| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9806451612903224, | |
| "eval_steps": 500, | |
| "global_step": 231, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012903225806451613, | |
| "grad_norm": 0.882150089808769, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.3191, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.025806451612903226, | |
| "grad_norm": 0.8369153094823952, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.249, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03870967741935484, | |
| "grad_norm": 0.8525103918091212, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.2775, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05161290322580645, | |
| "grad_norm": 0.8113130093304075, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.2577, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 0.7691226782403744, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 1.2275, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07741935483870968, | |
| "grad_norm": 0.5954210054804412, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1159, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09032258064516129, | |
| "grad_norm": 0.48189256930049384, | |
| "learning_rate": 5.833333333333334e-05, | |
| "loss": 1.0593, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1032258064516129, | |
| "grad_norm": 0.5241879927945232, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.0031, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11612903225806452, | |
| "grad_norm": 0.5751865259411146, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.9263, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 0.5686526755807603, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 0.8146, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14193548387096774, | |
| "grad_norm": 0.5156906474251192, | |
| "learning_rate": 9.166666666666667e-05, | |
| "loss": 0.7583, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15483870967741936, | |
| "grad_norm": 0.4901634328534619, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6686, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.16774193548387098, | |
| "grad_norm": 0.376084270046461, | |
| "learning_rate": 0.00010833333333333333, | |
| "loss": 0.6005, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.18064516129032257, | |
| "grad_norm": 0.2761318809240614, | |
| "learning_rate": 0.00011666666666666668, | |
| "loss": 0.5741, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 0.25038763704461725, | |
| "learning_rate": 0.000125, | |
| "loss": 0.5465, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2064516129032258, | |
| "grad_norm": 0.2214903977106201, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.5138, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.21935483870967742, | |
| "grad_norm": 0.28905541505099525, | |
| "learning_rate": 0.00014166666666666668, | |
| "loss": 0.5247, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23225806451612904, | |
| "grad_norm": 0.20699066633757193, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.4978, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.24516129032258063, | |
| "grad_norm": 0.219457528851344, | |
| "learning_rate": 0.00015833333333333332, | |
| "loss": 0.4924, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 0.16596853789220767, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 0.4759, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2709677419354839, | |
| "grad_norm": 0.13228412371333673, | |
| "learning_rate": 0.000175, | |
| "loss": 0.4613, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2838709677419355, | |
| "grad_norm": 0.1421107856190867, | |
| "learning_rate": 0.00018333333333333334, | |
| "loss": 0.4852, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2967741935483871, | |
| "grad_norm": 0.12552928984887968, | |
| "learning_rate": 0.00019166666666666667, | |
| "loss": 0.4786, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3096774193548387, | |
| "grad_norm": 0.11489463060846784, | |
| "learning_rate": 0.0002, | |
| "loss": 0.4532, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 0.11476879539402507, | |
| "learning_rate": 0.00019998848349441062, | |
| "loss": 0.4454, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.33548387096774196, | |
| "grad_norm": 0.1256602270101812, | |
| "learning_rate": 0.00019995393663024054, | |
| "loss": 0.4513, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.34838709677419355, | |
| "grad_norm": 0.11833482485698336, | |
| "learning_rate": 0.00019989636736467278, | |
| "loss": 0.44, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.36129032258064514, | |
| "grad_norm": 0.11124019681377781, | |
| "learning_rate": 0.00019981578895764273, | |
| "loss": 0.4439, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3741935483870968, | |
| "grad_norm": 0.10954971384477814, | |
| "learning_rate": 0.00019971221996878394, | |
| "loss": 0.4274, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 0.11422715129880294, | |
| "learning_rate": 0.00019958568425315314, | |
| "loss": 0.4254, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.11262310014016527, | |
| "learning_rate": 0.00019943621095573586, | |
| "loss": 0.4204, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4129032258064516, | |
| "grad_norm": 0.11143099554463408, | |
| "learning_rate": 0.00019926383450473344, | |
| "loss": 0.4105, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4258064516129032, | |
| "grad_norm": 0.1088260973247734, | |
| "learning_rate": 0.00019906859460363307, | |
| "loss": 0.4136, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.43870967741935485, | |
| "grad_norm": 0.10400753996611788, | |
| "learning_rate": 0.00019885053622206304, | |
| "loss": 0.4213, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 0.09587900896302251, | |
| "learning_rate": 0.0001986097095854347, | |
| "loss": 0.4085, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4645161290322581, | |
| "grad_norm": 0.10119603747308556, | |
| "learning_rate": 0.0001983461701633742, | |
| "loss": 0.4181, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4774193548387097, | |
| "grad_norm": 0.10062413136253176, | |
| "learning_rate": 0.00019805997865694614, | |
| "loss": 0.4098, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.49032258064516127, | |
| "grad_norm": 0.09162394941720846, | |
| "learning_rate": 0.0001977512009846721, | |
| "loss": 0.4085, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5032258064516129, | |
| "grad_norm": 0.09269316443279575, | |
| "learning_rate": 0.00019741990826734794, | |
| "loss": 0.3994, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 0.08782581803238095, | |
| "learning_rate": 0.00019706617681166218, | |
| "loss": 0.3983, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5290322580645161, | |
| "grad_norm": 0.08665646987756218, | |
| "learning_rate": 0.00019669008809262062, | |
| "loss": 0.3938, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5419354838709678, | |
| "grad_norm": 0.09289388957990503, | |
| "learning_rate": 0.00019629172873477995, | |
| "loss": 0.396, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5548387096774193, | |
| "grad_norm": 0.09203344649472522, | |
| "learning_rate": 0.00019587119049229557, | |
| "loss": 0.4052, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.567741935483871, | |
| "grad_norm": 0.08209774194723368, | |
| "learning_rate": 0.0001954285702277879, | |
| "loss": 0.3959, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 0.08595872863630391, | |
| "learning_rate": 0.00019496396989003193, | |
| "loss": 0.397, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5935483870967742, | |
| "grad_norm": 0.09041908237644536, | |
| "learning_rate": 0.00019447749649047542, | |
| "loss": 0.3992, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6064516129032258, | |
| "grad_norm": 0.08321976348844515, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.4095, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6193548387096774, | |
| "grad_norm": 0.07887604040253807, | |
| "learning_rate": 0.00019343938371606712, | |
| "loss": 0.3866, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.632258064516129, | |
| "grad_norm": 0.08329265943906447, | |
| "learning_rate": 0.00019288798344984672, | |
| "loss": 0.3985, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 0.08661703211305888, | |
| "learning_rate": 0.00019231518828401458, | |
| "loss": 0.3925, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6580645161290323, | |
| "grad_norm": 0.08382217550700771, | |
| "learning_rate": 0.00019172113015054532, | |
| "loss": 0.3862, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6709677419354839, | |
| "grad_norm": 0.08245124856491458, | |
| "learning_rate": 0.00019110594587891519, | |
| "loss": 0.3847, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6838709677419355, | |
| "grad_norm": 0.08319716279149986, | |
| "learning_rate": 0.00019046977716458626, | |
| "loss": 0.3775, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6967741935483871, | |
| "grad_norm": 0.08074648144423298, | |
| "learning_rate": 0.0001898127705363696, | |
| "loss": 0.3786, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 0.08472762376284584, | |
| "learning_rate": 0.0001891350773226754, | |
| "loss": 0.3923, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7225806451612903, | |
| "grad_norm": 0.08398076059437376, | |
| "learning_rate": 0.00018843685361665723, | |
| "loss": 0.3709, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7354838709677419, | |
| "grad_norm": 0.08465216102770419, | |
| "learning_rate": 0.00018771826024025946, | |
| "loss": 0.3818, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7483870967741936, | |
| "grad_norm": 0.09145572810056589, | |
| "learning_rate": 0.00018697946270717467, | |
| "loss": 0.39, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7612903225806451, | |
| "grad_norm": 0.08415188367023674, | |
| "learning_rate": 0.00018622063118472134, | |
| "loss": 0.3733, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 0.08576290382509591, | |
| "learning_rate": 0.00018544194045464886, | |
| "loss": 0.3878, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7870967741935484, | |
| "grad_norm": 0.0844142047859298, | |
| "learning_rate": 0.00018464356987288013, | |
| "loss": 0.3637, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.08918487261557899, | |
| "learning_rate": 0.00018382570332820043, | |
| "loss": 0.3775, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8129032258064516, | |
| "grad_norm": 0.0795181880669878, | |
| "learning_rate": 0.00018298852919990252, | |
| "loss": 0.3853, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8258064516129032, | |
| "grad_norm": 0.08173055996583302, | |
| "learning_rate": 0.0001821322403143969, | |
| "loss": 0.38, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 0.08525070031165603, | |
| "learning_rate": 0.0001812570339007983, | |
| "loss": 0.3778, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8516129032258064, | |
| "grad_norm": 0.08531235204546653, | |
| "learning_rate": 0.00018036311154549784, | |
| "loss": 0.3727, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.864516129032258, | |
| "grad_norm": 0.08169851479895494, | |
| "learning_rate": 0.00017945067914573146, | |
| "loss": 0.365, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8774193548387097, | |
| "grad_norm": 0.08463789046916101, | |
| "learning_rate": 0.0001785199468621559, | |
| "loss": 0.3752, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8903225806451613, | |
| "grad_norm": 0.09441843624235378, | |
| "learning_rate": 0.000177571129070442, | |
| "loss": 0.3665, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 0.08530939476149231, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.3926, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9161290322580645, | |
| "grad_norm": 0.0836606457284625, | |
| "learning_rate": 0.00017562011524313185, | |
| "loss": 0.3844, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9290322580645162, | |
| "grad_norm": 0.09868625782773943, | |
| "learning_rate": 0.00017461836858476856, | |
| "loss": 0.3835, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9419354838709677, | |
| "grad_norm": 0.082132336261239, | |
| "learning_rate": 0.00017359943506922774, | |
| "loss": 0.3792, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9548387096774194, | |
| "grad_norm": 0.08948965393301354, | |
| "learning_rate": 0.0001725635493875799, | |
| "loss": 0.3813, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 0.08539410389371488, | |
| "learning_rate": 0.00017151095013548994, | |
| "loss": 0.3774, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9806451612903225, | |
| "grad_norm": 0.08690404790165682, | |
| "learning_rate": 0.00017044187975826124, | |
| "loss": 0.3762, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "grad_norm": 0.09039522496805455, | |
| "learning_rate": 0.0001693565844949933, | |
| "loss": 0.3733, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9935483870967742, | |
| "eval_loss": 0.3743511736392975, | |
| "eval_runtime": 42.1339, | |
| "eval_samples_per_second": 24.66, | |
| "eval_steps_per_second": 0.783, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.0064516129032257, | |
| "grad_norm": 0.09165665911792642, | |
| "learning_rate": 0.00016825531432186543, | |
| "loss": 0.3532, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0193548387096774, | |
| "grad_norm": 0.0801922544260219, | |
| "learning_rate": 0.0001671383228945597, | |
| "loss": 0.347, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.032258064516129, | |
| "grad_norm": 0.08352186065175837, | |
| "learning_rate": 0.00016600586748983641, | |
| "loss": 0.3566, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0451612903225806, | |
| "grad_norm": 0.08793176795367076, | |
| "learning_rate": 0.0001648582089462756, | |
| "loss": 0.3473, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0580645161290323, | |
| "grad_norm": 0.08913951531063671, | |
| "learning_rate": 0.00016369561160419784, | |
| "loss": 0.342, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.070967741935484, | |
| "grad_norm": 0.08309712335786672, | |
| "learning_rate": 0.0001625183432447789, | |
| "loss": 0.345, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0838709677419356, | |
| "grad_norm": 0.08725330804483407, | |
| "learning_rate": 0.00016132667502837165, | |
| "loss": 0.3523, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.096774193548387, | |
| "grad_norm": 0.08680862762413778, | |
| "learning_rate": 0.00016012088143204953, | |
| "loss": 0.3554, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1096774193548387, | |
| "grad_norm": 0.0863782848559528, | |
| "learning_rate": 0.00015890124018638638, | |
| "loss": 0.364, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1225806451612903, | |
| "grad_norm": 0.08388848992116194, | |
| "learning_rate": 0.00015766803221148673, | |
| "loss": 0.3568, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.135483870967742, | |
| "grad_norm": 0.08226994751114965, | |
| "learning_rate": 0.00015642154155228122, | |
| "loss": 0.3489, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1483870967741936, | |
| "grad_norm": 0.08575965994905438, | |
| "learning_rate": 0.00015516205531310273, | |
| "loss": 0.3466, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1612903225806452, | |
| "grad_norm": 0.0895747440427046, | |
| "learning_rate": 0.00015388986359155758, | |
| "loss": 0.3488, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1741935483870969, | |
| "grad_norm": 0.08403222320010312, | |
| "learning_rate": 0.00015260525941170712, | |
| "loss": 0.356, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1870967741935483, | |
| "grad_norm": 0.08627434364043794, | |
| "learning_rate": 0.0001513085386565758, | |
| "loss": 0.3519, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.08925414655300028, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.3523, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2129032258064516, | |
| "grad_norm": 0.09120079741968923, | |
| "learning_rate": 0.00014867994483783485, | |
| "loss": 0.3555, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.2258064516129032, | |
| "grad_norm": 0.08519037826685563, | |
| "learning_rate": 0.0001473486772185334, | |
| "loss": 0.3551, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.238709677419355, | |
| "grad_norm": 0.08814591743170447, | |
| "learning_rate": 0.00014600650377311522, | |
| "loss": 0.3535, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2516129032258063, | |
| "grad_norm": 0.08812877093082108, | |
| "learning_rate": 0.00014465373364454001, | |
| "loss": 0.3498, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2645161290322582, | |
| "grad_norm": 0.08596197743921638, | |
| "learning_rate": 0.00014329067841650274, | |
| "loss": 0.3484, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2774193548387096, | |
| "grad_norm": 0.09025513346881896, | |
| "learning_rate": 0.00014191765204166643, | |
| "loss": 0.3465, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.2903225806451613, | |
| "grad_norm": 0.08665409616008209, | |
| "learning_rate": 0.00014053497076934948, | |
| "loss": 0.35, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.303225806451613, | |
| "grad_norm": 0.09012608398761074, | |
| "learning_rate": 0.00013914295307268396, | |
| "loss": 0.3516, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3161290322580645, | |
| "grad_norm": 0.09456407877563842, | |
| "learning_rate": 0.00013774191957526143, | |
| "loss": 0.3639, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3290322580645162, | |
| "grad_norm": 0.0888376260234129, | |
| "learning_rate": 0.00013633219297728416, | |
| "loss": 0.3396, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3419354838709676, | |
| "grad_norm": 0.08652600639054038, | |
| "learning_rate": 0.00013491409798123687, | |
| "loss": 0.3445, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3548387096774195, | |
| "grad_norm": 0.09269194410505097, | |
| "learning_rate": 0.00013348796121709862, | |
| "loss": 0.3555, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.367741935483871, | |
| "grad_norm": 0.09421096011594207, | |
| "learning_rate": 0.00013205411116710972, | |
| "loss": 0.3508, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3806451612903226, | |
| "grad_norm": 0.09286783444235318, | |
| "learning_rate": 0.00013061287809011242, | |
| "loss": 0.3571, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3935483870967742, | |
| "grad_norm": 0.08172852976047028, | |
| "learning_rate": 0.0001291645939454825, | |
| "loss": 0.3488, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.4064516129032258, | |
| "grad_norm": 0.09033973727962885, | |
| "learning_rate": 0.0001277095923166689, | |
| "loss": 0.3498, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.4193548387096775, | |
| "grad_norm": 0.09628933362833343, | |
| "learning_rate": 0.00012624820833435937, | |
| "loss": 0.3472, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.432258064516129, | |
| "grad_norm": 0.08471497514674803, | |
| "learning_rate": 0.00012478077859929, | |
| "loss": 0.3353, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4451612903225808, | |
| "grad_norm": 0.08976133324522119, | |
| "learning_rate": 0.00012330764110471566, | |
| "loss": 0.3468, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4580645161290322, | |
| "grad_norm": 0.09634877556737409, | |
| "learning_rate": 0.00012182913515856015, | |
| "loss": 0.3541, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4709677419354839, | |
| "grad_norm": 0.09348923296138459, | |
| "learning_rate": 0.0001203456013052634, | |
| "loss": 0.3521, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.4838709677419355, | |
| "grad_norm": 0.09437711091684706, | |
| "learning_rate": 0.00011885738124734358, | |
| "loss": 0.3566, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4967741935483871, | |
| "grad_norm": 0.08916702937111011, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.3458, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.5096774193548388, | |
| "grad_norm": 0.09100601467580355, | |
| "learning_rate": 0.00011586825464562514, | |
| "loss": 0.3593, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5225806451612902, | |
| "grad_norm": 0.08990470683690902, | |
| "learning_rate": 0.00011436803658769082, | |
| "loss": 0.3434, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.535483870967742, | |
| "grad_norm": 0.0932653393737011, | |
| "learning_rate": 0.00011286450913828312, | |
| "loss": 0.342, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.5483870967741935, | |
| "grad_norm": 0.08960531773257623, | |
| "learning_rate": 0.00011135801860504749, | |
| "loss": 0.3628, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5612903225806452, | |
| "grad_norm": 0.09275069273094473, | |
| "learning_rate": 0.00010984891197811687, | |
| "loss": 0.3513, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5741935483870968, | |
| "grad_norm": 0.09527469311088294, | |
| "learning_rate": 0.00010833753685018935, | |
| "loss": 0.3556, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5870967741935482, | |
| "grad_norm": 0.09323849659154124, | |
| "learning_rate": 0.0001068242413364671, | |
| "loss": 0.3448, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.08474554028292876, | |
| "learning_rate": 0.00010530937399447496, | |
| "loss": 0.3499, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6129032258064515, | |
| "grad_norm": 0.09382059811382143, | |
| "learning_rate": 0.00010379328374377715, | |
| "loss": 0.3384, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6258064516129034, | |
| "grad_norm": 0.09276702527842776, | |
| "learning_rate": 0.00010227631978561056, | |
| "loss": 0.3444, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6387096774193548, | |
| "grad_norm": 0.08750152088472078, | |
| "learning_rate": 0.00010075883152245334, | |
| "loss": 0.3569, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6516129032258065, | |
| "grad_norm": 0.08714445180642569, | |
| "learning_rate": 9.92411684775467e-05, | |
| "loss": 0.342, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.664516129032258, | |
| "grad_norm": 0.08469902272466831, | |
| "learning_rate": 9.772368021438943e-05, | |
| "loss": 0.3342, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6774193548387095, | |
| "grad_norm": 0.08724585745005611, | |
| "learning_rate": 9.620671625622288e-05, | |
| "loss": 0.3335, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6903225806451614, | |
| "grad_norm": 0.09087336723016343, | |
| "learning_rate": 9.469062600552509e-05, | |
| "loss": 0.3447, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.7032258064516128, | |
| "grad_norm": 0.08863278083042062, | |
| "learning_rate": 9.317575866353292e-05, | |
| "loss": 0.3487, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7161290322580647, | |
| "grad_norm": 0.08343459715762, | |
| "learning_rate": 9.166246314981066e-05, | |
| "loss": 0.3454, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.729032258064516, | |
| "grad_norm": 0.08837483796029806, | |
| "learning_rate": 9.015108802188313e-05, | |
| "loss": 0.3484, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.7419354838709677, | |
| "grad_norm": 0.08762249376974672, | |
| "learning_rate": 8.86419813949525e-05, | |
| "loss": 0.3447, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7548387096774194, | |
| "grad_norm": 0.08446853010895118, | |
| "learning_rate": 8.713549086171691e-05, | |
| "loss": 0.3466, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7677419354838708, | |
| "grad_norm": 0.08897676787603495, | |
| "learning_rate": 8.563196341230919e-05, | |
| "loss": 0.3434, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7806451612903227, | |
| "grad_norm": 0.09210810174866911, | |
| "learning_rate": 8.413174535437487e-05, | |
| "loss": 0.355, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.793548387096774, | |
| "grad_norm": 0.0877098792555575, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.3392, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.8064516129032258, | |
| "grad_norm": 0.09059259587839792, | |
| "learning_rate": 8.114261875265643e-05, | |
| "loss": 0.3465, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8193548387096774, | |
| "grad_norm": 0.09043152099082513, | |
| "learning_rate": 7.965439869473664e-05, | |
| "loss": 0.3409, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.832258064516129, | |
| "grad_norm": 0.08863483273837267, | |
| "learning_rate": 7.817086484143986e-05, | |
| "loss": 0.3497, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8451612903225807, | |
| "grad_norm": 0.08351509862847174, | |
| "learning_rate": 7.669235889528436e-05, | |
| "loss": 0.3484, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8580645161290321, | |
| "grad_norm": 0.08881689002413959, | |
| "learning_rate": 7.521922140071002e-05, | |
| "loss": 0.3428, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.870967741935484, | |
| "grad_norm": 0.08962413300366581, | |
| "learning_rate": 7.375179166564063e-05, | |
| "loss": 0.3353, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8838709677419354, | |
| "grad_norm": 0.08991947191225944, | |
| "learning_rate": 7.229040768333115e-05, | |
| "loss": 0.3366, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.896774193548387, | |
| "grad_norm": 0.0890545628104281, | |
| "learning_rate": 7.08354060545175e-05, | |
| "loss": 0.3381, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.9096774193548387, | |
| "grad_norm": 0.09306016588414409, | |
| "learning_rate": 6.93871219098876e-05, | |
| "loss": 0.3356, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9225806451612903, | |
| "grad_norm": 0.08816048934545212, | |
| "learning_rate": 6.79458888328903e-05, | |
| "loss": 0.3412, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.935483870967742, | |
| "grad_norm": 0.09006593042575502, | |
| "learning_rate": 6.651203878290139e-05, | |
| "loss": 0.3471, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9483870967741934, | |
| "grad_norm": 0.08499237638300171, | |
| "learning_rate": 6.508590201876317e-05, | |
| "loss": 0.335, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9612903225806453, | |
| "grad_norm": 0.09566747308379261, | |
| "learning_rate": 6.366780702271589e-05, | |
| "loss": 0.3395, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9741935483870967, | |
| "grad_norm": 0.0915253754596643, | |
| "learning_rate": 6.225808042473858e-05, | |
| "loss": 0.3488, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9870967741935484, | |
| "grad_norm": 0.08657357278603872, | |
| "learning_rate": 6.085704692731609e-05, | |
| "loss": 0.3344, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.08950726731743963, | |
| "learning_rate": 5.9465029230650534e-05, | |
| "loss": 0.33, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.35439133644104004, | |
| "eval_runtime": 36.1469, | |
| "eval_samples_per_second": 28.744, | |
| "eval_steps_per_second": 0.913, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.0129032258064514, | |
| "grad_norm": 0.08961232668946545, | |
| "learning_rate": 5.8082347958333625e-05, | |
| "loss": 0.3273, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0258064516129033, | |
| "grad_norm": 0.09402916213349197, | |
| "learning_rate": 5.670932158349731e-05, | |
| "loss": 0.3218, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0387096774193547, | |
| "grad_norm": 0.08520247695821515, | |
| "learning_rate": 5.5346266355459995e-05, | |
| "loss": 0.3089, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0516129032258066, | |
| "grad_norm": 0.08637288183919145, | |
| "learning_rate": 5.399349622688479e-05, | |
| "loss": 0.3266, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.064516129032258, | |
| "grad_norm": 0.08823864345930746, | |
| "learning_rate": 5.26513227814666e-05, | |
| "loss": 0.329, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.07741935483871, | |
| "grad_norm": 0.09384371931382793, | |
| "learning_rate": 5.1320055162165115e-05, | |
| "loss": 0.3275, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0903225806451613, | |
| "grad_norm": 0.09516405744887674, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.332, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.1032258064516127, | |
| "grad_norm": 0.08966279182804247, | |
| "learning_rate": 4.869146134342426e-05, | |
| "loss": 0.3247, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.1161290322580646, | |
| "grad_norm": 0.08700940402163973, | |
| "learning_rate": 4.739474058829289e-05, | |
| "loss": 0.3221, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.129032258064516, | |
| "grad_norm": 0.08984677102800173, | |
| "learning_rate": 4.611013640844245e-05, | |
| "loss": 0.3272, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.141935483870968, | |
| "grad_norm": 0.08964202186304891, | |
| "learning_rate": 4.483794468689728e-05, | |
| "loss": 0.3188, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1548387096774193, | |
| "grad_norm": 0.09997697429798251, | |
| "learning_rate": 4.357845844771881e-05, | |
| "loss": 0.3383, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.167741935483871, | |
| "grad_norm": 0.09510073376177604, | |
| "learning_rate": 4.2331967788513295e-05, | |
| "loss": 0.3252, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1806451612903226, | |
| "grad_norm": 0.09107612709336496, | |
| "learning_rate": 4.109875981361363e-05, | |
| "loss": 0.3217, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.193548387096774, | |
| "grad_norm": 0.08804927379783276, | |
| "learning_rate": 3.987911856795047e-05, | |
| "loss": 0.3173, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.206451612903226, | |
| "grad_norm": 0.0916081059987062, | |
| "learning_rate": 3.8673324971628357e-05, | |
| "loss": 0.3285, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2193548387096773, | |
| "grad_norm": 0.09226628432750343, | |
| "learning_rate": 3.7481656755221125e-05, | |
| "loss": 0.3154, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.232258064516129, | |
| "grad_norm": 0.09145015878266409, | |
| "learning_rate": 3.630438839580217e-05, | |
| "loss": 0.3087, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2451612903225806, | |
| "grad_norm": 0.08786201399591659, | |
| "learning_rate": 3.5141791053724405e-05, | |
| "loss": 0.3151, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.258064516129032, | |
| "grad_norm": 0.09259402512083086, | |
| "learning_rate": 3.399413251016359e-05, | |
| "loss": 0.3369, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.270967741935484, | |
| "grad_norm": 0.09311260751337232, | |
| "learning_rate": 3.2861677105440336e-05, | |
| "loss": 0.3051, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2838709677419353, | |
| "grad_norm": 0.09217712904693832, | |
| "learning_rate": 3.174468567813461e-05, | |
| "loss": 0.3199, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.296774193548387, | |
| "grad_norm": 0.09141877592974519, | |
| "learning_rate": 3.0643415505006735e-05, | |
| "loss": 0.3229, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.3096774193548386, | |
| "grad_norm": 0.09528833689903496, | |
| "learning_rate": 2.9558120241738784e-05, | |
| "loss": 0.3286, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.3225806451612905, | |
| "grad_norm": 0.09070636787107308, | |
| "learning_rate": 2.8489049864510054e-05, | |
| "loss": 0.3348, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.335483870967742, | |
| "grad_norm": 0.09307512327341362, | |
| "learning_rate": 2.7436450612420095e-05, | |
| "loss": 0.3256, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3483870967741938, | |
| "grad_norm": 0.09127823479306682, | |
| "learning_rate": 2.640056493077231e-05, | |
| "loss": 0.3181, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.361290322580645, | |
| "grad_norm": 0.09246009256113925, | |
| "learning_rate": 2.5381631415231454e-05, | |
| "loss": 0.3391, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3741935483870966, | |
| "grad_norm": 0.09095352379758655, | |
| "learning_rate": 2.4379884756868167e-05, | |
| "loss": 0.3172, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.3870967741935485, | |
| "grad_norm": 0.0926880163626768, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.3177, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.09094474131194094, | |
| "learning_rate": 2.242887092955801e-05, | |
| "loss": 0.3199, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.412903225806452, | |
| "grad_norm": 0.09106546035353981, | |
| "learning_rate": 2.1480053137844115e-05, | |
| "loss": 0.3222, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.425806451612903, | |
| "grad_norm": 0.08873018715134598, | |
| "learning_rate": 2.054932085426856e-05, | |
| "loss": 0.3118, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4387096774193546, | |
| "grad_norm": 0.0932765377498955, | |
| "learning_rate": 1.9636888454502178e-05, | |
| "loss": 0.3358, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.4516129032258065, | |
| "grad_norm": 0.09181586534157822, | |
| "learning_rate": 1.8742966099201697e-05, | |
| "loss": 0.3157, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.464516129032258, | |
| "grad_norm": 0.0929486436457203, | |
| "learning_rate": 1.7867759685603114e-05, | |
| "loss": 0.3154, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.47741935483871, | |
| "grad_norm": 0.09188630220285351, | |
| "learning_rate": 1.7011470800097496e-05, | |
| "loss": 0.3181, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.490322580645161, | |
| "grad_norm": 0.09574286894431329, | |
| "learning_rate": 1.6174296671799572e-05, | |
| "loss": 0.3222, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.5032258064516126, | |
| "grad_norm": 0.09145354457132104, | |
| "learning_rate": 1.5356430127119913e-05, | |
| "loss": 0.3222, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.5161290322580645, | |
| "grad_norm": 0.09039580690260736, | |
| "learning_rate": 1.4558059545351143e-05, | |
| "loss": 0.324, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5290322580645164, | |
| "grad_norm": 0.08979381831653434, | |
| "learning_rate": 1.3779368815278647e-05, | |
| "loss": 0.3107, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.541935483870968, | |
| "grad_norm": 0.09526292697431937, | |
| "learning_rate": 1.302053729282533e-05, | |
| "loss": 0.3219, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.554838709677419, | |
| "grad_norm": 0.09310358146453943, | |
| "learning_rate": 1.2281739759740574e-05, | |
| "loss": 0.3214, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.567741935483871, | |
| "grad_norm": 0.09212645063531479, | |
| "learning_rate": 1.1563146383342772e-05, | |
| "loss": 0.3154, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.5806451612903225, | |
| "grad_norm": 0.09533681862557382, | |
| "learning_rate": 1.0864922677324618e-05, | |
| "loss": 0.319, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5935483870967744, | |
| "grad_norm": 0.09551418366783314, | |
| "learning_rate": 1.01872294636304e-05, | |
| "loss": 0.3333, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.606451612903226, | |
| "grad_norm": 0.08930212325894361, | |
| "learning_rate": 9.530222835413738e-06, | |
| "loss": 0.3048, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.6193548387096772, | |
| "grad_norm": 0.09220378121771236, | |
| "learning_rate": 8.894054121084838e-06, | |
| "loss": 0.3146, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.632258064516129, | |
| "grad_norm": 0.09150774720724307, | |
| "learning_rate": 8.278869849454718e-06, | |
| "loss": 0.3311, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.6451612903225805, | |
| "grad_norm": 0.09261513270619316, | |
| "learning_rate": 7.684811715985429e-06, | |
| "loss": 0.3172, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6580645161290324, | |
| "grad_norm": 0.0941004102909483, | |
| "learning_rate": 7.1120165501533e-06, | |
| "loss": 0.3347, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.670967741935484, | |
| "grad_norm": 0.08707518610128166, | |
| "learning_rate": 6.560616283932897e-06, | |
| "loss": 0.3116, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6838709677419352, | |
| "grad_norm": 0.08648707636296159, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.3144, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.696774193548387, | |
| "grad_norm": 0.09169150101119816, | |
| "learning_rate": 5.52250350952459e-06, | |
| "loss": 0.3255, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.709677419354839, | |
| "grad_norm": 0.09060072523264334, | |
| "learning_rate": 5.036030109968082e-06, | |
| "loss": 0.3183, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7225806451612904, | |
| "grad_norm": 0.09077216490604942, | |
| "learning_rate": 4.5714297722121106e-06, | |
| "loss": 0.321, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.735483870967742, | |
| "grad_norm": 0.09088968433443333, | |
| "learning_rate": 4.128809507704445e-06, | |
| "loss": 0.3172, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7483870967741937, | |
| "grad_norm": 0.09191902683388614, | |
| "learning_rate": 3.7082712652200867e-06, | |
| "loss": 0.3261, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.761290322580645, | |
| "grad_norm": 0.08843215800144302, | |
| "learning_rate": 3.3099119073793928e-06, | |
| "loss": 0.3158, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.774193548387097, | |
| "grad_norm": 0.09079938334868655, | |
| "learning_rate": 2.9338231883378366e-06, | |
| "loss": 0.3178, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7870967741935484, | |
| "grad_norm": 0.09122789808454786, | |
| "learning_rate": 2.580091732652101e-06, | |
| "loss": 0.3282, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.09380292374109117, | |
| "learning_rate": 2.248799015327907e-06, | |
| "loss": 0.3359, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.8129032258064517, | |
| "grad_norm": 0.09035917420929797, | |
| "learning_rate": 1.9400213430538773e-06, | |
| "loss": 0.3169, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.825806451612903, | |
| "grad_norm": 0.09195121657817087, | |
| "learning_rate": 1.6538298366257976e-06, | |
| "loss": 0.3314, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.838709677419355, | |
| "grad_norm": 0.09166102367139951, | |
| "learning_rate": 1.3902904145653096e-06, | |
| "loss": 0.3258, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8516129032258064, | |
| "grad_norm": 0.0921992572010057, | |
| "learning_rate": 1.1494637779369766e-06, | |
| "loss": 0.3298, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.864516129032258, | |
| "grad_norm": 0.09068261067988724, | |
| "learning_rate": 9.314053963669245e-07, | |
| "loss": 0.3214, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8774193548387097, | |
| "grad_norm": 0.09417924199778298, | |
| "learning_rate": 7.361654952665609e-07, | |
| "loss": 0.3134, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8903225806451616, | |
| "grad_norm": 0.0901765977296441, | |
| "learning_rate": 5.637890442641402e-07, | |
| "loss": 0.3221, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.903225806451613, | |
| "grad_norm": 0.09094506589085496, | |
| "learning_rate": 4.143157468468717e-07, | |
| "loss": 0.3128, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.9161290322580644, | |
| "grad_norm": 0.08772549933058231, | |
| "learning_rate": 2.877800312160783e-07, | |
| "loss": 0.3248, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9290322580645163, | |
| "grad_norm": 0.09191883931659987, | |
| "learning_rate": 1.8421104235727405e-07, | |
| "loss": 0.3114, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9419354838709677, | |
| "grad_norm": 0.08876137430429, | |
| "learning_rate": 1.0363263532724432e-07, | |
| "loss": 0.3127, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9548387096774196, | |
| "grad_norm": 0.09157045134043748, | |
| "learning_rate": 4.606336975948589e-08, | |
| "loss": 0.3275, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.967741935483871, | |
| "grad_norm": 0.08940213355520302, | |
| "learning_rate": 1.1516505589381776e-08, | |
| "loss": 0.3246, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9806451612903224, | |
| "grad_norm": 0.0895898052255747, | |
| "learning_rate": 0.0, | |
| "loss": 0.3079, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9806451612903224, | |
| "eval_loss": 0.3507891595363617, | |
| "eval_runtime": 36.0777, | |
| "eval_samples_per_second": 28.799, | |
| "eval_steps_per_second": 0.915, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9806451612903224, | |
| "step": 231, | |
| "total_flos": 9.324729662937498e+16, | |
| "train_loss": 0.3951803825118325, | |
| "train_runtime": 2997.4381, | |
| "train_samples_per_second": 9.871, | |
| "train_steps_per_second": 0.077 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 231, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.324729662937498e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |