| { | |
| "best_metric": 0.038147661834955215, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-200", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 2.2591915130615234, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.8686, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.5639485120773315, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.4059, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.5393340587615967, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.2877, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 2.769625425338745, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.3688, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 3.951380729675293, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.1938, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 5.551625728607178, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.2132, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 10.370670318603516, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.1291, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.29360657930374146, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.1539, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.5713837146759033, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0606, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 4.661890983581543, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0847, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9699248120300752, | |
| "eval_loss": 0.10859738290309906, | |
| "eval_runtime": 1.8329, | |
| "eval_samples_per_second": 72.564, | |
| "eval_steps_per_second": 9.275, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.3584635853767395, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.0791, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.10284760594367981, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.1073, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.1232604905962944, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0534, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.27756789326667786, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0607, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 2.3795506954193115, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0334, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 10.455995559692383, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0451, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.562580943107605, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0184, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.10280799120664597, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0402, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.07935391366481781, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0513, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.07218295335769653, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0436, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9849624060150376, | |
| "eval_loss": 0.038147661834955215, | |
| "eval_runtime": 1.2372, | |
| "eval_samples_per_second": 107.501, | |
| "eval_steps_per_second": 13.741, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.0718994140625, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0145, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.06259436905384064, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0121, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.0606728158891201, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0115, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.2018558830022812, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0113, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.05980982258915901, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0115, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0552167184650898, | |
| "learning_rate": 0.0, | |
| "loss": 0.0165, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.12984647212120204, | |
| "train_runtime": 85.925, | |
| "train_samples_per_second": 48.135, | |
| "train_steps_per_second": 3.026 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |