| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007142857142857143, | |
| "grad_norm": 3.0706775188446045, | |
| "learning_rate": 9.92857142857143e-05, | |
| "loss": 1.2883, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014285714285714285, | |
| "grad_norm": 2.5454561710357666, | |
| "learning_rate": 9.857142857142858e-05, | |
| "loss": 1.1211, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02142857142857143, | |
| "grad_norm": 2.717937469482422, | |
| "learning_rate": 9.785714285714286e-05, | |
| "loss": 0.9927, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 1.2683441638946533, | |
| "learning_rate": 9.714285714285715e-05, | |
| "loss": 0.9406, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 7.902857303619385, | |
| "learning_rate": 9.642857142857143e-05, | |
| "loss": 0.7558, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04285714285714286, | |
| "grad_norm": 2.552997350692749, | |
| "learning_rate": 9.571428571428573e-05, | |
| "loss": 0.6445, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.5295348167419434, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.6534, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 1.8732738494873047, | |
| "learning_rate": 9.428571428571429e-05, | |
| "loss": 0.537, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06428571428571428, | |
| "grad_norm": 4.299726963043213, | |
| "learning_rate": 9.357142857142858e-05, | |
| "loss": 0.5508, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 2.1415882110595703, | |
| "learning_rate": 9.285714285714286e-05, | |
| "loss": 0.5438, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07857142857142857, | |
| "grad_norm": 3.7711620330810547, | |
| "learning_rate": 9.214285714285714e-05, | |
| "loss": 0.5906, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 1.4387797117233276, | |
| "learning_rate": 9.142857142857143e-05, | |
| "loss": 0.541, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09285714285714286, | |
| "grad_norm": 4.010119915008545, | |
| "learning_rate": 9.071428571428571e-05, | |
| "loss": 0.56, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.021274089813232, | |
| "learning_rate": 9e-05, | |
| "loss": 0.5323, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 2.138714551925659, | |
| "learning_rate": 8.92857142857143e-05, | |
| "loss": 0.4784, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 1.371328353881836, | |
| "learning_rate": 8.857142857142857e-05, | |
| "loss": 0.4861, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.12142857142857143, | |
| "grad_norm": 3.1777868270874023, | |
| "learning_rate": 8.785714285714286e-05, | |
| "loss": 0.504, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12857142857142856, | |
| "grad_norm": 2.0382461547851562, | |
| "learning_rate": 8.714285714285715e-05, | |
| "loss": 0.496, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1357142857142857, | |
| "grad_norm": 2.8927695751190186, | |
| "learning_rate": 8.642857142857143e-05, | |
| "loss": 0.5581, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 4.461900234222412, | |
| "learning_rate": 8.571428571428571e-05, | |
| "loss": 0.4643, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.028140068054199, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.4384, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.15714285714285714, | |
| "grad_norm": 2.4652626514434814, | |
| "learning_rate": 8.428571428571429e-05, | |
| "loss": 0.4705, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.16428571428571428, | |
| "grad_norm": 2.360128879547119, | |
| "learning_rate": 8.357142857142858e-05, | |
| "loss": 0.4238, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 1.5631648302078247, | |
| "learning_rate": 8.285714285714287e-05, | |
| "loss": 0.4364, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 1.9402961730957031, | |
| "learning_rate": 8.214285714285714e-05, | |
| "loss": 0.4472, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.18571428571428572, | |
| "grad_norm": 2.9574289321899414, | |
| "learning_rate": 8.142857142857143e-05, | |
| "loss": 0.4319, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.19285714285714287, | |
| "grad_norm": 1.7986479997634888, | |
| "learning_rate": 8.071428571428573e-05, | |
| "loss": 0.4873, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.539698123931885, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4222, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.20714285714285716, | |
| "grad_norm": 4.539310932159424, | |
| "learning_rate": 7.928571428571429e-05, | |
| "loss": 0.444, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 4.211949348449707, | |
| "learning_rate": 7.857142857142858e-05, | |
| "loss": 0.4713, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.22142857142857142, | |
| "grad_norm": 3.4612796306610107, | |
| "learning_rate": 7.785714285714286e-05, | |
| "loss": 0.421, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 2.1370456218719482, | |
| "learning_rate": 7.714285714285715e-05, | |
| "loss": 0.5117, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2357142857142857, | |
| "grad_norm": 2.022341012954712, | |
| "learning_rate": 7.642857142857143e-05, | |
| "loss": 0.3968, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.24285714285714285, | |
| "grad_norm": 1.5263112783432007, | |
| "learning_rate": 7.571428571428571e-05, | |
| "loss": 0.4032, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.152087450027466, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.4024, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 3.5288479328155518, | |
| "learning_rate": 7.428571428571429e-05, | |
| "loss": 0.4079, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2642857142857143, | |
| "grad_norm": 1.922340989112854, | |
| "learning_rate": 7.357142857142858e-05, | |
| "loss": 0.4839, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2714285714285714, | |
| "grad_norm": 2.6698594093322754, | |
| "learning_rate": 7.285714285714286e-05, | |
| "loss": 0.3772, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.2785714285714286, | |
| "grad_norm": 2.9632909297943115, | |
| "learning_rate": 7.214285714285714e-05, | |
| "loss": 0.4296, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 2.624523639678955, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.3931, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.29285714285714287, | |
| "grad_norm": 3.6488404273986816, | |
| "learning_rate": 7.071428571428573e-05, | |
| "loss": 0.3986, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.6739540100097656, | |
| "learning_rate": 7e-05, | |
| "loss": 0.3979, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.30714285714285716, | |
| "grad_norm": 2.866800546646118, | |
| "learning_rate": 6.928571428571429e-05, | |
| "loss": 0.4203, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 3.4130825996398926, | |
| "learning_rate": 6.857142857142858e-05, | |
| "loss": 0.498, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 1.6507424116134644, | |
| "learning_rate": 6.785714285714286e-05, | |
| "loss": 0.3672, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.32857142857142857, | |
| "grad_norm": 2.9661030769348145, | |
| "learning_rate": 6.714285714285714e-05, | |
| "loss": 0.372, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3357142857142857, | |
| "grad_norm": 1.8931576013565063, | |
| "learning_rate": 6.642857142857143e-05, | |
| "loss": 0.341, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 2.128786563873291, | |
| "learning_rate": 6.571428571428571e-05, | |
| "loss": 0.3235, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.9210121631622314, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.4568, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 2.2050564289093018, | |
| "learning_rate": 6.428571428571429e-05, | |
| "loss": 0.3468, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.36428571428571427, | |
| "grad_norm": 1.4351139068603516, | |
| "learning_rate": 6.357142857142857e-05, | |
| "loss": 0.3745, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 1.60938560962677, | |
| "learning_rate": 6.285714285714286e-05, | |
| "loss": 0.3944, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.37857142857142856, | |
| "grad_norm": 3.240180730819702, | |
| "learning_rate": 6.214285714285714e-05, | |
| "loss": 0.3701, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.38571428571428573, | |
| "grad_norm": 2.8629231452941895, | |
| "learning_rate": 6.142857142857143e-05, | |
| "loss": 0.3378, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 2.2353878021240234, | |
| "learning_rate": 6.0714285714285715e-05, | |
| "loss": 0.3783, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.000092029571533, | |
| "learning_rate": 6e-05, | |
| "loss": 0.3664, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.40714285714285714, | |
| "grad_norm": 1.9683377742767334, | |
| "learning_rate": 5.928571428571429e-05, | |
| "loss": 0.4366, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4142857142857143, | |
| "grad_norm": 1.6545411348342896, | |
| "learning_rate": 5.8571428571428575e-05, | |
| "loss": 0.4047, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.42142857142857143, | |
| "grad_norm": 6.393994331359863, | |
| "learning_rate": 5.785714285714287e-05, | |
| "loss": 0.3513, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 2.786479949951172, | |
| "learning_rate": 5.714285714285714e-05, | |
| "loss": 0.3973, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4357142857142857, | |
| "grad_norm": 2.1006624698638916, | |
| "learning_rate": 5.642857142857143e-05, | |
| "loss": 0.413, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.44285714285714284, | |
| "grad_norm": 2.376877784729004, | |
| "learning_rate": 5.571428571428572e-05, | |
| "loss": 0.427, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.311000347137451, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.4043, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 5.115108489990234, | |
| "learning_rate": 5.428571428571428e-05, | |
| "loss": 0.4501, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 1.7104153633117676, | |
| "learning_rate": 5.3571428571428575e-05, | |
| "loss": 0.3966, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4714285714285714, | |
| "grad_norm": 1.9817299842834473, | |
| "learning_rate": 5.285714285714286e-05, | |
| "loss": 0.3308, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.4785714285714286, | |
| "grad_norm": 2.1275410652160645, | |
| "learning_rate": 5.214285714285715e-05, | |
| "loss": 0.3219, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 2.4896678924560547, | |
| "learning_rate": 5.142857142857143e-05, | |
| "loss": 0.3084, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.4928571428571429, | |
| "grad_norm": 2.4593498706817627, | |
| "learning_rate": 5.0714285714285716e-05, | |
| "loss": 0.3695, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.4440925121307373, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3584, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5071428571428571, | |
| "grad_norm": 1.6439753770828247, | |
| "learning_rate": 4.928571428571429e-05, | |
| "loss": 0.4272, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 2.897373676300049, | |
| "learning_rate": 4.8571428571428576e-05, | |
| "loss": 0.3496, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5214285714285715, | |
| "grad_norm": 2.620026111602783, | |
| "learning_rate": 4.785714285714286e-05, | |
| "loss": 0.3745, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5285714285714286, | |
| "grad_norm": 2.081660032272339, | |
| "learning_rate": 4.714285714285714e-05, | |
| "loss": 0.3608, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 1.8209961652755737, | |
| "learning_rate": 4.642857142857143e-05, | |
| "loss": 0.3429, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 1.898141622543335, | |
| "learning_rate": 4.5714285714285716e-05, | |
| "loss": 0.3671, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.252241611480713, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.3843, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5571428571428572, | |
| "grad_norm": 2.201845407485962, | |
| "learning_rate": 4.428571428571428e-05, | |
| "loss": 0.3379, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5642857142857143, | |
| "grad_norm": 1.9457173347473145, | |
| "learning_rate": 4.3571428571428576e-05, | |
| "loss": 0.3366, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 3.7929930686950684, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.3091, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5785714285714286, | |
| "grad_norm": 1.8005012273788452, | |
| "learning_rate": 4.214285714285714e-05, | |
| "loss": 0.2944, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5857142857142857, | |
| "grad_norm": 6.003371715545654, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 0.3722, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.5928571428571429, | |
| "grad_norm": 2.1478536128997803, | |
| "learning_rate": 4.0714285714285717e-05, | |
| "loss": 0.3081, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.839242935180664, | |
| "learning_rate": 4e-05, | |
| "loss": 0.317, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 1.891158938407898, | |
| "learning_rate": 3.928571428571429e-05, | |
| "loss": 0.2995, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6142857142857143, | |
| "grad_norm": 5.433562278747559, | |
| "learning_rate": 3.857142857142858e-05, | |
| "loss": 0.3637, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6214285714285714, | |
| "grad_norm": 1.653617024421692, | |
| "learning_rate": 3.785714285714286e-05, | |
| "loss": 0.3285, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 2.9657585620880127, | |
| "learning_rate": 3.7142857142857143e-05, | |
| "loss": 0.3761, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6357142857142857, | |
| "grad_norm": 1.5580040216445923, | |
| "learning_rate": 3.642857142857143e-05, | |
| "loss": 0.4427, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 2.411190986633301, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 0.3322, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.375690937042236, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.3546, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 3.571958541870117, | |
| "learning_rate": 3.428571428571429e-05, | |
| "loss": 0.3742, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6642857142857143, | |
| "grad_norm": 3.132997512817383, | |
| "learning_rate": 3.357142857142857e-05, | |
| "loss": 0.3667, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.6714285714285714, | |
| "grad_norm": 2.296008586883545, | |
| "learning_rate": 3.285714285714286e-05, | |
| "loss": 0.2686, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 1.7339441776275635, | |
| "learning_rate": 3.2142857142857144e-05, | |
| "loss": 0.328, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 2.5679969787597656, | |
| "learning_rate": 3.142857142857143e-05, | |
| "loss": 0.3437, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6928571428571428, | |
| "grad_norm": 6.115095615386963, | |
| "learning_rate": 3.071428571428572e-05, | |
| "loss": 0.3264, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 9.149005889892578, | |
| "learning_rate": 3e-05, | |
| "loss": 0.339, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7071428571428572, | |
| "grad_norm": 3.723829746246338, | |
| "learning_rate": 2.9285714285714288e-05, | |
| "loss": 0.3932, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.2082722187042236, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 0.3854, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7214285714285714, | |
| "grad_norm": 2.192686080932617, | |
| "learning_rate": 2.785714285714286e-05, | |
| "loss": 0.352, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.7285714285714285, | |
| "grad_norm": 2.3141751289367676, | |
| "learning_rate": 2.714285714285714e-05, | |
| "loss": 0.3132, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.7357142857142858, | |
| "grad_norm": 4.0575995445251465, | |
| "learning_rate": 2.642857142857143e-05, | |
| "loss": 0.2929, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 2.5704588890075684, | |
| "learning_rate": 2.5714285714285714e-05, | |
| "loss": 0.3549, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.970313310623169, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3139, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.7571428571428571, | |
| "grad_norm": 3.143388271331787, | |
| "learning_rate": 2.4285714285714288e-05, | |
| "loss": 0.3732, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.7642857142857142, | |
| "grad_norm": 3.7672691345214844, | |
| "learning_rate": 2.357142857142857e-05, | |
| "loss": 0.3752, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 3.106049060821533, | |
| "learning_rate": 2.2857142857142858e-05, | |
| "loss": 0.3174, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.7785714285714286, | |
| "grad_norm": 3.5952601432800293, | |
| "learning_rate": 2.214285714285714e-05, | |
| "loss": 0.3445, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 2.764934778213501, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.3343, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7928571428571428, | |
| "grad_norm": 3.4910130500793457, | |
| "learning_rate": 2.0714285714285718e-05, | |
| "loss": 0.3265, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.074836254119873, | |
| "learning_rate": 2e-05, | |
| "loss": 0.334, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8071428571428572, | |
| "grad_norm": 3.0394229888916016, | |
| "learning_rate": 1.928571428571429e-05, | |
| "loss": 0.2932, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.8142857142857143, | |
| "grad_norm": 2.841371536254883, | |
| "learning_rate": 1.8571428571428572e-05, | |
| "loss": 0.3143, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 3.9328362941741943, | |
| "learning_rate": 1.785714285714286e-05, | |
| "loss": 0.3726, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 2.6354167461395264, | |
| "learning_rate": 1.7142857142857145e-05, | |
| "loss": 0.325, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.8357142857142857, | |
| "grad_norm": 4.113726615905762, | |
| "learning_rate": 1.642857142857143e-05, | |
| "loss": 0.3199, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8428571428571429, | |
| "grad_norm": 3.8376033306121826, | |
| "learning_rate": 1.5714285714285715e-05, | |
| "loss": 0.3299, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.71781063079834, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.317, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 3.487872362136841, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 0.3301, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8642857142857143, | |
| "grad_norm": 1.754288673400879, | |
| "learning_rate": 1.357142857142857e-05, | |
| "loss": 0.3469, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.8714285714285714, | |
| "grad_norm": 3.3238632678985596, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.3434, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.8785714285714286, | |
| "grad_norm": 2.276270627975464, | |
| "learning_rate": 1.2142857142857144e-05, | |
| "loss": 0.2825, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 4.0892720222473145, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.3364, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 2.1741929054260254, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 0.3592, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.3750619888305664, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4065, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.9071428571428571, | |
| "grad_norm": 2.2269678115844727, | |
| "learning_rate": 9.285714285714286e-06, | |
| "loss": 0.4073, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 2.025587320327759, | |
| "learning_rate": 8.571428571428573e-06, | |
| "loss": 0.3478, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.9214285714285714, | |
| "grad_norm": 5.870853900909424, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.386, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 2.9051337242126465, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 0.3257, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9357142857142857, | |
| "grad_norm": 3.7982888221740723, | |
| "learning_rate": 6.428571428571429e-06, | |
| "loss": 0.3405, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 5.413133144378662, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.2909, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.1331217288970947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3175, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.9571428571428572, | |
| "grad_norm": 4.568922519683838, | |
| "learning_rate": 4.285714285714286e-06, | |
| "loss": 0.3447, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 2.920402765274048, | |
| "learning_rate": 3.5714285714285714e-06, | |
| "loss": 0.3357, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 1.938262939453125, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.2942, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.9785714285714285, | |
| "grad_norm": 1.8624756336212158, | |
| "learning_rate": 2.142857142857143e-06, | |
| "loss": 0.3498, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.9857142857142858, | |
| "grad_norm": 2.570404291152954, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.3777, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9928571428571429, | |
| "grad_norm": 1.80099356174469, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.3356, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.955940008163452, | |
| "learning_rate": 0.0, | |
| "loss": 0.3044, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3177724778652191, | |
| "eval_runtime": 67.4116, | |
| "eval_samples_per_second": 2.967, | |
| "eval_steps_per_second": 0.371, | |
| "step": 1400 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8377941884928000.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |