| { | |
| "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": 3.2543029662772347, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.554, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007142857142857143, | |
| "grad_norm": 2.7624244157844706, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.4127, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010714285714285714, | |
| "grad_norm": 3.1002251906462344, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.4643, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014285714285714285, | |
| "grad_norm": 2.3755712686804804, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.3791, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 1.9308377348791768, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.3221, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02142857142857143, | |
| "grad_norm": 2.0118553962006427, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.2667, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 2.249667975836266, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 1.3315, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 3.3286044973535236, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 1.2336, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03214285714285714, | |
| "grad_norm": 2.607483213778446, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2549, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 1.9059436966580536, | |
| "learning_rate": 1.9999328066483867e-05, | |
| "loss": 1.1484, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.039285714285714285, | |
| "grad_norm": 2.351052203479045, | |
| "learning_rate": 1.9997312356234385e-05, | |
| "loss": 1.1006, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04285714285714286, | |
| "grad_norm": 2.287439785470941, | |
| "learning_rate": 1.999395314013622e-05, | |
| "loss": 1.0522, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04642857142857143, | |
| "grad_norm": 1.7697435082139328, | |
| "learning_rate": 1.998925086962334e-05, | |
| "loss": 1.0117, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.7809725059300023, | |
| "learning_rate": 1.998320617661839e-05, | |
| "loss": 0.9512, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 1.6904647102656667, | |
| "learning_rate": 1.997581987344772e-05, | |
| "loss": 0.9306, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 1.6895690138180741, | |
| "learning_rate": 1.9967092952732266e-05, | |
| "loss": 0.8694, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.060714285714285714, | |
| "grad_norm": 1.7669079061201738, | |
| "learning_rate": 1.9957026587254136e-05, | |
| "loss": 0.8984, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06428571428571428, | |
| "grad_norm": 1.476756594220327, | |
| "learning_rate": 1.9945622129799e-05, | |
| "loss": 0.8041, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06785714285714285, | |
| "grad_norm": 1.5474239491243957, | |
| "learning_rate": 1.9932881112974298e-05, | |
| "loss": 0.8109, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 1.449807734275577, | |
| "learning_rate": 1.9918805249003272e-05, | |
| "loss": 0.6632, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 1.4334867898446597, | |
| "learning_rate": 1.9903396429494882e-05, | |
| "loss": 0.6602, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07857142857142857, | |
| "grad_norm": 1.3803101146379801, | |
| "learning_rate": 1.9886656725189575e-05, | |
| "loss": 0.6428, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08214285714285714, | |
| "grad_norm": 1.3124607153366843, | |
| "learning_rate": 1.9868588385681035e-05, | |
| "loss": 0.6175, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 1.23178251089412, | |
| "learning_rate": 1.9849193839113833e-05, | |
| "loss": 0.5117, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 1.1688906258052285, | |
| "learning_rate": 1.9828475691857148e-05, | |
| "loss": 0.5335, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09285714285714286, | |
| "grad_norm": 1.1144465197033029, | |
| "learning_rate": 1.9806436728154484e-05, | |
| "loss": 0.5583, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09642857142857143, | |
| "grad_norm": 1.2143115091886203, | |
| "learning_rate": 1.9783079909749516e-05, | |
| "loss": 0.5482, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2235901463474592, | |
| "learning_rate": 1.975840837548807e-05, | |
| "loss": 0.4723, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10357142857142858, | |
| "grad_norm": 1.1827323503795928, | |
| "learning_rate": 1.9732425440896298e-05, | |
| "loss": 0.5831, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 1.1949009678747529, | |
| "learning_rate": 1.9705134597735113e-05, | |
| "loss": 0.4932, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11071428571428571, | |
| "grad_norm": 1.1096843684141346, | |
| "learning_rate": 1.9676539513530967e-05, | |
| "loss": 0.3598, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 1.0706468088750498, | |
| "learning_rate": 1.9646644031082948e-05, | |
| "loss": 0.4306, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11785714285714285, | |
| "grad_norm": 1.5246249464687178, | |
| "learning_rate": 1.9615452167946383e-05, | |
| "loss": 0.5073, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.12142857142857143, | |
| "grad_norm": 1.2556676622859617, | |
| "learning_rate": 1.958296811589293e-05, | |
| "loss": 0.4762, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 1.2909831452240883, | |
| "learning_rate": 1.954919624034725e-05, | |
| "loss": 0.4601, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12857142857142856, | |
| "grad_norm": 1.2644852139785117, | |
| "learning_rate": 1.951414107980036e-05, | |
| "loss": 0.3924, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13214285714285715, | |
| "grad_norm": 1.177541703534973, | |
| "learning_rate": 1.9477807345199717e-05, | |
| "loss": 0.4298, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1357142857142857, | |
| "grad_norm": 1.0788971626418669, | |
| "learning_rate": 1.9440199919316125e-05, | |
| "loss": 0.3866, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1392857142857143, | |
| "grad_norm": 1.198940400239735, | |
| "learning_rate": 1.9401323856087573e-05, | |
| "loss": 0.4065, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 1.120532645081093, | |
| "learning_rate": 1.936118437994003e-05, | |
| "loss": 0.4159, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14642857142857144, | |
| "grad_norm": 1.189893899476895, | |
| "learning_rate": 1.9319786885085366e-05, | |
| "loss": 0.3384, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.113585244588961, | |
| "learning_rate": 1.927713693479643e-05, | |
| "loss": 0.4105, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.15357142857142858, | |
| "grad_norm": 1.0519634925640304, | |
| "learning_rate": 1.923324026065944e-05, | |
| "loss": 0.3938, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.15714285714285714, | |
| "grad_norm": 1.0101223695618087, | |
| "learning_rate": 1.918810276180372e-05, | |
| "loss": 0.352, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 1.024763914116021, | |
| "learning_rate": 1.9141730504108923e-05, | |
| "loss": 0.3333, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16428571428571428, | |
| "grad_norm": 0.9967381517754978, | |
| "learning_rate": 1.9094129719389886e-05, | |
| "loss": 0.3412, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16785714285714284, | |
| "grad_norm": 1.2637178696514346, | |
| "learning_rate": 1.904530680455914e-05, | |
| "loss": 0.3622, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 1.0266452708065086, | |
| "learning_rate": 1.8995268320767254e-05, | |
| "loss": 0.3541, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 1.0740239915454661, | |
| "learning_rate": 1.894402099252109e-05, | |
| "loss": 0.2964, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 1.069730824561292, | |
| "learning_rate": 1.889157170678015e-05, | |
| "loss": 0.3421, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18214285714285713, | |
| "grad_norm": 0.9643772737738865, | |
| "learning_rate": 1.8837927512031022e-05, | |
| "loss": 0.3057, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.18571428571428572, | |
| "grad_norm": 1.0652997055685356, | |
| "learning_rate": 1.8783095617340193e-05, | |
| "loss": 0.3, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.18928571428571428, | |
| "grad_norm": 0.8166401050438583, | |
| "learning_rate": 1.872708339138522e-05, | |
| "loss": 0.2458, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19285714285714287, | |
| "grad_norm": 0.8876770564223524, | |
| "learning_rate": 1.866989836146449e-05, | |
| "loss": 0.2515, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 1.003831099607791, | |
| "learning_rate": 1.861154821248565e-05, | |
| "loss": 0.32, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.9005609017037866, | |
| "learning_rate": 1.8552040785932846e-05, | |
| "loss": 0.3109, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.20357142857142857, | |
| "grad_norm": 0.9196347354093852, | |
| "learning_rate": 1.849138407881296e-05, | |
| "loss": 0.2715, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.20714285714285716, | |
| "grad_norm": 1.1659775560675805, | |
| "learning_rate": 1.8429586242580884e-05, | |
| "loss": 0.3161, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21071428571428572, | |
| "grad_norm": 0.8229225999880141, | |
| "learning_rate": 1.8366655582044096e-05, | |
| "loss": 0.1971, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 1.0837498121968563, | |
| "learning_rate": 1.83026005542466e-05, | |
| "loss": 0.3534, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21785714285714286, | |
| "grad_norm": 0.9894720548109736, | |
| "learning_rate": 1.8237429767332407e-05, | |
| "loss": 0.2613, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.22142857142857142, | |
| "grad_norm": 0.9165029939399432, | |
| "learning_rate": 1.8171151979388715e-05, | |
| "loss": 0.232, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.8903286251798155, | |
| "learning_rate": 1.8103776097268942e-05, | |
| "loss": 0.2857, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 0.8053044878739753, | |
| "learning_rate": 1.803531117539577e-05, | |
| "loss": 0.2177, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 0.8248138749411966, | |
| "learning_rate": 1.7965766414544328e-05, | |
| "loss": 0.1982, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2357142857142857, | |
| "grad_norm": 0.9377249996982162, | |
| "learning_rate": 1.7895151160605758e-05, | |
| "loss": 0.2916, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2392857142857143, | |
| "grad_norm": 0.8817863744148133, | |
| "learning_rate": 1.782347490333123e-05, | |
| "loss": 0.2599, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.24285714285714285, | |
| "grad_norm": 0.6848196359492216, | |
| "learning_rate": 1.775074727505667e-05, | |
| "loss": 0.2374, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.24642857142857144, | |
| "grad_norm": 0.8834440854449724, | |
| "learning_rate": 1.7676978049408262e-05, | |
| "loss": 0.2134, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8655380809431652, | |
| "learning_rate": 1.7602177139989046e-05, | |
| "loss": 0.2456, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25357142857142856, | |
| "grad_norm": 0.9506232189055851, | |
| "learning_rate": 1.7526354599046637e-05, | |
| "loss": 0.2661, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 0.7978880305255668, | |
| "learning_rate": 1.7449520616122344e-05, | |
| "loss": 0.2355, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.26071428571428573, | |
| "grad_norm": 0.8250196602876038, | |
| "learning_rate": 1.7371685516681825e-05, | |
| "loss": 0.207, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2642857142857143, | |
| "grad_norm": 0.8117468052334055, | |
| "learning_rate": 1.7292859760727493e-05, | |
| "loss": 0.2303, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 0.7195318618659722, | |
| "learning_rate": 1.721305394139282e-05, | |
| "loss": 0.1655, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2714285714285714, | |
| "grad_norm": 0.6949862427137922, | |
| "learning_rate": 1.7132278783518756e-05, | |
| "loss": 0.2083, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.8514564789796272, | |
| "learning_rate": 1.7050545142212483e-05, | |
| "loss": 0.2242, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2785714285714286, | |
| "grad_norm": 0.7516508666701327, | |
| "learning_rate": 1.696786400138859e-05, | |
| "loss": 0.2437, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.28214285714285714, | |
| "grad_norm": 0.9714012683887667, | |
| "learning_rate": 1.6884246472293018e-05, | |
| "loss": 0.2375, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.7572091520879666, | |
| "learning_rate": 1.679970379200983e-05, | |
| "loss": 0.242, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2892857142857143, | |
| "grad_norm": 0.8297099751697198, | |
| "learning_rate": 1.6714247321951106e-05, | |
| "loss": 0.263, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.29285714285714287, | |
| "grad_norm": 0.9090423008791377, | |
| "learning_rate": 1.6627888546330136e-05, | |
| "loss": 0.2961, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.29642857142857143, | |
| "grad_norm": 0.8510800428736458, | |
| "learning_rate": 1.654063907061807e-05, | |
| "loss": 0.2321, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.7131209745137069, | |
| "learning_rate": 1.64525106199843e-05, | |
| "loss": 0.2288, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 0.6919526731968912, | |
| "learning_rate": 1.6363515037720774e-05, | |
| "loss": 0.2288, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30714285714285716, | |
| "grad_norm": 0.7958343122708701, | |
| "learning_rate": 1.6273664283650393e-05, | |
| "loss": 0.2323, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3107142857142857, | |
| "grad_norm": 0.6483638471772285, | |
| "learning_rate": 1.6182970432519772e-05, | |
| "loss": 0.1861, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 0.7300972630561462, | |
| "learning_rate": 1.609144567237658e-05, | |
| "loss": 0.2213, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.31785714285714284, | |
| "grad_norm": 0.6899792899268994, | |
| "learning_rate": 1.5999102302931585e-05, | |
| "loss": 0.2224, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 0.767512157488968, | |
| "learning_rate": 1.5905952733905777e-05, | |
| "loss": 0.2634, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.861421412557911, | |
| "learning_rate": 1.5812009483362643e-05, | |
| "loss": 0.2516, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.32857142857142857, | |
| "grad_norm": 1.2043142385923853, | |
| "learning_rate": 1.5717285176025913e-05, | |
| "loss": 0.2389, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.33214285714285713, | |
| "grad_norm": 1.0812423023626763, | |
| "learning_rate": 1.5621792541582968e-05, | |
| "loss": 0.2514, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3357142857142857, | |
| "grad_norm": 0.8297182880168162, | |
| "learning_rate": 1.552554441297413e-05, | |
| "loss": 0.245, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 0.7460278115544305, | |
| "learning_rate": 1.5428553724668103e-05, | |
| "loss": 0.2438, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 0.7162063436193922, | |
| "learning_rate": 1.533083351092372e-05, | |
| "loss": 0.1747, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3464285714285714, | |
| "grad_norm": 0.6456220159305923, | |
| "learning_rate": 1.5232396904038352e-05, | |
| "loss": 0.1534, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7094654735506336, | |
| "learning_rate": 1.5133257132583074e-05, | |
| "loss": 0.2545, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3535714285714286, | |
| "grad_norm": 0.8247734466190151, | |
| "learning_rate": 1.503342751962493e-05, | |
| "loss": 0.2848, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 0.7728186991716375, | |
| "learning_rate": 1.4932921480936491e-05, | |
| "loss": 0.2566, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3607142857142857, | |
| "grad_norm": 0.8642822181295595, | |
| "learning_rate": 1.4831752523192949e-05, | |
| "loss": 0.2583, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.36428571428571427, | |
| "grad_norm": 0.8060855146089662, | |
| "learning_rate": 1.4729934242157005e-05, | |
| "loss": 0.2754, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3678571428571429, | |
| "grad_norm": 0.8190230120369192, | |
| "learning_rate": 1.4627480320851775e-05, | |
| "loss": 0.2865, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 0.6600130809466522, | |
| "learning_rate": 1.4524404527721977e-05, | |
| "loss": 0.195, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.6884634046078892, | |
| "learning_rate": 1.4420720714783635e-05, | |
| "loss": 0.212, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.37857142857142856, | |
| "grad_norm": 0.7936195003334405, | |
| "learning_rate": 1.4316442815762543e-05, | |
| "loss": 0.2658, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3821428571428571, | |
| "grad_norm": 0.7025320097938863, | |
| "learning_rate": 1.4211584844221771e-05, | |
| "loss": 0.2087, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.38571428571428573, | |
| "grad_norm": 0.6172413774186932, | |
| "learning_rate": 1.4106160891678422e-05, | |
| "loss": 0.1727, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3892857142857143, | |
| "grad_norm": 0.5907547063659139, | |
| "learning_rate": 1.4000185125709919e-05, | |
| "loss": 0.1801, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 0.7794646496722246, | |
| "learning_rate": 1.3893671788050073e-05, | |
| "loss": 0.2018, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3964285714285714, | |
| "grad_norm": 0.5110684834921772, | |
| "learning_rate": 1.3786635192675184e-05, | |
| "loss": 0.1593, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.9463597697315387, | |
| "learning_rate": 1.3679089723880427e-05, | |
| "loss": 0.2699, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4035714285714286, | |
| "grad_norm": 0.6376009700059774, | |
| "learning_rate": 1.35710498343468e-05, | |
| "loss": 0.1562, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.40714285714285714, | |
| "grad_norm": 0.7904776444292465, | |
| "learning_rate": 1.3462530043198874e-05, | |
| "loss": 0.2702, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 0.5896293345349027, | |
| "learning_rate": 1.3353544934053618e-05, | |
| "loss": 0.2016, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4142857142857143, | |
| "grad_norm": 0.6475939911269051, | |
| "learning_rate": 1.324410915306055e-05, | |
| "loss": 0.2196, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.41785714285714287, | |
| "grad_norm": 0.5957948341177419, | |
| "learning_rate": 1.3134237406933493e-05, | |
| "loss": 0.1883, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.42142857142857143, | |
| "grad_norm": 0.7141704868499749, | |
| "learning_rate": 1.3023944460974183e-05, | |
| "loss": 0.1765, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 0.6190173244418413, | |
| "learning_rate": 1.2913245137088024e-05, | |
| "loss": 0.2338, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.5152892854036694, | |
| "learning_rate": 1.2802154311792196e-05, | |
| "loss": 0.1407, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.43214285714285716, | |
| "grad_norm": 0.6970455298668206, | |
| "learning_rate": 1.2690686914216475e-05, | |
| "loss": 0.2245, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4357142857142857, | |
| "grad_norm": 0.5424563727889877, | |
| "learning_rate": 1.2578857924096935e-05, | |
| "loss": 0.1642, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4392857142857143, | |
| "grad_norm": 0.5455879645989501, | |
| "learning_rate": 1.2466682369762883e-05, | |
| "loss": 0.1745, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.44285714285714284, | |
| "grad_norm": 0.7159501319130775, | |
| "learning_rate": 1.2354175326117252e-05, | |
| "loss": 0.2147, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 0.9747645380400841, | |
| "learning_rate": 1.2241351912610726e-05, | |
| "loss": 0.2842, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.663151474979954, | |
| "learning_rate": 1.212822729120989e-05, | |
| "loss": 0.1917, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.45357142857142857, | |
| "grad_norm": 0.539525979449454, | |
| "learning_rate": 1.2014816664359671e-05, | |
| "loss": 0.1929, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.7511233269985765, | |
| "learning_rate": 1.190113527294032e-05, | |
| "loss": 0.2415, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4607142857142857, | |
| "grad_norm": 0.6962189495102956, | |
| "learning_rate": 1.178719839421925e-05, | |
| "loss": 0.1904, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 0.5794963846183635, | |
| "learning_rate": 1.1673021339797967e-05, | |
| "loss": 0.2185, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.46785714285714286, | |
| "grad_norm": 0.5834120344196029, | |
| "learning_rate": 1.15586194535544e-05, | |
| "loss": 0.1648, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4714285714285714, | |
| "grad_norm": 0.42363171997494503, | |
| "learning_rate": 1.1444008109580884e-05, | |
| "loss": 0.1274, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.7339617718296433, | |
| "learning_rate": 1.1329202710118088e-05, | |
| "loss": 0.2408, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4785714285714286, | |
| "grad_norm": 0.7313570855021271, | |
| "learning_rate": 1.1214218683485159e-05, | |
| "loss": 0.2571, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 0.6413922491909684, | |
| "learning_rate": 1.1099071482006361e-05, | |
| "loss": 0.2049, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 0.6927763783969573, | |
| "learning_rate": 1.0983776579934483e-05, | |
| "loss": 0.2319, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.48928571428571427, | |
| "grad_norm": 0.8400234609027128, | |
| "learning_rate": 1.0868349471371316e-05, | |
| "loss": 0.2984, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4928571428571429, | |
| "grad_norm": 0.59355253258987, | |
| "learning_rate": 1.0752805668185442e-05, | |
| "loss": 0.1682, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.49642857142857144, | |
| "grad_norm": 0.6702328703095428, | |
| "learning_rate": 1.0637160697927651e-05, | |
| "loss": 0.2486, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.610663540942464, | |
| "learning_rate": 1.0521430101744238e-05, | |
| "loss": 0.1944, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5035714285714286, | |
| "grad_norm": 0.6670778255030475, | |
| "learning_rate": 1.040562943228849e-05, | |
| "loss": 0.2303, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5071428571428571, | |
| "grad_norm": 0.5298284958723746, | |
| "learning_rate": 1.0289774251630602e-05, | |
| "loss": 0.1324, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5107142857142857, | |
| "grad_norm": 0.7102449237713073, | |
| "learning_rate": 1.0173880129166358e-05, | |
| "loss": 0.2229, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 0.5146460843697134, | |
| "learning_rate": 1.0057962639524799e-05, | |
| "loss": 0.1508, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 0.5757523755087942, | |
| "learning_rate": 9.942037360475205e-06, | |
| "loss": 0.1728, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5214285714285715, | |
| "grad_norm": 0.6500936134711337, | |
| "learning_rate": 9.826119870833644e-06, | |
| "loss": 0.2437, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.5181133145906637, | |
| "learning_rate": 9.710225748369402e-06, | |
| "loss": 0.1865, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5285714285714286, | |
| "grad_norm": 0.495454322336506, | |
| "learning_rate": 9.594370567711512e-06, | |
| "loss": 0.1617, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5321428571428571, | |
| "grad_norm": 0.6228595031287888, | |
| "learning_rate": 9.478569898255765e-06, | |
| "loss": 0.2104, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.44664086678984055, | |
| "learning_rate": 9.362839302072354e-06, | |
| "loss": 0.1461, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5392857142857143, | |
| "grad_norm": 0.6712033221960267, | |
| "learning_rate": 9.247194331814561e-06, | |
| "loss": 0.2092, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 0.5187745738654083, | |
| "learning_rate": 9.131650528628688e-06, | |
| "loss": 0.1618, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5464285714285714, | |
| "grad_norm": 0.5742615521018551, | |
| "learning_rate": 9.016223420065519e-06, | |
| "loss": 0.1781, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.5996031168106218, | |
| "learning_rate": 8.900928517993644e-06, | |
| "loss": 0.1895, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 0.6328292684226152, | |
| "learning_rate": 8.785781316514841e-06, | |
| "loss": 0.2149, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5571428571428572, | |
| "grad_norm": 0.5387581480854878, | |
| "learning_rate": 8.670797289881915e-06, | |
| "loss": 0.1775, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5607142857142857, | |
| "grad_norm": 0.5572164434404691, | |
| "learning_rate": 8.555991890419116e-06, | |
| "loss": 0.1933, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5642857142857143, | |
| "grad_norm": 0.5649031155667329, | |
| "learning_rate": 8.441380546445603e-06, | |
| "loss": 0.181, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5678571428571428, | |
| "grad_norm": 0.6573973388367828, | |
| "learning_rate": 8.326978660202034e-06, | |
| "loss": 0.2374, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.6303978475727942, | |
| "learning_rate": 8.212801605780754e-06, | |
| "loss": 0.1735, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.4801400244382432, | |
| "learning_rate": 8.098864727059685e-06, | |
| "loss": 0.1412, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5785714285714286, | |
| "grad_norm": 0.6847676384869728, | |
| "learning_rate": 7.985183335640332e-06, | |
| "loss": 0.2142, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5821428571428572, | |
| "grad_norm": 0.42794916312384695, | |
| "learning_rate": 7.871772708790114e-06, | |
| "loss": 0.1167, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5857142857142857, | |
| "grad_norm": 0.7293980682002585, | |
| "learning_rate": 7.758648087389277e-06, | |
| "loss": 0.2031, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 0.6599403699786116, | |
| "learning_rate": 7.64582467388275e-06, | |
| "loss": 0.2415, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5928571428571429, | |
| "grad_norm": 0.48043926105927637, | |
| "learning_rate": 7.533317630237117e-06, | |
| "loss": 0.1888, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5964285714285714, | |
| "grad_norm": 0.6412422605963668, | |
| "learning_rate": 7.421142075903067e-06, | |
| "loss": 0.2089, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6809868730952757, | |
| "learning_rate": 7.3093130857835245e-06, | |
| "loss": 0.1805, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6035714285714285, | |
| "grad_norm": 0.6418069717776665, | |
| "learning_rate": 7.197845688207805e-06, | |
| "loss": 0.2438, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 0.5376357495185539, | |
| "learning_rate": 7.086754862911982e-06, | |
| "loss": 0.2239, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6107142857142858, | |
| "grad_norm": 0.682482573713786, | |
| "learning_rate": 6.976055539025819e-06, | |
| "loss": 0.2778, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6142857142857143, | |
| "grad_norm": 0.5510808352313203, | |
| "learning_rate": 6.865762593066514e-06, | |
| "loss": 0.1768, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6178571428571429, | |
| "grad_norm": 1.0898121691999565, | |
| "learning_rate": 6.755890846939454e-06, | |
| "loss": 0.2274, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6214285714285714, | |
| "grad_norm": 0.6833115682394764, | |
| "learning_rate": 6.646455065946386e-06, | |
| "loss": 0.209, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.5834718836130244, | |
| "learning_rate": 6.537469956801128e-06, | |
| "loss": 0.195, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 0.4059918209883239, | |
| "learning_rate": 6.428950165653204e-06, | |
| "loss": 0.1253, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6321428571428571, | |
| "grad_norm": 0.6159299376512053, | |
| "learning_rate": 6.320910276119576e-06, | |
| "loss": 0.2089, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6357142857142857, | |
| "grad_norm": 0.5708201106967846, | |
| "learning_rate": 6.213364807324817e-06, | |
| "loss": 0.1836, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6392857142857142, | |
| "grad_norm": 0.46753041275070745, | |
| "learning_rate": 6.106328211949928e-06, | |
| "loss": 0.151, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.46419983181684327, | |
| "learning_rate": 5.999814874290084e-06, | |
| "loss": 0.1424, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6464285714285715, | |
| "grad_norm": 0.6470305272135881, | |
| "learning_rate": 5.893839108321584e-06, | |
| "loss": 0.2592, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.582995072151267, | |
| "learning_rate": 5.7884151557782305e-06, | |
| "loss": 0.1394, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6535714285714286, | |
| "grad_norm": 0.6872272149230029, | |
| "learning_rate": 5.68355718423746e-06, | |
| "loss": 0.2273, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 0.7010753300233663, | |
| "learning_rate": 5.579279285216369e-06, | |
| "loss": 0.2578, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 0.6816215511289796, | |
| "learning_rate": 5.4755954722780236e-06, | |
| "loss": 0.2334, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6642857142857143, | |
| "grad_norm": 0.9305267527410732, | |
| "learning_rate": 5.372519679148227e-06, | |
| "loss": 0.3146, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6678571428571428, | |
| "grad_norm": 0.5828749443747008, | |
| "learning_rate": 5.270065757843e-06, | |
| "loss": 0.2424, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6714285714285714, | |
| "grad_norm": 0.7646561775047477, | |
| "learning_rate": 5.168247476807054e-06, | |
| "loss": 0.2737, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.5444525062063674, | |
| "learning_rate": 5.067078519063514e-06, | |
| "loss": 0.1996, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 0.5755758677832743, | |
| "learning_rate": 4.966572480375076e-06, | |
| "loss": 0.1933, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6821428571428572, | |
| "grad_norm": 0.6848265137867346, | |
| "learning_rate": 4.86674286741693e-06, | |
| "loss": 0.2527, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 0.8336972713921105, | |
| "learning_rate": 4.767603095961652e-06, | |
| "loss": 0.2904, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6892857142857143, | |
| "grad_norm": 0.7067542980447211, | |
| "learning_rate": 4.669166489076283e-06, | |
| "loss": 0.2754, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6928571428571428, | |
| "grad_norm": 0.5057729355169331, | |
| "learning_rate": 4.571446275331903e-06, | |
| "loss": 0.1903, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 0.6302783329150187, | |
| "learning_rate": 4.47445558702587e-06, | |
| "loss": 0.2433, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.7066408928533802, | |
| "learning_rate": 4.378207458417035e-06, | |
| "loss": 0.2295, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7035714285714286, | |
| "grad_norm": 0.5840045137089293, | |
| "learning_rate": 4.282714823974088e-06, | |
| "loss": 0.1797, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7071428571428572, | |
| "grad_norm": 0.5993859301367765, | |
| "learning_rate": 4.187990516637361e-06, | |
| "loss": 0.1884, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7107142857142857, | |
| "grad_norm": 0.5400595981826822, | |
| "learning_rate": 4.094047266094225e-06, | |
| "loss": 0.177, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.3473237225491981, | |
| "learning_rate": 4.000897697068418e-06, | |
| "loss": 0.1096, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7178571428571429, | |
| "grad_norm": 0.5910893499694284, | |
| "learning_rate": 3.908554327623425e-06, | |
| "loss": 0.228, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7214285714285714, | |
| "grad_norm": 0.60398021763633, | |
| "learning_rate": 3.817029567480228e-06, | |
| "loss": 0.2608, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.6200500316907418, | |
| "learning_rate": 3.7263357163496118e-06, | |
| "loss": 0.2154, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7285714285714285, | |
| "grad_norm": 0.5380162414775479, | |
| "learning_rate": 3.6364849622792262e-06, | |
| "loss": 0.1595, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 0.7246641798143522, | |
| "learning_rate": 3.5474893800157005e-06, | |
| "loss": 0.2228, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7357142857142858, | |
| "grad_norm": 0.45508651388658605, | |
| "learning_rate": 3.459360929381931e-06, | |
| "loss": 0.1591, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.7392857142857143, | |
| "grad_norm": 0.6429141963072342, | |
| "learning_rate": 3.372111453669864e-06, | |
| "loss": 0.2403, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 0.5855029330521654, | |
| "learning_rate": 3.2857526780488925e-06, | |
| "loss": 0.1934, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7464285714285714, | |
| "grad_norm": 0.5125660874685503, | |
| "learning_rate": 3.2002962079901743e-06, | |
| "loss": 0.1553, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6976778978872136, | |
| "learning_rate": 3.115753527706986e-06, | |
| "loss": 0.2384, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7535714285714286, | |
| "grad_norm": 0.47434866039192386, | |
| "learning_rate": 3.0321359986114096e-06, | |
| "loss": 0.1791, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7571428571428571, | |
| "grad_norm": 0.514888686551834, | |
| "learning_rate": 2.9494548577875195e-06, | |
| "loss": 0.1497, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7607142857142857, | |
| "grad_norm": 0.5887635651830888, | |
| "learning_rate": 2.8677212164812464e-06, | |
| "loss": 0.184, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7642857142857142, | |
| "grad_norm": 0.5947541481450019, | |
| "learning_rate": 2.786946058607187e-06, | |
| "loss": 0.1876, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 0.7110105563805584, | |
| "learning_rate": 2.70714023927251e-06, | |
| "loss": 0.2471, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 0.6416256915388147, | |
| "learning_rate": 2.628314483318178e-06, | |
| "loss": 0.2392, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.5903622152845527, | |
| "learning_rate": 2.5504793838776585e-06, | |
| "loss": 0.1842, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7785714285714286, | |
| "grad_norm": 0.6929636779116681, | |
| "learning_rate": 2.473645400953366e-06, | |
| "loss": 0.2768, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7821428571428571, | |
| "grad_norm": 0.5903013240273918, | |
| "learning_rate": 2.3978228600109564e-06, | |
| "loss": 0.2091, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.624843633893901, | |
| "learning_rate": 2.323021950591743e-06, | |
| "loss": 0.2183, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7892857142857143, | |
| "grad_norm": 0.626404755888219, | |
| "learning_rate": 2.249252724943336e-06, | |
| "loss": 0.236, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7928571428571428, | |
| "grad_norm": 0.6531550403690458, | |
| "learning_rate": 2.176525096668769e-06, | |
| "loss": 0.249, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7964285714285714, | |
| "grad_norm": 0.6861326233839365, | |
| "learning_rate": 2.1048488393942455e-06, | |
| "loss": 0.2167, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5361387918675531, | |
| "learning_rate": 2.0342335854556738e-06, | |
| "loss": 0.2032, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 0.5567786640700992, | |
| "learning_rate": 1.964688824604234e-06, | |
| "loss": 0.2286, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8071428571428572, | |
| "grad_norm": 0.44674720095893333, | |
| "learning_rate": 1.896223902731058e-06, | |
| "loss": 0.1147, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8107142857142857, | |
| "grad_norm": 0.41283033472542874, | |
| "learning_rate": 1.8288480206112879e-06, | |
| "loss": 0.1339, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.8142857142857143, | |
| "grad_norm": 0.7296847675489324, | |
| "learning_rate": 1.7625702326675952e-06, | |
| "loss": 0.2027, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.8178571428571428, | |
| "grad_norm": 0.5271072364545172, | |
| "learning_rate": 1.6973994457534026e-06, | |
| "loss": 0.1712, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 0.672253098189724, | |
| "learning_rate": 1.6333444179559078e-06, | |
| "loss": 0.2548, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.6715310371949257, | |
| "learning_rate": 1.5704137574191202e-06, | |
| "loss": 0.2507, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 0.6242146450408651, | |
| "learning_rate": 1.5086159211870445e-06, | |
| "loss": 0.2378, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8321428571428572, | |
| "grad_norm": 0.8815010954927127, | |
| "learning_rate": 1.447959214067155e-06, | |
| "loss": 0.2415, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8357142857142857, | |
| "grad_norm": 0.563249001616026, | |
| "learning_rate": 1.3884517875143544e-06, | |
| "loss": 0.188, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 0.6913123639322339, | |
| "learning_rate": 1.3301016385355093e-06, | |
| "loss": 0.2719, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8428571428571429, | |
| "grad_norm": 0.40615590184512884, | |
| "learning_rate": 1.2729166086147803e-06, | |
| "loss": 0.1136, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8464285714285714, | |
| "grad_norm": 0.5857754293443003, | |
| "learning_rate": 1.216904382659806e-06, | |
| "loss": 0.234, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.823607304030527, | |
| "learning_rate": 1.1620724879689793e-06, | |
| "loss": 0.2809, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8535714285714285, | |
| "grad_norm": 0.6169005264616082, | |
| "learning_rate": 1.1084282932198543e-06, | |
| "loss": 0.1763, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.6682010561482211, | |
| "learning_rate": 1.0559790074789134e-06, | |
| "loss": 0.2398, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8607142857142858, | |
| "grad_norm": 0.5418866168713049, | |
| "learning_rate": 1.00473167923275e-06, | |
| "loss": 0.1964, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8642857142857143, | |
| "grad_norm": 0.7106090644853337, | |
| "learning_rate": 9.546931954408622e-07, | |
| "loss": 0.2337, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8678571428571429, | |
| "grad_norm": 0.7417349503796012, | |
| "learning_rate": 9.058702806101172e-07, | |
| "loss": 0.2632, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8714285714285714, | |
| "grad_norm": 0.5175232713000997, | |
| "learning_rate": 8.582694958910809e-07, | |
| "loss": 0.193, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.026575312488047, | |
| "learning_rate": 8.118972381962853e-07, | |
| "loss": 0.3553, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8785714285714286, | |
| "grad_norm": 0.6308180674164335, | |
| "learning_rate": 7.667597393405602e-07, | |
| "loss": 0.1723, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8821428571428571, | |
| "grad_norm": 0.4590057669806479, | |
| "learning_rate": 7.228630652035717e-07, | |
| "loss": 0.1531, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 0.666809146038292, | |
| "learning_rate": 6.802131149146374e-07, | |
| "loss": 0.2449, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8892857142857142, | |
| "grad_norm": 0.5482805091255768, | |
| "learning_rate": 6.388156200599726e-07, | |
| "loss": 0.1861, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.5750571040464779, | |
| "learning_rate": 5.986761439124289e-07, | |
| "loss": 0.2074, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8964285714285715, | |
| "grad_norm": 0.8750045040878164, | |
| "learning_rate": 5.598000806838766e-07, | |
| "loss": 0.3201, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.48638037681505203, | |
| "learning_rate": 5.221926548002876e-07, | |
| "loss": 0.1644, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9035714285714286, | |
| "grad_norm": 0.49478576603633184, | |
| "learning_rate": 4.858589201996433e-07, | |
| "loss": 0.1918, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9071428571428571, | |
| "grad_norm": 0.598927975154658, | |
| "learning_rate": 4.5080375965275256e-07, | |
| "loss": 0.1711, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 0.6197004195423823, | |
| "learning_rate": 4.1703188410707087e-07, | |
| "loss": 0.1928, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.6313460016595605, | |
| "learning_rate": 3.845478320536178e-07, | |
| "loss": 0.2549, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9178571428571428, | |
| "grad_norm": 0.8956056686691475, | |
| "learning_rate": 3.5335596891705406e-07, | |
| "loss": 0.2841, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.9214285714285714, | |
| "grad_norm": 0.4910383596976099, | |
| "learning_rate": 3.2346048646903494e-07, | |
| "loss": 0.1827, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.6510508585080252, | |
| "learning_rate": 2.9486540226488556e-07, | |
| "loss": 0.2415, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.7480168465217824, | |
| "learning_rate": 2.6757455910370486e-07, | |
| "loss": 0.2374, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9321428571428572, | |
| "grad_norm": 0.47073582050435153, | |
| "learning_rate": 2.41591624511931e-07, | |
| "loss": 0.1982, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9357142857142857, | |
| "grad_norm": 0.5954530106280657, | |
| "learning_rate": 2.169200902504842e-07, | |
| "loss": 0.2321, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9392857142857143, | |
| "grad_norm": 0.5083864479298373, | |
| "learning_rate": 1.9356327184551716e-07, | |
| "loss": 0.2056, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 0.6176017928735918, | |
| "learning_rate": 1.7152430814285303e-07, | |
| "loss": 0.2248, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 0.8794551027561036, | |
| "learning_rate": 1.5080616088616884e-07, | |
| "loss": 0.276, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5963379475552784, | |
| "learning_rate": 1.3141161431896809e-07, | |
| "loss": 0.2682, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9535714285714286, | |
| "grad_norm": 0.6298840458046066, | |
| "learning_rate": 1.1334327481042573e-07, | |
| "loss": 0.1362, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9571428571428572, | |
| "grad_norm": 0.5539259365985325, | |
| "learning_rate": 9.660357050512158e-08, | |
| "loss": 0.2111, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9607142857142857, | |
| "grad_norm": 0.7343899826298071, | |
| "learning_rate": 8.119475099673035e-08, | |
| "loss": 0.2867, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.5275488999179762, | |
| "learning_rate": 6.711888702570556e-08, | |
| "loss": 0.1578, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9678571428571429, | |
| "grad_norm": 0.5471929852386093, | |
| "learning_rate": 5.437787020100116e-08, | |
| "loss": 0.1831, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 0.7137357035927524, | |
| "learning_rate": 4.297341274586475e-08, | |
| "loss": 0.2711, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.6861484126427937, | |
| "learning_rate": 3.290704726773619e-08, | |
| "loss": 0.2091, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9785714285714285, | |
| "grad_norm": 0.7660872222264061, | |
| "learning_rate": 2.4180126552284523e-08, | |
| "loss": 0.2517, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 0.6124965928901931, | |
| "learning_rate": 1.6793823381614506e-08, | |
| "loss": 0.2005, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9857142857142858, | |
| "grad_norm": 0.6082415014874585, | |
| "learning_rate": 1.0749130376659366e-08, | |
| "loss": 0.224, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9892857142857143, | |
| "grad_norm": 0.6973577336250109, | |
| "learning_rate": 6.046859863781951e-09, | |
| "loss": 0.2659, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9928571428571429, | |
| "grad_norm": 0.6364613016078383, | |
| "learning_rate": 2.6876437656153663e-09, | |
| "loss": 0.2133, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9964285714285714, | |
| "grad_norm": 0.5410559341159564, | |
| "learning_rate": 6.719335161364804e-10, | |
| "loss": 0.1768, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.684252146873541, | |
| "learning_rate": 0.0, | |
| "loss": 0.1917, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 280, | |
| "total_flos": 222415492415488.0, | |
| "train_loss": 0.3071911358673658, | |
| "train_runtime": 1347.2293, | |
| "train_samples_per_second": 26.594, | |
| "train_steps_per_second": 0.208 | |
| } | |
| ], | |
| "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": 222415492415488.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |