| { | |
| "best_metric": 0.927536231884058, | |
| "best_model_checkpoint": "videomae-base-finetuned-ssv2-finetuned-traffic-dataset-mae/checkpoint-456", | |
| "epoch": 7.125, | |
| "eval_steps": 500, | |
| "global_step": 608, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 10.86192798614502, | |
| "learning_rate": 8.196721311475409e-06, | |
| "loss": 0.6065, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.969607830047607, | |
| "learning_rate": 1.6393442622950818e-05, | |
| "loss": 0.5224, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.2913336753845215, | |
| "learning_rate": 2.459016393442623e-05, | |
| "loss": 0.5223, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.6693388223648071, | |
| "learning_rate": 3.2786885245901635e-05, | |
| "loss": 0.1888, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.9622797966003418, | |
| "learning_rate": 4.098360655737705e-05, | |
| "loss": 0.2521, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1212589740753174, | |
| "learning_rate": 4.918032786885246e-05, | |
| "loss": 0.0138, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.07918030023574829, | |
| "learning_rate": 4.917733089579525e-05, | |
| "loss": 0.0309, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_accuracy": 0.9130434782608695, | |
| "eval_loss": 0.30077555775642395, | |
| "eval_runtime": 121.2516, | |
| "eval_samples_per_second": 0.569, | |
| "eval_steps_per_second": 0.289, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.09372598677873611, | |
| "learning_rate": 4.826325411334552e-05, | |
| "loss": 0.0042, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.016796614974737167, | |
| "learning_rate": 4.73491773308958e-05, | |
| "loss": 0.0007, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.020755194127559662, | |
| "learning_rate": 4.643510054844607e-05, | |
| "loss": 0.3019, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.011669596657156944, | |
| "learning_rate": 4.5521023765996346e-05, | |
| "loss": 0.0003, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.003766011679545045, | |
| "learning_rate": 4.460694698354662e-05, | |
| "loss": 0.0002, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.08908724784851074, | |
| "learning_rate": 4.369287020109689e-05, | |
| "loss": 0.0002, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.005499520804733038, | |
| "learning_rate": 4.2778793418647164e-05, | |
| "loss": 0.0002, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.005470686126500368, | |
| "learning_rate": 4.1864716636197444e-05, | |
| "loss": 0.0002, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 2.1030023097991943, | |
| "eval_runtime": 100.6515, | |
| "eval_samples_per_second": 0.686, | |
| "eval_steps_per_second": 0.348, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.0026209524367004633, | |
| "learning_rate": 4.095063985374772e-05, | |
| "loss": 0.0001, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.009242965839803219, | |
| "learning_rate": 4.003656307129799e-05, | |
| "loss": 0.0001, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.0026453358586877584, | |
| "learning_rate": 3.912248628884826e-05, | |
| "loss": 0.0001, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.0033517158590257168, | |
| "learning_rate": 3.820840950639854e-05, | |
| "loss": 0.0001, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.003627925645560026, | |
| "learning_rate": 3.7294332723948815e-05, | |
| "loss": 0.0001, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.002879067324101925, | |
| "learning_rate": 3.638025594149909e-05, | |
| "loss": 0.0001, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.0036734691821038723, | |
| "learning_rate": 3.546617915904936e-05, | |
| "loss": 0.0001, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_accuracy": 0.7101449275362319, | |
| "eval_loss": 1.84578275680542, | |
| "eval_runtime": 105.6436, | |
| "eval_samples_per_second": 0.653, | |
| "eval_steps_per_second": 0.331, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.0011968453181907535, | |
| "learning_rate": 3.455210237659964e-05, | |
| "loss": 0.0001, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.0024677019100636244, | |
| "learning_rate": 3.3638025594149906e-05, | |
| "loss": 0.0001, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.0011525729205459356, | |
| "learning_rate": 3.2723948811700186e-05, | |
| "loss": 0.0001, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.003028567647561431, | |
| "learning_rate": 3.180987202925046e-05, | |
| "loss": 0.0001, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.0027707908302545547, | |
| "learning_rate": 3.089579524680074e-05, | |
| "loss": 0.0001, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.0017000396037474275, | |
| "learning_rate": 2.9981718464351004e-05, | |
| "loss": 0.0001, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.002608270151540637, | |
| "learning_rate": 2.906764168190128e-05, | |
| "loss": 0.0001, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.0012394236400723457, | |
| "learning_rate": 2.8153564899451557e-05, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "eval_accuracy": 0.7391304347826086, | |
| "eval_loss": 1.5200165510177612, | |
| "eval_runtime": 117.7366, | |
| "eval_samples_per_second": 0.586, | |
| "eval_steps_per_second": 0.297, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 0.00229161255992949, | |
| "learning_rate": 2.7239488117001826e-05, | |
| "loss": 0.0001, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.0013108043931424618, | |
| "learning_rate": 2.6325411334552102e-05, | |
| "loss": 0.0001, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.0012365940492600203, | |
| "learning_rate": 2.541133455210238e-05, | |
| "loss": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.002082814695313573, | |
| "learning_rate": 2.449725776965265e-05, | |
| "loss": 0.0001, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.0035416895989328623, | |
| "learning_rate": 2.3583180987202927e-05, | |
| "loss": 0.0001, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 0.0009142422350123525, | |
| "learning_rate": 2.26691042047532e-05, | |
| "loss": 0.0001, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 0.0014322904171422124, | |
| "learning_rate": 2.1755027422303476e-05, | |
| "loss": 0.0001, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 0.0017242878675460815, | |
| "learning_rate": 2.084095063985375e-05, | |
| "loss": 0.0001, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "eval_accuracy": 0.7536231884057971, | |
| "eval_loss": 1.4568772315979004, | |
| "eval_runtime": 106.8623, | |
| "eval_samples_per_second": 0.646, | |
| "eval_steps_per_second": 0.328, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 0.0011396174086257815, | |
| "learning_rate": 1.9926873857404025e-05, | |
| "loss": 0.0001, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 0.0009178342879749835, | |
| "learning_rate": 1.90127970749543e-05, | |
| "loss": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 0.0034428227227181196, | |
| "learning_rate": 1.809872029250457e-05, | |
| "loss": 0.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 0.0011987154139205813, | |
| "learning_rate": 1.7184643510054847e-05, | |
| "loss": 0.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 0.000779572525061667, | |
| "learning_rate": 1.627056672760512e-05, | |
| "loss": 0.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.0013000366743654013, | |
| "learning_rate": 1.5356489945155393e-05, | |
| "loss": 0.0003, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 0.0019034247379750013, | |
| "learning_rate": 1.4442413162705667e-05, | |
| "loss": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "eval_accuracy": 0.927536231884058, | |
| "eval_loss": 0.39408108592033386, | |
| "eval_runtime": 103.5083, | |
| "eval_samples_per_second": 0.667, | |
| "eval_steps_per_second": 0.338, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "grad_norm": 0.0016774114919826388, | |
| "learning_rate": 1.3528336380255944e-05, | |
| "loss": 0.0005, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 0.0013074069283902645, | |
| "learning_rate": 1.2614259597806216e-05, | |
| "loss": 0.0001, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 0.0015878825215622783, | |
| "learning_rate": 1.1700182815356491e-05, | |
| "loss": 0.0001, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "grad_norm": 0.0008045001304708421, | |
| "learning_rate": 1.0786106032906766e-05, | |
| "loss": 0.0001, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "grad_norm": 0.0012260900111868978, | |
| "learning_rate": 9.872029250457038e-06, | |
| "loss": 0.0002, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.09, | |
| "grad_norm": 0.001507502980530262, | |
| "learning_rate": 8.957952468007313e-06, | |
| "loss": 0.0001, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 6.11, | |
| "grad_norm": 0.001074848580174148, | |
| "learning_rate": 8.043875685557587e-06, | |
| "loss": 0.0001, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "grad_norm": 0.0012807522434741259, | |
| "learning_rate": 7.129798903107861e-06, | |
| "loss": 0.0001, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "eval_accuracy": 0.8695652173913043, | |
| "eval_loss": 0.9657596945762634, | |
| "eval_runtime": 105.1521, | |
| "eval_samples_per_second": 0.656, | |
| "eval_steps_per_second": 0.333, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 0.0008385140681639314, | |
| "learning_rate": 6.2157221206581355e-06, | |
| "loss": 0.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 0.001427665469236672, | |
| "learning_rate": 5.301645338208409e-06, | |
| "loss": 0.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "grad_norm": 0.001054079388268292, | |
| "learning_rate": 4.387568555758684e-06, | |
| "loss": 0.0001, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "grad_norm": 0.0016839707968756557, | |
| "learning_rate": 3.4734917733089582e-06, | |
| "loss": 0.0001, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 0.0011065697763115168, | |
| "learning_rate": 2.5594149908592323e-06, | |
| "loss": 0.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.0027244570665061474, | |
| "learning_rate": 1.6453382084095064e-06, | |
| "loss": 0.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "grad_norm": 0.002320781582966447, | |
| "learning_rate": 7.312614259597807e-07, | |
| "loss": 0.0001, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "eval_accuracy": 0.8405797101449275, | |
| "eval_loss": 0.983595609664917, | |
| "eval_runtime": 105.1841, | |
| "eval_samples_per_second": 0.656, | |
| "eval_steps_per_second": 0.333, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "step": 608, | |
| "total_flos": 1.5152145407976407e+18, | |
| "train_loss": 0.040272084387700156, | |
| "train_runtime": 2933.2373, | |
| "train_samples_per_second": 0.415, | |
| "train_steps_per_second": 0.207 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "eval_accuracy": 0.375, | |
| "eval_loss": 3.811370849609375, | |
| "eval_runtime": 15.0567, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 0.531, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "eval_accuracy": 0.375, | |
| "eval_loss": 3.811370849609375, | |
| "eval_runtime": 14.4555, | |
| "eval_samples_per_second": 1.107, | |
| "eval_steps_per_second": 0.553, | |
| "step": 608 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 608, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 1.5152145407976407e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |