| { | |
| "best_metric": 0.0027912899386137724, | |
| "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": 0.02252030000090599, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.0044, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.019246652722358704, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.0037, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.020115964114665985, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.0035, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.017674697563052177, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.0032, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.016097500920295715, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.003, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.015334432013332844, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.0029, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.015640193596482277, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.0027, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.014179477468132973, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0026, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.013374688103795052, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0024, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.014157657511532307, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0023, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.003015503752976656, | |
| "eval_runtime": 1.2677, | |
| "eval_samples_per_second": 104.914, | |
| "eval_steps_per_second": 13.41, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.01246692705899477, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.0023, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.012060105800628662, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.0022, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.011765383183956146, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0022, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.011269940994679928, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0021, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.011624449864029884, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.002, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.011689254082739353, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.002, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.010906362906098366, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0019, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.010598276741802692, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0019, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.010676290839910507, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0019, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.010374901816248894, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0018, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.0027912899386137724, | |
| "eval_runtime": 1.2657, | |
| "eval_samples_per_second": 105.081, | |
| "eval_steps_per_second": 13.431, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.01027376763522625, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0018, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.010400802828371525, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0018, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.01057503093034029, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0018, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.01048610545694828, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0018, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.01011856459081173, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0018, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.009878523647785187, | |
| "learning_rate": 0.0, | |
| "loss": 0.0018, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.0023681528221529266, | |
| "train_runtime": 85.2671, | |
| "train_samples_per_second": 48.506, | |
| "train_steps_per_second": 3.049 | |
| } | |
| ], | |
| "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 | |
| } | |