| { | |
| "best_metric": 0.9439287185668945, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_120-samples_config-2_full/checkpoint-60", | |
| "epoch": 18.0, | |
| "eval_steps": 500, | |
| "global_step": 99, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.47469812631607056, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.4189, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.4030732214450836, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.3486, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.44700008630752563, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.4237, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.3857043981552124, | |
| "eval_runtime": 7.7033, | |
| "eval_samples_per_second": 3.116, | |
| "eval_steps_per_second": 3.116, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.490207314491272, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.3998, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.44333764910697937, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.3779, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.38184991478919983, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3836, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.2988808155059814, | |
| "eval_runtime": 7.4559, | |
| "eval_samples_per_second": 3.219, | |
| "eval_steps_per_second": 3.219, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.3069492280483246, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.2702, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.26267415285110474, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2496, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.24956025183200836, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.2209, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 1.2032712697982788, | |
| "eval_runtime": 7.4655, | |
| "eval_samples_per_second": 3.215, | |
| "eval_steps_per_second": 3.215, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.24367348849773407, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.2022, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.25780192017555237, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1371, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.23961281776428223, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.129, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.1133126020431519, | |
| "eval_runtime": 7.4597, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.2551175355911255, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0838, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.3854904770851135, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.0353, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 1.033556580543518, | |
| "eval_runtime": 7.4616, | |
| "eval_samples_per_second": 3.216, | |
| "eval_steps_per_second": 3.216, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.21588054299354553, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 1.0051, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.21526531875133514, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.9757, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.18489542603492737, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.957, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.9890041351318359, | |
| "eval_runtime": 7.4597, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.17797501385211945, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.9378, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.18213433027267456, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.9193, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.14390909671783447, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.9251, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.97099369764328, | |
| "eval_runtime": 7.4646, | |
| "eval_samples_per_second": 3.215, | |
| "eval_steps_per_second": 3.215, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.14765390753746033, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.9268, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.13612937927246094, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.8768, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.13822229206562042, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.9012, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.957718551158905, | |
| "eval_runtime": 7.4596, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.13656413555145264, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.8686, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.16253499686717987, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.8677, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.9495101571083069, | |
| "eval_runtime": 7.4783, | |
| "eval_samples_per_second": 3.209, | |
| "eval_steps_per_second": 3.209, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.14867223799228668, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.8806, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.14953409135341644, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.8369, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.1416897475719452, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.8692, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.944615364074707, | |
| "eval_runtime": 7.4581, | |
| "eval_samples_per_second": 3.218, | |
| "eval_steps_per_second": 3.218, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.15356607735157013, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.8194, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.15293589234352112, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.817, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.15889222919940948, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.8262, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.9439287185668945, | |
| "eval_runtime": 7.4655, | |
| "eval_samples_per_second": 3.215, | |
| "eval_steps_per_second": 3.215, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.16184863448143005, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.8033, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.1613071709871292, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.8124, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.20444954931735992, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.7985, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9476203322410583, | |
| "eval_runtime": 7.4578, | |
| "eval_samples_per_second": 3.218, | |
| "eval_steps_per_second": 3.218, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.20746265351772308, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.7753, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.21244777739048004, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.7634, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.9546706080436707, | |
| "eval_runtime": 7.4607, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.24004337191581726, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.7568, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.478630393743515, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.784, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.2425345778465271, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.7465, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.9687718749046326, | |
| "eval_runtime": 7.456, | |
| "eval_samples_per_second": 3.219, | |
| "eval_steps_per_second": 3.219, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.25790467858314514, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.6986, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.3006918430328369, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.7486, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.4067930579185486, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.6551, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.9894356727600098, | |
| "eval_runtime": 7.4592, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.3998169004917145, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.6957, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.3630196154117584, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.6604, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.34696683287620544, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.6593, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.019141674041748, | |
| "eval_runtime": 7.4613, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.363636363636363, | |
| "grad_norm": 0.43112245202064514, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.6924, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.727272727272727, | |
| "grad_norm": 0.4381055533885956, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.599, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "eval_loss": 1.0529237985610962, | |
| "eval_runtime": 7.4607, | |
| "eval_samples_per_second": 3.217, | |
| "eval_steps_per_second": 3.217, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 17.09090909090909, | |
| "grad_norm": 0.4838111698627472, | |
| "learning_rate": 7.85356783842216e-05, | |
| "loss": 0.5833, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 17.454545454545453, | |
| "grad_norm": 0.4901573061943054, | |
| "learning_rate": 7.737816117462752e-05, | |
| "loss": 0.5777, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 17.818181818181817, | |
| "grad_norm": 0.566810667514801, | |
| "learning_rate": 7.619929529850397e-05, | |
| "loss": 0.5401, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.0785717964172363, | |
| "eval_runtime": 7.4647, | |
| "eval_samples_per_second": 3.215, | |
| "eval_steps_per_second": 3.215, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "step": 99, | |
| "total_flos": 4.256542434839757e+16, | |
| "train_loss": 0.9124392114504419, | |
| "train_runtime": 1803.5797, | |
| "train_samples_per_second": 2.44, | |
| "train_steps_per_second": 0.139 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.256542434839757e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |