| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "global_step": 19800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.747474747474748e-06, | |
| "loss": 4.718, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 9.494949494949497e-06, | |
| "loss": 4.0196, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4232175648212433, | |
| "eval_loss": 3.169081211090088, | |
| "eval_runtime": 2.1415, | |
| "eval_samples_per_second": 510.858, | |
| "eval_steps_per_second": 63.974, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 9.242424242424244e-06, | |
| "loss": 3.4554, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 8.98989898989899e-06, | |
| "loss": 2.8904, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 8.737373737373738e-06, | |
| "loss": 2.5409, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6435100436210632, | |
| "eval_loss": 2.04946231842041, | |
| "eval_runtime": 2.1238, | |
| "eval_samples_per_second": 515.114, | |
| "eval_steps_per_second": 64.507, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 2.1224, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 8.232323232323233e-06, | |
| "loss": 1.8396, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7861060500144958, | |
| "eval_loss": 1.3566279411315918, | |
| "eval_runtime": 2.0996, | |
| "eval_samples_per_second": 521.061, | |
| "eval_steps_per_second": 65.252, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 7.97979797979798e-06, | |
| "loss": 1.5406, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 1.2519, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 7.474747474747476e-06, | |
| "loss": 1.0986, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8555758595466614, | |
| "eval_loss": 0.9564884305000305, | |
| "eval_runtime": 2.1078, | |
| "eval_samples_per_second": 519.033, | |
| "eval_steps_per_second": 64.998, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 0.9206, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 0.7389, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 6.717171717171718e-06, | |
| "loss": 0.6711, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.880255937576294, | |
| "eval_loss": 0.7118942141532898, | |
| "eval_runtime": 2.0744, | |
| "eval_samples_per_second": 527.383, | |
| "eval_steps_per_second": 66.043, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "learning_rate": 6.464646464646466e-06, | |
| "loss": 0.5107, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "learning_rate": 6.212121212121213e-06, | |
| "loss": 0.4456, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8912248611450195, | |
| "eval_loss": 0.5804843902587891, | |
| "eval_runtime": 2.0875, | |
| "eval_samples_per_second": 524.077, | |
| "eval_steps_per_second": 65.629, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 5.95959595959596e-06, | |
| "loss": 0.3836, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 5.7070707070707075e-06, | |
| "loss": 0.2848, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.2563, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8994515538215637, | |
| "eval_loss": 0.5123510956764221, | |
| "eval_runtime": 2.1263, | |
| "eval_samples_per_second": 514.504, | |
| "eval_steps_per_second": 64.431, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 5.202020202020202e-06, | |
| "loss": 0.2322, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "learning_rate": 4.94949494949495e-06, | |
| "loss": 0.1742, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "learning_rate": 4.696969696969698e-06, | |
| "loss": 0.1686, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9021937847137451, | |
| "eval_loss": 0.49525973200798035, | |
| "eval_runtime": 2.0716, | |
| "eval_samples_per_second": 528.088, | |
| "eval_steps_per_second": 66.132, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1313, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "learning_rate": 4.191919191919192e-06, | |
| "loss": 0.1135, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9021937847137451, | |
| "eval_loss": 0.486285001039505, | |
| "eval_runtime": 2.0803, | |
| "eval_samples_per_second": 525.893, | |
| "eval_steps_per_second": 65.857, | |
| "step": 11880 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 3.93939393939394e-06, | |
| "loss": 0.1161, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "learning_rate": 3.686868686868687e-06, | |
| "loss": 0.0839, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "learning_rate": 3.4343434343434347e-06, | |
| "loss": 0.0846, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9049360156059265, | |
| "eval_loss": 0.48929038643836975, | |
| "eval_runtime": 2.0752, | |
| "eval_samples_per_second": 527.187, | |
| "eval_steps_per_second": 66.019, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 10.23, | |
| "learning_rate": 3.181818181818182e-06, | |
| "loss": 0.0658, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "learning_rate": 2.9292929292929295e-06, | |
| "loss": 0.0663, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "learning_rate": 2.676767676767677e-06, | |
| "loss": 0.07, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9076782464981079, | |
| "eval_loss": 0.485994428396225, | |
| "eval_runtime": 2.1075, | |
| "eval_samples_per_second": 519.1, | |
| "eval_steps_per_second": 65.006, | |
| "step": 14520 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 0.0489, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "learning_rate": 2.171717171717172e-06, | |
| "loss": 0.0487, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9058501124382019, | |
| "eval_loss": 0.4888700246810913, | |
| "eval_runtime": 2.0826, | |
| "eval_samples_per_second": 525.296, | |
| "eval_steps_per_second": 65.782, | |
| "step": 15840 | |
| }, | |
| { | |
| "epoch": 12.12, | |
| "learning_rate": 1.9191919191919192e-06, | |
| "loss": 0.049, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.0476, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "learning_rate": 1.4141414141414143e-06, | |
| "loss": 0.0342, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9067641496658325, | |
| "eval_loss": 0.4926171898841858, | |
| "eval_runtime": 2.3107, | |
| "eval_samples_per_second": 473.444, | |
| "eval_steps_per_second": 59.289, | |
| "step": 17160 | |
| }, | |
| { | |
| "epoch": 13.26, | |
| "learning_rate": 1.1616161616161617e-06, | |
| "loss": 0.0392, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 0.0332, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9067641496658325, | |
| "eval_loss": 0.495064914226532, | |
| "eval_runtime": 2.1748, | |
| "eval_samples_per_second": 503.039, | |
| "eval_steps_per_second": 62.995, | |
| "step": 18480 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 6.565656565656567e-07, | |
| "loss": 0.0411, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 14.39, | |
| "learning_rate": 4.040404040404041e-07, | |
| "loss": 0.0368, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 14.77, | |
| "learning_rate": 1.5151515151515152e-07, | |
| "loss": 0.029, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9067641496658325, | |
| "eval_loss": 0.4963526129722595, | |
| "eval_runtime": 2.0713, | |
| "eval_samples_per_second": 528.178, | |
| "eval_steps_per_second": 66.143, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 19800, | |
| "total_flos": 5215995096399360.0, | |
| "train_loss": 0.7935866661746093, | |
| "train_runtime": 2196.4523, | |
| "train_samples_per_second": 72.089, | |
| "train_steps_per_second": 9.015 | |
| } | |
| ], | |
| "max_steps": 19800, | |
| "num_train_epochs": 15, | |
| "total_flos": 5215995096399360.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |