| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 14.962593516209477, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004987531172069825, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 2.7726, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02493765586034913, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 2.8032, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04987531172069826, | |
| "grad_norm": 1.625, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 2.8059, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07481296758104738, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1e-05, | |
| "loss": 2.8169, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09975062344139651, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 2.7666, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12468827930174564, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.7631, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14962593516209477, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 2e-05, | |
| "loss": 2.6985, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1745635910224439, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 2.3333333333333336e-05, | |
| "loss": 2.6688, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.19950124688279303, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 2.558, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22443890274314215, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4229, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.24937655860349128, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.3437, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2743142144638404, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 2.237, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.29925187032418954, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1707, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.32418952618453867, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.3333333333333334e-05, | |
| "loss": 2.0941, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.3491271820448878, | |
| "grad_norm": 13.125, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 1.9843, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3740648379052369, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8983, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.39900249376558605, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 1.8418, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4239401496259352, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.666666666666667e-05, | |
| "loss": 1.7522, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4488778054862843, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 6e-05, | |
| "loss": 1.7016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.47381546134663344, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 6.333333333333333e-05, | |
| "loss": 1.6093, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.49875311720698257, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.5596, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5236907730673317, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5197, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5486284289276808, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 1.4761, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.57356608478803, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 7.666666666666667e-05, | |
| "loss": 1.4268, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5985037406483791, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3845, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6234413965087282, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 1.3622, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6483790523690773, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 1.3215, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6733167082294265, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 9e-05, | |
| "loss": 1.2959, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6982543640897756, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 1.2876, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7231920199501247, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.666666666666667e-05, | |
| "loss": 1.28, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7481296758104738, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2628, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.773067331670823, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 0.00010333333333333334, | |
| "loss": 1.2393, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.7980049875311721, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00010666666666666667, | |
| "loss": 1.2335, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8229426433915212, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 1.2125, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8478802992518704, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 0.00011333333333333334, | |
| "loss": 1.2182, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8728179551122195, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00011666666666666668, | |
| "loss": 1.1958, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.8977556109725686, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.00012, | |
| "loss": 1.2011, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9226932668329177, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00012333333333333334, | |
| "loss": 1.1778, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.9476309226932669, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00012666666666666666, | |
| "loss": 1.1628, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.972568578553616, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 1.1709, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.9975062344139651, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 1.163, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9975062344139651, | |
| "eval_loss": 2.5127053260803223, | |
| "eval_runtime": 0.5084, | |
| "eval_samples_per_second": 19.67, | |
| "eval_steps_per_second": 1.967, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0224438902743143, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00013666666666666666, | |
| "loss": 1.1675, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.0473815461346634, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00014, | |
| "loss": 1.1397, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.0723192019950125, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00014333333333333334, | |
| "loss": 1.143, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.0972568578553616, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00014666666666666666, | |
| "loss": 1.1466, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1221945137157108, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 1.1303, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.14713216957606, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.00015333333333333334, | |
| "loss": 1.1195, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.172069825436409, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00015666666666666666, | |
| "loss": 1.1156, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.1970074812967582, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00016, | |
| "loss": 1.1262, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.2219451371571073, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00016333333333333334, | |
| "loss": 1.1202, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.2468827930174564, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.1114, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2718204488778055, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00017, | |
| "loss": 1.1186, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.2967581047381547, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00017333333333333334, | |
| "loss": 1.1142, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3216957605985038, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00017666666666666666, | |
| "loss": 1.1062, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.346633416458853, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.00018, | |
| "loss": 1.0976, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.371571072319202, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00018333333333333334, | |
| "loss": 1.1058, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.3965087281795512, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.0001866666666666667, | |
| "loss": 1.0984, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4214463840399003, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00019, | |
| "loss": 1.0941, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.4463840399002494, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00019333333333333333, | |
| "loss": 1.0903, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.4713216957605986, | |
| "grad_norm": 0.875, | |
| "learning_rate": 0.00019666666666666666, | |
| "loss": 1.0919, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.4962593516209477, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0744, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5211970074812968, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00019999830768577443, | |
| "loss": 1.0881, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.546134663341646, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00019999323080037624, | |
| "loss": 1.0791, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.571072319201995, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00019998476951563915, | |
| "loss": 1.0633, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.5960099750623442, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00019997292411794618, | |
| "loss": 1.0775, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.6209476309226933, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.0001999576950082201, | |
| "loss": 1.0701, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.6458852867830425, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.0001999390827019096, | |
| "loss": 1.0794, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.6708229426433916, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00019991708782897213, | |
| "loss": 1.0662, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.6957605985037407, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001998917111338525, | |
| "loss": 1.0781, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.7206982543640899, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.0001998629534754574, | |
| "loss": 1.0677, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.745635910224439, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00019983081582712685, | |
| "loss": 1.072, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.770573566084788, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00019979529927660074, | |
| "loss": 1.0668, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.7955112219451372, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00019975640502598244, | |
| "loss": 1.0688, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8204488778054864, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.00019971413439169775, | |
| "loss": 1.0561, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.8453865336658355, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00019966848880445062, | |
| "loss": 1.0585, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.8703241895261846, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00019961946980917456, | |
| "loss": 1.0642, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.8952618453865338, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.00019956707906498044, | |
| "loss": 1.0538, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.9201995012468829, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.00019951131834510032, | |
| "loss": 1.0605, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.945137157107232, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00019945218953682734, | |
| "loss": 1.0631, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.9700748129675811, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.000199389694641452, | |
| "loss": 1.0573, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.9950124688279303, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00019932383577419432, | |
| "loss": 1.0647, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.4642767906188965, | |
| "eval_runtime": 0.528, | |
| "eval_samples_per_second": 18.94, | |
| "eval_steps_per_second": 1.894, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.0199501246882794, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.00019925461516413223, | |
| "loss": 1.0445, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.0448877805486285, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00019918203515412617, | |
| "loss": 1.0286, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0698254364089776, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00019910609820073986, | |
| "loss": 1.0285, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.0947630922693268, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00019902680687415705, | |
| "loss": 1.0393, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.119700748129676, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00019894416385809444, | |
| "loss": 1.0295, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.144638403990025, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00019885817194971117, | |
| "loss": 1.0329, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.169576059850374, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 1.0332, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.1945137157107233, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00019867615321125795, | |
| "loss": 1.0357, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.2194513715710724, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00019858013254184597, | |
| "loss": 1.0275, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.2443890274314215, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00019848077530122083, | |
| "loss": 1.0239, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.2693266832917707, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001983780848522559, | |
| "loss": 1.0292, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.29426433915212, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00019827206467064133, | |
| "loss": 1.0222, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.319201995012469, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00019816271834476642, | |
| "loss": 1.0256, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.344139650872818, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00019805004957559793, | |
| "loss": 1.0329, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.369077306733167, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00019793406217655517, | |
| "loss": 1.0253, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.3940149625935163, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.00019781476007338058, | |
| "loss": 1.0264, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.4189526184538654, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00019769214730400712, | |
| "loss": 1.0241, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.4438902743142146, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 0.00019756622801842143, | |
| "loss": 1.0198, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.4688279301745637, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00019743700647852354, | |
| "loss": 1.0253, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.493765586034913, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 1.0231, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.518703241895262, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00019716867424208806, | |
| "loss": 1.033, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.543640897755611, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00019702957262759965, | |
| "loss": 1.0258, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.56857855361596, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00019688718692259006, | |
| "loss": 1.003, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.5935162094763093, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00019674152194628638, | |
| "loss": 1.0126, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.6184538653366585, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00019659258262890683, | |
| "loss": 1.012, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.6433915211970076, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.0001964403740114939, | |
| "loss": 1.0175, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.6683291770573567, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00019628490124574377, | |
| "loss": 1.0117, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.693266832917706, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.0001961261695938319, | |
| "loss": 1.0187, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.718204488778055, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 0.00019596418442823494, | |
| "loss": 1.0184, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.743142144638404, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.0001957989512315489, | |
| "loss": 1.0146, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.7680798004987532, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.00019563047559630357, | |
| "loss": 1.0141, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.7930174563591024, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.0001954587632247732, | |
| "loss": 1.0209, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.8179551122194515, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00019528381992878362, | |
| "loss": 1.0165, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.8428927680798006, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 1.0161, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.8678304239401498, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 0.0001949242643573034, | |
| "loss": 1.0158, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.892768079800499, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00019473966425143292, | |
| "loss": 1.0127, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.917705735660848, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.0001945518575599317, | |
| "loss": 1.0037, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.942643391521197, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00019436085063935835, | |
| "loss": 1.0068, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.9675810473815463, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.00019416664995458756, | |
| "loss": 1.0061, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.9925187032418954, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 1.0051, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.997506234413965, | |
| "eval_loss": 2.4609696865081787, | |
| "eval_runtime": 0.6338, | |
| "eval_samples_per_second": 15.777, | |
| "eval_steps_per_second": 1.578, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 3.0174563591022445, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00019376869369221452, | |
| "loss": 0.9992, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 3.0423940149625937, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00019356495158395315, | |
| "loss": 0.9824, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.067331670822943, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00019335804264972018, | |
| "loss": 0.9911, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 3.092269326683292, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.00019314797389261424, | |
| "loss": 0.9921, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.117206982543641, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00019293475242268223, | |
| "loss": 0.9925, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.14214463840399, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00019271838545667876, | |
| "loss": 0.9767, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.1670822942643393, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.0001924988803178216, | |
| "loss": 0.9821, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 3.1920199501246884, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00019227624443554425, | |
| "loss": 0.9839, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.2169576059850375, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00019205048534524406, | |
| "loss": 0.9904, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 3.2418952618453867, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00019182161068802741, | |
| "loss": 0.987, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.266832917705736, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00019158962821045112, | |
| "loss": 0.9912, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 3.291770573566085, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.0001913545457642601, | |
| "loss": 0.9848, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.316708229426434, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.0001911163713061217, | |
| "loss": 0.9872, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 3.341645885286783, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.00019087511289735644, | |
| "loss": 0.989, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.3665835411471323, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.000190630778703665, | |
| "loss": 0.9968, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.3915211970074814, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00019038337699485208, | |
| "loss": 0.9839, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.4164588528678306, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00019013291614454621, | |
| "loss": 0.9919, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 3.4413965087281797, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.0001898794046299167, | |
| "loss": 0.9905, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.466334164588529, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00018962285103138636, | |
| "loss": 0.9932, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 3.491271820448878, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00018936326403234125, | |
| "loss": 0.982, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.516209476309227, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 0.9914, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 3.541147132169576, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00018883502507930042, | |
| "loss": 0.9809, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.5660847880299253, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.0001885663910042306, | |
| "loss": 0.9818, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 3.5910224438902745, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00018829475928589271, | |
| "loss": 0.9841, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.6159600997506236, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.00018802013911801112, | |
| "loss": 0.9796, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 3.6408977556109727, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.0001877425397954582, | |
| "loss": 0.9817, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.665835411471322, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00018746197071393958, | |
| "loss": 0.9848, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 3.690773067331671, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00018717844136967624, | |
| "loss": 0.9814, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.71571072319202, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00018689196135908304, | |
| "loss": 0.9831, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.7406483790523692, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.9829, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.765586034912718, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00018631018822358363, | |
| "loss": 0.9651, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.7905236907730675, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00018601491478953657, | |
| "loss": 0.9748, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.815461346633416, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00018571673007021123, | |
| "loss": 0.9896, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.8403990024937658, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00018541564415805258, | |
| "loss": 0.9782, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.8653366583541144, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00018511166724369997, | |
| "loss": 0.9824, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.890274314214464, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.0001848048096156426, | |
| "loss": 0.9884, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.9152119700748127, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00018449508165987105, | |
| "loss": 0.9872, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.9401496259351623, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00018418249385952575, | |
| "loss": 0.9833, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.965087281795511, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.00018386705679454242, | |
| "loss": 0.9718, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.9900249376558605, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00018354878114129367, | |
| "loss": 0.9807, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.4767043590545654, | |
| "eval_runtime": 0.5377, | |
| "eval_samples_per_second": 18.597, | |
| "eval_steps_per_second": 1.86, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 4.014962593516209, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.0001832276776722278, | |
| "loss": 0.9718, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 4.039900249376559, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00018290375725550417, | |
| "loss": 0.9596, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 4.0648379052369075, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.00018257703085462542, | |
| "loss": 0.9552, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 4.089775561097257, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00018224750952806624, | |
| "loss": 0.9614, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 4.114713216957606, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001819152044288992, | |
| "loss": 0.9702, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 4.139650872817955, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00018158012680441723, | |
| "loss": 0.9534, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 4.164588528678304, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00018124228799575295, | |
| "loss": 0.9643, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 4.1895261845386536, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.9464, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 4.214463840399002, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00018055837265729994, | |
| "loss": 0.9624, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 4.239401496259352, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.0001802123192755044, | |
| "loss": 0.9606, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.2643391521197005, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00017986355100472928, | |
| "loss": 0.9548, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 4.28927680798005, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.0001795120796494848, | |
| "loss": 0.9541, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 4.314214463840399, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00017915791710577033, | |
| "loss": 0.9537, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 4.339152119700748, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00017880107536067218, | |
| "loss": 0.9691, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 4.364089775561097, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00017844156649195759, | |
| "loss": 0.9597, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 4.389027431421447, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00017807940266766593, | |
| "loss": 0.9517, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.413965087281795, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.0001777145961456971, | |
| "loss": 0.9598, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 4.438902743142145, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.0001773471592733964, | |
| "loss": 0.9617, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 4.4638403990024935, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00017697710448713678, | |
| "loss": 0.9513, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 4.488778054862843, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.9637, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.513715710723192, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00017622919136084183, | |
| "loss": 0.9562, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 4.538653366583541, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00017585135833488692, | |
| "loss": 0.9632, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.56359102244389, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 0.00017547095802227723, | |
| "loss": 0.9448, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 4.58852867830424, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00017508800329814995, | |
| "loss": 0.963, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.613466334164588, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001747025071240996, | |
| "loss": 0.962, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 4.638403990024938, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00017431448254773944, | |
| "loss": 0.9627, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.6633416458852865, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.0001739239427022596, | |
| "loss": 0.9574, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 4.688279301745636, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001735309008059829, | |
| "loss": 0.9675, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.713216957605985, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00017313537016191706, | |
| "loss": 0.9546, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 4.738154613466334, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00017273736415730488, | |
| "loss": 0.9472, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.763092269326683, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001723368962631708, | |
| "loss": 0.9528, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 4.788029925187033, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.0001719339800338651, | |
| "loss": 0.962, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.812967581047381, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00017152862910660516, | |
| "loss": 0.9602, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 4.837905236907731, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.00017112085720101373, | |
| "loss": 0.9529, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.86284289276808, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.957, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 4.887780548628429, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001702981057425662, | |
| "loss": 0.9597, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.912718204488778, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00016988315403679, | |
| "loss": 0.9653, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 4.937655860349127, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00016946583704589973, | |
| "loss": 0.9486, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.962593516209476, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016904616889452497, | |
| "loss": 0.9587, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.987531172069826, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.9508, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.997506234413965, | |
| "eval_loss": 2.4788339138031006, | |
| "eval_runtime": 0.6532, | |
| "eval_samples_per_second": 15.309, | |
| "eval_steps_per_second": 1.531, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 5.012468827930174, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00016819983600624986, | |
| "loss": 0.9493, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 5.037406483790524, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00016777319991457325, | |
| "loss": 0.941, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 5.062344139650873, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00016734426995189004, | |
| "loss": 0.9364, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 5.087281795511222, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00016691306063588583, | |
| "loss": 0.9424, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 5.112219451371571, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00016647958656139378, | |
| "loss": 0.9365, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 5.13715710723192, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00016604386239990078, | |
| "loss": 0.9405, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 5.162094763092269, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00016560590289905073, | |
| "loss": 0.9274, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 5.187032418952619, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00016516572288214552, | |
| "loss": 0.9433, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 5.211970074812967, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016472333724764325, | |
| "loss": 0.9372, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 5.236907730673317, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.00016427876096865394, | |
| "loss": 0.948, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 5.261845386533666, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00016383200909243285, | |
| "loss": 0.9342, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 5.286783042394015, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00016338309673987101, | |
| "loss": 0.944, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 5.311720698254364, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00016293203910498376, | |
| "loss": 0.9438, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 5.3366583541147135, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.000162478851454396, | |
| "loss": 0.9319, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.361596009975062, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.000162023549126826, | |
| "loss": 0.9411, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 5.386533665835412, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001615661475325658, | |
| "loss": 0.9316, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 5.41147132169576, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00016110666215295998, | |
| "loss": 0.9386, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 5.43640897755611, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00016064510853988138, | |
| "loss": 0.9429, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 5.461346633416459, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00016018150231520486, | |
| "loss": 0.9378, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 5.486284289276808, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00015971585917027862, | |
| "loss": 0.9416, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.511221945137157, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00015924819486539307, | |
| "loss": 0.9239, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 5.5361596009975065, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.9264, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 5.561097256857855, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00015830686615841348, | |
| "loss": 0.9382, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 5.586034912718205, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00015783323361679864, | |
| "loss": 0.9405, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 5.610972568578553, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0001573576436351046, | |
| "loss": 0.9466, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 5.635910224438903, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00015688011231028518, | |
| "loss": 0.9366, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 5.660847880299252, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00015640065580500148, | |
| "loss": 0.9419, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 5.685785536159601, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001559192903470747, | |
| "loss": 0.9283, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 5.71072319201995, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00015543603222893716, | |
| "loss": 0.9384, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 5.7356608478802995, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.0001549508978070806, | |
| "loss": 0.9373, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.760598503740648, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00015446390350150273, | |
| "loss": 0.9268, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 5.785536159600998, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.0001539750657951513, | |
| "loss": 0.9381, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 5.8104738154613464, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.00015348440123336645, | |
| "loss": 0.9321, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 5.835411471321696, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.0001529919264233205, | |
| "loss": 0.9304, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 5.860349127182045, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.000152497658033456, | |
| "loss": 0.934, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 5.885286783042394, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00015200161279292155, | |
| "loss": 0.9493, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 5.910224438902743, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00015150380749100545, | |
| "loss": 0.9362, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 5.9351620947630925, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00015100425897656753, | |
| "loss": 0.9348, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 5.960099750623441, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.000150502984157469, | |
| "loss": 0.9406, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 5.985037406483791, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.9256, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.491248607635498, | |
| "eval_runtime": 0.5413, | |
| "eval_samples_per_second": 18.475, | |
| "eval_steps_per_second": 1.848, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 6.0099750623441395, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00014949532352830541, | |
| "loss": 0.9255, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 6.034912718204489, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001489889718238087, | |
| "loss": 0.9186, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 6.059850374064838, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00014848096202463372, | |
| "loss": 0.9201, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 6.084788029925187, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00014797131132502465, | |
| "loss": 0.9192, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 6.109725685785536, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00014746003697476404, | |
| "loss": 0.9178, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 6.134663341645886, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.00014694715627858908, | |
| "loss": 0.9112, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 6.159600997506234, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00014643268659560572, | |
| "loss": 0.9181, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 6.184538653366584, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00014591664533870118, | |
| "loss": 0.9302, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 6.2094763092269325, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 0.9257, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 6.234413965087282, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.9188, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 6.259351620947631, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00014435926704766362, | |
| "loss": 0.9218, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 6.28428927680798, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00014383711467890774, | |
| "loss": 0.9132, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 6.309226932668329, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00014331347858669632, | |
| "loss": 0.931, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 6.334164588528679, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00014278837649416544, | |
| "loss": 0.9187, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 6.359102244389027, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00014226182617406996, | |
| "loss": 0.9258, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 6.384039900249377, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001417338454481818, | |
| "loss": 0.9125, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 6.4089775561097255, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00014120445218668686, | |
| "loss": 0.9148, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 6.433915211970075, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00014067366430758004, | |
| "loss": 0.9149, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 6.458852867830424, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00014014149977605893, | |
| "loss": 0.9119, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 6.483790523690773, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.0001396079766039157, | |
| "loss": 0.9313, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.508728179551122, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00013907311284892736, | |
| "loss": 0.9148, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 6.533665835411472, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00013853692661424484, | |
| "loss": 0.9253, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 6.55860349127182, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00013799943604777992, | |
| "loss": 0.9204, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 6.58354114713217, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00013746065934159123, | |
| "loss": 0.9219, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 6.6084788029925186, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00013692061473126845, | |
| "loss": 0.9162, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 6.633416458852868, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00013637932049531516, | |
| "loss": 0.9239, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 6.658354114713217, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00013583679495453, | |
| "loss": 0.9217, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 6.683291770573566, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00013529305647138687, | |
| "loss": 0.9268, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 6.708229426433915, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00013474812344941315, | |
| "loss": 0.9238, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 6.733167082294265, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00013420201433256689, | |
| "loss": 0.9141, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 6.758104738154613, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00013365474760461266, | |
| "loss": 0.9219, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 6.783042394014963, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.0001331063417884958, | |
| "loss": 0.9177, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 6.807980049875312, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00013255681544571568, | |
| "loss": 0.9215, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 6.832917705735661, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00013200618717569714, | |
| "loss": 0.9087, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 6.85785536159601, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00013145447561516138, | |
| "loss": 0.9276, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 6.882793017456359, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.9147, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 6.907730673316708, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.0001303478773521171, | |
| "loss": 0.9209, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 6.932668329177058, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.0001297930281038482, | |
| "loss": 0.9248, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 6.957605985037406, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00012923717047227368, | |
| "loss": 0.9197, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 6.982543640897756, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00012868032327110904, | |
| "loss": 0.9216, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.997506234413965, | |
| "eval_loss": 2.5038018226623535, | |
| "eval_runtime": 0.6591, | |
| "eval_samples_per_second": 15.172, | |
| "eval_steps_per_second": 1.517, | |
| "step": 1403 | |
| }, | |
| { | |
| "epoch": 7.007481296758105, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00012812250534756308, | |
| "loss": 0.9115, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 7.032418952618454, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001275637355816999, | |
| "loss": 0.9038, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 7.057356608478803, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.0001270040328858001, | |
| "loss": 0.9088, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 7.082294264339152, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00012644341620372023, | |
| "loss": 0.9016, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 7.107231920199501, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00012588190451025207, | |
| "loss": 0.9112, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 7.132169576059851, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.0001253195168104802, | |
| "loss": 0.9081, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 7.157107231920199, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.0001247562721391386, | |
| "loss": 0.9094, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 7.182044887780549, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00012419218955996676, | |
| "loss": 0.8956, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 7.206982543640898, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00012362728816506417, | |
| "loss": 0.9044, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 7.231920199501247, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00012306158707424403, | |
| "loss": 0.9069, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 7.256857855361596, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.0001224951054343865, | |
| "loss": 0.8949, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 7.2817955112219455, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00012192786241879033, | |
| "loss": 0.901, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 7.306733167082294, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00012135987722652402, | |
| "loss": 0.8983, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 7.331670822942644, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00012079116908177593, | |
| "loss": 0.9113, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 7.356608478802992, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00012022175723320381, | |
| "loss": 0.9026, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 7.381546134663342, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00011965166095328301, | |
| "loss": 0.9075, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 7.406483790523691, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00011908089953765449, | |
| "loss": 0.9045, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 7.43142144638404, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00011850949230447145, | |
| "loss": 0.9099, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 7.456359102244389, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00011793745859374575, | |
| "loss": 0.9035, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 7.4812967581047385, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.9093, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.506234413965087, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00011679158920507774, | |
| "loss": 0.8925, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 7.531172069825437, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00011621779231055676, | |
| "loss": 0.9054, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 7.556109725685785, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 0.9017, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 7.581047381546135, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00011506857122494831, | |
| "loss": 0.9125, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 7.605985037406484, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00011449318593072466, | |
| "loss": 0.9051, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 7.630922693266833, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00011391731009600654, | |
| "loss": 0.894, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 7.655860349127182, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00011334096321205128, | |
| "loss": 0.9096, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 7.6807980049875315, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00011276416478605949, | |
| "loss": 0.908, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 7.70573566084788, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00011218693434051475, | |
| "loss": 0.9118, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 7.73067331670823, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00011160929141252303, | |
| "loss": 0.8973, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 7.7556109725685785, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.00011103125555315119, | |
| "loss": 0.9025, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 7.780548628428928, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00011045284632676536, | |
| "loss": 0.8963, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 7.805486284289277, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00010987408331036879, | |
| "loss": 0.9171, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 7.830423940149626, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00010929498609293924, | |
| "loss": 0.9225, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 7.855361596009975, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00010871557427476583, | |
| "loss": 0.9087, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 7.8802992518703245, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.00010813586746678583, | |
| "loss": 0.9119, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 7.905236907730673, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00010755588528992082, | |
| "loss": 0.8963, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 7.930174563591023, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00010697564737441252, | |
| "loss": 0.9115, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 7.9551122194513715, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00010639517335915856, | |
| "loss": 0.903, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 7.980049875311721, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00010581448289104758, | |
| "loss": 0.9094, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.512359380722046, | |
| "eval_runtime": 0.5399, | |
| "eval_samples_per_second": 18.52, | |
| "eval_steps_per_second": 1.852, | |
| "step": 1604 | |
| }, | |
| { | |
| "epoch": 8.00498753117207, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.0001052335956242944, | |
| "loss": 0.9066, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 8.029925187032418, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0001046525312197747, | |
| "loss": 0.8932, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 8.054862842892769, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.0001040713093443596, | |
| "loss": 0.8877, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 8.079800498753118, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00010348994967025012, | |
| "loss": 0.9041, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 8.104738154613466, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00010290847187431113, | |
| "loss": 0.8898, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 8.129675810473815, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00010232689563740563, | |
| "loss": 0.8814, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 8.154613466334165, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00010174524064372837, | |
| "loss": 0.8814, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 8.179551122194514, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00010116352658013973, | |
| "loss": 0.8934, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 8.204488778054863, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00010058177313549939, | |
| "loss": 0.8979, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 8.229426433915211, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9019, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 8.254364089775562, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 9.94182268645006e-05, | |
| "loss": 0.8859, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 8.27930174563591, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 9.883647341986032e-05, | |
| "loss": 0.8938, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 8.30423940149626, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 9.825475935627165e-05, | |
| "loss": 0.892, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 8.329177057356608, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 9.767310436259438e-05, | |
| "loss": 0.8955, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 8.354114713216958, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.709152812568886e-05, | |
| "loss": 0.8929, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 8.379052369077307, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 9.651005032974994e-05, | |
| "loss": 0.8945, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 8.403990024937656, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 9.592869065564043e-05, | |
| "loss": 0.888, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 8.428927680798004, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 9.534746878022534e-05, | |
| "loss": 0.8907, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 8.453865336658355, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 9.476640437570562e-05, | |
| "loss": 0.8885, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 8.478802992518704, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 9.418551710895243e-05, | |
| "loss": 0.8867, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.503740648379052, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 9.360482664084145e-05, | |
| "loss": 0.8881, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 8.528678304239401, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 9.302435262558747e-05, | |
| "loss": 0.8964, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 8.553615960099751, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 9.244411471007922e-05, | |
| "loss": 0.888, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 8.5785536159601, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 9.186413253321418e-05, | |
| "loss": 0.8806, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 8.603491271820449, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 9.128442572523417e-05, | |
| "loss": 0.8849, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 8.628428927680797, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 9.070501390706079e-05, | |
| "loss": 0.8924, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 8.653366583541148, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 9.012591668963122e-05, | |
| "loss": 0.8911, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 8.678304239401497, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 8.954715367323468e-05, | |
| "loss": 0.8956, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 8.703241895261845, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 8.896874444684883e-05, | |
| "loss": 0.8981, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 8.728179551122194, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.8989, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 8.753117206982544, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 8.781306565948528e-05, | |
| "loss": 0.9039, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 8.778054862842893, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 8.723583521394054e-05, | |
| "loss": 0.9023, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 8.802992518703242, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 8.665903678794873e-05, | |
| "loss": 0.9084, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 8.82793017456359, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 8.608268990399349e-05, | |
| "loss": 0.8976, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 8.85286783042394, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.550681406927535e-05, | |
| "loss": 0.8931, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 8.87780548628429, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 8.49314287750517e-05, | |
| "loss": 0.8893, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 8.902743142144638, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 0.9013, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 8.927680798004987, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 8.378220768944327e-05, | |
| "loss": 0.9058, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 8.952618453865338, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 8.32084107949223e-05, | |
| "loss": 0.9019, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 8.977556109725686, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.8961, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.997506234413965, | |
| "eval_loss": 2.5246224403381348, | |
| "eval_runtime": 0.6529, | |
| "eval_samples_per_second": 15.316, | |
| "eval_steps_per_second": 1.532, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 9.002493765586035, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 8.206254140625426e-05, | |
| "loss": 0.8949, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 9.027431421446384, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 8.149050769552856e-05, | |
| "loss": 0.8901, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 9.052369077306734, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 8.091910046234552e-05, | |
| "loss": 0.8822, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 9.077306733167083, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 8.034833904671698e-05, | |
| "loss": 0.8846, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 9.102244389027431, | |
| "grad_norm": 0.5, | |
| "learning_rate": 7.977824276679623e-05, | |
| "loss": 0.8819, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 9.12718204488778, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 7.920883091822408e-05, | |
| "loss": 0.888, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 9.15211970074813, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 7.864012277347602e-05, | |
| "loss": 0.8682, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 9.17705735660848, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 7.807213758120966e-05, | |
| "loss": 0.8772, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 9.201995012468828, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 7.750489456561352e-05, | |
| "loss": 0.8871, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 9.226932668329177, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 7.693841292575598e-05, | |
| "loss": 0.8856, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 9.251870324189527, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 7.637271183493586e-05, | |
| "loss": 0.884, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 9.276807980049876, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 7.580781044003324e-05, | |
| "loss": 0.8885, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 9.301745635910224, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 7.524372786086142e-05, | |
| "loss": 0.8825, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 9.326683291770573, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 7.468048318951983e-05, | |
| "loss": 0.8863, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 9.351620947630924, | |
| "grad_norm": 0.5, | |
| "learning_rate": 7.411809548974792e-05, | |
| "loss": 0.8877, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 9.376558603491272, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 7.35565837962798e-05, | |
| "loss": 0.8749, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 9.401496259351621, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 7.299596711419994e-05, | |
| "loss": 0.8851, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 9.42643391521197, | |
| "grad_norm": 0.5, | |
| "learning_rate": 7.243626441830009e-05, | |
| "loss": 0.8957, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 9.451371571072318, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 7.187749465243693e-05, | |
| "loss": 0.8904, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 9.476309226932669, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 7.131967672889101e-05, | |
| "loss": 0.8822, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 9.501246882793017, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 7.076282952772633e-05, | |
| "loss": 0.8771, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 9.526184538653366, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 7.02069718961518e-05, | |
| "loss": 0.8841, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 9.551122194513717, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 6.965212264788297e-05, | |
| "loss": 0.8857, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 9.576059850374065, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.8808, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 9.600997506234414, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 6.854552438483865e-05, | |
| "loss": 0.8872, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 9.625935162094763, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 6.799381282430284e-05, | |
| "loss": 0.893, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 9.650872817955111, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 6.744318455428436e-05, | |
| "loss": 0.87, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 9.675810473815462, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 6.68936582115042e-05, | |
| "loss": 0.9005, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 9.70074812967581, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 6.634525239538736e-05, | |
| "loss": 0.8936, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 9.72568578553616, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 6.579798566743314e-05, | |
| "loss": 0.898, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 9.75062344139651, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 6.525187655058686e-05, | |
| "loss": 0.8844, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 9.775561097256858, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 6.470694352861312e-05, | |
| "loss": 0.8849, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 9.800498753117207, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 6.416320504546997e-05, | |
| "loss": 0.881, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 9.825436408977556, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 6.362067950468489e-05, | |
| "loss": 0.8804, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 9.850374064837904, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 6.307938526873157e-05, | |
| "loss": 0.8794, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 9.875311720698255, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 6.25393406584088e-05, | |
| "loss": 0.8784, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 9.900249376558603, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 6.200056395222012e-05, | |
| "loss": 0.8875, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 9.925187032418952, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 6.146307338575519e-05, | |
| "loss": 0.8774, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 9.950124688279303, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 6.092688715107264e-05, | |
| "loss": 0.8789, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 9.975062344139651, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.8922, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 5.985850022394106e-05, | |
| "loss": 0.8816, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.5341641902923584, | |
| "eval_runtime": 0.5369, | |
| "eval_samples_per_second": 18.625, | |
| "eval_steps_per_second": 1.863, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 10.024937655860349, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 5.9326335692419995e-05, | |
| "loss": 0.8711, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 10.049875311720697, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 5.879554781331317e-05, | |
| "loss": 0.8889, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 10.074812967581048, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 5.8266154551818216e-05, | |
| "loss": 0.8789, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 10.099750623441397, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 5.773817382593008e-05, | |
| "loss": 0.8638, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 10.124688279301745, | |
| "grad_norm": 0.5, | |
| "learning_rate": 5.72116235058346e-05, | |
| "loss": 0.874, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 10.149625935162096, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 5.668652141330373e-05, | |
| "loss": 0.8764, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 10.174563591022444, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 5.616288532109225e-05, | |
| "loss": 0.8855, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 10.199501246882793, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 5.564073295233645e-05, | |
| "loss": 0.8802, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 10.224438902743142, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 5.5120081979953785e-05, | |
| "loss": 0.8738, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 10.24937655860349, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 0.8715, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 10.27431421446384, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 5.4083354661298814e-05, | |
| "loss": 0.878, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 10.29925187032419, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 5.356731340439431e-05, | |
| "loss": 0.8851, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 10.324189526184538, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 5.305284372141095e-05, | |
| "loss": 0.8741, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 10.349127182044889, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 5.253996302523596e-05, | |
| "loss": 0.8801, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 10.374064837905237, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 5.2028688674975415e-05, | |
| "loss": 0.8729, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 10.399002493765586, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 5.15190379753663e-05, | |
| "loss": 0.8726, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 10.423940149625935, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 5.101102817619131e-05, | |
| "loss": 0.8908, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 10.448877805486283, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 5.05046764716946e-05, | |
| "loss": 0.8868, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 10.473815461346634, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.8743, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 10.498753117206983, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 4.9497015842531026e-05, | |
| "loss": 0.8756, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 10.523690773067331, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 4.899574102343247e-05, | |
| "loss": 0.881, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 10.548628428927682, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 4.8496192508994576e-05, | |
| "loss": 0.8769, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 10.57356608478803, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 4.799838720707846e-05, | |
| "loss": 0.88, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 10.598503740648379, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 4.7502341966544e-05, | |
| "loss": 0.8787, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 10.623441396508728, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 4.700807357667952e-05, | |
| "loss": 0.8774, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 10.648379052369076, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 4.6515598766633597e-05, | |
| "loss": 0.8807, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 10.673316708229427, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 4.6024934204848745e-05, | |
| "loss": 0.8702, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 10.698254364089776, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 4.5536096498497295e-05, | |
| "loss": 0.873, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 10.723192019950124, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 4.50491021929194e-05, | |
| "loss": 0.8785, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 10.748129675810475, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 4.456396777106285e-05, | |
| "loss": 0.8851, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 10.773067331670823, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 4.4080709652925336e-05, | |
| "loss": 0.8827, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 10.798004987531172, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 4.359934419499858e-05, | |
| "loss": 0.8717, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 10.82294264339152, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 4.3119887689714844e-05, | |
| "loss": 0.8737, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 10.84788029925187, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 4.264235636489542e-05, | |
| "loss": 0.8703, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 10.87281795511222, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 4.216676638320135e-05, | |
| "loss": 0.8657, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 10.897755610972569, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 4.169313384158653e-05, | |
| "loss": 0.8798, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 10.922693266832917, | |
| "grad_norm": 0.5, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.873, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 10.947630922693268, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 4.0751805134606944e-05, | |
| "loss": 0.877, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 10.972568578553616, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 4.028414082972141e-05, | |
| "loss": 0.8892, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.997506234413965, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 3.981849768479517e-05, | |
| "loss": 0.8722, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 10.997506234413965, | |
| "eval_loss": 2.534595012664795, | |
| "eval_runtime": 0.6862, | |
| "eval_samples_per_second": 14.573, | |
| "eval_steps_per_second": 1.457, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 11.022443890274314, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 3.935489146011869e-05, | |
| "loss": 0.8622, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 11.047381546134662, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 3.8893337847040025e-05, | |
| "loss": 0.8802, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 11.072319201995013, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 3.843385246743417e-05, | |
| "loss": 0.8698, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 11.097256857855362, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 3.7976450873174005e-05, | |
| "loss": 0.8726, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 11.12219451371571, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 3.7521148545604e-05, | |
| "loss": 0.8788, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 11.147132169576059, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 3.7067960895016275e-05, | |
| "loss": 0.8712, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 11.17206982543641, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 3.661690326012897e-05, | |
| "loss": 0.8716, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 11.197007481296758, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.61679909075672e-05, | |
| "loss": 0.8845, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 11.221945137157107, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.874, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 11.246882793017456, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.527666275235677e-05, | |
| "loss": 0.8713, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 11.271820448877806, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 3.483427711785449e-05, | |
| "loss": 0.8764, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 11.296758104738155, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 3.439409710094929e-05, | |
| "loss": 0.8751, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 11.321695760598503, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 3.395613760009925e-05, | |
| "loss": 0.8752, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 11.346633416458852, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 3.352041343860621e-05, | |
| "loss": 0.8738, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 11.371571072319203, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 3.308693936411421e-05, | |
| "loss": 0.869, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 11.396508728179551, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 3.265573004810997e-05, | |
| "loss": 0.8755, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 11.4214463840399, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 3.222680008542678e-05, | |
| "loss": 0.8732, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 11.446384039900249, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 3.1800163993750166e-05, | |
| "loss": 0.8741, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 11.471321695760599, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 3.137583621312665e-05, | |
| "loss": 0.8636, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 11.496259351620948, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.095383110547506e-05, | |
| "loss": 0.8758, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 11.521197007481296, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 3.053416295410026e-05, | |
| "loss": 0.8626, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 11.546134663341645, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 3.0116845963209993e-05, | |
| "loss": 0.8797, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 11.571072319201996, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 2.9701894257433826e-05, | |
| "loss": 0.8623, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 11.596009975062344, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.8736, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 11.620947630922693, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 2.8879142798986292e-05, | |
| "loss": 0.8656, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 11.645885286783042, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 2.8471370893394866e-05, | |
| "loss": 0.8637, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 11.670822942643392, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 2.8066019966134904e-05, | |
| "loss": 0.875, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 11.69576059850374, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 2.7663103736829198e-05, | |
| "loss": 0.8727, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 11.72069825436409, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 2.7262635842695127e-05, | |
| "loss": 0.8767, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 11.745635910224438, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 2.6864629838082956e-05, | |
| "loss": 0.8718, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 11.770573566084789, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.6469099194017143e-05, | |
| "loss": 0.8805, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 11.795511221945137, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 2.6076057297740407e-05, | |
| "loss": 0.8701, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 11.820448877805486, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 2.5685517452260567e-05, | |
| "loss": 0.8661, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 11.845386533665835, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 2.529749287590042e-05, | |
| "loss": 0.8684, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 11.870324189526185, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.491199670185008e-05, | |
| "loss": 0.8673, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 11.895261845386534, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 2.45290419777228e-05, | |
| "loss": 0.8707, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 11.920199501246882, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 2.4148641665113113e-05, | |
| "loss": 0.8744, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 11.945137157107231, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 2.3770808639158216e-05, | |
| "loss": 0.8767, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 11.970074812967582, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.8761, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 11.99501246882793, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 2.302289551286321e-05, | |
| "loss": 0.8768, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 2.5409655570983887, | |
| "eval_runtime": 0.5391, | |
| "eval_samples_per_second": 18.549, | |
| "eval_steps_per_second": 1.855, | |
| "step": 2406 | |
| }, | |
| { | |
| "epoch": 12.019950124688279, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 2.265284072660362e-05, | |
| "loss": 0.8646, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 12.044887780548628, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 2.2285403854302912e-05, | |
| "loss": 0.8762, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 12.069825436408978, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 2.192059733233408e-05, | |
| "loss": 0.8769, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 12.094763092269327, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 2.155843350804243e-05, | |
| "loss": 0.865, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 12.119700748129675, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 2.119892463932781e-05, | |
| "loss": 0.8751, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 12.144638403990024, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 2.0842082894229676e-05, | |
| "loss": 0.8704, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 12.169576059850375, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 2.0487920350515212e-05, | |
| "loss": 0.8698, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 12.194513715710723, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 2.013644899527074e-05, | |
| "loss": 0.8773, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 12.219451371571072, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 1.9787680724495617e-05, | |
| "loss": 0.869, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 12.24438902743142, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 1.9441627342700065e-05, | |
| "loss": 0.8817, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 12.269326683291771, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.8671, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 12.29426433915212, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 1.8757712004247096e-05, | |
| "loss": 0.8663, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 12.319201995012468, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 1.8419873195582814e-05, | |
| "loss": 0.8712, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 12.344139650872817, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.808479557110081e-05, | |
| "loss": 0.8737, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 12.369077306733168, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 1.775249047193377e-05, | |
| "loss": 0.866, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 12.394014962593516, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 1.7422969145374592e-05, | |
| "loss": 0.8699, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 12.418952618453865, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 1.7096242744495837e-05, | |
| "loss": 0.875, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 12.443890274314214, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.677232232777224e-05, | |
| "loss": 0.8704, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 12.468827930174564, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.869, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 12.493765586034913, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 1.6132943205457606e-05, | |
| "loss": 0.8732, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 12.518703241895262, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 1.5817506140474247e-05, | |
| "loss": 0.8706, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 12.54364089775561, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 1.550491834012898e-05, | |
| "loss": 0.8665, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 12.56857855361596, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 1.5195190384357404e-05, | |
| "loss": 0.8715, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 12.59351620947631, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 1.4888332756300027e-05, | |
| "loss": 0.873, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 12.618453865336658, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 1.458435584194745e-05, | |
| "loss": 0.8663, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 12.643391521197007, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 1.4283269929788779e-05, | |
| "loss": 0.8715, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 12.668329177057357, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.3985085210463477e-05, | |
| "loss": 0.8625, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 12.693266832917706, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 1.3689811776416362e-05, | |
| "loss": 0.872, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 12.718204488778055, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 0.8621, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 12.743142144638403, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 1.3108038640916986e-05, | |
| "loss": 0.8709, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 12.768079800498754, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 1.2821558630323772e-05, | |
| "loss": 0.8675, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 12.793017456359102, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 1.2538029286060426e-05, | |
| "loss": 0.8683, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 12.817955112219451, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 1.2257460204541794e-05, | |
| "loss": 0.8646, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 12.8428927680798, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.1979860881988902e-05, | |
| "loss": 0.8703, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 12.86783042394015, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 1.1705240714107302e-05, | |
| "loss": 0.8732, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 12.892768079800499, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 1.1433608995769395e-05, | |
| "loss": 0.8722, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 12.917705735660848, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.116497492069961e-05, | |
| "loss": 0.8724, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 12.942643391521196, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 0.867, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 12.967581047381547, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.0636735967658784e-05, | |
| "loss": 0.8667, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 12.992518703241895, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 1.0377148968613658e-05, | |
| "loss": 0.8694, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 12.997506234413965, | |
| "eval_loss": 2.5414979457855225, | |
| "eval_runtime": 0.6532, | |
| "eval_samples_per_second": 15.31, | |
| "eval_steps_per_second": 1.531, | |
| "step": 2606 | |
| }, | |
| { | |
| "epoch": 13.017456359102244, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 1.0120595370083318e-05, | |
| "loss": 0.8778, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 13.042394014962593, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 9.867083855453774e-06, | |
| "loss": 0.8707, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 13.067331670822943, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.616623005147951e-06, | |
| "loss": 0.8675, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 13.092269326683292, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 9.369221296335006e-06, | |
| "loss": 0.8655, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 13.11720698254364, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 9.124887102643575e-06, | |
| "loss": 0.8662, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 13.14214463840399, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 8.883628693878298e-06, | |
| "loss": 0.8656, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 13.16708229426434, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 8.645454235739903e-06, | |
| "loss": 0.8677, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 13.192019950124688, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 8.41037178954891e-06, | |
| "loss": 0.8673, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 13.216957605985037, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 8.178389311972612e-06, | |
| "loss": 0.8726, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 13.241895261845386, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 7.949514654755962e-06, | |
| "loss": 0.8622, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 13.266832917705736, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 7.72375556445577e-06, | |
| "loss": 0.8703, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 13.291770573566085, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 7.501119682178393e-06, | |
| "loss": 0.8723, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 13.316708229426434, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 7.281614543321269e-06, | |
| "loss": 0.8673, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 13.341645885286782, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 7.0652475773177464e-06, | |
| "loss": 0.867, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 13.366583541147133, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 6.852026107385756e-06, | |
| "loss": 0.8627, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 13.391521197007481, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 6.6419573502798374e-06, | |
| "loss": 0.8693, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 13.41645885286783, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 6.435048416046863e-06, | |
| "loss": 0.8745, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 13.441396508728179, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 6.231306307785522e-06, | |
| "loss": 0.8654, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 13.46633416458853, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.8763, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 13.491271820448878, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 5.833350045412478e-06, | |
| "loss": 0.8773, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 13.516209476309227, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 5.639149360641649e-06, | |
| "loss": 0.8763, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 13.541147132169575, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 5.448142440068316e-06, | |
| "loss": 0.8732, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 13.566084788029926, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 5.26033574856708e-06, | |
| "loss": 0.8718, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 13.591022443890274, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 5.075735642696611e-06, | |
| "loss": 0.8776, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 13.615960099750623, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.8643, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 13.640897755610972, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 4.7161800712163806e-06, | |
| "loss": 0.8767, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 13.665835411471322, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 4.541236775226809e-06, | |
| "loss": 0.8656, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 13.690773067331671, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 4.369524403696457e-06, | |
| "loss": 0.8735, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 13.71571072319202, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.8678, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 13.740648379052368, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 4.0358155717650895e-06, | |
| "loss": 0.8591, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 13.765586034912719, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 3.873830406168111e-06, | |
| "loss": 0.8646, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 13.790523690773068, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 3.715098754256241e-06, | |
| "loss": 0.8643, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 13.815461346633416, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 3.5596259885061102e-06, | |
| "loss": 0.8702, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 13.840399002493765, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.40741737109318e-06, | |
| "loss": 0.8669, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 13.865336658354115, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 3.2584780537136207e-06, | |
| "loss": 0.8818, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 13.890274314214464, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.1128130774099262e-06, | |
| "loss": 0.8771, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 13.915211970074813, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 2.970427372400353e-06, | |
| "loss": 0.8643, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 13.940149625935161, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.8313257579119843e-06, | |
| "loss": 0.8569, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 13.965087281795512, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 2.6955129420176196e-06, | |
| "loss": 0.8664, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 13.99002493765586, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 2.5629935214764865e-06, | |
| "loss": 0.8709, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 2.5418429374694824, | |
| "eval_runtime": 0.5336, | |
| "eval_samples_per_second": 18.742, | |
| "eval_steps_per_second": 1.874, | |
| "step": 2807 | |
| }, | |
| { | |
| "epoch": 14.01496259351621, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 2.433771981578581e-06, | |
| "loss": 0.8642, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 14.039900249376558, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 2.307852695992907e-06, | |
| "loss": 0.8614, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 14.064837905236908, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 2.1852399266194314e-06, | |
| "loss": 0.8539, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 14.089775561097257, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 2.0659378234448525e-06, | |
| "loss": 0.867, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 14.114713216957606, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.9499504244020693e-06, | |
| "loss": 0.8612, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 14.139650872817954, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 1.8372816552336026e-06, | |
| "loss": 0.8633, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 14.164588528678305, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 1.7279353293586765e-06, | |
| "loss": 0.8609, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 14.189526184538654, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 1.621915147744124e-06, | |
| "loss": 0.8658, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 14.214463840399002, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 1.5192246987791981e-06, | |
| "loss": 0.8608, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 14.239401496259351, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 1.4198674581540339e-06, | |
| "loss": 0.8689, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 14.264339152119701, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 1.323846788742078e-06, | |
| "loss": 0.8778, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 14.28927680798005, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 0.8685, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 14.314214463840399, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.14182805028884e-06, | |
| "loss": 0.8658, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 14.339152119700747, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 1.055836141905553e-06, | |
| "loss": 0.8771, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 14.364089775561098, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 9.731931258429638e-07, | |
| "loss": 0.8816, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 14.389027431421447, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 8.939017992601328e-07, | |
| "loss": 0.8699, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 14.413965087281795, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 8.17964845873831e-07, | |
| "loss": 0.8734, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 14.438902743142144, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 7.453848358678017e-07, | |
| "loss": 0.8724, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 14.463840399002494, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 6.761642258056978e-07, | |
| "loss": 0.8784, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 14.488778054862843, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 6.103053585480023e-07, | |
| "loss": 0.8644, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 14.513715710723192, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 5.478104631726711e-07, | |
| "loss": 0.8615, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 14.53865336658354, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 4.88681654899692e-07, | |
| "loss": 0.8662, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 14.563591022443891, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 4.329209350195651e-07, | |
| "loss": 0.8694, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 14.58852867830424, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.805301908254455e-07, | |
| "loss": 0.8729, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 14.613466334164588, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 3.315111955493944e-07, | |
| "loss": 0.8758, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 14.638403990024937, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 2.858656083022604e-07, | |
| "loss": 0.8569, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 14.663341645885287, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 2.4359497401758024e-07, | |
| "loss": 0.8734, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 14.688279301745636, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 2.0470072339926484e-07, | |
| "loss": 0.868, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 14.713216957605985, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 1.6918417287318245e-07, | |
| "loss": 0.8797, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 14.738154613466333, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 1.3704652454261668e-07, | |
| "loss": 0.8734, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 14.763092269326684, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 1.0828886614754341e-07, | |
| "loss": 0.88, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 14.788029925187033, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 8.29121710278713e-08, | |
| "loss": 0.8667, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 14.812967581047381, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 6.09172980904238e-08, | |
| "loss": 0.8642, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 14.83790523690773, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 4.230499177994007e-08, | |
| "loss": 0.8733, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 14.86284289276808, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 2.7075882053828605e-08, | |
| "loss": 0.8704, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 14.88778054862843, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 1.5230484360873044e-08, | |
| "loss": 0.8678, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 14.912718204488778, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 6.769199623779532e-09, | |
| "loss": 0.8684, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 14.937655860349127, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 1.6923142255764746e-09, | |
| "loss": 0.8586, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 14.962593516209477, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0, | |
| "loss": 0.8781, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.962593516209477, | |
| "eval_loss": 2.542728900909424, | |
| "eval_runtime": 0.4926, | |
| "eval_samples_per_second": 20.301, | |
| "eval_steps_per_second": 2.03, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.962593516209477, | |
| "step": 3000, | |
| "total_flos": 1.7621200159784305e+18, | |
| "train_loss": 0.9822600702444713, | |
| "train_runtime": 10311.9233, | |
| "train_samples_per_second": 13.994, | |
| "train_steps_per_second": 0.291 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 3000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 100, | |
| "total_flos": 1.7621200159784305e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |