| { | |
| "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": 11.8125, | |
| "learning_rate": 0.0002999932345578783, | |
| "loss": 4.9688, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.000279703673635072, | |
| "loss": 3.1317, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "eval_loss": 2.590404748916626, | |
| "eval_runtime": 54.625, | |
| "eval_samples_per_second": 1728.238, | |
| "eval_steps_per_second": 6.755, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 0.0002594073472701441, | |
| "loss": 3.0056, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "eval_loss": 2.559239387512207, | |
| "eval_runtime": 54.5006, | |
| "eval_samples_per_second": 1732.184, | |
| "eval_steps_per_second": 6.771, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.0002391110209052161, | |
| "loss": 2.9796, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "eval_loss": 2.547764301300049, | |
| "eval_runtime": 54.3734, | |
| "eval_samples_per_second": 1736.236, | |
| "eval_steps_per_second": 6.786, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 0.0002188146945402882, | |
| "loss": 2.9712, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "eval_loss": 2.53904128074646, | |
| "eval_runtime": 54.3708, | |
| "eval_samples_per_second": 1736.317, | |
| "eval_steps_per_second": 6.787, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 0.00019851836817536025, | |
| "loss": 2.9638, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "eval_loss": 2.5364582538604736, | |
| "eval_runtime": 54.5162, | |
| "eval_samples_per_second": 1731.687, | |
| "eval_steps_per_second": 6.769, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.0001782220418104323, | |
| "loss": 2.9598, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "eval_loss": 2.5331978797912598, | |
| "eval_runtime": 54.5086, | |
| "eval_samples_per_second": 1731.927, | |
| "eval_steps_per_second": 6.77, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00015792571544550436, | |
| "loss": 2.958, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "eval_loss": 2.5318005084991455, | |
| "eval_runtime": 54.5162, | |
| "eval_samples_per_second": 1731.687, | |
| "eval_steps_per_second": 6.769, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.0001376293890805764, | |
| "loss": 2.9557, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "eval_loss": 2.53218150138855, | |
| "eval_runtime": 54.6541, | |
| "eval_samples_per_second": 1727.317, | |
| "eval_steps_per_second": 6.752, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.00011733306271564845, | |
| "loss": 2.9542, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "eval_loss": 2.5293021202087402, | |
| "eval_runtime": 54.5038, | |
| "eval_samples_per_second": 1732.08, | |
| "eval_steps_per_second": 6.77, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 9.703673635072052e-05, | |
| "loss": 2.9541, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "eval_loss": 2.5292599201202393, | |
| "eval_runtime": 54.4969, | |
| "eval_samples_per_second": 1732.3, | |
| "eval_steps_per_second": 6.771, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 7.674040998579256e-05, | |
| "loss": 2.9518, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "eval_loss": 2.5298101902008057, | |
| "eval_runtime": 54.3826, | |
| "eval_samples_per_second": 1735.942, | |
| "eval_steps_per_second": 6.785, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 5.644408362086462e-05, | |
| "loss": 2.955, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "eval_loss": 2.529937267303467, | |
| "eval_runtime": 54.3621, | |
| "eval_samples_per_second": 1736.595, | |
| "eval_steps_per_second": 6.788, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 3.614775725593667e-05, | |
| "loss": 2.9542, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "eval_loss": 2.5293867588043213, | |
| "eval_runtime": 54.3643, | |
| "eval_samples_per_second": 1736.525, | |
| "eval_steps_per_second": 6.788, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.5851430891008727e-05, | |
| "loss": 2.9536, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "eval_loss": 2.5295138359069824, | |
| "eval_runtime": 54.3667, | |
| "eval_samples_per_second": 1736.449, | |
| "eval_steps_per_second": 6.787, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 44343, | |
| "total_flos": 3.8956704097296384e+17, | |
| "train_loss": 2.9737832070450803, | |
| "train_runtime": 20072.1702, | |
| "train_samples_per_second": 565.532, | |
| "train_steps_per_second": 2.209 | |
| } | |
| ], | |
| "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": 3.8956704097296384e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |