| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.888125613346418, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03925417075564279, | |
| "grad_norm": 81.5, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.3818, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07850834151128558, | |
| "grad_norm": 322.0, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.367, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11776251226692837, | |
| "grad_norm": 784.0, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.4472, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15701668302257116, | |
| "grad_norm": 350.0, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.6806, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.19627085377821393, | |
| "grad_norm": 186.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9139, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.23552502453385674, | |
| "grad_norm": 199.0, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.8205, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2747791952894995, | |
| "grad_norm": 118.0, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.9017, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3140333660451423, | |
| "grad_norm": 616.0, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 3.1966, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.35328753680078506, | |
| "grad_norm": 692.0, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.0627, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.39254170755642787, | |
| "grad_norm": 75.5, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 7.3782, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.43179587831207067, | |
| "grad_norm": 908.0, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.1091, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.47105004906771347, | |
| "grad_norm": 61.5, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9262, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5103042198233563, | |
| "grad_norm": 45.25, | |
| "learning_rate": 1.04e-05, | |
| "loss": 0.6774, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.549558390578999, | |
| "grad_norm": 39.75, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.8204, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5888125613346418, | |
| "grad_norm": 38.0, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.6443, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6280667320902846, | |
| "grad_norm": 31.25, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 0.607, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6673209028459274, | |
| "grad_norm": 33.5, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 0.674, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7065750736015701, | |
| "grad_norm": 37.5, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 0.6806, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.745829244357213, | |
| "grad_norm": 36.25, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 0.6131, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7850834151128557, | |
| "grad_norm": 43.25, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.5914, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8243375858684985, | |
| "grad_norm": 27.75, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 0.5488, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8635917566241413, | |
| "grad_norm": 47.75, | |
| "learning_rate": 1.76e-05, | |
| "loss": 0.7107, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9028459273797841, | |
| "grad_norm": 16.375, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.2491, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9421000981354269, | |
| "grad_norm": 35.5, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 0.6448, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9813542688910697, | |
| "grad_norm": 22.875, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5682, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0206084396467126, | |
| "grad_norm": 32.75, | |
| "learning_rate": 1.9999025240093045e-05, | |
| "loss": 0.5689, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0598626104023552, | |
| "grad_norm": 28.375, | |
| "learning_rate": 1.9996101150403543e-05, | |
| "loss": 0.4836, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.099116781157998, | |
| "grad_norm": 18.25, | |
| "learning_rate": 1.9991228300988586e-05, | |
| "loss": 0.519, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.138370951913641, | |
| "grad_norm": 26.125, | |
| "learning_rate": 1.9984407641819812e-05, | |
| "loss": 0.4941, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1776251226692835, | |
| "grad_norm": 20.5, | |
| "learning_rate": 1.9975640502598243e-05, | |
| "loss": 0.4484, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2168792934249264, | |
| "grad_norm": 24.625, | |
| "learning_rate": 1.9964928592495046e-05, | |
| "loss": 0.5169, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2561334641805693, | |
| "grad_norm": 14.125, | |
| "learning_rate": 1.9952273999818312e-05, | |
| "loss": 0.4417, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.295387634936212, | |
| "grad_norm": 26.25, | |
| "learning_rate": 1.9937679191605964e-05, | |
| "loss": 0.4361, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3346418056918548, | |
| "grad_norm": 23.125, | |
| "learning_rate": 1.9921147013144782e-05, | |
| "loss": 0.4454, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3738959764474976, | |
| "grad_norm": 23.625, | |
| "learning_rate": 1.9902680687415704e-05, | |
| "loss": 0.446, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4131501472031402, | |
| "grad_norm": 13.9375, | |
| "learning_rate": 1.988228381446553e-05, | |
| "loss": 0.4006, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.452404317958783, | |
| "grad_norm": 31.5, | |
| "learning_rate": 1.985996037070505e-05, | |
| "loss": 0.4577, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.491658488714426, | |
| "grad_norm": 47.5, | |
| "learning_rate": 1.983571470813386e-05, | |
| "loss": 0.4207, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5309126594700686, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.9809551553491918e-05, | |
| "loss": 0.4251, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.5701668302257115, | |
| "grad_norm": 20.75, | |
| "learning_rate": 1.9781476007338058e-05, | |
| "loss": 0.3968, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.6094210009813543, | |
| "grad_norm": 15.0, | |
| "learning_rate": 1.9751493543055634e-05, | |
| "loss": 0.4019, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.648675171736997, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.9719610005785466e-05, | |
| "loss": 0.3894, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6879293424926398, | |
| "grad_norm": 20.375, | |
| "learning_rate": 1.9685831611286312e-05, | |
| "loss": 0.3923, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7271835132482827, | |
| "grad_norm": 16.75, | |
| "learning_rate": 1.9650164944723116e-05, | |
| "loss": 0.374, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.7664376840039253, | |
| "grad_norm": 21.375, | |
| "learning_rate": 1.961261695938319e-05, | |
| "loss": 0.402, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.8056918547595682, | |
| "grad_norm": 14.25, | |
| "learning_rate": 1.9573194975320672e-05, | |
| "loss": 0.3646, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.844946025515211, | |
| "grad_norm": 15.0625, | |
| "learning_rate": 1.9531906677929472e-05, | |
| "loss": 0.3451, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.8842001962708537, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.9488760116444966e-05, | |
| "loss": 0.3373, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9234543670264965, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.944376370237481e-05, | |
| "loss": 0.3693, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9627085377821394, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.3595, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.001962708537782, | |
| "grad_norm": 19.75, | |
| "learning_rate": 1.9348256763960146e-05, | |
| "loss": 0.3795, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.041216879293425, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.9297764858882516e-05, | |
| "loss": 0.3388, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.0804710500490677, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.9245460336123136e-05, | |
| "loss": 0.3306, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1197252208047104, | |
| "grad_norm": 25.25, | |
| "learning_rate": 1.9191353392552346e-05, | |
| "loss": 0.3471, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1589793915603535, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.913545457642601e-05, | |
| "loss": 0.3268, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.198233562315996, | |
| "grad_norm": 18.375, | |
| "learning_rate": 1.907777478532909e-05, | |
| "loss": 0.3138, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2374877330716387, | |
| "grad_norm": 11.375, | |
| "learning_rate": 1.901832526405114e-05, | |
| "loss": 0.3271, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.276741903827282, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 1.895711760239413e-05, | |
| "loss": 0.3189, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3159960745829244, | |
| "grad_norm": 21.0, | |
| "learning_rate": 1.889416373291298e-05, | |
| "loss": 0.3413, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.355250245338567, | |
| "grad_norm": 12.1875, | |
| "learning_rate": 1.8829475928589272e-05, | |
| "loss": 0.3222, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.39450441609421, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 1.8763066800438638e-05, | |
| "loss": 0.3238, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.433758586849853, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 1.869494929505219e-05, | |
| "loss": 0.3133, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.4730127576054954, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.8625136692072577e-05, | |
| "loss": 0.3337, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.5122669283611385, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 1.855364260160507e-05, | |
| "loss": 0.3016, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.551521099116781, | |
| "grad_norm": 8.0, | |
| "learning_rate": 1.848048096156426e-05, | |
| "loss": 0.3138, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.590775269872424, | |
| "grad_norm": 13.5, | |
| "learning_rate": 1.8405666034956842e-05, | |
| "loss": 0.295, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.630029440628067, | |
| "grad_norm": 18.125, | |
| "learning_rate": 1.8329212407100996e-05, | |
| "loss": 0.3002, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.6692836113837095, | |
| "grad_norm": 17.375, | |
| "learning_rate": 1.8251134982782952e-05, | |
| "loss": 0.3026, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.708537782139352, | |
| "grad_norm": 8.75, | |
| "learning_rate": 1.8171448983351284e-05, | |
| "loss": 0.2936, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7477919528949952, | |
| "grad_norm": 13.375, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.3181, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.787046123650638, | |
| "grad_norm": 22.0, | |
| "learning_rate": 1.8007313709487334e-05, | |
| "loss": 0.2988, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8263002944062805, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.792289643355191e-05, | |
| "loss": 0.3096, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.8655544651619236, | |
| "grad_norm": 21.25, | |
| "learning_rate": 1.78369345732584e-05, | |
| "loss": 0.2987, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.904808635917566, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 1.7749444887041797e-05, | |
| "loss": 0.2787, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.944062806673209, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.2763, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.983316977428852, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.7569950556517566e-05, | |
| "loss": 0.2841, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.0225711481844946, | |
| "grad_norm": 8.25, | |
| "learning_rate": 1.747798090498532e-05, | |
| "loss": 0.2799, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.061825318940137, | |
| "grad_norm": 15.5, | |
| "learning_rate": 1.7384553406258842e-05, | |
| "loss": 0.2664, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.1010794896957803, | |
| "grad_norm": 19.125, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 0.262, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.140333660451423, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.7193398003386514e-05, | |
| "loss": 0.2588, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.1795878312070656, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 1.709570736536521e-05, | |
| "loss": 0.264, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.2188420019627086, | |
| "grad_norm": 5.375, | |
| "learning_rate": 1.6996633405133656e-05, | |
| "loss": 0.2622, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.2580961727183513, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 1.68961954373567e-05, | |
| "loss": 0.2576, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.297350343473994, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 1.6794413042615168e-05, | |
| "loss": 0.2602, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.336604514229637, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 1.6691306063588583e-05, | |
| "loss": 0.2495, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.3758586849852796, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 1.6586894601186804e-05, | |
| "loss": 0.2568, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.4151128557409223, | |
| "grad_norm": 12.25, | |
| "learning_rate": 1.6481199010631312e-05, | |
| "loss": 0.2429, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.4543670264965654, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.63742398974869e-05, | |
| "loss": 0.2647, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.493621197252208, | |
| "grad_norm": 9.625, | |
| "learning_rate": 1.6266038113644605e-05, | |
| "loss": 0.2571, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.5328753680078506, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 1.6156614753256583e-05, | |
| "loss": 0.2565, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.5721295387634937, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1.6045991148623752e-05, | |
| "loss": 0.2591, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.6113837095191363, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 1.5934188866037017e-05, | |
| "loss": 0.2364, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.650637880274779, | |
| "grad_norm": 12.75, | |
| "learning_rate": 1.5821229701572897e-05, | |
| "loss": 0.2508, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.689892051030422, | |
| "grad_norm": 12.375, | |
| "learning_rate": 1.570713567684432e-05, | |
| "loss": 0.2251, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.7291462217860647, | |
| "grad_norm": 4.5, | |
| "learning_rate": 1.5591929034707468e-05, | |
| "loss": 0.2359, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.7684003925417073, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.5475632234925505e-05, | |
| "loss": 0.2498, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8076545632973504, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.5358267949789968e-05, | |
| "loss": 0.2337, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.846908734052993, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 1.5239859059700794e-05, | |
| "loss": 0.254, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.8861629048086357, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 1.5120428648705716e-05, | |
| "loss": 0.249, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.9254170755642788, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.2464, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.9646712463199214, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 1.4878596591387329e-05, | |
| "loss": 0.2452, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.003925417075564, | |
| "grad_norm": 20.0, | |
| "learning_rate": 1.4756242090702756e-05, | |
| "loss": 0.2338, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.043179587831207, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 1.463296035119862e-05, | |
| "loss": 0.2043, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.08243375858685, | |
| "grad_norm": 7.375, | |
| "learning_rate": 1.4508775406894308e-05, | |
| "loss": 0.2217, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.121687929342492, | |
| "grad_norm": 4.0, | |
| "learning_rate": 1.4383711467890776e-05, | |
| "loss": 0.2288, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.1609421000981355, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 1.4257792915650728e-05, | |
| "loss": 0.2023, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.200196270853779, | |
| "grad_norm": 9.25, | |
| "learning_rate": 1.413104429824542e-05, | |
| "loss": 0.2005, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.239450441609421, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.4003490325568953e-05, | |
| "loss": 0.2098, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.278704612365064, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 1.3875155864521031e-05, | |
| "loss": 0.2059, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.317958783120707, | |
| "grad_norm": 7.5, | |
| "learning_rate": 1.3746065934159123e-05, | |
| "loss": 0.1923, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.357212953876349, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 1.3616245700820922e-05, | |
| "loss": 0.2082, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.396467124631992, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.3485720473218153e-05, | |
| "loss": 0.213, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.435721295387635, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 1.3354515697502552e-05, | |
| "loss": 0.2213, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.4749754661432775, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.3222656952305113e-05, | |
| "loss": 0.1971, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.5142296368989205, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.2035, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.553483807654564, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 1.2957080500440469e-05, | |
| "loss": 0.1887, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.592737978410206, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 1.2823414568428767e-05, | |
| "loss": 0.1914, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.631992149165849, | |
| "grad_norm": 3.0, | |
| "learning_rate": 1.2689198206152657e-05, | |
| "loss": 0.2014, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.671246319921492, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 1.2554457579357906e-05, | |
| "loss": 0.1934, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.710500490677134, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 1.2419218955996677e-05, | |
| "loss": 0.1914, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.749754661432777, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 1.2283508701106559e-05, | |
| "loss": 0.2069, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.78900883218842, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.2147353271670634e-05, | |
| "loss": 0.2163, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.8282630029440625, | |
| "grad_norm": 5.25, | |
| "learning_rate": 1.2010779211459649e-05, | |
| "loss": 0.2, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.867517173699706, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.187381314585725e-05, | |
| "loss": 0.1848, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.906771344455349, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.1913, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.946025515210991, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.159881187691835e-05, | |
| "loss": 0.1963, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.985279685966634, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 1.1460830285624119e-05, | |
| "loss": 0.1911, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 5.024533856722277, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 1.1322563902571227e-05, | |
| "loss": 0.1775, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 5.063788027477919, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 1.1184039683065014e-05, | |
| "loss": 0.1645, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.103042198233562, | |
| "grad_norm": 5.40625, | |
| "learning_rate": 1.1045284632676535e-05, | |
| "loss": 0.1624, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.142296368989205, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.0906325801977804e-05, | |
| "loss": 0.1761, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.181550539744848, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.0767190281268187e-05, | |
| "loss": 0.1761, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.220804710500491, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 1.0627905195293135e-05, | |
| "loss": 0.1744, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.260058881256134, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.0488497697956134e-05, | |
| "loss": 0.1787, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.299313052011776, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 1.0348994967025012e-05, | |
| "loss": 0.1651, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.338567222767419, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.0209424198833571e-05, | |
| "loss": 0.1684, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 5.377821393523062, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 1.0069812602979617e-05, | |
| "loss": 0.1699, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 5.417075564278704, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 9.930187397020385e-06, | |
| "loss": 0.1677, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 5.456329735034347, | |
| "grad_norm": 3.0, | |
| "learning_rate": 9.790575801166432e-06, | |
| "loss": 0.1612, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 5.4955839057899905, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.651005032974994e-06, | |
| "loss": 0.1623, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.534838076545633, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 9.511502302043867e-06, | |
| "loss": 0.1674, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 5.574092247301276, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 9.372094804706867e-06, | |
| "loss": 0.1629, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 5.613346418056919, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 9.232809718731815e-06, | |
| "loss": 0.1625, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 5.652600588812561, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 9.093674198022201e-06, | |
| "loss": 0.1667, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 5.691854759568204, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 8.954715367323468e-06, | |
| "loss": 0.1702, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.731108930323847, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 8.815960316934991e-06, | |
| "loss": 0.1646, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 5.770363101079489, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.677436097428775e-06, | |
| "loss": 0.1635, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 5.809617271835132, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 8.539169714375885e-06, | |
| "loss": 0.1533, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.8488714425907755, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 8.401188123081653e-06, | |
| "loss": 0.1535, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 5.888125613346418, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.1586, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.73984932364288e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |