| { | |
| "best_metric": 0.5405952334403992, | |
| "best_model_checkpoint": "data/Mistral-7B_task-2_120-samples_config-1/checkpoint-44", | |
| "epoch": 11.0, | |
| "eval_steps": 500, | |
| "global_step": 121, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 2.696162700653076, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 1.0407, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.010056257247925, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 1.2022, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.4440057277679443, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 0.9285, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 2.539612293243408, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 0.9667, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 2.4596331119537354, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 1.0105, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.7177777290344238, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 0.8123, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7914873957633972, | |
| "eval_runtime": 30.0604, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 1.5904673337936401, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 0.7794, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 1.0533339977264404, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 0.736, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.8380967378616333, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 0.7123, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.8337122797966003, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 0.6878, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.6451910734176636, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 0.6072, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6075661182403564, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5536, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.6210684180259705, | |
| "eval_runtime": 30.034, | |
| "eval_samples_per_second": 0.799, | |
| "eval_steps_per_second": 0.799, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.5785300135612488, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 0.5834, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.5918728709220886, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.6466, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.5560950040817261, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.5513, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.6295881271362305, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.5324, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.5332650542259216, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.5094, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5669192671775818, | |
| "eval_runtime": 30.037, | |
| "eval_samples_per_second": 0.799, | |
| "eval_steps_per_second": 0.799, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.5463592410087585, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.5556, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.4856805205345154, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.5087, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 1.4090657234191895, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.4894, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.5057636499404907, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.4532, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.598127007484436, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.4761, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5398470163345337, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4475, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5405952334403992, | |
| "eval_runtime": 30.1274, | |
| "eval_samples_per_second": 0.797, | |
| "eval_steps_per_second": 0.797, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.6271505355834961, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.4387, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.591743528842926, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.3713, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.6342865228652954, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.4027, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.8849979639053345, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.3832, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.7518714070320129, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.392, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.545644223690033, | |
| "eval_runtime": 30.177, | |
| "eval_samples_per_second": 0.795, | |
| "eval_steps_per_second": 0.795, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.7579145431518555, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.3203, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 1.103210210800171, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.3053, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.8542510867118835, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.2632, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 1.2728869915008545, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.2499, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 1.0026756525039673, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.2527, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.128718614578247, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.25, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.6154773831367493, | |
| "eval_runtime": 30.0798, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.9723846316337585, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.1708, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 1.0810333490371704, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.1502, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 1.7670875787734985, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.134, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 1.2101190090179443, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.1234, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 1.163854956626892, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.1335, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.717187225818634, | |
| "eval_runtime": 30.0811, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.928467333316803, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.1112, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 1.090691328048706, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.0605, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 1.392821192741394, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.064, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 1.5230432748794556, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.0709, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 1.1392593383789062, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.0659, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.0564861297607422, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.0743, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.781478226184845, | |
| "eval_runtime": 30.074, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.8312259912490845, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.041, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 1.8428175449371338, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.0254, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 1.3573157787322998, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.0365, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 1.2816226482391357, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0354, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 1.1067599058151245, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.0326, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.8637485504150391, | |
| "eval_runtime": 30.0924, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.49380627274513245, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.022, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.5494066476821899, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.0157, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.4922139644622803, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.0176, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.9321619272232056, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0212, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.8448288440704346, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.0228, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.9266816973686218, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0175, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.943859875202179, | |
| "eval_runtime": 30.0503, | |
| "eval_samples_per_second": 0.799, | |
| "eval_steps_per_second": 0.799, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.6057682633399963, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.0092, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 1.122817873954773, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.013, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.7116965055465698, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.016, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 1.1520817279815674, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.02, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.42801597714424133, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0075, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.9337683320045471, | |
| "eval_runtime": 30.047, | |
| "eval_samples_per_second": 0.799, | |
| "eval_steps_per_second": 0.799, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "step": 121, | |
| "total_flos": 1.1411215447438131e+17, | |
| "train_loss": 0.32754210635927344, | |
| "train_runtime": 4160.4563, | |
| "train_samples_per_second": 1.058, | |
| "train_steps_per_second": 0.132 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 550, | |
| "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": 1.1411215447438131e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |