| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 3000, | |
| "global_step": 44343, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 6.76544212164265e-05, | |
| "grad_norm": 37.0, | |
| "learning_rate": 0.0002999932345578783, | |
| "loss": 9.25, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 0.000279703673635072, | |
| "loss": 4.2275, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "eval_loss": 3.5942580699920654, | |
| "eval_runtime": 50.1044, | |
| "eval_samples_per_second": 1884.165, | |
| "eval_steps_per_second": 7.365, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 0.0002594073472701441, | |
| "loss": 4.0104, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "eval_loss": 3.5292599201202393, | |
| "eval_runtime": 49.9718, | |
| "eval_samples_per_second": 1889.165, | |
| "eval_steps_per_second": 7.384, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 0.0002391110209052161, | |
| "loss": 3.9774, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "eval_loss": 3.5074524879455566, | |
| "eval_runtime": 49.8418, | |
| "eval_samples_per_second": 1894.092, | |
| "eval_steps_per_second": 7.403, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.0002188146945402882, | |
| "loss": 3.9649, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "eval_loss": 3.4984755516052246, | |
| "eval_runtime": 49.8451, | |
| "eval_samples_per_second": 1893.969, | |
| "eval_steps_per_second": 7.403, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.00019851836817536025, | |
| "loss": 3.9601, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "eval_loss": 3.492377996444702, | |
| "eval_runtime": 49.9766, | |
| "eval_samples_per_second": 1888.985, | |
| "eval_steps_per_second": 7.383, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 0.0001782220418104323, | |
| "loss": 3.9568, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "eval_loss": 3.489668130874634, | |
| "eval_runtime": 49.8577, | |
| "eval_samples_per_second": 1893.491, | |
| "eval_steps_per_second": 7.401, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.00015792571544550436, | |
| "loss": 3.9542, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "eval_loss": 3.488651752471924, | |
| "eval_runtime": 49.854, | |
| "eval_samples_per_second": 1893.63, | |
| "eval_steps_per_second": 7.402, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.0001376293890805764, | |
| "loss": 3.9522, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "eval_loss": 3.4861111640930176, | |
| "eval_runtime": 49.8638, | |
| "eval_samples_per_second": 1893.256, | |
| "eval_steps_per_second": 7.4, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 0.00011733306271564845, | |
| "loss": 3.9512, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "eval_loss": 3.4858570098876953, | |
| "eval_runtime": 49.8699, | |
| "eval_samples_per_second": 1893.027, | |
| "eval_steps_per_second": 7.399, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 9.703673635072052e-05, | |
| "loss": 3.9523, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "eval_loss": 3.485603094100952, | |
| "eval_runtime": 49.9876, | |
| "eval_samples_per_second": 1888.568, | |
| "eval_steps_per_second": 7.382, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 7.674040998579256e-05, | |
| "loss": 3.95, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "eval_loss": 3.485264301300049, | |
| "eval_runtime": 49.8575, | |
| "eval_samples_per_second": 1893.498, | |
| "eval_steps_per_second": 7.401, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 5.644408362086462e-05, | |
| "loss": 3.9525, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "eval_loss": 3.4850525856018066, | |
| "eval_runtime": 49.8563, | |
| "eval_samples_per_second": 1893.542, | |
| "eval_steps_per_second": 7.401, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 3.614775725593667e-05, | |
| "loss": 3.9509, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "eval_loss": 3.4850101470947266, | |
| "eval_runtime": 49.841, | |
| "eval_samples_per_second": 1894.125, | |
| "eval_steps_per_second": 7.404, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 1.5851430891008727e-05, | |
| "loss": 3.9511, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "eval_loss": 3.4849255084991455, | |
| "eval_runtime": 49.8566, | |
| "eval_samples_per_second": 1893.53, | |
| "eval_steps_per_second": 7.401, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 44343, | |
| "total_flos": 3.41551228634284e+17, | |
| "train_loss": 3.977961572288749, | |
| "train_runtime": 18296.5341, | |
| "train_samples_per_second": 620.416, | |
| "train_steps_per_second": 2.424 | |
| } | |
| ], | |
| "logging_steps": 3000, | |
| "max_steps": 44343, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.41551228634284e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |