|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.999921758860809, |
|
"eval_steps": 500, |
|
"global_step": 3195, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"grad_norm": 0.64453125, |
|
"learning_rate": 6.25e-07, |
|
"loss": 1.385, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 0.52734375, |
|
"learning_rate": 1.25e-06, |
|
"loss": 1.4608, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 0.255859375, |
|
"learning_rate": 1.8750000000000003e-06, |
|
"loss": 1.4275, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 0.54296875, |
|
"learning_rate": 2.5e-06, |
|
"loss": 1.4101, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 0.59765625, |
|
"learning_rate": 3.125e-06, |
|
"loss": 1.4565, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 0.236328125, |
|
"learning_rate": 3.7500000000000005e-06, |
|
"loss": 1.4078, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 0.1884765625, |
|
"learning_rate": 4.3750000000000005e-06, |
|
"loss": 1.3982, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 0.294921875, |
|
"learning_rate": 5e-06, |
|
"loss": 1.4799, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 0.138671875, |
|
"learning_rate": 5.625e-06, |
|
"loss": 1.4549, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 1.625, |
|
"learning_rate": 6.25e-06, |
|
"loss": 1.4205, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 0.33984375, |
|
"learning_rate": 6.875e-06, |
|
"loss": 1.4777, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 0.8359375, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 1.4071, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 0.35546875, |
|
"learning_rate": 8.125000000000001e-06, |
|
"loss": 1.3978, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 0.267578125, |
|
"learning_rate": 8.750000000000001e-06, |
|
"loss": 1.3935, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.7578125, |
|
"learning_rate": 9.375000000000001e-06, |
|
"loss": 1.3514, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.279296875, |
|
"learning_rate": 1e-05, |
|
"loss": 1.3446, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 0.1201171875, |
|
"learning_rate": 1.0625e-05, |
|
"loss": 1.2796, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.103515625, |
|
"learning_rate": 1.125e-05, |
|
"loss": 1.3715, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.08740234375, |
|
"learning_rate": 1.1875e-05, |
|
"loss": 1.2741, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 0.1552734375, |
|
"learning_rate": 1.25e-05, |
|
"loss": 1.2679, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.1357421875, |
|
"learning_rate": 1.3125e-05, |
|
"loss": 1.2512, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.1484375, |
|
"learning_rate": 1.375e-05, |
|
"loss": 1.1982, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.0947265625, |
|
"learning_rate": 1.4375e-05, |
|
"loss": 1.2303, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.48828125, |
|
"learning_rate": 1.5000000000000002e-05, |
|
"loss": 1.2074, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.28515625, |
|
"learning_rate": 1.5625e-05, |
|
"loss": 1.2086, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.1884765625, |
|
"learning_rate": 1.6250000000000002e-05, |
|
"loss": 1.1753, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 0.0751953125, |
|
"learning_rate": 1.6875e-05, |
|
"loss": 1.1524, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.07470703125, |
|
"learning_rate": 1.7500000000000002e-05, |
|
"loss": 1.174, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.10400390625, |
|
"learning_rate": 1.8125e-05, |
|
"loss": 1.1385, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 0.068359375, |
|
"learning_rate": 1.8750000000000002e-05, |
|
"loss": 1.1546, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.103515625, |
|
"learning_rate": 1.9375e-05, |
|
"loss": 1.1263, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.0625, |
|
"learning_rate": 2e-05, |
|
"loss": 1.122, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 0.2080078125, |
|
"learning_rate": 1.999940297883134e-05, |
|
"loss": 1.0784, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 0.0634765625, |
|
"learning_rate": 1.9997611986612203e-05, |
|
"loss": 1.0684, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 0.10205078125, |
|
"learning_rate": 1.9994627237194654e-05, |
|
"loss": 1.0772, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 0.1298828125, |
|
"learning_rate": 1.9990449086970404e-05, |
|
"loss": 1.0837, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.255859375, |
|
"learning_rate": 1.998507803482828e-05, |
|
"loss": 1.0269, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.059814453125, |
|
"learning_rate": 1.997851472209465e-05, |
|
"loss": 1.0229, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 0.060791015625, |
|
"learning_rate": 1.9970759932456836e-05, |
|
"loss": 1.0335, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 1.9961814591869558e-05, |
|
"loss": 1.031, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.058349609375, |
|
"learning_rate": 1.9951679768444346e-05, |
|
"loss": 1.0339, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.047119140625, |
|
"learning_rate": 1.9940356672322037e-05, |
|
"loss": 0.9754, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.07373046875, |
|
"learning_rate": 1.992784665552824e-05, |
|
"loss": 0.9955, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.051513671875, |
|
"learning_rate": 1.9914151211811924e-05, |
|
"loss": 1.0095, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.0625, |
|
"learning_rate": 1.9899271976467058e-05, |
|
"loss": 0.9758, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 0.08935546875, |
|
"learning_rate": 1.9883210726137324e-05, |
|
"loss": 0.9989, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.05029296875, |
|
"learning_rate": 1.9865969378604023e-05, |
|
"loss": 0.9932, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.13671875, |
|
"learning_rate": 1.984754999255704e-05, |
|
"loss": 0.9908, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 0.154296875, |
|
"learning_rate": 1.982795476734905e-05, |
|
"loss": 0.9849, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.058837890625, |
|
"learning_rate": 1.9807186042732908e-05, |
|
"loss": 0.9772, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.05908203125, |
|
"learning_rate": 1.9785246298582262e-05, |
|
"loss": 0.9763, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 0.0576171875, |
|
"learning_rate": 1.9762138154595448e-05, |
|
"loss": 0.975, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.057861328125, |
|
"learning_rate": 1.9737864369982695e-05, |
|
"loss": 0.9361, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.06396484375, |
|
"learning_rate": 1.971242784313665e-05, |
|
"loss": 0.9515, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 0.059814453125, |
|
"learning_rate": 1.9685831611286312e-05, |
|
"loss": 0.9349, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.087890625, |
|
"learning_rate": 1.965807885013437e-05, |
|
"loss": 1.0151, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.07080078125, |
|
"learning_rate": 1.9629172873477995e-05, |
|
"loss": 0.9818, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.08349609375, |
|
"learning_rate": 1.9599117132813187e-05, |
|
"loss": 0.9444, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 0.05908203125, |
|
"learning_rate": 1.9567915216922624e-05, |
|
"loss": 0.946, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.05078125, |
|
"learning_rate": 1.9535570851447166e-05, |
|
"loss": 0.9646, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.05078125, |
|
"learning_rate": 1.9502087898440988e-05, |
|
"loss": 1.0191, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 0.0673828125, |
|
"learning_rate": 1.9467470355910438e-05, |
|
"loss": 0.9789, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.058837890625, |
|
"learning_rate": 1.9431722357336657e-05, |
|
"loss": 0.9413, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.0986328125, |
|
"learning_rate": 1.939484817118202e-05, |
|
"loss": 0.974, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.181640625, |
|
"learning_rate": 1.9356852200380466e-05, |
|
"loss": 0.9416, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.04931640625, |
|
"learning_rate": 1.9317738981811776e-05, |
|
"loss": 0.9316, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.060791015625, |
|
"learning_rate": 1.9277513185759847e-05, |
|
"loss": 0.956, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 0.05078125, |
|
"learning_rate": 1.9236179615355026e-05, |
|
"loss": 0.9124, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.10302734375, |
|
"learning_rate": 1.9193743206000618e-05, |
|
"loss": 0.9405, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.048583984375, |
|
"learning_rate": 1.9150209024783564e-05, |
|
"loss": 0.9711, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 0.0732421875, |
|
"learning_rate": 1.9105582269869413e-05, |
|
"loss": 0.9114, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.1103515625, |
|
"learning_rate": 1.9059868269881637e-05, |
|
"loss": 0.9828, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.14453125, |
|
"learning_rate": 1.9013072483265377e-05, |
|
"loss": 0.9121, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.05615234375, |
|
"learning_rate": 1.896520049763568e-05, |
|
"loss": 0.9412, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 0.052490234375, |
|
"learning_rate": 1.8916258029110305e-05, |
|
"loss": 0.9624, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 1.88662509216272e-05, |
|
"loss": 0.926, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.060791015625, |
|
"learning_rate": 1.8815185146246718e-05, |
|
"loss": 0.9164, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.8763066800438638e-05, |
|
"loss": 0.9532, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 0.06494140625, |
|
"learning_rate": 1.87099021073541e-05, |
|
"loss": 0.944, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 0.11181640625, |
|
"learning_rate": 1.8655697415082556e-05, |
|
"loss": 0.9284, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 0.3984375, |
|
"learning_rate": 1.8600459195893737e-05, |
|
"loss": 0.9235, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 0.189453125, |
|
"learning_rate": 1.8544194045464888e-05, |
|
"loss": 0.9442, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.8486908682093175e-05, |
|
"loss": 0.9653, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 0.04736328125, |
|
"learning_rate": 1.842860994589352e-05, |
|
"loss": 0.9122, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 0.271484375, |
|
"learning_rate": 1.8369304797981843e-05, |
|
"loss": 0.9183, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 0.058837890625, |
|
"learning_rate": 1.8309000319643892e-05, |
|
"loss": 0.9078, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 0.04931640625, |
|
"learning_rate": 1.8247703711489684e-05, |
|
"loss": 0.9404, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 0.060546875, |
|
"learning_rate": 1.818542229259376e-05, |
|
"loss": 0.9584, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 0.056640625, |
|
"learning_rate": 1.8122163499621208e-05, |
|
"loss": 0.9051, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 0.053955078125, |
|
"learning_rate": 1.8057934885939734e-05, |
|
"loss": 0.9525, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 0.10791015625, |
|
"learning_rate": 1.7992744120717735e-05, |
|
"loss": 0.9301, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 0.055419921875, |
|
"learning_rate": 1.7926598988008584e-05, |
|
"loss": 0.9319, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 0.051513671875, |
|
"learning_rate": 1.7859507385821163e-05, |
|
"loss": 0.9408, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 0.0771484375, |
|
"learning_rate": 1.7791477325176824e-05, |
|
"loss": 0.9646, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 0.0625, |
|
"learning_rate": 1.7722516929152828e-05, |
|
"loss": 0.9263, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 0.05419921875, |
|
"learning_rate": 1.7652634431912417e-05, |
|
"loss": 0.9605, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 0.05224609375, |
|
"learning_rate": 1.758183817772163e-05, |
|
"loss": 0.9097, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.7510136619952947e-05, |
|
"loss": 0.9178, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 0.0595703125, |
|
"learning_rate": 1.743753832007593e-05, |
|
"loss": 0.8983, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 0.142578125, |
|
"learning_rate": 1.7364051946634953e-05, |
|
"loss": 0.9637, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 0.07275390625, |
|
"learning_rate": 1.7289686274214116e-05, |
|
"loss": 0.9142, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 0.091796875, |
|
"learning_rate": 1.721445018238956e-05, |
|
"loss": 0.9663, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 0.0732421875, |
|
"learning_rate": 1.713835265466917e-05, |
|
"loss": 0.9184, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 0.087890625, |
|
"learning_rate": 1.706140277741994e-05, |
|
"loss": 0.9207, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 0.07568359375, |
|
"learning_rate": 1.6983609738782993e-05, |
|
"loss": 0.917, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 0.427734375, |
|
"learning_rate": 1.6904982827576498e-05, |
|
"loss": 0.9255, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 0.08447265625, |
|
"learning_rate": 1.6825531432186545e-05, |
|
"loss": 0.9683, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"grad_norm": 0.054443359375, |
|
"learning_rate": 1.674526503944611e-05, |
|
"loss": 0.8988, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"grad_norm": 0.1357421875, |
|
"learning_rate": 1.6664193233502314e-05, |
|
"loss": 0.9114, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"grad_norm": 0.0556640625, |
|
"learning_rate": 1.6582325694672032e-05, |
|
"loss": 0.9341, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 0.291015625, |
|
"learning_rate": 1.6499672198285996e-05, |
|
"loss": 0.9284, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 0.076171875, |
|
"learning_rate": 1.6416242613521612e-05, |
|
"loss": 0.9461, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 0.052978515625, |
|
"learning_rate": 1.6332046902224518e-05, |
|
"loss": 0.9011, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 0.05322265625, |
|
"learning_rate": 1.6247095117719106e-05, |
|
"loss": 0.8884, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 0.0498046875, |
|
"learning_rate": 1.616139740360811e-05, |
|
"loss": 0.911, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 0.1767578125, |
|
"learning_rate": 1.607496399256141e-05, |
|
"loss": 0.9139, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 0.0517578125, |
|
"learning_rate": 1.5987805205094225e-05, |
|
"loss": 0.9277, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 0.0908203125, |
|
"learning_rate": 1.5899931448334788e-05, |
|
"loss": 0.8923, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 1.5703125, |
|
"learning_rate": 1.581135321478169e-05, |
|
"loss": 0.9193, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 0.052978515625, |
|
"learning_rate": 1.5722081081051032e-05, |
|
"loss": 0.8975, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 0.0947265625, |
|
"learning_rate": 1.5632125706613534e-05, |
|
"loss": 0.9291, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 0.054931640625, |
|
"learning_rate": 1.554149783252175e-05, |
|
"loss": 0.8887, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 0.04736328125, |
|
"learning_rate": 1.5450208280127543e-05, |
|
"loss": 0.8927, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 0.052978515625, |
|
"learning_rate": 1.5358267949789968e-05, |
|
"loss": 0.9219, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.526568781957371e-05, |
|
"loss": 0.8972, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 0.1259765625, |
|
"learning_rate": 1.5172478943938288e-05, |
|
"loss": 0.9107, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 1.5078652452418063e-05, |
|
"loss": 0.9162, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 0.06494140625, |
|
"learning_rate": 1.4984219548293361e-05, |
|
"loss": 0.9022, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 0.06201171875, |
|
"learning_rate": 1.4889191507252743e-05, |
|
"loss": 0.903, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 0.05322265625, |
|
"learning_rate": 1.479357967604663e-05, |
|
"loss": 0.8873, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 0.07421875, |
|
"learning_rate": 1.469739547113246e-05, |
|
"loss": 0.9103, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 0.055908203125, |
|
"learning_rate": 1.4600650377311523e-05, |
|
"loss": 0.919, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 0.08447265625, |
|
"learning_rate": 1.450335594635761e-05, |
|
"loss": 0.9349, |
|
"step": 1330 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 0.0908203125, |
|
"learning_rate": 1.44055237956377e-05, |
|
"loss": 0.8929, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 0.09423828125, |
|
"learning_rate": 1.4307165606724777e-05, |
|
"loss": 0.9082, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 0.052734375, |
|
"learning_rate": 1.4208293124003028e-05, |
|
"loss": 0.9079, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 0.05517578125, |
|
"learning_rate": 1.4108918153265485e-05, |
|
"loss": 0.9074, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 0.054931640625, |
|
"learning_rate": 1.400905256030438e-05, |
|
"loss": 0.8795, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.3908708269494318e-05, |
|
"loss": 0.9201, |
|
"step": 1390 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 0.1171875, |
|
"learning_rate": 1.3807897262368453e-05, |
|
"loss": 0.9424, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 0.056640625, |
|
"learning_rate": 1.3706631576187842e-05, |
|
"loss": 0.9165, |
|
"step": 1410 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 0.05224609375, |
|
"learning_rate": 1.3604923302504146e-05, |
|
"loss": 0.9164, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 0.2021484375, |
|
"learning_rate": 1.3502784585715853e-05, |
|
"loss": 0.9122, |
|
"step": 1430 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 0.050537109375, |
|
"learning_rate": 1.340022762161817e-05, |
|
"loss": 0.939, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 0.058837890625, |
|
"learning_rate": 1.3297264655946816e-05, |
|
"loss": 0.9257, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 0.052734375, |
|
"learning_rate": 1.31939079829158e-05, |
|
"loss": 0.9182, |
|
"step": 1460 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 0.09130859375, |
|
"learning_rate": 1.3090169943749475e-05, |
|
"loss": 0.9062, |
|
"step": 1470 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 0.2041015625, |
|
"learning_rate": 1.2986062925208913e-05, |
|
"loss": 0.9107, |
|
"step": 1480 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 0.06689453125, |
|
"learning_rate": 1.2881599358112888e-05, |
|
"loss": 0.931, |
|
"step": 1490 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 0.08251953125, |
|
"learning_rate": 1.2776791715853585e-05, |
|
"loss": 0.9473, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 0.10009765625, |
|
"learning_rate": 1.2671652512907213e-05, |
|
"loss": 0.917, |
|
"step": 1510 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 0.07080078125, |
|
"learning_rate": 1.2566194303339738e-05, |
|
"loss": 0.9566, |
|
"step": 1520 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 0.0615234375, |
|
"learning_rate": 1.2460429679307863e-05, |
|
"loss": 0.8839, |
|
"step": 1530 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 0.04931640625, |
|
"learning_rate": 1.2354371269555478e-05, |
|
"loss": 0.9068, |
|
"step": 1540 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 0.0947265625, |
|
"learning_rate": 1.2248031737905732e-05, |
|
"loss": 0.921, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 0.10400390625, |
|
"learning_rate": 1.2141423781748913e-05, |
|
"loss": 0.9177, |
|
"step": 1560 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 0.259765625, |
|
"learning_rate": 1.2034560130526341e-05, |
|
"loss": 0.9099, |
|
"step": 1570 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 0.060546875, |
|
"learning_rate": 1.1927453544210397e-05, |
|
"loss": 0.9166, |
|
"step": 1580 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 0.10498046875, |
|
"learning_rate": 1.182011681178095e-05, |
|
"loss": 0.9022, |
|
"step": 1590 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 0.109375, |
|
"learning_rate": 1.171256274969829e-05, |
|
"loss": 0.8913, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 0.054931640625, |
|
"learning_rate": 1.1604804200372786e-05, |
|
"loss": 0.9166, |
|
"step": 1610 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 0.06591796875, |
|
"learning_rate": 1.1496854030631443e-05, |
|
"loss": 0.9539, |
|
"step": 1620 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 0.0478515625, |
|
"learning_rate": 1.1388725130181566e-05, |
|
"loss": 0.8543, |
|
"step": 1630 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 0.10546875, |
|
"learning_rate": 1.1280430410071652e-05, |
|
"loss": 0.8803, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 0.07470703125, |
|
"learning_rate": 1.1171982801149774e-05, |
|
"loss": 0.8952, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 0.055419921875, |
|
"learning_rate": 1.106339525251958e-05, |
|
"loss": 0.8931, |
|
"step": 1660 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 0.1376953125, |
|
"learning_rate": 1.0954680729994103e-05, |
|
"loss": 0.9036, |
|
"step": 1670 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 0.059814453125, |
|
"learning_rate": 1.08458522145476e-05, |
|
"loss": 0.9455, |
|
"step": 1680 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 0.087890625, |
|
"learning_rate": 1.073692270076557e-05, |
|
"loss": 0.9097, |
|
"step": 1690 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 0.05029296875, |
|
"learning_rate": 1.0627905195293135e-05, |
|
"loss": 0.9132, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 0.049560546875, |
|
"learning_rate": 1.0518812715282001e-05, |
|
"loss": 0.8866, |
|
"step": 1710 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 0.06494140625, |
|
"learning_rate": 1.0409658286836144e-05, |
|
"loss": 0.9258, |
|
"step": 1720 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 0.07177734375, |
|
"learning_rate": 1.0300454943456457e-05, |
|
"loss": 0.8964, |
|
"step": 1730 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 1.0191215724484476e-05, |
|
"loss": 0.9028, |
|
"step": 1740 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 0.0712890625, |
|
"learning_rate": 1.0081953673545432e-05, |
|
"loss": 0.9022, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 0.10888671875, |
|
"learning_rate": 9.9726818369908e-06, |
|
"loss": 0.9018, |
|
"step": 1760 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 0.050537109375, |
|
"learning_rate": 9.863413262340491e-06, |
|
"loss": 0.9033, |
|
"step": 1770 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 0.1884765625, |
|
"learning_rate": 9.754160996724927e-06, |
|
"loss": 0.8838, |
|
"step": 1780 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 0.062255859375, |
|
"learning_rate": 9.644938085327174e-06, |
|
"loss": 0.9077, |
|
"step": 1790 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 0.1064453125, |
|
"learning_rate": 9.535757569825266e-06, |
|
"loss": 0.9251, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 0.33203125, |
|
"learning_rate": 9.426632486834998e-06, |
|
"loss": 0.908, |
|
"step": 1810 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 0.054443359375, |
|
"learning_rate": 9.317575866353293e-06, |
|
"loss": 0.9015, |
|
"step": 1820 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 0.2294921875, |
|
"learning_rate": 9.20860073020234e-06, |
|
"loss": 0.9201, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 0.55078125, |
|
"learning_rate": 9.099720090474779e-06, |
|
"loss": 0.9151, |
|
"step": 1840 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 0.064453125, |
|
"learning_rate": 8.990946947979955e-06, |
|
"loss": 0.9088, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 0.06787109375, |
|
"learning_rate": 8.882294290691609e-06, |
|
"loss": 0.9182, |
|
"step": 1860 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 0.234375, |
|
"learning_rate": 8.773775092197018e-06, |
|
"loss": 0.905, |
|
"step": 1870 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 0.17578125, |
|
"learning_rate": 8.665402310147924e-06, |
|
"loss": 0.8986, |
|
"step": 1880 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 0.06103515625, |
|
"learning_rate": 8.557188884713334e-06, |
|
"loss": 0.9175, |
|
"step": 1890 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 0.0517578125, |
|
"learning_rate": 8.44914773703438e-06, |
|
"loss": 0.926, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 0.1337890625, |
|
"learning_rate": 8.341291767681523e-06, |
|
"loss": 0.9322, |
|
"step": 1910 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 0.177734375, |
|
"learning_rate": 8.233633855114127e-06, |
|
"loss": 0.9372, |
|
"step": 1920 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 0.051025390625, |
|
"learning_rate": 8.126186854142752e-06, |
|
"loss": 0.9285, |
|
"step": 1930 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 0.07177734375, |
|
"learning_rate": 8.018963594394221e-06, |
|
"loss": 0.916, |
|
"step": 1940 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 0.07373046875, |
|
"learning_rate": 7.911976878779696e-06, |
|
"loss": 0.9359, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 0.0712890625, |
|
"learning_rate": 7.805239481965976e-06, |
|
"loss": 0.8861, |
|
"step": 1960 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 0.12890625, |
|
"learning_rate": 7.698764148850138e-06, |
|
"loss": 0.9203, |
|
"step": 1970 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 0.052001953125, |
|
"learning_rate": 7.592563593037746e-06, |
|
"loss": 0.9094, |
|
"step": 1980 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 0.08544921875, |
|
"learning_rate": 7.486650495324783e-06, |
|
"loss": 0.9085, |
|
"step": 1990 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 0.049560546875, |
|
"learning_rate": 7.3810375021835275e-06, |
|
"loss": 0.8981, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 7.275737224252504e-06, |
|
"loss": 0.9005, |
|
"step": 2010 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 0.11181640625, |
|
"learning_rate": 7.1707622348307e-06, |
|
"loss": 0.881, |
|
"step": 2020 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 0.05615234375, |
|
"learning_rate": 7.066125068376297e-06, |
|
"loss": 0.9161, |
|
"step": 2030 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 0.1630859375, |
|
"learning_rate": 6.961838219009968e-06, |
|
"loss": 0.9142, |
|
"step": 2040 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 0.051513671875, |
|
"learning_rate": 6.857914139023058e-06, |
|
"loss": 0.8743, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 0.06298828125, |
|
"learning_rate": 6.7543652373906966e-06, |
|
"loss": 0.9011, |
|
"step": 2060 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 0.0556640625, |
|
"learning_rate": 6.651203878290139e-06, |
|
"loss": 0.8941, |
|
"step": 2070 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 0.060791015625, |
|
"learning_rate": 6.548442379624425e-06, |
|
"loss": 0.9323, |
|
"step": 2080 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 0.0634765625, |
|
"learning_rate": 6.446093011551551e-06, |
|
"loss": 0.9249, |
|
"step": 2090 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 0.09130859375, |
|
"learning_rate": 6.344167995019395e-06, |
|
"loss": 0.9195, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 0.10595703125, |
|
"learning_rate": 6.242679500306443e-06, |
|
"loss": 0.9248, |
|
"step": 2110 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 0.12255859375, |
|
"learning_rate": 6.141639645568646e-06, |
|
"loss": 0.8919, |
|
"step": 2120 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 0.10400390625, |
|
"learning_rate": 6.041060495392437e-06, |
|
"loss": 0.9251, |
|
"step": 2130 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 0.09716796875, |
|
"learning_rate": 5.940954059354165e-06, |
|
"loss": 0.9018, |
|
"step": 2140 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 0.07568359375, |
|
"learning_rate": 5.841332290586126e-06, |
|
"loss": 0.9205, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 0.0546875, |
|
"learning_rate": 5.742207084349274e-06, |
|
"loss": 0.9046, |
|
"step": 2160 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 0.06689453125, |
|
"learning_rate": 5.643590276612909e-06, |
|
"loss": 0.8988, |
|
"step": 2170 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 0.05126953125, |
|
"learning_rate": 5.545493642641389e-06, |
|
"loss": 0.8806, |
|
"step": 2180 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 0.0615234375, |
|
"learning_rate": 5.447928895588128e-06, |
|
"loss": 0.8878, |
|
"step": 2190 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 0.08544921875, |
|
"learning_rate": 5.350907685096983e-06, |
|
"loss": 0.8997, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 0.05810546875, |
|
"learning_rate": 5.254441595911255e-06, |
|
"loss": 0.9127, |
|
"step": 2210 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 0.05419921875, |
|
"learning_rate": 5.1585421464904e-06, |
|
"loss": 0.8924, |
|
"step": 2220 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 0.0615234375, |
|
"learning_rate": 5.063220787634686e-06, |
|
"loss": 0.9403, |
|
"step": 2230 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 0.053955078125, |
|
"learning_rate": 4.9684889011179335e-06, |
|
"loss": 0.8869, |
|
"step": 2240 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 0.07275390625, |
|
"learning_rate": 4.874357798328464e-06, |
|
"loss": 0.9012, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"grad_norm": 0.050537109375, |
|
"learning_rate": 4.780838718918467e-06, |
|
"loss": 0.9102, |
|
"step": 2260 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"grad_norm": 0.051025390625, |
|
"learning_rate": 4.687942829461969e-06, |
|
"loss": 0.8802, |
|
"step": 2270 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"grad_norm": 0.05810546875, |
|
"learning_rate": 4.595681222121458e-06, |
|
"loss": 0.9363, |
|
"step": 2280 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 0.181640625, |
|
"learning_rate": 4.504064913323472e-06, |
|
"loss": 0.9008, |
|
"step": 2290 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 0.146484375, |
|
"learning_rate": 4.413104842443149e-06, |
|
"loss": 0.8748, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 0.06689453125, |
|
"learning_rate": 4.322811870498058e-06, |
|
"loss": 0.8809, |
|
"step": 2310 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 0.06591796875, |
|
"learning_rate": 4.2331967788513295e-06, |
|
"loss": 0.9136, |
|
"step": 2320 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 0.115234375, |
|
"learning_rate": 4.144270267924306e-06, |
|
"loss": 0.9095, |
|
"step": 2330 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 0.1162109375, |
|
"learning_rate": 4.05604295591889e-06, |
|
"loss": 0.9077, |
|
"step": 2340 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 0.2177734375, |
|
"learning_rate": 3.968525377549657e-06, |
|
"loss": 0.9191, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 0.06494140625, |
|
"learning_rate": 3.881727982785999e-06, |
|
"loss": 0.8953, |
|
"step": 2360 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 0.07568359375, |
|
"learning_rate": 3.7956611356043196e-06, |
|
"loss": 0.9204, |
|
"step": 2370 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 0.06884765625, |
|
"learning_rate": 3.7103351127505616e-06, |
|
"loss": 0.9004, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 0.06884765625, |
|
"learning_rate": 3.625760102513103e-06, |
|
"loss": 0.9206, |
|
"step": 2390 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 0.259765625, |
|
"learning_rate": 3.5419462035062313e-06, |
|
"loss": 0.8932, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 0.058837890625, |
|
"learning_rate": 3.4589034234643315e-06, |
|
"loss": 0.9136, |
|
"step": 2410 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 0.06494140625, |
|
"learning_rate": 3.376641678046926e-06, |
|
"loss": 0.8907, |
|
"step": 2420 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 0.05859375, |
|
"learning_rate": 3.2951707896546858e-06, |
|
"loss": 0.9091, |
|
"step": 2430 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 0.06689453125, |
|
"learning_rate": 3.2145004862566185e-06, |
|
"loss": 0.9153, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 0.059326171875, |
|
"learning_rate": 3.134640400228479e-06, |
|
"loss": 0.9054, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 0.06591796875, |
|
"learning_rate": 3.055600067202652e-06, |
|
"loss": 0.9191, |
|
"step": 2460 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 0.09326171875, |
|
"learning_rate": 2.9773889249295294e-06, |
|
"loss": 0.8951, |
|
"step": 2470 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 0.06298828125, |
|
"learning_rate": 2.900016312150631e-06, |
|
"loss": 0.9205, |
|
"step": 2480 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 0.0625, |
|
"learning_rate": 2.8234914674834913e-06, |
|
"loss": 0.9022, |
|
"step": 2490 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 0.15234375, |
|
"learning_rate": 2.7478235283185504e-06, |
|
"loss": 0.9253, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 0.06591796875, |
|
"learning_rate": 2.673021529728108e-06, |
|
"loss": 0.905, |
|
"step": 2510 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 0.064453125, |
|
"learning_rate": 2.599094403387481e-06, |
|
"loss": 0.8965, |
|
"step": 2520 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 0.0517578125, |
|
"learning_rate": 2.5260509765085474e-06, |
|
"loss": 0.9029, |
|
"step": 2530 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 0.1513671875, |
|
"learning_rate": 2.453899970785716e-06, |
|
"loss": 0.8841, |
|
"step": 2540 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 0.0556640625, |
|
"learning_rate": 2.382650001354543e-06, |
|
"loss": 0.8932, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 0.0771484375, |
|
"learning_rate": 2.3123095757630344e-06, |
|
"loss": 0.8846, |
|
"step": 2560 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 0.058349609375, |
|
"learning_rate": 2.2428870929558012e-06, |
|
"loss": 0.9301, |
|
"step": 2570 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 0.05712890625, |
|
"learning_rate": 2.1743908422712135e-06, |
|
"loss": 0.9335, |
|
"step": 2580 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 0.107421875, |
|
"learning_rate": 2.1068290024515925e-06, |
|
"loss": 0.9254, |
|
"step": 2590 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 0.1708984375, |
|
"learning_rate": 2.0402096406666562e-06, |
|
"loss": 0.9215, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 0.053955078125, |
|
"learning_rate": 1.974540711550248e-06, |
|
"loss": 0.9133, |
|
"step": 2610 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 0.0986328125, |
|
"learning_rate": 1.9098300562505266e-06, |
|
"loss": 0.9071, |
|
"step": 2620 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 0.050537109375, |
|
"learning_rate": 1.8460854014937068e-06, |
|
"loss": 0.8938, |
|
"step": 2630 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 0.0634765625, |
|
"learning_rate": 1.783314358661441e-06, |
|
"loss": 0.9061, |
|
"step": 2640 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 0.0595703125, |
|
"learning_rate": 1.7215244228820027e-06, |
|
"loss": 0.8908, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 0.09716796875, |
|
"learning_rate": 1.6607229721353202e-06, |
|
"loss": 0.9104, |
|
"step": 2660 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 0.06787109375, |
|
"learning_rate": 1.6009172663720352e-06, |
|
"loss": 0.9258, |
|
"step": 2670 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 0.08544921875, |
|
"learning_rate": 1.5421144466466164e-06, |
|
"loss": 0.8934, |
|
"step": 2680 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 0.09033203125, |
|
"learning_rate": 1.4843215342646922e-06, |
|
"loss": 0.9161, |
|
"step": 2690 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 0.138671875, |
|
"learning_rate": 1.4275454299446834e-06, |
|
"loss": 0.925, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 0.0654296875, |
|
"learning_rate": 1.3717929129938179e-06, |
|
"loss": 0.9128, |
|
"step": 2710 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 0.056640625, |
|
"learning_rate": 1.3170706404986645e-06, |
|
"loss": 0.8888, |
|
"step": 2720 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 0.0751953125, |
|
"learning_rate": 1.263385146530234e-06, |
|
"loss": 0.9086, |
|
"step": 2730 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 0.05322265625, |
|
"learning_rate": 1.2107428413637979e-06, |
|
"loss": 0.9183, |
|
"step": 2740 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 0.072265625, |
|
"learning_rate": 1.15915001071347e-06, |
|
"loss": 0.8933, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 0.05908203125, |
|
"learning_rate": 1.1086128149816544e-06, |
|
"loss": 0.91, |
|
"step": 2760 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 0.15625, |
|
"learning_rate": 1.0591372885234885e-06, |
|
"loss": 0.9167, |
|
"step": 2770 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 0.07373046875, |
|
"learning_rate": 1.0107293389262918e-06, |
|
"loss": 0.9172, |
|
"step": 2780 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 0.10498046875, |
|
"learning_rate": 9.63394746304198e-07, |
|
"loss": 0.89, |
|
"step": 2790 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 0.4140625, |
|
"learning_rate": 9.171391626079629e-07, |
|
"loss": 0.8683, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 0.10205078125, |
|
"learning_rate": 8.719681109501177e-07, |
|
"loss": 0.9123, |
|
"step": 2810 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 0.0576171875, |
|
"learning_rate": 8.278869849454718e-07, |
|
"loss": 0.9419, |
|
"step": 2820 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 0.06005859375, |
|
"learning_rate": 7.849010480670938e-07, |
|
"loss": 0.9269, |
|
"step": 2830 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 0.12451171875, |
|
"learning_rate": 7.43015433017844e-07, |
|
"loss": 0.9286, |
|
"step": 2840 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 0.056884765625, |
|
"learning_rate": 7.022351411174866e-07, |
|
"loss": 0.9317, |
|
"step": 2850 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 0.06396484375, |
|
"learning_rate": 6.625650417055296e-07, |
|
"loss": 0.9374, |
|
"step": 2860 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 0.05419921875, |
|
"learning_rate": 6.240098715597975e-07, |
|
"loss": 0.9221, |
|
"step": 2870 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 0.0556640625, |
|
"learning_rate": 5.865742343308345e-07, |
|
"loss": 0.8889, |
|
"step": 2880 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 0.07666015625, |
|
"learning_rate": 5.502625999922207e-07, |
|
"loss": 0.9015, |
|
"step": 2890 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 0.0693359375, |
|
"learning_rate": 5.150793043068269e-07, |
|
"loss": 0.9012, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 0.052001953125, |
|
"learning_rate": 4.810285483091181e-07, |
|
"loss": 0.8941, |
|
"step": 2910 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 0.083984375, |
|
"learning_rate": 4.481143978035196e-07, |
|
"loss": 0.9181, |
|
"step": 2920 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 0.060302734375, |
|
"learning_rate": 4.163407828789523e-07, |
|
"loss": 0.9184, |
|
"step": 2930 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 0.2578125, |
|
"learning_rate": 3.857114974395604e-07, |
|
"loss": 0.9085, |
|
"step": 2940 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 0.07470703125, |
|
"learning_rate": 3.5623019875169916e-07, |
|
"loss": 0.9153, |
|
"step": 2950 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 0.04736328125, |
|
"learning_rate": 3.2790040700725114e-07, |
|
"loss": 0.8967, |
|
"step": 2960 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 0.0810546875, |
|
"learning_rate": 3.0072550490328754e-07, |
|
"loss": 0.9105, |
|
"step": 2970 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 0.06298828125, |
|
"learning_rate": 2.7470873723817405e-07, |
|
"loss": 0.9336, |
|
"step": 2980 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"grad_norm": 0.06640625, |
|
"learning_rate": 2.498532105241158e-07, |
|
"loss": 0.9182, |
|
"step": 2990 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"grad_norm": 0.058349609375, |
|
"learning_rate": 2.2616189261623568e-07, |
|
"loss": 0.9152, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"grad_norm": 0.08935546875, |
|
"learning_rate": 2.0363761235819402e-07, |
|
"loss": 0.9392, |
|
"step": 3010 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 0.06591796875, |
|
"learning_rate": 1.8228305924441469e-07, |
|
"loss": 0.8772, |
|
"step": 3020 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 0.056396484375, |
|
"learning_rate": 1.621007830989496e-07, |
|
"loss": 0.9035, |
|
"step": 3030 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 0.0712890625, |
|
"learning_rate": 1.430931937710156e-07, |
|
"loss": 0.9264, |
|
"step": 3040 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 0.1015625, |
|
"learning_rate": 1.2526256084725351e-07, |
|
"loss": 0.9229, |
|
"step": 3050 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 0.07958984375, |
|
"learning_rate": 1.0861101338072499e-07, |
|
"loss": 0.8886, |
|
"step": 3060 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 0.05810546875, |
|
"learning_rate": 9.314053963669245e-08, |
|
"loss": 0.9022, |
|
"step": 3070 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 0.053466796875, |
|
"learning_rate": 7.885298685522235e-08, |
|
"loss": 0.902, |
|
"step": 3080 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 0.056884765625, |
|
"learning_rate": 6.575006103060388e-08, |
|
"loss": 0.9356, |
|
"step": 3090 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 0.058349609375, |
|
"learning_rate": 5.3833326707656284e-08, |
|
"loss": 0.9262, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 0.078125, |
|
"learning_rate": 4.310420679490945e-08, |
|
"loss": 0.8869, |
|
"step": 3110 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 0.07080078125, |
|
"learning_rate": 3.356398239470427e-08, |
|
"loss": 0.9479, |
|
"step": 3120 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 0.050537109375, |
|
"learning_rate": 2.5213792650227165e-08, |
|
"loss": 0.8915, |
|
"step": 3130 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 0.0888671875, |
|
"learning_rate": 1.8054634609484534e-08, |
|
"loss": 0.9069, |
|
"step": 3140 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 0.0634765625, |
|
"learning_rate": 1.20873631062568e-08, |
|
"loss": 0.8806, |
|
"step": 3150 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 0.3671875, |
|
"learning_rate": 7.312690658024535e-09, |
|
"loss": 0.9192, |
|
"step": 3160 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 0.11865234375, |
|
"learning_rate": 3.731187380893176e-09, |
|
"loss": 0.8848, |
|
"step": 3170 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 0.1455078125, |
|
"learning_rate": 1.343280921518586e-09, |
|
"loss": 0.9189, |
|
"step": 3180 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 0.232421875, |
|
"learning_rate": 1.4925640603902226e-10, |
|
"loss": 0.9215, |
|
"step": 3190 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 3195, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 1, |
|
"save_steps": 500, |
|
"total_flos": 9.706671003244954e+17, |
|
"train_batch_size": 2, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|