| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 100, | |
| "global_step": 1176, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 23.253705978393555, | |
| "learning_rate": 4.5748299319727895e-05, | |
| "loss": 3.0173, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 4.248861312866211, | |
| "eval_runtime": 33.2541, | |
| "eval_samples_per_second": 125.699, | |
| "eval_steps_per_second": 1.985, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 24.05482292175293, | |
| "learning_rate": 4.149659863945579e-05, | |
| "loss": 2.8146, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_loss": 4.2605061531066895, | |
| "eval_runtime": 33.1469, | |
| "eval_samples_per_second": 126.105, | |
| "eval_steps_per_second": 1.991, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 19.086837768554688, | |
| "learning_rate": 3.724489795918368e-05, | |
| "loss": 2.7177, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_loss": 4.254702091217041, | |
| "eval_runtime": 33.1625, | |
| "eval_samples_per_second": 126.046, | |
| "eval_steps_per_second": 1.99, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 19.842321395874023, | |
| "learning_rate": 3.2993197278911564e-05, | |
| "loss": 2.6338, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_loss": 4.386436462402344, | |
| "eval_runtime": 33.0306, | |
| "eval_samples_per_second": 126.549, | |
| "eval_steps_per_second": 1.998, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 25.054931640625, | |
| "learning_rate": 2.8741496598639456e-05, | |
| "loss": 2.102, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_loss": 4.353818893432617, | |
| "eval_runtime": 32.9823, | |
| "eval_samples_per_second": 126.734, | |
| "eval_steps_per_second": 2.001, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 22.865522384643555, | |
| "learning_rate": 2.448979591836735e-05, | |
| "loss": 2.1015, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_loss": 4.339511871337891, | |
| "eval_runtime": 33.0575, | |
| "eval_samples_per_second": 126.446, | |
| "eval_steps_per_second": 1.997, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 25.445798873901367, | |
| "learning_rate": 2.023809523809524e-05, | |
| "loss": 2.0691, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_loss": 4.336267471313477, | |
| "eval_runtime": 32.9895, | |
| "eval_samples_per_second": 126.707, | |
| "eval_steps_per_second": 2.001, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 22.133098602294922, | |
| "learning_rate": 1.5986394557823133e-05, | |
| "loss": 1.9422, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_loss": 4.633069038391113, | |
| "eval_runtime": 32.9021, | |
| "eval_samples_per_second": 127.044, | |
| "eval_steps_per_second": 2.006, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 31.207412719726562, | |
| "learning_rate": 1.1734693877551021e-05, | |
| "loss": 1.4405, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "eval_loss": 4.565266132354736, | |
| "eval_runtime": 33.1329, | |
| "eval_samples_per_second": 126.158, | |
| "eval_steps_per_second": 1.992, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 19.457901000976562, | |
| "learning_rate": 7.482993197278912e-06, | |
| "loss": 1.3896, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_loss": 4.576756477355957, | |
| "eval_runtime": 33.1251, | |
| "eval_samples_per_second": 126.188, | |
| "eval_steps_per_second": 1.992, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 20.708255767822266, | |
| "learning_rate": 3.231292517006803e-06, | |
| "loss": 1.3735, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_loss": 4.561042308807373, | |
| "eval_runtime": 33.6147, | |
| "eval_samples_per_second": 124.35, | |
| "eval_steps_per_second": 1.963, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1176, | |
| "total_flos": 4376310425967942.0, | |
| "train_loss": 2.093686363324016, | |
| "train_runtime": 1727.448, | |
| "train_samples_per_second": 43.554, | |
| "train_steps_per_second": 0.681 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1176, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 4376310425967942.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |