| { | |
| "best_metric": 0.7440476190476191, | |
| "best_model_checkpoint": "vivit-b-16x2-kinetics400-finetuned-kinectic/checkpoint-200", | |
| "epoch": 5.166666666666667, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 44.361122131347656, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.0408, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 41.430381774902344, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.9331, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 31.959531784057617, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5466, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 36.29765701293945, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.1729, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 43.942626953125, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.0732, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.47023809523809523, | |
| "eval_loss": 1.296799659729004, | |
| "eval_runtime": 89.7317, | |
| "eval_samples_per_second": 1.872, | |
| "eval_steps_per_second": 1.872, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 41.87215042114258, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.9702, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 51.68751525878906, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.6609, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 58.199771881103516, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.8115, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 24.880760192871094, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.7691, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 18.063873291015625, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.5128, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.6607142857142857, | |
| "eval_loss": 1.040852665901184, | |
| "eval_runtime": 89.5087, | |
| "eval_samples_per_second": 1.877, | |
| "eval_steps_per_second": 1.877, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 5.521175861358643, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.2528, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 2.120676279067993, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.1885, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.3402693271636963, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.1405, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 1.86380934715271, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.3616, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 25.8436336517334, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.1189, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.6607142857142857, | |
| "eval_loss": 0.9205451011657715, | |
| "eval_runtime": 89.0696, | |
| "eval_samples_per_second": 1.886, | |
| "eval_steps_per_second": 1.886, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.7615412473678589, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.2627, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.7165939211845398, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.5066, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.48741012811660767, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.3714, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.7969440221786499, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0071, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.039452001452445984, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.3398, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_accuracy": 0.7440476190476191, | |
| "eval_loss": 0.7473598718643188, | |
| "eval_runtime": 92.3069, | |
| "eval_samples_per_second": 1.82, | |
| "eval_steps_per_second": 1.82, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.09469372779130936, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0076, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 3.8569278717041016, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.0231, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 34.09035873413086, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.0509, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.07331710308790207, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.6114, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.16509070992469788, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.0102, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_accuracy": 0.6785714285714286, | |
| "eval_loss": 1.097224473953247, | |
| "eval_runtime": 89.5157, | |
| "eval_samples_per_second": 1.877, | |
| "eval_steps_per_second": 1.877, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 37.233829498291016, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.052, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 0.7301992177963257, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0158, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.07097156345844269, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.068, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 41.699851989746094, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.1216, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 0.10490654408931732, | |
| "learning_rate": 0.0, | |
| "loss": 0.0035, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.7142857142857143, | |
| "eval_loss": 0.9135054349899292, | |
| "eval_runtime": 90.4025, | |
| "eval_samples_per_second": 1.858, | |
| "eval_steps_per_second": 1.858, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "step": 300, | |
| "total_flos": 2.35711775029248e+17, | |
| "train_loss": 0.5335011211285988, | |
| "train_runtime": 666.5368, | |
| "train_samples_per_second": 0.45, | |
| "train_steps_per_second": 0.45 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.9263604879379272, | |
| "eval_runtime": 53.6181, | |
| "eval_samples_per_second": 2.947, | |
| "eval_steps_per_second": 2.947, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.7341772151898734, | |
| "eval_loss": 0.9263607263565063, | |
| "eval_runtime": 53.7885, | |
| "eval_samples_per_second": 2.937, | |
| "eval_steps_per_second": 2.937, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 2.35711775029248e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |