| { | |
| "best_metric": 0.8149010539054871, | |
| "best_model_checkpoint": "./output/training_results/C014_llama3-8b-base_instruct_20240428_005832/checkpoint-15", | |
| "epoch": 4.0, | |
| "eval_steps": 5, | |
| "global_step": 192, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.9832, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "eval_loss": 0.9730262160301208, | |
| "eval_runtime": 2.116, | |
| "eval_samples_per_second": 160.683, | |
| "eval_steps_per_second": 1.418, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 16.41232960271884, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.9463, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "eval_loss": 0.9420890212059021, | |
| "eval_runtime": 2.0786, | |
| "eval_samples_per_second": 163.572, | |
| "eval_steps_per_second": 1.443, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 6.185215383740627, | |
| "learning_rate": 5.25e-06, | |
| "loss": 0.8488, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "eval_loss": 0.8247124552726746, | |
| "eval_runtime": 2.0863, | |
| "eval_samples_per_second": 162.965, | |
| "eval_steps_per_second": 1.438, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 4.677780246708798, | |
| "learning_rate": 9e-06, | |
| "loss": 0.7833, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "eval_loss": 0.8149010539054871, | |
| "eval_runtime": 2.0708, | |
| "eval_samples_per_second": 164.186, | |
| "eval_steps_per_second": 1.449, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 4.282490348738236, | |
| "learning_rate": 1.275e-05, | |
| "loss": 0.7797, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "eval_loss": 0.8403318524360657, | |
| "eval_runtime": 2.076, | |
| "eval_samples_per_second": 163.776, | |
| "eval_steps_per_second": 1.445, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 4.312240371628775, | |
| "learning_rate": 1.3195176200175283e-05, | |
| "loss": 0.8542, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "eval_loss": 0.8670275807380676, | |
| "eval_runtime": 2.0781, | |
| "eval_samples_per_second": 163.608, | |
| "eval_steps_per_second": 1.444, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 4.2373297823136244, | |
| "learning_rate": 9.515676612044427e-06, | |
| "loss": 0.8895, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "eval_loss": 0.8718018531799316, | |
| "eval_runtime": 2.0707, | |
| "eval_samples_per_second": 164.196, | |
| "eval_steps_per_second": 1.449, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "grad_norm": 4.44083784051028, | |
| "learning_rate": 6.797580677308734e-06, | |
| "loss": 0.8519, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "eval_loss": 0.859227180480957, | |
| "eval_runtime": 2.0671, | |
| "eval_samples_per_second": 164.485, | |
| "eval_steps_per_second": 1.451, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 4.131620700380954, | |
| "learning_rate": 4.808575415542887e-06, | |
| "loss": 0.8224, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "eval_loss": 0.8491263389587402, | |
| "eval_runtime": 2.0743, | |
| "eval_samples_per_second": 163.912, | |
| "eval_steps_per_second": 1.446, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 4.319858409892453, | |
| "learning_rate": 3.3676619069852654e-06, | |
| "loss": 0.8538, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "eval_loss": 0.8384072780609131, | |
| "eval_runtime": 2.0776, | |
| "eval_samples_per_second": 163.653, | |
| "eval_steps_per_second": 1.444, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0416666666666667, | |
| "grad_norm": 3.80418376995363, | |
| "learning_rate": 2.334947896124909e-06, | |
| "loss": 0.6569, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0416666666666667, | |
| "eval_loss": 0.8294973373413086, | |
| "eval_runtime": 2.0689, | |
| "eval_samples_per_second": 164.335, | |
| "eval_steps_per_second": 1.45, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.1458333333333333, | |
| "grad_norm": 3.3894455787756694, | |
| "learning_rate": 1.603233215095547e-06, | |
| "loss": 0.437, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1458333333333333, | |
| "eval_loss": 0.8457258343696594, | |
| "eval_runtime": 2.0842, | |
| "eval_samples_per_second": 163.13, | |
| "eval_steps_per_second": 1.439, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 3.633089838413966, | |
| "learning_rate": 1.0911174606561334e-06, | |
| "loss": 0.4405, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.8668487071990967, | |
| "eval_runtime": 2.0758, | |
| "eval_samples_per_second": 163.796, | |
| "eval_steps_per_second": 1.445, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.3541666666666667, | |
| "grad_norm": 4.225857095854057, | |
| "learning_rate": 7.373930741131784e-07, | |
| "loss": 0.4331, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.3541666666666667, | |
| "eval_loss": 0.8670875430107117, | |
| "eval_runtime": 2.0785, | |
| "eval_samples_per_second": 163.58, | |
| "eval_steps_per_second": 1.443, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.4583333333333333, | |
| "grad_norm": 3.838684962723822, | |
| "learning_rate": 5.374210410959207e-07, | |
| "loss": 0.448, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.4583333333333333, | |
| "eval_loss": 0.8596971035003662, | |
| "eval_runtime": 2.0789, | |
| "eval_samples_per_second": 163.548, | |
| "eval_steps_per_second": 1.443, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 3.8823844934735114, | |
| "learning_rate": 3.6222476698215175e-07, | |
| "loss": 0.4673, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "eval_loss": 0.8513818383216858, | |
| "eval_runtime": 2.0778, | |
| "eval_samples_per_second": 163.638, | |
| "eval_steps_per_second": 1.444, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 3.2398255350683103, | |
| "learning_rate": 2.462755297384099e-07, | |
| "loss": 0.4298, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "eval_loss": 0.8474181294441223, | |
| "eval_runtime": 2.0907, | |
| "eval_samples_per_second": 162.623, | |
| "eval_steps_per_second": 1.435, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7708333333333335, | |
| "grad_norm": 3.153318195454539, | |
| "learning_rate": 1.7088740175034947e-07, | |
| "loss": 0.4252, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.7708333333333335, | |
| "eval_loss": 0.8457570672035217, | |
| "eval_runtime": 2.0841, | |
| "eval_samples_per_second": 163.139, | |
| "eval_steps_per_second": 1.439, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 3.9154872471233073, | |
| "learning_rate": 1.228102956599465e-07, | |
| "loss": 0.4429, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "eval_loss": 0.8451478481292725, | |
| "eval_runtime": 2.0694, | |
| "eval_samples_per_second": 164.3, | |
| "eval_steps_per_second": 1.45, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9791666666666665, | |
| "grad_norm": 4.304265882610879, | |
| "learning_rate": 9.279207916081227e-08, | |
| "loss": 0.4484, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.9791666666666665, | |
| "eval_loss": 0.8449902534484863, | |
| "eval_runtime": 2.0701, | |
| "eval_samples_per_second": 164.241, | |
| "eval_steps_per_second": 1.449, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0833333333333335, | |
| "grad_norm": 3.2728230120401633, | |
| "learning_rate": 7.448002404850094e-08, | |
| "loss": 0.3634, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.0833333333333335, | |
| "eval_loss": 0.8455283641815186, | |
| "eval_runtime": 2.0713, | |
| "eval_samples_per_second": 164.145, | |
| "eval_steps_per_second": 1.448, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "grad_norm": 3.660020107151519, | |
| "learning_rate": 6.35920070839697e-08, | |
| "loss": 0.3876, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "eval_loss": 0.8467428684234619, | |
| "eval_runtime": 2.0936, | |
| "eval_samples_per_second": 162.401, | |
| "eval_steps_per_second": 1.433, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.2916666666666665, | |
| "grad_norm": 3.8970751627622926, | |
| "learning_rate": 5.7299804687499997e-08, | |
| "loss": 0.3717, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.2916666666666665, | |
| "eval_loss": 0.8481121063232422, | |
| "eval_runtime": 2.0678, | |
| "eval_samples_per_second": 164.429, | |
| "eval_steps_per_second": 1.451, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.3958333333333335, | |
| "grad_norm": 3.386652715934595, | |
| "learning_rate": 5.37771434967624e-08, | |
| "loss": 0.387, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.3958333333333335, | |
| "eval_loss": 0.8493936061859131, | |
| "eval_runtime": 2.1051, | |
| "eval_samples_per_second": 161.51, | |
| "eval_steps_per_second": 1.425, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 3.4200942052169547, | |
| "learning_rate": 5.187403540619925e-08, | |
| "loss": 0.3561, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.85052889585495, | |
| "eval_runtime": 2.0652, | |
| "eval_samples_per_second": 164.632, | |
| "eval_steps_per_second": 1.453, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6041666666666665, | |
| "grad_norm": 3.268980501701993, | |
| "learning_rate": 5.088648238966908e-08, | |
| "loss": 0.4219, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6041666666666665, | |
| "eval_loss": 0.8516257405281067, | |
| "eval_runtime": 2.1146, | |
| "eval_samples_per_second": 160.788, | |
| "eval_steps_per_second": 1.419, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.7083333333333335, | |
| "grad_norm": 3.4285942542360806, | |
| "learning_rate": 5.039701925276604e-08, | |
| "loss": 0.3798, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.7083333333333335, | |
| "eval_loss": 0.8526514172554016, | |
| "eval_runtime": 2.1018, | |
| "eval_samples_per_second": 161.768, | |
| "eval_steps_per_second": 1.427, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "grad_norm": 3.438575160339058, | |
| "learning_rate": 5.0166900048082497e-08, | |
| "loss": 0.3551, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "eval_loss": 0.8536917567253113, | |
| "eval_runtime": 2.1025, | |
| "eval_samples_per_second": 161.713, | |
| "eval_steps_per_second": 1.427, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.9166666666666665, | |
| "grad_norm": 3.1199400563472683, | |
| "learning_rate": 5.0065147322870076e-08, | |
| "loss": 0.3827, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.9166666666666665, | |
| "eval_loss": 0.8546140193939209, | |
| "eval_runtime": 2.0898, | |
| "eval_samples_per_second": 162.691, | |
| "eval_steps_per_second": 1.436, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0208333333333335, | |
| "grad_norm": 3.1711921144705, | |
| "learning_rate": 5.002328628528332e-08, | |
| "loss": 0.3938, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.0208333333333335, | |
| "eval_loss": 0.8555943369865417, | |
| "eval_runtime": 2.0827, | |
| "eval_samples_per_second": 163.25, | |
| "eval_steps_per_second": 1.44, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 3.133782976096458, | |
| "learning_rate": 5.0007484528133236e-08, | |
| "loss": 0.3805, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "eval_loss": 0.8565306663513184, | |
| "eval_runtime": 2.1024, | |
| "eval_samples_per_second": 161.723, | |
| "eval_steps_per_second": 1.427, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.2291666666666665, | |
| "grad_norm": 3.7319435280210085, | |
| "learning_rate": 5.0002110817570477e-08, | |
| "loss": 0.3813, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.2291666666666665, | |
| "eval_loss": 0.8574034571647644, | |
| "eval_runtime": 2.0911, | |
| "eval_samples_per_second": 162.593, | |
| "eval_steps_per_second": 1.435, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 3.5844045117334833, | |
| "learning_rate": 5.0000504842356326e-08, | |
| "loss": 0.3894, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "eval_loss": 0.8581907153129578, | |
| "eval_runtime": 2.0963, | |
| "eval_samples_per_second": 162.194, | |
| "eval_steps_per_second": 1.431, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.4375, | |
| "grad_norm": 3.2964992218641544, | |
| "learning_rate": 5.000009745562451e-08, | |
| "loss": 0.3603, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.4375, | |
| "eval_loss": 0.8588598370552063, | |
| "eval_runtime": 2.0794, | |
| "eval_samples_per_second": 163.512, | |
| "eval_steps_per_second": 1.443, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.5416666666666665, | |
| "grad_norm": 3.307148767623163, | |
| "learning_rate": 5.0000014077810156e-08, | |
| "loss": 0.3515, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.5416666666666665, | |
| "eval_loss": 0.8596634864807129, | |
| "eval_runtime": 2.0755, | |
| "eval_samples_per_second": 163.816, | |
| "eval_steps_per_second": 1.445, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.6458333333333335, | |
| "grad_norm": 3.3334351206179402, | |
| "learning_rate": 5.0000001343508807e-08, | |
| "loss": 0.3433, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.6458333333333335, | |
| "eval_loss": 0.8604967594146729, | |
| "eval_runtime": 2.0699, | |
| "eval_samples_per_second": 164.261, | |
| "eval_steps_per_second": 1.449, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 3.196293836404165, | |
| "learning_rate": 5.000000006747581e-08, | |
| "loss": 0.3511, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "eval_loss": 0.861361026763916, | |
| "eval_runtime": 2.0796, | |
| "eval_samples_per_second": 163.491, | |
| "eval_steps_per_second": 1.443, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.8541666666666665, | |
| "grad_norm": 3.472738636185267, | |
| "learning_rate": 5.0000000001094325e-08, | |
| "loss": 0.3599, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.8541666666666665, | |
| "eval_loss": 0.8619682192802429, | |
| "eval_runtime": 2.0705, | |
| "eval_samples_per_second": 164.215, | |
| "eval_steps_per_second": 1.449, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.9583333333333335, | |
| "grad_norm": 3.6408101963860187, | |
| "learning_rate": 5.000000000000139e-08, | |
| "loss": 0.3994, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.9583333333333335, | |
| "eval_loss": 0.8621244430541992, | |
| "eval_runtime": 2.0725, | |
| "eval_samples_per_second": 164.052, | |
| "eval_steps_per_second": 1.448, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 192, | |
| "total_flos": 5360548577280.0, | |
| "train_loss": 0.507965192819635, | |
| "train_runtime": 5963.6843, | |
| "train_samples_per_second": 2.048, | |
| "train_steps_per_second": 0.032 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 192, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 5, | |
| "total_flos": 5360548577280.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |