| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 21.900392532348633, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.4267, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8409196624360422, | |
| "eval_loss": 0.35864609479904175, | |
| "eval_precision": 0.8892469089546646, | |
| "eval_recall": 0.8148754318967084, | |
| "eval_runtime": 4.9024, | |
| "eval_samples_per_second": 81.389, | |
| "eval_steps_per_second": 10.199, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 23.599449157714844, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2234, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8696741854636592, | |
| "eval_f1": 0.8539284708532808, | |
| "eval_loss": 0.36683306097984314, | |
| "eval_precision": 0.8394707327494125, | |
| "eval_recall": 0.8852973267866885, | |
| "eval_runtime": 4.9237, | |
| "eval_samples_per_second": 81.037, | |
| "eval_steps_per_second": 10.155, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 8.989884376525879, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.126, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8922305764411027, | |
| "eval_f1": 0.8755702215614461, | |
| "eval_loss": 0.4554330110549927, | |
| "eval_precision": 0.8632157235517781, | |
| "eval_recall": 0.8937534097108566, | |
| "eval_runtime": 4.9271, | |
| "eval_samples_per_second": 80.981, | |
| "eval_steps_per_second": 10.148, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 15.863175392150879, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0886, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8854915648632926, | |
| "eval_loss": 0.44412538409233093, | |
| "eval_precision": 0.8956662848415425, | |
| "eval_recall": 0.8768867066739408, | |
| "eval_runtime": 4.9398, | |
| "eval_samples_per_second": 80.773, | |
| "eval_steps_per_second": 10.122, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.03718271106481552, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0611, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8839406001224739, | |
| "eval_loss": 0.4922772943973541, | |
| "eval_precision": 0.8880654743486602, | |
| "eval_recall": 0.880114566284779, | |
| "eval_runtime": 4.9142, | |
| "eval_samples_per_second": 81.193, | |
| "eval_steps_per_second": 10.175, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.006236851681023836, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.0366, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8813841488792438, | |
| "eval_loss": 0.6796092391014099, | |
| "eval_precision": 0.8748029197080291, | |
| "eval_recall": 0.8890707401345699, | |
| "eval_runtime": 4.9691, | |
| "eval_samples_per_second": 80.296, | |
| "eval_steps_per_second": 10.062, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.006574722938239574, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.0358, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8820775261324042, | |
| "eval_loss": 0.5746134519577026, | |
| "eval_precision": 0.8934835488413775, | |
| "eval_recall": 0.8726132024004365, | |
| "eval_runtime": 4.9771, | |
| "eval_samples_per_second": 80.167, | |
| "eval_steps_per_second": 10.046, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.0049128723330795765, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0272, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8947368421052632, | |
| "eval_f1": 0.8736504011098378, | |
| "eval_loss": 0.5952700972557068, | |
| "eval_precision": 0.8718487394957983, | |
| "eval_recall": 0.8755228223313329, | |
| "eval_runtime": 4.9556, | |
| "eval_samples_per_second": 80.515, | |
| "eval_steps_per_second": 10.09, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.006197785492986441, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.0231, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8751876876876876, | |
| "eval_loss": 0.6506014466285706, | |
| "eval_precision": 0.8891156462585035, | |
| "eval_recall": 0.864066193853428, | |
| "eval_runtime": 4.9505, | |
| "eval_samples_per_second": 80.598, | |
| "eval_steps_per_second": 10.1, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.009196682833135128, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0141, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8823853973772722, | |
| "eval_loss": 0.6854431629180908, | |
| "eval_precision": 0.8814464081066409, | |
| "eval_recall": 0.8833424258956174, | |
| "eval_runtime": 4.9291, | |
| "eval_samples_per_second": 80.948, | |
| "eval_steps_per_second": 10.144, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.01001653354614973, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.023, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8823853973772722, | |
| "eval_loss": 0.7218220829963684, | |
| "eval_precision": 0.8814464081066409, | |
| "eval_recall": 0.8833424258956174, | |
| "eval_runtime": 4.9437, | |
| "eval_samples_per_second": 80.709, | |
| "eval_steps_per_second": 10.114, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.0031623237300664186, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0067, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8823853973772722, | |
| "eval_loss": 0.7694610357284546, | |
| "eval_precision": 0.8814464081066409, | |
| "eval_recall": 0.8833424258956174, | |
| "eval_runtime": 4.9712, | |
| "eval_samples_per_second": 80.263, | |
| "eval_steps_per_second": 10.058, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.002341507002711296, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.0064, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8796992481203008, | |
| "eval_f1": 0.8602043795620438, | |
| "eval_loss": 0.9004446268081665, | |
| "eval_precision": 0.849624060150376, | |
| "eval_recall": 0.8748863429714493, | |
| "eval_runtime": 4.9523, | |
| "eval_samples_per_second": 80.569, | |
| "eval_steps_per_second": 10.096, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.0026841196231544018, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0103, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8835263209107715, | |
| "eval_loss": 0.7978267073631287, | |
| "eval_precision": 0.8791501449961532, | |
| "eval_recall": 0.8883433351518457, | |
| "eval_runtime": 4.9622, | |
| "eval_samples_per_second": 80.407, | |
| "eval_steps_per_second": 10.076, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.0031805976759642363, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0072, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8790689216221131, | |
| "eval_loss": 0.8251467943191528, | |
| "eval_precision": 0.8790689216221131, | |
| "eval_recall": 0.8790689216221131, | |
| "eval_runtime": 4.9558, | |
| "eval_samples_per_second": 80.512, | |
| "eval_steps_per_second": 10.089, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.0014117677928879857, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0054, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8823853973772722, | |
| "eval_loss": 0.7715300917625427, | |
| "eval_precision": 0.8814464081066409, | |
| "eval_recall": 0.8833424258956174, | |
| "eval_runtime": 4.96, | |
| "eval_samples_per_second": 80.444, | |
| "eval_steps_per_second": 10.081, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.0014807094121351838, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.0038, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8867007927797945, | |
| "eval_loss": 0.7821339964866638, | |
| "eval_precision": 0.89198606271777, | |
| "eval_recall": 0.8818876159301692, | |
| "eval_runtime": 4.9491, | |
| "eval_samples_per_second": 80.62, | |
| "eval_steps_per_second": 10.103, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.0014356797328218818, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0021, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8972431077694235, | |
| "eval_f1": 0.8763538792940554, | |
| "eval_loss": 0.8211472630500793, | |
| "eval_precision": 0.8754297605404427, | |
| "eval_recall": 0.877295871976723, | |
| "eval_runtime": 4.9631, | |
| "eval_samples_per_second": 80.393, | |
| "eval_steps_per_second": 10.074, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.0011136590037494898, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0022, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8790689216221131, | |
| "eval_loss": 0.8161725401878357, | |
| "eval_precision": 0.8790689216221131, | |
| "eval_recall": 0.8790689216221131, | |
| "eval_runtime": 4.9339, | |
| "eval_samples_per_second": 80.869, | |
| "eval_steps_per_second": 10.134, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.0013613783521577716, | |
| "learning_rate": 0.0, | |
| "loss": 0.0027, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8817957385392532, | |
| "eval_loss": 0.8104944229125977, | |
| "eval_precision": 0.8827677592299257, | |
| "eval_recall": 0.8808419712675032, | |
| "eval_runtime": 4.9425, | |
| "eval_samples_per_second": 80.729, | |
| "eval_steps_per_second": 10.116, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 7584162436176000.0, | |
| "train_loss": 0.05662053943168922, | |
| "train_runtime": 2712.8409, | |
| "train_samples_per_second": 26.821, | |
| "train_steps_per_second": 0.899 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 7584162436176000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |