| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 1000, | |
| "global_step": 16509, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00018171906232963838, | |
| "grad_norm": 7.594760894775391, | |
| "learning_rate": 0.0, | |
| "loss": 3.1229, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1817190623296384, | |
| "grad_norm": 6.193614959716797, | |
| "learning_rate": 9.076867163365438e-06, | |
| "loss": 2.6857, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1817190623296384, | |
| "eval_cosine_accuracy": 0.9440529942512512, | |
| "eval_loss": 1.6310006380081177, | |
| "eval_runtime": 21.3914, | |
| "eval_samples_per_second": 444.525, | |
| "eval_steps_per_second": 3.506, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3634381246592768, | |
| "grad_norm": 5.20527458190918, | |
| "learning_rate": 1.8153734326730876e-05, | |
| "loss": 2.0541, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3634381246592768, | |
| "eval_cosine_accuracy": 0.9472079277038574, | |
| "eval_loss": 1.544768214225769, | |
| "eval_runtime": 21.6963, | |
| "eval_samples_per_second": 438.277, | |
| "eval_steps_per_second": 3.457, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5451571869889151, | |
| "grad_norm": 4.887779712677002, | |
| "learning_rate": 2.7239687443212796e-05, | |
| "loss": 1.7335, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5451571869889151, | |
| "eval_cosine_accuracy": 0.948469877243042, | |
| "eval_loss": 1.5235735177993774, | |
| "eval_runtime": 21.262, | |
| "eval_samples_per_second": 447.229, | |
| "eval_steps_per_second": 3.527, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7268762493185535, | |
| "grad_norm": 6.290886878967285, | |
| "learning_rate": 3.632564055969471e-05, | |
| "loss": 1.2495, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7268762493185535, | |
| "eval_cosine_accuracy": 0.9433168768882751, | |
| "eval_loss": 1.5551835298538208, | |
| "eval_runtime": 21.377, | |
| "eval_samples_per_second": 444.824, | |
| "eval_steps_per_second": 3.508, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9085953116481919, | |
| "grad_norm": 8.564718246459961, | |
| "learning_rate": 4.5411593676176635e-05, | |
| "loss": 0.813, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9085953116481919, | |
| "eval_cosine_accuracy": 0.9472079277038574, | |
| "eval_loss": 1.5794354677200317, | |
| "eval_runtime": 21.6842, | |
| "eval_samples_per_second": 438.522, | |
| "eval_steps_per_second": 3.459, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0903143739778303, | |
| "grad_norm": 5.288447856903076, | |
| "learning_rate": 4.8875613301835365e-05, | |
| "loss": 1.0512, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.0903143739778303, | |
| "eval_cosine_accuracy": 0.9566726088523865, | |
| "eval_loss": 1.4543931484222412, | |
| "eval_runtime": 23.0837, | |
| "eval_samples_per_second": 411.936, | |
| "eval_steps_per_second": 3.249, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.2720334363074688, | |
| "grad_norm": 4.507662296295166, | |
| "learning_rate": 4.660412502271488e-05, | |
| "loss": 1.2912, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2720334363074688, | |
| "eval_cosine_accuracy": 0.9562519788742065, | |
| "eval_loss": 1.4492090940475464, | |
| "eval_runtime": 21.5059, | |
| "eval_samples_per_second": 442.159, | |
| "eval_steps_per_second": 3.487, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.453752498637107, | |
| "grad_norm": 4.300741672515869, | |
| "learning_rate": 4.433263674359441e-05, | |
| "loss": 1.1994, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.453752498637107, | |
| "eval_cosine_accuracy": 0.956777811050415, | |
| "eval_loss": 1.4519122838974, | |
| "eval_runtime": 21.4819, | |
| "eval_samples_per_second": 442.651, | |
| "eval_steps_per_second": 3.491, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.6354715609667454, | |
| "grad_norm": 4.133264064788818, | |
| "learning_rate": 4.206569144103217e-05, | |
| "loss": 1.0662, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.6354715609667454, | |
| "eval_cosine_accuracy": 0.9544641971588135, | |
| "eval_loss": 1.4635158777236938, | |
| "eval_runtime": 21.6809, | |
| "eval_samples_per_second": 438.588, | |
| "eval_steps_per_second": 3.459, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.817190623296384, | |
| "grad_norm": 5.683165073394775, | |
| "learning_rate": 3.9794203161911684e-05, | |
| "loss": 0.6724, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.817190623296384, | |
| "eval_cosine_accuracy": 0.9454201459884644, | |
| "eval_loss": 1.571725845336914, | |
| "eval_runtime": 21.3873, | |
| "eval_samples_per_second": 444.61, | |
| "eval_steps_per_second": 3.507, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.998909685626022, | |
| "grad_norm": 7.759120464324951, | |
| "learning_rate": 3.752498637107033e-05, | |
| "loss": 0.4761, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.998909685626022, | |
| "eval_cosine_accuracy": 0.9502576589584351, | |
| "eval_loss": 1.550882339477539, | |
| "eval_runtime": 22.2267, | |
| "eval_samples_per_second": 427.818, | |
| "eval_steps_per_second": 3.374, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.1806287479556605, | |
| "grad_norm": 4.08748197555542, | |
| "learning_rate": 3.525349809194985e-05, | |
| "loss": 1.0468, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.1806287479556605, | |
| "eval_cosine_accuracy": 0.9590913653373718, | |
| "eval_loss": 1.4509832859039307, | |
| "eval_runtime": 21.4477, | |
| "eval_samples_per_second": 443.358, | |
| "eval_steps_per_second": 3.497, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.362347810285299, | |
| "grad_norm": 4.058324813842773, | |
| "learning_rate": 3.2982009812829365e-05, | |
| "loss": 0.9871, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.362347810285299, | |
| "eval_cosine_accuracy": 0.9607740044593811, | |
| "eval_loss": 1.4624947309494019, | |
| "eval_runtime": 21.4174, | |
| "eval_samples_per_second": 443.985, | |
| "eval_steps_per_second": 3.502, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.5440668726149376, | |
| "grad_norm": 4.1556572914123535, | |
| "learning_rate": 3.071279302198801e-05, | |
| "loss": 0.9596, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5440668726149376, | |
| "eval_cosine_accuracy": 0.9605636596679688, | |
| "eval_loss": 1.4531124830245972, | |
| "eval_runtime": 21.3445, | |
| "eval_samples_per_second": 445.501, | |
| "eval_steps_per_second": 3.514, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.7257859349445757, | |
| "grad_norm": 6.094749450683594, | |
| "learning_rate": 2.8443576231146652e-05, | |
| "loss": 0.7272, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.7257859349445757, | |
| "eval_cosine_accuracy": 0.9588810801506042, | |
| "eval_loss": 1.4685230255126953, | |
| "eval_runtime": 21.4386, | |
| "eval_samples_per_second": 443.546, | |
| "eval_steps_per_second": 3.498, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.907504997274214, | |
| "grad_norm": 4.291028022766113, | |
| "learning_rate": 2.6172087952026168e-05, | |
| "loss": 0.4716, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.907504997274214, | |
| "eval_cosine_accuracy": 0.9548848271369934, | |
| "eval_loss": 1.5063105821609497, | |
| "eval_runtime": 21.5765, | |
| "eval_samples_per_second": 440.711, | |
| "eval_steps_per_second": 3.476, | |
| "step": 16000 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 27515, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |