| { | |
| "best_metric": 0.9857142857142858, | |
| "best_model_checkpoint": "videomae-small-finetuned-kinetics-finetuned-ucf101-subset/checkpoint-300", | |
| "epoch": 3.25, | |
| "eval_steps": 500, | |
| "global_step": 600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 10.107406616210938, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 2.3241, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 9.185954093933105, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.3329, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 9.057476043701172, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.249, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 10.952166557312012, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.1501, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 12.320801734924316, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 1.9362, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 9.885478973388672, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9542, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 11.586761474609375, | |
| "learning_rate": 4.9074074074074075e-05, | |
| "loss": 1.7627, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 14.895970344543457, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.4858, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 12.182781219482422, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 1.5468, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 11.095314979553223, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.3455, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 15.226543426513672, | |
| "learning_rate": 4.5370370370370374e-05, | |
| "loss": 1.2495, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 7.219595432281494, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.2843, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 8.71719741821289, | |
| "learning_rate": 4.351851851851852e-05, | |
| "loss": 1.0719, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 10.525912284851074, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 1.1183, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 11.814860343933105, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 1.0253, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.9131744503974915, | |
| "eval_runtime": 11.1598, | |
| "eval_samples_per_second": 6.273, | |
| "eval_steps_per_second": 3.136, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 8.825668334960938, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.7737, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 7.7434401512146, | |
| "learning_rate": 3.981481481481482e-05, | |
| "loss": 0.8706, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 8.683493614196777, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.6771, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 10.121947288513184, | |
| "learning_rate": 3.7962962962962964e-05, | |
| "loss": 0.5947, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 12.381387710571289, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.5723, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 6.67233419418335, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.517, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 14.212167739868164, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.5525, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 6.088913440704346, | |
| "learning_rate": 3.425925925925926e-05, | |
| "loss": 0.5391, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 3.7612547874450684, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.5433, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 19.784332275390625, | |
| "learning_rate": 3.240740740740741e-05, | |
| "loss": 0.5314, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 3.197401523590088, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.3408, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 17.882139205932617, | |
| "learning_rate": 3.055555555555556e-05, | |
| "loss": 0.4107, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 3.921525239944458, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.2736, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 3.377349615097046, | |
| "learning_rate": 2.8703703703703706e-05, | |
| "loss": 0.3586, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.598513603210449, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.4108, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_accuracy": 0.9857142857142858, | |
| "eval_loss": 0.35705387592315674, | |
| "eval_runtime": 11.0323, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 3.173, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 2.1947929859161377, | |
| "learning_rate": 2.6851851851851855e-05, | |
| "loss": 0.1908, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 5.048274040222168, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.1773, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 27.889989852905273, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3617, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 7.93351936340332, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.1977, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 5.7223687171936035, | |
| "learning_rate": 2.314814814814815e-05, | |
| "loss": 0.294, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.6956955194473267, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.1919, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 1.247668981552124, | |
| "learning_rate": 2.1296296296296296e-05, | |
| "loss": 0.1648, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 2.6172192096710205, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.1725, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.0564138889312744, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.1309, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.1969168186187744, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.1417, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 4.431863784790039, | |
| "learning_rate": 1.7592592592592595e-05, | |
| "loss": 0.1201, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.5482935905456543, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.1148, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 1.0731251239776611, | |
| "learning_rate": 1.574074074074074e-05, | |
| "loss": 0.1323, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 10.92293643951416, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.1361, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 3.3526909351348877, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.1051, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "eval_accuracy": 0.8714285714285714, | |
| "eval_loss": 0.39677634835243225, | |
| "eval_runtime": 10.9089, | |
| "eval_samples_per_second": 6.417, | |
| "eval_steps_per_second": 3.208, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.9733325839042664, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.0886, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 3.1266071796417236, | |
| "learning_rate": 1.2037037037037037e-05, | |
| "loss": 0.084, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.8183128237724304, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0884, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.6514427065849304, | |
| "learning_rate": 1.0185185185185185e-05, | |
| "loss": 0.084, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.8285560011863708, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.1815, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.0948920249938965, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.0902, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.9514235854148865, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.1251, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 3.346337080001831, | |
| "learning_rate": 6.481481481481481e-06, | |
| "loss": 0.0929, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.788431704044342, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.1483, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.8212040662765503, | |
| "learning_rate": 4.6296296296296296e-06, | |
| "loss": 0.0753, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.6672505736351013, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0918, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7973439693450928, | |
| "learning_rate": 2.777777777777778e-06, | |
| "loss": 0.0858, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.8263640999794006, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0764, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.7566823363304138, | |
| "learning_rate": 9.259259259259259e-07, | |
| "loss": 0.0951, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 3.032885789871216, | |
| "learning_rate": 0.0, | |
| "loss": 0.1796, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_accuracy": 0.9857142857142858, | |
| "eval_loss": 0.12802168726921082, | |
| "eval_runtime": 11.5146, | |
| "eval_samples_per_second": 6.079, | |
| "eval_steps_per_second": 3.04, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "step": 600, | |
| "total_flos": 3.794829164937216e+17, | |
| "train_loss": 0.6170218646526336, | |
| "train_runtime": 427.577, | |
| "train_samples_per_second": 2.807, | |
| "train_steps_per_second": 1.403 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.4149034917354584, | |
| "eval_runtime": 25.1037, | |
| "eval_samples_per_second": 6.174, | |
| "eval_steps_per_second": 3.107, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.414903461933136, | |
| "eval_runtime": 24.3949, | |
| "eval_samples_per_second": 6.354, | |
| "eval_steps_per_second": 3.197, | |
| "step": 600 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 3.794829164937216e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |