| { | |
| "best_metric": 0.32094342968611456, | |
| "best_model_checkpoint": "Mrohit01/cards-vit-base-patch16-224-finetuned-v1/checkpoint-3404", | |
| "epoch": 9.993390614672835, | |
| "eval_steps": 500, | |
| "global_step": 3780, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.026437541308658295, | |
| "grad_norm": 1.8425437211990356, | |
| "learning_rate": 1.3227513227513228e-06, | |
| "loss": 1.9993, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05287508261731659, | |
| "grad_norm": 1.9051584005355835, | |
| "learning_rate": 2.6455026455026455e-06, | |
| "loss": 2.0327, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07931262392597488, | |
| "grad_norm": 1.936784267425537, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 2.0055, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10575016523463318, | |
| "grad_norm": 1.9875764846801758, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 1.9807, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13218770654329148, | |
| "grad_norm": 2.0385992527008057, | |
| "learning_rate": 6.613756613756614e-06, | |
| "loss": 1.9866, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15862524785194976, | |
| "grad_norm": 1.9623510837554932, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 1.9612, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18506278916060806, | |
| "grad_norm": 1.8609628677368164, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 1.9499, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21150033046926636, | |
| "grad_norm": 2.279550552368164, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 1.9277, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23793787177792466, | |
| "grad_norm": 2.0092594623565674, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 1.9246, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26437541308658297, | |
| "grad_norm": 1.8894586563110352, | |
| "learning_rate": 1.3227513227513228e-05, | |
| "loss": 1.9187, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29081295439524124, | |
| "grad_norm": 1.841611623764038, | |
| "learning_rate": 1.455026455026455e-05, | |
| "loss": 1.9279, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3172504957038995, | |
| "grad_norm": 1.9487497806549072, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 1.8989, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34368803701255785, | |
| "grad_norm": 1.8522062301635742, | |
| "learning_rate": 1.7195767195767195e-05, | |
| "loss": 1.9041, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3701255783212161, | |
| "grad_norm": 1.6527143716812134, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 1.9107, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3965631196298744, | |
| "grad_norm": 1.6777476072311401, | |
| "learning_rate": 1.984126984126984e-05, | |
| "loss": 1.8915, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4230006609385327, | |
| "grad_norm": 2.1537675857543945, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 1.8814, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.449438202247191, | |
| "grad_norm": 2.119074583053589, | |
| "learning_rate": 2.2486772486772487e-05, | |
| "loss": 1.8767, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47587574355584933, | |
| "grad_norm": 2.027825355529785, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 1.8542, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5023132848645075, | |
| "grad_norm": 1.9427117109298706, | |
| "learning_rate": 2.5132275132275137e-05, | |
| "loss": 1.8614, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5287508261731659, | |
| "grad_norm": 2.057250499725342, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 1.8578, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5551883674818242, | |
| "grad_norm": 1.760541558265686, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.8507, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5816259087904825, | |
| "grad_norm": 2.5953147411346436, | |
| "learning_rate": 2.91005291005291e-05, | |
| "loss": 1.8325, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6080634500991408, | |
| "grad_norm": 2.2065205574035645, | |
| "learning_rate": 3.0423280423280425e-05, | |
| "loss": 1.8418, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.634500991407799, | |
| "grad_norm": 1.930132269859314, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 1.8295, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6609385327164574, | |
| "grad_norm": 2.2985873222351074, | |
| "learning_rate": 3.306878306878307e-05, | |
| "loss": 1.8367, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6873760740251157, | |
| "grad_norm": 2.229893207550049, | |
| "learning_rate": 3.439153439153439e-05, | |
| "loss": 1.8298, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.713813615333774, | |
| "grad_norm": 2.627453088760376, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 1.8321, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7402511566424322, | |
| "grad_norm": 2.1309328079223633, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.8377, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7666886979510905, | |
| "grad_norm": 1.6566647291183472, | |
| "learning_rate": 3.835978835978836e-05, | |
| "loss": 1.8391, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7931262392597488, | |
| "grad_norm": 3.22412109375, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 1.825, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8195637805684072, | |
| "grad_norm": 2.1069350242614746, | |
| "learning_rate": 4.100529100529101e-05, | |
| "loss": 1.8074, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8460013218770654, | |
| "grad_norm": 2.3734729290008545, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 1.8281, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8724388631857237, | |
| "grad_norm": 1.7937417030334473, | |
| "learning_rate": 4.3650793650793655e-05, | |
| "loss": 1.8065, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.898876404494382, | |
| "grad_norm": 2.0130672454833984, | |
| "learning_rate": 4.4973544973544974e-05, | |
| "loss": 1.8133, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9253139458030403, | |
| "grad_norm": 1.8942523002624512, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.8097, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9517514871116987, | |
| "grad_norm": 2.3066070079803467, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.8325, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9781890284203569, | |
| "grad_norm": 1.563308596611023, | |
| "learning_rate": 4.894179894179895e-05, | |
| "loss": 1.7866, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9993390614672836, | |
| "eval_accuracy": 0.2745522255719099, | |
| "eval_loss": 1.7676516771316528, | |
| "eval_runtime": 664.7291, | |
| "eval_samples_per_second": 42.416, | |
| "eval_steps_per_second": 0.663, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.004626569729015, | |
| "grad_norm": 1.7081835269927979, | |
| "learning_rate": 4.9970605526161084e-05, | |
| "loss": 1.7884, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0310641110376735, | |
| "grad_norm": 2.7630808353424072, | |
| "learning_rate": 4.982363315696649e-05, | |
| "loss": 1.7739, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0575016523463319, | |
| "grad_norm": 3.5567917823791504, | |
| "learning_rate": 4.96766607877719e-05, | |
| "loss": 1.7764, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.08393919365499, | |
| "grad_norm": 2.181006669998169, | |
| "learning_rate": 4.952968841857731e-05, | |
| "loss": 1.8154, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.1103767349636484, | |
| "grad_norm": 2.285545587539673, | |
| "learning_rate": 4.938271604938271e-05, | |
| "loss": 1.7998, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1368142762723066, | |
| "grad_norm": 2.4093916416168213, | |
| "learning_rate": 4.923574368018813e-05, | |
| "loss": 1.7643, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.163251817580965, | |
| "grad_norm": 1.6129083633422852, | |
| "learning_rate": 4.908877131099354e-05, | |
| "loss": 1.7932, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.1896893588896233, | |
| "grad_norm": 1.6316031217575073, | |
| "learning_rate": 4.894179894179895e-05, | |
| "loss": 1.7851, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.2161269001982815, | |
| "grad_norm": 1.6007944345474243, | |
| "learning_rate": 4.879482657260435e-05, | |
| "loss": 1.795, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.24256444150694, | |
| "grad_norm": 1.69334077835083, | |
| "learning_rate": 4.864785420340976e-05, | |
| "loss": 1.7523, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.269001982815598, | |
| "grad_norm": 1.9214173555374146, | |
| "learning_rate": 4.850088183421517e-05, | |
| "loss": 1.7595, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.2954395241242564, | |
| "grad_norm": 1.5973252058029175, | |
| "learning_rate": 4.835390946502058e-05, | |
| "loss": 1.7542, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.3218770654329148, | |
| "grad_norm": 1.8921444416046143, | |
| "learning_rate": 4.820693709582599e-05, | |
| "loss": 1.7466, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.348314606741573, | |
| "grad_norm": 1.84861159324646, | |
| "learning_rate": 4.8059964726631394e-05, | |
| "loss": 1.7763, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.3747521480502314, | |
| "grad_norm": 2.476146936416626, | |
| "learning_rate": 4.79129923574368e-05, | |
| "loss": 1.7577, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.4011896893588895, | |
| "grad_norm": 1.5669139623641968, | |
| "learning_rate": 4.776601998824221e-05, | |
| "loss": 1.7697, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.427627230667548, | |
| "grad_norm": 2.1005804538726807, | |
| "learning_rate": 4.761904761904762e-05, | |
| "loss": 1.7737, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.4540647719762063, | |
| "grad_norm": 1.8329302072525024, | |
| "learning_rate": 4.747207524985303e-05, | |
| "loss": 1.7476, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.4805023132848645, | |
| "grad_norm": 1.7490884065628052, | |
| "learning_rate": 4.732510288065844e-05, | |
| "loss": 1.7695, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.5069398545935226, | |
| "grad_norm": 1.7790248394012451, | |
| "learning_rate": 4.717813051146385e-05, | |
| "loss": 1.7696, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.533377395902181, | |
| "grad_norm": 1.4908485412597656, | |
| "learning_rate": 4.7031158142269256e-05, | |
| "loss": 1.7809, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.5598149372108394, | |
| "grad_norm": 1.4122120141983032, | |
| "learning_rate": 4.6884185773074665e-05, | |
| "loss": 1.7281, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.5862524785194978, | |
| "grad_norm": 1.5507616996765137, | |
| "learning_rate": 4.673721340388007e-05, | |
| "loss": 1.7906, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.612690019828156, | |
| "grad_norm": 1.7312538623809814, | |
| "learning_rate": 4.659024103468548e-05, | |
| "loss": 1.7863, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.6391275611368141, | |
| "grad_norm": 1.9535393714904785, | |
| "learning_rate": 4.644326866549089e-05, | |
| "loss": 1.7963, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.6655651024454725, | |
| "grad_norm": 2.0952374935150146, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.784, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.692002643754131, | |
| "grad_norm": 1.8771756887435913, | |
| "learning_rate": 4.614932392710171e-05, | |
| "loss": 1.7717, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.7184401850627893, | |
| "grad_norm": 1.9628280401229858, | |
| "learning_rate": 4.600235155790711e-05, | |
| "loss": 1.7156, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.7448777263714474, | |
| "grad_norm": 1.7401697635650635, | |
| "learning_rate": 4.585537918871252e-05, | |
| "loss": 1.7748, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.7713152676801056, | |
| "grad_norm": 1.9188202619552612, | |
| "learning_rate": 4.5708406819517937e-05, | |
| "loss": 1.7759, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.797752808988764, | |
| "grad_norm": 1.7092291116714478, | |
| "learning_rate": 4.5561434450323345e-05, | |
| "loss": 1.7368, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.8241903502974224, | |
| "grad_norm": 1.9781934022903442, | |
| "learning_rate": 4.541446208112875e-05, | |
| "loss": 1.7743, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.8506278916060808, | |
| "grad_norm": 2.159959077835083, | |
| "learning_rate": 4.5267489711934157e-05, | |
| "loss": 1.7723, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.877065432914739, | |
| "grad_norm": 1.9522385597229004, | |
| "learning_rate": 4.5120517342739565e-05, | |
| "loss": 1.7669, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.903502974223397, | |
| "grad_norm": 2.2479653358459473, | |
| "learning_rate": 4.4973544973544974e-05, | |
| "loss": 1.7479, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.9299405155320555, | |
| "grad_norm": 2.117433547973633, | |
| "learning_rate": 4.482657260435038e-05, | |
| "loss": 1.7363, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.9563780568407139, | |
| "grad_norm": 1.7201601266860962, | |
| "learning_rate": 4.467960023515579e-05, | |
| "loss": 1.7551, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.9828155981493722, | |
| "grad_norm": 1.7983427047729492, | |
| "learning_rate": 4.45326278659612e-05, | |
| "loss": 1.7457, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.998678122934567, | |
| "eval_accuracy": 0.29902464976059584, | |
| "eval_loss": 1.7163423299789429, | |
| "eval_runtime": 618.2518, | |
| "eval_samples_per_second": 45.604, | |
| "eval_steps_per_second": 0.713, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 2.00925313945803, | |
| "grad_norm": 1.3950998783111572, | |
| "learning_rate": 4.438565549676661e-05, | |
| "loss": 1.7106, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.0356906807666886, | |
| "grad_norm": 1.764172077178955, | |
| "learning_rate": 4.423868312757202e-05, | |
| "loss": 1.7251, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.062128222075347, | |
| "grad_norm": 1.5473984479904175, | |
| "learning_rate": 4.409171075837743e-05, | |
| "loss": 1.7107, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.0885657633840053, | |
| "grad_norm": 1.851131558418274, | |
| "learning_rate": 4.394473838918284e-05, | |
| "loss": 1.7104, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.1150033046926637, | |
| "grad_norm": 1.5341060161590576, | |
| "learning_rate": 4.3797766019988246e-05, | |
| "loss": 1.7051, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.1414408460013217, | |
| "grad_norm": 1.7939796447753906, | |
| "learning_rate": 4.3650793650793655e-05, | |
| "loss": 1.7393, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.16787838730998, | |
| "grad_norm": 1.713605284690857, | |
| "learning_rate": 4.3503821281599064e-05, | |
| "loss": 1.6997, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.1943159286186384, | |
| "grad_norm": 1.7437527179718018, | |
| "learning_rate": 4.3356848912404466e-05, | |
| "loss": 1.7004, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.220753469927297, | |
| "grad_norm": 1.8171709775924683, | |
| "learning_rate": 4.3209876543209875e-05, | |
| "loss": 1.7275, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.247191011235955, | |
| "grad_norm": 1.628079891204834, | |
| "learning_rate": 4.306290417401529e-05, | |
| "loss": 1.7378, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.273628552544613, | |
| "grad_norm": 1.7622318267822266, | |
| "learning_rate": 4.29159318048207e-05, | |
| "loss": 1.7247, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.3000660938532715, | |
| "grad_norm": 1.4866081476211548, | |
| "learning_rate": 4.27689594356261e-05, | |
| "loss": 1.7166, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.32650363516193, | |
| "grad_norm": 1.7897216081619263, | |
| "learning_rate": 4.262198706643151e-05, | |
| "loss": 1.7096, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 1.7703666687011719, | |
| "learning_rate": 4.247501469723692e-05, | |
| "loss": 1.7096, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.3793787177792467, | |
| "grad_norm": 1.9003839492797852, | |
| "learning_rate": 4.232804232804233e-05, | |
| "loss": 1.7172, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.4058162590879046, | |
| "grad_norm": 2.338798999786377, | |
| "learning_rate": 4.2181069958847744e-05, | |
| "loss": 1.7209, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.432253800396563, | |
| "grad_norm": 2.1922528743743896, | |
| "learning_rate": 4.2034097589653146e-05, | |
| "loss": 1.7195, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.4586913417052214, | |
| "grad_norm": 1.7596076726913452, | |
| "learning_rate": 4.1887125220458555e-05, | |
| "loss": 1.7316, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.48512888301388, | |
| "grad_norm": 1.8625160455703735, | |
| "learning_rate": 4.1740152851263964e-05, | |
| "loss": 1.7224, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.511566424322538, | |
| "grad_norm": 2.2473673820495605, | |
| "learning_rate": 4.159318048206937e-05, | |
| "loss": 1.7198, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.538003965631196, | |
| "grad_norm": 1.8248813152313232, | |
| "learning_rate": 4.144620811287478e-05, | |
| "loss": 1.7246, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.5644415069398545, | |
| "grad_norm": 1.8088594675064087, | |
| "learning_rate": 4.129923574368019e-05, | |
| "loss": 1.707, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.590879048248513, | |
| "grad_norm": 1.818827748298645, | |
| "learning_rate": 4.11522633744856e-05, | |
| "loss": 1.7038, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.6173165895571713, | |
| "grad_norm": 1.716060757637024, | |
| "learning_rate": 4.100529100529101e-05, | |
| "loss": 1.7066, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.6437541308658297, | |
| "grad_norm": 1.7306545972824097, | |
| "learning_rate": 4.085831863609642e-05, | |
| "loss": 1.6767, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.6701916721744876, | |
| "grad_norm": 1.7304011583328247, | |
| "learning_rate": 4.071134626690182e-05, | |
| "loss": 1.6969, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.696629213483146, | |
| "grad_norm": 2.024599552154541, | |
| "learning_rate": 4.056437389770723e-05, | |
| "loss": 1.7181, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.7230667547918044, | |
| "grad_norm": 2.146998643875122, | |
| "learning_rate": 4.0417401528512645e-05, | |
| "loss": 1.7311, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.7495042961004628, | |
| "grad_norm": 1.935433030128479, | |
| "learning_rate": 4.0270429159318054e-05, | |
| "loss": 1.7194, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.775941837409121, | |
| "grad_norm": 1.6351971626281738, | |
| "learning_rate": 4.012345679012346e-05, | |
| "loss": 1.7519, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.802379378717779, | |
| "grad_norm": 1.971442461013794, | |
| "learning_rate": 3.9976484420928865e-05, | |
| "loss": 1.7377, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.8288169200264375, | |
| "grad_norm": 1.8716827630996704, | |
| "learning_rate": 3.9829512051734274e-05, | |
| "loss": 1.7066, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.855254461335096, | |
| "grad_norm": 1.6349095106124878, | |
| "learning_rate": 3.968253968253968e-05, | |
| "loss": 1.7, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.8816920026437542, | |
| "grad_norm": 1.7663626670837402, | |
| "learning_rate": 3.95355673133451e-05, | |
| "loss": 1.6906, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.9081295439524126, | |
| "grad_norm": 2.1080522537231445, | |
| "learning_rate": 3.93885949441505e-05, | |
| "loss": 1.7245, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.9345670852610706, | |
| "grad_norm": 1.7439042329788208, | |
| "learning_rate": 3.924162257495591e-05, | |
| "loss": 1.6977, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.961004626569729, | |
| "grad_norm": 1.6347275972366333, | |
| "learning_rate": 3.909465020576132e-05, | |
| "loss": 1.7225, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.9874421678783873, | |
| "grad_norm": 2.7075507640838623, | |
| "learning_rate": 3.894767783656673e-05, | |
| "loss": 1.7123, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.9980171844018506, | |
| "eval_accuracy": 0.30069161198794114, | |
| "eval_loss": 1.6861532926559448, | |
| "eval_runtime": 594.266, | |
| "eval_samples_per_second": 47.445, | |
| "eval_steps_per_second": 0.742, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 3.0138797091870457, | |
| "grad_norm": 2.001418113708496, | |
| "learning_rate": 3.8800705467372136e-05, | |
| "loss": 1.7138, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.040317250495704, | |
| "grad_norm": 1.5465753078460693, | |
| "learning_rate": 3.8653733098177545e-05, | |
| "loss": 1.6694, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.066754791804362, | |
| "grad_norm": 2.0662121772766113, | |
| "learning_rate": 3.8506760728982954e-05, | |
| "loss": 1.6635, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.0931923331130204, | |
| "grad_norm": 1.6004307270050049, | |
| "learning_rate": 3.835978835978836e-05, | |
| "loss": 1.6928, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.119629874421679, | |
| "grad_norm": 2.330245018005371, | |
| "learning_rate": 3.821281599059377e-05, | |
| "loss": 1.6512, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.146067415730337, | |
| "grad_norm": 2.421938896179199, | |
| "learning_rate": 3.806584362139918e-05, | |
| "loss": 1.6904, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.1725049570389956, | |
| "grad_norm": 1.943880319595337, | |
| "learning_rate": 3.791887125220458e-05, | |
| "loss": 1.6807, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.1989424983476535, | |
| "grad_norm": 1.9737601280212402, | |
| "learning_rate": 3.777189888301e-05, | |
| "loss": 1.6763, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.225380039656312, | |
| "grad_norm": 1.5722044706344604, | |
| "learning_rate": 3.762492651381541e-05, | |
| "loss": 1.6449, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.2518175809649703, | |
| "grad_norm": 1.8357861042022705, | |
| "learning_rate": 3.7477954144620817e-05, | |
| "loss": 1.6637, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.2782551222736287, | |
| "grad_norm": 2.0803420543670654, | |
| "learning_rate": 3.733098177542622e-05, | |
| "loss": 1.6715, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.304692663582287, | |
| "grad_norm": 1.9865127801895142, | |
| "learning_rate": 3.718400940623163e-05, | |
| "loss": 1.6841, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.331130204890945, | |
| "grad_norm": 1.953385353088379, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.6699, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.3575677461996034, | |
| "grad_norm": 4.0031867027282715, | |
| "learning_rate": 3.689006466784245e-05, | |
| "loss": 1.6795, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.384005287508262, | |
| "grad_norm": 2.2489538192749023, | |
| "learning_rate": 3.6743092298647854e-05, | |
| "loss": 1.6705, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.41044282881692, | |
| "grad_norm": 1.7907918691635132, | |
| "learning_rate": 3.659611992945326e-05, | |
| "loss": 1.6754, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.4368803701255786, | |
| "grad_norm": 1.7390344142913818, | |
| "learning_rate": 3.644914756025867e-05, | |
| "loss": 1.6621, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.4633179114342365, | |
| "grad_norm": 1.6002901792526245, | |
| "learning_rate": 3.630217519106408e-05, | |
| "loss": 1.679, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.489755452742895, | |
| "grad_norm": 1.9788098335266113, | |
| "learning_rate": 3.615520282186949e-05, | |
| "loss": 1.6633, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.5161929940515533, | |
| "grad_norm": 2.1224050521850586, | |
| "learning_rate": 3.60082304526749e-05, | |
| "loss": 1.679, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.5426305353602117, | |
| "grad_norm": 2.0570385456085205, | |
| "learning_rate": 3.586125808348031e-05, | |
| "loss": 1.6752, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.56906807666887, | |
| "grad_norm": 2.158527374267578, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 1.6624, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.595505617977528, | |
| "grad_norm": 2.440808057785034, | |
| "learning_rate": 3.5567313345091126e-05, | |
| "loss": 1.66, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.6219431592861864, | |
| "grad_norm": 2.058363914489746, | |
| "learning_rate": 3.5420340975896535e-05, | |
| "loss": 1.6985, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.6483807005948448, | |
| "grad_norm": 1.795248031616211, | |
| "learning_rate": 3.527336860670194e-05, | |
| "loss": 1.7037, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.6748182419035027, | |
| "grad_norm": 1.7252378463745117, | |
| "learning_rate": 3.512639623750735e-05, | |
| "loss": 1.6633, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.7012557832121615, | |
| "grad_norm": 1.8798094987869263, | |
| "learning_rate": 3.497942386831276e-05, | |
| "loss": 1.6881, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.7276933245208195, | |
| "grad_norm": 2.317443609237671, | |
| "learning_rate": 3.483245149911817e-05, | |
| "loss": 1.6916, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.754130865829478, | |
| "grad_norm": 2.2357330322265625, | |
| "learning_rate": 3.468547912992357e-05, | |
| "loss": 1.665, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.7805684071381362, | |
| "grad_norm": 1.8255904912948608, | |
| "learning_rate": 3.453850676072898e-05, | |
| "loss": 1.6839, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.807005948446794, | |
| "grad_norm": 1.95132315158844, | |
| "learning_rate": 3.439153439153439e-05, | |
| "loss": 1.6977, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.833443489755453, | |
| "grad_norm": 1.942299723625183, | |
| "learning_rate": 3.4244562022339806e-05, | |
| "loss": 1.6648, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.859881031064111, | |
| "grad_norm": 2.280670642852783, | |
| "learning_rate": 3.4097589653145215e-05, | |
| "loss": 1.6621, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.8863185723727693, | |
| "grad_norm": 2.0500681400299072, | |
| "learning_rate": 3.395061728395062e-05, | |
| "loss": 1.6395, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.9127561136814277, | |
| "grad_norm": 1.9235775470733643, | |
| "learning_rate": 3.3803644914756026e-05, | |
| "loss": 1.6907, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.9391936549900857, | |
| "grad_norm": 1.8194037675857544, | |
| "learning_rate": 3.3656672545561435e-05, | |
| "loss": 1.6793, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.965631196298744, | |
| "grad_norm": 2.047736167907715, | |
| "learning_rate": 3.3509700176366844e-05, | |
| "loss": 1.682, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.9920687376074024, | |
| "grad_norm": 2.2198452949523926, | |
| "learning_rate": 3.336272780717225e-05, | |
| "loss": 1.6607, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.3080688065259798, | |
| "eval_loss": 1.6822971105575562, | |
| "eval_runtime": 603.9647, | |
| "eval_samples_per_second": 46.683, | |
| "eval_steps_per_second": 0.73, | |
| "step": 1513 | |
| }, | |
| { | |
| "epoch": 4.01850627891606, | |
| "grad_norm": 1.416684627532959, | |
| "learning_rate": 3.321575543797766e-05, | |
| "loss": 1.6447, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.044943820224719, | |
| "grad_norm": 2.129450798034668, | |
| "learning_rate": 3.306878306878307e-05, | |
| "loss": 1.6303, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.071381361533377, | |
| "grad_norm": 1.7213213443756104, | |
| "learning_rate": 3.292181069958848e-05, | |
| "loss": 1.6126, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 4.097818902842036, | |
| "grad_norm": 2.0465450286865234, | |
| "learning_rate": 3.277483833039389e-05, | |
| "loss": 1.6351, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.124256444150694, | |
| "grad_norm": 2.259282350540161, | |
| "learning_rate": 3.262786596119929e-05, | |
| "loss": 1.6437, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.150693985459352, | |
| "grad_norm": 2.2626054286956787, | |
| "learning_rate": 3.24808935920047e-05, | |
| "loss": 1.6523, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.177131526768011, | |
| "grad_norm": 2.269216537475586, | |
| "learning_rate": 3.2333921222810116e-05, | |
| "loss": 1.6369, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.203569068076669, | |
| "grad_norm": 2.1457483768463135, | |
| "learning_rate": 3.2186948853615525e-05, | |
| "loss": 1.6211, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.2300066093853275, | |
| "grad_norm": 1.9659942388534546, | |
| "learning_rate": 3.2039976484420934e-05, | |
| "loss": 1.639, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.256444150693985, | |
| "grad_norm": 2.447136402130127, | |
| "learning_rate": 3.1893004115226336e-05, | |
| "loss": 1.6548, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.282881692002643, | |
| "grad_norm": 2.2011499404907227, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 1.6042, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.309319233311302, | |
| "grad_norm": 1.9171946048736572, | |
| "learning_rate": 3.1599059376837154e-05, | |
| "loss": 1.6273, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.33575677461996, | |
| "grad_norm": 1.9522180557250977, | |
| "learning_rate": 3.145208700764257e-05, | |
| "loss": 1.6479, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.362194315928619, | |
| "grad_norm": 1.8249017000198364, | |
| "learning_rate": 3.130511463844797e-05, | |
| "loss": 1.633, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.388631857237277, | |
| "grad_norm": 2.093296766281128, | |
| "learning_rate": 3.115814226925338e-05, | |
| "loss": 1.6519, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.415069398545935, | |
| "grad_norm": 2.218247652053833, | |
| "learning_rate": 3.101116990005879e-05, | |
| "loss": 1.6478, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.441506939854594, | |
| "grad_norm": 1.9361261129379272, | |
| "learning_rate": 3.08641975308642e-05, | |
| "loss": 1.6473, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.467944481163252, | |
| "grad_norm": 1.9687649011611938, | |
| "learning_rate": 3.071722516166961e-05, | |
| "loss": 1.6263, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.49438202247191, | |
| "grad_norm": 1.8882747888565063, | |
| "learning_rate": 3.0570252792475016e-05, | |
| "loss": 1.6531, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.520819563780568, | |
| "grad_norm": 2.0464086532592773, | |
| "learning_rate": 3.0423280423280425e-05, | |
| "loss": 1.6481, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.547257105089226, | |
| "grad_norm": 2.1452088356018066, | |
| "learning_rate": 3.0276308054085834e-05, | |
| "loss": 1.6126, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.573694646397885, | |
| "grad_norm": 1.7108988761901855, | |
| "learning_rate": 3.012933568489124e-05, | |
| "loss": 1.6365, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.600132187706543, | |
| "grad_norm": 1.7739413976669312, | |
| "learning_rate": 2.998236331569665e-05, | |
| "loss": 1.6373, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.626569729015202, | |
| "grad_norm": 1.9384492635726929, | |
| "learning_rate": 2.9835390946502057e-05, | |
| "loss": 1.6212, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.65300727032386, | |
| "grad_norm": 2.0029444694519043, | |
| "learning_rate": 2.968841857730747e-05, | |
| "loss": 1.6185, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.679444811632518, | |
| "grad_norm": 2.2173917293548584, | |
| "learning_rate": 2.954144620811288e-05, | |
| "loss": 1.6441, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 2.4079232215881348, | |
| "learning_rate": 2.9394473838918284e-05, | |
| "loss": 1.6441, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.732319894249835, | |
| "grad_norm": 2.1316912174224854, | |
| "learning_rate": 2.9247501469723693e-05, | |
| "loss": 1.6464, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.758757435558493, | |
| "grad_norm": 2.2552201747894287, | |
| "learning_rate": 2.91005291005291e-05, | |
| "loss": 1.6321, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.785194976867151, | |
| "grad_norm": 2.0181891918182373, | |
| "learning_rate": 2.8953556731334508e-05, | |
| "loss": 1.6584, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.811632518175809, | |
| "grad_norm": 1.912116527557373, | |
| "learning_rate": 2.880658436213992e-05, | |
| "loss": 1.6618, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.838070059484468, | |
| "grad_norm": 1.916869878768921, | |
| "learning_rate": 2.865961199294533e-05, | |
| "loss": 1.6719, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.864507600793126, | |
| "grad_norm": 2.0317203998565674, | |
| "learning_rate": 2.8512639623750738e-05, | |
| "loss": 1.6258, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.890945142101785, | |
| "grad_norm": 2.080411195755005, | |
| "learning_rate": 2.8365667254556143e-05, | |
| "loss": 1.6518, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.917382683410443, | |
| "grad_norm": 1.9478580951690674, | |
| "learning_rate": 2.8218694885361552e-05, | |
| "loss": 1.614, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.943820224719101, | |
| "grad_norm": 1.889710545539856, | |
| "learning_rate": 2.8071722516166958e-05, | |
| "loss": 1.6221, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.97025776602776, | |
| "grad_norm": 2.255192756652832, | |
| "learning_rate": 2.7924750146972374e-05, | |
| "loss": 1.6271, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.9966953073364175, | |
| "grad_norm": 2.026352643966675, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.6188, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.999339061467284, | |
| "eval_accuracy": 0.3107997871963114, | |
| "eval_loss": 1.6907494068145752, | |
| "eval_runtime": 587.6675, | |
| "eval_samples_per_second": 47.978, | |
| "eval_steps_per_second": 0.75, | |
| "step": 1891 | |
| }, | |
| { | |
| "epoch": 5.023132848645076, | |
| "grad_norm": 1.7152390480041504, | |
| "learning_rate": 2.7630805408583188e-05, | |
| "loss": 1.6156, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.049570389953734, | |
| "grad_norm": 2.2232937812805176, | |
| "learning_rate": 2.7483833039388597e-05, | |
| "loss": 1.618, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 5.076007931262392, | |
| "grad_norm": 1.9228154420852661, | |
| "learning_rate": 2.7336860670194003e-05, | |
| "loss": 1.6137, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 5.102445472571051, | |
| "grad_norm": 1.9937798976898193, | |
| "learning_rate": 2.718988830099941e-05, | |
| "loss": 1.6139, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 5.128883013879709, | |
| "grad_norm": 2.4573240280151367, | |
| "learning_rate": 2.7042915931804824e-05, | |
| "loss": 1.5976, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 5.155320555188368, | |
| "grad_norm": 2.107956647872925, | |
| "learning_rate": 2.6895943562610233e-05, | |
| "loss": 1.6046, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 5.181758096497026, | |
| "grad_norm": 1.9604800939559937, | |
| "learning_rate": 2.6748971193415638e-05, | |
| "loss": 1.6001, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 5.208195637805684, | |
| "grad_norm": 2.3065364360809326, | |
| "learning_rate": 2.6601998824221047e-05, | |
| "loss": 1.5818, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 5.234633179114343, | |
| "grad_norm": 2.3068606853485107, | |
| "learning_rate": 2.6455026455026456e-05, | |
| "loss": 1.629, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 5.2610707204230005, | |
| "grad_norm": 1.9791010618209839, | |
| "learning_rate": 2.630805408583186e-05, | |
| "loss": 1.5831, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 5.287508261731659, | |
| "grad_norm": 2.128002643585205, | |
| "learning_rate": 2.6161081716637274e-05, | |
| "loss": 1.6208, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.313945803040317, | |
| "grad_norm": 1.999395728111267, | |
| "learning_rate": 2.6014109347442683e-05, | |
| "loss": 1.5791, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.340383344348975, | |
| "grad_norm": 2.4271738529205322, | |
| "learning_rate": 2.5867136978248092e-05, | |
| "loss": 1.6145, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.366820885657634, | |
| "grad_norm": 2.0685036182403564, | |
| "learning_rate": 2.5720164609053497e-05, | |
| "loss": 1.6013, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 5.393258426966292, | |
| "grad_norm": 2.3582699298858643, | |
| "learning_rate": 2.5573192239858906e-05, | |
| "loss": 1.5878, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 5.419695968274951, | |
| "grad_norm": 2.0049946308135986, | |
| "learning_rate": 2.5426219870664315e-05, | |
| "loss": 1.6114, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.446133509583609, | |
| "grad_norm": 1.902218222618103, | |
| "learning_rate": 2.5279247501469728e-05, | |
| "loss": 1.5934, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 5.472571050892267, | |
| "grad_norm": 2.1240365505218506, | |
| "learning_rate": 2.5132275132275137e-05, | |
| "loss": 1.561, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.4990085922009255, | |
| "grad_norm": 2.0185890197753906, | |
| "learning_rate": 2.4985302763080542e-05, | |
| "loss": 1.5912, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.5254461335095835, | |
| "grad_norm": 2.5916292667388916, | |
| "learning_rate": 2.483833039388595e-05, | |
| "loss": 1.6176, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.551883674818242, | |
| "grad_norm": 2.4469070434570312, | |
| "learning_rate": 2.4691358024691357e-05, | |
| "loss": 1.6373, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.5783212161269, | |
| "grad_norm": 1.9966007471084595, | |
| "learning_rate": 2.454438565549677e-05, | |
| "loss": 1.6049, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 5.604758757435558, | |
| "grad_norm": 2.5291779041290283, | |
| "learning_rate": 2.4397413286302174e-05, | |
| "loss": 1.595, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 5.631196298744217, | |
| "grad_norm": 2.031622886657715, | |
| "learning_rate": 2.4250440917107583e-05, | |
| "loss": 1.6069, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 5.657633840052875, | |
| "grad_norm": 1.9951566457748413, | |
| "learning_rate": 2.4103468547912996e-05, | |
| "loss": 1.583, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 5.684071381361534, | |
| "grad_norm": 2.609140634536743, | |
| "learning_rate": 2.39564961787184e-05, | |
| "loss": 1.5821, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.710508922670192, | |
| "grad_norm": 2.389698028564453, | |
| "learning_rate": 2.380952380952381e-05, | |
| "loss": 1.6114, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 5.73694646397885, | |
| "grad_norm": 2.4471077919006348, | |
| "learning_rate": 2.366255144032922e-05, | |
| "loss": 1.5919, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 5.7633840052875085, | |
| "grad_norm": 2.0070338249206543, | |
| "learning_rate": 2.3515579071134628e-05, | |
| "loss": 1.5963, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 5.789821546596166, | |
| "grad_norm": 2.1658718585968018, | |
| "learning_rate": 2.3368606701940034e-05, | |
| "loss": 1.6038, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.816259087904825, | |
| "grad_norm": 2.08571457862854, | |
| "learning_rate": 2.3221634332745446e-05, | |
| "loss": 1.5962, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.842696629213483, | |
| "grad_norm": 2.118481397628784, | |
| "learning_rate": 2.3074661963550855e-05, | |
| "loss": 1.5945, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.869134170522141, | |
| "grad_norm": 2.135956048965454, | |
| "learning_rate": 2.292768959435626e-05, | |
| "loss": 1.5785, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.8955717118308, | |
| "grad_norm": 2.1876909732818604, | |
| "learning_rate": 2.2780717225161673e-05, | |
| "loss": 1.5769, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.922009253139458, | |
| "grad_norm": 2.121159315109253, | |
| "learning_rate": 2.2633744855967078e-05, | |
| "loss": 1.5869, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.948446794448117, | |
| "grad_norm": 2.0962424278259277, | |
| "learning_rate": 2.2486772486772487e-05, | |
| "loss": 1.6203, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.974884335756775, | |
| "grad_norm": 1.9238280057907104, | |
| "learning_rate": 2.2339800117577896e-05, | |
| "loss": 1.6009, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.998678122934567, | |
| "eval_accuracy": 0.3150203936868239, | |
| "eval_loss": 1.677284598350525, | |
| "eval_runtime": 825.1974, | |
| "eval_samples_per_second": 34.168, | |
| "eval_steps_per_second": 0.534, | |
| "step": 2269 | |
| }, | |
| { | |
| "epoch": 6.001321877065433, | |
| "grad_norm": 3.6250228881835938, | |
| "learning_rate": 2.2192827748383305e-05, | |
| "loss": 1.6085, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 6.0277594183740915, | |
| "grad_norm": 2.6421561241149902, | |
| "learning_rate": 2.2045855379188714e-05, | |
| "loss": 1.5612, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 6.054196959682749, | |
| "grad_norm": 1.9773502349853516, | |
| "learning_rate": 2.1898883009994123e-05, | |
| "loss": 1.558, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 6.080634500991408, | |
| "grad_norm": 2.200873374938965, | |
| "learning_rate": 2.1751910640799532e-05, | |
| "loss": 1.5701, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 6.107072042300066, | |
| "grad_norm": 2.346799373626709, | |
| "learning_rate": 2.1604938271604937e-05, | |
| "loss": 1.5605, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 6.133509583608724, | |
| "grad_norm": 2.3931312561035156, | |
| "learning_rate": 2.145796590241035e-05, | |
| "loss": 1.5498, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 6.159947124917383, | |
| "grad_norm": 2.662827968597412, | |
| "learning_rate": 2.1310993533215755e-05, | |
| "loss": 1.5782, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 6.186384666226041, | |
| "grad_norm": 2.2345123291015625, | |
| "learning_rate": 2.1164021164021164e-05, | |
| "loss": 1.5672, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 6.2128222075347, | |
| "grad_norm": 2.2579140663146973, | |
| "learning_rate": 2.1017048794826573e-05, | |
| "loss": 1.5323, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 6.239259748843358, | |
| "grad_norm": 2.2183802127838135, | |
| "learning_rate": 2.0870076425631982e-05, | |
| "loss": 1.5499, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 6.265697290152016, | |
| "grad_norm": 2.227612257003784, | |
| "learning_rate": 2.072310405643739e-05, | |
| "loss": 1.5522, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 6.292134831460674, | |
| "grad_norm": 2.7729430198669434, | |
| "learning_rate": 2.05761316872428e-05, | |
| "loss": 1.5614, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 6.318572372769332, | |
| "grad_norm": 2.166227340698242, | |
| "learning_rate": 2.042915931804821e-05, | |
| "loss": 1.5358, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 6.345009914077991, | |
| "grad_norm": 2.1141440868377686, | |
| "learning_rate": 2.0282186948853614e-05, | |
| "loss": 1.554, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.371447455386649, | |
| "grad_norm": 2.1744093894958496, | |
| "learning_rate": 2.0135214579659027e-05, | |
| "loss": 1.5583, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 6.397884996695307, | |
| "grad_norm": 2.0957577228546143, | |
| "learning_rate": 1.9988242210464432e-05, | |
| "loss": 1.5573, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 6.424322538003966, | |
| "grad_norm": 2.6140244007110596, | |
| "learning_rate": 1.984126984126984e-05, | |
| "loss": 1.5597, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 6.450760079312624, | |
| "grad_norm": 2.0459976196289062, | |
| "learning_rate": 1.969429747207525e-05, | |
| "loss": 1.5439, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 6.477197620621283, | |
| "grad_norm": 2.2509853839874268, | |
| "learning_rate": 1.954732510288066e-05, | |
| "loss": 1.5478, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 6.503635161929941, | |
| "grad_norm": 2.0103518962860107, | |
| "learning_rate": 1.9400352733686068e-05, | |
| "loss": 1.555, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.530072703238599, | |
| "grad_norm": 2.5956430435180664, | |
| "learning_rate": 1.9253380364491477e-05, | |
| "loss": 1.5604, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 6.556510244547257, | |
| "grad_norm": 2.4196279048919678, | |
| "learning_rate": 1.9106407995296886e-05, | |
| "loss": 1.554, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 6.582947785855915, | |
| "grad_norm": 2.2963523864746094, | |
| "learning_rate": 1.895943562610229e-05, | |
| "loss": 1.5345, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 6.609385327164574, | |
| "grad_norm": 2.4753236770629883, | |
| "learning_rate": 1.8812463256907704e-05, | |
| "loss": 1.5698, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.635822868473232, | |
| "grad_norm": 2.251317024230957, | |
| "learning_rate": 1.866549088771311e-05, | |
| "loss": 1.5625, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 6.66226040978189, | |
| "grad_norm": 2.4385430812835693, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 1.5649, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 6.688697951090549, | |
| "grad_norm": 2.1294140815734863, | |
| "learning_rate": 1.8371546149323927e-05, | |
| "loss": 1.5214, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 6.715135492399207, | |
| "grad_norm": 2.234025478363037, | |
| "learning_rate": 1.8224573780129336e-05, | |
| "loss": 1.5928, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 6.741573033707866, | |
| "grad_norm": 2.6647934913635254, | |
| "learning_rate": 1.8077601410934745e-05, | |
| "loss": 1.5519, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.768010575016524, | |
| "grad_norm": 2.2540123462677, | |
| "learning_rate": 1.7930629041740154e-05, | |
| "loss": 1.5642, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 6.7944481163251815, | |
| "grad_norm": 2.701176404953003, | |
| "learning_rate": 1.7783656672545563e-05, | |
| "loss": 1.5733, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 6.82088565763384, | |
| "grad_norm": 2.6911590099334717, | |
| "learning_rate": 1.763668430335097e-05, | |
| "loss": 1.5651, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 6.847323198942498, | |
| "grad_norm": 2.3102941513061523, | |
| "learning_rate": 1.748971193415638e-05, | |
| "loss": 1.5716, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 6.873760740251157, | |
| "grad_norm": 2.4161031246185303, | |
| "learning_rate": 1.7342739564961786e-05, | |
| "loss": 1.5558, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.900198281559815, | |
| "grad_norm": 2.4778518676757812, | |
| "learning_rate": 1.7195767195767195e-05, | |
| "loss": 1.5766, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 6.926635822868473, | |
| "grad_norm": 2.188422918319702, | |
| "learning_rate": 1.7048794826572608e-05, | |
| "loss": 1.5593, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 6.953073364177132, | |
| "grad_norm": 2.146476984024048, | |
| "learning_rate": 1.6901822457378013e-05, | |
| "loss": 1.6021, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 6.97951090548579, | |
| "grad_norm": 2.322460174560547, | |
| "learning_rate": 1.6754850088183422e-05, | |
| "loss": 1.5485, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.998017184401851, | |
| "eval_accuracy": 0.31984394396169535, | |
| "eval_loss": 1.6719752550125122, | |
| "eval_runtime": 687.5385, | |
| "eval_samples_per_second": 41.009, | |
| "eval_steps_per_second": 0.641, | |
| "step": 2647 | |
| }, | |
| { | |
| "epoch": 7.005948446794448, | |
| "grad_norm": 2.46777606010437, | |
| "learning_rate": 1.660787771898883e-05, | |
| "loss": 1.5634, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 7.0323859881031066, | |
| "grad_norm": 2.2213032245635986, | |
| "learning_rate": 1.646090534979424e-05, | |
| "loss": 1.5148, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 2.3514413833618164, | |
| "learning_rate": 1.6313932980599646e-05, | |
| "loss": 1.4979, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 7.085261070720423, | |
| "grad_norm": 2.6042823791503906, | |
| "learning_rate": 1.6166960611405058e-05, | |
| "loss": 1.5066, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 7.111698612029081, | |
| "grad_norm": 2.6137919425964355, | |
| "learning_rate": 1.6019988242210467e-05, | |
| "loss": 1.5394, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 7.138136153337739, | |
| "grad_norm": 2.8564794063568115, | |
| "learning_rate": 1.5873015873015872e-05, | |
| "loss": 1.5452, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 7.164573694646398, | |
| "grad_norm": 2.685819149017334, | |
| "learning_rate": 1.5726043503821285e-05, | |
| "loss": 1.5514, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 7.191011235955056, | |
| "grad_norm": 2.6131200790405273, | |
| "learning_rate": 1.557907113462669e-05, | |
| "loss": 1.5208, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 7.217448777263715, | |
| "grad_norm": 2.341998338699341, | |
| "learning_rate": 1.54320987654321e-05, | |
| "loss": 1.525, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 7.243886318572373, | |
| "grad_norm": 2.3170387744903564, | |
| "learning_rate": 1.5285126396237508e-05, | |
| "loss": 1.5363, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 7.270323859881031, | |
| "grad_norm": 2.702817440032959, | |
| "learning_rate": 1.5138154027042917e-05, | |
| "loss": 1.5114, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 7.2967614011896895, | |
| "grad_norm": 2.3331921100616455, | |
| "learning_rate": 1.4991181657848324e-05, | |
| "loss": 1.5078, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 7.3231989424983475, | |
| "grad_norm": 3.1344053745269775, | |
| "learning_rate": 1.4844209288653735e-05, | |
| "loss": 1.5251, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 7.349636483807006, | |
| "grad_norm": 2.3613967895507812, | |
| "learning_rate": 1.4697236919459142e-05, | |
| "loss": 1.5444, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 7.376074025115664, | |
| "grad_norm": 2.8878893852233887, | |
| "learning_rate": 1.455026455026455e-05, | |
| "loss": 1.5078, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 7.402511566424322, | |
| "grad_norm": 2.3623602390289307, | |
| "learning_rate": 1.440329218106996e-05, | |
| "loss": 1.5497, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 7.428949107732981, | |
| "grad_norm": 2.708502769470215, | |
| "learning_rate": 1.4256319811875369e-05, | |
| "loss": 1.5249, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 7.455386649041639, | |
| "grad_norm": 2.521937847137451, | |
| "learning_rate": 1.4109347442680776e-05, | |
| "loss": 1.5556, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 7.481824190350298, | |
| "grad_norm": 2.3380672931671143, | |
| "learning_rate": 1.3962375073486187e-05, | |
| "loss": 1.5295, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 7.508261731658956, | |
| "grad_norm": 2.609221935272217, | |
| "learning_rate": 1.3815402704291594e-05, | |
| "loss": 1.5173, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 7.534699272967614, | |
| "grad_norm": 2.692725658416748, | |
| "learning_rate": 1.3668430335097001e-05, | |
| "loss": 1.5256, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 7.5611368142762725, | |
| "grad_norm": 2.8412363529205322, | |
| "learning_rate": 1.3521457965902412e-05, | |
| "loss": 1.5213, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 7.58757435558493, | |
| "grad_norm": 2.349609375, | |
| "learning_rate": 1.3374485596707819e-05, | |
| "loss": 1.5245, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.614011896893589, | |
| "grad_norm": 2.3491156101226807, | |
| "learning_rate": 1.3227513227513228e-05, | |
| "loss": 1.5397, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 7.640449438202247, | |
| "grad_norm": 2.2094242572784424, | |
| "learning_rate": 1.3080540858318637e-05, | |
| "loss": 1.5225, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 7.666886979510905, | |
| "grad_norm": 2.7071151733398438, | |
| "learning_rate": 1.2933568489124046e-05, | |
| "loss": 1.5397, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.693324520819564, | |
| "grad_norm": 2.4890148639678955, | |
| "learning_rate": 1.2786596119929453e-05, | |
| "loss": 1.5188, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 7.719762062128222, | |
| "grad_norm": 2.607391595840454, | |
| "learning_rate": 1.2639623750734864e-05, | |
| "loss": 1.5243, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 7.746199603436881, | |
| "grad_norm": 2.255218505859375, | |
| "learning_rate": 1.2492651381540271e-05, | |
| "loss": 1.4923, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 7.772637144745539, | |
| "grad_norm": 2.095276355743408, | |
| "learning_rate": 1.2345679012345678e-05, | |
| "loss": 1.5154, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 7.799074686054197, | |
| "grad_norm": 2.3163740634918213, | |
| "learning_rate": 1.2198706643151087e-05, | |
| "loss": 1.5273, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 7.8255122273628555, | |
| "grad_norm": 2.356632709503174, | |
| "learning_rate": 1.2051734273956498e-05, | |
| "loss": 1.5434, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 7.851949768671513, | |
| "grad_norm": 2.6374380588531494, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 1.5443, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 7.878387309980172, | |
| "grad_norm": 2.324978828430176, | |
| "learning_rate": 1.1757789535567314e-05, | |
| "loss": 1.5277, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 7.90482485128883, | |
| "grad_norm": 2.326624870300293, | |
| "learning_rate": 1.1610817166372723e-05, | |
| "loss": 1.5392, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 7.931262392597488, | |
| "grad_norm": 2.3055641651153564, | |
| "learning_rate": 1.146384479717813e-05, | |
| "loss": 1.5304, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.957699933906147, | |
| "grad_norm": 2.652165174484253, | |
| "learning_rate": 1.1316872427983539e-05, | |
| "loss": 1.4925, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 7.984137475214805, | |
| "grad_norm": 2.6190438270568848, | |
| "learning_rate": 1.1169900058788948e-05, | |
| "loss": 1.5133, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.3199148785245611, | |
| "eval_loss": 1.6810516119003296, | |
| "eval_runtime": 655.8832, | |
| "eval_samples_per_second": 42.988, | |
| "eval_steps_per_second": 0.672, | |
| "step": 3026 | |
| }, | |
| { | |
| "epoch": 8.010575016523463, | |
| "grad_norm": 2.4034554958343506, | |
| "learning_rate": 1.1022927689594357e-05, | |
| "loss": 1.4946, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 8.03701255783212, | |
| "grad_norm": 2.3963074684143066, | |
| "learning_rate": 1.0875955320399766e-05, | |
| "loss": 1.4916, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 8.06345009914078, | |
| "grad_norm": 2.228837728500366, | |
| "learning_rate": 1.0728982951205175e-05, | |
| "loss": 1.4966, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 8.089887640449438, | |
| "grad_norm": 2.434828042984009, | |
| "learning_rate": 1.0582010582010582e-05, | |
| "loss": 1.476, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 8.116325181758096, | |
| "grad_norm": 3.0598857402801514, | |
| "learning_rate": 1.0435038212815991e-05, | |
| "loss": 1.509, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 8.142762723066754, | |
| "grad_norm": 2.3611867427825928, | |
| "learning_rate": 1.02880658436214e-05, | |
| "loss": 1.5034, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 8.169200264375412, | |
| "grad_norm": 2.5760085582733154, | |
| "learning_rate": 1.0141093474426807e-05, | |
| "loss": 1.4722, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 8.195637805684072, | |
| "grad_norm": 2.364185333251953, | |
| "learning_rate": 9.994121105232216e-06, | |
| "loss": 1.503, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 8.22207534699273, | |
| "grad_norm": 2.9226784706115723, | |
| "learning_rate": 9.847148736037625e-06, | |
| "loss": 1.5095, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 8.248512888301388, | |
| "grad_norm": 2.599740505218506, | |
| "learning_rate": 9.700176366843034e-06, | |
| "loss": 1.5113, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 8.274950429610046, | |
| "grad_norm": 2.4345662593841553, | |
| "learning_rate": 9.553203997648443e-06, | |
| "loss": 1.4932, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 8.301387970918704, | |
| "grad_norm": 2.821202516555786, | |
| "learning_rate": 9.406231628453852e-06, | |
| "loss": 1.4873, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 8.327825512227363, | |
| "grad_norm": 2.3733325004577637, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 1.4714, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 8.354263053536021, | |
| "grad_norm": 2.5024616718292236, | |
| "learning_rate": 9.112286890064668e-06, | |
| "loss": 1.4943, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 8.38070059484468, | |
| "grad_norm": 2.6110079288482666, | |
| "learning_rate": 8.965314520870077e-06, | |
| "loss": 1.4889, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 8.407138136153337, | |
| "grad_norm": 2.432016611099243, | |
| "learning_rate": 8.818342151675484e-06, | |
| "loss": 1.498, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 8.433575677461995, | |
| "grad_norm": 2.565856695175171, | |
| "learning_rate": 8.671369782480893e-06, | |
| "loss": 1.5063, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 8.460013218770655, | |
| "grad_norm": 2.3665668964385986, | |
| "learning_rate": 8.524397413286304e-06, | |
| "loss": 1.4984, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 8.486450760079313, | |
| "grad_norm": 2.3754770755767822, | |
| "learning_rate": 8.377425044091711e-06, | |
| "loss": 1.4884, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 8.51288830138797, | |
| "grad_norm": 2.3668394088745117, | |
| "learning_rate": 8.23045267489712e-06, | |
| "loss": 1.5134, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 8.539325842696629, | |
| "grad_norm": 2.5414023399353027, | |
| "learning_rate": 8.083480305702529e-06, | |
| "loss": 1.4838, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 8.565763384005287, | |
| "grad_norm": 3.5699350833892822, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 1.5143, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 8.592200925313946, | |
| "grad_norm": 2.8063623905181885, | |
| "learning_rate": 7.789535567313345e-06, | |
| "loss": 1.496, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 8.618638466622604, | |
| "grad_norm": 2.588224172592163, | |
| "learning_rate": 7.642563198118754e-06, | |
| "loss": 1.4791, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 8.645076007931262, | |
| "grad_norm": 2.4756646156311035, | |
| "learning_rate": 7.495590828924162e-06, | |
| "loss": 1.5166, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 8.67151354923992, | |
| "grad_norm": 2.575335741043091, | |
| "learning_rate": 7.348618459729571e-06, | |
| "loss": 1.4806, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.697951090548578, | |
| "grad_norm": 2.5106089115142822, | |
| "learning_rate": 7.20164609053498e-06, | |
| "loss": 1.517, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 8.724388631857238, | |
| "grad_norm": 2.419461727142334, | |
| "learning_rate": 7.054673721340388e-06, | |
| "loss": 1.5046, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.750826173165896, | |
| "grad_norm": 2.6011338233947754, | |
| "learning_rate": 6.907701352145797e-06, | |
| "loss": 1.51, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 8.777263714474554, | |
| "grad_norm": 3.790093183517456, | |
| "learning_rate": 6.760728982951206e-06, | |
| "loss": 1.4779, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 8.803701255783212, | |
| "grad_norm": 2.2048144340515137, | |
| "learning_rate": 6.613756613756614e-06, | |
| "loss": 1.5019, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 8.83013879709187, | |
| "grad_norm": 2.383060932159424, | |
| "learning_rate": 6.466784244562023e-06, | |
| "loss": 1.4841, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 8.85657633840053, | |
| "grad_norm": 2.647857666015625, | |
| "learning_rate": 6.319811875367432e-06, | |
| "loss": 1.486, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 8.883013879709187, | |
| "grad_norm": 2.314814805984497, | |
| "learning_rate": 6.172839506172839e-06, | |
| "loss": 1.4847, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 8.909451421017845, | |
| "grad_norm": 2.521620988845825, | |
| "learning_rate": 6.025867136978249e-06, | |
| "loss": 1.4983, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 8.935888962326503, | |
| "grad_norm": 2.53841495513916, | |
| "learning_rate": 5.878894767783657e-06, | |
| "loss": 1.5085, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 8.962326503635161, | |
| "grad_norm": 3.0474183559417725, | |
| "learning_rate": 5.731922398589065e-06, | |
| "loss": 1.5159, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 8.98876404494382, | |
| "grad_norm": 2.3993120193481445, | |
| "learning_rate": 5.584950029394474e-06, | |
| "loss": 1.5001, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.999339061467284, | |
| "eval_accuracy": 0.32094342968611456, | |
| "eval_loss": 1.6820588111877441, | |
| "eval_runtime": 644.8259, | |
| "eval_samples_per_second": 43.725, | |
| "eval_steps_per_second": 0.684, | |
| "step": 3404 | |
| }, | |
| { | |
| "epoch": 9.015201586252479, | |
| "grad_norm": 2.3832859992980957, | |
| "learning_rate": 5.437977660199883e-06, | |
| "loss": 1.4755, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 9.041639127561137, | |
| "grad_norm": 2.3615214824676514, | |
| "learning_rate": 5.291005291005291e-06, | |
| "loss": 1.4527, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 9.068076668869795, | |
| "grad_norm": 2.408276319503784, | |
| "learning_rate": 5.1440329218107e-06, | |
| "loss": 1.5007, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 9.094514210178453, | |
| "grad_norm": 3.026998281478882, | |
| "learning_rate": 4.997060552616108e-06, | |
| "loss": 1.4759, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 9.120951751487112, | |
| "grad_norm": 2.7928555011749268, | |
| "learning_rate": 4.850088183421517e-06, | |
| "loss": 1.5078, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 9.14738929279577, | |
| "grad_norm": 2.3969271183013916, | |
| "learning_rate": 4.703115814226926e-06, | |
| "loss": 1.4786, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 9.173826834104428, | |
| "grad_norm": 2.5589447021484375, | |
| "learning_rate": 4.556143445032334e-06, | |
| "loss": 1.4857, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 9.200264375413086, | |
| "grad_norm": 2.3872599601745605, | |
| "learning_rate": 4.409171075837742e-06, | |
| "loss": 1.5131, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 9.226701916721744, | |
| "grad_norm": 3.470109462738037, | |
| "learning_rate": 4.262198706643152e-06, | |
| "loss": 1.4878, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 9.253139458030404, | |
| "grad_norm": 2.545947790145874, | |
| "learning_rate": 4.11522633744856e-06, | |
| "loss": 1.4973, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 9.279576999339062, | |
| "grad_norm": 2.713286876678467, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 1.4771, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 9.30601454064772, | |
| "grad_norm": 2.5866544246673584, | |
| "learning_rate": 3.821281599059377e-06, | |
| "loss": 1.4987, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 9.332452081956378, | |
| "grad_norm": 2.608093023300171, | |
| "learning_rate": 3.6743092298647855e-06, | |
| "loss": 1.4855, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 9.358889623265036, | |
| "grad_norm": 2.363469362258911, | |
| "learning_rate": 3.527336860670194e-06, | |
| "loss": 1.4903, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 9.385327164573695, | |
| "grad_norm": 2.4162986278533936, | |
| "learning_rate": 3.380364491475603e-06, | |
| "loss": 1.4944, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 2.531153440475464, | |
| "learning_rate": 3.2333921222810115e-06, | |
| "loss": 1.4974, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 9.438202247191011, | |
| "grad_norm": 2.674409866333008, | |
| "learning_rate": 3.0864197530864196e-06, | |
| "loss": 1.4544, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 9.46463978849967, | |
| "grad_norm": 2.8534953594207764, | |
| "learning_rate": 2.9394473838918285e-06, | |
| "loss": 1.5179, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 9.491077329808327, | |
| "grad_norm": 2.373920202255249, | |
| "learning_rate": 2.792475014697237e-06, | |
| "loss": 1.487, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 9.517514871116987, | |
| "grad_norm": 2.327130079269409, | |
| "learning_rate": 2.6455026455026455e-06, | |
| "loss": 1.464, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 9.543952412425645, | |
| "grad_norm": 2.580538749694824, | |
| "learning_rate": 2.498530276308054e-06, | |
| "loss": 1.474, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 9.570389953734303, | |
| "grad_norm": 2.5735127925872803, | |
| "learning_rate": 2.351557907113463e-06, | |
| "loss": 1.4443, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 9.59682749504296, | |
| "grad_norm": 2.6845741271972656, | |
| "learning_rate": 2.204585537918871e-06, | |
| "loss": 1.4728, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 9.623265036351619, | |
| "grad_norm": 2.7224180698394775, | |
| "learning_rate": 2.05761316872428e-06, | |
| "loss": 1.4737, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 9.649702577660278, | |
| "grad_norm": 2.6456120014190674, | |
| "learning_rate": 1.9106407995296885e-06, | |
| "loss": 1.4572, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 9.676140118968936, | |
| "grad_norm": 2.3827290534973145, | |
| "learning_rate": 1.763668430335097e-06, | |
| "loss": 1.4906, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 9.702577660277594, | |
| "grad_norm": 2.7378439903259277, | |
| "learning_rate": 1.6166960611405057e-06, | |
| "loss": 1.4844, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 9.729015201586252, | |
| "grad_norm": 2.330042839050293, | |
| "learning_rate": 1.4697236919459143e-06, | |
| "loss": 1.4552, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 9.75545274289491, | |
| "grad_norm": 2.9327447414398193, | |
| "learning_rate": 1.3227513227513228e-06, | |
| "loss": 1.4699, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 9.78189028420357, | |
| "grad_norm": 3.050356149673462, | |
| "learning_rate": 1.1757789535567315e-06, | |
| "loss": 1.4831, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.808327825512228, | |
| "grad_norm": 2.769808053970337, | |
| "learning_rate": 1.02880658436214e-06, | |
| "loss": 1.4886, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 9.834765366820886, | |
| "grad_norm": 2.5177111625671387, | |
| "learning_rate": 8.818342151675485e-07, | |
| "loss": 1.5074, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 9.861202908129544, | |
| "grad_norm": 2.4207041263580322, | |
| "learning_rate": 7.348618459729571e-07, | |
| "loss": 1.4792, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 9.887640449438202, | |
| "grad_norm": 2.842231512069702, | |
| "learning_rate": 5.878894767783657e-07, | |
| "loss": 1.4696, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 9.914077990746861, | |
| "grad_norm": 2.7768394947052, | |
| "learning_rate": 4.4091710758377425e-07, | |
| "loss": 1.4642, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 9.94051553205552, | |
| "grad_norm": 2.7544069290161133, | |
| "learning_rate": 2.9394473838918287e-07, | |
| "loss": 1.4824, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 9.966953073364177, | |
| "grad_norm": 2.5182807445526123, | |
| "learning_rate": 1.4697236919459144e-07, | |
| "loss": 1.4768, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 9.993390614672835, | |
| "grad_norm": 2.31466007232666, | |
| "learning_rate": 0.0, | |
| "loss": 1.4303, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 9.993390614672835, | |
| "eval_accuracy": 0.3192410001773364, | |
| "eval_loss": 1.6835154294967651, | |
| "eval_runtime": 626.4861, | |
| "eval_samples_per_second": 45.005, | |
| "eval_steps_per_second": 0.704, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 9.993390614672835, | |
| "step": 3780, | |
| "total_flos": 7.511248143138256e+19, | |
| "train_loss": 1.6334815527396227, | |
| "train_runtime": 44633.0248, | |
| "train_samples_per_second": 21.683, | |
| "train_steps_per_second": 0.085 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3780, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7.511248143138256e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |