| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0035714285714285713, | |
| "grad_norm": 2.8542327447781655, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.2459, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007142857142857143, | |
| "grad_norm": 3.2008343511486888, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.2209, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010714285714285714, | |
| "grad_norm": 2.5713840881424925, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.2089, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014285714285714285, | |
| "grad_norm": 1.955191718766848, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.0984, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 2.117930426021548, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.001, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02142857142857143, | |
| "grad_norm": 5.368009843070649, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.9729, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 2.6030409421004315, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 0.9481, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 2.4961632055644256, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 0.8393, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03214285714285714, | |
| "grad_norm": 2.366915137111893, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7796, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 2.323220568477207, | |
| "learning_rate": 1.9999328066483867e-05, | |
| "loss": 0.7788, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.039285714285714285, | |
| "grad_norm": 1.936852006950086, | |
| "learning_rate": 1.9997312356234385e-05, | |
| "loss": 0.7202, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04285714285714286, | |
| "grad_norm": 2.0208748078757965, | |
| "learning_rate": 1.999395314013622e-05, | |
| "loss": 0.5943, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04642857142857143, | |
| "grad_norm": 2.0052413798661766, | |
| "learning_rate": 1.998925086962334e-05, | |
| "loss": 0.5392, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.1438278959706687, | |
| "learning_rate": 1.998320617661839e-05, | |
| "loss": 0.5334, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 1.7068775254192634, | |
| "learning_rate": 1.997581987344772e-05, | |
| "loss": 0.5014, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 1.9194172128721236, | |
| "learning_rate": 1.9967092952732266e-05, | |
| "loss": 0.4964, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.060714285714285714, | |
| "grad_norm": 1.858301483386355, | |
| "learning_rate": 1.9957026587254136e-05, | |
| "loss": 0.4843, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06428571428571428, | |
| "grad_norm": 1.608800232721517, | |
| "learning_rate": 1.9945622129799e-05, | |
| "loss": 0.4586, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06785714285714285, | |
| "grad_norm": 1.6662313175415042, | |
| "learning_rate": 1.9932881112974298e-05, | |
| "loss": 0.4879, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 1.743091779375127, | |
| "learning_rate": 1.9918805249003272e-05, | |
| "loss": 0.4174, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 1.7071719867098498, | |
| "learning_rate": 1.9903396429494882e-05, | |
| "loss": 0.423, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07857142857142857, | |
| "grad_norm": 1.4872510145523106, | |
| "learning_rate": 1.9886656725189575e-05, | |
| "loss": 0.4247, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08214285714285714, | |
| "grad_norm": 1.4662548028627673, | |
| "learning_rate": 1.9868588385681035e-05, | |
| "loss": 0.353, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 1.2910547643569992, | |
| "learning_rate": 1.9849193839113833e-05, | |
| "loss": 0.2913, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 1.2204351230084813, | |
| "learning_rate": 1.9828475691857148e-05, | |
| "loss": 0.2707, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09285714285714286, | |
| "grad_norm": 1.3717201569579562, | |
| "learning_rate": 1.9806436728154484e-05, | |
| "loss": 0.3425, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09642857142857143, | |
| "grad_norm": 1.17629244354678, | |
| "learning_rate": 1.9783079909749516e-05, | |
| "loss": 0.2771, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1945359382636211, | |
| "learning_rate": 1.975840837548807e-05, | |
| "loss": 0.3211, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10357142857142858, | |
| "grad_norm": 1.3376592216303969, | |
| "learning_rate": 1.9732425440896298e-05, | |
| "loss": 0.3301, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 1.3450447765698734, | |
| "learning_rate": 1.9705134597735113e-05, | |
| "loss": 0.3195, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11071428571428571, | |
| "grad_norm": 0.9886589945902243, | |
| "learning_rate": 1.9676539513530967e-05, | |
| "loss": 0.246, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 1.2275546257818077, | |
| "learning_rate": 1.9646644031082948e-05, | |
| "loss": 0.3145, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11785714285714285, | |
| "grad_norm": 1.3539593128551837, | |
| "learning_rate": 1.9615452167946383e-05, | |
| "loss": 0.3316, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.12142857142857143, | |
| "grad_norm": 1.283776358581948, | |
| "learning_rate": 1.958296811589293e-05, | |
| "loss": 0.2939, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 1.7849067225456108, | |
| "learning_rate": 1.954919624034725e-05, | |
| "loss": 0.3578, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12857142857142856, | |
| "grad_norm": 1.0954369777387916, | |
| "learning_rate": 1.951414107980036e-05, | |
| "loss": 0.265, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13214285714285715, | |
| "grad_norm": 1.1139647009348417, | |
| "learning_rate": 1.9477807345199717e-05, | |
| "loss": 0.2516, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1357142857142857, | |
| "grad_norm": 1.0058557775589556, | |
| "learning_rate": 1.9440199919316125e-05, | |
| "loss": 0.2823, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1392857142857143, | |
| "grad_norm": 1.3874195280252404, | |
| "learning_rate": 1.9401323856087573e-05, | |
| "loss": 0.2589, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 1.1059724872637073, | |
| "learning_rate": 1.936118437994003e-05, | |
| "loss": 0.3127, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14642857142857144, | |
| "grad_norm": 1.4336617434314778, | |
| "learning_rate": 1.9319786885085366e-05, | |
| "loss": 0.2788, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.184578032516573, | |
| "learning_rate": 1.927713693479643e-05, | |
| "loss": 0.2891, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.15357142857142858, | |
| "grad_norm": 1.1818501699215336, | |
| "learning_rate": 1.923324026065944e-05, | |
| "loss": 0.2951, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.15714285714285714, | |
| "grad_norm": 1.0975108683871662, | |
| "learning_rate": 1.918810276180372e-05, | |
| "loss": 0.2184, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 1.1218315646897867, | |
| "learning_rate": 1.9141730504108923e-05, | |
| "loss": 0.2297, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16428571428571428, | |
| "grad_norm": 1.056967251433103, | |
| "learning_rate": 1.9094129719389886e-05, | |
| "loss": 0.2118, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16785714285714284, | |
| "grad_norm": 1.0992433056470872, | |
| "learning_rate": 1.904530680455914e-05, | |
| "loss": 0.2941, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 1.233231717888253, | |
| "learning_rate": 1.8995268320767254e-05, | |
| "loss": 0.2458, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 1.063969135221994, | |
| "learning_rate": 1.894402099252109e-05, | |
| "loss": 0.2752, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 1.2463378053837864, | |
| "learning_rate": 1.889157170678015e-05, | |
| "loss": 0.2654, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18214285714285713, | |
| "grad_norm": 1.142139802192778, | |
| "learning_rate": 1.8837927512031022e-05, | |
| "loss": 0.2991, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.18571428571428572, | |
| "grad_norm": 1.3979732581315991, | |
| "learning_rate": 1.8783095617340193e-05, | |
| "loss": 0.2277, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.18928571428571428, | |
| "grad_norm": 0.945360454505982, | |
| "learning_rate": 1.872708339138522e-05, | |
| "loss": 0.1957, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19285714285714287, | |
| "grad_norm": 1.1124113891661973, | |
| "learning_rate": 1.866989836146449e-05, | |
| "loss": 0.254, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 1.223628442154201, | |
| "learning_rate": 1.861154821248565e-05, | |
| "loss": 0.2185, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.0447822547318244, | |
| "learning_rate": 1.8552040785932846e-05, | |
| "loss": 0.2167, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.20357142857142857, | |
| "grad_norm": 1.0215491299481245, | |
| "learning_rate": 1.849138407881296e-05, | |
| "loss": 0.2341, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.20714285714285716, | |
| "grad_norm": 1.04110956091982, | |
| "learning_rate": 1.8429586242580884e-05, | |
| "loss": 0.257, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21071428571428572, | |
| "grad_norm": 1.0970335540731302, | |
| "learning_rate": 1.8366655582044096e-05, | |
| "loss": 0.2612, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 1.5414206823739205, | |
| "learning_rate": 1.83026005542466e-05, | |
| "loss": 0.3208, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21785714285714286, | |
| "grad_norm": 1.1115383818307014, | |
| "learning_rate": 1.8237429767332407e-05, | |
| "loss": 0.286, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.22142857142857142, | |
| "grad_norm": 0.9397263766552314, | |
| "learning_rate": 1.8171151979388715e-05, | |
| "loss": 0.2484, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 1.0153689250082296, | |
| "learning_rate": 1.8103776097268942e-05, | |
| "loss": 0.2659, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 0.9284042621070391, | |
| "learning_rate": 1.803531117539577e-05, | |
| "loss": 0.1991, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 0.9342315130686223, | |
| "learning_rate": 1.7965766414544328e-05, | |
| "loss": 0.2032, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2357142857142857, | |
| "grad_norm": 1.072773106151265, | |
| "learning_rate": 1.7895151160605758e-05, | |
| "loss": 0.2448, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2392857142857143, | |
| "grad_norm": 1.1178318721781217, | |
| "learning_rate": 1.782347490333123e-05, | |
| "loss": 0.2496, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.24285714285714285, | |
| "grad_norm": 1.023124180885441, | |
| "learning_rate": 1.775074727505667e-05, | |
| "loss": 0.2168, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.24642857142857144, | |
| "grad_norm": 0.9996638909558939, | |
| "learning_rate": 1.7676978049408262e-05, | |
| "loss": 0.2362, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8713422474197986, | |
| "learning_rate": 1.7602177139989046e-05, | |
| "loss": 0.1897, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25357142857142856, | |
| "grad_norm": 1.0498129217931838, | |
| "learning_rate": 1.7526354599046637e-05, | |
| "loss": 0.2514, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 0.8958335737383101, | |
| "learning_rate": 1.7449520616122344e-05, | |
| "loss": 0.2137, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.26071428571428573, | |
| "grad_norm": 0.9491324981957817, | |
| "learning_rate": 1.7371685516681825e-05, | |
| "loss": 0.1908, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2642857142857143, | |
| "grad_norm": 0.8474993073986217, | |
| "learning_rate": 1.7292859760727493e-05, | |
| "loss": 0.2113, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 0.9763010793347767, | |
| "learning_rate": 1.721305394139282e-05, | |
| "loss": 0.1616, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2714285714285714, | |
| "grad_norm": 0.8756798542172813, | |
| "learning_rate": 1.7132278783518756e-05, | |
| "loss": 0.2389, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.84872169147024, | |
| "learning_rate": 1.7050545142212483e-05, | |
| "loss": 0.2337, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2785714285714286, | |
| "grad_norm": 0.8331599870610199, | |
| "learning_rate": 1.696786400138859e-05, | |
| "loss": 0.1975, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.28214285714285714, | |
| "grad_norm": 0.9769440268150504, | |
| "learning_rate": 1.6884246472293018e-05, | |
| "loss": 0.2096, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.8822887283843762, | |
| "learning_rate": 1.679970379200983e-05, | |
| "loss": 0.2178, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2892857142857143, | |
| "grad_norm": 1.040368028081877, | |
| "learning_rate": 1.6714247321951106e-05, | |
| "loss": 0.2668, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.29285714285714287, | |
| "grad_norm": 1.0378300096521034, | |
| "learning_rate": 1.6627888546330136e-05, | |
| "loss": 0.2742, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.29642857142857143, | |
| "grad_norm": 1.0081956925343618, | |
| "learning_rate": 1.654063907061807e-05, | |
| "loss": 0.2317, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.9295295553823995, | |
| "learning_rate": 1.64525106199843e-05, | |
| "loss": 0.2155, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 1.1067271100643354, | |
| "learning_rate": 1.6363515037720774e-05, | |
| "loss": 0.2173, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30714285714285716, | |
| "grad_norm": 0.8455456352551043, | |
| "learning_rate": 1.6273664283650393e-05, | |
| "loss": 0.2296, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3107142857142857, | |
| "grad_norm": 0.7511551059374534, | |
| "learning_rate": 1.6182970432519772e-05, | |
| "loss": 0.2117, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 1.27007129583062, | |
| "learning_rate": 1.609144567237658e-05, | |
| "loss": 0.2458, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.31785714285714284, | |
| "grad_norm": 0.8492452536821289, | |
| "learning_rate": 1.5999102302931585e-05, | |
| "loss": 0.2187, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 1.1577254689751963, | |
| "learning_rate": 1.5905952733905777e-05, | |
| "loss": 0.2655, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 1.0482100964602592, | |
| "learning_rate": 1.5812009483362643e-05, | |
| "loss": 0.308, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.32857142857142857, | |
| "grad_norm": 1.1600584410906734, | |
| "learning_rate": 1.5717285176025913e-05, | |
| "loss": 0.2034, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.33214285714285713, | |
| "grad_norm": 1.0208452713577167, | |
| "learning_rate": 1.5621792541582968e-05, | |
| "loss": 0.2886, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3357142857142857, | |
| "grad_norm": 0.949057228666076, | |
| "learning_rate": 1.552554441297413e-05, | |
| "loss": 0.2233, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 1.1702785572108525, | |
| "learning_rate": 1.5428553724668103e-05, | |
| "loss": 0.2553, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.8302656215189927, | |
| "learning_rate": 1.533083351092372e-05, | |
| "loss": 0.2366, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3464285714285714, | |
| "grad_norm": 0.7111981015865874, | |
| "learning_rate": 1.5232396904038352e-05, | |
| "loss": 0.1443, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.2280877546433842, | |
| "learning_rate": 1.5133257132583074e-05, | |
| "loss": 0.2611, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3535714285714286, | |
| "grad_norm": 0.8800901835863867, | |
| "learning_rate": 1.503342751962493e-05, | |
| "loss": 0.2847, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.9166060274088821, | |
| "learning_rate": 1.4932921480936491e-05, | |
| "loss": 0.1499, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3607142857142857, | |
| "grad_norm": 1.0017986138200043, | |
| "learning_rate": 1.4831752523192949e-05, | |
| "loss": 0.2172, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.36428571428571427, | |
| "grad_norm": 1.095933130496475, | |
| "learning_rate": 1.4729934242157005e-05, | |
| "loss": 0.2998, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3678571428571429, | |
| "grad_norm": 0.9962381045121098, | |
| "learning_rate": 1.4627480320851775e-05, | |
| "loss": 0.2737, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 0.8702970432647809, | |
| "learning_rate": 1.4524404527721977e-05, | |
| "loss": 0.2522, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.985504195424588, | |
| "learning_rate": 1.4420720714783635e-05, | |
| "loss": 0.2358, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.37857142857142856, | |
| "grad_norm": 1.2118284574664568, | |
| "learning_rate": 1.4316442815762543e-05, | |
| "loss": 0.2662, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3821428571428571, | |
| "grad_norm": 0.82084189003784, | |
| "learning_rate": 1.4211584844221771e-05, | |
| "loss": 0.1882, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.38571428571428573, | |
| "grad_norm": 0.7069507409286877, | |
| "learning_rate": 1.4106160891678422e-05, | |
| "loss": 0.1951, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3892857142857143, | |
| "grad_norm": 0.7689161582894709, | |
| "learning_rate": 1.4000185125709919e-05, | |
| "loss": 0.1725, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 0.7676679280673471, | |
| "learning_rate": 1.3893671788050073e-05, | |
| "loss": 0.1585, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3964285714285714, | |
| "grad_norm": 0.77471395404216, | |
| "learning_rate": 1.3786635192675184e-05, | |
| "loss": 0.2206, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.9984901548871592, | |
| "learning_rate": 1.3679089723880427e-05, | |
| "loss": 0.2617, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4035714285714286, | |
| "grad_norm": 1.0400909733079975, | |
| "learning_rate": 1.35710498343468e-05, | |
| "loss": 0.1769, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.40714285714285714, | |
| "grad_norm": 1.002292537016592, | |
| "learning_rate": 1.3462530043198874e-05, | |
| "loss": 0.2586, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 0.6551106931796162, | |
| "learning_rate": 1.3353544934053618e-05, | |
| "loss": 0.1673, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4142857142857143, | |
| "grad_norm": 0.9060227081683486, | |
| "learning_rate": 1.324410915306055e-05, | |
| "loss": 0.228, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.41785714285714287, | |
| "grad_norm": 0.785686533390065, | |
| "learning_rate": 1.3134237406933493e-05, | |
| "loss": 0.2083, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.42142857142857143, | |
| "grad_norm": 0.8497309507466154, | |
| "learning_rate": 1.3023944460974183e-05, | |
| "loss": 0.2009, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.8470251087419485, | |
| "learning_rate": 1.2913245137088024e-05, | |
| "loss": 0.2061, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.9589921778817803, | |
| "learning_rate": 1.2802154311792196e-05, | |
| "loss": 0.1537, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.43214285714285716, | |
| "grad_norm": 0.9947480480371786, | |
| "learning_rate": 1.2690686914216475e-05, | |
| "loss": 0.2773, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4357142857142857, | |
| "grad_norm": 0.8212329827096161, | |
| "learning_rate": 1.2578857924096935e-05, | |
| "loss": 0.2148, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4392857142857143, | |
| "grad_norm": 1.0285447504055023, | |
| "learning_rate": 1.2466682369762883e-05, | |
| "loss": 0.2163, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.44285714285714284, | |
| "grad_norm": 1.0445202414513237, | |
| "learning_rate": 1.2354175326117252e-05, | |
| "loss": 0.2626, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 0.9733002033825501, | |
| "learning_rate": 1.2241351912610726e-05, | |
| "loss": 0.3042, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.7648220446696808, | |
| "learning_rate": 1.212822729120989e-05, | |
| "loss": 0.228, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.45357142857142857, | |
| "grad_norm": 0.7866538639716522, | |
| "learning_rate": 1.2014816664359671e-05, | |
| "loss": 0.2181, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.8064363725412049, | |
| "learning_rate": 1.190113527294032e-05, | |
| "loss": 0.2415, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4607142857142857, | |
| "grad_norm": 0.7803983228981857, | |
| "learning_rate": 1.178719839421925e-05, | |
| "loss": 0.1576, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 0.9439425043263437, | |
| "learning_rate": 1.1673021339797967e-05, | |
| "loss": 0.2844, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.46785714285714286, | |
| "grad_norm": 0.8150618613103979, | |
| "learning_rate": 1.15586194535544e-05, | |
| "loss": 0.1969, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4714285714285714, | |
| "grad_norm": 0.7772114342406427, | |
| "learning_rate": 1.1444008109580884e-05, | |
| "loss": 0.1383, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 1.0738272204622934, | |
| "learning_rate": 1.1329202710118088e-05, | |
| "loss": 0.2481, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4785714285714286, | |
| "grad_norm": 0.7852428793449633, | |
| "learning_rate": 1.1214218683485159e-05, | |
| "loss": 0.2216, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 0.7378172524312823, | |
| "learning_rate": 1.1099071482006361e-05, | |
| "loss": 0.1732, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 0.7273259562797488, | |
| "learning_rate": 1.0983776579934483e-05, | |
| "loss": 0.1985, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.48928571428571427, | |
| "grad_norm": 1.0411454605240655, | |
| "learning_rate": 1.0868349471371316e-05, | |
| "loss": 0.3302, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4928571428571429, | |
| "grad_norm": 1.1149497038574403, | |
| "learning_rate": 1.0752805668185442e-05, | |
| "loss": 0.2199, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.49642857142857144, | |
| "grad_norm": 0.9853475126131891, | |
| "learning_rate": 1.0637160697927651e-05, | |
| "loss": 0.2238, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.7975162036724801, | |
| "learning_rate": 1.0521430101744238e-05, | |
| "loss": 0.2225, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5035714285714286, | |
| "grad_norm": 0.7778773518681362, | |
| "learning_rate": 1.040562943228849e-05, | |
| "loss": 0.2107, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5071428571428571, | |
| "grad_norm": 0.7016828398202349, | |
| "learning_rate": 1.0289774251630602e-05, | |
| "loss": 0.1646, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5107142857142857, | |
| "grad_norm": 0.8960386929571689, | |
| "learning_rate": 1.0173880129166358e-05, | |
| "loss": 0.2033, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 0.6604103212473279, | |
| "learning_rate": 1.0057962639524799e-05, | |
| "loss": 0.1605, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 1.383210847230254, | |
| "learning_rate": 9.942037360475205e-06, | |
| "loss": 0.2616, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5214285714285715, | |
| "grad_norm": 0.9233268580117602, | |
| "learning_rate": 9.826119870833644e-06, | |
| "loss": 0.2682, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.9098722370866834, | |
| "learning_rate": 9.710225748369402e-06, | |
| "loss": 0.2005, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5285714285714286, | |
| "grad_norm": 0.7043270611487241, | |
| "learning_rate": 9.594370567711512e-06, | |
| "loss": 0.1361, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5321428571428571, | |
| "grad_norm": 0.945912907239272, | |
| "learning_rate": 9.478569898255765e-06, | |
| "loss": 0.27, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.7227418929195129, | |
| "learning_rate": 9.362839302072354e-06, | |
| "loss": 0.1523, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5392857142857143, | |
| "grad_norm": 0.901696864239785, | |
| "learning_rate": 9.247194331814561e-06, | |
| "loss": 0.2297, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 0.6952433582976846, | |
| "learning_rate": 9.131650528628688e-06, | |
| "loss": 0.1879, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5464285714285714, | |
| "grad_norm": 0.8968896375753643, | |
| "learning_rate": 9.016223420065519e-06, | |
| "loss": 0.1998, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.0357021485704796, | |
| "learning_rate": 8.900928517993644e-06, | |
| "loss": 0.289, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 0.9263285065897571, | |
| "learning_rate": 8.785781316514841e-06, | |
| "loss": 0.2387, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5571428571428572, | |
| "grad_norm": 0.8130477337572882, | |
| "learning_rate": 8.670797289881915e-06, | |
| "loss": 0.2393, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5607142857142857, | |
| "grad_norm": 0.8664747661731167, | |
| "learning_rate": 8.555991890419116e-06, | |
| "loss": 0.2295, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5642857142857143, | |
| "grad_norm": 0.7139524830413835, | |
| "learning_rate": 8.441380546445603e-06, | |
| "loss": 0.1683, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5678571428571428, | |
| "grad_norm": 1.1700171170245692, | |
| "learning_rate": 8.326978660202034e-06, | |
| "loss": 0.3086, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.772880515075922, | |
| "learning_rate": 8.212801605780754e-06, | |
| "loss": 0.1811, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.7715213885473983, | |
| "learning_rate": 8.098864727059685e-06, | |
| "loss": 0.2182, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5785714285714286, | |
| "grad_norm": 1.2255567186716825, | |
| "learning_rate": 7.985183335640332e-06, | |
| "loss": 0.2629, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5821428571428572, | |
| "grad_norm": 0.7057680167932003, | |
| "learning_rate": 7.871772708790114e-06, | |
| "loss": 0.1414, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5857142857142857, | |
| "grad_norm": 0.7076516916675438, | |
| "learning_rate": 7.758648087389277e-06, | |
| "loss": 0.1953, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 0.9193199539199463, | |
| "learning_rate": 7.64582467388275e-06, | |
| "loss": 0.2648, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5928571428571429, | |
| "grad_norm": 0.7694356287549676, | |
| "learning_rate": 7.533317630237117e-06, | |
| "loss": 0.2015, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5964285714285714, | |
| "grad_norm": 0.9020767096753005, | |
| "learning_rate": 7.421142075903067e-06, | |
| "loss": 0.2151, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.9010708295786384, | |
| "learning_rate": 7.3093130857835245e-06, | |
| "loss": 0.1934, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6035714285714285, | |
| "grad_norm": 0.9781518043714732, | |
| "learning_rate": 7.197845688207805e-06, | |
| "loss": 0.2783, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 0.8062846336754722, | |
| "learning_rate": 7.086754862911982e-06, | |
| "loss": 0.1876, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6107142857142858, | |
| "grad_norm": 0.8993288898038415, | |
| "learning_rate": 6.976055539025819e-06, | |
| "loss": 0.2765, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6142857142857143, | |
| "grad_norm": 0.9544638260763084, | |
| "learning_rate": 6.865762593066514e-06, | |
| "loss": 0.2224, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6178571428571429, | |
| "grad_norm": 0.7861637636734311, | |
| "learning_rate": 6.755890846939454e-06, | |
| "loss": 0.2351, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6214285714285714, | |
| "grad_norm": 1.526261650246375, | |
| "learning_rate": 6.646455065946386e-06, | |
| "loss": 0.3024, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.8243808604395686, | |
| "learning_rate": 6.537469956801128e-06, | |
| "loss": 0.2193, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 0.9068394658105672, | |
| "learning_rate": 6.428950165653204e-06, | |
| "loss": 0.1525, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6321428571428571, | |
| "grad_norm": 0.8066101085303891, | |
| "learning_rate": 6.320910276119576e-06, | |
| "loss": 0.232, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6357142857142857, | |
| "grad_norm": 0.849276264490868, | |
| "learning_rate": 6.213364807324817e-06, | |
| "loss": 0.2112, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6392857142857142, | |
| "grad_norm": 0.9467657676513234, | |
| "learning_rate": 6.106328211949928e-06, | |
| "loss": 0.257, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.8690174893226397, | |
| "learning_rate": 5.999814874290084e-06, | |
| "loss": 0.2214, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6464285714285715, | |
| "grad_norm": 0.875769130127647, | |
| "learning_rate": 5.893839108321584e-06, | |
| "loss": 0.2033, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.6946362310360622, | |
| "learning_rate": 5.7884151557782305e-06, | |
| "loss": 0.1731, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6535714285714286, | |
| "grad_norm": 1.059540577971593, | |
| "learning_rate": 5.68355718423746e-06, | |
| "loss": 0.2825, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 0.871973114594604, | |
| "learning_rate": 5.579279285216369e-06, | |
| "loss": 0.266, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 0.9362769886838416, | |
| "learning_rate": 5.4755954722780236e-06, | |
| "loss": 0.2261, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6642857142857143, | |
| "grad_norm": 1.0156195729999553, | |
| "learning_rate": 5.372519679148227e-06, | |
| "loss": 0.2772, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6678571428571428, | |
| "grad_norm": 1.1707981800497014, | |
| "learning_rate": 5.270065757843e-06, | |
| "loss": 0.2357, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6714285714285714, | |
| "grad_norm": 1.1200974625591849, | |
| "learning_rate": 5.168247476807054e-06, | |
| "loss": 0.2955, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.7302141966698424, | |
| "learning_rate": 5.067078519063514e-06, | |
| "loss": 0.2351, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 0.9547345238061237, | |
| "learning_rate": 4.966572480375076e-06, | |
| "loss": 0.2201, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6821428571428572, | |
| "grad_norm": 0.9859252903039029, | |
| "learning_rate": 4.86674286741693e-06, | |
| "loss": 0.2634, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 1.2303560929176436, | |
| "learning_rate": 4.767603095961652e-06, | |
| "loss": 0.3092, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6892857142857143, | |
| "grad_norm": 0.902957152854701, | |
| "learning_rate": 4.669166489076283e-06, | |
| "loss": 0.2614, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6928571428571428, | |
| "grad_norm": 0.7887211722266535, | |
| "learning_rate": 4.571446275331903e-06, | |
| "loss": 0.1946, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 0.9681771399551783, | |
| "learning_rate": 4.47445558702587e-06, | |
| "loss": 0.2515, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.8387940708324988, | |
| "learning_rate": 4.378207458417035e-06, | |
| "loss": 0.2442, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7035714285714286, | |
| "grad_norm": 1.2905572299013393, | |
| "learning_rate": 4.282714823974088e-06, | |
| "loss": 0.2412, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7071428571428572, | |
| "grad_norm": 0.9660020289811414, | |
| "learning_rate": 4.187990516637361e-06, | |
| "loss": 0.2183, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7107142857142857, | |
| "grad_norm": 0.9987611883772257, | |
| "learning_rate": 4.094047266094225e-06, | |
| "loss": 0.2111, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.6971224340534188, | |
| "learning_rate": 4.000897697068418e-06, | |
| "loss": 0.1147, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7178571428571429, | |
| "grad_norm": 0.8499318496594657, | |
| "learning_rate": 3.908554327623425e-06, | |
| "loss": 0.2028, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7214285714285714, | |
| "grad_norm": 0.9757223338840971, | |
| "learning_rate": 3.817029567480228e-06, | |
| "loss": 0.2687, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.9724129336265752, | |
| "learning_rate": 3.7263357163496118e-06, | |
| "loss": 0.2951, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7285714285714285, | |
| "grad_norm": 0.846692248077101, | |
| "learning_rate": 3.6364849622792262e-06, | |
| "loss": 0.2116, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 0.7244565494778341, | |
| "learning_rate": 3.5474893800157005e-06, | |
| "loss": 0.1824, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7357142857142858, | |
| "grad_norm": 1.0573449443528709, | |
| "learning_rate": 3.459360929381931e-06, | |
| "loss": 0.1647, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.7392857142857143, | |
| "grad_norm": 1.3741948680591016, | |
| "learning_rate": 3.372111453669864e-06, | |
| "loss": 0.29, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 1.0221267171956534, | |
| "learning_rate": 3.2857526780488925e-06, | |
| "loss": 0.2258, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7464285714285714, | |
| "grad_norm": 0.7776596440463766, | |
| "learning_rate": 3.2002962079901743e-06, | |
| "loss": 0.2101, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.8390204592490403, | |
| "learning_rate": 3.115753527706986e-06, | |
| "loss": 0.2355, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7535714285714286, | |
| "grad_norm": 0.6624410313645264, | |
| "learning_rate": 3.0321359986114096e-06, | |
| "loss": 0.1756, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7571428571428571, | |
| "grad_norm": 1.0623528998074427, | |
| "learning_rate": 2.9494548577875195e-06, | |
| "loss": 0.1945, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7607142857142857, | |
| "grad_norm": 1.1302403323388186, | |
| "learning_rate": 2.8677212164812464e-06, | |
| "loss": 0.2234, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7642857142857142, | |
| "grad_norm": 0.8609060708419913, | |
| "learning_rate": 2.786946058607187e-06, | |
| "loss": 0.2657, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 0.9847057932243088, | |
| "learning_rate": 2.70714023927251e-06, | |
| "loss": 0.2727, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 1.3326482318392308, | |
| "learning_rate": 2.628314483318178e-06, | |
| "loss": 0.2287, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.9202719280439493, | |
| "learning_rate": 2.5504793838776585e-06, | |
| "loss": 0.2034, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7785714285714286, | |
| "grad_norm": 0.9299327013502574, | |
| "learning_rate": 2.473645400953366e-06, | |
| "loss": 0.2584, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7821428571428571, | |
| "grad_norm": 0.9697882855786745, | |
| "learning_rate": 2.3978228600109564e-06, | |
| "loss": 0.2023, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 1.0868782541786108, | |
| "learning_rate": 2.323021950591743e-06, | |
| "loss": 0.2423, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7892857142857143, | |
| "grad_norm": 0.9406345195120313, | |
| "learning_rate": 2.249252724943336e-06, | |
| "loss": 0.2093, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7928571428571428, | |
| "grad_norm": 0.8299606398290135, | |
| "learning_rate": 2.176525096668769e-06, | |
| "loss": 0.2024, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7964285714285714, | |
| "grad_norm": 0.9557718556324724, | |
| "learning_rate": 2.1048488393942455e-06, | |
| "loss": 0.2651, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.020716425877873, | |
| "learning_rate": 2.0342335854556738e-06, | |
| "loss": 0.218, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 0.8247593195528904, | |
| "learning_rate": 1.964688824604234e-06, | |
| "loss": 0.2224, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8071428571428572, | |
| "grad_norm": 0.7141792575779932, | |
| "learning_rate": 1.896223902731058e-06, | |
| "loss": 0.148, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8107142857142857, | |
| "grad_norm": 0.6798152535391203, | |
| "learning_rate": 1.8288480206112879e-06, | |
| "loss": 0.1693, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.8142857142857143, | |
| "grad_norm": 0.7865744423081975, | |
| "learning_rate": 1.7625702326675952e-06, | |
| "loss": 0.2307, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.8178571428571428, | |
| "grad_norm": 0.8961338272400713, | |
| "learning_rate": 1.6973994457534026e-06, | |
| "loss": 0.1763, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 1.0891294721112785, | |
| "learning_rate": 1.6333444179559078e-06, | |
| "loss": 0.251, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.9895358178842029, | |
| "learning_rate": 1.5704137574191202e-06, | |
| "loss": 0.2457, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 0.8797961625322609, | |
| "learning_rate": 1.5086159211870445e-06, | |
| "loss": 0.2974, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8321428571428572, | |
| "grad_norm": 1.0022630637461174, | |
| "learning_rate": 1.447959214067155e-06, | |
| "loss": 0.2911, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8357142857142857, | |
| "grad_norm": 0.9292460119394966, | |
| "learning_rate": 1.3884517875143544e-06, | |
| "loss": 0.253, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 0.8525100768549407, | |
| "learning_rate": 1.3301016385355093e-06, | |
| "loss": 0.2574, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8428571428571429, | |
| "grad_norm": 0.7616448289314827, | |
| "learning_rate": 1.2729166086147803e-06, | |
| "loss": 0.1935, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8464285714285714, | |
| "grad_norm": 0.9032528436222605, | |
| "learning_rate": 1.216904382659806e-06, | |
| "loss": 0.2136, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.9221094277183024, | |
| "learning_rate": 1.1620724879689793e-06, | |
| "loss": 0.2366, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8535714285714285, | |
| "grad_norm": 0.8128301362566547, | |
| "learning_rate": 1.1084282932198543e-06, | |
| "loss": 0.205, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.8351463450189537, | |
| "learning_rate": 1.0559790074789134e-06, | |
| "loss": 0.2451, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8607142857142858, | |
| "grad_norm": 0.8132844496379037, | |
| "learning_rate": 1.00473167923275e-06, | |
| "loss": 0.1986, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8642857142857143, | |
| "grad_norm": 0.8413773702315226, | |
| "learning_rate": 9.546931954408622e-07, | |
| "loss": 0.2568, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8678571428571429, | |
| "grad_norm": 0.864722633477625, | |
| "learning_rate": 9.058702806101172e-07, | |
| "loss": 0.2647, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8714285714285714, | |
| "grad_norm": 0.8564326198019813, | |
| "learning_rate": 8.582694958910809e-07, | |
| "loss": 0.2082, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.0551157545813, | |
| "learning_rate": 8.118972381962853e-07, | |
| "loss": 0.3371, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8785714285714286, | |
| "grad_norm": 0.7703975629143833, | |
| "learning_rate": 7.667597393405602e-07, | |
| "loss": 0.2341, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8821428571428571, | |
| "grad_norm": 0.6061411407576762, | |
| "learning_rate": 7.228630652035717e-07, | |
| "loss": 0.1415, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 1.1824816837626368, | |
| "learning_rate": 6.802131149146374e-07, | |
| "loss": 0.2504, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8892857142857142, | |
| "grad_norm": 0.7388040182293693, | |
| "learning_rate": 6.388156200599726e-07, | |
| "loss": 0.1832, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.7414017170453198, | |
| "learning_rate": 5.986761439124289e-07, | |
| "loss": 0.175, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8964285714285715, | |
| "grad_norm": 0.9715948557202785, | |
| "learning_rate": 5.598000806838766e-07, | |
| "loss": 0.3161, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6724183434948179, | |
| "learning_rate": 5.221926548002876e-07, | |
| "loss": 0.1827, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9035714285714286, | |
| "grad_norm": 0.7818282617469353, | |
| "learning_rate": 4.858589201996433e-07, | |
| "loss": 0.2202, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9071428571428571, | |
| "grad_norm": 0.8204974648639228, | |
| "learning_rate": 4.5080375965275256e-07, | |
| "loss": 0.1967, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 1.0856778296057459, | |
| "learning_rate": 4.1703188410707087e-07, | |
| "loss": 0.2444, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.8449959915498736, | |
| "learning_rate": 3.845478320536178e-07, | |
| "loss": 0.2371, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9178571428571428, | |
| "grad_norm": 0.9495844646371367, | |
| "learning_rate": 3.5335596891705406e-07, | |
| "loss": 0.2335, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.9214285714285714, | |
| "grad_norm": 0.8901968360351575, | |
| "learning_rate": 3.2346048646903494e-07, | |
| "loss": 0.2204, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.8923998294274479, | |
| "learning_rate": 2.9486540226488556e-07, | |
| "loss": 0.2567, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.815717327935455, | |
| "learning_rate": 2.6757455910370486e-07, | |
| "loss": 0.2578, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9321428571428572, | |
| "grad_norm": 0.7291953617908891, | |
| "learning_rate": 2.41591624511931e-07, | |
| "loss": 0.1857, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9357142857142857, | |
| "grad_norm": 1.8534979443451833, | |
| "learning_rate": 2.169200902504842e-07, | |
| "loss": 0.2899, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9392857142857143, | |
| "grad_norm": 0.8961982311920613, | |
| "learning_rate": 1.9356327184551716e-07, | |
| "loss": 0.1946, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 0.8300414662763644, | |
| "learning_rate": 1.7152430814285303e-07, | |
| "loss": 0.253, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 1.1075650320032002, | |
| "learning_rate": 1.5080616088616884e-07, | |
| "loss": 0.2717, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.16743264805072, | |
| "learning_rate": 1.3141161431896809e-07, | |
| "loss": 0.3116, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9535714285714286, | |
| "grad_norm": 0.8257674956219426, | |
| "learning_rate": 1.1334327481042573e-07, | |
| "loss": 0.1874, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9571428571428572, | |
| "grad_norm": 0.9283778527162151, | |
| "learning_rate": 9.660357050512158e-08, | |
| "loss": 0.2219, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9607142857142857, | |
| "grad_norm": 0.956684846701526, | |
| "learning_rate": 8.119475099673035e-08, | |
| "loss": 0.2465, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.6856186450599495, | |
| "learning_rate": 6.711888702570556e-08, | |
| "loss": 0.2051, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9678571428571429, | |
| "grad_norm": 0.8013460800051045, | |
| "learning_rate": 5.437787020100116e-08, | |
| "loss": 0.2108, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 0.9855143079178192, | |
| "learning_rate": 4.297341274586475e-08, | |
| "loss": 0.2287, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.9501486308995364, | |
| "learning_rate": 3.290704726773619e-08, | |
| "loss": 0.2572, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9785714285714285, | |
| "grad_norm": 0.8054069093536848, | |
| "learning_rate": 2.4180126552284523e-08, | |
| "loss": 0.2522, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 1.2803070465419621, | |
| "learning_rate": 1.6793823381614506e-08, | |
| "loss": 0.2266, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9857142857142858, | |
| "grad_norm": 0.8212197677595763, | |
| "learning_rate": 1.0749130376659366e-08, | |
| "loss": 0.2241, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9892857142857143, | |
| "grad_norm": 1.3359306690870951, | |
| "learning_rate": 6.046859863781951e-09, | |
| "loss": 0.3035, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9928571428571429, | |
| "grad_norm": 0.8933949606817133, | |
| "learning_rate": 2.6876437656153663e-09, | |
| "loss": 0.2186, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9964285714285714, | |
| "grad_norm": 0.7916708816942223, | |
| "learning_rate": 6.719335161364804e-10, | |
| "loss": 0.2021, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8075329735012263, | |
| "learning_rate": 0.0, | |
| "loss": 0.2782, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 280, | |
| "total_flos": 478920998780928.0, | |
| "train_loss": 0.2742149087467364, | |
| "train_runtime": 2790.475, | |
| "train_samples_per_second": 12.839, | |
| "train_steps_per_second": 0.1 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 478920998780928.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |