| { | |
| "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": 4.908648608990839, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.7961, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007142857142857143, | |
| "grad_norm": 4.592180104184582, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.713, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010714285714285714, | |
| "grad_norm": 4.69308510471444, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.6831, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014285714285714285, | |
| "grad_norm": 3.653347608269976, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.5785, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.017857142857142856, | |
| "grad_norm": 2.804649047110185, | |
| "learning_rate": 1.1111111111111113e-05, | |
| "loss": 1.5988, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02142857142857143, | |
| "grad_norm": 3.1675123269741516, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.4795, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 3.6958086333820566, | |
| "learning_rate": 1.555555555555556e-05, | |
| "loss": 1.5327, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 5.724480820109564, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 1.4916, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03214285714285714, | |
| "grad_norm": 4.073008923111407, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4973, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 2.6759891152038526, | |
| "learning_rate": 1.9999328066483867e-05, | |
| "loss": 1.3998, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.039285714285714285, | |
| "grad_norm": 3.2603665327453277, | |
| "learning_rate": 1.9997312356234385e-05, | |
| "loss": 1.3604, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04285714285714286, | |
| "grad_norm": 3.6130206264673457, | |
| "learning_rate": 1.999395314013622e-05, | |
| "loss": 1.3048, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04642857142857143, | |
| "grad_norm": 2.661927084764187, | |
| "learning_rate": 1.998925086962334e-05, | |
| "loss": 1.2912, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.576973416368585, | |
| "learning_rate": 1.998320617661839e-05, | |
| "loss": 1.2325, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05357142857142857, | |
| "grad_norm": 2.4728531527823385, | |
| "learning_rate": 1.997581987344772e-05, | |
| "loss": 1.2227, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 2.6919633979798023, | |
| "learning_rate": 1.9967092952732266e-05, | |
| "loss": 1.1903, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.060714285714285714, | |
| "grad_norm": 2.8589949375920414, | |
| "learning_rate": 1.9957026587254136e-05, | |
| "loss": 1.1843, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06428571428571428, | |
| "grad_norm": 2.3469600295209823, | |
| "learning_rate": 1.9945622129799e-05, | |
| "loss": 1.1089, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06785714285714285, | |
| "grad_norm": 2.5116022382668968, | |
| "learning_rate": 1.9932881112974298e-05, | |
| "loss": 1.0887, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 2.2064320702567413, | |
| "learning_rate": 1.9918805249003272e-05, | |
| "loss": 1.0124, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 2.5511459947910557, | |
| "learning_rate": 1.9903396429494882e-05, | |
| "loss": 0.9848, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07857142857142857, | |
| "grad_norm": 2.074497064278477, | |
| "learning_rate": 1.9886656725189575e-05, | |
| "loss": 0.9706, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08214285714285714, | |
| "grad_norm": 2.0029563935827714, | |
| "learning_rate": 1.9868588385681035e-05, | |
| "loss": 0.9629, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 2.166398531324486, | |
| "learning_rate": 1.9849193839113833e-05, | |
| "loss": 0.8597, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08928571428571429, | |
| "grad_norm": 2.034388942107051, | |
| "learning_rate": 1.9828475691857148e-05, | |
| "loss": 0.8356, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09285714285714286, | |
| "grad_norm": 1.8616897229035658, | |
| "learning_rate": 1.9806436728154484e-05, | |
| "loss": 0.7967, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09642857142857143, | |
| "grad_norm": 1.937793646144307, | |
| "learning_rate": 1.9783079909749516e-05, | |
| "loss": 0.7902, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.03802418267336, | |
| "learning_rate": 1.975840837548807e-05, | |
| "loss": 0.8302, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10357142857142858, | |
| "grad_norm": 1.7299975214673873, | |
| "learning_rate": 1.9732425440896298e-05, | |
| "loss": 0.8622, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 1.806743882141901, | |
| "learning_rate": 1.9705134597735113e-05, | |
| "loss": 0.7735, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11071428571428571, | |
| "grad_norm": 1.8544287022749248, | |
| "learning_rate": 1.9676539513530967e-05, | |
| "loss": 0.6243, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 1.8369289759240388, | |
| "learning_rate": 1.9646644031082948e-05, | |
| "loss": 0.6859, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11785714285714285, | |
| "grad_norm": 1.941599747726953, | |
| "learning_rate": 1.9615452167946383e-05, | |
| "loss": 0.7323, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.12142857142857143, | |
| "grad_norm": 1.9016904671057466, | |
| "learning_rate": 1.958296811589293e-05, | |
| "loss": 0.6924, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 2.0994171001442754, | |
| "learning_rate": 1.954919624034725e-05, | |
| "loss": 0.7021, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12857142857142856, | |
| "grad_norm": 1.8461885488185878, | |
| "learning_rate": 1.951414107980036e-05, | |
| "loss": 0.6776, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13214285714285715, | |
| "grad_norm": 1.697750537080288, | |
| "learning_rate": 1.9477807345199717e-05, | |
| "loss": 0.7172, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1357142857142857, | |
| "grad_norm": 1.8267671959875706, | |
| "learning_rate": 1.9440199919316125e-05, | |
| "loss": 0.6439, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1392857142857143, | |
| "grad_norm": 1.6993602930346685, | |
| "learning_rate": 1.9401323856087573e-05, | |
| "loss": 0.618, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 1.6727471871735518, | |
| "learning_rate": 1.936118437994003e-05, | |
| "loss": 0.6352, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14642857142857144, | |
| "grad_norm": 1.754512057986924, | |
| "learning_rate": 1.9319786885085366e-05, | |
| "loss": 0.5894, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.740151549924287, | |
| "learning_rate": 1.927713693479643e-05, | |
| "loss": 0.6899, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.15357142857142858, | |
| "grad_norm": 1.594981649980205, | |
| "learning_rate": 1.923324026065944e-05, | |
| "loss": 0.5648, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.15714285714285714, | |
| "grad_norm": 1.5576937266056854, | |
| "learning_rate": 1.918810276180372e-05, | |
| "loss": 0.5596, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.16071428571428573, | |
| "grad_norm": 1.7216253449123322, | |
| "learning_rate": 1.9141730504108923e-05, | |
| "loss": 0.6042, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16428571428571428, | |
| "grad_norm": 1.6040787546322597, | |
| "learning_rate": 1.9094129719389886e-05, | |
| "loss": 0.5168, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16785714285714284, | |
| "grad_norm": 1.731120413859831, | |
| "learning_rate": 1.904530680455914e-05, | |
| "loss": 0.6288, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 1.626971149833354, | |
| "learning_rate": 1.8995268320767254e-05, | |
| "loss": 0.5288, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 1.4082904642307315, | |
| "learning_rate": 1.894402099252109e-05, | |
| "loss": 0.5477, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 1.6671028870240991, | |
| "learning_rate": 1.889157170678015e-05, | |
| "loss": 0.5541, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18214285714285713, | |
| "grad_norm": 1.629789549819335, | |
| "learning_rate": 1.8837927512031022e-05, | |
| "loss": 0.4901, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.18571428571428572, | |
| "grad_norm": 1.565700755496585, | |
| "learning_rate": 1.8783095617340193e-05, | |
| "loss": 0.4624, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.18928571428571428, | |
| "grad_norm": 1.4678500727905517, | |
| "learning_rate": 1.872708339138522e-05, | |
| "loss": 0.4271, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19285714285714287, | |
| "grad_norm": 1.4609229980872802, | |
| "learning_rate": 1.866989836146449e-05, | |
| "loss": 0.4376, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.19642857142857142, | |
| "grad_norm": 1.6797655019359938, | |
| "learning_rate": 1.861154821248565e-05, | |
| "loss": 0.5878, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.577613646211982, | |
| "learning_rate": 1.8552040785932846e-05, | |
| "loss": 0.4783, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.20357142857142857, | |
| "grad_norm": 1.7027221419545615, | |
| "learning_rate": 1.849138407881296e-05, | |
| "loss": 0.3975, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.20714285714285716, | |
| "grad_norm": 1.670046977724538, | |
| "learning_rate": 1.8429586242580884e-05, | |
| "loss": 0.4367, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21071428571428572, | |
| "grad_norm": 1.465983705393668, | |
| "learning_rate": 1.8366655582044096e-05, | |
| "loss": 0.3857, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 1.619649836022893, | |
| "learning_rate": 1.83026005542466e-05, | |
| "loss": 0.4711, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21785714285714286, | |
| "grad_norm": 1.570708639647691, | |
| "learning_rate": 1.8237429767332407e-05, | |
| "loss": 0.4524, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.22142857142857142, | |
| "grad_norm": 1.476746454110955, | |
| "learning_rate": 1.8171151979388715e-05, | |
| "loss": 0.3994, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 1.6497165796916549, | |
| "learning_rate": 1.8103776097268942e-05, | |
| "loss": 0.4169, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 1.30798054852489, | |
| "learning_rate": 1.803531117539577e-05, | |
| "loss": 0.385, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.23214285714285715, | |
| "grad_norm": 1.3237517166357549, | |
| "learning_rate": 1.7965766414544328e-05, | |
| "loss": 0.348, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2357142857142857, | |
| "grad_norm": 1.6055539374817416, | |
| "learning_rate": 1.7895151160605758e-05, | |
| "loss": 0.4277, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2392857142857143, | |
| "grad_norm": 1.6951560958360459, | |
| "learning_rate": 1.782347490333123e-05, | |
| "loss": 0.4142, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.24285714285714285, | |
| "grad_norm": 1.2191039708896538, | |
| "learning_rate": 1.775074727505667e-05, | |
| "loss": 0.2954, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.24642857142857144, | |
| "grad_norm": 1.5607149280052393, | |
| "learning_rate": 1.7676978049408262e-05, | |
| "loss": 0.3828, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.5038483401175802, | |
| "learning_rate": 1.7602177139989046e-05, | |
| "loss": 0.405, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25357142857142856, | |
| "grad_norm": 1.6682214568606215, | |
| "learning_rate": 1.7526354599046637e-05, | |
| "loss": 0.3953, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 1.347036705570398, | |
| "learning_rate": 1.7449520616122344e-05, | |
| "loss": 0.3629, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.26071428571428573, | |
| "grad_norm": 1.2414625150772796, | |
| "learning_rate": 1.7371685516681825e-05, | |
| "loss": 0.3161, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2642857142857143, | |
| "grad_norm": 1.3436886422108463, | |
| "learning_rate": 1.7292859760727493e-05, | |
| "loss": 0.3269, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.26785714285714285, | |
| "grad_norm": 1.3530913321365612, | |
| "learning_rate": 1.721305394139282e-05, | |
| "loss": 0.2728, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2714285714285714, | |
| "grad_norm": 1.222920831545463, | |
| "learning_rate": 1.7132278783518756e-05, | |
| "loss": 0.2876, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 1.5358149513341914, | |
| "learning_rate": 1.7050545142212483e-05, | |
| "loss": 0.3957, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2785714285714286, | |
| "grad_norm": 1.4149548491141548, | |
| "learning_rate": 1.696786400138859e-05, | |
| "loss": 0.3104, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.28214285714285714, | |
| "grad_norm": 1.7669148749254266, | |
| "learning_rate": 1.6884246472293018e-05, | |
| "loss": 0.3743, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 1.265504102534398, | |
| "learning_rate": 1.679970379200983e-05, | |
| "loss": 0.317, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2892857142857143, | |
| "grad_norm": 1.2888395180660637, | |
| "learning_rate": 1.6714247321951106e-05, | |
| "loss": 0.342, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.29285714285714287, | |
| "grad_norm": 1.422200915278466, | |
| "learning_rate": 1.6627888546330136e-05, | |
| "loss": 0.4185, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.29642857142857143, | |
| "grad_norm": 1.4661329628560202, | |
| "learning_rate": 1.654063907061807e-05, | |
| "loss": 0.3713, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.4678935830745146, | |
| "learning_rate": 1.64525106199843e-05, | |
| "loss": 0.3066, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.30357142857142855, | |
| "grad_norm": 1.2825311493245366, | |
| "learning_rate": 1.6363515037720774e-05, | |
| "loss": 0.3134, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30714285714285716, | |
| "grad_norm": 1.3008599464968365, | |
| "learning_rate": 1.6273664283650393e-05, | |
| "loss": 0.3824, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3107142857142857, | |
| "grad_norm": 1.5609153598311494, | |
| "learning_rate": 1.6182970432519772e-05, | |
| "loss": 0.2641, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 1.321725446631543, | |
| "learning_rate": 1.609144567237658e-05, | |
| "loss": 0.3433, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.31785714285714284, | |
| "grad_norm": 1.230389929756339, | |
| "learning_rate": 1.5999102302931585e-05, | |
| "loss": 0.2882, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 1.282470788355045, | |
| "learning_rate": 1.5905952733905777e-05, | |
| "loss": 0.4001, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 1.244760100205609, | |
| "learning_rate": 1.5812009483362643e-05, | |
| "loss": 0.3303, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.32857142857142857, | |
| "grad_norm": 1.3077612857336067, | |
| "learning_rate": 1.5717285176025913e-05, | |
| "loss": 0.3009, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.33214285714285713, | |
| "grad_norm": 1.5007751444825745, | |
| "learning_rate": 1.5621792541582968e-05, | |
| "loss": 0.4037, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3357142857142857, | |
| "grad_norm": 1.2889530196987555, | |
| "learning_rate": 1.552554441297413e-05, | |
| "loss": 0.3166, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3392857142857143, | |
| "grad_norm": 1.471774386098153, | |
| "learning_rate": 1.5428553724668103e-05, | |
| "loss": 0.3141, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 1.2425633719745615, | |
| "learning_rate": 1.533083351092372e-05, | |
| "loss": 0.3032, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3464285714285714, | |
| "grad_norm": 1.105289454224738, | |
| "learning_rate": 1.5232396904038352e-05, | |
| "loss": 0.2762, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.2970218695315465, | |
| "learning_rate": 1.5133257132583074e-05, | |
| "loss": 0.3101, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3535714285714286, | |
| "grad_norm": 1.3017528350777372, | |
| "learning_rate": 1.503342751962493e-05, | |
| "loss": 0.3731, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 1.3494066259729216, | |
| "learning_rate": 1.4932921480936491e-05, | |
| "loss": 0.2738, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3607142857142857, | |
| "grad_norm": 1.2452542674259262, | |
| "learning_rate": 1.4831752523192949e-05, | |
| "loss": 0.2732, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.36428571428571427, | |
| "grad_norm": 1.1826372513240704, | |
| "learning_rate": 1.4729934242157005e-05, | |
| "loss": 0.2772, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3678571428571429, | |
| "grad_norm": 1.3655889492363933, | |
| "learning_rate": 1.4627480320851775e-05, | |
| "loss": 0.3117, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 1.0905245915838764, | |
| "learning_rate": 1.4524404527721977e-05, | |
| "loss": 0.2421, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 1.1892908660572088, | |
| "learning_rate": 1.4420720714783635e-05, | |
| "loss": 0.3403, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.37857142857142856, | |
| "grad_norm": 1.3612551275939382, | |
| "learning_rate": 1.4316442815762543e-05, | |
| "loss": 0.3111, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3821428571428571, | |
| "grad_norm": 1.0293414732881423, | |
| "learning_rate": 1.4211584844221771e-05, | |
| "loss": 0.2198, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.38571428571428573, | |
| "grad_norm": 1.0973339441554495, | |
| "learning_rate": 1.4106160891678422e-05, | |
| "loss": 0.2438, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3892857142857143, | |
| "grad_norm": 0.9728591339955907, | |
| "learning_rate": 1.4000185125709919e-05, | |
| "loss": 0.2189, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 1.0719702208838044, | |
| "learning_rate": 1.3893671788050073e-05, | |
| "loss": 0.2782, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3964285714285714, | |
| "grad_norm": 0.9706227479864399, | |
| "learning_rate": 1.3786635192675184e-05, | |
| "loss": 0.227, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.4196235328043891, | |
| "learning_rate": 1.3679089723880427e-05, | |
| "loss": 0.3279, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4035714285714286, | |
| "grad_norm": 0.9875713875583682, | |
| "learning_rate": 1.35710498343468e-05, | |
| "loss": 0.2223, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.40714285714285714, | |
| "grad_norm": 1.2683114513213385, | |
| "learning_rate": 1.3462530043198874e-05, | |
| "loss": 0.3193, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4107142857142857, | |
| "grad_norm": 1.1548967936561843, | |
| "learning_rate": 1.3353544934053618e-05, | |
| "loss": 0.2418, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4142857142857143, | |
| "grad_norm": 1.1578835473542461, | |
| "learning_rate": 1.324410915306055e-05, | |
| "loss": 0.2741, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.41785714285714287, | |
| "grad_norm": 0.867810988436804, | |
| "learning_rate": 1.3134237406933493e-05, | |
| "loss": 0.2182, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.42142857142857143, | |
| "grad_norm": 1.226765456766979, | |
| "learning_rate": 1.3023944460974183e-05, | |
| "loss": 0.2486, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 1.2047266305046154, | |
| "learning_rate": 1.2913245137088024e-05, | |
| "loss": 0.2654, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 1.0309280957839033, | |
| "learning_rate": 1.2802154311792196e-05, | |
| "loss": 0.2194, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.43214285714285716, | |
| "grad_norm": 1.1216683723429337, | |
| "learning_rate": 1.2690686914216475e-05, | |
| "loss": 0.2471, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4357142857142857, | |
| "grad_norm": 0.7992505679628088, | |
| "learning_rate": 1.2578857924096935e-05, | |
| "loss": 0.2053, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4392857142857143, | |
| "grad_norm": 0.9521815104123159, | |
| "learning_rate": 1.2466682369762883e-05, | |
| "loss": 0.218, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.44285714285714284, | |
| "grad_norm": 1.0131891476301753, | |
| "learning_rate": 1.2354175326117252e-05, | |
| "loss": 0.2797, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 1.1514536611512578, | |
| "learning_rate": 1.2241351912610726e-05, | |
| "loss": 0.326, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.4037867589116553, | |
| "learning_rate": 1.212822729120989e-05, | |
| "loss": 0.2137, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.45357142857142857, | |
| "grad_norm": 0.9592307976271547, | |
| "learning_rate": 1.2014816664359671e-05, | |
| "loss": 0.2489, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 1.239268404867852, | |
| "learning_rate": 1.190113527294032e-05, | |
| "loss": 0.3135, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4607142857142857, | |
| "grad_norm": 0.9664072810151023, | |
| "learning_rate": 1.178719839421925e-05, | |
| "loss": 0.2025, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 1.000615478826764, | |
| "learning_rate": 1.1673021339797967e-05, | |
| "loss": 0.2701, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.46785714285714286, | |
| "grad_norm": 1.0284186941827786, | |
| "learning_rate": 1.15586194535544e-05, | |
| "loss": 0.2448, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4714285714285714, | |
| "grad_norm": 0.7308129021048215, | |
| "learning_rate": 1.1444008109580884e-05, | |
| "loss": 0.1569, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 0.8601356644168414, | |
| "learning_rate": 1.1329202710118088e-05, | |
| "loss": 0.2752, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4785714285714286, | |
| "grad_norm": 1.1174847463212987, | |
| "learning_rate": 1.1214218683485159e-05, | |
| "loss": 0.2813, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.48214285714285715, | |
| "grad_norm": 0.9222986025138733, | |
| "learning_rate": 1.1099071482006361e-05, | |
| "loss": 0.2127, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 1.0153753324402643, | |
| "learning_rate": 1.0983776579934483e-05, | |
| "loss": 0.3222, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.48928571428571427, | |
| "grad_norm": 1.1498734203876837, | |
| "learning_rate": 1.0868349471371316e-05, | |
| "loss": 0.3267, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4928571428571429, | |
| "grad_norm": 1.0327891450611766, | |
| "learning_rate": 1.0752805668185442e-05, | |
| "loss": 0.2247, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.49642857142857144, | |
| "grad_norm": 1.04721921113995, | |
| "learning_rate": 1.0637160697927651e-05, | |
| "loss": 0.285, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.9226664091190546, | |
| "learning_rate": 1.0521430101744238e-05, | |
| "loss": 0.2138, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5035714285714286, | |
| "grad_norm": 1.2238486576835246, | |
| "learning_rate": 1.040562943228849e-05, | |
| "loss": 0.2742, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5071428571428571, | |
| "grad_norm": 0.8784185700324445, | |
| "learning_rate": 1.0289774251630602e-05, | |
| "loss": 0.1844, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5107142857142857, | |
| "grad_norm": 1.105225096168923, | |
| "learning_rate": 1.0173880129166358e-05, | |
| "loss": 0.2664, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 0.7220119823796761, | |
| "learning_rate": 1.0057962639524799e-05, | |
| "loss": 0.1911, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5178571428571429, | |
| "grad_norm": 0.9901039928611911, | |
| "learning_rate": 9.942037360475205e-06, | |
| "loss": 0.2047, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5214285714285715, | |
| "grad_norm": 0.990343263513173, | |
| "learning_rate": 9.826119870833644e-06, | |
| "loss": 0.2482, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 1.0814804344690816, | |
| "learning_rate": 9.710225748369402e-06, | |
| "loss": 0.2791, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5285714285714286, | |
| "grad_norm": 0.6699567812276747, | |
| "learning_rate": 9.594370567711512e-06, | |
| "loss": 0.1689, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5321428571428571, | |
| "grad_norm": 1.193514972349772, | |
| "learning_rate": 9.478569898255765e-06, | |
| "loss": 0.3267, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 0.718259276227398, | |
| "learning_rate": 9.362839302072354e-06, | |
| "loss": 0.133, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5392857142857143, | |
| "grad_norm": 1.071572674237938, | |
| "learning_rate": 9.247194331814561e-06, | |
| "loss": 0.2434, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 1.0026565147511441, | |
| "learning_rate": 9.131650528628688e-06, | |
| "loss": 0.2248, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5464285714285714, | |
| "grad_norm": 0.7606699824597489, | |
| "learning_rate": 9.016223420065519e-06, | |
| "loss": 0.2294, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.9509831030302495, | |
| "learning_rate": 8.900928517993644e-06, | |
| "loss": 0.2921, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5535714285714286, | |
| "grad_norm": 0.9871198309746722, | |
| "learning_rate": 8.785781316514841e-06, | |
| "loss": 0.2715, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5571428571428572, | |
| "grad_norm": 0.9057094151212118, | |
| "learning_rate": 8.670797289881915e-06, | |
| "loss": 0.2449, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5607142857142857, | |
| "grad_norm": 0.8772737026647174, | |
| "learning_rate": 8.555991890419116e-06, | |
| "loss": 0.2564, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5642857142857143, | |
| "grad_norm": 0.8695626353738833, | |
| "learning_rate": 8.441380546445603e-06, | |
| "loss": 0.2229, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5678571428571428, | |
| "grad_norm": 1.073152464539624, | |
| "learning_rate": 8.326978660202034e-06, | |
| "loss": 0.3386, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.012416344543051, | |
| "learning_rate": 8.212801605780754e-06, | |
| "loss": 0.2207, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.7691240257824403, | |
| "learning_rate": 8.098864727059685e-06, | |
| "loss": 0.2051, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5785714285714286, | |
| "grad_norm": 1.1387099994481609, | |
| "learning_rate": 7.985183335640332e-06, | |
| "loss": 0.3062, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5821428571428572, | |
| "grad_norm": 0.6548715665719999, | |
| "learning_rate": 7.871772708790114e-06, | |
| "loss": 0.1671, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5857142857142857, | |
| "grad_norm": 1.0074068488831938, | |
| "learning_rate": 7.758648087389277e-06, | |
| "loss": 0.2457, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5892857142857143, | |
| "grad_norm": 1.3590905269175524, | |
| "learning_rate": 7.64582467388275e-06, | |
| "loss": 0.2763, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5928571428571429, | |
| "grad_norm": 0.6688733178759887, | |
| "learning_rate": 7.533317630237117e-06, | |
| "loss": 0.2107, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5964285714285714, | |
| "grad_norm": 1.1521044685448598, | |
| "learning_rate": 7.421142075903067e-06, | |
| "loss": 0.3182, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.8663913856359603, | |
| "learning_rate": 7.3093130857835245e-06, | |
| "loss": 0.1907, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6035714285714285, | |
| "grad_norm": 1.0039720576945341, | |
| "learning_rate": 7.197845688207805e-06, | |
| "loss": 0.2518, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 0.8382491705573645, | |
| "learning_rate": 7.086754862911982e-06, | |
| "loss": 0.2304, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6107142857142858, | |
| "grad_norm": 1.102459135393568, | |
| "learning_rate": 6.976055539025819e-06, | |
| "loss": 0.2607, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6142857142857143, | |
| "grad_norm": 0.8659998101978996, | |
| "learning_rate": 6.865762593066514e-06, | |
| "loss": 0.2173, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6178571428571429, | |
| "grad_norm": 1.038371053979899, | |
| "learning_rate": 6.755890846939454e-06, | |
| "loss": 0.2449, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6214285714285714, | |
| "grad_norm": 0.9112242172551936, | |
| "learning_rate": 6.646455065946386e-06, | |
| "loss": 0.2446, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.8416570702077658, | |
| "learning_rate": 6.537469956801128e-06, | |
| "loss": 0.2074, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 0.7222497910641669, | |
| "learning_rate": 6.428950165653204e-06, | |
| "loss": 0.2037, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6321428571428571, | |
| "grad_norm": 0.9350757855602984, | |
| "learning_rate": 6.320910276119576e-06, | |
| "loss": 0.26, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6357142857142857, | |
| "grad_norm": 0.8975477713777471, | |
| "learning_rate": 6.213364807324817e-06, | |
| "loss": 0.2336, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6392857142857142, | |
| "grad_norm": 0.713247964590351, | |
| "learning_rate": 6.106328211949928e-06, | |
| "loss": 0.1881, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 0.7432676487695248, | |
| "learning_rate": 5.999814874290084e-06, | |
| "loss": 0.188, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6464285714285715, | |
| "grad_norm": 0.9375433226321466, | |
| "learning_rate": 5.893839108321584e-06, | |
| "loss": 0.2229, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.9035178928866511, | |
| "learning_rate": 5.7884151557782305e-06, | |
| "loss": 0.268, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6535714285714286, | |
| "grad_norm": 1.0861928031153827, | |
| "learning_rate": 5.68355718423746e-06, | |
| "loss": 0.2536, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 0.9249874842289382, | |
| "learning_rate": 5.579279285216369e-06, | |
| "loss": 0.2984, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6607142857142857, | |
| "grad_norm": 0.9641440586255716, | |
| "learning_rate": 5.4755954722780236e-06, | |
| "loss": 0.2843, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6642857142857143, | |
| "grad_norm": 1.047518256416677, | |
| "learning_rate": 5.372519679148227e-06, | |
| "loss": 0.2846, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6678571428571428, | |
| "grad_norm": 0.8899779907249936, | |
| "learning_rate": 5.270065757843e-06, | |
| "loss": 0.2814, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6714285714285714, | |
| "grad_norm": 1.257826620638329, | |
| "learning_rate": 5.168247476807054e-06, | |
| "loss": 0.3616, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.8378870170037112, | |
| "learning_rate": 5.067078519063514e-06, | |
| "loss": 0.206, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 0.8499116136055077, | |
| "learning_rate": 4.966572480375076e-06, | |
| "loss": 0.2655, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6821428571428572, | |
| "grad_norm": 1.0500985457733107, | |
| "learning_rate": 4.86674286741693e-06, | |
| "loss": 0.2423, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 1.1824577867055674, | |
| "learning_rate": 4.767603095961652e-06, | |
| "loss": 0.3178, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6892857142857143, | |
| "grad_norm": 1.0135768392531344, | |
| "learning_rate": 4.669166489076283e-06, | |
| "loss": 0.3272, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6928571428571428, | |
| "grad_norm": 0.7599653832936717, | |
| "learning_rate": 4.571446275331903e-06, | |
| "loss": 0.2039, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6964285714285714, | |
| "grad_norm": 1.0212451468927062, | |
| "learning_rate": 4.47445558702587e-06, | |
| "loss": 0.285, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.9471963980854589, | |
| "learning_rate": 4.378207458417035e-06, | |
| "loss": 0.2724, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7035714285714286, | |
| "grad_norm": 0.8971393901363991, | |
| "learning_rate": 4.282714823974088e-06, | |
| "loss": 0.1893, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7071428571428572, | |
| "grad_norm": 0.9659760236519076, | |
| "learning_rate": 4.187990516637361e-06, | |
| "loss": 0.2423, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7107142857142857, | |
| "grad_norm": 0.8988943829563382, | |
| "learning_rate": 4.094047266094225e-06, | |
| "loss": 0.2479, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.6167861266259378, | |
| "learning_rate": 4.000897697068418e-06, | |
| "loss": 0.186, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7178571428571429, | |
| "grad_norm": 0.9931841997154047, | |
| "learning_rate": 3.908554327623425e-06, | |
| "loss": 0.2703, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7214285714285714, | |
| "grad_norm": 1.0515423629697425, | |
| "learning_rate": 3.817029567480228e-06, | |
| "loss": 0.2872, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 1.0078364499059038, | |
| "learning_rate": 3.7263357163496118e-06, | |
| "loss": 0.2771, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7285714285714285, | |
| "grad_norm": 0.7946329327315834, | |
| "learning_rate": 3.6364849622792262e-06, | |
| "loss": 0.2636, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7321428571428571, | |
| "grad_norm": 1.0961952413792404, | |
| "learning_rate": 3.5474893800157005e-06, | |
| "loss": 0.2644, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7357142857142858, | |
| "grad_norm": 0.8310362218238262, | |
| "learning_rate": 3.459360929381931e-06, | |
| "loss": 0.1621, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.7392857142857143, | |
| "grad_norm": 0.9112572267088846, | |
| "learning_rate": 3.372111453669864e-06, | |
| "loss": 0.2583, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 0.8328762351773727, | |
| "learning_rate": 3.2857526780488925e-06, | |
| "loss": 0.252, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7464285714285714, | |
| "grad_norm": 0.7790083845874031, | |
| "learning_rate": 3.2002962079901743e-06, | |
| "loss": 0.1739, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.8937099611163849, | |
| "learning_rate": 3.115753527706986e-06, | |
| "loss": 0.2641, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7535714285714286, | |
| "grad_norm": 0.665235869076356, | |
| "learning_rate": 3.0321359986114096e-06, | |
| "loss": 0.1929, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7571428571428571, | |
| "grad_norm": 0.9975910293748517, | |
| "learning_rate": 2.9494548577875195e-06, | |
| "loss": 0.2071, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7607142857142857, | |
| "grad_norm": 0.8635252751329144, | |
| "learning_rate": 2.8677212164812464e-06, | |
| "loss": 0.2123, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7642857142857142, | |
| "grad_norm": 0.9914555320649461, | |
| "learning_rate": 2.786946058607187e-06, | |
| "loss": 0.3021, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7678571428571429, | |
| "grad_norm": 1.067677741156891, | |
| "learning_rate": 2.70714023927251e-06, | |
| "loss": 0.3197, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 1.0127321773951852, | |
| "learning_rate": 2.628314483318178e-06, | |
| "loss": 0.2979, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.8879989714883786, | |
| "learning_rate": 2.5504793838776585e-06, | |
| "loss": 0.2181, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7785714285714286, | |
| "grad_norm": 1.0116373979132076, | |
| "learning_rate": 2.473645400953366e-06, | |
| "loss": 0.3007, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7821428571428571, | |
| "grad_norm": 0.8314841547996328, | |
| "learning_rate": 2.3978228600109564e-06, | |
| "loss": 0.275, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 0.9162454344653038, | |
| "learning_rate": 2.323021950591743e-06, | |
| "loss": 0.2748, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7892857142857143, | |
| "grad_norm": 1.0145382352758672, | |
| "learning_rate": 2.249252724943336e-06, | |
| "loss": 0.2701, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7928571428571428, | |
| "grad_norm": 0.8666753235531185, | |
| "learning_rate": 2.176525096668769e-06, | |
| "loss": 0.254, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7964285714285714, | |
| "grad_norm": 0.8928533620858623, | |
| "learning_rate": 2.1048488393942455e-06, | |
| "loss": 0.2623, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.7977864862031556, | |
| "learning_rate": 2.0342335854556738e-06, | |
| "loss": 0.2331, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8035714285714286, | |
| "grad_norm": 0.9396848764074734, | |
| "learning_rate": 1.964688824604234e-06, | |
| "loss": 0.2283, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8071428571428572, | |
| "grad_norm": 0.7043691856433903, | |
| "learning_rate": 1.896223902731058e-06, | |
| "loss": 0.1487, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8107142857142857, | |
| "grad_norm": 0.8028110829154207, | |
| "learning_rate": 1.8288480206112879e-06, | |
| "loss": 0.178, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.8142857142857143, | |
| "grad_norm": 0.9551250104442663, | |
| "learning_rate": 1.7625702326675952e-06, | |
| "loss": 0.2512, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.8178571428571428, | |
| "grad_norm": 0.8542011521297516, | |
| "learning_rate": 1.6973994457534026e-06, | |
| "loss": 0.2858, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 0.8529278139532056, | |
| "learning_rate": 1.6333444179559078e-06, | |
| "loss": 0.2351, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.9391405907710427, | |
| "learning_rate": 1.5704137574191202e-06, | |
| "loss": 0.2612, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 0.9898771118350357, | |
| "learning_rate": 1.5086159211870445e-06, | |
| "loss": 0.2806, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8321428571428572, | |
| "grad_norm": 1.3372111053303188, | |
| "learning_rate": 1.447959214067155e-06, | |
| "loss": 0.3463, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8357142857142857, | |
| "grad_norm": 0.8321272930723775, | |
| "learning_rate": 1.3884517875143544e-06, | |
| "loss": 0.2537, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8392857142857143, | |
| "grad_norm": 1.2134366545999935, | |
| "learning_rate": 1.3301016385355093e-06, | |
| "loss": 0.3719, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8428571428571429, | |
| "grad_norm": 0.6334539224525021, | |
| "learning_rate": 1.2729166086147803e-06, | |
| "loss": 0.1432, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8464285714285714, | |
| "grad_norm": 0.876155106277133, | |
| "learning_rate": 1.216904382659806e-06, | |
| "loss": 0.2743, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.3009308060492257, | |
| "learning_rate": 1.1620724879689793e-06, | |
| "loss": 0.3375, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8535714285714285, | |
| "grad_norm": 0.8890031161074461, | |
| "learning_rate": 1.1084282932198543e-06, | |
| "loss": 0.2675, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.9590774137550018, | |
| "learning_rate": 1.0559790074789134e-06, | |
| "loss": 0.2847, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8607142857142858, | |
| "grad_norm": 0.9091813907584589, | |
| "learning_rate": 1.00473167923275e-06, | |
| "loss": 0.2142, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8642857142857143, | |
| "grad_norm": 1.1904931385168855, | |
| "learning_rate": 9.546931954408622e-07, | |
| "loss": 0.2657, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8678571428571429, | |
| "grad_norm": 1.0331566604474673, | |
| "learning_rate": 9.058702806101172e-07, | |
| "loss": 0.2839, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8714285714285714, | |
| "grad_norm": 0.8801867878946903, | |
| "learning_rate": 8.582694958910809e-07, | |
| "loss": 0.2513, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.2746454104318332, | |
| "learning_rate": 8.118972381962853e-07, | |
| "loss": 0.475, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8785714285714286, | |
| "grad_norm": 0.8860673740110999, | |
| "learning_rate": 7.667597393405602e-07, | |
| "loss": 0.2575, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8821428571428571, | |
| "grad_norm": 0.7619213943421504, | |
| "learning_rate": 7.228630652035717e-07, | |
| "loss": 0.2073, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 1.024455123403684, | |
| "learning_rate": 6.802131149146374e-07, | |
| "loss": 0.2807, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8892857142857142, | |
| "grad_norm": 0.8835926745414224, | |
| "learning_rate": 6.388156200599726e-07, | |
| "loss": 0.2636, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 0.7925007490357969, | |
| "learning_rate": 5.986761439124289e-07, | |
| "loss": 0.2308, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8964285714285715, | |
| "grad_norm": 1.3499511845536636, | |
| "learning_rate": 5.598000806838766e-07, | |
| "loss": 0.3801, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.7300222807825056, | |
| "learning_rate": 5.221926548002876e-07, | |
| "loss": 0.2236, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9035714285714286, | |
| "grad_norm": 0.7658414820534243, | |
| "learning_rate": 4.858589201996433e-07, | |
| "loss": 0.2408, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9071428571428571, | |
| "grad_norm": 0.8451965731847849, | |
| "learning_rate": 4.5080375965275256e-07, | |
| "loss": 0.241, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9107142857142857, | |
| "grad_norm": 0.9763077207541743, | |
| "learning_rate": 4.1703188410707087e-07, | |
| "loss": 0.2934, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 1.0450416449938653, | |
| "learning_rate": 3.845478320536178e-07, | |
| "loss": 0.3037, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9178571428571428, | |
| "grad_norm": 1.0924428398328787, | |
| "learning_rate": 3.5335596891705406e-07, | |
| "loss": 0.2839, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.9214285714285714, | |
| "grad_norm": 0.7920945565609422, | |
| "learning_rate": 3.2346048646903494e-07, | |
| "loss": 0.2368, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.9915364248763068, | |
| "learning_rate": 2.9486540226488556e-07, | |
| "loss": 0.2638, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 0.9701488675520441, | |
| "learning_rate": 2.6757455910370486e-07, | |
| "loss": 0.2348, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9321428571428572, | |
| "grad_norm": 0.7205020345952264, | |
| "learning_rate": 2.41591624511931e-07, | |
| "loss": 0.2275, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9357142857142857, | |
| "grad_norm": 0.9145542664276822, | |
| "learning_rate": 2.169200902504842e-07, | |
| "loss": 0.2504, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9392857142857143, | |
| "grad_norm": 0.9366813037254895, | |
| "learning_rate": 1.9356327184551716e-07, | |
| "loss": 0.2359, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 0.7940417127042732, | |
| "learning_rate": 1.7152430814285303e-07, | |
| "loss": 0.2657, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9464285714285714, | |
| "grad_norm": 1.0837046668744013, | |
| "learning_rate": 1.5080616088616884e-07, | |
| "loss": 0.2948, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.9318179117611772, | |
| "learning_rate": 1.3141161431896809e-07, | |
| "loss": 0.3247, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9535714285714286, | |
| "grad_norm": 0.8452839653435972, | |
| "learning_rate": 1.1334327481042573e-07, | |
| "loss": 0.185, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9571428571428572, | |
| "grad_norm": 0.9398904285158833, | |
| "learning_rate": 9.660357050512158e-08, | |
| "loss": 0.2707, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9607142857142857, | |
| "grad_norm": 1.242666900837473, | |
| "learning_rate": 8.119475099673035e-08, | |
| "loss": 0.3251, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 0.788276270917447, | |
| "learning_rate": 6.711888702570556e-08, | |
| "loss": 0.2321, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9678571428571429, | |
| "grad_norm": 0.8206210297411658, | |
| "learning_rate": 5.437787020100116e-08, | |
| "loss": 0.2181, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 1.0131859227874498, | |
| "learning_rate": 4.297341274586475e-08, | |
| "loss": 0.3155, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 1.0890650915672193, | |
| "learning_rate": 3.290704726773619e-08, | |
| "loss": 0.2678, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9785714285714285, | |
| "grad_norm": 1.0442519996924977, | |
| "learning_rate": 2.4180126552284523e-08, | |
| "loss": 0.2663, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9821428571428571, | |
| "grad_norm": 1.0298701388046845, | |
| "learning_rate": 1.6793823381614506e-08, | |
| "loss": 0.2908, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9857142857142858, | |
| "grad_norm": 0.9020170998081513, | |
| "learning_rate": 1.0749130376659366e-08, | |
| "loss": 0.2456, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9892857142857143, | |
| "grad_norm": 0.9361316642559707, | |
| "learning_rate": 6.046859863781951e-09, | |
| "loss": 0.3191, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9928571428571429, | |
| "grad_norm": 0.9664382332584418, | |
| "learning_rate": 2.6876437656153663e-09, | |
| "loss": 0.2245, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9964285714285714, | |
| "grad_norm": 0.8441442143857916, | |
| "learning_rate": 6.719335161364804e-10, | |
| "loss": 0.2086, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.085337268493026, | |
| "learning_rate": 0.0, | |
| "loss": 0.2965, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 280, | |
| "total_flos": 146328534351872.0, | |
| "train_loss": 0.4051302823637213, | |
| "train_runtime": 1179.9378, | |
| "train_samples_per_second": 30.364, | |
| "train_steps_per_second": 0.237 | |
| } | |
| ], | |
| "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": 146328534351872.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |