|
{ |
|
"best_metric": 0.9475982532751092, |
|
"best_model_checkpoint": "videomae-base-finetuned-good-gesturePhaseV10/checkpoint-630", |
|
"epoch": 4.193650793650794, |
|
"eval_steps": 500, |
|
"global_step": 630, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.015873015873015872, |
|
"grad_norm": 10.510711669921875, |
|
"learning_rate": 7.936507936507936e-06, |
|
"loss": 1.5865, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.031746031746031744, |
|
"grad_norm": 5.081316947937012, |
|
"learning_rate": 1.5873015873015872e-05, |
|
"loss": 1.19, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.047619047619047616, |
|
"grad_norm": 4.605384349822998, |
|
"learning_rate": 2.380952380952381e-05, |
|
"loss": 0.9454, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.06349206349206349, |
|
"grad_norm": 6.820161819458008, |
|
"learning_rate": 3.1746031746031745e-05, |
|
"loss": 0.905, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.07936507936507936, |
|
"grad_norm": 3.678572654724121, |
|
"learning_rate": 3.968253968253968e-05, |
|
"loss": 0.9165, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.09523809523809523, |
|
"grad_norm": 6.76840353012085, |
|
"learning_rate": 4.761904761904762e-05, |
|
"loss": 1.061, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.1111111111111111, |
|
"grad_norm": 3.580310344696045, |
|
"learning_rate": 4.938271604938271e-05, |
|
"loss": 0.835, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.12698412698412698, |
|
"grad_norm": 6.999443054199219, |
|
"learning_rate": 4.850088183421517e-05, |
|
"loss": 0.9638, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.14285714285714285, |
|
"grad_norm": 3.5926501750946045, |
|
"learning_rate": 4.761904761904762e-05, |
|
"loss": 0.9482, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.15873015873015872, |
|
"grad_norm": 3.4658665657043457, |
|
"learning_rate": 4.673721340388007e-05, |
|
"loss": 0.852, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.1746031746031746, |
|
"grad_norm": 5.933653354644775, |
|
"learning_rate": 4.585537918871252e-05, |
|
"loss": 0.9297, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.19047619047619047, |
|
"grad_norm": 3.1261682510375977, |
|
"learning_rate": 4.4973544973544974e-05, |
|
"loss": 1.1344, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.2015873015873016, |
|
"eval_accuracy": 0.6899563318777293, |
|
"eval_accuracy_hold": 0.0, |
|
"eval_accuracy_preparation": 1.0, |
|
"eval_accuracy_recovery": 0.0, |
|
"eval_accuracy_stroke": 0.0, |
|
"eval_accuracy_unknown": 0.0, |
|
"eval_loss": 1.0020831823349, |
|
"eval_runtime": 21.4274, |
|
"eval_samples_per_second": 10.687, |
|
"eval_steps_per_second": 1.353, |
|
"step": 127 |
|
}, |
|
{ |
|
"epoch": 1.0047619047619047, |
|
"grad_norm": 1.5987485647201538, |
|
"learning_rate": 4.409171075837743e-05, |
|
"loss": 0.9775, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 1.0206349206349206, |
|
"grad_norm": 6.478776931762695, |
|
"learning_rate": 4.3209876543209875e-05, |
|
"loss": 0.9226, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 1.0365079365079366, |
|
"grad_norm": 2.5407347679138184, |
|
"learning_rate": 4.232804232804233e-05, |
|
"loss": 0.9558, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 1.0523809523809524, |
|
"grad_norm": 2.95082426071167, |
|
"learning_rate": 4.144620811287478e-05, |
|
"loss": 0.8218, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 1.0682539682539682, |
|
"grad_norm": 8.649039268493652, |
|
"learning_rate": 4.056437389770723e-05, |
|
"loss": 0.9146, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 1.084126984126984, |
|
"grad_norm": 5.632199764251709, |
|
"learning_rate": 3.968253968253968e-05, |
|
"loss": 0.7657, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 1.1, |
|
"grad_norm": 3.284162998199463, |
|
"learning_rate": 3.8800705467372136e-05, |
|
"loss": 0.979, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 1.115873015873016, |
|
"grad_norm": 3.5784647464752197, |
|
"learning_rate": 3.791887125220458e-05, |
|
"loss": 0.74, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.1317460317460317, |
|
"grad_norm": 9.924901008605957, |
|
"learning_rate": 3.7037037037037037e-05, |
|
"loss": 0.8304, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 1.1476190476190475, |
|
"grad_norm": 7.721616744995117, |
|
"learning_rate": 3.615520282186949e-05, |
|
"loss": 0.7795, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 1.1634920634920636, |
|
"grad_norm": 6.16989278793335, |
|
"learning_rate": 3.527336860670194e-05, |
|
"loss": 0.6725, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 1.1793650793650794, |
|
"grad_norm": 5.634149074554443, |
|
"learning_rate": 3.439153439153439e-05, |
|
"loss": 0.5051, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 1.1952380952380952, |
|
"grad_norm": 10.187247276306152, |
|
"learning_rate": 3.3509700176366844e-05, |
|
"loss": 0.5961, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.2015873015873015, |
|
"eval_accuracy": 0.7947598253275109, |
|
"eval_accuracy_hold": 0.2692307692307692, |
|
"eval_accuracy_preparation": 0.9873417721518988, |
|
"eval_accuracy_recovery": 0.058823529411764705, |
|
"eval_accuracy_stroke": 0.0, |
|
"eval_accuracy_unknown": 0.8181818181818182, |
|
"eval_loss": 0.6022052764892578, |
|
"eval_runtime": 18.9231, |
|
"eval_samples_per_second": 12.102, |
|
"eval_steps_per_second": 1.533, |
|
"step": 254 |
|
}, |
|
{ |
|
"epoch": 2.0095238095238095, |
|
"grad_norm": 2.109996795654297, |
|
"learning_rate": 3.262786596119929e-05, |
|
"loss": 0.4873, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 2.0253968253968253, |
|
"grad_norm": 13.33204460144043, |
|
"learning_rate": 3.1746031746031745e-05, |
|
"loss": 0.5243, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 2.041269841269841, |
|
"grad_norm": 6.260931491851807, |
|
"learning_rate": 3.08641975308642e-05, |
|
"loss": 0.4785, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 2.057142857142857, |
|
"grad_norm": 26.649112701416016, |
|
"learning_rate": 2.998236331569665e-05, |
|
"loss": 0.4677, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 2.073015873015873, |
|
"grad_norm": 7.0665283203125, |
|
"learning_rate": 2.91005291005291e-05, |
|
"loss": 0.3441, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 2.088888888888889, |
|
"grad_norm": 9.628989219665527, |
|
"learning_rate": 2.8218694885361552e-05, |
|
"loss": 0.7396, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 2.104761904761905, |
|
"grad_norm": 1.708268165588379, |
|
"learning_rate": 2.7336860670194003e-05, |
|
"loss": 0.5972, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 2.1206349206349207, |
|
"grad_norm": 8.236438751220703, |
|
"learning_rate": 2.6455026455026456e-05, |
|
"loss": 0.4421, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 2.1365079365079365, |
|
"grad_norm": 3.947949171066284, |
|
"learning_rate": 2.5573192239858906e-05, |
|
"loss": 0.3602, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 2.1523809523809523, |
|
"grad_norm": 4.807738304138184, |
|
"learning_rate": 2.4691358024691357e-05, |
|
"loss": 0.3214, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 2.168253968253968, |
|
"grad_norm": 9.600074768066406, |
|
"learning_rate": 2.380952380952381e-05, |
|
"loss": 0.4783, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 2.1841269841269844, |
|
"grad_norm": 19.80223846435547, |
|
"learning_rate": 2.292768959435626e-05, |
|
"loss": 0.2683, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 2.2, |
|
"grad_norm": 0.9186673164367676, |
|
"learning_rate": 2.2045855379188714e-05, |
|
"loss": 0.3453, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 2.2015873015873018, |
|
"eval_accuracy": 0.8777292576419214, |
|
"eval_accuracy_hold": 0.8076923076923077, |
|
"eval_accuracy_preparation": 0.9746835443037974, |
|
"eval_accuracy_recovery": 0.4117647058823529, |
|
"eval_accuracy_stroke": 0.0, |
|
"eval_accuracy_unknown": 0.8636363636363636, |
|
"eval_loss": 0.3925057649612427, |
|
"eval_runtime": 17.6116, |
|
"eval_samples_per_second": 13.003, |
|
"eval_steps_per_second": 1.647, |
|
"step": 381 |
|
}, |
|
{ |
|
"epoch": 3.0142857142857142, |
|
"grad_norm": 11.65360164642334, |
|
"learning_rate": 2.1164021164021164e-05, |
|
"loss": 0.3585, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 3.03015873015873, |
|
"grad_norm": 4.8670454025268555, |
|
"learning_rate": 2.0282186948853614e-05, |
|
"loss": 0.4365, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 3.046031746031746, |
|
"grad_norm": 7.012415409088135, |
|
"learning_rate": 1.9400352733686068e-05, |
|
"loss": 0.183, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 3.0619047619047617, |
|
"grad_norm": 0.5204718112945557, |
|
"learning_rate": 1.8518518518518518e-05, |
|
"loss": 0.3634, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 3.077777777777778, |
|
"grad_norm": 13.590949058532715, |
|
"learning_rate": 1.763668430335097e-05, |
|
"loss": 0.1989, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 3.0936507936507938, |
|
"grad_norm": 7.4863691329956055, |
|
"learning_rate": 1.6754850088183422e-05, |
|
"loss": 0.1979, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 3.1095238095238096, |
|
"grad_norm": 4.570287704467773, |
|
"learning_rate": 1.5873015873015872e-05, |
|
"loss": 0.3445, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 3.1253968253968254, |
|
"grad_norm": 2.023911952972412, |
|
"learning_rate": 1.4991181657848324e-05, |
|
"loss": 0.1738, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 3.141269841269841, |
|
"grad_norm": 0.15230882167816162, |
|
"learning_rate": 1.4109347442680776e-05, |
|
"loss": 0.1889, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 3.157142857142857, |
|
"grad_norm": 0.4252125322818756, |
|
"learning_rate": 1.3227513227513228e-05, |
|
"loss": 0.1824, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 3.173015873015873, |
|
"grad_norm": 23.363100051879883, |
|
"learning_rate": 1.2345679012345678e-05, |
|
"loss": 0.2642, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 3.188888888888889, |
|
"grad_norm": 2.36000394821167, |
|
"learning_rate": 1.146384479717813e-05, |
|
"loss": 0.1551, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 3.2015873015873018, |
|
"eval_accuracy": 0.9432314410480349, |
|
"eval_accuracy_hold": 0.9615384615384616, |
|
"eval_accuracy_preparation": 0.9936708860759493, |
|
"eval_accuracy_recovery": 0.7058823529411765, |
|
"eval_accuracy_stroke": 0.16666666666666666, |
|
"eval_accuracy_unknown": 0.9545454545454546, |
|
"eval_loss": 0.2178172767162323, |
|
"eval_runtime": 17.2686, |
|
"eval_samples_per_second": 13.261, |
|
"eval_steps_per_second": 1.679, |
|
"step": 508 |
|
}, |
|
{ |
|
"epoch": 4.003174603174603, |
|
"grad_norm": 0.14277657866477966, |
|
"learning_rate": 1.0582010582010582e-05, |
|
"loss": 0.1695, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 4.019047619047619, |
|
"grad_norm": 0.6571791768074036, |
|
"learning_rate": 9.700176366843034e-06, |
|
"loss": 0.0663, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 4.034920634920635, |
|
"grad_norm": 40.71855926513672, |
|
"learning_rate": 8.818342151675484e-06, |
|
"loss": 0.2735, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 4.050793650793651, |
|
"grad_norm": 12.299385070800781, |
|
"learning_rate": 7.936507936507936e-06, |
|
"loss": 0.0465, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 4.066666666666666, |
|
"grad_norm": 0.19348235428333282, |
|
"learning_rate": 7.054673721340388e-06, |
|
"loss": 0.0676, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 4.082539682539682, |
|
"grad_norm": 0.1367979794740677, |
|
"learning_rate": 6.172839506172839e-06, |
|
"loss": 0.0643, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 4.098412698412698, |
|
"grad_norm": 10.41691780090332, |
|
"learning_rate": 5.291005291005291e-06, |
|
"loss": 0.0992, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 4.114285714285714, |
|
"grad_norm": 21.60604476928711, |
|
"learning_rate": 4.409171075837742e-06, |
|
"loss": 0.2483, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 4.130158730158731, |
|
"grad_norm": 0.17013657093048096, |
|
"learning_rate": 3.527336860670194e-06, |
|
"loss": 0.0683, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 4.146031746031746, |
|
"grad_norm": 0.12723946571350098, |
|
"learning_rate": 2.6455026455026455e-06, |
|
"loss": 0.0116, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 4.161904761904762, |
|
"grad_norm": 0.1051618903875351, |
|
"learning_rate": 1.763668430335097e-06, |
|
"loss": 0.1836, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 4.177777777777778, |
|
"grad_norm": 8.298240661621094, |
|
"learning_rate": 8.818342151675485e-07, |
|
"loss": 0.0978, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 4.193650793650794, |
|
"grad_norm": 0.7603585124015808, |
|
"learning_rate": 0.0, |
|
"loss": 0.1213, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 4.193650793650794, |
|
"eval_accuracy": 0.9475982532751092, |
|
"eval_accuracy_hold": 0.9615384615384616, |
|
"eval_accuracy_preparation": 1.0, |
|
"eval_accuracy_recovery": 0.7058823529411765, |
|
"eval_accuracy_stroke": 0.6666666666666666, |
|
"eval_accuracy_unknown": 0.8181818181818182, |
|
"eval_loss": 0.2031552791595459, |
|
"eval_runtime": 17.2621, |
|
"eval_samples_per_second": 13.266, |
|
"eval_steps_per_second": 1.68, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 4.193650793650794, |
|
"step": 630, |
|
"total_flos": 6.275349161424323e+18, |
|
"train_loss": 0.5261434914100738, |
|
"train_runtime": 1007.5783, |
|
"train_samples_per_second": 5.002, |
|
"train_steps_per_second": 0.625 |
|
}, |
|
{ |
|
"epoch": 4.193650793650794, |
|
"eval_accuracy": 0.9253112033195021, |
|
"eval_accuracy_hold": 1.0, |
|
"eval_accuracy_preparation": 1.0, |
|
"eval_accuracy_recovery": 0.7894736842105263, |
|
"eval_accuracy_stroke": 0.42857142857142855, |
|
"eval_accuracy_unknown": 0.6428571428571429, |
|
"eval_loss": 0.31221458315849304, |
|
"eval_runtime": 24.503, |
|
"eval_samples_per_second": 9.836, |
|
"eval_steps_per_second": 1.265, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 4.193650793650794, |
|
"eval_accuracy": 0.9253112033195021, |
|
"eval_accuracy_hold": 1.0, |
|
"eval_accuracy_preparation": 1.0, |
|
"eval_accuracy_recovery": 0.7894736842105263, |
|
"eval_accuracy_stroke": 0.42857142857142855, |
|
"eval_accuracy_unknown": 0.6428571428571429, |
|
"eval_loss": 0.31221458315849304, |
|
"eval_runtime": 17.9771, |
|
"eval_samples_per_second": 13.406, |
|
"eval_steps_per_second": 1.724, |
|
"step": 630 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 630, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 9223372036854775807, |
|
"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": 6.275349161424323e+18, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|