| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 10560, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2754160165786743, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.7665, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8710623074629832, | |
| "eval_f1": 0.1887287024901704, | |
| "eval_loss": 0.42895129323005676, | |
| "eval_precision": 0.38028169014084506, | |
| "eval_recall": 0.12550842533410808, | |
| "eval_runtime": 4.4417, | |
| "eval_samples_per_second": 210.505, | |
| "eval_steps_per_second": 3.377, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.9667028784751892, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.336, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9335188313624168, | |
| "eval_f1": 0.6881609491875161, | |
| "eval_loss": 0.21774514019489288, | |
| "eval_precision": 0.6187384044526901, | |
| "eval_recall": 0.7751307379430563, | |
| "eval_runtime": 4.5594, | |
| "eval_samples_per_second": 205.071, | |
| "eval_steps_per_second": 3.29, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0499272346496582, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.2067, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9410215641458809, | |
| "eval_f1": 0.7261015202267457, | |
| "eval_loss": 0.17432348430156708, | |
| "eval_precision": 0.6523148148148148, | |
| "eval_recall": 0.8187100522951772, | |
| "eval_runtime": 4.8137, | |
| "eval_samples_per_second": 194.236, | |
| "eval_steps_per_second": 3.116, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.841440200805664, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1734, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9499652191195469, | |
| "eval_f1": 0.7669569807337029, | |
| "eval_loss": 0.15249943733215332, | |
| "eval_precision": 0.7026112185686654, | |
| "eval_recall": 0.8442765833817548, | |
| "eval_runtime": 4.8347, | |
| "eval_samples_per_second": 193.394, | |
| "eval_steps_per_second": 3.103, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.0053447484970093, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.1557, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9523501937791911, | |
| "eval_f1": 0.7757479481069631, | |
| "eval_loss": 0.14416351914405823, | |
| "eval_precision": 0.7125486381322957, | |
| "eval_recall": 0.8512492736780941, | |
| "eval_runtime": 4.5552, | |
| "eval_samples_per_second": 205.259, | |
| "eval_steps_per_second": 3.293, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.2348506450653076, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.146, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9519526980025838, | |
| "eval_f1": 0.7780979827089337, | |
| "eval_loss": 0.14452077448368073, | |
| "eval_precision": 0.7084923664122137, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.5529, | |
| "eval_samples_per_second": 205.363, | |
| "eval_steps_per_second": 3.295, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.4902212619781494, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.1397, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9525489416674948, | |
| "eval_f1": 0.7873728150273936, | |
| "eval_loss": 0.14437128603458405, | |
| "eval_precision": 0.7144886363636364, | |
| "eval_recall": 0.8768158047646717, | |
| "eval_runtime": 4.5644, | |
| "eval_samples_per_second": 204.845, | |
| "eval_steps_per_second": 3.286, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.4951108992099762, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1338, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9545364205505317, | |
| "eval_f1": 0.7905745300503045, | |
| "eval_loss": 0.1385926455259323, | |
| "eval_precision": 0.7261673151750972, | |
| "eval_recall": 0.8675188843695526, | |
| "eval_runtime": 4.5189, | |
| "eval_samples_per_second": 206.907, | |
| "eval_steps_per_second": 3.319, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.3395795822143555, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.1277, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9561264036569611, | |
| "eval_f1": 0.7964601769911505, | |
| "eval_loss": 0.13648679852485657, | |
| "eval_precision": 0.7395418326693227, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.4933, | |
| "eval_samples_per_second": 208.086, | |
| "eval_steps_per_second": 3.338, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.1327812671661377, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1255, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9562754645731889, | |
| "eval_f1": 0.7936932121859969, | |
| "eval_loss": 0.1332310438156128, | |
| "eval_precision": 0.7347847600197922, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.519, | |
| "eval_samples_per_second": 206.904, | |
| "eval_steps_per_second": 3.319, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 11.329936027526855, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1215, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9557289078803538, | |
| "eval_f1": 0.7884564469155414, | |
| "eval_loss": 0.1330229640007019, | |
| "eval_precision": 0.7242217898832685, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5033, | |
| "eval_samples_per_second": 207.627, | |
| "eval_steps_per_second": 3.331, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.500403642654419, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1189, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9561264036569611, | |
| "eval_f1": 0.794345158708989, | |
| "eval_loss": 0.13398447632789612, | |
| "eval_precision": 0.7342209072978304, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5341, | |
| "eval_samples_per_second": 206.215, | |
| "eval_steps_per_second": 3.308, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.2570128440856934, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1179, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9571201430984796, | |
| "eval_f1": 0.7973009446693656, | |
| "eval_loss": 0.12948381900787354, | |
| "eval_precision": 0.7444556451612904, | |
| "eval_recall": 0.8582219639744335, | |
| "eval_runtime": 4.5379, | |
| "eval_samples_per_second": 206.041, | |
| "eval_steps_per_second": 3.305, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.9663475751876831, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.114, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9578654476796185, | |
| "eval_f1": 0.8013955984970478, | |
| "eval_loss": 0.1294805407524109, | |
| "eval_precision": 0.7446384039900249, | |
| "eval_recall": 0.8675188843695526, | |
| "eval_runtime": 4.5196, | |
| "eval_samples_per_second": 206.877, | |
| "eval_steps_per_second": 3.319, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.1781103610992432, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1128, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9571201430984796, | |
| "eval_f1": 0.7960438385458433, | |
| "eval_loss": 0.13167551159858704, | |
| "eval_precision": 0.7371287128712871, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.523, | |
| "eval_samples_per_second": 206.722, | |
| "eval_steps_per_second": 3.316, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.9357908964157104, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1115, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.957517638875087, | |
| "eval_f1": 0.7973297730307075, | |
| "eval_loss": 0.12998828291893005, | |
| "eval_precision": 0.7376482213438735, | |
| "eval_recall": 0.8675188843695526, | |
| "eval_runtime": 4.5035, | |
| "eval_samples_per_second": 207.616, | |
| "eval_steps_per_second": 3.331, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.123679280281067, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1109, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9577163867633907, | |
| "eval_f1": 0.7951935914552738, | |
| "eval_loss": 0.13070179522037506, | |
| "eval_precision": 0.7356719367588933, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5394, | |
| "eval_samples_per_second": 205.975, | |
| "eval_steps_per_second": 3.304, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.571700096130371, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1097, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.957517638875087, | |
| "eval_f1": 0.7968959058067969, | |
| "eval_loss": 0.1318700611591339, | |
| "eval_precision": 0.7385912698412699, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5083, | |
| "eval_samples_per_second": 207.394, | |
| "eval_steps_per_second": 3.327, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 2.028625726699829, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1086, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9573188909867832, | |
| "eval_f1": 0.7943148297130599, | |
| "eval_loss": 0.12963609397411346, | |
| "eval_precision": 0.7375498007968128, | |
| "eval_recall": 0.8605461940732132, | |
| "eval_runtime": 4.5495, | |
| "eval_samples_per_second": 205.519, | |
| "eval_steps_per_second": 3.297, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.7686336040496826, | |
| "learning_rate": 0.0, | |
| "loss": 0.1094, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9573188909867832, | |
| "eval_f1": 0.7943148297130599, | |
| "eval_loss": 0.13019020855426788, | |
| "eval_precision": 0.7375498007968128, | |
| "eval_recall": 0.8605461940732132, | |
| "eval_runtime": 4.5125, | |
| "eval_samples_per_second": 207.2, | |
| "eval_steps_per_second": 3.324, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10560, | |
| "total_flos": 4545096690358590.0, | |
| "train_loss": 0.17230618686387034, | |
| "train_runtime": 1231.892, | |
| "train_samples_per_second": 136.976, | |
| "train_steps_per_second": 8.572 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4545096690358590.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |