| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 200, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.03862990066409111, | |
| "eval_runtime": 37.7121, | |
| "eval_samples_per_second": 424.267, | |
| "eval_steps_per_second": 6.629, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.0523061603307724, | |
| "eval_runtime": 37.8838, | |
| "eval_samples_per_second": 422.344, | |
| "eval_steps_per_second": 6.599, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 72862.21875, | |
| "learning_rate": 4.9692208514878444e-05, | |
| "loss": 0.1425, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 1.0542395114898682, | |
| "eval_runtime": 37.8634, | |
| "eval_samples_per_second": 422.571, | |
| "eval_steps_per_second": 6.603, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 1.0459295511245728, | |
| "eval_runtime": 37.8174, | |
| "eval_samples_per_second": 423.086, | |
| "eval_steps_per_second": 6.611, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.641770839691162, | |
| "learning_rate": 4.877641290737884e-05, | |
| "loss": 0.9177, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3677258789539337, | |
| "eval_runtime": 37.6734, | |
| "eval_samples_per_second": 424.703, | |
| "eval_steps_per_second": 6.636, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.02963975816965103, | |
| "eval_runtime": 37.6097, | |
| "eval_samples_per_second": 425.422, | |
| "eval_steps_per_second": 6.647, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_loss": 0.028412258252501488, | |
| "eval_runtime": 37.7795, | |
| "eval_samples_per_second": 423.51, | |
| "eval_steps_per_second": 6.617, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.17703795433044434, | |
| "learning_rate": 4.72751631047092e-05, | |
| "loss": 0.0421, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.02752041630446911, | |
| "eval_runtime": 37.7317, | |
| "eval_samples_per_second": 424.046, | |
| "eval_steps_per_second": 6.626, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 0.027357231825590134, | |
| "eval_runtime": 37.9192, | |
| "eval_samples_per_second": 421.95, | |
| "eval_steps_per_second": 6.593, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.14987020194530487, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 0.029, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.026410279795527458, | |
| "eval_runtime": 37.883, | |
| "eval_samples_per_second": 422.353, | |
| "eval_steps_per_second": 6.599, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_loss": 0.025838496163487434, | |
| "eval_runtime": 37.6759, | |
| "eval_samples_per_second": 424.674, | |
| "eval_steps_per_second": 6.636, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.025557253509759903, | |
| "eval_runtime": 37.6691, | |
| "eval_samples_per_second": 424.751, | |
| "eval_steps_per_second": 6.637, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.21350397169589996, | |
| "learning_rate": 4.267766952966369e-05, | |
| "loss": 0.0276, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "eval_loss": 0.025391312316060066, | |
| "eval_runtime": 37.9185, | |
| "eval_samples_per_second": 421.958, | |
| "eval_steps_per_second": 6.593, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 0.025234265252947807, | |
| "eval_runtime": 37.6569, | |
| "eval_samples_per_second": 424.889, | |
| "eval_steps_per_second": 6.639, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.09507149457931519, | |
| "learning_rate": 3.969463130731183e-05, | |
| "loss": 0.0265, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.025119660422205925, | |
| "eval_runtime": 37.7071, | |
| "eval_samples_per_second": 424.323, | |
| "eval_steps_per_second": 6.63, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 0.02474472112953663, | |
| "eval_runtime": 37.8472, | |
| "eval_samples_per_second": 422.753, | |
| "eval_steps_per_second": 6.606, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "eval_loss": 0.02474530041217804, | |
| "eval_runtime": 37.964, | |
| "eval_samples_per_second": 421.452, | |
| "eval_steps_per_second": 6.585, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.21493718028068542, | |
| "learning_rate": 3.634976249348867e-05, | |
| "loss": 0.0256, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "eval_loss": 0.024641884490847588, | |
| "eval_runtime": 37.7028, | |
| "eval_samples_per_second": 424.372, | |
| "eval_steps_per_second": 6.631, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_loss": 0.02516881749033928, | |
| "eval_runtime": 37.6697, | |
| "eval_samples_per_second": 424.745, | |
| "eval_steps_per_second": 6.637, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.11537094414234161, | |
| "learning_rate": 3.272542485937369e-05, | |
| "loss": 0.0262, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.024907398968935013, | |
| "eval_runtime": 37.8134, | |
| "eval_samples_per_second": 423.13, | |
| "eval_steps_per_second": 6.611, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "eval_loss": 0.024411499500274658, | |
| "eval_runtime": 37.768, | |
| "eval_samples_per_second": 423.639, | |
| "eval_steps_per_second": 6.619, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "eval_loss": 0.02421731874346733, | |
| "eval_runtime": 37.8841, | |
| "eval_samples_per_second": 422.34, | |
| "eval_steps_per_second": 6.599, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.15358753502368927, | |
| "learning_rate": 2.8910861626005776e-05, | |
| "loss": 0.0255, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "eval_loss": 0.0241916012018919, | |
| "eval_runtime": 37.7974, | |
| "eval_samples_per_second": 423.31, | |
| "eval_steps_per_second": 6.614, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_loss": 0.024809282273054123, | |
| "eval_runtime": 37.9668, | |
| "eval_samples_per_second": 421.42, | |
| "eval_steps_per_second": 6.585, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.21434639394283295, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0251, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.023995770141482353, | |
| "eval_runtime": 37.8979, | |
| "eval_samples_per_second": 422.187, | |
| "eval_steps_per_second": 6.597, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "eval_loss": 0.024207767099142075, | |
| "eval_runtime": 37.6737, | |
| "eval_samples_per_second": 424.7, | |
| "eval_steps_per_second": 6.636, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "eval_loss": 0.023853810504078865, | |
| "eval_runtime": 37.6942, | |
| "eval_samples_per_second": 424.469, | |
| "eval_steps_per_second": 6.632, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.13689687848091125, | |
| "learning_rate": 2.1089138373994223e-05, | |
| "loss": 0.0254, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "eval_loss": 0.02390187606215477, | |
| "eval_runtime": 37.8063, | |
| "eval_samples_per_second": 423.21, | |
| "eval_steps_per_second": 6.613, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "eval_loss": 0.02373598702251911, | |
| "eval_runtime": 37.708, | |
| "eval_samples_per_second": 424.313, | |
| "eval_steps_per_second": 6.63, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.14604102075099945, | |
| "learning_rate": 1.7274575140626318e-05, | |
| "loss": 0.0244, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.02371423877775669, | |
| "eval_runtime": 37.6971, | |
| "eval_samples_per_second": 424.436, | |
| "eval_steps_per_second": 6.632, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "eval_loss": 0.02387085184454918, | |
| "eval_runtime": 37.6715, | |
| "eval_samples_per_second": 424.724, | |
| "eval_steps_per_second": 6.636, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_loss": 0.023735951632261276, | |
| "eval_runtime": 37.9359, | |
| "eval_samples_per_second": 421.764, | |
| "eval_steps_per_second": 6.59, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.12262556701898575, | |
| "learning_rate": 1.3650237506511331e-05, | |
| "loss": 0.0244, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "eval_loss": 0.023663455620408058, | |
| "eval_runtime": 37.8483, | |
| "eval_samples_per_second": 422.74, | |
| "eval_steps_per_second": 6.605, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "eval_loss": 0.023848820477724075, | |
| "eval_runtime": 37.7446, | |
| "eval_samples_per_second": 423.902, | |
| "eval_steps_per_second": 6.623, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.28356611728668213, | |
| "learning_rate": 1.0305368692688174e-05, | |
| "loss": 0.0246, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.02359016053378582, | |
| "eval_runtime": 37.7474, | |
| "eval_samples_per_second": 423.87, | |
| "eval_steps_per_second": 6.623, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "eval_loss": 0.023543963208794594, | |
| "eval_runtime": 37.7642, | |
| "eval_samples_per_second": 423.681, | |
| "eval_steps_per_second": 6.62, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "eval_loss": 0.023517215624451637, | |
| "eval_runtime": 37.9453, | |
| "eval_samples_per_second": 421.66, | |
| "eval_steps_per_second": 6.588, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.10998739302158356, | |
| "learning_rate": 7.3223304703363135e-06, | |
| "loss": 0.0242, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "eval_loss": 0.023526180535554886, | |
| "eval_runtime": 37.925, | |
| "eval_samples_per_second": 421.885, | |
| "eval_steps_per_second": 6.592, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "eval_loss": 0.02347410097718239, | |
| "eval_runtime": 37.6793, | |
| "eval_samples_per_second": 424.636, | |
| "eval_steps_per_second": 6.635, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.0942414253950119, | |
| "learning_rate": 4.7745751406263165e-06, | |
| "loss": 0.0244, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.023570595309138298, | |
| "eval_runtime": 37.876, | |
| "eval_samples_per_second": 422.432, | |
| "eval_steps_per_second": 6.6, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "eval_loss": 0.023448683321475983, | |
| "eval_runtime": 37.7532, | |
| "eval_samples_per_second": 423.805, | |
| "eval_steps_per_second": 6.622, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "eval_loss": 0.02345592901110649, | |
| "eval_runtime": 37.6951, | |
| "eval_samples_per_second": 424.459, | |
| "eval_steps_per_second": 6.632, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.1675085425376892, | |
| "learning_rate": 2.7248368952908053e-06, | |
| "loss": 0.0246, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "eval_loss": 0.023442383855581284, | |
| "eval_runtime": 37.8284, | |
| "eval_samples_per_second": 422.963, | |
| "eval_steps_per_second": 6.609, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "eval_loss": 0.023424193263053894, | |
| "eval_runtime": 37.8096, | |
| "eval_samples_per_second": 423.173, | |
| "eval_steps_per_second": 6.612, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.19085237383842468, | |
| "learning_rate": 1.2235870926211619e-06, | |
| "loss": 0.0237, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.023412322625517845, | |
| "eval_runtime": 37.9142, | |
| "eval_samples_per_second": 422.006, | |
| "eval_steps_per_second": 6.594, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "eval_loss": 0.023400841280817986, | |
| "eval_runtime": 37.7841, | |
| "eval_samples_per_second": 423.458, | |
| "eval_steps_per_second": 6.617, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "eval_loss": 0.023405231535434723, | |
| "eval_runtime": 37.8666, | |
| "eval_samples_per_second": 422.536, | |
| "eval_steps_per_second": 6.602, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.16063953936100006, | |
| "learning_rate": 3.077914851215585e-07, | |
| "loss": 0.0241, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "eval_loss": 0.023399699479341507, | |
| "eval_runtime": 37.8663, | |
| "eval_samples_per_second": 422.539, | |
| "eval_steps_per_second": 6.602, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "eval_loss": 0.023402543738484383, | |
| "eval_runtime": 37.6977, | |
| "eval_samples_per_second": 424.429, | |
| "eval_steps_per_second": 6.632, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.11751583963632584, | |
| "learning_rate": 0.0, | |
| "loss": 0.024, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.023403601720929146, | |
| "eval_runtime": 37.692, | |
| "eval_samples_per_second": 424.493, | |
| "eval_steps_per_second": 6.633, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000, | |
| "total_flos": 4.180672512e+16, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |