| { | |
| "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": 4.7663, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 9.494949494949497e-06, | |
| "loss": 4.1505, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.37934187054634094, | |
| "eval_loss": 3.329326629638672, | |
| "eval_runtime": 3.3016, | |
| "eval_samples_per_second": 331.353, | |
| "eval_steps_per_second": 41.495, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 9.242424242424244e-06, | |
| "loss": 3.6225, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 8.98989898989899e-06, | |
| "loss": 3.0767, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 8.737373737373738e-06, | |
| "loss": 2.7333, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6133455038070679, | |
| "eval_loss": 2.2294859886169434, | |
| "eval_runtime": 2.0812, | |
| "eval_samples_per_second": 525.664, | |
| "eval_steps_per_second": 65.828, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 2.3064, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 8.232323232323233e-06, | |
| "loss": 2.0189, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7586837410926819, | |
| "eval_loss": 1.5133757591247559, | |
| "eval_runtime": 2.0858, | |
| "eval_samples_per_second": 524.509, | |
| "eval_steps_per_second": 65.683, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 7.97979797979798e-06, | |
| "loss": 1.7059, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 1.4008, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 7.474747474747476e-06, | |
| "loss": 1.2504, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8281535506248474, | |
| "eval_loss": 1.0764867067337036, | |
| "eval_runtime": 2.3812, | |
| "eval_samples_per_second": 459.44, | |
| "eval_steps_per_second": 57.535, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 1.0472, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 0.8521, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 6.717171717171718e-06, | |
| "loss": 0.7733, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8628884553909302, | |
| "eval_loss": 0.7936682105064392, | |
| "eval_runtime": 2.4487, | |
| "eval_samples_per_second": 446.77, | |
| "eval_steps_per_second": 55.948, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "learning_rate": 6.464646464646466e-06, | |
| "loss": 0.6059, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "learning_rate": 6.212121212121213e-06, | |
| "loss": 0.5217, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8784278035163879, | |
| "eval_loss": 0.643801212310791, | |
| "eval_runtime": 2.8497, | |
| "eval_samples_per_second": 383.9, | |
| "eval_steps_per_second": 48.075, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 5.95959595959596e-06, | |
| "loss": 0.4521, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 5.7070707070707075e-06, | |
| "loss": 0.3498, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.3148, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8857403993606567, | |
| "eval_loss": 0.5733386874198914, | |
| "eval_runtime": 2.4195, | |
| "eval_samples_per_second": 452.164, | |
| "eval_steps_per_second": 56.624, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 5.202020202020202e-06, | |
| "loss": 0.2811, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "learning_rate": 4.94949494949495e-06, | |
| "loss": 0.2148, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "learning_rate": 4.696969696969698e-06, | |
| "loss": 0.2067, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8912248611450195, | |
| "eval_loss": 0.536188542842865, | |
| "eval_runtime": 2.4481, | |
| "eval_samples_per_second": 446.874, | |
| "eval_steps_per_second": 55.961, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1633, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "learning_rate": 4.191919191919192e-06, | |
| "loss": 0.1507, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8903107643127441, | |
| "eval_loss": 0.5098453164100647, | |
| "eval_runtime": 2.5274, | |
| "eval_samples_per_second": 432.848, | |
| "eval_steps_per_second": 54.205, | |
| "step": 11880 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 3.93939393939394e-06, | |
| "loss": 0.1385, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "learning_rate": 3.686868686868687e-06, | |
| "loss": 0.1079, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "learning_rate": 3.4343434343434347e-06, | |
| "loss": 0.1024, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8976234197616577, | |
| "eval_loss": 0.5077915191650391, | |
| "eval_runtime": 2.4776, | |
| "eval_samples_per_second": 441.549, | |
| "eval_steps_per_second": 55.295, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 10.23, | |
| "learning_rate": 3.181818181818182e-06, | |
| "loss": 0.083, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "learning_rate": 2.9292929292929295e-06, | |
| "loss": 0.0784, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "learning_rate": 2.676767676767677e-06, | |
| "loss": 0.0837, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8967093229293823, | |
| "eval_loss": 0.5054203271865845, | |
| "eval_runtime": 2.4447, | |
| "eval_samples_per_second": 447.49, | |
| "eval_steps_per_second": 56.039, | |
| "step": 14520 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 0.0587, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "learning_rate": 2.171717171717172e-06, | |
| "loss": 0.0608, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8957952260971069, | |
| "eval_loss": 0.5061630606651306, | |
| "eval_runtime": 2.4836, | |
| "eval_samples_per_second": 440.493, | |
| "eval_steps_per_second": 55.162, | |
| "step": 15840 | |
| }, | |
| { | |
| "epoch": 12.12, | |
| "learning_rate": 1.9191919191919192e-06, | |
| "loss": 0.0601, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.0561, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "learning_rate": 1.4141414141414143e-06, | |
| "loss": 0.0426, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9012796878814697, | |
| "eval_loss": 0.5072430968284607, | |
| "eval_runtime": 2.5695, | |
| "eval_samples_per_second": 425.758, | |
| "eval_steps_per_second": 53.317, | |
| "step": 17160 | |
| }, | |
| { | |
| "epoch": 13.26, | |
| "learning_rate": 1.1616161616161617e-06, | |
| "loss": 0.0493, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 0.0374, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9040219187736511, | |
| "eval_loss": 0.5110482573509216, | |
| "eval_runtime": 2.4737, | |
| "eval_samples_per_second": 442.248, | |
| "eval_steps_per_second": 55.382, | |
| "step": 18480 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 6.565656565656567e-07, | |
| "loss": 0.0491, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 14.39, | |
| "learning_rate": 4.040404040404041e-07, | |
| "loss": 0.041, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 14.77, | |
| "learning_rate": 1.5151515151515152e-07, | |
| "loss": 0.0346, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9003656506538391, | |
| "eval_loss": 0.5111713409423828, | |
| "eval_runtime": 2.5194, | |
| "eval_samples_per_second": 434.227, | |
| "eval_steps_per_second": 54.378, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 19800, | |
| "total_flos": 5215995096399360.0, | |
| "train_loss": 0.8604840965945312, | |
| "train_runtime": 2276.2445, | |
| "train_samples_per_second": 69.562, | |
| "train_steps_per_second": 8.699 | |
| } | |
| ], | |
| "max_steps": 19800, | |
| "num_train_epochs": 15, | |
| "total_flos": 5215995096399360.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |