| { | |
| "best_metric": 0.3927362263202667, | |
| "best_model_checkpoint": "limb_classification_person_crop_seq/t4_8heads_1layers_1e-4lr/checkpoint-1776", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2960, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16891891891891891, | |
| "grad_norm": 358243.25, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3942, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.33783783783783783, | |
| "grad_norm": 251778.359375, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3603, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5067567567567568, | |
| "grad_norm": 132885.078125, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2986, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 586557.4375, | |
| "learning_rate": 4e-05, | |
| "loss": 1.1933, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8445945945945946, | |
| "grad_norm": 172983.421875, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0563, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8754491017964072, | |
| "eval_loss": 0.5863036513328552, | |
| "eval_runtime": 60.375, | |
| "eval_samples_per_second": 13.83, | |
| "eval_steps_per_second": 0.447, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0135135135135136, | |
| "grad_norm": 176986.609375, | |
| "learning_rate": 6e-05, | |
| "loss": 0.7536, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1824324324324325, | |
| "grad_norm": 141175.546875, | |
| "learning_rate": 7e-05, | |
| "loss": 0.5917, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 171763.8125, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5229, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5202702702702702, | |
| "grad_norm": 235346.625, | |
| "learning_rate": 9e-05, | |
| "loss": 0.4497, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.689189189189189, | |
| "grad_norm": 91142.6953125, | |
| "learning_rate": 0.0001, | |
| "loss": 0.443, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.8581081081081081, | |
| "grad_norm": 174055.0, | |
| "learning_rate": 9.907749077490776e-05, | |
| "loss": 0.4235, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8982035928143712, | |
| "eval_loss": 0.4248776435852051, | |
| "eval_runtime": 60.0285, | |
| "eval_samples_per_second": 13.91, | |
| "eval_steps_per_second": 0.45, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 457668.3125, | |
| "learning_rate": 9.81549815498155e-05, | |
| "loss": 0.4702, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.195945945945946, | |
| "grad_norm": 132193.171875, | |
| "learning_rate": 9.723247232472326e-05, | |
| "loss": 0.4457, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.364864864864865, | |
| "grad_norm": 96091.71875, | |
| "learning_rate": 9.6309963099631e-05, | |
| "loss": 0.4403, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.5337837837837838, | |
| "grad_norm": 161375.140625, | |
| "learning_rate": 9.538745387453874e-05, | |
| "loss": 0.423, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 130612.0078125, | |
| "learning_rate": 9.44649446494465e-05, | |
| "loss": 0.43, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.8716216216216215, | |
| "grad_norm": 238536.953125, | |
| "learning_rate": 9.354243542435425e-05, | |
| "loss": 0.4494, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9089820359281438, | |
| "eval_loss": 0.42482611536979675, | |
| "eval_runtime": 59.7234, | |
| "eval_samples_per_second": 13.981, | |
| "eval_steps_per_second": 0.452, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.0405405405405403, | |
| "grad_norm": 60061.19921875, | |
| "learning_rate": 9.2619926199262e-05, | |
| "loss": 0.4362, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.2094594594594597, | |
| "grad_norm": 96977.171875, | |
| "learning_rate": 9.169741697416975e-05, | |
| "loss": 0.4302, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 98531.53125, | |
| "learning_rate": 9.077490774907749e-05, | |
| "loss": 0.409, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.5472972972972974, | |
| "grad_norm": 101922.609375, | |
| "learning_rate": 8.985239852398525e-05, | |
| "loss": 0.3979, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.7162162162162162, | |
| "grad_norm": 108008.8671875, | |
| "learning_rate": 8.892988929889299e-05, | |
| "loss": 0.4353, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.885135135135135, | |
| "grad_norm": 93226.5390625, | |
| "learning_rate": 8.800738007380073e-05, | |
| "loss": 0.4183, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8934131736526946, | |
| "eval_loss": 0.43285271525382996, | |
| "eval_runtime": 60.1768, | |
| "eval_samples_per_second": 13.876, | |
| "eval_steps_per_second": 0.449, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 4.054054054054054, | |
| "grad_norm": 104521.6328125, | |
| "learning_rate": 8.708487084870849e-05, | |
| "loss": 0.4753, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.222972972972973, | |
| "grad_norm": 67537.2265625, | |
| "learning_rate": 8.616236162361624e-05, | |
| "loss": 0.3917, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.391891891891892, | |
| "grad_norm": 150912.1875, | |
| "learning_rate": 8.523985239852399e-05, | |
| "loss": 0.4119, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.5608108108108105, | |
| "grad_norm": 73587.5078125, | |
| "learning_rate": 8.431734317343174e-05, | |
| "loss": 0.4247, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.72972972972973, | |
| "grad_norm": 115034.75, | |
| "learning_rate": 8.339483394833948e-05, | |
| "loss": 0.4285, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.898648648648649, | |
| "grad_norm": 99327.8828125, | |
| "learning_rate": 8.247232472324724e-05, | |
| "loss": 0.4203, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.918562874251497, | |
| "eval_loss": 0.39791449904441833, | |
| "eval_runtime": 60.3518, | |
| "eval_samples_per_second": 13.836, | |
| "eval_steps_per_second": 0.447, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.0675675675675675, | |
| "grad_norm": 198293.546875, | |
| "learning_rate": 8.154981549815498e-05, | |
| "loss": 0.4155, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 5.236486486486487, | |
| "grad_norm": 82787.34375, | |
| "learning_rate": 8.062730627306274e-05, | |
| "loss": 0.3911, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 5.405405405405405, | |
| "grad_norm": 114961.9765625, | |
| "learning_rate": 7.970479704797048e-05, | |
| "loss": 0.4102, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.574324324324325, | |
| "grad_norm": 153899.546875, | |
| "learning_rate": 7.878228782287823e-05, | |
| "loss": 0.419, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.743243243243243, | |
| "grad_norm": 108072.7109375, | |
| "learning_rate": 7.785977859778598e-05, | |
| "loss": 0.4064, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.912162162162162, | |
| "grad_norm": 73436.2421875, | |
| "learning_rate": 7.693726937269373e-05, | |
| "loss": 0.4073, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9149700598802395, | |
| "eval_loss": 0.39525091648101807, | |
| "eval_runtime": 59.9209, | |
| "eval_samples_per_second": 13.935, | |
| "eval_steps_per_second": 0.451, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 6.081081081081081, | |
| "grad_norm": 109643.8359375, | |
| "learning_rate": 7.601476014760149e-05, | |
| "loss": 0.4227, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 103467.765625, | |
| "learning_rate": 7.509225092250923e-05, | |
| "loss": 0.3968, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 6.418918918918919, | |
| "grad_norm": 114479.0859375, | |
| "learning_rate": 7.416974169741697e-05, | |
| "loss": 0.4118, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.587837837837838, | |
| "grad_norm": 171495.734375, | |
| "learning_rate": 7.324723247232473e-05, | |
| "loss": 0.4134, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.756756756756757, | |
| "grad_norm": 166568.453125, | |
| "learning_rate": 7.232472324723247e-05, | |
| "loss": 0.3839, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.925675675675675, | |
| "grad_norm": 68976.8828125, | |
| "learning_rate": 7.140221402214023e-05, | |
| "loss": 0.3927, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9209580838323354, | |
| "eval_loss": 0.39880236983299255, | |
| "eval_runtime": 60.3856, | |
| "eval_samples_per_second": 13.828, | |
| "eval_steps_per_second": 0.447, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 7.094594594594595, | |
| "grad_norm": 101431.9921875, | |
| "learning_rate": 7.047970479704797e-05, | |
| "loss": 0.4122, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 7.263513513513513, | |
| "grad_norm": 44332.38671875, | |
| "learning_rate": 6.955719557195572e-05, | |
| "loss": 0.4072, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 7.4324324324324325, | |
| "grad_norm": 120717.484375, | |
| "learning_rate": 6.863468634686348e-05, | |
| "loss": 0.4207, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.601351351351351, | |
| "grad_norm": 141817.53125, | |
| "learning_rate": 6.771217712177122e-05, | |
| "loss": 0.3748, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.77027027027027, | |
| "grad_norm": 108992.875, | |
| "learning_rate": 6.678966789667896e-05, | |
| "loss": 0.3973, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.9391891891891895, | |
| "grad_norm": 125266.5390625, | |
| "learning_rate": 6.586715867158672e-05, | |
| "loss": 0.3931, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9173652694610779, | |
| "eval_loss": 0.4065897464752197, | |
| "eval_runtime": 59.0702, | |
| "eval_samples_per_second": 14.136, | |
| "eval_steps_per_second": 0.457, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 8.108108108108109, | |
| "grad_norm": 79784.984375, | |
| "learning_rate": 6.494464944649446e-05, | |
| "loss": 0.3869, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.277027027027026, | |
| "grad_norm": 58690.61328125, | |
| "learning_rate": 6.402214022140222e-05, | |
| "loss": 0.3796, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 8.445945945945946, | |
| "grad_norm": 79699.40625, | |
| "learning_rate": 6.309963099630997e-05, | |
| "loss": 0.4066, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.614864864864865, | |
| "grad_norm": 116672.609375, | |
| "learning_rate": 6.217712177121771e-05, | |
| "loss": 0.3723, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 8.783783783783784, | |
| "grad_norm": 84205.125, | |
| "learning_rate": 6.125461254612547e-05, | |
| "loss": 0.419, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.952702702702704, | |
| "grad_norm": 194455.796875, | |
| "learning_rate": 6.033210332103322e-05, | |
| "loss": 0.4225, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9221556886227545, | |
| "eval_loss": 0.39390280842781067, | |
| "eval_runtime": 59.2665, | |
| "eval_samples_per_second": 14.089, | |
| "eval_steps_per_second": 0.456, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 9.121621621621621, | |
| "grad_norm": 124840.765625, | |
| "learning_rate": 5.940959409594096e-05, | |
| "loss": 0.3988, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 9.29054054054054, | |
| "grad_norm": 166515.8125, | |
| "learning_rate": 5.848708487084871e-05, | |
| "loss": 0.4045, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 9.45945945945946, | |
| "grad_norm": 69126.875, | |
| "learning_rate": 5.756457564575646e-05, | |
| "loss": 0.3878, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.628378378378379, | |
| "grad_norm": 93407.296875, | |
| "learning_rate": 5.664206642066421e-05, | |
| "loss": 0.3846, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 9.797297297297296, | |
| "grad_norm": 70069.578125, | |
| "learning_rate": 5.5719557195571956e-05, | |
| "loss": 0.4031, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.966216216216216, | |
| "grad_norm": 89726.5859375, | |
| "learning_rate": 5.479704797047971e-05, | |
| "loss": 0.3768, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.911377245508982, | |
| "eval_loss": 0.404273122549057, | |
| "eval_runtime": 59.2046, | |
| "eval_samples_per_second": 14.104, | |
| "eval_steps_per_second": 0.456, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 10.135135135135135, | |
| "grad_norm": 50900.421875, | |
| "learning_rate": 5.387453874538746e-05, | |
| "loss": 0.3971, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.304054054054054, | |
| "grad_norm": 195171.375, | |
| "learning_rate": 5.295202952029521e-05, | |
| "loss": 0.4006, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 10.472972972972974, | |
| "grad_norm": 133812.78125, | |
| "learning_rate": 5.202952029520295e-05, | |
| "loss": 0.3527, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.641891891891891, | |
| "grad_norm": 77933.1328125, | |
| "learning_rate": 5.11070110701107e-05, | |
| "loss": 0.377, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 10.81081081081081, | |
| "grad_norm": 141872.90625, | |
| "learning_rate": 5.018450184501845e-05, | |
| "loss": 0.4182, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.97972972972973, | |
| "grad_norm": 105366.984375, | |
| "learning_rate": 4.92619926199262e-05, | |
| "loss": 0.3918, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9101796407185628, | |
| "eval_loss": 0.4043692350387573, | |
| "eval_runtime": 58.8135, | |
| "eval_samples_per_second": 14.197, | |
| "eval_steps_per_second": 0.459, | |
| "step": 1628 | |
| }, | |
| { | |
| "epoch": 11.14864864864865, | |
| "grad_norm": 102680.625, | |
| "learning_rate": 4.833948339483395e-05, | |
| "loss": 0.4068, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 11.317567567567568, | |
| "grad_norm": 54038.0625, | |
| "learning_rate": 4.74169741697417e-05, | |
| "loss": 0.3948, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 11.486486486486486, | |
| "grad_norm": 104827.5078125, | |
| "learning_rate": 4.6494464944649444e-05, | |
| "loss": 0.3606, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 11.655405405405405, | |
| "grad_norm": 120384.34375, | |
| "learning_rate": 4.55719557195572e-05, | |
| "loss": 0.3815, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 11.824324324324325, | |
| "grad_norm": 236821.96875, | |
| "learning_rate": 4.464944649446495e-05, | |
| "loss": 0.3613, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.993243243243244, | |
| "grad_norm": 83007.3359375, | |
| "learning_rate": 4.37269372693727e-05, | |
| "loss": 0.398, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9173652694610779, | |
| "eval_loss": 0.3927362263202667, | |
| "eval_runtime": 59.8526, | |
| "eval_samples_per_second": 13.951, | |
| "eval_steps_per_second": 0.451, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 12.162162162162161, | |
| "grad_norm": 114086.703125, | |
| "learning_rate": 4.280442804428044e-05, | |
| "loss": 0.3695, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.33108108108108, | |
| "grad_norm": 189440.21875, | |
| "learning_rate": 4.1881918819188195e-05, | |
| "loss": 0.3917, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 77842.9921875, | |
| "learning_rate": 4.0959409594095944e-05, | |
| "loss": 0.3908, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 12.66891891891892, | |
| "grad_norm": 106204.4609375, | |
| "learning_rate": 4.003690036900369e-05, | |
| "loss": 0.4054, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 12.837837837837839, | |
| "grad_norm": 79211.3125, | |
| "learning_rate": 3.911439114391144e-05, | |
| "loss": 0.381, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9173652694610779, | |
| "eval_loss": 0.39785751700401306, | |
| "eval_runtime": 59.3643, | |
| "eval_samples_per_second": 14.066, | |
| "eval_steps_per_second": 0.455, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 13.006756756756756, | |
| "grad_norm": 96323.5703125, | |
| "learning_rate": 3.819188191881919e-05, | |
| "loss": 0.3419, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 13.175675675675675, | |
| "grad_norm": 179315.421875, | |
| "learning_rate": 3.726937269372694e-05, | |
| "loss": 0.3987, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 13.344594594594595, | |
| "grad_norm": 99669.9609375, | |
| "learning_rate": 3.634686346863469e-05, | |
| "loss": 0.3626, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 13.513513513513514, | |
| "grad_norm": 90959.078125, | |
| "learning_rate": 3.542435424354244e-05, | |
| "loss": 0.3744, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.682432432432432, | |
| "grad_norm": 123356.265625, | |
| "learning_rate": 3.4501845018450186e-05, | |
| "loss": 0.3613, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 13.85135135135135, | |
| "grad_norm": 62648.0234375, | |
| "learning_rate": 3.3579335793357934e-05, | |
| "loss": 0.3817, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9137724550898204, | |
| "eval_loss": 0.4134121239185333, | |
| "eval_runtime": 60.0106, | |
| "eval_samples_per_second": 13.914, | |
| "eval_steps_per_second": 0.45, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 14.02027027027027, | |
| "grad_norm": 104165.796875, | |
| "learning_rate": 3.265682656826568e-05, | |
| "loss": 0.3751, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 14.18918918918919, | |
| "grad_norm": 69535.90625, | |
| "learning_rate": 3.173431734317343e-05, | |
| "loss": 0.3675, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.358108108108109, | |
| "grad_norm": 103158.5, | |
| "learning_rate": 3.081180811808118e-05, | |
| "loss": 0.39, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 14.527027027027026, | |
| "grad_norm": 68135.5859375, | |
| "learning_rate": 2.9889298892988933e-05, | |
| "loss": 0.3658, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 14.695945945945946, | |
| "grad_norm": 167864.78125, | |
| "learning_rate": 2.8966789667896682e-05, | |
| "loss": 0.3857, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 14.864864864864865, | |
| "grad_norm": 98201.515625, | |
| "learning_rate": 2.8044280442804427e-05, | |
| "loss": 0.3815, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9173652694610779, | |
| "eval_loss": 0.40676751732826233, | |
| "eval_runtime": 58.4707, | |
| "eval_samples_per_second": 14.281, | |
| "eval_steps_per_second": 0.462, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 15.033783783783784, | |
| "grad_norm": 186007.90625, | |
| "learning_rate": 2.7121771217712176e-05, | |
| "loss": 0.3717, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 15.202702702702704, | |
| "grad_norm": 157017.15625, | |
| "learning_rate": 2.619926199261993e-05, | |
| "loss": 0.3717, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 15.371621621621621, | |
| "grad_norm": 74874.7109375, | |
| "learning_rate": 2.5276752767527677e-05, | |
| "loss": 0.3537, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 15.54054054054054, | |
| "grad_norm": 92439.0078125, | |
| "learning_rate": 2.4354243542435426e-05, | |
| "loss": 0.3535, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 15.70945945945946, | |
| "grad_norm": 164555.34375, | |
| "learning_rate": 2.3431734317343175e-05, | |
| "loss": 0.3645, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 15.878378378378379, | |
| "grad_norm": 91601.59375, | |
| "learning_rate": 2.2509225092250924e-05, | |
| "loss": 0.3774, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9233532934131736, | |
| "eval_loss": 0.39439886808395386, | |
| "eval_runtime": 58.9596, | |
| "eval_samples_per_second": 14.162, | |
| "eval_steps_per_second": 0.458, | |
| "step": 2368 | |
| }, | |
| { | |
| "epoch": 16.0472972972973, | |
| "grad_norm": 82947.125, | |
| "learning_rate": 2.1586715867158673e-05, | |
| "loss": 0.3644, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 16.216216216216218, | |
| "grad_norm": 77509.625, | |
| "learning_rate": 2.066420664206642e-05, | |
| "loss": 0.3418, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.385135135135137, | |
| "grad_norm": 63176.84765625, | |
| "learning_rate": 1.974169741697417e-05, | |
| "loss": 0.3502, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 16.554054054054053, | |
| "grad_norm": 194086.15625, | |
| "learning_rate": 1.881918819188192e-05, | |
| "loss": 0.3788, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 16.722972972972972, | |
| "grad_norm": 81391.1171875, | |
| "learning_rate": 1.7896678966789668e-05, | |
| "loss": 0.347, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 16.89189189189189, | |
| "grad_norm": 34088.36328125, | |
| "learning_rate": 1.6974169741697417e-05, | |
| "loss": 0.3499, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9209580838323354, | |
| "eval_loss": 0.3971590995788574, | |
| "eval_runtime": 59.2527, | |
| "eval_samples_per_second": 14.092, | |
| "eval_steps_per_second": 0.456, | |
| "step": 2516 | |
| }, | |
| { | |
| "epoch": 17.06081081081081, | |
| "grad_norm": 151931.296875, | |
| "learning_rate": 1.6051660516605166e-05, | |
| "loss": 0.3927, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 17.22972972972973, | |
| "grad_norm": 76409.84375, | |
| "learning_rate": 1.5129151291512916e-05, | |
| "loss": 0.3497, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 17.39864864864865, | |
| "grad_norm": 65936.53125, | |
| "learning_rate": 1.4206642066420663e-05, | |
| "loss": 0.363, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 17.56756756756757, | |
| "grad_norm": 50398.3203125, | |
| "learning_rate": 1.3284132841328414e-05, | |
| "loss": 0.3499, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 17.736486486486488, | |
| "grad_norm": 109634.046875, | |
| "learning_rate": 1.2361623616236164e-05, | |
| "loss": 0.337, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 17.905405405405407, | |
| "grad_norm": 40496.69921875, | |
| "learning_rate": 1.1439114391143913e-05, | |
| "loss": 0.3825, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9197604790419162, | |
| "eval_loss": 0.4014655351638794, | |
| "eval_runtime": 59.3898, | |
| "eval_samples_per_second": 14.06, | |
| "eval_steps_per_second": 0.455, | |
| "step": 2664 | |
| }, | |
| { | |
| "epoch": 18.074324324324323, | |
| "grad_norm": 117493.03125, | |
| "learning_rate": 1.0516605166051662e-05, | |
| "loss": 0.3519, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 18.243243243243242, | |
| "grad_norm": 46973.1875, | |
| "learning_rate": 9.59409594095941e-06, | |
| "loss": 0.3237, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 18.41216216216216, | |
| "grad_norm": 114745.4609375, | |
| "learning_rate": 8.67158671586716e-06, | |
| "loss": 0.3714, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 18.58108108108108, | |
| "grad_norm": 38971.9296875, | |
| "learning_rate": 7.749077490774908e-06, | |
| "loss": 0.3231, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "grad_norm": 156186.84375, | |
| "learning_rate": 6.826568265682657e-06, | |
| "loss": 0.3416, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 18.91891891891892, | |
| "grad_norm": 76312.9921875, | |
| "learning_rate": 5.904059040590406e-06, | |
| "loss": 0.3731, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.918562874251497, | |
| "eval_loss": 0.39593929052352905, | |
| "eval_runtime": 58.8099, | |
| "eval_samples_per_second": 14.198, | |
| "eval_steps_per_second": 0.459, | |
| "step": 2812 | |
| }, | |
| { | |
| "epoch": 19.08783783783784, | |
| "grad_norm": 91183.2109375, | |
| "learning_rate": 4.981549815498155e-06, | |
| "loss": 0.3723, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 19.256756756756758, | |
| "grad_norm": 80640.0234375, | |
| "learning_rate": 4.059040590405904e-06, | |
| "loss": 0.3349, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 19.425675675675677, | |
| "grad_norm": 89848.125, | |
| "learning_rate": 3.136531365313653e-06, | |
| "loss": 0.3679, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 19.594594594594593, | |
| "grad_norm": 107155.53125, | |
| "learning_rate": 2.2140221402214023e-06, | |
| "loss": 0.3463, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 19.763513513513512, | |
| "grad_norm": 128204.28125, | |
| "learning_rate": 1.2915129151291513e-06, | |
| "loss": 0.3549, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 19.93243243243243, | |
| "grad_norm": 91720.7734375, | |
| "learning_rate": 3.690036900369004e-07, | |
| "loss": 0.323, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9209580838323354, | |
| "eval_loss": 0.3986169397830963, | |
| "eval_runtime": 58.2229, | |
| "eval_samples_per_second": 14.341, | |
| "eval_steps_per_second": 0.464, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2960, | |
| "total_flos": 0.0, | |
| "train_loss": 0.43191270973231344, | |
| "train_runtime": 11059.7547, | |
| "train_samples_per_second": 8.552, | |
| "train_steps_per_second": 0.268 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2960, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |