| { | |
| "best_metric": 0.8611111111111112, | |
| "best_model_checkpoint": "timesformer-base-finetuned-k400-finetuned-kinetic400-subset-epoch6-num_frame_10_spatial_only/checkpoint-100", | |
| "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": 32.31907272338867, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.6709, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 31.60603904724121, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.4509, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 26.731109619140625, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2292, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 28.381160736083984, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.938, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 8.430110931396484, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.7735, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.7222222222222222, | |
| "eval_loss": 0.7761979699134827, | |
| "eval_runtime": 58.8413, | |
| "eval_samples_per_second": 1.224, | |
| "eval_steps_per_second": 1.224, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 7.695086479187012, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.0949, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.16361366212368011, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.2414, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.19437631964683533, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.1627, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 23.58306121826172, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.674, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.047477759420871735, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.1355, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.8611111111111112, | |
| "eval_loss": 0.4255567789077759, | |
| "eval_runtime": 57.9566, | |
| "eval_samples_per_second": 1.242, | |
| "eval_steps_per_second": 1.242, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.36428090929985046, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.2783, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.09045802056789398, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0191, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.11298607289791107, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.0392, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.1291336566209793, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0113, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.06459435075521469, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0014, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.8194444444444444, | |
| "eval_loss": 0.5817016363143921, | |
| "eval_runtime": 58.0229, | |
| "eval_samples_per_second": 1.241, | |
| "eval_steps_per_second": 1.241, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.023863501846790314, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.071, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.033055633306503296, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.0103, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.06424976140260696, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0079, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.019295834004878998, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0015, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.02555103413760662, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.001, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_accuracy": 0.8611111111111112, | |
| "eval_loss": 0.4200752079486847, | |
| "eval_runtime": 58.4996, | |
| "eval_samples_per_second": 1.231, | |
| "eval_steps_per_second": 1.231, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 5.690987586975098, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0113, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 0.04108642414212227, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.0018, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.013796760700643063, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.001, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.015395864844322205, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0212, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.038583505898714066, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.0013, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_accuracy": 0.8472222222222222, | |
| "eval_loss": 0.38439834117889404, | |
| "eval_runtime": 57.5706, | |
| "eval_samples_per_second": 1.251, | |
| "eval_steps_per_second": 1.251, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 0.7331423759460449, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0018, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 1.2412992715835571, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0087, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.016475042328238487, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0006, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 0.012038950808346272, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.047, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 0.020063335075974464, | |
| "learning_rate": 0.0, | |
| "loss": 0.0007, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.8611111111111112, | |
| "eval_loss": 0.36808598041534424, | |
| "eval_runtime": 58.1898, | |
| "eval_samples_per_second": 1.237, | |
| "eval_steps_per_second": 1.237, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "step": 300, | |
| "total_flos": 2.32503028660224e+17, | |
| "train_loss": 0.26357639746895684, | |
| "train_runtime": 573.7148, | |
| "train_samples_per_second": 0.523, | |
| "train_steps_per_second": 0.523 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.8235294117647058, | |
| "eval_loss": 0.536857008934021, | |
| "eval_runtime": 34.3107, | |
| "eval_samples_per_second": 1.982, | |
| "eval_steps_per_second": 1.982, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.8235294117647058, | |
| "eval_loss": 0.536857008934021, | |
| "eval_runtime": 31.8618, | |
| "eval_samples_per_second": 2.134, | |
| "eval_steps_per_second": 2.134, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 2.32503028660224e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |