| { | |
| "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": 4.800751686096191, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5455, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7543859649122807, | |
| "eval_f1": 0.6639335808580858, | |
| "eval_loss": 0.48765844106674194, | |
| "eval_precision": 0.7053086419753087, | |
| "eval_recall": 0.6512093107837789, | |
| "eval_runtime": 1.7695, | |
| "eval_samples_per_second": 225.487, | |
| "eval_steps_per_second": 28.257, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.828364849090576, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4356, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8209674617878647, | |
| "eval_loss": 0.35374370217323303, | |
| "eval_precision": 0.810288627625787, | |
| "eval_recall": 0.8375613747954174, | |
| "eval_runtime": 1.7671, | |
| "eval_samples_per_second": 225.791, | |
| "eval_steps_per_second": 28.295, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 7.537293910980225, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.3468, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.849624060150376, | |
| "eval_f1": 0.8072960267885891, | |
| "eval_loss": 0.3415985107421875, | |
| "eval_precision": 0.8325657894736842, | |
| "eval_recall": 0.7910983815239134, | |
| "eval_runtime": 1.7644, | |
| "eval_samples_per_second": 226.134, | |
| "eval_steps_per_second": 28.338, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.427431106567383, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3049, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8546365914786967, | |
| "eval_f1": 0.8180088078011953, | |
| "eval_loss": 0.3125925660133362, | |
| "eval_precision": 0.8323930726843348, | |
| "eval_recall": 0.8071467539552646, | |
| "eval_runtime": 1.7656, | |
| "eval_samples_per_second": 225.985, | |
| "eval_steps_per_second": 28.319, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.060295820236206, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2673, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8596491228070176, | |
| "eval_f1": 0.8315338681464504, | |
| "eval_loss": 0.29185813665390015, | |
| "eval_precision": 0.8299369747899159, | |
| "eval_recall": 0.8331969448990726, | |
| "eval_runtime": 1.7632, | |
| "eval_samples_per_second": 226.299, | |
| "eval_steps_per_second": 28.358, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.6104700565338135, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.2516, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8646616541353384, | |
| "eval_f1": 0.8350734843845683, | |
| "eval_loss": 0.2822675108909607, | |
| "eval_precision": 0.8386812302741506, | |
| "eval_recall": 0.8317421349336243, | |
| "eval_runtime": 1.761, | |
| "eval_samples_per_second": 226.577, | |
| "eval_steps_per_second": 28.393, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.407837837934494, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.2243, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8822055137844611, | |
| "eval_f1": 0.8622085718274466, | |
| "eval_loss": 0.26883605122566223, | |
| "eval_precision": 0.8530168716042322, | |
| "eval_recall": 0.8741589379887251, | |
| "eval_runtime": 1.7626, | |
| "eval_samples_per_second": 226.37, | |
| "eval_steps_per_second": 28.367, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6773055791854858, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2157, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8947368421052632, | |
| "eval_f1": 0.8696646341463414, | |
| "eval_loss": 0.2641172409057617, | |
| "eval_precision": 0.8807130969146075, | |
| "eval_recall": 0.8605200945626478, | |
| "eval_runtime": 1.7614, | |
| "eval_samples_per_second": 226.526, | |
| "eval_steps_per_second": 28.387, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 3.9705936908721924, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2052, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8572517421602788, | |
| "eval_loss": 0.26267343759536743, | |
| "eval_precision": 0.8679426449878376, | |
| "eval_recall": 0.8484269867248591, | |
| "eval_runtime": 1.7598, | |
| "eval_samples_per_second": 226.73, | |
| "eval_steps_per_second": 28.412, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 13.646815299987793, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1864, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8847117794486216, | |
| "eval_f1": 0.8548402404302435, | |
| "eval_loss": 0.288084477186203, | |
| "eval_precision": 0.8737291514386338, | |
| "eval_recall": 0.8409256228405164, | |
| "eval_runtime": 1.7593, | |
| "eval_samples_per_second": 226.79, | |
| "eval_steps_per_second": 28.42, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 4.454102039337158, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1928, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8872180451127819, | |
| "eval_f1": 0.867476104365797, | |
| "eval_loss": 0.2784619927406311, | |
| "eval_precision": 0.8592596863468634, | |
| "eval_recall": 0.8777050372795053, | |
| "eval_runtime": 1.7608, | |
| "eval_samples_per_second": 226.608, | |
| "eval_steps_per_second": 28.397, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 4.11415958404541, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1804, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8758710801393728, | |
| "eval_loss": 0.2506408393383026, | |
| "eval_precision": 0.8870983228779925, | |
| "eval_recall": 0.8665666484815421, | |
| "eval_runtime": 1.7598, | |
| "eval_samples_per_second": 226.734, | |
| "eval_steps_per_second": 28.413, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.9822263717651367, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1654, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8790689216221131, | |
| "eval_loss": 0.26640743017196655, | |
| "eval_precision": 0.8790689216221131, | |
| "eval_recall": 0.8790689216221131, | |
| "eval_runtime": 1.7621, | |
| "eval_samples_per_second": 226.429, | |
| "eval_steps_per_second": 28.375, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.247542381286621, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1567, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8878351186601172, | |
| "eval_loss": 0.2660583257675171, | |
| "eval_precision": 0.879776516905975, | |
| "eval_recall": 0.8976177486815784, | |
| "eval_runtime": 1.7643, | |
| "eval_samples_per_second": 226.146, | |
| "eval_steps_per_second": 28.339, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.1535786390304565, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1438, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9097744360902256, | |
| "eval_f1": 0.8917003438084323, | |
| "eval_loss": 0.2614538073539734, | |
| "eval_precision": 0.8898109243697478, | |
| "eval_recall": 0.893662484088016, | |
| "eval_runtime": 1.7625, | |
| "eval_samples_per_second": 226.385, | |
| "eval_steps_per_second": 28.369, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.5205868482589722, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1472, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8856836962422341, | |
| "eval_loss": 0.25552284717559814, | |
| "eval_precision": 0.8838235294117647, | |
| "eval_recall": 0.8876159301691217, | |
| "eval_runtime": 1.7611, | |
| "eval_samples_per_second": 226.564, | |
| "eval_steps_per_second": 28.392, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 6.205023288726807, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1394, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8790689216221131, | |
| "eval_loss": 0.2647987902164459, | |
| "eval_precision": 0.8790689216221131, | |
| "eval_recall": 0.8790689216221131, | |
| "eval_runtime": 1.7605, | |
| "eval_samples_per_second": 226.638, | |
| "eval_steps_per_second": 28.401, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.5985056161880493, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1387, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8862394957983193, | |
| "eval_loss": 0.26299694180488586, | |
| "eval_precision": 0.8826476009275225, | |
| "eval_recall": 0.8901163847972358, | |
| "eval_runtime": 1.7607, | |
| "eval_samples_per_second": 226.62, | |
| "eval_steps_per_second": 28.399, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 1.7635153532028198, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1378, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8845345436822225, | |
| "eval_loss": 0.26893848180770874, | |
| "eval_precision": 0.8865278628291489, | |
| "eval_recall": 0.8826150209128933, | |
| "eval_runtime": 1.7631, | |
| "eval_samples_per_second": 226.305, | |
| "eval_steps_per_second": 28.359, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 3.3988428115844727, | |
| "learning_rate": 0.0, | |
| "loss": 0.1365, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8817957385392532, | |
| "eval_loss": 0.2682853937149048, | |
| "eval_precision": 0.8827677592299257, | |
| "eval_recall": 0.8808419712675032, | |
| "eval_runtime": 1.7608, | |
| "eval_samples_per_second": 226.596, | |
| "eval_steps_per_second": 28.396, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 8444128359504000.0, | |
| "train_loss": 0.22609603365913766, | |
| "train_runtime": 618.465, | |
| "train_samples_per_second": 117.646, | |
| "train_steps_per_second": 3.945 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 8444128359504000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |