| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005, | |
| "grad_norm": 0.07553869201667132, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.277, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.07142321749040205, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1731, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.015, | |
| "grad_norm": 0.04004541380672789, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.2066, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.03135466258081705, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1541, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 0.08107204814222205, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.3396, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.02826228780285631, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0857, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.035, | |
| "grad_norm": 0.04534872682356579, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1668, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.08062512421235096, | |
| "learning_rate": 2e-05, | |
| "loss": 0.307, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.045, | |
| "grad_norm": 0.09855642598201231, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.3094, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.08223399648677109, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3169, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.055, | |
| "grad_norm": 0.036306345795043035, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.156, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0446200343132829, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1902, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.065, | |
| "grad_norm": 0.0394950521156506, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.1665, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0779690335299316, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.3144, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 0.0646426759636167, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2018, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.08984226365886303, | |
| "learning_rate": 4e-05, | |
| "loss": 0.2255, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.085, | |
| "grad_norm": 0.036674267927634084, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.1264, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.09371135142216022, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.2776, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.095, | |
| "grad_norm": 0.04744368538635616, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.149, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.11052474076313488, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3257, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.105, | |
| "grad_norm": 0.12235330629107824, | |
| "learning_rate": 5.25e-05, | |
| "loss": 0.2811, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.09475772803286764, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.2566, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.115, | |
| "grad_norm": 0.07682060292069236, | |
| "learning_rate": 5.7499999999999995e-05, | |
| "loss": 0.194, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.10253318753424252, | |
| "learning_rate": 6e-05, | |
| "loss": 0.1742, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.15452516312260753, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.2426, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.06980713124278756, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.1524, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.135, | |
| "grad_norm": 0.17169291075340687, | |
| "learning_rate": 6.750000000000001e-05, | |
| "loss": 0.2302, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.12097272092886306, | |
| "learning_rate": 7e-05, | |
| "loss": 0.1751, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.145, | |
| "grad_norm": 0.04894250463264245, | |
| "learning_rate": 7.25e-05, | |
| "loss": 0.0896, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.16842709573150588, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.3866, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.155, | |
| "grad_norm": 0.10293501286832617, | |
| "learning_rate": 7.75e-05, | |
| "loss": 0.166, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.06255095679441007, | |
| "learning_rate": 8e-05, | |
| "loss": 0.1388, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.165, | |
| "grad_norm": 0.09759267630202532, | |
| "learning_rate": 8.25e-05, | |
| "loss": 0.1998, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.03725031482525615, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.0808, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 0.06548370439282793, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.1329, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.06363300533898003, | |
| "learning_rate": 9e-05, | |
| "loss": 0.0765, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.185, | |
| "grad_norm": 0.1319969217444984, | |
| "learning_rate": 9.250000000000001e-05, | |
| "loss": 0.1583, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.1570940458256374, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.1703, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.195, | |
| "grad_norm": 0.17816543762213244, | |
| "learning_rate": 9.75e-05, | |
| "loss": 0.1677, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.06814489472510708, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0813, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.205, | |
| "grad_norm": 0.12234791580019863, | |
| "learning_rate": 9.999809615320856e-05, | |
| "loss": 0.067, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.11898491410474155, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.0832, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.215, | |
| "grad_norm": 0.11599178809474275, | |
| "learning_rate": 9.998286624877786e-05, | |
| "loss": 0.1261, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.18501882395233082, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.1908, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.06856454706988713, | |
| "learning_rate": 9.99524110790929e-05, | |
| "loss": 0.0672, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.05947222817167692, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.0745, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.235, | |
| "grad_norm": 0.07765190202488637, | |
| "learning_rate": 9.990673992109335e-05, | |
| "loss": 0.0756, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.1359994526177061, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.0331, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.245, | |
| "grad_norm": 0.1373406192095304, | |
| "learning_rate": 9.98458666866564e-05, | |
| "loss": 0.0806, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.05473571476326683, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.0522, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.255, | |
| "grad_norm": 0.1169729050564611, | |
| "learning_rate": 9.976980991835894e-05, | |
| "loss": 0.0762, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.05579810258223367, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.0345, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.265, | |
| "grad_norm": 0.09381823707624487, | |
| "learning_rate": 9.967859278382938e-05, | |
| "loss": 0.0897, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.07938166641979207, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.0654, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.13087865330833595, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 0.1015, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.263321695571175, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.0519, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.285, | |
| "grad_norm": 0.14202447905316087, | |
| "learning_rate": 9.945079316809585e-05, | |
| "loss": 0.0368, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.21296861819635188, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.0699, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.295, | |
| "grad_norm": 0.17479932116014785, | |
| "learning_rate": 9.931428007686158e-05, | |
| "loss": 0.0383, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.19384996245201125, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.0977, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.305, | |
| "grad_norm": 0.3132458011578011, | |
| "learning_rate": 9.916274537819775e-05, | |
| "loss": 0.0783, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.24343342969724435, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.1309, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.315, | |
| "grad_norm": 0.06358039936658727, | |
| "learning_rate": 9.899623523104149e-05, | |
| "loss": 0.0341, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.2220768799042642, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.1244, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.09877556684729816, | |
| "learning_rate": 9.881480035599667e-05, | |
| "loss": 0.0584, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.07125822760195885, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.0099, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.335, | |
| "grad_norm": 0.19991568339646582, | |
| "learning_rate": 9.861849601988383e-05, | |
| "loss": 0.0532, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.18115871524580987, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.1011, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.345, | |
| "grad_norm": 0.07567960238386034, | |
| "learning_rate": 9.84073820189054e-05, | |
| "loss": 0.034, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.15451087209839287, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.0348, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.355, | |
| "grad_norm": 0.08313091042199042, | |
| "learning_rate": 9.818152266043114e-05, | |
| "loss": 0.0422, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.07302944049536657, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0417, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.365, | |
| "grad_norm": 0.041309166149277605, | |
| "learning_rate": 9.794098674340965e-05, | |
| "loss": 0.0231, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.05571619251756244, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.0517, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.07780175573372976, | |
| "learning_rate": 9.768584753741134e-05, | |
| "loss": 0.0312, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.046226872712960713, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.0418, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.385, | |
| "grad_norm": 0.08057703631440487, | |
| "learning_rate": 9.741618276030997e-05, | |
| "loss": 0.0419, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.05050489987091237, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.0163, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.395, | |
| "grad_norm": 0.0959178937528156, | |
| "learning_rate": 9.713207455460894e-05, | |
| "loss": 0.0428, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.08603826631317996, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0674, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.405, | |
| "grad_norm": 0.044173037654917646, | |
| "learning_rate": 9.683360946241989e-05, | |
| "loss": 0.0425, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.026290493271921827, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.0053, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.415, | |
| "grad_norm": 0.09325040701230163, | |
| "learning_rate": 9.652087839910124e-05, | |
| "loss": 0.0505, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.11764293119485195, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.0305, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.18654991334295415, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.011, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.08098006638815154, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.0183, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.435, | |
| "grad_norm": 0.07659797438155175, | |
| "learning_rate": 9.58530037192562e-05, | |
| "loss": 0.0274, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.16347900939631616, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.0859, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.445, | |
| "grad_norm": 0.15150817750296133, | |
| "learning_rate": 9.549806354382717e-05, | |
| "loss": 0.0333, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.07123424851384792, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.0405, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.455, | |
| "grad_norm": 0.11026163678942051, | |
| "learning_rate": 9.512926421749304e-05, | |
| "loss": 0.0462, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.07331190265762698, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.0069, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.465, | |
| "grad_norm": 0.06418964403024571, | |
| "learning_rate": 9.474671808010126e-05, | |
| "loss": 0.0441, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.04366811050044381, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.0321, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.06891688314378318, | |
| "learning_rate": 9.435054165891109e-05, | |
| "loss": 0.0197, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0791719137172165, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0211, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.485, | |
| "grad_norm": 0.09055479977019616, | |
| "learning_rate": 9.394085563309827e-05, | |
| "loss": 0.0368, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.1078342902713965, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.0402, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.495, | |
| "grad_norm": 0.04234438785284557, | |
| "learning_rate": 9.351778479699499e-05, | |
| "loss": 0.0155, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.32259634488597166, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.0234, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.505, | |
| "grad_norm": 0.057346774722805624, | |
| "learning_rate": 9.308145802207629e-05, | |
| "loss": 0.0057, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.06845584491597821, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.0239, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.515, | |
| "grad_norm": 0.027458411814212238, | |
| "learning_rate": 9.263200821770461e-05, | |
| "loss": 0.0046, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.04119439564225103, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.0067, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.25272039680541203, | |
| "learning_rate": 9.21695722906443e-05, | |
| "loss": 0.1068, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.15642301404795694, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 0.0487, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.535, | |
| "grad_norm": 0.020234092439841273, | |
| "learning_rate": 9.169429110335841e-05, | |
| "loss": 0.0044, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.021594134995858247, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.0023, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.545, | |
| "grad_norm": 0.11956523872488599, | |
| "learning_rate": 9.120630943110077e-05, | |
| "loss": 0.0647, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.06453797114816638, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.0378, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.555, | |
| "grad_norm": 0.040619091199357006, | |
| "learning_rate": 9.070577591781597e-05, | |
| "loss": 0.0101, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.012631286216702108, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.0025, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.565, | |
| "grad_norm": 0.11250625624010586, | |
| "learning_rate": 9.019284303086087e-05, | |
| "loss": 0.0326, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.008498513085188694, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 0.0019, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.06549524971538588, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 0.0144, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.1243637807598006, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.0427, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.585, | |
| "grad_norm": 0.4551297436540559, | |
| "learning_rate": 8.91304078426207e-05, | |
| "loss": 0.1364, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.14895611964997935, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.0516, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.595, | |
| "grad_norm": 0.03168114639692374, | |
| "learning_rate": 8.858122916938601e-05, | |
| "loss": 0.0038, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.08370376943417981, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.0068, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.605, | |
| "grad_norm": 0.07476474940113496, | |
| "learning_rate": 8.802029828000156e-05, | |
| "loss": 0.0182, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.13159976173852048, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 0.0352, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.615, | |
| "grad_norm": 0.022707973634618142, | |
| "learning_rate": 8.744778603945011e-05, | |
| "loss": 0.0033, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.12482944752372985, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.0202, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.39302174930812633, | |
| "learning_rate": 8.68638668405062e-05, | |
| "loss": 0.0758, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.19383130479510832, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 0.0325, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.635, | |
| "grad_norm": 0.03039287191356843, | |
| "learning_rate": 8.626871855061438e-05, | |
| "loss": 0.0041, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.048215778020076296, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0273, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.645, | |
| "grad_norm": 0.1392522953938403, | |
| "learning_rate": 8.566252245770909e-05, | |
| "loss": 0.0622, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.07411597583466574, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.0116, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.655, | |
| "grad_norm": 0.08394509913329051, | |
| "learning_rate": 8.504546321499255e-05, | |
| "loss": 0.0454, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.04530702990326196, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 0.0185, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.665, | |
| "grad_norm": 0.03765639754984064, | |
| "learning_rate": 8.44177287846877e-05, | |
| "loss": 0.0065, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.031586807052073985, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 0.0064, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.08297489591002068, | |
| "learning_rate": 8.377951038078302e-05, | |
| "loss": 0.0086, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.03562288518144821, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.0117, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.685, | |
| "grad_norm": 0.031042156897553914, | |
| "learning_rate": 8.313100241078689e-05, | |
| "loss": 0.0032, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.02483745350649097, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 0.0116, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.695, | |
| "grad_norm": 0.009338647544721107, | |
| "learning_rate": 8.247240241650918e-05, | |
| "loss": 0.0015, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.34270075427008806, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.0789, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.705, | |
| "grad_norm": 0.024842574848788327, | |
| "learning_rate": 8.18039110138882e-05, | |
| "loss": 0.0039, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.12421960639233866, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 0.0398, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.715, | |
| "grad_norm": 0.056187242189785436, | |
| "learning_rate": 8.112573183188099e-05, | |
| "loss": 0.0103, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.023766584635126216, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.0049, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.04512771394673685, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 0.0219, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.0274526387971708, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 0.0033, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.735, | |
| "grad_norm": 0.19532934746359332, | |
| "learning_rate": 7.974113933756707e-05, | |
| "loss": 0.0734, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.032852874131568864, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.0042, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.745, | |
| "grad_norm": 0.0640117280012966, | |
| "learning_rate": 7.903514778554699e-05, | |
| "loss": 0.0156, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.047410567888246616, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 0.0074, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.755, | |
| "grad_norm": 0.16842524790743002, | |
| "learning_rate": 7.832031184624164e-05, | |
| "loss": 0.0398, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.006147996063681932, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.001, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.765, | |
| "grad_norm": 0.04008165019001057, | |
| "learning_rate": 7.75968492656029e-05, | |
| "loss": 0.0059, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.007292282430214955, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 0.0012, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.14138265996109384, | |
| "learning_rate": 7.68649804173412e-05, | |
| "loss": 0.0259, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.07468592471462034, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 0.0192, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.785, | |
| "grad_norm": 0.014395933526058188, | |
| "learning_rate": 7.612492823579745e-05, | |
| "loss": 0.0025, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.0038502942774920196, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 0.0005, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.795, | |
| "grad_norm": 0.10226426219442211, | |
| "learning_rate": 7.537691814803521e-05, | |
| "loss": 0.0123, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.012762312166514576, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0019, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.805, | |
| "grad_norm": 0.01630848631501936, | |
| "learning_rate": 7.462117800517336e-05, | |
| "loss": 0.0019, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.09121920855984286, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 0.0252, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.815, | |
| "grad_norm": 0.044779381951263734, | |
| "learning_rate": 7.385793801298042e-05, | |
| "loss": 0.0049, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0069701432925383874, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 0.0005, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.04833507572943274, | |
| "learning_rate": 7.308743066175172e-05, | |
| "loss": 0.0127, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.001782479532615867, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 0.0003, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.835, | |
| "grad_norm": 0.1033970098216112, | |
| "learning_rate": 7.230989065549044e-05, | |
| "loss": 0.0331, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.050963861639285846, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.0078, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.845, | |
| "grad_norm": 0.025336676884289144, | |
| "learning_rate": 7.152555484041476e-05, | |
| "loss": 0.0037, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.04631516809520592, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 0.0026, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.855, | |
| "grad_norm": 0.024526402505777593, | |
| "learning_rate": 7.073466213281196e-05, | |
| "loss": 0.0032, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.12794480290400578, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 0.0286, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.865, | |
| "grad_norm": 0.05388064509201144, | |
| "learning_rate": 6.993745344626231e-05, | |
| "loss": 0.019, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.029754195193650093, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 0.0127, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.11351797783250586, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0276, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.04238335598804278, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.0044, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.885, | |
| "grad_norm": 0.07192092013374167, | |
| "learning_rate": 6.832506133621487e-05, | |
| "loss": 0.0033, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.086298683194853, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 0.0255, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.895, | |
| "grad_norm": 0.010115510174389889, | |
| "learning_rate": 6.751036906297337e-05, | |
| "loss": 0.0005, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.11331127094042127, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 0.0334, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.905, | |
| "grad_norm": 0.00359905539491798, | |
| "learning_rate": 6.669034296168855e-05, | |
| "loss": 0.0005, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.003786666725911578, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 0.0004, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.915, | |
| "grad_norm": 0.1178858284146152, | |
| "learning_rate": 6.586523282025462e-05, | |
| "loss": 0.0059, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0028456285153728687, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.0004, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.005713089625735049, | |
| "learning_rate": 6.503528997521366e-05, | |
| "loss": 0.0012, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.0030406845588623957, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 0.0003, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.935, | |
| "grad_norm": 0.0028405909976545836, | |
| "learning_rate": 6.420076723519614e-05, | |
| "loss": 0.0003, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.2531768340842614, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 0.0043, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.945, | |
| "grad_norm": 0.05452635559394549, | |
| "learning_rate": 6.336191880391284e-05, | |
| "loss": 0.0288, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.009714923881054332, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 0.0012, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.955, | |
| "grad_norm": 0.004170292898958103, | |
| "learning_rate": 6.251900020272208e-05, | |
| "loss": 0.0004, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.06545271449024238, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.0316, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.965, | |
| "grad_norm": 0.002583766103443765, | |
| "learning_rate": 6.167226819279528e-05, | |
| "loss": 0.0005, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.05335515958281829, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 0.0248, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.11111023851134917, | |
| "learning_rate": 6.0821980696905146e-05, | |
| "loss": 0.015, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0237783109632992, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 0.0053, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.985, | |
| "grad_norm": 0.04143947097844851, | |
| "learning_rate": 5.9968396720859864e-05, | |
| "loss": 0.0051, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.02380124179400986, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 0.0035, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.995, | |
| "grad_norm": 0.01783821356262124, | |
| "learning_rate": 5.911177627460739e-05, | |
| "loss": 0.0019, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.012578119490010059, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.0009, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.07603452354669571, | |
| "eval_runtime": 0.4047, | |
| "eval_samples_per_second": 2.471, | |
| "eval_steps_per_second": 2.471, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.005, | |
| "grad_norm": 0.014839351675487417, | |
| "learning_rate": 5.8252380293033884e-05, | |
| "loss": 0.0006, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.012314267589470526, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 0.0029, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.015, | |
| "grad_norm": 0.003125090751647675, | |
| "learning_rate": 5.7390470556480545e-05, | |
| "loss": 0.0004, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.007073162659384819, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 0.001, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.025, | |
| "grad_norm": 0.31379793587721094, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 0.0057, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.0011295952381376085, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 0.0001, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.035, | |
| "grad_norm": 0.0046728430974052725, | |
| "learning_rate": 5.566016068839535e-05, | |
| "loss": 0.0003, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.0055708348232121965, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.0006, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.045, | |
| "grad_norm": 0.0465350248080129, | |
| "learning_rate": 5.4792287626011204e-05, | |
| "loss": 0.0012, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.0008151862390823683, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 0.0001, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.055, | |
| "grad_norm": 0.0017764023093753461, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 0.0003, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.028033413085950722, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 0.0065, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.065, | |
| "grad_norm": 0.00419815372280561, | |
| "learning_rate": 5.3052426976742855e-05, | |
| "loss": 0.0004, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.0102127421097092, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 0.0021, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.075, | |
| "grad_norm": 0.0015729817142162348, | |
| "learning_rate": 5.218096936826681e-05, | |
| "loss": 0.0003, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.004004133543726314, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.0002, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.085, | |
| "grad_norm": 0.0012454184210680926, | |
| "learning_rate": 5.1308847415393666e-05, | |
| "loss": 0.0002, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.0005409115041498422, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 0.0001, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.095, | |
| "grad_norm": 0.001726174471896587, | |
| "learning_rate": 5.04363267749187e-05, | |
| "loss": 0.0002, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.0010814524798174311, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0001, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.105, | |
| "grad_norm": 0.009560946952238247, | |
| "learning_rate": 4.9563673225081314e-05, | |
| "loss": 0.0008, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.01441628431952276, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 0.0009, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.115, | |
| "grad_norm": 0.0023532660086251268, | |
| "learning_rate": 4.869115258460635e-05, | |
| "loss": 0.0003, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.057060873899210225, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.0041, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 0.020282977300679454, | |
| "learning_rate": 4.781903063173321e-05, | |
| "loss": 0.002, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.02046603304331513, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 0.0075, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.135, | |
| "grad_norm": 0.013990089439389753, | |
| "learning_rate": 4.694757302325715e-05, | |
| "loss": 0.0014, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 0.211826851392036, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 0.0054, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.145, | |
| "grad_norm": 0.0064942039937938645, | |
| "learning_rate": 4.607704521360776e-05, | |
| "loss": 0.0011, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.028422826475287776, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 0.0109, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.155, | |
| "grad_norm": 0.0002586519968099824, | |
| "learning_rate": 4.52077123739888e-05, | |
| "loss": 0.0, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.13429703504492657, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.0067, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.165, | |
| "grad_norm": 0.0009260106161090144, | |
| "learning_rate": 4.433983931160467e-05, | |
| "loss": 0.0001, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.2565611405752989, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 0.0069, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.175, | |
| "grad_norm": 0.026210448606146238, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 0.002, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.0012302259036583447, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 0.0002, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.185, | |
| "grad_norm": 0.02625494458907875, | |
| "learning_rate": 4.260952944351947e-05, | |
| "loss": 0.0011, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.000887755762843119, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 0.0001, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.195, | |
| "grad_norm": 0.003911862624404058, | |
| "learning_rate": 4.174761970696612e-05, | |
| "loss": 0.0002, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.03517802115885231, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.0019, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.205, | |
| "grad_norm": 0.020725215351612047, | |
| "learning_rate": 4.088822372539263e-05, | |
| "loss": 0.001, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.0002202527290049613, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 0.0, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.215, | |
| "grad_norm": 0.004526177203391082, | |
| "learning_rate": 4.003160327914015e-05, | |
| "loss": 0.0005, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.018305572530131792, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 0.0005, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.225, | |
| "grad_norm": 0.0007986471410436935, | |
| "learning_rate": 3.917801930309486e-05, | |
| "loss": 0.0001, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.0259210834529961, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 0.0015, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.2349999999999999, | |
| "grad_norm": 0.007431515309959642, | |
| "learning_rate": 3.832773180720475e-05, | |
| "loss": 0.0008, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.21370639527155416, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.0296, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.245, | |
| "grad_norm": 0.0017749674756192052, | |
| "learning_rate": 3.748099979727792e-05, | |
| "loss": 0.0002, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.06429161818678288, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.004, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.255, | |
| "grad_norm": 0.4999355458993204, | |
| "learning_rate": 3.663808119608716e-05, | |
| "loss": 0.0118, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.004634517265489021, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 0.0002, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.2650000000000001, | |
| "grad_norm": 0.10035773018671189, | |
| "learning_rate": 3.579923276480387e-05, | |
| "loss": 0.0132, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.014262309364614909, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 0.0006, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.275, | |
| "grad_norm": 0.0006263407491042736, | |
| "learning_rate": 3.4964710024786354e-05, | |
| "loss": 0.0001, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.0018123347337400675, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.0003, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.285, | |
| "grad_norm": 0.0180314931604898, | |
| "learning_rate": 3.4134767179745406e-05, | |
| "loss": 0.0056, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.005834199925826024, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 0.0007, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.295, | |
| "grad_norm": 0.008235805677850457, | |
| "learning_rate": 3.330965703831146e-05, | |
| "loss": 0.0014, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.01238354074430894, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.0027, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.305, | |
| "grad_norm": 0.01737694828749958, | |
| "learning_rate": 3.248963093702663e-05, | |
| "loss": 0.0022, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.0019556398997595315, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 0.0002, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.315, | |
| "grad_norm": 0.011691972118475544, | |
| "learning_rate": 3.167493866378514e-05, | |
| "loss": 0.0009, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.004834944193413171, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.0004, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.325, | |
| "grad_norm": 0.00037022329873905106, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.017585669474956853, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 0.0034, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.335, | |
| "grad_norm": 0.0015713497036280754, | |
| "learning_rate": 3.006254655373769e-05, | |
| "loss": 0.0002, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.006894231638829978, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.0012, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.345, | |
| "grad_norm": 0.0024889803245449374, | |
| "learning_rate": 2.926533786718806e-05, | |
| "loss": 0.0003, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.196326209710552, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 0.0108, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.355, | |
| "grad_norm": 0.0006519053745859627, | |
| "learning_rate": 2.8474445159585235e-05, | |
| "loss": 0.0001, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.017091849830244706, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.0017, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.365, | |
| "grad_norm": 0.000303651475823324, | |
| "learning_rate": 2.7690109344509563e-05, | |
| "loss": 0.0001, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.026097473151970325, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 0.0078, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.00044755689357070186, | |
| "learning_rate": 2.6912569338248315e-05, | |
| "loss": 0.0001, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.0005042582378594239, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 0.0, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.385, | |
| "grad_norm": 0.003238731213585818, | |
| "learning_rate": 2.6142061987019577e-05, | |
| "loss": 0.0001, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.3900000000000001, | |
| "grad_norm": 0.014755686528500189, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 0.0013, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.395, | |
| "grad_norm": 0.00553162806351944, | |
| "learning_rate": 2.537882199482665e-05, | |
| "loss": 0.0005, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.0005184577628493705, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0001, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.405, | |
| "grad_norm": 0.0037574278569336966, | |
| "learning_rate": 2.4623081851964806e-05, | |
| "loss": 0.0002, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.0013208423085230652, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 0.0001, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.415, | |
| "grad_norm": 0.008749035967488314, | |
| "learning_rate": 2.3875071764202563e-05, | |
| "loss": 0.0006, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.1377433041169159, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 0.0129, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.425, | |
| "grad_norm": 0.0010673749455521134, | |
| "learning_rate": 2.3135019582658802e-05, | |
| "loss": 0.0001, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.0014952971304103484, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 0.0001, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.435, | |
| "grad_norm": 0.0037047451799289005, | |
| "learning_rate": 2.2403150734397094e-05, | |
| "loss": 0.0003, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.0016251406693602976, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 0.0002, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.445, | |
| "grad_norm": 0.00028939999776533094, | |
| "learning_rate": 2.167968815375837e-05, | |
| "loss": 0.0, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.001349118937608808, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 0.0001, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.455, | |
| "grad_norm": 0.0067421886978353855, | |
| "learning_rate": 2.0964852214453013e-05, | |
| "loss": 0.0008, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.0011089997483127705, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.0002, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.465, | |
| "grad_norm": 0.0028423391897432755, | |
| "learning_rate": 2.0258860662432942e-05, | |
| "loss": 0.0002, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.01152426090888302, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 0.0005, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.475, | |
| "grad_norm": 0.0674872940289676, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 0.0089, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.0014300801418590321, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 0.0001, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.4849999999999999, | |
| "grad_norm": 0.0009000382689770731, | |
| "learning_rate": 1.887426816811903e-05, | |
| "loss": 0.0001, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.0037992575880660263, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 0.0005, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.495, | |
| "grad_norm": 0.22933461676514044, | |
| "learning_rate": 1.8196088986111797e-05, | |
| "loss": 0.0094, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.042524052580419906, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 0.0092, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.505, | |
| "grad_norm": 0.0023691615908781444, | |
| "learning_rate": 1.7527597583490822e-05, | |
| "loss": 0.0001, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.019487602890200652, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 0.0033, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.5150000000000001, | |
| "grad_norm": 0.0016513291656156698, | |
| "learning_rate": 1.6868997589213136e-05, | |
| "loss": 0.0003, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.0014481350536131741, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.0002, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.525, | |
| "grad_norm": 0.030859003830790453, | |
| "learning_rate": 1.622048961921699e-05, | |
| "loss": 0.0047, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.0013227578422165872, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 0.0002, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.5350000000000001, | |
| "grad_norm": 0.040650034143242364, | |
| "learning_rate": 1.5582271215312294e-05, | |
| "loss": 0.0058, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.010752068089793936, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 0.0004, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.545, | |
| "grad_norm": 0.0007336372616353323, | |
| "learning_rate": 1.4954536785007456e-05, | |
| "loss": 0.0001, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.0006185076286575476, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0001, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.5550000000000002, | |
| "grad_norm": 0.0012108807503260438, | |
| "learning_rate": 1.4337477542290928e-05, | |
| "loss": 0.0001, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.014653087670058953, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 0.0038, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.565, | |
| "grad_norm": 0.00070106922262732, | |
| "learning_rate": 1.373128144938563e-05, | |
| "loss": 0.0001, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.5699999999999998, | |
| "grad_norm": 0.022894525200356754, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 0.0034, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.575, | |
| "grad_norm": 0.15319546037835274, | |
| "learning_rate": 1.3136133159493802e-05, | |
| "loss": 0.005, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.02298142323498265, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 0.0036, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.585, | |
| "grad_norm": 0.000645097633470432, | |
| "learning_rate": 1.2552213960549891e-05, | |
| "loss": 0.0001, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.5899999999999999, | |
| "grad_norm": 0.08053780392470371, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 0.0249, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.595, | |
| "grad_norm": 0.00023214176871761666, | |
| "learning_rate": 1.1979701719998453e-05, | |
| "loss": 0.0, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.0008201475435205818, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.0001, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.605, | |
| "grad_norm": 0.00476872158921415, | |
| "learning_rate": 1.1418770830614013e-05, | |
| "loss": 0.0002, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.6099999999999999, | |
| "grad_norm": 0.001393468029081844, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 0.0001, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.615, | |
| "grad_norm": 0.003859464277860662, | |
| "learning_rate": 1.0869592157379304e-05, | |
| "loss": 0.0001, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.004353663579813155, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 0.0003, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 0.013200524828428624, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 0.0018, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.0024888356543417843, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 0.0001, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.635, | |
| "grad_norm": 0.02372982952309258, | |
| "learning_rate": 9.807156969139136e-06, | |
| "loss": 0.0031, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.0227337977382863, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.0032, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.645, | |
| "grad_norm": 0.0006468911312178352, | |
| "learning_rate": 9.294224082184045e-06, | |
| "loss": 0.0001, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.00031790071448860156, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.655, | |
| "grad_norm": 0.014069307621453687, | |
| "learning_rate": 8.793690568899216e-06, | |
| "loss": 0.001, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.6600000000000001, | |
| "grad_norm": 0.002913520292376533, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 0.0001, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.665, | |
| "grad_norm": 0.010269178065459858, | |
| "learning_rate": 8.305708896641594e-06, | |
| "loss": 0.0008, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.004202310016346467, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 0.0002, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.675, | |
| "grad_norm": 0.06791650915734657, | |
| "learning_rate": 7.830427709355725e-06, | |
| "loss": 0.0089, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.00023193543187296342, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 0.0, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.685, | |
| "grad_norm": 0.009522222092425871, | |
| "learning_rate": 7.367991782295391e-06, | |
| "loss": 0.0005, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.0445881491075994, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 0.0016, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.6949999999999998, | |
| "grad_norm": 0.000599146744339273, | |
| "learning_rate": 6.918541977923709e-06, | |
| "loss": 0.0001, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.027249503153038576, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.0021, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.705, | |
| "grad_norm": 0.002016792070660535, | |
| "learning_rate": 6.482215203005015e-06, | |
| "loss": 0.0002, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.007484120928358062, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 0.0003, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.7149999999999999, | |
| "grad_norm": 0.0005063050902788136, | |
| "learning_rate": 6.059144366901736e-06, | |
| "loss": 0.0, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.0009197549253660609, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 0.0001, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.725, | |
| "grad_norm": 0.003191114724532278, | |
| "learning_rate": 5.649458341088915e-06, | |
| "loss": 0.0005, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.0014447492382083482, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 0.0002, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.7349999999999999, | |
| "grad_norm": 0.004059252437141361, | |
| "learning_rate": 5.2532819198987506e-06, | |
| "loss": 0.0006, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.02120955388832761, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 0.0007, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.745, | |
| "grad_norm": 0.0011651360195809543, | |
| "learning_rate": 4.87073578250698e-06, | |
| "loss": 0.0001, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.0010159633731216518, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 0.0001, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.755, | |
| "grad_norm": 0.04624949210450388, | |
| "learning_rate": 4.501936456172845e-06, | |
| "loss": 0.0023, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.013865761867621462, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 0.0008, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.7650000000000001, | |
| "grad_norm": 0.0006593659250274612, | |
| "learning_rate": 4.146996280743798e-06, | |
| "loss": 0.0001, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.0006699609878223177, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 0.0001, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.775, | |
| "grad_norm": 0.010073962298842973, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0014, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.00046017499160588973, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 0.0001, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.7850000000000001, | |
| "grad_norm": 0.001412185181637751, | |
| "learning_rate": 3.479121600898777e-06, | |
| "loss": 0.0001, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.0010948577791496526, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 0.0001, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.795, | |
| "grad_norm": 0.0027495650503305998, | |
| "learning_rate": 3.1663905375801216e-06, | |
| "loss": 0.0003, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.00497037863878421, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.0004, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8050000000000002, | |
| "grad_norm": 0.04826775060358748, | |
| "learning_rate": 2.8679254453910785e-06, | |
| "loss": 0.0052, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.0002190733399899643, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 0.0, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.815, | |
| "grad_norm": 0.03983428885724064, | |
| "learning_rate": 2.583817239690034e-06, | |
| "loss": 0.0108, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.8199999999999998, | |
| "grad_norm": 0.005353636861069711, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.0004, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.825, | |
| "grad_norm": 0.007365720628728885, | |
| "learning_rate": 2.314152462588659e-06, | |
| "loss": 0.0009, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.0020943736140563596, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 0.0002, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.835, | |
| "grad_norm": 0.049181169958885725, | |
| "learning_rate": 2.0590132565903476e-06, | |
| "loss": 0.0066, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.008505272014854959, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 0.001, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.845, | |
| "grad_norm": 0.015026960520249318, | |
| "learning_rate": 1.8184773395688526e-06, | |
| "loss": 0.0027, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.0012138520031968582, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 0.0002, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.855, | |
| "grad_norm": 0.00024125978466275796, | |
| "learning_rate": 1.5926179810946184e-06, | |
| "loss": 0.0001, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.8599999999999999, | |
| "grad_norm": 0.07024226815561596, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 0.004, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.865, | |
| "grad_norm": 0.0021247803242464055, | |
| "learning_rate": 1.3815039801161721e-06, | |
| "loss": 0.0002, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.04502769124604199, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 0.0122, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.007452172380573964, | |
| "learning_rate": 1.1851996440033319e-06, | |
| "loss": 0.001, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.0015481238834281481, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.0002, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.885, | |
| "grad_norm": 0.005530884948723974, | |
| "learning_rate": 1.0037647689585206e-06, | |
| "loss": 0.0009, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.8900000000000001, | |
| "grad_norm": 0.0067523371656823305, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 0.0005, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.895, | |
| "grad_norm": 0.003506175781494403, | |
| "learning_rate": 8.372546218022747e-07, | |
| "loss": 0.0002, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.0019742849650606926, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 0.0003, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.905, | |
| "grad_norm": 0.0009738360342464184, | |
| "learning_rate": 6.857199231384282e-07, | |
| "loss": 0.0001, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.9100000000000001, | |
| "grad_norm": 0.0004408482841272174, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 0.0, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.915, | |
| "grad_norm": 0.010244955433424012, | |
| "learning_rate": 5.492068319041588e-07, | |
| "loss": 0.0013, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.0521719111078745, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 0.0071, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.925, | |
| "grad_norm": 0.0012864130857160657, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 0.0002, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.9300000000000002, | |
| "grad_norm": 0.2423177287343396, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 0.028, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.935, | |
| "grad_norm": 0.010456316918486782, | |
| "learning_rate": 3.214072161706272e-07, | |
| "loss": 0.0007, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.0005706094269043214, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 0.0001, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.9449999999999998, | |
| "grad_norm": 0.013608987991299646, | |
| "learning_rate": 2.3019008164105738e-07, | |
| "loss": 0.0026, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.01259450581057638, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 0.0006, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.955, | |
| "grad_norm": 0.006309175090935663, | |
| "learning_rate": 1.5413331334360182e-07, | |
| "loss": 0.0006, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.00027718336636984805, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 0.0001, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.9649999999999999, | |
| "grad_norm": 0.01973445145502157, | |
| "learning_rate": 9.3260078906654e-08, | |
| "loss": 0.0058, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.0050223114988680664, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 0.0007, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.975, | |
| "grad_norm": 0.041307491240042314, | |
| "learning_rate": 4.7588920907110094e-08, | |
| "loss": 0.004, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.0019487309746442478, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 0.0001, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.9849999999999999, | |
| "grad_norm": 0.00051503727945184, | |
| "learning_rate": 1.7133751222137007e-08, | |
| "loss": 0.0001, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.02902500982586163, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 0.003, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.995, | |
| "grad_norm": 0.02095204026685112, | |
| "learning_rate": 1.903846791434516e-09, | |
| "loss": 0.0024, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.006401152408197018, | |
| "learning_rate": 0.0, | |
| "loss": 0.0004, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.0046582599170506, | |
| "eval_runtime": 1.5245, | |
| "eval_samples_per_second": 0.656, | |
| "eval_steps_per_second": 0.656, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 400, | |
| "total_flos": 6612300644352.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 1.3861, | |
| "train_samples_per_second": 577.159, | |
| "train_steps_per_second": 288.579 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 6612300644352.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |