step": 14399 + }, + { + "epoch": 0.8749012698219819, + "grad_norm": 0.1492416262626648, + "learning_rate": 3.838602200753183e-06, + "loss": 1.0135, + "step": 14400 + }, + { + "epoch": 0.8749620268546084, + "grad_norm": 0.1612376719713211, + "learning_rate": 3.83492468287146e-06, + "loss": 1.0681, + "step": 14401 + }, + { + "epoch": 0.875022783887235, + "grad_norm": 0.180460125207901, + "learning_rate": 3.8312488571794816e-06, + "loss": 1.0367, + "step": 14402 + }, + { + "epoch": 0.8750835409198615, + "grad_norm": 0.25058209896087646, + "learning_rate": 3.827574723811994e-06, + "loss": 1.2664, + "step": 14403 + }, + { + "epoch": 0.875144297952488, + "grad_norm": 0.1555749475955963, + "learning_rate": 3.8239022829036905e-06, + "loss": 1.1071, + "step": 14404 + }, + { + "epoch": 0.8752050549851145, + "grad_norm": 0.13740970194339752, + "learning_rate": 3.8202315345891696e-06, + "loss": 1.013, + "step": 14405 + }, + { + "epoch": 0.875265812017741, + "grad_norm": 0.2238846868276596, + "learning_rate": 3.816562479002994e-06, + "loss": 1.0522, + "step": 14406 + }, + { + "epoch": 0.8753265690503675, + "grad_norm": 2.239617109298706, + "learning_rate": 3.812895116279641e-06, + "loss": 1.0063, + "step": 14407 + }, + { + "epoch": 0.8753873260829941, + "grad_norm": 0.17144718766212463, + "learning_rate": 3.8092294465535527e-06, + "loss": 1.0604, + "step": 14408 + }, + { + "epoch": 0.8754480831156206, + "grad_norm": 0.1408739686012268, + "learning_rate": 3.8055654699590827e-06, + "loss": 1.0456, + "step": 14409 + }, + { + "epoch": 0.8755088401482471, + "grad_norm": 0.24737276136875153, + "learning_rate": 3.80190318663054e-06, + "loss": 1.1275, + "step": 14410 + }, + { + "epoch": 0.8755695971808737, + "grad_norm": 0.21512773633003235, + "learning_rate": 3.7982425967021673e-06, + "loss": 1.1051, + "step": 14411 + }, + { + "epoch": 0.8756303542135002, + "grad_norm": 0.14475232362747192, + "learning_rate": 3.7945837003081406e-06, + "loss": 1.0259, + "step": 14412 + }, + { + "epoch": 0.8756911112461268, + "grad_norm": 0.20246560871601105, + "learning_rate": 3.790926497582581e-06, + "loss": 1.1335, + "step": 14413 + }, + { + "epoch": 0.8757518682787533, + "grad_norm": 0.16499514877796173, + "learning_rate": 3.7872709886595413e-06, + "loss": 1.0706, + "step": 14414 + }, + { + "epoch": 0.8758126253113798, + "grad_norm": 0.18121904134750366, + "learning_rate": 3.7836171736730206e-06, + "loss": 1.0291, + "step": 14415 + }, + { + "epoch": 0.8758733823440064, + "grad_norm": 0.13060136139392853, + "learning_rate": 3.7799650527569396e-06, + "loss": 1.0493, + "step": 14416 + }, + { + "epoch": 0.8759341393766329, + "grad_norm": 0.1665709912776947, + "learning_rate": 3.77631462604518e-06, + "loss": 1.0385, + "step": 14417 + }, + { + "epoch": 0.8759948964092594, + "grad_norm": 0.16085350513458252, + "learning_rate": 3.7726658936715397e-06, + "loss": 0.9977, + "step": 14418 + }, + { + "epoch": 0.8760556534418859, + "grad_norm": 0.12469112128019333, + "learning_rate": 3.769018855769757e-06, + "loss": 1.0264, + "step": 14419 + }, + { + "epoch": 0.8761164104745124, + "grad_norm": 0.12992802262306213, + "learning_rate": 3.7653735124735355e-06, + "loss": 1.082, + "step": 14420 + }, + { + "epoch": 0.8761771675071389, + "grad_norm": 0.14491324126720428, + "learning_rate": 3.7617298639164913e-06, + "loss": 1.0243, + "step": 14421 + }, + { + "epoch": 0.8762379245397655, + "grad_norm": 0.13997085392475128, + "learning_rate": 3.7580879102321663e-06, + "loss": 1.0351, + "step": 14422 + }, + { + "epoch": 0.876298681572392, + "grad_norm": 5.346330642700195, + "learning_rate": 3.7544476515540715e-06, + "loss": 1.1506, + "step": 14423 + }, + { + "epoch": 0.8763594386050185, + "grad_norm": 0.2220321148633957, + "learning_rate": 3.750809088015628e-06, + "loss": 1.0673, + "step": 14424 + }, + { + "epoch": 0.8764201956376451, + "grad_norm": 0.1224537119269371, + "learning_rate": 3.747172219750222e-06, + "loss": 1.0099, + "step": 14425 + }, + { + "epoch": 0.8764809526702716, + "grad_norm": 0.17345163226127625, + "learning_rate": 3.74353704689116e-06, + "loss": 1.0565, + "step": 14426 + }, + { + "epoch": 0.8765417097028981, + "grad_norm": 0.15073923766613007, + "learning_rate": 3.73990356957169e-06, + "loss": 1.038, + "step": 14427 + }, + { + "epoch": 0.8766024667355247, + "grad_norm": 0.20939861238002777, + "learning_rate": 3.7362717879250053e-06, + "loss": 1.1027, + "step": 14428 + }, + { + "epoch": 0.8766632237681512, + "grad_norm": 0.2016695737838745, + "learning_rate": 3.732641702084205e-06, + "loss": 1.1292, + "step": 14429 + }, + { + "epoch": 0.8767239808007777, + "grad_norm": 1.3114008903503418, + "learning_rate": 3.7290133121823723e-06, + "loss": 1.0004, + "step": 14430 + }, + { + "epoch": 0.8767847378334043, + "grad_norm": 0.15456895530223846, + "learning_rate": 3.7253866183525e-06, + "loss": 1.0441, + "step": 14431 + }, + { + "epoch": 0.8768454948660307, + "grad_norm": 0.10794993489980698, + "learning_rate": 3.721761620727532e-06, + "loss": 1.0077, + "step": 14432 + }, + { + "epoch": 0.8769062518986572, + "grad_norm": 0.18258923292160034, + "learning_rate": 3.7181383194403286e-06, + "loss": 1.0925, + "step": 14433 + }, + { + "epoch": 0.8769670089312838, + "grad_norm": 1.8274368047714233, + "learning_rate": 3.714516714623717e-06, + "loss": 1.1112, + "step": 14434 + }, + { + "epoch": 0.8770277659639103, + "grad_norm": 0.16921208798885345, + "learning_rate": 3.7108968064104412e-06, + "loss": 1.0513, + "step": 14435 + }, + { + "epoch": 0.8770885229965368, + "grad_norm": 0.12198228389024734, + "learning_rate": 3.707278594933189e-06, + "loss": 1.0541, + "step": 14436 + }, + { + "epoch": 0.8771492800291634, + "grad_norm": 0.23565281927585602, + "learning_rate": 3.703662080324588e-06, + "loss": 1.1421, + "step": 14437 + }, + { + "epoch": 0.8772100370617899, + "grad_norm": 0.15964752435684204, + "learning_rate": 3.7000472627172047e-06, + "loss": 1.0136, + "step": 14438 + }, + { + "epoch": 0.8772707940944164, + "grad_norm": 0.19400636851787567, + "learning_rate": 3.696434142243538e-06, + "loss": 1.1468, + "step": 14439 + }, + { + "epoch": 0.877331551127043, + "grad_norm": 0.18440289795398712, + "learning_rate": 3.6928227190360154e-06, + "loss": 1.0475, + "step": 14440 + }, + { + "epoch": 0.8773923081596695, + "grad_norm": 0.15126103162765503, + "learning_rate": 3.6892129932270482e-06, + "loss": 1.057, + "step": 14441 + }, + { + "epoch": 0.877453065192296, + "grad_norm": 0.11365754157304764, + "learning_rate": 3.685604964948919e-06, + "loss": 1.0107, + "step": 14442 + }, + { + "epoch": 0.8775138222249226, + "grad_norm": 0.16070865094661713, + "learning_rate": 3.681998634333894e-06, + "loss": 1.0632, + "step": 14443 + }, + { + "epoch": 0.8775745792575491, + "grad_norm": 0.15649114549160004, + "learning_rate": 3.6783940015141625e-06, + "loss": 1.0783, + "step": 14444 + }, + { + "epoch": 0.8776353362901755, + "grad_norm": 0.13479791581630707, + "learning_rate": 3.6747910666218466e-06, + "loss": 1.0532, + "step": 14445 + }, + { + "epoch": 0.8776960933228021, + "grad_norm": 0.12779037654399872, + "learning_rate": 3.6711898297890234e-06, + "loss": 1.0385, + "step": 14446 + }, + { + "epoch": 0.8777568503554286, + "grad_norm": 0.13625748455524445, + "learning_rate": 3.667590291147693e-06, + "loss": 1.0617, + "step": 14447 + }, + { + "epoch": 0.8778176073880551, + "grad_norm": 0.1561240404844284, + "learning_rate": 3.663992450829801e-06, + "loss": 1.0506, + "step": 14448 + }, + { + "epoch": 0.8778783644206817, + "grad_norm": 0.17477624118328094, + "learning_rate": 3.660396308967218e-06, + "loss": 0.9843, + "step": 14449 + }, + { + "epoch": 0.8779391214533082, + "grad_norm": 0.1726529896259308, + "learning_rate": 3.6568018656917614e-06, + "loss": 1.0388, + "step": 14450 + }, + { + "epoch": 0.8779998784859347, + "grad_norm": 0.9136340022087097, + "learning_rate": 3.6532091211351928e-06, + "loss": 0.98, + "step": 14451 + }, + { + "epoch": 0.8780606355185613, + "grad_norm": 0.2696666419506073, + "learning_rate": 3.6496180754292007e-06, + "loss": 1.135, + "step": 14452 + }, + { + "epoch": 0.8781213925511878, + "grad_norm": 0.1500103622674942, + "learning_rate": 3.6460287287054196e-06, + "loss": 1.0599, + "step": 14453 + }, + { + "epoch": 0.8781821495838144, + "grad_norm": 0.14111120998859406, + "learning_rate": 3.642441081095421e-06, + "loss": 1.0058, + "step": 14454 + }, + { + "epoch": 0.8782429066164409, + "grad_norm": 0.17225222289562225, + "learning_rate": 3.638855132730701e-06, + "loss": 1.0601, + "step": 14455 + }, + { + "epoch": 0.8783036636490674, + "grad_norm": 0.1802489310503006, + "learning_rate": 3.635270883742703e-06, + "loss": 1.0125, + "step": 14456 + }, + { + "epoch": 0.878364420681694, + "grad_norm": 0.2624325752258301, + "learning_rate": 3.631688334262817e-06, + "loss": 1.0315, + "step": 14457 + }, + { + "epoch": 0.8784251777143204, + "grad_norm": 0.16279882192611694, + "learning_rate": 3.6281074844223608e-06, + "loss": 1.0215, + "step": 14458 + }, + { + "epoch": 0.8784859347469469, + "grad_norm": 0.34861084818840027, + "learning_rate": 3.6245283343525837e-06, + "loss": 1.1188, + "step": 14459 + }, + { + "epoch": 0.8785466917795735, + "grad_norm": 0.1743268072605133, + "learning_rate": 3.6209508841846873e-06, + "loss": 1.0528, + "step": 14460 + }, + { + "epoch": 0.8786074488122, + "grad_norm": 0.14961734414100647, + "learning_rate": 3.6173751340497997e-06, + "loss": 1.0329, + "step": 14461 + }, + { + "epoch": 0.8786682058448265, + "grad_norm": 0.16821926832199097, + "learning_rate": 3.6138010840789936e-06, + "loss": 0.9814, + "step": 14462 + }, + { + "epoch": 0.8787289628774531, + "grad_norm": 0.20660807192325592, + "learning_rate": 3.6102287344032703e-06, + "loss": 1.0986, + "step": 14463 + }, + { + "epoch": 0.8787897199100796, + "grad_norm": 0.22808469831943512, + "learning_rate": 3.6066580851535857e-06, + "loss": 1.1827, + "step": 14464 + }, + { + "epoch": 0.8788504769427061, + "grad_norm": 0.15190352499485016, + "learning_rate": 3.6030891364608133e-06, + "loss": 1.0073, + "step": 14465 + }, + { + "epoch": 0.8789112339753327, + "grad_norm": 0.1339408904314041, + "learning_rate": 3.59952188845577e-06, + "loss": 1.0506, + "step": 14466 + }, + { + "epoch": 0.8789719910079592, + "grad_norm": 0.3095022439956665, + "learning_rate": 3.5959563412692297e-06, + "loss": 1.2806, + "step": 14467 + }, + { + "epoch": 0.8790327480405857, + "grad_norm": 0.24397894740104675, + "learning_rate": 3.5923924950318875e-06, + "loss": 1.1573, + "step": 14468 + }, + { + "epoch": 0.8790935050732123, + "grad_norm": 0.17128486931324005, + "learning_rate": 3.588830349874356e-06, + "loss": 1.0287, + "step": 14469 + }, + { + "epoch": 0.8791542621058388, + "grad_norm": 0.1929018646478653, + "learning_rate": 3.5852699059272243e-06, + "loss": 1.0719, + "step": 14470 + }, + { + "epoch": 0.8792150191384652, + "grad_norm": 0.22124391794204712, + "learning_rate": 3.581711163320989e-06, + "loss": 1.0577, + "step": 14471 + }, + { + "epoch": 0.8792757761710918, + "grad_norm": 0.16079503297805786, + "learning_rate": 3.578154122186106e-06, + "loss": 1.037, + "step": 14472 + }, + { + "epoch": 0.8793365332037183, + "grad_norm": 0.1346602737903595, + "learning_rate": 3.5745987826529604e-06, + "loss": 1.033, + "step": 14473 + }, + { + "epoch": 0.8793972902363448, + "grad_norm": 0.2628200054168701, + "learning_rate": 3.57104514485187e-06, + "loss": 1.1706, + "step": 14474 + }, + { + "epoch": 0.8794580472689714, + "grad_norm": 0.15217271447181702, + "learning_rate": 3.567493208913103e-06, + "loss": 1.0516, + "step": 14475 + }, + { + "epoch": 0.8795188043015979, + "grad_norm": 0.12688978016376495, + "learning_rate": 3.5639429749668272e-06, + "loss": 0.9768, + "step": 14476 + }, + { + "epoch": 0.8795795613342244, + "grad_norm": 0.29704996943473816, + "learning_rate": 3.5603944431432114e-06, + "loss": 1.0748, + "step": 14477 + }, + { + "epoch": 0.879640318366851, + "grad_norm": 0.14897190034389496, + "learning_rate": 3.5568476135723116e-06, + "loss": 1.0344, + "step": 14478 + }, + { + "epoch": 0.8797010753994775, + "grad_norm": 0.15824462473392487, + "learning_rate": 3.553302486384141e-06, + "loss": 1.0482, + "step": 14479 + }, + { + "epoch": 0.879761832432104, + "grad_norm": 0.19568955898284912, + "learning_rate": 3.549759061708641e-06, + "loss": 1.1169, + "step": 14480 + }, + { + "epoch": 0.8798225894647306, + "grad_norm": 0.1629011631011963, + "learning_rate": 3.546217339675706e-06, + "loss": 1.0985, + "step": 14481 + }, + { + "epoch": 0.8798833464973571, + "grad_norm": 0.2627391517162323, + "learning_rate": 3.5426773204151498e-06, + "loss": 1.0819, + "step": 14482 + }, + { + "epoch": 0.8799441035299836, + "grad_norm": 0.4110495448112488, + "learning_rate": 3.5391390040567408e-06, + "loss": 1.2318, + "step": 14483 + }, + { + "epoch": 0.8800048605626101, + "grad_norm": 0.12778513133525848, + "learning_rate": 3.5356023907301696e-06, + "loss": 1.0074, + "step": 14484 + }, + { + "epoch": 0.8800656175952366, + "grad_norm": 0.20158414542675018, + "learning_rate": 3.5320674805650767e-06, + "loss": 1.0897, + "step": 14485 + }, + { + "epoch": 0.8801263746278631, + "grad_norm": 0.7970670461654663, + "learning_rate": 3.528534273691031e-06, + "loss": 1.0442, + "step": 14486 + }, + { + "epoch": 0.8801871316604897, + "grad_norm": 0.14079803228378296, + "learning_rate": 3.5250027702375345e-06, + "loss": 1.0379, + "step": 14487 + }, + { + "epoch": 0.8802478886931162, + "grad_norm": 0.22049300372600555, + "learning_rate": 3.521472970334061e-06, + "loss": 1.1366, + "step": 14488 + }, + { + "epoch": 0.8803086457257427, + "grad_norm": 0.4296678304672241, + "learning_rate": 3.5179448741099743e-06, + "loss": 1.1443, + "step": 14489 + }, + { + "epoch": 0.8803694027583693, + "grad_norm": 0.1529468446969986, + "learning_rate": 3.5144184816946036e-06, + "loss": 0.9768, + "step": 14490 + }, + { + "epoch": 0.8804301597909958, + "grad_norm": 0.17840777337551117, + "learning_rate": 3.510893793217207e-06, + "loss": 1.1046, + "step": 14491 + }, + { + "epoch": 0.8804909168236223, + "grad_norm": 0.13200390338897705, + "learning_rate": 3.5073708088069756e-06, + "loss": 1.0193, + "step": 14492 + }, + { + "epoch": 0.8805516738562489, + "grad_norm": 0.19987930357456207, + "learning_rate": 3.503849528593062e-06, + "loss": 1.0637, + "step": 14493 + }, + { + "epoch": 0.8806124308888754, + "grad_norm": 0.27242833375930786, + "learning_rate": 3.500329952704534e-06, + "loss": 1.0872, + "step": 14494 + }, + { + "epoch": 0.880673187921502, + "grad_norm": 0.12059976905584335, + "learning_rate": 3.496812081270406e-06, + "loss": 1.0542, + "step": 14495 + }, + { + "epoch": 0.8807339449541285, + "grad_norm": 0.13918662071228027, + "learning_rate": 3.493295914419603e-06, + "loss": 1.0192, + "step": 14496 + }, + { + "epoch": 0.8807947019867549, + "grad_norm": 0.13058942556381226, + "learning_rate": 3.489781452281038e-06, + "loss": 1.0291, + "step": 14497 + }, + { + "epoch": 0.8808554590193814, + "grad_norm": 3.391538619995117, + "learning_rate": 3.4862686949835244e-06, + "loss": 1.0504, + "step": 14498 + }, + { + "epoch": 0.880916216052008, + "grad_norm": 8.77546215057373, + "learning_rate": 3.4827576426558206e-06, + "loss": 1.0135, + "step": 14499 + }, + { + "epoch": 0.8809769730846345, + "grad_norm": 0.3346589207649231, + "learning_rate": 3.479248295426624e-06, + "loss": 1.0977, + "step": 14500 + }, + { + "epoch": 0.881037730117261, + "grad_norm": 0.21148189902305603, + "learning_rate": 3.4757406534245808e-06, + "loss": 1.0874, + "step": 14501 + }, + { + "epoch": 0.8810984871498876, + "grad_norm": 0.18310369551181793, + "learning_rate": 3.4722347167782498e-06, + "loss": 1.015, + "step": 14502 + }, + { + "epoch": 0.8811592441825141, + "grad_norm": 0.21790918707847595, + "learning_rate": 3.468730485616156e-06, + "loss": 1.1849, + "step": 14503 + }, + { + "epoch": 0.8812200012151407, + "grad_norm": 0.1728108525276184, + "learning_rate": 3.4652279600667357e-06, + "loss": 1.083, + "step": 14504 + }, + { + "epoch": 0.8812807582477672, + "grad_norm": 0.252174973487854, + "learning_rate": 3.4617271402583806e-06, + "loss": 1.0949, + "step": 14505 + }, + { + "epoch": 0.8813415152803937, + "grad_norm": 0.16570855677127838, + "learning_rate": 3.4582280263194157e-06, + "loss": 1.0575, + "step": 14506 + }, + { + "epoch": 0.8814022723130203, + "grad_norm": 0.21254824101924896, + "learning_rate": 3.4547306183780825e-06, + "loss": 1.0304, + "step": 14507 + }, + { + "epoch": 0.8814630293456468, + "grad_norm": 0.13818056881427765, + "learning_rate": 3.451234916562618e-06, + "loss": 1.0663, + "step": 14508 + }, + { + "epoch": 0.8815237863782733, + "grad_norm": 1.3638237714767456, + "learning_rate": 3.44774092100113e-06, + "loss": 1.0285, + "step": 14509 + }, + { + "epoch": 0.8815845434108998, + "grad_norm": 0.2002485692501068, + "learning_rate": 3.444248631821689e-06, + "loss": 1.1395, + "step": 14510 + }, + { + "epoch": 0.8816453004435263, + "grad_norm": 0.14149457216262817, + "learning_rate": 3.4407580491523206e-06, + "loss": 1.0355, + "step": 14511 + }, + { + "epoch": 0.8817060574761528, + "grad_norm": 0.11646909266710281, + "learning_rate": 3.437269173120955e-06, + "loss": 0.9818, + "step": 14512 + }, + { + "epoch": 0.8817668145087794, + "grad_norm": 0.20628085732460022, + "learning_rate": 3.4337820038554958e-06, + "loss": 1.0231, + "step": 14513 + }, + { + "epoch": 0.8818275715414059, + "grad_norm": 0.18940769135951996, + "learning_rate": 3.4302965414837575e-06, + "loss": 1.0885, + "step": 14514 + }, + { + "epoch": 0.8818883285740324, + "grad_norm": 0.4171278476715088, + "learning_rate": 3.4268127861335097e-06, + "loss": 1.0083, + "step": 14515 + }, + { + "epoch": 0.881949085606659, + "grad_norm": 0.28989920020103455, + "learning_rate": 3.423330737932434e-06, + "loss": 1.1958, + "step": 14516 + }, + { + "epoch": 0.8820098426392855, + "grad_norm": 14.782523155212402, + "learning_rate": 3.4198503970081675e-06, + "loss": 1.0862, + "step": 14517 + }, + { + "epoch": 0.882070599671912, + "grad_norm": 0.12632082402706146, + "learning_rate": 3.416371763488291e-06, + "loss": 1.0274, + "step": 14518 + }, + { + "epoch": 0.8821313567045386, + "grad_norm": 0.14843371510505676, + "learning_rate": 3.412894837500319e-06, + "loss": 1.0584, + "step": 14519 + }, + { + "epoch": 0.8821921137371651, + "grad_norm": 0.15209273993968964, + "learning_rate": 3.409419619171683e-06, + "loss": 1.0621, + "step": 14520 + }, + { + "epoch": 0.8822528707697916, + "grad_norm": 0.11612807959318161, + "learning_rate": 3.4059461086297815e-06, + "loss": 1.0012, + "step": 14521 + }, + { + "epoch": 0.8823136278024182, + "grad_norm": 0.15618036687374115, + "learning_rate": 3.402474306001935e-06, + "loss": 1.0715, + "step": 14522 + }, + { + "epoch": 0.8823743848350447, + "grad_norm": 0.15520496666431427, + "learning_rate": 3.3990042114153965e-06, + "loss": 1.0499, + "step": 14523 + }, + { + "epoch": 0.8824351418676711, + "grad_norm": 0.3711283504962921, + "learning_rate": 3.3955358249973646e-06, + "loss": 1.0399, + "step": 14524 + }, + { + "epoch": 0.8824958989002977, + "grad_norm": 0.2275630086660385, + "learning_rate": 3.3920691468749768e-06, + "loss": 1.0859, + "step": 14525 + }, + { + "epoch": 0.8825566559329242, + "grad_norm": 0.11444173753261566, + "learning_rate": 3.3886041771753086e-06, + "loss": 1.0139, + "step": 14526 + }, + { + "epoch": 0.8826174129655507, + "grad_norm": 0.15336456894874573, + "learning_rate": 3.385140916025359e-06, + "loss": 1.0974, + "step": 14527 + }, + { + "epoch": 0.8826781699981773, + "grad_norm": 0.1284293681383133, + "learning_rate": 3.3816793635520816e-06, + "loss": 1.0094, + "step": 14528 + }, + { + "epoch": 0.8827389270308038, + "grad_norm": 0.13089114427566528, + "learning_rate": 3.378219519882353e-06, + "loss": 1.0532, + "step": 14529 + }, + { + "epoch": 0.8827996840634303, + "grad_norm": 0.6996250748634338, + "learning_rate": 3.374761385143005e-06, + "loss": 1.1825, + "step": 14530 + }, + { + "epoch": 0.8828604410960569, + "grad_norm": 0.14399175345897675, + "learning_rate": 3.3713049594607915e-06, + "loss": 1.0587, + "step": 14531 + }, + { + "epoch": 0.8829211981286834, + "grad_norm": 0.27685341238975525, + "learning_rate": 3.3678502429624057e-06, + "loss": 1.109, + "step": 14532 + }, + { + "epoch": 0.88298195516131, + "grad_norm": 0.11646721512079239, + "learning_rate": 3.3643972357744746e-06, + "loss": 1.0496, + "step": 14533 + }, + { + "epoch": 0.8830427121939365, + "grad_norm": 0.20729738473892212, + "learning_rate": 3.360945938023585e-06, + "loss": 1.2421, + "step": 14534 + }, + { + "epoch": 0.883103469226563, + "grad_norm": 0.27463585138320923, + "learning_rate": 3.357496349836248e-06, + "loss": 1.2531, + "step": 14535 + }, + { + "epoch": 0.8831642262591896, + "grad_norm": 0.2670561969280243, + "learning_rate": 3.3540484713388832e-06, + "loss": 1.1609, + "step": 14536 + }, + { + "epoch": 0.883224983291816, + "grad_norm": 0.26559165120124817, + "learning_rate": 3.350602302657896e-06, + "loss": 1.1951, + "step": 14537 + }, + { + "epoch": 0.8832857403244425, + "grad_norm": 0.12087835371494293, + "learning_rate": 3.347157843919585e-06, + "loss": 1.005, + "step": 14538 + }, + { + "epoch": 0.883346497357069, + "grad_norm": 0.2109571248292923, + "learning_rate": 3.3437150952502327e-06, + "loss": 1.1503, + "step": 14539 + }, + { + "epoch": 0.8834072543896956, + "grad_norm": 0.14600321650505066, + "learning_rate": 3.3402740567760214e-06, + "loss": 1.0564, + "step": 14540 + }, + { + "epoch": 0.8834680114223221, + "grad_norm": 0.17097905278205872, + "learning_rate": 3.336834728623084e-06, + "loss": 1.1199, + "step": 14541 + }, + { + "epoch": 0.8835287684549487, + "grad_norm": 0.20755547285079956, + "learning_rate": 3.333397110917497e-06, + "loss": 1.1611, + "step": 14542 + }, + { + "epoch": 0.8835895254875752, + "grad_norm": 0.15270714461803436, + "learning_rate": 3.329961203785242e-06, + "loss": 1.0544, + "step": 14543 + }, + { + "epoch": 0.8836502825202017, + "grad_norm": 0.1690043956041336, + "learning_rate": 3.3265270073522925e-06, + "loss": 0.9828, + "step": 14544 + }, + { + "epoch": 0.8837110395528283, + "grad_norm": 0.1836870014667511, + "learning_rate": 3.323094521744513e-06, + "loss": 1.1341, + "step": 14545 + }, + { + "epoch": 0.8837717965854548, + "grad_norm": 0.505225658416748, + "learning_rate": 3.3196637470877256e-06, + "loss": 1.2481, + "step": 14546 + }, + { + "epoch": 0.8838325536180813, + "grad_norm": 0.29924362897872925, + "learning_rate": 3.3162346835076906e-06, + "loss": 1.1829, + "step": 14547 + }, + { + "epoch": 0.8838933106507079, + "grad_norm": 0.18640224635601044, + "learning_rate": 3.312807331130091e-06, + "loss": 1.1324, + "step": 14548 + }, + { + "epoch": 0.8839540676833344, + "grad_norm": 0.5874704718589783, + "learning_rate": 3.309381690080571e-06, + "loss": 1.1831, + "step": 14549 + }, + { + "epoch": 0.8840148247159608, + "grad_norm": 0.14716625213623047, + "learning_rate": 3.305957760484685e-06, + "loss": 1.1191, + "step": 14550 + }, + { + "epoch": 0.8840755817485874, + "grad_norm": 0.2093965709209442, + "learning_rate": 3.302535542467944e-06, + "loss": 1.0699, + "step": 14551 + }, + { + "epoch": 0.8841363387812139, + "grad_norm": 0.19039620459079742, + "learning_rate": 3.2991150361557865e-06, + "loss": 1.0948, + "step": 14552 + }, + { + "epoch": 0.8841970958138404, + "grad_norm": 0.3159107267856598, + "learning_rate": 3.2956962416736015e-06, + "loss": 1.0536, + "step": 14553 + }, + { + "epoch": 0.884257852846467, + "grad_norm": 0.2243555784225464, + "learning_rate": 3.292279159146683e-06, + "loss": 1.0561, + "step": 14554 + }, + { + "epoch": 0.8843186098790935, + "grad_norm": 0.14930252730846405, + "learning_rate": 3.288863788700319e-06, + "loss": 1.0336, + "step": 14555 + }, + { + "epoch": 0.88437936691172, + "grad_norm": 0.15131725370883942, + "learning_rate": 3.2854501304596774e-06, + "loss": 1.0738, + "step": 14556 + }, + { + "epoch": 0.8844401239443466, + "grad_norm": 1.4232388734817505, + "learning_rate": 3.282038184549885e-06, + "loss": 1.0201, + "step": 14557 + }, + { + "epoch": 0.8845008809769731, + "grad_norm": 0.24283523857593536, + "learning_rate": 3.2786279510960193e-06, + "loss": 1.0175, + "step": 14558 + }, + { + "epoch": 0.8845616380095996, + "grad_norm": 0.15136750042438507, + "learning_rate": 3.2752194302230644e-06, + "loss": 1.0595, + "step": 14559 + }, + { + "epoch": 0.8846223950422262, + "grad_norm": 0.12167925387620926, + "learning_rate": 3.271812622055981e-06, + "loss": 0.9965, + "step": 14560 + }, + { + "epoch": 0.8846831520748527, + "grad_norm": 0.14213620126247406, + "learning_rate": 3.268407526719641e-06, + "loss": 0.9988, + "step": 14561 + }, + { + "epoch": 0.8847439091074792, + "grad_norm": 0.20265522599220276, + "learning_rate": 3.265004144338862e-06, + "loss": 1.0939, + "step": 14562 + }, + { + "epoch": 0.8848046661401057, + "grad_norm": 0.19795045256614685, + "learning_rate": 3.2616024750383833e-06, + "loss": 1.0322, + "step": 14563 + }, + { + "epoch": 0.8848654231727322, + "grad_norm": 0.22269441187381744, + "learning_rate": 3.2582025189428934e-06, + "loss": 1.1659, + "step": 14564 + }, + { + "epoch": 0.8849261802053587, + "grad_norm": 0.1275341957807541, + "learning_rate": 3.2548042761770316e-06, + "loss": 1.029, + "step": 14565 + }, + { + "epoch": 0.8849869372379853, + "grad_norm": 0.30392855405807495, + "learning_rate": 3.2514077468653537e-06, + "loss": 1.0965, + "step": 14566 + }, + { + "epoch": 0.8850476942706118, + "grad_norm": 0.20868369936943054, + "learning_rate": 3.248012931132366e-06, + "loss": 1.1226, + "step": 14567 + }, + { + "epoch": 0.8851084513032383, + "grad_norm": 0.17310146987438202, + "learning_rate": 3.244619829102502e-06, + "loss": 1.035, + "step": 14568 + }, + { + "epoch": 0.8851692083358649, + "grad_norm": 0.16774041950702667, + "learning_rate": 3.2412284409001235e-06, + "loss": 0.9967, + "step": 14569 + }, + { + "epoch": 0.8852299653684914, + "grad_norm": 1.005616307258606, + "learning_rate": 3.2378387666495645e-06, + "loss": 1.1563, + "step": 14570 + }, + { + "epoch": 0.885290722401118, + "grad_norm": 0.1913168728351593, + "learning_rate": 3.234450806475059e-06, + "loss": 1.1321, + "step": 14571 + }, + { + "epoch": 0.8853514794337445, + "grad_norm": 0.1384090781211853, + "learning_rate": 3.2310645605008015e-06, + "loss": 1.0309, + "step": 14572 + }, + { + "epoch": 0.885412236466371, + "grad_norm": 0.34611019492149353, + "learning_rate": 3.2276800288509157e-06, + "loss": 1.212, + "step": 14573 + }, + { + "epoch": 0.8854729934989976, + "grad_norm": 0.18774652481079102, + "learning_rate": 3.2242972116494574e-06, + "loss": 1.0243, + "step": 14574 + }, + { + "epoch": 0.8855337505316241, + "grad_norm": 0.13275499641895294, + "learning_rate": 3.2209161090204277e-06, + "loss": 1.0286, + "step": 14575 + }, + { + "epoch": 0.8855945075642505, + "grad_norm": 0.16403929889202118, + "learning_rate": 3.217536721087755e-06, + "loss": 1.0421, + "step": 14576 + }, + { + "epoch": 0.885655264596877, + "grad_norm": 0.20940187573432922, + "learning_rate": 3.2141590479753236e-06, + "loss": 1.0071, + "step": 14577 + }, + { + "epoch": 0.8857160216295036, + "grad_norm": 0.2083963304758072, + "learning_rate": 3.210783089806929e-06, + "loss": 1.112, + "step": 14578 + }, + { + "epoch": 0.8857767786621301, + "grad_norm": 0.23006758093833923, + "learning_rate": 3.2074088467063277e-06, + "loss": 1.0323, + "step": 14579 + }, + { + "epoch": 0.8858375356947567, + "grad_norm": 0.2136506885290146, + "learning_rate": 3.2040363187971935e-06, + "loss": 1.1265, + "step": 14580 + }, + { + "epoch": 0.8858982927273832, + "grad_norm": 0.466183602809906, + "learning_rate": 3.2006655062031597e-06, + "loss": 1.0446, + "step": 14581 + }, + { + "epoch": 0.8859590497600097, + "grad_norm": 0.2476292997598648, + "learning_rate": 3.197296409047784e-06, + "loss": 1.2173, + "step": 14582 + }, + { + "epoch": 0.8860198067926363, + "grad_norm": 0.16957299411296844, + "learning_rate": 3.193929027454551e-06, + "loss": 1.0453, + "step": 14583 + }, + { + "epoch": 0.8860805638252628, + "grad_norm": 0.11056461930274963, + "learning_rate": 3.1905633615468944e-06, + "loss": 1.0366, + "step": 14584 + }, + { + "epoch": 0.8861413208578893, + "grad_norm": 0.1865938901901245, + "learning_rate": 3.1871994114481772e-06, + "loss": 1.113, + "step": 14585 + }, + { + "epoch": 0.8862020778905159, + "grad_norm": 0.15404826402664185, + "learning_rate": 3.1838371772817286e-06, + "loss": 1.054, + "step": 14586 + }, + { + "epoch": 0.8862628349231424, + "grad_norm": 0.1660851538181305, + "learning_rate": 3.1804766591707724e-06, + "loss": 1.0693, + "step": 14587 + }, + { + "epoch": 0.8863235919557689, + "grad_norm": 0.7072659730911255, + "learning_rate": 3.177117857238493e-06, + "loss": 1.1371, + "step": 14588 + }, + { + "epoch": 0.8863843489883954, + "grad_norm": 0.18539051711559296, + "learning_rate": 3.173760771608009e-06, + "loss": 1.1068, + "step": 14589 + }, + { + "epoch": 0.8864451060210219, + "grad_norm": 0.19305945932865143, + "learning_rate": 3.1704054024023767e-06, + "loss": 1.0679, + "step": 14590 + }, + { + "epoch": 0.8865058630536484, + "grad_norm": 0.19097889959812164, + "learning_rate": 3.167051749744587e-06, + "loss": 1.0772, + "step": 14591 + }, + { + "epoch": 0.886566620086275, + "grad_norm": 0.18765684962272644, + "learning_rate": 3.16369981375757e-06, + "loss": 1.0267, + "step": 14592 + }, + { + "epoch": 0.8866273771189015, + "grad_norm": 0.20550034940242767, + "learning_rate": 3.160349594564188e-06, + "loss": 1.0768, + "step": 14593 + }, + { + "epoch": 0.886688134151528, + "grad_norm": 0.8764232993125916, + "learning_rate": 3.157001092287243e-06, + "loss": 1.0957, + "step": 14594 + }, + { + "epoch": 0.8867488911841546, + "grad_norm": 1.1572489738464355, + "learning_rate": 3.1536543070494806e-06, + "loss": 1.0592, + "step": 14595 + }, + { + "epoch": 0.8868096482167811, + "grad_norm": 0.28691262006759644, + "learning_rate": 3.1503092389735754e-06, + "loss": 1.0445, + "step": 14596 + }, + { + "epoch": 0.8868704052494076, + "grad_norm": 0.23771090805530548, + "learning_rate": 3.146965888182135e-06, + "loss": 1.1818, + "step": 14597 + }, + { + "epoch": 0.8869311622820342, + "grad_norm": 0.45283403992652893, + "learning_rate": 3.1436242547977225e-06, + "loss": 1.2117, + "step": 14598 + }, + { + "epoch": 0.8869919193146607, + "grad_norm": 0.25190654397010803, + "learning_rate": 3.140284338942817e-06, + "loss": 1.1286, + "step": 14599 + }, + { + "epoch": 0.8870526763472872, + "grad_norm": 1.2471692562103271, + "learning_rate": 3.1369461407398495e-06, + "loss": 1.086, + "step": 14600 + }, + { + "epoch": 0.8871134333799138, + "grad_norm": 0.14307160675525665, + "learning_rate": 3.133609660311171e-06, + "loss": 0.9591, + "step": 14601 + }, + { + "epoch": 0.8871741904125402, + "grad_norm": 0.13872161507606506, + "learning_rate": 3.1302748977791007e-06, + "loss": 1.0018, + "step": 14602 + }, + { + "epoch": 0.8872349474451667, + "grad_norm": 0.15229426324367523, + "learning_rate": 3.1269418532658635e-06, + "loss": 1.1261, + "step": 14603 + }, + { + "epoch": 0.8872957044777933, + "grad_norm": 0.19084076583385468, + "learning_rate": 3.1236105268936277e-06, + "loss": 1.1312, + "step": 14604 + }, + { + "epoch": 0.8873564615104198, + "grad_norm": 0.12702147662639618, + "learning_rate": 3.1202809187845016e-06, + "loss": 0.9949, + "step": 14605 + }, + { + "epoch": 0.8874172185430463, + "grad_norm": 0.29680004715919495, + "learning_rate": 3.1169530290605486e-06, + "loss": 1.258, + "step": 14606 + }, + { + "epoch": 0.8874779755756729, + "grad_norm": 0.13634715974330902, + "learning_rate": 3.113626857843743e-06, + "loss": 1.0111, + "step": 14607 + }, + { + "epoch": 0.8875387326082994, + "grad_norm": 0.20294725894927979, + "learning_rate": 3.1103024052560105e-06, + "loss": 1.1181, + "step": 14608 + }, + { + "epoch": 0.8875994896409259, + "grad_norm": 0.4525616765022278, + "learning_rate": 3.1069796714192132e-06, + "loss": 1.051, + "step": 14609 + }, + { + "epoch": 0.8876602466735525, + "grad_norm": 0.1778891235589981, + "learning_rate": 3.1036586564551273e-06, + "loss": 0.9815, + "step": 14610 + }, + { + "epoch": 0.887721003706179, + "grad_norm": 0.18693867325782776, + "learning_rate": 3.1003393604855047e-06, + "loss": 1.0867, + "step": 14611 + }, + { + "epoch": 0.8877817607388055, + "grad_norm": 0.16011813282966614, + "learning_rate": 3.097021783632009e-06, + "loss": 1.019, + "step": 14612 + }, + { + "epoch": 0.8878425177714321, + "grad_norm": 0.1639193445444107, + "learning_rate": 3.0937059260162438e-06, + "loss": 1.0231, + "step": 14613 + }, + { + "epoch": 0.8879032748040586, + "grad_norm": 0.157588392496109, + "learning_rate": 3.0903917877597556e-06, + "loss": 1.0542, + "step": 14614 + }, + { + "epoch": 0.887964031836685, + "grad_norm": 0.27615031599998474, + "learning_rate": 3.087079368984025e-06, + "loss": 1.1033, + "step": 14615 + }, + { + "epoch": 0.8880247888693116, + "grad_norm": 0.25622040033340454, + "learning_rate": 3.0837686698104663e-06, + "loss": 1.1849, + "step": 14616 + }, + { + "epoch": 0.8880855459019381, + "grad_norm": 0.25379499793052673, + "learning_rate": 3.0804596903604376e-06, + "loss": 1.0419, + "step": 14617 + }, + { + "epoch": 0.8881463029345646, + "grad_norm": 0.20628035068511963, + "learning_rate": 3.0771524307552256e-06, + "loss": 1.0537, + "step": 14618 + }, + { + "epoch": 0.8882070599671912, + "grad_norm": 0.3097895383834839, + "learning_rate": 3.0738468911160614e-06, + "loss": 1.0353, + "step": 14619 + }, + { + "epoch": 0.8882678169998177, + "grad_norm": 0.12198124080896378, + "learning_rate": 3.0705430715641135e-06, + "loss": 0.9991, + "step": 14620 + }, + { + "epoch": 0.8883285740324443, + "grad_norm": 0.18162652850151062, + "learning_rate": 3.06724097222047e-06, + "loss": 1.0787, + "step": 14621 + }, + { + "epoch": 0.8883893310650708, + "grad_norm": 0.1640348583459854, + "learning_rate": 3.0639405932061994e-06, + "loss": 0.9964, + "step": 14622 + }, + { + "epoch": 0.8884500880976973, + "grad_norm": 0.13682705163955688, + "learning_rate": 3.0606419346422444e-06, + "loss": 1.0282, + "step": 14623 + }, + { + "epoch": 0.8885108451303239, + "grad_norm": 0.23875309526920319, + "learning_rate": 3.057344996649536e-06, + "loss": 1.1158, + "step": 14624 + }, + { + "epoch": 0.8885716021629504, + "grad_norm": 0.15083160996437073, + "learning_rate": 3.054049779348922e-06, + "loss": 1.0208, + "step": 14625 + }, + { + "epoch": 0.8886323591955769, + "grad_norm": 0.21252503991127014, + "learning_rate": 3.050756282861178e-06, + "loss": 1.0655, + "step": 14626 + }, + { + "epoch": 0.8886931162282035, + "grad_norm": 1.007609248161316, + "learning_rate": 3.0474645073070517e-06, + "loss": 1.1828, + "step": 14627 + }, + { + "epoch": 0.88875387326083, + "grad_norm": 0.1680469512939453, + "learning_rate": 3.04417445280718e-06, + "loss": 1.0828, + "step": 14628 + }, + { + "epoch": 0.8888146302934564, + "grad_norm": 0.20705044269561768, + "learning_rate": 3.040886119482184e-06, + "loss": 0.967, + "step": 14629 + }, + { + "epoch": 0.888875387326083, + "grad_norm": 0.21193023025989532, + "learning_rate": 3.0375995074525764e-06, + "loss": 1.0842, + "step": 14630 + }, + { + "epoch": 0.8889361443587095, + "grad_norm": 0.16255468130111694, + "learning_rate": 3.0343146168388235e-06, + "loss": 1.0606, + "step": 14631 + }, + { + "epoch": 0.888996901391336, + "grad_norm": 0.14345093071460724, + "learning_rate": 3.031031447761362e-06, + "loss": 1.0364, + "step": 14632 + }, + { + "epoch": 0.8890576584239626, + "grad_norm": 0.24496330320835114, + "learning_rate": 3.0277500003405178e-06, + "loss": 1.1021, + "step": 14633 + }, + { + "epoch": 0.8891184154565891, + "grad_norm": 0.26732370257377625, + "learning_rate": 3.0244702746965723e-06, + "loss": 1.0169, + "step": 14634 + }, + { + "epoch": 0.8891791724892156, + "grad_norm": 0.21916088461875916, + "learning_rate": 3.021192270949763e-06, + "loss": 0.9834, + "step": 14635 + }, + { + "epoch": 0.8892399295218422, + "grad_norm": 0.18948239088058472, + "learning_rate": 3.0179159892202157e-06, + "loss": 1.0255, + "step": 14636 + }, + { + "epoch": 0.8893006865544687, + "grad_norm": 0.24425023794174194, + "learning_rate": 3.0146414296280454e-06, + "loss": 1.1384, + "step": 14637 + }, + { + "epoch": 0.8893614435870952, + "grad_norm": 0.18844151496887207, + "learning_rate": 3.0113685922932734e-06, + "loss": 1.0203, + "step": 14638 + }, + { + "epoch": 0.8894222006197218, + "grad_norm": 0.13320310413837433, + "learning_rate": 3.008097477335875e-06, + "loss": 1.0279, + "step": 14639 + }, + { + "epoch": 0.8894829576523483, + "grad_norm": 0.13494965434074402, + "learning_rate": 3.004828084875744e-06, + "loss": 1.0038, + "step": 14640 + }, + { + "epoch": 0.8895437146849748, + "grad_norm": 0.35109448432922363, + "learning_rate": 3.001560415032717e-06, + "loss": 1.1135, + "step": 14641 + }, + { + "epoch": 0.8896044717176013, + "grad_norm": 0.48480701446533203, + "learning_rate": 2.9982944679265877e-06, + "loss": 1.1258, + "step": 14642 + }, + { + "epoch": 0.8896652287502278, + "grad_norm": 0.2916924059391022, + "learning_rate": 2.995030243677055e-06, + "loss": 1.2358, + "step": 14643 + }, + { + "epoch": 0.8897259857828543, + "grad_norm": 0.18133772909641266, + "learning_rate": 2.9917677424037725e-06, + "loss": 1.0717, + "step": 14644 + }, + { + "epoch": 0.8897867428154809, + "grad_norm": 0.11694946885108948, + "learning_rate": 2.988506964226334e-06, + "loss": 1.0092, + "step": 14645 + }, + { + "epoch": 0.8898474998481074, + "grad_norm": 0.16095174849033356, + "learning_rate": 2.985247909264255e-06, + "loss": 1.0185, + "step": 14646 + }, + { + "epoch": 0.8899082568807339, + "grad_norm": 0.29381459951400757, + "learning_rate": 2.9819905776370006e-06, + "loss": 1.0949, + "step": 14647 + }, + { + "epoch": 0.8899690139133605, + "grad_norm": 0.12396300584077835, + "learning_rate": 2.97873496946397e-06, + "loss": 0.9819, + "step": 14648 + }, + { + "epoch": 0.890029770945987, + "grad_norm": 0.226085364818573, + "learning_rate": 2.975481084864512e-06, + "loss": 1.0771, + "step": 14649 + }, + { + "epoch": 0.8900905279786135, + "grad_norm": 0.136358842253685, + "learning_rate": 2.972228923957876e-06, + "loss": 0.9803, + "step": 14650 + }, + { + "epoch": 0.8901512850112401, + "grad_norm": 0.12932565808296204, + "learning_rate": 2.9689784868632773e-06, + "loss": 0.9882, + "step": 14651 + }, + { + "epoch": 0.8902120420438666, + "grad_norm": 0.16891126334667206, + "learning_rate": 2.9657297736998548e-06, + "loss": 1.068, + "step": 14652 + }, + { + "epoch": 0.8902727990764931, + "grad_norm": 0.18841731548309326, + "learning_rate": 2.9624827845867065e-06, + "loss": 1.1065, + "step": 14653 + }, + { + "epoch": 0.8903335561091197, + "grad_norm": 0.29139041900634766, + "learning_rate": 2.9592375196428436e-06, + "loss": 1.0975, + "step": 14654 + }, + { + "epoch": 0.8903943131417461, + "grad_norm": 0.16195866465568542, + "learning_rate": 2.955993978987226e-06, + "loss": 1.0268, + "step": 14655 + }, + { + "epoch": 0.8904550701743726, + "grad_norm": 0.2590111196041107, + "learning_rate": 2.9527521627387476e-06, + "loss": 1.1959, + "step": 14656 + }, + { + "epoch": 0.8905158272069992, + "grad_norm": 0.5179414749145508, + "learning_rate": 2.949512071016214e-06, + "loss": 1.1273, + "step": 14657 + }, + { + "epoch": 0.8905765842396257, + "grad_norm": 0.34335848689079285, + "learning_rate": 2.9462737039384236e-06, + "loss": 1.0471, + "step": 14658 + }, + { + "epoch": 0.8906373412722522, + "grad_norm": 0.3361416161060333, + "learning_rate": 2.9430370616240656e-06, + "loss": 1.0358, + "step": 14659 + }, + { + "epoch": 0.8906980983048788, + "grad_norm": 0.2579076588153839, + "learning_rate": 2.9398021441917724e-06, + "loss": 1.0991, + "step": 14660 + }, + { + "epoch": 0.8907588553375053, + "grad_norm": 0.1831299364566803, + "learning_rate": 2.936568951760138e-06, + "loss": 0.9916, + "step": 14661 + }, + { + "epoch": 0.8908196123701319, + "grad_norm": 0.212103933095932, + "learning_rate": 2.933337484447657e-06, + "loss": 1.1712, + "step": 14662 + }, + { + "epoch": 0.8908803694027584, + "grad_norm": 0.14105215668678284, + "learning_rate": 2.9301077423727953e-06, + "loss": 1.0342, + "step": 14663 + }, + { + "epoch": 0.8909411264353849, + "grad_norm": 1.9712475538253784, + "learning_rate": 2.9268797256539305e-06, + "loss": 1.0534, + "step": 14664 + }, + { + "epoch": 0.8910018834680115, + "grad_norm": 0.11939506232738495, + "learning_rate": 2.9236534344093904e-06, + "loss": 1.0031, + "step": 14665 + }, + { + "epoch": 0.891062640500638, + "grad_norm": 0.12716372311115265, + "learning_rate": 2.92042886875743e-06, + "loss": 1.0399, + "step": 14666 + }, + { + "epoch": 0.8911233975332645, + "grad_norm": 0.1809227168560028, + "learning_rate": 2.9172060288162493e-06, + "loss": 1.1505, + "step": 14667 + }, + { + "epoch": 0.891184154565891, + "grad_norm": 0.1774737536907196, + "learning_rate": 2.913984914703977e-06, + "loss": 1.1326, + "step": 14668 + }, + { + "epoch": 0.8912449115985175, + "grad_norm": 0.3677569031715393, + "learning_rate": 2.9107655265387e-06, + "loss": 1.2055, + "step": 14669 + }, + { + "epoch": 0.891305668631144, + "grad_norm": 0.19066473841667175, + "learning_rate": 2.9075478644384146e-06, + "loss": 1.105, + "step": 14670 + }, + { + "epoch": 0.8913664256637706, + "grad_norm": 0.1584310084581375, + "learning_rate": 2.904331928521059e-06, + "loss": 1.0031, + "step": 14671 + }, + { + "epoch": 0.8914271826963971, + "grad_norm": 0.6721485257148743, + "learning_rate": 2.9011177189045225e-06, + "loss": 1.0726, + "step": 14672 + }, + { + "epoch": 0.8914879397290236, + "grad_norm": 0.1390460729598999, + "learning_rate": 2.8979052357066104e-06, + "loss": 1.0487, + "step": 14673 + }, + { + "epoch": 0.8915486967616502, + "grad_norm": 0.1984427273273468, + "learning_rate": 2.894694479045096e-06, + "loss": 1.0478, + "step": 14674 + }, + { + "epoch": 0.8916094537942767, + "grad_norm": 0.12097391486167908, + "learning_rate": 2.891485449037662e-06, + "loss": 1.028, + "step": 14675 + }, + { + "epoch": 0.8916702108269032, + "grad_norm": 0.16790391504764557, + "learning_rate": 2.8882781458019426e-06, + "loss": 1.0489, + "step": 14676 + }, + { + "epoch": 0.8917309678595298, + "grad_norm": 0.1218056008219719, + "learning_rate": 2.8850725694554827e-06, + "loss": 1.0118, + "step": 14677 + }, + { + "epoch": 0.8917917248921563, + "grad_norm": 0.15531152486801147, + "learning_rate": 2.881868720115788e-06, + "loss": 1.1056, + "step": 14678 + }, + { + "epoch": 0.8918524819247828, + "grad_norm": 0.1883462518453598, + "learning_rate": 2.8786665979003146e-06, + "loss": 0.9999, + "step": 14679 + }, + { + "epoch": 0.8919132389574094, + "grad_norm": 0.12468085438013077, + "learning_rate": 2.8754662029264248e-06, + "loss": 1.0507, + "step": 14680 + }, + { + "epoch": 0.8919739959900358, + "grad_norm": 0.2806541919708252, + "learning_rate": 2.8722675353114236e-06, + "loss": 1.1313, + "step": 14681 + }, + { + "epoch": 0.8920347530226623, + "grad_norm": 0.11890199780464172, + "learning_rate": 2.869070595172574e-06, + "loss": 0.9651, + "step": 14682 + }, + { + "epoch": 0.8920955100552889, + "grad_norm": 0.25475019216537476, + "learning_rate": 2.8658753826270477e-06, + "loss": 1.084, + "step": 14683 + }, + { + "epoch": 0.8921562670879154, + "grad_norm": 0.18632656335830688, + "learning_rate": 2.862681897791969e-06, + "loss": 1.1254, + "step": 14684 + }, + { + "epoch": 0.8922170241205419, + "grad_norm": 0.13907262682914734, + "learning_rate": 2.859490140784399e-06, + "loss": 1.0247, + "step": 14685 + }, + { + "epoch": 0.8922777811531685, + "grad_norm": 0.15254278481006622, + "learning_rate": 2.856300111721327e-06, + "loss": 1.1379, + "step": 14686 + }, + { + "epoch": 0.892338538185795, + "grad_norm": 0.19251669943332672, + "learning_rate": 2.853111810719694e-06, + "loss": 1.0787, + "step": 14687 + }, + { + "epoch": 0.8923992952184215, + "grad_norm": 0.17783202230930328, + "learning_rate": 2.849925237896356e-06, + "loss": 1.0844, + "step": 14688 + }, + { + "epoch": 0.8924600522510481, + "grad_norm": 0.17330402135849, + "learning_rate": 2.8467403933681193e-06, + "loss": 0.9935, + "step": 14689 + }, + { + "epoch": 0.8925208092836746, + "grad_norm": 0.1208832785487175, + "learning_rate": 2.843557277251735e-06, + "loss": 1.0162, + "step": 14690 + }, + { + "epoch": 0.8925815663163011, + "grad_norm": 0.21591010689735413, + "learning_rate": 2.840375889663871e-06, + "loss": 1.0866, + "step": 14691 + }, + { + "epoch": 0.8926423233489277, + "grad_norm": 0.20209206640720367, + "learning_rate": 2.8371962307211395e-06, + "loss": 1.0766, + "step": 14692 + }, + { + "epoch": 0.8927030803815542, + "grad_norm": 0.18249891698360443, + "learning_rate": 2.834018300540103e-06, + "loss": 1.0187, + "step": 14693 + }, + { + "epoch": 0.8927638374141806, + "grad_norm": 0.12931036949157715, + "learning_rate": 2.8308420992372298e-06, + "loss": 1.0431, + "step": 14694 + }, + { + "epoch": 0.8928245944468072, + "grad_norm": 0.24115125834941864, + "learning_rate": 2.8276676269289704e-06, + "loss": 1.0558, + "step": 14695 + }, + { + "epoch": 0.8928853514794337, + "grad_norm": 0.20680926740169525, + "learning_rate": 2.8244948837316764e-06, + "loss": 1.0751, + "step": 14696 + }, + { + "epoch": 0.8929461085120602, + "grad_norm": 0.2144317328929901, + "learning_rate": 2.821323869761633e-06, + "loss": 1.0799, + "step": 14697 + }, + { + "epoch": 0.8930068655446868, + "grad_norm": 0.14680685102939606, + "learning_rate": 2.81815458513508e-06, + "loss": 1.0375, + "step": 14698 + }, + { + "epoch": 0.8930676225773133, + "grad_norm": 0.21481108665466309, + "learning_rate": 2.8149870299681913e-06, + "loss": 1.1855, + "step": 14699 + }, + { + "epoch": 0.8931283796099398, + "grad_norm": 0.14212164282798767, + "learning_rate": 2.8118212043770798e-06, + "loss": 1.0576, + "step": 14700 + }, + { + "epoch": 0.8931891366425664, + "grad_norm": 0.2129441797733307, + "learning_rate": 2.80865710847778e-06, + "loss": 1.1025, + "step": 14701 + }, + { + "epoch": 0.8932498936751929, + "grad_norm": 0.20770376920700073, + "learning_rate": 2.8054947423862833e-06, + "loss": 1.0618, + "step": 14702 + }, + { + "epoch": 0.8933106507078195, + "grad_norm": 0.1531221717596054, + "learning_rate": 2.802334106218485e-06, + "loss": 1.0818, + "step": 14703 + }, + { + "epoch": 0.893371407740446, + "grad_norm": 0.2183615267276764, + "learning_rate": 2.7991752000902595e-06, + "loss": 1.035, + "step": 14704 + }, + { + "epoch": 0.8934321647730725, + "grad_norm": 0.29008588194847107, + "learning_rate": 2.7960180241173917e-06, + "loss": 1.0783, + "step": 14705 + }, + { + "epoch": 0.8934929218056991, + "grad_norm": 0.13818205893039703, + "learning_rate": 2.7928625784156115e-06, + "loss": 1.0345, + "step": 14706 + }, + { + "epoch": 0.8935536788383255, + "grad_norm": 0.10830472409725189, + "learning_rate": 2.789708863100571e-06, + "loss": 1.0509, + "step": 14707 + }, + { + "epoch": 0.893614435870952, + "grad_norm": 0.23321640491485596, + "learning_rate": 2.7865568782878826e-06, + "loss": 1.0403, + "step": 14708 + }, + { + "epoch": 0.8936751929035786, + "grad_norm": 0.1814018189907074, + "learning_rate": 2.783406624093077e-06, + "loss": 1.084, + "step": 14709 + }, + { + "epoch": 0.8937359499362051, + "grad_norm": 1.3585861921310425, + "learning_rate": 2.7802581006316273e-06, + "loss": 1.0353, + "step": 14710 + }, + { + "epoch": 0.8937967069688316, + "grad_norm": 0.1805654615163803, + "learning_rate": 2.777111308018948e-06, + "loss": 1.0885, + "step": 14711 + }, + { + "epoch": 0.8938574640014582, + "grad_norm": 0.27119383215904236, + "learning_rate": 2.7739662463703743e-06, + "loss": 1.1525, + "step": 14712 + }, + { + "epoch": 0.8939182210340847, + "grad_norm": 0.19312453269958496, + "learning_rate": 2.770822915801202e-06, + "loss": 1.0375, + "step": 14713 + }, + { + "epoch": 0.8939789780667112, + "grad_norm": 0.1984638273715973, + "learning_rate": 2.76768131642664e-06, + "loss": 1.0195, + "step": 14714 + }, + { + "epoch": 0.8940397350993378, + "grad_norm": 0.15525318682193756, + "learning_rate": 2.7645414483618516e-06, + "loss": 1.0397, + "step": 14715 + }, + { + "epoch": 0.8941004921319643, + "grad_norm": 0.14937272667884827, + "learning_rate": 2.7614033117219327e-06, + "loss": 1.0667, + "step": 14716 + }, + { + "epoch": 0.8941612491645908, + "grad_norm": 0.13210071623325348, + "learning_rate": 2.7582669066219034e-06, + "loss": 1.0014, + "step": 14717 + }, + { + "epoch": 0.8942220061972174, + "grad_norm": 0.16811108589172363, + "learning_rate": 2.7551322331767326e-06, + "loss": 1.0456, + "step": 14718 + }, + { + "epoch": 0.8942827632298439, + "grad_norm": 0.8649924397468567, + "learning_rate": 2.7519992915013115e-06, + "loss": 0.9627, + "step": 14719 + }, + { + "epoch": 0.8943435202624703, + "grad_norm": 1.4538787603378296, + "learning_rate": 2.7488680817104982e-06, + "loss": 1.0653, + "step": 14720 + }, + { + "epoch": 0.8944042772950969, + "grad_norm": 0.2882120907306671, + "learning_rate": 2.7457386039190568e-06, + "loss": 1.0487, + "step": 14721 + }, + { + "epoch": 0.8944650343277234, + "grad_norm": 0.1604241579771042, + "learning_rate": 2.742610858241712e-06, + "loss": 1.0092, + "step": 14722 + }, + { + "epoch": 0.8945257913603499, + "grad_norm": 0.1570860743522644, + "learning_rate": 2.7394848447930887e-06, + "loss": 1.2029, + "step": 14723 + }, + { + "epoch": 0.8945865483929765, + "grad_norm": 0.2321135252714157, + "learning_rate": 2.736360563687779e-06, + "loss": 1.1652, + "step": 14724 + }, + { + "epoch": 0.894647305425603, + "grad_norm": 0.1689106971025467, + "learning_rate": 2.7332380150403182e-06, + "loss": 1.1016, + "step": 14725 + }, + { + "epoch": 0.8947080624582295, + "grad_norm": 0.2418767362833023, + "learning_rate": 2.730117198965154e-06, + "loss": 1.093, + "step": 14726 + }, + { + "epoch": 0.8947688194908561, + "grad_norm": 0.5828955173492432, + "learning_rate": 2.726998115576679e-06, + "loss": 1.2434, + "step": 14727 + }, + { + "epoch": 0.8948295765234826, + "grad_norm": 0.18136967718601227, + "learning_rate": 2.7238807649892226e-06, + "loss": 1.0632, + "step": 14728 + }, + { + "epoch": 0.8948903335561091, + "grad_norm": 0.14071328938007355, + "learning_rate": 2.7207651473170615e-06, + "loss": 1.0689, + "step": 14729 + }, + { + "epoch": 0.8949510905887357, + "grad_norm": 0.6226828694343567, + "learning_rate": 2.7176512626743867e-06, + "loss": 1.0211, + "step": 14730 + }, + { + "epoch": 0.8950118476213622, + "grad_norm": 1.1389577388763428, + "learning_rate": 2.714539111175346e-06, + "loss": 1.0293, + "step": 14731 + }, + { + "epoch": 0.8950726046539887, + "grad_norm": 0.2055598348379135, + "learning_rate": 2.7114286929340206e-06, + "loss": 1.1713, + "step": 14732 + }, + { + "epoch": 0.8951333616866153, + "grad_norm": 0.2627527713775635, + "learning_rate": 2.7083200080644077e-06, + "loss": 1.1033, + "step": 14733 + }, + { + "epoch": 0.8951941187192417, + "grad_norm": 0.18842647969722748, + "learning_rate": 2.7052130566804723e-06, + "loss": 1.0335, + "step": 14734 + }, + { + "epoch": 0.8952548757518682, + "grad_norm": 0.14938762784004211, + "learning_rate": 2.702107838896084e-06, + "loss": 1.0739, + "step": 14735 + }, + { + "epoch": 0.8953156327844948, + "grad_norm": 0.20101863145828247, + "learning_rate": 2.699004354825091e-06, + "loss": 1.0463, + "step": 14736 + }, + { + "epoch": 0.8953763898171213, + "grad_norm": 0.20890556275844574, + "learning_rate": 2.69590260458123e-06, + "loss": 1.0924, + "step": 14737 + }, + { + "epoch": 0.8954371468497478, + "grad_norm": 0.1667727828025818, + "learning_rate": 2.6928025882782037e-06, + "loss": 1.0282, + "step": 14738 + }, + { + "epoch": 0.8954979038823744, + "grad_norm": 0.6812107563018799, + "learning_rate": 2.689704306029639e-06, + "loss": 1.2909, + "step": 14739 + }, + { + "epoch": 0.8955586609150009, + "grad_norm": 0.17004570364952087, + "learning_rate": 2.6866077579491054e-06, + "loss": 1.0464, + "step": 14740 + }, + { + "epoch": 0.8956194179476274, + "grad_norm": 0.12152669578790665, + "learning_rate": 2.6835129441501126e-06, + "loss": 1.0038, + "step": 14741 + }, + { + "epoch": 0.895680174980254, + "grad_norm": 0.14214549958705902, + "learning_rate": 2.6804198647461033e-06, + "loss": 1.0504, + "step": 14742 + }, + { + "epoch": 0.8957409320128805, + "grad_norm": 0.19423016905784607, + "learning_rate": 2.677328519850453e-06, + "loss": 1.1029, + "step": 14743 + }, + { + "epoch": 0.895801689045507, + "grad_norm": 0.12794411182403564, + "learning_rate": 2.674238909576471e-06, + "loss": 1.0113, + "step": 14744 + }, + { + "epoch": 0.8958624460781336, + "grad_norm": 1.454732894897461, + "learning_rate": 2.671151034037395e-06, + "loss": 1.075, + "step": 14745 + }, + { + "epoch": 0.8959232031107601, + "grad_norm": 0.6005421876907349, + "learning_rate": 2.668064893346439e-06, + "loss": 1.192, + "step": 14746 + }, + { + "epoch": 0.8959839601433865, + "grad_norm": 0.2230108678340912, + "learning_rate": 2.664980487616714e-06, + "loss": 1.1067, + "step": 14747 + }, + { + "epoch": 0.8960447171760131, + "grad_norm": 0.1403961479663849, + "learning_rate": 2.6618978169612786e-06, + "loss": 1.0047, + "step": 14748 + }, + { + "epoch": 0.8961054742086396, + "grad_norm": 0.14917153120040894, + "learning_rate": 2.658816881493131e-06, + "loss": 1.0603, + "step": 14749 + }, + { + "epoch": 0.8961662312412662, + "grad_norm": 0.18496359884738922, + "learning_rate": 2.6557376813251934e-06, + "loss": 1.0266, + "step": 14750 + }, + { + "epoch": 0.8962269882738927, + "grad_norm": 0.12762930989265442, + "learning_rate": 2.6526602165703464e-06, + "loss": 1.0206, + "step": 14751 + }, + { + "epoch": 0.8962877453065192, + "grad_norm": 0.14046984910964966, + "learning_rate": 2.6495844873413943e-06, + "loss": 1.0399, + "step": 14752 + }, + { + "epoch": 0.8963485023391458, + "grad_norm": 0.19891126453876495, + "learning_rate": 2.64651049375107e-06, + "loss": 1.1557, + "step": 14753 + }, + { + "epoch": 0.8964092593717723, + "grad_norm": 0.2858726680278778, + "learning_rate": 2.6434382359120604e-06, + "loss": 1.0686, + "step": 14754 + }, + { + "epoch": 0.8964700164043988, + "grad_norm": 0.15749414265155792, + "learning_rate": 2.6403677139369754e-06, + "loss": 1.0697, + "step": 14755 + }, + { + "epoch": 0.8965307734370254, + "grad_norm": 0.1980634480714798, + "learning_rate": 2.63729892793837e-06, + "loss": 1.0718, + "step": 14756 + }, + { + "epoch": 0.8965915304696519, + "grad_norm": 0.20071861147880554, + "learning_rate": 2.6342318780287203e-06, + "loss": 1.1328, + "step": 14757 + }, + { + "epoch": 0.8966522875022784, + "grad_norm": 0.21230880916118622, + "learning_rate": 2.6311665643204586e-06, + "loss": 1.0346, + "step": 14758 + }, + { + "epoch": 0.896713044534905, + "grad_norm": 1.0529778003692627, + "learning_rate": 2.6281029869259454e-06, + "loss": 1.0907, + "step": 14759 + }, + { + "epoch": 0.8967738015675314, + "grad_norm": 0.11792907863855362, + "learning_rate": 2.625041145957474e-06, + "loss": 1.0112, + "step": 14760 + }, + { + "epoch": 0.8968345586001579, + "grad_norm": 0.18067672848701477, + "learning_rate": 2.6219810415272714e-06, + "loss": 1.0771, + "step": 14761 + }, + { + "epoch": 0.8968953156327845, + "grad_norm": 0.1860075294971466, + "learning_rate": 2.6189226737475146e-06, + "loss": 1.0318, + "step": 14762 + }, + { + "epoch": 0.896956072665411, + "grad_norm": 0.13522951304912567, + "learning_rate": 2.615866042730314e-06, + "loss": 1.0049, + "step": 14763 + }, + { + "epoch": 0.8970168296980375, + "grad_norm": 0.802241861820221, + "learning_rate": 2.612811148587696e-06, + "loss": 1.0559, + "step": 14764 + }, + { + "epoch": 0.8970775867306641, + "grad_norm": 0.11852771043777466, + "learning_rate": 2.609757991431644e-06, + "loss": 0.975, + "step": 14765 + }, + { + "epoch": 0.8971383437632906, + "grad_norm": 0.11643736064434052, + "learning_rate": 2.6067065713740677e-06, + "loss": 1.0598, + "step": 14766 + }, + { + "epoch": 0.8971991007959171, + "grad_norm": 0.2370849847793579, + "learning_rate": 2.6036568885268287e-06, + "loss": 1.1074, + "step": 14767 + }, + { + "epoch": 0.8972598578285437, + "grad_norm": 0.18365634977817535, + "learning_rate": 2.600608943001709e-06, + "loss": 1.0971, + "step": 14768 + }, + { + "epoch": 0.8973206148611702, + "grad_norm": 0.131282240152359, + "learning_rate": 2.597562734910436e-06, + "loss": 1.0038, + "step": 14769 + }, + { + "epoch": 0.8973813718937967, + "grad_norm": 0.26499810814857483, + "learning_rate": 2.594518264364659e-06, + "loss": 1.0808, + "step": 14770 + }, + { + "epoch": 0.8974421289264233, + "grad_norm": 2.061115026473999, + "learning_rate": 2.591475531475973e-06, + "loss": 1.21, + "step": 14771 + }, + { + "epoch": 0.8975028859590498, + "grad_norm": 0.2766638398170471, + "learning_rate": 2.588434536355916e-06, + "loss": 1.1796, + "step": 14772 + }, + { + "epoch": 0.8975636429916762, + "grad_norm": 0.2073231190443039, + "learning_rate": 2.58539527911596e-06, + "loss": 1.1233, + "step": 14773 + }, + { + "epoch": 0.8976244000243028, + "grad_norm": 0.5310015082359314, + "learning_rate": 2.5823577598675043e-06, + "loss": 1.2761, + "step": 14774 + }, + { + "epoch": 0.8976851570569293, + "grad_norm": 0.17726966738700867, + "learning_rate": 2.579321978721888e-06, + "loss": 1.0997, + "step": 14775 + }, + { + "epoch": 0.8977459140895558, + "grad_norm": 0.18836133182048798, + "learning_rate": 2.5762879357903946e-06, + "loss": 1.0742, + "step": 14776 + }, + { + "epoch": 0.8978066711221824, + "grad_norm": 0.1266496479511261, + "learning_rate": 2.573255631184235e-06, + "loss": 1.0104, + "step": 14777 + }, + { + "epoch": 0.8978674281548089, + "grad_norm": 2.7757680416107178, + "learning_rate": 2.570225065014553e-06, + "loss": 0.9968, + "step": 14778 + }, + { + "epoch": 0.8979281851874354, + "grad_norm": 2.23264479637146, + "learning_rate": 2.5671962373924384e-06, + "loss": 1.1416, + "step": 14779 + }, + { + "epoch": 0.897988942220062, + "grad_norm": 0.217926025390625, + "learning_rate": 2.5641691484289187e-06, + "loss": 1.0716, + "step": 14780 + }, + { + "epoch": 0.8980496992526885, + "grad_norm": 0.15174643695354462, + "learning_rate": 2.5611437982349493e-06, + "loss": 1.0065, + "step": 14781 + }, + { + "epoch": 0.898110456285315, + "grad_norm": 0.18801817297935486, + "learning_rate": 2.5581201869214087e-06, + "loss": 1.0526, + "step": 14782 + }, + { + "epoch": 0.8981712133179416, + "grad_norm": 0.25780627131462097, + "learning_rate": 2.5550983145991635e-06, + "loss": 1.0354, + "step": 14783 + }, + { + "epoch": 0.8982319703505681, + "grad_norm": 0.220517098903656, + "learning_rate": 2.552078181378953e-06, + "loss": 1.0671, + "step": 14784 + }, + { + "epoch": 0.8982927273831947, + "grad_norm": 0.21774059534072876, + "learning_rate": 2.549059787371488e-06, + "loss": 1.0813, + "step": 14785 + }, + { + "epoch": 0.8983534844158211, + "grad_norm": 0.17260508239269257, + "learning_rate": 2.546043132687409e-06, + "loss": 0.9978, + "step": 14786 + }, + { + "epoch": 0.8984142414484476, + "grad_norm": 4.440997123718262, + "learning_rate": 2.5430282174372823e-06, + "loss": 1.0123, + "step": 14787 + }, + { + "epoch": 0.8984749984810741, + "grad_norm": 0.15748123824596405, + "learning_rate": 2.5400150417316358e-06, + "loss": 1.0513, + "step": 14788 + }, + { + "epoch": 0.8985357555137007, + "grad_norm": 0.128719300031662, + "learning_rate": 2.537003605680921e-06, + "loss": 1.0407, + "step": 14789 + }, + { + "epoch": 0.8985965125463272, + "grad_norm": 0.12868525087833405, + "learning_rate": 2.5339939093955047e-06, + "loss": 1.0327, + "step": 14790 + }, + { + "epoch": 0.8986572695789538, + "grad_norm": 0.30143582820892334, + "learning_rate": 2.5309859529857095e-06, + "loss": 1.0426, + "step": 14791 + }, + { + "epoch": 0.8987180266115803, + "grad_norm": 1.6469806432724, + "learning_rate": 2.5279797365618085e-06, + "loss": 1.0, + "step": 14792 + }, + { + "epoch": 0.8987787836442068, + "grad_norm": 0.19774304330348969, + "learning_rate": 2.524975260233986e-06, + "loss": 1.0557, + "step": 14793 + }, + { + "epoch": 0.8988395406768334, + "grad_norm": 0.1630176305770874, + "learning_rate": 2.521972524112376e-06, + "loss": 1.0953, + "step": 14794 + }, + { + "epoch": 0.8989002977094599, + "grad_norm": 0.45816394686698914, + "learning_rate": 2.5189715283070346e-06, + "loss": 1.1193, + "step": 14795 + }, + { + "epoch": 0.8989610547420864, + "grad_norm": 0.1805506944656372, + "learning_rate": 2.515972272927969e-06, + "loss": 1.0684, + "step": 14796 + }, + { + "epoch": 0.899021811774713, + "grad_norm": 0.1315872222185135, + "learning_rate": 2.512974758085124e-06, + "loss": 1.0367, + "step": 14797 + }, + { + "epoch": 0.8990825688073395, + "grad_norm": 0.1244564801454544, + "learning_rate": 2.509978983888367e-06, + "loss": 1.0269, + "step": 14798 + }, + { + "epoch": 0.8991433258399659, + "grad_norm": 0.18802234530448914, + "learning_rate": 2.506984950447505e-06, + "loss": 1.0793, + "step": 14799 + }, + { + "epoch": 0.8992040828725925, + "grad_norm": 0.17557960748672485, + "learning_rate": 2.503992657872295e-06, + "loss": 1.0156, + "step": 14800 + }, + { + "epoch": 0.899264839905219, + "grad_norm": 0.7172213792800903, + "learning_rate": 2.5010021062724097e-06, + "loss": 1.1081, + "step": 14801 + }, + { + "epoch": 0.8993255969378455, + "grad_norm": 5.102267265319824, + "learning_rate": 2.498013295757479e-06, + "loss": 1.1648, + "step": 14802 + }, + { + "epoch": 0.8993863539704721, + "grad_norm": 0.297062486410141, + "learning_rate": 2.4950262264370473e-06, + "loss": 1.2159, + "step": 14803 + }, + { + "epoch": 0.8994471110030986, + "grad_norm": 0.1991564929485321, + "learning_rate": 2.4920408984206166e-06, + "loss": 1.0785, + "step": 14804 + }, + { + "epoch": 0.8995078680357251, + "grad_norm": 0.18376772105693817, + "learning_rate": 2.4890573118176107e-06, + "loss": 1.1003, + "step": 14805 + }, + { + "epoch": 0.8995686250683517, + "grad_norm": 0.31689199805259705, + "learning_rate": 2.4860754667373866e-06, + "loss": 1.1167, + "step": 14806 + }, + { + "epoch": 0.8996293821009782, + "grad_norm": 0.21827276051044464, + "learning_rate": 2.4830953632892505e-06, + "loss": 1.2769, + "step": 14807 + }, + { + "epoch": 0.8996901391336047, + "grad_norm": 0.13514092564582825, + "learning_rate": 2.4801170015824384e-06, + "loss": 1.0388, + "step": 14808 + }, + { + "epoch": 0.8997508961662313, + "grad_norm": 0.14156198501586914, + "learning_rate": 2.4771403817261285e-06, + "loss": 1.0079, + "step": 14809 + }, + { + "epoch": 0.8998116531988578, + "grad_norm": 0.130026176571846, + "learning_rate": 2.4741655038294285e-06, + "loss": 1.0591, + "step": 14810 + }, + { + "epoch": 0.8998724102314843, + "grad_norm": 0.15922057628631592, + "learning_rate": 2.471192368001374e-06, + "loss": 1.0708, + "step": 14811 + }, + { + "epoch": 0.8999331672641108, + "grad_norm": 0.26900923252105713, + "learning_rate": 2.4682209743509376e-06, + "loss": 1.0693, + "step": 14812 + }, + { + "epoch": 0.8999939242967373, + "grad_norm": 0.796972930431366, + "learning_rate": 2.4652513229870612e-06, + "loss": 1.0367, + "step": 14813 + }, + { + "epoch": 0.9000546813293638, + "grad_norm": 0.18521910905838013, + "learning_rate": 2.4622834140185846e-06, + "loss": 1.0798, + "step": 14814 + }, + { + "epoch": 0.9001154383619904, + "grad_norm": 0.2889372706413269, + "learning_rate": 2.459317247554299e-06, + "loss": 1.1317, + "step": 14815 + }, + { + "epoch": 0.9001761953946169, + "grad_norm": 0.1447586715221405, + "learning_rate": 2.4563528237029344e-06, + "loss": 1.0533, + "step": 14816 + }, + { + "epoch": 0.9002369524272434, + "grad_norm": 0.1541641354560852, + "learning_rate": 2.4533901425731365e-06, + "loss": 1.0713, + "step": 14817 + }, + { + "epoch": 0.90029770945987, + "grad_norm": 0.1650155931711197, + "learning_rate": 2.4504292042735187e-06, + "loss": 1.0672, + "step": 14818 + }, + { + "epoch": 0.9003584664924965, + "grad_norm": 0.19674576818943024, + "learning_rate": 2.4474700089126055e-06, + "loss": 1.0527, + "step": 14819 + }, + { + "epoch": 0.900419223525123, + "grad_norm": 0.31124231219291687, + "learning_rate": 2.4445125565988767e-06, + "loss": 1.1993, + "step": 14820 + }, + { + "epoch": 0.9004799805577496, + "grad_norm": 0.22260890901088715, + "learning_rate": 2.441556847440729e-06, + "loss": 1.0871, + "step": 14821 + }, + { + "epoch": 0.9005407375903761, + "grad_norm": 0.1261293888092041, + "learning_rate": 2.438602881546509e-06, + "loss": 1.0368, + "step": 14822 + }, + { + "epoch": 0.9006014946230027, + "grad_norm": 0.11667251586914062, + "learning_rate": 2.4356506590244967e-06, + "loss": 0.9958, + "step": 14823 + }, + { + "epoch": 0.9006622516556292, + "grad_norm": 0.12213541567325592, + "learning_rate": 2.4327001799828995e-06, + "loss": 1.0294, + "step": 14824 + }, + { + "epoch": 0.9007230086882556, + "grad_norm": 0.23025371134281158, + "learning_rate": 2.429751444529871e-06, + "loss": 1.0897, + "step": 14825 + }, + { + "epoch": 0.9007837657208821, + "grad_norm": 0.1594216525554657, + "learning_rate": 2.426804452773501e-06, + "loss": 1.031, + "step": 14826 + }, + { + "epoch": 0.9008445227535087, + "grad_norm": 0.1775679886341095, + "learning_rate": 2.42385920482181e-06, + "loss": 1.0848, + "step": 14827 + }, + { + "epoch": 0.9009052797861352, + "grad_norm": 0.27050432562828064, + "learning_rate": 2.420915700782744e-06, + "loss": 1.2472, + "step": 14828 + }, + { + "epoch": 0.9009660368187618, + "grad_norm": 0.1611764281988144, + "learning_rate": 2.417973940764223e-06, + "loss": 1.0956, + "step": 14829 + }, + { + "epoch": 0.9010267938513883, + "grad_norm": 0.20790152251720428, + "learning_rate": 2.415033924874066e-06, + "loss": 1.0869, + "step": 14830 + }, + { + "epoch": 0.9010875508840148, + "grad_norm": 0.19813285768032074, + "learning_rate": 2.4120956532200367e-06, + "loss": 1.02, + "step": 14831 + }, + { + "epoch": 0.9011483079166414, + "grad_norm": 0.14312927424907684, + "learning_rate": 2.409159125909838e-06, + "loss": 1.0734, + "step": 14832 + }, + { + "epoch": 0.9012090649492679, + "grad_norm": 0.34180277585983276, + "learning_rate": 2.4062243430511056e-06, + "loss": 1.0878, + "step": 14833 + }, + { + "epoch": 0.9012698219818944, + "grad_norm": 0.1445678323507309, + "learning_rate": 2.4032913047514204e-06, + "loss": 1.0229, + "step": 14834 + }, + { + "epoch": 0.901330579014521, + "grad_norm": 0.14330722391605377, + "learning_rate": 2.400360011118297e-06, + "loss": 0.9911, + "step": 14835 + }, + { + "epoch": 0.9013913360471475, + "grad_norm": 0.1386917531490326, + "learning_rate": 2.397430462259187e-06, + "loss": 1.0074, + "step": 14836 + }, + { + "epoch": 0.901452093079774, + "grad_norm": 0.3769070506095886, + "learning_rate": 2.3945026582814557e-06, + "loss": 1.1689, + "step": 14837 + }, + { + "epoch": 0.9015128501124006, + "grad_norm": 0.19499348104000092, + "learning_rate": 2.391576599292422e-06, + "loss": 1.1456, + "step": 14838 + }, + { + "epoch": 0.901573607145027, + "grad_norm": 0.19439934194087982, + "learning_rate": 2.388652285399362e-06, + "loss": 1.0836, + "step": 14839 + }, + { + "epoch": 0.9016343641776535, + "grad_norm": 0.20006129145622253, + "learning_rate": 2.3857297167094508e-06, + "loss": 1.0429, + "step": 14840 + }, + { + "epoch": 0.9016951212102801, + "grad_norm": 0.17779572308063507, + "learning_rate": 2.3828088933298243e-06, + "loss": 1.0434, + "step": 14841 + }, + { + "epoch": 0.9017558782429066, + "grad_norm": 0.24300448596477509, + "learning_rate": 2.379889815367542e-06, + "loss": 1.2262, + "step": 14842 + }, + { + "epoch": 0.9018166352755331, + "grad_norm": 0.1332509070634842, + "learning_rate": 2.3769724829296015e-06, + "loss": 1.04, + "step": 14843 + }, + { + "epoch": 0.9018773923081597, + "grad_norm": 0.1924302577972412, + "learning_rate": 2.3740568961229394e-06, + "loss": 1.0788, + "step": 14844 + }, + { + "epoch": 0.9019381493407862, + "grad_norm": 2.8674087524414062, + "learning_rate": 2.3711430550544257e-06, + "loss": 1.0419, + "step": 14845 + }, + { + "epoch": 0.9019989063734127, + "grad_norm": 0.5314818620681763, + "learning_rate": 2.3682309598308747e-06, + "loss": 1.1501, + "step": 14846 + }, + { + "epoch": 0.9020596634060393, + "grad_norm": 0.22032330930233002, + "learning_rate": 2.365320610559024e-06, + "loss": 1.0764, + "step": 14847 + }, + { + "epoch": 0.9021204204386658, + "grad_norm": 0.13184747099876404, + "learning_rate": 2.3624120073455548e-06, + "loss": 1.0399, + "step": 14848 + }, + { + "epoch": 0.9021811774712923, + "grad_norm": 0.1483260989189148, + "learning_rate": 2.35950515029707e-06, + "loss": 1.0372, + "step": 14849 + }, + { + "epoch": 0.9022419345039189, + "grad_norm": 0.13248421251773834, + "learning_rate": 2.3566000395201513e-06, + "loss": 1.0405, + "step": 14850 + }, + { + "epoch": 0.9023026915365454, + "grad_norm": 0.1914897859096527, + "learning_rate": 2.3536966751212586e-06, + "loss": 1.0933, + "step": 14851 + }, + { + "epoch": 0.9023634485691718, + "grad_norm": 0.14552147686481476, + "learning_rate": 2.3507950572068228e-06, + "loss": 1.103, + "step": 14852 + }, + { + "epoch": 0.9024242056017984, + "grad_norm": 0.1307748258113861, + "learning_rate": 2.347895185883203e-06, + "loss": 1.0045, + "step": 14853 + }, + { + "epoch": 0.9024849626344249, + "grad_norm": 0.21948905289173126, + "learning_rate": 2.344997061256693e-06, + "loss": 1.0464, + "step": 14854 + }, + { + "epoch": 0.9025457196670514, + "grad_norm": 0.16389919817447662, + "learning_rate": 2.3421006834335348e-06, + "loss": 1.0771, + "step": 14855 + }, + { + "epoch": 0.902606476699678, + "grad_norm": 0.3150233328342438, + "learning_rate": 2.3392060525198935e-06, + "loss": 1.0909, + "step": 14856 + }, + { + "epoch": 0.9026672337323045, + "grad_norm": 0.14308902621269226, + "learning_rate": 2.3363131686218563e-06, + "loss": 1.034, + "step": 14857 + }, + { + "epoch": 0.902727990764931, + "grad_norm": 0.1331058293581009, + "learning_rate": 2.333422031845478e-06, + "loss": 0.9895, + "step": 14858 + }, + { + "epoch": 0.9027887477975576, + "grad_norm": 0.2687436640262604, + "learning_rate": 2.3305326422967178e-06, + "loss": 1.0244, + "step": 14859 + }, + { + "epoch": 0.9028495048301841, + "grad_norm": 0.12817716598510742, + "learning_rate": 2.3276450000815076e-06, + "loss": 1.0132, + "step": 14860 + }, + { + "epoch": 0.9029102618628106, + "grad_norm": 0.2683033049106598, + "learning_rate": 2.324759105305685e-06, + "loss": 1.1597, + "step": 14861 + }, + { + "epoch": 0.9029710188954372, + "grad_norm": 0.1235908567905426, + "learning_rate": 2.321874958075032e-06, + "loss": 1.0275, + "step": 14862 + }, + { + "epoch": 0.9030317759280637, + "grad_norm": 0.18241475522518158, + "learning_rate": 2.318992558495281e-06, + "loss": 1.1253, + "step": 14863 + }, + { + "epoch": 0.9030925329606903, + "grad_norm": 0.12272278219461441, + "learning_rate": 2.316111906672053e-06, + "loss": 1.0431, + "step": 14864 + }, + { + "epoch": 0.9031532899933167, + "grad_norm": 0.17401783168315887, + "learning_rate": 2.3132330027109683e-06, + "loss": 1.0353, + "step": 14865 + }, + { + "epoch": 0.9032140470259432, + "grad_norm": 0.1804550439119339, + "learning_rate": 2.310355846717549e-06, + "loss": 1.0331, + "step": 14866 + }, + { + "epoch": 0.9032748040585697, + "grad_norm": 3.958571672439575, + "learning_rate": 2.3074804387972493e-06, + "loss": 1.0882, + "step": 14867 + }, + { + "epoch": 0.9033355610911963, + "grad_norm": 0.21868796646595, + "learning_rate": 2.3046067790554794e-06, + "loss": 1.1304, + "step": 14868 + }, + { + "epoch": 0.9033963181238228, + "grad_norm": 0.4179894030094147, + "learning_rate": 2.3017348675975604e-06, + "loss": 1.2843, + "step": 14869 + }, + { + "epoch": 0.9034570751564494, + "grad_norm": 0.1966385692358017, + "learning_rate": 2.2988647045287747e-06, + "loss": 1.0945, + "step": 14870 + }, + { + "epoch": 0.9035178321890759, + "grad_norm": 0.13428857922554016, + "learning_rate": 2.295996289954322e-06, + "loss": 1.0505, + "step": 14871 + }, + { + "epoch": 0.9035785892217024, + "grad_norm": 0.15862739086151123, + "learning_rate": 2.2931296239793455e-06, + "loss": 1.0884, + "step": 14872 + }, + { + "epoch": 0.903639346254329, + "grad_norm": 0.11974199116230011, + "learning_rate": 2.290264706708922e-06, + "loss": 1.0463, + "step": 14873 + }, + { + "epoch": 0.9037001032869555, + "grad_norm": 0.14039455354213715, + "learning_rate": 2.287401538248074e-06, + "loss": 1.0181, + "step": 14874 + }, + { + "epoch": 0.903760860319582, + "grad_norm": 0.24471774697303772, + "learning_rate": 2.2845401187017333e-06, + "loss": 1.2533, + "step": 14875 + }, + { + "epoch": 0.9038216173522086, + "grad_norm": 0.265040785074234, + "learning_rate": 2.2816804481748055e-06, + "loss": 1.131, + "step": 14876 + }, + { + "epoch": 0.9038823743848351, + "grad_norm": 0.2298332005739212, + "learning_rate": 2.278822526772112e-06, + "loss": 1.0913, + "step": 14877 + }, + { + "epoch": 0.9039431314174615, + "grad_norm": 0.1614462435245514, + "learning_rate": 2.2759663545983966e-06, + "loss": 1.0523, + "step": 14878 + }, + { + "epoch": 0.9040038884500881, + "grad_norm": 0.12145266681909561, + "learning_rate": 2.2731119317583593e-06, + "loss": 0.985, + "step": 14879 + }, + { + "epoch": 0.9040646454827146, + "grad_norm": 0.14107373356819153, + "learning_rate": 2.270259258356622e-06, + "loss": 1.0527, + "step": 14880 + }, + { + "epoch": 0.9041254025153411, + "grad_norm": 0.20524485409259796, + "learning_rate": 2.2674083344977615e-06, + "loss": 1.1039, + "step": 14881 + }, + { + "epoch": 0.9041861595479677, + "grad_norm": 0.16556096076965332, + "learning_rate": 2.2645591602862836e-06, + "loss": 1.0479, + "step": 14882 + }, + { + "epoch": 0.9042469165805942, + "grad_norm": 0.24907484650611877, + "learning_rate": 2.261711735826616e-06, + "loss": 1.1737, + "step": 14883 + }, + { + "epoch": 0.9043076736132207, + "grad_norm": 0.10683733224868774, + "learning_rate": 2.258866061223119e-06, + "loss": 0.9885, + "step": 14884 + }, + { + "epoch": 0.9043684306458473, + "grad_norm": 0.1292402148246765, + "learning_rate": 2.2560221365801273e-06, + "loss": 0.9811, + "step": 14885 + }, + { + "epoch": 0.9044291876784738, + "grad_norm": 0.21259796619415283, + "learning_rate": 2.2531799620018667e-06, + "loss": 1.0269, + "step": 14886 + }, + { + "epoch": 0.9044899447111003, + "grad_norm": 0.17320990562438965, + "learning_rate": 2.2503395375925275e-06, + "loss": 0.9793, + "step": 14887 + }, + { + "epoch": 0.9045507017437269, + "grad_norm": 0.2419673502445221, + "learning_rate": 2.24750086345622e-06, + "loss": 1.0805, + "step": 14888 + }, + { + "epoch": 0.9046114587763534, + "grad_norm": 0.23656171560287476, + "learning_rate": 2.2446639396970006e-06, + "loss": 1.1383, + "step": 14889 + }, + { + "epoch": 0.9046722158089799, + "grad_norm": 0.6495109796524048, + "learning_rate": 2.241828766418852e-06, + "loss": 1.1065, + "step": 14890 + }, + { + "epoch": 0.9047329728416064, + "grad_norm": 4.310469627380371, + "learning_rate": 2.238995343725703e-06, + "loss": 1.0513, + "step": 14891 + }, + { + "epoch": 0.9047937298742329, + "grad_norm": 0.1579255312681198, + "learning_rate": 2.236163671721414e-06, + "loss": 1.0577, + "step": 14892 + }, + { + "epoch": 0.9048544869068594, + "grad_norm": 0.17000891268253326, + "learning_rate": 2.233333750509775e-06, + "loss": 1.0502, + "step": 14893 + }, + { + "epoch": 0.904915243939486, + "grad_norm": 0.15944992005825043, + "learning_rate": 2.2305055801945253e-06, + "loss": 0.987, + "step": 14894 + }, + { + "epoch": 0.9049760009721125, + "grad_norm": 0.2205081284046173, + "learning_rate": 2.2276791608793257e-06, + "loss": 1.128, + "step": 14895 + }, + { + "epoch": 0.905036758004739, + "grad_norm": 0.14026787877082825, + "learning_rate": 2.2248544926677717e-06, + "loss": 1.0283, + "step": 14896 + }, + { + "epoch": 0.9050975150373656, + "grad_norm": 0.14874908328056335, + "learning_rate": 2.2220315756634245e-06, + "loss": 1.0223, + "step": 14897 + }, + { + "epoch": 0.9051582720699921, + "grad_norm": 0.20054081082344055, + "learning_rate": 2.2192104099697407e-06, + "loss": 1.0611, + "step": 14898 + }, + { + "epoch": 0.9052190291026186, + "grad_norm": 0.15943153202533722, + "learning_rate": 2.216390995690132e-06, + "loss": 1.0115, + "step": 14899 + }, + { + "epoch": 0.9052797861352452, + "grad_norm": 0.1489144116640091, + "learning_rate": 2.213573332927943e-06, + "loss": 0.9954, + "step": 14900 + }, + { + "epoch": 0.9053405431678717, + "grad_norm": 0.19145680963993073, + "learning_rate": 2.2107574217864634e-06, + "loss": 1.0781, + "step": 14901 + }, + { + "epoch": 0.9054013002004982, + "grad_norm": 0.13404956459999084, + "learning_rate": 2.2079432623689113e-06, + "loss": 0.9995, + "step": 14902 + }, + { + "epoch": 0.9054620572331248, + "grad_norm": 0.1382082998752594, + "learning_rate": 2.205130854778442e-06, + "loss": 1.0039, + "step": 14903 + }, + { + "epoch": 0.9055228142657512, + "grad_norm": 0.11300737410783768, + "learning_rate": 2.2023201991181354e-06, + "loss": 0.9973, + "step": 14904 + }, + { + "epoch": 0.9055835712983777, + "grad_norm": 0.1429503709077835, + "learning_rate": 2.1995112954910134e-06, + "loss": 1.0327, + "step": 14905 + }, + { + "epoch": 0.9056443283310043, + "grad_norm": 0.13290822505950928, + "learning_rate": 2.1967041440000558e-06, + "loss": 1.0409, + "step": 14906 + }, + { + "epoch": 0.9057050853636308, + "grad_norm": 0.12945635616779327, + "learning_rate": 2.1938987447481405e-06, + "loss": 1.0684, + "step": 14907 + }, + { + "epoch": 0.9057658423962573, + "grad_norm": 0.18463639914989471, + "learning_rate": 2.1910950978381086e-06, + "loss": 1.1517, + "step": 14908 + }, + { + "epoch": 0.9058265994288839, + "grad_norm": 0.2140742689371109, + "learning_rate": 2.188293203372732e-06, + "loss": 1.0901, + "step": 14909 + }, + { + "epoch": 0.9058873564615104, + "grad_norm": 0.19834370911121368, + "learning_rate": 2.1854930614547074e-06, + "loss": 1.0445, + "step": 14910 + }, + { + "epoch": 0.905948113494137, + "grad_norm": 10.710871696472168, + "learning_rate": 2.1826946721866747e-06, + "loss": 1.0575, + "step": 14911 + }, + { + "epoch": 0.9060088705267635, + "grad_norm": 0.23960964381694794, + "learning_rate": 2.179898035671213e-06, + "loss": 1.0535, + "step": 14912 + }, + { + "epoch": 0.90606962755939, + "grad_norm": 2.6178269386291504, + "learning_rate": 2.1771031520108343e-06, + "loss": 1.0034, + "step": 14913 + }, + { + "epoch": 0.9061303845920166, + "grad_norm": 0.2317754179239273, + "learning_rate": 2.1743100213079794e-06, + "loss": 1.1524, + "step": 14914 + }, + { + "epoch": 0.9061911416246431, + "grad_norm": 0.13025236129760742, + "learning_rate": 2.171518643665038e-06, + "loss": 1.0342, + "step": 14915 + }, + { + "epoch": 0.9062518986572696, + "grad_norm": 0.11753544956445694, + "learning_rate": 2.1687290191843178e-06, + "loss": 1.0081, + "step": 14916 + }, + { + "epoch": 0.906312655689896, + "grad_norm": 0.15852797031402588, + "learning_rate": 2.165941147968098e-06, + "loss": 1.0613, + "step": 14917 + }, + { + "epoch": 0.9063734127225226, + "grad_norm": 0.14190050959587097, + "learning_rate": 2.163155030118541e-06, + "loss": 1.0049, + "step": 14918 + }, + { + "epoch": 0.9064341697551491, + "grad_norm": 0.12560412287712097, + "learning_rate": 2.160370665737782e-06, + "loss": 1.0205, + "step": 14919 + }, + { + "epoch": 0.9064949267877757, + "grad_norm": 0.17282907664775848, + "learning_rate": 2.157588054927884e-06, + "loss": 1.092, + "step": 14920 + }, + { + "epoch": 0.9065556838204022, + "grad_norm": 0.1233811005949974, + "learning_rate": 2.1548071977908423e-06, + "loss": 1.0317, + "step": 14921 + }, + { + "epoch": 0.9066164408530287, + "grad_norm": 0.1354319155216217, + "learning_rate": 2.152028094428593e-06, + "loss": 1.012, + "step": 14922 + }, + { + "epoch": 0.9066771978856553, + "grad_norm": 0.1330280750989914, + "learning_rate": 2.1492507449430098e-06, + "loss": 1.012, + "step": 14923 + }, + { + "epoch": 0.9067379549182818, + "grad_norm": 0.229650616645813, + "learning_rate": 2.1464751494358837e-06, + "loss": 1.1302, + "step": 14924 + }, + { + "epoch": 0.9067987119509083, + "grad_norm": 0.16664597392082214, + "learning_rate": 2.1437013080089608e-06, + "loss": 1.0109, + "step": 14925 + }, + { + "epoch": 0.9068594689835349, + "grad_norm": 0.4418676197528839, + "learning_rate": 2.1409292207639096e-06, + "loss": 1.0667, + "step": 14926 + }, + { + "epoch": 0.9069202260161614, + "grad_norm": 0.5708197355270386, + "learning_rate": 2.138158887802355e-06, + "loss": 1.0158, + "step": 14927 + }, + { + "epoch": 0.9069809830487879, + "grad_norm": 0.20245254039764404, + "learning_rate": 2.1353903092258374e-06, + "loss": 1.1559, + "step": 14928 + }, + { + "epoch": 0.9070417400814145, + "grad_norm": 0.18634402751922607, + "learning_rate": 2.132623485135843e-06, + "loss": 1.051, + "step": 14929 + }, + { + "epoch": 0.9071024971140409, + "grad_norm": 0.21216164529323578, + "learning_rate": 2.1298584156337896e-06, + "loss": 1.0844, + "step": 14930 + }, + { + "epoch": 0.9071632541466674, + "grad_norm": 0.19077926874160767, + "learning_rate": 2.1270951008210192e-06, + "loss": 1.102, + "step": 14931 + }, + { + "epoch": 0.907224011179294, + "grad_norm": 0.1437380015850067, + "learning_rate": 2.1243335407988396e-06, + "loss": 1.0329, + "step": 14932 + }, + { + "epoch": 0.9072847682119205, + "grad_norm": 0.1513354480266571, + "learning_rate": 2.121573735668464e-06, + "loss": 0.9998, + "step": 14933 + }, + { + "epoch": 0.907345525244547, + "grad_norm": 0.12943045794963837, + "learning_rate": 2.1188156855310613e-06, + "loss": 1.0073, + "step": 14934 + }, + { + "epoch": 0.9074062822771736, + "grad_norm": 0.18649598956108093, + "learning_rate": 2.1160593904877235e-06, + "loss": 1.1118, + "step": 14935 + }, + { + "epoch": 0.9074670393098001, + "grad_norm": 0.16009528934955597, + "learning_rate": 2.113304850639486e-06, + "loss": 1.0233, + "step": 14936 + }, + { + "epoch": 0.9075277963424266, + "grad_norm": 0.17618104815483093, + "learning_rate": 2.1105520660873123e-06, + "loss": 1.0857, + "step": 14937 + }, + { + "epoch": 0.9075885533750532, + "grad_norm": 0.12416744977235794, + "learning_rate": 2.107801036932111e-06, + "loss": 1.0515, + "step": 14938 + }, + { + "epoch": 0.9076493104076797, + "grad_norm": 0.1735563725233078, + "learning_rate": 2.1050517632747178e-06, + "loss": 1.0798, + "step": 14939 + }, + { + "epoch": 0.9077100674403062, + "grad_norm": 0.20261690020561218, + "learning_rate": 2.1023042452159136e-06, + "loss": 1.1262, + "step": 14940 + }, + { + "epoch": 0.9077708244729328, + "grad_norm": 0.17964822053909302, + "learning_rate": 2.0995584828564064e-06, + "loss": 1.0816, + "step": 14941 + }, + { + "epoch": 0.9078315815055593, + "grad_norm": 0.1563953012228012, + "learning_rate": 2.096814476296832e-06, + "loss": 0.9894, + "step": 14942 + }, + { + "epoch": 0.9078923385381859, + "grad_norm": 0.21762873232364655, + "learning_rate": 2.094072225637794e-06, + "loss": 1.1621, + "step": 14943 + }, + { + "epoch": 0.9079530955708123, + "grad_norm": 0.8375523090362549, + "learning_rate": 2.091331730979801e-06, + "loss": 1.0422, + "step": 14944 + }, + { + "epoch": 0.9080138526034388, + "grad_norm": 0.20646756887435913, + "learning_rate": 2.0885929924232994e-06, + "loss": 1.0813, + "step": 14945 + }, + { + "epoch": 0.9080746096360653, + "grad_norm": 0.16158004105091095, + "learning_rate": 2.0858560100686763e-06, + "loss": 1.0633, + "step": 14946 + }, + { + "epoch": 0.9081353666686919, + "grad_norm": 0.1544608622789383, + "learning_rate": 2.0831207840162626e-06, + "loss": 1.0533, + "step": 14947 + }, + { + "epoch": 0.9081961237013184, + "grad_norm": 0.14952722191810608, + "learning_rate": 2.0803873143663277e-06, + "loss": 1.0279, + "step": 14948 + }, + { + "epoch": 0.908256880733945, + "grad_norm": 0.1187838688492775, + "learning_rate": 2.0776556012190527e-06, + "loss": 1.0002, + "step": 14949 + }, + { + "epoch": 0.9083176377665715, + "grad_norm": 0.43056565523147583, + "learning_rate": 2.07492564467458e-06, + "loss": 1.1801, + "step": 14950 + }, + { + "epoch": 0.908378394799198, + "grad_norm": 0.15497423708438873, + "learning_rate": 2.0721974448329683e-06, + "loss": 1.0306, + "step": 14951 + }, + { + "epoch": 0.9084391518318246, + "grad_norm": 0.43166497349739075, + "learning_rate": 2.069471001794221e-06, + "loss": 1.2053, + "step": 14952 + }, + { + "epoch": 0.9084999088644511, + "grad_norm": 0.13769285380840302, + "learning_rate": 2.06674631565828e-06, + "loss": 1.0159, + "step": 14953 + }, + { + "epoch": 0.9085606658970776, + "grad_norm": 0.1438886970281601, + "learning_rate": 2.0640233865250215e-06, + "loss": 1.0365, + "step": 14954 + }, + { + "epoch": 0.9086214229297042, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.0613022144942482e-06, + "loss": 1.0273, + "step": 14955 + }, + { + "epoch": 0.9086821799623307, + "grad_norm": 0.21020103991031647, + "learning_rate": 2.058582799665709e-06, + "loss": 1.0475, + "step": 14956 + }, + { + "epoch": 0.9087429369949571, + "grad_norm": 0.1225581094622612, + "learning_rate": 2.055865142139085e-06, + "loss": 1.0248, + "step": 14957 + }, + { + "epoch": 0.9088036940275837, + "grad_norm": 0.21765778958797455, + "learning_rate": 2.053149242013991e-06, + "loss": 0.9795, + "step": 14958 + }, + { + "epoch": 0.9088644510602102, + "grad_norm": 0.2051234245300293, + "learning_rate": 2.05043509938998e-06, + "loss": 1.0325, + "step": 14959 + }, + { + "epoch": 0.9089252080928367, + "grad_norm": 0.29569265246391296, + "learning_rate": 2.04772271436654e-06, + "loss": 1.0694, + "step": 14960 + }, + { + "epoch": 0.9089859651254633, + "grad_norm": 0.1336507946252823, + "learning_rate": 2.045012087043091e-06, + "loss": 1.0367, + "step": 14961 + }, + { + "epoch": 0.9090467221580898, + "grad_norm": 5.41617488861084, + "learning_rate": 2.0423032175189983e-06, + "loss": 1.1387, + "step": 14962 + }, + { + "epoch": 0.9091074791907163, + "grad_norm": 0.25564202666282654, + "learning_rate": 2.0395961058935377e-06, + "loss": 1.0193, + "step": 14963 + }, + { + "epoch": 0.9091682362233429, + "grad_norm": 0.15987765789031982, + "learning_rate": 2.0368907522659686e-06, + "loss": 1.0603, + "step": 14964 + }, + { + "epoch": 0.9092289932559694, + "grad_norm": 0.14195716381072998, + "learning_rate": 2.034187156735434e-06, + "loss": 1.0424, + "step": 14965 + }, + { + "epoch": 0.9092897502885959, + "grad_norm": 0.16375279426574707, + "learning_rate": 2.0314853194010385e-06, + "loss": 1.1074, + "step": 14966 + }, + { + "epoch": 0.9093505073212225, + "grad_norm": 0.18077027797698975, + "learning_rate": 2.0287852403618247e-06, + "loss": 1.0004, + "step": 14967 + }, + { + "epoch": 0.909411264353849, + "grad_norm": 0.16770777106285095, + "learning_rate": 2.026086919716752e-06, + "loss": 1.0496, + "step": 14968 + }, + { + "epoch": 0.9094720213864755, + "grad_norm": 0.16731670498847961, + "learning_rate": 2.023390357564742e-06, + "loss": 1.0317, + "step": 14969 + }, + { + "epoch": 0.909532778419102, + "grad_norm": 0.7469571828842163, + "learning_rate": 2.020695554004637e-06, + "loss": 1.0651, + "step": 14970 + }, + { + "epoch": 0.9095935354517285, + "grad_norm": 0.1581275314092636, + "learning_rate": 2.0180025091352086e-06, + "loss": 1.0518, + "step": 14971 + }, + { + "epoch": 0.909654292484355, + "grad_norm": 0.17387378215789795, + "learning_rate": 2.0153112230551717e-06, + "loss": 1.0917, + "step": 14972 + }, + { + "epoch": 0.9097150495169816, + "grad_norm": 0.21704690158367157, + "learning_rate": 2.0126216958631707e-06, + "loss": 1.0029, + "step": 14973 + }, + { + "epoch": 0.9097758065496081, + "grad_norm": 0.18038031458854675, + "learning_rate": 2.009933927657803e-06, + "loss": 1.086, + "step": 14974 + }, + { + "epoch": 0.9098365635822346, + "grad_norm": 0.11200353503227234, + "learning_rate": 2.0072479185375805e-06, + "loss": 1.026, + "step": 14975 + }, + { + "epoch": 0.9098973206148612, + "grad_norm": 0.17705635726451874, + "learning_rate": 2.004563668600967e-06, + "loss": 1.0738, + "step": 14976 + }, + { + "epoch": 0.9099580776474877, + "grad_norm": 4.210912227630615, + "learning_rate": 2.0018811779463464e-06, + "loss": 1.0902, + "step": 14977 + }, + { + "epoch": 0.9100188346801142, + "grad_norm": 0.15570750832557678, + "learning_rate": 1.9992004466720503e-06, + "loss": 1.0237, + "step": 14978 + }, + { + "epoch": 0.9100795917127408, + "grad_norm": 0.11191708594560623, + "learning_rate": 1.9965214748763395e-06, + "loss": 0.9908, + "step": 14979 + }, + { + "epoch": 0.9101403487453673, + "grad_norm": 0.20823174715042114, + "learning_rate": 1.993844262657413e-06, + "loss": 1.0798, + "step": 14980 + }, + { + "epoch": 0.9102011057779938, + "grad_norm": 0.22196254134178162, + "learning_rate": 1.991168810113403e-06, + "loss": 1.0299, + "step": 14981 + }, + { + "epoch": 0.9102618628106204, + "grad_norm": 0.15476521849632263, + "learning_rate": 1.9884951173423814e-06, + "loss": 1.0657, + "step": 14982 + }, + { + "epoch": 0.9103226198432468, + "grad_norm": 0.11805330216884613, + "learning_rate": 1.9858231844423536e-06, + "loss": 1.017, + "step": 14983 + }, + { + "epoch": 0.9103833768758733, + "grad_norm": 0.12542356550693512, + "learning_rate": 1.9831530115112575e-06, + "loss": 1.0464, + "step": 14984 + }, + { + "epoch": 0.9104441339084999, + "grad_norm": 1.0208297967910767, + "learning_rate": 1.98048459864697e-06, + "loss": 1.0898, + "step": 14985 + }, + { + "epoch": 0.9105048909411264, + "grad_norm": 0.13925513625144958, + "learning_rate": 1.9778179459472968e-06, + "loss": 1.0219, + "step": 14986 + }, + { + "epoch": 0.910565647973753, + "grad_norm": 7.218045234680176, + "learning_rate": 1.9751530535099926e-06, + "loss": 1.1154, + "step": 14987 + }, + { + "epoch": 0.9106264050063795, + "grad_norm": 0.1373661607503891, + "learning_rate": 1.972489921432735e-06, + "loss": 1.0218, + "step": 14988 + }, + { + "epoch": 0.910687162039006, + "grad_norm": 0.21538372337818146, + "learning_rate": 1.969828549813141e-06, + "loss": 1.0728, + "step": 14989 + }, + { + "epoch": 0.9107479190716326, + "grad_norm": 0.22167865931987762, + "learning_rate": 1.9671689387487756e-06, + "loss": 1.1039, + "step": 14990 + }, + { + "epoch": 0.9108086761042591, + "grad_norm": 0.267124205827713, + "learning_rate": 1.9645110883371064e-06, + "loss": 1.0513, + "step": 14991 + }, + { + "epoch": 0.9108694331368856, + "grad_norm": 0.15103879570960999, + "learning_rate": 1.9618549986755717e-06, + "loss": 1.0347, + "step": 14992 + }, + { + "epoch": 0.9109301901695122, + "grad_norm": 0.12741953134536743, + "learning_rate": 1.959200669861522e-06, + "loss": 1.0545, + "step": 14993 + }, + { + "epoch": 0.9109909472021387, + "grad_norm": 0.19085583090782166, + "learning_rate": 1.956548101992267e-06, + "loss": 1.0601, + "step": 14994 + }, + { + "epoch": 0.9110517042347652, + "grad_norm": 0.3262154459953308, + "learning_rate": 1.9538972951650245e-06, + "loss": 1.0594, + "step": 14995 + }, + { + "epoch": 0.9111124612673916, + "grad_norm": 3.455380916595459, + "learning_rate": 1.9512482494769613e-06, + "loss": 1.0466, + "step": 14996 + }, + { + "epoch": 0.9111732183000182, + "grad_norm": 0.186110720038414, + "learning_rate": 1.948600965025188e-06, + "loss": 1.0191, + "step": 14997 + }, + { + "epoch": 0.9112339753326447, + "grad_norm": 0.18384554982185364, + "learning_rate": 1.9459554419067273e-06, + "loss": 1.1084, + "step": 14998 + }, + { + "epoch": 0.9112947323652713, + "grad_norm": 0.15885309875011444, + "learning_rate": 1.9433116802185626e-06, + "loss": 1.0449, + "step": 14999 + }, + { + "epoch": 0.9113554893978978, + "grad_norm": 0.24170510470867157, + "learning_rate": 1.940669680057594e-06, + "loss": 1.0335, + "step": 15000 + }, + { + "epoch": 0.9114162464305243, + "grad_norm": 0.143033966422081, + "learning_rate": 1.9380294415206723e-06, + "loss": 1.0683, + "step": 15001 + }, + { + "epoch": 0.9114770034631509, + "grad_norm": 0.13504737615585327, + "learning_rate": 1.935390964704564e-06, + "loss": 1.0603, + "step": 15002 + }, + { + "epoch": 0.9115377604957774, + "grad_norm": 0.16544565558433533, + "learning_rate": 1.932754249705998e-06, + "loss": 0.9978, + "step": 15003 + }, + { + "epoch": 0.9115985175284039, + "grad_norm": 0.28749409317970276, + "learning_rate": 1.930119296621613e-06, + "loss": 1.0397, + "step": 15004 + }, + { + "epoch": 0.9116592745610305, + "grad_norm": 0.2119385451078415, + "learning_rate": 1.9274861055479986e-06, + "loss": 1.1093, + "step": 15005 + }, + { + "epoch": 0.911720031593657, + "grad_norm": 0.12908820807933807, + "learning_rate": 1.9248546765816722e-06, + "loss": 1.039, + "step": 15006 + }, + { + "epoch": 0.9117807886262835, + "grad_norm": 0.13592183589935303, + "learning_rate": 1.9222250098190897e-06, + "loss": 1.0738, + "step": 15007 + }, + { + "epoch": 0.9118415456589101, + "grad_norm": 0.18259616196155548, + "learning_rate": 1.9195971053566464e-06, + "loss": 1.1101, + "step": 15008 + }, + { + "epoch": 0.9119023026915365, + "grad_norm": 0.25272610783576965, + "learning_rate": 1.916970963290654e-06, + "loss": 1.0926, + "step": 15009 + }, + { + "epoch": 0.911963059724163, + "grad_norm": 0.13104963302612305, + "learning_rate": 1.9143465837174023e-06, + "loss": 1.0216, + "step": 15010 + }, + { + "epoch": 0.9120238167567896, + "grad_norm": 0.18271933495998383, + "learning_rate": 1.911723966733059e-06, + "loss": 1.0232, + "step": 15011 + }, + { + "epoch": 0.9120845737894161, + "grad_norm": 0.11080977320671082, + "learning_rate": 1.9091031124337744e-06, + "loss": 1.0678, + "step": 15012 + }, + { + "epoch": 0.9121453308220426, + "grad_norm": 0.18342067301273346, + "learning_rate": 1.906484020915611e-06, + "loss": 0.9768, + "step": 15013 + }, + { + "epoch": 0.9122060878546692, + "grad_norm": 0.17402873933315277, + "learning_rate": 1.903866692274564e-06, + "loss": 1.1066, + "step": 15014 + }, + { + "epoch": 0.9122668448872957, + "grad_norm": 0.2296784669160843, + "learning_rate": 1.9012511266065847e-06, + "loss": 1.0522, + "step": 15015 + }, + { + "epoch": 0.9123276019199222, + "grad_norm": 0.13235192000865936, + "learning_rate": 1.8986373240075462e-06, + "loss": 1.0281, + "step": 15016 + }, + { + "epoch": 0.9123883589525488, + "grad_norm": 0.11587842553853989, + "learning_rate": 1.8960252845732551e-06, + "loss": 1.0577, + "step": 15017 + }, + { + "epoch": 0.9124491159851753, + "grad_norm": 0.11673952639102936, + "learning_rate": 1.8934150083994518e-06, + "loss": 0.9874, + "step": 15018 + }, + { + "epoch": 0.9125098730178018, + "grad_norm": 0.20689044892787933, + "learning_rate": 1.890806495581815e-06, + "loss": 1.0433, + "step": 15019 + }, + { + "epoch": 0.9125706300504284, + "grad_norm": 0.26666173338890076, + "learning_rate": 1.888199746215974e-06, + "loss": 1.0583, + "step": 15020 + }, + { + "epoch": 0.9126313870830549, + "grad_norm": 0.17910990118980408, + "learning_rate": 1.8855947603974688e-06, + "loss": 1.0696, + "step": 15021 + }, + { + "epoch": 0.9126921441156813, + "grad_norm": 0.14266815781593323, + "learning_rate": 1.8829915382217899e-06, + "loss": 1.0679, + "step": 15022 + }, + { + "epoch": 0.9127529011483079, + "grad_norm": 0.23071017861366272, + "learning_rate": 1.88039007978435e-06, + "loss": 1.1425, + "step": 15023 + }, + { + "epoch": 0.9128136581809344, + "grad_norm": 0.577202320098877, + "learning_rate": 1.8777903851805222e-06, + "loss": 1.1423, + "step": 15024 + }, + { + "epoch": 0.9128744152135609, + "grad_norm": 0.20211109519004822, + "learning_rate": 1.875192454505581e-06, + "loss": 1.0242, + "step": 15025 + }, + { + "epoch": 0.9129351722461875, + "grad_norm": 0.12228650599718094, + "learning_rate": 1.8725962878547721e-06, + "loss": 1.0318, + "step": 15026 + }, + { + "epoch": 0.912995929278814, + "grad_norm": 0.3720133304595947, + "learning_rate": 1.8700018853232416e-06, + "loss": 1.1938, + "step": 15027 + }, + { + "epoch": 0.9130566863114405, + "grad_norm": 0.6146543622016907, + "learning_rate": 1.8674092470060966e-06, + "loss": 1.0233, + "step": 15028 + }, + { + "epoch": 0.9131174433440671, + "grad_norm": 0.14590871334075928, + "learning_rate": 1.8648183729983726e-06, + "loss": 1.0417, + "step": 15029 + }, + { + "epoch": 0.9131782003766936, + "grad_norm": 0.2145708203315735, + "learning_rate": 1.8622292633950266e-06, + "loss": 1.1358, + "step": 15030 + }, + { + "epoch": 0.9132389574093202, + "grad_norm": 0.1720024198293686, + "learning_rate": 1.859641918290983e-06, + "loss": 1.0725, + "step": 15031 + }, + { + "epoch": 0.9132997144419467, + "grad_norm": 0.2165413200855255, + "learning_rate": 1.8570563377810657e-06, + "loss": 1.0698, + "step": 15032 + }, + { + "epoch": 0.9133604714745732, + "grad_norm": 0.14329174160957336, + "learning_rate": 1.8544725219600544e-06, + "loss": 1.0482, + "step": 15033 + }, + { + "epoch": 0.9134212285071998, + "grad_norm": 0.19281533360481262, + "learning_rate": 1.8518904709226626e-06, + "loss": 1.1215, + "step": 15034 + }, + { + "epoch": 0.9134819855398262, + "grad_norm": 0.12573176622390747, + "learning_rate": 1.8493101847635196e-06, + "loss": 1.0513, + "step": 15035 + }, + { + "epoch": 0.9135427425724527, + "grad_norm": 0.16342419385910034, + "learning_rate": 1.846731663577228e-06, + "loss": 1.0941, + "step": 15036 + }, + { + "epoch": 0.9136034996050793, + "grad_norm": 0.12883926928043365, + "learning_rate": 1.8441549074583008e-06, + "loss": 1.0344, + "step": 15037 + }, + { + "epoch": 0.9136642566377058, + "grad_norm": 0.12481041997671127, + "learning_rate": 1.8415799165011793e-06, + "loss": 1.0003, + "step": 15038 + }, + { + "epoch": 0.9137250136703323, + "grad_norm": 0.15143145620822906, + "learning_rate": 1.8390066908002545e-06, + "loss": 1.0286, + "step": 15039 + }, + { + "epoch": 0.9137857707029589, + "grad_norm": 0.3528098165988922, + "learning_rate": 1.83643523044984e-06, + "loss": 1.1992, + "step": 15040 + }, + { + "epoch": 0.9138465277355854, + "grad_norm": 0.13306164741516113, + "learning_rate": 1.8338655355442104e-06, + "loss": 1.0118, + "step": 15041 + }, + { + "epoch": 0.9139072847682119, + "grad_norm": 0.1697445660829544, + "learning_rate": 1.8312976061775512e-06, + "loss": 1.1018, + "step": 15042 + }, + { + "epoch": 0.9139680418008385, + "grad_norm": 0.22958295047283173, + "learning_rate": 1.8287314424439872e-06, + "loss": 1.1353, + "step": 15043 + }, + { + "epoch": 0.914028798833465, + "grad_norm": 0.26398932933807373, + "learning_rate": 1.8261670444375934e-06, + "loss": 1.2653, + "step": 15044 + }, + { + "epoch": 0.9140895558660915, + "grad_norm": 0.20388229191303253, + "learning_rate": 1.8236044122523389e-06, + "loss": 1.0811, + "step": 15045 + }, + { + "epoch": 0.9141503128987181, + "grad_norm": 0.31976118683815, + "learning_rate": 1.8210435459821873e-06, + "loss": 1.0453, + "step": 15046 + }, + { + "epoch": 0.9142110699313446, + "grad_norm": 0.12748916447162628, + "learning_rate": 1.818484445720997e-06, + "loss": 1.0575, + "step": 15047 + }, + { + "epoch": 0.9142718269639711, + "grad_norm": 1.5186364650726318, + "learning_rate": 1.815927111562571e-06, + "loss": 1.1186, + "step": 15048 + }, + { + "epoch": 0.9143325839965976, + "grad_norm": 0.1612519919872284, + "learning_rate": 1.813371543600656e-06, + "loss": 1.0476, + "step": 15049 + }, + { + "epoch": 0.9143933410292241, + "grad_norm": 1.7542378902435303, + "learning_rate": 1.8108177419289162e-06, + "loss": 1.0526, + "step": 15050 + }, + { + "epoch": 0.9144540980618506, + "grad_norm": 0.18711532652378082, + "learning_rate": 1.808265706640966e-06, + "loss": 1.0986, + "step": 15051 + }, + { + "epoch": 0.9145148550944772, + "grad_norm": 0.17725831270217896, + "learning_rate": 1.805715437830352e-06, + "loss": 1.0884, + "step": 15052 + }, + { + "epoch": 0.9145756121271037, + "grad_norm": 0.19067347049713135, + "learning_rate": 1.8031669355905556e-06, + "loss": 1.0884, + "step": 15053 + }, + { + "epoch": 0.9146363691597302, + "grad_norm": 0.3660125434398651, + "learning_rate": 1.8006202000149963e-06, + "loss": 1.0729, + "step": 15054 + }, + { + "epoch": 0.9146971261923568, + "grad_norm": 0.14986911416053772, + "learning_rate": 1.7980752311970106e-06, + "loss": 1.0625, + "step": 15055 + }, + { + "epoch": 0.9147578832249833, + "grad_norm": 0.2169409543275833, + "learning_rate": 1.795532029229896e-06, + "loss": 1.1234, + "step": 15056 + }, + { + "epoch": 0.9148186402576098, + "grad_norm": 0.1255958378314972, + "learning_rate": 1.7929905942068836e-06, + "loss": 1.0116, + "step": 15057 + }, + { + "epoch": 0.9148793972902364, + "grad_norm": 0.26168501377105713, + "learning_rate": 1.79045092622111e-06, + "loss": 1.0815, + "step": 15058 + }, + { + "epoch": 0.9149401543228629, + "grad_norm": 0.33262747526168823, + "learning_rate": 1.7879130253656783e-06, + "loss": 1.2009, + "step": 15059 + }, + { + "epoch": 0.9150009113554894, + "grad_norm": 0.1548096388578415, + "learning_rate": 1.7853768917336144e-06, + "loss": 0.9868, + "step": 15060 + }, + { + "epoch": 0.915061668388116, + "grad_norm": 0.21516329050064087, + "learning_rate": 1.7828425254178715e-06, + "loss": 1.1521, + "step": 15061 + }, + { + "epoch": 0.9151224254207424, + "grad_norm": 0.29394906759262085, + "learning_rate": 1.7803099265113643e-06, + "loss": 1.0762, + "step": 15062 + }, + { + "epoch": 0.9151831824533689, + "grad_norm": 0.12318924814462662, + "learning_rate": 1.7777790951069185e-06, + "loss": 1.0033, + "step": 15063 + }, + { + "epoch": 0.9152439394859955, + "grad_norm": 0.2595454454421997, + "learning_rate": 1.7752500312973042e-06, + "loss": 1.1231, + "step": 15064 + }, + { + "epoch": 0.915304696518622, + "grad_norm": 0.12680596113204956, + "learning_rate": 1.7727227351752195e-06, + "loss": 1.0558, + "step": 15065 + }, + { + "epoch": 0.9153654535512485, + "grad_norm": 0.12583443522453308, + "learning_rate": 1.7701972068332962e-06, + "loss": 1.0177, + "step": 15066 + }, + { + "epoch": 0.9154262105838751, + "grad_norm": 0.2510320246219635, + "learning_rate": 1.7676734463641209e-06, + "loss": 1.1955, + "step": 15067 + }, + { + "epoch": 0.9154869676165016, + "grad_norm": 0.1292337030172348, + "learning_rate": 1.7651514538602033e-06, + "loss": 0.9919, + "step": 15068 + }, + { + "epoch": 0.9155477246491281, + "grad_norm": 0.2651914954185486, + "learning_rate": 1.7626312294139801e-06, + "loss": 1.2944, + "step": 15069 + }, + { + "epoch": 0.9156084816817547, + "grad_norm": 0.18186278641223907, + "learning_rate": 1.7601127731178335e-06, + "loss": 1.0591, + "step": 15070 + }, + { + "epoch": 0.9156692387143812, + "grad_norm": 0.1142735406756401, + "learning_rate": 1.7575960850640783e-06, + "loss": 0.993, + "step": 15071 + }, + { + "epoch": 0.9157299957470078, + "grad_norm": 0.13044516742229462, + "learning_rate": 1.7550811653449628e-06, + "loss": 1.0376, + "step": 15072 + }, + { + "epoch": 0.9157907527796343, + "grad_norm": 7.756257057189941, + "learning_rate": 1.7525680140526746e-06, + "loss": 1.0318, + "step": 15073 + }, + { + "epoch": 0.9158515098122608, + "grad_norm": 0.1589779257774353, + "learning_rate": 1.7500566312793287e-06, + "loss": 1.1437, + "step": 15074 + }, + { + "epoch": 0.9159122668448872, + "grad_norm": 0.4525679349899292, + "learning_rate": 1.747547017116985e-06, + "loss": 1.1593, + "step": 15075 + }, + { + "epoch": 0.9159730238775138, + "grad_norm": 0.17991384863853455, + "learning_rate": 1.7450391716576308e-06, + "loss": 1.0188, + "step": 15076 + }, + { + "epoch": 0.9160337809101403, + "grad_norm": 0.11373403668403625, + "learning_rate": 1.7425330949931985e-06, + "loss": 1.0778, + "step": 15077 + }, + { + "epoch": 0.9160945379427669, + "grad_norm": 0.378408282995224, + "learning_rate": 1.7400287872155362e-06, + "loss": 1.1413, + "step": 15078 + }, + { + "epoch": 0.9161552949753934, + "grad_norm": 0.1763552874326706, + "learning_rate": 1.7375262484164546e-06, + "loss": 1.0958, + "step": 15079 + }, + { + "epoch": 0.9162160520080199, + "grad_norm": 0.13120415806770325, + "learning_rate": 1.735025478687674e-06, + "loss": 1.0611, + "step": 15080 + }, + { + "epoch": 0.9162768090406465, + "grad_norm": 0.14163728058338165, + "learning_rate": 1.7325264781208606e-06, + "loss": 1.0099, + "step": 15081 + }, + { + "epoch": 0.916337566073273, + "grad_norm": 0.21266256272792816, + "learning_rate": 1.7300292468076186e-06, + "loss": 1.0135, + "step": 15082 + }, + { + "epoch": 0.9163983231058995, + "grad_norm": 0.13033802807331085, + "learning_rate": 1.7275337848394856e-06, + "loss": 1.0884, + "step": 15083 + }, + { + "epoch": 0.9164590801385261, + "grad_norm": 0.17783619463443756, + "learning_rate": 1.7250400923079447e-06, + "loss": 1.0989, + "step": 15084 + }, + { + "epoch": 0.9165198371711526, + "grad_norm": 0.1254071444272995, + "learning_rate": 1.7225481693043776e-06, + "loss": 1.0636, + "step": 15085 + }, + { + "epoch": 0.9165805942037791, + "grad_norm": 0.11680983752012253, + "learning_rate": 1.7200580159201396e-06, + "loss": 1.0247, + "step": 15086 + }, + { + "epoch": 0.9166413512364057, + "grad_norm": 0.21700836718082428, + "learning_rate": 1.7175696322465129e-06, + "loss": 1.1087, + "step": 15087 + }, + { + "epoch": 0.9167021082690321, + "grad_norm": 0.26445820927619934, + "learning_rate": 1.7150830183746968e-06, + "loss": 1.0935, + "step": 15088 + }, + { + "epoch": 0.9167628653016586, + "grad_norm": 0.13498102128505707, + "learning_rate": 1.7125981743958518e-06, + "loss": 1.0495, + "step": 15089 + }, + { + "epoch": 0.9168236223342852, + "grad_norm": 0.19517460465431213, + "learning_rate": 1.7101151004010551e-06, + "loss": 1.0657, + "step": 15090 + }, + { + "epoch": 0.9168843793669117, + "grad_norm": 0.1458859145641327, + "learning_rate": 1.7076337964813226e-06, + "loss": 1.0412, + "step": 15091 + }, + { + "epoch": 0.9169451363995382, + "grad_norm": 0.16978909075260162, + "learning_rate": 1.7051542627276097e-06, + "loss": 1.0942, + "step": 15092 + }, + { + "epoch": 0.9170058934321648, + "grad_norm": 0.994243860244751, + "learning_rate": 1.7026764992307987e-06, + "loss": 1.0376, + "step": 15093 + }, + { + "epoch": 0.9170666504647913, + "grad_norm": 0.11853274703025818, + "learning_rate": 1.7002005060817173e-06, + "loss": 0.9993, + "step": 15094 + }, + { + "epoch": 0.9171274074974178, + "grad_norm": 0.17867101728916168, + "learning_rate": 1.6977262833711205e-06, + "loss": 1.1206, + "step": 15095 + }, + { + "epoch": 0.9171881645300444, + "grad_norm": 0.1290816068649292, + "learning_rate": 1.695253831189708e-06, + "loss": 1.0065, + "step": 15096 + }, + { + "epoch": 0.9172489215626709, + "grad_norm": 0.18111705780029297, + "learning_rate": 1.6927831496281021e-06, + "loss": 1.2928, + "step": 15097 + }, + { + "epoch": 0.9173096785952974, + "grad_norm": 0.1601625233888626, + "learning_rate": 1.6903142387768634e-06, + "loss": 1.0642, + "step": 15098 + }, + { + "epoch": 0.917370435627924, + "grad_norm": 0.893388569355011, + "learning_rate": 1.6878470987264971e-06, + "loss": 1.0212, + "step": 15099 + }, + { + "epoch": 0.9174311926605505, + "grad_norm": 0.1660187691450119, + "learning_rate": 1.6853817295674368e-06, + "loss": 1.059, + "step": 15100 + }, + { + "epoch": 0.9174919496931769, + "grad_norm": 0.19975721836090088, + "learning_rate": 1.6829181313900488e-06, + "loss": 1.1633, + "step": 15101 + }, + { + "epoch": 0.9175527067258035, + "grad_norm": 0.1976107954978943, + "learning_rate": 1.6804563042846223e-06, + "loss": 1.1413, + "step": 15102 + }, + { + "epoch": 0.91761346375843, + "grad_norm": 0.12046632915735245, + "learning_rate": 1.6779962483414235e-06, + "loss": 1.0435, + "step": 15103 + }, + { + "epoch": 0.9176742207910565, + "grad_norm": 0.7159797549247742, + "learning_rate": 1.6755379636506141e-06, + "loss": 1.1342, + "step": 15104 + }, + { + "epoch": 0.9177349778236831, + "grad_norm": 0.17540772259235382, + "learning_rate": 1.6730814503022997e-06, + "loss": 1.0165, + "step": 15105 + }, + { + "epoch": 0.9177957348563096, + "grad_norm": 0.2008938491344452, + "learning_rate": 1.6706267083865245e-06, + "loss": 1.1411, + "step": 15106 + }, + { + "epoch": 0.9178564918889361, + "grad_norm": 0.1839478760957718, + "learning_rate": 1.6681737379932616e-06, + "loss": 1.0119, + "step": 15107 + }, + { + "epoch": 0.9179172489215627, + "grad_norm": 0.14366011321544647, + "learning_rate": 1.6657225392124387e-06, + "loss": 1.037, + "step": 15108 + }, + { + "epoch": 0.9179780059541892, + "grad_norm": 0.20343852043151855, + "learning_rate": 1.6632731121339008e-06, + "loss": 1.129, + "step": 15109 + }, + { + "epoch": 0.9180387629868157, + "grad_norm": 0.1504216194152832, + "learning_rate": 1.6608254568474313e-06, + "loss": 1.0112, + "step": 15110 + }, + { + "epoch": 0.9180995200194423, + "grad_norm": 0.16716989874839783, + "learning_rate": 1.6583795734427477e-06, + "loss": 1.0564, + "step": 15111 + }, + { + "epoch": 0.9181602770520688, + "grad_norm": 0.11961951851844788, + "learning_rate": 1.6559354620095003e-06, + "loss": 1.0459, + "step": 15112 + }, + { + "epoch": 0.9182210340846954, + "grad_norm": 0.1276145726442337, + "learning_rate": 1.6534931226372896e-06, + "loss": 1.0216, + "step": 15113 + }, + { + "epoch": 0.9182817911173218, + "grad_norm": 0.21451596915721893, + "learning_rate": 1.6510525554156275e-06, + "loss": 1.074, + "step": 15114 + }, + { + "epoch": 0.9183425481499483, + "grad_norm": 0.21007023751735687, + "learning_rate": 1.6486137604339813e-06, + "loss": 1.0953, + "step": 15115 + }, + { + "epoch": 0.9184033051825748, + "grad_norm": 0.17110972106456757, + "learning_rate": 1.646176737781746e-06, + "loss": 1.0971, + "step": 15116 + }, + { + "epoch": 0.9184640622152014, + "grad_norm": 0.17734146118164062, + "learning_rate": 1.6437414875482504e-06, + "loss": 1.091, + "step": 15117 + }, + { + "epoch": 0.9185248192478279, + "grad_norm": 0.1987987607717514, + "learning_rate": 1.6413080098227561e-06, + "loss": 1.0659, + "step": 15118 + }, + { + "epoch": 0.9185855762804545, + "grad_norm": 0.2392287254333496, + "learning_rate": 1.6388763046944644e-06, + "loss": 1.2041, + "step": 15119 + }, + { + "epoch": 0.918646333313081, + "grad_norm": 0.13121603429317474, + "learning_rate": 1.6364463722525092e-06, + "loss": 1.084, + "step": 15120 + }, + { + "epoch": 0.9187070903457075, + "grad_norm": 3.2505829334259033, + "learning_rate": 1.6340182125859583e-06, + "loss": 1.1064, + "step": 15121 + }, + { + "epoch": 0.9187678473783341, + "grad_norm": 0.13934729993343353, + "learning_rate": 1.6315918257838236e-06, + "loss": 1.0146, + "step": 15122 + }, + { + "epoch": 0.9188286044109606, + "grad_norm": 0.11236473172903061, + "learning_rate": 1.6291672119350288e-06, + "loss": 1.056, + "step": 15123 + }, + { + "epoch": 0.9188893614435871, + "grad_norm": 0.131972536444664, + "learning_rate": 1.6267443711284747e-06, + "loss": 1.0287, + "step": 15124 + }, + { + "epoch": 0.9189501184762137, + "grad_norm": 0.24614328145980835, + "learning_rate": 1.6243233034529515e-06, + "loss": 1.2599, + "step": 15125 + }, + { + "epoch": 0.9190108755088402, + "grad_norm": 0.5775460600852966, + "learning_rate": 1.621904008997205e-06, + "loss": 1.1743, + "step": 15126 + }, + { + "epoch": 0.9190716325414666, + "grad_norm": 0.1951819658279419, + "learning_rate": 1.6194864878499193e-06, + "loss": 1.0421, + "step": 15127 + }, + { + "epoch": 0.9191323895740932, + "grad_norm": 0.1609693169593811, + "learning_rate": 1.617070740099702e-06, + "loss": 0.9718, + "step": 15128 + }, + { + "epoch": 0.9191931466067197, + "grad_norm": 0.3872741758823395, + "learning_rate": 1.6146567658351154e-06, + "loss": 1.1553, + "step": 15129 + }, + { + "epoch": 0.9192539036393462, + "grad_norm": 0.17179414629936218, + "learning_rate": 1.6122445651446383e-06, + "loss": 1.0288, + "step": 15130 + }, + { + "epoch": 0.9193146606719728, + "grad_norm": 0.17508447170257568, + "learning_rate": 1.6098341381166949e-06, + "loss": 1.0432, + "step": 15131 + }, + { + "epoch": 0.9193754177045993, + "grad_norm": 0.10342410206794739, + "learning_rate": 1.607425484839631e-06, + "loss": 1.0284, + "step": 15132 + }, + { + "epoch": 0.9194361747372258, + "grad_norm": 0.23633724451065063, + "learning_rate": 1.6050186054017313e-06, + "loss": 1.1395, + "step": 15133 + }, + { + "epoch": 0.9194969317698524, + "grad_norm": 0.19651401042938232, + "learning_rate": 1.6026134998912367e-06, + "loss": 1.0769, + "step": 15134 + }, + { + "epoch": 0.9195576888024789, + "grad_norm": 0.12753118574619293, + "learning_rate": 1.600210168396299e-06, + "loss": 1.0414, + "step": 15135 + }, + { + "epoch": 0.9196184458351054, + "grad_norm": 0.17867004871368408, + "learning_rate": 1.597808611005014e-06, + "loss": 1.0953, + "step": 15136 + }, + { + "epoch": 0.919679202867732, + "grad_norm": 0.11915569752454758, + "learning_rate": 1.5954088278054114e-06, + "loss": 1.0208, + "step": 15137 + }, + { + "epoch": 0.9197399599003585, + "grad_norm": 0.1314152628183365, + "learning_rate": 1.5930108188854487e-06, + "loss": 1.0652, + "step": 15138 + }, + { + "epoch": 0.919800716932985, + "grad_norm": 0.1397184133529663, + "learning_rate": 1.590614584333039e-06, + "loss": 1.0043, + "step": 15139 + }, + { + "epoch": 0.9198614739656115, + "grad_norm": 0.2914765775203705, + "learning_rate": 1.5882201242360006e-06, + "loss": 1.138, + "step": 15140 + }, + { + "epoch": 0.919922230998238, + "grad_norm": 0.18411117792129517, + "learning_rate": 1.5858274386821192e-06, + "loss": 1.1203, + "step": 15141 + }, + { + "epoch": 0.9199829880308645, + "grad_norm": 1.07609224319458, + "learning_rate": 1.5834365277590857e-06, + "loss": 1.0464, + "step": 15142 + }, + { + "epoch": 0.9200437450634911, + "grad_norm": 0.18319430947303772, + "learning_rate": 1.5810473915545466e-06, + "loss": 1.0668, + "step": 15143 + }, + { + "epoch": 0.9201045020961176, + "grad_norm": 0.24377134442329407, + "learning_rate": 1.578660030156076e-06, + "loss": 1.1191, + "step": 15144 + }, + { + "epoch": 0.9201652591287441, + "grad_norm": 0.20553335547447205, + "learning_rate": 1.5762744436511822e-06, + "loss": 1.0746, + "step": 15145 + }, + { + "epoch": 0.9202260161613707, + "grad_norm": 0.3222634196281433, + "learning_rate": 1.573890632127306e-06, + "loss": 1.2391, + "step": 15146 + }, + { + "epoch": 0.9202867731939972, + "grad_norm": 0.16117174923419952, + "learning_rate": 1.5715085956718278e-06, + "loss": 1.0757, + "step": 15147 + }, + { + "epoch": 0.9203475302266237, + "grad_norm": 0.13106529414653778, + "learning_rate": 1.5691283343720664e-06, + "loss": 1.0204, + "step": 15148 + }, + { + "epoch": 0.9204082872592503, + "grad_norm": 0.14375187456607819, + "learning_rate": 1.566749848315263e-06, + "loss": 1.0034, + "step": 15149 + }, + { + "epoch": 0.9204690442918768, + "grad_norm": 0.31181424856185913, + "learning_rate": 1.5643731375886094e-06, + "loss": 1.039, + "step": 15150 + }, + { + "epoch": 0.9205298013245033, + "grad_norm": 0.2209201157093048, + "learning_rate": 1.5619982022792245e-06, + "loss": 1.1472, + "step": 15151 + }, + { + "epoch": 0.9205905583571299, + "grad_norm": 0.21888791024684906, + "learning_rate": 1.5596250424741553e-06, + "loss": 1.1261, + "step": 15152 + }, + { + "epoch": 0.9206513153897564, + "grad_norm": 0.21244072914123535, + "learning_rate": 1.5572536582603935e-06, + "loss": 1.0396, + "step": 15153 + }, + { + "epoch": 0.9207120724223828, + "grad_norm": 0.2582434415817261, + "learning_rate": 1.5548840497248584e-06, + "loss": 1.1326, + "step": 15154 + }, + { + "epoch": 0.9207728294550094, + "grad_norm": 0.24324634671211243, + "learning_rate": 1.5525162169544138e-06, + "loss": 1.0008, + "step": 15155 + }, + { + "epoch": 0.9208335864876359, + "grad_norm": 1.2240279912948608, + "learning_rate": 1.550150160035857e-06, + "loss": 1.0174, + "step": 15156 + }, + { + "epoch": 0.9208943435202624, + "grad_norm": 0.17303971946239471, + "learning_rate": 1.5477858790559075e-06, + "loss": 1.0349, + "step": 15157 + }, + { + "epoch": 0.920955100552889, + "grad_norm": 0.1432923674583435, + "learning_rate": 1.5454233741012403e-06, + "loss": 1.0517, + "step": 15158 + }, + { + "epoch": 0.9210158575855155, + "grad_norm": 0.15205706655979156, + "learning_rate": 1.5430626452584306e-06, + "loss": 1.0632, + "step": 15159 + }, + { + "epoch": 0.921076614618142, + "grad_norm": 0.13573546707630157, + "learning_rate": 1.5407036926140372e-06, + "loss": 0.9742, + "step": 15160 + }, + { + "epoch": 0.9211373716507686, + "grad_norm": 0.16809168457984924, + "learning_rate": 1.538346516254513e-06, + "loss": 1.092, + "step": 15161 + }, + { + "epoch": 0.9211981286833951, + "grad_norm": 6.831033706665039, + "learning_rate": 1.5359911162662667e-06, + "loss": 1.0172, + "step": 15162 + }, + { + "epoch": 0.9212588857160217, + "grad_norm": 0.12193295359611511, + "learning_rate": 1.5336374927356344e-06, + "loss": 1.0078, + "step": 15163 + }, + { + "epoch": 0.9213196427486482, + "grad_norm": 0.29597926139831543, + "learning_rate": 1.5312856457488866e-06, + "loss": 1.1437, + "step": 15164 + }, + { + "epoch": 0.9213803997812747, + "grad_norm": 0.38507261872291565, + "learning_rate": 1.5289355753922318e-06, + "loss": 1.2842, + "step": 15165 + }, + { + "epoch": 0.9214411568139013, + "grad_norm": 0.11700275540351868, + "learning_rate": 1.5265872817518123e-06, + "loss": 0.9809, + "step": 15166 + }, + { + "epoch": 0.9215019138465277, + "grad_norm": 0.12514102458953857, + "learning_rate": 1.5242407649137036e-06, + "loss": 1.0475, + "step": 15167 + }, + { + "epoch": 0.9215626708791542, + "grad_norm": 0.1709546148777008, + "learning_rate": 1.5218960249639258e-06, + "loss": 1.0766, + "step": 15168 + }, + { + "epoch": 0.9216234279117808, + "grad_norm": 0.18118377029895782, + "learning_rate": 1.5195530619884158e-06, + "loss": 1.0614, + "step": 15169 + }, + { + "epoch": 0.9216841849444073, + "grad_norm": 0.15356643497943878, + "learning_rate": 1.5172118760730546e-06, + "loss": 1.0035, + "step": 15170 + }, + { + "epoch": 0.9217449419770338, + "grad_norm": 0.7493870258331299, + "learning_rate": 1.514872467303674e-06, + "loss": 1.0752, + "step": 15171 + }, + { + "epoch": 0.9218056990096604, + "grad_norm": 0.11947408318519592, + "learning_rate": 1.5125348357660108e-06, + "loss": 1.0735, + "step": 15172 + }, + { + "epoch": 0.9218664560422869, + "grad_norm": 0.17510126531124115, + "learning_rate": 1.5101989815457573e-06, + "loss": 1.0577, + "step": 15173 + }, + { + "epoch": 0.9219272130749134, + "grad_norm": 0.1464775800704956, + "learning_rate": 1.5078649047285288e-06, + "loss": 1.0218, + "step": 15174 + }, + { + "epoch": 0.92198797010754, + "grad_norm": 0.156769260764122, + "learning_rate": 1.505532605399884e-06, + "loss": 1.1414, + "step": 15175 + }, + { + "epoch": 0.9220487271401665, + "grad_norm": 0.21014265716075897, + "learning_rate": 1.5032020836453164e-06, + "loss": 1.0673, + "step": 15176 + }, + { + "epoch": 0.922109484172793, + "grad_norm": 0.22088074684143066, + "learning_rate": 1.5008733395502517e-06, + "loss": 1.0532, + "step": 15177 + }, + { + "epoch": 0.9221702412054196, + "grad_norm": 0.14391463994979858, + "learning_rate": 1.4985463732000604e-06, + "loss": 1.0484, + "step": 15178 + }, + { + "epoch": 0.9222309982380461, + "grad_norm": 0.1382591277360916, + "learning_rate": 1.4962211846800077e-06, + "loss": 0.9969, + "step": 15179 + }, + { + "epoch": 0.9222917552706725, + "grad_norm": 0.1192982941865921, + "learning_rate": 1.4938977740753535e-06, + "loss": 1.0256, + "step": 15180 + }, + { + "epoch": 0.9223525123032991, + "grad_norm": 0.13911142945289612, + "learning_rate": 1.4915761414712514e-06, + "loss": 0.9561, + "step": 15181 + }, + { + "epoch": 0.9224132693359256, + "grad_norm": 0.2410011887550354, + "learning_rate": 1.4892562869528059e-06, + "loss": 1.0953, + "step": 15182 + }, + { + "epoch": 0.9224740263685521, + "grad_norm": 0.18647785484790802, + "learning_rate": 1.4869382106050433e-06, + "loss": 1.0527, + "step": 15183 + }, + { + "epoch": 0.9225347834011787, + "grad_norm": 0.16710509359836578, + "learning_rate": 1.48462191251294e-06, + "loss": 1.0373, + "step": 15184 + }, + { + "epoch": 0.9225955404338052, + "grad_norm": 0.7772679924964905, + "learning_rate": 1.4823073927614006e-06, + "loss": 1.1927, + "step": 15185 + }, + { + "epoch": 0.9226562974664317, + "grad_norm": 0.17850013077259064, + "learning_rate": 1.479994651435257e-06, + "loss": 1.1474, + "step": 15186 + }, + { + "epoch": 0.9227170544990583, + "grad_norm": 0.14692606031894684, + "learning_rate": 1.4776836886192914e-06, + "loss": 1.0602, + "step": 15187 + }, + { + "epoch": 0.9227778115316848, + "grad_norm": 0.1816050112247467, + "learning_rate": 1.4753745043982081e-06, + "loss": 1.0285, + "step": 15188 + }, + { + "epoch": 0.9228385685643113, + "grad_norm": 0.16985581815242767, + "learning_rate": 1.4730670988566508e-06, + "loss": 1.04, + "step": 15189 + }, + { + "epoch": 0.9228993255969379, + "grad_norm": 0.13880155980587006, + "learning_rate": 1.4707614720792018e-06, + "loss": 1.037, + "step": 15190 + }, + { + "epoch": 0.9229600826295644, + "grad_norm": 0.5344361662864685, + "learning_rate": 1.4684576241503712e-06, + "loss": 1.1645, + "step": 15191 + }, + { + "epoch": 0.923020839662191, + "grad_norm": 0.17235684394836426, + "learning_rate": 1.4661555551546024e-06, + "loss": 1.0643, + "step": 15192 + }, + { + "epoch": 0.9230815966948174, + "grad_norm": 0.13637882471084595, + "learning_rate": 1.4638552651762893e-06, + "loss": 1.0568, + "step": 15193 + }, + { + "epoch": 0.9231423537274439, + "grad_norm": 0.13152460753917694, + "learning_rate": 1.4615567542997422e-06, + "loss": 0.997, + "step": 15194 + }, + { + "epoch": 0.9232031107600704, + "grad_norm": 0.21165142953395844, + "learning_rate": 1.4592600226092045e-06, + "loss": 1.0048, + "step": 15195 + }, + { + "epoch": 0.923263867792697, + "grad_norm": 0.13127923011779785, + "learning_rate": 1.4569650701888815e-06, + "loss": 1.0678, + "step": 15196 + }, + { + "epoch": 0.9233246248253235, + "grad_norm": 0.12129757553339005, + "learning_rate": 1.454671897122889e-06, + "loss": 1.0101, + "step": 15197 + }, + { + "epoch": 0.92338538185795, + "grad_norm": 0.5389415621757507, + "learning_rate": 1.4523805034952874e-06, + "loss": 1.4256, + "step": 15198 + }, + { + "epoch": 0.9234461388905766, + "grad_norm": 0.6383952498435974, + "learning_rate": 1.4500908893900544e-06, + "loss": 1.0294, + "step": 15199 + }, + { + "epoch": 0.9235068959232031, + "grad_norm": 0.1868608593940735, + "learning_rate": 1.447803054891117e-06, + "loss": 1.0813, + "step": 15200 + }, + { + "epoch": 0.9235676529558297, + "grad_norm": 0.1167590469121933, + "learning_rate": 1.4455170000823526e-06, + "loss": 0.9951, + "step": 15201 + }, + { + "epoch": 0.9236284099884562, + "grad_norm": 0.17671693861484528, + "learning_rate": 1.4432327250475497e-06, + "loss": 1.0928, + "step": 15202 + }, + { + "epoch": 0.9236891670210827, + "grad_norm": 0.18732228875160217, + "learning_rate": 1.4409502298704359e-06, + "loss": 1.0801, + "step": 15203 + }, + { + "epoch": 0.9237499240537093, + "grad_norm": 0.11604892462491989, + "learning_rate": 1.4386695146346774e-06, + "loss": 0.9682, + "step": 15204 + }, + { + "epoch": 0.9238106810863358, + "grad_norm": 0.13154380023479462, + "learning_rate": 1.4363905794238686e-06, + "loss": 1.0138, + "step": 15205 + }, + { + "epoch": 0.9238714381189622, + "grad_norm": 0.12546613812446594, + "learning_rate": 1.4341134243215593e-06, + "loss": 1.0273, + "step": 15206 + }, + { + "epoch": 0.9239321951515888, + "grad_norm": 0.15136907994747162, + "learning_rate": 1.431838049411205e-06, + "loss": 1.0624, + "step": 15207 + }, + { + "epoch": 0.9239929521842153, + "grad_norm": 0.10266900062561035, + "learning_rate": 1.4295644547762166e-06, + "loss": 1.0208, + "step": 15208 + }, + { + "epoch": 0.9240537092168418, + "grad_norm": 0.1891787350177765, + "learning_rate": 1.4272926404999276e-06, + "loss": 1.0613, + "step": 15209 + }, + { + "epoch": 0.9241144662494684, + "grad_norm": 0.21260635554790497, + "learning_rate": 1.4250226066656213e-06, + "loss": 0.9779, + "step": 15210 + }, + { + "epoch": 0.9241752232820949, + "grad_norm": 0.13082045316696167, + "learning_rate": 1.422754353356498e-06, + "loss": 1.0622, + "step": 15211 + }, + { + "epoch": 0.9242359803147214, + "grad_norm": 0.17848241329193115, + "learning_rate": 1.420487880655702e-06, + "loss": 1.0666, + "step": 15212 + }, + { + "epoch": 0.924296737347348, + "grad_norm": 0.17924031615257263, + "learning_rate": 1.4182231886463171e-06, + "loss": 1.0111, + "step": 15213 + }, + { + "epoch": 0.9243574943799745, + "grad_norm": 0.21587426960468292, + "learning_rate": 1.415960277411349e-06, + "loss": 1.127, + "step": 15214 + }, + { + "epoch": 0.924418251412601, + "grad_norm": 0.18544095754623413, + "learning_rate": 1.413699147033748e-06, + "loss": 1.0166, + "step": 15215 + }, + { + "epoch": 0.9244790084452276, + "grad_norm": 0.17474770545959473, + "learning_rate": 1.4114397975963923e-06, + "loss": 1.0492, + "step": 15216 + }, + { + "epoch": 0.9245397654778541, + "grad_norm": 0.23304995894432068, + "learning_rate": 1.40918222918211e-06, + "loss": 1.0695, + "step": 15217 + }, + { + "epoch": 0.9246005225104806, + "grad_norm": 0.11473589390516281, + "learning_rate": 1.406926441873646e-06, + "loss": 0.9995, + "step": 15218 + }, + { + "epoch": 0.9246612795431071, + "grad_norm": 0.25928956270217896, + "learning_rate": 1.4046724357536846e-06, + "loss": 1.1416, + "step": 15219 + }, + { + "epoch": 0.9247220365757336, + "grad_norm": 0.11540583521127701, + "learning_rate": 1.402420210904848e-06, + "loss": 1.0044, + "step": 15220 + }, + { + "epoch": 0.9247827936083601, + "grad_norm": 0.1562364101409912, + "learning_rate": 1.400169767409687e-06, + "loss": 1.0658, + "step": 15221 + }, + { + "epoch": 0.9248435506409867, + "grad_norm": 0.12526695430278778, + "learning_rate": 1.3979211053506968e-06, + "loss": 0.969, + "step": 15222 + }, + { + "epoch": 0.9249043076736132, + "grad_norm": 0.17435210943222046, + "learning_rate": 1.3956742248103117e-06, + "loss": 1.072, + "step": 15223 + }, + { + "epoch": 0.9249650647062397, + "grad_norm": 0.35668209195137024, + "learning_rate": 1.3934291258708765e-06, + "loss": 1.2325, + "step": 15224 + }, + { + "epoch": 0.9250258217388663, + "grad_norm": 0.17788204550743103, + "learning_rate": 1.3911858086147033e-06, + "loss": 1.1005, + "step": 15225 + }, + { + "epoch": 0.9250865787714928, + "grad_norm": 0.4734276831150055, + "learning_rate": 1.3889442731239933e-06, + "loss": 1.1451, + "step": 15226 + }, + { + "epoch": 0.9251473358041193, + "grad_norm": 0.15550793707370758, + "learning_rate": 1.3867045194809358e-06, + "loss": 1.0456, + "step": 15227 + }, + { + "epoch": 0.9252080928367459, + "grad_norm": 0.16887140274047852, + "learning_rate": 1.384466547767621e-06, + "loss": 0.9943, + "step": 15228 + }, + { + "epoch": 0.9252688498693724, + "grad_norm": 0.23503762483596802, + "learning_rate": 1.3822303580660833e-06, + "loss": 1.064, + "step": 15229 + }, + { + "epoch": 0.925329606901999, + "grad_norm": 0.16279903054237366, + "learning_rate": 1.379995950458285e-06, + "loss": 1.1902, + "step": 15230 + }, + { + "epoch": 0.9253903639346255, + "grad_norm": 0.14893780648708344, + "learning_rate": 1.377763325026138e-06, + "loss": 1.0406, + "step": 15231 + }, + { + "epoch": 0.9254511209672519, + "grad_norm": 0.17709410190582275, + "learning_rate": 1.3755324818514714e-06, + "loss": 1.1205, + "step": 15232 + }, + { + "epoch": 0.9255118779998784, + "grad_norm": 0.23320560157299042, + "learning_rate": 1.373303421016059e-06, + "loss": 1.0574, + "step": 15233 + }, + { + "epoch": 0.925572635032505, + "grad_norm": 0.14793838560581207, + "learning_rate": 1.3710761426016127e-06, + "loss": 1.0547, + "step": 15234 + }, + { + "epoch": 0.9256333920651315, + "grad_norm": 0.20497393608093262, + "learning_rate": 1.3688506466897733e-06, + "loss": 1.0988, + "step": 15235 + }, + { + "epoch": 0.925694149097758, + "grad_norm": 0.23386915028095245, + "learning_rate": 1.3666269333621084e-06, + "loss": 1.1049, + "step": 15236 + }, + { + "epoch": 0.9257549061303846, + "grad_norm": 0.12716856598854065, + "learning_rate": 1.364405002700131e-06, + "loss": 1.0253, + "step": 15237 + }, + { + "epoch": 0.9258156631630111, + "grad_norm": 0.13896603882312775, + "learning_rate": 1.3621848547852978e-06, + "loss": 1.0932, + "step": 15238 + }, + { + "epoch": 0.9258764201956377, + "grad_norm": 0.15205805003643036, + "learning_rate": 1.3599664896989773e-06, + "loss": 1.0258, + "step": 15239 + }, + { + "epoch": 0.9259371772282642, + "grad_norm": 0.12381820380687714, + "learning_rate": 1.357749907522482e-06, + "loss": 1.0005, + "step": 15240 + }, + { + "epoch": 0.9259979342608907, + "grad_norm": 0.21158760786056519, + "learning_rate": 1.3555351083370694e-06, + "loss": 1.0803, + "step": 15241 + }, + { + "epoch": 0.9260586912935173, + "grad_norm": 0.12812484800815582, + "learning_rate": 1.3533220922239188e-06, + "loss": 1.0447, + "step": 15242 + }, + { + "epoch": 0.9261194483261438, + "grad_norm": 0.30226337909698486, + "learning_rate": 1.3511108592641487e-06, + "loss": 1.2135, + "step": 15243 + }, + { + "epoch": 0.9261802053587703, + "grad_norm": 0.1271791011095047, + "learning_rate": 1.3489014095388163e-06, + "loss": 1.0491, + "step": 15244 + }, + { + "epoch": 0.9262409623913967, + "grad_norm": 0.12615853548049927, + "learning_rate": 1.346693743128913e-06, + "loss": 1.0118, + "step": 15245 + }, + { + "epoch": 0.9263017194240233, + "grad_norm": 0.12861666083335876, + "learning_rate": 1.344487860115351e-06, + "loss": 1.0484, + "step": 15246 + }, + { + "epoch": 0.9263624764566498, + "grad_norm": 8.732900619506836, + "learning_rate": 1.3422837605789828e-06, + "loss": 1.1386, + "step": 15247 + }, + { + "epoch": 0.9264232334892764, + "grad_norm": 0.17969848215579987, + "learning_rate": 1.340081444600616e-06, + "loss": 1.046, + "step": 15248 + }, + { + "epoch": 0.9264839905219029, + "grad_norm": 0.23171931505203247, + "learning_rate": 1.3378809122609693e-06, + "loss": 1.0682, + "step": 15249 + }, + { + "epoch": 0.9265447475545294, + "grad_norm": 0.18004456162452698, + "learning_rate": 1.3356821636407059e-06, + "loss": 1.0766, + "step": 15250 + }, + { + "epoch": 0.926605504587156, + "grad_norm": 0.27219358086586, + "learning_rate": 1.3334851988204167e-06, + "loss": 1.1391, + "step": 15251 + }, + { + "epoch": 0.9266662616197825, + "grad_norm": 0.1879502832889557, + "learning_rate": 1.331290017880632e-06, + "loss": 1.0354, + "step": 15252 + }, + { + "epoch": 0.926727018652409, + "grad_norm": 0.12063165009021759, + "learning_rate": 1.3290966209018264e-06, + "loss": 1.0224, + "step": 15253 + }, + { + "epoch": 0.9267877756850356, + "grad_norm": 0.1658681482076645, + "learning_rate": 1.3269050079643851e-06, + "loss": 1.0844, + "step": 15254 + }, + { + "epoch": 0.9268485327176621, + "grad_norm": 0.2422177642583847, + "learning_rate": 1.3247151791486557e-06, + "loss": 1.0735, + "step": 15255 + }, + { + "epoch": 0.9269092897502886, + "grad_norm": 8.390303611755371, + "learning_rate": 1.3225271345348956e-06, + "loss": 1.1252, + "step": 15256 + }, + { + "epoch": 0.9269700467829152, + "grad_norm": 0.12028484046459198, + "learning_rate": 1.320340874203313e-06, + "loss": 0.9978, + "step": 15257 + }, + { + "epoch": 0.9270308038155417, + "grad_norm": 0.18053553998470306, + "learning_rate": 1.3181563982340438e-06, + "loss": 1.0979, + "step": 15258 + }, + { + "epoch": 0.9270915608481681, + "grad_norm": 5.1018877029418945, + "learning_rate": 1.3159737067071631e-06, + "loss": 1.0889, + "step": 15259 + }, + { + "epoch": 0.9271523178807947, + "grad_norm": 0.2123270481824875, + "learning_rate": 1.313792799702679e-06, + "loss": 1.0766, + "step": 15260 + }, + { + "epoch": 0.9272130749134212, + "grad_norm": 0.2917413115501404, + "learning_rate": 1.311613677300527e-06, + "loss": 1.1377, + "step": 15261 + }, + { + "epoch": 0.9272738319460477, + "grad_norm": 0.2830723226070404, + "learning_rate": 1.3094363395805887e-06, + "loss": 1.0432, + "step": 15262 + }, + { + "epoch": 0.9273345889786743, + "grad_norm": 0.14120881259441376, + "learning_rate": 1.307260786622666e-06, + "loss": 1.0159, + "step": 15263 + }, + { + "epoch": 0.9273953460113008, + "grad_norm": 0.26278871297836304, + "learning_rate": 1.3050870185065178e-06, + "loss": 0.9978, + "step": 15264 + }, + { + "epoch": 0.9274561030439273, + "grad_norm": 0.20695659518241882, + "learning_rate": 1.3029150353118247e-06, + "loss": 1.0616, + "step": 15265 + }, + { + "epoch": 0.9275168600765539, + "grad_norm": 0.1511428952217102, + "learning_rate": 1.3007448371181897e-06, + "loss": 1.0786, + "step": 15266 + }, + { + "epoch": 0.9275776171091804, + "grad_norm": 0.13203196227550507, + "learning_rate": 1.29857642400516e-06, + "loss": 1.0573, + "step": 15267 + }, + { + "epoch": 0.9276383741418069, + "grad_norm": 0.8210073709487915, + "learning_rate": 1.2964097960522225e-06, + "loss": 1.0564, + "step": 15268 + }, + { + "epoch": 0.9276991311744335, + "grad_norm": 0.24519607424736023, + "learning_rate": 1.2942449533388078e-06, + "loss": 1.0818, + "step": 15269 + }, + { + "epoch": 0.92775988820706, + "grad_norm": 0.1717200130224228, + "learning_rate": 1.2920818959442526e-06, + "loss": 1.0477, + "step": 15270 + }, + { + "epoch": 0.9278206452396865, + "grad_norm": 0.20985811948776245, + "learning_rate": 1.2899206239478545e-06, + "loss": 1.0737, + "step": 15271 + }, + { + "epoch": 0.927881402272313, + "grad_norm": 0.13843272626399994, + "learning_rate": 1.287761137428828e-06, + "loss": 1.023, + "step": 15272 + }, + { + "epoch": 0.9279421593049395, + "grad_norm": 0.1551545113325119, + "learning_rate": 1.2856034364663372e-06, + "loss": 1.0642, + "step": 15273 + }, + { + "epoch": 0.928002916337566, + "grad_norm": 0.13297893106937408, + "learning_rate": 1.2834475211394692e-06, + "loss": 1.0408, + "step": 15274 + }, + { + "epoch": 0.9280636733701926, + "grad_norm": 0.12439602613449097, + "learning_rate": 1.281293391527244e-06, + "loss": 1.0696, + "step": 15275 + }, + { + "epoch": 0.9281244304028191, + "grad_norm": 0.1411406397819519, + "learning_rate": 1.2791410477086319e-06, + "loss": 1.0355, + "step": 15276 + }, + { + "epoch": 0.9281851874354456, + "grad_norm": 0.260945200920105, + "learning_rate": 1.276990489762514e-06, + "loss": 0.9934, + "step": 15277 + }, + { + "epoch": 0.9282459444680722, + "grad_norm": 0.28933802247047424, + "learning_rate": 1.274841717767733e-06, + "loss": 1.076, + "step": 15278 + }, + { + "epoch": 0.9283067015006987, + "grad_norm": 0.17655979096889496, + "learning_rate": 1.272694731803048e-06, + "loss": 1.1035, + "step": 15279 + }, + { + "epoch": 0.9283674585333253, + "grad_norm": 0.1425919383764267, + "learning_rate": 1.2705495319471573e-06, + "loss": 0.9845, + "step": 15280 + }, + { + "epoch": 0.9284282155659518, + "grad_norm": 0.3008064031600952, + "learning_rate": 1.2684061182786922e-06, + "loss": 1.0777, + "step": 15281 + }, + { + "epoch": 0.9284889725985783, + "grad_norm": 0.15926264226436615, + "learning_rate": 1.2662644908762179e-06, + "loss": 1.0634, + "step": 15282 + }, + { + "epoch": 0.9285497296312049, + "grad_norm": 0.5845902562141418, + "learning_rate": 1.264124649818238e-06, + "loss": 1.0828, + "step": 15283 + }, + { + "epoch": 0.9286104866638314, + "grad_norm": 0.21071211993694305, + "learning_rate": 1.26198659518319e-06, + "loss": 1.1023, + "step": 15284 + }, + { + "epoch": 0.9286712436964578, + "grad_norm": 0.1412813514471054, + "learning_rate": 1.25985032704945e-06, + "loss": 0.9527, + "step": 15285 + }, + { + "epoch": 0.9287320007290844, + "grad_norm": 0.30660557746887207, + "learning_rate": 1.2577158454953108e-06, + "loss": 1.0897, + "step": 15286 + }, + { + "epoch": 0.9287927577617109, + "grad_norm": 0.2128531038761139, + "learning_rate": 1.2555831505990267e-06, + "loss": 1.056, + "step": 15287 + }, + { + "epoch": 0.9288535147943374, + "grad_norm": 0.19072996079921722, + "learning_rate": 1.2534522424387518e-06, + "loss": 1.0366, + "step": 15288 + }, + { + "epoch": 0.928914271826964, + "grad_norm": 4.715278625488281, + "learning_rate": 1.2513231210926124e-06, + "loss": 1.0219, + "step": 15289 + }, + { + "epoch": 0.9289750288595905, + "grad_norm": 0.23550324141979218, + "learning_rate": 1.2491957866386516e-06, + "loss": 1.1874, + "step": 15290 + }, + { + "epoch": 0.929035785892217, + "grad_norm": 0.17378301918506622, + "learning_rate": 1.2470702391548406e-06, + "loss": 1.0992, + "step": 15291 + }, + { + "epoch": 0.9290965429248436, + "grad_norm": 0.2621119022369385, + "learning_rate": 1.2449464787191e-06, + "loss": 1.2124, + "step": 15292 + }, + { + "epoch": 0.9291572999574701, + "grad_norm": 0.2640044689178467, + "learning_rate": 1.242824505409257e-06, + "loss": 1.2164, + "step": 15293 + }, + { + "epoch": 0.9292180569900966, + "grad_norm": 0.1392028033733368, + "learning_rate": 1.2407043193031098e-06, + "loss": 1.0484, + "step": 15294 + }, + { + "epoch": 0.9292788140227232, + "grad_norm": 0.3325469493865967, + "learning_rate": 1.2385859204783746e-06, + "loss": 1.0596, + "step": 15295 + }, + { + "epoch": 0.9293395710553497, + "grad_norm": 0.17319820821285248, + "learning_rate": 1.2364693090126999e-06, + "loss": 1.0401, + "step": 15296 + }, + { + "epoch": 0.9294003280879762, + "grad_norm": 0.11336430162191391, + "learning_rate": 1.234354484983663e-06, + "loss": 1.0366, + "step": 15297 + }, + { + "epoch": 0.9294610851206027, + "grad_norm": 0.3455331325531006, + "learning_rate": 1.2322414484687904e-06, + "loss": 1.0043, + "step": 15298 + }, + { + "epoch": 0.9295218421532292, + "grad_norm": 0.1402764767408371, + "learning_rate": 1.2301301995455371e-06, + "loss": 1.0682, + "step": 15299 + }, + { + "epoch": 0.9295825991858557, + "grad_norm": 0.2419901192188263, + "learning_rate": 1.2280207382912855e-06, + "loss": 1.0466, + "step": 15300 + }, + { + "epoch": 0.9296433562184823, + "grad_norm": 0.19777609407901764, + "learning_rate": 1.2259130647833627e-06, + "loss": 1.1526, + "step": 15301 + }, + { + "epoch": 0.9297041132511088, + "grad_norm": 1.3311214447021484, + "learning_rate": 1.2238071790990236e-06, + "loss": 1.0021, + "step": 15302 + }, + { + "epoch": 0.9297648702837353, + "grad_norm": 0.16920223832130432, + "learning_rate": 1.2217030813154617e-06, + "loss": 1.1232, + "step": 15303 + }, + { + "epoch": 0.9298256273163619, + "grad_norm": 0.16616949439048767, + "learning_rate": 1.2196007715097934e-06, + "loss": 1.0522, + "step": 15304 + }, + { + "epoch": 0.9298863843489884, + "grad_norm": 0.12175856530666351, + "learning_rate": 1.2175002497591016e-06, + "loss": 1.0645, + "step": 15305 + }, + { + "epoch": 0.9299471413816149, + "grad_norm": 0.1192983090877533, + "learning_rate": 1.2154015161403631e-06, + "loss": 0.9593, + "step": 15306 + }, + { + "epoch": 0.9300078984142415, + "grad_norm": 0.22115953266620636, + "learning_rate": 1.2133045707305112e-06, + "loss": 1.1983, + "step": 15307 + }, + { + "epoch": 0.930068655446868, + "grad_norm": 0.17767119407653809, + "learning_rate": 1.2112094136064067e-06, + "loss": 1.1515, + "step": 15308 + }, + { + "epoch": 0.9301294124794945, + "grad_norm": 0.276507705450058, + "learning_rate": 1.209116044844849e-06, + "loss": 0.9947, + "step": 15309 + }, + { + "epoch": 0.9301901695121211, + "grad_norm": 2.3183846473693848, + "learning_rate": 1.2070244645225825e-06, + "loss": 1.243, + "step": 15310 + }, + { + "epoch": 0.9302509265447475, + "grad_norm": 0.11932019144296646, + "learning_rate": 1.2049346727162626e-06, + "loss": 0.9842, + "step": 15311 + }, + { + "epoch": 0.930311683577374, + "grad_norm": 0.2639768421649933, + "learning_rate": 1.2028466695024942e-06, + "loss": 1.0957, + "step": 15312 + }, + { + "epoch": 0.9303724406100006, + "grad_norm": 0.15221168100833893, + "learning_rate": 1.2007604549578167e-06, + "loss": 1.0686, + "step": 15313 + }, + { + "epoch": 0.9304331976426271, + "grad_norm": 0.3506404459476471, + "learning_rate": 1.1986760291586908e-06, + "loss": 1.1335, + "step": 15314 + }, + { + "epoch": 0.9304939546752536, + "grad_norm": 0.13175009191036224, + "learning_rate": 1.1965933921815275e-06, + "loss": 1.0291, + "step": 15315 + }, + { + "epoch": 0.9305547117078802, + "grad_norm": 0.1252862513065338, + "learning_rate": 1.1945125441026718e-06, + "loss": 1.0159, + "step": 15316 + }, + { + "epoch": 0.9306154687405067, + "grad_norm": 0.13322541117668152, + "learning_rate": 1.1924334849983898e-06, + "loss": 1.0098, + "step": 15317 + }, + { + "epoch": 0.9306762257731332, + "grad_norm": 0.14481836557388306, + "learning_rate": 1.1903562149448988e-06, + "loss": 1.0275, + "step": 15318 + }, + { + "epoch": 0.9307369828057598, + "grad_norm": 0.12059970200061798, + "learning_rate": 1.1882807340183265e-06, + "loss": 0.9779, + "step": 15319 + }, + { + "epoch": 0.9307977398383863, + "grad_norm": 0.12137799710035324, + "learning_rate": 1.186207042294768e-06, + "loss": 1.0554, + "step": 15320 + }, + { + "epoch": 0.9308584968710129, + "grad_norm": 0.40280088782310486, + "learning_rate": 1.1841351398502177e-06, + "loss": 1.0383, + "step": 15321 + }, + { + "epoch": 0.9309192539036394, + "grad_norm": 0.18132294714450836, + "learning_rate": 1.182065026760637e-06, + "loss": 1.0198, + "step": 15322 + }, + { + "epoch": 0.9309800109362659, + "grad_norm": 0.33644869923591614, + "learning_rate": 1.1799967031018933e-06, + "loss": 1.2517, + "step": 15323 + }, + { + "epoch": 0.9310407679688923, + "grad_norm": 0.24785423278808594, + "learning_rate": 1.1779301689498091e-06, + "loss": 1.1376, + "step": 15324 + }, + { + "epoch": 0.9311015250015189, + "grad_norm": 0.10992396622896194, + "learning_rate": 1.1758654243801348e-06, + "loss": 1.002, + "step": 15325 + }, + { + "epoch": 0.9311622820341454, + "grad_norm": 0.1956585943698883, + "learning_rate": 1.1738024694685435e-06, + "loss": 0.9923, + "step": 15326 + }, + { + "epoch": 0.931223039066772, + "grad_norm": 0.839550793170929, + "learning_rate": 1.171741304290669e-06, + "loss": 1.217, + "step": 15327 + }, + { + "epoch": 0.9312837960993985, + "grad_norm": 0.13427482545375824, + "learning_rate": 1.1696819289220507e-06, + "loss": 1.018, + "step": 15328 + }, + { + "epoch": 0.931344553132025, + "grad_norm": 0.146847665309906, + "learning_rate": 1.1676243434381783e-06, + "loss": 1.0775, + "step": 15329 + }, + { + "epoch": 0.9314053101646516, + "grad_norm": 0.1301151067018509, + "learning_rate": 1.165568547914475e-06, + "loss": 1.0591, + "step": 15330 + }, + { + "epoch": 0.9314660671972781, + "grad_norm": 0.19895371794700623, + "learning_rate": 1.1635145424262973e-06, + "loss": 1.1097, + "step": 15331 + }, + { + "epoch": 0.9315268242299046, + "grad_norm": 0.29034051299095154, + "learning_rate": 1.16146232704894e-06, + "loss": 1.1231, + "step": 15332 + }, + { + "epoch": 0.9315875812625312, + "grad_norm": 0.1754167377948761, + "learning_rate": 1.1594119018576155e-06, + "loss": 1.0845, + "step": 15333 + }, + { + "epoch": 0.9316483382951577, + "grad_norm": 0.1740913689136505, + "learning_rate": 1.157363266927486e-06, + "loss": 1.0567, + "step": 15334 + }, + { + "epoch": 0.9317090953277842, + "grad_norm": 0.12388742715120316, + "learning_rate": 1.1553164223336465e-06, + "loss": 1.0355, + "step": 15335 + }, + { + "epoch": 0.9317698523604108, + "grad_norm": 0.11986780166625977, + "learning_rate": 1.153271368151132e-06, + "loss": 1.0162, + "step": 15336 + }, + { + "epoch": 0.9318306093930372, + "grad_norm": 0.25391507148742676, + "learning_rate": 1.1512281044548933e-06, + "loss": 1.1407, + "step": 15337 + }, + { + "epoch": 0.9318913664256637, + "grad_norm": 0.1758277267217636, + "learning_rate": 1.1491866313198318e-06, + "loss": 1.027, + "step": 15338 + }, + { + "epoch": 0.9319521234582903, + "grad_norm": 0.20846906304359436, + "learning_rate": 1.1471469488207765e-06, + "loss": 1.0785, + "step": 15339 + }, + { + "epoch": 0.9320128804909168, + "grad_norm": 0.14032472670078278, + "learning_rate": 1.1451090570324896e-06, + "loss": 1.0173, + "step": 15340 + }, + { + "epoch": 0.9320736375235433, + "grad_norm": 0.2942488491535187, + "learning_rate": 1.1430729560296782e-06, + "loss": 1.1245, + "step": 15341 + }, + { + "epoch": 0.9321343945561699, + "grad_norm": 0.1435425877571106, + "learning_rate": 1.1410386458869715e-06, + "loss": 1.0243, + "step": 15342 + }, + { + "epoch": 0.9321951515887964, + "grad_norm": 0.17792147397994995, + "learning_rate": 1.139006126678932e-06, + "loss": 1.0173, + "step": 15343 + }, + { + "epoch": 0.9322559086214229, + "grad_norm": 0.15136665105819702, + "learning_rate": 1.1369753984800724e-06, + "loss": 1.0954, + "step": 15344 + }, + { + "epoch": 0.9323166656540495, + "grad_norm": 0.14888280630111694, + "learning_rate": 1.1349464613648275e-06, + "loss": 1.0254, + "step": 15345 + }, + { + "epoch": 0.932377422686676, + "grad_norm": 0.19266605377197266, + "learning_rate": 1.1329193154075601e-06, + "loss": 1.0597, + "step": 15346 + }, + { + "epoch": 0.9324381797193025, + "grad_norm": 0.16566236317157745, + "learning_rate": 1.130893960682583e-06, + "loss": 1.0536, + "step": 15347 + }, + { + "epoch": 0.9324989367519291, + "grad_norm": 0.12564612925052643, + "learning_rate": 1.1288703972641313e-06, + "loss": 1.0228, + "step": 15348 + }, + { + "epoch": 0.9325596937845556, + "grad_norm": 0.12340943515300751, + "learning_rate": 1.1268486252263844e-06, + "loss": 0.9816, + "step": 15349 + }, + { + "epoch": 0.932620450817182, + "grad_norm": 0.1276017278432846, + "learning_rate": 1.1248286446434498e-06, + "loss": 1.0498, + "step": 15350 + }, + { + "epoch": 0.9326812078498086, + "grad_norm": 0.19445984065532684, + "learning_rate": 1.1228104555893625e-06, + "loss": 1.1042, + "step": 15351 + }, + { + "epoch": 0.9327419648824351, + "grad_norm": 0.17888957262039185, + "learning_rate": 1.1207940581381138e-06, + "loss": 1.1153, + "step": 15352 + }, + { + "epoch": 0.9328027219150616, + "grad_norm": 0.1140437126159668, + "learning_rate": 1.1187794523635997e-06, + "loss": 1.0131, + "step": 15353 + }, + { + "epoch": 0.9328634789476882, + "grad_norm": 0.13291214406490326, + "learning_rate": 1.116766638339678e-06, + "loss": 1.0463, + "step": 15354 + }, + { + "epoch": 0.9329242359803147, + "grad_norm": 0.16485002636909485, + "learning_rate": 1.114755616140123e-06, + "loss": 0.994, + "step": 15355 + }, + { + "epoch": 0.9329849930129412, + "grad_norm": 0.15001343190670013, + "learning_rate": 1.1127463858386477e-06, + "loss": 1.0587, + "step": 15356 + }, + { + "epoch": 0.9330457500455678, + "grad_norm": 0.2237502634525299, + "learning_rate": 1.1107389475089048e-06, + "loss": 1.066, + "step": 15357 + }, + { + "epoch": 0.9331065070781943, + "grad_norm": 0.13696837425231934, + "learning_rate": 1.1087333012244737e-06, + "loss": 1.0073, + "step": 15358 + }, + { + "epoch": 0.9331672641108208, + "grad_norm": 0.2033979594707489, + "learning_rate": 1.1067294470588852e-06, + "loss": 1.0609, + "step": 15359 + }, + { + "epoch": 0.9332280211434474, + "grad_norm": 0.16134701669216156, + "learning_rate": 1.1047273850855689e-06, + "loss": 1.0453, + "step": 15360 + }, + { + "epoch": 0.9332887781760739, + "grad_norm": 0.12504303455352783, + "learning_rate": 1.1027271153779218e-06, + "loss": 0.9526, + "step": 15361 + }, + { + "epoch": 0.9333495352087005, + "grad_norm": 0.13349880278110504, + "learning_rate": 1.100728638009263e-06, + "loss": 0.9987, + "step": 15362 + }, + { + "epoch": 0.933410292241327, + "grad_norm": 0.14257459342479706, + "learning_rate": 1.0987319530528506e-06, + "loss": 1.0035, + "step": 15363 + }, + { + "epoch": 0.9334710492739534, + "grad_norm": 0.11717060208320618, + "learning_rate": 1.0967370605818705e-06, + "loss": 1.0023, + "step": 15364 + }, + { + "epoch": 0.93353180630658, + "grad_norm": 0.22379514575004578, + "learning_rate": 1.0947439606694532e-06, + "loss": 1.1685, + "step": 15365 + }, + { + "epoch": 0.9335925633392065, + "grad_norm": 0.22433388233184814, + "learning_rate": 1.0927526533886401e-06, + "loss": 1.1448, + "step": 15366 + }, + { + "epoch": 0.933653320371833, + "grad_norm": 0.12259352952241898, + "learning_rate": 1.0907631388124339e-06, + "loss": 1.0192, + "step": 15367 + }, + { + "epoch": 0.9337140774044596, + "grad_norm": 1.8735300302505493, + "learning_rate": 1.0887754170137598e-06, + "loss": 1.169, + "step": 15368 + }, + { + "epoch": 0.9337748344370861, + "grad_norm": 0.1701839119195938, + "learning_rate": 1.0867894880654762e-06, + "loss": 1.0716, + "step": 15369 + }, + { + "epoch": 0.9338355914697126, + "grad_norm": 0.3106153607368469, + "learning_rate": 1.08480535204038e-06, + "loss": 1.0912, + "step": 15370 + }, + { + "epoch": 0.9338963485023392, + "grad_norm": 0.12908470630645752, + "learning_rate": 1.0828230090111969e-06, + "loss": 1.0567, + "step": 15371 + }, + { + "epoch": 0.9339571055349657, + "grad_norm": 0.1727006882429123, + "learning_rate": 1.080842459050596e-06, + "loss": 1.0378, + "step": 15372 + }, + { + "epoch": 0.9340178625675922, + "grad_norm": 0.19954133033752441, + "learning_rate": 1.0788637022311642e-06, + "loss": 1.0974, + "step": 15373 + }, + { + "epoch": 0.9340786196002188, + "grad_norm": 0.8826844096183777, + "learning_rate": 1.0768867386254434e-06, + "loss": 1.066, + "step": 15374 + }, + { + "epoch": 0.9341393766328453, + "grad_norm": 0.3567865192890167, + "learning_rate": 1.0749115683058974e-06, + "loss": 1.2216, + "step": 15375 + }, + { + "epoch": 0.9342001336654718, + "grad_norm": 0.25660440325737, + "learning_rate": 1.0729381913449244e-06, + "loss": 0.9833, + "step": 15376 + }, + { + "epoch": 0.9342608906980983, + "grad_norm": 0.17670774459838867, + "learning_rate": 1.0709666078148551e-06, + "loss": 1.0702, + "step": 15377 + }, + { + "epoch": 0.9343216477307248, + "grad_norm": 0.23658257722854614, + "learning_rate": 1.068996817787965e-06, + "loss": 1.2546, + "step": 15378 + }, + { + "epoch": 0.9343824047633513, + "grad_norm": 0.12839379906654358, + "learning_rate": 1.0670288213364577e-06, + "loss": 1.0043, + "step": 15379 + }, + { + "epoch": 0.9344431617959779, + "grad_norm": 0.1378110647201538, + "learning_rate": 1.065062618532464e-06, + "loss": 0.9782, + "step": 15380 + }, + { + "epoch": 0.9345039188286044, + "grad_norm": 0.11447496712207794, + "learning_rate": 1.06309820944806e-06, + "loss": 1.0217, + "step": 15381 + }, + { + "epoch": 0.9345646758612309, + "grad_norm": 0.17230872809886932, + "learning_rate": 1.0611355941552491e-06, + "loss": 1.0043, + "step": 15382 + }, + { + "epoch": 0.9346254328938575, + "grad_norm": 0.15745168924331665, + "learning_rate": 1.0591747727259739e-06, + "loss": 1.0615, + "step": 15383 + }, + { + "epoch": 0.934686189926484, + "grad_norm": 0.15704792737960815, + "learning_rate": 1.0572157452321097e-06, + "loss": 1.0481, + "step": 15384 + }, + { + "epoch": 0.9347469469591105, + "grad_norm": 0.37411150336265564, + "learning_rate": 1.0552585117454606e-06, + "loss": 1.0615, + "step": 15385 + }, + { + "epoch": 0.9348077039917371, + "grad_norm": 0.14910778403282166, + "learning_rate": 1.053303072337769e-06, + "loss": 1.089, + "step": 15386 + }, + { + "epoch": 0.9348684610243636, + "grad_norm": 0.2524835765361786, + "learning_rate": 1.0513494270807168e-06, + "loss": 1.139, + "step": 15387 + }, + { + "epoch": 0.9349292180569901, + "grad_norm": 0.11682143062353134, + "learning_rate": 1.049397576045913e-06, + "loss": 0.9846, + "step": 15388 + }, + { + "epoch": 0.9349899750896167, + "grad_norm": 0.3085375130176544, + "learning_rate": 1.047447519304906e-06, + "loss": 1.1235, + "step": 15389 + }, + { + "epoch": 0.9350507321222431, + "grad_norm": 0.36753910779953003, + "learning_rate": 1.0454992569291666e-06, + "loss": 1.1021, + "step": 15390 + }, + { + "epoch": 0.9351114891548696, + "grad_norm": 0.17073282599449158, + "learning_rate": 1.0435527889901208e-06, + "loss": 1.0533, + "step": 15391 + }, + { + "epoch": 0.9351722461874962, + "grad_norm": 0.2875780165195465, + "learning_rate": 1.0416081155591061e-06, + "loss": 1.1757, + "step": 15392 + }, + { + "epoch": 0.9352330032201227, + "grad_norm": 0.25547468662261963, + "learning_rate": 1.03966523670741e-06, + "loss": 1.1772, + "step": 15393 + }, + { + "epoch": 0.9352937602527492, + "grad_norm": 0.16250906884670258, + "learning_rate": 1.0377241525062531e-06, + "loss": 1.0432, + "step": 15394 + }, + { + "epoch": 0.9353545172853758, + "grad_norm": 0.22641822695732117, + "learning_rate": 1.035784863026773e-06, + "loss": 1.1896, + "step": 15395 + }, + { + "epoch": 0.9354152743180023, + "grad_norm": 0.2214493453502655, + "learning_rate": 1.0338473683400685e-06, + "loss": 1.0903, + "step": 15396 + }, + { + "epoch": 0.9354760313506288, + "grad_norm": 0.20306697487831116, + "learning_rate": 1.0319116685171492e-06, + "loss": 1.1025, + "step": 15397 + }, + { + "epoch": 0.9355367883832554, + "grad_norm": 0.27541497349739075, + "learning_rate": 1.0299777636289808e-06, + "loss": 1.1118, + "step": 15398 + }, + { + "epoch": 0.9355975454158819, + "grad_norm": 0.199677512049675, + "learning_rate": 1.02804565374644e-06, + "loss": 1.0805, + "step": 15399 + }, + { + "epoch": 0.9356583024485084, + "grad_norm": 0.14143909513950348, + "learning_rate": 1.0261153389403533e-06, + "loss": 0.9654, + "step": 15400 + }, + { + "epoch": 0.935719059481135, + "grad_norm": 0.17483264207839966, + "learning_rate": 1.0241868192814752e-06, + "loss": 1.2425, + "step": 15401 + }, + { + "epoch": 0.9357798165137615, + "grad_norm": 1.6043829917907715, + "learning_rate": 1.0222600948404882e-06, + "loss": 1.0577, + "step": 15402 + }, + { + "epoch": 0.9358405735463879, + "grad_norm": 2.448054075241089, + "learning_rate": 1.02033516568803e-06, + "loss": 1.0773, + "step": 15403 + }, + { + "epoch": 0.9359013305790145, + "grad_norm": 0.10662398487329483, + "learning_rate": 1.0184120318946554e-06, + "loss": 1.0026, + "step": 15404 + }, + { + "epoch": 0.935962087611641, + "grad_norm": 0.2379944622516632, + "learning_rate": 1.0164906935308526e-06, + "loss": 1.0839, + "step": 15405 + }, + { + "epoch": 0.9360228446442675, + "grad_norm": 0.13646939396858215, + "learning_rate": 1.0145711506670596e-06, + "loss": 0.9935, + "step": 15406 + }, + { + "epoch": 0.9360836016768941, + "grad_norm": 0.15874788165092468, + "learning_rate": 1.0126534033736202e-06, + "loss": 1.1358, + "step": 15407 + }, + { + "epoch": 0.9361443587095206, + "grad_norm": 0.15469656884670258, + "learning_rate": 1.0107374517208447e-06, + "loss": 1.0905, + "step": 15408 + }, + { + "epoch": 0.9362051157421472, + "grad_norm": 0.3145158886909485, + "learning_rate": 1.0088232957789545e-06, + "loss": 1.0909, + "step": 15409 + }, + { + "epoch": 0.9362658727747737, + "grad_norm": 0.17362472414970398, + "learning_rate": 1.0069109356181162e-06, + "loss": 1.104, + "step": 15410 + }, + { + "epoch": 0.9363266298074002, + "grad_norm": 5.211677074432373, + "learning_rate": 1.005000371308429e-06, + "loss": 1.0267, + "step": 15411 + }, + { + "epoch": 0.9363873868400268, + "grad_norm": 5.238958835601807, + "learning_rate": 1.0030916029199256e-06, + "loss": 1.018, + "step": 15412 + }, + { + "epoch": 0.9364481438726533, + "grad_norm": 0.1375567615032196, + "learning_rate": 1.001184630522567e-06, + "loss": 1.0258, + "step": 15413 + }, + { + "epoch": 0.9365089009052798, + "grad_norm": 0.19185861945152283, + "learning_rate": 9.992794541862582e-07, + "loss": 1.0496, + "step": 15414 + }, + { + "epoch": 0.9365696579379064, + "grad_norm": 0.36689046025276184, + "learning_rate": 9.973760739808325e-07, + "loss": 1.1128, + "step": 15415 + }, + { + "epoch": 0.9366304149705328, + "grad_norm": 0.21147724986076355, + "learning_rate": 9.954744899760616e-07, + "loss": 1.2285, + "step": 15416 + }, + { + "epoch": 0.9366911720031593, + "grad_norm": 8.476731300354004, + "learning_rate": 9.935747022416397e-07, + "loss": 1.0618, + "step": 15417 + }, + { + "epoch": 0.9367519290357859, + "grad_norm": 0.6054396629333496, + "learning_rate": 9.916767108472114e-07, + "loss": 1.0795, + "step": 15418 + }, + { + "epoch": 0.9368126860684124, + "grad_norm": 0.14368443191051483, + "learning_rate": 9.89780515862354e-07, + "loss": 1.0486, + "step": 15419 + }, + { + "epoch": 0.9368734431010389, + "grad_norm": 0.14726805686950684, + "learning_rate": 9.878861173565569e-07, + "loss": 1.027, + "step": 15420 + }, + { + "epoch": 0.9369342001336655, + "grad_norm": 0.12706002593040466, + "learning_rate": 9.859935153992694e-07, + "loss": 1.0121, + "step": 15421 + }, + { + "epoch": 0.936994957166292, + "grad_norm": 0.1864461898803711, + "learning_rate": 9.841027100598699e-07, + "loss": 1.1049, + "step": 15422 + }, + { + "epoch": 0.9370557141989185, + "grad_norm": 0.2949415147304535, + "learning_rate": 9.82213701407647e-07, + "loss": 1.1487, + "step": 15423 + }, + { + "epoch": 0.9371164712315451, + "grad_norm": 0.2535955011844635, + "learning_rate": 9.80326489511868e-07, + "loss": 1.2345, + "step": 15424 + }, + { + "epoch": 0.9371772282641716, + "grad_norm": 0.1851813644170761, + "learning_rate": 9.784410744416882e-07, + "loss": 1.1196, + "step": 15425 + }, + { + "epoch": 0.9372379852967981, + "grad_norm": 0.12779666483402252, + "learning_rate": 9.765574562662362e-07, + "loss": 1.0502, + "step": 15426 + }, + { + "epoch": 0.9372987423294247, + "grad_norm": 0.2776857912540436, + "learning_rate": 9.746756350545449e-07, + "loss": 1.1455, + "step": 15427 + }, + { + "epoch": 0.9373594993620512, + "grad_norm": 0.11265168339014053, + "learning_rate": 9.727956108755875e-07, + "loss": 1.0099, + "step": 15428 + }, + { + "epoch": 0.9374202563946776, + "grad_norm": 0.11672201007604599, + "learning_rate": 9.709173837982866e-07, + "loss": 1.0261, + "step": 15429 + }, + { + "epoch": 0.9374810134273042, + "grad_norm": 0.18230339884757996, + "learning_rate": 9.690409538914869e-07, + "loss": 1.0254, + "step": 15430 + }, + { + "epoch": 0.9375417704599307, + "grad_norm": 0.1740657091140747, + "learning_rate": 9.671663212239667e-07, + "loss": 1.0743, + "step": 15431 + }, + { + "epoch": 0.9376025274925572, + "grad_norm": 0.1717427521944046, + "learning_rate": 9.65293485864449e-07, + "loss": 1.0887, + "step": 15432 + }, + { + "epoch": 0.9376632845251838, + "grad_norm": 0.11930928379297256, + "learning_rate": 9.63422447881568e-07, + "loss": 1.0236, + "step": 15433 + }, + { + "epoch": 0.9377240415578103, + "grad_norm": 0.21351349353790283, + "learning_rate": 9.615532073439238e-07, + "loss": 1.1546, + "step": 15434 + }, + { + "epoch": 0.9377847985904368, + "grad_norm": 0.12577441334724426, + "learning_rate": 9.59685764320023e-07, + "loss": 1.0495, + "step": 15435 + }, + { + "epoch": 0.9378455556230634, + "grad_norm": 0.18328459560871124, + "learning_rate": 9.578201188783165e-07, + "loss": 1.1303, + "step": 15436 + }, + { + "epoch": 0.9379063126556899, + "grad_norm": 0.3981226682662964, + "learning_rate": 9.55956271087194e-07, + "loss": 1.0598, + "step": 15437 + }, + { + "epoch": 0.9379670696883164, + "grad_norm": 0.16804806888103485, + "learning_rate": 9.540942210149783e-07, + "loss": 1.0608, + "step": 15438 + }, + { + "epoch": 0.938027826720943, + "grad_norm": 0.6524078249931335, + "learning_rate": 9.522339687299153e-07, + "loss": 1.1534, + "step": 15439 + }, + { + "epoch": 0.9380885837535695, + "grad_norm": 10.553890228271484, + "learning_rate": 9.503755143002002e-07, + "loss": 1.0307, + "step": 15440 + }, + { + "epoch": 0.938149340786196, + "grad_norm": 0.18413898348808289, + "learning_rate": 9.485188577939508e-07, + "loss": 1.1319, + "step": 15441 + }, + { + "epoch": 0.9382100978188225, + "grad_norm": 0.2837417721748352, + "learning_rate": 9.466639992792237e-07, + "loss": 1.0247, + "step": 15442 + }, + { + "epoch": 0.938270854851449, + "grad_norm": 0.2064703106880188, + "learning_rate": 9.448109388240089e-07, + "loss": 1.0797, + "step": 15443 + }, + { + "epoch": 0.9383316118840755, + "grad_norm": 0.15010347962379456, + "learning_rate": 9.429596764962245e-07, + "loss": 1.0277, + "step": 15444 + }, + { + "epoch": 0.9383923689167021, + "grad_norm": 0.3248666226863861, + "learning_rate": 9.411102123637439e-07, + "loss": 1.0495, + "step": 15445 + }, + { + "epoch": 0.9384531259493286, + "grad_norm": 0.14208881556987762, + "learning_rate": 9.392625464943572e-07, + "loss": 1.0088, + "step": 15446 + }, + { + "epoch": 0.9385138829819551, + "grad_norm": 0.35373157262802124, + "learning_rate": 9.37416678955777e-07, + "loss": 1.0941, + "step": 15447 + }, + { + "epoch": 0.9385746400145817, + "grad_norm": 0.36616086959838867, + "learning_rate": 9.355726098156714e-07, + "loss": 1.3264, + "step": 15448 + }, + { + "epoch": 0.9386353970472082, + "grad_norm": 0.13608017563819885, + "learning_rate": 9.337303391416363e-07, + "loss": 1.0515, + "step": 15449 + }, + { + "epoch": 0.9386961540798348, + "grad_norm": 0.18329325318336487, + "learning_rate": 9.318898670012011e-07, + "loss": 1.008, + "step": 15450 + }, + { + "epoch": 0.9387569111124613, + "grad_norm": 0.10860113054513931, + "learning_rate": 9.300511934618284e-07, + "loss": 0.9916, + "step": 15451 + }, + { + "epoch": 0.9388176681450878, + "grad_norm": 2.0383851528167725, + "learning_rate": 9.282143185909087e-07, + "loss": 1.1347, + "step": 15452 + }, + { + "epoch": 0.9388784251777144, + "grad_norm": 0.23310232162475586, + "learning_rate": 9.263792424557882e-07, + "loss": 1.1035, + "step": 15453 + }, + { + "epoch": 0.9389391822103409, + "grad_norm": 0.11067614704370499, + "learning_rate": 9.245459651237132e-07, + "loss": 0.9974, + "step": 15454 + }, + { + "epoch": 0.9389999392429673, + "grad_norm": 0.13938288390636444, + "learning_rate": 9.227144866618964e-07, + "loss": 1.0337, + "step": 15455 + }, + { + "epoch": 0.9390606962755939, + "grad_norm": 0.1712646633386612, + "learning_rate": 9.208848071374676e-07, + "loss": 1.0581, + "step": 15456 + }, + { + "epoch": 0.9391214533082204, + "grad_norm": 0.1511811465024948, + "learning_rate": 9.190569266174898e-07, + "loss": 1.05, + "step": 15457 + }, + { + "epoch": 0.9391822103408469, + "grad_norm": 0.14525629580020905, + "learning_rate": 9.172308451689704e-07, + "loss": 1.0115, + "step": 15458 + }, + { + "epoch": 0.9392429673734735, + "grad_norm": 0.12309926003217697, + "learning_rate": 9.154065628588392e-07, + "loss": 1.0085, + "step": 15459 + }, + { + "epoch": 0.9393037244061, + "grad_norm": 0.21768274903297424, + "learning_rate": 9.135840797539708e-07, + "loss": 1.0539, + "step": 15460 + }, + { + "epoch": 0.9393644814387265, + "grad_norm": 0.12994812428951263, + "learning_rate": 9.117633959211668e-07, + "loss": 1.0117, + "step": 15461 + }, + { + "epoch": 0.9394252384713531, + "grad_norm": 0.1788455992937088, + "learning_rate": 9.099445114271632e-07, + "loss": 1.0437, + "step": 15462 + }, + { + "epoch": 0.9394859955039796, + "grad_norm": 0.1905161589384079, + "learning_rate": 9.081274263386285e-07, + "loss": 1.1006, + "step": 15463 + }, + { + "epoch": 0.9395467525366061, + "grad_norm": 0.13060259819030762, + "learning_rate": 9.063121407221764e-07, + "loss": 1.0334, + "step": 15464 + }, + { + "epoch": 0.9396075095692327, + "grad_norm": 0.16830359399318695, + "learning_rate": 9.044986546443368e-07, + "loss": 1.0654, + "step": 15465 + }, + { + "epoch": 0.9396682666018592, + "grad_norm": 0.17378097772598267, + "learning_rate": 9.026869681715954e-07, + "loss": 1.0543, + "step": 15466 + }, + { + "epoch": 0.9397290236344857, + "grad_norm": 0.2641183137893677, + "learning_rate": 9.008770813703549e-07, + "loss": 1.039, + "step": 15467 + }, + { + "epoch": 0.9397897806671123, + "grad_norm": 0.16849763691425323, + "learning_rate": 8.990689943069508e-07, + "loss": 1.0944, + "step": 15468 + }, + { + "epoch": 0.9398505376997387, + "grad_norm": 0.21001361310482025, + "learning_rate": 8.972627070476636e-07, + "loss": 1.1702, + "step": 15469 + }, + { + "epoch": 0.9399112947323652, + "grad_norm": 0.17092666029930115, + "learning_rate": 8.954582196587013e-07, + "loss": 1.0784, + "step": 15470 + }, + { + "epoch": 0.9399720517649918, + "grad_norm": 0.12919259071350098, + "learning_rate": 8.936555322062168e-07, + "loss": 1.0161, + "step": 15471 + }, + { + "epoch": 0.9400328087976183, + "grad_norm": 0.20022866129875183, + "learning_rate": 8.918546447562737e-07, + "loss": 1.0956, + "step": 15472 + }, + { + "epoch": 0.9400935658302448, + "grad_norm": 0.1509554088115692, + "learning_rate": 8.90055557374897e-07, + "loss": 1.0316, + "step": 15473 + }, + { + "epoch": 0.9401543228628714, + "grad_norm": 0.13477730751037598, + "learning_rate": 8.882582701280228e-07, + "loss": 1.0284, + "step": 15474 + }, + { + "epoch": 0.9402150798954979, + "grad_norm": 0.12174800783395767, + "learning_rate": 8.864627830815375e-07, + "loss": 1.0, + "step": 15475 + }, + { + "epoch": 0.9402758369281244, + "grad_norm": 0.1383143812417984, + "learning_rate": 8.846690963012494e-07, + "loss": 1.0293, + "step": 15476 + }, + { + "epoch": 0.940336593960751, + "grad_norm": 0.13279202580451965, + "learning_rate": 8.828772098529114e-07, + "loss": 1.0395, + "step": 15477 + }, + { + "epoch": 0.9403973509933775, + "grad_norm": 0.49114689230918884, + "learning_rate": 8.81087123802199e-07, + "loss": 1.1242, + "step": 15478 + }, + { + "epoch": 0.940458108026004, + "grad_norm": 0.17501495778560638, + "learning_rate": 8.792988382147372e-07, + "loss": 1.0155, + "step": 15479 + }, + { + "epoch": 0.9405188650586306, + "grad_norm": 0.6761817336082458, + "learning_rate": 8.775123531560736e-07, + "loss": 1.0715, + "step": 15480 + }, + { + "epoch": 0.9405796220912571, + "grad_norm": 0.15254837274551392, + "learning_rate": 8.757276686916837e-07, + "loss": 1.0701, + "step": 15481 + }, + { + "epoch": 0.9406403791238835, + "grad_norm": 0.17896534502506256, + "learning_rate": 8.739447848869987e-07, + "loss": 1.002, + "step": 15482 + }, + { + "epoch": 0.9407011361565101, + "grad_norm": 0.1767841875553131, + "learning_rate": 8.721637018073548e-07, + "loss": 1.0676, + "step": 15483 + }, + { + "epoch": 0.9407618931891366, + "grad_norm": 0.5318757891654968, + "learning_rate": 8.703844195180555e-07, + "loss": 1.0463, + "step": 15484 + }, + { + "epoch": 0.9408226502217631, + "grad_norm": 0.8863385915756226, + "learning_rate": 8.686069380843042e-07, + "loss": 1.1241, + "step": 15485 + }, + { + "epoch": 0.9408834072543897, + "grad_norm": 0.19632409512996674, + "learning_rate": 8.668312575712656e-07, + "loss": 0.9902, + "step": 15486 + }, + { + "epoch": 0.9409441642870162, + "grad_norm": 0.1418273150920868, + "learning_rate": 8.650573780440263e-07, + "loss": 1.0626, + "step": 15487 + }, + { + "epoch": 0.9410049213196428, + "grad_norm": 0.23336203396320343, + "learning_rate": 8.632852995676066e-07, + "loss": 1.0447, + "step": 15488 + }, + { + "epoch": 0.9410656783522693, + "grad_norm": 2.9486753940582275, + "learning_rate": 8.615150222069601e-07, + "loss": 1.1105, + "step": 15489 + }, + { + "epoch": 0.9411264353848958, + "grad_norm": 0.18078917264938354, + "learning_rate": 8.597465460269738e-07, + "loss": 0.9999, + "step": 15490 + }, + { + "epoch": 0.9411871924175224, + "grad_norm": 0.2550185024738312, + "learning_rate": 8.579798710924791e-07, + "loss": 1.1219, + "step": 15491 + }, + { + "epoch": 0.9412479494501489, + "grad_norm": 0.13539738953113556, + "learning_rate": 8.562149974682354e-07, + "loss": 1.0744, + "step": 15492 + }, + { + "epoch": 0.9413087064827754, + "grad_norm": 0.19891247153282166, + "learning_rate": 8.544519252189354e-07, + "loss": 1.0696, + "step": 15493 + }, + { + "epoch": 0.941369463515402, + "grad_norm": 0.2045857459306717, + "learning_rate": 8.526906544091884e-07, + "loss": 1.1248, + "step": 15494 + }, + { + "epoch": 0.9414302205480284, + "grad_norm": 0.2038622498512268, + "learning_rate": 8.50931185103565e-07, + "loss": 1.0693, + "step": 15495 + }, + { + "epoch": 0.9414909775806549, + "grad_norm": 1.6396106481552124, + "learning_rate": 8.491735173665693e-07, + "loss": 1.1543, + "step": 15496 + }, + { + "epoch": 0.9415517346132815, + "grad_norm": 0.12587201595306396, + "learning_rate": 8.474176512626108e-07, + "loss": 1.0288, + "step": 15497 + }, + { + "epoch": 0.941612491645908, + "grad_norm": 0.17208772897720337, + "learning_rate": 8.456635868560603e-07, + "loss": 1.1561, + "step": 15498 + }, + { + "epoch": 0.9416732486785345, + "grad_norm": 0.19065725803375244, + "learning_rate": 8.439113242112217e-07, + "loss": 1.1126, + "step": 15499 + }, + { + "epoch": 0.9417340057111611, + "grad_norm": 0.1341850757598877, + "learning_rate": 8.421608633923051e-07, + "loss": 1.0539, + "step": 15500 + }, + { + "epoch": 0.9417947627437876, + "grad_norm": 0.15588371455669403, + "learning_rate": 8.404122044634865e-07, + "loss": 0.9698, + "step": 15501 + }, + { + "epoch": 0.9418555197764141, + "grad_norm": 0.17041058838367462, + "learning_rate": 8.386653474888595e-07, + "loss": 1.0974, + "step": 15502 + }, + { + "epoch": 0.9419162768090407, + "grad_norm": 0.1772596538066864, + "learning_rate": 8.369202925324559e-07, + "loss": 1.0331, + "step": 15503 + }, + { + "epoch": 0.9419770338416672, + "grad_norm": 0.14430458843708038, + "learning_rate": 8.35177039658247e-07, + "loss": 0.9915, + "step": 15504 + }, + { + "epoch": 0.9420377908742937, + "grad_norm": 0.18160651624202728, + "learning_rate": 8.33435588930126e-07, + "loss": 1.0731, + "step": 15505 + }, + { + "epoch": 0.9420985479069203, + "grad_norm": 0.4123280942440033, + "learning_rate": 8.316959404119252e-07, + "loss": 1.2548, + "step": 15506 + }, + { + "epoch": 0.9421593049395468, + "grad_norm": 0.16502569615840912, + "learning_rate": 8.299580941674157e-07, + "loss": 1.0902, + "step": 15507 + }, + { + "epoch": 0.9422200619721732, + "grad_norm": 0.12947624921798706, + "learning_rate": 8.282220502602911e-07, + "loss": 1.0419, + "step": 15508 + }, + { + "epoch": 0.9422808190047998, + "grad_norm": 0.14080999791622162, + "learning_rate": 8.264878087542005e-07, + "loss": 1.0501, + "step": 15509 + }, + { + "epoch": 0.9423415760374263, + "grad_norm": 0.19941137731075287, + "learning_rate": 8.247553697126986e-07, + "loss": 1.135, + "step": 15510 + }, + { + "epoch": 0.9424023330700528, + "grad_norm": 0.10916775465011597, + "learning_rate": 8.230247331992902e-07, + "loss": 1.0112, + "step": 15511 + }, + { + "epoch": 0.9424630901026794, + "grad_norm": 0.24294036626815796, + "learning_rate": 8.212958992774189e-07, + "loss": 1.0589, + "step": 15512 + }, + { + "epoch": 0.9425238471353059, + "grad_norm": 0.16311386227607727, + "learning_rate": 8.195688680104618e-07, + "loss": 1.1213, + "step": 15513 + }, + { + "epoch": 0.9425846041679324, + "grad_norm": 0.41804632544517517, + "learning_rate": 8.178436394617073e-07, + "loss": 1.1659, + "step": 15514 + }, + { + "epoch": 0.942645361200559, + "grad_norm": 0.14955374598503113, + "learning_rate": 8.16120213694399e-07, + "loss": 1.0857, + "step": 15515 + }, + { + "epoch": 0.9427061182331855, + "grad_norm": 0.12781886756420135, + "learning_rate": 8.143985907717089e-07, + "loss": 0.995, + "step": 15516 + }, + { + "epoch": 0.942766875265812, + "grad_norm": 0.17817473411560059, + "learning_rate": 8.126787707567473e-07, + "loss": 1.09, + "step": 15517 + }, + { + "epoch": 0.9428276322984386, + "grad_norm": 0.2069421410560608, + "learning_rate": 8.109607537125585e-07, + "loss": 1.1165, + "step": 15518 + }, + { + "epoch": 0.9428883893310651, + "grad_norm": 0.7390443682670593, + "learning_rate": 8.092445397021142e-07, + "loss": 1.071, + "step": 15519 + }, + { + "epoch": 0.9429491463636916, + "grad_norm": 0.18742093443870544, + "learning_rate": 8.075301287883141e-07, + "loss": 1.059, + "step": 15520 + }, + { + "epoch": 0.9430099033963181, + "grad_norm": 0.19499041140079498, + "learning_rate": 8.058175210339968e-07, + "loss": 1.0803, + "step": 15521 + }, + { + "epoch": 0.9430706604289446, + "grad_norm": 0.11112088710069656, + "learning_rate": 8.041067165019567e-07, + "loss": 0.9858, + "step": 15522 + }, + { + "epoch": 0.9431314174615711, + "grad_norm": 0.1163569763302803, + "learning_rate": 8.023977152548934e-07, + "loss": 1.0201, + "step": 15523 + }, + { + "epoch": 0.9431921744941977, + "grad_norm": 0.1632387638092041, + "learning_rate": 8.006905173554513e-07, + "loss": 0.9859, + "step": 15524 + }, + { + "epoch": 0.9432529315268242, + "grad_norm": 0.21261055767536163, + "learning_rate": 7.989851228662082e-07, + "loss": 1.0552, + "step": 15525 + }, + { + "epoch": 0.9433136885594507, + "grad_norm": 0.15866491198539734, + "learning_rate": 7.972815318496807e-07, + "loss": 1.0862, + "step": 15526 + }, + { + "epoch": 0.9433744455920773, + "grad_norm": 1.0427333116531372, + "learning_rate": 7.955797443683022e-07, + "loss": 1.0448, + "step": 15527 + }, + { + "epoch": 0.9434352026247038, + "grad_norm": 0.11727310717105865, + "learning_rate": 7.938797604844616e-07, + "loss": 0.9988, + "step": 15528 + }, + { + "epoch": 0.9434959596573304, + "grad_norm": 0.13976772129535675, + "learning_rate": 7.921815802604704e-07, + "loss": 1.0153, + "step": 15529 + }, + { + "epoch": 0.9435567166899569, + "grad_norm": 0.1581905037164688, + "learning_rate": 7.904852037585731e-07, + "loss": 1.0577, + "step": 15530 + }, + { + "epoch": 0.9436174737225834, + "grad_norm": 0.15511280298233032, + "learning_rate": 7.887906310409588e-07, + "loss": 1.0262, + "step": 15531 + }, + { + "epoch": 0.94367823075521, + "grad_norm": 0.14707230031490326, + "learning_rate": 7.870978621697222e-07, + "loss": 1.0427, + "step": 15532 + }, + { + "epoch": 0.9437389877878365, + "grad_norm": 0.15507178008556366, + "learning_rate": 7.854068972069418e-07, + "loss": 1.2066, + "step": 15533 + }, + { + "epoch": 0.9437997448204629, + "grad_norm": 0.1287851184606552, + "learning_rate": 7.837177362145786e-07, + "loss": 1.0295, + "step": 15534 + }, + { + "epoch": 0.9438605018530895, + "grad_norm": 0.13537128269672394, + "learning_rate": 7.820303792545558e-07, + "loss": 1.0744, + "step": 15535 + }, + { + "epoch": 0.943921258885716, + "grad_norm": 0.14298313856124878, + "learning_rate": 7.803448263887237e-07, + "loss": 1.0133, + "step": 15536 + }, + { + "epoch": 0.9439820159183425, + "grad_norm": 0.449429452419281, + "learning_rate": 7.78661077678855e-07, + "loss": 1.1283, + "step": 15537 + }, + { + "epoch": 0.9440427729509691, + "grad_norm": 0.1932600438594818, + "learning_rate": 7.769791331866894e-07, + "loss": 1.1196, + "step": 15538 + }, + { + "epoch": 0.9441035299835956, + "grad_norm": 0.15955592691898346, + "learning_rate": 7.752989929738663e-07, + "loss": 1.067, + "step": 15539 + }, + { + "epoch": 0.9441642870162221, + "grad_norm": 0.16048896312713623, + "learning_rate": 7.73620657101981e-07, + "loss": 1.0601, + "step": 15540 + }, + { + "epoch": 0.9442250440488487, + "grad_norm": 0.32528069615364075, + "learning_rate": 7.719441256325399e-07, + "loss": 1.1432, + "step": 15541 + }, + { + "epoch": 0.9442858010814752, + "grad_norm": 0.17604315280914307, + "learning_rate": 7.702693986269938e-07, + "loss": 1.1464, + "step": 15542 + }, + { + "epoch": 0.9443465581141017, + "grad_norm": 0.18795163929462433, + "learning_rate": 7.685964761467491e-07, + "loss": 1.0728, + "step": 15543 + }, + { + "epoch": 0.9444073151467283, + "grad_norm": 0.15914508700370789, + "learning_rate": 7.669253582531122e-07, + "loss": 1.0491, + "step": 15544 + }, + { + "epoch": 0.9444680721793548, + "grad_norm": 0.21286284923553467, + "learning_rate": 7.652560450073454e-07, + "loss": 1.1173, + "step": 15545 + }, + { + "epoch": 0.9445288292119813, + "grad_norm": 0.12124127894639969, + "learning_rate": 7.635885364706385e-07, + "loss": 1.0634, + "step": 15546 + }, + { + "epoch": 0.9445895862446078, + "grad_norm": 0.15473072230815887, + "learning_rate": 7.61922832704104e-07, + "loss": 1.0519, + "step": 15547 + }, + { + "epoch": 0.9446503432772343, + "grad_norm": 0.11602921783924103, + "learning_rate": 7.602589337688093e-07, + "loss": 1.0087, + "step": 15548 + }, + { + "epoch": 0.9447111003098608, + "grad_norm": 0.21966996788978577, + "learning_rate": 7.585968397257392e-07, + "loss": 1.0951, + "step": 15549 + }, + { + "epoch": 0.9447718573424874, + "grad_norm": 0.1505543440580368, + "learning_rate": 7.569365506358228e-07, + "loss": 1.0419, + "step": 15550 + }, + { + "epoch": 0.9448326143751139, + "grad_norm": 0.12142287939786911, + "learning_rate": 7.552780665599113e-07, + "loss": 0.9829, + "step": 15551 + }, + { + "epoch": 0.9448933714077404, + "grad_norm": 0.18807706236839294, + "learning_rate": 7.536213875588061e-07, + "loss": 1.0623, + "step": 15552 + }, + { + "epoch": 0.944954128440367, + "grad_norm": 0.19684258103370667, + "learning_rate": 7.519665136932252e-07, + "loss": 1.0529, + "step": 15553 + }, + { + "epoch": 0.9450148854729935, + "grad_norm": 0.7728477120399475, + "learning_rate": 7.503134450238314e-07, + "loss": 1.3065, + "step": 15554 + }, + { + "epoch": 0.94507564250562, + "grad_norm": 0.8436038494110107, + "learning_rate": 7.486621816112149e-07, + "loss": 1.0897, + "step": 15555 + }, + { + "epoch": 0.9451363995382466, + "grad_norm": 0.19557011127471924, + "learning_rate": 7.470127235159108e-07, + "loss": 1.0953, + "step": 15556 + }, + { + "epoch": 0.9451971565708731, + "grad_norm": 0.19505125284194946, + "learning_rate": 7.453650707983761e-07, + "loss": 1.0418, + "step": 15557 + }, + { + "epoch": 0.9452579136034996, + "grad_norm": 0.826728880405426, + "learning_rate": 7.437192235190016e-07, + "loss": 1.0439, + "step": 15558 + }, + { + "epoch": 0.9453186706361262, + "grad_norm": 0.20165561139583588, + "learning_rate": 7.420751817381222e-07, + "loss": 1.1216, + "step": 15559 + }, + { + "epoch": 0.9453794276687526, + "grad_norm": 0.19503562152385712, + "learning_rate": 7.404329455160008e-07, + "loss": 1.0456, + "step": 15560 + }, + { + "epoch": 0.9454401847013791, + "grad_norm": 0.2051047831773758, + "learning_rate": 7.387925149128283e-07, + "loss": 1.0879, + "step": 15561 + }, + { + "epoch": 0.9455009417340057, + "grad_norm": 0.2587934136390686, + "learning_rate": 7.371538899887398e-07, + "loss": 1.1162, + "step": 15562 + }, + { + "epoch": 0.9455616987666322, + "grad_norm": 0.2388824075460434, + "learning_rate": 7.355170708037928e-07, + "loss": 1.1122, + "step": 15563 + }, + { + "epoch": 0.9456224557992587, + "grad_norm": 0.25501003861427307, + "learning_rate": 7.33882057417995e-07, + "loss": 1.2129, + "step": 15564 + }, + { + "epoch": 0.9456832128318853, + "grad_norm": 0.13585162162780762, + "learning_rate": 7.322488498912705e-07, + "loss": 1.0476, + "step": 15565 + }, + { + "epoch": 0.9457439698645118, + "grad_norm": 0.22533878684043884, + "learning_rate": 7.306174482834938e-07, + "loss": 1.1002, + "step": 15566 + }, + { + "epoch": 0.9458047268971383, + "grad_norm": 1.702141523361206, + "learning_rate": 7.289878526544503e-07, + "loss": 1.2204, + "step": 15567 + }, + { + "epoch": 0.9458654839297649, + "grad_norm": 3.5431642532348633, + "learning_rate": 7.273600630638866e-07, + "loss": 1.3854, + "step": 15568 + }, + { + "epoch": 0.9459262409623914, + "grad_norm": 0.2770073711872101, + "learning_rate": 7.257340795714607e-07, + "loss": 1.0711, + "step": 15569 + }, + { + "epoch": 0.945986997995018, + "grad_norm": 0.37590470910072327, + "learning_rate": 7.241099022367803e-07, + "loss": 1.1181, + "step": 15570 + }, + { + "epoch": 0.9460477550276445, + "grad_norm": 0.18525652587413788, + "learning_rate": 7.224875311193757e-07, + "loss": 1.067, + "step": 15571 + }, + { + "epoch": 0.946108512060271, + "grad_norm": 0.6307420134544373, + "learning_rate": 7.208669662787104e-07, + "loss": 1.0794, + "step": 15572 + }, + { + "epoch": 0.9461692690928976, + "grad_norm": 0.14861057698726654, + "learning_rate": 7.19248207774198e-07, + "loss": 1.0373, + "step": 15573 + }, + { + "epoch": 0.946230026125524, + "grad_norm": 0.21414567530155182, + "learning_rate": 7.176312556651632e-07, + "loss": 1.0921, + "step": 15574 + }, + { + "epoch": 0.9462907831581505, + "grad_norm": 0.17335575819015503, + "learning_rate": 7.160161100108864e-07, + "loss": 1.0939, + "step": 15575 + }, + { + "epoch": 0.946351540190777, + "grad_norm": 0.1607893705368042, + "learning_rate": 7.144027708705592e-07, + "loss": 1.0514, + "step": 15576 + }, + { + "epoch": 0.9464122972234036, + "grad_norm": 0.1302529126405716, + "learning_rate": 7.127912383033286e-07, + "loss": 1.0724, + "step": 15577 + }, + { + "epoch": 0.9464730542560301, + "grad_norm": 0.16422320902347565, + "learning_rate": 7.111815123682641e-07, + "loss": 1.0085, + "step": 15578 + }, + { + "epoch": 0.9465338112886567, + "grad_norm": 0.14539048075675964, + "learning_rate": 7.09573593124363e-07, + "loss": 1.0355, + "step": 15579 + }, + { + "epoch": 0.9465945683212832, + "grad_norm": 0.17410799860954285, + "learning_rate": 7.07967480630578e-07, + "loss": 1.0275, + "step": 15580 + }, + { + "epoch": 0.9466553253539097, + "grad_norm": 0.12437879294157028, + "learning_rate": 7.063631749457677e-07, + "loss": 1.0399, + "step": 15581 + }, + { + "epoch": 0.9467160823865363, + "grad_norm": 0.21448512375354767, + "learning_rate": 7.047606761287406e-07, + "loss": 1.1122, + "step": 15582 + }, + { + "epoch": 0.9467768394191628, + "grad_norm": 9.690837860107422, + "learning_rate": 7.031599842382386e-07, + "loss": 1.0382, + "step": 15583 + }, + { + "epoch": 0.9468375964517893, + "grad_norm": 0.44579043984413147, + "learning_rate": 7.015610993329424e-07, + "loss": 1.1191, + "step": 15584 + }, + { + "epoch": 0.9468983534844159, + "grad_norm": 0.1512700915336609, + "learning_rate": 6.999640214714498e-07, + "loss": 1.069, + "step": 15585 + }, + { + "epoch": 0.9469591105170424, + "grad_norm": 0.18287241458892822, + "learning_rate": 6.983687507123138e-07, + "loss": 1.2102, + "step": 15586 + }, + { + "epoch": 0.9470198675496688, + "grad_norm": 0.15826700627803802, + "learning_rate": 6.967752871139932e-07, + "loss": 1.0331, + "step": 15587 + }, + { + "epoch": 0.9470806245822954, + "grad_norm": 0.1521591991186142, + "learning_rate": 6.951836307349025e-07, + "loss": 1.0033, + "step": 15588 + }, + { + "epoch": 0.9471413816149219, + "grad_norm": 0.15993905067443848, + "learning_rate": 6.935937816333837e-07, + "loss": 1.0148, + "step": 15589 + }, + { + "epoch": 0.9472021386475484, + "grad_norm": 0.2883405387401581, + "learning_rate": 6.920057398677238e-07, + "loss": 1.1705, + "step": 15590 + }, + { + "epoch": 0.947262895680175, + "grad_norm": 0.17201805114746094, + "learning_rate": 6.904195054961204e-07, + "loss": 1.1171, + "step": 15591 + }, + { + "epoch": 0.9473236527128015, + "grad_norm": 0.22900508344173431, + "learning_rate": 6.888350785767217e-07, + "loss": 1.1923, + "step": 15592 + }, + { + "epoch": 0.947384409745428, + "grad_norm": 4.1723527908325195, + "learning_rate": 6.872524591676089e-07, + "loss": 1.0097, + "step": 15593 + }, + { + "epoch": 0.9474451667780546, + "grad_norm": 0.17712073028087616, + "learning_rate": 6.856716473267854e-07, + "loss": 1.0713, + "step": 15594 + }, + { + "epoch": 0.9475059238106811, + "grad_norm": 0.350065141916275, + "learning_rate": 6.840926431121997e-07, + "loss": 1.1572, + "step": 15595 + }, + { + "epoch": 0.9475666808433076, + "grad_norm": 0.1414537876844406, + "learning_rate": 6.82515446581733e-07, + "loss": 1.042, + "step": 15596 + }, + { + "epoch": 0.9476274378759342, + "grad_norm": 0.3665454387664795, + "learning_rate": 6.809400577931946e-07, + "loss": 1.1275, + "step": 15597 + }, + { + "epoch": 0.9476881949085607, + "grad_norm": 0.4751051664352417, + "learning_rate": 6.793664768043273e-07, + "loss": 1.0972, + "step": 15598 + }, + { + "epoch": 0.9477489519411872, + "grad_norm": 0.20352542400360107, + "learning_rate": 6.777947036728182e-07, + "loss": 1.0484, + "step": 15599 + }, + { + "epoch": 0.9478097089738137, + "grad_norm": 0.1469539850950241, + "learning_rate": 6.762247384562825e-07, + "loss": 1.0618, + "step": 15600 + }, + { + "epoch": 0.9478704660064402, + "grad_norm": 0.21546193957328796, + "learning_rate": 6.746565812122574e-07, + "loss": 1.1412, + "step": 15601 + }, + { + "epoch": 0.9479312230390667, + "grad_norm": 0.22022108733654022, + "learning_rate": 6.730902319982302e-07, + "loss": 1.1942, + "step": 15602 + }, + { + "epoch": 0.9479919800716933, + "grad_norm": 0.17324426770210266, + "learning_rate": 6.715256908716161e-07, + "loss": 1.0459, + "step": 15603 + }, + { + "epoch": 0.9480527371043198, + "grad_norm": 0.3780989646911621, + "learning_rate": 6.699629578897637e-07, + "loss": 0.9785, + "step": 15604 + }, + { + "epoch": 0.9481134941369463, + "grad_norm": 0.30604737997055054, + "learning_rate": 6.684020331099494e-07, + "loss": 1.1966, + "step": 15605 + }, + { + "epoch": 0.9481742511695729, + "grad_norm": 0.18010807037353516, + "learning_rate": 6.668429165893997e-07, + "loss": 1.0952, + "step": 15606 + }, + { + "epoch": 0.9482350082021994, + "grad_norm": 0.29873424768447876, + "learning_rate": 6.652856083852632e-07, + "loss": 1.0788, + "step": 15607 + }, + { + "epoch": 0.948295765234826, + "grad_norm": 0.12083946168422699, + "learning_rate": 6.637301085546111e-07, + "loss": 1.0261, + "step": 15608 + }, + { + "epoch": 0.9483565222674525, + "grad_norm": 0.1257619559764862, + "learning_rate": 6.621764171544698e-07, + "loss": 1.0093, + "step": 15609 + }, + { + "epoch": 0.948417279300079, + "grad_norm": 0.12082166969776154, + "learning_rate": 6.606245342417882e-07, + "loss": 1.0237, + "step": 15610 + }, + { + "epoch": 0.9484780363327056, + "grad_norm": 0.354981929063797, + "learning_rate": 6.590744598734544e-07, + "loss": 0.9899, + "step": 15611 + }, + { + "epoch": 0.9485387933653321, + "grad_norm": 0.5237496495246887, + "learning_rate": 6.575261941062838e-07, + "loss": 1.1712, + "step": 15612 + }, + { + "epoch": 0.9485995503979585, + "grad_norm": 8.183042526245117, + "learning_rate": 6.559797369970256e-07, + "loss": 0.9617, + "step": 15613 + }, + { + "epoch": 0.948660307430585, + "grad_norm": 0.11655563861131668, + "learning_rate": 6.544350886023675e-07, + "loss": 1.0589, + "step": 15614 + }, + { + "epoch": 0.9487210644632116, + "grad_norm": 0.2279948592185974, + "learning_rate": 6.528922489789313e-07, + "loss": 1.1071, + "step": 15615 + }, + { + "epoch": 0.9487818214958381, + "grad_norm": 0.7739730477333069, + "learning_rate": 6.513512181832659e-07, + "loss": 1.1233, + "step": 15616 + }, + { + "epoch": 0.9488425785284647, + "grad_norm": 0.13703100383281708, + "learning_rate": 6.498119962718652e-07, + "loss": 1.0484, + "step": 15617 + }, + { + "epoch": 0.9489033355610912, + "grad_norm": 0.13347269594669342, + "learning_rate": 6.482745833011394e-07, + "loss": 1.0262, + "step": 15618 + }, + { + "epoch": 0.9489640925937177, + "grad_norm": 0.14487218856811523, + "learning_rate": 6.467389793274547e-07, + "loss": 1.0553, + "step": 15619 + }, + { + "epoch": 0.9490248496263443, + "grad_norm": 0.1882328987121582, + "learning_rate": 6.452051844070883e-07, + "loss": 0.9908, + "step": 15620 + }, + { + "epoch": 0.9490856066589708, + "grad_norm": 0.2310590296983719, + "learning_rate": 6.436731985962674e-07, + "loss": 1.0799, + "step": 15621 + }, + { + "epoch": 0.9491463636915973, + "grad_norm": 0.24600329995155334, + "learning_rate": 6.421430219511415e-07, + "loss": 1.0867, + "step": 15622 + }, + { + "epoch": 0.9492071207242239, + "grad_norm": 0.22335784137248993, + "learning_rate": 6.406146545278102e-07, + "loss": 1.0792, + "step": 15623 + }, + { + "epoch": 0.9492678777568504, + "grad_norm": 0.2036837637424469, + "learning_rate": 6.390880963822842e-07, + "loss": 1.1073, + "step": 15624 + }, + { + "epoch": 0.9493286347894769, + "grad_norm": 0.11619195342063904, + "learning_rate": 6.375633475705245e-07, + "loss": 0.9886, + "step": 15625 + }, + { + "epoch": 0.9493893918221034, + "grad_norm": 0.2378149926662445, + "learning_rate": 6.36040408148425e-07, + "loss": 1.1461, + "step": 15626 + }, + { + "epoch": 0.9494501488547299, + "grad_norm": 0.13937169313430786, + "learning_rate": 6.345192781718134e-07, + "loss": 1.0201, + "step": 15627 + }, + { + "epoch": 0.9495109058873564, + "grad_norm": 3.907757043838501, + "learning_rate": 6.329999576964341e-07, + "loss": 1.0186, + "step": 15628 + }, + { + "epoch": 0.949571662919983, + "grad_norm": 0.8481959104537964, + "learning_rate": 6.314824467779812e-07, + "loss": 1.0139, + "step": 15629 + }, + { + "epoch": 0.9496324199526095, + "grad_norm": 0.1354498267173767, + "learning_rate": 6.29966745472077e-07, + "loss": 1.0141, + "step": 15630 + }, + { + "epoch": 0.949693176985236, + "grad_norm": 2.6860108375549316, + "learning_rate": 6.284528538342937e-07, + "loss": 1.0564, + "step": 15631 + }, + { + "epoch": 0.9497539340178626, + "grad_norm": 3.7938923835754395, + "learning_rate": 6.269407719201092e-07, + "loss": 1.0462, + "step": 15632 + }, + { + "epoch": 0.9498146910504891, + "grad_norm": 0.13726535439491272, + "learning_rate": 6.254304997849624e-07, + "loss": 1.0431, + "step": 15633 + }, + { + "epoch": 0.9498754480831156, + "grad_norm": 0.237648606300354, + "learning_rate": 6.239220374841981e-07, + "loss": 1.1135, + "step": 15634 + }, + { + "epoch": 0.9499362051157422, + "grad_norm": 0.11887457966804504, + "learning_rate": 6.224153850731107e-07, + "loss": 0.9887, + "step": 15635 + }, + { + "epoch": 0.9499969621483687, + "grad_norm": 3.0066018104553223, + "learning_rate": 6.209105426069395e-07, + "loss": 1.2417, + "step": 15636 + }, + { + "epoch": 0.9500577191809952, + "grad_norm": 0.13448326289653778, + "learning_rate": 6.194075101408348e-07, + "loss": 1.0255, + "step": 15637 + }, + { + "epoch": 0.9501184762136218, + "grad_norm": 0.6042473912239075, + "learning_rate": 6.179062877298969e-07, + "loss": 1.0119, + "step": 15638 + }, + { + "epoch": 0.9501792332462482, + "grad_norm": 0.1876097321510315, + "learning_rate": 6.16406875429143e-07, + "loss": 1.092, + "step": 15639 + }, + { + "epoch": 0.9502399902788747, + "grad_norm": 0.11830680072307587, + "learning_rate": 6.149092732935457e-07, + "loss": 1.0253, + "step": 15640 + }, + { + "epoch": 0.9503007473115013, + "grad_norm": 0.5267478823661804, + "learning_rate": 6.13413481377989e-07, + "loss": 1.0838, + "step": 15641 + }, + { + "epoch": 0.9503615043441278, + "grad_norm": 0.1767970770597458, + "learning_rate": 6.119194997373123e-07, + "loss": 1.0509, + "step": 15642 + }, + { + "epoch": 0.9504222613767543, + "grad_norm": 0.16034843027591705, + "learning_rate": 6.104273284262718e-07, + "loss": 1.0834, + "step": 15643 + }, + { + "epoch": 0.9504830184093809, + "grad_norm": 0.11588355153799057, + "learning_rate": 6.089369674995626e-07, + "loss": 1.0389, + "step": 15644 + }, + { + "epoch": 0.9505437754420074, + "grad_norm": 0.12805397808551788, + "learning_rate": 6.074484170118188e-07, + "loss": 1.0428, + "step": 15645 + }, + { + "epoch": 0.950604532474634, + "grad_norm": 0.12749528884887695, + "learning_rate": 6.059616770175969e-07, + "loss": 1.0047, + "step": 15646 + }, + { + "epoch": 0.9506652895072605, + "grad_norm": 0.18255400657653809, + "learning_rate": 6.044767475714031e-07, + "loss": 1.162, + "step": 15647 + }, + { + "epoch": 0.950726046539887, + "grad_norm": 0.11360357701778412, + "learning_rate": 6.029936287276605e-07, + "loss": 0.9616, + "step": 15648 + }, + { + "epoch": 0.9507868035725136, + "grad_norm": 0.18397608399391174, + "learning_rate": 6.015123205407313e-07, + "loss": 1.0617, + "step": 15649 + }, + { + "epoch": 0.9508475606051401, + "grad_norm": 0.11834269016981125, + "learning_rate": 6.000328230649166e-07, + "loss": 1.0284, + "step": 15650 + }, + { + "epoch": 0.9509083176377666, + "grad_norm": 0.12310358881950378, + "learning_rate": 5.98555136354445e-07, + "loss": 0.9992, + "step": 15651 + }, + { + "epoch": 0.950969074670393, + "grad_norm": 0.13682183623313904, + "learning_rate": 5.970792604634901e-07, + "loss": 1.0454, + "step": 15652 + }, + { + "epoch": 0.9510298317030196, + "grad_norm": 0.1247885525226593, + "learning_rate": 5.956051954461472e-07, + "loss": 1.0485, + "step": 15653 + }, + { + "epoch": 0.9510905887356461, + "grad_norm": 0.18931463360786438, + "learning_rate": 5.941329413564401e-07, + "loss": 1.108, + "step": 15654 + }, + { + "epoch": 0.9511513457682726, + "grad_norm": 0.21694165468215942, + "learning_rate": 5.926624982483364e-07, + "loss": 1.0325, + "step": 15655 + }, + { + "epoch": 0.9512121028008992, + "grad_norm": 0.11654436588287354, + "learning_rate": 5.911938661757433e-07, + "loss": 0.9885, + "step": 15656 + }, + { + "epoch": 0.9512728598335257, + "grad_norm": 0.13640858232975006, + "learning_rate": 5.897270451924897e-07, + "loss": 1.036, + "step": 15657 + }, + { + "epoch": 0.9513336168661523, + "grad_norm": 0.10796596109867096, + "learning_rate": 5.882620353523383e-07, + "loss": 0.9636, + "step": 15658 + }, + { + "epoch": 0.9513943738987788, + "grad_norm": 0.12118857353925705, + "learning_rate": 5.867988367089961e-07, + "loss": 1.0353, + "step": 15659 + }, + { + "epoch": 0.9514551309314053, + "grad_norm": 3.663520097732544, + "learning_rate": 5.853374493160979e-07, + "loss": 1.0862, + "step": 15660 + }, + { + "epoch": 0.9515158879640319, + "grad_norm": 0.31335803866386414, + "learning_rate": 5.83877873227201e-07, + "loss": 1.0196, + "step": 15661 + }, + { + "epoch": 0.9515766449966584, + "grad_norm": 0.15128333866596222, + "learning_rate": 5.824201084958181e-07, + "loss": 1.0838, + "step": 15662 + }, + { + "epoch": 0.9516374020292849, + "grad_norm": 0.13150979578495026, + "learning_rate": 5.809641551753731e-07, + "loss": 1.0726, + "step": 15663 + }, + { + "epoch": 0.9516981590619115, + "grad_norm": 0.11511262506246567, + "learning_rate": 5.7951001331924e-07, + "loss": 1.0191, + "step": 15664 + }, + { + "epoch": 0.9517589160945379, + "grad_norm": 0.20673958957195282, + "learning_rate": 5.78057682980726e-07, + "loss": 1.1013, + "step": 15665 + }, + { + "epoch": 0.9518196731271644, + "grad_norm": 0.15394078195095062, + "learning_rate": 5.766071642130555e-07, + "loss": 0.9901, + "step": 15666 + }, + { + "epoch": 0.951880430159791, + "grad_norm": 0.17244891822338104, + "learning_rate": 5.751584570694024e-07, + "loss": 1.0379, + "step": 15667 + }, + { + "epoch": 0.9519411871924175, + "grad_norm": 0.11609045416116714, + "learning_rate": 5.737115616028687e-07, + "loss": 1.0012, + "step": 15668 + }, + { + "epoch": 0.952001944225044, + "grad_norm": 0.15334002673625946, + "learning_rate": 5.722664778664955e-07, + "loss": 1.0571, + "step": 15669 + }, + { + "epoch": 0.9520627012576706, + "grad_norm": 0.14956983923912048, + "learning_rate": 5.708232059132456e-07, + "loss": 1.0253, + "step": 15670 + }, + { + "epoch": 0.9521234582902971, + "grad_norm": 0.12743394076824188, + "learning_rate": 5.693817457960271e-07, + "loss": 1.0306, + "step": 15671 + }, + { + "epoch": 0.9521842153229236, + "grad_norm": 0.11378083378076553, + "learning_rate": 5.679420975676753e-07, + "loss": 1.0239, + "step": 15672 + }, + { + "epoch": 0.9522449723555502, + "grad_norm": 0.1648378223180771, + "learning_rate": 5.66504261280959e-07, + "loss": 1.0233, + "step": 15673 + }, + { + "epoch": 0.9523057293881767, + "grad_norm": 0.48625320196151733, + "learning_rate": 5.650682369885918e-07, + "loss": 1.1734, + "step": 15674 + }, + { + "epoch": 0.9523664864208032, + "grad_norm": 0.15202584862709045, + "learning_rate": 5.636340247431981e-07, + "loss": 1.0153, + "step": 15675 + }, + { + "epoch": 0.9524272434534298, + "grad_norm": 0.22682605683803558, + "learning_rate": 5.622016245973527e-07, + "loss": 1.161, + "step": 15676 + }, + { + "epoch": 0.9524880004860563, + "grad_norm": 0.21841900050640106, + "learning_rate": 5.607710366035634e-07, + "loss": 1.1463, + "step": 15677 + }, + { + "epoch": 0.9525487575186828, + "grad_norm": 1.9286853075027466, + "learning_rate": 5.593422608142718e-07, + "loss": 1.1782, + "step": 15678 + }, + { + "epoch": 0.9526095145513093, + "grad_norm": 0.15244348347187042, + "learning_rate": 5.579152972818469e-07, + "loss": 1.0617, + "step": 15679 + }, + { + "epoch": 0.9526702715839358, + "grad_norm": 0.2120814323425293, + "learning_rate": 5.564901460585914e-07, + "loss": 1.1145, + "step": 15680 + }, + { + "epoch": 0.9527310286165623, + "grad_norm": 0.2721669673919678, + "learning_rate": 5.550668071967469e-07, + "loss": 1.077, + "step": 15681 + }, + { + "epoch": 0.9527917856491889, + "grad_norm": 0.23098333179950714, + "learning_rate": 5.536452807484827e-07, + "loss": 1.0787, + "step": 15682 + }, + { + "epoch": 0.9528525426818154, + "grad_norm": 0.16313430666923523, + "learning_rate": 5.522255667659127e-07, + "loss": 1.0076, + "step": 15683 + }, + { + "epoch": 0.9529132997144419, + "grad_norm": 0.21127358078956604, + "learning_rate": 5.508076653010729e-07, + "loss": 1.1678, + "step": 15684 + }, + { + "epoch": 0.9529740567470685, + "grad_norm": 0.17304788529872894, + "learning_rate": 5.493915764059332e-07, + "loss": 1.0346, + "step": 15685 + }, + { + "epoch": 0.953034813779695, + "grad_norm": 0.20408296585083008, + "learning_rate": 5.479773001324074e-07, + "loss": 1.245, + "step": 15686 + }, + { + "epoch": 0.9530955708123215, + "grad_norm": 0.1257437765598297, + "learning_rate": 5.465648365323317e-07, + "loss": 0.9772, + "step": 15687 + }, + { + "epoch": 0.9531563278449481, + "grad_norm": 0.23594969511032104, + "learning_rate": 5.451541856574815e-07, + "loss": 1.1301, + "step": 15688 + }, + { + "epoch": 0.9532170848775746, + "grad_norm": 0.20644228160381317, + "learning_rate": 5.4374534755956e-07, + "loss": 1.1526, + "step": 15689 + }, + { + "epoch": 0.9532778419102012, + "grad_norm": 0.5140280723571777, + "learning_rate": 5.4233832229022e-07, + "loss": 1.2274, + "step": 15690 + }, + { + "epoch": 0.9533385989428277, + "grad_norm": 0.21546581387519836, + "learning_rate": 5.409331099010206e-07, + "loss": 1.1021, + "step": 15691 + }, + { + "epoch": 0.9533993559754541, + "grad_norm": 3.047247886657715, + "learning_rate": 5.395297104434815e-07, + "loss": 1.2332, + "step": 15692 + }, + { + "epoch": 0.9534601130080806, + "grad_norm": 0.2695804536342621, + "learning_rate": 5.381281239690394e-07, + "loss": 1.1973, + "step": 15693 + }, + { + "epoch": 0.9535208700407072, + "grad_norm": 0.15590772032737732, + "learning_rate": 5.36728350529081e-07, + "loss": 1.0401, + "step": 15694 + }, + { + "epoch": 0.9535816270733337, + "grad_norm": 0.11646371334791183, + "learning_rate": 5.353303901748985e-07, + "loss": 0.9768, + "step": 15695 + }, + { + "epoch": 0.9536423841059603, + "grad_norm": 0.17041917145252228, + "learning_rate": 5.339342429577398e-07, + "loss": 1.0795, + "step": 15696 + }, + { + "epoch": 0.9537031411385868, + "grad_norm": 0.16379816830158234, + "learning_rate": 5.325399089287809e-07, + "loss": 1.0727, + "step": 15697 + }, + { + "epoch": 0.9537638981712133, + "grad_norm": 0.2524796426296234, + "learning_rate": 5.311473881391415e-07, + "loss": 1.1394, + "step": 15698 + }, + { + "epoch": 0.9538246552038399, + "grad_norm": 0.17920872569084167, + "learning_rate": 5.297566806398535e-07, + "loss": 1.0103, + "step": 15699 + }, + { + "epoch": 0.9538854122364664, + "grad_norm": 0.11340802907943726, + "learning_rate": 5.283677864818981e-07, + "loss": 1.0594, + "step": 15700 + }, + { + "epoch": 0.9539461692690929, + "grad_norm": 14.428019523620605, + "learning_rate": 5.269807057161847e-07, + "loss": 1.0336, + "step": 15701 + }, + { + "epoch": 0.9540069263017195, + "grad_norm": 0.16854970157146454, + "learning_rate": 5.255954383935502e-07, + "loss": 1.0722, + "step": 15702 + }, + { + "epoch": 0.954067683334346, + "grad_norm": 0.37928712368011475, + "learning_rate": 5.242119845647819e-07, + "loss": 1.0325, + "step": 15703 + }, + { + "epoch": 0.9541284403669725, + "grad_norm": 0.15976952016353607, + "learning_rate": 5.228303442805893e-07, + "loss": 1.0752, + "step": 15704 + }, + { + "epoch": 0.954189197399599, + "grad_norm": 0.1680813580751419, + "learning_rate": 5.214505175916096e-07, + "loss": 1.074, + "step": 15705 + }, + { + "epoch": 0.9542499544322255, + "grad_norm": 0.18805907666683197, + "learning_rate": 5.2007250454843e-07, + "loss": 1.0487, + "step": 15706 + }, + { + "epoch": 0.954310711464852, + "grad_norm": 0.12236303091049194, + "learning_rate": 5.186963052015547e-07, + "loss": 1.0034, + "step": 15707 + }, + { + "epoch": 0.9543714684974786, + "grad_norm": 0.2763650715351105, + "learning_rate": 5.173219196014267e-07, + "loss": 1.0497, + "step": 15708 + }, + { + "epoch": 0.9544322255301051, + "grad_norm": 0.22458645701408386, + "learning_rate": 5.159493477984335e-07, + "loss": 1.0427, + "step": 15709 + }, + { + "epoch": 0.9544929825627316, + "grad_norm": 0.11679227650165558, + "learning_rate": 5.145785898428846e-07, + "loss": 1.1822, + "step": 15710 + }, + { + "epoch": 0.9545537395953582, + "grad_norm": 0.21832974255084991, + "learning_rate": 5.132096457850177e-07, + "loss": 1.0752, + "step": 15711 + }, + { + "epoch": 0.9546144966279847, + "grad_norm": 0.1513703465461731, + "learning_rate": 5.118425156750206e-07, + "loss": 1.0209, + "step": 15712 + }, + { + "epoch": 0.9546752536606112, + "grad_norm": 0.13795199990272522, + "learning_rate": 5.10477199562992e-07, + "loss": 1.0251, + "step": 15713 + }, + { + "epoch": 0.9547360106932378, + "grad_norm": 0.29039427638053894, + "learning_rate": 5.091136974990029e-07, + "loss": 1.1877, + "step": 15714 + }, + { + "epoch": 0.9547967677258643, + "grad_norm": 0.1210121288895607, + "learning_rate": 5.07752009533008e-07, + "loss": 1.0221, + "step": 15715 + }, + { + "epoch": 0.9548575247584908, + "grad_norm": 0.12347547709941864, + "learning_rate": 5.063921357149337e-07, + "loss": 1.0145, + "step": 15716 + }, + { + "epoch": 0.9549182817911174, + "grad_norm": 0.13484084606170654, + "learning_rate": 5.050340760946182e-07, + "loss": 1.0126, + "step": 15717 + }, + { + "epoch": 0.9549790388237438, + "grad_norm": 0.12226403504610062, + "learning_rate": 5.036778307218493e-07, + "loss": 0.9951, + "step": 15718 + }, + { + "epoch": 0.9550397958563703, + "grad_norm": 0.1258535087108612, + "learning_rate": 5.023233996463317e-07, + "loss": 1.0145, + "step": 15719 + }, + { + "epoch": 0.9551005528889969, + "grad_norm": 0.10771535336971283, + "learning_rate": 5.009707829177257e-07, + "loss": 0.9564, + "step": 15720 + }, + { + "epoch": 0.9551613099216234, + "grad_norm": 0.23255352675914764, + "learning_rate": 4.996199805856028e-07, + "loss": 1.0908, + "step": 15721 + }, + { + "epoch": 0.9552220669542499, + "grad_norm": 0.16564537584781647, + "learning_rate": 4.982709926994733e-07, + "loss": 1.0731, + "step": 15722 + }, + { + "epoch": 0.9552828239868765, + "grad_norm": 0.22524182498455048, + "learning_rate": 4.969238193087867e-07, + "loss": 1.1857, + "step": 15723 + }, + { + "epoch": 0.955343581019503, + "grad_norm": 0.27950331568717957, + "learning_rate": 4.95578460462931e-07, + "loss": 1.1037, + "step": 15724 + }, + { + "epoch": 0.9554043380521295, + "grad_norm": 0.19370266795158386, + "learning_rate": 4.94234916211217e-07, + "loss": 1.0277, + "step": 15725 + }, + { + "epoch": 0.9554650950847561, + "grad_norm": 0.15008951723575592, + "learning_rate": 4.928931866028885e-07, + "loss": 1.0273, + "step": 15726 + }, + { + "epoch": 0.9555258521173826, + "grad_norm": 0.1643773913383484, + "learning_rate": 4.915532716871341e-07, + "loss": 1.1098, + "step": 15727 + }, + { + "epoch": 0.9555866091500091, + "grad_norm": 0.41293343901634216, + "learning_rate": 4.902151715130587e-07, + "loss": 1.1261, + "step": 15728 + }, + { + "epoch": 0.9556473661826357, + "grad_norm": 0.12163228541612625, + "learning_rate": 4.888788861297178e-07, + "loss": 1.0259, + "step": 15729 + }, + { + "epoch": 0.9557081232152622, + "grad_norm": 0.13683412969112396, + "learning_rate": 4.875444155860886e-07, + "loss": 0.9766, + "step": 15730 + }, + { + "epoch": 0.9557688802478886, + "grad_norm": 0.17898204922676086, + "learning_rate": 4.862117599310933e-07, + "loss": 1.0822, + "step": 15731 + }, + { + "epoch": 0.9558296372805152, + "grad_norm": 0.11954610794782639, + "learning_rate": 4.848809192135761e-07, + "loss": 1.0395, + "step": 15732 + }, + { + "epoch": 0.9558903943131417, + "grad_norm": 0.6467819213867188, + "learning_rate": 4.835518934823203e-07, + "loss": 1.0874, + "step": 15733 + }, + { + "epoch": 0.9559511513457682, + "grad_norm": 0.18230871856212616, + "learning_rate": 4.822246827860422e-07, + "loss": 0.9803, + "step": 15734 + }, + { + "epoch": 0.9560119083783948, + "grad_norm": 0.132160946726799, + "learning_rate": 4.80899287173392e-07, + "loss": 1.0227, + "step": 15735 + }, + { + "epoch": 0.9560726654110213, + "grad_norm": 0.6312487721443176, + "learning_rate": 4.795757066929419e-07, + "loss": 1.3558, + "step": 15736 + }, + { + "epoch": 0.9561334224436479, + "grad_norm": 0.38087332248687744, + "learning_rate": 4.782539413932252e-07, + "loss": 1.1269, + "step": 15737 + }, + { + "epoch": 0.9561941794762744, + "grad_norm": 0.17015869915485382, + "learning_rate": 4.769339913226756e-07, + "loss": 1.0804, + "step": 15738 + }, + { + "epoch": 0.9562549365089009, + "grad_norm": 0.16035455465316772, + "learning_rate": 4.756158565296875e-07, + "loss": 1.0475, + "step": 15739 + }, + { + "epoch": 0.9563156935415275, + "grad_norm": 0.135816290974617, + "learning_rate": 4.7429953706257246e-07, + "loss": 1.0099, + "step": 15740 + }, + { + "epoch": 0.956376450574154, + "grad_norm": 0.13799645006656647, + "learning_rate": 4.7298503296958065e-07, + "loss": 1.0306, + "step": 15741 + }, + { + "epoch": 0.9564372076067805, + "grad_norm": 0.20718713104724884, + "learning_rate": 4.716723442988957e-07, + "loss": 1.1252, + "step": 15742 + }, + { + "epoch": 0.9564979646394071, + "grad_norm": 0.1449251025915146, + "learning_rate": 4.7036147109863484e-07, + "loss": 1.0768, + "step": 15743 + }, + { + "epoch": 0.9565587216720335, + "grad_norm": 0.21829570829868317, + "learning_rate": 4.690524134168428e-07, + "loss": 1.0981, + "step": 15744 + }, + { + "epoch": 0.95661947870466, + "grad_norm": 0.12830974161624908, + "learning_rate": 4.677451713015146e-07, + "loss": 1.0191, + "step": 15745 + }, + { + "epoch": 0.9566802357372866, + "grad_norm": 0.12319254130125046, + "learning_rate": 4.66439744800562e-07, + "loss": 1.0043, + "step": 15746 + }, + { + "epoch": 0.9567409927699131, + "grad_norm": 0.21136261522769928, + "learning_rate": 4.651361339618354e-07, + "loss": 1.0438, + "step": 15747 + }, + { + "epoch": 0.9568017498025396, + "grad_norm": 0.1658204048871994, + "learning_rate": 4.63834338833119e-07, + "loss": 1.0351, + "step": 15748 + }, + { + "epoch": 0.9568625068351662, + "grad_norm": 5.667989253997803, + "learning_rate": 4.6253435946212454e-07, + "loss": 1.119, + "step": 15749 + }, + { + "epoch": 0.9569232638677927, + "grad_norm": 0.14644819498062134, + "learning_rate": 4.612361958965139e-07, + "loss": 1.0287, + "step": 15750 + }, + { + "epoch": 0.9569840209004192, + "grad_norm": 0.20605774223804474, + "learning_rate": 4.5993984818386015e-07, + "loss": 1.0756, + "step": 15751 + }, + { + "epoch": 0.9570447779330458, + "grad_norm": 0.399675190448761, + "learning_rate": 4.586453163716919e-07, + "loss": 1.2111, + "step": 15752 + }, + { + "epoch": 0.9571055349656723, + "grad_norm": 0.19715170562267303, + "learning_rate": 4.5735260050745464e-07, + "loss": 1.0098, + "step": 15753 + }, + { + "epoch": 0.9571662919982988, + "grad_norm": 0.12527599930763245, + "learning_rate": 4.5606170063853816e-07, + "loss": 0.9665, + "step": 15754 + }, + { + "epoch": 0.9572270490309254, + "grad_norm": 0.26160117983818054, + "learning_rate": 4.54772616812249e-07, + "loss": 1.1422, + "step": 15755 + }, + { + "epoch": 0.9572878060635519, + "grad_norm": 0.15000265836715698, + "learning_rate": 4.5348534907585507e-07, + "loss": 1.0177, + "step": 15756 + }, + { + "epoch": 0.9573485630961783, + "grad_norm": 0.14371396601200104, + "learning_rate": 4.5219989747652404e-07, + "loss": 1.059, + "step": 15757 + }, + { + "epoch": 0.9574093201288049, + "grad_norm": 0.20526671409606934, + "learning_rate": 4.5091626206139047e-07, + "loss": 1.125, + "step": 15758 + }, + { + "epoch": 0.9574700771614314, + "grad_norm": 0.1375073343515396, + "learning_rate": 4.496344428774946e-07, + "loss": 1.0483, + "step": 15759 + }, + { + "epoch": 0.9575308341940579, + "grad_norm": 0.14336751401424408, + "learning_rate": 4.483544399718265e-07, + "loss": 1.0559, + "step": 15760 + }, + { + "epoch": 0.9575915912266845, + "grad_norm": 0.20404881238937378, + "learning_rate": 4.4707625339130976e-07, + "loss": 1.0493, + "step": 15761 + }, + { + "epoch": 0.957652348259311, + "grad_norm": 0.12977959215641022, + "learning_rate": 4.4579988318278477e-07, + "loss": 1.0136, + "step": 15762 + }, + { + "epoch": 0.9577131052919375, + "grad_norm": 0.145528182387352, + "learning_rate": 4.4452532939304734e-07, + "loss": 1.0169, + "step": 15763 + }, + { + "epoch": 0.9577738623245641, + "grad_norm": 0.38733938336372375, + "learning_rate": 4.432525920688102e-07, + "loss": 1.0292, + "step": 15764 + }, + { + "epoch": 0.9578346193571906, + "grad_norm": 0.16327251493930817, + "learning_rate": 4.4198167125672485e-07, + "loss": 1.0509, + "step": 15765 + }, + { + "epoch": 0.9578953763898171, + "grad_norm": 0.195705384016037, + "learning_rate": 4.4071256700338737e-07, + "loss": 1.0762, + "step": 15766 + }, + { + "epoch": 0.9579561334224437, + "grad_norm": 0.1258038580417633, + "learning_rate": 4.3944527935531053e-07, + "loss": 1.0067, + "step": 15767 + }, + { + "epoch": 0.9580168904550702, + "grad_norm": 0.2192327231168747, + "learning_rate": 4.381798083589461e-07, + "loss": 1.1395, + "step": 15768 + }, + { + "epoch": 0.9580776474876967, + "grad_norm": 0.18069805204868317, + "learning_rate": 4.3691615406067365e-07, + "loss": 1.0874, + "step": 15769 + }, + { + "epoch": 0.9581384045203232, + "grad_norm": 0.1759243607521057, + "learning_rate": 4.356543165068283e-07, + "loss": 1.0588, + "step": 15770 + }, + { + "epoch": 0.9581991615529497, + "grad_norm": 0.1160273402929306, + "learning_rate": 4.3439429574365086e-07, + "loss": 0.9949, + "step": 15771 + }, + { + "epoch": 0.9582599185855762, + "grad_norm": 0.1819160133600235, + "learning_rate": 4.331360918173322e-07, + "loss": 1.1189, + "step": 15772 + }, + { + "epoch": 0.9583206756182028, + "grad_norm": 0.155807763338089, + "learning_rate": 4.3187970477399087e-07, + "loss": 0.9957, + "step": 15773 + }, + { + "epoch": 0.9583814326508293, + "grad_norm": 0.13829417526721954, + "learning_rate": 4.306251346596846e-07, + "loss": 1.0861, + "step": 15774 + }, + { + "epoch": 0.9584421896834558, + "grad_norm": 0.12061191350221634, + "learning_rate": 4.293723815203876e-07, + "loss": 1.0278, + "step": 15775 + }, + { + "epoch": 0.9585029467160824, + "grad_norm": 0.12152165174484253, + "learning_rate": 4.2812144540203545e-07, + "loss": 0.9885, + "step": 15776 + }, + { + "epoch": 0.9585637037487089, + "grad_norm": 0.14399969577789307, + "learning_rate": 4.2687232635046923e-07, + "loss": 1.0732, + "step": 15777 + }, + { + "epoch": 0.9586244607813355, + "grad_norm": 0.17510230839252472, + "learning_rate": 4.256250244114801e-07, + "loss": 1.1073, + "step": 15778 + }, + { + "epoch": 0.958685217813962, + "grad_norm": 0.29502254724502563, + "learning_rate": 4.2437953963078704e-07, + "loss": 1.1557, + "step": 15779 + }, + { + "epoch": 0.9587459748465885, + "grad_norm": 0.11508319526910782, + "learning_rate": 4.231358720540479e-07, + "loss": 1.0098, + "step": 15780 + }, + { + "epoch": 0.9588067318792151, + "grad_norm": 0.33305883407592773, + "learning_rate": 4.2189402172684856e-07, + "loss": 1.3078, + "step": 15781 + }, + { + "epoch": 0.9588674889118416, + "grad_norm": 2.4770612716674805, + "learning_rate": 4.206539886947025e-07, + "loss": 1.035, + "step": 15782 + }, + { + "epoch": 0.9589282459444681, + "grad_norm": 0.12621639668941498, + "learning_rate": 4.194157730030679e-07, + "loss": 1.0384, + "step": 15783 + }, + { + "epoch": 0.9589890029770946, + "grad_norm": 0.13871322572231293, + "learning_rate": 4.1817937469733613e-07, + "loss": 1.0388, + "step": 15784 + }, + { + "epoch": 0.9590497600097211, + "grad_norm": 0.15086989104747772, + "learning_rate": 4.169447938228155e-07, + "loss": 1.0, + "step": 15785 + }, + { + "epoch": 0.9591105170423476, + "grad_norm": 0.12772372364997864, + "learning_rate": 4.1571203042477523e-07, + "loss": 1.0617, + "step": 15786 + }, + { + "epoch": 0.9591712740749742, + "grad_norm": 0.19866453111171722, + "learning_rate": 4.144810845483904e-07, + "loss": 1.0816, + "step": 15787 + }, + { + "epoch": 0.9592320311076007, + "grad_norm": 0.13530512154102325, + "learning_rate": 4.1325195623879153e-07, + "loss": 1.0467, + "step": 15788 + }, + { + "epoch": 0.9592927881402272, + "grad_norm": 0.15202058851718903, + "learning_rate": 4.120246455410204e-07, + "loss": 1.0308, + "step": 15789 + }, + { + "epoch": 0.9593535451728538, + "grad_norm": 0.22073738276958466, + "learning_rate": 4.1079915250007427e-07, + "loss": 1.0349, + "step": 15790 + }, + { + "epoch": 0.9594143022054803, + "grad_norm": 0.12134106457233429, + "learning_rate": 4.0957547716086733e-07, + "loss": 1.0339, + "step": 15791 + }, + { + "epoch": 0.9594750592381068, + "grad_norm": 0.2011730670928955, + "learning_rate": 4.083536195682636e-07, + "loss": 1.0904, + "step": 15792 + }, + { + "epoch": 0.9595358162707334, + "grad_norm": 0.1641150712966919, + "learning_rate": 4.0713357976703834e-07, + "loss": 1.0559, + "step": 15793 + }, + { + "epoch": 0.9595965733033599, + "grad_norm": 0.18348631262779236, + "learning_rate": 4.059153578019226e-07, + "loss": 1.0914, + "step": 15794 + }, + { + "epoch": 0.9596573303359864, + "grad_norm": 0.27773988246917725, + "learning_rate": 4.0469895371756383e-07, + "loss": 0.975, + "step": 15795 + }, + { + "epoch": 0.959718087368613, + "grad_norm": 0.25764867663383484, + "learning_rate": 4.034843675585487e-07, + "loss": 1.0826, + "step": 15796 + }, + { + "epoch": 0.9597788444012394, + "grad_norm": 0.12363221496343613, + "learning_rate": 4.022715993694026e-07, + "loss": 1.057, + "step": 15797 + }, + { + "epoch": 0.9598396014338659, + "grad_norm": 0.2588735818862915, + "learning_rate": 4.0106064919457897e-07, + "loss": 1.0151, + "step": 15798 + }, + { + "epoch": 0.9599003584664925, + "grad_norm": 0.12394178658723831, + "learning_rate": 3.9985151707846444e-07, + "loss": 0.991, + "step": 15799 + }, + { + "epoch": 0.959961115499119, + "grad_norm": 0.22941401600837708, + "learning_rate": 3.986442030653792e-07, + "loss": 1.1236, + "step": 15800 + }, + { + "epoch": 0.9600218725317455, + "grad_norm": 0.12507542967796326, + "learning_rate": 3.9743870719957664e-07, + "loss": 1.0402, + "step": 15801 + }, + { + "epoch": 0.9600826295643721, + "grad_norm": 3.131091833114624, + "learning_rate": 3.962350295252493e-07, + "loss": 1.0321, + "step": 15802 + }, + { + "epoch": 0.9601433865969986, + "grad_norm": 0.21746250987052917, + "learning_rate": 3.9503317008651173e-07, + "loss": 1.1772, + "step": 15803 + }, + { + "epoch": 0.9602041436296251, + "grad_norm": 0.21517686545848846, + "learning_rate": 3.938331289274233e-07, + "loss": 1.1082, + "step": 15804 + }, + { + "epoch": 0.9602649006622517, + "grad_norm": 0.2129553258419037, + "learning_rate": 3.9263490609197094e-07, + "loss": 1.14, + "step": 15805 + }, + { + "epoch": 0.9603256576948782, + "grad_norm": 0.16317510604858398, + "learning_rate": 3.914385016240751e-07, + "loss": 1.0739, + "step": 15806 + }, + { + "epoch": 0.9603864147275047, + "grad_norm": 0.1503862589597702, + "learning_rate": 3.902439155675897e-07, + "loss": 1.0419, + "step": 15807 + }, + { + "epoch": 0.9604471717601313, + "grad_norm": 0.2891089618206024, + "learning_rate": 3.8905114796630724e-07, + "loss": 1.0874, + "step": 15808 + }, + { + "epoch": 0.9605079287927578, + "grad_norm": 0.11986003816127777, + "learning_rate": 3.878601988639374e-07, + "loss": 1.0426, + "step": 15809 + }, + { + "epoch": 0.9605686858253842, + "grad_norm": 0.3048224151134491, + "learning_rate": 3.8667106830414525e-07, + "loss": 1.2976, + "step": 15810 + }, + { + "epoch": 0.9606294428580108, + "grad_norm": 0.18673205375671387, + "learning_rate": 3.8548375633051247e-07, + "loss": 1.122, + "step": 15811 + }, + { + "epoch": 0.9606901998906373, + "grad_norm": 0.1308933049440384, + "learning_rate": 3.8429826298656544e-07, + "loss": 1.0346, + "step": 15812 + }, + { + "epoch": 0.9607509569232638, + "grad_norm": 0.12401796132326126, + "learning_rate": 3.831145883157528e-07, + "loss": 1.072, + "step": 15813 + }, + { + "epoch": 0.9608117139558904, + "grad_norm": 2.4950575828552246, + "learning_rate": 3.81932732361473e-07, + "loss": 1.0276, + "step": 15814 + }, + { + "epoch": 0.9608724709885169, + "grad_norm": 0.169499933719635, + "learning_rate": 3.8075269516703596e-07, + "loss": 0.9904, + "step": 15815 + }, + { + "epoch": 0.9609332280211434, + "grad_norm": 0.14131030440330505, + "learning_rate": 3.795744767756959e-07, + "loss": 1.0556, + "step": 15816 + }, + { + "epoch": 0.96099398505377, + "grad_norm": 0.18881265819072723, + "learning_rate": 3.783980772306517e-07, + "loss": 1.1305, + "step": 15817 + }, + { + "epoch": 0.9610547420863965, + "grad_norm": 0.23628680408000946, + "learning_rate": 3.772234965750132e-07, + "loss": 1.1097, + "step": 15818 + }, + { + "epoch": 0.961115499119023, + "grad_norm": 0.12054502964019775, + "learning_rate": 3.760507348518405e-07, + "loss": 1.0154, + "step": 15819 + }, + { + "epoch": 0.9611762561516496, + "grad_norm": 0.27647507190704346, + "learning_rate": 3.748797921041214e-07, + "loss": 1.0405, + "step": 15820 + }, + { + "epoch": 0.9612370131842761, + "grad_norm": 0.22636257112026215, + "learning_rate": 3.7371066837477154e-07, + "loss": 1.0242, + "step": 15821 + }, + { + "epoch": 0.9612977702169027, + "grad_norm": 0.2907763123512268, + "learning_rate": 3.7254336370665666e-07, + "loss": 1.0105, + "step": 15822 + }, + { + "epoch": 0.9613585272495291, + "grad_norm": 0.132111594080925, + "learning_rate": 3.7137787814255365e-07, + "loss": 1.0495, + "step": 15823 + }, + { + "epoch": 0.9614192842821556, + "grad_norm": 0.2163810133934021, + "learning_rate": 3.702142117251839e-07, + "loss": 1.0672, + "step": 15824 + }, + { + "epoch": 0.9614800413147822, + "grad_norm": 0.22730477154254913, + "learning_rate": 3.690523644972077e-07, + "loss": 1.1675, + "step": 15825 + }, + { + "epoch": 0.9615407983474087, + "grad_norm": 0.1352747529745102, + "learning_rate": 3.678923365012132e-07, + "loss": 0.9764, + "step": 15826 + }, + { + "epoch": 0.9616015553800352, + "grad_norm": 0.1476755142211914, + "learning_rate": 3.667341277797165e-07, + "loss": 1.0243, + "step": 15827 + }, + { + "epoch": 0.9616623124126618, + "grad_norm": 0.11062777042388916, + "learning_rate": 3.6557773837517797e-07, + "loss": 1.0428, + "step": 15828 + }, + { + "epoch": 0.9617230694452883, + "grad_norm": 0.16851942241191864, + "learning_rate": 3.6442316832998035e-07, + "loss": 1.0558, + "step": 15829 + }, + { + "epoch": 0.9617838264779148, + "grad_norm": 0.48340800404548645, + "learning_rate": 3.6327041768643987e-07, + "loss": 1.0603, + "step": 15830 + }, + { + "epoch": 0.9618445835105414, + "grad_norm": 0.30548959970474243, + "learning_rate": 3.621194864868227e-07, + "loss": 1.0708, + "step": 15831 + }, + { + "epoch": 0.9619053405431679, + "grad_norm": 0.13712218403816223, + "learning_rate": 3.609703747733062e-07, + "loss": 1.0301, + "step": 15832 + }, + { + "epoch": 0.9619660975757944, + "grad_norm": 0.15498583018779755, + "learning_rate": 3.598230825880178e-07, + "loss": 1.0874, + "step": 15833 + }, + { + "epoch": 0.962026854608421, + "grad_norm": 0.12120738625526428, + "learning_rate": 3.586776099730127e-07, + "loss": 0.9587, + "step": 15834 + }, + { + "epoch": 0.9620876116410475, + "grad_norm": 0.17440903186798096, + "learning_rate": 3.57533956970274e-07, + "loss": 1.0678, + "step": 15835 + }, + { + "epoch": 0.9621483686736739, + "grad_norm": 0.11958181858062744, + "learning_rate": 3.5639212362172383e-07, + "loss": 1.0087, + "step": 15836 + }, + { + "epoch": 0.9622091257063005, + "grad_norm": 0.2171083390712738, + "learning_rate": 3.5525210996921187e-07, + "loss": 1.1934, + "step": 15837 + }, + { + "epoch": 0.962269882738927, + "grad_norm": 0.475858598947525, + "learning_rate": 3.5411391605453257e-07, + "loss": 1.1015, + "step": 15838 + }, + { + "epoch": 0.9623306397715535, + "grad_norm": 0.2285282164812088, + "learning_rate": 3.529775419194081e-07, + "loss": 1.0889, + "step": 15839 + }, + { + "epoch": 0.9623913968041801, + "grad_norm": 0.1661931425333023, + "learning_rate": 3.5184298760548296e-07, + "loss": 1.0624, + "step": 15840 + }, + { + "epoch": 0.9624521538368066, + "grad_norm": 0.2599220275878906, + "learning_rate": 3.507102531543516e-07, + "loss": 1.1811, + "step": 15841 + }, + { + "epoch": 0.9625129108694331, + "grad_norm": 0.2042168527841568, + "learning_rate": 3.495793386075308e-07, + "loss": 1.0857, + "step": 15842 + }, + { + "epoch": 0.9625736679020597, + "grad_norm": 0.16105598211288452, + "learning_rate": 3.484502440064818e-07, + "loss": 0.9852, + "step": 15843 + }, + { + "epoch": 0.9626344249346862, + "grad_norm": 0.11752396821975708, + "learning_rate": 3.473229693925828e-07, + "loss": 1.0467, + "step": 15844 + }, + { + "epoch": 0.9626951819673127, + "grad_norm": 0.14757245779037476, + "learning_rate": 3.461975148071561e-07, + "loss": 0.9821, + "step": 15845 + }, + { + "epoch": 0.9627559389999393, + "grad_norm": 0.1908593326807022, + "learning_rate": 3.4507388029146324e-07, + "loss": 1.0787, + "step": 15846 + }, + { + "epoch": 0.9628166960325658, + "grad_norm": 0.2143988460302353, + "learning_rate": 3.439520658866824e-07, + "loss": 1.0125, + "step": 15847 + }, + { + "epoch": 0.9628774530651923, + "grad_norm": 0.1700141578912735, + "learning_rate": 3.4283207163393614e-07, + "loss": 1.039, + "step": 15848 + }, + { + "epoch": 0.9629382100978188, + "grad_norm": 0.2083440124988556, + "learning_rate": 3.417138975742806e-07, + "loss": 1.0455, + "step": 15849 + }, + { + "epoch": 0.9629989671304453, + "grad_norm": 0.14064471423625946, + "learning_rate": 3.405975437486997e-07, + "loss": 0.9384, + "step": 15850 + }, + { + "epoch": 0.9630597241630718, + "grad_norm": 0.18559369444847107, + "learning_rate": 3.3948301019811056e-07, + "loss": 1.1417, + "step": 15851 + }, + { + "epoch": 0.9631204811956984, + "grad_norm": 0.172892764210701, + "learning_rate": 3.383702969633751e-07, + "loss": 0.9979, + "step": 15852 + }, + { + "epoch": 0.9631812382283249, + "grad_norm": 0.22538983821868896, + "learning_rate": 3.372594040852772e-07, + "loss": 1.0469, + "step": 15853 + }, + { + "epoch": 0.9632419952609514, + "grad_norm": 0.2507227063179016, + "learning_rate": 3.3615033160453444e-07, + "loss": 1.238, + "step": 15854 + }, + { + "epoch": 0.963302752293578, + "grad_norm": 0.13666996359825134, + "learning_rate": 3.350430795617976e-07, + "loss": 0.9974, + "step": 15855 + }, + { + "epoch": 0.9633635093262045, + "grad_norm": 4.167539119720459, + "learning_rate": 3.339376479976619e-07, + "loss": 1.0794, + "step": 15856 + }, + { + "epoch": 0.963424266358831, + "grad_norm": 0.18577080965042114, + "learning_rate": 3.328340369526395e-07, + "loss": 1.1481, + "step": 15857 + }, + { + "epoch": 0.9634850233914576, + "grad_norm": 0.5210777521133423, + "learning_rate": 3.317322464671868e-07, + "loss": 1.1981, + "step": 15858 + }, + { + "epoch": 0.9635457804240841, + "grad_norm": 0.16906839609146118, + "learning_rate": 3.3063227658168827e-07, + "loss": 1.1268, + "step": 15859 + }, + { + "epoch": 0.9636065374567107, + "grad_norm": 0.13034486770629883, + "learning_rate": 3.295341273364616e-07, + "loss": 1.052, + "step": 15860 + }, + { + "epoch": 0.9636672944893372, + "grad_norm": 0.14959298074245453, + "learning_rate": 3.2843779877176904e-07, + "loss": 1.0489, + "step": 15861 + }, + { + "epoch": 0.9637280515219636, + "grad_norm": 0.5510388016700745, + "learning_rate": 3.273432909277896e-07, + "loss": 1.0964, + "step": 15862 + }, + { + "epoch": 0.9637888085545901, + "grad_norm": 0.22286456823349, + "learning_rate": 3.262506038446356e-07, + "loss": 1.1755, + "step": 15863 + }, + { + "epoch": 0.9638495655872167, + "grad_norm": 0.17665845155715942, + "learning_rate": 3.2515973756237496e-07, + "loss": 1.0878, + "step": 15864 + }, + { + "epoch": 0.9639103226198432, + "grad_norm": 1.990303635597229, + "learning_rate": 3.240706921209813e-07, + "loss": 1.0642, + "step": 15865 + }, + { + "epoch": 0.9639710796524698, + "grad_norm": 0.14564980566501617, + "learning_rate": 3.229834675603838e-07, + "loss": 1.0247, + "step": 15866 + }, + { + "epoch": 0.9640318366850963, + "grad_norm": 0.11650662869215012, + "learning_rate": 3.2189806392042275e-07, + "loss": 1.018, + "step": 15867 + }, + { + "epoch": 0.9640925937177228, + "grad_norm": 0.162004753947258, + "learning_rate": 3.208144812408942e-07, + "loss": 1.0814, + "step": 15868 + }, + { + "epoch": 0.9641533507503494, + "grad_norm": 0.14691366255283356, + "learning_rate": 3.197327195615163e-07, + "loss": 1.0482, + "step": 15869 + }, + { + "epoch": 0.9642141077829759, + "grad_norm": 0.2144773155450821, + "learning_rate": 3.1865277892193515e-07, + "loss": 1.149, + "step": 15870 + }, + { + "epoch": 0.9642748648156024, + "grad_norm": 0.11379552632570267, + "learning_rate": 3.1757465936174127e-07, + "loss": 1.0077, + "step": 15871 + }, + { + "epoch": 0.964335621848229, + "grad_norm": 0.492879182100296, + "learning_rate": 3.1649836092045306e-07, + "loss": 1.0923, + "step": 15872 + }, + { + "epoch": 0.9643963788808555, + "grad_norm": 0.15862998366355896, + "learning_rate": 3.1542388363751676e-07, + "loss": 1.0467, + "step": 15873 + }, + { + "epoch": 0.964457135913482, + "grad_norm": 0.19025181233882904, + "learning_rate": 3.143512275523286e-07, + "loss": 1.1135, + "step": 15874 + }, + { + "epoch": 0.9645178929461085, + "grad_norm": 0.20708706974983215, + "learning_rate": 3.132803927041961e-07, + "loss": 1.0869, + "step": 15875 + }, + { + "epoch": 0.964578649978735, + "grad_norm": 0.23568205535411835, + "learning_rate": 3.1221137913237107e-07, + "loss": 1.005, + "step": 15876 + }, + { + "epoch": 0.9646394070113615, + "grad_norm": 0.19224685430526733, + "learning_rate": 3.111441868760445e-07, + "loss": 1.0825, + "step": 15877 + }, + { + "epoch": 0.9647001640439881, + "grad_norm": 0.19791728258132935, + "learning_rate": 3.100788159743351e-07, + "loss": 1.121, + "step": 15878 + }, + { + "epoch": 0.9647609210766146, + "grad_norm": 0.16504329442977905, + "learning_rate": 3.090152664662893e-07, + "loss": 1.2296, + "step": 15879 + }, + { + "epoch": 0.9648216781092411, + "grad_norm": 0.13588619232177734, + "learning_rate": 3.0795353839089826e-07, + "loss": 0.9904, + "step": 15880 + }, + { + "epoch": 0.9648824351418677, + "grad_norm": 0.23128920793533325, + "learning_rate": 3.0689363178706964e-07, + "loss": 1.0358, + "step": 15881 + }, + { + "epoch": 0.9649431921744942, + "grad_norm": 0.11999993026256561, + "learning_rate": 3.0583554669366687e-07, + "loss": 0.9945, + "step": 15882 + }, + { + "epoch": 0.9650039492071207, + "grad_norm": 0.18417789041996002, + "learning_rate": 3.047792831494589e-07, + "loss": 1.0702, + "step": 15883 + }, + { + "epoch": 0.9650647062397473, + "grad_norm": 0.18872062861919403, + "learning_rate": 3.0372484119318143e-07, + "loss": 1.0388, + "step": 15884 + }, + { + "epoch": 0.9651254632723738, + "grad_norm": 2.216291666030884, + "learning_rate": 3.0267222086347026e-07, + "loss": 1.0472, + "step": 15885 + }, + { + "epoch": 0.9651862203050003, + "grad_norm": 0.11928977817296982, + "learning_rate": 3.0162142219891663e-07, + "loss": 1.0468, + "step": 15886 + }, + { + "epoch": 0.9652469773376269, + "grad_norm": 0.39532384276390076, + "learning_rate": 3.005724452380343e-07, + "loss": 1.1224, + "step": 15887 + }, + { + "epoch": 0.9653077343702534, + "grad_norm": 0.18231981992721558, + "learning_rate": 2.995252900192813e-07, + "loss": 1.1174, + "step": 15888 + }, + { + "epoch": 0.9653684914028798, + "grad_norm": 0.20653830468654633, + "learning_rate": 2.984799565810326e-07, + "loss": 1.1004, + "step": 15889 + }, + { + "epoch": 0.9654292484355064, + "grad_norm": 0.14395245909690857, + "learning_rate": 2.974364449616074e-07, + "loss": 1.0153, + "step": 15890 + }, + { + "epoch": 0.9654900054681329, + "grad_norm": 0.1345723271369934, + "learning_rate": 2.9639475519925297e-07, + "loss": 1.079, + "step": 15891 + }, + { + "epoch": 0.9655507625007594, + "grad_norm": 0.1250736564397812, + "learning_rate": 2.95354887332161e-07, + "loss": 1.0239, + "step": 15892 + }, + { + "epoch": 0.965611519533386, + "grad_norm": 0.2912229895591736, + "learning_rate": 2.943168413984454e-07, + "loss": 1.2278, + "step": 15893 + }, + { + "epoch": 0.9656722765660125, + "grad_norm": 0.11981916427612305, + "learning_rate": 2.9328061743614796e-07, + "loss": 1.0843, + "step": 15894 + }, + { + "epoch": 0.965733033598639, + "grad_norm": 0.6097171902656555, + "learning_rate": 2.922462154832606e-07, + "loss": 1.2638, + "step": 15895 + }, + { + "epoch": 0.9657937906312656, + "grad_norm": 0.1612081676721573, + "learning_rate": 2.912136355776973e-07, + "loss": 1.0832, + "step": 15896 + }, + { + "epoch": 0.9658545476638921, + "grad_norm": 2.9336392879486084, + "learning_rate": 2.9018287775730566e-07, + "loss": 0.9901, + "step": 15897 + }, + { + "epoch": 0.9659153046965187, + "grad_norm": 0.12422700226306915, + "learning_rate": 2.891539420598721e-07, + "loss": 0.9899, + "step": 15898 + }, + { + "epoch": 0.9659760617291452, + "grad_norm": 2.7571585178375244, + "learning_rate": 2.881268285231054e-07, + "loss": 1.0217, + "step": 15899 + }, + { + "epoch": 0.9660368187617717, + "grad_norm": 0.12333909422159195, + "learning_rate": 2.871015371846586e-07, + "loss": 1.0397, + "step": 15900 + }, + { + "epoch": 0.9660975757943983, + "grad_norm": 0.16778922080993652, + "learning_rate": 2.860780680821185e-07, + "loss": 0.9936, + "step": 15901 + }, + { + "epoch": 0.9661583328270247, + "grad_norm": 0.13864189386367798, + "learning_rate": 2.850564212529938e-07, + "loss": 1.0293, + "step": 15902 + }, + { + "epoch": 0.9662190898596512, + "grad_norm": 0.21494293212890625, + "learning_rate": 2.84036596734738e-07, + "loss": 1.0528, + "step": 15903 + }, + { + "epoch": 0.9662798468922777, + "grad_norm": 0.22247940301895142, + "learning_rate": 2.830185945647268e-07, + "loss": 1.1675, + "step": 15904 + }, + { + "epoch": 0.9663406039249043, + "grad_norm": 0.1271168440580368, + "learning_rate": 2.820024147802802e-07, + "loss": 1.0403, + "step": 15905 + }, + { + "epoch": 0.9664013609575308, + "grad_norm": 2.177945852279663, + "learning_rate": 2.809880574186463e-07, + "loss": 1.0354, + "step": 15906 + }, + { + "epoch": 0.9664621179901574, + "grad_norm": 0.15773336589336395, + "learning_rate": 2.799755225170064e-07, + "loss": 1.0955, + "step": 15907 + }, + { + "epoch": 0.9665228750227839, + "grad_norm": 0.15277719497680664, + "learning_rate": 2.789648101124753e-07, + "loss": 1.0425, + "step": 15908 + }, + { + "epoch": 0.9665836320554104, + "grad_norm": 0.14750215411186218, + "learning_rate": 2.7795592024209563e-07, + "loss": 1.0726, + "step": 15909 + }, + { + "epoch": 0.966644389088037, + "grad_norm": 1.0825926065444946, + "learning_rate": 2.769488529428543e-07, + "loss": 1.0486, + "step": 15910 + }, + { + "epoch": 0.9667051461206635, + "grad_norm": 0.25789040327072144, + "learning_rate": 2.7594360825166644e-07, + "loss": 1.1341, + "step": 15911 + }, + { + "epoch": 0.96676590315329, + "grad_norm": 0.1634395718574524, + "learning_rate": 2.7494018620537466e-07, + "loss": 1.0502, + "step": 15912 + }, + { + "epoch": 0.9668266601859166, + "grad_norm": 0.12434323132038116, + "learning_rate": 2.739385868407607e-07, + "loss": 0.9876, + "step": 15913 + }, + { + "epoch": 0.9668874172185431, + "grad_norm": 0.21771100163459778, + "learning_rate": 2.7293881019454513e-07, + "loss": 1.159, + "step": 15914 + }, + { + "epoch": 0.9669481742511695, + "grad_norm": 0.15703751146793365, + "learning_rate": 2.719408563033654e-07, + "loss": 1.0354, + "step": 15915 + }, + { + "epoch": 0.9670089312837961, + "grad_norm": 0.1883440464735031, + "learning_rate": 2.709447252038089e-07, + "loss": 1.0717, + "step": 15916 + }, + { + "epoch": 0.9670696883164226, + "grad_norm": 0.148812398314476, + "learning_rate": 2.6995041693237965e-07, + "loss": 1.0004, + "step": 15917 + }, + { + "epoch": 0.9671304453490491, + "grad_norm": 0.7239001393318176, + "learning_rate": 2.689579315255375e-07, + "loss": 1.3432, + "step": 15918 + }, + { + "epoch": 0.9671912023816757, + "grad_norm": 0.29771092534065247, + "learning_rate": 2.679672690196533e-07, + "loss": 1.1206, + "step": 15919 + }, + { + "epoch": 0.9672519594143022, + "grad_norm": 0.9613197445869446, + "learning_rate": 2.669784294510369e-07, + "loss": 1.1884, + "step": 15920 + }, + { + "epoch": 0.9673127164469287, + "grad_norm": 0.17974425852298737, + "learning_rate": 2.659914128559426e-07, + "loss": 1.0232, + "step": 15921 + }, + { + "epoch": 0.9673734734795553, + "grad_norm": 0.210346519947052, + "learning_rate": 2.6500621927054715e-07, + "loss": 1.1113, + "step": 15922 + }, + { + "epoch": 0.9674342305121818, + "grad_norm": 0.12528258562088013, + "learning_rate": 2.6402284873096594e-07, + "loss": 1.0279, + "step": 15923 + }, + { + "epoch": 0.9674949875448083, + "grad_norm": 0.1716786026954651, + "learning_rate": 2.630413012732369e-07, + "loss": 1.0987, + "step": 15924 + }, + { + "epoch": 0.9675557445774349, + "grad_norm": 0.13459090888500214, + "learning_rate": 2.6206157693333677e-07, + "loss": 1.0327, + "step": 15925 + }, + { + "epoch": 0.9676165016100614, + "grad_norm": 0.19681993126869202, + "learning_rate": 2.610836757471924e-07, + "loss": 1.1052, + "step": 15926 + }, + { + "epoch": 0.9676772586426879, + "grad_norm": 0.17902269959449768, + "learning_rate": 2.601075977506362e-07, + "loss": 1.0386, + "step": 15927 + }, + { + "epoch": 0.9677380156753144, + "grad_norm": 0.19848330318927765, + "learning_rate": 2.5913334297945624e-07, + "loss": 1.1204, + "step": 15928 + }, + { + "epoch": 0.9677987727079409, + "grad_norm": 0.2038920372724533, + "learning_rate": 2.5816091146935176e-07, + "loss": 1.08, + "step": 15929 + }, + { + "epoch": 0.9678595297405674, + "grad_norm": 0.14603574573993683, + "learning_rate": 2.5719030325597746e-07, + "loss": 1.0593, + "step": 15930 + }, + { + "epoch": 0.967920286773194, + "grad_norm": 0.12619628012180328, + "learning_rate": 2.5622151837490504e-07, + "loss": 1.0147, + "step": 15931 + }, + { + "epoch": 0.9679810438058205, + "grad_norm": 0.25650787353515625, + "learning_rate": 2.5525455686165043e-07, + "loss": 1.1723, + "step": 15932 + }, + { + "epoch": 0.968041800838447, + "grad_norm": 0.1254999190568924, + "learning_rate": 2.54289418751652e-07, + "loss": 0.9941, + "step": 15933 + }, + { + "epoch": 0.9681025578710736, + "grad_norm": 0.29422008991241455, + "learning_rate": 2.533261040802981e-07, + "loss": 1.3069, + "step": 15934 + }, + { + "epoch": 0.9681633149037001, + "grad_norm": 0.14384031295776367, + "learning_rate": 2.5236461288288274e-07, + "loss": 1.0193, + "step": 15935 + }, + { + "epoch": 0.9682240719363266, + "grad_norm": 0.11932938545942307, + "learning_rate": 2.514049451946665e-07, + "loss": 1.0077, + "step": 15936 + }, + { + "epoch": 0.9682848289689532, + "grad_norm": 0.20866437256336212, + "learning_rate": 2.5044710105081027e-07, + "loss": 1.1336, + "step": 15937 + }, + { + "epoch": 0.9683455860015797, + "grad_norm": 0.4732510447502136, + "learning_rate": 2.494910804864359e-07, + "loss": 1.0189, + "step": 15938 + }, + { + "epoch": 0.9684063430342063, + "grad_norm": 0.14153967797756195, + "learning_rate": 2.4853688353658753e-07, + "loss": 1.0201, + "step": 15939 + }, + { + "epoch": 0.9684671000668328, + "grad_norm": 0.833027184009552, + "learning_rate": 2.475845102362262e-07, + "loss": 1.0606, + "step": 15940 + }, + { + "epoch": 0.9685278570994592, + "grad_norm": 0.44892945885658264, + "learning_rate": 2.466339606202794e-07, + "loss": 1.0646, + "step": 15941 + }, + { + "epoch": 0.9685886141320857, + "grad_norm": 0.1269439160823822, + "learning_rate": 2.4568523472358054e-07, + "loss": 1.0412, + "step": 15942 + }, + { + "epoch": 0.9686493711647123, + "grad_norm": 0.20885975658893585, + "learning_rate": 2.4473833258090715e-07, + "loss": 1.052, + "step": 15943 + }, + { + "epoch": 0.9687101281973388, + "grad_norm": 0.2538444697856903, + "learning_rate": 2.4379325422696495e-07, + "loss": 1.1564, + "step": 15944 + }, + { + "epoch": 0.9687708852299654, + "grad_norm": 0.11493953317403793, + "learning_rate": 2.4284999969639844e-07, + "loss": 0.9835, + "step": 15945 + }, + { + "epoch": 0.9688316422625919, + "grad_norm": 0.16870814561843872, + "learning_rate": 2.419085690237799e-07, + "loss": 1.0855, + "step": 15946 + }, + { + "epoch": 0.9688923992952184, + "grad_norm": 0.1368991881608963, + "learning_rate": 2.4096896224362617e-07, + "loss": 0.9903, + "step": 15947 + }, + { + "epoch": 0.968953156327845, + "grad_norm": 10.489834785461426, + "learning_rate": 2.4003117939037647e-07, + "loss": 1.1431, + "step": 15948 + }, + { + "epoch": 0.9690139133604715, + "grad_norm": 0.21081164479255676, + "learning_rate": 2.390952204983976e-07, + "loss": 1.1556, + "step": 15949 + }, + { + "epoch": 0.969074670393098, + "grad_norm": 0.40293940901756287, + "learning_rate": 2.3816108560200111e-07, + "loss": 1.0857, + "step": 15950 + }, + { + "epoch": 0.9691354274257246, + "grad_norm": 0.18464644253253937, + "learning_rate": 2.3722877473543182e-07, + "loss": 1.1472, + "step": 15951 + }, + { + "epoch": 0.9691961844583511, + "grad_norm": 0.12940700352191925, + "learning_rate": 2.362982879328568e-07, + "loss": 0.981, + "step": 15952 + }, + { + "epoch": 0.9692569414909776, + "grad_norm": 1.41521155834198, + "learning_rate": 2.3536962522838768e-07, + "loss": 1.0854, + "step": 15953 + }, + { + "epoch": 0.969317698523604, + "grad_norm": 0.1375926285982132, + "learning_rate": 2.3444278665606944e-07, + "loss": 0.9967, + "step": 15954 + }, + { + "epoch": 0.9693784555562306, + "grad_norm": 0.17588137090206146, + "learning_rate": 2.3351777224986938e-07, + "loss": 1.1185, + "step": 15955 + }, + { + "epoch": 0.9694392125888571, + "grad_norm": 0.15536680817604065, + "learning_rate": 2.3259458204368812e-07, + "loss": 1.068, + "step": 15956 + }, + { + "epoch": 0.9694999696214837, + "grad_norm": 0.2362576425075531, + "learning_rate": 2.3167321607137637e-07, + "loss": 1.1963, + "step": 15957 + }, + { + "epoch": 0.9695607266541102, + "grad_norm": 1.8998051881790161, + "learning_rate": 2.3075367436670715e-07, + "loss": 1.0531, + "step": 15958 + }, + { + "epoch": 0.9696214836867367, + "grad_norm": 0.13165737688541412, + "learning_rate": 2.2983595696337567e-07, + "loss": 1.0784, + "step": 15959 + }, + { + "epoch": 0.9696822407193633, + "grad_norm": 0.18202240765094757, + "learning_rate": 2.2892006389503285e-07, + "loss": 1.1026, + "step": 15960 + }, + { + "epoch": 0.9697429977519898, + "grad_norm": 0.12575796246528625, + "learning_rate": 2.280059951952407e-07, + "loss": 0.9856, + "step": 15961 + }, + { + "epoch": 0.9698037547846163, + "grad_norm": 0.4692280888557434, + "learning_rate": 2.2709375089751127e-07, + "loss": 1.1448, + "step": 15962 + }, + { + "epoch": 0.9698645118172429, + "grad_norm": 0.23251740634441376, + "learning_rate": 2.2618333103528455e-07, + "loss": 1.1115, + "step": 15963 + }, + { + "epoch": 0.9699252688498694, + "grad_norm": 2.102827787399292, + "learning_rate": 2.252747356419227e-07, + "loss": 1.0543, + "step": 15964 + }, + { + "epoch": 0.9699860258824959, + "grad_norm": 0.13372790813446045, + "learning_rate": 2.243679647507435e-07, + "loss": 1.0196, + "step": 15965 + }, + { + "epoch": 0.9700467829151225, + "grad_norm": 0.2305283397436142, + "learning_rate": 2.234630183949704e-07, + "loss": 1.0928, + "step": 15966 + }, + { + "epoch": 0.9701075399477489, + "grad_norm": 0.13396549224853516, + "learning_rate": 2.2255989660778798e-07, + "loss": 1.0194, + "step": 15967 + }, + { + "epoch": 0.9701682969803754, + "grad_norm": 0.20233449339866638, + "learning_rate": 2.2165859942229194e-07, + "loss": 1.0129, + "step": 15968 + }, + { + "epoch": 0.970229054013002, + "grad_norm": 0.11968764662742615, + "learning_rate": 2.2075912687152255e-07, + "loss": 1.0441, + "step": 15969 + }, + { + "epoch": 0.9702898110456285, + "grad_norm": 0.11924891173839569, + "learning_rate": 2.1986147898844788e-07, + "loss": 0.9888, + "step": 15970 + }, + { + "epoch": 0.970350568078255, + "grad_norm": 0.17051726579666138, + "learning_rate": 2.1896565580597495e-07, + "loss": 1.0159, + "step": 15971 + }, + { + "epoch": 0.9704113251108816, + "grad_norm": 0.1339452564716339, + "learning_rate": 2.1807165735693857e-07, + "loss": 0.9731, + "step": 15972 + }, + { + "epoch": 0.9704720821435081, + "grad_norm": 0.1252250075340271, + "learning_rate": 2.17179483674107e-07, + "loss": 1.0237, + "step": 15973 + }, + { + "epoch": 0.9705328391761346, + "grad_norm": 4.455218315124512, + "learning_rate": 2.1628913479018186e-07, + "loss": 1.0854, + "step": 15974 + }, + { + "epoch": 0.9705935962087612, + "grad_norm": 0.1265498250722885, + "learning_rate": 2.1540061073780927e-07, + "loss": 1.0599, + "step": 15975 + }, + { + "epoch": 0.9706543532413877, + "grad_norm": 1.06244957447052, + "learning_rate": 2.1451391154954093e-07, + "loss": 1.0036, + "step": 15976 + }, + { + "epoch": 0.9707151102740142, + "grad_norm": 4.255451679229736, + "learning_rate": 2.136290372578953e-07, + "loss": 1.1253, + "step": 15977 + }, + { + "epoch": 0.9707758673066408, + "grad_norm": 0.18407650291919708, + "learning_rate": 2.1274598789529643e-07, + "loss": 1.0871, + "step": 15978 + }, + { + "epoch": 0.9708366243392673, + "grad_norm": 0.17677772045135498, + "learning_rate": 2.1186476349411843e-07, + "loss": 1.0541, + "step": 15979 + }, + { + "epoch": 0.9708973813718937, + "grad_norm": 0.14362943172454834, + "learning_rate": 2.1098536408666325e-07, + "loss": 1.0388, + "step": 15980 + }, + { + "epoch": 0.9709581384045203, + "grad_norm": 0.8759345412254333, + "learning_rate": 2.1010778970516064e-07, + "loss": 1.0381, + "step": 15981 + }, + { + "epoch": 0.9710188954371468, + "grad_norm": 0.10893037170171738, + "learning_rate": 2.092320403817849e-07, + "loss": 1.0404, + "step": 15982 + }, + { + "epoch": 0.9710796524697733, + "grad_norm": 0.5004994869232178, + "learning_rate": 2.0835811614863254e-07, + "loss": 1.0245, + "step": 15983 + }, + { + "epoch": 0.9711404095023999, + "grad_norm": 0.20230625569820404, + "learning_rate": 2.074860170377335e-07, + "loss": 1.0877, + "step": 15984 + }, + { + "epoch": 0.9712011665350264, + "grad_norm": 0.11668587476015091, + "learning_rate": 2.0661574308106224e-07, + "loss": 1.0185, + "step": 15985 + }, + { + "epoch": 0.971261923567653, + "grad_norm": 0.1800309270620346, + "learning_rate": 2.0574729431051543e-07, + "loss": 1.1306, + "step": 15986 + }, + { + "epoch": 0.9713226806002795, + "grad_norm": 0.20712333917617798, + "learning_rate": 2.0488067075792872e-07, + "loss": 1.1126, + "step": 15987 + }, + { + "epoch": 0.971383437632906, + "grad_norm": 0.11909375339746475, + "learning_rate": 2.040158724550656e-07, + "loss": 0.9822, + "step": 15988 + }, + { + "epoch": 0.9714441946655326, + "grad_norm": 0.43867915868759155, + "learning_rate": 2.0315289943362292e-07, + "loss": 1.0761, + "step": 15989 + }, + { + "epoch": 0.9715049516981591, + "grad_norm": 0.15452586114406586, + "learning_rate": 2.022917517252365e-07, + "loss": 1.0725, + "step": 15990 + }, + { + "epoch": 0.9715657087307856, + "grad_norm": 0.4701095223426819, + "learning_rate": 2.0143242936147554e-07, + "loss": 1.1149, + "step": 15991 + }, + { + "epoch": 0.9716264657634122, + "grad_norm": 0.27840083837509155, + "learning_rate": 2.005749323738315e-07, + "loss": 1.1833, + "step": 15992 + }, + { + "epoch": 0.9716872227960387, + "grad_norm": 0.21867714822292328, + "learning_rate": 1.9971926079374037e-07, + "loss": 1.156, + "step": 15993 + }, + { + "epoch": 0.9717479798286651, + "grad_norm": 4.6639909744262695, + "learning_rate": 1.9886541465256593e-07, + "loss": 1.0509, + "step": 15994 + }, + { + "epoch": 0.9718087368612917, + "grad_norm": 0.1124323308467865, + "learning_rate": 1.9801339398160534e-07, + "loss": 1.0625, + "step": 15995 + }, + { + "epoch": 0.9718694938939182, + "grad_norm": 0.14120090007781982, + "learning_rate": 1.9716319881208922e-07, + "loss": 1.0225, + "step": 15996 + }, + { + "epoch": 0.9719302509265447, + "grad_norm": 0.13158804178237915, + "learning_rate": 1.9631482917518707e-07, + "loss": 1.0449, + "step": 15997 + }, + { + "epoch": 0.9719910079591713, + "grad_norm": 0.14530204236507416, + "learning_rate": 1.954682851019851e-07, + "loss": 1.057, + "step": 15998 + }, + { + "epoch": 0.9720517649917978, + "grad_norm": 0.17193961143493652, + "learning_rate": 1.9462356662352522e-07, + "loss": 1.0902, + "step": 15999 + }, + { + "epoch": 0.9721125220244243, + "grad_norm": 0.20645564794540405, + "learning_rate": 1.9378067377076593e-07, + "loss": 1.2907, + "step": 16000 + }, + { + "epoch": 0.9721732790570509, + "grad_norm": 0.14517450332641602, + "learning_rate": 1.929396065745992e-07, + "loss": 1.0217, + "step": 16001 + }, + { + "epoch": 0.9722340360896774, + "grad_norm": 0.8415945172309875, + "learning_rate": 1.9210036506586148e-07, + "loss": 1.019, + "step": 16002 + }, + { + "epoch": 0.9722947931223039, + "grad_norm": 0.1590428352355957, + "learning_rate": 1.9126294927531153e-07, + "loss": 1.0652, + "step": 16003 + }, + { + "epoch": 0.9723555501549305, + "grad_norm": 0.17362003028392792, + "learning_rate": 1.9042735923364696e-07, + "loss": 1.1099, + "step": 16004 + }, + { + "epoch": 0.972416307187557, + "grad_norm": 0.20347914099693298, + "learning_rate": 1.8959359497149886e-07, + "loss": 1.1079, + "step": 16005 + }, + { + "epoch": 0.9724770642201835, + "grad_norm": 0.12873537838459015, + "learning_rate": 1.8876165651942056e-07, + "loss": 1.0039, + "step": 16006 + }, + { + "epoch": 0.97253782125281, + "grad_norm": 0.1304507851600647, + "learning_rate": 1.8793154390791545e-07, + "loss": 1.0667, + "step": 16007 + }, + { + "epoch": 0.9725985782854365, + "grad_norm": 0.13589982688426971, + "learning_rate": 1.8710325716740363e-07, + "loss": 1.0206, + "step": 16008 + }, + { + "epoch": 0.972659335318063, + "grad_norm": 0.1820124238729477, + "learning_rate": 1.8627679632825523e-07, + "loss": 1.1197, + "step": 16009 + }, + { + "epoch": 0.9727200923506896, + "grad_norm": 0.21318185329437256, + "learning_rate": 1.8545216142075717e-07, + "loss": 1.1231, + "step": 16010 + }, + { + "epoch": 0.9727808493833161, + "grad_norm": 0.12348305433988571, + "learning_rate": 1.846293524751408e-07, + "loss": 1.006, + "step": 16011 + }, + { + "epoch": 0.9728416064159426, + "grad_norm": 0.16331954300403595, + "learning_rate": 1.8380836952155979e-07, + "loss": 1.1292, + "step": 16012 + }, + { + "epoch": 0.9729023634485692, + "grad_norm": 0.19831925630569458, + "learning_rate": 1.8298921259011782e-07, + "loss": 1.1073, + "step": 16013 + }, + { + "epoch": 0.9729631204811957, + "grad_norm": 0.41348716616630554, + "learning_rate": 1.8217188171082977e-07, + "loss": 1.167, + "step": 16014 + }, + { + "epoch": 0.9730238775138222, + "grad_norm": 0.2028099000453949, + "learning_rate": 1.8135637691366613e-07, + "loss": 1.0648, + "step": 16015 + }, + { + "epoch": 0.9730846345464488, + "grad_norm": 0.12847863137722015, + "learning_rate": 1.805426982285141e-07, + "loss": 1.02, + "step": 16016 + }, + { + "epoch": 0.9731453915790753, + "grad_norm": 0.12162615358829498, + "learning_rate": 1.7973084568519982e-07, + "loss": 1.0011, + "step": 16017 + }, + { + "epoch": 0.9732061486117018, + "grad_norm": 0.13329797983169556, + "learning_rate": 1.7892081931347726e-07, + "loss": 1.0343, + "step": 16018 + }, + { + "epoch": 0.9732669056443284, + "grad_norm": 0.4047829806804657, + "learning_rate": 1.7811261914304488e-07, + "loss": 1.0825, + "step": 16019 + }, + { + "epoch": 0.9733276626769548, + "grad_norm": 0.1531861126422882, + "learning_rate": 1.7730624520352346e-07, + "loss": 1.0118, + "step": 16020 + }, + { + "epoch": 0.9733884197095813, + "grad_norm": 0.11587836593389511, + "learning_rate": 1.7650169752447265e-07, + "loss": 0.98, + "step": 16021 + }, + { + "epoch": 0.9734491767422079, + "grad_norm": 0.1757090538740158, + "learning_rate": 1.7569897613538555e-07, + "loss": 1.0405, + "step": 16022 + }, + { + "epoch": 0.9735099337748344, + "grad_norm": 0.178058460354805, + "learning_rate": 1.7489808106568305e-07, + "loss": 1.0707, + "step": 16023 + }, + { + "epoch": 0.973570690807461, + "grad_norm": 0.14034578204154968, + "learning_rate": 1.7409901234471948e-07, + "loss": 1.0215, + "step": 16024 + }, + { + "epoch": 0.9736314478400875, + "grad_norm": 0.2031918168067932, + "learning_rate": 1.7330177000178805e-07, + "loss": 1.1124, + "step": 16025 + }, + { + "epoch": 0.973692204872714, + "grad_norm": 0.13193994760513306, + "learning_rate": 1.7250635406611538e-07, + "loss": 0.9781, + "step": 16026 + }, + { + "epoch": 0.9737529619053406, + "grad_norm": 0.16582059860229492, + "learning_rate": 1.7171276456685037e-07, + "loss": 1.0571, + "step": 16027 + }, + { + "epoch": 0.9738137189379671, + "grad_norm": 0.13737717270851135, + "learning_rate": 1.7092100153308643e-07, + "loss": 1.037, + "step": 16028 + }, + { + "epoch": 0.9738744759705936, + "grad_norm": 0.12876541912555695, + "learning_rate": 1.7013106499383924e-07, + "loss": 1.0596, + "step": 16029 + }, + { + "epoch": 0.9739352330032202, + "grad_norm": 0.1493791937828064, + "learning_rate": 1.6934295497808006e-07, + "loss": 1.005, + "step": 16030 + }, + { + "epoch": 0.9739959900358467, + "grad_norm": 0.24591226875782013, + "learning_rate": 1.6855667151468024e-07, + "loss": 1.1842, + "step": 16031 + }, + { + "epoch": 0.9740567470684732, + "grad_norm": 0.1159437969326973, + "learning_rate": 1.6777221463246673e-07, + "loss": 0.9591, + "step": 16032 + }, + { + "epoch": 0.9741175041010997, + "grad_norm": 7.241575241088867, + "learning_rate": 1.6698958436019985e-07, + "loss": 1.1088, + "step": 16033 + }, + { + "epoch": 0.9741782611337262, + "grad_norm": 0.1758609265089035, + "learning_rate": 1.6620878072655664e-07, + "loss": 1.0628, + "step": 16034 + }, + { + "epoch": 0.9742390181663527, + "grad_norm": 0.25762495398521423, + "learning_rate": 1.6542980376016982e-07, + "loss": 1.043, + "step": 16035 + }, + { + "epoch": 0.9742997751989793, + "grad_norm": 0.16597811877727509, + "learning_rate": 1.6465265348958314e-07, + "loss": 1.0745, + "step": 16036 + }, + { + "epoch": 0.9743605322316058, + "grad_norm": 0.18467192351818085, + "learning_rate": 1.6387732994328496e-07, + "loss": 1.065, + "step": 16037 + }, + { + "epoch": 0.9744212892642323, + "grad_norm": 0.3706793487071991, + "learning_rate": 1.63103833149697e-07, + "loss": 1.0749, + "step": 16038 + }, + { + "epoch": 0.9744820462968589, + "grad_norm": 0.14396531879901886, + "learning_rate": 1.6233216313717437e-07, + "loss": 1.0409, + "step": 16039 + }, + { + "epoch": 0.9745428033294854, + "grad_norm": 0.19907397031784058, + "learning_rate": 1.615623199339944e-07, + "loss": 1.1127, + "step": 16040 + }, + { + "epoch": 0.9746035603621119, + "grad_norm": 0.11393120884895325, + "learning_rate": 1.6079430356838454e-07, + "loss": 1.0324, + "step": 16041 + }, + { + "epoch": 0.9746643173947385, + "grad_norm": 0.11399836093187332, + "learning_rate": 1.600281140684945e-07, + "loss": 1.0005, + "step": 16042 + }, + { + "epoch": 0.974725074427365, + "grad_norm": 0.19599370658397675, + "learning_rate": 1.5926375146240734e-07, + "loss": 1.1074, + "step": 16043 + }, + { + "epoch": 0.9747858314599915, + "grad_norm": 0.18766307830810547, + "learning_rate": 1.5850121577813959e-07, + "loss": 1.0216, + "step": 16044 + }, + { + "epoch": 0.9748465884926181, + "grad_norm": 0.21851354837417603, + "learning_rate": 1.5774050704364106e-07, + "loss": 1.0803, + "step": 16045 + }, + { + "epoch": 0.9749073455252445, + "grad_norm": 5.629420757293701, + "learning_rate": 1.5698162528680616e-07, + "loss": 1.1353, + "step": 16046 + }, + { + "epoch": 0.974968102557871, + "grad_norm": 0.2156076729297638, + "learning_rate": 1.5622457053544038e-07, + "loss": 1.0279, + "step": 16047 + }, + { + "epoch": 0.9750288595904976, + "grad_norm": 0.26076892018318176, + "learning_rate": 1.5546934281729375e-07, + "loss": 1.2636, + "step": 16048 + }, + { + "epoch": 0.9750896166231241, + "grad_norm": 0.21939487755298615, + "learning_rate": 1.547159421600608e-07, + "loss": 1.2405, + "step": 16049 + }, + { + "epoch": 0.9751503736557506, + "grad_norm": 0.18778149783611298, + "learning_rate": 1.5396436859134168e-07, + "loss": 1.0851, + "step": 16050 + }, + { + "epoch": 0.9752111306883772, + "grad_norm": 0.1266818791627884, + "learning_rate": 1.5321462213869764e-07, + "loss": 1.0551, + "step": 16051 + }, + { + "epoch": 0.9752718877210037, + "grad_norm": 0.2293151468038559, + "learning_rate": 1.5246670282960674e-07, + "loss": 1.1178, + "step": 16052 + }, + { + "epoch": 0.9753326447536302, + "grad_norm": 0.18465852737426758, + "learning_rate": 1.517206106914859e-07, + "loss": 1.1228, + "step": 16053 + }, + { + "epoch": 0.9753934017862568, + "grad_norm": 0.7020722031593323, + "learning_rate": 1.5097634575167996e-07, + "loss": 1.2234, + "step": 16054 + }, + { + "epoch": 0.9754541588188833, + "grad_norm": 0.1345306783914566, + "learning_rate": 1.502339080374726e-07, + "loss": 1.0253, + "step": 16055 + }, + { + "epoch": 0.9755149158515098, + "grad_norm": 0.13883526623249054, + "learning_rate": 1.4949329757607544e-07, + "loss": 1.0587, + "step": 16056 + }, + { + "epoch": 0.9755756728841364, + "grad_norm": 0.7106316089630127, + "learning_rate": 1.4875451439463894e-07, + "loss": 1.1146, + "step": 16057 + }, + { + "epoch": 0.9756364299167629, + "grad_norm": 0.16689079999923706, + "learning_rate": 1.4801755852024145e-07, + "loss": 1.1009, + "step": 16058 + }, + { + "epoch": 0.9756971869493893, + "grad_norm": 0.13701871037483215, + "learning_rate": 1.4728242997989473e-07, + "loss": 1.0936, + "step": 16059 + }, + { + "epoch": 0.9757579439820159, + "grad_norm": 0.40464407205581665, + "learning_rate": 1.4654912880054382e-07, + "loss": 1.1281, + "step": 16060 + }, + { + "epoch": 0.9758187010146424, + "grad_norm": 0.22826455533504486, + "learning_rate": 1.458176550090784e-07, + "loss": 1.0917, + "step": 16061 + }, + { + "epoch": 0.9758794580472689, + "grad_norm": 0.22348837554454803, + "learning_rate": 1.450880086322992e-07, + "loss": 1.1496, + "step": 16062 + }, + { + "epoch": 0.9759402150798955, + "grad_norm": 0.15505599975585938, + "learning_rate": 1.4436018969695708e-07, + "loss": 1.0594, + "step": 16063 + }, + { + "epoch": 0.976000972112522, + "grad_norm": 1.9686870574951172, + "learning_rate": 1.4363419822973067e-07, + "loss": 1.0664, + "step": 16064 + }, + { + "epoch": 0.9760617291451485, + "grad_norm": 0.17271870374679565, + "learning_rate": 1.4291003425722648e-07, + "loss": 1.1058, + "step": 16065 + }, + { + "epoch": 0.9761224861777751, + "grad_norm": 0.16958126425743103, + "learning_rate": 1.4218769780598995e-07, + "loss": 1.1381, + "step": 16066 + }, + { + "epoch": 0.9761832432104016, + "grad_norm": 0.16132481396198273, + "learning_rate": 1.4146718890250543e-07, + "loss": 1.0086, + "step": 16067 + }, + { + "epoch": 0.9762440002430282, + "grad_norm": 0.3277038037776947, + "learning_rate": 1.407485075731796e-07, + "loss": 1.2296, + "step": 16068 + }, + { + "epoch": 0.9763047572756547, + "grad_norm": 0.16135403513908386, + "learning_rate": 1.4003165384435246e-07, + "loss": 0.9989, + "step": 16069 + }, + { + "epoch": 0.9763655143082812, + "grad_norm": 0.19204193353652954, + "learning_rate": 1.3931662774230304e-07, + "loss": 1.0639, + "step": 16070 + }, + { + "epoch": 0.9764262713409078, + "grad_norm": 0.2071891725063324, + "learning_rate": 1.386034292932381e-07, + "loss": 1.1753, + "step": 16071 + }, + { + "epoch": 0.9764870283735342, + "grad_norm": 0.14060404896736145, + "learning_rate": 1.37892058523309e-07, + "loss": 1.0056, + "step": 16072 + }, + { + "epoch": 0.9765477854061607, + "grad_norm": 0.3894897699356079, + "learning_rate": 1.371825154585782e-07, + "loss": 1.1638, + "step": 16073 + }, + { + "epoch": 0.9766085424387873, + "grad_norm": 0.22857226431369781, + "learning_rate": 1.3647480012506375e-07, + "loss": 1.0623, + "step": 16074 + }, + { + "epoch": 0.9766692994714138, + "grad_norm": 0.19580797851085663, + "learning_rate": 1.357689125487005e-07, + "loss": 1.0634, + "step": 16075 + }, + { + "epoch": 0.9767300565040403, + "grad_norm": 0.1338660717010498, + "learning_rate": 1.3506485275536772e-07, + "loss": 1.01, + "step": 16076 + }, + { + "epoch": 0.9767908135366669, + "grad_norm": 0.15864011645317078, + "learning_rate": 1.3436262077087258e-07, + "loss": 1.2374, + "step": 16077 + }, + { + "epoch": 0.9768515705692934, + "grad_norm": 0.2069341093301773, + "learning_rate": 1.3366221662095558e-07, + "loss": 1.0603, + "step": 16078 + }, + { + "epoch": 0.9769123276019199, + "grad_norm": 0.22489233314990997, + "learning_rate": 1.3296364033128506e-07, + "loss": 1.102, + "step": 16079 + }, + { + "epoch": 0.9769730846345465, + "grad_norm": 0.17519886791706085, + "learning_rate": 1.3226689192747387e-07, + "loss": 1.1061, + "step": 16080 + }, + { + "epoch": 0.977033841667173, + "grad_norm": 0.23372021317481995, + "learning_rate": 1.3157197143506273e-07, + "loss": 1.0909, + "step": 16081 + }, + { + "epoch": 0.9770945986997995, + "grad_norm": 0.12334156781435013, + "learning_rate": 1.3087887887951457e-07, + "loss": 1.0384, + "step": 16082 + }, + { + "epoch": 0.9771553557324261, + "grad_norm": 0.11817861348390579, + "learning_rate": 1.3018761428624237e-07, + "loss": 0.9871, + "step": 16083 + }, + { + "epoch": 0.9772161127650526, + "grad_norm": 0.2089742124080658, + "learning_rate": 1.294981776805815e-07, + "loss": 1.0758, + "step": 16084 + }, + { + "epoch": 0.977276869797679, + "grad_norm": 0.6454282402992249, + "learning_rate": 1.2881056908780608e-07, + "loss": 1.0176, + "step": 16085 + }, + { + "epoch": 0.9773376268303056, + "grad_norm": 0.18240317702293396, + "learning_rate": 1.2812478853311826e-07, + "loss": 1.1306, + "step": 16086 + }, + { + "epoch": 0.9773983838629321, + "grad_norm": 4.919723987579346, + "learning_rate": 1.27440836041659e-07, + "loss": 1.0693, + "step": 16087 + }, + { + "epoch": 0.9774591408955586, + "grad_norm": 0.13495619595050812, + "learning_rate": 1.2675871163849718e-07, + "loss": 1.0729, + "step": 16088 + }, + { + "epoch": 0.9775198979281852, + "grad_norm": 0.32928404211997986, + "learning_rate": 1.260784153486294e-07, + "loss": 1.2433, + "step": 16089 + }, + { + "epoch": 0.9775806549608117, + "grad_norm": 0.1618475317955017, + "learning_rate": 1.2539994719700244e-07, + "loss": 1.045, + "step": 16090 + }, + { + "epoch": 0.9776414119934382, + "grad_norm": 0.1313403844833374, + "learning_rate": 1.247233072084797e-07, + "loss": 0.9732, + "step": 16091 + }, + { + "epoch": 0.9777021690260648, + "grad_norm": 0.13734275102615356, + "learning_rate": 1.2404849540786357e-07, + "loss": 1.0804, + "step": 16092 + }, + { + "epoch": 0.9777629260586913, + "grad_norm": 0.17345871031284332, + "learning_rate": 1.2337551181988983e-07, + "loss": 1.058, + "step": 16093 + }, + { + "epoch": 0.9778236830913178, + "grad_norm": 0.7297492027282715, + "learning_rate": 1.2270435646922762e-07, + "loss": 1.2106, + "step": 16094 + }, + { + "epoch": 0.9778844401239444, + "grad_norm": 0.1376461386680603, + "learning_rate": 1.2203502938047952e-07, + "loss": 1.0073, + "step": 16095 + }, + { + "epoch": 0.9779451971565709, + "grad_norm": 0.3606899380683899, + "learning_rate": 1.213675305781814e-07, + "loss": 1.0342, + "step": 16096 + }, + { + "epoch": 0.9780059541891974, + "grad_norm": 0.13359619677066803, + "learning_rate": 1.2070186008679707e-07, + "loss": 1.0064, + "step": 16097 + }, + { + "epoch": 0.978066711221824, + "grad_norm": 0.17910289764404297, + "learning_rate": 1.2003801793072366e-07, + "loss": 1.0899, + "step": 16098 + }, + { + "epoch": 0.9781274682544504, + "grad_norm": 0.1508798748254776, + "learning_rate": 1.1937600413430283e-07, + "loss": 1.0582, + "step": 16099 + }, + { + "epoch": 0.9781882252870769, + "grad_norm": 0.259865403175354, + "learning_rate": 1.1871581872179294e-07, + "loss": 1.0576, + "step": 16100 + }, + { + "epoch": 0.9782489823197035, + "grad_norm": 0.1672934889793396, + "learning_rate": 1.1805746171739685e-07, + "loss": 1.0648, + "step": 16101 + }, + { + "epoch": 0.97830973935233, + "grad_norm": 0.11790236085653305, + "learning_rate": 1.1740093314524525e-07, + "loss": 1.0554, + "step": 16102 + }, + { + "epoch": 0.9783704963849565, + "grad_norm": 0.16731683909893036, + "learning_rate": 1.167462330294078e-07, + "loss": 1.083, + "step": 16103 + }, + { + "epoch": 0.9784312534175831, + "grad_norm": 0.35459843277931213, + "learning_rate": 1.1609336139387639e-07, + "loss": 1.1223, + "step": 16104 + }, + { + "epoch": 0.9784920104502096, + "grad_norm": 0.12305504828691483, + "learning_rate": 1.1544231826258745e-07, + "loss": 1.0064, + "step": 16105 + }, + { + "epoch": 0.9785527674828361, + "grad_norm": 0.2740165591239929, + "learning_rate": 1.1479310365939966e-07, + "loss": 1.1251, + "step": 16106 + }, + { + "epoch": 0.9786135245154627, + "grad_norm": 0.8507675528526306, + "learning_rate": 1.1414571760811621e-07, + "loss": 1.0879, + "step": 16107 + }, + { + "epoch": 0.9786742815480892, + "grad_norm": 0.1523904800415039, + "learning_rate": 1.1350016013246256e-07, + "loss": 1.0667, + "step": 16108 + }, + { + "epoch": 0.9787350385807158, + "grad_norm": 0.3108605444431305, + "learning_rate": 1.1285643125610867e-07, + "loss": 1.1983, + "step": 16109 + }, + { + "epoch": 0.9787957956133423, + "grad_norm": 0.13928180932998657, + "learning_rate": 1.1221453100264123e-07, + "loss": 1.028, + "step": 16110 + }, + { + "epoch": 0.9788565526459688, + "grad_norm": 0.4610697031021118, + "learning_rate": 1.1157445939559141e-07, + "loss": 1.0898, + "step": 16111 + }, + { + "epoch": 0.9789173096785952, + "grad_norm": 0.12443111091852188, + "learning_rate": 1.1093621645842378e-07, + "loss": 0.9927, + "step": 16112 + }, + { + "epoch": 0.9789780667112218, + "grad_norm": 0.12096915394067764, + "learning_rate": 1.1029980221453073e-07, + "loss": 1.0404, + "step": 16113 + }, + { + "epoch": 0.9790388237438483, + "grad_norm": 0.14651578664779663, + "learning_rate": 1.0966521668724361e-07, + "loss": 0.9869, + "step": 16114 + }, + { + "epoch": 0.9790995807764749, + "grad_norm": 0.12659388780593872, + "learning_rate": 1.0903245989982714e-07, + "loss": 1.0106, + "step": 16115 + }, + { + "epoch": 0.9791603378091014, + "grad_norm": 0.43814608454704285, + "learning_rate": 1.0840153187546276e-07, + "loss": 1.109, + "step": 16116 + }, + { + "epoch": 0.9792210948417279, + "grad_norm": 0.2830072343349457, + "learning_rate": 1.0777243263728754e-07, + "loss": 1.0165, + "step": 16117 + }, + { + "epoch": 0.9792818518743545, + "grad_norm": 0.16310855746269226, + "learning_rate": 1.0714516220835525e-07, + "loss": 1.0579, + "step": 16118 + }, + { + "epoch": 0.979342608906981, + "grad_norm": 0.19141477346420288, + "learning_rate": 1.0651972061166415e-07, + "loss": 1.0883, + "step": 16119 + }, + { + "epoch": 0.9794033659396075, + "grad_norm": 0.2186070680618286, + "learning_rate": 1.0589610787013482e-07, + "loss": 1.1433, + "step": 16120 + }, + { + "epoch": 0.9794641229722341, + "grad_norm": 0.18501213192939758, + "learning_rate": 1.0527432400662674e-07, + "loss": 1.048, + "step": 16121 + }, + { + "epoch": 0.9795248800048606, + "grad_norm": 0.1150391697883606, + "learning_rate": 1.0465436904393277e-07, + "loss": 1.0205, + "step": 16122 + }, + { + "epoch": 0.9795856370374871, + "grad_norm": 0.26747676730155945, + "learning_rate": 1.0403624300477921e-07, + "loss": 1.1386, + "step": 16123 + }, + { + "epoch": 0.9796463940701137, + "grad_norm": 0.23971505463123322, + "learning_rate": 1.0341994591182014e-07, + "loss": 1.035, + "step": 16124 + }, + { + "epoch": 0.9797071511027401, + "grad_norm": 0.26782485842704773, + "learning_rate": 1.028054777876486e-07, + "loss": 1.0545, + "step": 16125 + }, + { + "epoch": 0.9797679081353666, + "grad_norm": 0.10993538796901703, + "learning_rate": 1.0219283865479101e-07, + "loss": 1.0422, + "step": 16126 + }, + { + "epoch": 0.9798286651679932, + "grad_norm": 0.11925152689218521, + "learning_rate": 1.015820285356961e-07, + "loss": 0.9993, + "step": 16127 + }, + { + "epoch": 0.9798894222006197, + "grad_norm": 0.20034094154834747, + "learning_rate": 1.0097304745275704e-07, + "loss": 1.1167, + "step": 16128 + }, + { + "epoch": 0.9799501792332462, + "grad_norm": 0.17866800725460052, + "learning_rate": 1.0036589542829488e-07, + "loss": 1.0656, + "step": 16129 + }, + { + "epoch": 0.9800109362658728, + "grad_norm": 0.1093328446149826, + "learning_rate": 9.97605724845696e-08, + "loss": 1.0216, + "step": 16130 + }, + { + "epoch": 0.9800716932984993, + "grad_norm": 0.20997603237628937, + "learning_rate": 9.915707864376345e-08, + "loss": 1.1226, + "step": 16131 + }, + { + "epoch": 0.9801324503311258, + "grad_norm": 0.11471472680568695, + "learning_rate": 9.855541392800316e-08, + "loss": 1.0577, + "step": 16132 + }, + { + "epoch": 0.9801932073637524, + "grad_norm": 0.12546268105506897, + "learning_rate": 9.795557835933778e-08, + "loss": 1.0621, + "step": 16133 + }, + { + "epoch": 0.9802539643963789, + "grad_norm": 0.14443178474903107, + "learning_rate": 9.735757195975525e-08, + "loss": 1.0331, + "step": 16134 + }, + { + "epoch": 0.9803147214290054, + "grad_norm": 0.12066084146499634, + "learning_rate": 9.676139475117696e-08, + "loss": 1.0279, + "step": 16135 + }, + { + "epoch": 0.980375478461632, + "grad_norm": 0.13670536875724792, + "learning_rate": 9.616704675545762e-08, + "loss": 0.9979, + "step": 16136 + }, + { + "epoch": 0.9804362354942585, + "grad_norm": 0.224175825715065, + "learning_rate": 9.557452799437982e-08, + "loss": 1.1086, + "step": 16137 + }, + { + "epoch": 0.9804969925268849, + "grad_norm": 0.11802633106708527, + "learning_rate": 9.498383848966508e-08, + "loss": 1.0054, + "step": 16138 + }, + { + "epoch": 0.9805577495595115, + "grad_norm": 0.18890927731990814, + "learning_rate": 9.439497826296828e-08, + "loss": 1.0244, + "step": 16139 + }, + { + "epoch": 0.980618506592138, + "grad_norm": 0.14303602278232574, + "learning_rate": 9.380794733586107e-08, + "loss": 1.0662, + "step": 16140 + }, + { + "epoch": 0.9806792636247645, + "grad_norm": 0.1844303458929062, + "learning_rate": 9.322274572987621e-08, + "loss": 1.0847, + "step": 16141 + }, + { + "epoch": 0.9807400206573911, + "grad_norm": 0.20522867143154144, + "learning_rate": 9.263937346645769e-08, + "loss": 1.0403, + "step": 16142 + }, + { + "epoch": 0.9808007776900176, + "grad_norm": 0.21198460459709167, + "learning_rate": 9.205783056698835e-08, + "loss": 1.0586, + "step": 16143 + }, + { + "epoch": 0.9808615347226441, + "grad_norm": 0.21410316228866577, + "learning_rate": 9.147811705279008e-08, + "loss": 1.1461, + "step": 16144 + }, + { + "epoch": 0.9809222917552707, + "grad_norm": 0.2493596374988556, + "learning_rate": 9.090023294510697e-08, + "loss": 1.1257, + "step": 16145 + }, + { + "epoch": 0.9809830487878972, + "grad_norm": 0.16352759301662445, + "learning_rate": 9.032417826512207e-08, + "loss": 1.0529, + "step": 16146 + }, + { + "epoch": 0.9810438058205238, + "grad_norm": 0.11248141527175903, + "learning_rate": 8.974995303395184e-08, + "loss": 1.0347, + "step": 16147 + }, + { + "epoch": 0.9811045628531503, + "grad_norm": 0.13883396983146667, + "learning_rate": 8.917755727264054e-08, + "loss": 1.0491, + "step": 16148 + }, + { + "epoch": 0.9811653198857768, + "grad_norm": 0.1515413522720337, + "learning_rate": 8.860699100217696e-08, + "loss": 1.0206, + "step": 16149 + }, + { + "epoch": 0.9812260769184034, + "grad_norm": 0.1263536810874939, + "learning_rate": 8.803825424347212e-08, + "loss": 0.9641, + "step": 16150 + }, + { + "epoch": 0.9812868339510298, + "grad_norm": 0.1624402105808258, + "learning_rate": 8.747134701737602e-08, + "loss": 1.044, + "step": 16151 + }, + { + "epoch": 0.9813475909836563, + "grad_norm": 0.13232336938381195, + "learning_rate": 8.690626934466096e-08, + "loss": 1.0724, + "step": 16152 + }, + { + "epoch": 0.9814083480162828, + "grad_norm": 0.1404111534357071, + "learning_rate": 8.634302124604366e-08, + "loss": 1.0697, + "step": 16153 + }, + { + "epoch": 0.9814691050489094, + "grad_norm": 0.17977111041545868, + "learning_rate": 8.578160274217428e-08, + "loss": 1.0495, + "step": 16154 + }, + { + "epoch": 0.9815298620815359, + "grad_norm": 0.6686611175537109, + "learning_rate": 8.522201385362527e-08, + "loss": 1.0929, + "step": 16155 + }, + { + "epoch": 0.9815906191141625, + "grad_norm": 0.1799343079328537, + "learning_rate": 8.466425460091909e-08, + "loss": 1.0742, + "step": 16156 + }, + { + "epoch": 0.981651376146789, + "grad_norm": 0.16793352365493774, + "learning_rate": 8.410832500448385e-08, + "loss": 0.9929, + "step": 16157 + }, + { + "epoch": 0.9817121331794155, + "grad_norm": 0.18645429611206055, + "learning_rate": 8.355422508470878e-08, + "loss": 1.077, + "step": 16158 + }, + { + "epoch": 0.9817728902120421, + "grad_norm": 0.13322800397872925, + "learning_rate": 8.300195486190543e-08, + "loss": 1.0423, + "step": 16159 + }, + { + "epoch": 0.9818336472446686, + "grad_norm": 0.1837165355682373, + "learning_rate": 8.245151435631316e-08, + "loss": 1.0091, + "step": 16160 + }, + { + "epoch": 0.9818944042772951, + "grad_norm": 0.13260650634765625, + "learning_rate": 8.190290358811026e-08, + "loss": 0.9794, + "step": 16161 + }, + { + "epoch": 0.9819551613099217, + "grad_norm": 0.11007736623287201, + "learning_rate": 8.135612257740289e-08, + "loss": 1.043, + "step": 16162 + }, + { + "epoch": 0.9820159183425482, + "grad_norm": 0.2954457998275757, + "learning_rate": 8.08111713442361e-08, + "loss": 1.0426, + "step": 16163 + }, + { + "epoch": 0.9820766753751746, + "grad_norm": 0.16975396871566772, + "learning_rate": 8.026804990858283e-08, + "loss": 1.0009, + "step": 16164 + }, + { + "epoch": 0.9821374324078012, + "grad_norm": 0.09751985222101212, + "learning_rate": 7.972675829036046e-08, + "loss": 0.9393, + "step": 16165 + }, + { + "epoch": 0.9821981894404277, + "grad_norm": 0.25893545150756836, + "learning_rate": 7.91872965093976e-08, + "loss": 1.1018, + "step": 16166 + }, + { + "epoch": 0.9822589464730542, + "grad_norm": 0.17480066418647766, + "learning_rate": 7.86496645854784e-08, + "loss": 1.0576, + "step": 16167 + }, + { + "epoch": 0.9823197035056808, + "grad_norm": 0.13282382488250732, + "learning_rate": 7.811386253829822e-08, + "loss": 1.0456, + "step": 16168 + }, + { + "epoch": 0.9823804605383073, + "grad_norm": 0.21847635507583618, + "learning_rate": 7.757989038750801e-08, + "loss": 1.1308, + "step": 16169 + }, + { + "epoch": 0.9824412175709338, + "grad_norm": 0.18405449390411377, + "learning_rate": 7.704774815268101e-08, + "loss": 1.1418, + "step": 16170 + }, + { + "epoch": 0.9825019746035604, + "grad_norm": 0.12723439931869507, + "learning_rate": 7.651743585331273e-08, + "loss": 1.0435, + "step": 16171 + }, + { + "epoch": 0.9825627316361869, + "grad_norm": 0.38326382637023926, + "learning_rate": 7.598895350884871e-08, + "loss": 1.0533, + "step": 16172 + }, + { + "epoch": 0.9826234886688134, + "grad_norm": 0.29351603984832764, + "learning_rate": 7.546230113866237e-08, + "loss": 1.1519, + "step": 16173 + }, + { + "epoch": 0.98268424570144, + "grad_norm": 0.19449828565120697, + "learning_rate": 7.493747876204937e-08, + "loss": 1.1017, + "step": 16174 + }, + { + "epoch": 0.9827450027340665, + "grad_norm": 0.19991861283779144, + "learning_rate": 7.441448639826099e-08, + "loss": 1.1511, + "step": 16175 + }, + { + "epoch": 0.982805759766693, + "grad_norm": 0.14351055026054382, + "learning_rate": 7.389332406645966e-08, + "loss": 1.044, + "step": 16176 + }, + { + "epoch": 0.9828665167993195, + "grad_norm": 0.11192145198583603, + "learning_rate": 7.337399178574678e-08, + "loss": 1.0288, + "step": 16177 + }, + { + "epoch": 0.982927273831946, + "grad_norm": 0.12917964160442352, + "learning_rate": 7.285648957515712e-08, + "loss": 1.0076, + "step": 16178 + }, + { + "epoch": 0.9829880308645725, + "grad_norm": 0.19620993733406067, + "learning_rate": 7.234081745366438e-08, + "loss": 0.9933, + "step": 16179 + }, + { + "epoch": 0.9830487878971991, + "grad_norm": 0.19788406789302826, + "learning_rate": 7.182697544017569e-08, + "loss": 1.0904, + "step": 16180 + }, + { + "epoch": 0.9831095449298256, + "grad_norm": 0.2257850170135498, + "learning_rate": 7.131496355351485e-08, + "loss": 1.2234, + "step": 16181 + }, + { + "epoch": 0.9831703019624521, + "grad_norm": 0.13064563274383545, + "learning_rate": 7.080478181245576e-08, + "loss": 1.0637, + "step": 16182 + }, + { + "epoch": 0.9832310589950787, + "grad_norm": 0.2112591713666916, + "learning_rate": 7.02964302357001e-08, + "loss": 1.0469, + "step": 16183 + }, + { + "epoch": 0.9832918160277052, + "grad_norm": 0.21161852777004242, + "learning_rate": 6.978990884187742e-08, + "loss": 1.0902, + "step": 16184 + }, + { + "epoch": 0.9833525730603317, + "grad_norm": 0.12644535303115845, + "learning_rate": 6.928521764956176e-08, + "loss": 1.0306, + "step": 16185 + }, + { + "epoch": 0.9834133300929583, + "grad_norm": 0.11552881449460983, + "learning_rate": 6.878235667724386e-08, + "loss": 1.0626, + "step": 16186 + }, + { + "epoch": 0.9834740871255848, + "grad_norm": 0.2546122670173645, + "learning_rate": 6.828132594336456e-08, + "loss": 1.2017, + "step": 16187 + }, + { + "epoch": 0.9835348441582114, + "grad_norm": 0.14141373336315155, + "learning_rate": 6.778212546628138e-08, + "loss": 1.0425, + "step": 16188 + }, + { + "epoch": 0.9835956011908379, + "grad_norm": 0.12583577632904053, + "learning_rate": 6.728475526429634e-08, + "loss": 1.0489, + "step": 16189 + }, + { + "epoch": 0.9836563582234643, + "grad_norm": 0.1947242021560669, + "learning_rate": 6.67892153556393e-08, + "loss": 1.1533, + "step": 16190 + }, + { + "epoch": 0.9837171152560908, + "grad_norm": 0.12350840121507645, + "learning_rate": 6.629550575847354e-08, + "loss": 1.0031, + "step": 16191 + }, + { + "epoch": 0.9837778722887174, + "grad_norm": 0.41339102387428284, + "learning_rate": 6.580362649090122e-08, + "loss": 0.9906, + "step": 16192 + }, + { + "epoch": 0.9838386293213439, + "grad_norm": 0.211782768368721, + "learning_rate": 6.531357757095235e-08, + "loss": 1.0994, + "step": 16193 + }, + { + "epoch": 0.9838993863539705, + "grad_norm": 0.15947555005550385, + "learning_rate": 6.482535901657927e-08, + "loss": 0.9875, + "step": 16194 + }, + { + "epoch": 0.983960143386597, + "grad_norm": 0.21115639805793762, + "learning_rate": 6.433897084568985e-08, + "loss": 1.0179, + "step": 16195 + }, + { + "epoch": 0.9840209004192235, + "grad_norm": 0.23701371252536774, + "learning_rate": 6.385441307610873e-08, + "loss": 1.0782, + "step": 16196 + }, + { + "epoch": 0.9840816574518501, + "grad_norm": 0.5121179819107056, + "learning_rate": 6.337168572559393e-08, + "loss": 1.0968, + "step": 16197 + }, + { + "epoch": 0.9841424144844766, + "grad_norm": 0.7124260067939758, + "learning_rate": 6.289078881184241e-08, + "loss": 1.1134, + "step": 16198 + }, + { + "epoch": 0.9842031715171031, + "grad_norm": 0.22166576981544495, + "learning_rate": 6.241172235247894e-08, + "loss": 1.0769, + "step": 16199 + }, + { + "epoch": 0.9842639285497297, + "grad_norm": 0.17180852591991425, + "learning_rate": 6.193448636506726e-08, + "loss": 1.085, + "step": 16200 + }, + { + "epoch": 0.9843246855823562, + "grad_norm": 0.17914564907550812, + "learning_rate": 6.14590808671045e-08, + "loss": 1.0043, + "step": 16201 + }, + { + "epoch": 0.9843854426149827, + "grad_norm": 0.2384634017944336, + "learning_rate": 6.098550587601004e-08, + "loss": 1.1174, + "step": 16202 + }, + { + "epoch": 0.9844461996476093, + "grad_norm": 0.19638222455978394, + "learning_rate": 6.051376140914778e-08, + "loss": 1.0701, + "step": 16203 + }, + { + "epoch": 0.9845069566802357, + "grad_norm": 0.2662593424320221, + "learning_rate": 6.004384748379832e-08, + "loss": 1.0212, + "step": 16204 + }, + { + "epoch": 0.9845677137128622, + "grad_norm": 0.1751398742198944, + "learning_rate": 5.957576411720345e-08, + "loss": 1.0259, + "step": 16205 + }, + { + "epoch": 0.9846284707454888, + "grad_norm": 0.26714932918548584, + "learning_rate": 5.9109511326504994e-08, + "loss": 1.0808, + "step": 16206 + }, + { + "epoch": 0.9846892277781153, + "grad_norm": 0.34504997730255127, + "learning_rate": 5.864508912880595e-08, + "loss": 1.0581, + "step": 16207 + }, + { + "epoch": 0.9847499848107418, + "grad_norm": 0.2686702311038971, + "learning_rate": 5.818249754112603e-08, + "loss": 1.0506, + "step": 16208 + }, + { + "epoch": 0.9848107418433684, + "grad_norm": 0.25470781326293945, + "learning_rate": 5.77217365804128e-08, + "loss": 1.053, + "step": 16209 + }, + { + "epoch": 0.9848714988759949, + "grad_norm": 0.19792455434799194, + "learning_rate": 5.726280626356939e-08, + "loss": 1.1038, + "step": 16210 + }, + { + "epoch": 0.9849322559086214, + "grad_norm": 0.2467641830444336, + "learning_rate": 5.680570660741014e-08, + "loss": 1.2001, + "step": 16211 + }, + { + "epoch": 0.984993012941248, + "grad_norm": 0.3211309313774109, + "learning_rate": 5.6350437628688303e-08, + "loss": 1.1551, + "step": 16212 + }, + { + "epoch": 0.9850537699738745, + "grad_norm": 1.2136353254318237, + "learning_rate": 5.589699934410164e-08, + "loss": 0.9635, + "step": 16213 + }, + { + "epoch": 0.985114527006501, + "grad_norm": 0.49433276057243347, + "learning_rate": 5.544539177025909e-08, + "loss": 1.1899, + "step": 16214 + }, + { + "epoch": 0.9851752840391276, + "grad_norm": 0.16868828237056732, + "learning_rate": 5.499561492371963e-08, + "loss": 1.0993, + "step": 16215 + }, + { + "epoch": 0.9852360410717541, + "grad_norm": 0.2784973978996277, + "learning_rate": 5.454766882097007e-08, + "loss": 1.0048, + "step": 16216 + }, + { + "epoch": 0.9852967981043805, + "grad_norm": 1.2006316184997559, + "learning_rate": 5.410155347843615e-08, + "loss": 1.1268, + "step": 16217 + }, + { + "epoch": 0.9853575551370071, + "grad_norm": 0.5824905633926392, + "learning_rate": 5.3657268912454815e-08, + "loss": 1.087, + "step": 16218 + }, + { + "epoch": 0.9854183121696336, + "grad_norm": 0.2584362328052521, + "learning_rate": 5.321481513932969e-08, + "loss": 0.9878, + "step": 16219 + }, + { + "epoch": 0.9854790692022601, + "grad_norm": 3.3345394134521484, + "learning_rate": 5.277419217526447e-08, + "loss": 1.0993, + "step": 16220 + }, + { + "epoch": 0.9855398262348867, + "grad_norm": 0.18873542547225952, + "learning_rate": 5.233540003641291e-08, + "loss": 1.0113, + "step": 16221 + }, + { + "epoch": 0.9856005832675132, + "grad_norm": 0.20059317350387573, + "learning_rate": 5.189843873886768e-08, + "loss": 1.1398, + "step": 16222 + }, + { + "epoch": 0.9856613403001397, + "grad_norm": 0.12081721425056458, + "learning_rate": 5.1463308298638216e-08, + "loss": 1.0278, + "step": 16223 + }, + { + "epoch": 0.9857220973327663, + "grad_norm": 0.16078169643878937, + "learning_rate": 5.1030008731678406e-08, + "loss": 1.0763, + "step": 16224 + }, + { + "epoch": 0.9857828543653928, + "grad_norm": 0.17297233641147614, + "learning_rate": 5.059854005386444e-08, + "loss": 1.0816, + "step": 16225 + }, + { + "epoch": 0.9858436113980193, + "grad_norm": 0.15695062279701233, + "learning_rate": 5.0168902281017003e-08, + "loss": 0.9904, + "step": 16226 + }, + { + "epoch": 0.9859043684306459, + "grad_norm": 0.31816115975379944, + "learning_rate": 4.974109542888461e-08, + "loss": 1.0323, + "step": 16227 + }, + { + "epoch": 0.9859651254632724, + "grad_norm": 0.2413288801908493, + "learning_rate": 4.931511951314915e-08, + "loss": 1.1828, + "step": 16228 + }, + { + "epoch": 0.986025882495899, + "grad_norm": 0.11147577315568924, + "learning_rate": 4.889097454942037e-08, + "loss": 0.9848, + "step": 16229 + }, + { + "epoch": 0.9860866395285254, + "grad_norm": 0.1732899248600006, + "learning_rate": 4.846866055324695e-08, + "loss": 1.0015, + "step": 16230 + }, + { + "epoch": 0.9861473965611519, + "grad_norm": 1.1415661573410034, + "learning_rate": 4.8048177540116476e-08, + "loss": 1.058, + "step": 16231 + }, + { + "epoch": 0.9862081535937784, + "grad_norm": 0.21308070421218872, + "learning_rate": 4.7629525525427767e-08, + "loss": 1.146, + "step": 16232 + }, + { + "epoch": 0.986268910626405, + "grad_norm": 0.23377543687820435, + "learning_rate": 4.721270452454074e-08, + "loss": 1.1567, + "step": 16233 + }, + { + "epoch": 0.9863296676590315, + "grad_norm": 0.10501943528652191, + "learning_rate": 4.679771455272653e-08, + "loss": 1.0038, + "step": 16234 + }, + { + "epoch": 0.986390424691658, + "grad_norm": 0.16991972923278809, + "learning_rate": 4.638455562519517e-08, + "loss": 1.0109, + "step": 16235 + }, + { + "epoch": 0.9864511817242846, + "grad_norm": 0.15158405900001526, + "learning_rate": 4.597322775709567e-08, + "loss": 1.0279, + "step": 16236 + }, + { + "epoch": 0.9865119387569111, + "grad_norm": 0.2010205239057541, + "learning_rate": 4.5563730963499306e-08, + "loss": 1.0414, + "step": 16237 + }, + { + "epoch": 0.9865726957895377, + "grad_norm": 0.11476615071296692, + "learning_rate": 4.515606525942739e-08, + "loss": 1.0176, + "step": 16238 + }, + { + "epoch": 0.9866334528221642, + "grad_norm": 0.13501101732254028, + "learning_rate": 4.475023065981243e-08, + "loss": 1.0315, + "step": 16239 + }, + { + "epoch": 0.9866942098547907, + "grad_norm": 0.20598869025707245, + "learning_rate": 4.434622717953141e-08, + "loss": 1.0242, + "step": 16240 + }, + { + "epoch": 0.9867549668874173, + "grad_norm": 0.6349784135818481, + "learning_rate": 4.394405483340025e-08, + "loss": 1.2812, + "step": 16241 + }, + { + "epoch": 0.9868157239200438, + "grad_norm": 0.27103686332702637, + "learning_rate": 4.3543713636151615e-08, + "loss": 1.1479, + "step": 16242 + }, + { + "epoch": 0.9868764809526702, + "grad_norm": 0.10845721513032913, + "learning_rate": 4.3145203602468207e-08, + "loss": 0.9955, + "step": 16243 + }, + { + "epoch": 0.9869372379852968, + "grad_norm": 0.16707192361354828, + "learning_rate": 4.274852474694946e-08, + "loss": 1.1265, + "step": 16244 + }, + { + "epoch": 0.9869979950179233, + "grad_norm": 0.31152719259262085, + "learning_rate": 4.235367708414484e-08, + "loss": 1.31, + "step": 16245 + }, + { + "epoch": 0.9870587520505498, + "grad_norm": 0.22817759215831757, + "learning_rate": 4.1960660628520554e-08, + "loss": 1.1039, + "step": 16246 + }, + { + "epoch": 0.9871195090831764, + "grad_norm": 0.15281790494918823, + "learning_rate": 4.15694753944873e-08, + "loss": 1.0942, + "step": 16247 + }, + { + "epoch": 0.9871802661158029, + "grad_norm": 0.13286717236042023, + "learning_rate": 4.118012139637806e-08, + "loss": 0.9933, + "step": 16248 + }, + { + "epoch": 0.9872410231484294, + "grad_norm": 0.20295105874538422, + "learning_rate": 4.079259864847029e-08, + "loss": 1.0923, + "step": 16249 + }, + { + "epoch": 0.987301780181056, + "grad_norm": 0.14469251036643982, + "learning_rate": 4.040690716496376e-08, + "loss": 1.0028, + "step": 16250 + }, + { + "epoch": 0.9873625372136825, + "grad_norm": 0.1670244336128235, + "learning_rate": 4.00230469600027e-08, + "loss": 1.0676, + "step": 16251 + }, + { + "epoch": 0.987423294246309, + "grad_norm": 0.14847491681575775, + "learning_rate": 3.9641018047653633e-08, + "loss": 1.0144, + "step": 16252 + }, + { + "epoch": 0.9874840512789356, + "grad_norm": 0.12235371023416519, + "learning_rate": 3.9260820441916477e-08, + "loss": 1.0064, + "step": 16253 + }, + { + "epoch": 0.9875448083115621, + "grad_norm": 0.1961022913455963, + "learning_rate": 3.8882454156735634e-08, + "loss": 1.0909, + "step": 16254 + }, + { + "epoch": 0.9876055653441886, + "grad_norm": 0.480579137802124, + "learning_rate": 3.850591920597224e-08, + "loss": 1.0784, + "step": 16255 + }, + { + "epoch": 0.9876663223768151, + "grad_norm": 0.21070148050785065, + "learning_rate": 3.8131215603437467e-08, + "loss": 1.095, + "step": 16256 + }, + { + "epoch": 0.9877270794094416, + "grad_norm": 0.13883410394191742, + "learning_rate": 3.775834336285367e-08, + "loss": 1.0646, + "step": 16257 + }, + { + "epoch": 0.9877878364420681, + "grad_norm": 0.184866800904274, + "learning_rate": 3.738730249789879e-08, + "loss": 1.1712, + "step": 16258 + }, + { + "epoch": 0.9878485934746947, + "grad_norm": 0.17700117826461792, + "learning_rate": 3.701809302216752e-08, + "loss": 1.1152, + "step": 16259 + }, + { + "epoch": 0.9879093505073212, + "grad_norm": 0.22738327085971832, + "learning_rate": 3.665071494919903e-08, + "loss": 1.0326, + "step": 16260 + }, + { + "epoch": 0.9879701075399477, + "grad_norm": 0.16676084697246552, + "learning_rate": 3.6285168292454766e-08, + "loss": 1.0756, + "step": 16261 + }, + { + "epoch": 0.9880308645725743, + "grad_norm": 0.12049511075019836, + "learning_rate": 3.592145306533512e-08, + "loss": 1.0164, + "step": 16262 + }, + { + "epoch": 0.9880916216052008, + "grad_norm": 0.14225143194198608, + "learning_rate": 3.555956928117388e-08, + "loss": 0.9645, + "step": 16263 + }, + { + "epoch": 0.9881523786378273, + "grad_norm": 0.15711703896522522, + "learning_rate": 3.519951695323265e-08, + "loss": 1.0225, + "step": 16264 + }, + { + "epoch": 0.9882131356704539, + "grad_norm": 0.2732267379760742, + "learning_rate": 3.4841296094711984e-08, + "loss": 1.0211, + "step": 16265 + }, + { + "epoch": 0.9882738927030804, + "grad_norm": 0.23263387382030487, + "learning_rate": 3.4484906718740276e-08, + "loss": 1.0039, + "step": 16266 + }, + { + "epoch": 0.988334649735707, + "grad_norm": 0.2152697890996933, + "learning_rate": 3.413034883838484e-08, + "loss": 1.1156, + "step": 16267 + }, + { + "epoch": 0.9883954067683335, + "grad_norm": 0.16583624482154846, + "learning_rate": 3.377762246663529e-08, + "loss": 1.1449, + "step": 16268 + }, + { + "epoch": 0.9884561638009599, + "grad_norm": 0.3638359308242798, + "learning_rate": 3.3426727616431286e-08, + "loss": 1.0223, + "step": 16269 + }, + { + "epoch": 0.9885169208335864, + "grad_norm": 0.11269981414079666, + "learning_rate": 3.3077664300623645e-08, + "loss": 0.9996, + "step": 16270 + }, + { + "epoch": 0.988577677866213, + "grad_norm": 0.13369394838809967, + "learning_rate": 3.273043253201324e-08, + "loss": 0.9756, + "step": 16271 + }, + { + "epoch": 0.9886384348988395, + "grad_norm": 0.1481218785047531, + "learning_rate": 3.238503232333434e-08, + "loss": 1.0451, + "step": 16272 + }, + { + "epoch": 0.988699191931466, + "grad_norm": 0.13871608674526215, + "learning_rate": 3.2041463687232374e-08, + "loss": 0.9801, + "step": 16273 + }, + { + "epoch": 0.9887599489640926, + "grad_norm": 0.2543506324291229, + "learning_rate": 3.169972663631393e-08, + "loss": 1.2039, + "step": 16274 + }, + { + "epoch": 0.9888207059967191, + "grad_norm": 0.34224602580070496, + "learning_rate": 3.135982118309677e-08, + "loss": 1.0979, + "step": 16275 + }, + { + "epoch": 0.9888814630293457, + "grad_norm": 0.13415755331516266, + "learning_rate": 3.10217473400487e-08, + "loss": 1.0475, + "step": 16276 + }, + { + "epoch": 0.9889422200619722, + "grad_norm": 0.14049571752548218, + "learning_rate": 3.068550511955426e-08, + "loss": 1.0334, + "step": 16277 + }, + { + "epoch": 0.9890029770945987, + "grad_norm": 0.12107331305742264, + "learning_rate": 3.035109453393692e-08, + "loss": 1.051, + "step": 16278 + }, + { + "epoch": 0.9890637341272253, + "grad_norm": 0.13931992650032043, + "learning_rate": 3.0018515595464645e-08, + "loss": 1.0236, + "step": 16279 + }, + { + "epoch": 0.9891244911598518, + "grad_norm": 0.15619860589504242, + "learning_rate": 2.968776831632214e-08, + "loss": 1.0172, + "step": 16280 + }, + { + "epoch": 0.9891852481924783, + "grad_norm": 0.15940599143505096, + "learning_rate": 2.9358852708633032e-08, + "loss": 1.044, + "step": 16281 + }, + { + "epoch": 0.9892460052251048, + "grad_norm": 0.16640304028987885, + "learning_rate": 2.903176878445435e-08, + "loss": 1.0422, + "step": 16282 + }, + { + "epoch": 0.9893067622577313, + "grad_norm": 0.10971435904502869, + "learning_rate": 2.8706516555776497e-08, + "loss": 0.9941, + "step": 16283 + }, + { + "epoch": 0.9893675192903578, + "grad_norm": 0.14190451800823212, + "learning_rate": 2.838309603452327e-08, + "loss": 1.0077, + "step": 16284 + }, + { + "epoch": 0.9894282763229844, + "grad_norm": 0.7089543342590332, + "learning_rate": 2.806150723254075e-08, + "loss": 1.1447, + "step": 16285 + }, + { + "epoch": 0.9894890333556109, + "grad_norm": 0.18077917397022247, + "learning_rate": 2.7741750161625057e-08, + "loss": 1.0835, + "step": 16286 + }, + { + "epoch": 0.9895497903882374, + "grad_norm": 0.23958849906921387, + "learning_rate": 2.7423824833494593e-08, + "loss": 1.0762, + "step": 16287 + }, + { + "epoch": 0.989610547420864, + "grad_norm": 0.15780402719974518, + "learning_rate": 2.71077312598067e-08, + "loss": 1.0197, + "step": 16288 + }, + { + "epoch": 0.9896713044534905, + "grad_norm": 0.12371399253606796, + "learning_rate": 2.6793469452141007e-08, + "loss": 1.0342, + "step": 16289 + }, + { + "epoch": 0.989732061486117, + "grad_norm": 0.18142110109329224, + "learning_rate": 2.6481039422021626e-08, + "loss": 1.0723, + "step": 16290 + }, + { + "epoch": 0.9897928185187436, + "grad_norm": 0.28373855352401733, + "learning_rate": 2.6170441180900506e-08, + "loss": 1.1045, + "step": 16291 + }, + { + "epoch": 0.9898535755513701, + "grad_norm": 0.25553131103515625, + "learning_rate": 2.586167474015744e-08, + "loss": 1.0593, + "step": 16292 + }, + { + "epoch": 0.9899143325839966, + "grad_norm": 0.1502576470375061, + "learning_rate": 2.5554740111116693e-08, + "loss": 1.0535, + "step": 16293 + }, + { + "epoch": 0.9899750896166232, + "grad_norm": 1.3029146194458008, + "learning_rate": 2.5249637305030383e-08, + "loss": 1.0132, + "step": 16294 + }, + { + "epoch": 0.9900358466492496, + "grad_norm": 0.18011648952960968, + "learning_rate": 2.49463663330729e-08, + "loss": 1.0742, + "step": 16295 + }, + { + "epoch": 0.9900966036818761, + "grad_norm": 0.14972196519374847, + "learning_rate": 2.4644927206368685e-08, + "loss": 1.0488, + "step": 16296 + }, + { + "epoch": 0.9901573607145027, + "grad_norm": 0.20186147093772888, + "learning_rate": 2.434531993596445e-08, + "loss": 1.1355, + "step": 16297 + }, + { + "epoch": 0.9902181177471292, + "grad_norm": 1.093990445137024, + "learning_rate": 2.4047544532845857e-08, + "loss": 1.1244, + "step": 16298 + }, + { + "epoch": 0.9902788747797557, + "grad_norm": 0.23552975058555603, + "learning_rate": 2.3751601007920842e-08, + "loss": 1.0685, + "step": 16299 + }, + { + "epoch": 0.9903396318123823, + "grad_norm": 0.138398215174675, + "learning_rate": 2.345748937204184e-08, + "loss": 1.0291, + "step": 16300 + }, + { + "epoch": 0.9904003888450088, + "grad_norm": 0.21474500000476837, + "learning_rate": 2.3165209635989116e-08, + "loss": 1.0638, + "step": 16301 + }, + { + "epoch": 0.9904611458776353, + "grad_norm": 0.1341409832239151, + "learning_rate": 2.287476181047632e-08, + "loss": 1.0131, + "step": 16302 + }, + { + "epoch": 0.9905219029102619, + "grad_norm": 0.12806721031665802, + "learning_rate": 2.258614590615049e-08, + "loss": 1.0123, + "step": 16303 + }, + { + "epoch": 0.9905826599428884, + "grad_norm": 0.17192089557647705, + "learning_rate": 2.2299361933586505e-08, + "loss": 1.072, + "step": 16304 + }, + { + "epoch": 0.990643416975515, + "grad_norm": 0.30139848589897156, + "learning_rate": 2.2014409903303724e-08, + "loss": 1.1974, + "step": 16305 + }, + { + "epoch": 0.9907041740081415, + "grad_norm": 0.14943064749240875, + "learning_rate": 2.1731289825738245e-08, + "loss": 1.0374, + "step": 16306 + }, + { + "epoch": 0.990764931040768, + "grad_norm": 0.2403365522623062, + "learning_rate": 2.14500017112762e-08, + "loss": 1.0905, + "step": 16307 + }, + { + "epoch": 0.9908256880733946, + "grad_norm": 0.11420606076717377, + "learning_rate": 2.117054557022602e-08, + "loss": 1.0361, + "step": 16308 + }, + { + "epoch": 0.990886445106021, + "grad_norm": 0.1579413115978241, + "learning_rate": 2.0892921412829504e-08, + "loss": 1.0321, + "step": 16309 + }, + { + "epoch": 0.9909472021386475, + "grad_norm": 0.12303187698125839, + "learning_rate": 2.0617129249261845e-08, + "loss": 1.0271, + "step": 16310 + }, + { + "epoch": 0.991007959171274, + "grad_norm": 0.16822737455368042, + "learning_rate": 2.0343169089637182e-08, + "loss": 1.0209, + "step": 16311 + }, + { + "epoch": 0.9910687162039006, + "grad_norm": 0.11190731823444366, + "learning_rate": 2.0071040943991926e-08, + "loss": 1.0368, + "step": 16312 + }, + { + "epoch": 0.9911294732365271, + "grad_norm": 0.11619444936513901, + "learning_rate": 1.980074482230698e-08, + "loss": 1.0097, + "step": 16313 + }, + { + "epoch": 0.9911902302691536, + "grad_norm": 0.22781704366207123, + "learning_rate": 1.953228073447999e-08, + "loss": 1.0518, + "step": 16314 + }, + { + "epoch": 0.9912509873017802, + "grad_norm": 0.12319988012313843, + "learning_rate": 1.926564869036418e-08, + "loss": 1.0312, + "step": 16315 + }, + { + "epoch": 0.9913117443344067, + "grad_norm": 0.17771238088607788, + "learning_rate": 1.9000848699723962e-08, + "loss": 1.1125, + "step": 16316 + }, + { + "epoch": 0.9913725013670333, + "grad_norm": 0.18642158806324005, + "learning_rate": 1.8737880772273786e-08, + "loss": 1.0905, + "step": 16317 + }, + { + "epoch": 0.9914332583996598, + "grad_norm": 0.24748654663562775, + "learning_rate": 1.847674491763929e-08, + "loss": 1.1001, + "step": 16318 + }, + { + "epoch": 0.9914940154322863, + "grad_norm": 0.19767725467681885, + "learning_rate": 1.8217441145407243e-08, + "loss": 1.1628, + "step": 16319 + }, + { + "epoch": 0.9915547724649129, + "grad_norm": 0.18430787324905396, + "learning_rate": 1.795996946507561e-08, + "loss": 1.0822, + "step": 16320 + }, + { + "epoch": 0.9916155294975394, + "grad_norm": 0.19113914668560028, + "learning_rate": 1.7704329886081282e-08, + "loss": 1.0817, + "step": 16321 + }, + { + "epoch": 0.9916762865301658, + "grad_norm": 0.1252918690443039, + "learning_rate": 1.7450522417794546e-08, + "loss": 1.0286, + "step": 16322 + }, + { + "epoch": 0.9917370435627924, + "grad_norm": 0.2603912353515625, + "learning_rate": 1.719854706952462e-08, + "loss": 1.1539, + "step": 16323 + }, + { + "epoch": 0.9917978005954189, + "grad_norm": 20.486133575439453, + "learning_rate": 1.6948403850503003e-08, + "loss": 1.0674, + "step": 16324 + }, + { + "epoch": 0.9918585576280454, + "grad_norm": 0.10905192047357559, + "learning_rate": 1.6700092769894593e-08, + "loss": 1.0116, + "step": 16325 + }, + { + "epoch": 0.991919314660672, + "grad_norm": 0.12687237560749054, + "learning_rate": 1.645361383680877e-08, + "loss": 0.9756, + "step": 16326 + }, + { + "epoch": 0.9919800716932985, + "grad_norm": 0.15872901678085327, + "learning_rate": 1.6208967060277193e-08, + "loss": 1.0335, + "step": 16327 + }, + { + "epoch": 0.992040828725925, + "grad_norm": 0.1922800987958908, + "learning_rate": 1.5966152449264916e-08, + "loss": 1.0966, + "step": 16328 + }, + { + "epoch": 0.9921015857585516, + "grad_norm": 0.6323729753494263, + "learning_rate": 1.5725170012681478e-08, + "loss": 1.3105, + "step": 16329 + }, + { + "epoch": 0.9921623427911781, + "grad_norm": 0.6023107171058655, + "learning_rate": 1.5486019759347603e-08, + "loss": 1.1476, + "step": 16330 + }, + { + "epoch": 0.9922230998238046, + "grad_norm": 0.17939308285713196, + "learning_rate": 1.5248701698039604e-08, + "loss": 0.9949, + "step": 16331 + }, + { + "epoch": 0.9922838568564312, + "grad_norm": 0.1638331413269043, + "learning_rate": 1.5013215837450522e-08, + "loss": 1.0595, + "step": 16332 + }, + { + "epoch": 0.9923446138890577, + "grad_norm": 0.19300232827663422, + "learning_rate": 1.4779562186212348e-08, + "loss": 1.1704, + "step": 16333 + }, + { + "epoch": 0.9924053709216842, + "grad_norm": 0.15505912899971008, + "learning_rate": 1.4547740752890449e-08, + "loss": 1.0378, + "step": 16334 + }, + { + "epoch": 0.9924661279543107, + "grad_norm": 1.0033138990402222, + "learning_rate": 1.4317751545983582e-08, + "loss": 1.0825, + "step": 16335 + }, + { + "epoch": 0.9925268849869372, + "grad_norm": 0.13367217779159546, + "learning_rate": 1.4089594573923893e-08, + "loss": 1.0142, + "step": 16336 + }, + { + "epoch": 0.9925876420195637, + "grad_norm": 0.29016193747520447, + "learning_rate": 1.3863269845071359e-08, + "loss": 1.1481, + "step": 16337 + }, + { + "epoch": 0.9926483990521903, + "grad_norm": 0.1312548965215683, + "learning_rate": 1.3638777367724897e-08, + "loss": 1.0968, + "step": 16338 + }, + { + "epoch": 0.9927091560848168, + "grad_norm": 0.2199987769126892, + "learning_rate": 1.3416117150105711e-08, + "loss": 1.1472, + "step": 16339 + }, + { + "epoch": 0.9927699131174433, + "grad_norm": 0.36868685483932495, + "learning_rate": 1.3195289200385042e-08, + "loss": 1.1217, + "step": 16340 + }, + { + "epoch": 0.9928306701500699, + "grad_norm": 0.6374727487564087, + "learning_rate": 1.2976293526656414e-08, + "loss": 1.1409, + "step": 16341 + }, + { + "epoch": 0.9928914271826964, + "grad_norm": 0.2003149688243866, + "learning_rate": 1.2759130136935638e-08, + "loss": 1.094, + "step": 16342 + }, + { + "epoch": 0.9929521842153229, + "grad_norm": 0.13444620370864868, + "learning_rate": 1.2543799039199666e-08, + "loss": 1.0041, + "step": 16343 + }, + { + "epoch": 0.9930129412479495, + "grad_norm": 0.17183896899223328, + "learning_rate": 1.2330300241325531e-08, + "loss": 1.0734, + "step": 16344 + }, + { + "epoch": 0.993073698280576, + "grad_norm": 0.15526849031448364, + "learning_rate": 1.211863375115141e-08, + "loss": 1.0663, + "step": 16345 + }, + { + "epoch": 0.9931344553132025, + "grad_norm": 1.6956391334533691, + "learning_rate": 1.1908799576432207e-08, + "loss": 1.0471, + "step": 16346 + }, + { + "epoch": 0.9931952123458291, + "grad_norm": 0.2077009230852127, + "learning_rate": 1.1700797724856217e-08, + "loss": 1.1557, + "step": 16347 + }, + { + "epoch": 0.9932559693784555, + "grad_norm": 1.7512750625610352, + "learning_rate": 1.1494628204050672e-08, + "loss": 1.0464, + "step": 16348 + }, + { + "epoch": 0.993316726411082, + "grad_norm": 0.13495981693267822, + "learning_rate": 1.1290291021570643e-08, + "loss": 0.9651, + "step": 16349 + }, + { + "epoch": 0.9933774834437086, + "grad_norm": 0.15675342082977295, + "learning_rate": 1.1087786184910132e-08, + "loss": 1.0758, + "step": 16350 + }, + { + "epoch": 0.9934382404763351, + "grad_norm": 0.5898312330245972, + "learning_rate": 1.088711370148543e-08, + "loss": 1.1145, + "step": 16351 + }, + { + "epoch": 0.9934989975089616, + "grad_norm": 0.11548783630132675, + "learning_rate": 1.0688273578657315e-08, + "loss": 1.0113, + "step": 16352 + }, + { + "epoch": 0.9935597545415882, + "grad_norm": 0.13474421203136444, + "learning_rate": 1.0491265823714403e-08, + "loss": 1.06, + "step": 16353 + }, + { + "epoch": 0.9936205115742147, + "grad_norm": 0.14343956112861633, + "learning_rate": 1.0296090443878691e-08, + "loss": 1.0764, + "step": 16354 + }, + { + "epoch": 0.9936812686068413, + "grad_norm": 0.13664895296096802, + "learning_rate": 1.010274744630002e-08, + "loss": 1.0117, + "step": 16355 + }, + { + "epoch": 0.9937420256394678, + "grad_norm": 0.18116681277751923, + "learning_rate": 9.911236838067162e-09, + "loss": 1.1071, + "step": 16356 + }, + { + "epoch": 0.9938027826720943, + "grad_norm": 0.11483743041753769, + "learning_rate": 9.721558626207827e-09, + "loss": 1.0195, + "step": 16357 + }, + { + "epoch": 0.9938635397047209, + "grad_norm": 0.22149419784545898, + "learning_rate": 9.533712817660912e-09, + "loss": 1.206, + "step": 16358 + }, + { + "epoch": 0.9939242967373474, + "grad_norm": 0.21469362080097198, + "learning_rate": 9.347699419320898e-09, + "loss": 1.0504, + "step": 16359 + }, + { + "epoch": 0.9939850537699739, + "grad_norm": 0.1521887332201004, + "learning_rate": 9.163518437999009e-09, + "loss": 1.0553, + "step": 16360 + }, + { + "epoch": 0.9940458108026003, + "grad_norm": 0.2015298455953598, + "learning_rate": 8.981169880456497e-09, + "loss": 1.064, + "step": 16361 + }, + { + "epoch": 0.9941065678352269, + "grad_norm": 0.195998415350914, + "learning_rate": 8.800653753371357e-09, + "loss": 1.0647, + "step": 16362 + }, + { + "epoch": 0.9941673248678534, + "grad_norm": 0.23869946599006653, + "learning_rate": 8.621970063360518e-09, + "loss": 1.0829, + "step": 16363 + }, + { + "epoch": 0.99422808190048, + "grad_norm": 0.184474378824234, + "learning_rate": 8.445118816979846e-09, + "loss": 1.0474, + "step": 16364 + }, + { + "epoch": 0.9942888389331065, + "grad_norm": 0.15587466955184937, + "learning_rate": 8.270100020701943e-09, + "loss": 1.0446, + "step": 16365 + }, + { + "epoch": 0.994349595965733, + "grad_norm": 0.4585663974285126, + "learning_rate": 8.096913680943896e-09, + "loss": 1.0697, + "step": 16366 + }, + { + "epoch": 0.9944103529983596, + "grad_norm": 0.1373085379600525, + "learning_rate": 7.925559804061733e-09, + "loss": 1.0187, + "step": 16367 + }, + { + "epoch": 0.9944711100309861, + "grad_norm": 0.18975766003131866, + "learning_rate": 7.756038396328214e-09, + "loss": 1.0614, + "step": 16368 + }, + { + "epoch": 0.9945318670636126, + "grad_norm": 0.20069445669651031, + "learning_rate": 7.588349463960587e-09, + "loss": 1.1382, + "step": 16369 + }, + { + "epoch": 0.9945926240962392, + "grad_norm": 0.1429833024740219, + "learning_rate": 7.422493013103937e-09, + "loss": 1.013, + "step": 16370 + }, + { + "epoch": 0.9946533811288657, + "grad_norm": 0.1351323127746582, + "learning_rate": 7.258469049842287e-09, + "loss": 1.0286, + "step": 16371 + }, + { + "epoch": 0.9947141381614922, + "grad_norm": 0.19215433299541473, + "learning_rate": 7.0962775801874935e-09, + "loss": 1.0979, + "step": 16372 + }, + { + "epoch": 0.9947748951941188, + "grad_norm": 0.21135517954826355, + "learning_rate": 6.9359186100737e-09, + "loss": 1.066, + "step": 16373 + }, + { + "epoch": 0.9948356522267452, + "grad_norm": 0.11406291276216507, + "learning_rate": 6.777392145396188e-09, + "loss": 1.0089, + "step": 16374 + }, + { + "epoch": 0.9948964092593717, + "grad_norm": 0.1268865019083023, + "learning_rate": 6.620698191950325e-09, + "loss": 0.986, + "step": 16375 + }, + { + "epoch": 0.9949571662919983, + "grad_norm": 0.15171509981155396, + "learning_rate": 6.4658367554870644e-09, + "loss": 1.0069, + "step": 16376 + }, + { + "epoch": 0.9950179233246248, + "grad_norm": 0.18591687083244324, + "learning_rate": 6.312807841685198e-09, + "loss": 1.1299, + "step": 16377 + }, + { + "epoch": 0.9950786803572513, + "grad_norm": 0.1494932472705841, + "learning_rate": 6.16161145615135e-09, + "loss": 1.0462, + "step": 16378 + }, + { + "epoch": 0.9951394373898779, + "grad_norm": 1.8639997243881226, + "learning_rate": 6.012247604425536e-09, + "loss": 1.0452, + "step": 16379 + }, + { + "epoch": 0.9952001944225044, + "grad_norm": 0.21134072542190552, + "learning_rate": 5.864716291981154e-09, + "loss": 1.1125, + "step": 16380 + }, + { + "epoch": 0.9952609514551309, + "grad_norm": 0.4260047674179077, + "learning_rate": 5.719017524236092e-09, + "loss": 1.1417, + "step": 16381 + }, + { + "epoch": 0.9953217084877575, + "grad_norm": 0.12688510119915009, + "learning_rate": 5.575151306519421e-09, + "loss": 1.0, + "step": 16382 + }, + { + "epoch": 0.995382465520384, + "grad_norm": 0.12574705481529236, + "learning_rate": 5.433117644110252e-09, + "loss": 1.0649, + "step": 16383 + }, + { + "epoch": 0.9954432225530105, + "grad_norm": 0.12928487360477448, + "learning_rate": 5.29291654220998e-09, + "loss": 1.0618, + "step": 16384 + }, + { + "epoch": 0.9955039795856371, + "grad_norm": 0.17470209300518036, + "learning_rate": 5.154548005964488e-09, + "loss": 1.0545, + "step": 16385 + }, + { + "epoch": 0.9955647366182636, + "grad_norm": 0.16413837671279907, + "learning_rate": 5.018012040447495e-09, + "loss": 1.0631, + "step": 16386 + }, + { + "epoch": 0.99562549365089, + "grad_norm": 0.15903984010219574, + "learning_rate": 4.883308650655005e-09, + "loss": 1.0381, + "step": 16387 + }, + { + "epoch": 0.9956862506835166, + "grad_norm": 0.1863747239112854, + "learning_rate": 4.750437841527511e-09, + "loss": 1.049, + "step": 16388 + }, + { + "epoch": 0.9957470077161431, + "grad_norm": 0.19599860906600952, + "learning_rate": 4.619399617938891e-09, + "loss": 1.074, + "step": 16389 + }, + { + "epoch": 0.9958077647487696, + "grad_norm": 0.2714446485042572, + "learning_rate": 4.49019398469086e-09, + "loss": 1.2105, + "step": 16390 + }, + { + "epoch": 0.9958685217813962, + "grad_norm": 0.16383177042007446, + "learning_rate": 4.362820946512969e-09, + "loss": 0.9899, + "step": 16391 + }, + { + "epoch": 0.9959292788140227, + "grad_norm": 0.17515195906162262, + "learning_rate": 4.2372805080848065e-09, + "loss": 1.0524, + "step": 16392 + }, + { + "epoch": 0.9959900358466492, + "grad_norm": 0.24984948337078094, + "learning_rate": 4.113572674002697e-09, + "loss": 1.0519, + "step": 16393 + }, + { + "epoch": 0.9960507928792758, + "grad_norm": 0.19413872063159943, + "learning_rate": 3.991697448801901e-09, + "loss": 1.0671, + "step": 16394 + }, + { + "epoch": 0.9961115499119023, + "grad_norm": 0.2320142388343811, + "learning_rate": 3.871654836951066e-09, + "loss": 1.0197, + "step": 16395 + }, + { + "epoch": 0.9961723069445289, + "grad_norm": 0.1180267259478569, + "learning_rate": 3.753444842841125e-09, + "loss": 0.9772, + "step": 16396 + }, + { + "epoch": 0.9962330639771554, + "grad_norm": 0.16709129512310028, + "learning_rate": 3.6370674708186005e-09, + "loss": 1.0374, + "step": 16397 + }, + { + "epoch": 0.9962938210097819, + "grad_norm": 2.01326847076416, + "learning_rate": 3.5225227251467484e-09, + "loss": 1.0211, + "step": 16398 + }, + { + "epoch": 0.9963545780424085, + "grad_norm": 0.10694006085395813, + "learning_rate": 3.4098106100166618e-09, + "loss": 1.0666, + "step": 16399 + }, + { + "epoch": 0.9964153350750349, + "grad_norm": 0.19048918783664703, + "learning_rate": 3.29893112956392e-09, + "loss": 1.0482, + "step": 16400 + }, + { + "epoch": 0.9964760921076614, + "grad_norm": 0.10791581869125366, + "learning_rate": 3.189884287851941e-09, + "loss": 0.9857, + "step": 16401 + }, + { + "epoch": 0.996536849140288, + "grad_norm": 0.12205059826374054, + "learning_rate": 3.0826700888830773e-09, + "loss": 1.0259, + "step": 16402 + }, + { + "epoch": 0.9965976061729145, + "grad_norm": 0.13100798428058624, + "learning_rate": 2.9772885365764168e-09, + "loss": 0.9968, + "step": 16403 + }, + { + "epoch": 0.996658363205541, + "grad_norm": 2.4321212768554688, + "learning_rate": 2.873739634806638e-09, + "loss": 1.178, + "step": 16404 + }, + { + "epoch": 0.9967191202381676, + "grad_norm": 0.21445173025131226, + "learning_rate": 2.7720233873651523e-09, + "loss": 1.1548, + "step": 16405 + }, + { + "epoch": 0.9967798772707941, + "grad_norm": 0.13345611095428467, + "learning_rate": 2.672139797976758e-09, + "loss": 1.0102, + "step": 16406 + }, + { + "epoch": 0.9968406343034206, + "grad_norm": 7.372500419616699, + "learning_rate": 2.5740888703051915e-09, + "loss": 1.0778, + "step": 16407 + }, + { + "epoch": 0.9969013913360472, + "grad_norm": 0.10448794066905975, + "learning_rate": 2.4778706079475745e-09, + "loss": 0.96, + "step": 16408 + }, + { + "epoch": 0.9969621483686737, + "grad_norm": 0.17675848305225372, + "learning_rate": 2.3834850144233143e-09, + "loss": 1.0874, + "step": 16409 + }, + { + "epoch": 0.9970229054013002, + "grad_norm": 0.16304606199264526, + "learning_rate": 2.2909320932018586e-09, + "loss": 1.0489, + "step": 16410 + }, + { + "epoch": 0.9970836624339268, + "grad_norm": 0.1152791902422905, + "learning_rate": 2.2002118476638355e-09, + "loss": 1.0456, + "step": 16411 + }, + { + "epoch": 0.9971444194665533, + "grad_norm": 6.010976791381836, + "learning_rate": 2.111324281145466e-09, + "loss": 1.0665, + "step": 16412 + }, + { + "epoch": 0.9972051764991798, + "grad_norm": 0.2412978708744049, + "learning_rate": 2.024269396899703e-09, + "loss": 1.0614, + "step": 16413 + }, + { + "epoch": 0.9972659335318063, + "grad_norm": 0.21254907548427582, + "learning_rate": 1.939047198123989e-09, + "loss": 1.1304, + "step": 16414 + }, + { + "epoch": 0.9973266905644328, + "grad_norm": 0.3980441391468048, + "learning_rate": 1.8556576879324993e-09, + "loss": 1.1139, + "step": 16415 + }, + { + "epoch": 0.9973874475970593, + "grad_norm": 0.1165439710021019, + "learning_rate": 1.774100869383899e-09, + "loss": 0.9655, + "step": 16416 + }, + { + "epoch": 0.9974482046296859, + "grad_norm": 0.1297926902770996, + "learning_rate": 1.6943767454757897e-09, + "loss": 1.0175, + "step": 16417 + }, + { + "epoch": 0.9975089616623124, + "grad_norm": 0.13189701735973358, + "learning_rate": 1.616485319122507e-09, + "loss": 1.0233, + "step": 16418 + }, + { + "epoch": 0.9975697186949389, + "grad_norm": 0.12016113102436066, + "learning_rate": 1.5404265931773243e-09, + "loss": 1.0037, + "step": 16419 + }, + { + "epoch": 0.9976304757275655, + "grad_norm": 0.138640359044075, + "learning_rate": 1.4662005704380033e-09, + "loss": 1.0485, + "step": 16420 + }, + { + "epoch": 0.997691232760192, + "grad_norm": 0.1815258413553238, + "learning_rate": 1.3938072536134884e-09, + "loss": 1.0808, + "step": 16421 + }, + { + "epoch": 0.9977519897928185, + "grad_norm": 0.22147329151630402, + "learning_rate": 1.3232466453683146e-09, + "loss": 1.1333, + "step": 16422 + }, + { + "epoch": 0.9978127468254451, + "grad_norm": 0.18007372319698334, + "learning_rate": 1.2545187482837505e-09, + "loss": 1.0904, + "step": 16423 + }, + { + "epoch": 0.9978735038580716, + "grad_norm": 0.3677292466163635, + "learning_rate": 1.1876235648800027e-09, + "loss": 1.1522, + "step": 16424 + }, + { + "epoch": 0.9979342608906981, + "grad_norm": 0.14289970695972443, + "learning_rate": 1.1225610976051126e-09, + "loss": 1.0438, + "step": 16425 + }, + { + "epoch": 0.9979950179233247, + "grad_norm": 0.11188235878944397, + "learning_rate": 1.059331348851611e-09, + "loss": 1.0626, + "step": 16426 + }, + { + "epoch": 0.9980557749559511, + "grad_norm": 0.11805826425552368, + "learning_rate": 9.979343209287618e-10, + "loss": 1.0131, + "step": 16427 + }, + { + "epoch": 0.9981165319885776, + "grad_norm": 0.19771507382392883, + "learning_rate": 9.383700160958686e-10, + "loss": 1.1611, + "step": 16428 + }, + { + "epoch": 0.9981772890212042, + "grad_norm": 0.13267900049686432, + "learning_rate": 8.806384365289688e-10, + "loss": 1.0557, + "step": 16429 + }, + { + "epoch": 0.9982380460538307, + "grad_norm": 0.251427561044693, + "learning_rate": 8.247395843430372e-10, + "loss": 1.0424, + "step": 16430 + }, + { + "epoch": 0.9982988030864572, + "grad_norm": 0.14963307976722717, + "learning_rate": 7.706734615975375e-10, + "loss": 1.0525, + "step": 16431 + }, + { + "epoch": 0.9983595601190838, + "grad_norm": 0.16894052922725677, + "learning_rate": 7.184400702631156e-10, + "loss": 1.0778, + "step": 16432 + }, + { + "epoch": 0.9984203171517103, + "grad_norm": 0.1591864824295044, + "learning_rate": 6.680394122604572e-10, + "loss": 1.1138, + "step": 16433 + }, + { + "epoch": 0.9984810741843368, + "grad_norm": 0.119852215051651, + "learning_rate": 6.194714894380838e-10, + "loss": 0.9852, + "step": 16434 + }, + { + "epoch": 0.9985418312169634, + "grad_norm": 0.24688120186328888, + "learning_rate": 5.727363035668009e-10, + "loss": 1.1265, + "step": 16435 + }, + { + "epoch": 0.9986025882495899, + "grad_norm": 0.4413137137889862, + "learning_rate": 5.278338563730056e-10, + "loss": 1.1105, + "step": 16436 + }, + { + "epoch": 0.9986633452822165, + "grad_norm": 0.11139416694641113, + "learning_rate": 4.847641494942767e-10, + "loss": 1.0001, + "step": 16437 + }, + { + "epoch": 0.998724102314843, + "grad_norm": 0.1553596705198288, + "learning_rate": 4.4352718451268204e-10, + "loss": 1.0795, + "step": 16438 + }, + { + "epoch": 0.9987848593474695, + "grad_norm": 0.19915704429149628, + "learning_rate": 4.041229629325738e-10, + "loss": 1.1049, + "step": 16439 + }, + { + "epoch": 0.998845616380096, + "grad_norm": 0.12232493609189987, + "learning_rate": 3.6655148620834413e-10, + "loss": 1.03, + "step": 16440 + }, + { + "epoch": 0.9989063734127225, + "grad_norm": 0.1777585744857788, + "learning_rate": 3.308127557111185e-10, + "loss": 1.0395, + "step": 16441 + }, + { + "epoch": 0.998967130445349, + "grad_norm": 0.23771603405475616, + "learning_rate": 2.9690677275651114e-10, + "loss": 1.1803, + "step": 16442 + }, + { + "epoch": 0.9990278874779756, + "grad_norm": 0.29386481642723083, + "learning_rate": 2.648335385824208e-10, + "loss": 1.1175, + "step": 16443 + }, + { + "epoch": 0.9990886445106021, + "grad_norm": 0.24861106276512146, + "learning_rate": 2.345930543601327e-10, + "loss": 1.1544, + "step": 16444 + }, + { + "epoch": 0.9991494015432286, + "grad_norm": 0.8491747379302979, + "learning_rate": 2.0618532121097213e-10, + "loss": 1.0517, + "step": 16445 + }, + { + "epoch": 0.9992101585758552, + "grad_norm": 0.1639053374528885, + "learning_rate": 1.7961034016189538e-10, + "loss": 1.1966, + "step": 16446 + }, + { + "epoch": 0.9992709156084817, + "grad_norm": 0.21290753781795502, + "learning_rate": 1.5486811220100096e-10, + "loss": 1.0967, + "step": 16447 + }, + { + "epoch": 0.9993316726411082, + "grad_norm": 0.23757170140743256, + "learning_rate": 1.3195863822201838e-10, + "loss": 1.0508, + "step": 16448 + }, + { + "epoch": 0.9993924296737348, + "grad_norm": 0.7063068151473999, + "learning_rate": 1.1088191907426825e-10, + "loss": 1.0297, + "step": 16449 + }, + { + "epoch": 0.9994531867063613, + "grad_norm": 0.24706414341926575, + "learning_rate": 9.163795552935561e-11, + "loss": 1.1474, + "step": 16450 + }, + { + "epoch": 0.9995139437389878, + "grad_norm": 0.13747864961624146, + "learning_rate": 7.422674828672093e-11, + "loss": 1.0289, + "step": 16451 + }, + { + "epoch": 0.9995747007716144, + "grad_norm": 0.138253852725029, + "learning_rate": 5.864829798474247e-11, + "loss": 1.0515, + "step": 16452 + }, + { + "epoch": 0.9996354578042408, + "grad_norm": 0.17477886378765106, + "learning_rate": 4.490260520628731e-11, + "loss": 1.0291, + "step": 16453 + }, + { + "epoch": 0.9996962148368673, + "grad_norm": 0.39978620409965515, + "learning_rate": 3.298967043985357e-11, + "loss": 1.0578, + "step": 16454 + }, + { + "epoch": 0.9997569718694939, + "grad_norm": 0.17506404221057892, + "learning_rate": 2.2909494129530474e-11, + "loss": 1.1475, + "step": 16455 + }, + { + "epoch": 0.9998177289021204, + "grad_norm": 0.26539742946624756, + "learning_rate": 1.466207664724273e-11, + "loss": 1.0442, + "step": 16456 + }, + { + "epoch": 0.9998784859347469, + "grad_norm": 0.14767298102378845, + "learning_rate": 8.247418292750553e-12, + "loss": 1.0678, + "step": 16457 + }, + { + "epoch": 0.9999392429673735, + "grad_norm": 3.146981716156006, + "learning_rate": 3.665519293649666e-12, + "loss": 1.0601, + "step": 16458 + }, + { + "epoch": 1.0, + "grad_norm": 0.18900837004184723, + "learning_rate": 9.16379833126868e-13, + "loss": 1.1165, + "step": 16459 + } + ], + "logging_steps": 1.0, + "max_steps": 16459, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 9.881425348891194e+19, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}