| { | |
| "best_metric": 0.049635350704193115, | |
| "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": 0.055947333574295044, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.013, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.04776504635810852, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.0105, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 4.179445266723633, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.0658, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.04383094608783722, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.0468, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.04349808394908905, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.0089, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.036882348358631134, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.0417, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.31804150342941284, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.0229, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.039565134793519974, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0808, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.05417277663946152, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0721, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.035959456115961075, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0077, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9323308270676691, | |
| "eval_loss": 0.24353064596652985, | |
| "eval_runtime": 1.2699, | |
| "eval_samples_per_second": 104.732, | |
| "eval_steps_per_second": 13.387, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.03680484741926193, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.0748, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.03853499889373779, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.0551, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.026426393538713455, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0056, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 1.0170791149139404, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0055, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.04796629026532173, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0046, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.023357488214969635, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.02236676774919033, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0045, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.0214188601821661, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.004, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.050203777849674225, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.004, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.021328149363398552, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0397, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9849624060150376, | |
| "eval_loss": 0.049635350704193115, | |
| "eval_runtime": 1.208, | |
| "eval_samples_per_second": 110.099, | |
| "eval_steps_per_second": 14.073, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.020989395678043365, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.004, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.023574335500597954, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0039, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.022276753559708595, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0039, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.023986103013157845, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0038, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.021982692182064056, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0037, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.022230053320527077, | |
| "learning_rate": 0.0, | |
| "loss": 0.0038, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.022900577806509458, | |
| "train_runtime": 90.8343, | |
| "train_samples_per_second": 45.533, | |
| "train_steps_per_second": 2.862 | |
| } | |
| ], | |
| "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 | |
| } | |