| { | |
| "best_metric": 0.02237132005393505, | |
| "best_model_checkpoint": "hurricane_model/checkpoint-1000", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 1252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03194888178913738, | |
| "grad_norm": 1.788046956062317, | |
| "learning_rate": 0.00019840255591054313, | |
| "loss": 0.5871, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06389776357827476, | |
| "grad_norm": 1.5475027561187744, | |
| "learning_rate": 0.00019680511182108628, | |
| "loss": 0.283, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 6.500663757324219, | |
| "learning_rate": 0.0001952076677316294, | |
| "loss": 0.1913, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12779552715654952, | |
| "grad_norm": 11.902435302734375, | |
| "learning_rate": 0.00019361022364217253, | |
| "loss": 0.2797, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1597444089456869, | |
| "grad_norm": 0.10494810342788696, | |
| "learning_rate": 0.00019201277955271565, | |
| "loss": 0.1443, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 3.824563980102539, | |
| "learning_rate": 0.0001904153354632588, | |
| "loss": 0.1807, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22364217252396165, | |
| "grad_norm": 5.14697790145874, | |
| "learning_rate": 0.00018881789137380192, | |
| "loss": 0.2322, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25559105431309903, | |
| "grad_norm": 1.0805331468582153, | |
| "learning_rate": 0.00018722044728434505, | |
| "loss": 0.1309, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 0.8772349953651428, | |
| "learning_rate": 0.0001856230031948882, | |
| "loss": 0.2684, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "grad_norm": 0.09347983449697495, | |
| "learning_rate": 0.00018402555910543132, | |
| "loss": 0.1118, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "eval_accuracy": 0.9476, | |
| "eval_loss": 0.14858174324035645, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 14.0146, | |
| "eval_samples_per_second": 178.385, | |
| "eval_steps_per_second": 22.334, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3514376996805112, | |
| "grad_norm": 0.8523539304733276, | |
| "learning_rate": 0.00018242811501597444, | |
| "loss": 0.1224, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 26.86109733581543, | |
| "learning_rate": 0.00018083067092651756, | |
| "loss": 0.1597, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.41533546325878595, | |
| "grad_norm": 0.07714489102363586, | |
| "learning_rate": 0.00017923322683706071, | |
| "loss": 0.3385, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4472843450479233, | |
| "grad_norm": 0.46879705786705017, | |
| "learning_rate": 0.00017763578274760384, | |
| "loss": 0.1338, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.19195349514484406, | |
| "learning_rate": 0.000176038338658147, | |
| "loss": 0.1792, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5111821086261981, | |
| "grad_norm": 1.775911569595337, | |
| "learning_rate": 0.0001744408945686901, | |
| "loss": 0.142, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5431309904153354, | |
| "grad_norm": 3.7460222244262695, | |
| "learning_rate": 0.00017284345047923323, | |
| "loss": 0.2113, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 1.6586716175079346, | |
| "learning_rate": 0.00017124600638977638, | |
| "loss": 0.2379, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6070287539936102, | |
| "grad_norm": 0.37085050344467163, | |
| "learning_rate": 0.00016964856230031948, | |
| "loss": 0.0626, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "grad_norm": 1.5378907918930054, | |
| "learning_rate": 0.00016805111821086263, | |
| "loss": 0.1112, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "eval_accuracy": 0.9752, | |
| "eval_loss": 0.0701194703578949, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.3233, | |
| "eval_samples_per_second": 202.868, | |
| "eval_steps_per_second": 25.399, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.3154517114162445, | |
| "learning_rate": 0.00016645367412140575, | |
| "loss": 0.051, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7028753993610224, | |
| "grad_norm": 0.05547064542770386, | |
| "learning_rate": 0.0001648562300319489, | |
| "loss": 0.0512, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7348242811501597, | |
| "grad_norm": 0.0428142175078392, | |
| "learning_rate": 0.00016325878594249202, | |
| "loss": 0.1203, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.07227767258882523, | |
| "learning_rate": 0.00016166134185303515, | |
| "loss": 0.1993, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7987220447284346, | |
| "grad_norm": 4.76859188079834, | |
| "learning_rate": 0.0001600638977635783, | |
| "loss": 0.0644, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8306709265175719, | |
| "grad_norm": 0.06700027734041214, | |
| "learning_rate": 0.00015846645367412142, | |
| "loss": 0.0898, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.14176583290100098, | |
| "learning_rate": 0.00015686900958466454, | |
| "loss": 0.1146, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8945686900958466, | |
| "grad_norm": 0.23704954981803894, | |
| "learning_rate": 0.00015527156549520767, | |
| "loss": 0.1933, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9265175718849841, | |
| "grad_norm": 0.32876431941986084, | |
| "learning_rate": 0.00015367412140575082, | |
| "loss": 0.1404, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.569324791431427, | |
| "learning_rate": 0.00015207667731629394, | |
| "loss": 0.0694, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "eval_accuracy": 0.9808, | |
| "eval_loss": 0.06084931641817093, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.4406, | |
| "eval_samples_per_second": 200.956, | |
| "eval_steps_per_second": 25.16, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9904153354632588, | |
| "grad_norm": 2.446889877319336, | |
| "learning_rate": 0.00015047923322683706, | |
| "loss": 0.0315, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0223642172523961, | |
| "grad_norm": 0.13555611670017242, | |
| "learning_rate": 0.0001488817891373802, | |
| "loss": 0.0476, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.076418437063694, | |
| "learning_rate": 0.00014728434504792333, | |
| "loss": 0.0388, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0862619808306708, | |
| "grad_norm": 0.034216273576021194, | |
| "learning_rate": 0.00014568690095846646, | |
| "loss": 0.0514, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1182108626198084, | |
| "grad_norm": 0.8253958821296692, | |
| "learning_rate": 0.00014408945686900958, | |
| "loss": 0.0734, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.5345448851585388, | |
| "learning_rate": 0.00014249201277955273, | |
| "loss": 0.0159, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.182108626198083, | |
| "grad_norm": 0.031145131215453148, | |
| "learning_rate": 0.00014089456869009585, | |
| "loss": 0.1411, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2140575079872205, | |
| "grad_norm": 2.8567800521850586, | |
| "learning_rate": 0.000139297124600639, | |
| "loss": 0.0221, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.06364869326353073, | |
| "learning_rate": 0.00013769968051118212, | |
| "loss": 0.0326, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "grad_norm": 0.024139605462551117, | |
| "learning_rate": 0.00013610223642172525, | |
| "loss": 0.0048, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "eval_accuracy": 0.9744, | |
| "eval_loss": 0.09166774898767471, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.5348, | |
| "eval_samples_per_second": 199.445, | |
| "eval_steps_per_second": 24.971, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3099041533546325, | |
| "grad_norm": 0.061418768018484116, | |
| "learning_rate": 0.00013450479233226837, | |
| "loss": 0.0336, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.04032498970627785, | |
| "learning_rate": 0.0001329073482428115, | |
| "loss": 0.0272, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3738019169329074, | |
| "grad_norm": 0.05345893278717995, | |
| "learning_rate": 0.00013130990415335464, | |
| "loss": 0.0232, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4057507987220448, | |
| "grad_norm": 0.1362222582101822, | |
| "learning_rate": 0.00012971246006389777, | |
| "loss": 0.0234, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 2.3406131267547607, | |
| "learning_rate": 0.00012811501597444092, | |
| "loss": 0.0339, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4696485623003195, | |
| "grad_norm": 0.01726055145263672, | |
| "learning_rate": 0.00012651757188498404, | |
| "loss": 0.0329, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5015974440894568, | |
| "grad_norm": 0.021831806749105453, | |
| "learning_rate": 0.00012492012779552716, | |
| "loss": 0.0539, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.01614479534327984, | |
| "learning_rate": 0.00012332268370607028, | |
| "loss": 0.071, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5654952076677318, | |
| "grad_norm": 0.2092825472354889, | |
| "learning_rate": 0.00012172523961661342, | |
| "loss": 0.0503, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "grad_norm": 0.02242388017475605, | |
| "learning_rate": 0.00012012779552715656, | |
| "loss": 0.036, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "eval_accuracy": 0.9836, | |
| "eval_loss": 0.05519802123308182, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.5164, | |
| "eval_samples_per_second": 199.738, | |
| "eval_steps_per_second": 25.007, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 1.8705270290374756, | |
| "learning_rate": 0.00011853035143769968, | |
| "loss": 0.0691, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6613418530351438, | |
| "grad_norm": 0.12965114414691925, | |
| "learning_rate": 0.00011693290734824283, | |
| "loss": 0.0254, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6932907348242812, | |
| "grad_norm": 0.2740132212638855, | |
| "learning_rate": 0.00011533546325878595, | |
| "loss": 0.0508, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.01778402552008629, | |
| "learning_rate": 0.00011373801916932908, | |
| "loss": 0.0145, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7571884984025559, | |
| "grad_norm": 0.014663617126643658, | |
| "learning_rate": 0.00011214057507987221, | |
| "loss": 0.0315, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7891373801916934, | |
| "grad_norm": 0.015493140555918217, | |
| "learning_rate": 0.00011054313099041533, | |
| "loss": 0.1852, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.05500736087560654, | |
| "learning_rate": 0.00010894568690095847, | |
| "loss": 0.0292, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8530351437699681, | |
| "grad_norm": 0.014051590114831924, | |
| "learning_rate": 0.0001073482428115016, | |
| "loss": 0.0193, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.8849840255591053, | |
| "grad_norm": 0.050858091562986374, | |
| "learning_rate": 0.00010575079872204474, | |
| "loss": 0.0825, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.08435127884149551, | |
| "learning_rate": 0.00010415335463258787, | |
| "loss": 0.0594, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "eval_accuracy": 0.9808, | |
| "eval_loss": 0.05469144880771637, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.5149, | |
| "eval_samples_per_second": 199.761, | |
| "eval_steps_per_second": 25.01, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9488817891373802, | |
| "grad_norm": 0.016737263649702072, | |
| "learning_rate": 0.000102555910543131, | |
| "loss": 0.0463, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9808306709265175, | |
| "grad_norm": 0.018503542989492416, | |
| "learning_rate": 0.00010095846645367413, | |
| "loss": 0.0213, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.11741246283054352, | |
| "learning_rate": 9.936102236421726e-05, | |
| "loss": 0.006, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0447284345047922, | |
| "grad_norm": 0.015154512599110603, | |
| "learning_rate": 9.77635782747604e-05, | |
| "loss": 0.003, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.07667731629393, | |
| "grad_norm": 0.017609527334570885, | |
| "learning_rate": 9.616613418530351e-05, | |
| "loss": 0.0017, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.012800313532352448, | |
| "learning_rate": 9.456869009584664e-05, | |
| "loss": 0.0042, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1405750798722045, | |
| "grad_norm": 1.593684434890747, | |
| "learning_rate": 9.297124600638978e-05, | |
| "loss": 0.0132, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.1725239616613417, | |
| "grad_norm": 0.01244130078703165, | |
| "learning_rate": 9.137380191693292e-05, | |
| "loss": 0.0015, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.011427606455981731, | |
| "learning_rate": 8.977635782747604e-05, | |
| "loss": 0.0355, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.236421725239617, | |
| "grad_norm": 0.017172975465655327, | |
| "learning_rate": 8.817891373801918e-05, | |
| "loss": 0.0115, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.236421725239617, | |
| "eval_accuracy": 0.9844, | |
| "eval_loss": 0.0627300888299942, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.3578, | |
| "eval_samples_per_second": 202.301, | |
| "eval_steps_per_second": 25.328, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.268370607028754, | |
| "grad_norm": 0.010955709032714367, | |
| "learning_rate": 8.658146964856231e-05, | |
| "loss": 0.0135, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.013627716340124607, | |
| "learning_rate": 8.498402555910544e-05, | |
| "loss": 0.0328, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.3322683706070286, | |
| "grad_norm": 0.024327559396624565, | |
| "learning_rate": 8.338658146964856e-05, | |
| "loss": 0.0049, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.364217252396166, | |
| "grad_norm": 0.012321406975388527, | |
| "learning_rate": 8.17891373801917e-05, | |
| "loss": 0.0096, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.010169831104576588, | |
| "learning_rate": 8.019169329073483e-05, | |
| "loss": 0.0014, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.428115015974441, | |
| "grad_norm": 0.010039562359452248, | |
| "learning_rate": 7.859424920127795e-05, | |
| "loss": 0.0482, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.460063897763578, | |
| "grad_norm": 0.010927636176347733, | |
| "learning_rate": 7.699680511182109e-05, | |
| "loss": 0.0149, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.010604580864310265, | |
| "learning_rate": 7.539936102236423e-05, | |
| "loss": 0.0028, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.523961661341853, | |
| "grad_norm": 0.009957361966371536, | |
| "learning_rate": 7.380191693290735e-05, | |
| "loss": 0.0042, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.5559105431309903, | |
| "grad_norm": 0.01096680574119091, | |
| "learning_rate": 7.220447284345049e-05, | |
| "loss": 0.0016, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.5559105431309903, | |
| "eval_accuracy": 0.9936, | |
| "eval_loss": 0.02957286313176155, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.3992, | |
| "eval_samples_per_second": 201.627, | |
| "eval_steps_per_second": 25.244, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.010381845757365227, | |
| "learning_rate": 7.060702875399361e-05, | |
| "loss": 0.0013, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.619808306709265, | |
| "grad_norm": 0.03203699365258217, | |
| "learning_rate": 6.900958466453674e-05, | |
| "loss": 0.0209, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.6517571884984026, | |
| "grad_norm": 0.009094555862247944, | |
| "learning_rate": 6.741214057507987e-05, | |
| "loss": 0.0011, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.008862023241817951, | |
| "learning_rate": 6.5814696485623e-05, | |
| "loss": 0.0287, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.7156549520766773, | |
| "grad_norm": 0.00914779119193554, | |
| "learning_rate": 6.421725239616614e-05, | |
| "loss": 0.0012, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.747603833865815, | |
| "grad_norm": 0.009615966118872166, | |
| "learning_rate": 6.261980830670928e-05, | |
| "loss": 0.0011, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.035701438784599304, | |
| "learning_rate": 6.1022364217252406e-05, | |
| "loss": 0.0089, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8115015974440896, | |
| "grad_norm": 0.012455513700842857, | |
| "learning_rate": 5.942492012779552e-05, | |
| "loss": 0.0011, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.8434504792332267, | |
| "grad_norm": 0.009292890317738056, | |
| "learning_rate": 5.782747603833866e-05, | |
| "loss": 0.0091, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.008451340720057487, | |
| "learning_rate": 5.623003194888179e-05, | |
| "loss": 0.004, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "eval_accuracy": 0.9916, | |
| "eval_loss": 0.032514333724975586, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.4929, | |
| "eval_samples_per_second": 200.114, | |
| "eval_steps_per_second": 25.054, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.9073482428115014, | |
| "grad_norm": 0.7223944067955017, | |
| "learning_rate": 5.4632587859424925e-05, | |
| "loss": 0.0019, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.939297124600639, | |
| "grad_norm": 0.007820590399205685, | |
| "learning_rate": 5.3035143769968054e-05, | |
| "loss": 0.001, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.011096654459834099, | |
| "learning_rate": 5.1437699680511184e-05, | |
| "loss": 0.0011, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0031948881789137, | |
| "grad_norm": 0.007654030807316303, | |
| "learning_rate": 4.984025559105431e-05, | |
| "loss": 0.0011, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.0351437699680512, | |
| "grad_norm": 0.010331162251532078, | |
| "learning_rate": 4.824281150159744e-05, | |
| "loss": 0.0009, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0670926517571884, | |
| "grad_norm": 0.007492523640394211, | |
| "learning_rate": 4.664536741214058e-05, | |
| "loss": 0.0009, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.099041533546326, | |
| "grad_norm": 0.007593994960188866, | |
| "learning_rate": 4.504792332268371e-05, | |
| "loss": 0.001, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.130990415335463, | |
| "grad_norm": 0.007749509997665882, | |
| "learning_rate": 4.345047923322684e-05, | |
| "loss": 0.0009, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.1629392971246006, | |
| "grad_norm": 0.007726697251200676, | |
| "learning_rate": 4.185303514376997e-05, | |
| "loss": 0.0009, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.194888178913738, | |
| "grad_norm": 0.007175728678703308, | |
| "learning_rate": 4.0255591054313104e-05, | |
| "loss": 0.0009, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.194888178913738, | |
| "eval_accuracy": 0.9948, | |
| "eval_loss": 0.02237132005393505, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.3213, | |
| "eval_samples_per_second": 202.9, | |
| "eval_steps_per_second": 25.403, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2268370607028753, | |
| "grad_norm": 0.007404172793030739, | |
| "learning_rate": 3.8658146964856234e-05, | |
| "loss": 0.0019, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.258785942492013, | |
| "grad_norm": 0.007146772928535938, | |
| "learning_rate": 3.7060702875399364e-05, | |
| "loss": 0.0009, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.29073482428115, | |
| "grad_norm": 0.02829296886920929, | |
| "learning_rate": 3.546325878594249e-05, | |
| "loss": 0.0009, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3226837060702876, | |
| "grad_norm": 0.007179423235356808, | |
| "learning_rate": 3.386581469648562e-05, | |
| "loss": 0.0009, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.3546325878594248, | |
| "grad_norm": 0.0070446510799229145, | |
| "learning_rate": 3.226837060702875e-05, | |
| "loss": 0.0009, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.3865814696485623, | |
| "grad_norm": 0.007006002124398947, | |
| "learning_rate": 3.067092651757188e-05, | |
| "loss": 0.0009, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.4185303514377, | |
| "grad_norm": 0.006977152545005083, | |
| "learning_rate": 2.907348242811502e-05, | |
| "loss": 0.0008, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.450479233226837, | |
| "grad_norm": 0.0067046028561890125, | |
| "learning_rate": 2.747603833865815e-05, | |
| "loss": 0.0009, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.4824281150159746, | |
| "grad_norm": 0.0067406343296170235, | |
| "learning_rate": 2.5878594249201278e-05, | |
| "loss": 0.0012, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.5143769968051117, | |
| "grad_norm": 0.006645725108683109, | |
| "learning_rate": 2.428115015974441e-05, | |
| "loss": 0.0008, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.5143769968051117, | |
| "eval_accuracy": 0.9936, | |
| "eval_loss": 0.027039185166358948, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 13.3116, | |
| "eval_samples_per_second": 187.807, | |
| "eval_steps_per_second": 23.513, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.5463258785942493, | |
| "grad_norm": 0.0066969566978514194, | |
| "learning_rate": 2.268370607028754e-05, | |
| "loss": 0.0008, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.5782747603833864, | |
| "grad_norm": 0.0066523440182209015, | |
| "learning_rate": 2.108626198083067e-05, | |
| "loss": 0.0008, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.610223642172524, | |
| "grad_norm": 0.0067514642141759396, | |
| "learning_rate": 1.9488817891373803e-05, | |
| "loss": 0.0008, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.642172523961661, | |
| "grad_norm": 0.0066239056177437305, | |
| "learning_rate": 1.7891373801916932e-05, | |
| "loss": 0.001, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.6741214057507987, | |
| "grad_norm": 0.006579084321856499, | |
| "learning_rate": 1.6293929712460065e-05, | |
| "loss": 0.0008, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.7060702875399363, | |
| "grad_norm": 0.006953445728868246, | |
| "learning_rate": 1.4696485623003195e-05, | |
| "loss": 0.0009, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.7380191693290734, | |
| "grad_norm": 0.006420870777219534, | |
| "learning_rate": 1.3099041533546328e-05, | |
| "loss": 0.0008, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.769968051118211, | |
| "grad_norm": 0.006650346331298351, | |
| "learning_rate": 1.1501597444089457e-05, | |
| "loss": 0.0008, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.801916932907348, | |
| "grad_norm": 0.006530832499265671, | |
| "learning_rate": 9.904153354632589e-06, | |
| "loss": 0.0018, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.8338658146964857, | |
| "grad_norm": 0.007318571675568819, | |
| "learning_rate": 8.306709265175718e-06, | |
| "loss": 0.0008, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8338658146964857, | |
| "eval_accuracy": 0.994, | |
| "eval_loss": 0.02559490129351616, | |
| "eval_model_preparation_time": 0.0051, | |
| "eval_runtime": 12.5362, | |
| "eval_samples_per_second": 199.423, | |
| "eval_steps_per_second": 24.968, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8658146964856233, | |
| "grad_norm": 0.007055082358419895, | |
| "learning_rate": 6.70926517571885e-06, | |
| "loss": 0.001, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.8977635782747604, | |
| "grad_norm": 0.006507499609142542, | |
| "learning_rate": 5.111821086261981e-06, | |
| "loss": 0.0008, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.9297124600638975, | |
| "grad_norm": 0.006634117104113102, | |
| "learning_rate": 3.5143769968051118e-06, | |
| "loss": 0.0007, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.961661341853035, | |
| "grad_norm": 0.006404239218682051, | |
| "learning_rate": 1.9169329073482426e-06, | |
| "loss": 0.0008, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.9936102236421727, | |
| "grad_norm": 0.006795979104936123, | |
| "learning_rate": 3.194888178913738e-07, | |
| "loss": 0.0008, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 1252, | |
| "total_flos": 1.54983979229184e+18, | |
| "train_loss": 0.055932876091605174, | |
| "train_runtime": 533.4054, | |
| "train_samples_per_second": 37.495, | |
| "train_steps_per_second": 2.347 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.54983979229184e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |