| { | |
| "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": 4.7768, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 9.494949494949497e-06, | |
| "loss": 4.0751, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4085923135280609, | |
| "eval_loss": 3.1674067974090576, | |
| "eval_runtime": 2.3527, | |
| "eval_samples_per_second": 464.995, | |
| "eval_steps_per_second": 58.231, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 9.242424242424244e-06, | |
| "loss": 3.4974, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 8.98989898989899e-06, | |
| "loss": 2.9213, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 8.737373737373738e-06, | |
| "loss": 2.5619, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6425960063934326, | |
| "eval_loss": 2.033519983291626, | |
| "eval_runtime": 2.6883, | |
| "eval_samples_per_second": 406.954, | |
| "eval_steps_per_second": 50.962, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 2.1398, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 8.232323232323233e-06, | |
| "loss": 1.8549, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7861060500144958, | |
| "eval_loss": 1.3536856174468994, | |
| "eval_runtime": 2.7119, | |
| "eval_samples_per_second": 403.407, | |
| "eval_steps_per_second": 50.518, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 7.97979797979798e-06, | |
| "loss": 1.5512, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 7.727272727272727e-06, | |
| "loss": 1.2539, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 7.474747474747476e-06, | |
| "loss": 1.106, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8519195318222046, | |
| "eval_loss": 0.9515264630317688, | |
| "eval_runtime": 2.0985, | |
| "eval_samples_per_second": 521.321, | |
| "eval_steps_per_second": 65.284, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 0.9313, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 0.7451, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 6.717171717171718e-06, | |
| "loss": 0.6698, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8756855726242065, | |
| "eval_loss": 0.7151774764060974, | |
| "eval_runtime": 2.6359, | |
| "eval_samples_per_second": 415.04, | |
| "eval_steps_per_second": 51.975, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "learning_rate": 6.464646464646466e-06, | |
| "loss": 0.5083, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "learning_rate": 6.212121212121213e-06, | |
| "loss": 0.4497, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8921389579772949, | |
| "eval_loss": 0.5837920904159546, | |
| "eval_runtime": 2.4768, | |
| "eval_samples_per_second": 441.702, | |
| "eval_steps_per_second": 55.314, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 5.95959595959596e-06, | |
| "loss": 0.3833, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "learning_rate": 5.7070707070707075e-06, | |
| "loss": 0.2833, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 0.2626, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8939670920372009, | |
| "eval_loss": 0.5299845933914185, | |
| "eval_runtime": 2.4222, | |
| "eval_samples_per_second": 451.662, | |
| "eval_steps_per_second": 56.561, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 5.202020202020202e-06, | |
| "loss": 0.2341, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "learning_rate": 4.94949494949495e-06, | |
| "loss": 0.1797, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "learning_rate": 4.696969696969698e-06, | |
| "loss": 0.1762, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8957952260971069, | |
| "eval_loss": 0.4984068274497986, | |
| "eval_runtime": 2.3921, | |
| "eval_samples_per_second": 457.333, | |
| "eval_steps_per_second": 57.271, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1334, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "learning_rate": 4.191919191919192e-06, | |
| "loss": 0.119, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9058501124382019, | |
| "eval_loss": 0.4905829429626465, | |
| "eval_runtime": 2.4787, | |
| "eval_samples_per_second": 441.362, | |
| "eval_steps_per_second": 55.271, | |
| "step": 11880 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 3.93939393939394e-06, | |
| "loss": 0.1192, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "learning_rate": 3.686868686868687e-06, | |
| "loss": 0.0868, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "learning_rate": 3.4343434343434347e-06, | |
| "loss": 0.0919, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8994515538215637, | |
| "eval_loss": 0.48957258462905884, | |
| "eval_runtime": 2.08, | |
| "eval_samples_per_second": 525.969, | |
| "eval_steps_per_second": 65.866, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 10.23, | |
| "learning_rate": 3.181818181818182e-06, | |
| "loss": 0.0706, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "learning_rate": 2.9292929292929295e-06, | |
| "loss": 0.0686, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "learning_rate": 2.676767676767677e-06, | |
| "loss": 0.0722, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9021937847137451, | |
| "eval_loss": 0.501167356967926, | |
| "eval_runtime": 2.5009, | |
| "eval_samples_per_second": 437.434, | |
| "eval_steps_per_second": 54.779, | |
| "step": 14520 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 0.0512, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "learning_rate": 2.171717171717172e-06, | |
| "loss": 0.0517, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9040219187736511, | |
| "eval_loss": 0.4950941503047943, | |
| "eval_runtime": 2.5035, | |
| "eval_samples_per_second": 436.984, | |
| "eval_steps_per_second": 54.723, | |
| "step": 15840 | |
| }, | |
| { | |
| "epoch": 12.12, | |
| "learning_rate": 1.9191919191919192e-06, | |
| "loss": 0.0518, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.0493, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "learning_rate": 1.4141414141414143e-06, | |
| "loss": 0.0353, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9040219187736511, | |
| "eval_loss": 0.49883490800857544, | |
| "eval_runtime": 2.4613, | |
| "eval_samples_per_second": 444.472, | |
| "eval_steps_per_second": 55.661, | |
| "step": 17160 | |
| }, | |
| { | |
| "epoch": 13.26, | |
| "learning_rate": 1.1616161616161617e-06, | |
| "loss": 0.0453, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 0.0334, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9049360156059265, | |
| "eval_loss": 0.5035073161125183, | |
| "eval_runtime": 2.3997, | |
| "eval_samples_per_second": 455.895, | |
| "eval_steps_per_second": 57.091, | |
| "step": 18480 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 6.565656565656567e-07, | |
| "loss": 0.0455, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 14.39, | |
| "learning_rate": 4.040404040404041e-07, | |
| "loss": 0.0378, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 14.77, | |
| "learning_rate": 1.5151515151515152e-07, | |
| "loss": 0.0304, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9067641496658325, | |
| "eval_loss": 0.5008062124252319, | |
| "eval_runtime": 2.0764, | |
| "eval_samples_per_second": 526.877, | |
| "eval_steps_per_second": 65.98, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 19800, | |
| "total_flos": 5215995096399360.0, | |
| "train_loss": 0.8025478858658762, | |
| "train_runtime": 2288.0689, | |
| "train_samples_per_second": 69.202, | |
| "train_steps_per_second": 8.654 | |
| } | |
| ], | |
| "max_steps": 19800, | |
| "num_train_epochs": 15, | |
| "total_flos": 5215995096399360.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |