| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.9878213802435725, |
| "eval_steps": 500, |
| "global_step": 276, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.010825439783491205, |
| "grad_norm": 5.8088322268939026, |
| "learning_rate": 2.8571428571428573e-06, |
| "loss": 0.8566, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.02165087956698241, |
| "grad_norm": 5.813170605377498, |
| "learning_rate": 5.7142857142857145e-06, |
| "loss": 0.8603, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.03247631935047361, |
| "grad_norm": 5.442220189869692, |
| "learning_rate": 8.571428571428571e-06, |
| "loss": 0.8525, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.04330175913396482, |
| "grad_norm": 2.305473908583789, |
| "learning_rate": 1.1428571428571429e-05, |
| "loss": 0.7759, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.05412719891745602, |
| "grad_norm": 3.92748540597831, |
| "learning_rate": 1.4285714285714287e-05, |
| "loss": 0.7543, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.06495263870094722, |
| "grad_norm": 4.2082400956680965, |
| "learning_rate": 1.7142857142857142e-05, |
| "loss": 0.7509, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.07577807848443843, |
| "grad_norm": 4.355204221488866, |
| "learning_rate": 2e-05, |
| "loss": 0.7343, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.08660351826792964, |
| "grad_norm": 2.910526629158842, |
| "learning_rate": 2.2857142857142858e-05, |
| "loss": 0.7289, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.09742895805142084, |
| "grad_norm": 2.7507600752385195, |
| "learning_rate": 2.5714285714285718e-05, |
| "loss": 0.6721, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.10825439783491204, |
| "grad_norm": 2.1232327521684957, |
| "learning_rate": 2.8571428571428574e-05, |
| "loss": 0.6659, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.11907983761840325, |
| "grad_norm": 1.5541574165991885, |
| "learning_rate": 3.142857142857143e-05, |
| "loss": 0.6379, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.12990527740189445, |
| "grad_norm": 1.3622424203539478, |
| "learning_rate": 3.4285714285714284e-05, |
| "loss": 0.6302, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.14073071718538566, |
| "grad_norm": 1.5695142530632886, |
| "learning_rate": 3.714285714285715e-05, |
| "loss": 0.6141, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.15155615696887687, |
| "grad_norm": 1.203554594979693, |
| "learning_rate": 4e-05, |
| "loss": 0.6243, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.16238159675236807, |
| "grad_norm": 1.4054219874433034, |
| "learning_rate": 4.2857142857142856e-05, |
| "loss": 0.6073, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.17320703653585928, |
| "grad_norm": 1.200808072103448, |
| "learning_rate": 4.5714285714285716e-05, |
| "loss": 0.5995, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.18403247631935046, |
| "grad_norm": 1.7138206517858583, |
| "learning_rate": 4.857142857142857e-05, |
| "loss": 0.5943, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.19485791610284167, |
| "grad_norm": 1.135374128767807, |
| "learning_rate": 5.1428571428571436e-05, |
| "loss": 0.577, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.20568335588633288, |
| "grad_norm": 1.5358674471797358, |
| "learning_rate": 5.4285714285714295e-05, |
| "loss": 0.5696, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.2165087956698241, |
| "grad_norm": 1.457820215468553, |
| "learning_rate": 5.714285714285715e-05, |
| "loss": 0.5736, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.2273342354533153, |
| "grad_norm": 1.836451551117814, |
| "learning_rate": 6.000000000000001e-05, |
| "loss": 0.57, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.2381596752368065, |
| "grad_norm": 1.2877464793886966, |
| "learning_rate": 6.285714285714286e-05, |
| "loss": 0.5679, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.2489851150202977, |
| "grad_norm": 2.1045382011992273, |
| "learning_rate": 6.571428571428571e-05, |
| "loss": 0.5641, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.2598105548037889, |
| "grad_norm": 1.6437806308258178, |
| "learning_rate": 6.857142857142857e-05, |
| "loss": 0.5693, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.2706359945872801, |
| "grad_norm": 2.107089475881658, |
| "learning_rate": 7.142857142857143e-05, |
| "loss": 0.5614, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.2814614343707713, |
| "grad_norm": 1.7285732683115125, |
| "learning_rate": 7.42857142857143e-05, |
| "loss": 0.5415, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.2922868741542625, |
| "grad_norm": 1.5027616655977198, |
| "learning_rate": 7.714285714285715e-05, |
| "loss": 0.5441, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.30311231393775373, |
| "grad_norm": 1.2345678257838402, |
| "learning_rate": 8e-05, |
| "loss": 0.5507, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.31393775372124494, |
| "grad_norm": 2.149518135585814, |
| "learning_rate": 7.999679062421315e-05, |
| "loss": 0.5553, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.32476319350473615, |
| "grad_norm": 1.664754613353701, |
| "learning_rate": 7.998716301185722e-05, |
| "loss": 0.5428, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.33558863328822736, |
| "grad_norm": 1.6950629811866913, |
| "learning_rate": 7.997111870786354e-05, |
| "loss": 0.5483, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.34641407307171856, |
| "grad_norm": 1.3054323430772967, |
| "learning_rate": 7.994866028684212e-05, |
| "loss": 0.5382, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.3572395128552097, |
| "grad_norm": 1.5518136444133512, |
| "learning_rate": 7.991979135266861e-05, |
| "loss": 0.5401, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.3680649526387009, |
| "grad_norm": 1.852237456765202, |
| "learning_rate": 7.98845165379059e-05, |
| "loss": 0.5439, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.37889039242219213, |
| "grad_norm": 1.0576494308389435, |
| "learning_rate": 7.984284150306085e-05, |
| "loss": 0.5242, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.38971583220568334, |
| "grad_norm": 2.3304856182379767, |
| "learning_rate": 7.97947729356758e-05, |
| "loss": 0.5224, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.40054127198917455, |
| "grad_norm": 1.3906984159228295, |
| "learning_rate": 7.974031854925562e-05, |
| "loss": 0.5338, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.41136671177266576, |
| "grad_norm": 2.6841681286686248, |
| "learning_rate": 7.967948708202972e-05, |
| "loss": 0.5236, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.42219215155615697, |
| "grad_norm": 2.1338153954371606, |
| "learning_rate": 7.961228829555003e-05, |
| "loss": 0.5387, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.4330175913396482, |
| "grad_norm": 1.8619373516339954, |
| "learning_rate": 7.953873297312447e-05, |
| "loss": 0.5227, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.4438430311231394, |
| "grad_norm": 1.548983471469427, |
| "learning_rate": 7.945883291808655e-05, |
| "loss": 0.5192, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.4546684709066306, |
| "grad_norm": 1.4352471105396338, |
| "learning_rate": 7.937260095190137e-05, |
| "loss": 0.5137, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.4654939106901218, |
| "grad_norm": 1.1595063012586966, |
| "learning_rate": 7.928005091210817e-05, |
| "loss": 0.51, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.476319350473613, |
| "grad_norm": 1.4726656876357218, |
| "learning_rate": 7.918119765009979e-05, |
| "loss": 0.5121, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.4871447902571042, |
| "grad_norm": 1.487100581482714, |
| "learning_rate": 7.907605702873948e-05, |
| "loss": 0.5167, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.4979702300405954, |
| "grad_norm": 1.0124772477505708, |
| "learning_rate": 7.896464591981549e-05, |
| "loss": 0.5081, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.5087956698240866, |
| "grad_norm": 0.9076344394213861, |
| "learning_rate": 7.884698220133357e-05, |
| "loss": 0.507, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.5196211096075778, |
| "grad_norm": 0.948018821557288, |
| "learning_rate": 7.872308475464818e-05, |
| "loss": 0.5088, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.530446549391069, |
| "grad_norm": 2.021248061293733, |
| "learning_rate": 7.859297346143258e-05, |
| "loss": 0.5186, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.5412719891745602, |
| "grad_norm": 1.294430989189167, |
| "learning_rate": 7.84566692004885e-05, |
| "loss": 0.5073, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.5520974289580515, |
| "grad_norm": 1.3483179710969726, |
| "learning_rate": 7.831419384439565e-05, |
| "loss": 0.5044, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.5629228687415426, |
| "grad_norm": 1.5375274937154497, |
| "learning_rate": 7.816557025600196e-05, |
| "loss": 0.4979, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.5737483085250338, |
| "grad_norm": 1.138539066191574, |
| "learning_rate": 7.80108222847547e-05, |
| "loss": 0.4982, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.584573748308525, |
| "grad_norm": 0.9455835571587606, |
| "learning_rate": 7.784997476287349e-05, |
| "loss": 0.5068, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.5953991880920162, |
| "grad_norm": 2.1391708993694123, |
| "learning_rate": 7.76830535013654e-05, |
| "loss": 0.5015, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.6062246278755075, |
| "grad_norm": 0.9963202038399334, |
| "learning_rate": 7.751008528588322e-05, |
| "loss": 0.5049, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.6170500676589986, |
| "grad_norm": 2.6002608947302868, |
| "learning_rate": 7.733109787242708e-05, |
| "loss": 0.5159, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.6278755074424899, |
| "grad_norm": 1.9341514663010684, |
| "learning_rate": 7.71461199828905e-05, |
| "loss": 0.5205, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.638700947225981, |
| "grad_norm": 2.105857537142511, |
| "learning_rate": 7.695518130045147e-05, |
| "loss": 0.4968, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.6495263870094723, |
| "grad_norm": 2.0226793353731045, |
| "learning_rate": 7.675831246480923e-05, |
| "loss": 0.5059, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.6603518267929634, |
| "grad_norm": 1.7011609522379212, |
| "learning_rate": 7.655554506726747e-05, |
| "loss": 0.5164, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.6711772665764547, |
| "grad_norm": 1.5029227881331557, |
| "learning_rate": 7.6346911645665e-05, |
| "loss": 0.4984, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.6820027063599459, |
| "grad_norm": 1.3409493889398278, |
| "learning_rate": 7.61324456791544e-05, |
| "loss": 0.5061, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.6928281461434371, |
| "grad_norm": 1.4193456880899875, |
| "learning_rate": 7.591218158282968e-05, |
| "loss": 0.5009, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.7036535859269283, |
| "grad_norm": 0.8915479657327203, |
| "learning_rate": 7.568615470220369e-05, |
| "loss": 0.4897, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.7144790257104194, |
| "grad_norm": 1.6232415324265208, |
| "learning_rate": 7.545440130753634e-05, |
| "loss": 0.5005, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.7253044654939107, |
| "grad_norm": 0.9447005484764661, |
| "learning_rate": 7.52169585880143e-05, |
| "loss": 0.4972, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.7361299052774019, |
| "grad_norm": 1.377203093352462, |
| "learning_rate": 7.497386464578329e-05, |
| "loss": 0.5065, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.7469553450608931, |
| "grad_norm": 1.3669620434145435, |
| "learning_rate": 7.472515848983394e-05, |
| "loss": 0.4997, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.7577807848443843, |
| "grad_norm": 0.8734490608081458, |
| "learning_rate": 7.447088002974199e-05, |
| "loss": 0.4792, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.7686062246278755, |
| "grad_norm": 1.3481324332197542, |
| "learning_rate": 7.421107006926408e-05, |
| "loss": 0.4832, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.7794316644113667, |
| "grad_norm": 0.7943982329371713, |
| "learning_rate": 7.394577029979004e-05, |
| "loss": 0.4936, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.790257104194858, |
| "grad_norm": 1.134585813685754, |
| "learning_rate": 7.367502329365268e-05, |
| "loss": 0.5008, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.8010825439783491, |
| "grad_norm": 1.1444689248464068, |
| "learning_rate": 7.33988724972963e-05, |
| "loss": 0.4998, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.8119079837618404, |
| "grad_norm": 1.570355010614296, |
| "learning_rate": 7.311736222430487e-05, |
| "loss": 0.4969, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.8227334235453315, |
| "grad_norm": 0.7307027465767969, |
| "learning_rate": 7.283053764829106e-05, |
| "loss": 0.4834, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.8335588633288228, |
| "grad_norm": 1.5156782350119813, |
| "learning_rate": 7.253844479564737e-05, |
| "loss": 0.4887, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.8443843031123139, |
| "grad_norm": 0.9627261310112593, |
| "learning_rate": 7.224113053816021e-05, |
| "loss": 0.4871, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.8552097428958051, |
| "grad_norm": 1.7769819863059997, |
| "learning_rate": 7.193864258548855e-05, |
| "loss": 0.4956, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.8660351826792964, |
| "grad_norm": 1.3057566887617058, |
| "learning_rate": 7.163102947750794e-05, |
| "loss": 0.4904, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.8768606224627875, |
| "grad_norm": 1.55736756271743, |
| "learning_rate": 7.131834057652142e-05, |
| "loss": 0.4968, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.8876860622462788, |
| "grad_norm": 1.3726270786588717, |
| "learning_rate": 7.100062605933835e-05, |
| "loss": 0.4841, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.8985115020297699, |
| "grad_norm": 1.2631231071305447, |
| "learning_rate": 7.067793690922268e-05, |
| "loss": 0.4875, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.9093369418132612, |
| "grad_norm": 1.4697742131141704, |
| "learning_rate": 7.035032490771165e-05, |
| "loss": 0.4857, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.9201623815967523, |
| "grad_norm": 0.9130895779617919, |
| "learning_rate": 7.001784262630652e-05, |
| "loss": 0.4822, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.9309878213802436, |
| "grad_norm": 1.3805866452426423, |
| "learning_rate": 6.968054341803644e-05, |
| "loss": 0.4846, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.9418132611637348, |
| "grad_norm": 0.8781175020763933, |
| "learning_rate": 6.933848140889705e-05, |
| "loss": 0.4818, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.952638700947226, |
| "grad_norm": 1.5493397081142941, |
| "learning_rate": 6.89917114891648e-05, |
| "loss": 0.4666, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.9634641407307172, |
| "grad_norm": 1.0670620246568012, |
| "learning_rate": 6.864028930458892e-05, |
| "loss": 0.4838, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.9742895805142084, |
| "grad_norm": 1.7422601069891361, |
| "learning_rate": 6.828427124746191e-05, |
| "loss": 0.4806, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.9851150202976996, |
| "grad_norm": 1.4949185468809587, |
| "learning_rate": 6.792371444757037e-05, |
| "loss": 0.4833, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.9959404600811907, |
| "grad_norm": 1.3684573222022662, |
| "learning_rate": 6.755867676302747e-05, |
| "loss": 0.4831, |
| "step": 92 |
| }, |
| { |
| "epoch": 1.006765899864682, |
| "grad_norm": 1.594260543224364, |
| "learning_rate": 6.718921677098853e-05, |
| "loss": 0.7498, |
| "step": 93 |
| }, |
| { |
| "epoch": 1.0175913396481733, |
| "grad_norm": 1.1919567789012502, |
| "learning_rate": 6.681539375825115e-05, |
| "loss": 0.4597, |
| "step": 94 |
| }, |
| { |
| "epoch": 1.0284167794316643, |
| "grad_norm": 0.9843773524280992, |
| "learning_rate": 6.643726771174164e-05, |
| "loss": 0.4578, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.0392422192151556, |
| "grad_norm": 1.8047666421292219, |
| "learning_rate": 6.60548993088889e-05, |
| "loss": 0.4595, |
| "step": 96 |
| }, |
| { |
| "epoch": 1.0500676589986468, |
| "grad_norm": 1.150709829354635, |
| "learning_rate": 6.56683499078876e-05, |
| "loss": 0.4487, |
| "step": 97 |
| }, |
| { |
| "epoch": 1.060893098782138, |
| "grad_norm": 1.6015028118294863, |
| "learning_rate": 6.527768153785216e-05, |
| "loss": 0.458, |
| "step": 98 |
| }, |
| { |
| "epoch": 1.0717185385656292, |
| "grad_norm": 1.4811422857284644, |
| "learning_rate": 6.488295688886295e-05, |
| "loss": 0.4566, |
| "step": 99 |
| }, |
| { |
| "epoch": 1.0825439783491204, |
| "grad_norm": 0.7699742468198911, |
| "learning_rate": 6.448423930190653e-05, |
| "loss": 0.4493, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.0933694181326117, |
| "grad_norm": 1.5306435492691353, |
| "learning_rate": 6.408159275871132e-05, |
| "loss": 0.4531, |
| "step": 101 |
| }, |
| { |
| "epoch": 1.104194857916103, |
| "grad_norm": 0.9802106555752499, |
| "learning_rate": 6.36750818714807e-05, |
| "loss": 0.4498, |
| "step": 102 |
| }, |
| { |
| "epoch": 1.115020297699594, |
| "grad_norm": 1.0366612460408087, |
| "learning_rate": 6.326477187252455e-05, |
| "loss": 0.4536, |
| "step": 103 |
| }, |
| { |
| "epoch": 1.1258457374830853, |
| "grad_norm": 0.7928329971928261, |
| "learning_rate": 6.28507286037917e-05, |
| "loss": 0.4471, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.1366711772665765, |
| "grad_norm": 1.1009775639540247, |
| "learning_rate": 6.243301850630419e-05, |
| "loss": 0.4529, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.1474966170500678, |
| "grad_norm": 1.2338692549823334, |
| "learning_rate": 6.201170860949565e-05, |
| "loss": 0.4577, |
| "step": 106 |
| }, |
| { |
| "epoch": 1.1583220568335588, |
| "grad_norm": 0.921356197831234, |
| "learning_rate": 6.15868665204552e-05, |
| "loss": 0.4503, |
| "step": 107 |
| }, |
| { |
| "epoch": 1.16914749661705, |
| "grad_norm": 0.7945396358653812, |
| "learning_rate": 6.11585604130785e-05, |
| "loss": 0.4438, |
| "step": 108 |
| }, |
| { |
| "epoch": 1.1799729364005414, |
| "grad_norm": 0.8214761344740851, |
| "learning_rate": 6.072685901712808e-05, |
| "loss": 0.4496, |
| "step": 109 |
| }, |
| { |
| "epoch": 1.1907983761840324, |
| "grad_norm": 0.7869742134352299, |
| "learning_rate": 6.02918316072043e-05, |
| "loss": 0.4547, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.2016238159675237, |
| "grad_norm": 0.8177171451163405, |
| "learning_rate": 5.9853547991628967e-05, |
| "loss": 0.4535, |
| "step": 111 |
| }, |
| { |
| "epoch": 1.212449255751015, |
| "grad_norm": 0.8293137987156753, |
| "learning_rate": 5.941207850124325e-05, |
| "loss": 0.4469, |
| "step": 112 |
| }, |
| { |
| "epoch": 1.2232746955345062, |
| "grad_norm": 0.6243958292154561, |
| "learning_rate": 5.896749397812181e-05, |
| "loss": 0.442, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.2341001353179972, |
| "grad_norm": 0.7692802100906287, |
| "learning_rate": 5.8519865764204834e-05, |
| "loss": 0.4505, |
| "step": 114 |
| }, |
| { |
| "epoch": 1.2449255751014885, |
| "grad_norm": 0.5627604101482472, |
| "learning_rate": 5.8069265689849884e-05, |
| "loss": 0.4555, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.2557510148849798, |
| "grad_norm": 0.6076684710899567, |
| "learning_rate": 5.761576606230538e-05, |
| "loss": 0.4474, |
| "step": 116 |
| }, |
| { |
| "epoch": 1.266576454668471, |
| "grad_norm": 0.6176591408605276, |
| "learning_rate": 5.7159439654107506e-05, |
| "loss": 0.4447, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.277401894451962, |
| "grad_norm": 0.6151330849946537, |
| "learning_rate": 5.6700359691402533e-05, |
| "loss": 0.444, |
| "step": 118 |
| }, |
| { |
| "epoch": 1.2882273342354533, |
| "grad_norm": 0.5945893866102415, |
| "learning_rate": 5.6238599842196285e-05, |
| "loss": 0.4486, |
| "step": 119 |
| }, |
| { |
| "epoch": 1.2990527740189446, |
| "grad_norm": 0.5719852552849911, |
| "learning_rate": 5.5774234204532746e-05, |
| "loss": 0.4399, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.3098782138024356, |
| "grad_norm": 0.5625173050589614, |
| "learning_rate": 5.5307337294603595e-05, |
| "loss": 0.4471, |
| "step": 121 |
| }, |
| { |
| "epoch": 1.320703653585927, |
| "grad_norm": 0.41901601585146925, |
| "learning_rate": 5.483798403479072e-05, |
| "loss": 0.4428, |
| "step": 122 |
| }, |
| { |
| "epoch": 1.3315290933694182, |
| "grad_norm": 0.5542457308518791, |
| "learning_rate": 5.436624974164349e-05, |
| "loss": 0.4454, |
| "step": 123 |
| }, |
| { |
| "epoch": 1.3423545331529092, |
| "grad_norm": 0.5622447138663931, |
| "learning_rate": 5.389221011379281e-05, |
| "loss": 0.4429, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.3531799729364005, |
| "grad_norm": 0.39891990504185004, |
| "learning_rate": 5.3415941219803895e-05, |
| "loss": 0.4438, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.3640054127198917, |
| "grad_norm": 0.46251409626010687, |
| "learning_rate": 5.2937519485969525e-05, |
| "loss": 0.4425, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.374830852503383, |
| "grad_norm": 0.41442268928594483, |
| "learning_rate": 5.245702168404616e-05, |
| "loss": 0.4354, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.3856562922868743, |
| "grad_norm": 0.46080666475487947, |
| "learning_rate": 5.1974524918934336e-05, |
| "loss": 0.4382, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.3964817320703653, |
| "grad_norm": 0.40205953759789465, |
| "learning_rate": 5.14901066163058e-05, |
| "loss": 0.4368, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.4073071718538566, |
| "grad_norm": 0.5302628542951422, |
| "learning_rate": 5.1003844510179126e-05, |
| "loss": 0.4421, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.4181326116373478, |
| "grad_norm": 0.5643727896360857, |
| "learning_rate": 5.0515816630445795e-05, |
| "loss": 0.4432, |
| "step": 131 |
| }, |
| { |
| "epoch": 1.4289580514208389, |
| "grad_norm": 0.567898277228873, |
| "learning_rate": 5.002610129034883e-05, |
| "loss": 0.4452, |
| "step": 132 |
| }, |
| { |
| "epoch": 1.4397834912043301, |
| "grad_norm": 0.4879893942135545, |
| "learning_rate": 4.953477707391597e-05, |
| "loss": 0.4387, |
| "step": 133 |
| }, |
| { |
| "epoch": 1.4506089309878214, |
| "grad_norm": 0.44580188517360675, |
| "learning_rate": 4.90419228233494e-05, |
| "loss": 0.4402, |
| "step": 134 |
| }, |
| { |
| "epoch": 1.4614343707713127, |
| "grad_norm": 0.34312001042074763, |
| "learning_rate": 4.854761762637403e-05, |
| "loss": 0.4403, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.472259810554804, |
| "grad_norm": 0.27073851961143425, |
| "learning_rate": 4.805194080354641e-05, |
| "loss": 0.4392, |
| "step": 136 |
| }, |
| { |
| "epoch": 1.483085250338295, |
| "grad_norm": 0.3831138691864976, |
| "learning_rate": 4.7554971895526175e-05, |
| "loss": 0.4364, |
| "step": 137 |
| }, |
| { |
| "epoch": 1.4939106901217862, |
| "grad_norm": 0.35968897365441027, |
| "learning_rate": 4.705679065031235e-05, |
| "loss": 0.4425, |
| "step": 138 |
| }, |
| { |
| "epoch": 1.5047361299052775, |
| "grad_norm": 0.3921891739790258, |
| "learning_rate": 4.6557477010446206e-05, |
| "loss": 0.4418, |
| "step": 139 |
| }, |
| { |
| "epoch": 1.5155615696887685, |
| "grad_norm": 0.32519376313428694, |
| "learning_rate": 4.605711110018307e-05, |
| "loss": 0.4409, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.5263870094722598, |
| "grad_norm": 0.2889746876592647, |
| "learning_rate": 4.555577321263477e-05, |
| "loss": 0.4369, |
| "step": 141 |
| }, |
| { |
| "epoch": 1.537212449255751, |
| "grad_norm": 0.34060322005606364, |
| "learning_rate": 4.505354379688518e-05, |
| "loss": 0.4379, |
| "step": 142 |
| }, |
| { |
| "epoch": 1.548037889039242, |
| "grad_norm": 0.3544782098609683, |
| "learning_rate": 4.4550503445080606e-05, |
| "loss": 0.4441, |
| "step": 143 |
| }, |
| { |
| "epoch": 1.5588633288227334, |
| "grad_norm": 0.32503145323454175, |
| "learning_rate": 4.4046732879497295e-05, |
| "loss": 0.4361, |
| "step": 144 |
| }, |
| { |
| "epoch": 1.5696887686062246, |
| "grad_norm": 0.33433760347689245, |
| "learning_rate": 4.354231293958801e-05, |
| "loss": 0.4342, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.5805142083897157, |
| "grad_norm": 0.27822404907197074, |
| "learning_rate": 4.3037324569009854e-05, |
| "loss": 0.4341, |
| "step": 146 |
| }, |
| { |
| "epoch": 1.5913396481732072, |
| "grad_norm": 0.21433947236043074, |
| "learning_rate": 4.2531848802635264e-05, |
| "loss": 0.4305, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.6021650879566982, |
| "grad_norm": 0.2693726552613991, |
| "learning_rate": 4.202596675354851e-05, |
| "loss": 0.4505, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.6129905277401895, |
| "grad_norm": 0.31839490626531153, |
| "learning_rate": 4.151975960002958e-05, |
| "loss": 0.4329, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.6238159675236807, |
| "grad_norm": 0.29869132543027865, |
| "learning_rate": 4.101330857252752e-05, |
| "loss": 0.439, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.6346414073071718, |
| "grad_norm": 0.30548534307447794, |
| "learning_rate": 4.050669494062561e-05, |
| "loss": 0.4395, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.645466847090663, |
| "grad_norm": 0.22533097566731655, |
| "learning_rate": 4e-05, |
| "loss": 0.433, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.6562922868741543, |
| "grad_norm": 0.2857809612843107, |
| "learning_rate": 3.9493305059374405e-05, |
| "loss": 0.4461, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.6671177266576453, |
| "grad_norm": 0.2969380754852739, |
| "learning_rate": 3.8986691427472496e-05, |
| "loss": 0.4387, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.6779431664411368, |
| "grad_norm": 0.26858616293926957, |
| "learning_rate": 3.8480240399970436e-05, |
| "loss": 0.4315, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.6887686062246279, |
| "grad_norm": 0.22376872247908408, |
| "learning_rate": 3.7974033246451496e-05, |
| "loss": 0.4399, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.699594046008119, |
| "grad_norm": 0.2147935202073495, |
| "learning_rate": 3.746815119736475e-05, |
| "loss": 0.4258, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.7104194857916104, |
| "grad_norm": 0.21407307646532672, |
| "learning_rate": 3.696267543099016e-05, |
| "loss": 0.4369, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.7212449255751014, |
| "grad_norm": 0.22040203612960246, |
| "learning_rate": 3.6457687060412e-05, |
| "loss": 0.4437, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.7320703653585927, |
| "grad_norm": 0.19541670594304736, |
| "learning_rate": 3.595326712050272e-05, |
| "loss": 0.4375, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.742895805142084, |
| "grad_norm": 0.2638985177983092, |
| "learning_rate": 3.5449496554919414e-05, |
| "loss": 0.434, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.753721244925575, |
| "grad_norm": 0.1936379675806606, |
| "learning_rate": 3.494645620311484e-05, |
| "loss": 0.4399, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.7645466847090663, |
| "grad_norm": 0.23151389721476723, |
| "learning_rate": 3.444422678736525e-05, |
| "loss": 0.4334, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.7753721244925575, |
| "grad_norm": 0.18589997021167423, |
| "learning_rate": 3.394288889981695e-05, |
| "loss": 0.431, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.7861975642760486, |
| "grad_norm": 0.19699335160697973, |
| "learning_rate": 3.34425229895538e-05, |
| "loss": 0.438, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.79702300405954, |
| "grad_norm": 0.24249716814306238, |
| "learning_rate": 3.294320934968768e-05, |
| "loss": 0.431, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.8078484438430311, |
| "grad_norm": 0.20087415994991933, |
| "learning_rate": 3.2445028104473845e-05, |
| "loss": 0.4365, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.8186738836265224, |
| "grad_norm": 0.22587462647538303, |
| "learning_rate": 3.194805919645359e-05, |
| "loss": 0.4331, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.8294993234100136, |
| "grad_norm": 0.18648214190719595, |
| "learning_rate": 3.145238237362596e-05, |
| "loss": 0.4409, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.8403247631935047, |
| "grad_norm": 0.1837847468858913, |
| "learning_rate": 3.0958077176650606e-05, |
| "loss": 0.4323, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.851150202976996, |
| "grad_norm": 0.16790249180327427, |
| "learning_rate": 3.0465222926084036e-05, |
| "loss": 0.4286, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.8619756427604872, |
| "grad_norm": 0.18791171767972534, |
| "learning_rate": 2.997389870965118e-05, |
| "loss": 0.4294, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.8728010825439783, |
| "grad_norm": 0.20052428698047692, |
| "learning_rate": 2.948418336955421e-05, |
| "loss": 0.4236, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.8836265223274695, |
| "grad_norm": 0.21006309241348184, |
| "learning_rate": 2.899615548982088e-05, |
| "loss": 0.4362, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.8944519621109608, |
| "grad_norm": 0.17507475038657377, |
| "learning_rate": 2.8509893383694213e-05, |
| "loss": 0.43, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.9052774018944518, |
| "grad_norm": 0.20449273058242617, |
| "learning_rate": 2.8025475081065684e-05, |
| "loss": 0.425, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.9161028416779433, |
| "grad_norm": 0.20025584198847277, |
| "learning_rate": 2.754297831595385e-05, |
| "loss": 0.4263, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.9269282814614344, |
| "grad_norm": 0.19675907045215793, |
| "learning_rate": 2.7062480514030478e-05, |
| "loss": 0.4367, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.9377537212449256, |
| "grad_norm": 0.19647692180533832, |
| "learning_rate": 2.658405878019612e-05, |
| "loss": 0.4307, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.9485791610284169, |
| "grad_norm": 0.17640852809639837, |
| "learning_rate": 2.6107789886207195e-05, |
| "loss": 0.4332, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.959404600811908, |
| "grad_norm": 0.16434805778304712, |
| "learning_rate": 2.563375025835652e-05, |
| "loss": 0.4316, |
| "step": 181 |
| }, |
| { |
| "epoch": 1.9702300405953992, |
| "grad_norm": 0.1622795215969799, |
| "learning_rate": 2.5162015965209295e-05, |
| "loss": 0.4286, |
| "step": 182 |
| }, |
| { |
| "epoch": 1.9810554803788905, |
| "grad_norm": 0.18207876344495097, |
| "learning_rate": 2.4692662705396412e-05, |
| "loss": 0.4358, |
| "step": 183 |
| }, |
| { |
| "epoch": 1.9918809201623815, |
| "grad_norm": 0.15940281960579916, |
| "learning_rate": 2.4225765795467267e-05, |
| "loss": 0.4362, |
| "step": 184 |
| }, |
| { |
| "epoch": 2.002706359945873, |
| "grad_norm": 0.30510551239734873, |
| "learning_rate": 2.376140015780372e-05, |
| "loss": 0.6859, |
| "step": 185 |
| }, |
| { |
| "epoch": 2.013531799729364, |
| "grad_norm": 0.2889608940417901, |
| "learning_rate": 2.3299640308597487e-05, |
| "loss": 0.4062, |
| "step": 186 |
| }, |
| { |
| "epoch": 2.024357239512855, |
| "grad_norm": 0.18715978836716335, |
| "learning_rate": 2.2840560345892518e-05, |
| "loss": 0.405, |
| "step": 187 |
| }, |
| { |
| "epoch": 2.0351826792963466, |
| "grad_norm": 0.26618002533258645, |
| "learning_rate": 2.2384233937694626e-05, |
| "loss": 0.4055, |
| "step": 188 |
| }, |
| { |
| "epoch": 2.0460081190798376, |
| "grad_norm": 0.24432220208503594, |
| "learning_rate": 2.1930734310150116e-05, |
| "loss": 0.3998, |
| "step": 189 |
| }, |
| { |
| "epoch": 2.0568335588633286, |
| "grad_norm": 0.2438307934593196, |
| "learning_rate": 2.1480134235795173e-05, |
| "loss": 0.4043, |
| "step": 190 |
| }, |
| { |
| "epoch": 2.06765899864682, |
| "grad_norm": 0.21418084374643698, |
| "learning_rate": 2.10325060218782e-05, |
| "loss": 0.4092, |
| "step": 191 |
| }, |
| { |
| "epoch": 2.078484438430311, |
| "grad_norm": 0.23406949090678664, |
| "learning_rate": 2.0587921498756768e-05, |
| "loss": 0.4128, |
| "step": 192 |
| }, |
| { |
| "epoch": 2.089309878213802, |
| "grad_norm": 0.23241158373316004, |
| "learning_rate": 2.014645200837105e-05, |
| "loss": 0.4059, |
| "step": 193 |
| }, |
| { |
| "epoch": 2.1001353179972937, |
| "grad_norm": 0.21329568927152176, |
| "learning_rate": 1.9708168392795718e-05, |
| "loss": 0.3973, |
| "step": 194 |
| }, |
| { |
| "epoch": 2.1109607577807847, |
| "grad_norm": 0.21956497402247424, |
| "learning_rate": 1.9273140982871936e-05, |
| "loss": 0.4068, |
| "step": 195 |
| }, |
| { |
| "epoch": 2.121786197564276, |
| "grad_norm": 0.24174655161577632, |
| "learning_rate": 1.8841439586921515e-05, |
| "loss": 0.4055, |
| "step": 196 |
| }, |
| { |
| "epoch": 2.1326116373477673, |
| "grad_norm": 0.2226235613223951, |
| "learning_rate": 1.841313347954482e-05, |
| "loss": 0.4008, |
| "step": 197 |
| }, |
| { |
| "epoch": 2.1434370771312583, |
| "grad_norm": 0.20041392895663887, |
| "learning_rate": 1.7988291390504348e-05, |
| "loss": 0.4, |
| "step": 198 |
| }, |
| { |
| "epoch": 2.15426251691475, |
| "grad_norm": 0.17447148351618208, |
| "learning_rate": 1.7566981493695828e-05, |
| "loss": 0.4068, |
| "step": 199 |
| }, |
| { |
| "epoch": 2.165087956698241, |
| "grad_norm": 0.17488318983541223, |
| "learning_rate": 1.71492713962083e-05, |
| "loss": 0.4088, |
| "step": 200 |
| }, |
| { |
| "epoch": 2.175913396481732, |
| "grad_norm": 0.1663846156373835, |
| "learning_rate": 1.673522812747544e-05, |
| "loss": 0.4045, |
| "step": 201 |
| }, |
| { |
| "epoch": 2.1867388362652234, |
| "grad_norm": 0.13359463787477008, |
| "learning_rate": 1.6324918128519306e-05, |
| "loss": 0.4017, |
| "step": 202 |
| }, |
| { |
| "epoch": 2.1975642760487144, |
| "grad_norm": 0.19122503362612153, |
| "learning_rate": 1.5918407241288678e-05, |
| "loss": 0.4082, |
| "step": 203 |
| }, |
| { |
| "epoch": 2.208389715832206, |
| "grad_norm": 0.14102405846996677, |
| "learning_rate": 1.5515760698093485e-05, |
| "loss": 0.4066, |
| "step": 204 |
| }, |
| { |
| "epoch": 2.219215155615697, |
| "grad_norm": 0.152536968570073, |
| "learning_rate": 1.511704311113705e-05, |
| "loss": 0.4044, |
| "step": 205 |
| }, |
| { |
| "epoch": 2.230040595399188, |
| "grad_norm": 0.14601541059029532, |
| "learning_rate": 1.4722318462147844e-05, |
| "loss": 0.4043, |
| "step": 206 |
| }, |
| { |
| "epoch": 2.2408660351826795, |
| "grad_norm": 0.14846049098921832, |
| "learning_rate": 1.4331650092112406e-05, |
| "loss": 0.4092, |
| "step": 207 |
| }, |
| { |
| "epoch": 2.2516914749661705, |
| "grad_norm": 0.15154485636442028, |
| "learning_rate": 1.394510069111112e-05, |
| "loss": 0.3992, |
| "step": 208 |
| }, |
| { |
| "epoch": 2.2625169147496615, |
| "grad_norm": 0.12922166696259543, |
| "learning_rate": 1.3562732288258377e-05, |
| "loss": 0.4052, |
| "step": 209 |
| }, |
| { |
| "epoch": 2.273342354533153, |
| "grad_norm": 0.13187880371612812, |
| "learning_rate": 1.3184606241748857e-05, |
| "loss": 0.4002, |
| "step": 210 |
| }, |
| { |
| "epoch": 2.284167794316644, |
| "grad_norm": 0.14862969647085414, |
| "learning_rate": 1.2810783229011486e-05, |
| "loss": 0.4013, |
| "step": 211 |
| }, |
| { |
| "epoch": 2.2949932341001356, |
| "grad_norm": 0.14118764670077505, |
| "learning_rate": 1.2441323236972536e-05, |
| "loss": 0.402, |
| "step": 212 |
| }, |
| { |
| "epoch": 2.3058186738836266, |
| "grad_norm": 0.12840893339773712, |
| "learning_rate": 1.2076285552429642e-05, |
| "loss": 0.4094, |
| "step": 213 |
| }, |
| { |
| "epoch": 2.3166441136671176, |
| "grad_norm": 0.12817010450870242, |
| "learning_rate": 1.1715728752538103e-05, |
| "loss": 0.4039, |
| "step": 214 |
| }, |
| { |
| "epoch": 2.3274695534506087, |
| "grad_norm": 0.13249387545369556, |
| "learning_rate": 1.1359710695411086e-05, |
| "loss": 0.4049, |
| "step": 215 |
| }, |
| { |
| "epoch": 2.3382949932341, |
| "grad_norm": 0.12426069210681286, |
| "learning_rate": 1.100828851083521e-05, |
| "loss": 0.4103, |
| "step": 216 |
| }, |
| { |
| "epoch": 2.349120433017591, |
| "grad_norm": 0.12731813593787142, |
| "learning_rate": 1.0661518591102973e-05, |
| "loss": 0.4042, |
| "step": 217 |
| }, |
| { |
| "epoch": 2.3599458728010827, |
| "grad_norm": 0.1279464788506675, |
| "learning_rate": 1.0319456581963578e-05, |
| "loss": 0.4119, |
| "step": 218 |
| }, |
| { |
| "epoch": 2.3707713125845737, |
| "grad_norm": 0.12186728200221489, |
| "learning_rate": 9.982157373693502e-06, |
| "loss": 0.4091, |
| "step": 219 |
| }, |
| { |
| "epoch": 2.381596752368065, |
| "grad_norm": 0.13287274650519393, |
| "learning_rate": 9.649675092288366e-06, |
| "loss": 0.3958, |
| "step": 220 |
| }, |
| { |
| "epoch": 2.3924221921515563, |
| "grad_norm": 0.12611811960255165, |
| "learning_rate": 9.322063090777331e-06, |
| "loss": 0.4042, |
| "step": 221 |
| }, |
| { |
| "epoch": 2.4032476319350473, |
| "grad_norm": 0.12692566385279155, |
| "learning_rate": 8.99937394066165e-06, |
| "loss": 0.4101, |
| "step": 222 |
| }, |
| { |
| "epoch": 2.414073071718539, |
| "grad_norm": 0.1368999128690929, |
| "learning_rate": 8.681659423478587e-06, |
| "loss": 0.4098, |
| "step": 223 |
| }, |
| { |
| "epoch": 2.42489851150203, |
| "grad_norm": 0.1338830307854189, |
| "learning_rate": 8.368970522492064e-06, |
| "loss": 0.406, |
| "step": 224 |
| }, |
| { |
| "epoch": 2.435723951285521, |
| "grad_norm": 0.11649450453634784, |
| "learning_rate": 8.06135741451146e-06, |
| "loss": 0.4047, |
| "step": 225 |
| }, |
| { |
| "epoch": 2.4465493910690124, |
| "grad_norm": 0.13539242030434118, |
| "learning_rate": 7.758869461839808e-06, |
| "loss": 0.4073, |
| "step": 226 |
| }, |
| { |
| "epoch": 2.4573748308525034, |
| "grad_norm": 0.11657804263433023, |
| "learning_rate": 7.461555204352655e-06, |
| "loss": 0.4008, |
| "step": 227 |
| }, |
| { |
| "epoch": 2.4682002706359945, |
| "grad_norm": 0.13670292033952086, |
| "learning_rate": 7.169462351708958e-06, |
| "loss": 0.403, |
| "step": 228 |
| }, |
| { |
| "epoch": 2.479025710419486, |
| "grad_norm": 0.1082875144964409, |
| "learning_rate": 6.882637775695147e-06, |
| "loss": 0.3992, |
| "step": 229 |
| }, |
| { |
| "epoch": 2.489851150202977, |
| "grad_norm": 0.11137336492715712, |
| "learning_rate": 6.60112750270371e-06, |
| "loss": 0.4003, |
| "step": 230 |
| }, |
| { |
| "epoch": 2.500676589986468, |
| "grad_norm": 0.11702322580920509, |
| "learning_rate": 6.324976706347317e-06, |
| "loss": 0.4099, |
| "step": 231 |
| }, |
| { |
| "epoch": 2.5115020297699595, |
| "grad_norm": 0.11497417897845069, |
| "learning_rate": 6.054229700209959e-06, |
| "loss": 0.4073, |
| "step": 232 |
| }, |
| { |
| "epoch": 2.5223274695534506, |
| "grad_norm": 0.11154688705295221, |
| "learning_rate": 5.788929930735916e-06, |
| "loss": 0.4019, |
| "step": 233 |
| }, |
| { |
| "epoch": 2.533152909336942, |
| "grad_norm": 0.1018877557973943, |
| "learning_rate": 5.529119970258014e-06, |
| "loss": 0.4053, |
| "step": 234 |
| }, |
| { |
| "epoch": 2.543978349120433, |
| "grad_norm": 0.10031563066170543, |
| "learning_rate": 5.274841510166062e-06, |
| "loss": 0.398, |
| "step": 235 |
| }, |
| { |
| "epoch": 2.554803788903924, |
| "grad_norm": 0.13448752014975668, |
| "learning_rate": 5.026135354216717e-06, |
| "loss": 0.3976, |
| "step": 236 |
| }, |
| { |
| "epoch": 2.565629228687415, |
| "grad_norm": 0.10140684042198074, |
| "learning_rate": 4.783041411985716e-06, |
| "loss": 0.4048, |
| "step": 237 |
| }, |
| { |
| "epoch": 2.5764546684709067, |
| "grad_norm": 0.10212279572915171, |
| "learning_rate": 4.545598692463675e-06, |
| "loss": 0.4032, |
| "step": 238 |
| }, |
| { |
| "epoch": 2.5872801082543977, |
| "grad_norm": 0.10158019283857284, |
| "learning_rate": 4.3138452977963266e-06, |
| "loss": 0.4004, |
| "step": 239 |
| }, |
| { |
| "epoch": 2.598105548037889, |
| "grad_norm": 0.1022041865541248, |
| "learning_rate": 4.087818417170337e-06, |
| "loss": 0.4077, |
| "step": 240 |
| }, |
| { |
| "epoch": 2.60893098782138, |
| "grad_norm": 0.09996787610145384, |
| "learning_rate": 3.867554320845601e-06, |
| "loss": 0.4103, |
| "step": 241 |
| }, |
| { |
| "epoch": 2.6197564276048713, |
| "grad_norm": 0.10270168841542693, |
| "learning_rate": 3.6530883543350038e-06, |
| "loss": 0.4095, |
| "step": 242 |
| }, |
| { |
| "epoch": 2.6305818673883627, |
| "grad_norm": 0.10267888944347095, |
| "learning_rate": 3.4444549327325325e-06, |
| "loss": 0.4007, |
| "step": 243 |
| }, |
| { |
| "epoch": 2.641407307171854, |
| "grad_norm": 0.11093242523851488, |
| "learning_rate": 3.241687535190776e-06, |
| "loss": 0.3988, |
| "step": 244 |
| }, |
| { |
| "epoch": 2.6522327469553453, |
| "grad_norm": 0.10212024928320693, |
| "learning_rate": 3.0448186995485307e-06, |
| "loss": 0.3997, |
| "step": 245 |
| }, |
| { |
| "epoch": 2.6630581867388363, |
| "grad_norm": 0.1003530665770803, |
| "learning_rate": 2.853880017109516e-06, |
| "loss": 0.4004, |
| "step": 246 |
| }, |
| { |
| "epoch": 2.6738836265223274, |
| "grad_norm": 0.10260969949658093, |
| "learning_rate": 2.6689021275729366e-06, |
| "loss": 0.4066, |
| "step": 247 |
| }, |
| { |
| "epoch": 2.6847090663058184, |
| "grad_norm": 0.10613900755046507, |
| "learning_rate": 2.489914714116788e-06, |
| "loss": 0.4017, |
| "step": 248 |
| }, |
| { |
| "epoch": 2.69553450608931, |
| "grad_norm": 0.10360749903576409, |
| "learning_rate": 2.316946498634605e-06, |
| "loss": 0.3981, |
| "step": 249 |
| }, |
| { |
| "epoch": 2.706359945872801, |
| "grad_norm": 0.09968866548090832, |
| "learning_rate": 2.1500252371265253e-06, |
| "loss": 0.4089, |
| "step": 250 |
| }, |
| { |
| "epoch": 2.7171853856562924, |
| "grad_norm": 0.0957249635009425, |
| "learning_rate": 1.989177715245307e-06, |
| "loss": 0.4046, |
| "step": 251 |
| }, |
| { |
| "epoch": 2.7280108254397835, |
| "grad_norm": 0.09073907320840835, |
| "learning_rate": 1.8344297439980475e-06, |
| "loss": 0.4085, |
| "step": 252 |
| }, |
| { |
| "epoch": 2.7388362652232745, |
| "grad_norm": 0.09354963060185198, |
| "learning_rate": 1.685806155604346e-06, |
| "loss": 0.4075, |
| "step": 253 |
| }, |
| { |
| "epoch": 2.749661705006766, |
| "grad_norm": 0.09424129764450478, |
| "learning_rate": 1.5433307995115043e-06, |
| "loss": 0.4018, |
| "step": 254 |
| }, |
| { |
| "epoch": 2.760487144790257, |
| "grad_norm": 0.08984579907714713, |
| "learning_rate": 1.4070265385674176e-06, |
| "loss": 0.3972, |
| "step": 255 |
| }, |
| { |
| "epoch": 2.7713125845737485, |
| "grad_norm": 0.09324286361746678, |
| "learning_rate": 1.276915245351833e-06, |
| "loss": 0.3955, |
| "step": 256 |
| }, |
| { |
| "epoch": 2.7821380243572396, |
| "grad_norm": 0.08530192588749626, |
| "learning_rate": 1.1530177986664425e-06, |
| "loss": 0.4097, |
| "step": 257 |
| }, |
| { |
| "epoch": 2.7929634641407306, |
| "grad_norm": 0.08424868002952222, |
| "learning_rate": 1.0353540801845229e-06, |
| "loss": 0.4022, |
| "step": 258 |
| }, |
| { |
| "epoch": 2.803788903924222, |
| "grad_norm": 0.08732410562716998, |
| "learning_rate": 9.239429712605274e-07, |
| "loss": 0.4011, |
| "step": 259 |
| }, |
| { |
| "epoch": 2.814614343707713, |
| "grad_norm": 0.08380906207530975, |
| "learning_rate": 8.188023499002206e-07, |
| "loss": 0.4062, |
| "step": 260 |
| }, |
| { |
| "epoch": 2.825439783491204, |
| "grad_norm": 0.08291463845385402, |
| "learning_rate": 7.199490878918314e-07, |
| "loss": 0.3963, |
| "step": 261 |
| }, |
| { |
| "epoch": 2.8362652232746957, |
| "grad_norm": 0.08312318907245991, |
| "learning_rate": 6.273990480986314e-07, |
| "loss": 0.4012, |
| "step": 262 |
| }, |
| { |
| "epoch": 2.8470906630581867, |
| "grad_norm": 0.08687005437995408, |
| "learning_rate": 5.411670819134651e-07, |
| "loss": 0.4055, |
| "step": 263 |
| }, |
| { |
| "epoch": 2.8579161028416777, |
| "grad_norm": 0.09210152893251213, |
| "learning_rate": 4.6126702687554483e-07, |
| "loss": 0.4008, |
| "step": 264 |
| }, |
| { |
| "epoch": 2.8687415426251692, |
| "grad_norm": 0.08608240792550054, |
| "learning_rate": 3.8771170444996895e-07, |
| "loss": 0.4053, |
| "step": 265 |
| }, |
| { |
| "epoch": 2.8795669824086603, |
| "grad_norm": 0.08553485327981258, |
| "learning_rate": 3.2051291797027925e-07, |
| "loss": 0.4039, |
| "step": 266 |
| }, |
| { |
| "epoch": 2.8903924221921518, |
| "grad_norm": 0.08477721494442002, |
| "learning_rate": 2.59681450744389e-07, |
| "loss": 0.3994, |
| "step": 267 |
| }, |
| { |
| "epoch": 2.901217861975643, |
| "grad_norm": 0.08227208962416807, |
| "learning_rate": 2.0522706432419382e-07, |
| "loss": 0.4011, |
| "step": 268 |
| }, |
| { |
| "epoch": 2.912043301759134, |
| "grad_norm": 0.08157490844159622, |
| "learning_rate": 1.5715849693916264e-07, |
| "loss": 0.4027, |
| "step": 269 |
| }, |
| { |
| "epoch": 2.9228687415426253, |
| "grad_norm": 0.11408958016274344, |
| "learning_rate": 1.1548346209410366e-07, |
| "loss": 0.4043, |
| "step": 270 |
| }, |
| { |
| "epoch": 2.9336941813261164, |
| "grad_norm": 0.08386767173187845, |
| "learning_rate": 8.020864733140343e-08, |
| "loss": 0.4018, |
| "step": 271 |
| }, |
| { |
| "epoch": 2.944519621109608, |
| "grad_norm": 0.08255025974503309, |
| "learning_rate": 5.133971315788966e-08, |
| "loss": 0.407, |
| "step": 272 |
| }, |
| { |
| "epoch": 2.955345060893099, |
| "grad_norm": 0.08499202773599712, |
| "learning_rate": 2.8881292136468952e-08, |
| "loss": 0.4048, |
| "step": 273 |
| }, |
| { |
| "epoch": 2.96617050067659, |
| "grad_norm": 0.08679835116294055, |
| "learning_rate": 1.2836988142779228e-08, |
| "loss": 0.4032, |
| "step": 274 |
| }, |
| { |
| "epoch": 2.976995940460081, |
| "grad_norm": 0.08245268279088012, |
| "learning_rate": 3.209375786856761e-09, |
| "loss": 0.4115, |
| "step": 275 |
| }, |
| { |
| "epoch": 2.9878213802435725, |
| "grad_norm": 0.0798253456947544, |
| "learning_rate": 0.0, |
| "loss": 0.3986, |
| "step": 276 |
| }, |
| { |
| "epoch": 2.9878213802435725, |
| "step": 276, |
| "total_flos": 7.10230108051551e+18, |
| "train_loss": 0.4666922677686249, |
| "train_runtime": 60539.5618, |
| "train_samples_per_second": 2.343, |
| "train_steps_per_second": 0.005 |
| } |
| ], |
| "logging_steps": 1.0, |
| "max_steps": 276, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "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": 7.10230108051551e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
