| { | |
| "best_metric": 0.02550239861011505, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v5/checkpoint-100", | |
| "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": 0.026073064655065536, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.0059, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.02703898586332798, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.0053, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.025978663936257362, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.0049, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.02143310010433197, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.0039, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.3948848247528076, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.0045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.01829284243285656, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.0047, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.6177886724472046, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.1699, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.0340842679142952, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0802, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.028572693467140198, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0117, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.036357320845127106, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0419, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9924812030075187, | |
| "eval_loss": 0.02550239861011505, | |
| "eval_runtime": 1.9733, | |
| "eval_samples_per_second": 67.398, | |
| "eval_steps_per_second": 8.615, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 9.704477310180664, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.0174, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.05334135890007019, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.0547, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 20.598960876464844, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0576, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.018562769517302513, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0154, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.01950875297188759, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0615, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.017226506024599075, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0123, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.02454870566725731, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.003, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.020279334858059883, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0028, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.013689412735402584, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0031, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.013543753884732723, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0119, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.16051983833312988, | |
| "eval_runtime": 1.2648, | |
| "eval_samples_per_second": 105.151, | |
| "eval_steps_per_second": 13.44, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.01208597794175148, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0024, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.014367583207786083, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0025, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.013976923190057278, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0023, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.013237027451395988, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0023, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.011933455243706703, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0022, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.011974948458373547, | |
| "learning_rate": 0.0, | |
| "loss": 0.0418, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.02407647935816875, | |
| "train_runtime": 86.142, | |
| "train_samples_per_second": 48.014, | |
| "train_steps_per_second": 3.018 | |
| } | |
| ], | |
| "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 | |
| } | |