| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.416666666666666, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.0434980392456055, | |
| "learning_rate": 2.2963272120200336e-05, | |
| "loss": 0.4992, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.8384565711021423, | |
| "learning_rate": 2.087646076794658e-05, | |
| "loss": 0.0871, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "eval_loss": 0.06118601933121681, | |
| "eval_runtime": 1.5053, | |
| "eval_samples_per_second": 29.231, | |
| "eval_steps_per_second": 3.986, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 3.239577054977417, | |
| "learning_rate": 1.878964941569282e-05, | |
| "loss": 0.0646, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 1.6355056762695312, | |
| "learning_rate": 1.6702838063439067e-05, | |
| "loss": 0.0531, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "eval_loss": 0.04738747701048851, | |
| "eval_runtime": 1.5005, | |
| "eval_samples_per_second": 29.323, | |
| "eval_steps_per_second": 3.999, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 1.0931097269058228, | |
| "learning_rate": 1.461602671118531e-05, | |
| "loss": 0.0524, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 0.6507939696311951, | |
| "learning_rate": 1.2529215358931554e-05, | |
| "loss": 0.0493, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "eval_loss": 0.04496830329298973, | |
| "eval_runtime": 1.5011, | |
| "eval_samples_per_second": 29.312, | |
| "eval_steps_per_second": 3.997, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "grad_norm": 0.834311842918396, | |
| "learning_rate": 1.0442404006677797e-05, | |
| "loss": 0.0496, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 6.94, | |
| "grad_norm": 1.1006280183792114, | |
| "learning_rate": 8.355592654424042e-06, | |
| "loss": 0.0484, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.94, | |
| "eval_loss": 0.04376932606101036, | |
| "eval_runtime": 1.5008, | |
| "eval_samples_per_second": 29.318, | |
| "eval_steps_per_second": 3.998, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "grad_norm": 0.6833261251449585, | |
| "learning_rate": 6.268781302170284e-06, | |
| "loss": 0.049, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 8.68, | |
| "grad_norm": 0.6349731087684631, | |
| "learning_rate": 4.181969949916528e-06, | |
| "loss": 0.0483, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.68, | |
| "eval_loss": 0.04366318881511688, | |
| "eval_runtime": 1.5018, | |
| "eval_samples_per_second": 29.298, | |
| "eval_steps_per_second": 3.995, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.55, | |
| "grad_norm": 0.7335835695266724, | |
| "learning_rate": 2.0951585976627714e-06, | |
| "loss": 0.0461, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 10.42, | |
| "grad_norm": 0.8647783398628235, | |
| "learning_rate": 8.347245409015025e-09, | |
| "loss": 0.048, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 10.42, | |
| "eval_loss": 0.04346328228712082, | |
| "eval_runtime": 1.4995, | |
| "eval_samples_per_second": 29.342, | |
| "eval_steps_per_second": 4.001, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 3000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 11, | |
| "save_steps": 500, | |
| "total_flos": 6.4761311232e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |