| { | |
| "best_metric": 0.30522507429122925, | |
| "best_model_checkpoint": "finetuned-ai-real/checkpoint-25", | |
| "epoch": 4.0, | |
| "eval_steps": 25, | |
| "global_step": 176, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 7.896988868713379, | |
| "learning_rate": 0.00019090909090909092, | |
| "loss": 0.5879, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 9.905610084533691, | |
| "learning_rate": 0.00017954545454545456, | |
| "loss": 0.5817, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "eval_accuracy": 0.8911290322580645, | |
| "eval_loss": 0.30522507429122925, | |
| "eval_runtime": 4.9056, | |
| "eval_samples_per_second": 50.555, | |
| "eval_steps_per_second": 6.319, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 7.3549628257751465, | |
| "learning_rate": 0.0001681818181818182, | |
| "loss": 0.3464, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 2.119649887084961, | |
| "learning_rate": 0.00015681818181818182, | |
| "loss": 0.2422, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 10.291028022766113, | |
| "learning_rate": 0.00014545454545454546, | |
| "loss": 0.2006, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "eval_accuracy": 0.8911290322580645, | |
| "eval_loss": 0.3756392002105713, | |
| "eval_runtime": 4.8724, | |
| "eval_samples_per_second": 50.899, | |
| "eval_steps_per_second": 6.362, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 1.0470197200775146, | |
| "learning_rate": 0.0001340909090909091, | |
| "loss": 0.2003, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.5909090909090908, | |
| "grad_norm": 2.770779848098755, | |
| "learning_rate": 0.00012272727272727272, | |
| "loss": 0.2175, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.7045454545454546, | |
| "eval_accuracy": 0.8830645161290323, | |
| "eval_loss": 0.3168693780899048, | |
| "eval_runtime": 4.5227, | |
| "eval_samples_per_second": 54.834, | |
| "eval_steps_per_second": 6.854, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 6.933122158050537, | |
| "learning_rate": 0.00011136363636363636, | |
| "loss": 0.1467, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.0454545454545454, | |
| "grad_norm": 2.787731885910034, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1328, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 2.367295980453491, | |
| "learning_rate": 8.863636363636364e-05, | |
| "loss": 0.07, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "eval_accuracy": 0.9112903225806451, | |
| "eval_loss": 0.363909512758255, | |
| "eval_runtime": 5.2997, | |
| "eval_samples_per_second": 46.795, | |
| "eval_steps_per_second": 5.849, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.1476662158966064, | |
| "learning_rate": 7.727272727272727e-05, | |
| "loss": 0.1031, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 1.065962553024292, | |
| "learning_rate": 6.59090909090909e-05, | |
| "loss": 0.0568, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.840909090909091, | |
| "eval_accuracy": 0.9233870967741935, | |
| "eval_loss": 0.34076112508773804, | |
| "eval_runtime": 4.5072, | |
| "eval_samples_per_second": 55.023, | |
| "eval_steps_per_second": 6.878, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.9545454545454546, | |
| "grad_norm": 10.271200180053711, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.1321, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 1.2485915422439575, | |
| "learning_rate": 4.318181818181819e-05, | |
| "loss": 0.0432, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.409090909090909, | |
| "grad_norm": 6.485482215881348, | |
| "learning_rate": 3.181818181818182e-05, | |
| "loss": 0.0793, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.409090909090909, | |
| "eval_accuracy": 0.907258064516129, | |
| "eval_loss": 0.3802954852581024, | |
| "eval_runtime": 5.5626, | |
| "eval_samples_per_second": 44.583, | |
| "eval_steps_per_second": 5.573, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.10474708676338196, | |
| "learning_rate": 2.0454545454545457e-05, | |
| "loss": 0.0421, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.8636363636363638, | |
| "grad_norm": 0.7910298109054565, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.0292, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.9772727272727275, | |
| "eval_accuracy": 0.9193548387096774, | |
| "eval_loss": 0.3216198682785034, | |
| "eval_runtime": 4.5692, | |
| "eval_samples_per_second": 54.277, | |
| "eval_steps_per_second": 6.785, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 176, | |
| "total_flos": 1.0966148095175885e+17, | |
| "train_loss": 0.1836181137372147, | |
| "train_runtime": 150.901, | |
| "train_samples_per_second": 37.137, | |
| "train_steps_per_second": 1.166 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 176, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0966148095175885e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |