| { | |
| "best_metric": 0.0630112811923027, | |
| "best_model_checkpoint": "beans_model/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": 1.528293251991272, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.7841, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.525876522064209, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.3219, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.951483964920044, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.1955, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.8248862624168396, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.1905, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 4.840185642242432, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.2207, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 3.156651496887207, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.248, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.3212275505065918, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.1409, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.14757445454597473, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0797, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.16617465019226074, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.112, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 1.4776158332824707, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0983, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9624060150375939, | |
| "eval_loss": 0.11257314682006836, | |
| "eval_model_preparation_time": 0.006, | |
| "eval_runtime": 1.2635, | |
| "eval_samples_per_second": 105.261, | |
| "eval_steps_per_second": 13.454, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.17706085741519928, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.1054, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 7.802417278289795, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.1234, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.10370013862848282, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0904, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.1017041876912117, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0363, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.5070149302482605, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0429, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 7.553247928619385, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0324, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 6.309382438659668, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0308, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.35512471199035645, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.019, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.062438324093818665, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0194, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.06177965924143791, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0283, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.0630112811923027, | |
| "eval_model_preparation_time": 0.006, | |
| "eval_runtime": 1.262, | |
| "eval_samples_per_second": 105.388, | |
| "eval_steps_per_second": 13.471, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.05896027013659477, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0116, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.06045134365558624, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0133, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.06165293976664543, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0113, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.057042431086301804, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0115, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.05782578885555267, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0117, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.06298615038394928, | |
| "learning_rate": 0.0, | |
| "loss": 0.0109, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.11501647262619091, | |
| "train_runtime": 105.1039, | |
| "train_samples_per_second": 39.352, | |
| "train_steps_per_second": 2.474 | |
| } | |
| ], | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |