|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.02, |
|
"eval_steps": 2000, |
|
"global_step": 4000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 5e-05, |
|
"grad_norm": 31.625, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 11.0902, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.0001, |
|
"grad_norm": 46.0, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 11.0927, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.00015, |
|
"grad_norm": 30.25, |
|
"learning_rate": 1.5e-06, |
|
"loss": 11.0934, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.0002, |
|
"grad_norm": 50.5, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 11.0834, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.00025, |
|
"grad_norm": 20.25, |
|
"learning_rate": 2.5e-06, |
|
"loss": 11.082, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.0003, |
|
"grad_norm": 53.0, |
|
"learning_rate": 3e-06, |
|
"loss": 11.096, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.00035, |
|
"grad_norm": 32.75, |
|
"learning_rate": 3.5000000000000004e-06, |
|
"loss": 11.0645, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.0004, |
|
"grad_norm": 36.0, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 11.0728, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.00045, |
|
"grad_norm": 20.25, |
|
"learning_rate": 4.5e-06, |
|
"loss": 11.0674, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.0005, |
|
"grad_norm": 25.125, |
|
"learning_rate": 5e-06, |
|
"loss": 11.0666, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.00055, |
|
"grad_norm": 16.75, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 11.0664, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.0006, |
|
"grad_norm": 22.125, |
|
"learning_rate": 6e-06, |
|
"loss": 11.0592, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.00065, |
|
"grad_norm": 23.625, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 11.049, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.0007, |
|
"grad_norm": 13.3125, |
|
"learning_rate": 7.000000000000001e-06, |
|
"loss": 11.0685, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.00075, |
|
"grad_norm": 15.0, |
|
"learning_rate": 7.5e-06, |
|
"loss": 11.0337, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.0008, |
|
"grad_norm": 30.125, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 11.0335, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.00085, |
|
"grad_norm": 14.75, |
|
"learning_rate": 8.500000000000002e-06, |
|
"loss": 11.0266, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.0009, |
|
"grad_norm": 17.375, |
|
"learning_rate": 9e-06, |
|
"loss": 11.0171, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.00095, |
|
"grad_norm": 42.0, |
|
"learning_rate": 9.5e-06, |
|
"loss": 11.0091, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.001, |
|
"grad_norm": 29.25, |
|
"learning_rate": 1e-05, |
|
"loss": 11.0014, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.00105, |
|
"grad_norm": 18.375, |
|
"learning_rate": 1.05e-05, |
|
"loss": 10.993, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.0011, |
|
"grad_norm": 40.25, |
|
"learning_rate": 1.1000000000000001e-05, |
|
"loss": 10.973, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.00115, |
|
"grad_norm": 39.0, |
|
"learning_rate": 1.1500000000000002e-05, |
|
"loss": 10.9513, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.0012, |
|
"grad_norm": 23.375, |
|
"learning_rate": 1.2e-05, |
|
"loss": 10.9342, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.00125, |
|
"grad_norm": 33.25, |
|
"learning_rate": 1.25e-05, |
|
"loss": 10.9172, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.0013, |
|
"grad_norm": 42.25, |
|
"learning_rate": 1.3000000000000001e-05, |
|
"loss": 10.9073, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.00135, |
|
"grad_norm": 10.625, |
|
"learning_rate": 1.3500000000000001e-05, |
|
"loss": 10.8946, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.0014, |
|
"grad_norm": 10.6875, |
|
"learning_rate": 1.4000000000000001e-05, |
|
"loss": 10.8823, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.00145, |
|
"grad_norm": 63.75, |
|
"learning_rate": 1.45e-05, |
|
"loss": 10.867, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.0015, |
|
"grad_norm": 38.5, |
|
"learning_rate": 1.5e-05, |
|
"loss": 10.8479, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.00155, |
|
"grad_norm": 17.375, |
|
"learning_rate": 1.55e-05, |
|
"loss": 10.8308, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.0016, |
|
"grad_norm": 42.0, |
|
"learning_rate": 1.6000000000000003e-05, |
|
"loss": 10.8119, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.00165, |
|
"grad_norm": 22.25, |
|
"learning_rate": 1.65e-05, |
|
"loss": 10.8032, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.0017, |
|
"grad_norm": 13.0, |
|
"learning_rate": 1.7000000000000003e-05, |
|
"loss": 10.7896, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.00175, |
|
"grad_norm": 32.5, |
|
"learning_rate": 1.75e-05, |
|
"loss": 10.7853, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.0018, |
|
"grad_norm": 12.75, |
|
"learning_rate": 1.8e-05, |
|
"loss": 10.7559, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.00185, |
|
"grad_norm": 16.75, |
|
"learning_rate": 1.85e-05, |
|
"loss": 10.7292, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.0019, |
|
"grad_norm": 13.5, |
|
"learning_rate": 1.9e-05, |
|
"loss": 10.7117, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.00195, |
|
"grad_norm": 26.625, |
|
"learning_rate": 1.9500000000000003e-05, |
|
"loss": 10.6682, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.002, |
|
"grad_norm": 27.625, |
|
"learning_rate": 2e-05, |
|
"loss": 10.6554, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.00205, |
|
"grad_norm": 18.5, |
|
"learning_rate": 2.05e-05, |
|
"loss": 10.6264, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.0021, |
|
"grad_norm": 22.375, |
|
"learning_rate": 2.1e-05, |
|
"loss": 10.5963, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.00215, |
|
"grad_norm": 16.625, |
|
"learning_rate": 2.15e-05, |
|
"loss": 10.5767, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.0022, |
|
"grad_norm": 20.5, |
|
"learning_rate": 2.2000000000000003e-05, |
|
"loss": 10.5003, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.00225, |
|
"grad_norm": 128.0, |
|
"learning_rate": 2.25e-05, |
|
"loss": 10.4183, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.0023, |
|
"grad_norm": 35.25, |
|
"learning_rate": 2.3000000000000003e-05, |
|
"loss": 10.4039, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.00235, |
|
"grad_norm": 46.5, |
|
"learning_rate": 2.35e-05, |
|
"loss": 10.337, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.0024, |
|
"grad_norm": 31.75, |
|
"learning_rate": 2.4e-05, |
|
"loss": 10.287, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.00245, |
|
"grad_norm": 15.5625, |
|
"learning_rate": 2.45e-05, |
|
"loss": 10.258, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.0025, |
|
"grad_norm": 30.375, |
|
"learning_rate": 2.5e-05, |
|
"loss": 10.2216, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.00255, |
|
"grad_norm": 14.125, |
|
"learning_rate": 2.5500000000000003e-05, |
|
"loss": 10.1201, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.0026, |
|
"grad_norm": 28.5, |
|
"learning_rate": 2.6000000000000002e-05, |
|
"loss": 10.054, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.00265, |
|
"grad_norm": 14.0625, |
|
"learning_rate": 2.6500000000000004e-05, |
|
"loss": 9.9927, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.0027, |
|
"grad_norm": 10.375, |
|
"learning_rate": 2.7000000000000002e-05, |
|
"loss": 9.8655, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.00275, |
|
"grad_norm": 30.5, |
|
"learning_rate": 2.7500000000000004e-05, |
|
"loss": 9.7691, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.0028, |
|
"grad_norm": 20.125, |
|
"learning_rate": 2.8000000000000003e-05, |
|
"loss": 9.612, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.00285, |
|
"grad_norm": 12.0625, |
|
"learning_rate": 2.8499999999999998e-05, |
|
"loss": 9.4771, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.0029, |
|
"grad_norm": 37.75, |
|
"learning_rate": 2.9e-05, |
|
"loss": 9.3691, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.00295, |
|
"grad_norm": 9.5625, |
|
"learning_rate": 2.95e-05, |
|
"loss": 9.1528, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.003, |
|
"grad_norm": 62.0, |
|
"learning_rate": 3e-05, |
|
"loss": 9.0495, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.00305, |
|
"grad_norm": 25.125, |
|
"learning_rate": 3.05e-05, |
|
"loss": 8.8528, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.0031, |
|
"grad_norm": 32.75, |
|
"learning_rate": 3.1e-05, |
|
"loss": 8.7636, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.00315, |
|
"grad_norm": 25.75, |
|
"learning_rate": 3.15e-05, |
|
"loss": 8.6855, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.0032, |
|
"grad_norm": 18.625, |
|
"learning_rate": 3.2000000000000005e-05, |
|
"loss": 8.615, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.00325, |
|
"grad_norm": 20.5, |
|
"learning_rate": 3.2500000000000004e-05, |
|
"loss": 8.5321, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.0033, |
|
"grad_norm": 188.0, |
|
"learning_rate": 3.3e-05, |
|
"loss": 8.4743, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.00335, |
|
"grad_norm": 20.5, |
|
"learning_rate": 3.35e-05, |
|
"loss": 8.4076, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.0034, |
|
"grad_norm": 58.25, |
|
"learning_rate": 3.4000000000000007e-05, |
|
"loss": 8.3754, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.00345, |
|
"grad_norm": 26.5, |
|
"learning_rate": 3.45e-05, |
|
"loss": 8.3274, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.0035, |
|
"grad_norm": 11.875, |
|
"learning_rate": 3.5e-05, |
|
"loss": 8.2515, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.00355, |
|
"grad_norm": 39.5, |
|
"learning_rate": 3.55e-05, |
|
"loss": 8.2337, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.0036, |
|
"grad_norm": 30.625, |
|
"learning_rate": 3.6e-05, |
|
"loss": 8.1032, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.00365, |
|
"grad_norm": 9.5, |
|
"learning_rate": 3.65e-05, |
|
"loss": 8.0804, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.0037, |
|
"grad_norm": 24.125, |
|
"learning_rate": 3.7e-05, |
|
"loss": 8.0145, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.00375, |
|
"grad_norm": 11.375, |
|
"learning_rate": 3.7500000000000003e-05, |
|
"loss": 7.9738, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.0038, |
|
"grad_norm": 76.0, |
|
"learning_rate": 3.8e-05, |
|
"loss": 7.943, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.00385, |
|
"grad_norm": 28.625, |
|
"learning_rate": 3.85e-05, |
|
"loss": 7.9069, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.0039, |
|
"grad_norm": 35.75, |
|
"learning_rate": 3.9000000000000006e-05, |
|
"loss": 7.8638, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.00395, |
|
"grad_norm": 26.25, |
|
"learning_rate": 3.9500000000000005e-05, |
|
"loss": 7.8659, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.004, |
|
"grad_norm": 64.0, |
|
"learning_rate": 4e-05, |
|
"loss": 7.8315, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.00405, |
|
"grad_norm": 11.0625, |
|
"learning_rate": 4.05e-05, |
|
"loss": 7.8273, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.0041, |
|
"grad_norm": 18.875, |
|
"learning_rate": 4.1e-05, |
|
"loss": 7.7964, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.00415, |
|
"grad_norm": 30.375, |
|
"learning_rate": 4.15e-05, |
|
"loss": 7.7627, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.0042, |
|
"grad_norm": 12.1875, |
|
"learning_rate": 4.2e-05, |
|
"loss": 7.7935, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.00425, |
|
"grad_norm": 27.25, |
|
"learning_rate": 4.25e-05, |
|
"loss": 7.7644, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.0043, |
|
"grad_norm": 12.0625, |
|
"learning_rate": 4.3e-05, |
|
"loss": 7.7387, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.00435, |
|
"grad_norm": 24.25, |
|
"learning_rate": 4.35e-05, |
|
"loss": 7.7656, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.0044, |
|
"grad_norm": 75.0, |
|
"learning_rate": 4.4000000000000006e-05, |
|
"loss": 7.7289, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.00445, |
|
"grad_norm": 49.25, |
|
"learning_rate": 4.4500000000000004e-05, |
|
"loss": 7.7205, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.0045, |
|
"grad_norm": 7.4375, |
|
"learning_rate": 4.5e-05, |
|
"loss": 7.721, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.00455, |
|
"grad_norm": 23.875, |
|
"learning_rate": 4.55e-05, |
|
"loss": 7.6778, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.0046, |
|
"grad_norm": 27.125, |
|
"learning_rate": 4.600000000000001e-05, |
|
"loss": 7.6931, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.00465, |
|
"grad_norm": 3.5625, |
|
"learning_rate": 4.6500000000000005e-05, |
|
"loss": 7.6755, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.0047, |
|
"grad_norm": 12.1875, |
|
"learning_rate": 4.7e-05, |
|
"loss": 7.6832, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.00475, |
|
"grad_norm": 76.5, |
|
"learning_rate": 4.75e-05, |
|
"loss": 7.6759, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.0048, |
|
"grad_norm": 6.125, |
|
"learning_rate": 4.8e-05, |
|
"loss": 7.6717, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.00485, |
|
"grad_norm": 82.0, |
|
"learning_rate": 4.85e-05, |
|
"loss": 7.6602, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.0049, |
|
"grad_norm": 107.5, |
|
"learning_rate": 4.9e-05, |
|
"loss": 7.638, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.00495, |
|
"grad_norm": 14.1875, |
|
"learning_rate": 4.9500000000000004e-05, |
|
"loss": 7.6483, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.005, |
|
"grad_norm": 13.5, |
|
"learning_rate": 5e-05, |
|
"loss": 7.6629, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.00505, |
|
"grad_norm": 15.4375, |
|
"learning_rate": 4.9997487437185933e-05, |
|
"loss": 7.6448, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 0.0051, |
|
"grad_norm": 26.125, |
|
"learning_rate": 4.9994974874371864e-05, |
|
"loss": 7.6476, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 0.00515, |
|
"grad_norm": 12.3125, |
|
"learning_rate": 4.999246231155779e-05, |
|
"loss": 7.6406, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 0.0052, |
|
"grad_norm": 58.5, |
|
"learning_rate": 4.998994974874372e-05, |
|
"loss": 7.6537, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 0.00525, |
|
"grad_norm": 31.5, |
|
"learning_rate": 4.998743718592965e-05, |
|
"loss": 7.6293, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.0053, |
|
"grad_norm": 25.5, |
|
"learning_rate": 4.998492462311558e-05, |
|
"loss": 7.6318, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 0.00535, |
|
"grad_norm": 8.5625, |
|
"learning_rate": 4.9982412060301506e-05, |
|
"loss": 7.6348, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 0.0054, |
|
"grad_norm": 43.25, |
|
"learning_rate": 4.997989949748744e-05, |
|
"loss": 7.6476, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 0.00545, |
|
"grad_norm": 30.75, |
|
"learning_rate": 4.997738693467337e-05, |
|
"loss": 7.6541, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 0.0055, |
|
"grad_norm": 3.796875, |
|
"learning_rate": 4.99748743718593e-05, |
|
"loss": 7.599, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.00555, |
|
"grad_norm": 3.0, |
|
"learning_rate": 4.997236180904522e-05, |
|
"loss": 7.6384, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 0.0056, |
|
"grad_norm": 26.125, |
|
"learning_rate": 4.996984924623116e-05, |
|
"loss": 7.6353, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 0.00565, |
|
"grad_norm": 11.375, |
|
"learning_rate": 4.9967336683417085e-05, |
|
"loss": 7.6023, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 0.0057, |
|
"grad_norm": 33.5, |
|
"learning_rate": 4.9964824120603016e-05, |
|
"loss": 7.6397, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 0.00575, |
|
"grad_norm": 2.734375, |
|
"learning_rate": 4.996231155778895e-05, |
|
"loss": 7.6122, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.0058, |
|
"grad_norm": 11.75, |
|
"learning_rate": 4.995979899497488e-05, |
|
"loss": 7.6448, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 0.00585, |
|
"grad_norm": 3.65625, |
|
"learning_rate": 4.99572864321608e-05, |
|
"loss": 7.6119, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 0.0059, |
|
"grad_norm": 101.5, |
|
"learning_rate": 4.995477386934674e-05, |
|
"loss": 7.6335, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 0.00595, |
|
"grad_norm": 46.0, |
|
"learning_rate": 4.9952261306532665e-05, |
|
"loss": 7.6137, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 0.006, |
|
"grad_norm": 3.828125, |
|
"learning_rate": 4.9949748743718596e-05, |
|
"loss": 7.6085, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.00605, |
|
"grad_norm": 26.125, |
|
"learning_rate": 4.994723618090453e-05, |
|
"loss": 7.6161, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 0.0061, |
|
"grad_norm": 12.125, |
|
"learning_rate": 4.994472361809046e-05, |
|
"loss": 7.6085, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 0.00615, |
|
"grad_norm": 5.40625, |
|
"learning_rate": 4.994221105527638e-05, |
|
"loss": 7.6353, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 0.0062, |
|
"grad_norm": 2.46875, |
|
"learning_rate": 4.993969849246231e-05, |
|
"loss": 7.6219, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 0.00625, |
|
"grad_norm": 5.375, |
|
"learning_rate": 4.9937185929648244e-05, |
|
"loss": 7.6237, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.0063, |
|
"grad_norm": 45.5, |
|
"learning_rate": 4.9934673366834175e-05, |
|
"loss": 7.6058, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 0.00635, |
|
"grad_norm": 70.0, |
|
"learning_rate": 4.99321608040201e-05, |
|
"loss": 7.5939, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 0.0064, |
|
"grad_norm": 102.5, |
|
"learning_rate": 4.992964824120604e-05, |
|
"loss": 7.6313, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 0.00645, |
|
"grad_norm": 31.0, |
|
"learning_rate": 4.992713567839196e-05, |
|
"loss": 7.5911, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 0.0065, |
|
"grad_norm": 34.0, |
|
"learning_rate": 4.992462311557789e-05, |
|
"loss": 7.6197, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.00655, |
|
"grad_norm": 3.78125, |
|
"learning_rate": 4.992211055276382e-05, |
|
"loss": 7.5896, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 0.0066, |
|
"grad_norm": 9.5, |
|
"learning_rate": 4.9919597989949754e-05, |
|
"loss": 7.5665, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 0.00665, |
|
"grad_norm": 13.3125, |
|
"learning_rate": 4.991708542713568e-05, |
|
"loss": 7.6148, |
|
"step": 1330 |
|
}, |
|
{ |
|
"epoch": 0.0067, |
|
"grad_norm": 3.796875, |
|
"learning_rate": 4.991457286432161e-05, |
|
"loss": 7.5974, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 0.00675, |
|
"grad_norm": 13.9375, |
|
"learning_rate": 4.991206030150754e-05, |
|
"loss": 7.6231, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.0068, |
|
"grad_norm": 4.9375, |
|
"learning_rate": 4.990954773869347e-05, |
|
"loss": 7.573, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 0.00685, |
|
"grad_norm": 27.375, |
|
"learning_rate": 4.99070351758794e-05, |
|
"loss": 7.6299, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 0.0069, |
|
"grad_norm": 9.6875, |
|
"learning_rate": 4.990452261306533e-05, |
|
"loss": 7.6229, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 0.00695, |
|
"grad_norm": 4.28125, |
|
"learning_rate": 4.990201005025126e-05, |
|
"loss": 7.6073, |
|
"step": 1390 |
|
}, |
|
{ |
|
"epoch": 0.007, |
|
"grad_norm": 53.0, |
|
"learning_rate": 4.989949748743719e-05, |
|
"loss": 7.6114, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.00705, |
|
"grad_norm": 7.15625, |
|
"learning_rate": 4.989698492462312e-05, |
|
"loss": 7.5919, |
|
"step": 1410 |
|
}, |
|
{ |
|
"epoch": 0.0071, |
|
"grad_norm": 22.75, |
|
"learning_rate": 4.9894472361809044e-05, |
|
"loss": 7.6078, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 0.00715, |
|
"grad_norm": 1.6640625, |
|
"learning_rate": 4.9891959798994975e-05, |
|
"loss": 7.5975, |
|
"step": 1430 |
|
}, |
|
{ |
|
"epoch": 0.0072, |
|
"grad_norm": 5.375, |
|
"learning_rate": 4.9889447236180906e-05, |
|
"loss": 7.5817, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 0.00725, |
|
"grad_norm": 25.625, |
|
"learning_rate": 4.988693467336684e-05, |
|
"loss": 7.6091, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 0.0073, |
|
"grad_norm": 5.5625, |
|
"learning_rate": 4.988442211055276e-05, |
|
"loss": 7.5941, |
|
"step": 1460 |
|
}, |
|
{ |
|
"epoch": 0.00735, |
|
"grad_norm": 7.4375, |
|
"learning_rate": 4.98819095477387e-05, |
|
"loss": 7.6132, |
|
"step": 1470 |
|
}, |
|
{ |
|
"epoch": 0.0074, |
|
"grad_norm": 6.1875, |
|
"learning_rate": 4.9879396984924623e-05, |
|
"loss": 7.5393, |
|
"step": 1480 |
|
}, |
|
{ |
|
"epoch": 0.00745, |
|
"grad_norm": 1.5390625, |
|
"learning_rate": 4.9876884422110554e-05, |
|
"loss": 7.5867, |
|
"step": 1490 |
|
}, |
|
{ |
|
"epoch": 0.0075, |
|
"grad_norm": 16.375, |
|
"learning_rate": 4.9874371859296486e-05, |
|
"loss": 7.581, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.00755, |
|
"grad_norm": 17.125, |
|
"learning_rate": 4.9871859296482417e-05, |
|
"loss": 7.6056, |
|
"step": 1510 |
|
}, |
|
{ |
|
"epoch": 0.0076, |
|
"grad_norm": 41.0, |
|
"learning_rate": 4.986934673366834e-05, |
|
"loss": 7.6051, |
|
"step": 1520 |
|
}, |
|
{ |
|
"epoch": 0.00765, |
|
"grad_norm": 2.84375, |
|
"learning_rate": 4.986683417085428e-05, |
|
"loss": 7.6122, |
|
"step": 1530 |
|
}, |
|
{ |
|
"epoch": 0.0077, |
|
"grad_norm": 1.9296875, |
|
"learning_rate": 4.98643216080402e-05, |
|
"loss": 7.5703, |
|
"step": 1540 |
|
}, |
|
{ |
|
"epoch": 0.00775, |
|
"grad_norm": 8.5, |
|
"learning_rate": 4.9861809045226134e-05, |
|
"loss": 7.6179, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 0.0078, |
|
"grad_norm": 28.875, |
|
"learning_rate": 4.985929648241206e-05, |
|
"loss": 7.5747, |
|
"step": 1560 |
|
}, |
|
{ |
|
"epoch": 0.00785, |
|
"grad_norm": 13.25, |
|
"learning_rate": 4.9856783919597996e-05, |
|
"loss": 7.5935, |
|
"step": 1570 |
|
}, |
|
{ |
|
"epoch": 0.0079, |
|
"grad_norm": 13.75, |
|
"learning_rate": 4.985427135678392e-05, |
|
"loss": 7.6138, |
|
"step": 1580 |
|
}, |
|
{ |
|
"epoch": 0.00795, |
|
"grad_norm": 3.703125, |
|
"learning_rate": 4.985175879396985e-05, |
|
"loss": 7.5876, |
|
"step": 1590 |
|
}, |
|
{ |
|
"epoch": 0.008, |
|
"grad_norm": 4.4375, |
|
"learning_rate": 4.984924623115578e-05, |
|
"loss": 7.5742, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 0.00805, |
|
"grad_norm": 3.28125, |
|
"learning_rate": 4.984673366834171e-05, |
|
"loss": 7.5962, |
|
"step": 1610 |
|
}, |
|
{ |
|
"epoch": 0.0081, |
|
"grad_norm": 4.78125, |
|
"learning_rate": 4.984422110552764e-05, |
|
"loss": 7.5845, |
|
"step": 1620 |
|
}, |
|
{ |
|
"epoch": 0.00815, |
|
"grad_norm": 2.109375, |
|
"learning_rate": 4.9841708542713575e-05, |
|
"loss": 7.5543, |
|
"step": 1630 |
|
}, |
|
{ |
|
"epoch": 0.0082, |
|
"grad_norm": 6.25, |
|
"learning_rate": 4.98391959798995e-05, |
|
"loss": 7.62, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 0.00825, |
|
"grad_norm": 4.9375, |
|
"learning_rate": 4.983668341708543e-05, |
|
"loss": 7.5663, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 0.0083, |
|
"grad_norm": 2.0625, |
|
"learning_rate": 4.983417085427136e-05, |
|
"loss": 7.614, |
|
"step": 1660 |
|
}, |
|
{ |
|
"epoch": 0.00835, |
|
"grad_norm": 9.5, |
|
"learning_rate": 4.983165829145729e-05, |
|
"loss": 7.552, |
|
"step": 1670 |
|
}, |
|
{ |
|
"epoch": 0.0084, |
|
"grad_norm": 13.9375, |
|
"learning_rate": 4.982914572864322e-05, |
|
"loss": 7.5656, |
|
"step": 1680 |
|
}, |
|
{ |
|
"epoch": 0.00845, |
|
"grad_norm": 11.4375, |
|
"learning_rate": 4.982663316582915e-05, |
|
"loss": 7.6072, |
|
"step": 1690 |
|
}, |
|
{ |
|
"epoch": 0.0085, |
|
"grad_norm": 17.5, |
|
"learning_rate": 4.982412060301508e-05, |
|
"loss": 7.5712, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 0.00855, |
|
"grad_norm": 50.25, |
|
"learning_rate": 4.982160804020101e-05, |
|
"loss": 7.5918, |
|
"step": 1710 |
|
}, |
|
{ |
|
"epoch": 0.0086, |
|
"grad_norm": 1.9296875, |
|
"learning_rate": 4.9819095477386934e-05, |
|
"loss": 7.5785, |
|
"step": 1720 |
|
}, |
|
{ |
|
"epoch": 0.00865, |
|
"grad_norm": 5.40625, |
|
"learning_rate": 4.9816582914572865e-05, |
|
"loss": 7.5781, |
|
"step": 1730 |
|
}, |
|
{ |
|
"epoch": 0.0087, |
|
"grad_norm": 1.4921875, |
|
"learning_rate": 4.9814070351758796e-05, |
|
"loss": 7.5902, |
|
"step": 1740 |
|
}, |
|
{ |
|
"epoch": 0.00875, |
|
"grad_norm": 7.8125, |
|
"learning_rate": 4.981155778894473e-05, |
|
"loss": 7.6011, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 0.0088, |
|
"grad_norm": 17.25, |
|
"learning_rate": 4.980904522613066e-05, |
|
"loss": 7.5952, |
|
"step": 1760 |
|
}, |
|
{ |
|
"epoch": 0.00885, |
|
"grad_norm": 7.6875, |
|
"learning_rate": 4.980653266331658e-05, |
|
"loss": 7.588, |
|
"step": 1770 |
|
}, |
|
{ |
|
"epoch": 0.0089, |
|
"grad_norm": 12.1875, |
|
"learning_rate": 4.980402010050251e-05, |
|
"loss": 7.5903, |
|
"step": 1780 |
|
}, |
|
{ |
|
"epoch": 0.00895, |
|
"grad_norm": 31.875, |
|
"learning_rate": 4.9801507537688444e-05, |
|
"loss": 7.5423, |
|
"step": 1790 |
|
}, |
|
{ |
|
"epoch": 0.009, |
|
"grad_norm": 7.46875, |
|
"learning_rate": 4.9798994974874375e-05, |
|
"loss": 7.5941, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 0.00905, |
|
"grad_norm": 18.25, |
|
"learning_rate": 4.97964824120603e-05, |
|
"loss": 7.6326, |
|
"step": 1810 |
|
}, |
|
{ |
|
"epoch": 0.0091, |
|
"grad_norm": 4.0, |
|
"learning_rate": 4.979396984924624e-05, |
|
"loss": 7.5826, |
|
"step": 1820 |
|
}, |
|
{ |
|
"epoch": 0.00915, |
|
"grad_norm": 2.296875, |
|
"learning_rate": 4.979145728643216e-05, |
|
"loss": 7.5959, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 0.0092, |
|
"grad_norm": 18.0, |
|
"learning_rate": 4.978894472361809e-05, |
|
"loss": 7.6069, |
|
"step": 1840 |
|
}, |
|
{ |
|
"epoch": 0.00925, |
|
"grad_norm": 11.1875, |
|
"learning_rate": 4.9786432160804024e-05, |
|
"loss": 7.6083, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 0.0093, |
|
"grad_norm": 17.25, |
|
"learning_rate": 4.9783919597989955e-05, |
|
"loss": 7.5944, |
|
"step": 1860 |
|
}, |
|
{ |
|
"epoch": 0.00935, |
|
"grad_norm": 6.59375, |
|
"learning_rate": 4.978140703517588e-05, |
|
"loss": 7.5846, |
|
"step": 1870 |
|
}, |
|
{ |
|
"epoch": 0.0094, |
|
"grad_norm": 16.75, |
|
"learning_rate": 4.977889447236181e-05, |
|
"loss": 7.6238, |
|
"step": 1880 |
|
}, |
|
{ |
|
"epoch": 0.00945, |
|
"grad_norm": 2.890625, |
|
"learning_rate": 4.977638190954774e-05, |
|
"loss": 7.5763, |
|
"step": 1890 |
|
}, |
|
{ |
|
"epoch": 0.0095, |
|
"grad_norm": 13.0625, |
|
"learning_rate": 4.977386934673367e-05, |
|
"loss": 7.5709, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 0.00955, |
|
"grad_norm": 23.25, |
|
"learning_rate": 4.9771356783919596e-05, |
|
"loss": 7.5941, |
|
"step": 1910 |
|
}, |
|
{ |
|
"epoch": 0.0096, |
|
"grad_norm": 27.5, |
|
"learning_rate": 4.9768844221105534e-05, |
|
"loss": 7.5912, |
|
"step": 1920 |
|
}, |
|
{ |
|
"epoch": 0.00965, |
|
"grad_norm": 21.25, |
|
"learning_rate": 4.976633165829146e-05, |
|
"loss": 7.6247, |
|
"step": 1930 |
|
}, |
|
{ |
|
"epoch": 0.0097, |
|
"grad_norm": 13.125, |
|
"learning_rate": 4.976381909547739e-05, |
|
"loss": 7.5757, |
|
"step": 1940 |
|
}, |
|
{ |
|
"epoch": 0.00975, |
|
"grad_norm": 5.1875, |
|
"learning_rate": 4.976130653266332e-05, |
|
"loss": 7.5845, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 0.0098, |
|
"grad_norm": 56.25, |
|
"learning_rate": 4.975879396984925e-05, |
|
"loss": 7.5584, |
|
"step": 1960 |
|
}, |
|
{ |
|
"epoch": 0.00985, |
|
"grad_norm": 1.390625, |
|
"learning_rate": 4.9756281407035176e-05, |
|
"loss": 7.5904, |
|
"step": 1970 |
|
}, |
|
{ |
|
"epoch": 0.0099, |
|
"grad_norm": 2.34375, |
|
"learning_rate": 4.975376884422111e-05, |
|
"loss": 7.5794, |
|
"step": 1980 |
|
}, |
|
{ |
|
"epoch": 0.00995, |
|
"grad_norm": 71.0, |
|
"learning_rate": 4.975125628140704e-05, |
|
"loss": 7.5853, |
|
"step": 1990 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 26.75, |
|
"learning_rate": 4.974874371859297e-05, |
|
"loss": 7.591, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"eval_loss": 7.591722011566162, |
|
"eval_runtime": 87.3606, |
|
"eval_samples_per_second": 28.617, |
|
"eval_steps_per_second": 0.458, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.01005, |
|
"grad_norm": 4.40625, |
|
"learning_rate": 4.97462311557789e-05, |
|
"loss": 7.5937, |
|
"step": 2010 |
|
}, |
|
{ |
|
"epoch": 0.0101, |
|
"grad_norm": 42.0, |
|
"learning_rate": 4.974371859296483e-05, |
|
"loss": 7.6135, |
|
"step": 2020 |
|
}, |
|
{ |
|
"epoch": 0.01015, |
|
"grad_norm": 2.25, |
|
"learning_rate": 4.9741206030150755e-05, |
|
"loss": 7.6028, |
|
"step": 2030 |
|
}, |
|
{ |
|
"epoch": 0.0102, |
|
"grad_norm": 4.4375, |
|
"learning_rate": 4.9738693467336686e-05, |
|
"loss": 7.593, |
|
"step": 2040 |
|
}, |
|
{ |
|
"epoch": 0.01025, |
|
"grad_norm": 3.921875, |
|
"learning_rate": 4.973618090452262e-05, |
|
"loss": 7.6155, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 0.0103, |
|
"grad_norm": 9.8125, |
|
"learning_rate": 4.973366834170855e-05, |
|
"loss": 7.5792, |
|
"step": 2060 |
|
}, |
|
{ |
|
"epoch": 0.01035, |
|
"grad_norm": 1.78125, |
|
"learning_rate": 4.973115577889447e-05, |
|
"loss": 7.6247, |
|
"step": 2070 |
|
}, |
|
{ |
|
"epoch": 0.0104, |
|
"grad_norm": 2.6875, |
|
"learning_rate": 4.97286432160804e-05, |
|
"loss": 7.5797, |
|
"step": 2080 |
|
}, |
|
{ |
|
"epoch": 0.01045, |
|
"grad_norm": 33.75, |
|
"learning_rate": 4.9726130653266334e-05, |
|
"loss": 7.6301, |
|
"step": 2090 |
|
}, |
|
{ |
|
"epoch": 0.0105, |
|
"grad_norm": 4.40625, |
|
"learning_rate": 4.9723618090452265e-05, |
|
"loss": 7.5729, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 0.01055, |
|
"grad_norm": 16.75, |
|
"learning_rate": 4.9721105527638196e-05, |
|
"loss": 7.5846, |
|
"step": 2110 |
|
}, |
|
{ |
|
"epoch": 0.0106, |
|
"grad_norm": 78.0, |
|
"learning_rate": 4.971859296482412e-05, |
|
"loss": 7.5739, |
|
"step": 2120 |
|
}, |
|
{ |
|
"epoch": 0.01065, |
|
"grad_norm": 3.125, |
|
"learning_rate": 4.971608040201005e-05, |
|
"loss": 7.5262, |
|
"step": 2130 |
|
}, |
|
{ |
|
"epoch": 0.0107, |
|
"grad_norm": 1.84375, |
|
"learning_rate": 4.971356783919598e-05, |
|
"loss": 7.614, |
|
"step": 2140 |
|
}, |
|
{ |
|
"epoch": 0.01075, |
|
"grad_norm": 60.5, |
|
"learning_rate": 4.9711055276381914e-05, |
|
"loss": 7.5534, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 0.0108, |
|
"grad_norm": 2.03125, |
|
"learning_rate": 4.970854271356784e-05, |
|
"loss": 7.5945, |
|
"step": 2160 |
|
}, |
|
{ |
|
"epoch": 0.01085, |
|
"grad_norm": 180.0, |
|
"learning_rate": 4.970603015075377e-05, |
|
"loss": 7.5861, |
|
"step": 2170 |
|
}, |
|
{ |
|
"epoch": 0.0109, |
|
"grad_norm": 66.0, |
|
"learning_rate": 4.97035175879397e-05, |
|
"loss": 7.5985, |
|
"step": 2180 |
|
}, |
|
{ |
|
"epoch": 0.01095, |
|
"grad_norm": 2.875, |
|
"learning_rate": 4.970100502512563e-05, |
|
"loss": 7.601, |
|
"step": 2190 |
|
}, |
|
{ |
|
"epoch": 0.011, |
|
"grad_norm": 4.0625, |
|
"learning_rate": 4.9698492462311555e-05, |
|
"loss": 7.5665, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 0.01105, |
|
"grad_norm": 10.1875, |
|
"learning_rate": 4.969597989949749e-05, |
|
"loss": 7.5843, |
|
"step": 2210 |
|
}, |
|
{ |
|
"epoch": 0.0111, |
|
"grad_norm": 26.5, |
|
"learning_rate": 4.969346733668342e-05, |
|
"loss": 7.5829, |
|
"step": 2220 |
|
}, |
|
{ |
|
"epoch": 0.01115, |
|
"grad_norm": 4.875, |
|
"learning_rate": 4.969095477386935e-05, |
|
"loss": 7.6214, |
|
"step": 2230 |
|
}, |
|
{ |
|
"epoch": 0.0112, |
|
"grad_norm": 27.375, |
|
"learning_rate": 4.968844221105528e-05, |
|
"loss": 7.5561, |
|
"step": 2240 |
|
}, |
|
{ |
|
"epoch": 0.01125, |
|
"grad_norm": 58.75, |
|
"learning_rate": 4.968592964824121e-05, |
|
"loss": 7.5848, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 0.0113, |
|
"grad_norm": 29.5, |
|
"learning_rate": 4.9683417085427134e-05, |
|
"loss": 7.5941, |
|
"step": 2260 |
|
}, |
|
{ |
|
"epoch": 0.01135, |
|
"grad_norm": 1.40625, |
|
"learning_rate": 4.968090452261307e-05, |
|
"loss": 7.5891, |
|
"step": 2270 |
|
}, |
|
{ |
|
"epoch": 0.0114, |
|
"grad_norm": 25.0, |
|
"learning_rate": 4.9678391959798996e-05, |
|
"loss": 7.5454, |
|
"step": 2280 |
|
}, |
|
{ |
|
"epoch": 0.01145, |
|
"grad_norm": 1.265625, |
|
"learning_rate": 4.967587939698493e-05, |
|
"loss": 7.5377, |
|
"step": 2290 |
|
}, |
|
{ |
|
"epoch": 0.0115, |
|
"grad_norm": 18.625, |
|
"learning_rate": 4.967336683417086e-05, |
|
"loss": 7.5584, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 0.01155, |
|
"grad_norm": 25.375, |
|
"learning_rate": 4.967085427135679e-05, |
|
"loss": 7.5637, |
|
"step": 2310 |
|
}, |
|
{ |
|
"epoch": 0.0116, |
|
"grad_norm": 34.75, |
|
"learning_rate": 4.9668341708542714e-05, |
|
"loss": 7.5909, |
|
"step": 2320 |
|
}, |
|
{ |
|
"epoch": 0.01165, |
|
"grad_norm": 2.171875, |
|
"learning_rate": 4.9665829145728645e-05, |
|
"loss": 7.6102, |
|
"step": 2330 |
|
}, |
|
{ |
|
"epoch": 0.0117, |
|
"grad_norm": 10.4375, |
|
"learning_rate": 4.9663316582914576e-05, |
|
"loss": 7.5496, |
|
"step": 2340 |
|
}, |
|
{ |
|
"epoch": 0.01175, |
|
"grad_norm": 1.2890625, |
|
"learning_rate": 4.966080402010051e-05, |
|
"loss": 7.5872, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 0.0118, |
|
"grad_norm": 7.0, |
|
"learning_rate": 4.965829145728643e-05, |
|
"loss": 7.5669, |
|
"step": 2360 |
|
}, |
|
{ |
|
"epoch": 0.01185, |
|
"grad_norm": 1.71875, |
|
"learning_rate": 4.965577889447237e-05, |
|
"loss": 7.5733, |
|
"step": 2370 |
|
}, |
|
{ |
|
"epoch": 0.0119, |
|
"grad_norm": 3.453125, |
|
"learning_rate": 4.965326633165829e-05, |
|
"loss": 7.5912, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 0.01195, |
|
"grad_norm": 4.21875, |
|
"learning_rate": 4.9650753768844224e-05, |
|
"loss": 7.5514, |
|
"step": 2390 |
|
}, |
|
{ |
|
"epoch": 0.012, |
|
"grad_norm": 2.5, |
|
"learning_rate": 4.9648241206030155e-05, |
|
"loss": 7.5824, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 0.01205, |
|
"grad_norm": 20.375, |
|
"learning_rate": 4.9645728643216086e-05, |
|
"loss": 7.5583, |
|
"step": 2410 |
|
}, |
|
{ |
|
"epoch": 0.0121, |
|
"grad_norm": 9.0625, |
|
"learning_rate": 4.964321608040201e-05, |
|
"loss": 7.5876, |
|
"step": 2420 |
|
}, |
|
{ |
|
"epoch": 0.01215, |
|
"grad_norm": 14.875, |
|
"learning_rate": 4.964070351758794e-05, |
|
"loss": 7.5762, |
|
"step": 2430 |
|
}, |
|
{ |
|
"epoch": 0.0122, |
|
"grad_norm": 21.375, |
|
"learning_rate": 4.963819095477387e-05, |
|
"loss": 7.6275, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 0.01225, |
|
"grad_norm": 19.625, |
|
"learning_rate": 4.9635678391959803e-05, |
|
"loss": 7.5706, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 0.0123, |
|
"grad_norm": 3.4375, |
|
"learning_rate": 4.9633165829145734e-05, |
|
"loss": 7.5645, |
|
"step": 2460 |
|
}, |
|
{ |
|
"epoch": 0.01235, |
|
"grad_norm": 7.53125, |
|
"learning_rate": 4.963065326633166e-05, |
|
"loss": 7.5835, |
|
"step": 2470 |
|
}, |
|
{ |
|
"epoch": 0.0124, |
|
"grad_norm": 8.75, |
|
"learning_rate": 4.962814070351759e-05, |
|
"loss": 7.5419, |
|
"step": 2480 |
|
}, |
|
{ |
|
"epoch": 0.01245, |
|
"grad_norm": 2.625, |
|
"learning_rate": 4.9625628140703514e-05, |
|
"loss": 7.5884, |
|
"step": 2490 |
|
}, |
|
{ |
|
"epoch": 0.0125, |
|
"grad_norm": 2.4375, |
|
"learning_rate": 4.962311557788945e-05, |
|
"loss": 7.5708, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.01255, |
|
"grad_norm": 3.5, |
|
"learning_rate": 4.9620603015075376e-05, |
|
"loss": 7.5626, |
|
"step": 2510 |
|
}, |
|
{ |
|
"epoch": 0.0126, |
|
"grad_norm": 23.375, |
|
"learning_rate": 4.961809045226131e-05, |
|
"loss": 7.5763, |
|
"step": 2520 |
|
}, |
|
{ |
|
"epoch": 0.01265, |
|
"grad_norm": 3.765625, |
|
"learning_rate": 4.961557788944724e-05, |
|
"loss": 7.556, |
|
"step": 2530 |
|
}, |
|
{ |
|
"epoch": 0.0127, |
|
"grad_norm": 1.875, |
|
"learning_rate": 4.961306532663317e-05, |
|
"loss": 7.5764, |
|
"step": 2540 |
|
}, |
|
{ |
|
"epoch": 0.01275, |
|
"grad_norm": 21.5, |
|
"learning_rate": 4.961055276381909e-05, |
|
"loss": 7.5888, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 0.0128, |
|
"grad_norm": 3.0625, |
|
"learning_rate": 4.960804020100503e-05, |
|
"loss": 7.5852, |
|
"step": 2560 |
|
}, |
|
{ |
|
"epoch": 0.01285, |
|
"grad_norm": 10.625, |
|
"learning_rate": 4.9605527638190955e-05, |
|
"loss": 7.5849, |
|
"step": 2570 |
|
}, |
|
{ |
|
"epoch": 0.0129, |
|
"grad_norm": 4.1875, |
|
"learning_rate": 4.9603015075376886e-05, |
|
"loss": 7.5777, |
|
"step": 2580 |
|
}, |
|
{ |
|
"epoch": 0.01295, |
|
"grad_norm": 16.75, |
|
"learning_rate": 4.960050251256282e-05, |
|
"loss": 7.5524, |
|
"step": 2590 |
|
}, |
|
{ |
|
"epoch": 0.013, |
|
"grad_norm": 15.25, |
|
"learning_rate": 4.959798994974875e-05, |
|
"loss": 7.5561, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 0.01305, |
|
"grad_norm": 5.40625, |
|
"learning_rate": 4.959547738693467e-05, |
|
"loss": 7.5688, |
|
"step": 2610 |
|
}, |
|
{ |
|
"epoch": 0.0131, |
|
"grad_norm": 1.515625, |
|
"learning_rate": 4.959296482412061e-05, |
|
"loss": 7.6088, |
|
"step": 2620 |
|
}, |
|
{ |
|
"epoch": 0.01315, |
|
"grad_norm": 5.90625, |
|
"learning_rate": 4.9590452261306535e-05, |
|
"loss": 7.5727, |
|
"step": 2630 |
|
}, |
|
{ |
|
"epoch": 0.0132, |
|
"grad_norm": 1.421875, |
|
"learning_rate": 4.9587939698492466e-05, |
|
"loss": 7.5454, |
|
"step": 2640 |
|
}, |
|
{ |
|
"epoch": 0.01325, |
|
"grad_norm": 1.53125, |
|
"learning_rate": 4.958542713567839e-05, |
|
"loss": 7.5715, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 0.0133, |
|
"grad_norm": 13.0625, |
|
"learning_rate": 4.958291457286433e-05, |
|
"loss": 7.5807, |
|
"step": 2660 |
|
}, |
|
{ |
|
"epoch": 0.01335, |
|
"grad_norm": 1.65625, |
|
"learning_rate": 4.958040201005025e-05, |
|
"loss": 7.5653, |
|
"step": 2670 |
|
}, |
|
{ |
|
"epoch": 0.0134, |
|
"grad_norm": 33.5, |
|
"learning_rate": 4.957788944723618e-05, |
|
"loss": 7.5593, |
|
"step": 2680 |
|
}, |
|
{ |
|
"epoch": 0.01345, |
|
"grad_norm": 2.546875, |
|
"learning_rate": 4.9575376884422114e-05, |
|
"loss": 7.5758, |
|
"step": 2690 |
|
}, |
|
{ |
|
"epoch": 0.0135, |
|
"grad_norm": 1.8046875, |
|
"learning_rate": 4.9572864321608045e-05, |
|
"loss": 7.5581, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 0.01355, |
|
"grad_norm": 20.625, |
|
"learning_rate": 4.957035175879397e-05, |
|
"loss": 7.5494, |
|
"step": 2710 |
|
}, |
|
{ |
|
"epoch": 0.0136, |
|
"grad_norm": 24.25, |
|
"learning_rate": 4.956783919597991e-05, |
|
"loss": 7.5785, |
|
"step": 2720 |
|
}, |
|
{ |
|
"epoch": 0.01365, |
|
"grad_norm": 12.0, |
|
"learning_rate": 4.956532663316583e-05, |
|
"loss": 7.5553, |
|
"step": 2730 |
|
}, |
|
{ |
|
"epoch": 0.0137, |
|
"grad_norm": 1.5078125, |
|
"learning_rate": 4.956281407035176e-05, |
|
"loss": 7.5827, |
|
"step": 2740 |
|
}, |
|
{ |
|
"epoch": 0.01375, |
|
"grad_norm": 32.0, |
|
"learning_rate": 4.956030150753769e-05, |
|
"loss": 7.5814, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 0.0138, |
|
"grad_norm": 110.5, |
|
"learning_rate": 4.9557788944723624e-05, |
|
"loss": 7.5819, |
|
"step": 2760 |
|
}, |
|
{ |
|
"epoch": 0.01385, |
|
"grad_norm": 7.625, |
|
"learning_rate": 4.955527638190955e-05, |
|
"loss": 7.5767, |
|
"step": 2770 |
|
}, |
|
{ |
|
"epoch": 0.0139, |
|
"grad_norm": 54.25, |
|
"learning_rate": 4.955276381909548e-05, |
|
"loss": 7.5621, |
|
"step": 2780 |
|
}, |
|
{ |
|
"epoch": 0.01395, |
|
"grad_norm": 7.8125, |
|
"learning_rate": 4.955025125628141e-05, |
|
"loss": 7.5634, |
|
"step": 2790 |
|
}, |
|
{ |
|
"epoch": 0.014, |
|
"grad_norm": 1.96875, |
|
"learning_rate": 4.954773869346734e-05, |
|
"loss": 7.5808, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 0.01405, |
|
"grad_norm": 8.9375, |
|
"learning_rate": 4.9545226130653266e-05, |
|
"loss": 7.6055, |
|
"step": 2810 |
|
}, |
|
{ |
|
"epoch": 0.0141, |
|
"grad_norm": 1.2890625, |
|
"learning_rate": 4.95427135678392e-05, |
|
"loss": 7.5899, |
|
"step": 2820 |
|
}, |
|
{ |
|
"epoch": 0.01415, |
|
"grad_norm": 4.875, |
|
"learning_rate": 4.954020100502513e-05, |
|
"loss": 7.5589, |
|
"step": 2830 |
|
}, |
|
{ |
|
"epoch": 0.0142, |
|
"grad_norm": 16.375, |
|
"learning_rate": 4.953768844221105e-05, |
|
"loss": 7.6175, |
|
"step": 2840 |
|
}, |
|
{ |
|
"epoch": 0.01425, |
|
"grad_norm": 1.7734375, |
|
"learning_rate": 4.953517587939699e-05, |
|
"loss": 7.577, |
|
"step": 2850 |
|
}, |
|
{ |
|
"epoch": 0.0143, |
|
"grad_norm": 2.515625, |
|
"learning_rate": 4.9532663316582914e-05, |
|
"loss": 7.5966, |
|
"step": 2860 |
|
}, |
|
{ |
|
"epoch": 0.01435, |
|
"grad_norm": 39.0, |
|
"learning_rate": 4.9530150753768845e-05, |
|
"loss": 7.5393, |
|
"step": 2870 |
|
}, |
|
{ |
|
"epoch": 0.0144, |
|
"grad_norm": 1.171875, |
|
"learning_rate": 4.9527638190954776e-05, |
|
"loss": 7.5757, |
|
"step": 2880 |
|
}, |
|
{ |
|
"epoch": 0.01445, |
|
"grad_norm": 1.8203125, |
|
"learning_rate": 4.952512562814071e-05, |
|
"loss": 7.5779, |
|
"step": 2890 |
|
}, |
|
{ |
|
"epoch": 0.0145, |
|
"grad_norm": 1.7578125, |
|
"learning_rate": 4.952261306532663e-05, |
|
"loss": 7.6028, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 0.01455, |
|
"grad_norm": 15.8125, |
|
"learning_rate": 4.952010050251257e-05, |
|
"loss": 7.53, |
|
"step": 2910 |
|
}, |
|
{ |
|
"epoch": 0.0146, |
|
"grad_norm": 1.8203125, |
|
"learning_rate": 4.9517587939698493e-05, |
|
"loss": 7.5656, |
|
"step": 2920 |
|
}, |
|
{ |
|
"epoch": 0.01465, |
|
"grad_norm": 197.0, |
|
"learning_rate": 4.9515075376884424e-05, |
|
"loss": 7.559, |
|
"step": 2930 |
|
}, |
|
{ |
|
"epoch": 0.0147, |
|
"grad_norm": 2.53125, |
|
"learning_rate": 4.9512562814070355e-05, |
|
"loss": 7.5449, |
|
"step": 2940 |
|
}, |
|
{ |
|
"epoch": 0.01475, |
|
"grad_norm": 1.703125, |
|
"learning_rate": 4.9510050251256287e-05, |
|
"loss": 7.5569, |
|
"step": 2950 |
|
}, |
|
{ |
|
"epoch": 0.0148, |
|
"grad_norm": 1.5703125, |
|
"learning_rate": 4.950753768844221e-05, |
|
"loss": 7.5683, |
|
"step": 2960 |
|
}, |
|
{ |
|
"epoch": 0.01485, |
|
"grad_norm": 65.0, |
|
"learning_rate": 4.950502512562814e-05, |
|
"loss": 7.566, |
|
"step": 2970 |
|
}, |
|
{ |
|
"epoch": 0.0149, |
|
"grad_norm": 1.5, |
|
"learning_rate": 4.950251256281407e-05, |
|
"loss": 7.5829, |
|
"step": 2980 |
|
}, |
|
{ |
|
"epoch": 0.01495, |
|
"grad_norm": 44.25, |
|
"learning_rate": 4.9500000000000004e-05, |
|
"loss": 7.5524, |
|
"step": 2990 |
|
}, |
|
{ |
|
"epoch": 0.015, |
|
"grad_norm": 73.5, |
|
"learning_rate": 4.949748743718593e-05, |
|
"loss": 7.5774, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.01505, |
|
"grad_norm": 76.0, |
|
"learning_rate": 4.9494974874371866e-05, |
|
"loss": 7.5661, |
|
"step": 3010 |
|
}, |
|
{ |
|
"epoch": 0.0151, |
|
"grad_norm": 1.9375, |
|
"learning_rate": 4.949246231155779e-05, |
|
"loss": 7.5823, |
|
"step": 3020 |
|
}, |
|
{ |
|
"epoch": 0.01515, |
|
"grad_norm": 6.8125, |
|
"learning_rate": 4.948994974874372e-05, |
|
"loss": 7.5367, |
|
"step": 3030 |
|
}, |
|
{ |
|
"epoch": 0.0152, |
|
"grad_norm": 13.5625, |
|
"learning_rate": 4.948743718592965e-05, |
|
"loss": 7.5738, |
|
"step": 3040 |
|
}, |
|
{ |
|
"epoch": 0.01525, |
|
"grad_norm": 1.5, |
|
"learning_rate": 4.948492462311558e-05, |
|
"loss": 7.5641, |
|
"step": 3050 |
|
}, |
|
{ |
|
"epoch": 0.0153, |
|
"grad_norm": 6.28125, |
|
"learning_rate": 4.948241206030151e-05, |
|
"loss": 7.5391, |
|
"step": 3060 |
|
}, |
|
{ |
|
"epoch": 0.01535, |
|
"grad_norm": 2.8125, |
|
"learning_rate": 4.9479899497487445e-05, |
|
"loss": 7.5468, |
|
"step": 3070 |
|
}, |
|
{ |
|
"epoch": 0.0154, |
|
"grad_norm": 1.6015625, |
|
"learning_rate": 4.947738693467337e-05, |
|
"loss": 7.5947, |
|
"step": 3080 |
|
}, |
|
{ |
|
"epoch": 0.01545, |
|
"grad_norm": 82.5, |
|
"learning_rate": 4.94748743718593e-05, |
|
"loss": 7.5516, |
|
"step": 3090 |
|
}, |
|
{ |
|
"epoch": 0.0155, |
|
"grad_norm": 2.546875, |
|
"learning_rate": 4.947236180904523e-05, |
|
"loss": 7.5146, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 0.01555, |
|
"grad_norm": 1.9140625, |
|
"learning_rate": 4.946984924623116e-05, |
|
"loss": 7.5449, |
|
"step": 3110 |
|
}, |
|
{ |
|
"epoch": 0.0156, |
|
"grad_norm": 2.5625, |
|
"learning_rate": 4.946733668341709e-05, |
|
"loss": 7.6047, |
|
"step": 3120 |
|
}, |
|
{ |
|
"epoch": 0.01565, |
|
"grad_norm": 1.84375, |
|
"learning_rate": 4.946482412060302e-05, |
|
"loss": 7.5818, |
|
"step": 3130 |
|
}, |
|
{ |
|
"epoch": 0.0157, |
|
"grad_norm": 10.9375, |
|
"learning_rate": 4.946231155778895e-05, |
|
"loss": 7.5757, |
|
"step": 3140 |
|
}, |
|
{ |
|
"epoch": 0.01575, |
|
"grad_norm": 15.375, |
|
"learning_rate": 4.945979899497487e-05, |
|
"loss": 7.5958, |
|
"step": 3150 |
|
}, |
|
{ |
|
"epoch": 0.0158, |
|
"grad_norm": 1.796875, |
|
"learning_rate": 4.9457286432160804e-05, |
|
"loss": 7.537, |
|
"step": 3160 |
|
}, |
|
{ |
|
"epoch": 0.01585, |
|
"grad_norm": 1.5859375, |
|
"learning_rate": 4.9454773869346735e-05, |
|
"loss": 7.5929, |
|
"step": 3170 |
|
}, |
|
{ |
|
"epoch": 0.0159, |
|
"grad_norm": 15.8125, |
|
"learning_rate": 4.9452261306532666e-05, |
|
"loss": 7.5747, |
|
"step": 3180 |
|
}, |
|
{ |
|
"epoch": 0.01595, |
|
"grad_norm": 17.375, |
|
"learning_rate": 4.944974874371859e-05, |
|
"loss": 7.5971, |
|
"step": 3190 |
|
}, |
|
{ |
|
"epoch": 0.016, |
|
"grad_norm": 10.0, |
|
"learning_rate": 4.944723618090453e-05, |
|
"loss": 7.5508, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 0.01605, |
|
"grad_norm": 41.5, |
|
"learning_rate": 4.944472361809045e-05, |
|
"loss": 7.5781, |
|
"step": 3210 |
|
}, |
|
{ |
|
"epoch": 0.0161, |
|
"grad_norm": 39.25, |
|
"learning_rate": 4.944221105527638e-05, |
|
"loss": 7.5582, |
|
"step": 3220 |
|
}, |
|
{ |
|
"epoch": 0.01615, |
|
"grad_norm": 19.75, |
|
"learning_rate": 4.9439698492462314e-05, |
|
"loss": 7.5801, |
|
"step": 3230 |
|
}, |
|
{ |
|
"epoch": 0.0162, |
|
"grad_norm": 1.328125, |
|
"learning_rate": 4.9437185929648245e-05, |
|
"loss": 7.6125, |
|
"step": 3240 |
|
}, |
|
{ |
|
"epoch": 0.01625, |
|
"grad_norm": 1.5234375, |
|
"learning_rate": 4.943467336683417e-05, |
|
"loss": 7.5894, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 0.0163, |
|
"grad_norm": 7.34375, |
|
"learning_rate": 4.943216080402011e-05, |
|
"loss": 7.5851, |
|
"step": 3260 |
|
}, |
|
{ |
|
"epoch": 0.01635, |
|
"grad_norm": 7.59375, |
|
"learning_rate": 4.942964824120603e-05, |
|
"loss": 7.5881, |
|
"step": 3270 |
|
}, |
|
{ |
|
"epoch": 0.0164, |
|
"grad_norm": 4.34375, |
|
"learning_rate": 4.942713567839196e-05, |
|
"loss": 7.6143, |
|
"step": 3280 |
|
}, |
|
{ |
|
"epoch": 0.01645, |
|
"grad_norm": 40.25, |
|
"learning_rate": 4.942462311557789e-05, |
|
"loss": 7.5958, |
|
"step": 3290 |
|
}, |
|
{ |
|
"epoch": 0.0165, |
|
"grad_norm": 5.96875, |
|
"learning_rate": 4.9422110552763825e-05, |
|
"loss": 7.5495, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 0.01655, |
|
"grad_norm": 1.671875, |
|
"learning_rate": 4.941959798994975e-05, |
|
"loss": 7.5791, |
|
"step": 3310 |
|
}, |
|
{ |
|
"epoch": 0.0166, |
|
"grad_norm": 10.25, |
|
"learning_rate": 4.941708542713568e-05, |
|
"loss": 7.579, |
|
"step": 3320 |
|
}, |
|
{ |
|
"epoch": 0.01665, |
|
"grad_norm": 2.75, |
|
"learning_rate": 4.941457286432161e-05, |
|
"loss": 7.5623, |
|
"step": 3330 |
|
}, |
|
{ |
|
"epoch": 0.0167, |
|
"grad_norm": 1.5, |
|
"learning_rate": 4.941206030150754e-05, |
|
"loss": 7.5906, |
|
"step": 3340 |
|
}, |
|
{ |
|
"epoch": 0.01675, |
|
"grad_norm": 38.75, |
|
"learning_rate": 4.9409547738693466e-05, |
|
"loss": 7.5825, |
|
"step": 3350 |
|
}, |
|
{ |
|
"epoch": 0.0168, |
|
"grad_norm": 5.28125, |
|
"learning_rate": 4.9407035175879404e-05, |
|
"loss": 7.5844, |
|
"step": 3360 |
|
}, |
|
{ |
|
"epoch": 0.01685, |
|
"grad_norm": 1.5703125, |
|
"learning_rate": 4.940452261306533e-05, |
|
"loss": 7.5995, |
|
"step": 3370 |
|
}, |
|
{ |
|
"epoch": 0.0169, |
|
"grad_norm": 5.84375, |
|
"learning_rate": 4.940201005025126e-05, |
|
"loss": 7.5426, |
|
"step": 3380 |
|
}, |
|
{ |
|
"epoch": 0.01695, |
|
"grad_norm": 1.6796875, |
|
"learning_rate": 4.939949748743719e-05, |
|
"loss": 7.5876, |
|
"step": 3390 |
|
}, |
|
{ |
|
"epoch": 0.017, |
|
"grad_norm": 1.25, |
|
"learning_rate": 4.939698492462312e-05, |
|
"loss": 7.564, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 0.01705, |
|
"grad_norm": 22.5, |
|
"learning_rate": 4.9394472361809046e-05, |
|
"loss": 7.567, |
|
"step": 3410 |
|
}, |
|
{ |
|
"epoch": 0.0171, |
|
"grad_norm": 11.9375, |
|
"learning_rate": 4.9391959798994977e-05, |
|
"loss": 7.5689, |
|
"step": 3420 |
|
}, |
|
{ |
|
"epoch": 0.01715, |
|
"grad_norm": 2.234375, |
|
"learning_rate": 4.938944723618091e-05, |
|
"loss": 7.54, |
|
"step": 3430 |
|
}, |
|
{ |
|
"epoch": 0.0172, |
|
"grad_norm": 24.375, |
|
"learning_rate": 4.938693467336684e-05, |
|
"loss": 7.5887, |
|
"step": 3440 |
|
}, |
|
{ |
|
"epoch": 0.01725, |
|
"grad_norm": 57.25, |
|
"learning_rate": 4.938442211055276e-05, |
|
"loss": 7.5857, |
|
"step": 3450 |
|
}, |
|
{ |
|
"epoch": 0.0173, |
|
"grad_norm": 44.0, |
|
"learning_rate": 4.93819095477387e-05, |
|
"loss": 7.5939, |
|
"step": 3460 |
|
}, |
|
{ |
|
"epoch": 0.01735, |
|
"grad_norm": 12.4375, |
|
"learning_rate": 4.9379396984924625e-05, |
|
"loss": 7.5695, |
|
"step": 3470 |
|
}, |
|
{ |
|
"epoch": 0.0174, |
|
"grad_norm": 5.71875, |
|
"learning_rate": 4.9376884422110556e-05, |
|
"loss": 7.5632, |
|
"step": 3480 |
|
}, |
|
{ |
|
"epoch": 0.01745, |
|
"grad_norm": 2.1875, |
|
"learning_rate": 4.937437185929649e-05, |
|
"loss": 7.5836, |
|
"step": 3490 |
|
}, |
|
{ |
|
"epoch": 0.0175, |
|
"grad_norm": 2.421875, |
|
"learning_rate": 4.937185929648241e-05, |
|
"loss": 7.5457, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 0.01755, |
|
"grad_norm": 6.15625, |
|
"learning_rate": 4.936934673366834e-05, |
|
"loss": 7.5696, |
|
"step": 3510 |
|
}, |
|
{ |
|
"epoch": 0.0176, |
|
"grad_norm": 3.671875, |
|
"learning_rate": 4.936683417085427e-05, |
|
"loss": 7.5662, |
|
"step": 3520 |
|
}, |
|
{ |
|
"epoch": 0.01765, |
|
"grad_norm": 1.84375, |
|
"learning_rate": 4.9364321608040204e-05, |
|
"loss": 7.5914, |
|
"step": 3530 |
|
}, |
|
{ |
|
"epoch": 0.0177, |
|
"grad_norm": 4.3125, |
|
"learning_rate": 4.936180904522613e-05, |
|
"loss": 7.5754, |
|
"step": 3540 |
|
}, |
|
{ |
|
"epoch": 0.01775, |
|
"grad_norm": 8.0, |
|
"learning_rate": 4.9359296482412066e-05, |
|
"loss": 7.614, |
|
"step": 3550 |
|
}, |
|
{ |
|
"epoch": 0.0178, |
|
"grad_norm": 1.890625, |
|
"learning_rate": 4.935678391959799e-05, |
|
"loss": 7.5519, |
|
"step": 3560 |
|
}, |
|
{ |
|
"epoch": 0.01785, |
|
"grad_norm": 1.625, |
|
"learning_rate": 4.935427135678392e-05, |
|
"loss": 7.588, |
|
"step": 3570 |
|
}, |
|
{ |
|
"epoch": 0.0179, |
|
"grad_norm": 45.5, |
|
"learning_rate": 4.9351758793969846e-05, |
|
"loss": 7.5629, |
|
"step": 3580 |
|
}, |
|
{ |
|
"epoch": 0.01795, |
|
"grad_norm": 1.8046875, |
|
"learning_rate": 4.9349246231155784e-05, |
|
"loss": 7.5556, |
|
"step": 3590 |
|
}, |
|
{ |
|
"epoch": 0.018, |
|
"grad_norm": 1.203125, |
|
"learning_rate": 4.934673366834171e-05, |
|
"loss": 7.5878, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 0.01805, |
|
"grad_norm": 7.34375, |
|
"learning_rate": 4.934422110552764e-05, |
|
"loss": 7.5983, |
|
"step": 3610 |
|
}, |
|
{ |
|
"epoch": 0.0181, |
|
"grad_norm": 4.28125, |
|
"learning_rate": 4.934170854271357e-05, |
|
"loss": 7.6071, |
|
"step": 3620 |
|
}, |
|
{ |
|
"epoch": 0.01815, |
|
"grad_norm": 1.703125, |
|
"learning_rate": 4.93391959798995e-05, |
|
"loss": 7.5621, |
|
"step": 3630 |
|
}, |
|
{ |
|
"epoch": 0.0182, |
|
"grad_norm": 1.5, |
|
"learning_rate": 4.9336683417085425e-05, |
|
"loss": 7.582, |
|
"step": 3640 |
|
}, |
|
{ |
|
"epoch": 0.01825, |
|
"grad_norm": 2.078125, |
|
"learning_rate": 4.933417085427136e-05, |
|
"loss": 7.5727, |
|
"step": 3650 |
|
}, |
|
{ |
|
"epoch": 0.0183, |
|
"grad_norm": 67.5, |
|
"learning_rate": 4.933165829145729e-05, |
|
"loss": 7.572, |
|
"step": 3660 |
|
}, |
|
{ |
|
"epoch": 0.01835, |
|
"grad_norm": 103.5, |
|
"learning_rate": 4.932914572864322e-05, |
|
"loss": 7.5624, |
|
"step": 3670 |
|
}, |
|
{ |
|
"epoch": 0.0184, |
|
"grad_norm": 1.5703125, |
|
"learning_rate": 4.932663316582915e-05, |
|
"loss": 7.5899, |
|
"step": 3680 |
|
}, |
|
{ |
|
"epoch": 0.01845, |
|
"grad_norm": 1.6484375, |
|
"learning_rate": 4.932412060301508e-05, |
|
"loss": 7.5524, |
|
"step": 3690 |
|
}, |
|
{ |
|
"epoch": 0.0185, |
|
"grad_norm": 48.5, |
|
"learning_rate": 4.9321608040201004e-05, |
|
"loss": 7.6027, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 0.01855, |
|
"grad_norm": 2.09375, |
|
"learning_rate": 4.931909547738694e-05, |
|
"loss": 7.5536, |
|
"step": 3710 |
|
}, |
|
{ |
|
"epoch": 0.0186, |
|
"grad_norm": 1.84375, |
|
"learning_rate": 4.9316582914572866e-05, |
|
"loss": 7.5719, |
|
"step": 3720 |
|
}, |
|
{ |
|
"epoch": 0.01865, |
|
"grad_norm": 1.9765625, |
|
"learning_rate": 4.93140703517588e-05, |
|
"loss": 7.5592, |
|
"step": 3730 |
|
}, |
|
{ |
|
"epoch": 0.0187, |
|
"grad_norm": 3.0625, |
|
"learning_rate": 4.931155778894472e-05, |
|
"loss": 7.6041, |
|
"step": 3740 |
|
}, |
|
{ |
|
"epoch": 0.01875, |
|
"grad_norm": 1.515625, |
|
"learning_rate": 4.930904522613066e-05, |
|
"loss": 7.574, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 0.0188, |
|
"grad_norm": 122.5, |
|
"learning_rate": 4.9306532663316584e-05, |
|
"loss": 7.5402, |
|
"step": 3760 |
|
}, |
|
{ |
|
"epoch": 0.01885, |
|
"grad_norm": 1.6171875, |
|
"learning_rate": 4.9304020100502515e-05, |
|
"loss": 7.5854, |
|
"step": 3770 |
|
}, |
|
{ |
|
"epoch": 0.0189, |
|
"grad_norm": 22.125, |
|
"learning_rate": 4.9301507537688446e-05, |
|
"loss": 7.5882, |
|
"step": 3780 |
|
}, |
|
{ |
|
"epoch": 0.01895, |
|
"grad_norm": 1.28125, |
|
"learning_rate": 4.929899497487438e-05, |
|
"loss": 7.5971, |
|
"step": 3790 |
|
}, |
|
{ |
|
"epoch": 0.019, |
|
"grad_norm": 2.03125, |
|
"learning_rate": 4.92964824120603e-05, |
|
"loss": 7.5531, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 0.01905, |
|
"grad_norm": 1.3125, |
|
"learning_rate": 4.929396984924623e-05, |
|
"loss": 7.5697, |
|
"step": 3810 |
|
}, |
|
{ |
|
"epoch": 0.0191, |
|
"grad_norm": 3.703125, |
|
"learning_rate": 4.929145728643216e-05, |
|
"loss": 7.5603, |
|
"step": 3820 |
|
}, |
|
{ |
|
"epoch": 0.01915, |
|
"grad_norm": 4.3125, |
|
"learning_rate": 4.9288944723618094e-05, |
|
"loss": 7.5705, |
|
"step": 3830 |
|
}, |
|
{ |
|
"epoch": 0.0192, |
|
"grad_norm": 22.125, |
|
"learning_rate": 4.9286432160804025e-05, |
|
"loss": 7.585, |
|
"step": 3840 |
|
}, |
|
{ |
|
"epoch": 0.01925, |
|
"grad_norm": 13.4375, |
|
"learning_rate": 4.928391959798995e-05, |
|
"loss": 7.5508, |
|
"step": 3850 |
|
}, |
|
{ |
|
"epoch": 0.0193, |
|
"grad_norm": 19.125, |
|
"learning_rate": 4.928140703517588e-05, |
|
"loss": 7.6242, |
|
"step": 3860 |
|
}, |
|
{ |
|
"epoch": 0.01935, |
|
"grad_norm": 30.375, |
|
"learning_rate": 4.927889447236181e-05, |
|
"loss": 7.5297, |
|
"step": 3870 |
|
}, |
|
{ |
|
"epoch": 0.0194, |
|
"grad_norm": 26.625, |
|
"learning_rate": 4.927638190954774e-05, |
|
"loss": 7.6138, |
|
"step": 3880 |
|
}, |
|
{ |
|
"epoch": 0.01945, |
|
"grad_norm": 15.9375, |
|
"learning_rate": 4.9273869346733667e-05, |
|
"loss": 7.5671, |
|
"step": 3890 |
|
}, |
|
{ |
|
"epoch": 0.0195, |
|
"grad_norm": 17.75, |
|
"learning_rate": 4.92713567839196e-05, |
|
"loss": 7.6151, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 0.01955, |
|
"grad_norm": 1.234375, |
|
"learning_rate": 4.926884422110553e-05, |
|
"loss": 7.6086, |
|
"step": 3910 |
|
}, |
|
{ |
|
"epoch": 0.0196, |
|
"grad_norm": 1.359375, |
|
"learning_rate": 4.926633165829146e-05, |
|
"loss": 7.5837, |
|
"step": 3920 |
|
}, |
|
{ |
|
"epoch": 0.01965, |
|
"grad_norm": 61.0, |
|
"learning_rate": 4.9263819095477384e-05, |
|
"loss": 7.5868, |
|
"step": 3930 |
|
}, |
|
{ |
|
"epoch": 0.0197, |
|
"grad_norm": 1.984375, |
|
"learning_rate": 4.926130653266332e-05, |
|
"loss": 7.5785, |
|
"step": 3940 |
|
}, |
|
{ |
|
"epoch": 0.01975, |
|
"grad_norm": 37.5, |
|
"learning_rate": 4.9258793969849246e-05, |
|
"loss": 7.615, |
|
"step": 3950 |
|
}, |
|
{ |
|
"epoch": 0.0198, |
|
"grad_norm": 2.03125, |
|
"learning_rate": 4.925628140703518e-05, |
|
"loss": 7.5894, |
|
"step": 3960 |
|
}, |
|
{ |
|
"epoch": 0.01985, |
|
"grad_norm": 1.8203125, |
|
"learning_rate": 4.925376884422111e-05, |
|
"loss": 7.6036, |
|
"step": 3970 |
|
}, |
|
{ |
|
"epoch": 0.0199, |
|
"grad_norm": 1.453125, |
|
"learning_rate": 4.925125628140704e-05, |
|
"loss": 7.5874, |
|
"step": 3980 |
|
}, |
|
{ |
|
"epoch": 0.01995, |
|
"grad_norm": 57.0, |
|
"learning_rate": 4.924874371859296e-05, |
|
"loss": 7.5653, |
|
"step": 3990 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 16.625, |
|
"learning_rate": 4.92462311557789e-05, |
|
"loss": 7.5676, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"eval_loss": 7.588551998138428, |
|
"eval_runtime": 89.0965, |
|
"eval_samples_per_second": 28.059, |
|
"eval_steps_per_second": 0.449, |
|
"step": 4000 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 200000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 9223372036854775807, |
|
"save_steps": 2000, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 1.586817140785152e+17, |
|
"train_batch_size": 64, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|