{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 3125, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0032, "grad_norm": 366.0841096744857, "learning_rate": 1.437699680511182e-07, "loss": 23.85431823730469, "step": 10 }, { "epoch": 0.0064, "grad_norm": 367.47333882445946, "learning_rate": 3.0351437699680514e-07, "loss": 23.65589599609375, "step": 20 }, { "epoch": 0.0096, "grad_norm": 367.96579270464326, "learning_rate": 4.6325878594249205e-07, "loss": 22.780029296875, "step": 30 }, { "epoch": 0.0128, "grad_norm": 332.5732884154056, "learning_rate": 6.230031948881789e-07, "loss": 20.279689025878906, "step": 40 }, { "epoch": 0.016, "grad_norm": 219.53674756423746, "learning_rate": 7.82747603833866e-07, "loss": 15.498806762695313, "step": 50 }, { "epoch": 0.0192, "grad_norm": 156.1487544830451, "learning_rate": 9.424920127795528e-07, "loss": 10.388201904296874, "step": 60 }, { "epoch": 0.0224, "grad_norm": 37.96869040917498, "learning_rate": 1.1022364217252397e-06, "loss": 3.7560958862304688, "step": 70 }, { "epoch": 0.0256, "grad_norm": 16.783464772614202, "learning_rate": 1.2619808306709266e-06, "loss": 2.033830261230469, "step": 80 }, { "epoch": 0.0288, "grad_norm": 5.438256169634593, "learning_rate": 1.4217252396166134e-06, "loss": 1.0431390762329102, "step": 90 }, { "epoch": 0.032, "grad_norm": 3.6935246150045775, "learning_rate": 1.5814696485623005e-06, "loss": 0.8069572448730469, "step": 100 }, { "epoch": 0.0352, "grad_norm": 9.218312544625562, "learning_rate": 1.7412140575079875e-06, "loss": 0.7057615280151367, "step": 110 }, { "epoch": 0.0384, "grad_norm": 5.394484238866305, "learning_rate": 1.9009584664536742e-06, "loss": 0.6301750183105469, "step": 120 }, { "epoch": 0.0416, "grad_norm": 6.577481237217732, "learning_rate": 2.060702875399361e-06, "loss": 0.5898516654968262, "step": 130 }, { "epoch": 0.0448, "grad_norm": 3.4158074483641068, "learning_rate": 2.220447284345048e-06, "loss": 0.5524418830871582, "step": 140 }, { "epoch": 0.048, "grad_norm": 4.032046521040006, "learning_rate": 2.380191693290735e-06, "loss": 0.5317594051361084, "step": 150 }, { "epoch": 0.0512, "grad_norm": 5.468634675306576, "learning_rate": 2.539936102236422e-06, "loss": 0.5184277534484864, "step": 160 }, { "epoch": 0.0544, "grad_norm": 3.4313124951156424, "learning_rate": 2.699680511182109e-06, "loss": 0.5204483985900878, "step": 170 }, { "epoch": 0.0576, "grad_norm": 5.13400179254009, "learning_rate": 2.8594249201277955e-06, "loss": 0.5058025360107422, "step": 180 }, { "epoch": 0.0608, "grad_norm": 5.9183424216837786, "learning_rate": 3.0191693290734825e-06, "loss": 0.5073411941528321, "step": 190 }, { "epoch": 0.064, "grad_norm": 5.625073986187664, "learning_rate": 3.17891373801917e-06, "loss": 0.5000103950500489, "step": 200 }, { "epoch": 0.0672, "grad_norm": 5.050603467051007, "learning_rate": 3.3386581469648564e-06, "loss": 0.488192081451416, "step": 210 }, { "epoch": 0.0704, "grad_norm": 11.776866822937645, "learning_rate": 3.4984025559105434e-06, "loss": 0.48699202537536623, "step": 220 }, { "epoch": 0.0736, "grad_norm": 7.438900018795585, "learning_rate": 3.6581469648562303e-06, "loss": 0.4820102691650391, "step": 230 }, { "epoch": 0.0768, "grad_norm": 4.3491840646532065, "learning_rate": 3.817891373801918e-06, "loss": 0.47640199661254884, "step": 240 }, { "epoch": 0.08, "grad_norm": 3.472565426233091, "learning_rate": 3.977635782747604e-06, "loss": 0.4729574203491211, "step": 250 }, { "epoch": 0.0832, "grad_norm": 3.1912744148161942, "learning_rate": 4.137380191693291e-06, "loss": 0.4786433219909668, "step": 260 }, { "epoch": 0.0864, "grad_norm": 3.9698013424470777, "learning_rate": 4.297124600638978e-06, "loss": 0.4748369216918945, "step": 270 }, { "epoch": 0.0896, "grad_norm": 8.11949393489321, "learning_rate": 4.456869009584665e-06, "loss": 0.4681865692138672, "step": 280 }, { "epoch": 0.0928, "grad_norm": 4.7349566199381234, "learning_rate": 4.616613418530352e-06, "loss": 0.46743001937866213, "step": 290 }, { "epoch": 0.096, "grad_norm": 4.756427284033883, "learning_rate": 4.776357827476039e-06, "loss": 0.46964178085327146, "step": 300 }, { "epoch": 0.0992, "grad_norm": 4.86570605379029, "learning_rate": 4.936102236421725e-06, "loss": 0.45612516403198244, "step": 310 }, { "epoch": 0.1024, "grad_norm": 5.762654788054032, "learning_rate": 4.999943833158769e-06, "loss": 0.45009474754333495, "step": 320 }, { "epoch": 0.1056, "grad_norm": 3.501477346053355, "learning_rate": 4.999600600490783e-06, "loss": 0.4523477554321289, "step": 330 }, { "epoch": 0.1088, "grad_norm": 7.957279740713588, "learning_rate": 4.9989453817439345e-06, "loss": 0.44190473556518556, "step": 340 }, { "epoch": 0.112, "grad_norm": 7.660308885793361, "learning_rate": 4.997978258698942e-06, "loss": 0.43758931159973147, "step": 350 }, { "epoch": 0.1152, "grad_norm": 5.8839479464224205, "learning_rate": 4.996699352066659e-06, "loss": 0.4371060371398926, "step": 360 }, { "epoch": 0.1184, "grad_norm": 3.452842882877267, "learning_rate": 4.995108821473014e-06, "loss": 0.42999753952026365, "step": 370 }, { "epoch": 0.1216, "grad_norm": 4.825810317520427, "learning_rate": 4.993206865439084e-06, "loss": 0.4285894393920898, "step": 380 }, { "epoch": 0.1248, "grad_norm": 5.379766821254966, "learning_rate": 4.990993721356317e-06, "loss": 0.42139811515808107, "step": 390 }, { "epoch": 0.128, "grad_norm": 4.854730410799869, "learning_rate": 4.988469665456901e-06, "loss": 0.42040281295776366, "step": 400 }, { "epoch": 0.1312, "grad_norm": 4.6616615938661745, "learning_rate": 4.985635012779288e-06, "loss": 0.4207456588745117, "step": 410 }, { "epoch": 0.1344, "grad_norm": 4.5341296475975605, "learning_rate": 4.98249011712887e-06, "loss": 0.414472770690918, "step": 420 }, { "epoch": 0.1376, "grad_norm": 5.217437981869656, "learning_rate": 4.979035371033824e-06, "loss": 0.41441006660461427, "step": 430 }, { "epoch": 0.1408, "grad_norm": 3.561516924716779, "learning_rate": 4.975271205696115e-06, "loss": 0.40767755508422854, "step": 440 }, { "epoch": 0.144, "grad_norm": 3.815692337476438, "learning_rate": 4.971198090937671e-06, "loss": 0.3997596263885498, "step": 450 }, { "epoch": 0.1472, "grad_norm": 4.559242371997167, "learning_rate": 4.966816535141756e-06, "loss": 0.39360842704772947, "step": 460 }, { "epoch": 0.1504, "grad_norm": 3.432229350472061, "learning_rate": 4.9621270851895035e-06, "loss": 0.40289998054504395, "step": 470 }, { "epoch": 0.1536, "grad_norm": 5.375227134041046, "learning_rate": 4.957130326391662e-06, "loss": 0.3982266664505005, "step": 480 }, { "epoch": 0.1568, "grad_norm": 5.539585521677851, "learning_rate": 4.951826882415544e-06, "loss": 0.39270691871643065, "step": 490 }, { "epoch": 0.16, "grad_norm": 3.4147092253345743, "learning_rate": 4.946217415207177e-06, "loss": 0.3853750705718994, "step": 500 }, { "epoch": 0.1632, "grad_norm": 4.444175842440995, "learning_rate": 4.940302624908689e-06, "loss": 0.38694162368774415, "step": 510 }, { "epoch": 0.1664, "grad_norm": 3.3493207902303475, "learning_rate": 4.934083249770912e-06, "loss": 0.3797153949737549, "step": 520 }, { "epoch": 0.1696, "grad_norm": 3.0499194254019097, "learning_rate": 4.927560066061251e-06, "loss": 0.38063654899597166, "step": 530 }, { "epoch": 0.1728, "grad_norm": 3.141871281336489, "learning_rate": 4.920733887966783e-06, "loss": 0.39005699157714846, "step": 540 }, { "epoch": 0.176, "grad_norm": 3.979297184951908, "learning_rate": 4.913605567492636e-06, "loss": 0.38013472557067873, "step": 550 }, { "epoch": 0.1792, "grad_norm": 3.7669251986704113, "learning_rate": 4.906175994355656e-06, "loss": 0.37832577228546144, "step": 560 }, { "epoch": 0.1824, "grad_norm": 2.983798431857085, "learning_rate": 4.898446095873345e-06, "loss": 0.38150479793548586, "step": 570 }, { "epoch": 0.1856, "grad_norm": 3.657787030439589, "learning_rate": 4.890416836848128e-06, "loss": 0.3775670528411865, "step": 580 }, { "epoch": 0.1888, "grad_norm": 3.551048022748126, "learning_rate": 4.882089219446925e-06, "loss": 0.37199065685272215, "step": 590 }, { "epoch": 0.192, "grad_norm": 4.750977601329729, "learning_rate": 4.873464283076074e-06, "loss": 0.3790221452713013, "step": 600 }, { "epoch": 0.1952, "grad_norm": 7.684545118387627, "learning_rate": 4.864543104251587e-06, "loss": 0.37508673667907716, "step": 610 }, { "epoch": 0.1984, "grad_norm": 5.872575231845199, "learning_rate": 4.855326796464798e-06, "loss": 0.3811868906021118, "step": 620 }, { "epoch": 0.2016, "grad_norm": 3.9960144706794316, "learning_rate": 4.8458165100433725e-06, "loss": 0.37326750755310056, "step": 630 }, { "epoch": 0.2048, "grad_norm": 3.9998452581157657, "learning_rate": 4.836013432007738e-06, "loss": 0.3709099769592285, "step": 640 }, { "epoch": 0.208, "grad_norm": 2.6973135018594343, "learning_rate": 4.825918785922921e-06, "loss": 0.3728507995605469, "step": 650 }, { "epoch": 0.2112, "grad_norm": 4.478756132604264, "learning_rate": 4.8155338317458315e-06, "loss": 0.36782591342926024, "step": 660 }, { "epoch": 0.2144, "grad_norm": 2.5620662799375378, "learning_rate": 4.804859865668002e-06, "loss": 0.36416780948638916, "step": 670 }, { "epoch": 0.2176, "grad_norm": 2.9398359151969884, "learning_rate": 4.793898219953804e-06, "loss": 0.36772732734680175, "step": 680 }, { "epoch": 0.2208, "grad_norm": 3.404020172068192, "learning_rate": 4.782650262774164e-06, "loss": 0.3651688575744629, "step": 690 }, { "epoch": 0.224, "grad_norm": 2.588678061474319, "learning_rate": 4.7711173980357886e-06, "loss": 0.3649880409240723, "step": 700 }, { "epoch": 0.2272, "grad_norm": 3.5390276900279773, "learning_rate": 4.759301065205947e-06, "loss": 0.3612825870513916, "step": 710 }, { "epoch": 0.2304, "grad_norm": 3.8670986814196473, "learning_rate": 4.7472027391328e-06, "loss": 0.3657612085342407, "step": 720 }, { "epoch": 0.2336, "grad_norm": 3.0276354554801217, "learning_rate": 4.734823929861317e-06, "loss": 0.36682844161987305, "step": 730 }, { "epoch": 0.2368, "grad_norm": 5.205227283770371, "learning_rate": 4.722166182444801e-06, "loss": 0.3605961322784424, "step": 740 }, { "epoch": 0.24, "grad_norm": 3.1037248816470737, "learning_rate": 4.709231076752045e-06, "loss": 0.3625338554382324, "step": 750 }, { "epoch": 0.2432, "grad_norm": 3.827009314178272, "learning_rate": 4.696020227270142e-06, "loss": 0.36273531913757323, "step": 760 }, { "epoch": 0.2464, "grad_norm": 2.553717481812464, "learning_rate": 4.6825352829029705e-06, "loss": 0.35740270614624026, "step": 770 }, { "epoch": 0.2496, "grad_norm": 2.8273485176739563, "learning_rate": 4.668777926765392e-06, "loss": 0.3613132953643799, "step": 780 }, { "epoch": 0.2528, "grad_norm": 3.242165291552063, "learning_rate": 4.6547498759731725e-06, "loss": 0.3525214672088623, "step": 790 }, { "epoch": 0.256, "grad_norm": 2.607635187753211, "learning_rate": 4.6404528814286575e-06, "loss": 0.3569283723831177, "step": 800 }, { "epoch": 0.2592, "grad_norm": 3.2439792578606204, "learning_rate": 4.6258887276022425e-06, "loss": 0.357681941986084, "step": 810 }, { "epoch": 0.2624, "grad_norm": 2.9728036180938284, "learning_rate": 4.611059232309639e-06, "loss": 0.3537192106246948, "step": 820 }, { "epoch": 0.2656, "grad_norm": 2.556165398739607, "learning_rate": 4.595966246484986e-06, "loss": 0.3528641700744629, "step": 830 }, { "epoch": 0.2688, "grad_norm": 2.593548528246384, "learning_rate": 4.580611653949829e-06, "loss": 0.3564203500747681, "step": 840 }, { "epoch": 0.272, "grad_norm": 3.428440109671292, "learning_rate": 4.564997371177992e-06, "loss": 0.3518026828765869, "step": 850 }, { "epoch": 0.2752, "grad_norm": 4.993564850548027, "learning_rate": 4.54912534705637e-06, "loss": 0.35079920291900635, "step": 860 }, { "epoch": 0.2784, "grad_norm": 3.340510283095063, "learning_rate": 4.532997562641683e-06, "loss": 0.3466078042984009, "step": 870 }, { "epoch": 0.2816, "grad_norm": 2.6894615056191644, "learning_rate": 4.516616030913214e-06, "loss": 0.3472653865814209, "step": 880 }, { "epoch": 0.2848, "grad_norm": 1.891440124594712, "learning_rate": 4.499982796521556e-06, "loss": 0.34483723640441893, "step": 890 }, { "epoch": 0.288, "grad_norm": 3.223309297530686, "learning_rate": 4.48309993553341e-06, "loss": 0.3444544553756714, "step": 900 }, { "epoch": 0.2912, "grad_norm": 3.1032077209020468, "learning_rate": 4.465969555172468e-06, "loss": 0.34571564197540283, "step": 910 }, { "epoch": 0.2944, "grad_norm": 2.5407458837926638, "learning_rate": 4.448593793556391e-06, "loss": 0.3534140110015869, "step": 920 }, { "epoch": 0.2976, "grad_norm": 3.1253686498979123, "learning_rate": 4.430974819429954e-06, "loss": 0.3445676326751709, "step": 930 }, { "epoch": 0.3008, "grad_norm": 3.740083740472538, "learning_rate": 4.413114831894344e-06, "loss": 0.33962287902832033, "step": 940 }, { "epoch": 0.304, "grad_norm": 4.724023923665093, "learning_rate": 4.3950160601326865e-06, "loss": 0.3363780498504639, "step": 950 }, { "epoch": 0.3072, "grad_norm": 3.597276867142834, "learning_rate": 4.376680763131811e-06, "loss": 0.3429840087890625, "step": 960 }, { "epoch": 0.3104, "grad_norm": 2.97998267012516, "learning_rate": 4.358111229400296e-06, "loss": 0.3470882177352905, "step": 970 }, { "epoch": 0.3136, "grad_norm": 3.1405275857331856, "learning_rate": 4.33930977668283e-06, "loss": 0.35235731601715087, "step": 980 }, { "epoch": 0.3168, "grad_norm": 3.774584318253359, "learning_rate": 4.320278751670922e-06, "loss": 0.3418004512786865, "step": 990 }, { "epoch": 0.32, "grad_norm": 3.4325438208492605, "learning_rate": 4.301020529710009e-06, "loss": 0.3456583499908447, "step": 1000 }, { "epoch": 0.3232, "grad_norm": 3.1407187711443916, "learning_rate": 4.281537514502962e-06, "loss": 0.3446167469024658, "step": 1010 }, { "epoch": 0.3264, "grad_norm": 2.6154317834679226, "learning_rate": 4.261832137810093e-06, "loss": 0.34354138374328613, "step": 1020 }, { "epoch": 0.3296, "grad_norm": 2.8993376261822648, "learning_rate": 4.241906859145611e-06, "loss": 0.3451784372329712, "step": 1030 }, { "epoch": 0.3328, "grad_norm": 2.3351853591260574, "learning_rate": 4.221764165470661e-06, "loss": 0.33875834941864014, "step": 1040 }, { "epoch": 0.336, "grad_norm": 3.4295735539049605, "learning_rate": 4.201406570882898e-06, "loss": 0.33980226516723633, "step": 1050 }, { "epoch": 0.3392, "grad_norm": 2.6388634367096735, "learning_rate": 4.180836616302704e-06, "loss": 0.3395829200744629, "step": 1060 }, { "epoch": 0.3424, "grad_norm": 3.211009486395674, "learning_rate": 4.160056869156041e-06, "loss": 0.3433471441268921, "step": 1070 }, { "epoch": 0.3456, "grad_norm": 3.4377414857289317, "learning_rate": 4.139069923053995e-06, "loss": 0.34047765731811525, "step": 1080 }, { "epoch": 0.3488, "grad_norm": 3.131466112366247, "learning_rate": 4.117878397469062e-06, "loss": 0.3420018434524536, "step": 1090 }, { "epoch": 0.352, "grad_norm": 2.388207923072635, "learning_rate": 4.096484937408195e-06, "loss": 0.3351470470428467, "step": 1100 }, { "epoch": 0.3552, "grad_norm": 2.2910707329028117, "learning_rate": 4.074892213082676e-06, "loss": 0.33539299964904784, "step": 1110 }, { "epoch": 0.3584, "grad_norm": 2.156244058261874, "learning_rate": 4.0531029195748265e-06, "loss": 0.33862009048461916, "step": 1120 }, { "epoch": 0.3616, "grad_norm": 2.6382644444406296, "learning_rate": 4.03111977650163e-06, "loss": 0.34041495323181153, "step": 1130 }, { "epoch": 0.3648, "grad_norm": 2.5960896388831545, "learning_rate": 4.008945527675281e-06, "loss": 0.3390871524810791, "step": 1140 }, { "epoch": 0.368, "grad_norm": 3.657074741484568, "learning_rate": 3.986582940760717e-06, "loss": 0.3278806209564209, "step": 1150 }, { "epoch": 0.3712, "grad_norm": 2.9587401358526075, "learning_rate": 3.9640348069301785e-06, "loss": 0.3368961334228516, "step": 1160 }, { "epoch": 0.3744, "grad_norm": 1.965300565427372, "learning_rate": 3.941303940514826e-06, "loss": 0.3339808464050293, "step": 1170 }, { "epoch": 0.3776, "grad_norm": 2.90985435283837, "learning_rate": 3.918393178653472e-06, "loss": 0.3376065969467163, "step": 1180 }, { "epoch": 0.3808, "grad_norm": 3.27190473511409, "learning_rate": 3.895305380938468e-06, "loss": 0.3342454433441162, "step": 1190 }, { "epoch": 0.384, "grad_norm": 2.0468253424433165, "learning_rate": 3.872043429058783e-06, "loss": 0.32965447902679446, "step": 1200 }, { "epoch": 0.3872, "grad_norm": 2.5123150680001576, "learning_rate": 3.84861022644033e-06, "loss": 0.3357837677001953, "step": 1210 }, { "epoch": 0.3904, "grad_norm": 3.148104290988529, "learning_rate": 3.825008697883574e-06, "loss": 0.34343953132629396, "step": 1220 }, { "epoch": 0.3936, "grad_norm": 2.488823913942074, "learning_rate": 3.8012417891984776e-06, "loss": 0.333116340637207, "step": 1230 }, { "epoch": 0.3968, "grad_norm": 3.0225259799028645, "learning_rate": 3.777312466836819e-06, "loss": 0.3318933486938477, "step": 1240 }, { "epoch": 0.4, "grad_norm": 3.3439153363899115, "learning_rate": 3.7532237175219378e-06, "loss": 0.32833037376403806, "step": 1250 }, { "epoch": 0.4032, "grad_norm": 2.72884090647899, "learning_rate": 3.728978547875948e-06, "loss": 0.3360243082046509, "step": 1260 }, { "epoch": 0.4064, "grad_norm": 2.5999080124511966, "learning_rate": 3.7045799840444712e-06, "loss": 0.33025145530700684, "step": 1270 }, { "epoch": 0.4096, "grad_norm": 3.0518346526448488, "learning_rate": 3.6800310713189258e-06, "loss": 0.3306798219680786, "step": 1280 }, { "epoch": 0.4128, "grad_norm": 2.0509087709244507, "learning_rate": 3.6553348737564328e-06, "loss": 0.33091559410095217, "step": 1290 }, { "epoch": 0.416, "grad_norm": 2.908137390744499, "learning_rate": 3.6304944737973794e-06, "loss": 0.33455810546875, "step": 1300 }, { "epoch": 0.4192, "grad_norm": 3.0396312942670796, "learning_rate": 3.6055129718806836e-06, "loss": 0.331624960899353, "step": 1310 }, { "epoch": 0.4224, "grad_norm": 3.282462978283218, "learning_rate": 3.5803934860568134e-06, "loss": 0.32364490032196047, "step": 1320 }, { "epoch": 0.4256, "grad_norm": 2.2269456751164727, "learning_rate": 3.5551391515986163e-06, "loss": 0.3319955348968506, "step": 1330 }, { "epoch": 0.4288, "grad_norm": 2.8364899461485527, "learning_rate": 3.529753120609982e-06, "loss": 0.3252741813659668, "step": 1340 }, { "epoch": 0.432, "grad_norm": 2.89515974439621, "learning_rate": 3.5042385616324243e-06, "loss": 0.3287111520767212, "step": 1350 }, { "epoch": 0.4352, "grad_norm": 2.311001238312573, "learning_rate": 3.4785986592495934e-06, "loss": 0.32939796447753905, "step": 1360 }, { "epoch": 0.4384, "grad_norm": 2.4126049139350734, "learning_rate": 3.452836613689803e-06, "loss": 0.32168779373168943, "step": 1370 }, { "epoch": 0.4416, "grad_norm": 3.1765584413022254, "learning_rate": 3.426955640426584e-06, "loss": 0.32864985466003416, "step": 1380 }, { "epoch": 0.4448, "grad_norm": 3.154206643410634, "learning_rate": 3.4009589697773605e-06, "loss": 0.3260640621185303, "step": 1390 }, { "epoch": 0.448, "grad_norm": 3.4230687653412564, "learning_rate": 3.3748498465002475e-06, "loss": 0.32304584980010986, "step": 1400 }, { "epoch": 0.4512, "grad_norm": 2.6276396964869684, "learning_rate": 3.3486315293890693e-06, "loss": 0.33318138122558594, "step": 1410 }, { "epoch": 0.4544, "grad_norm": 2.754821177049362, "learning_rate": 3.3223072908666053e-06, "loss": 0.32256054878234863, "step": 1420 }, { "epoch": 0.4576, "grad_norm": 2.881952130772473, "learning_rate": 3.295880416576153e-06, "loss": 0.33387539386749265, "step": 1430 }, { "epoch": 0.4608, "grad_norm": 2.5217047707442966, "learning_rate": 3.269354204971427e-06, "loss": 0.32321481704711913, "step": 1440 }, { "epoch": 0.464, "grad_norm": 2.976679985492794, "learning_rate": 3.242731966904865e-06, "loss": 0.32245721817016604, "step": 1450 }, { "epoch": 0.4672, "grad_norm": 2.527563459090948, "learning_rate": 3.2160170252143913e-06, "loss": 0.32239205837249757, "step": 1460 }, { "epoch": 0.4704, "grad_norm": 1.997832889519553, "learning_rate": 3.1892127143086716e-06, "loss": 0.32758924961090086, "step": 1470 }, { "epoch": 0.4736, "grad_norm": 2.299101703675196, "learning_rate": 3.1623223797509347e-06, "loss": 0.31891183853149413, "step": 1480 }, { "epoch": 0.4768, "grad_norm": 2.9210746413068907, "learning_rate": 3.135349377841396e-06, "loss": 0.32430353164672854, "step": 1490 }, { "epoch": 0.48, "grad_norm": 2.6265609696149146, "learning_rate": 3.1082970751983497e-06, "loss": 0.3312281608581543, "step": 1500 }, { "epoch": 0.4832, "grad_norm": 2.5956160397204786, "learning_rate": 3.0811688483379546e-06, "loss": 0.3238035202026367, "step": 1510 }, { "epoch": 0.4864, "grad_norm": 2.231793404952503, "learning_rate": 3.0539680832528074e-06, "loss": 0.32330875396728515, "step": 1520 }, { "epoch": 0.4896, "grad_norm": 2.5723097920479763, "learning_rate": 3.026698174989316e-06, "loss": 0.32520170211791993, "step": 1530 }, { "epoch": 0.4928, "grad_norm": 2.691498291676849, "learning_rate": 2.999362527223952e-06, "loss": 0.3273704290390015, "step": 1540 }, { "epoch": 0.496, "grad_norm": 2.0511124933056375, "learning_rate": 2.9719645518384194e-06, "loss": 0.3250606536865234, "step": 1550 }, { "epoch": 0.4992, "grad_norm": 2.872290392112785, "learning_rate": 2.944507668493807e-06, "loss": 0.3281686782836914, "step": 1560 }, { "epoch": 0.5024, "grad_norm": 2.330246614888919, "learning_rate": 2.9169953042037623e-06, "loss": 0.32374157905578616, "step": 1570 }, { "epoch": 0.5056, "grad_norm": 2.0520711406500394, "learning_rate": 2.889430892906754e-06, "loss": 0.3169667720794678, "step": 1580 }, { "epoch": 0.5088, "grad_norm": 2.048670737699487, "learning_rate": 2.861817875037462e-06, "loss": 0.3160442590713501, "step": 1590 }, { "epoch": 0.512, "grad_norm": 2.8695840695234303, "learning_rate": 2.8341596970973683e-06, "loss": 0.32544608116149903, "step": 1600 }, { "epoch": 0.5152, "grad_norm": 1.976397223627746, "learning_rate": 2.80645981122458e-06, "loss": 0.3229134798049927, "step": 1610 }, { "epoch": 0.5184, "grad_norm": 2.7070609575351807, "learning_rate": 2.7787216747629508e-06, "loss": 0.32655487060546873, "step": 1620 }, { "epoch": 0.5216, "grad_norm": 2.6027463070090993, "learning_rate": 2.7509487498305615e-06, "loss": 0.31430754661560056, "step": 1630 }, { "epoch": 0.5248, "grad_norm": 2.4274539931656585, "learning_rate": 2.7231445028875924e-06, "loss": 0.3237884759902954, "step": 1640 }, { "epoch": 0.528, "grad_norm": 1.9308598632845329, "learning_rate": 2.6953124043036604e-06, "loss": 0.32111692428588867, "step": 1650 }, { "epoch": 0.5312, "grad_norm": 2.1321964485217784, "learning_rate": 2.667455927924667e-06, "loss": 0.3178241729736328, "step": 1660 }, { "epoch": 0.5344, "grad_norm": 3.1390388403682534, "learning_rate": 2.6395785506392164e-06, "loss": 0.31754770278930666, "step": 1670 }, { "epoch": 0.5376, "grad_norm": 2.137535651695072, "learning_rate": 2.6116837519446407e-06, "loss": 0.3183767795562744, "step": 1680 }, { "epoch": 0.5408, "grad_norm": 2.353751591087722, "learning_rate": 2.5837750135127192e-06, "loss": 0.31382954120635986, "step": 1690 }, { "epoch": 0.544, "grad_norm": 2.58704039056448, "learning_rate": 2.555855818755108e-06, "loss": 0.3226866483688354, "step": 1700 }, { "epoch": 0.5472, "grad_norm": 2.709677414439902, "learning_rate": 2.5279296523885636e-06, "loss": 0.3166576623916626, "step": 1710 }, { "epoch": 0.5504, "grad_norm": 2.0859245317104107, "learning_rate": 2.5e-06, "loss": 0.3218212127685547, "step": 1720 }, { "epoch": 0.5536, "grad_norm": 2.3347357869338436, "learning_rate": 2.472070347611437e-06, "loss": 0.31246294975280764, "step": 1730 }, { "epoch": 0.5568, "grad_norm": 2.5799420800617106, "learning_rate": 2.444144181244893e-06, "loss": 0.31868853569030764, "step": 1740 }, { "epoch": 0.56, "grad_norm": 2.8867509619529406, "learning_rate": 2.416224986487282e-06, "loss": 0.31381807327270506, "step": 1750 }, { "epoch": 0.5632, "grad_norm": 2.625660671305278, "learning_rate": 2.3883162480553605e-06, "loss": 0.31146280765533446, "step": 1760 }, { "epoch": 0.5664, "grad_norm": 2.8862495653341544, "learning_rate": 2.3604214493607844e-06, "loss": 0.3111546993255615, "step": 1770 }, { "epoch": 0.5696, "grad_norm": 2.267020272744141, "learning_rate": 2.332544072075333e-06, "loss": 0.32178173065185545, "step": 1780 }, { "epoch": 0.5728, "grad_norm": 2.073205643473978, "learning_rate": 2.30468759569634e-06, "loss": 0.31751441955566406, "step": 1790 }, { "epoch": 0.576, "grad_norm": 2.232045258362397, "learning_rate": 2.276855497112408e-06, "loss": 0.3135702610015869, "step": 1800 }, { "epoch": 0.5792, "grad_norm": 3.4632505976937744, "learning_rate": 2.2490512501694394e-06, "loss": 0.3126095771789551, "step": 1810 }, { "epoch": 0.5824, "grad_norm": 2.7008114205550022, "learning_rate": 2.2212783252370496e-06, "loss": 0.31725611686706545, "step": 1820 }, { "epoch": 0.5856, "grad_norm": 2.640110404643157, "learning_rate": 2.1935401887754213e-06, "loss": 0.3210929870605469, "step": 1830 }, { "epoch": 0.5888, "grad_norm": 2.9154181525967924, "learning_rate": 2.165840302902632e-06, "loss": 0.31817543506622314, "step": 1840 }, { "epoch": 0.592, "grad_norm": 2.3435756622683916, "learning_rate": 2.1381821249625383e-06, "loss": 0.3186073303222656, "step": 1850 }, { "epoch": 0.5952, "grad_norm": 2.391868801860604, "learning_rate": 2.1105691070932465e-06, "loss": 0.3081700563430786, "step": 1860 }, { "epoch": 0.5984, "grad_norm": 2.27033295147997, "learning_rate": 2.083004695796238e-06, "loss": 0.30403599739074705, "step": 1870 }, { "epoch": 0.6016, "grad_norm": 2.1095837820360157, "learning_rate": 2.055492331506194e-06, "loss": 0.31353535652160647, "step": 1880 }, { "epoch": 0.6048, "grad_norm": 2.284519052184323, "learning_rate": 2.0280354481615814e-06, "loss": 0.31677517890930174, "step": 1890 }, { "epoch": 0.608, "grad_norm": 2.237766836173548, "learning_rate": 2.000637472776049e-06, "loss": 0.3152945041656494, "step": 1900 }, { "epoch": 0.6112, "grad_norm": 2.7842715157490434, "learning_rate": 1.973301825010685e-06, "loss": 0.30818216800689696, "step": 1910 }, { "epoch": 0.6144, "grad_norm": 2.4813744091778784, "learning_rate": 1.9460319167471934e-06, "loss": 0.31820502281188967, "step": 1920 }, { "epoch": 0.6176, "grad_norm": 2.0821248606030887, "learning_rate": 1.9188311516620466e-06, "loss": 0.31040709018707274, "step": 1930 }, { "epoch": 0.6208, "grad_norm": 2.9336859566866975, "learning_rate": 1.891702924801651e-06, "loss": 0.31292426586151123, "step": 1940 }, { "epoch": 0.624, "grad_norm": 2.511253012965921, "learning_rate": 1.864650622158604e-06, "loss": 0.32196660041809083, "step": 1950 }, { "epoch": 0.6272, "grad_norm": 2.4545922236833455, "learning_rate": 1.8376776202490666e-06, "loss": 0.31464810371398927, "step": 1960 }, { "epoch": 0.6304, "grad_norm": 2.277913414668649, "learning_rate": 1.8107872856913293e-06, "loss": 0.30748977661132815, "step": 1970 }, { "epoch": 0.6336, "grad_norm": 3.6960663974743273, "learning_rate": 1.7839829747856096e-06, "loss": 0.31303911209106444, "step": 1980 }, { "epoch": 0.6368, "grad_norm": 2.5169048193896844, "learning_rate": 1.7572680330951359e-06, "loss": 0.309541130065918, "step": 1990 }, { "epoch": 0.64, "grad_norm": 2.625801312197355, "learning_rate": 1.7306457950285747e-06, "loss": 0.31228773593902587, "step": 2000 }, { "epoch": 0.6432, "grad_norm": 3.166705714244592, "learning_rate": 1.704119583423848e-06, "loss": 0.30709683895111084, "step": 2010 }, { "epoch": 0.6464, "grad_norm": 2.7529448920288755, "learning_rate": 1.677692709133396e-06, "loss": 0.3121641159057617, "step": 2020 }, { "epoch": 0.6496, "grad_norm": 2.4164266641009386, "learning_rate": 1.6513684706109311e-06, "loss": 0.31612191200256345, "step": 2030 }, { "epoch": 0.6528, "grad_norm": 2.1475852674178486, "learning_rate": 1.6251501534997529e-06, "loss": 0.30926761627197263, "step": 2040 }, { "epoch": 0.656, "grad_norm": 3.027937409819003, "learning_rate": 1.5990410302226405e-06, "loss": 0.3059820652008057, "step": 2050 }, { "epoch": 0.6592, "grad_norm": 2.3663528005893575, "learning_rate": 1.5730443595734162e-06, "loss": 0.30960190296173096, "step": 2060 }, { "epoch": 0.6624, "grad_norm": 2.5495655090650806, "learning_rate": 1.5471633863101982e-06, "loss": 0.3146512508392334, "step": 2070 }, { "epoch": 0.6656, "grad_norm": 2.563871195645732, "learning_rate": 1.521401340750407e-06, "loss": 0.3116560935974121, "step": 2080 }, { "epoch": 0.6688, "grad_norm": 2.4316488926893314, "learning_rate": 1.495761438367577e-06, "loss": 0.31447796821594237, "step": 2090 }, { "epoch": 0.672, "grad_norm": 2.446980089200077, "learning_rate": 1.4702468793900187e-06, "loss": 0.3153538703918457, "step": 2100 }, { "epoch": 0.6752, "grad_norm": 2.2511595317617283, "learning_rate": 1.444860848401384e-06, "loss": 0.31273808479309084, "step": 2110 }, { "epoch": 0.6784, "grad_norm": 2.459748219135552, "learning_rate": 1.4196065139431866e-06, "loss": 0.31059865951538085, "step": 2120 }, { "epoch": 0.6816, "grad_norm": 2.4570005490031805, "learning_rate": 1.3944870281193178e-06, "loss": 0.31122384071350095, "step": 2130 }, { "epoch": 0.6848, "grad_norm": 2.5940034157380447, "learning_rate": 1.3695055262026208e-06, "loss": 0.3145638704299927, "step": 2140 }, { "epoch": 0.688, "grad_norm": 2.8940635665298644, "learning_rate": 1.3446651262435679e-06, "loss": 0.31133465766906737, "step": 2150 }, { "epoch": 0.6912, "grad_norm": 2.2603444512196216, "learning_rate": 1.3199689286810746e-06, "loss": 0.31110968589782717, "step": 2160 }, { "epoch": 0.6944, "grad_norm": 2.3697248986342223, "learning_rate": 1.2954200159555294e-06, "loss": 0.3046250820159912, "step": 2170 }, { "epoch": 0.6976, "grad_norm": 2.9149559965372083, "learning_rate": 1.2710214521240527e-06, "loss": 0.3056375503540039, "step": 2180 }, { "epoch": 0.7008, "grad_norm": 2.785583016537511, "learning_rate": 1.246776282478063e-06, "loss": 0.3074607849121094, "step": 2190 }, { "epoch": 0.704, "grad_norm": 2.238483419316128, "learning_rate": 1.222687533163181e-06, "loss": 0.30821986198425294, "step": 2200 }, { "epoch": 0.7072, "grad_norm": 2.0963873111402225, "learning_rate": 1.1987582108015228e-06, "loss": 0.31098227500915526, "step": 2210 }, { "epoch": 0.7104, "grad_norm": 2.3511311934322725, "learning_rate": 1.1749913021164255e-06, "loss": 0.3125911712646484, "step": 2220 }, { "epoch": 0.7136, "grad_norm": 2.0182013166602735, "learning_rate": 1.1513897735596702e-06, "loss": 0.30506420135498047, "step": 2230 }, { "epoch": 0.7168, "grad_norm": 2.0904990978865654, "learning_rate": 1.127956570941218e-06, "loss": 0.30170474052429197, "step": 2240 }, { "epoch": 0.72, "grad_norm": 2.3591898483151525, "learning_rate": 1.104694619061533e-06, "loss": 0.3140627145767212, "step": 2250 }, { "epoch": 0.7232, "grad_norm": 2.3874798738589553, "learning_rate": 1.0816068213465295e-06, "loss": 0.3148207187652588, "step": 2260 }, { "epoch": 0.7264, "grad_norm": 2.462173136321867, "learning_rate": 1.0586960594851762e-06, "loss": 0.30828402042388914, "step": 2270 }, { "epoch": 0.7296, "grad_norm": 2.2877287929832946, "learning_rate": 1.0359651930698217e-06, "loss": 0.30725433826446535, "step": 2280 }, { "epoch": 0.7328, "grad_norm": 2.5585705908550413, "learning_rate": 1.0134170592392837e-06, "loss": 0.30991530418395996, "step": 2290 }, { "epoch": 0.736, "grad_norm": 2.415441399008779, "learning_rate": 9.910544723247204e-07, "loss": 0.31087689399719237, "step": 2300 }, { "epoch": 0.7392, "grad_norm": 2.6450690086623285, "learning_rate": 9.688802234983706e-07, "loss": 0.3067446231842041, "step": 2310 }, { "epoch": 0.7424, "grad_norm": 2.363123649822279, "learning_rate": 9.468970804251742e-07, "loss": 0.30767192840576174, "step": 2320 }, { "epoch": 0.7456, "grad_norm": 2.245412676348008, "learning_rate": 9.251077869173244e-07, "loss": 0.30107917785644533, "step": 2330 }, { "epoch": 0.7488, "grad_norm": 2.5736642361970503, "learning_rate": 9.035150625918054e-07, "loss": 0.303986120223999, "step": 2340 }, { "epoch": 0.752, "grad_norm": 2.6844109007429138, "learning_rate": 8.821216025309395e-07, "loss": 0.3074802875518799, "step": 2350 }, { "epoch": 0.7552, "grad_norm": 2.412670568786912, "learning_rate": 8.609300769460055e-07, "loss": 0.30130510330200194, "step": 2360 }, { "epoch": 0.7584, "grad_norm": 3.176069141824472, "learning_rate": 8.399431308439592e-07, "loss": 0.3105806827545166, "step": 2370 }, { "epoch": 0.7616, "grad_norm": 2.23339472526297, "learning_rate": 8.191633836972962e-07, "loss": 0.3084972620010376, "step": 2380 }, { "epoch": 0.7648, "grad_norm": 2.6912839020724175, "learning_rate": 7.985934291171024e-07, "loss": 0.3067460536956787, "step": 2390 }, { "epoch": 0.768, "grad_norm": 2.5426618104677976, "learning_rate": 7.7823583452934e-07, "loss": 0.30809898376464845, "step": 2400 }, { "epoch": 0.7712, "grad_norm": 2.55531536817282, "learning_rate": 7.58093140854389e-07, "loss": 0.3071744441986084, "step": 2410 }, { "epoch": 0.7744, "grad_norm": 2.285863017236424, "learning_rate": 7.381678621899077e-07, "loss": 0.3093477725982666, "step": 2420 }, { "epoch": 0.7776, "grad_norm": 2.3600405361881767, "learning_rate": 7.184624854970379e-07, "loss": 0.30798888206481934, "step": 2430 }, { "epoch": 0.7808, "grad_norm": 2.0247328579355726, "learning_rate": 6.989794702899932e-07, "loss": 0.3048464298248291, "step": 2440 }, { "epoch": 0.784, "grad_norm": 2.7079172300622334, "learning_rate": 6.797212483290777e-07, "loss": 0.3093360424041748, "step": 2450 }, { "epoch": 0.7872, "grad_norm": 2.8011999237207967, "learning_rate": 6.60690223317171e-07, "loss": 0.30233092308044435, "step": 2460 }, { "epoch": 0.7904, "grad_norm": 2.202966089641912, "learning_rate": 6.418887705997046e-07, "loss": 0.3048731327056885, "step": 2470 }, { "epoch": 0.7936, "grad_norm": 2.6510546903467755, "learning_rate": 6.23319236868189e-07, "loss": 0.3104764461517334, "step": 2480 }, { "epoch": 0.7968, "grad_norm": 2.510992490322273, "learning_rate": 6.049839398673141e-07, "loss": 0.31223044395446775, "step": 2490 }, { "epoch": 0.8, "grad_norm": 2.7988283248607604, "learning_rate": 5.868851681056567e-07, "loss": 0.3109541893005371, "step": 2500 }, { "epoch": 0.8032, "grad_norm": 2.370572243788772, "learning_rate": 5.690251805700467e-07, "loss": 0.3075347900390625, "step": 2510 }, { "epoch": 0.8064, "grad_norm": 2.057318428676814, "learning_rate": 5.514062064436096e-07, "loss": 0.30944228172302246, "step": 2520 }, { "epoch": 0.8096, "grad_norm": 2.9526395601791937, "learning_rate": 5.34030444827533e-07, "loss": 0.30773684978485105, "step": 2530 }, { "epoch": 0.8128, "grad_norm": 2.1808951881567165, "learning_rate": 5.169000644665895e-07, "loss": 0.30281686782836914, "step": 2540 }, { "epoch": 0.816, "grad_norm": 2.501184820191482, "learning_rate": 5.000172034784442e-07, "loss": 0.30731327533721925, "step": 2550 }, { "epoch": 0.8192, "grad_norm": 2.4433836822113304, "learning_rate": 4.833839690867853e-07, "loss": 0.30861892700195315, "step": 2560 }, { "epoch": 0.8224, "grad_norm": 2.482955525732734, "learning_rate": 4.6700243735831705e-07, "loss": 0.3014340400695801, "step": 2570 }, { "epoch": 0.8256, "grad_norm": 2.516375989369738, "learning_rate": 4.508746529436311e-07, "loss": 0.302032995223999, "step": 2580 }, { "epoch": 0.8288, "grad_norm": 2.2676227598264926, "learning_rate": 4.350026288220083e-07, "loss": 0.30550131797790525, "step": 2590 }, { "epoch": 0.832, "grad_norm": 2.3829531066293126, "learning_rate": 4.1938834605017133e-07, "loss": 0.3046237945556641, "step": 2600 }, { "epoch": 0.8352, "grad_norm": 2.0018887466739548, "learning_rate": 4.0403375351501515e-07, "loss": 0.3024258852005005, "step": 2610 }, { "epoch": 0.8384, "grad_norm": 2.5182571334882597, "learning_rate": 3.88940767690362e-07, "loss": 0.3063870906829834, "step": 2620 }, { "epoch": 0.8416, "grad_norm": 2.7441991027074355, "learning_rate": 3.7411127239775774e-07, "loss": 0.30306272506713866, "step": 2630 }, { "epoch": 0.8448, "grad_norm": 2.161963722714269, "learning_rate": 3.595471185713431e-07, "loss": 0.3009947299957275, "step": 2640 }, { "epoch": 0.848, "grad_norm": 2.7694143698141285, "learning_rate": 3.4525012402682826e-07, "loss": 0.30188300609588625, "step": 2650 }, { "epoch": 0.8512, "grad_norm": 2.6814413975784217, "learning_rate": 3.3122207323460804e-07, "loss": 0.3024703025817871, "step": 2660 }, { "epoch": 0.8544, "grad_norm": 2.4444711671869306, "learning_rate": 3.1746471709702963e-07, "loss": 0.3008608102798462, "step": 2670 }, { "epoch": 0.8576, "grad_norm": 2.6886622183433015, "learning_rate": 3.039797727298585e-07, "loss": 0.30821614265441893, "step": 2680 }, { "epoch": 0.8608, "grad_norm": 2.641784614909192, "learning_rate": 2.9076892324795546e-07, "loss": 0.30515303611755373, "step": 2690 }, { "epoch": 0.864, "grad_norm": 2.5595370943122444, "learning_rate": 2.778338175551995e-07, "loss": 0.3007267236709595, "step": 2700 }, { "epoch": 0.8672, "grad_norm": 2.283872628964803, "learning_rate": 2.6517607013868326e-07, "loss": 0.30617167949676516, "step": 2710 }, { "epoch": 0.8704, "grad_norm": 2.558413840419693, "learning_rate": 2.527972608672002e-07, "loss": 0.3038905143737793, "step": 2720 }, { "epoch": 0.8736, "grad_norm": 2.4952676522317567, "learning_rate": 2.40698934794053e-07, "loss": 0.3054081201553345, "step": 2730 }, { "epoch": 0.8768, "grad_norm": 2.247637838190116, "learning_rate": 2.2888260196421237e-07, "loss": 0.3028261661529541, "step": 2740 }, { "epoch": 0.88, "grad_norm": 2.5035963414447804, "learning_rate": 2.1734973722583735e-07, "loss": 0.3062435626983643, "step": 2750 }, { "epoch": 0.8832, "grad_norm": 1.918923632238423, "learning_rate": 2.0610178004619564e-07, "loss": 0.2972743034362793, "step": 2760 }, { "epoch": 0.8864, "grad_norm": 2.4603002546330845, "learning_rate": 1.9514013433199834e-07, "loss": 0.3119321346282959, "step": 2770 }, { "epoch": 0.8896, "grad_norm": 2.1315709346733667, "learning_rate": 1.8446616825416958e-07, "loss": 0.30900893211364744, "step": 2780 }, { "epoch": 0.8928, "grad_norm": 2.3753122188061218, "learning_rate": 1.7408121407708007e-07, "loss": 0.3069151401519775, "step": 2790 }, { "epoch": 0.896, "grad_norm": 2.207415755325001, "learning_rate": 1.6398656799226253e-07, "loss": 0.2986165523529053, "step": 2800 }, { "epoch": 0.8992, "grad_norm": 2.178561452169741, "learning_rate": 1.5418348995662773e-07, "loss": 0.3010268688201904, "step": 2810 }, { "epoch": 0.9024, "grad_norm": 2.5082064593439393, "learning_rate": 1.4467320353520275e-07, "loss": 0.2984073877334595, "step": 2820 }, { "epoch": 0.9056, "grad_norm": 2.366814729694057, "learning_rate": 1.3545689574841341e-07, "loss": 0.3026757001876831, "step": 2830 }, { "epoch": 0.9088, "grad_norm": 2.380709246306716, "learning_rate": 1.26535716923927e-07, "loss": 0.310437536239624, "step": 2840 }, { "epoch": 0.912, "grad_norm": 2.484246324702375, "learning_rate": 1.1791078055307493e-07, "loss": 0.30369887351989744, "step": 2850 }, { "epoch": 0.9152, "grad_norm": 2.6412244000001786, "learning_rate": 1.0958316315187289e-07, "loss": 0.3044759750366211, "step": 2860 }, { "epoch": 0.9184, "grad_norm": 2.4542916560781967, "learning_rate": 1.0155390412665528e-07, "loss": 0.30136928558349607, "step": 2870 }, { "epoch": 0.9216, "grad_norm": 2.631911471911446, "learning_rate": 9.38240056443443e-08, "loss": 0.30144243240356444, "step": 2880 }, { "epoch": 0.9248, "grad_norm": 2.2530200189747243, "learning_rate": 8.639443250736402e-08, "loss": 0.3027902603149414, "step": 2890 }, { "epoch": 0.928, "grad_norm": 3.1331936934174123, "learning_rate": 7.926611203321777e-08, "loss": 0.30441856384277344, "step": 2900 }, { "epoch": 0.9312, "grad_norm": 2.5134219010551067, "learning_rate": 7.243993393874882e-08, "loss": 0.306389307975769, "step": 2910 }, { "epoch": 0.9344, "grad_norm": 2.372785201514508, "learning_rate": 6.591675022908805e-08, "loss": 0.30292179584503176, "step": 2920 }, { "epoch": 0.9376, "grad_norm": 2.407913531878434, "learning_rate": 5.969737509131241e-08, "loss": 0.29895825386047364, "step": 2930 }, { "epoch": 0.9408, "grad_norm": 2.2376435379528865, "learning_rate": 5.3782584792823334e-08, "loss": 0.30271134376525877, "step": 2940 }, { "epoch": 0.944, "grad_norm": 2.653290725438786, "learning_rate": 4.817311758445686e-08, "loss": 0.3062829732894897, "step": 2950 }, { "epoch": 0.9472, "grad_norm": 2.42511171945876, "learning_rate": 4.286967360833866e-08, "loss": 0.3066932439804077, "step": 2960 }, { "epoch": 0.9504, "grad_norm": 2.1534299736877895, "learning_rate": 3.787291481049754e-08, "loss": 0.3068870544433594, "step": 2970 }, { "epoch": 0.9536, "grad_norm": 2.209956884835794, "learning_rate": 3.3183464858244364e-08, "loss": 0.31453580856323243, "step": 2980 }, { "epoch": 0.9568, "grad_norm": 2.5928568899987017, "learning_rate": 2.8801909062328992e-08, "loss": 0.2991969108581543, "step": 2990 }, { "epoch": 0.96, "grad_norm": 2.385980918167846, "learning_rate": 2.4728794303886248e-08, "loss": 0.2963397026062012, "step": 3000 }, { "epoch": 0.9632, "grad_norm": 2.374100986684654, "learning_rate": 2.0964628966175794e-08, "loss": 0.30301966667175295, "step": 3010 }, { "epoch": 0.9664, "grad_norm": 2.094256605734986, "learning_rate": 1.750988287113009e-08, "loss": 0.2994666576385498, "step": 3020 }, { "epoch": 0.9696, "grad_norm": 1.916185239441286, "learning_rate": 1.4364987220713278e-08, "loss": 0.3080729007720947, "step": 3030 }, { "epoch": 0.9728, "grad_norm": 2.3446521041543207, "learning_rate": 1.1530334543099763e-08, "loss": 0.3026130199432373, "step": 3040 }, { "epoch": 0.976, "grad_norm": 2.5854178252734323, "learning_rate": 9.006278643683697e-09, "loss": 0.309655499458313, "step": 3050 }, { "epoch": 0.9792, "grad_norm": 1.9908162772434517, "learning_rate": 6.793134560916514e-09, "loss": 0.31186389923095703, "step": 3060 }, { "epoch": 0.9824, "grad_norm": 2.1977962094508534, "learning_rate": 4.891178526986451e-09, "loss": 0.30645883083343506, "step": 3070 }, { "epoch": 0.9856, "grad_norm": 2.2397406638818147, "learning_rate": 3.3006479333413943e-09, "loss": 0.3090504169464111, "step": 3080 }, { "epoch": 0.9888, "grad_norm": 2.0435901319475036, "learning_rate": 2.021741301058422e-09, "loss": 0.3049570322036743, "step": 3090 }, { "epoch": 0.992, "grad_norm": 2.371036869409615, "learning_rate": 1.0546182560652872e-09, "loss": 0.3073274612426758, "step": 3100 }, { "epoch": 0.9952, "grad_norm": 2.2551729202130457, "learning_rate": 3.9939950921774607e-10, "loss": 0.30047030448913575, "step": 3110 }, { "epoch": 0.9984, "grad_norm": 2.2067081414460827, "learning_rate": 5.616684123160854e-11, "loss": 0.3023503065109253, "step": 3120 }, { "epoch": 1.0, "step": 3125, "total_flos": 1.0913057758773248e+16, "train_loss": 0.7292402684783935, "train_runtime": 30167.0559, "train_samples_per_second": 6.63, "train_steps_per_second": 0.104 } ], "logging_steps": 10, "max_steps": 3125, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.0913057758773248e+16, "train_batch_size": 4, "trial_name": null, "trial_params": null }