{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 4864, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00020559210526315788, "grad_norm": 157.54771423339844, "learning_rate": 9.997944078947369e-06, "loss": 3.4844, "step": 1 }, { "epoch": 0.00041118421052631577, "grad_norm": 123.90846252441406, "learning_rate": 9.995888157894738e-06, "loss": 2.75, "step": 2 }, { "epoch": 0.0006167763157894736, "grad_norm": 86.50556182861328, "learning_rate": 9.993832236842106e-06, "loss": 2.1914, "step": 3 }, { "epoch": 0.0008223684210526315, "grad_norm": 55.82136154174805, "learning_rate": 9.991776315789474e-06, "loss": 1.0469, "step": 4 }, { "epoch": 0.0010279605263157894, "grad_norm": 101.13475036621094, "learning_rate": 9.989720394736842e-06, "loss": 1.6758, "step": 5 }, { "epoch": 0.0012335526315789473, "grad_norm": 140.59400939941406, "learning_rate": 9.987664473684211e-06, "loss": 2.1523, "step": 6 }, { "epoch": 0.0014391447368421052, "grad_norm": 120.80261993408203, "learning_rate": 9.98560855263158e-06, "loss": 2.9141, "step": 7 }, { "epoch": 0.001644736842105263, "grad_norm": 87.47180938720703, "learning_rate": 9.983552631578947e-06, "loss": 2.0801, "step": 8 }, { "epoch": 0.001850328947368421, "grad_norm": 57.03076171875, "learning_rate": 9.981496710526317e-06, "loss": 1.125, "step": 9 }, { "epoch": 0.002055921052631579, "grad_norm": 79.38758850097656, "learning_rate": 9.979440789473685e-06, "loss": 2.1133, "step": 10 }, { "epoch": 0.002261513157894737, "grad_norm": 54.57301712036133, "learning_rate": 9.977384868421054e-06, "loss": 1.3398, "step": 11 }, { "epoch": 0.0024671052631578946, "grad_norm": 63.4300537109375, "learning_rate": 9.975328947368422e-06, "loss": 1.5781, "step": 12 }, { "epoch": 0.0026726973684210527, "grad_norm": 95.71305847167969, "learning_rate": 9.97327302631579e-06, "loss": 1.6484, "step": 13 }, { "epoch": 0.0028782894736842104, "grad_norm": 65.47756958007812, "learning_rate": 9.97121710526316e-06, "loss": 1.207, "step": 14 }, { "epoch": 0.0030838815789473685, "grad_norm": 63.16679763793945, "learning_rate": 9.969161184210527e-06, "loss": 1.1953, "step": 15 }, { "epoch": 0.003289473684210526, "grad_norm": 43.10338592529297, "learning_rate": 9.967105263157895e-06, "loss": 0.8438, "step": 16 }, { "epoch": 0.0034950657894736842, "grad_norm": 64.31382751464844, "learning_rate": 9.965049342105263e-06, "loss": 1.0273, "step": 17 }, { "epoch": 0.003700657894736842, "grad_norm": 47.334022521972656, "learning_rate": 9.962993421052633e-06, "loss": 1.375, "step": 18 }, { "epoch": 0.00390625, "grad_norm": 68.7382583618164, "learning_rate": 9.9609375e-06, "loss": 1.3086, "step": 19 }, { "epoch": 0.004111842105263158, "grad_norm": 44.792762756347656, "learning_rate": 9.958881578947369e-06, "loss": 1.3555, "step": 20 }, { "epoch": 0.004317434210526316, "grad_norm": 74.220703125, "learning_rate": 9.956825657894738e-06, "loss": 1.5508, "step": 21 }, { "epoch": 0.004523026315789474, "grad_norm": 46.97074890136719, "learning_rate": 9.954769736842106e-06, "loss": 1.1289, "step": 22 }, { "epoch": 0.0047286184210526315, "grad_norm": 50.19340133666992, "learning_rate": 9.952713815789474e-06, "loss": 1.2148, "step": 23 }, { "epoch": 0.004934210526315789, "grad_norm": 30.05093765258789, "learning_rate": 9.950657894736842e-06, "loss": 0.9565, "step": 24 }, { "epoch": 0.005139802631578948, "grad_norm": 34.52531051635742, "learning_rate": 9.948601973684211e-06, "loss": 0.9551, "step": 25 }, { "epoch": 0.005345394736842105, "grad_norm": 33.772029876708984, "learning_rate": 9.94654605263158e-06, "loss": 1.3809, "step": 26 }, { "epoch": 0.005550986842105263, "grad_norm": 28.9645938873291, "learning_rate": 9.944490131578947e-06, "loss": 0.9395, "step": 27 }, { "epoch": 0.005756578947368421, "grad_norm": 45.0798454284668, "learning_rate": 9.942434210526317e-06, "loss": 1.7773, "step": 28 }, { "epoch": 0.005962171052631579, "grad_norm": 38.91462326049805, "learning_rate": 9.940378289473686e-06, "loss": 1.3398, "step": 29 }, { "epoch": 0.006167763157894737, "grad_norm": 26.340608596801758, "learning_rate": 9.938322368421054e-06, "loss": 0.4863, "step": 30 }, { "epoch": 0.006373355263157895, "grad_norm": 55.48283004760742, "learning_rate": 9.936266447368422e-06, "loss": 1.457, "step": 31 }, { "epoch": 0.006578947368421052, "grad_norm": 37.73257827758789, "learning_rate": 9.93421052631579e-06, "loss": 1.3809, "step": 32 }, { "epoch": 0.006784539473684211, "grad_norm": 31.30984878540039, "learning_rate": 9.93215460526316e-06, "loss": 0.9766, "step": 33 }, { "epoch": 0.0069901315789473685, "grad_norm": 42.240867614746094, "learning_rate": 9.930098684210527e-06, "loss": 1.5117, "step": 34 }, { "epoch": 0.007195723684210526, "grad_norm": 45.23456573486328, "learning_rate": 9.928042763157895e-06, "loss": 1.5117, "step": 35 }, { "epoch": 0.007401315789473684, "grad_norm": 29.15300178527832, "learning_rate": 9.925986842105265e-06, "loss": 1.3789, "step": 36 }, { "epoch": 0.007606907894736842, "grad_norm": 37.042938232421875, "learning_rate": 9.923930921052633e-06, "loss": 1.2812, "step": 37 }, { "epoch": 0.0078125, "grad_norm": 41.20448684692383, "learning_rate": 9.921875e-06, "loss": 1.3223, "step": 38 }, { "epoch": 0.008018092105263159, "grad_norm": 30.049509048461914, "learning_rate": 9.919819078947368e-06, "loss": 0.9062, "step": 39 }, { "epoch": 0.008223684210526315, "grad_norm": 63.39948654174805, "learning_rate": 9.917763157894738e-06, "loss": 2.1406, "step": 40 }, { "epoch": 0.008429276315789474, "grad_norm": 36.27692413330078, "learning_rate": 9.915707236842106e-06, "loss": 1.1875, "step": 41 }, { "epoch": 0.008634868421052632, "grad_norm": 52.56294631958008, "learning_rate": 9.913651315789474e-06, "loss": 1.0762, "step": 42 }, { "epoch": 0.00884046052631579, "grad_norm": 40.016929626464844, "learning_rate": 9.911595394736843e-06, "loss": 1.5547, "step": 43 }, { "epoch": 0.009046052631578948, "grad_norm": 58.32114791870117, "learning_rate": 9.909539473684211e-06, "loss": 2.0078, "step": 44 }, { "epoch": 0.009251644736842105, "grad_norm": 39.912681579589844, "learning_rate": 9.90748355263158e-06, "loss": 1.1621, "step": 45 }, { "epoch": 0.009457236842105263, "grad_norm": 25.986845016479492, "learning_rate": 9.905427631578947e-06, "loss": 0.7734, "step": 46 }, { "epoch": 0.009662828947368422, "grad_norm": 34.828182220458984, "learning_rate": 9.903371710526317e-06, "loss": 0.9609, "step": 47 }, { "epoch": 0.009868421052631578, "grad_norm": 32.21487045288086, "learning_rate": 9.901315789473686e-06, "loss": 1.4023, "step": 48 }, { "epoch": 0.010074013157894737, "grad_norm": 44.35273361206055, "learning_rate": 9.899259868421054e-06, "loss": 1.5039, "step": 49 }, { "epoch": 0.010279605263157895, "grad_norm": 33.314334869384766, "learning_rate": 9.897203947368422e-06, "loss": 0.8555, "step": 50 }, { "epoch": 0.010485197368421052, "grad_norm": 45.395896911621094, "learning_rate": 9.89514802631579e-06, "loss": 1.6367, "step": 51 }, { "epoch": 0.01069078947368421, "grad_norm": 36.70625305175781, "learning_rate": 9.89309210526316e-06, "loss": 0.6953, "step": 52 }, { "epoch": 0.010896381578947368, "grad_norm": 24.804597854614258, "learning_rate": 9.891036184210527e-06, "loss": 0.665, "step": 53 }, { "epoch": 0.011101973684210526, "grad_norm": 59.51494216918945, "learning_rate": 9.888980263157895e-06, "loss": 1.6484, "step": 54 }, { "epoch": 0.011307565789473685, "grad_norm": 23.145305633544922, "learning_rate": 9.886924342105265e-06, "loss": 0.6006, "step": 55 }, { "epoch": 0.011513157894736841, "grad_norm": 29.955137252807617, "learning_rate": 9.884868421052633e-06, "loss": 0.4746, "step": 56 }, { "epoch": 0.01171875, "grad_norm": 30.172256469726562, "learning_rate": 9.8828125e-06, "loss": 0.6562, "step": 57 }, { "epoch": 0.011924342105263159, "grad_norm": 27.06200408935547, "learning_rate": 9.880756578947368e-06, "loss": 0.3457, "step": 58 }, { "epoch": 0.012129934210526315, "grad_norm": 59.511497497558594, "learning_rate": 9.878700657894738e-06, "loss": 0.7905, "step": 59 }, { "epoch": 0.012335526315789474, "grad_norm": 78.72785186767578, "learning_rate": 9.876644736842106e-06, "loss": 0.7719, "step": 60 }, { "epoch": 0.012541118421052632, "grad_norm": 35.399600982666016, "learning_rate": 9.874588815789474e-06, "loss": 0.3203, "step": 61 }, { "epoch": 0.01274671052631579, "grad_norm": 109.25142669677734, "learning_rate": 9.872532894736843e-06, "loss": 2.3086, "step": 62 }, { "epoch": 0.012952302631578948, "grad_norm": 87.57012176513672, "learning_rate": 9.870476973684211e-06, "loss": 1.1602, "step": 63 }, { "epoch": 0.013157894736842105, "grad_norm": 77.046142578125, "learning_rate": 9.868421052631579e-06, "loss": 1.5234, "step": 64 }, { "epoch": 0.013363486842105263, "grad_norm": 150.69921875, "learning_rate": 9.866365131578947e-06, "loss": 2.7227, "step": 65 }, { "epoch": 0.013569078947368422, "grad_norm": 51.12712097167969, "learning_rate": 9.864309210526317e-06, "loss": 0.4189, "step": 66 }, { "epoch": 0.013774671052631578, "grad_norm": 136.12965393066406, "learning_rate": 9.862253289473686e-06, "loss": 3.1953, "step": 67 }, { "epoch": 0.013980263157894737, "grad_norm": 51.5265998840332, "learning_rate": 9.860197368421054e-06, "loss": 0.4653, "step": 68 }, { "epoch": 0.014185855263157895, "grad_norm": 72.97129821777344, "learning_rate": 9.858141447368422e-06, "loss": 0.5615, "step": 69 }, { "epoch": 0.014391447368421052, "grad_norm": 110.23238372802734, "learning_rate": 9.856085526315791e-06, "loss": 0.8633, "step": 70 }, { "epoch": 0.01459703947368421, "grad_norm": 123.81166076660156, "learning_rate": 9.85402960526316e-06, "loss": 0.9282, "step": 71 }, { "epoch": 0.014802631578947368, "grad_norm": 63.15595626831055, "learning_rate": 9.851973684210527e-06, "loss": 0.873, "step": 72 }, { "epoch": 0.015008223684210526, "grad_norm": 81.80257415771484, "learning_rate": 9.849917763157895e-06, "loss": 1.3926, "step": 73 }, { "epoch": 0.015213815789473685, "grad_norm": 90.5693359375, "learning_rate": 9.847861842105265e-06, "loss": 1.4871, "step": 74 }, { "epoch": 0.015419407894736841, "grad_norm": 66.70879364013672, "learning_rate": 9.845805921052633e-06, "loss": 1.1562, "step": 75 }, { "epoch": 0.015625, "grad_norm": 182.38619995117188, "learning_rate": 9.84375e-06, "loss": 3.4375, "step": 76 }, { "epoch": 0.015830592105263157, "grad_norm": 155.50247192382812, "learning_rate": 9.84169407894737e-06, "loss": 3.373, "step": 77 }, { "epoch": 0.016036184210526317, "grad_norm": 99.62799835205078, "learning_rate": 9.839638157894738e-06, "loss": 0.7852, "step": 78 }, { "epoch": 0.016241776315789474, "grad_norm": 292.9690246582031, "learning_rate": 9.837582236842106e-06, "loss": 1.3457, "step": 79 }, { "epoch": 0.01644736842105263, "grad_norm": 98.01781463623047, "learning_rate": 9.835526315789474e-06, "loss": 1.2559, "step": 80 }, { "epoch": 0.01665296052631579, "grad_norm": 164.79798889160156, "learning_rate": 9.833470394736843e-06, "loss": 1.6875, "step": 81 }, { "epoch": 0.016858552631578948, "grad_norm": 119.11803436279297, "learning_rate": 9.831414473684211e-06, "loss": 1.3652, "step": 82 }, { "epoch": 0.017064144736842105, "grad_norm": 136.28712463378906, "learning_rate": 9.829358552631579e-06, "loss": 1.3008, "step": 83 }, { "epoch": 0.017269736842105265, "grad_norm": 272.5860900878906, "learning_rate": 9.827302631578949e-06, "loss": 2.3555, "step": 84 }, { "epoch": 0.01747532894736842, "grad_norm": 160.2485809326172, "learning_rate": 9.825246710526316e-06, "loss": 2.0137, "step": 85 }, { "epoch": 0.01768092105263158, "grad_norm": 96.88226318359375, "learning_rate": 9.823190789473686e-06, "loss": 1.9307, "step": 86 }, { "epoch": 0.017886513157894735, "grad_norm": 207.39341735839844, "learning_rate": 9.821134868421054e-06, "loss": 3.3203, "step": 87 }, { "epoch": 0.018092105263157895, "grad_norm": 117.29885864257812, "learning_rate": 9.819078947368422e-06, "loss": 1.3848, "step": 88 }, { "epoch": 0.018297697368421052, "grad_norm": 105.456787109375, "learning_rate": 9.817023026315791e-06, "loss": 1.2188, "step": 89 }, { "epoch": 0.01850328947368421, "grad_norm": 117.52583312988281, "learning_rate": 9.81496710526316e-06, "loss": 1.7539, "step": 90 }, { "epoch": 0.01870888157894737, "grad_norm": 308.0257873535156, "learning_rate": 9.812911184210527e-06, "loss": 3.5156, "step": 91 }, { "epoch": 0.018914473684210526, "grad_norm": 266.7477722167969, "learning_rate": 9.810855263157895e-06, "loss": 2.3945, "step": 92 }, { "epoch": 0.019120065789473683, "grad_norm": 208.29061889648438, "learning_rate": 9.808799342105265e-06, "loss": 4.2344, "step": 93 }, { "epoch": 0.019325657894736843, "grad_norm": 48.40181350708008, "learning_rate": 9.806743421052632e-06, "loss": 0.8242, "step": 94 }, { "epoch": 0.01953125, "grad_norm": 95.56977844238281, "learning_rate": 9.8046875e-06, "loss": 1.4141, "step": 95 }, { "epoch": 0.019736842105263157, "grad_norm": 158.22152709960938, "learning_rate": 9.80263157894737e-06, "loss": 1.707, "step": 96 }, { "epoch": 0.019942434210526317, "grad_norm": 147.1932830810547, "learning_rate": 9.800575657894738e-06, "loss": 1.2461, "step": 97 }, { "epoch": 0.020148026315789474, "grad_norm": 226.3417205810547, "learning_rate": 9.798519736842106e-06, "loss": 1.1777, "step": 98 }, { "epoch": 0.02035361842105263, "grad_norm": 67.17134094238281, "learning_rate": 9.796463815789474e-06, "loss": 0.8662, "step": 99 }, { "epoch": 0.02055921052631579, "grad_norm": 138.85198974609375, "learning_rate": 9.794407894736843e-06, "loss": 2.252, "step": 100 }, { "epoch": 0.020764802631578948, "grad_norm": 129.0947723388672, "learning_rate": 9.792351973684211e-06, "loss": 1.8945, "step": 101 }, { "epoch": 0.020970394736842105, "grad_norm": 53.47102355957031, "learning_rate": 9.790296052631579e-06, "loss": 0.7793, "step": 102 }, { "epoch": 0.021175986842105265, "grad_norm": 130.3709716796875, "learning_rate": 9.788240131578948e-06, "loss": 2.8359, "step": 103 }, { "epoch": 0.02138157894736842, "grad_norm": 108.55887603759766, "learning_rate": 9.786184210526316e-06, "loss": 2.1211, "step": 104 }, { "epoch": 0.02158717105263158, "grad_norm": 159.06735229492188, "learning_rate": 9.784128289473686e-06, "loss": 2.479, "step": 105 }, { "epoch": 0.021792763157894735, "grad_norm": 172.36964416503906, "learning_rate": 9.782072368421054e-06, "loss": 2.4141, "step": 106 }, { "epoch": 0.021998355263157895, "grad_norm": 68.53754425048828, "learning_rate": 9.780016447368422e-06, "loss": 0.7852, "step": 107 }, { "epoch": 0.022203947368421052, "grad_norm": 214.4906768798828, "learning_rate": 9.777960526315791e-06, "loss": 1.5742, "step": 108 }, { "epoch": 0.02240953947368421, "grad_norm": 95.62648010253906, "learning_rate": 9.775904605263159e-06, "loss": 1.4492, "step": 109 }, { "epoch": 0.02261513157894737, "grad_norm": 79.51062774658203, "learning_rate": 9.773848684210527e-06, "loss": 0.7949, "step": 110 }, { "epoch": 0.022820723684210526, "grad_norm": 47.98078155517578, "learning_rate": 9.771792763157895e-06, "loss": 0.5938, "step": 111 }, { "epoch": 0.023026315789473683, "grad_norm": 100.1364974975586, "learning_rate": 9.769736842105264e-06, "loss": 1.9219, "step": 112 }, { "epoch": 0.023231907894736843, "grad_norm": 235.1077880859375, "learning_rate": 9.767680921052632e-06, "loss": 2.293, "step": 113 }, { "epoch": 0.0234375, "grad_norm": 99.74748229980469, "learning_rate": 9.765625e-06, "loss": 1.9375, "step": 114 }, { "epoch": 0.023643092105263157, "grad_norm": 167.4914093017578, "learning_rate": 9.76356907894737e-06, "loss": 1.0508, "step": 115 }, { "epoch": 0.023848684210526317, "grad_norm": 234.4880828857422, "learning_rate": 9.761513157894738e-06, "loss": 2.0957, "step": 116 }, { "epoch": 0.024054276315789474, "grad_norm": 61.61326599121094, "learning_rate": 9.759457236842106e-06, "loss": 1.0498, "step": 117 }, { "epoch": 0.02425986842105263, "grad_norm": 120.47391510009766, "learning_rate": 9.757401315789473e-06, "loss": 1.2539, "step": 118 }, { "epoch": 0.02446546052631579, "grad_norm": 49.18326950073242, "learning_rate": 9.755345394736843e-06, "loss": 0.8896, "step": 119 }, { "epoch": 0.024671052631578948, "grad_norm": 115.40419006347656, "learning_rate": 9.753289473684211e-06, "loss": 2.0859, "step": 120 }, { "epoch": 0.024876644736842105, "grad_norm": 88.2777099609375, "learning_rate": 9.751233552631579e-06, "loss": 1.1738, "step": 121 }, { "epoch": 0.025082236842105265, "grad_norm": 163.45523071289062, "learning_rate": 9.749177631578948e-06, "loss": 1.4062, "step": 122 }, { "epoch": 0.02528782894736842, "grad_norm": 112.65428161621094, "learning_rate": 9.747121710526316e-06, "loss": 1.7891, "step": 123 }, { "epoch": 0.02549342105263158, "grad_norm": 133.77169799804688, "learning_rate": 9.745065789473686e-06, "loss": 1.5869, "step": 124 }, { "epoch": 0.025699013157894735, "grad_norm": 62.263702392578125, "learning_rate": 9.743009868421054e-06, "loss": 1.4766, "step": 125 }, { "epoch": 0.025904605263157895, "grad_norm": 164.55999755859375, "learning_rate": 9.740953947368422e-06, "loss": 2.8359, "step": 126 }, { "epoch": 0.026110197368421052, "grad_norm": 122.83229064941406, "learning_rate": 9.738898026315791e-06, "loss": 2.625, "step": 127 }, { "epoch": 0.02631578947368421, "grad_norm": 106.13177490234375, "learning_rate": 9.736842105263159e-06, "loss": 1.7217, "step": 128 }, { "epoch": 0.02652138157894737, "grad_norm": 97.46070098876953, "learning_rate": 9.734786184210527e-06, "loss": 1.1816, "step": 129 }, { "epoch": 0.026726973684210526, "grad_norm": 68.71578979492188, "learning_rate": 9.732730263157896e-06, "loss": 0.6479, "step": 130 }, { "epoch": 0.026932565789473683, "grad_norm": 81.17630767822266, "learning_rate": 9.730674342105264e-06, "loss": 1.3281, "step": 131 }, { "epoch": 0.027138157894736843, "grad_norm": 206.6467742919922, "learning_rate": 9.728618421052632e-06, "loss": 2.4062, "step": 132 }, { "epoch": 0.02734375, "grad_norm": 71.69466400146484, "learning_rate": 9.7265625e-06, "loss": 0.7285, "step": 133 }, { "epoch": 0.027549342105263157, "grad_norm": 66.50535583496094, "learning_rate": 9.72450657894737e-06, "loss": 1.043, "step": 134 }, { "epoch": 0.027754934210526317, "grad_norm": 167.72650146484375, "learning_rate": 9.722450657894738e-06, "loss": 1.3613, "step": 135 }, { "epoch": 0.027960526315789474, "grad_norm": 56.72340393066406, "learning_rate": 9.720394736842105e-06, "loss": 0.7168, "step": 136 }, { "epoch": 0.02816611842105263, "grad_norm": 64.05767059326172, "learning_rate": 9.718338815789475e-06, "loss": 0.7246, "step": 137 }, { "epoch": 0.02837171052631579, "grad_norm": 87.97325134277344, "learning_rate": 9.716282894736843e-06, "loss": 1.1445, "step": 138 }, { "epoch": 0.028577302631578948, "grad_norm": 124.7219467163086, "learning_rate": 9.71422697368421e-06, "loss": 2.3203, "step": 139 }, { "epoch": 0.028782894736842105, "grad_norm": 83.45574951171875, "learning_rate": 9.712171052631579e-06, "loss": 1.3047, "step": 140 }, { "epoch": 0.028988486842105265, "grad_norm": 65.3621826171875, "learning_rate": 9.710115131578948e-06, "loss": 1.5898, "step": 141 }, { "epoch": 0.02919407894736842, "grad_norm": 80.9415054321289, "learning_rate": 9.708059210526316e-06, "loss": 1.334, "step": 142 }, { "epoch": 0.02939967105263158, "grad_norm": 106.05699157714844, "learning_rate": 9.706003289473686e-06, "loss": 0.79, "step": 143 }, { "epoch": 0.029605263157894735, "grad_norm": 87.05155944824219, "learning_rate": 9.703947368421054e-06, "loss": 0.6367, "step": 144 }, { "epoch": 0.029810855263157895, "grad_norm": 102.74181365966797, "learning_rate": 9.701891447368421e-06, "loss": 1.8594, "step": 145 }, { "epoch": 0.030016447368421052, "grad_norm": 209.3494110107422, "learning_rate": 9.699835526315791e-06, "loss": 3.1709, "step": 146 }, { "epoch": 0.03022203947368421, "grad_norm": 75.75135040283203, "learning_rate": 9.697779605263159e-06, "loss": 2.1523, "step": 147 }, { "epoch": 0.03042763157894737, "grad_norm": 59.05912399291992, "learning_rate": 9.695723684210527e-06, "loss": 1.1328, "step": 148 }, { "epoch": 0.030633223684210526, "grad_norm": 74.14704132080078, "learning_rate": 9.693667763157896e-06, "loss": 1.0117, "step": 149 }, { "epoch": 0.030838815789473683, "grad_norm": 71.49958038330078, "learning_rate": 9.691611842105264e-06, "loss": 1.8398, "step": 150 }, { "epoch": 0.031044407894736843, "grad_norm": 43.915565490722656, "learning_rate": 9.689555921052632e-06, "loss": 0.832, "step": 151 }, { "epoch": 0.03125, "grad_norm": 42.98212814331055, "learning_rate": 9.6875e-06, "loss": 1.0879, "step": 152 }, { "epoch": 0.03145559210526316, "grad_norm": 37.05778121948242, "learning_rate": 9.68544407894737e-06, "loss": 0.3398, "step": 153 }, { "epoch": 0.031661184210526314, "grad_norm": 85.47862243652344, "learning_rate": 9.683388157894737e-06, "loss": 2.1328, "step": 154 }, { "epoch": 0.03186677631578947, "grad_norm": 114.26785278320312, "learning_rate": 9.681332236842105e-06, "loss": 1.9102, "step": 155 }, { "epoch": 0.032072368421052634, "grad_norm": 112.1761703491211, "learning_rate": 9.679276315789475e-06, "loss": 2.8516, "step": 156 }, { "epoch": 0.03227796052631579, "grad_norm": 65.07213592529297, "learning_rate": 9.677220394736843e-06, "loss": 0.8994, "step": 157 }, { "epoch": 0.03248355263157895, "grad_norm": 71.04268646240234, "learning_rate": 9.67516447368421e-06, "loss": 1.5742, "step": 158 }, { "epoch": 0.032689144736842105, "grad_norm": 121.01145935058594, "learning_rate": 9.673108552631579e-06, "loss": 2.1758, "step": 159 }, { "epoch": 0.03289473684210526, "grad_norm": 217.50210571289062, "learning_rate": 9.671052631578948e-06, "loss": 3.1562, "step": 160 }, { "epoch": 0.03310032894736842, "grad_norm": 97.17508697509766, "learning_rate": 9.668996710526316e-06, "loss": 2.168, "step": 161 }, { "epoch": 0.03330592105263158, "grad_norm": 86.92736053466797, "learning_rate": 9.666940789473686e-06, "loss": 1.2695, "step": 162 }, { "epoch": 0.03351151315789474, "grad_norm": 85.95757293701172, "learning_rate": 9.664884868421053e-06, "loss": 1.6289, "step": 163 }, { "epoch": 0.033717105263157895, "grad_norm": 63.56707000732422, "learning_rate": 9.662828947368423e-06, "loss": 1.6445, "step": 164 }, { "epoch": 0.03392269736842105, "grad_norm": 90.49592590332031, "learning_rate": 9.660773026315791e-06, "loss": 2.3516, "step": 165 }, { "epoch": 0.03412828947368421, "grad_norm": 73.68732452392578, "learning_rate": 9.658717105263159e-06, "loss": 2.0234, "step": 166 }, { "epoch": 0.034333881578947366, "grad_norm": 164.30149841308594, "learning_rate": 9.656661184210527e-06, "loss": 3.1406, "step": 167 }, { "epoch": 0.03453947368421053, "grad_norm": 47.69912338256836, "learning_rate": 9.654605263157896e-06, "loss": 0.9639, "step": 168 }, { "epoch": 0.034745065789473686, "grad_norm": 102.29673767089844, "learning_rate": 9.652549342105264e-06, "loss": 2.6875, "step": 169 }, { "epoch": 0.03495065789473684, "grad_norm": 87.81875610351562, "learning_rate": 9.650493421052632e-06, "loss": 2.4844, "step": 170 }, { "epoch": 0.03515625, "grad_norm": 69.05582427978516, "learning_rate": 9.648437500000002e-06, "loss": 1.5898, "step": 171 }, { "epoch": 0.03536184210526316, "grad_norm": 179.32859802246094, "learning_rate": 9.64638157894737e-06, "loss": 2.125, "step": 172 }, { "epoch": 0.035567434210526314, "grad_norm": 47.69493103027344, "learning_rate": 9.644325657894737e-06, "loss": 1.9727, "step": 173 }, { "epoch": 0.03577302631578947, "grad_norm": 88.21089935302734, "learning_rate": 9.642269736842105e-06, "loss": 1.7969, "step": 174 }, { "epoch": 0.035978618421052634, "grad_norm": 91.75403594970703, "learning_rate": 9.640213815789475e-06, "loss": 2.1836, "step": 175 }, { "epoch": 0.03618421052631579, "grad_norm": 66.17277526855469, "learning_rate": 9.638157894736843e-06, "loss": 0.9453, "step": 176 }, { "epoch": 0.03638980263157895, "grad_norm": 54.707027435302734, "learning_rate": 9.63610197368421e-06, "loss": 1.4688, "step": 177 }, { "epoch": 0.036595394736842105, "grad_norm": 41.72214126586914, "learning_rate": 9.63404605263158e-06, "loss": 0.9414, "step": 178 }, { "epoch": 0.03680098684210526, "grad_norm": 53.67431640625, "learning_rate": 9.631990131578948e-06, "loss": 1.8672, "step": 179 }, { "epoch": 0.03700657894736842, "grad_norm": 45.96166229248047, "learning_rate": 9.629934210526316e-06, "loss": 0.6289, "step": 180 }, { "epoch": 0.03721217105263158, "grad_norm": 109.84236145019531, "learning_rate": 9.627878289473685e-06, "loss": 2.8438, "step": 181 }, { "epoch": 0.03741776315789474, "grad_norm": 81.16616821289062, "learning_rate": 9.625822368421053e-06, "loss": 1.8906, "step": 182 }, { "epoch": 0.037623355263157895, "grad_norm": 30.04365348815918, "learning_rate": 9.623766447368423e-06, "loss": 0.8574, "step": 183 }, { "epoch": 0.03782894736842105, "grad_norm": 40.17713165283203, "learning_rate": 9.62171052631579e-06, "loss": 1.3086, "step": 184 }, { "epoch": 0.03803453947368421, "grad_norm": 34.26838302612305, "learning_rate": 9.619654605263159e-06, "loss": 0.8145, "step": 185 }, { "epoch": 0.038240131578947366, "grad_norm": 44.376197814941406, "learning_rate": 9.617598684210527e-06, "loss": 1.1406, "step": 186 }, { "epoch": 0.03844572368421053, "grad_norm": 171.97088623046875, "learning_rate": 9.615542763157896e-06, "loss": 3.1055, "step": 187 }, { "epoch": 0.038651315789473686, "grad_norm": 50.11237335205078, "learning_rate": 9.613486842105264e-06, "loss": 0.7734, "step": 188 }, { "epoch": 0.03885690789473684, "grad_norm": 52.43278121948242, "learning_rate": 9.611430921052632e-06, "loss": 0.6348, "step": 189 }, { "epoch": 0.0390625, "grad_norm": 73.75617980957031, "learning_rate": 9.609375000000001e-06, "loss": 1.8906, "step": 190 }, { "epoch": 0.03926809210526316, "grad_norm": 65.51220703125, "learning_rate": 9.60731907894737e-06, "loss": 1.8984, "step": 191 }, { "epoch": 0.039473684210526314, "grad_norm": 66.25209045410156, "learning_rate": 9.605263157894737e-06, "loss": 1.0625, "step": 192 }, { "epoch": 0.03967927631578947, "grad_norm": 38.14455795288086, "learning_rate": 9.603207236842105e-06, "loss": 0.959, "step": 193 }, { "epoch": 0.039884868421052634, "grad_norm": 72.10670471191406, "learning_rate": 9.601151315789475e-06, "loss": 1.3789, "step": 194 }, { "epoch": 0.04009046052631579, "grad_norm": 49.085689544677734, "learning_rate": 9.599095394736843e-06, "loss": 0.9277, "step": 195 }, { "epoch": 0.04029605263157895, "grad_norm": 50.40119934082031, "learning_rate": 9.59703947368421e-06, "loss": 1.5977, "step": 196 }, { "epoch": 0.040501644736842105, "grad_norm": 73.01385498046875, "learning_rate": 9.59498355263158e-06, "loss": 1.4922, "step": 197 }, { "epoch": 0.04070723684210526, "grad_norm": 83.19461822509766, "learning_rate": 9.592927631578948e-06, "loss": 0.916, "step": 198 }, { "epoch": 0.04091282894736842, "grad_norm": 35.14956283569336, "learning_rate": 9.590871710526316e-06, "loss": 0.8887, "step": 199 }, { "epoch": 0.04111842105263158, "grad_norm": 53.86555862426758, "learning_rate": 9.588815789473685e-06, "loss": 1.1133, "step": 200 }, { "epoch": 0.04132401315789474, "grad_norm": 47.63760757446289, "learning_rate": 9.586759868421053e-06, "loss": 1.2695, "step": 201 }, { "epoch": 0.041529605263157895, "grad_norm": 67.27269744873047, "learning_rate": 9.584703947368423e-06, "loss": 1.2266, "step": 202 }, { "epoch": 0.04173519736842105, "grad_norm": 56.93370056152344, "learning_rate": 9.58264802631579e-06, "loss": 1.1953, "step": 203 }, { "epoch": 0.04194078947368421, "grad_norm": 58.086280822753906, "learning_rate": 9.580592105263159e-06, "loss": 0.9863, "step": 204 }, { "epoch": 0.042146381578947366, "grad_norm": 62.98575210571289, "learning_rate": 9.578536184210528e-06, "loss": 0.9629, "step": 205 }, { "epoch": 0.04235197368421053, "grad_norm": 35.1092643737793, "learning_rate": 9.576480263157896e-06, "loss": 0.7334, "step": 206 }, { "epoch": 0.042557565789473686, "grad_norm": 160.66604614257812, "learning_rate": 9.574424342105264e-06, "loss": 2.2383, "step": 207 }, { "epoch": 0.04276315789473684, "grad_norm": 43.58798599243164, "learning_rate": 9.572368421052632e-06, "loss": 0.5254, "step": 208 }, { "epoch": 0.04296875, "grad_norm": 55.71181106567383, "learning_rate": 9.570312500000001e-06, "loss": 1.2734, "step": 209 }, { "epoch": 0.04317434210526316, "grad_norm": 46.19253158569336, "learning_rate": 9.56825657894737e-06, "loss": 0.667, "step": 210 }, { "epoch": 0.043379934210526314, "grad_norm": 47.46964645385742, "learning_rate": 9.566200657894737e-06, "loss": 0.9707, "step": 211 }, { "epoch": 0.04358552631578947, "grad_norm": 86.06796264648438, "learning_rate": 9.564144736842105e-06, "loss": 1.3555, "step": 212 }, { "epoch": 0.043791118421052634, "grad_norm": 37.15574264526367, "learning_rate": 9.562088815789475e-06, "loss": 0.666, "step": 213 }, { "epoch": 0.04399671052631579, "grad_norm": 86.48974609375, "learning_rate": 9.560032894736842e-06, "loss": 0.9688, "step": 214 }, { "epoch": 0.04420230263157895, "grad_norm": 57.31755828857422, "learning_rate": 9.55797697368421e-06, "loss": 0.7656, "step": 215 }, { "epoch": 0.044407894736842105, "grad_norm": 93.53360748291016, "learning_rate": 9.55592105263158e-06, "loss": 2.5156, "step": 216 }, { "epoch": 0.04461348684210526, "grad_norm": 40.321632385253906, "learning_rate": 9.553865131578948e-06, "loss": 1.1743, "step": 217 }, { "epoch": 0.04481907894736842, "grad_norm": 49.70573425292969, "learning_rate": 9.551809210526316e-06, "loss": 1.375, "step": 218 }, { "epoch": 0.04502467105263158, "grad_norm": 97.36224365234375, "learning_rate": 9.549753289473685e-06, "loss": 1.918, "step": 219 }, { "epoch": 0.04523026315789474, "grad_norm": 48.46147918701172, "learning_rate": 9.547697368421053e-06, "loss": 1.0898, "step": 220 }, { "epoch": 0.045435855263157895, "grad_norm": 60.99446105957031, "learning_rate": 9.545641447368423e-06, "loss": 0.6816, "step": 221 }, { "epoch": 0.04564144736842105, "grad_norm": 63.48299789428711, "learning_rate": 9.54358552631579e-06, "loss": 0.8184, "step": 222 }, { "epoch": 0.04584703947368421, "grad_norm": 185.73202514648438, "learning_rate": 9.541529605263158e-06, "loss": 3.6875, "step": 223 }, { "epoch": 0.046052631578947366, "grad_norm": 55.041709899902344, "learning_rate": 9.539473684210528e-06, "loss": 0.6814, "step": 224 }, { "epoch": 0.04625822368421053, "grad_norm": 106.05138397216797, "learning_rate": 9.537417763157896e-06, "loss": 2.3945, "step": 225 }, { "epoch": 0.046463815789473686, "grad_norm": 50.3757209777832, "learning_rate": 9.535361842105264e-06, "loss": 0.8789, "step": 226 }, { "epoch": 0.04666940789473684, "grad_norm": 111.6707992553711, "learning_rate": 9.533305921052632e-06, "loss": 1.5859, "step": 227 }, { "epoch": 0.046875, "grad_norm": 35.787353515625, "learning_rate": 9.531250000000001e-06, "loss": 0.6138, "step": 228 }, { "epoch": 0.04708059210526316, "grad_norm": 74.34663391113281, "learning_rate": 9.52919407894737e-06, "loss": 1.2539, "step": 229 }, { "epoch": 0.047286184210526314, "grad_norm": 43.01491165161133, "learning_rate": 9.527138157894737e-06, "loss": 0.7812, "step": 230 }, { "epoch": 0.04749177631578947, "grad_norm": 106.63488006591797, "learning_rate": 9.525082236842107e-06, "loss": 1.5625, "step": 231 }, { "epoch": 0.047697368421052634, "grad_norm": 87.31039428710938, "learning_rate": 9.523026315789474e-06, "loss": 1.5801, "step": 232 }, { "epoch": 0.04790296052631579, "grad_norm": 68.3610610961914, "learning_rate": 9.520970394736842e-06, "loss": 1.249, "step": 233 }, { "epoch": 0.04810855263157895, "grad_norm": 93.34453582763672, "learning_rate": 9.51891447368421e-06, "loss": 2.0039, "step": 234 }, { "epoch": 0.048314144736842105, "grad_norm": 100.5185546875, "learning_rate": 9.51685855263158e-06, "loss": 2.3008, "step": 235 }, { "epoch": 0.04851973684210526, "grad_norm": 115.72514343261719, "learning_rate": 9.514802631578948e-06, "loss": 2.7812, "step": 236 }, { "epoch": 0.04872532894736842, "grad_norm": 99.58477783203125, "learning_rate": 9.512746710526316e-06, "loss": 2.3125, "step": 237 }, { "epoch": 0.04893092105263158, "grad_norm": 59.35429763793945, "learning_rate": 9.510690789473685e-06, "loss": 0.918, "step": 238 }, { "epoch": 0.04913651315789474, "grad_norm": 45.29487609863281, "learning_rate": 9.508634868421053e-06, "loss": 0.541, "step": 239 }, { "epoch": 0.049342105263157895, "grad_norm": 66.3087387084961, "learning_rate": 9.506578947368423e-06, "loss": 1.7227, "step": 240 }, { "epoch": 0.04954769736842105, "grad_norm": 72.6855239868164, "learning_rate": 9.50452302631579e-06, "loss": 2.1211, "step": 241 }, { "epoch": 0.04975328947368421, "grad_norm": 40.03398132324219, "learning_rate": 9.502467105263158e-06, "loss": 0.7852, "step": 242 }, { "epoch": 0.049958881578947366, "grad_norm": 156.98919677734375, "learning_rate": 9.500411184210528e-06, "loss": 1.9512, "step": 243 }, { "epoch": 0.05016447368421053, "grad_norm": 48.3450813293457, "learning_rate": 9.498355263157896e-06, "loss": 0.6489, "step": 244 }, { "epoch": 0.050370065789473686, "grad_norm": 46.85075378417969, "learning_rate": 9.496299342105264e-06, "loss": 0.8697, "step": 245 }, { "epoch": 0.05057565789473684, "grad_norm": 67.99323272705078, "learning_rate": 9.494243421052632e-06, "loss": 1.0044, "step": 246 }, { "epoch": 0.05078125, "grad_norm": 86.78130340576172, "learning_rate": 9.492187500000001e-06, "loss": 0.666, "step": 247 }, { "epoch": 0.05098684210526316, "grad_norm": 88.6587142944336, "learning_rate": 9.490131578947369e-06, "loss": 0.999, "step": 248 }, { "epoch": 0.051192434210526314, "grad_norm": 64.42821502685547, "learning_rate": 9.488075657894737e-06, "loss": 0.8789, "step": 249 }, { "epoch": 0.05139802631578947, "grad_norm": 42.16807556152344, "learning_rate": 9.486019736842107e-06, "loss": 0.8184, "step": 250 }, { "epoch": 0.051603618421052634, "grad_norm": 34.90488815307617, "learning_rate": 9.483963815789474e-06, "loss": 0.4902, "step": 251 }, { "epoch": 0.05180921052631579, "grad_norm": 23.641334533691406, "learning_rate": 9.481907894736842e-06, "loss": 0.2803, "step": 252 }, { "epoch": 0.05201480263157895, "grad_norm": 40.338623046875, "learning_rate": 9.47985197368421e-06, "loss": 0.9009, "step": 253 }, { "epoch": 0.052220394736842105, "grad_norm": 46.449832916259766, "learning_rate": 9.47779605263158e-06, "loss": 1.4141, "step": 254 }, { "epoch": 0.05242598684210526, "grad_norm": 47.70627975463867, "learning_rate": 9.475740131578948e-06, "loss": 0.6895, "step": 255 }, { "epoch": 0.05263157894736842, "grad_norm": 117.47351837158203, "learning_rate": 9.473684210526315e-06, "loss": 2.812, "step": 256 }, { "epoch": 0.05283717105263158, "grad_norm": 41.12131881713867, "learning_rate": 9.471628289473685e-06, "loss": 0.9609, "step": 257 }, { "epoch": 0.05304276315789474, "grad_norm": 71.14788818359375, "learning_rate": 9.469572368421055e-06, "loss": 1.1211, "step": 258 }, { "epoch": 0.053248355263157895, "grad_norm": 86.0069808959961, "learning_rate": 9.467516447368423e-06, "loss": 1.9844, "step": 259 }, { "epoch": 0.05345394736842105, "grad_norm": 41.394737243652344, "learning_rate": 9.46546052631579e-06, "loss": 0.5605, "step": 260 }, { "epoch": 0.05365953947368421, "grad_norm": 70.2432861328125, "learning_rate": 9.463404605263158e-06, "loss": 0.71, "step": 261 }, { "epoch": 0.053865131578947366, "grad_norm": 64.5775146484375, "learning_rate": 9.461348684210528e-06, "loss": 2.2734, "step": 262 }, { "epoch": 0.05407072368421053, "grad_norm": 32.41098403930664, "learning_rate": 9.459292763157896e-06, "loss": 0.46, "step": 263 }, { "epoch": 0.054276315789473686, "grad_norm": 58.78820037841797, "learning_rate": 9.457236842105264e-06, "loss": 1.4531, "step": 264 }, { "epoch": 0.05448190789473684, "grad_norm": 47.887908935546875, "learning_rate": 9.455180921052633e-06, "loss": 0.8535, "step": 265 }, { "epoch": 0.0546875, "grad_norm": 115.64647674560547, "learning_rate": 9.453125000000001e-06, "loss": 1.6113, "step": 266 }, { "epoch": 0.05489309210526316, "grad_norm": 45.19354248046875, "learning_rate": 9.451069078947369e-06, "loss": 0.9551, "step": 267 }, { "epoch": 0.055098684210526314, "grad_norm": 35.74041748046875, "learning_rate": 9.449013157894737e-06, "loss": 0.5996, "step": 268 }, { "epoch": 0.05530427631578947, "grad_norm": 63.58024215698242, "learning_rate": 9.446957236842106e-06, "loss": 1.0918, "step": 269 }, { "epoch": 0.055509868421052634, "grad_norm": 25.927593231201172, "learning_rate": 9.444901315789474e-06, "loss": 0.417, "step": 270 }, { "epoch": 0.05571546052631579, "grad_norm": 31.22515869140625, "learning_rate": 9.442845394736842e-06, "loss": 0.3122, "step": 271 }, { "epoch": 0.05592105263157895, "grad_norm": 195.76246643066406, "learning_rate": 9.440789473684212e-06, "loss": 1.6523, "step": 272 }, { "epoch": 0.056126644736842105, "grad_norm": 97.8608169555664, "learning_rate": 9.43873355263158e-06, "loss": 2.5933, "step": 273 }, { "epoch": 0.05633223684210526, "grad_norm": 85.54328155517578, "learning_rate": 9.436677631578948e-06, "loss": 1.1611, "step": 274 }, { "epoch": 0.05653782894736842, "grad_norm": 98.65076446533203, "learning_rate": 9.434621710526315e-06, "loss": 2.1243, "step": 275 }, { "epoch": 0.05674342105263158, "grad_norm": 116.48287200927734, "learning_rate": 9.432565789473685e-06, "loss": 3.0078, "step": 276 }, { "epoch": 0.05694901315789474, "grad_norm": 18.842750549316406, "learning_rate": 9.430509868421055e-06, "loss": 0.2344, "step": 277 }, { "epoch": 0.057154605263157895, "grad_norm": 50.00761413574219, "learning_rate": 9.428453947368422e-06, "loss": 1.0957, "step": 278 }, { "epoch": 0.05736019736842105, "grad_norm": 84.39752960205078, "learning_rate": 9.42639802631579e-06, "loss": 1.8945, "step": 279 }, { "epoch": 0.05756578947368421, "grad_norm": 149.30723571777344, "learning_rate": 9.424342105263158e-06, "loss": 1.7671, "step": 280 }, { "epoch": 0.057771381578947366, "grad_norm": 106.67811584472656, "learning_rate": 9.422286184210528e-06, "loss": 1.4805, "step": 281 }, { "epoch": 0.05797697368421053, "grad_norm": 36.05638885498047, "learning_rate": 9.420230263157896e-06, "loss": 1.2012, "step": 282 }, { "epoch": 0.058182565789473686, "grad_norm": 93.82987976074219, "learning_rate": 9.418174342105264e-06, "loss": 2.5156, "step": 283 }, { "epoch": 0.05838815789473684, "grad_norm": 109.50467681884766, "learning_rate": 9.416118421052633e-06, "loss": 2.1641, "step": 284 }, { "epoch": 0.05859375, "grad_norm": 128.9050750732422, "learning_rate": 9.414062500000001e-06, "loss": 0.957, "step": 285 }, { "epoch": 0.05879934210526316, "grad_norm": 56.430686950683594, "learning_rate": 9.412006578947369e-06, "loss": 1.332, "step": 286 }, { "epoch": 0.059004934210526314, "grad_norm": 68.03912353515625, "learning_rate": 9.409950657894737e-06, "loss": 1.0732, "step": 287 }, { "epoch": 0.05921052631578947, "grad_norm": 62.56614685058594, "learning_rate": 9.407894736842106e-06, "loss": 1.6328, "step": 288 }, { "epoch": 0.059416118421052634, "grad_norm": 44.94523239135742, "learning_rate": 9.405838815789474e-06, "loss": 0.8242, "step": 289 }, { "epoch": 0.05962171052631579, "grad_norm": 59.16824722290039, "learning_rate": 9.403782894736842e-06, "loss": 1.6523, "step": 290 }, { "epoch": 0.05982730263157895, "grad_norm": 75.50399017333984, "learning_rate": 9.401726973684212e-06, "loss": 1.8945, "step": 291 }, { "epoch": 0.060032894736842105, "grad_norm": 79.97955322265625, "learning_rate": 9.39967105263158e-06, "loss": 2.3359, "step": 292 }, { "epoch": 0.06023848684210526, "grad_norm": 105.5020523071289, "learning_rate": 9.397615131578947e-06, "loss": 1.7773, "step": 293 }, { "epoch": 0.06044407894736842, "grad_norm": 53.01664352416992, "learning_rate": 9.395559210526315e-06, "loss": 0.5361, "step": 294 }, { "epoch": 0.06064967105263158, "grad_norm": 33.497154235839844, "learning_rate": 9.393503289473685e-06, "loss": 0.7871, "step": 295 }, { "epoch": 0.06085526315789474, "grad_norm": 23.027793884277344, "learning_rate": 9.391447368421054e-06, "loss": 0.2915, "step": 296 }, { "epoch": 0.061060855263157895, "grad_norm": 58.62592697143555, "learning_rate": 9.389391447368422e-06, "loss": 1.1426, "step": 297 }, { "epoch": 0.06126644736842105, "grad_norm": 57.419654846191406, "learning_rate": 9.38733552631579e-06, "loss": 1.6133, "step": 298 }, { "epoch": 0.06147203947368421, "grad_norm": 40.22806930541992, "learning_rate": 9.38527960526316e-06, "loss": 0.6152, "step": 299 }, { "epoch": 0.061677631578947366, "grad_norm": 64.0289535522461, "learning_rate": 9.383223684210528e-06, "loss": 1.0547, "step": 300 }, { "epoch": 0.06188322368421053, "grad_norm": 38.185264587402344, "learning_rate": 9.381167763157896e-06, "loss": 0.3254, "step": 301 }, { "epoch": 0.062088815789473686, "grad_norm": 34.86383819580078, "learning_rate": 9.379111842105263e-06, "loss": 0.793, "step": 302 }, { "epoch": 0.06229440789473684, "grad_norm": 42.52946472167969, "learning_rate": 9.377055921052633e-06, "loss": 1.0645, "step": 303 }, { "epoch": 0.0625, "grad_norm": 78.47978973388672, "learning_rate": 9.375000000000001e-06, "loss": 0.7502, "step": 304 }, { "epoch": 0.06270559210526316, "grad_norm": 53.0689582824707, "learning_rate": 9.372944078947369e-06, "loss": 1.0254, "step": 305 }, { "epoch": 0.06291118421052631, "grad_norm": 68.66390228271484, "learning_rate": 9.370888157894738e-06, "loss": 1.7969, "step": 306 }, { "epoch": 0.06311677631578948, "grad_norm": 40.7020378112793, "learning_rate": 9.368832236842106e-06, "loss": 0.4561, "step": 307 }, { "epoch": 0.06332236842105263, "grad_norm": 16.405012130737305, "learning_rate": 9.366776315789474e-06, "loss": 0.1624, "step": 308 }, { "epoch": 0.06352796052631579, "grad_norm": 66.62581634521484, "learning_rate": 9.364720394736842e-06, "loss": 1.2793, "step": 309 }, { "epoch": 0.06373355263157894, "grad_norm": 38.57311248779297, "learning_rate": 9.362664473684212e-06, "loss": 0.6895, "step": 310 }, { "epoch": 0.0639391447368421, "grad_norm": 85.42110443115234, "learning_rate": 9.36060855263158e-06, "loss": 1.4336, "step": 311 }, { "epoch": 0.06414473684210527, "grad_norm": 92.61295318603516, "learning_rate": 9.358552631578947e-06, "loss": 2.7227, "step": 312 }, { "epoch": 0.06435032894736842, "grad_norm": 90.53781127929688, "learning_rate": 9.356496710526315e-06, "loss": 1.4766, "step": 313 }, { "epoch": 0.06455592105263158, "grad_norm": 132.5380096435547, "learning_rate": 9.354440789473685e-06, "loss": 1.1553, "step": 314 }, { "epoch": 0.06476151315789473, "grad_norm": 67.5123519897461, "learning_rate": 9.352384868421054e-06, "loss": 1.167, "step": 315 }, { "epoch": 0.0649671052631579, "grad_norm": 146.13992309570312, "learning_rate": 9.350328947368422e-06, "loss": 3.5391, "step": 316 }, { "epoch": 0.06517269736842106, "grad_norm": 27.957015991210938, "learning_rate": 9.34827302631579e-06, "loss": 0.2771, "step": 317 }, { "epoch": 0.06537828947368421, "grad_norm": 101.00431823730469, "learning_rate": 9.34621710526316e-06, "loss": 2.3125, "step": 318 }, { "epoch": 0.06558388157894737, "grad_norm": 29.328088760375977, "learning_rate": 9.344161184210528e-06, "loss": 0.207, "step": 319 }, { "epoch": 0.06578947368421052, "grad_norm": 79.87821960449219, "learning_rate": 9.342105263157895e-06, "loss": 1.1992, "step": 320 }, { "epoch": 0.06599506578947369, "grad_norm": 61.61282730102539, "learning_rate": 9.340049342105263e-06, "loss": 0.8828, "step": 321 }, { "epoch": 0.06620065789473684, "grad_norm": 70.58615112304688, "learning_rate": 9.337993421052633e-06, "loss": 1.6484, "step": 322 }, { "epoch": 0.06640625, "grad_norm": 102.12861633300781, "learning_rate": 9.3359375e-06, "loss": 1.5273, "step": 323 }, { "epoch": 0.06661184210526316, "grad_norm": 99.98857116699219, "learning_rate": 9.333881578947369e-06, "loss": 2.3301, "step": 324 }, { "epoch": 0.06681743421052631, "grad_norm": 47.745689392089844, "learning_rate": 9.331825657894738e-06, "loss": 0.9277, "step": 325 }, { "epoch": 0.06702302631578948, "grad_norm": 56.94204330444336, "learning_rate": 9.329769736842106e-06, "loss": 1.5234, "step": 326 }, { "epoch": 0.06722861842105263, "grad_norm": 88.049072265625, "learning_rate": 9.327713815789474e-06, "loss": 1.4062, "step": 327 }, { "epoch": 0.06743421052631579, "grad_norm": 36.29627990722656, "learning_rate": 9.325657894736842e-06, "loss": 0.5876, "step": 328 }, { "epoch": 0.06763980263157894, "grad_norm": 69.29109954833984, "learning_rate": 9.323601973684211e-06, "loss": 1.4072, "step": 329 }, { "epoch": 0.0678453947368421, "grad_norm": 77.83673858642578, "learning_rate": 9.32154605263158e-06, "loss": 1.9531, "step": 330 }, { "epoch": 0.06805098684210527, "grad_norm": 42.96986389160156, "learning_rate": 9.319490131578947e-06, "loss": 0.6543, "step": 331 }, { "epoch": 0.06825657894736842, "grad_norm": 58.11347198486328, "learning_rate": 9.317434210526317e-06, "loss": 0.542, "step": 332 }, { "epoch": 0.06846217105263158, "grad_norm": 63.39133071899414, "learning_rate": 9.315378289473685e-06, "loss": 1.4297, "step": 333 }, { "epoch": 0.06866776315789473, "grad_norm": 172.48423767089844, "learning_rate": 9.313322368421054e-06, "loss": 1.5586, "step": 334 }, { "epoch": 0.0688733552631579, "grad_norm": 71.94248962402344, "learning_rate": 9.311266447368422e-06, "loss": 1.543, "step": 335 }, { "epoch": 0.06907894736842106, "grad_norm": 86.32141876220703, "learning_rate": 9.30921052631579e-06, "loss": 1.7969, "step": 336 }, { "epoch": 0.06928453947368421, "grad_norm": 42.6099739074707, "learning_rate": 9.30715460526316e-06, "loss": 0.7114, "step": 337 }, { "epoch": 0.06949013157894737, "grad_norm": 104.55664825439453, "learning_rate": 9.305098684210527e-06, "loss": 1.8364, "step": 338 }, { "epoch": 0.06969572368421052, "grad_norm": 63.93256378173828, "learning_rate": 9.303042763157895e-06, "loss": 1.7109, "step": 339 }, { "epoch": 0.06990131578947369, "grad_norm": 58.4498405456543, "learning_rate": 9.300986842105263e-06, "loss": 0.9355, "step": 340 }, { "epoch": 0.07010690789473684, "grad_norm": 52.1556510925293, "learning_rate": 9.298930921052633e-06, "loss": 1.1992, "step": 341 }, { "epoch": 0.0703125, "grad_norm": 37.667076110839844, "learning_rate": 9.296875e-06, "loss": 0.5352, "step": 342 }, { "epoch": 0.07051809210526316, "grad_norm": 53.36186218261719, "learning_rate": 9.294819078947369e-06, "loss": 1.4238, "step": 343 }, { "epoch": 0.07072368421052631, "grad_norm": 11.160712242126465, "learning_rate": 9.292763157894738e-06, "loss": 0.1106, "step": 344 }, { "epoch": 0.07092927631578948, "grad_norm": 168.68495178222656, "learning_rate": 9.290707236842106e-06, "loss": 2.5527, "step": 345 }, { "epoch": 0.07113486842105263, "grad_norm": 25.288623809814453, "learning_rate": 9.288651315789474e-06, "loss": 0.1942, "step": 346 }, { "epoch": 0.07134046052631579, "grad_norm": 109.98817443847656, "learning_rate": 9.286595394736842e-06, "loss": 1.5703, "step": 347 }, { "epoch": 0.07154605263157894, "grad_norm": 51.776451110839844, "learning_rate": 9.284539473684211e-06, "loss": 0.8633, "step": 348 }, { "epoch": 0.0717516447368421, "grad_norm": 238.97979736328125, "learning_rate": 9.28248355263158e-06, "loss": 2.415, "step": 349 }, { "epoch": 0.07195723684210527, "grad_norm": 71.90460205078125, "learning_rate": 9.280427631578947e-06, "loss": 1.4844, "step": 350 }, { "epoch": 0.07216282894736842, "grad_norm": 50.335166931152344, "learning_rate": 9.278371710526317e-06, "loss": 1.4941, "step": 351 }, { "epoch": 0.07236842105263158, "grad_norm": 72.93436431884766, "learning_rate": 9.276315789473686e-06, "loss": 1.7852, "step": 352 }, { "epoch": 0.07257401315789473, "grad_norm": 46.42775344848633, "learning_rate": 9.274259868421054e-06, "loss": 0.791, "step": 353 }, { "epoch": 0.0727796052631579, "grad_norm": 137.2333984375, "learning_rate": 9.272203947368422e-06, "loss": 3.2969, "step": 354 }, { "epoch": 0.07298519736842106, "grad_norm": 44.80105209350586, "learning_rate": 9.27014802631579e-06, "loss": 1.1365, "step": 355 }, { "epoch": 0.07319078947368421, "grad_norm": 58.527339935302734, "learning_rate": 9.26809210526316e-06, "loss": 1.1318, "step": 356 }, { "epoch": 0.07339638157894737, "grad_norm": 137.2529296875, "learning_rate": 9.266036184210527e-06, "loss": 4.1172, "step": 357 }, { "epoch": 0.07360197368421052, "grad_norm": 54.234554290771484, "learning_rate": 9.263980263157895e-06, "loss": 1.4062, "step": 358 }, { "epoch": 0.07380756578947369, "grad_norm": 104.55722045898438, "learning_rate": 9.261924342105265e-06, "loss": 2.0781, "step": 359 }, { "epoch": 0.07401315789473684, "grad_norm": 42.222511291503906, "learning_rate": 9.259868421052633e-06, "loss": 1.4023, "step": 360 }, { "epoch": 0.07421875, "grad_norm": 61.9639892578125, "learning_rate": 9.2578125e-06, "loss": 1.3516, "step": 361 }, { "epoch": 0.07442434210526316, "grad_norm": 65.20064544677734, "learning_rate": 9.255756578947368e-06, "loss": 1.668, "step": 362 }, { "epoch": 0.07462993421052631, "grad_norm": 47.0842399597168, "learning_rate": 9.253700657894738e-06, "loss": 1.0605, "step": 363 }, { "epoch": 0.07483552631578948, "grad_norm": 36.10401916503906, "learning_rate": 9.251644736842106e-06, "loss": 0.3901, "step": 364 }, { "epoch": 0.07504111842105263, "grad_norm": 57.47480010986328, "learning_rate": 9.249588815789474e-06, "loss": 1.0176, "step": 365 }, { "epoch": 0.07524671052631579, "grad_norm": 56.571815490722656, "learning_rate": 9.247532894736843e-06, "loss": 1.167, "step": 366 }, { "epoch": 0.07545230263157894, "grad_norm": 52.939090728759766, "learning_rate": 9.245476973684211e-06, "loss": 1.1289, "step": 367 }, { "epoch": 0.0756578947368421, "grad_norm": 109.11473846435547, "learning_rate": 9.243421052631579e-06, "loss": 2.5703, "step": 368 }, { "epoch": 0.07586348684210527, "grad_norm": 28.54892921447754, "learning_rate": 9.241365131578947e-06, "loss": 0.9155, "step": 369 }, { "epoch": 0.07606907894736842, "grad_norm": 56.282127380371094, "learning_rate": 9.239309210526317e-06, "loss": 1.834, "step": 370 }, { "epoch": 0.07627467105263158, "grad_norm": 51.99695587158203, "learning_rate": 9.237253289473686e-06, "loss": 1.2344, "step": 371 }, { "epoch": 0.07648026315789473, "grad_norm": 88.74778747558594, "learning_rate": 9.235197368421054e-06, "loss": 1.3418, "step": 372 }, { "epoch": 0.0766858552631579, "grad_norm": 66.2579116821289, "learning_rate": 9.233141447368422e-06, "loss": 1.5742, "step": 373 }, { "epoch": 0.07689144736842106, "grad_norm": 46.30642318725586, "learning_rate": 9.23108552631579e-06, "loss": 1.6699, "step": 374 }, { "epoch": 0.07709703947368421, "grad_norm": 68.09831237792969, "learning_rate": 9.22902960526316e-06, "loss": 1.9609, "step": 375 }, { "epoch": 0.07730263157894737, "grad_norm": 35.26036834716797, "learning_rate": 9.226973684210527e-06, "loss": 0.7422, "step": 376 }, { "epoch": 0.07750822368421052, "grad_norm": 32.12290573120117, "learning_rate": 9.224917763157895e-06, "loss": 0.603, "step": 377 }, { "epoch": 0.07771381578947369, "grad_norm": 63.565208435058594, "learning_rate": 9.222861842105265e-06, "loss": 2.0977, "step": 378 }, { "epoch": 0.07791940789473684, "grad_norm": 47.623435974121094, "learning_rate": 9.220805921052633e-06, "loss": 2.0742, "step": 379 }, { "epoch": 0.078125, "grad_norm": 133.25247192382812, "learning_rate": 9.21875e-06, "loss": 4.207, "step": 380 }, { "epoch": 0.07833059210526316, "grad_norm": 63.67714309692383, "learning_rate": 9.216694078947368e-06, "loss": 1.6055, "step": 381 }, { "epoch": 0.07853618421052631, "grad_norm": 27.7446231842041, "learning_rate": 9.214638157894738e-06, "loss": 0.5469, "step": 382 }, { "epoch": 0.07874177631578948, "grad_norm": 61.205806732177734, "learning_rate": 9.212582236842106e-06, "loss": 1.75, "step": 383 }, { "epoch": 0.07894736842105263, "grad_norm": 43.072532653808594, "learning_rate": 9.210526315789474e-06, "loss": 0.8701, "step": 384 }, { "epoch": 0.07915296052631579, "grad_norm": 51.761295318603516, "learning_rate": 9.208470394736843e-06, "loss": 0.9922, "step": 385 }, { "epoch": 0.07935855263157894, "grad_norm": 46.98358154296875, "learning_rate": 9.206414473684211e-06, "loss": 1.6484, "step": 386 }, { "epoch": 0.0795641447368421, "grad_norm": 114.98619842529297, "learning_rate": 9.204358552631579e-06, "loss": 1.9062, "step": 387 }, { "epoch": 0.07976973684210527, "grad_norm": 35.20199203491211, "learning_rate": 9.202302631578947e-06, "loss": 1.3188, "step": 388 }, { "epoch": 0.07997532894736842, "grad_norm": 60.974517822265625, "learning_rate": 9.200246710526316e-06, "loss": 1.6797, "step": 389 }, { "epoch": 0.08018092105263158, "grad_norm": 33.2817497253418, "learning_rate": 9.198190789473686e-06, "loss": 0.2773, "step": 390 }, { "epoch": 0.08038651315789473, "grad_norm": 34.50399398803711, "learning_rate": 9.196134868421054e-06, "loss": 0.5178, "step": 391 }, { "epoch": 0.0805921052631579, "grad_norm": 67.05232238769531, "learning_rate": 9.194078947368422e-06, "loss": 1.5312, "step": 392 }, { "epoch": 0.08079769736842106, "grad_norm": 53.91796112060547, "learning_rate": 9.192023026315791e-06, "loss": 0.9692, "step": 393 }, { "epoch": 0.08100328947368421, "grad_norm": 64.1976089477539, "learning_rate": 9.18996710526316e-06, "loss": 1.5898, "step": 394 }, { "epoch": 0.08120888157894737, "grad_norm": 52.68128967285156, "learning_rate": 9.187911184210527e-06, "loss": 1.3516, "step": 395 }, { "epoch": 0.08141447368421052, "grad_norm": 59.86910629272461, "learning_rate": 9.185855263157895e-06, "loss": 1.6484, "step": 396 }, { "epoch": 0.08162006578947369, "grad_norm": 55.737464904785156, "learning_rate": 9.183799342105265e-06, "loss": 0.9902, "step": 397 }, { "epoch": 0.08182565789473684, "grad_norm": 29.606901168823242, "learning_rate": 9.181743421052632e-06, "loss": 0.7256, "step": 398 }, { "epoch": 0.08203125, "grad_norm": 43.858131408691406, "learning_rate": 9.1796875e-06, "loss": 0.7412, "step": 399 }, { "epoch": 0.08223684210526316, "grad_norm": 25.700777053833008, "learning_rate": 9.17763157894737e-06, "loss": 1.0117, "step": 400 }, { "epoch": 0.08244243421052631, "grad_norm": 48.33639144897461, "learning_rate": 9.175575657894738e-06, "loss": 1.4414, "step": 401 }, { "epoch": 0.08264802631578948, "grad_norm": 39.56529998779297, "learning_rate": 9.173519736842106e-06, "loss": 1.2148, "step": 402 }, { "epoch": 0.08285361842105263, "grad_norm": 75.14580535888672, "learning_rate": 9.171463815789474e-06, "loss": 1.2109, "step": 403 }, { "epoch": 0.08305921052631579, "grad_norm": 91.44287872314453, "learning_rate": 9.169407894736843e-06, "loss": 1.7344, "step": 404 }, { "epoch": 0.08326480263157894, "grad_norm": 119.52699279785156, "learning_rate": 9.167351973684211e-06, "loss": 0.7402, "step": 405 }, { "epoch": 0.0834703947368421, "grad_norm": 44.99494171142578, "learning_rate": 9.165296052631579e-06, "loss": 0.9219, "step": 406 }, { "epoch": 0.08367598684210527, "grad_norm": 66.73922729492188, "learning_rate": 9.163240131578948e-06, "loss": 1.4141, "step": 407 }, { "epoch": 0.08388157894736842, "grad_norm": 41.1556282043457, "learning_rate": 9.161184210526316e-06, "loss": 1.6133, "step": 408 }, { "epoch": 0.08408717105263158, "grad_norm": 60.30184555053711, "learning_rate": 9.159128289473686e-06, "loss": 1.6016, "step": 409 }, { "epoch": 0.08429276315789473, "grad_norm": 60.43864440917969, "learning_rate": 9.157072368421054e-06, "loss": 1.2188, "step": 410 }, { "epoch": 0.0844983552631579, "grad_norm": 74.89896392822266, "learning_rate": 9.155016447368422e-06, "loss": 1.8867, "step": 411 }, { "epoch": 0.08470394736842106, "grad_norm": 82.18927764892578, "learning_rate": 9.152960526315791e-06, "loss": 1.0254, "step": 412 }, { "epoch": 0.08490953947368421, "grad_norm": 97.62265014648438, "learning_rate": 9.15090460526316e-06, "loss": 0.8662, "step": 413 }, { "epoch": 0.08511513157894737, "grad_norm": 55.48585891723633, "learning_rate": 9.148848684210527e-06, "loss": 1.2227, "step": 414 }, { "epoch": 0.08532072368421052, "grad_norm": 41.93524169921875, "learning_rate": 9.146792763157895e-06, "loss": 0.8457, "step": 415 }, { "epoch": 0.08552631578947369, "grad_norm": 32.4094352722168, "learning_rate": 9.144736842105264e-06, "loss": 0.7056, "step": 416 }, { "epoch": 0.08573190789473684, "grad_norm": 44.75262451171875, "learning_rate": 9.142680921052632e-06, "loss": 1.0684, "step": 417 }, { "epoch": 0.0859375, "grad_norm": 82.04063415527344, "learning_rate": 9.140625e-06, "loss": 2.375, "step": 418 }, { "epoch": 0.08614309210526316, "grad_norm": 39.31158447265625, "learning_rate": 9.13856907894737e-06, "loss": 1.5234, "step": 419 }, { "epoch": 0.08634868421052631, "grad_norm": 53.82627868652344, "learning_rate": 9.136513157894738e-06, "loss": 1.4297, "step": 420 }, { "epoch": 0.08655427631578948, "grad_norm": 44.05521011352539, "learning_rate": 9.134457236842106e-06, "loss": 0.7344, "step": 421 }, { "epoch": 0.08675986842105263, "grad_norm": 28.152185440063477, "learning_rate": 9.132401315789473e-06, "loss": 0.2146, "step": 422 }, { "epoch": 0.08696546052631579, "grad_norm": 80.3797607421875, "learning_rate": 9.130345394736843e-06, "loss": 1.7539, "step": 423 }, { "epoch": 0.08717105263157894, "grad_norm": 116.70099639892578, "learning_rate": 9.128289473684211e-06, "loss": 2.0969, "step": 424 }, { "epoch": 0.0873766447368421, "grad_norm": 55.04074478149414, "learning_rate": 9.126233552631579e-06, "loss": 1.0703, "step": 425 }, { "epoch": 0.08758223684210527, "grad_norm": 36.19375991821289, "learning_rate": 9.124177631578948e-06, "loss": 1.0, "step": 426 }, { "epoch": 0.08778782894736842, "grad_norm": 79.45502471923828, "learning_rate": 9.122121710526316e-06, "loss": 2.3125, "step": 427 }, { "epoch": 0.08799342105263158, "grad_norm": 26.826318740844727, "learning_rate": 9.120065789473686e-06, "loss": 0.4482, "step": 428 }, { "epoch": 0.08819901315789473, "grad_norm": 54.794654846191406, "learning_rate": 9.118009868421054e-06, "loss": 1.0332, "step": 429 }, { "epoch": 0.0884046052631579, "grad_norm": 41.12156677246094, "learning_rate": 9.115953947368422e-06, "loss": 1.0996, "step": 430 }, { "epoch": 0.08861019736842106, "grad_norm": 57.06736373901367, "learning_rate": 9.113898026315791e-06, "loss": 0.7393, "step": 431 }, { "epoch": 0.08881578947368421, "grad_norm": 38.87145233154297, "learning_rate": 9.111842105263159e-06, "loss": 0.5869, "step": 432 }, { "epoch": 0.08902138157894737, "grad_norm": 39.577945709228516, "learning_rate": 9.109786184210527e-06, "loss": 0.6436, "step": 433 }, { "epoch": 0.08922697368421052, "grad_norm": 40.98161315917969, "learning_rate": 9.107730263157895e-06, "loss": 0.9902, "step": 434 }, { "epoch": 0.08943256578947369, "grad_norm": 103.50910949707031, "learning_rate": 9.105674342105264e-06, "loss": 2.1289, "step": 435 }, { "epoch": 0.08963815789473684, "grad_norm": 46.874149322509766, "learning_rate": 9.103618421052632e-06, "loss": 0.9551, "step": 436 }, { "epoch": 0.08984375, "grad_norm": 55.521419525146484, "learning_rate": 9.1015625e-06, "loss": 0.9272, "step": 437 }, { "epoch": 0.09004934210526316, "grad_norm": 27.40374183654785, "learning_rate": 9.09950657894737e-06, "loss": 0.8047, "step": 438 }, { "epoch": 0.09025493421052631, "grad_norm": 85.0200424194336, "learning_rate": 9.097450657894738e-06, "loss": 1.5146, "step": 439 }, { "epoch": 0.09046052631578948, "grad_norm": 60.651512145996094, "learning_rate": 9.095394736842105e-06, "loss": 0.9167, "step": 440 }, { "epoch": 0.09066611842105263, "grad_norm": 31.83222007751465, "learning_rate": 9.093338815789473e-06, "loss": 0.6914, "step": 441 }, { "epoch": 0.09087171052631579, "grad_norm": 72.23406219482422, "learning_rate": 9.091282894736843e-06, "loss": 1.4297, "step": 442 }, { "epoch": 0.09107730263157894, "grad_norm": 41.55168914794922, "learning_rate": 9.08922697368421e-06, "loss": 1.2402, "step": 443 }, { "epoch": 0.0912828947368421, "grad_norm": 32.14000701904297, "learning_rate": 9.087171052631579e-06, "loss": 0.7275, "step": 444 }, { "epoch": 0.09148848684210527, "grad_norm": 90.45081329345703, "learning_rate": 9.085115131578948e-06, "loss": 1.5938, "step": 445 }, { "epoch": 0.09169407894736842, "grad_norm": 160.26524353027344, "learning_rate": 9.083059210526316e-06, "loss": 1.793, "step": 446 }, { "epoch": 0.09189967105263158, "grad_norm": 34.58512878417969, "learning_rate": 9.081003289473686e-06, "loss": 0.3506, "step": 447 }, { "epoch": 0.09210526315789473, "grad_norm": 89.3415298461914, "learning_rate": 9.078947368421054e-06, "loss": 2.0391, "step": 448 }, { "epoch": 0.0923108552631579, "grad_norm": 21.30736541748047, "learning_rate": 9.076891447368422e-06, "loss": 0.3647, "step": 449 }, { "epoch": 0.09251644736842106, "grad_norm": 43.312625885009766, "learning_rate": 9.074835526315791e-06, "loss": 0.6055, "step": 450 }, { "epoch": 0.09272203947368421, "grad_norm": 114.7179183959961, "learning_rate": 9.072779605263159e-06, "loss": 1.3359, "step": 451 }, { "epoch": 0.09292763157894737, "grad_norm": 95.91043853759766, "learning_rate": 9.070723684210527e-06, "loss": 1.5234, "step": 452 }, { "epoch": 0.09313322368421052, "grad_norm": 32.4932861328125, "learning_rate": 9.068667763157896e-06, "loss": 0.2995, "step": 453 }, { "epoch": 0.09333881578947369, "grad_norm": 62.683502197265625, "learning_rate": 9.066611842105264e-06, "loss": 1.4727, "step": 454 }, { "epoch": 0.09354440789473684, "grad_norm": 112.28824615478516, "learning_rate": 9.064555921052632e-06, "loss": 1.3711, "step": 455 }, { "epoch": 0.09375, "grad_norm": 46.08243942260742, "learning_rate": 9.0625e-06, "loss": 1.5508, "step": 456 }, { "epoch": 0.09395559210526316, "grad_norm": 6.434681415557861, "learning_rate": 9.06044407894737e-06, "loss": 0.0301, "step": 457 }, { "epoch": 0.09416118421052631, "grad_norm": 58.43425750732422, "learning_rate": 9.058388157894738e-06, "loss": 1.2012, "step": 458 }, { "epoch": 0.09436677631578948, "grad_norm": 52.69418716430664, "learning_rate": 9.056332236842105e-06, "loss": 1.123, "step": 459 }, { "epoch": 0.09457236842105263, "grad_norm": 116.84672546386719, "learning_rate": 9.054276315789475e-06, "loss": 4.7471, "step": 460 }, { "epoch": 0.09477796052631579, "grad_norm": 65.88794708251953, "learning_rate": 9.052220394736843e-06, "loss": 1.1025, "step": 461 }, { "epoch": 0.09498355263157894, "grad_norm": 44.127254486083984, "learning_rate": 9.05016447368421e-06, "loss": 0.5254, "step": 462 }, { "epoch": 0.0951891447368421, "grad_norm": 51.21589279174805, "learning_rate": 9.048108552631579e-06, "loss": 0.5181, "step": 463 }, { "epoch": 0.09539473684210527, "grad_norm": 75.06077575683594, "learning_rate": 9.046052631578948e-06, "loss": 1.3242, "step": 464 }, { "epoch": 0.09560032894736842, "grad_norm": 24.185739517211914, "learning_rate": 9.043996710526316e-06, "loss": 0.1935, "step": 465 }, { "epoch": 0.09580592105263158, "grad_norm": 52.345916748046875, "learning_rate": 9.041940789473686e-06, "loss": 1.3125, "step": 466 }, { "epoch": 0.09601151315789473, "grad_norm": 182.30845642089844, "learning_rate": 9.039884868421054e-06, "loss": 2.4414, "step": 467 }, { "epoch": 0.0962171052631579, "grad_norm": 46.852943420410156, "learning_rate": 9.037828947368421e-06, "loss": 0.7529, "step": 468 }, { "epoch": 0.09642269736842106, "grad_norm": 81.45367431640625, "learning_rate": 9.035773026315791e-06, "loss": 0.7734, "step": 469 }, { "epoch": 0.09662828947368421, "grad_norm": 75.91923522949219, "learning_rate": 9.033717105263159e-06, "loss": 1.4463, "step": 470 }, { "epoch": 0.09683388157894737, "grad_norm": 69.4185791015625, "learning_rate": 9.031661184210527e-06, "loss": 1.1992, "step": 471 }, { "epoch": 0.09703947368421052, "grad_norm": 113.5587387084961, "learning_rate": 9.029605263157896e-06, "loss": 2.5977, "step": 472 }, { "epoch": 0.09724506578947369, "grad_norm": 118.1943588256836, "learning_rate": 9.027549342105264e-06, "loss": 2.1533, "step": 473 }, { "epoch": 0.09745065789473684, "grad_norm": 36.538936614990234, "learning_rate": 9.025493421052632e-06, "loss": 0.8215, "step": 474 }, { "epoch": 0.09765625, "grad_norm": 28.604766845703125, "learning_rate": 9.0234375e-06, "loss": 0.2812, "step": 475 }, { "epoch": 0.09786184210526316, "grad_norm": 55.23013687133789, "learning_rate": 9.02138157894737e-06, "loss": 0.5586, "step": 476 }, { "epoch": 0.09806743421052631, "grad_norm": 26.155681610107422, "learning_rate": 9.019325657894737e-06, "loss": 0.2393, "step": 477 }, { "epoch": 0.09827302631578948, "grad_norm": 53.908836364746094, "learning_rate": 9.017269736842105e-06, "loss": 0.8281, "step": 478 }, { "epoch": 0.09847861842105263, "grad_norm": 79.45370483398438, "learning_rate": 9.015213815789475e-06, "loss": 1.8281, "step": 479 }, { "epoch": 0.09868421052631579, "grad_norm": 18.244407653808594, "learning_rate": 9.013157894736843e-06, "loss": 0.1193, "step": 480 }, { "epoch": 0.09888980263157894, "grad_norm": 132.9093017578125, "learning_rate": 9.01110197368421e-06, "loss": 2.8789, "step": 481 }, { "epoch": 0.0990953947368421, "grad_norm": 94.24540710449219, "learning_rate": 9.009046052631579e-06, "loss": 1.2266, "step": 482 }, { "epoch": 0.09930098684210527, "grad_norm": 96.02490234375, "learning_rate": 9.006990131578948e-06, "loss": 1.5781, "step": 483 }, { "epoch": 0.09950657894736842, "grad_norm": 65.41879272460938, "learning_rate": 9.004934210526316e-06, "loss": 1.002, "step": 484 }, { "epoch": 0.09971217105263158, "grad_norm": 49.90369415283203, "learning_rate": 9.002878289473686e-06, "loss": 0.5996, "step": 485 }, { "epoch": 0.09991776315789473, "grad_norm": 23.771255493164062, "learning_rate": 9.000822368421053e-06, "loss": 0.1675, "step": 486 }, { "epoch": 0.1001233552631579, "grad_norm": 34.21731185913086, "learning_rate": 8.998766447368423e-06, "loss": 1.123, "step": 487 }, { "epoch": 0.10032894736842106, "grad_norm": 54.1513671875, "learning_rate": 8.996710526315791e-06, "loss": 0.9023, "step": 488 }, { "epoch": 0.10053453947368421, "grad_norm": 95.0126724243164, "learning_rate": 8.994654605263159e-06, "loss": 0.998, "step": 489 }, { "epoch": 0.10074013157894737, "grad_norm": 54.945533752441406, "learning_rate": 8.992598684210527e-06, "loss": 0.7588, "step": 490 }, { "epoch": 0.10094572368421052, "grad_norm": 62.35584259033203, "learning_rate": 8.990542763157896e-06, "loss": 1.0547, "step": 491 }, { "epoch": 0.10115131578947369, "grad_norm": 74.60970306396484, "learning_rate": 8.988486842105264e-06, "loss": 1.3125, "step": 492 }, { "epoch": 0.10135690789473684, "grad_norm": 57.24214172363281, "learning_rate": 8.986430921052632e-06, "loss": 0.5615, "step": 493 }, { "epoch": 0.1015625, "grad_norm": 109.57658386230469, "learning_rate": 8.984375000000002e-06, "loss": 2.0859, "step": 494 }, { "epoch": 0.10176809210526316, "grad_norm": 23.76524543762207, "learning_rate": 8.98231907894737e-06, "loss": 0.4209, "step": 495 }, { "epoch": 0.10197368421052631, "grad_norm": 79.11624145507812, "learning_rate": 8.980263157894737e-06, "loss": 1.0879, "step": 496 }, { "epoch": 0.10217927631578948, "grad_norm": 51.99403762817383, "learning_rate": 8.978207236842105e-06, "loss": 0.4878, "step": 497 }, { "epoch": 0.10238486842105263, "grad_norm": 47.00723648071289, "learning_rate": 8.976151315789475e-06, "loss": 0.5952, "step": 498 }, { "epoch": 0.10259046052631579, "grad_norm": 49.25107955932617, "learning_rate": 8.974095394736843e-06, "loss": 0.832, "step": 499 }, { "epoch": 0.10279605263157894, "grad_norm": 24.258941650390625, "learning_rate": 8.97203947368421e-06, "loss": 0.2119, "step": 500 }, { "epoch": 0.1030016447368421, "grad_norm": 106.48870849609375, "learning_rate": 8.96998355263158e-06, "loss": 3.0547, "step": 501 }, { "epoch": 0.10320723684210527, "grad_norm": 14.608073234558105, "learning_rate": 8.967927631578948e-06, "loss": 0.0911, "step": 502 }, { "epoch": 0.10341282894736842, "grad_norm": 30.091787338256836, "learning_rate": 8.965871710526316e-06, "loss": 0.2014, "step": 503 }, { "epoch": 0.10361842105263158, "grad_norm": 74.08060455322266, "learning_rate": 8.963815789473685e-06, "loss": 1.1445, "step": 504 }, { "epoch": 0.10382401315789473, "grad_norm": 14.84162712097168, "learning_rate": 8.961759868421053e-06, "loss": 0.2082, "step": 505 }, { "epoch": 0.1040296052631579, "grad_norm": 65.01032257080078, "learning_rate": 8.959703947368423e-06, "loss": 1.8828, "step": 506 }, { "epoch": 0.10423519736842106, "grad_norm": 52.74924850463867, "learning_rate": 8.95764802631579e-06, "loss": 0.7656, "step": 507 }, { "epoch": 0.10444078947368421, "grad_norm": 96.07575988769531, "learning_rate": 8.955592105263159e-06, "loss": 2.0938, "step": 508 }, { "epoch": 0.10464638157894737, "grad_norm": 72.13784790039062, "learning_rate": 8.953536184210527e-06, "loss": 1.707, "step": 509 }, { "epoch": 0.10485197368421052, "grad_norm": 30.789823532104492, "learning_rate": 8.951480263157896e-06, "loss": 0.2675, "step": 510 }, { "epoch": 0.10505756578947369, "grad_norm": 55.09064865112305, "learning_rate": 8.949424342105264e-06, "loss": 0.5625, "step": 511 }, { "epoch": 0.10526315789473684, "grad_norm": 50.57075500488281, "learning_rate": 8.947368421052632e-06, "loss": 0.8511, "step": 512 }, { "epoch": 0.10546875, "grad_norm": 68.86466979980469, "learning_rate": 8.945312500000001e-06, "loss": 1.2227, "step": 513 }, { "epoch": 0.10567434210526316, "grad_norm": 49.604087829589844, "learning_rate": 8.94325657894737e-06, "loss": 0.4844, "step": 514 }, { "epoch": 0.10587993421052631, "grad_norm": 101.70752716064453, "learning_rate": 8.941200657894737e-06, "loss": 1.1963, "step": 515 }, { "epoch": 0.10608552631578948, "grad_norm": 20.733449935913086, "learning_rate": 8.939144736842105e-06, "loss": 0.1188, "step": 516 }, { "epoch": 0.10629111842105263, "grad_norm": 28.02437973022461, "learning_rate": 8.937088815789475e-06, "loss": 0.2214, "step": 517 }, { "epoch": 0.10649671052631579, "grad_norm": 51.40896224975586, "learning_rate": 8.935032894736843e-06, "loss": 0.5186, "step": 518 }, { "epoch": 0.10670230263157894, "grad_norm": 179.7532501220703, "learning_rate": 8.93297697368421e-06, "loss": 2.4844, "step": 519 }, { "epoch": 0.1069078947368421, "grad_norm": 59.62074279785156, "learning_rate": 8.93092105263158e-06, "loss": 1.1396, "step": 520 }, { "epoch": 0.10711348684210527, "grad_norm": 34.643524169921875, "learning_rate": 8.928865131578948e-06, "loss": 0.3428, "step": 521 }, { "epoch": 0.10731907894736842, "grad_norm": 18.8770751953125, "learning_rate": 8.926809210526316e-06, "loss": 0.1699, "step": 522 }, { "epoch": 0.10752467105263158, "grad_norm": 32.40262985229492, "learning_rate": 8.924753289473685e-06, "loss": 0.2402, "step": 523 }, { "epoch": 0.10773026315789473, "grad_norm": 30.51153564453125, "learning_rate": 8.922697368421053e-06, "loss": 0.48, "step": 524 }, { "epoch": 0.1079358552631579, "grad_norm": 24.62276268005371, "learning_rate": 8.920641447368423e-06, "loss": 0.1769, "step": 525 }, { "epoch": 0.10814144736842106, "grad_norm": 92.70629119873047, "learning_rate": 8.91858552631579e-06, "loss": 1.5762, "step": 526 }, { "epoch": 0.10834703947368421, "grad_norm": 185.435302734375, "learning_rate": 8.916529605263159e-06, "loss": 3.9141, "step": 527 }, { "epoch": 0.10855263157894737, "grad_norm": 75.98495483398438, "learning_rate": 8.914473684210528e-06, "loss": 1.0703, "step": 528 }, { "epoch": 0.10875822368421052, "grad_norm": 19.31216812133789, "learning_rate": 8.912417763157896e-06, "loss": 0.1071, "step": 529 }, { "epoch": 0.10896381578947369, "grad_norm": 46.42657470703125, "learning_rate": 8.910361842105264e-06, "loss": 0.9297, "step": 530 }, { "epoch": 0.10916940789473684, "grad_norm": 12.118858337402344, "learning_rate": 8.908305921052632e-06, "loss": 0.0625, "step": 531 }, { "epoch": 0.109375, "grad_norm": 37.3786506652832, "learning_rate": 8.906250000000001e-06, "loss": 0.3389, "step": 532 }, { "epoch": 0.10958059210526316, "grad_norm": 46.39356994628906, "learning_rate": 8.90419407894737e-06, "loss": 1.2969, "step": 533 }, { "epoch": 0.10978618421052631, "grad_norm": 84.6044692993164, "learning_rate": 8.902138157894737e-06, "loss": 2.5625, "step": 534 }, { "epoch": 0.10999177631578948, "grad_norm": 42.14690017700195, "learning_rate": 8.900082236842105e-06, "loss": 0.6123, "step": 535 }, { "epoch": 0.11019736842105263, "grad_norm": 71.29698944091797, "learning_rate": 8.898026315789475e-06, "loss": 1.2266, "step": 536 }, { "epoch": 0.11040296052631579, "grad_norm": 343.23724365234375, "learning_rate": 8.895970394736842e-06, "loss": 2.1415, "step": 537 }, { "epoch": 0.11060855263157894, "grad_norm": 60.57801055908203, "learning_rate": 8.89391447368421e-06, "loss": 0.4893, "step": 538 }, { "epoch": 0.1108141447368421, "grad_norm": 30.034034729003906, "learning_rate": 8.89185855263158e-06, "loss": 0.2221, "step": 539 }, { "epoch": 0.11101973684210527, "grad_norm": 92.22252655029297, "learning_rate": 8.889802631578948e-06, "loss": 1.8867, "step": 540 }, { "epoch": 0.11122532894736842, "grad_norm": 63.38169479370117, "learning_rate": 8.887746710526316e-06, "loss": 1.7246, "step": 541 }, { "epoch": 0.11143092105263158, "grad_norm": 50.7624397277832, "learning_rate": 8.885690789473685e-06, "loss": 0.6309, "step": 542 }, { "epoch": 0.11163651315789473, "grad_norm": 62.52885437011719, "learning_rate": 8.883634868421053e-06, "loss": 0.6044, "step": 543 }, { "epoch": 0.1118421052631579, "grad_norm": 47.585899353027344, "learning_rate": 8.881578947368423e-06, "loss": 0.3184, "step": 544 }, { "epoch": 0.11204769736842106, "grad_norm": 75.10714721679688, "learning_rate": 8.87952302631579e-06, "loss": 1.3496, "step": 545 }, { "epoch": 0.11225328947368421, "grad_norm": 78.04474639892578, "learning_rate": 8.877467105263158e-06, "loss": 1.3945, "step": 546 }, { "epoch": 0.11245888157894737, "grad_norm": 55.10225296020508, "learning_rate": 8.875411184210528e-06, "loss": 0.7305, "step": 547 }, { "epoch": 0.11266447368421052, "grad_norm": 41.95757293701172, "learning_rate": 8.873355263157896e-06, "loss": 0.3457, "step": 548 }, { "epoch": 0.11287006578947369, "grad_norm": 80.20491027832031, "learning_rate": 8.871299342105264e-06, "loss": 1.9297, "step": 549 }, { "epoch": 0.11307565789473684, "grad_norm": 51.45291519165039, "learning_rate": 8.869243421052632e-06, "loss": 0.6519, "step": 550 }, { "epoch": 0.11328125, "grad_norm": 82.49842834472656, "learning_rate": 8.867187500000001e-06, "loss": 1.7031, "step": 551 }, { "epoch": 0.11348684210526316, "grad_norm": 122.59345245361328, "learning_rate": 8.865131578947369e-06, "loss": 2.0476, "step": 552 }, { "epoch": 0.11369243421052631, "grad_norm": 37.445518493652344, "learning_rate": 8.863075657894737e-06, "loss": 0.2871, "step": 553 }, { "epoch": 0.11389802631578948, "grad_norm": 33.076988220214844, "learning_rate": 8.861019736842107e-06, "loss": 0.7402, "step": 554 }, { "epoch": 0.11410361842105263, "grad_norm": 76.73178100585938, "learning_rate": 8.858963815789474e-06, "loss": 1.2407, "step": 555 }, { "epoch": 0.11430921052631579, "grad_norm": 50.65952682495117, "learning_rate": 8.856907894736842e-06, "loss": 1.0508, "step": 556 }, { "epoch": 0.11451480263157894, "grad_norm": 113.38887786865234, "learning_rate": 8.85485197368421e-06, "loss": 1.7344, "step": 557 }, { "epoch": 0.1147203947368421, "grad_norm": 80.32279968261719, "learning_rate": 8.85279605263158e-06, "loss": 1.4419, "step": 558 }, { "epoch": 0.11492598684210527, "grad_norm": 12.285173416137695, "learning_rate": 8.850740131578948e-06, "loss": 0.0535, "step": 559 }, { "epoch": 0.11513157894736842, "grad_norm": 57.930416107177734, "learning_rate": 8.848684210526316e-06, "loss": 1.1719, "step": 560 }, { "epoch": 0.11533717105263158, "grad_norm": 35.41834259033203, "learning_rate": 8.846628289473685e-06, "loss": 0.2417, "step": 561 }, { "epoch": 0.11554276315789473, "grad_norm": 31.536388397216797, "learning_rate": 8.844572368421053e-06, "loss": 0.8828, "step": 562 }, { "epoch": 0.1157483552631579, "grad_norm": 57.997840881347656, "learning_rate": 8.842516447368423e-06, "loss": 0.9727, "step": 563 }, { "epoch": 0.11595394736842106, "grad_norm": 74.02205657958984, "learning_rate": 8.84046052631579e-06, "loss": 1.1699, "step": 564 }, { "epoch": 0.11615953947368421, "grad_norm": 46.27134323120117, "learning_rate": 8.838404605263158e-06, "loss": 0.6475, "step": 565 }, { "epoch": 0.11636513157894737, "grad_norm": 80.5855712890625, "learning_rate": 8.836348684210528e-06, "loss": 1.5039, "step": 566 }, { "epoch": 0.11657072368421052, "grad_norm": 35.01012420654297, "learning_rate": 8.834292763157896e-06, "loss": 0.2402, "step": 567 }, { "epoch": 0.11677631578947369, "grad_norm": 35.913021087646484, "learning_rate": 8.832236842105264e-06, "loss": 0.353, "step": 568 }, { "epoch": 0.11698190789473684, "grad_norm": 36.755977630615234, "learning_rate": 8.830180921052632e-06, "loss": 0.876, "step": 569 }, { "epoch": 0.1171875, "grad_norm": 71.32474517822266, "learning_rate": 8.828125000000001e-06, "loss": 1.2324, "step": 570 }, { "epoch": 0.11739309210526316, "grad_norm": 65.75851440429688, "learning_rate": 8.826069078947369e-06, "loss": 0.667, "step": 571 }, { "epoch": 0.11759868421052631, "grad_norm": 37.135982513427734, "learning_rate": 8.824013157894737e-06, "loss": 0.6621, "step": 572 }, { "epoch": 0.11780427631578948, "grad_norm": 156.21673583984375, "learning_rate": 8.821957236842106e-06, "loss": 2.7266, "step": 573 }, { "epoch": 0.11800986842105263, "grad_norm": 56.560516357421875, "learning_rate": 8.819901315789474e-06, "loss": 1.3672, "step": 574 }, { "epoch": 0.11821546052631579, "grad_norm": 57.547367095947266, "learning_rate": 8.817845394736842e-06, "loss": 1.1875, "step": 575 }, { "epoch": 0.11842105263157894, "grad_norm": 118.61968231201172, "learning_rate": 8.81578947368421e-06, "loss": 1.3984, "step": 576 }, { "epoch": 0.1186266447368421, "grad_norm": 125.99451446533203, "learning_rate": 8.81373355263158e-06, "loss": 1.5625, "step": 577 }, { "epoch": 0.11883223684210527, "grad_norm": 41.52878189086914, "learning_rate": 8.811677631578948e-06, "loss": 0.9238, "step": 578 }, { "epoch": 0.11903782894736842, "grad_norm": 11.081133842468262, "learning_rate": 8.809621710526315e-06, "loss": 0.051, "step": 579 }, { "epoch": 0.11924342105263158, "grad_norm": 24.236080169677734, "learning_rate": 8.807565789473685e-06, "loss": 0.2139, "step": 580 }, { "epoch": 0.11944901315789473, "grad_norm": 45.85995101928711, "learning_rate": 8.805509868421055e-06, "loss": 0.8804, "step": 581 }, { "epoch": 0.1196546052631579, "grad_norm": 33.53053665161133, "learning_rate": 8.803453947368422e-06, "loss": 0.3623, "step": 582 }, { "epoch": 0.11986019736842106, "grad_norm": 25.469436645507812, "learning_rate": 8.80139802631579e-06, "loss": 0.269, "step": 583 }, { "epoch": 0.12006578947368421, "grad_norm": 96.12547302246094, "learning_rate": 8.799342105263158e-06, "loss": 1.875, "step": 584 }, { "epoch": 0.12027138157894737, "grad_norm": 42.33710861206055, "learning_rate": 8.797286184210528e-06, "loss": 1.1343, "step": 585 }, { "epoch": 0.12047697368421052, "grad_norm": 44.909976959228516, "learning_rate": 8.795230263157896e-06, "loss": 0.8984, "step": 586 }, { "epoch": 0.12068256578947369, "grad_norm": 75.26776123046875, "learning_rate": 8.793174342105264e-06, "loss": 0.9482, "step": 587 }, { "epoch": 0.12088815789473684, "grad_norm": 168.76397705078125, "learning_rate": 8.791118421052633e-06, "loss": 2.2422, "step": 588 }, { "epoch": 0.12109375, "grad_norm": 55.70532989501953, "learning_rate": 8.789062500000001e-06, "loss": 0.5195, "step": 589 }, { "epoch": 0.12129934210526316, "grad_norm": 54.19007110595703, "learning_rate": 8.787006578947369e-06, "loss": 0.8438, "step": 590 }, { "epoch": 0.12150493421052631, "grad_norm": 62.38780212402344, "learning_rate": 8.784950657894737e-06, "loss": 1.9219, "step": 591 }, { "epoch": 0.12171052631578948, "grad_norm": 14.971415519714355, "learning_rate": 8.782894736842106e-06, "loss": 0.2568, "step": 592 }, { "epoch": 0.12191611842105263, "grad_norm": 38.33452606201172, "learning_rate": 8.780838815789474e-06, "loss": 0.2642, "step": 593 }, { "epoch": 0.12212171052631579, "grad_norm": 60.936344146728516, "learning_rate": 8.778782894736842e-06, "loss": 0.9395, "step": 594 }, { "epoch": 0.12232730263157894, "grad_norm": 84.47057342529297, "learning_rate": 8.776726973684212e-06, "loss": 1.5018, "step": 595 }, { "epoch": 0.1225328947368421, "grad_norm": 55.65741729736328, "learning_rate": 8.77467105263158e-06, "loss": 1.332, "step": 596 }, { "epoch": 0.12273848684210527, "grad_norm": 56.13993453979492, "learning_rate": 8.772615131578947e-06, "loss": 1.4336, "step": 597 }, { "epoch": 0.12294407894736842, "grad_norm": 284.2030334472656, "learning_rate": 8.770559210526315e-06, "loss": 3.4707, "step": 598 }, { "epoch": 0.12314967105263158, "grad_norm": 52.934810638427734, "learning_rate": 8.768503289473685e-06, "loss": 0.668, "step": 599 }, { "epoch": 0.12335526315789473, "grad_norm": 27.108388900756836, "learning_rate": 8.766447368421054e-06, "loss": 0.1794, "step": 600 }, { "epoch": 0.1235608552631579, "grad_norm": 66.6955795288086, "learning_rate": 8.764391447368422e-06, "loss": 0.5298, "step": 601 }, { "epoch": 0.12376644736842106, "grad_norm": 78.07063293457031, "learning_rate": 8.76233552631579e-06, "loss": 0.9844, "step": 602 }, { "epoch": 0.12397203947368421, "grad_norm": 26.64858627319336, "learning_rate": 8.760279605263158e-06, "loss": 0.6559, "step": 603 }, { "epoch": 0.12417763157894737, "grad_norm": 43.06453323364258, "learning_rate": 8.758223684210528e-06, "loss": 1.1235, "step": 604 }, { "epoch": 0.12438322368421052, "grad_norm": 17.36416244506836, "learning_rate": 8.756167763157896e-06, "loss": 0.4379, "step": 605 }, { "epoch": 0.12458881578947369, "grad_norm": 55.48896026611328, "learning_rate": 8.754111842105263e-06, "loss": 0.7324, "step": 606 }, { "epoch": 0.12479440789473684, "grad_norm": 189.31414794921875, "learning_rate": 8.752055921052633e-06, "loss": 0.8223, "step": 607 }, { "epoch": 0.125, "grad_norm": 120.24112701416016, "learning_rate": 8.750000000000001e-06, "loss": 1.793, "step": 608 }, { "epoch": 0.12520559210526316, "grad_norm": 114.48238372802734, "learning_rate": 8.747944078947369e-06, "loss": 1.5381, "step": 609 }, { "epoch": 0.12541118421052633, "grad_norm": 45.41645431518555, "learning_rate": 8.745888157894737e-06, "loss": 1.1777, "step": 610 }, { "epoch": 0.12561677631578946, "grad_norm": 96.90277099609375, "learning_rate": 8.743832236842106e-06, "loss": 2.0415, "step": 611 }, { "epoch": 0.12582236842105263, "grad_norm": 124.80281829833984, "learning_rate": 8.741776315789474e-06, "loss": 2.9736, "step": 612 }, { "epoch": 0.1260279605263158, "grad_norm": 65.52892303466797, "learning_rate": 8.739720394736842e-06, "loss": 0.8369, "step": 613 }, { "epoch": 0.12623355263157895, "grad_norm": 67.1312026977539, "learning_rate": 8.737664473684212e-06, "loss": 1.1914, "step": 614 }, { "epoch": 0.12643914473684212, "grad_norm": 43.768394470214844, "learning_rate": 8.73560855263158e-06, "loss": 0.3994, "step": 615 }, { "epoch": 0.12664473684210525, "grad_norm": 61.6808967590332, "learning_rate": 8.733552631578947e-06, "loss": 2.1484, "step": 616 }, { "epoch": 0.12685032894736842, "grad_norm": 94.35441589355469, "learning_rate": 8.731496710526315e-06, "loss": 2.2148, "step": 617 }, { "epoch": 0.12705592105263158, "grad_norm": 52.543277740478516, "learning_rate": 8.729440789473685e-06, "loss": 0.875, "step": 618 }, { "epoch": 0.12726151315789475, "grad_norm": 146.04537963867188, "learning_rate": 8.727384868421054e-06, "loss": 3.8672, "step": 619 }, { "epoch": 0.12746710526315788, "grad_norm": 83.7697525024414, "learning_rate": 8.725328947368422e-06, "loss": 2.4922, "step": 620 }, { "epoch": 0.12767269736842105, "grad_norm": 46.227142333984375, "learning_rate": 8.72327302631579e-06, "loss": 1.4004, "step": 621 }, { "epoch": 0.1278782894736842, "grad_norm": 46.344058990478516, "learning_rate": 8.72121710526316e-06, "loss": 1.1797, "step": 622 }, { "epoch": 0.12808388157894737, "grad_norm": 45.84075164794922, "learning_rate": 8.719161184210528e-06, "loss": 1.6641, "step": 623 }, { "epoch": 0.12828947368421054, "grad_norm": 64.3900146484375, "learning_rate": 8.717105263157895e-06, "loss": 0.8457, "step": 624 }, { "epoch": 0.12849506578947367, "grad_norm": 32.062469482421875, "learning_rate": 8.715049342105263e-06, "loss": 0.8042, "step": 625 }, { "epoch": 0.12870065789473684, "grad_norm": 84.95153045654297, "learning_rate": 8.712993421052633e-06, "loss": 2.3135, "step": 626 }, { "epoch": 0.12890625, "grad_norm": 18.51488494873047, "learning_rate": 8.7109375e-06, "loss": 0.293, "step": 627 }, { "epoch": 0.12911184210526316, "grad_norm": 26.460599899291992, "learning_rate": 8.708881578947369e-06, "loss": 0.7979, "step": 628 }, { "epoch": 0.12931743421052633, "grad_norm": 24.95766830444336, "learning_rate": 8.706825657894738e-06, "loss": 0.2036, "step": 629 }, { "epoch": 0.12952302631578946, "grad_norm": 6.404128074645996, "learning_rate": 8.704769736842106e-06, "loss": 0.0811, "step": 630 }, { "epoch": 0.12972861842105263, "grad_norm": 43.55191421508789, "learning_rate": 8.702713815789474e-06, "loss": 1.207, "step": 631 }, { "epoch": 0.1299342105263158, "grad_norm": 33.68545913696289, "learning_rate": 8.700657894736842e-06, "loss": 0.7256, "step": 632 }, { "epoch": 0.13013980263157895, "grad_norm": 87.67304992675781, "learning_rate": 8.698601973684211e-06, "loss": 0.9199, "step": 633 }, { "epoch": 0.13034539473684212, "grad_norm": 16.775917053222656, "learning_rate": 8.69654605263158e-06, "loss": 0.3315, "step": 634 }, { "epoch": 0.13055098684210525, "grad_norm": 62.16676712036133, "learning_rate": 8.694490131578947e-06, "loss": 1.8594, "step": 635 }, { "epoch": 0.13075657894736842, "grad_norm": 29.703746795654297, "learning_rate": 8.692434210526315e-06, "loss": 0.5332, "step": 636 }, { "epoch": 0.13096217105263158, "grad_norm": 48.01938247680664, "learning_rate": 8.690378289473685e-06, "loss": 1.3516, "step": 637 }, { "epoch": 0.13116776315789475, "grad_norm": 77.99559783935547, "learning_rate": 8.688322368421054e-06, "loss": 1.4902, "step": 638 }, { "epoch": 0.13137335526315788, "grad_norm": 30.925325393676758, "learning_rate": 8.686266447368422e-06, "loss": 0.416, "step": 639 }, { "epoch": 0.13157894736842105, "grad_norm": 54.03066635131836, "learning_rate": 8.68421052631579e-06, "loss": 0.4864, "step": 640 }, { "epoch": 0.1317845394736842, "grad_norm": 46.27168655395508, "learning_rate": 8.68215460526316e-06, "loss": 1.125, "step": 641 }, { "epoch": 0.13199013157894737, "grad_norm": 11.973187446594238, "learning_rate": 8.680098684210528e-06, "loss": 0.127, "step": 642 }, { "epoch": 0.13219572368421054, "grad_norm": 68.3950424194336, "learning_rate": 8.678042763157895e-06, "loss": 1.6426, "step": 643 }, { "epoch": 0.13240131578947367, "grad_norm": 46.79332733154297, "learning_rate": 8.675986842105263e-06, "loss": 0.4893, "step": 644 }, { "epoch": 0.13260690789473684, "grad_norm": 26.141590118408203, "learning_rate": 8.673930921052633e-06, "loss": 0.2128, "step": 645 }, { "epoch": 0.1328125, "grad_norm": 28.37802505493164, "learning_rate": 8.671875e-06, "loss": 0.6501, "step": 646 }, { "epoch": 0.13301809210526316, "grad_norm": 89.68912506103516, "learning_rate": 8.669819078947369e-06, "loss": 0.7046, "step": 647 }, { "epoch": 0.13322368421052633, "grad_norm": 48.829551696777344, "learning_rate": 8.667763157894738e-06, "loss": 0.7109, "step": 648 }, { "epoch": 0.13342927631578946, "grad_norm": 63.6287727355957, "learning_rate": 8.665707236842106e-06, "loss": 1.5625, "step": 649 }, { "epoch": 0.13363486842105263, "grad_norm": 16.44005584716797, "learning_rate": 8.663651315789474e-06, "loss": 0.0898, "step": 650 }, { "epoch": 0.1338404605263158, "grad_norm": 51.655845642089844, "learning_rate": 8.661595394736842e-06, "loss": 1.124, "step": 651 }, { "epoch": 0.13404605263157895, "grad_norm": 118.97359466552734, "learning_rate": 8.659539473684211e-06, "loss": 1.918, "step": 652 }, { "epoch": 0.13425164473684212, "grad_norm": 40.05059814453125, "learning_rate": 8.65748355263158e-06, "loss": 0.8218, "step": 653 }, { "epoch": 0.13445723684210525, "grad_norm": 184.9432830810547, "learning_rate": 8.655427631578947e-06, "loss": 1.1129, "step": 654 }, { "epoch": 0.13466282894736842, "grad_norm": 80.51482391357422, "learning_rate": 8.653371710526317e-06, "loss": 2.6094, "step": 655 }, { "epoch": 0.13486842105263158, "grad_norm": 64.6722183227539, "learning_rate": 8.651315789473685e-06, "loss": 1.4238, "step": 656 }, { "epoch": 0.13507401315789475, "grad_norm": 40.85758972167969, "learning_rate": 8.649259868421054e-06, "loss": 0.8198, "step": 657 }, { "epoch": 0.13527960526315788, "grad_norm": 63.886749267578125, "learning_rate": 8.647203947368422e-06, "loss": 1.0645, "step": 658 }, { "epoch": 0.13548519736842105, "grad_norm": 35.2276611328125, "learning_rate": 8.64514802631579e-06, "loss": 0.3193, "step": 659 }, { "epoch": 0.1356907894736842, "grad_norm": 50.52458953857422, "learning_rate": 8.64309210526316e-06, "loss": 1.1855, "step": 660 }, { "epoch": 0.13589638157894737, "grad_norm": 54.2060432434082, "learning_rate": 8.641036184210527e-06, "loss": 0.5552, "step": 661 }, { "epoch": 0.13610197368421054, "grad_norm": 61.133514404296875, "learning_rate": 8.638980263157895e-06, "loss": 0.8965, "step": 662 }, { "epoch": 0.13630756578947367, "grad_norm": 118.91746520996094, "learning_rate": 8.636924342105263e-06, "loss": 1.6973, "step": 663 }, { "epoch": 0.13651315789473684, "grad_norm": 118.81941223144531, "learning_rate": 8.634868421052633e-06, "loss": 1.4705, "step": 664 }, { "epoch": 0.13671875, "grad_norm": 68.5114517211914, "learning_rate": 8.6328125e-06, "loss": 1.1084, "step": 665 }, { "epoch": 0.13692434210526316, "grad_norm": 97.8519287109375, "learning_rate": 8.630756578947369e-06, "loss": 1.1406, "step": 666 }, { "epoch": 0.13712993421052633, "grad_norm": 41.25439453125, "learning_rate": 8.628700657894738e-06, "loss": 0.3616, "step": 667 }, { "epoch": 0.13733552631578946, "grad_norm": 145.345947265625, "learning_rate": 8.626644736842106e-06, "loss": 4.0703, "step": 668 }, { "epoch": 0.13754111842105263, "grad_norm": 63.660888671875, "learning_rate": 8.624588815789474e-06, "loss": 0.5547, "step": 669 }, { "epoch": 0.1377467105263158, "grad_norm": 42.98672866821289, "learning_rate": 8.622532894736842e-06, "loss": 0.3481, "step": 670 }, { "epoch": 0.13795230263157895, "grad_norm": 11.528818130493164, "learning_rate": 8.620476973684211e-06, "loss": 0.1072, "step": 671 }, { "epoch": 0.13815789473684212, "grad_norm": 52.6195182800293, "learning_rate": 8.61842105263158e-06, "loss": 1.3226, "step": 672 }, { "epoch": 0.13836348684210525, "grad_norm": 14.427908897399902, "learning_rate": 8.616365131578947e-06, "loss": 0.0542, "step": 673 }, { "epoch": 0.13856907894736842, "grad_norm": 44.66035461425781, "learning_rate": 8.614309210526317e-06, "loss": 0.853, "step": 674 }, { "epoch": 0.13877467105263158, "grad_norm": 68.77999877929688, "learning_rate": 8.612253289473686e-06, "loss": 1.3213, "step": 675 }, { "epoch": 0.13898026315789475, "grad_norm": 112.95706939697266, "learning_rate": 8.610197368421054e-06, "loss": 1.5469, "step": 676 }, { "epoch": 0.13918585526315788, "grad_norm": 93.5947265625, "learning_rate": 8.608141447368422e-06, "loss": 1.2578, "step": 677 }, { "epoch": 0.13939144736842105, "grad_norm": 35.734012603759766, "learning_rate": 8.60608552631579e-06, "loss": 0.9463, "step": 678 }, { "epoch": 0.1395970394736842, "grad_norm": 20.29075050354004, "learning_rate": 8.60402960526316e-06, "loss": 0.5652, "step": 679 }, { "epoch": 0.13980263157894737, "grad_norm": 58.36338424682617, "learning_rate": 8.601973684210527e-06, "loss": 1.0527, "step": 680 }, { "epoch": 0.14000822368421054, "grad_norm": 61.02059555053711, "learning_rate": 8.599917763157895e-06, "loss": 0.9668, "step": 681 }, { "epoch": 0.14021381578947367, "grad_norm": 43.32736587524414, "learning_rate": 8.597861842105265e-06, "loss": 0.7578, "step": 682 }, { "epoch": 0.14041940789473684, "grad_norm": 117.1735610961914, "learning_rate": 8.595805921052633e-06, "loss": 1.7148, "step": 683 }, { "epoch": 0.140625, "grad_norm": 32.301666259765625, "learning_rate": 8.59375e-06, "loss": 0.9375, "step": 684 }, { "epoch": 0.14083059210526316, "grad_norm": 34.357521057128906, "learning_rate": 8.591694078947368e-06, "loss": 0.2314, "step": 685 }, { "epoch": 0.14103618421052633, "grad_norm": 37.894615173339844, "learning_rate": 8.589638157894738e-06, "loss": 0.3457, "step": 686 }, { "epoch": 0.14124177631578946, "grad_norm": 56.571983337402344, "learning_rate": 8.587582236842106e-06, "loss": 0.5234, "step": 687 }, { "epoch": 0.14144736842105263, "grad_norm": 69.14144134521484, "learning_rate": 8.585526315789474e-06, "loss": 1.2891, "step": 688 }, { "epoch": 0.1416529605263158, "grad_norm": 69.99906921386719, "learning_rate": 8.583470394736843e-06, "loss": 1.5834, "step": 689 }, { "epoch": 0.14185855263157895, "grad_norm": 23.907567977905273, "learning_rate": 8.581414473684211e-06, "loss": 0.7519, "step": 690 }, { "epoch": 0.14206414473684212, "grad_norm": 28.495101928710938, "learning_rate": 8.579358552631579e-06, "loss": 0.2227, "step": 691 }, { "epoch": 0.14226973684210525, "grad_norm": 94.09139251708984, "learning_rate": 8.577302631578947e-06, "loss": 1.9951, "step": 692 }, { "epoch": 0.14247532894736842, "grad_norm": 34.167503356933594, "learning_rate": 8.575246710526317e-06, "loss": 0.2659, "step": 693 }, { "epoch": 0.14268092105263158, "grad_norm": 8.988431930541992, "learning_rate": 8.573190789473686e-06, "loss": 0.0779, "step": 694 }, { "epoch": 0.14288651315789475, "grad_norm": 44.86157989501953, "learning_rate": 8.571134868421054e-06, "loss": 1.3359, "step": 695 }, { "epoch": 0.14309210526315788, "grad_norm": 111.48028564453125, "learning_rate": 8.569078947368422e-06, "loss": 2.0742, "step": 696 }, { "epoch": 0.14329769736842105, "grad_norm": 78.90316772460938, "learning_rate": 8.56702302631579e-06, "loss": 1.3281, "step": 697 }, { "epoch": 0.1435032894736842, "grad_norm": 47.11545944213867, "learning_rate": 8.56496710526316e-06, "loss": 0.5965, "step": 698 }, { "epoch": 0.14370888157894737, "grad_norm": 65.02469635009766, "learning_rate": 8.562911184210527e-06, "loss": 1.4043, "step": 699 }, { "epoch": 0.14391447368421054, "grad_norm": 125.16715240478516, "learning_rate": 8.560855263157895e-06, "loss": 1.1948, "step": 700 }, { "epoch": 0.14412006578947367, "grad_norm": 234.9722442626953, "learning_rate": 8.558799342105265e-06, "loss": 1.202, "step": 701 }, { "epoch": 0.14432565789473684, "grad_norm": 188.00404357910156, "learning_rate": 8.556743421052633e-06, "loss": 4.8086, "step": 702 }, { "epoch": 0.14453125, "grad_norm": 57.805328369140625, "learning_rate": 8.5546875e-06, "loss": 0.4912, "step": 703 }, { "epoch": 0.14473684210526316, "grad_norm": 17.65337562561035, "learning_rate": 8.552631578947368e-06, "loss": 0.3257, "step": 704 }, { "epoch": 0.14494243421052633, "grad_norm": 17.71002197265625, "learning_rate": 8.550575657894738e-06, "loss": 0.1665, "step": 705 }, { "epoch": 0.14514802631578946, "grad_norm": 63.479217529296875, "learning_rate": 8.548519736842106e-06, "loss": 0.8652, "step": 706 }, { "epoch": 0.14535361842105263, "grad_norm": 69.96138000488281, "learning_rate": 8.546463815789474e-06, "loss": 1.7773, "step": 707 }, { "epoch": 0.1455592105263158, "grad_norm": 36.51927947998047, "learning_rate": 8.544407894736843e-06, "loss": 0.417, "step": 708 }, { "epoch": 0.14576480263157895, "grad_norm": 73.3127670288086, "learning_rate": 8.542351973684211e-06, "loss": 1.1934, "step": 709 }, { "epoch": 0.14597039473684212, "grad_norm": 17.67814064025879, "learning_rate": 8.540296052631579e-06, "loss": 0.2007, "step": 710 }, { "epoch": 0.14617598684210525, "grad_norm": 93.25010681152344, "learning_rate": 8.538240131578947e-06, "loss": 1.2227, "step": 711 }, { "epoch": 0.14638157894736842, "grad_norm": 47.098793029785156, "learning_rate": 8.536184210526316e-06, "loss": 0.7549, "step": 712 }, { "epoch": 0.14658717105263158, "grad_norm": 28.055419921875, "learning_rate": 8.534128289473686e-06, "loss": 0.3501, "step": 713 }, { "epoch": 0.14679276315789475, "grad_norm": 147.5088653564453, "learning_rate": 8.532072368421054e-06, "loss": 2.5859, "step": 714 }, { "epoch": 0.14699835526315788, "grad_norm": 30.32993507385254, "learning_rate": 8.530016447368422e-06, "loss": 0.3438, "step": 715 }, { "epoch": 0.14720394736842105, "grad_norm": 85.01473236083984, "learning_rate": 8.527960526315791e-06, "loss": 0.7856, "step": 716 }, { "epoch": 0.1474095394736842, "grad_norm": 148.74844360351562, "learning_rate": 8.52590460526316e-06, "loss": 2.5762, "step": 717 }, { "epoch": 0.14761513157894737, "grad_norm": 24.058984756469727, "learning_rate": 8.523848684210527e-06, "loss": 0.866, "step": 718 }, { "epoch": 0.14782072368421054, "grad_norm": 189.75238037109375, "learning_rate": 8.521792763157895e-06, "loss": 3.1016, "step": 719 }, { "epoch": 0.14802631578947367, "grad_norm": 54.72446060180664, "learning_rate": 8.519736842105265e-06, "loss": 1.5684, "step": 720 }, { "epoch": 0.14823190789473684, "grad_norm": 5.390693664550781, "learning_rate": 8.517680921052632e-06, "loss": 0.0195, "step": 721 }, { "epoch": 0.1484375, "grad_norm": 73.91433715820312, "learning_rate": 8.515625e-06, "loss": 1.7812, "step": 722 }, { "epoch": 0.14864309210526316, "grad_norm": 112.03641510009766, "learning_rate": 8.51356907894737e-06, "loss": 4.0195, "step": 723 }, { "epoch": 0.14884868421052633, "grad_norm": 102.6246566772461, "learning_rate": 8.511513157894738e-06, "loss": 2.1274, "step": 724 }, { "epoch": 0.14905427631578946, "grad_norm": 98.91804504394531, "learning_rate": 8.509457236842106e-06, "loss": 2.0234, "step": 725 }, { "epoch": 0.14925986842105263, "grad_norm": 62.544925689697266, "learning_rate": 8.507401315789474e-06, "loss": 1.1094, "step": 726 }, { "epoch": 0.1494654605263158, "grad_norm": 54.08305740356445, "learning_rate": 8.505345394736843e-06, "loss": 0.4775, "step": 727 }, { "epoch": 0.14967105263157895, "grad_norm": 43.37862014770508, "learning_rate": 8.503289473684211e-06, "loss": 1.0181, "step": 728 }, { "epoch": 0.14987664473684212, "grad_norm": 26.713764190673828, "learning_rate": 8.501233552631579e-06, "loss": 0.9658, "step": 729 }, { "epoch": 0.15008223684210525, "grad_norm": 113.76016235351562, "learning_rate": 8.499177631578948e-06, "loss": 2.0, "step": 730 }, { "epoch": 0.15028782894736842, "grad_norm": 39.05321502685547, "learning_rate": 8.497121710526316e-06, "loss": 1.332, "step": 731 }, { "epoch": 0.15049342105263158, "grad_norm": 31.36034393310547, "learning_rate": 8.495065789473686e-06, "loss": 0.8984, "step": 732 }, { "epoch": 0.15069901315789475, "grad_norm": 48.43738555908203, "learning_rate": 8.493009868421054e-06, "loss": 0.7139, "step": 733 }, { "epoch": 0.15090460526315788, "grad_norm": 47.14624786376953, "learning_rate": 8.490953947368422e-06, "loss": 0.8633, "step": 734 }, { "epoch": 0.15111019736842105, "grad_norm": 26.767778396606445, "learning_rate": 8.488898026315791e-06, "loss": 0.9062, "step": 735 }, { "epoch": 0.1513157894736842, "grad_norm": 16.452152252197266, "learning_rate": 8.486842105263159e-06, "loss": 0.0793, "step": 736 }, { "epoch": 0.15152138157894737, "grad_norm": 26.655061721801758, "learning_rate": 8.484786184210527e-06, "loss": 0.9209, "step": 737 }, { "epoch": 0.15172697368421054, "grad_norm": 47.059226989746094, "learning_rate": 8.482730263157895e-06, "loss": 0.5972, "step": 738 }, { "epoch": 0.15193256578947367, "grad_norm": 93.81578063964844, "learning_rate": 8.480674342105264e-06, "loss": 3.1289, "step": 739 }, { "epoch": 0.15213815789473684, "grad_norm": 36.891265869140625, "learning_rate": 8.478618421052632e-06, "loss": 0.8848, "step": 740 }, { "epoch": 0.15234375, "grad_norm": 48.324066162109375, "learning_rate": 8.4765625e-06, "loss": 1.0083, "step": 741 }, { "epoch": 0.15254934210526316, "grad_norm": 46.1828498840332, "learning_rate": 8.47450657894737e-06, "loss": 1.5195, "step": 742 }, { "epoch": 0.15275493421052633, "grad_norm": 108.768798828125, "learning_rate": 8.472450657894738e-06, "loss": 4.1914, "step": 743 }, { "epoch": 0.15296052631578946, "grad_norm": 111.224609375, "learning_rate": 8.470394736842106e-06, "loss": 2.7852, "step": 744 }, { "epoch": 0.15316611842105263, "grad_norm": 96.24161529541016, "learning_rate": 8.468338815789473e-06, "loss": 2.293, "step": 745 }, { "epoch": 0.1533717105263158, "grad_norm": 56.32974624633789, "learning_rate": 8.466282894736843e-06, "loss": 2.0742, "step": 746 }, { "epoch": 0.15357730263157895, "grad_norm": 32.66743850708008, "learning_rate": 8.464226973684211e-06, "loss": 0.8521, "step": 747 }, { "epoch": 0.15378289473684212, "grad_norm": 32.8243522644043, "learning_rate": 8.462171052631579e-06, "loss": 0.7788, "step": 748 }, { "epoch": 0.15398848684210525, "grad_norm": 77.74825286865234, "learning_rate": 8.460115131578948e-06, "loss": 1.6016, "step": 749 }, { "epoch": 0.15419407894736842, "grad_norm": 31.36269187927246, "learning_rate": 8.458059210526316e-06, "loss": 0.9375, "step": 750 }, { "epoch": 0.15439967105263158, "grad_norm": 31.363245010375977, "learning_rate": 8.456003289473686e-06, "loss": 0.6338, "step": 751 }, { "epoch": 0.15460526315789475, "grad_norm": 91.40959167480469, "learning_rate": 8.453947368421054e-06, "loss": 1.3477, "step": 752 }, { "epoch": 0.15481085526315788, "grad_norm": 52.79136657714844, "learning_rate": 8.451891447368422e-06, "loss": 1.1543, "step": 753 }, { "epoch": 0.15501644736842105, "grad_norm": 119.48922729492188, "learning_rate": 8.449835526315791e-06, "loss": 1.9727, "step": 754 }, { "epoch": 0.1552220394736842, "grad_norm": 34.76116180419922, "learning_rate": 8.447779605263159e-06, "loss": 0.6621, "step": 755 }, { "epoch": 0.15542763157894737, "grad_norm": 68.73171997070312, "learning_rate": 8.445723684210527e-06, "loss": 1.5234, "step": 756 }, { "epoch": 0.15563322368421054, "grad_norm": 43.29606628417969, "learning_rate": 8.443667763157895e-06, "loss": 1.5508, "step": 757 }, { "epoch": 0.15583881578947367, "grad_norm": 46.02786636352539, "learning_rate": 8.441611842105264e-06, "loss": 1.4434, "step": 758 }, { "epoch": 0.15604440789473684, "grad_norm": 59.896602630615234, "learning_rate": 8.439555921052632e-06, "loss": 0.7656, "step": 759 }, { "epoch": 0.15625, "grad_norm": 50.118228912353516, "learning_rate": 8.4375e-06, "loss": 1.5898, "step": 760 }, { "epoch": 0.15645559210526316, "grad_norm": 47.21403121948242, "learning_rate": 8.43544407894737e-06, "loss": 1.3887, "step": 761 }, { "epoch": 0.15666118421052633, "grad_norm": 33.9365119934082, "learning_rate": 8.433388157894738e-06, "loss": 0.7063, "step": 762 }, { "epoch": 0.15686677631578946, "grad_norm": 40.72140884399414, "learning_rate": 8.431332236842105e-06, "loss": 0.9429, "step": 763 }, { "epoch": 0.15707236842105263, "grad_norm": 45.16706085205078, "learning_rate": 8.429276315789473e-06, "loss": 1.6309, "step": 764 }, { "epoch": 0.1572779605263158, "grad_norm": 72.04479217529297, "learning_rate": 8.427220394736843e-06, "loss": 1.8262, "step": 765 }, { "epoch": 0.15748355263157895, "grad_norm": 29.45753288269043, "learning_rate": 8.42516447368421e-06, "loss": 0.4854, "step": 766 }, { "epoch": 0.15768914473684212, "grad_norm": 22.632686614990234, "learning_rate": 8.423108552631579e-06, "loss": 0.3516, "step": 767 }, { "epoch": 0.15789473684210525, "grad_norm": 30.07136344909668, "learning_rate": 8.421052631578948e-06, "loss": 1.0918, "step": 768 }, { "epoch": 0.15810032894736842, "grad_norm": 27.62939453125, "learning_rate": 8.418996710526316e-06, "loss": 0.8623, "step": 769 }, { "epoch": 0.15830592105263158, "grad_norm": 70.21210479736328, "learning_rate": 8.416940789473686e-06, "loss": 1.7773, "step": 770 }, { "epoch": 0.15851151315789475, "grad_norm": 57.31368637084961, "learning_rate": 8.414884868421054e-06, "loss": 1.3477, "step": 771 }, { "epoch": 0.15871710526315788, "grad_norm": 44.01366424560547, "learning_rate": 8.412828947368421e-06, "loss": 0.6689, "step": 772 }, { "epoch": 0.15892269736842105, "grad_norm": 65.52149963378906, "learning_rate": 8.410773026315791e-06, "loss": 0.7314, "step": 773 }, { "epoch": 0.1591282894736842, "grad_norm": 26.66001319885254, "learning_rate": 8.408717105263159e-06, "loss": 0.3623, "step": 774 }, { "epoch": 0.15933388157894737, "grad_norm": 58.062747955322266, "learning_rate": 8.406661184210527e-06, "loss": 1.2773, "step": 775 }, { "epoch": 0.15953947368421054, "grad_norm": 48.5870361328125, "learning_rate": 8.404605263157896e-06, "loss": 0.9023, "step": 776 }, { "epoch": 0.15974506578947367, "grad_norm": 39.26237487792969, "learning_rate": 8.402549342105264e-06, "loss": 0.8398, "step": 777 }, { "epoch": 0.15995065789473684, "grad_norm": 25.91082191467285, "learning_rate": 8.400493421052632e-06, "loss": 0.264, "step": 778 }, { "epoch": 0.16015625, "grad_norm": 40.69412612915039, "learning_rate": 8.3984375e-06, "loss": 1.1953, "step": 779 }, { "epoch": 0.16036184210526316, "grad_norm": 58.23236846923828, "learning_rate": 8.39638157894737e-06, "loss": 1.1602, "step": 780 }, { "epoch": 0.16056743421052633, "grad_norm": 25.991674423217773, "learning_rate": 8.394325657894737e-06, "loss": 0.5002, "step": 781 }, { "epoch": 0.16077302631578946, "grad_norm": 21.386781692504883, "learning_rate": 8.392269736842105e-06, "loss": 0.1313, "step": 782 }, { "epoch": 0.16097861842105263, "grad_norm": 42.25035095214844, "learning_rate": 8.390213815789475e-06, "loss": 0.9863, "step": 783 }, { "epoch": 0.1611842105263158, "grad_norm": 43.930721282958984, "learning_rate": 8.388157894736843e-06, "loss": 0.6719, "step": 784 }, { "epoch": 0.16138980263157895, "grad_norm": 49.7684440612793, "learning_rate": 8.38610197368421e-06, "loss": 0.5375, "step": 785 }, { "epoch": 0.16159539473684212, "grad_norm": 95.18107604980469, "learning_rate": 8.384046052631579e-06, "loss": 2.0, "step": 786 }, { "epoch": 0.16180098684210525, "grad_norm": 101.97228240966797, "learning_rate": 8.381990131578948e-06, "loss": 1.7109, "step": 787 }, { "epoch": 0.16200657894736842, "grad_norm": 24.941986083984375, "learning_rate": 8.379934210526316e-06, "loss": 0.1838, "step": 788 }, { "epoch": 0.16221217105263158, "grad_norm": 62.146080017089844, "learning_rate": 8.377878289473686e-06, "loss": 1.6724, "step": 789 }, { "epoch": 0.16241776315789475, "grad_norm": 41.657981872558594, "learning_rate": 8.375822368421053e-06, "loss": 0.8779, "step": 790 }, { "epoch": 0.16262335526315788, "grad_norm": 53.92509078979492, "learning_rate": 8.373766447368421e-06, "loss": 0.9043, "step": 791 }, { "epoch": 0.16282894736842105, "grad_norm": 44.05774688720703, "learning_rate": 8.371710526315791e-06, "loss": 0.8276, "step": 792 }, { "epoch": 0.1630345394736842, "grad_norm": 92.7431640625, "learning_rate": 8.369654605263159e-06, "loss": 0.8672, "step": 793 }, { "epoch": 0.16324013157894737, "grad_norm": 65.08289337158203, "learning_rate": 8.367598684210527e-06, "loss": 1.4961, "step": 794 }, { "epoch": 0.16344572368421054, "grad_norm": 52.59501647949219, "learning_rate": 8.365542763157896e-06, "loss": 1.2812, "step": 795 }, { "epoch": 0.16365131578947367, "grad_norm": 55.36376190185547, "learning_rate": 8.363486842105264e-06, "loss": 1.1953, "step": 796 }, { "epoch": 0.16385690789473684, "grad_norm": 14.310023307800293, "learning_rate": 8.361430921052632e-06, "loss": 0.0894, "step": 797 }, { "epoch": 0.1640625, "grad_norm": 55.169273376464844, "learning_rate": 8.359375e-06, "loss": 1.2734, "step": 798 }, { "epoch": 0.16426809210526316, "grad_norm": 16.220945358276367, "learning_rate": 8.35731907894737e-06, "loss": 0.0417, "step": 799 }, { "epoch": 0.16447368421052633, "grad_norm": 50.0037841796875, "learning_rate": 8.355263157894737e-06, "loss": 0.6593, "step": 800 }, { "epoch": 0.16467927631578946, "grad_norm": 51.58594512939453, "learning_rate": 8.353207236842105e-06, "loss": 1.3516, "step": 801 }, { "epoch": 0.16488486842105263, "grad_norm": 132.88796997070312, "learning_rate": 8.351151315789475e-06, "loss": 2.2148, "step": 802 }, { "epoch": 0.1650904605263158, "grad_norm": 12.37620735168457, "learning_rate": 8.349095394736843e-06, "loss": 0.0721, "step": 803 }, { "epoch": 0.16529605263157895, "grad_norm": 45.504676818847656, "learning_rate": 8.34703947368421e-06, "loss": 0.6914, "step": 804 }, { "epoch": 0.16550164473684212, "grad_norm": 89.21688842773438, "learning_rate": 8.344983552631578e-06, "loss": 0.8594, "step": 805 }, { "epoch": 0.16570723684210525, "grad_norm": 127.39305114746094, "learning_rate": 8.342927631578948e-06, "loss": 2.873, "step": 806 }, { "epoch": 0.16591282894736842, "grad_norm": 38.8594856262207, "learning_rate": 8.340871710526316e-06, "loss": 0.5596, "step": 807 }, { "epoch": 0.16611842105263158, "grad_norm": 149.3431854248047, "learning_rate": 8.338815789473685e-06, "loss": 3.9004, "step": 808 }, { "epoch": 0.16632401315789475, "grad_norm": 91.25466918945312, "learning_rate": 8.336759868421053e-06, "loss": 1.9258, "step": 809 }, { "epoch": 0.16652960526315788, "grad_norm": 45.718666076660156, "learning_rate": 8.334703947368423e-06, "loss": 0.8047, "step": 810 }, { "epoch": 0.16673519736842105, "grad_norm": 13.24303150177002, "learning_rate": 8.33264802631579e-06, "loss": 0.129, "step": 811 }, { "epoch": 0.1669407894736842, "grad_norm": 63.98895263671875, "learning_rate": 8.330592105263159e-06, "loss": 1.0742, "step": 812 }, { "epoch": 0.16714638157894737, "grad_norm": 40.018070220947266, "learning_rate": 8.328536184210527e-06, "loss": 0.6943, "step": 813 }, { "epoch": 0.16735197368421054, "grad_norm": 42.91875076293945, "learning_rate": 8.326480263157896e-06, "loss": 0.5542, "step": 814 }, { "epoch": 0.16755756578947367, "grad_norm": 132.8564453125, "learning_rate": 8.324424342105264e-06, "loss": 2.3281, "step": 815 }, { "epoch": 0.16776315789473684, "grad_norm": 42.07183837890625, "learning_rate": 8.322368421052632e-06, "loss": 0.3662, "step": 816 }, { "epoch": 0.16796875, "grad_norm": 71.09544372558594, "learning_rate": 8.320312500000001e-06, "loss": 1.4785, "step": 817 }, { "epoch": 0.16817434210526316, "grad_norm": 65.64195251464844, "learning_rate": 8.31825657894737e-06, "loss": 1.0322, "step": 818 }, { "epoch": 0.16837993421052633, "grad_norm": 49.48176574707031, "learning_rate": 8.316200657894737e-06, "loss": 1.0352, "step": 819 }, { "epoch": 0.16858552631578946, "grad_norm": 123.81171417236328, "learning_rate": 8.314144736842105e-06, "loss": 2.7148, "step": 820 }, { "epoch": 0.16879111842105263, "grad_norm": 35.62040328979492, "learning_rate": 8.312088815789475e-06, "loss": 0.2793, "step": 821 }, { "epoch": 0.1689967105263158, "grad_norm": 155.6439208984375, "learning_rate": 8.310032894736843e-06, "loss": 4.5625, "step": 822 }, { "epoch": 0.16920230263157895, "grad_norm": 42.95448684692383, "learning_rate": 8.30797697368421e-06, "loss": 0.7607, "step": 823 }, { "epoch": 0.16940789473684212, "grad_norm": 12.156123161315918, "learning_rate": 8.30592105263158e-06, "loss": 0.0815, "step": 824 }, { "epoch": 0.16961348684210525, "grad_norm": 66.42166900634766, "learning_rate": 8.303865131578948e-06, "loss": 1.168, "step": 825 }, { "epoch": 0.16981907894736842, "grad_norm": 107.2586898803711, "learning_rate": 8.301809210526316e-06, "loss": 1.8555, "step": 826 }, { "epoch": 0.17002467105263158, "grad_norm": 81.77083587646484, "learning_rate": 8.299753289473685e-06, "loss": 1.6641, "step": 827 }, { "epoch": 0.17023026315789475, "grad_norm": 62.40121078491211, "learning_rate": 8.297697368421053e-06, "loss": 0.9258, "step": 828 }, { "epoch": 0.17043585526315788, "grad_norm": 137.39935302734375, "learning_rate": 8.295641447368423e-06, "loss": 4.1562, "step": 829 }, { "epoch": 0.17064144736842105, "grad_norm": 59.91421127319336, "learning_rate": 8.29358552631579e-06, "loss": 0.978, "step": 830 }, { "epoch": 0.1708470394736842, "grad_norm": 44.41098403930664, "learning_rate": 8.291529605263159e-06, "loss": 0.9502, "step": 831 }, { "epoch": 0.17105263157894737, "grad_norm": 8.190799713134766, "learning_rate": 8.289473684210526e-06, "loss": 0.0553, "step": 832 }, { "epoch": 0.17125822368421054, "grad_norm": 78.97562408447266, "learning_rate": 8.287417763157896e-06, "loss": 1.2754, "step": 833 }, { "epoch": 0.17146381578947367, "grad_norm": 71.33646392822266, "learning_rate": 8.285361842105264e-06, "loss": 1.0488, "step": 834 }, { "epoch": 0.17166940789473684, "grad_norm": 55.12741470336914, "learning_rate": 8.283305921052632e-06, "loss": 0.8789, "step": 835 }, { "epoch": 0.171875, "grad_norm": 36.55903244018555, "learning_rate": 8.281250000000001e-06, "loss": 0.9355, "step": 836 }, { "epoch": 0.17208059210526316, "grad_norm": 28.95254898071289, "learning_rate": 8.27919407894737e-06, "loss": 0.1787, "step": 837 }, { "epoch": 0.17228618421052633, "grad_norm": 30.96012306213379, "learning_rate": 8.277138157894737e-06, "loss": 0.3838, "step": 838 }, { "epoch": 0.17249177631578946, "grad_norm": 43.353660583496094, "learning_rate": 8.275082236842105e-06, "loss": 0.9937, "step": 839 }, { "epoch": 0.17269736842105263, "grad_norm": 34.2910041809082, "learning_rate": 8.273026315789475e-06, "loss": 0.5039, "step": 840 }, { "epoch": 0.1729029605263158, "grad_norm": 39.16138458251953, "learning_rate": 8.270970394736842e-06, "loss": 0.8574, "step": 841 }, { "epoch": 0.17310855263157895, "grad_norm": 39.40522766113281, "learning_rate": 8.26891447368421e-06, "loss": 1.0059, "step": 842 }, { "epoch": 0.17331414473684212, "grad_norm": 87.07376098632812, "learning_rate": 8.26685855263158e-06, "loss": 1.7539, "step": 843 }, { "epoch": 0.17351973684210525, "grad_norm": 43.97018814086914, "learning_rate": 8.264802631578948e-06, "loss": 1.125, "step": 844 }, { "epoch": 0.17372532894736842, "grad_norm": 84.031494140625, "learning_rate": 8.262746710526316e-06, "loss": 1.2324, "step": 845 }, { "epoch": 0.17393092105263158, "grad_norm": 111.75521087646484, "learning_rate": 8.260690789473685e-06, "loss": 1.5, "step": 846 }, { "epoch": 0.17413651315789475, "grad_norm": 39.29418182373047, "learning_rate": 8.258634868421053e-06, "loss": 0.6724, "step": 847 }, { "epoch": 0.17434210526315788, "grad_norm": 94.5953140258789, "learning_rate": 8.256578947368423e-06, "loss": 2.0977, "step": 848 }, { "epoch": 0.17454769736842105, "grad_norm": 36.831321716308594, "learning_rate": 8.25452302631579e-06, "loss": 0.3071, "step": 849 }, { "epoch": 0.1747532894736842, "grad_norm": 55.60489273071289, "learning_rate": 8.252467105263158e-06, "loss": 0.9414, "step": 850 }, { "epoch": 0.17495888157894737, "grad_norm": 61.15327835083008, "learning_rate": 8.250411184210526e-06, "loss": 0.8867, "step": 851 }, { "epoch": 0.17516447368421054, "grad_norm": 76.65862274169922, "learning_rate": 8.248355263157896e-06, "loss": 1.4414, "step": 852 }, { "epoch": 0.17537006578947367, "grad_norm": 27.289533615112305, "learning_rate": 8.246299342105264e-06, "loss": 0.4521, "step": 853 }, { "epoch": 0.17557565789473684, "grad_norm": 34.95981216430664, "learning_rate": 8.244243421052632e-06, "loss": 0.708, "step": 854 }, { "epoch": 0.17578125, "grad_norm": 23.47173500061035, "learning_rate": 8.242187500000001e-06, "loss": 0.2603, "step": 855 }, { "epoch": 0.17598684210526316, "grad_norm": 45.35258865356445, "learning_rate": 8.24013157894737e-06, "loss": 0.9253, "step": 856 }, { "epoch": 0.17619243421052633, "grad_norm": 30.798206329345703, "learning_rate": 8.238075657894737e-06, "loss": 0.8408, "step": 857 }, { "epoch": 0.17639802631578946, "grad_norm": 79.47782135009766, "learning_rate": 8.236019736842105e-06, "loss": 2.1992, "step": 858 }, { "epoch": 0.17660361842105263, "grad_norm": 40.13658142089844, "learning_rate": 8.233963815789475e-06, "loss": 0.6973, "step": 859 }, { "epoch": 0.1768092105263158, "grad_norm": 70.75611877441406, "learning_rate": 8.231907894736842e-06, "loss": 1.7266, "step": 860 }, { "epoch": 0.17701480263157895, "grad_norm": 47.003387451171875, "learning_rate": 8.22985197368421e-06, "loss": 0.5254, "step": 861 }, { "epoch": 0.17722039473684212, "grad_norm": 41.44414520263672, "learning_rate": 8.22779605263158e-06, "loss": 1.0793, "step": 862 }, { "epoch": 0.17742598684210525, "grad_norm": 47.15561294555664, "learning_rate": 8.225740131578948e-06, "loss": 0.7764, "step": 863 }, { "epoch": 0.17763157894736842, "grad_norm": 46.422882080078125, "learning_rate": 8.223684210526316e-06, "loss": 0.9028, "step": 864 }, { "epoch": 0.17783717105263158, "grad_norm": 54.42194366455078, "learning_rate": 8.221628289473685e-06, "loss": 1.5117, "step": 865 }, { "epoch": 0.17804276315789475, "grad_norm": 25.571990966796875, "learning_rate": 8.219572368421053e-06, "loss": 0.5361, "step": 866 }, { "epoch": 0.17824835526315788, "grad_norm": 16.561800003051758, "learning_rate": 8.217516447368423e-06, "loss": 0.1489, "step": 867 }, { "epoch": 0.17845394736842105, "grad_norm": 46.0649299621582, "learning_rate": 8.21546052631579e-06, "loss": 0.8242, "step": 868 }, { "epoch": 0.1786595394736842, "grad_norm": 58.212669372558594, "learning_rate": 8.213404605263158e-06, "loss": 0.6738, "step": 869 }, { "epoch": 0.17886513157894737, "grad_norm": 61.01872634887695, "learning_rate": 8.211348684210528e-06, "loss": 0.8691, "step": 870 }, { "epoch": 0.17907072368421054, "grad_norm": 115.90441131591797, "learning_rate": 8.209292763157896e-06, "loss": 3.7578, "step": 871 }, { "epoch": 0.17927631578947367, "grad_norm": 228.1526641845703, "learning_rate": 8.207236842105264e-06, "loss": 2.2461, "step": 872 }, { "epoch": 0.17948190789473684, "grad_norm": 13.343188285827637, "learning_rate": 8.205180921052632e-06, "loss": 0.1885, "step": 873 }, { "epoch": 0.1796875, "grad_norm": 44.549922943115234, "learning_rate": 8.203125000000001e-06, "loss": 0.4927, "step": 874 }, { "epoch": 0.17989309210526316, "grad_norm": 30.92880630493164, "learning_rate": 8.201069078947369e-06, "loss": 1.0312, "step": 875 }, { "epoch": 0.18009868421052633, "grad_norm": 53.3516845703125, "learning_rate": 8.199013157894737e-06, "loss": 1.1309, "step": 876 }, { "epoch": 0.18030427631578946, "grad_norm": 26.752775192260742, "learning_rate": 8.196957236842107e-06, "loss": 0.2852, "step": 877 }, { "epoch": 0.18050986842105263, "grad_norm": 64.98573303222656, "learning_rate": 8.194901315789474e-06, "loss": 1.4297, "step": 878 }, { "epoch": 0.1807154605263158, "grad_norm": 58.76778030395508, "learning_rate": 8.192845394736842e-06, "loss": 0.7368, "step": 879 }, { "epoch": 0.18092105263157895, "grad_norm": 155.8352813720703, "learning_rate": 8.19078947368421e-06, "loss": 3.6797, "step": 880 }, { "epoch": 0.18112664473684212, "grad_norm": 31.67601203918457, "learning_rate": 8.18873355263158e-06, "loss": 0.7402, "step": 881 }, { "epoch": 0.18133223684210525, "grad_norm": 28.075942993164062, "learning_rate": 8.186677631578948e-06, "loss": 0.3684, "step": 882 }, { "epoch": 0.18153782894736842, "grad_norm": 45.93857192993164, "learning_rate": 8.184621710526316e-06, "loss": 0.8311, "step": 883 }, { "epoch": 0.18174342105263158, "grad_norm": 92.395751953125, "learning_rate": 8.182565789473685e-06, "loss": 1.4531, "step": 884 }, { "epoch": 0.18194901315789475, "grad_norm": 56.68762969970703, "learning_rate": 8.180509868421053e-06, "loss": 0.5934, "step": 885 }, { "epoch": 0.18215460526315788, "grad_norm": 75.53089904785156, "learning_rate": 8.178453947368423e-06, "loss": 1.2148, "step": 886 }, { "epoch": 0.18236019736842105, "grad_norm": 36.817115783691406, "learning_rate": 8.17639802631579e-06, "loss": 0.8789, "step": 887 }, { "epoch": 0.1825657894736842, "grad_norm": 53.36594772338867, "learning_rate": 8.174342105263158e-06, "loss": 0.4741, "step": 888 }, { "epoch": 0.18277138157894737, "grad_norm": 62.18541717529297, "learning_rate": 8.172286184210528e-06, "loss": 1.127, "step": 889 }, { "epoch": 0.18297697368421054, "grad_norm": 123.99641418457031, "learning_rate": 8.170230263157896e-06, "loss": 1.0146, "step": 890 }, { "epoch": 0.18318256578947367, "grad_norm": 71.2369613647461, "learning_rate": 8.168174342105264e-06, "loss": 0.9062, "step": 891 }, { "epoch": 0.18338815789473684, "grad_norm": 72.19764709472656, "learning_rate": 8.166118421052632e-06, "loss": 1.3711, "step": 892 }, { "epoch": 0.18359375, "grad_norm": 37.888301849365234, "learning_rate": 8.164062500000001e-06, "loss": 0.8438, "step": 893 }, { "epoch": 0.18379934210526316, "grad_norm": 49.95079040527344, "learning_rate": 8.162006578947369e-06, "loss": 1.3398, "step": 894 }, { "epoch": 0.18400493421052633, "grad_norm": 70.93107604980469, "learning_rate": 8.159950657894737e-06, "loss": 1.5586, "step": 895 }, { "epoch": 0.18421052631578946, "grad_norm": 42.17367172241211, "learning_rate": 8.157894736842106e-06, "loss": 0.627, "step": 896 }, { "epoch": 0.18441611842105263, "grad_norm": 79.28921508789062, "learning_rate": 8.155838815789474e-06, "loss": 1.1582, "step": 897 }, { "epoch": 0.1846217105263158, "grad_norm": 42.4417610168457, "learning_rate": 8.153782894736842e-06, "loss": 0.6836, "step": 898 }, { "epoch": 0.18482730263157895, "grad_norm": 39.043724060058594, "learning_rate": 8.15172697368421e-06, "loss": 0.5908, "step": 899 }, { "epoch": 0.18503289473684212, "grad_norm": 114.31543731689453, "learning_rate": 8.14967105263158e-06, "loss": 2.4609, "step": 900 }, { "epoch": 0.18523848684210525, "grad_norm": 30.667861938476562, "learning_rate": 8.147615131578948e-06, "loss": 0.2983, "step": 901 }, { "epoch": 0.18544407894736842, "grad_norm": 17.16651725769043, "learning_rate": 8.145559210526315e-06, "loss": 0.104, "step": 902 }, { "epoch": 0.18564967105263158, "grad_norm": 152.54837036132812, "learning_rate": 8.143503289473685e-06, "loss": 1.7227, "step": 903 }, { "epoch": 0.18585526315789475, "grad_norm": 50.75700759887695, "learning_rate": 8.141447368421055e-06, "loss": 1.1133, "step": 904 }, { "epoch": 0.18606085526315788, "grad_norm": 92.75712585449219, "learning_rate": 8.139391447368422e-06, "loss": 1.168, "step": 905 }, { "epoch": 0.18626644736842105, "grad_norm": 37.260677337646484, "learning_rate": 8.13733552631579e-06, "loss": 0.7383, "step": 906 }, { "epoch": 0.1864720394736842, "grad_norm": 96.64717864990234, "learning_rate": 8.135279605263158e-06, "loss": 1.4246, "step": 907 }, { "epoch": 0.18667763157894737, "grad_norm": 84.24066925048828, "learning_rate": 8.133223684210528e-06, "loss": 1.4131, "step": 908 }, { "epoch": 0.18688322368421054, "grad_norm": 80.53427124023438, "learning_rate": 8.131167763157896e-06, "loss": 1.3711, "step": 909 }, { "epoch": 0.18708881578947367, "grad_norm": 86.02105712890625, "learning_rate": 8.129111842105264e-06, "loss": 1.0488, "step": 910 }, { "epoch": 0.18729440789473684, "grad_norm": 33.86091613769531, "learning_rate": 8.127055921052633e-06, "loss": 0.2502, "step": 911 }, { "epoch": 0.1875, "grad_norm": 15.04715633392334, "learning_rate": 8.125000000000001e-06, "loss": 0.1973, "step": 912 }, { "epoch": 0.18770559210526316, "grad_norm": 72.32575225830078, "learning_rate": 8.122944078947369e-06, "loss": 0.832, "step": 913 }, { "epoch": 0.18791118421052633, "grad_norm": 93.72522735595703, "learning_rate": 8.120888157894737e-06, "loss": 1.5005, "step": 914 }, { "epoch": 0.18811677631578946, "grad_norm": 69.71036529541016, "learning_rate": 8.118832236842106e-06, "loss": 1.6484, "step": 915 }, { "epoch": 0.18832236842105263, "grad_norm": 55.752525329589844, "learning_rate": 8.116776315789474e-06, "loss": 1.1523, "step": 916 }, { "epoch": 0.1885279605263158, "grad_norm": 32.2404670715332, "learning_rate": 8.114720394736842e-06, "loss": 0.7207, "step": 917 }, { "epoch": 0.18873355263157895, "grad_norm": 39.2967414855957, "learning_rate": 8.112664473684212e-06, "loss": 0.2998, "step": 918 }, { "epoch": 0.18893914473684212, "grad_norm": 54.65935516357422, "learning_rate": 8.11060855263158e-06, "loss": 1.2461, "step": 919 }, { "epoch": 0.18914473684210525, "grad_norm": 87.1291275024414, "learning_rate": 8.108552631578947e-06, "loss": 0.6895, "step": 920 }, { "epoch": 0.18935032894736842, "grad_norm": 34.256752014160156, "learning_rate": 8.106496710526315e-06, "loss": 0.2402, "step": 921 }, { "epoch": 0.18955592105263158, "grad_norm": 20.202226638793945, "learning_rate": 8.104440789473685e-06, "loss": 0.394, "step": 922 }, { "epoch": 0.18976151315789475, "grad_norm": 72.20425415039062, "learning_rate": 8.102384868421054e-06, "loss": 0.9316, "step": 923 }, { "epoch": 0.18996710526315788, "grad_norm": 63.2459716796875, "learning_rate": 8.100328947368422e-06, "loss": 0.8672, "step": 924 }, { "epoch": 0.19017269736842105, "grad_norm": 62.19611740112305, "learning_rate": 8.09827302631579e-06, "loss": 0.6621, "step": 925 }, { "epoch": 0.1903782894736842, "grad_norm": 59.0235481262207, "learning_rate": 8.096217105263158e-06, "loss": 0.5015, "step": 926 }, { "epoch": 0.19058388157894737, "grad_norm": 62.96659469604492, "learning_rate": 8.094161184210528e-06, "loss": 0.5781, "step": 927 }, { "epoch": 0.19078947368421054, "grad_norm": 20.47020149230957, "learning_rate": 8.092105263157896e-06, "loss": 0.231, "step": 928 }, { "epoch": 0.19099506578947367, "grad_norm": 49.3768310546875, "learning_rate": 8.090049342105263e-06, "loss": 1.1748, "step": 929 }, { "epoch": 0.19120065789473684, "grad_norm": 40.16157531738281, "learning_rate": 8.087993421052633e-06, "loss": 0.8936, "step": 930 }, { "epoch": 0.19140625, "grad_norm": 149.476806640625, "learning_rate": 8.085937500000001e-06, "loss": 2.1406, "step": 931 }, { "epoch": 0.19161184210526316, "grad_norm": 35.28090286254883, "learning_rate": 8.083881578947369e-06, "loss": 0.3828, "step": 932 }, { "epoch": 0.19181743421052633, "grad_norm": 88.84160614013672, "learning_rate": 8.081825657894737e-06, "loss": 1.2383, "step": 933 }, { "epoch": 0.19202302631578946, "grad_norm": 114.96373748779297, "learning_rate": 8.079769736842106e-06, "loss": 1.6714, "step": 934 }, { "epoch": 0.19222861842105263, "grad_norm": 25.39069175720215, "learning_rate": 8.077713815789474e-06, "loss": 0.6859, "step": 935 }, { "epoch": 0.1924342105263158, "grad_norm": 72.97904968261719, "learning_rate": 8.075657894736842e-06, "loss": 1.8828, "step": 936 }, { "epoch": 0.19263980263157895, "grad_norm": 144.6828155517578, "learning_rate": 8.073601973684212e-06, "loss": 4.6172, "step": 937 }, { "epoch": 0.19284539473684212, "grad_norm": 23.279582977294922, "learning_rate": 8.07154605263158e-06, "loss": 0.2949, "step": 938 }, { "epoch": 0.19305098684210525, "grad_norm": 23.883325576782227, "learning_rate": 8.069490131578947e-06, "loss": 0.11, "step": 939 }, { "epoch": 0.19325657894736842, "grad_norm": 26.210596084594727, "learning_rate": 8.067434210526315e-06, "loss": 0.2138, "step": 940 }, { "epoch": 0.19346217105263158, "grad_norm": 30.82142448425293, "learning_rate": 8.065378289473685e-06, "loss": 0.5137, "step": 941 }, { "epoch": 0.19366776315789475, "grad_norm": 110.36473846435547, "learning_rate": 8.063322368421054e-06, "loss": 1.4531, "step": 942 }, { "epoch": 0.19387335526315788, "grad_norm": 53.813785552978516, "learning_rate": 8.061266447368422e-06, "loss": 0.3589, "step": 943 }, { "epoch": 0.19407894736842105, "grad_norm": 30.58393669128418, "learning_rate": 8.05921052631579e-06, "loss": 1.1172, "step": 944 }, { "epoch": 0.1942845394736842, "grad_norm": 48.89240646362305, "learning_rate": 8.05715460526316e-06, "loss": 1.4414, "step": 945 }, { "epoch": 0.19449013157894737, "grad_norm": 46.165985107421875, "learning_rate": 8.055098684210528e-06, "loss": 0.6162, "step": 946 }, { "epoch": 0.19469572368421054, "grad_norm": 93.2582015991211, "learning_rate": 8.053042763157895e-06, "loss": 1.5371, "step": 947 }, { "epoch": 0.19490131578947367, "grad_norm": 194.70730590820312, "learning_rate": 8.050986842105263e-06, "loss": 2.8281, "step": 948 }, { "epoch": 0.19510690789473684, "grad_norm": 112.63994598388672, "learning_rate": 8.048930921052633e-06, "loss": 1.1191, "step": 949 }, { "epoch": 0.1953125, "grad_norm": 22.64583969116211, "learning_rate": 8.046875e-06, "loss": 0.6582, "step": 950 }, { "epoch": 0.19551809210526316, "grad_norm": 48.50675964355469, "learning_rate": 8.044819078947369e-06, "loss": 1.0757, "step": 951 }, { "epoch": 0.19572368421052633, "grad_norm": 30.7310848236084, "learning_rate": 8.042763157894737e-06, "loss": 0.7217, "step": 952 }, { "epoch": 0.19592927631578946, "grad_norm": 86.075927734375, "learning_rate": 8.040707236842106e-06, "loss": 1.4551, "step": 953 }, { "epoch": 0.19613486842105263, "grad_norm": 61.13302993774414, "learning_rate": 8.038651315789474e-06, "loss": 0.9824, "step": 954 }, { "epoch": 0.1963404605263158, "grad_norm": 17.352359771728516, "learning_rate": 8.036595394736842e-06, "loss": 0.1653, "step": 955 }, { "epoch": 0.19654605263157895, "grad_norm": 94.20052337646484, "learning_rate": 8.034539473684211e-06, "loss": 1.7871, "step": 956 }, { "epoch": 0.19675164473684212, "grad_norm": 38.36785125732422, "learning_rate": 8.03248355263158e-06, "loss": 0.4541, "step": 957 }, { "epoch": 0.19695723684210525, "grad_norm": 74.19889068603516, "learning_rate": 8.030427631578947e-06, "loss": 1.4727, "step": 958 }, { "epoch": 0.19716282894736842, "grad_norm": 31.721586227416992, "learning_rate": 8.028371710526315e-06, "loss": 0.3801, "step": 959 }, { "epoch": 0.19736842105263158, "grad_norm": 24.497468948364258, "learning_rate": 8.026315789473685e-06, "loss": 0.2446, "step": 960 }, { "epoch": 0.19757401315789475, "grad_norm": 52.174476623535156, "learning_rate": 8.024259868421054e-06, "loss": 0.5271, "step": 961 }, { "epoch": 0.19777960526315788, "grad_norm": 32.147159576416016, "learning_rate": 8.022203947368422e-06, "loss": 0.5195, "step": 962 }, { "epoch": 0.19798519736842105, "grad_norm": 186.4349822998047, "learning_rate": 8.02014802631579e-06, "loss": 1.6455, "step": 963 }, { "epoch": 0.1981907894736842, "grad_norm": 80.9439926147461, "learning_rate": 8.01809210526316e-06, "loss": 1.6538, "step": 964 }, { "epoch": 0.19839638157894737, "grad_norm": 22.226125717163086, "learning_rate": 8.016036184210527e-06, "loss": 0.344, "step": 965 }, { "epoch": 0.19860197368421054, "grad_norm": 66.10164642333984, "learning_rate": 8.013980263157895e-06, "loss": 0.7598, "step": 966 }, { "epoch": 0.19880756578947367, "grad_norm": 52.738468170166016, "learning_rate": 8.011924342105263e-06, "loss": 0.5898, "step": 967 }, { "epoch": 0.19901315789473684, "grad_norm": 67.19638061523438, "learning_rate": 8.009868421052633e-06, "loss": 1.2246, "step": 968 }, { "epoch": 0.19921875, "grad_norm": 36.44961166381836, "learning_rate": 8.0078125e-06, "loss": 0.7139, "step": 969 }, { "epoch": 0.19942434210526316, "grad_norm": 151.35704040527344, "learning_rate": 8.005756578947369e-06, "loss": 1.668, "step": 970 }, { "epoch": 0.19962993421052633, "grad_norm": 82.3469009399414, "learning_rate": 8.003700657894738e-06, "loss": 0.5459, "step": 971 }, { "epoch": 0.19983552631578946, "grad_norm": 114.07935333251953, "learning_rate": 8.001644736842106e-06, "loss": 1.6621, "step": 972 }, { "epoch": 0.20004111842105263, "grad_norm": 40.399749755859375, "learning_rate": 7.999588815789474e-06, "loss": 0.3606, "step": 973 }, { "epoch": 0.2002467105263158, "grad_norm": 109.84380340576172, "learning_rate": 7.997532894736842e-06, "loss": 1.3398, "step": 974 }, { "epoch": 0.20045230263157895, "grad_norm": 64.76303100585938, "learning_rate": 7.995476973684211e-06, "loss": 1.8789, "step": 975 }, { "epoch": 0.20065789473684212, "grad_norm": 33.260284423828125, "learning_rate": 7.99342105263158e-06, "loss": 0.3716, "step": 976 }, { "epoch": 0.20086348684210525, "grad_norm": 42.699405670166016, "learning_rate": 7.991365131578947e-06, "loss": 0.9629, "step": 977 }, { "epoch": 0.20106907894736842, "grad_norm": 24.7950382232666, "learning_rate": 7.989309210526317e-06, "loss": 0.5544, "step": 978 }, { "epoch": 0.20127467105263158, "grad_norm": 145.00035095214844, "learning_rate": 7.987253289473685e-06, "loss": 3.1094, "step": 979 }, { "epoch": 0.20148026315789475, "grad_norm": 25.82627296447754, "learning_rate": 7.985197368421054e-06, "loss": 0.5847, "step": 980 }, { "epoch": 0.20168585526315788, "grad_norm": 93.4742431640625, "learning_rate": 7.983141447368422e-06, "loss": 1.374, "step": 981 }, { "epoch": 0.20189144736842105, "grad_norm": 60.8559684753418, "learning_rate": 7.98108552631579e-06, "loss": 1.6914, "step": 982 }, { "epoch": 0.2020970394736842, "grad_norm": 83.05685424804688, "learning_rate": 7.97902960526316e-06, "loss": 0.9932, "step": 983 }, { "epoch": 0.20230263157894737, "grad_norm": 19.37748908996582, "learning_rate": 7.976973684210527e-06, "loss": 0.115, "step": 984 }, { "epoch": 0.20250822368421054, "grad_norm": 53.78739929199219, "learning_rate": 7.974917763157895e-06, "loss": 1.2688, "step": 985 }, { "epoch": 0.20271381578947367, "grad_norm": 41.915672302246094, "learning_rate": 7.972861842105263e-06, "loss": 1.0254, "step": 986 }, { "epoch": 0.20291940789473684, "grad_norm": 55.22368240356445, "learning_rate": 7.970805921052633e-06, "loss": 0.5078, "step": 987 }, { "epoch": 0.203125, "grad_norm": 57.66218185424805, "learning_rate": 7.96875e-06, "loss": 0.8337, "step": 988 }, { "epoch": 0.20333059210526316, "grad_norm": 49.647911071777344, "learning_rate": 7.966694078947368e-06, "loss": 1.6133, "step": 989 }, { "epoch": 0.20353618421052633, "grad_norm": 15.947494506835938, "learning_rate": 7.964638157894738e-06, "loss": 0.0812, "step": 990 }, { "epoch": 0.20374177631578946, "grad_norm": 101.75177001953125, "learning_rate": 7.962582236842106e-06, "loss": 2.4258, "step": 991 }, { "epoch": 0.20394736842105263, "grad_norm": 75.89444732666016, "learning_rate": 7.960526315789474e-06, "loss": 1.7227, "step": 992 }, { "epoch": 0.2041529605263158, "grad_norm": 47.2090950012207, "learning_rate": 7.958470394736842e-06, "loss": 1.1606, "step": 993 }, { "epoch": 0.20435855263157895, "grad_norm": 88.2051010131836, "learning_rate": 7.956414473684211e-06, "loss": 1.9922, "step": 994 }, { "epoch": 0.20456414473684212, "grad_norm": 18.619991302490234, "learning_rate": 7.954358552631579e-06, "loss": 0.1071, "step": 995 }, { "epoch": 0.20476973684210525, "grad_norm": 87.63352966308594, "learning_rate": 7.952302631578947e-06, "loss": 1.625, "step": 996 }, { "epoch": 0.20497532894736842, "grad_norm": 28.964557647705078, "learning_rate": 7.950246710526317e-06, "loss": 0.8989, "step": 997 }, { "epoch": 0.20518092105263158, "grad_norm": 47.482086181640625, "learning_rate": 7.948190789473686e-06, "loss": 1.3398, "step": 998 }, { "epoch": 0.20538651315789475, "grad_norm": 18.82389259338379, "learning_rate": 7.946134868421054e-06, "loss": 0.3926, "step": 999 }, { "epoch": 0.20559210526315788, "grad_norm": 170.39083862304688, "learning_rate": 7.944078947368422e-06, "loss": 3.7465, "step": 1000 }, { "epoch": 0.20579769736842105, "grad_norm": 61.24917221069336, "learning_rate": 7.94202302631579e-06, "loss": 1.293, "step": 1001 }, { "epoch": 0.2060032894736842, "grad_norm": 45.527103424072266, "learning_rate": 7.93996710526316e-06, "loss": 0.5435, "step": 1002 }, { "epoch": 0.20620888157894737, "grad_norm": 44.50002670288086, "learning_rate": 7.937911184210527e-06, "loss": 0.3887, "step": 1003 }, { "epoch": 0.20641447368421054, "grad_norm": 141.9852294921875, "learning_rate": 7.935855263157895e-06, "loss": 3.0352, "step": 1004 }, { "epoch": 0.20662006578947367, "grad_norm": 25.379989624023438, "learning_rate": 7.933799342105265e-06, "loss": 0.4785, "step": 1005 }, { "epoch": 0.20682565789473684, "grad_norm": 41.58266830444336, "learning_rate": 7.931743421052633e-06, "loss": 0.5273, "step": 1006 }, { "epoch": 0.20703125, "grad_norm": 31.876428604125977, "learning_rate": 7.9296875e-06, "loss": 1.375, "step": 1007 }, { "epoch": 0.20723684210526316, "grad_norm": 102.0283432006836, "learning_rate": 7.927631578947368e-06, "loss": 1.3789, "step": 1008 }, { "epoch": 0.20744243421052633, "grad_norm": 116.41083526611328, "learning_rate": 7.925575657894738e-06, "loss": 2.2539, "step": 1009 }, { "epoch": 0.20764802631578946, "grad_norm": 19.452754974365234, "learning_rate": 7.923519736842106e-06, "loss": 0.5361, "step": 1010 }, { "epoch": 0.20785361842105263, "grad_norm": 54.097415924072266, "learning_rate": 7.921463815789474e-06, "loss": 1.0068, "step": 1011 }, { "epoch": 0.2080592105263158, "grad_norm": 43.8322868347168, "learning_rate": 7.919407894736843e-06, "loss": 0.7988, "step": 1012 }, { "epoch": 0.20826480263157895, "grad_norm": 90.24042510986328, "learning_rate": 7.917351973684211e-06, "loss": 1.7188, "step": 1013 }, { "epoch": 0.20847039473684212, "grad_norm": 42.9840202331543, "learning_rate": 7.915296052631579e-06, "loss": 0.6504, "step": 1014 }, { "epoch": 0.20867598684210525, "grad_norm": 69.71157836914062, "learning_rate": 7.913240131578947e-06, "loss": 1.8438, "step": 1015 }, { "epoch": 0.20888157894736842, "grad_norm": 50.517295837402344, "learning_rate": 7.911184210526316e-06, "loss": 0.834, "step": 1016 }, { "epoch": 0.20908717105263158, "grad_norm": 24.43018341064453, "learning_rate": 7.909128289473686e-06, "loss": 0.7959, "step": 1017 }, { "epoch": 0.20929276315789475, "grad_norm": 89.54730224609375, "learning_rate": 7.907072368421054e-06, "loss": 1.7383, "step": 1018 }, { "epoch": 0.20949835526315788, "grad_norm": 47.35501480102539, "learning_rate": 7.905016447368422e-06, "loss": 0.8301, "step": 1019 }, { "epoch": 0.20970394736842105, "grad_norm": 23.631650924682617, "learning_rate": 7.90296052631579e-06, "loss": 0.2646, "step": 1020 }, { "epoch": 0.2099095394736842, "grad_norm": 59.05613708496094, "learning_rate": 7.90090460526316e-06, "loss": 0.9375, "step": 1021 }, { "epoch": 0.21011513157894737, "grad_norm": 31.860857009887695, "learning_rate": 7.898848684210527e-06, "loss": 0.3574, "step": 1022 }, { "epoch": 0.21032072368421054, "grad_norm": 63.234100341796875, "learning_rate": 7.896792763157895e-06, "loss": 1.0371, "step": 1023 }, { "epoch": 0.21052631578947367, "grad_norm": 162.39967346191406, "learning_rate": 7.894736842105265e-06, "loss": 1.6836, "step": 1024 }, { "epoch": 0.21073190789473684, "grad_norm": 62.477725982666016, "learning_rate": 7.892680921052632e-06, "loss": 0.7417, "step": 1025 }, { "epoch": 0.2109375, "grad_norm": 48.06618118286133, "learning_rate": 7.890625e-06, "loss": 0.5898, "step": 1026 }, { "epoch": 0.21114309210526316, "grad_norm": 114.32039642333984, "learning_rate": 7.888569078947368e-06, "loss": 1.7969, "step": 1027 }, { "epoch": 0.21134868421052633, "grad_norm": 66.91285705566406, "learning_rate": 7.886513157894738e-06, "loss": 1.1406, "step": 1028 }, { "epoch": 0.21155427631578946, "grad_norm": 52.01984786987305, "learning_rate": 7.884457236842106e-06, "loss": 0.6787, "step": 1029 }, { "epoch": 0.21175986842105263, "grad_norm": 25.37289810180664, "learning_rate": 7.882401315789474e-06, "loss": 0.557, "step": 1030 }, { "epoch": 0.2119654605263158, "grad_norm": 102.43817901611328, "learning_rate": 7.880345394736843e-06, "loss": 1.0381, "step": 1031 }, { "epoch": 0.21217105263157895, "grad_norm": 133.42950439453125, "learning_rate": 7.878289473684211e-06, "loss": 4.1797, "step": 1032 }, { "epoch": 0.21237664473684212, "grad_norm": 27.53020668029785, "learning_rate": 7.876233552631579e-06, "loss": 0.4442, "step": 1033 }, { "epoch": 0.21258223684210525, "grad_norm": 54.91940689086914, "learning_rate": 7.874177631578947e-06, "loss": 2.1797, "step": 1034 }, { "epoch": 0.21278782894736842, "grad_norm": 13.80307388305664, "learning_rate": 7.872121710526316e-06, "loss": 0.332, "step": 1035 }, { "epoch": 0.21299342105263158, "grad_norm": 46.513240814208984, "learning_rate": 7.870065789473686e-06, "loss": 0.3536, "step": 1036 }, { "epoch": 0.21319901315789475, "grad_norm": 36.82777786254883, "learning_rate": 7.868009868421054e-06, "loss": 0.96, "step": 1037 }, { "epoch": 0.21340460526315788, "grad_norm": 133.81712341308594, "learning_rate": 7.865953947368422e-06, "loss": 3.9766, "step": 1038 }, { "epoch": 0.21361019736842105, "grad_norm": 155.23001098632812, "learning_rate": 7.863898026315791e-06, "loss": 3.2891, "step": 1039 }, { "epoch": 0.2138157894736842, "grad_norm": 94.83102416992188, "learning_rate": 7.86184210526316e-06, "loss": 1.4316, "step": 1040 }, { "epoch": 0.21402138157894737, "grad_norm": 27.875537872314453, "learning_rate": 7.859786184210527e-06, "loss": 0.3867, "step": 1041 }, { "epoch": 0.21422697368421054, "grad_norm": 7.562720775604248, "learning_rate": 7.857730263157895e-06, "loss": 0.0515, "step": 1042 }, { "epoch": 0.21443256578947367, "grad_norm": 37.16962814331055, "learning_rate": 7.855674342105265e-06, "loss": 0.3643, "step": 1043 }, { "epoch": 0.21463815789473684, "grad_norm": 112.35224914550781, "learning_rate": 7.853618421052632e-06, "loss": 1.4307, "step": 1044 }, { "epoch": 0.21484375, "grad_norm": 39.44952392578125, "learning_rate": 7.8515625e-06, "loss": 1.7305, "step": 1045 }, { "epoch": 0.21504934210526316, "grad_norm": 108.35917663574219, "learning_rate": 7.84950657894737e-06, "loss": 2.1719, "step": 1046 }, { "epoch": 0.21525493421052633, "grad_norm": 41.99906921386719, "learning_rate": 7.847450657894738e-06, "loss": 0.8291, "step": 1047 }, { "epoch": 0.21546052631578946, "grad_norm": 20.198474884033203, "learning_rate": 7.845394736842106e-06, "loss": 0.1177, "step": 1048 }, { "epoch": 0.21566611842105263, "grad_norm": 40.171852111816406, "learning_rate": 7.843338815789473e-06, "loss": 0.9678, "step": 1049 }, { "epoch": 0.2158717105263158, "grad_norm": 28.797210693359375, "learning_rate": 7.841282894736843e-06, "loss": 0.3945, "step": 1050 }, { "epoch": 0.21607730263157895, "grad_norm": 53.02735137939453, "learning_rate": 7.839226973684211e-06, "loss": 0.3492, "step": 1051 }, { "epoch": 0.21628289473684212, "grad_norm": 87.19514465332031, "learning_rate": 7.837171052631579e-06, "loss": 1.3452, "step": 1052 }, { "epoch": 0.21648848684210525, "grad_norm": 26.628854751586914, "learning_rate": 7.835115131578947e-06, "loss": 0.1826, "step": 1053 }, { "epoch": 0.21669407894736842, "grad_norm": 19.074636459350586, "learning_rate": 7.833059210526316e-06, "loss": 0.6792, "step": 1054 }, { "epoch": 0.21689967105263158, "grad_norm": 6.944777488708496, "learning_rate": 7.831003289473686e-06, "loss": 0.0664, "step": 1055 }, { "epoch": 0.21710526315789475, "grad_norm": 105.94255065917969, "learning_rate": 7.828947368421054e-06, "loss": 1.4844, "step": 1056 }, { "epoch": 0.21731085526315788, "grad_norm": 79.083740234375, "learning_rate": 7.826891447368422e-06, "loss": 1.0366, "step": 1057 }, { "epoch": 0.21751644736842105, "grad_norm": 47.30836868286133, "learning_rate": 7.824835526315791e-06, "loss": 1.4375, "step": 1058 }, { "epoch": 0.2177220394736842, "grad_norm": 60.837493896484375, "learning_rate": 7.822779605263159e-06, "loss": 1.0403, "step": 1059 }, { "epoch": 0.21792763157894737, "grad_norm": 34.3600959777832, "learning_rate": 7.820723684210527e-06, "loss": 0.2991, "step": 1060 }, { "epoch": 0.21813322368421054, "grad_norm": 31.08753776550293, "learning_rate": 7.818667763157895e-06, "loss": 0.3975, "step": 1061 }, { "epoch": 0.21833881578947367, "grad_norm": 169.1045379638672, "learning_rate": 7.816611842105264e-06, "loss": 4.4922, "step": 1062 }, { "epoch": 0.21854440789473684, "grad_norm": 49.69659423828125, "learning_rate": 7.814555921052632e-06, "loss": 0.6875, "step": 1063 }, { "epoch": 0.21875, "grad_norm": 58.87510681152344, "learning_rate": 7.8125e-06, "loss": 1.375, "step": 1064 }, { "epoch": 0.21895559210526316, "grad_norm": 37.57533264160156, "learning_rate": 7.81044407894737e-06, "loss": 0.4729, "step": 1065 }, { "epoch": 0.21916118421052633, "grad_norm": 24.772003173828125, "learning_rate": 7.808388157894738e-06, "loss": 0.6234, "step": 1066 }, { "epoch": 0.21936677631578946, "grad_norm": 213.3047332763672, "learning_rate": 7.806332236842105e-06, "loss": 1.7812, "step": 1067 }, { "epoch": 0.21957236842105263, "grad_norm": 106.89733123779297, "learning_rate": 7.804276315789473e-06, "loss": 2.0586, "step": 1068 }, { "epoch": 0.2197779605263158, "grad_norm": 57.08930587768555, "learning_rate": 7.802220394736843e-06, "loss": 1.9219, "step": 1069 }, { "epoch": 0.21998355263157895, "grad_norm": 52.6435661315918, "learning_rate": 7.80016447368421e-06, "loss": 0.4998, "step": 1070 }, { "epoch": 0.22018914473684212, "grad_norm": 39.55086135864258, "learning_rate": 7.798108552631579e-06, "loss": 0.627, "step": 1071 }, { "epoch": 0.22039473684210525, "grad_norm": 16.066463470458984, "learning_rate": 7.796052631578948e-06, "loss": 0.1041, "step": 1072 }, { "epoch": 0.22060032894736842, "grad_norm": 38.09550476074219, "learning_rate": 7.793996710526316e-06, "loss": 0.4255, "step": 1073 }, { "epoch": 0.22080592105263158, "grad_norm": 45.24042510986328, "learning_rate": 7.791940789473686e-06, "loss": 0.5576, "step": 1074 }, { "epoch": 0.22101151315789475, "grad_norm": 28.552001953125, "learning_rate": 7.789884868421054e-06, "loss": 0.3423, "step": 1075 }, { "epoch": 0.22121710526315788, "grad_norm": 92.75128173828125, "learning_rate": 7.787828947368422e-06, "loss": 2.0195, "step": 1076 }, { "epoch": 0.22142269736842105, "grad_norm": 9.256317138671875, "learning_rate": 7.785773026315791e-06, "loss": 0.0739, "step": 1077 }, { "epoch": 0.2216282894736842, "grad_norm": 53.657432556152344, "learning_rate": 7.783717105263159e-06, "loss": 0.6777, "step": 1078 }, { "epoch": 0.22183388157894737, "grad_norm": 8.292916297912598, "learning_rate": 7.781661184210527e-06, "loss": 0.0427, "step": 1079 }, { "epoch": 0.22203947368421054, "grad_norm": 108.16395568847656, "learning_rate": 7.779605263157895e-06, "loss": 1.2935, "step": 1080 }, { "epoch": 0.22224506578947367, "grad_norm": 199.14163208007812, "learning_rate": 7.777549342105264e-06, "loss": 2.5625, "step": 1081 }, { "epoch": 0.22245065789473684, "grad_norm": 189.4644775390625, "learning_rate": 7.775493421052632e-06, "loss": 5.382, "step": 1082 }, { "epoch": 0.22265625, "grad_norm": 72.38389587402344, "learning_rate": 7.7734375e-06, "loss": 0.6626, "step": 1083 }, { "epoch": 0.22286184210526316, "grad_norm": 56.37211608886719, "learning_rate": 7.77138157894737e-06, "loss": 0.5645, "step": 1084 }, { "epoch": 0.22306743421052633, "grad_norm": 31.613569259643555, "learning_rate": 7.769325657894738e-06, "loss": 0.1621, "step": 1085 }, { "epoch": 0.22327302631578946, "grad_norm": 128.200439453125, "learning_rate": 7.767269736842105e-06, "loss": 3.3164, "step": 1086 }, { "epoch": 0.22347861842105263, "grad_norm": 121.77963256835938, "learning_rate": 7.765213815789473e-06, "loss": 4.9099, "step": 1087 }, { "epoch": 0.2236842105263158, "grad_norm": 35.31592559814453, "learning_rate": 7.763157894736843e-06, "loss": 0.9346, "step": 1088 }, { "epoch": 0.22388980263157895, "grad_norm": 52.60261917114258, "learning_rate": 7.76110197368421e-06, "loss": 0.9023, "step": 1089 }, { "epoch": 0.22409539473684212, "grad_norm": 105.51455688476562, "learning_rate": 7.759046052631579e-06, "loss": 2.3281, "step": 1090 }, { "epoch": 0.22430098684210525, "grad_norm": 55.43037414550781, "learning_rate": 7.756990131578948e-06, "loss": 1.3579, "step": 1091 }, { "epoch": 0.22450657894736842, "grad_norm": 63.85261535644531, "learning_rate": 7.754934210526316e-06, "loss": 1.418, "step": 1092 }, { "epoch": 0.22471217105263158, "grad_norm": 41.611446380615234, "learning_rate": 7.752878289473686e-06, "loss": 0.4336, "step": 1093 }, { "epoch": 0.22491776315789475, "grad_norm": 27.610265731811523, "learning_rate": 7.750822368421054e-06, "loss": 1.1738, "step": 1094 }, { "epoch": 0.22512335526315788, "grad_norm": 32.40602493286133, "learning_rate": 7.748766447368421e-06, "loss": 0.5986, "step": 1095 }, { "epoch": 0.22532894736842105, "grad_norm": 46.989253997802734, "learning_rate": 7.746710526315791e-06, "loss": 0.5059, "step": 1096 }, { "epoch": 0.2255345394736842, "grad_norm": 36.69861602783203, "learning_rate": 7.744654605263159e-06, "loss": 0.4351, "step": 1097 }, { "epoch": 0.22574013157894737, "grad_norm": 50.10142517089844, "learning_rate": 7.742598684210527e-06, "loss": 0.6016, "step": 1098 }, { "epoch": 0.22594572368421054, "grad_norm": 24.425554275512695, "learning_rate": 7.740542763157896e-06, "loss": 0.7695, "step": 1099 }, { "epoch": 0.22615131578947367, "grad_norm": 11.09209156036377, "learning_rate": 7.738486842105264e-06, "loss": 0.1138, "step": 1100 }, { "epoch": 0.22635690789473684, "grad_norm": 11.106768608093262, "learning_rate": 7.736430921052632e-06, "loss": 0.0679, "step": 1101 }, { "epoch": 0.2265625, "grad_norm": 54.8348388671875, "learning_rate": 7.734375e-06, "loss": 1.1387, "step": 1102 }, { "epoch": 0.22676809210526316, "grad_norm": 22.478654861450195, "learning_rate": 7.73231907894737e-06, "loss": 0.4481, "step": 1103 }, { "epoch": 0.22697368421052633, "grad_norm": 12.69533634185791, "learning_rate": 7.730263157894737e-06, "loss": 0.1768, "step": 1104 }, { "epoch": 0.22717927631578946, "grad_norm": 124.11457061767578, "learning_rate": 7.728207236842105e-06, "loss": 2.2031, "step": 1105 }, { "epoch": 0.22738486842105263, "grad_norm": 94.24560546875, "learning_rate": 7.726151315789475e-06, "loss": 1.6133, "step": 1106 }, { "epoch": 0.2275904605263158, "grad_norm": 105.40461730957031, "learning_rate": 7.724095394736843e-06, "loss": 1.1523, "step": 1107 }, { "epoch": 0.22779605263157895, "grad_norm": 76.28590393066406, "learning_rate": 7.72203947368421e-06, "loss": 1.8047, "step": 1108 }, { "epoch": 0.22800164473684212, "grad_norm": 90.71025085449219, "learning_rate": 7.719983552631579e-06, "loss": 1.2793, "step": 1109 }, { "epoch": 0.22820723684210525, "grad_norm": 27.842134475708008, "learning_rate": 7.717927631578948e-06, "loss": 0.5977, "step": 1110 }, { "epoch": 0.22841282894736842, "grad_norm": 35.75087356567383, "learning_rate": 7.715871710526316e-06, "loss": 1.0645, "step": 1111 }, { "epoch": 0.22861842105263158, "grad_norm": 59.92615509033203, "learning_rate": 7.713815789473686e-06, "loss": 1.0762, "step": 1112 }, { "epoch": 0.22882401315789475, "grad_norm": 125.94391632080078, "learning_rate": 7.711759868421053e-06, "loss": 1.8311, "step": 1113 }, { "epoch": 0.22902960526315788, "grad_norm": 195.8531494140625, "learning_rate": 7.709703947368421e-06, "loss": 2.8535, "step": 1114 }, { "epoch": 0.22923519736842105, "grad_norm": 28.551443099975586, "learning_rate": 7.707648026315791e-06, "loss": 0.1191, "step": 1115 }, { "epoch": 0.2294407894736842, "grad_norm": 39.70603942871094, "learning_rate": 7.705592105263159e-06, "loss": 0.2466, "step": 1116 }, { "epoch": 0.22964638157894737, "grad_norm": 68.98139953613281, "learning_rate": 7.703536184210527e-06, "loss": 1.2305, "step": 1117 }, { "epoch": 0.22985197368421054, "grad_norm": 26.285722732543945, "learning_rate": 7.701480263157896e-06, "loss": 0.1694, "step": 1118 }, { "epoch": 0.23005756578947367, "grad_norm": 58.49002456665039, "learning_rate": 7.699424342105264e-06, "loss": 0.8862, "step": 1119 }, { "epoch": 0.23026315789473684, "grad_norm": 27.220056533813477, "learning_rate": 7.697368421052632e-06, "loss": 0.1167, "step": 1120 }, { "epoch": 0.23046875, "grad_norm": 42.44319152832031, "learning_rate": 7.6953125e-06, "loss": 0.6373, "step": 1121 }, { "epoch": 0.23067434210526316, "grad_norm": 79.46936798095703, "learning_rate": 7.69325657894737e-06, "loss": 2.1445, "step": 1122 }, { "epoch": 0.23087993421052633, "grad_norm": 97.77277374267578, "learning_rate": 7.691200657894737e-06, "loss": 1.8379, "step": 1123 }, { "epoch": 0.23108552631578946, "grad_norm": 21.820125579833984, "learning_rate": 7.689144736842105e-06, "loss": 0.166, "step": 1124 }, { "epoch": 0.23129111842105263, "grad_norm": 144.39422607421875, "learning_rate": 7.687088815789475e-06, "loss": 3.0391, "step": 1125 }, { "epoch": 0.2314967105263158, "grad_norm": 84.17731475830078, "learning_rate": 7.685032894736843e-06, "loss": 1.2402, "step": 1126 }, { "epoch": 0.23170230263157895, "grad_norm": 52.6409912109375, "learning_rate": 7.68297697368421e-06, "loss": 1.5898, "step": 1127 }, { "epoch": 0.23190789473684212, "grad_norm": 145.22825622558594, "learning_rate": 7.680921052631578e-06, "loss": 3.9844, "step": 1128 }, { "epoch": 0.23211348684210525, "grad_norm": 65.86983489990234, "learning_rate": 7.678865131578948e-06, "loss": 1.2578, "step": 1129 }, { "epoch": 0.23231907894736842, "grad_norm": 21.269628524780273, "learning_rate": 7.676809210526316e-06, "loss": 0.1592, "step": 1130 }, { "epoch": 0.23252467105263158, "grad_norm": 36.935726165771484, "learning_rate": 7.674753289473685e-06, "loss": 0.6695, "step": 1131 }, { "epoch": 0.23273026315789475, "grad_norm": 29.519136428833008, "learning_rate": 7.672697368421053e-06, "loss": 0.208, "step": 1132 }, { "epoch": 0.23293585526315788, "grad_norm": 106.15299987792969, "learning_rate": 7.670641447368423e-06, "loss": 3.2664, "step": 1133 }, { "epoch": 0.23314144736842105, "grad_norm": 112.33860778808594, "learning_rate": 7.66858552631579e-06, "loss": 2.2355, "step": 1134 }, { "epoch": 0.2333470394736842, "grad_norm": 92.2876968383789, "learning_rate": 7.666529605263159e-06, "loss": 2.0195, "step": 1135 }, { "epoch": 0.23355263157894737, "grad_norm": 52.371028900146484, "learning_rate": 7.664473684210527e-06, "loss": 1.0488, "step": 1136 }, { "epoch": 0.23375822368421054, "grad_norm": 56.68132400512695, "learning_rate": 7.662417763157896e-06, "loss": 1.543, "step": 1137 }, { "epoch": 0.23396381578947367, "grad_norm": 77.58499145507812, "learning_rate": 7.660361842105264e-06, "loss": 1.7578, "step": 1138 }, { "epoch": 0.23416940789473684, "grad_norm": 156.91082763671875, "learning_rate": 7.658305921052632e-06, "loss": 3.8984, "step": 1139 }, { "epoch": 0.234375, "grad_norm": 49.3909797668457, "learning_rate": 7.656250000000001e-06, "loss": 0.875, "step": 1140 }, { "epoch": 0.23458059210526316, "grad_norm": 50.13317108154297, "learning_rate": 7.65419407894737e-06, "loss": 0.6816, "step": 1141 }, { "epoch": 0.23478618421052633, "grad_norm": 49.6068229675293, "learning_rate": 7.652138157894737e-06, "loss": 1.2891, "step": 1142 }, { "epoch": 0.23499177631578946, "grad_norm": 49.893714904785156, "learning_rate": 7.650082236842105e-06, "loss": 0.6836, "step": 1143 }, { "epoch": 0.23519736842105263, "grad_norm": 49.55369567871094, "learning_rate": 7.648026315789475e-06, "loss": 0.9219, "step": 1144 }, { "epoch": 0.2354029605263158, "grad_norm": 19.917491912841797, "learning_rate": 7.645970394736843e-06, "loss": 0.585, "step": 1145 }, { "epoch": 0.23560855263157895, "grad_norm": 27.70393180847168, "learning_rate": 7.64391447368421e-06, "loss": 0.4824, "step": 1146 }, { "epoch": 0.23581414473684212, "grad_norm": 23.954113006591797, "learning_rate": 7.64185855263158e-06, "loss": 0.9746, "step": 1147 }, { "epoch": 0.23601973684210525, "grad_norm": 38.5436897277832, "learning_rate": 7.639802631578948e-06, "loss": 0.71, "step": 1148 }, { "epoch": 0.23622532894736842, "grad_norm": 62.097171783447266, "learning_rate": 7.637746710526316e-06, "loss": 1.7227, "step": 1149 }, { "epoch": 0.23643092105263158, "grad_norm": 54.48320388793945, "learning_rate": 7.635690789473685e-06, "loss": 1.2227, "step": 1150 }, { "epoch": 0.23663651315789475, "grad_norm": 28.890033721923828, "learning_rate": 7.633634868421053e-06, "loss": 0.728, "step": 1151 }, { "epoch": 0.23684210526315788, "grad_norm": 50.08293533325195, "learning_rate": 7.631578947368423e-06, "loss": 1.2268, "step": 1152 }, { "epoch": 0.23704769736842105, "grad_norm": 77.17365264892578, "learning_rate": 7.62952302631579e-06, "loss": 1.6406, "step": 1153 }, { "epoch": 0.2372532894736842, "grad_norm": 83.27348327636719, "learning_rate": 7.6274671052631586e-06, "loss": 1.8477, "step": 1154 }, { "epoch": 0.23745888157894737, "grad_norm": 19.41888427734375, "learning_rate": 7.625411184210527e-06, "loss": 0.1929, "step": 1155 }, { "epoch": 0.23766447368421054, "grad_norm": 73.23312377929688, "learning_rate": 7.623355263157895e-06, "loss": 0.9492, "step": 1156 }, { "epoch": 0.23787006578947367, "grad_norm": 23.510608673095703, "learning_rate": 7.621299342105264e-06, "loss": 0.2559, "step": 1157 }, { "epoch": 0.23807565789473684, "grad_norm": 21.0023136138916, "learning_rate": 7.619243421052633e-06, "loss": 0.3711, "step": 1158 }, { "epoch": 0.23828125, "grad_norm": 87.44369506835938, "learning_rate": 7.6171875000000005e-06, "loss": 1.8672, "step": 1159 }, { "epoch": 0.23848684210526316, "grad_norm": 44.81456756591797, "learning_rate": 7.615131578947369e-06, "loss": 0.7812, "step": 1160 }, { "epoch": 0.23869243421052633, "grad_norm": 89.71466827392578, "learning_rate": 7.613075657894737e-06, "loss": 2.0166, "step": 1161 }, { "epoch": 0.23889802631578946, "grad_norm": 110.26764678955078, "learning_rate": 7.611019736842106e-06, "loss": 0.7715, "step": 1162 }, { "epoch": 0.23910361842105263, "grad_norm": 51.89344024658203, "learning_rate": 7.608963815789474e-06, "loss": 1.3652, "step": 1163 }, { "epoch": 0.2393092105263158, "grad_norm": 53.0599479675293, "learning_rate": 7.6069078947368424e-06, "loss": 0.7266, "step": 1164 }, { "epoch": 0.23951480263157895, "grad_norm": 40.53295135498047, "learning_rate": 7.604851973684211e-06, "loss": 0.3408, "step": 1165 }, { "epoch": 0.23972039473684212, "grad_norm": 122.67205810546875, "learning_rate": 7.602796052631579e-06, "loss": 1.668, "step": 1166 }, { "epoch": 0.23992598684210525, "grad_norm": 48.47007369995117, "learning_rate": 7.600740131578948e-06, "loss": 0.5752, "step": 1167 }, { "epoch": 0.24013157894736842, "grad_norm": 34.36094665527344, "learning_rate": 7.598684210526316e-06, "loss": 0.2427, "step": 1168 }, { "epoch": 0.24033717105263158, "grad_norm": 16.486661911010742, "learning_rate": 7.596628289473685e-06, "loss": 0.137, "step": 1169 }, { "epoch": 0.24054276315789475, "grad_norm": 41.22781753540039, "learning_rate": 7.594572368421054e-06, "loss": 1.5117, "step": 1170 }, { "epoch": 0.24074835526315788, "grad_norm": 25.72682762145996, "learning_rate": 7.592516447368422e-06, "loss": 0.333, "step": 1171 }, { "epoch": 0.24095394736842105, "grad_norm": 45.21040344238281, "learning_rate": 7.5904605263157906e-06, "loss": 0.3853, "step": 1172 }, { "epoch": 0.2411595394736842, "grad_norm": 15.82118034362793, "learning_rate": 7.5884046052631585e-06, "loss": 0.0531, "step": 1173 }, { "epoch": 0.24136513157894737, "grad_norm": 31.240829467773438, "learning_rate": 7.586348684210527e-06, "loss": 0.5869, "step": 1174 }, { "epoch": 0.24157072368421054, "grad_norm": 34.102230072021484, "learning_rate": 7.584292763157896e-06, "loss": 0.6426, "step": 1175 }, { "epoch": 0.24177631578947367, "grad_norm": 124.72824096679688, "learning_rate": 7.582236842105264e-06, "loss": 4.3115, "step": 1176 }, { "epoch": 0.24198190789473684, "grad_norm": 71.93131256103516, "learning_rate": 7.5801809210526325e-06, "loss": 0.592, "step": 1177 }, { "epoch": 0.2421875, "grad_norm": 29.20132064819336, "learning_rate": 7.578125e-06, "loss": 0.6234, "step": 1178 }, { "epoch": 0.24239309210526316, "grad_norm": 30.472171783447266, "learning_rate": 7.576069078947369e-06, "loss": 0.3251, "step": 1179 }, { "epoch": 0.24259868421052633, "grad_norm": 54.66229248046875, "learning_rate": 7.574013157894737e-06, "loss": 1.2266, "step": 1180 }, { "epoch": 0.24280427631578946, "grad_norm": 132.42901611328125, "learning_rate": 7.571957236842106e-06, "loss": 1.5066, "step": 1181 }, { "epoch": 0.24300986842105263, "grad_norm": 100.47991180419922, "learning_rate": 7.5699013157894745e-06, "loss": 2.0352, "step": 1182 }, { "epoch": 0.2432154605263158, "grad_norm": 52.75249481201172, "learning_rate": 7.567845394736842e-06, "loss": 0.5088, "step": 1183 }, { "epoch": 0.24342105263157895, "grad_norm": 29.94684410095215, "learning_rate": 7.565789473684211e-06, "loss": 0.8418, "step": 1184 }, { "epoch": 0.24362664473684212, "grad_norm": 40.22248458862305, "learning_rate": 7.563733552631579e-06, "loss": 1.0498, "step": 1185 }, { "epoch": 0.24383223684210525, "grad_norm": 77.0259017944336, "learning_rate": 7.561677631578948e-06, "loss": 1.3984, "step": 1186 }, { "epoch": 0.24403782894736842, "grad_norm": 132.8969268798828, "learning_rate": 7.5596217105263156e-06, "loss": 1.4336, "step": 1187 }, { "epoch": 0.24424342105263158, "grad_norm": 83.6072769165039, "learning_rate": 7.557565789473685e-06, "loss": 1.2148, "step": 1188 }, { "epoch": 0.24444901315789475, "grad_norm": 69.39627075195312, "learning_rate": 7.555509868421054e-06, "loss": 0.9453, "step": 1189 }, { "epoch": 0.24465460526315788, "grad_norm": 103.031982421875, "learning_rate": 7.553453947368422e-06, "loss": 1.4028, "step": 1190 }, { "epoch": 0.24486019736842105, "grad_norm": 31.25803565979004, "learning_rate": 7.5513980263157905e-06, "loss": 0.1851, "step": 1191 }, { "epoch": 0.2450657894736842, "grad_norm": 5.241404056549072, "learning_rate": 7.549342105263159e-06, "loss": 0.0701, "step": 1192 }, { "epoch": 0.24527138157894737, "grad_norm": 14.483295440673828, "learning_rate": 7.547286184210527e-06, "loss": 0.1223, "step": 1193 }, { "epoch": 0.24547697368421054, "grad_norm": 77.95394134521484, "learning_rate": 7.545230263157896e-06, "loss": 1.9844, "step": 1194 }, { "epoch": 0.24568256578947367, "grad_norm": 29.298973083496094, "learning_rate": 7.543174342105264e-06, "loss": 0.3503, "step": 1195 }, { "epoch": 0.24588815789473684, "grad_norm": 12.329487800598145, "learning_rate": 7.541118421052632e-06, "loss": 0.0963, "step": 1196 }, { "epoch": 0.24609375, "grad_norm": 23.38083839416504, "learning_rate": 7.5390625e-06, "loss": 0.2107, "step": 1197 }, { "epoch": 0.24629934210526316, "grad_norm": 20.118961334228516, "learning_rate": 7.537006578947369e-06, "loss": 0.1987, "step": 1198 }, { "epoch": 0.24650493421052633, "grad_norm": 19.375211715698242, "learning_rate": 7.534950657894738e-06, "loss": 0.1729, "step": 1199 }, { "epoch": 0.24671052631578946, "grad_norm": 24.0430908203125, "learning_rate": 7.532894736842106e-06, "loss": 0.1379, "step": 1200 }, { "epoch": 0.24691611842105263, "grad_norm": 35.60968017578125, "learning_rate": 7.530838815789474e-06, "loss": 0.6753, "step": 1201 }, { "epoch": 0.2471217105263158, "grad_norm": 57.855777740478516, "learning_rate": 7.528782894736842e-06, "loss": 0.5957, "step": 1202 }, { "epoch": 0.24732730263157895, "grad_norm": 39.349281311035156, "learning_rate": 7.526726973684211e-06, "loss": 0.5742, "step": 1203 }, { "epoch": 0.24753289473684212, "grad_norm": 134.94505310058594, "learning_rate": 7.524671052631579e-06, "loss": 4.5742, "step": 1204 }, { "epoch": 0.24773848684210525, "grad_norm": 54.461570739746094, "learning_rate": 7.5226151315789476e-06, "loss": 0.3975, "step": 1205 }, { "epoch": 0.24794407894736842, "grad_norm": 48.77339553833008, "learning_rate": 7.520559210526316e-06, "loss": 0.5283, "step": 1206 }, { "epoch": 0.24814967105263158, "grad_norm": 143.7308807373047, "learning_rate": 7.518503289473685e-06, "loss": 2.585, "step": 1207 }, { "epoch": 0.24835526315789475, "grad_norm": 137.59080505371094, "learning_rate": 7.516447368421054e-06, "loss": 2.2969, "step": 1208 }, { "epoch": 0.24856085526315788, "grad_norm": 35.400997161865234, "learning_rate": 7.5143914473684225e-06, "loss": 0.3398, "step": 1209 }, { "epoch": 0.24876644736842105, "grad_norm": 58.61001968383789, "learning_rate": 7.51233552631579e-06, "loss": 0.6111, "step": 1210 }, { "epoch": 0.2489720394736842, "grad_norm": 73.6641845703125, "learning_rate": 7.510279605263159e-06, "loss": 2.168, "step": 1211 }, { "epoch": 0.24917763157894737, "grad_norm": 17.001388549804688, "learning_rate": 7.508223684210527e-06, "loss": 0.1689, "step": 1212 }, { "epoch": 0.24938322368421054, "grad_norm": 48.171836853027344, "learning_rate": 7.506167763157896e-06, "loss": 0.4543, "step": 1213 }, { "epoch": 0.24958881578947367, "grad_norm": 83.96248626708984, "learning_rate": 7.504111842105264e-06, "loss": 0.9766, "step": 1214 }, { "epoch": 0.24979440789473684, "grad_norm": 50.66523361206055, "learning_rate": 7.502055921052632e-06, "loss": 0.7232, "step": 1215 }, { "epoch": 0.25, "grad_norm": 37.854671478271484, "learning_rate": 7.500000000000001e-06, "loss": 0.3311, "step": 1216 }, { "epoch": 0.25020559210526316, "grad_norm": 41.629581451416016, "learning_rate": 7.497944078947369e-06, "loss": 1.1172, "step": 1217 }, { "epoch": 0.2504111842105263, "grad_norm": 106.35006713867188, "learning_rate": 7.495888157894738e-06, "loss": 1.668, "step": 1218 }, { "epoch": 0.2506167763157895, "grad_norm": 72.49296569824219, "learning_rate": 7.4938322368421055e-06, "loss": 0.8691, "step": 1219 }, { "epoch": 0.25082236842105265, "grad_norm": 76.32626342773438, "learning_rate": 7.491776315789474e-06, "loss": 1.3105, "step": 1220 }, { "epoch": 0.25102796052631576, "grad_norm": 118.467041015625, "learning_rate": 7.489720394736842e-06, "loss": 1.126, "step": 1221 }, { "epoch": 0.2512335526315789, "grad_norm": 87.3779067993164, "learning_rate": 7.487664473684211e-06, "loss": 1.8594, "step": 1222 }, { "epoch": 0.2514391447368421, "grad_norm": 49.890628814697266, "learning_rate": 7.485608552631579e-06, "loss": 0.3291, "step": 1223 }, { "epoch": 0.25164473684210525, "grad_norm": 44.99102020263672, "learning_rate": 7.4835526315789475e-06, "loss": 0.5073, "step": 1224 }, { "epoch": 0.2518503289473684, "grad_norm": 39.726139068603516, "learning_rate": 7.481496710526316e-06, "loss": 0.3013, "step": 1225 }, { "epoch": 0.2520559210526316, "grad_norm": 65.66204071044922, "learning_rate": 7.479440789473685e-06, "loss": 1.4775, "step": 1226 }, { "epoch": 0.25226151315789475, "grad_norm": 18.49059295654297, "learning_rate": 7.477384868421054e-06, "loss": 0.2192, "step": 1227 }, { "epoch": 0.2524671052631579, "grad_norm": 33.54808044433594, "learning_rate": 7.475328947368422e-06, "loss": 0.3604, "step": 1228 }, { "epoch": 0.2526726973684211, "grad_norm": 19.183290481567383, "learning_rate": 7.47327302631579e-06, "loss": 0.2203, "step": 1229 }, { "epoch": 0.25287828947368424, "grad_norm": 25.13044548034668, "learning_rate": 7.471217105263159e-06, "loss": 0.4413, "step": 1230 }, { "epoch": 0.25308388157894735, "grad_norm": 39.525146484375, "learning_rate": 7.469161184210527e-06, "loss": 1.0107, "step": 1231 }, { "epoch": 0.2532894736842105, "grad_norm": 22.03388214111328, "learning_rate": 7.467105263157896e-06, "loss": 0.7463, "step": 1232 }, { "epoch": 0.2534950657894737, "grad_norm": 79.5975341796875, "learning_rate": 7.4650493421052635e-06, "loss": 1.2437, "step": 1233 }, { "epoch": 0.25370065789473684, "grad_norm": 26.237741470336914, "learning_rate": 7.462993421052632e-06, "loss": 0.406, "step": 1234 }, { "epoch": 0.25390625, "grad_norm": 113.60836029052734, "learning_rate": 7.460937500000001e-06, "loss": 0.8315, "step": 1235 }, { "epoch": 0.25411184210526316, "grad_norm": 30.453826904296875, "learning_rate": 7.458881578947369e-06, "loss": 0.3271, "step": 1236 }, { "epoch": 0.2543174342105263, "grad_norm": 52.26222610473633, "learning_rate": 7.4568256578947375e-06, "loss": 0.4961, "step": 1237 }, { "epoch": 0.2545230263157895, "grad_norm": 122.3512191772461, "learning_rate": 7.4547697368421054e-06, "loss": 1.1066, "step": 1238 }, { "epoch": 0.25472861842105265, "grad_norm": 181.7263641357422, "learning_rate": 7.452713815789474e-06, "loss": 2.8826, "step": 1239 }, { "epoch": 0.25493421052631576, "grad_norm": 54.304080963134766, "learning_rate": 7.450657894736842e-06, "loss": 0.9629, "step": 1240 }, { "epoch": 0.2551398026315789, "grad_norm": 136.34446716308594, "learning_rate": 7.448601973684211e-06, "loss": 1.9316, "step": 1241 }, { "epoch": 0.2553453947368421, "grad_norm": 38.374183654785156, "learning_rate": 7.4465460526315795e-06, "loss": 0.2349, "step": 1242 }, { "epoch": 0.25555098684210525, "grad_norm": 59.54668045043945, "learning_rate": 7.444490131578947e-06, "loss": 1.1915, "step": 1243 }, { "epoch": 0.2557565789473684, "grad_norm": 66.34526062011719, "learning_rate": 7.442434210526316e-06, "loss": 0.7422, "step": 1244 }, { "epoch": 0.2559621710526316, "grad_norm": 43.354148864746094, "learning_rate": 7.440378289473686e-06, "loss": 0.4536, "step": 1245 }, { "epoch": 0.25616776315789475, "grad_norm": 87.89765167236328, "learning_rate": 7.4383223684210536e-06, "loss": 0.77, "step": 1246 }, { "epoch": 0.2563733552631579, "grad_norm": 185.0482635498047, "learning_rate": 7.436266447368422e-06, "loss": 4.2266, "step": 1247 }, { "epoch": 0.2565789473684211, "grad_norm": 99.60914611816406, "learning_rate": 7.43421052631579e-06, "loss": 2.1797, "step": 1248 }, { "epoch": 0.25678453947368424, "grad_norm": 43.287837982177734, "learning_rate": 7.432154605263159e-06, "loss": 0.4199, "step": 1249 }, { "epoch": 0.25699013157894735, "grad_norm": 74.04180908203125, "learning_rate": 7.430098684210527e-06, "loss": 1.4746, "step": 1250 }, { "epoch": 0.2571957236842105, "grad_norm": 20.656484603881836, "learning_rate": 7.4280427631578955e-06, "loss": 0.0858, "step": 1251 }, { "epoch": 0.2574013157894737, "grad_norm": 32.7215576171875, "learning_rate": 7.425986842105264e-06, "loss": 0.4014, "step": 1252 }, { "epoch": 0.25760690789473684, "grad_norm": 29.713714599609375, "learning_rate": 7.423930921052632e-06, "loss": 0.1762, "step": 1253 }, { "epoch": 0.2578125, "grad_norm": 57.218772888183594, "learning_rate": 7.421875000000001e-06, "loss": 0.5876, "step": 1254 }, { "epoch": 0.25801809210526316, "grad_norm": 46.91562271118164, "learning_rate": 7.419819078947369e-06, "loss": 0.4717, "step": 1255 }, { "epoch": 0.2582236842105263, "grad_norm": 55.86995315551758, "learning_rate": 7.4177631578947374e-06, "loss": 1.2266, "step": 1256 }, { "epoch": 0.2584292763157895, "grad_norm": 32.99650955200195, "learning_rate": 7.415707236842105e-06, "loss": 0.5381, "step": 1257 }, { "epoch": 0.25863486842105265, "grad_norm": 85.31369018554688, "learning_rate": 7.413651315789474e-06, "loss": 0.4628, "step": 1258 }, { "epoch": 0.25884046052631576, "grad_norm": 12.302230834960938, "learning_rate": 7.411595394736843e-06, "loss": 0.095, "step": 1259 }, { "epoch": 0.2590460526315789, "grad_norm": 37.45296859741211, "learning_rate": 7.409539473684211e-06, "loss": 0.4629, "step": 1260 }, { "epoch": 0.2592516447368421, "grad_norm": 39.98808288574219, "learning_rate": 7.407483552631579e-06, "loss": 0.374, "step": 1261 }, { "epoch": 0.25945723684210525, "grad_norm": 39.38546371459961, "learning_rate": 7.405427631578947e-06, "loss": 0.5374, "step": 1262 }, { "epoch": 0.2596628289473684, "grad_norm": 67.47021484375, "learning_rate": 7.403371710526316e-06, "loss": 0.5946, "step": 1263 }, { "epoch": 0.2598684210526316, "grad_norm": 112.15380859375, "learning_rate": 7.4013157894736856e-06, "loss": 1.6484, "step": 1264 }, { "epoch": 0.26007401315789475, "grad_norm": 89.34825134277344, "learning_rate": 7.3992598684210534e-06, "loss": 0.8848, "step": 1265 }, { "epoch": 0.2602796052631579, "grad_norm": 84.31710052490234, "learning_rate": 7.397203947368422e-06, "loss": 1.0361, "step": 1266 }, { "epoch": 0.2604851973684211, "grad_norm": 109.64484405517578, "learning_rate": 7.39514802631579e-06, "loss": 2.5234, "step": 1267 }, { "epoch": 0.26069078947368424, "grad_norm": 42.871604919433594, "learning_rate": 7.393092105263159e-06, "loss": 0.5735, "step": 1268 }, { "epoch": 0.26089638157894735, "grad_norm": 206.36097717285156, "learning_rate": 7.3910361842105275e-06, "loss": 3.0977, "step": 1269 }, { "epoch": 0.2611019736842105, "grad_norm": 53.33968734741211, "learning_rate": 7.388980263157895e-06, "loss": 0.6309, "step": 1270 }, { "epoch": 0.2613075657894737, "grad_norm": 95.10393524169922, "learning_rate": 7.386924342105264e-06, "loss": 0.8113, "step": 1271 }, { "epoch": 0.26151315789473684, "grad_norm": 18.8743953704834, "learning_rate": 7.384868421052632e-06, "loss": 0.2646, "step": 1272 }, { "epoch": 0.26171875, "grad_norm": 76.83868408203125, "learning_rate": 7.382812500000001e-06, "loss": 1.1016, "step": 1273 }, { "epoch": 0.26192434210526316, "grad_norm": 22.942625045776367, "learning_rate": 7.380756578947369e-06, "loss": 0.1899, "step": 1274 }, { "epoch": 0.2621299342105263, "grad_norm": 43.74995422363281, "learning_rate": 7.378700657894737e-06, "loss": 0.4121, "step": 1275 }, { "epoch": 0.2623355263157895, "grad_norm": 64.34101104736328, "learning_rate": 7.376644736842106e-06, "loss": 0.7715, "step": 1276 }, { "epoch": 0.26254111842105265, "grad_norm": 86.35357666015625, "learning_rate": 7.374588815789474e-06, "loss": 0.7102, "step": 1277 }, { "epoch": 0.26274671052631576, "grad_norm": 85.45638275146484, "learning_rate": 7.372532894736843e-06, "loss": 0.9609, "step": 1278 }, { "epoch": 0.2629523026315789, "grad_norm": 106.51177215576172, "learning_rate": 7.3704769736842106e-06, "loss": 1.4297, "step": 1279 }, { "epoch": 0.2631578947368421, "grad_norm": 7.436676502227783, "learning_rate": 7.368421052631579e-06, "loss": 0.0837, "step": 1280 }, { "epoch": 0.26336348684210525, "grad_norm": 72.91498565673828, "learning_rate": 7.366365131578947e-06, "loss": 0.6953, "step": 1281 }, { "epoch": 0.2635690789473684, "grad_norm": 47.08981704711914, "learning_rate": 7.364309210526316e-06, "loss": 0.793, "step": 1282 }, { "epoch": 0.2637746710526316, "grad_norm": 45.55181884765625, "learning_rate": 7.3622532894736855e-06, "loss": 0.8208, "step": 1283 }, { "epoch": 0.26398026315789475, "grad_norm": 42.72093200683594, "learning_rate": 7.360197368421053e-06, "loss": 0.9224, "step": 1284 }, { "epoch": 0.2641858552631579, "grad_norm": 25.799732208251953, "learning_rate": 7.358141447368422e-06, "loss": 0.1406, "step": 1285 }, { "epoch": 0.2643914473684211, "grad_norm": 96.98381805419922, "learning_rate": 7.356085526315791e-06, "loss": 1.957, "step": 1286 }, { "epoch": 0.26459703947368424, "grad_norm": 175.0242462158203, "learning_rate": 7.354029605263159e-06, "loss": 1.4172, "step": 1287 }, { "epoch": 0.26480263157894735, "grad_norm": 15.866820335388184, "learning_rate": 7.351973684210527e-06, "loss": 0.0486, "step": 1288 }, { "epoch": 0.2650082236842105, "grad_norm": 32.13304138183594, "learning_rate": 7.349917763157895e-06, "loss": 0.2832, "step": 1289 }, { "epoch": 0.2652138157894737, "grad_norm": 277.136474609375, "learning_rate": 7.347861842105264e-06, "loss": 3.2422, "step": 1290 }, { "epoch": 0.26541940789473684, "grad_norm": 96.4759750366211, "learning_rate": 7.345805921052632e-06, "loss": 1.4297, "step": 1291 }, { "epoch": 0.265625, "grad_norm": 193.43252563476562, "learning_rate": 7.343750000000001e-06, "loss": 2.3867, "step": 1292 }, { "epoch": 0.26583059210526316, "grad_norm": 44.23900604248047, "learning_rate": 7.341694078947369e-06, "loss": 0.7319, "step": 1293 }, { "epoch": 0.2660361842105263, "grad_norm": 63.61642837524414, "learning_rate": 7.339638157894737e-06, "loss": 1.0179, "step": 1294 }, { "epoch": 0.2662417763157895, "grad_norm": 43.21895980834961, "learning_rate": 7.337582236842106e-06, "loss": 0.916, "step": 1295 }, { "epoch": 0.26644736842105265, "grad_norm": 41.42702102661133, "learning_rate": 7.335526315789474e-06, "loss": 0.8555, "step": 1296 }, { "epoch": 0.26665296052631576, "grad_norm": 16.094329833984375, "learning_rate": 7.3334703947368426e-06, "loss": 0.4221, "step": 1297 }, { "epoch": 0.2668585526315789, "grad_norm": 30.071882247924805, "learning_rate": 7.3314144736842105e-06, "loss": 0.1567, "step": 1298 }, { "epoch": 0.2670641447368421, "grad_norm": 45.848148345947266, "learning_rate": 7.329358552631579e-06, "loss": 0.4645, "step": 1299 }, { "epoch": 0.26726973684210525, "grad_norm": 8.830588340759277, "learning_rate": 7.327302631578948e-06, "loss": 0.0637, "step": 1300 }, { "epoch": 0.2674753289473684, "grad_norm": 35.70171356201172, "learning_rate": 7.325246710526316e-06, "loss": 0.2539, "step": 1301 }, { "epoch": 0.2676809210526316, "grad_norm": 128.7365264892578, "learning_rate": 7.323190789473685e-06, "loss": 4.3418, "step": 1302 }, { "epoch": 0.26788651315789475, "grad_norm": 98.72674560546875, "learning_rate": 7.321134868421054e-06, "loss": 0.9297, "step": 1303 }, { "epoch": 0.2680921052631579, "grad_norm": 166.69039916992188, "learning_rate": 7.319078947368422e-06, "loss": 3.25, "step": 1304 }, { "epoch": 0.2682976973684211, "grad_norm": 92.06898498535156, "learning_rate": 7.317023026315791e-06, "loss": 1.1396, "step": 1305 }, { "epoch": 0.26850328947368424, "grad_norm": 73.20455932617188, "learning_rate": 7.314967105263159e-06, "loss": 0.6597, "step": 1306 }, { "epoch": 0.26870888157894735, "grad_norm": 93.51766967773438, "learning_rate": 7.312911184210527e-06, "loss": 1.1895, "step": 1307 }, { "epoch": 0.2689144736842105, "grad_norm": 50.09158706665039, "learning_rate": 7.310855263157895e-06, "loss": 0.5072, "step": 1308 }, { "epoch": 0.2691200657894737, "grad_norm": 21.838245391845703, "learning_rate": 7.308799342105264e-06, "loss": 0.1921, "step": 1309 }, { "epoch": 0.26932565789473684, "grad_norm": 14.81062126159668, "learning_rate": 7.306743421052633e-06, "loss": 0.1133, "step": 1310 }, { "epoch": 0.26953125, "grad_norm": 31.906169891357422, "learning_rate": 7.3046875000000005e-06, "loss": 0.4277, "step": 1311 }, { "epoch": 0.26973684210526316, "grad_norm": 24.204654693603516, "learning_rate": 7.302631578947369e-06, "loss": 0.5178, "step": 1312 }, { "epoch": 0.2699424342105263, "grad_norm": 56.27023696899414, "learning_rate": 7.300575657894737e-06, "loss": 0.8086, "step": 1313 }, { "epoch": 0.2701480263157895, "grad_norm": 199.65264892578125, "learning_rate": 7.298519736842106e-06, "loss": 2.3682, "step": 1314 }, { "epoch": 0.27035361842105265, "grad_norm": 37.48896408081055, "learning_rate": 7.296463815789474e-06, "loss": 0.2783, "step": 1315 }, { "epoch": 0.27055921052631576, "grad_norm": 171.03179931640625, "learning_rate": 7.2944078947368425e-06, "loss": 2.6699, "step": 1316 }, { "epoch": 0.2707648026315789, "grad_norm": 32.48602294921875, "learning_rate": 7.29235197368421e-06, "loss": 0.2344, "step": 1317 }, { "epoch": 0.2709703947368421, "grad_norm": 25.04925537109375, "learning_rate": 7.290296052631579e-06, "loss": 0.7266, "step": 1318 }, { "epoch": 0.27117598684210525, "grad_norm": 66.02865600585938, "learning_rate": 7.288240131578948e-06, "loss": 0.9722, "step": 1319 }, { "epoch": 0.2713815789473684, "grad_norm": 61.98065185546875, "learning_rate": 7.286184210526316e-06, "loss": 1.1855, "step": 1320 }, { "epoch": 0.2715871710526316, "grad_norm": 143.75376892089844, "learning_rate": 7.284128289473685e-06, "loss": 2.3984, "step": 1321 }, { "epoch": 0.27179276315789475, "grad_norm": 129.5109100341797, "learning_rate": 7.282072368421054e-06, "loss": 1.2129, "step": 1322 }, { "epoch": 0.2719983552631579, "grad_norm": 80.25537872314453, "learning_rate": 7.280016447368422e-06, "loss": 1.011, "step": 1323 }, { "epoch": 0.2722039473684211, "grad_norm": 79.6856689453125, "learning_rate": 7.277960526315791e-06, "loss": 0.4846, "step": 1324 }, { "epoch": 0.27240953947368424, "grad_norm": 50.54153060913086, "learning_rate": 7.2759046052631585e-06, "loss": 0.4358, "step": 1325 }, { "epoch": 0.27261513157894735, "grad_norm": 42.60911560058594, "learning_rate": 7.273848684210527e-06, "loss": 1.1719, "step": 1326 }, { "epoch": 0.2728207236842105, "grad_norm": 140.49679565429688, "learning_rate": 7.271792763157895e-06, "loss": 3.1719, "step": 1327 }, { "epoch": 0.2730263157894737, "grad_norm": 34.19179153442383, "learning_rate": 7.269736842105264e-06, "loss": 0.6943, "step": 1328 }, { "epoch": 0.27323190789473684, "grad_norm": 32.9194221496582, "learning_rate": 7.2676809210526325e-06, "loss": 0.223, "step": 1329 }, { "epoch": 0.2734375, "grad_norm": 82.0766830444336, "learning_rate": 7.265625e-06, "loss": 1.293, "step": 1330 }, { "epoch": 0.27364309210526316, "grad_norm": 46.51140594482422, "learning_rate": 7.263569078947369e-06, "loss": 0.8672, "step": 1331 }, { "epoch": 0.2738486842105263, "grad_norm": 38.3123664855957, "learning_rate": 7.261513157894737e-06, "loss": 0.4355, "step": 1332 }, { "epoch": 0.2740542763157895, "grad_norm": 110.29784393310547, "learning_rate": 7.259457236842106e-06, "loss": 2.6914, "step": 1333 }, { "epoch": 0.27425986842105265, "grad_norm": 45.59654235839844, "learning_rate": 7.257401315789474e-06, "loss": 0.6299, "step": 1334 }, { "epoch": 0.27446546052631576, "grad_norm": 168.8142852783203, "learning_rate": 7.255345394736842e-06, "loss": 3.2817, "step": 1335 }, { "epoch": 0.2746710526315789, "grad_norm": 53.27945327758789, "learning_rate": 7.253289473684211e-06, "loss": 0.6641, "step": 1336 }, { "epoch": 0.2748766447368421, "grad_norm": 26.350004196166992, "learning_rate": 7.251233552631579e-06, "loss": 0.3687, "step": 1337 }, { "epoch": 0.27508223684210525, "grad_norm": 49.09874725341797, "learning_rate": 7.249177631578948e-06, "loss": 0.4453, "step": 1338 }, { "epoch": 0.2752878289473684, "grad_norm": 124.85372924804688, "learning_rate": 7.247121710526316e-06, "loss": 1.7773, "step": 1339 }, { "epoch": 0.2754934210526316, "grad_norm": 98.42041778564453, "learning_rate": 7.245065789473685e-06, "loss": 1.4692, "step": 1340 }, { "epoch": 0.27569901315789475, "grad_norm": 131.2499542236328, "learning_rate": 7.243009868421054e-06, "loss": 2.3086, "step": 1341 }, { "epoch": 0.2759046052631579, "grad_norm": 147.41195678710938, "learning_rate": 7.240953947368422e-06, "loss": 3.9688, "step": 1342 }, { "epoch": 0.2761101973684211, "grad_norm": 69.39802551269531, "learning_rate": 7.2388980263157905e-06, "loss": 1.0352, "step": 1343 }, { "epoch": 0.27631578947368424, "grad_norm": 141.8444366455078, "learning_rate": 7.236842105263158e-06, "loss": 1.6719, "step": 1344 }, { "epoch": 0.27652138157894735, "grad_norm": 19.314552307128906, "learning_rate": 7.234786184210527e-06, "loss": 0.1628, "step": 1345 }, { "epoch": 0.2767269736842105, "grad_norm": 48.27471160888672, "learning_rate": 7.232730263157896e-06, "loss": 0.5776, "step": 1346 }, { "epoch": 0.2769325657894737, "grad_norm": 50.435272216796875, "learning_rate": 7.230674342105264e-06, "loss": 0.9199, "step": 1347 }, { "epoch": 0.27713815789473684, "grad_norm": 72.74262237548828, "learning_rate": 7.2286184210526324e-06, "loss": 1.2598, "step": 1348 }, { "epoch": 0.27734375, "grad_norm": 150.28378295898438, "learning_rate": 7.2265625e-06, "loss": 2.626, "step": 1349 }, { "epoch": 0.27754934210526316, "grad_norm": 44.10841369628906, "learning_rate": 7.224506578947369e-06, "loss": 0.4365, "step": 1350 }, { "epoch": 0.2777549342105263, "grad_norm": 34.24583435058594, "learning_rate": 7.222450657894737e-06, "loss": 0.6748, "step": 1351 }, { "epoch": 0.2779605263157895, "grad_norm": 59.492671966552734, "learning_rate": 7.220394736842106e-06, "loss": 1.0742, "step": 1352 }, { "epoch": 0.27816611842105265, "grad_norm": 29.16636085510254, "learning_rate": 7.218338815789474e-06, "loss": 0.3945, "step": 1353 }, { "epoch": 0.27837171052631576, "grad_norm": 63.58211135864258, "learning_rate": 7.216282894736842e-06, "loss": 0.9727, "step": 1354 }, { "epoch": 0.2785773026315789, "grad_norm": 56.00787353515625, "learning_rate": 7.214226973684211e-06, "loss": 0.6826, "step": 1355 }, { "epoch": 0.2787828947368421, "grad_norm": 78.60523223876953, "learning_rate": 7.212171052631579e-06, "loss": 0.9902, "step": 1356 }, { "epoch": 0.27898848684210525, "grad_norm": 7.072010517120361, "learning_rate": 7.210115131578948e-06, "loss": 0.0641, "step": 1357 }, { "epoch": 0.2791940789473684, "grad_norm": 93.62564849853516, "learning_rate": 7.2080592105263155e-06, "loss": 2.0834, "step": 1358 }, { "epoch": 0.2793996710526316, "grad_norm": 56.16436767578125, "learning_rate": 7.206003289473685e-06, "loss": 0.502, "step": 1359 }, { "epoch": 0.27960526315789475, "grad_norm": 32.820743560791016, "learning_rate": 7.203947368421054e-06, "loss": 0.5156, "step": 1360 }, { "epoch": 0.2798108552631579, "grad_norm": 94.68168640136719, "learning_rate": 7.201891447368422e-06, "loss": 1.0078, "step": 1361 }, { "epoch": 0.2800164473684211, "grad_norm": 142.84051513671875, "learning_rate": 7.19983552631579e-06, "loss": 2.7031, "step": 1362 }, { "epoch": 0.28022203947368424, "grad_norm": 54.99312210083008, "learning_rate": 7.197779605263159e-06, "loss": 1.0068, "step": 1363 }, { "epoch": 0.28042763157894735, "grad_norm": 45.040870666503906, "learning_rate": 7.195723684210527e-06, "loss": 0.416, "step": 1364 }, { "epoch": 0.2806332236842105, "grad_norm": 43.667327880859375, "learning_rate": 7.193667763157896e-06, "loss": 0.2539, "step": 1365 }, { "epoch": 0.2808388157894737, "grad_norm": 61.859100341796875, "learning_rate": 7.191611842105264e-06, "loss": 1.7539, "step": 1366 }, { "epoch": 0.28104440789473684, "grad_norm": 98.35357666015625, "learning_rate": 7.189555921052632e-06, "loss": 1.4609, "step": 1367 }, { "epoch": 0.28125, "grad_norm": 50.56107711791992, "learning_rate": 7.1875e-06, "loss": 0.9043, "step": 1368 }, { "epoch": 0.28145559210526316, "grad_norm": 68.0674819946289, "learning_rate": 7.185444078947369e-06, "loss": 0.8307, "step": 1369 }, { "epoch": 0.2816611842105263, "grad_norm": 26.283954620361328, "learning_rate": 7.183388157894738e-06, "loss": 0.1935, "step": 1370 }, { "epoch": 0.2818667763157895, "grad_norm": 31.897598266601562, "learning_rate": 7.1813322368421056e-06, "loss": 0.2374, "step": 1371 }, { "epoch": 0.28207236842105265, "grad_norm": 103.25392150878906, "learning_rate": 7.179276315789474e-06, "loss": 1.4457, "step": 1372 }, { "epoch": 0.28227796052631576, "grad_norm": 73.79505157470703, "learning_rate": 7.177220394736842e-06, "loss": 0.9746, "step": 1373 }, { "epoch": 0.2824835526315789, "grad_norm": 30.540573120117188, "learning_rate": 7.175164473684211e-06, "loss": 0.3926, "step": 1374 }, { "epoch": 0.2826891447368421, "grad_norm": 16.28680992126465, "learning_rate": 7.173108552631579e-06, "loss": 0.135, "step": 1375 }, { "epoch": 0.28289473684210525, "grad_norm": 28.812646865844727, "learning_rate": 7.1710526315789475e-06, "loss": 0.4277, "step": 1376 }, { "epoch": 0.2831003289473684, "grad_norm": 107.8692855834961, "learning_rate": 7.168996710526316e-06, "loss": 0.5078, "step": 1377 }, { "epoch": 0.2833059210526316, "grad_norm": 74.16141510009766, "learning_rate": 7.166940789473685e-06, "loss": 1.2656, "step": 1378 }, { "epoch": 0.28351151315789475, "grad_norm": 5.15885066986084, "learning_rate": 7.164884868421054e-06, "loss": 0.0269, "step": 1379 }, { "epoch": 0.2837171052631579, "grad_norm": 32.029693603515625, "learning_rate": 7.162828947368422e-06, "loss": 0.6143, "step": 1380 }, { "epoch": 0.2839226973684211, "grad_norm": 88.78885650634766, "learning_rate": 7.16077302631579e-06, "loss": 1.022, "step": 1381 }, { "epoch": 0.28412828947368424, "grad_norm": 41.907005310058594, "learning_rate": 7.158717105263159e-06, "loss": 0.5532, "step": 1382 }, { "epoch": 0.28433388157894735, "grad_norm": 62.21117401123047, "learning_rate": 7.156661184210527e-06, "loss": 0.8015, "step": 1383 }, { "epoch": 0.2845394736842105, "grad_norm": 60.173606872558594, "learning_rate": 7.154605263157896e-06, "loss": 0.6602, "step": 1384 }, { "epoch": 0.2847450657894737, "grad_norm": 74.30499267578125, "learning_rate": 7.1525493421052635e-06, "loss": 1.4672, "step": 1385 }, { "epoch": 0.28495065789473684, "grad_norm": 14.422243118286133, "learning_rate": 7.150493421052632e-06, "loss": 0.1035, "step": 1386 }, { "epoch": 0.28515625, "grad_norm": 77.6588363647461, "learning_rate": 7.148437500000001e-06, "loss": 0.9015, "step": 1387 }, { "epoch": 0.28536184210526316, "grad_norm": 42.39907455444336, "learning_rate": 7.146381578947369e-06, "loss": 0.585, "step": 1388 }, { "epoch": 0.2855674342105263, "grad_norm": 81.23006439208984, "learning_rate": 7.1443256578947376e-06, "loss": 1.7891, "step": 1389 }, { "epoch": 0.2857730263157895, "grad_norm": 177.25648498535156, "learning_rate": 7.1422697368421055e-06, "loss": 2.5078, "step": 1390 }, { "epoch": 0.28597861842105265, "grad_norm": 116.13312530517578, "learning_rate": 7.140213815789474e-06, "loss": 2.0195, "step": 1391 }, { "epoch": 0.28618421052631576, "grad_norm": 42.76192092895508, "learning_rate": 7.138157894736842e-06, "loss": 0.7383, "step": 1392 }, { "epoch": 0.2863898026315789, "grad_norm": 52.15032958984375, "learning_rate": 7.136101973684211e-06, "loss": 0.9717, "step": 1393 }, { "epoch": 0.2865953947368421, "grad_norm": 171.4599151611328, "learning_rate": 7.1340460526315795e-06, "loss": 4.2578, "step": 1394 }, { "epoch": 0.28680098684210525, "grad_norm": 143.01983642578125, "learning_rate": 7.131990131578947e-06, "loss": 0.7827, "step": 1395 }, { "epoch": 0.2870065789473684, "grad_norm": 104.63043212890625, "learning_rate": 7.129934210526316e-06, "loss": 1.3008, "step": 1396 }, { "epoch": 0.2872121710526316, "grad_norm": 10.88134765625, "learning_rate": 7.127878289473686e-06, "loss": 0.0338, "step": 1397 }, { "epoch": 0.28741776315789475, "grad_norm": 39.55748748779297, "learning_rate": 7.125822368421054e-06, "loss": 0.4819, "step": 1398 }, { "epoch": 0.2876233552631579, "grad_norm": 43.5048713684082, "learning_rate": 7.123766447368422e-06, "loss": 0.3257, "step": 1399 }, { "epoch": 0.2878289473684211, "grad_norm": 22.466638565063477, "learning_rate": 7.12171052631579e-06, "loss": 0.2803, "step": 1400 }, { "epoch": 0.28803453947368424, "grad_norm": 135.86322021484375, "learning_rate": 7.119654605263159e-06, "loss": 2.5391, "step": 1401 }, { "epoch": 0.28824013157894735, "grad_norm": 13.339402198791504, "learning_rate": 7.117598684210527e-06, "loss": 0.1455, "step": 1402 }, { "epoch": 0.2884457236842105, "grad_norm": 122.66312408447266, "learning_rate": 7.1155427631578955e-06, "loss": 4.2323, "step": 1403 }, { "epoch": 0.2886513157894737, "grad_norm": 84.91791534423828, "learning_rate": 7.113486842105264e-06, "loss": 1.4219, "step": 1404 }, { "epoch": 0.28885690789473684, "grad_norm": 104.05675506591797, "learning_rate": 7.111430921052632e-06, "loss": 1.2305, "step": 1405 }, { "epoch": 0.2890625, "grad_norm": 39.96343231201172, "learning_rate": 7.109375000000001e-06, "loss": 0.4219, "step": 1406 }, { "epoch": 0.28926809210526316, "grad_norm": 64.79421997070312, "learning_rate": 7.107319078947369e-06, "loss": 1.1172, "step": 1407 }, { "epoch": 0.2894736842105263, "grad_norm": 138.5366668701172, "learning_rate": 7.1052631578947375e-06, "loss": 4.0591, "step": 1408 }, { "epoch": 0.2896792763157895, "grad_norm": 61.63038635253906, "learning_rate": 7.103207236842105e-06, "loss": 0.7734, "step": 1409 }, { "epoch": 0.28988486842105265, "grad_norm": 139.41616821289062, "learning_rate": 7.101151315789474e-06, "loss": 1.5669, "step": 1410 }, { "epoch": 0.29009046052631576, "grad_norm": 68.10540008544922, "learning_rate": 7.099095394736843e-06, "loss": 1.4355, "step": 1411 }, { "epoch": 0.2902960526315789, "grad_norm": 84.15885162353516, "learning_rate": 7.097039473684211e-06, "loss": 0.6113, "step": 1412 }, { "epoch": 0.2905016447368421, "grad_norm": 33.299949645996094, "learning_rate": 7.094983552631579e-06, "loss": 0.3635, "step": 1413 }, { "epoch": 0.29070723684210525, "grad_norm": 42.442832946777344, "learning_rate": 7.092927631578947e-06, "loss": 0.4824, "step": 1414 }, { "epoch": 0.2909128289473684, "grad_norm": 40.567352294921875, "learning_rate": 7.090871710526316e-06, "loss": 0.6602, "step": 1415 }, { "epoch": 0.2911184210526316, "grad_norm": 32.84964370727539, "learning_rate": 7.088815789473686e-06, "loss": 0.3486, "step": 1416 }, { "epoch": 0.29132401315789475, "grad_norm": 103.05779266357422, "learning_rate": 7.0867598684210535e-06, "loss": 0.832, "step": 1417 }, { "epoch": 0.2915296052631579, "grad_norm": 48.85757827758789, "learning_rate": 7.084703947368422e-06, "loss": 0.4727, "step": 1418 }, { "epoch": 0.2917351973684211, "grad_norm": 15.71153736114502, "learning_rate": 7.08264802631579e-06, "loss": 0.0649, "step": 1419 }, { "epoch": 0.29194078947368424, "grad_norm": 56.21582794189453, "learning_rate": 7.080592105263159e-06, "loss": 1.1426, "step": 1420 }, { "epoch": 0.29214638157894735, "grad_norm": 57.05986785888672, "learning_rate": 7.0785361842105275e-06, "loss": 0.6406, "step": 1421 }, { "epoch": 0.2923519736842105, "grad_norm": 30.14991569519043, "learning_rate": 7.076480263157895e-06, "loss": 0.2773, "step": 1422 }, { "epoch": 0.2925575657894737, "grad_norm": 39.81775665283203, "learning_rate": 7.074424342105264e-06, "loss": 0.3296, "step": 1423 }, { "epoch": 0.29276315789473684, "grad_norm": 94.3897933959961, "learning_rate": 7.072368421052632e-06, "loss": 0.9351, "step": 1424 }, { "epoch": 0.29296875, "grad_norm": 50.50587844848633, "learning_rate": 7.070312500000001e-06, "loss": 0.8203, "step": 1425 }, { "epoch": 0.29317434210526316, "grad_norm": 16.34046745300293, "learning_rate": 7.068256578947369e-06, "loss": 0.1301, "step": 1426 }, { "epoch": 0.2933799342105263, "grad_norm": 78.02462005615234, "learning_rate": 7.066200657894737e-06, "loss": 1.5742, "step": 1427 }, { "epoch": 0.2935855263157895, "grad_norm": 42.30268478393555, "learning_rate": 7.064144736842105e-06, "loss": 0.7217, "step": 1428 }, { "epoch": 0.29379111842105265, "grad_norm": 35.656280517578125, "learning_rate": 7.062088815789474e-06, "loss": 0.2187, "step": 1429 }, { "epoch": 0.29399671052631576, "grad_norm": 43.5911865234375, "learning_rate": 7.060032894736843e-06, "loss": 0.439, "step": 1430 }, { "epoch": 0.2942023026315789, "grad_norm": 67.81900787353516, "learning_rate": 7.057976973684211e-06, "loss": 0.6631, "step": 1431 }, { "epoch": 0.2944078947368421, "grad_norm": 107.8806381225586, "learning_rate": 7.055921052631579e-06, "loss": 1.3901, "step": 1432 }, { "epoch": 0.29461348684210525, "grad_norm": 105.67761993408203, "learning_rate": 7.053865131578947e-06, "loss": 1.0527, "step": 1433 }, { "epoch": 0.2948190789473684, "grad_norm": 115.29364776611328, "learning_rate": 7.051809210526316e-06, "loss": 1.0996, "step": 1434 }, { "epoch": 0.2950246710526316, "grad_norm": 215.97531127929688, "learning_rate": 7.0497532894736855e-06, "loss": 1.6179, "step": 1435 }, { "epoch": 0.29523026315789475, "grad_norm": 95.88826751708984, "learning_rate": 7.047697368421053e-06, "loss": 1.0166, "step": 1436 }, { "epoch": 0.2954358552631579, "grad_norm": 126.1279296875, "learning_rate": 7.045641447368422e-06, "loss": 1.5977, "step": 1437 }, { "epoch": 0.2956414473684211, "grad_norm": 20.201759338378906, "learning_rate": 7.04358552631579e-06, "loss": 0.1865, "step": 1438 }, { "epoch": 0.29584703947368424, "grad_norm": 117.72960662841797, "learning_rate": 7.041529605263159e-06, "loss": 1.4492, "step": 1439 }, { "epoch": 0.29605263157894735, "grad_norm": 46.51884841918945, "learning_rate": 7.0394736842105274e-06, "loss": 0.3906, "step": 1440 }, { "epoch": 0.2962582236842105, "grad_norm": 83.38851165771484, "learning_rate": 7.037417763157895e-06, "loss": 0.8193, "step": 1441 }, { "epoch": 0.2964638157894737, "grad_norm": 41.4078254699707, "learning_rate": 7.035361842105264e-06, "loss": 0.6133, "step": 1442 }, { "epoch": 0.29666940789473684, "grad_norm": 9.15371036529541, "learning_rate": 7.033305921052632e-06, "loss": 0.0958, "step": 1443 }, { "epoch": 0.296875, "grad_norm": 152.39598083496094, "learning_rate": 7.031250000000001e-06, "loss": 1.0645, "step": 1444 }, { "epoch": 0.29708059210526316, "grad_norm": 20.20143699645996, "learning_rate": 7.0291940789473685e-06, "loss": 0.1558, "step": 1445 }, { "epoch": 0.2972861842105263, "grad_norm": 27.154861450195312, "learning_rate": 7.027138157894737e-06, "loss": 0.2683, "step": 1446 }, { "epoch": 0.2974917763157895, "grad_norm": 28.920854568481445, "learning_rate": 7.025082236842106e-06, "loss": 0.1536, "step": 1447 }, { "epoch": 0.29769736842105265, "grad_norm": 49.804718017578125, "learning_rate": 7.023026315789474e-06, "loss": 0.4238, "step": 1448 }, { "epoch": 0.29790296052631576, "grad_norm": 72.40191650390625, "learning_rate": 7.020970394736843e-06, "loss": 0.7603, "step": 1449 }, { "epoch": 0.2981085526315789, "grad_norm": 39.28083419799805, "learning_rate": 7.0189144736842105e-06, "loss": 0.2944, "step": 1450 }, { "epoch": 0.2983141447368421, "grad_norm": 99.3991928100586, "learning_rate": 7.016858552631579e-06, "loss": 1.4102, "step": 1451 }, { "epoch": 0.29851973684210525, "grad_norm": 38.61861801147461, "learning_rate": 7.014802631578947e-06, "loss": 0.3828, "step": 1452 }, { "epoch": 0.2987253289473684, "grad_norm": 110.9914779663086, "learning_rate": 7.012746710526316e-06, "loss": 2.0078, "step": 1453 }, { "epoch": 0.2989309210526316, "grad_norm": 144.4355926513672, "learning_rate": 7.010690789473685e-06, "loss": 1.2119, "step": 1454 }, { "epoch": 0.29913651315789475, "grad_norm": 105.60731506347656, "learning_rate": 7.008634868421053e-06, "loss": 1.0439, "step": 1455 }, { "epoch": 0.2993421052631579, "grad_norm": 181.28378295898438, "learning_rate": 7.006578947368422e-06, "loss": 1.7383, "step": 1456 }, { "epoch": 0.2995476973684211, "grad_norm": 53.406166076660156, "learning_rate": 7.004523026315791e-06, "loss": 0.625, "step": 1457 }, { "epoch": 0.29975328947368424, "grad_norm": 28.806137084960938, "learning_rate": 7.002467105263159e-06, "loss": 0.1833, "step": 1458 }, { "epoch": 0.29995888157894735, "grad_norm": 14.96003532409668, "learning_rate": 7.000411184210527e-06, "loss": 0.0547, "step": 1459 }, { "epoch": 0.3001644736842105, "grad_norm": 145.9787139892578, "learning_rate": 6.998355263157895e-06, "loss": 1.6836, "step": 1460 }, { "epoch": 0.3003700657894737, "grad_norm": 43.882755279541016, "learning_rate": 6.996299342105264e-06, "loss": 0.4136, "step": 1461 }, { "epoch": 0.30057565789473684, "grad_norm": 36.68684005737305, "learning_rate": 6.994243421052632e-06, "loss": 0.3564, "step": 1462 }, { "epoch": 0.30078125, "grad_norm": 106.03460693359375, "learning_rate": 6.9921875000000006e-06, "loss": 1.1719, "step": 1463 }, { "epoch": 0.30098684210526316, "grad_norm": 43.88253402709961, "learning_rate": 6.990131578947369e-06, "loss": 0.7305, "step": 1464 }, { "epoch": 0.3011924342105263, "grad_norm": 60.29634475708008, "learning_rate": 6.988075657894737e-06, "loss": 0.9824, "step": 1465 }, { "epoch": 0.3013980263157895, "grad_norm": 99.68193817138672, "learning_rate": 6.986019736842106e-06, "loss": 1.5127, "step": 1466 }, { "epoch": 0.30160361842105265, "grad_norm": 68.05619049072266, "learning_rate": 6.983963815789474e-06, "loss": 0.8984, "step": 1467 }, { "epoch": 0.30180921052631576, "grad_norm": 20.977848052978516, "learning_rate": 6.9819078947368425e-06, "loss": 0.1038, "step": 1468 }, { "epoch": 0.3020148026315789, "grad_norm": 150.08819580078125, "learning_rate": 6.97985197368421e-06, "loss": 2.4326, "step": 1469 }, { "epoch": 0.3022203947368421, "grad_norm": 118.20062255859375, "learning_rate": 6.977796052631579e-06, "loss": 1.1533, "step": 1470 }, { "epoch": 0.30242598684210525, "grad_norm": 3.627790689468384, "learning_rate": 6.975740131578948e-06, "loss": 0.0282, "step": 1471 }, { "epoch": 0.3026315789473684, "grad_norm": 76.19268035888672, "learning_rate": 6.973684210526316e-06, "loss": 0.6191, "step": 1472 }, { "epoch": 0.3028371710526316, "grad_norm": 60.10389709472656, "learning_rate": 6.971628289473685e-06, "loss": 0.6406, "step": 1473 }, { "epoch": 0.30304276315789475, "grad_norm": 41.81595230102539, "learning_rate": 6.969572368421054e-06, "loss": 0.3799, "step": 1474 }, { "epoch": 0.3032483552631579, "grad_norm": 99.0905532836914, "learning_rate": 6.967516447368422e-06, "loss": 1.3281, "step": 1475 }, { "epoch": 0.3034539473684211, "grad_norm": 222.89308166503906, "learning_rate": 6.965460526315791e-06, "loss": 2.9141, "step": 1476 }, { "epoch": 0.30365953947368424, "grad_norm": 53.5798454284668, "learning_rate": 6.9634046052631585e-06, "loss": 0.5603, "step": 1477 }, { "epoch": 0.30386513157894735, "grad_norm": 62.619850158691406, "learning_rate": 6.961348684210527e-06, "loss": 0.5886, "step": 1478 }, { "epoch": 0.3040707236842105, "grad_norm": 19.665138244628906, "learning_rate": 6.959292763157895e-06, "loss": 0.1564, "step": 1479 }, { "epoch": 0.3042763157894737, "grad_norm": 27.344942092895508, "learning_rate": 6.957236842105264e-06, "loss": 0.271, "step": 1480 }, { "epoch": 0.30448190789473684, "grad_norm": 124.30885314941406, "learning_rate": 6.9551809210526326e-06, "loss": 1.8398, "step": 1481 }, { "epoch": 0.3046875, "grad_norm": 84.9030532836914, "learning_rate": 6.9531250000000004e-06, "loss": 1.5391, "step": 1482 }, { "epoch": 0.30489309210526316, "grad_norm": 105.01704406738281, "learning_rate": 6.951069078947369e-06, "loss": 1.0898, "step": 1483 }, { "epoch": 0.3050986842105263, "grad_norm": 69.89837646484375, "learning_rate": 6.949013157894737e-06, "loss": 0.6504, "step": 1484 }, { "epoch": 0.3053042763157895, "grad_norm": 50.45353317260742, "learning_rate": 6.946957236842106e-06, "loss": 0.4653, "step": 1485 }, { "epoch": 0.30550986842105265, "grad_norm": 9.4555082321167, "learning_rate": 6.944901315789474e-06, "loss": 0.0703, "step": 1486 }, { "epoch": 0.30571546052631576, "grad_norm": 188.95675659179688, "learning_rate": 6.942845394736842e-06, "loss": 1.6504, "step": 1487 }, { "epoch": 0.3059210526315789, "grad_norm": 79.68267822265625, "learning_rate": 6.940789473684211e-06, "loss": 0.9805, "step": 1488 }, { "epoch": 0.3061266447368421, "grad_norm": 122.16349792480469, "learning_rate": 6.938733552631579e-06, "loss": 1.949, "step": 1489 }, { "epoch": 0.30633223684210525, "grad_norm": 58.515533447265625, "learning_rate": 6.936677631578948e-06, "loss": 1.2681, "step": 1490 }, { "epoch": 0.3065378289473684, "grad_norm": 96.61385345458984, "learning_rate": 6.934621710526316e-06, "loss": 1.0645, "step": 1491 }, { "epoch": 0.3067434210526316, "grad_norm": 41.27313232421875, "learning_rate": 6.932565789473685e-06, "loss": 1.0518, "step": 1492 }, { "epoch": 0.30694901315789475, "grad_norm": 26.7445125579834, "learning_rate": 6.930509868421054e-06, "loss": 0.3097, "step": 1493 }, { "epoch": 0.3071546052631579, "grad_norm": 82.87490844726562, "learning_rate": 6.928453947368422e-06, "loss": 1.3242, "step": 1494 }, { "epoch": 0.3073601973684211, "grad_norm": 108.46993255615234, "learning_rate": 6.9263980263157905e-06, "loss": 1.1824, "step": 1495 }, { "epoch": 0.30756578947368424, "grad_norm": 46.71767807006836, "learning_rate": 6.924342105263158e-06, "loss": 0.8564, "step": 1496 }, { "epoch": 0.30777138157894735, "grad_norm": 30.394990921020508, "learning_rate": 6.922286184210527e-06, "loss": 0.4014, "step": 1497 }, { "epoch": 0.3079769736842105, "grad_norm": 50.294315338134766, "learning_rate": 6.920230263157896e-06, "loss": 0.5107, "step": 1498 }, { "epoch": 0.3081825657894737, "grad_norm": 94.36602783203125, "learning_rate": 6.918174342105264e-06, "loss": 0.8809, "step": 1499 }, { "epoch": 0.30838815789473684, "grad_norm": 69.90850830078125, "learning_rate": 6.9161184210526325e-06, "loss": 0.9473, "step": 1500 }, { "epoch": 0.30859375, "grad_norm": 45.8420524597168, "learning_rate": 6.9140625e-06, "loss": 0.3345, "step": 1501 }, { "epoch": 0.30879934210526316, "grad_norm": 71.61844635009766, "learning_rate": 6.912006578947369e-06, "loss": 1.0813, "step": 1502 }, { "epoch": 0.3090049342105263, "grad_norm": 58.11845016479492, "learning_rate": 6.909950657894737e-06, "loss": 0.5104, "step": 1503 }, { "epoch": 0.3092105263157895, "grad_norm": 26.336774826049805, "learning_rate": 6.907894736842106e-06, "loss": 0.2444, "step": 1504 }, { "epoch": 0.30941611842105265, "grad_norm": 34.764617919921875, "learning_rate": 6.905838815789474e-06, "loss": 0.4514, "step": 1505 }, { "epoch": 0.30962171052631576, "grad_norm": 139.77304077148438, "learning_rate": 6.903782894736842e-06, "loss": 2.6917, "step": 1506 }, { "epoch": 0.3098273026315789, "grad_norm": 72.95701599121094, "learning_rate": 6.901726973684211e-06, "loss": 1.2891, "step": 1507 }, { "epoch": 0.3100328947368421, "grad_norm": 58.5020751953125, "learning_rate": 6.899671052631579e-06, "loss": 1.0944, "step": 1508 }, { "epoch": 0.31023848684210525, "grad_norm": 14.310029029846191, "learning_rate": 6.897615131578948e-06, "loss": 0.1224, "step": 1509 }, { "epoch": 0.3104440789473684, "grad_norm": 90.10921478271484, "learning_rate": 6.8955592105263155e-06, "loss": 1.4648, "step": 1510 }, { "epoch": 0.3106496710526316, "grad_norm": 43.4324836730957, "learning_rate": 6.893503289473685e-06, "loss": 0.8301, "step": 1511 }, { "epoch": 0.31085526315789475, "grad_norm": 83.97347259521484, "learning_rate": 6.891447368421054e-06, "loss": 1.207, "step": 1512 }, { "epoch": 0.3110608552631579, "grad_norm": 38.65486145019531, "learning_rate": 6.889391447368422e-06, "loss": 0.334, "step": 1513 }, { "epoch": 0.3112664473684211, "grad_norm": 47.9266357421875, "learning_rate": 6.88733552631579e-06, "loss": 0.4673, "step": 1514 }, { "epoch": 0.31147203947368424, "grad_norm": 91.01927185058594, "learning_rate": 6.885279605263159e-06, "loss": 1.5547, "step": 1515 }, { "epoch": 0.31167763157894735, "grad_norm": 187.78387451171875, "learning_rate": 6.883223684210527e-06, "loss": 2.5977, "step": 1516 }, { "epoch": 0.3118832236842105, "grad_norm": 26.70853042602539, "learning_rate": 6.881167763157896e-06, "loss": 0.3025, "step": 1517 }, { "epoch": 0.3120888157894737, "grad_norm": 152.0382080078125, "learning_rate": 6.879111842105264e-06, "loss": 3.3521, "step": 1518 }, { "epoch": 0.31229440789473684, "grad_norm": 124.91081237792969, "learning_rate": 6.877055921052632e-06, "loss": 1.8784, "step": 1519 }, { "epoch": 0.3125, "grad_norm": 46.93314743041992, "learning_rate": 6.875e-06, "loss": 1.2227, "step": 1520 }, { "epoch": 0.31270559210526316, "grad_norm": 38.073890686035156, "learning_rate": 6.872944078947369e-06, "loss": 0.6426, "step": 1521 }, { "epoch": 0.3129111842105263, "grad_norm": 45.50654983520508, "learning_rate": 6.870888157894738e-06, "loss": 1.3828, "step": 1522 }, { "epoch": 0.3131167763157895, "grad_norm": 101.17218780517578, "learning_rate": 6.868832236842106e-06, "loss": 2.832, "step": 1523 }, { "epoch": 0.31332236842105265, "grad_norm": 49.302059173583984, "learning_rate": 6.866776315789474e-06, "loss": 0.4756, "step": 1524 }, { "epoch": 0.31352796052631576, "grad_norm": 60.79299545288086, "learning_rate": 6.864720394736842e-06, "loss": 1.0391, "step": 1525 }, { "epoch": 0.3137335526315789, "grad_norm": 42.17608642578125, "learning_rate": 6.862664473684211e-06, "loss": 0.9014, "step": 1526 }, { "epoch": 0.3139391447368421, "grad_norm": 54.91154479980469, "learning_rate": 6.860608552631579e-06, "loss": 1.3594, "step": 1527 }, { "epoch": 0.31414473684210525, "grad_norm": 65.04061126708984, "learning_rate": 6.8585526315789475e-06, "loss": 0.7207, "step": 1528 }, { "epoch": 0.3143503289473684, "grad_norm": 23.02989387512207, "learning_rate": 6.856496710526315e-06, "loss": 0.1338, "step": 1529 }, { "epoch": 0.3145559210526316, "grad_norm": 56.46159362792969, "learning_rate": 6.854440789473685e-06, "loss": 0.5869, "step": 1530 }, { "epoch": 0.31476151315789475, "grad_norm": 45.26396179199219, "learning_rate": 6.852384868421054e-06, "loss": 0.7188, "step": 1531 }, { "epoch": 0.3149671052631579, "grad_norm": 110.18689727783203, "learning_rate": 6.8503289473684224e-06, "loss": 1.1063, "step": 1532 }, { "epoch": 0.3151726973684211, "grad_norm": 39.656742095947266, "learning_rate": 6.84827302631579e-06, "loss": 0.9961, "step": 1533 }, { "epoch": 0.31537828947368424, "grad_norm": 59.094886779785156, "learning_rate": 6.846217105263159e-06, "loss": 0.9473, "step": 1534 }, { "epoch": 0.31558388157894735, "grad_norm": 39.746124267578125, "learning_rate": 6.844161184210527e-06, "loss": 0.7778, "step": 1535 }, { "epoch": 0.3157894736842105, "grad_norm": 98.23493957519531, "learning_rate": 6.842105263157896e-06, "loss": 1.7371, "step": 1536 }, { "epoch": 0.3159950657894737, "grad_norm": 56.139991760253906, "learning_rate": 6.8400493421052635e-06, "loss": 0.6309, "step": 1537 }, { "epoch": 0.31620065789473684, "grad_norm": 21.36074447631836, "learning_rate": 6.837993421052632e-06, "loss": 0.251, "step": 1538 }, { "epoch": 0.31640625, "grad_norm": 39.55891799926758, "learning_rate": 6.8359375e-06, "loss": 0.5723, "step": 1539 }, { "epoch": 0.31661184210526316, "grad_norm": 75.90860748291016, "learning_rate": 6.833881578947369e-06, "loss": 0.7378, "step": 1540 }, { "epoch": 0.3168174342105263, "grad_norm": 30.383079528808594, "learning_rate": 6.831825657894738e-06, "loss": 0.2056, "step": 1541 }, { "epoch": 0.3170230263157895, "grad_norm": 17.931198120117188, "learning_rate": 6.8297697368421055e-06, "loss": 0.144, "step": 1542 }, { "epoch": 0.31722861842105265, "grad_norm": 14.403411865234375, "learning_rate": 6.827713815789474e-06, "loss": 0.1492, "step": 1543 }, { "epoch": 0.31743421052631576, "grad_norm": 68.28717803955078, "learning_rate": 6.825657894736842e-06, "loss": 0.7285, "step": 1544 }, { "epoch": 0.3176398026315789, "grad_norm": 32.50103759765625, "learning_rate": 6.823601973684211e-06, "loss": 0.4609, "step": 1545 }, { "epoch": 0.3178453947368421, "grad_norm": 36.7113151550293, "learning_rate": 6.821546052631579e-06, "loss": 0.6035, "step": 1546 }, { "epoch": 0.31805098684210525, "grad_norm": 8.89301872253418, "learning_rate": 6.819490131578947e-06, "loss": 0.0741, "step": 1547 }, { "epoch": 0.3182565789473684, "grad_norm": 78.57244110107422, "learning_rate": 6.817434210526316e-06, "loss": 1.0557, "step": 1548 }, { "epoch": 0.3184621710526316, "grad_norm": 76.08383178710938, "learning_rate": 6.815378289473685e-06, "loss": 1.1797, "step": 1549 }, { "epoch": 0.31866776315789475, "grad_norm": 63.21186065673828, "learning_rate": 6.813322368421054e-06, "loss": 0.835, "step": 1550 }, { "epoch": 0.3188733552631579, "grad_norm": 27.894285202026367, "learning_rate": 6.811266447368422e-06, "loss": 0.2403, "step": 1551 }, { "epoch": 0.3190789473684211, "grad_norm": 110.40480041503906, "learning_rate": 6.80921052631579e-06, "loss": 1.8906, "step": 1552 }, { "epoch": 0.31928453947368424, "grad_norm": 54.31237030029297, "learning_rate": 6.807154605263159e-06, "loss": 1.1709, "step": 1553 }, { "epoch": 0.31949013157894735, "grad_norm": 30.399330139160156, "learning_rate": 6.805098684210527e-06, "loss": 0.2402, "step": 1554 }, { "epoch": 0.3196957236842105, "grad_norm": 35.94065856933594, "learning_rate": 6.8030427631578956e-06, "loss": 0.5664, "step": 1555 }, { "epoch": 0.3199013157894737, "grad_norm": 70.88258361816406, "learning_rate": 6.8009868421052634e-06, "loss": 1.4219, "step": 1556 }, { "epoch": 0.32010690789473684, "grad_norm": 125.91007995605469, "learning_rate": 6.798930921052632e-06, "loss": 1.532, "step": 1557 }, { "epoch": 0.3203125, "grad_norm": 22.812734603881836, "learning_rate": 6.796875000000001e-06, "loss": 0.2255, "step": 1558 }, { "epoch": 0.32051809210526316, "grad_norm": 82.59606170654297, "learning_rate": 6.794819078947369e-06, "loss": 0.8477, "step": 1559 }, { "epoch": 0.3207236842105263, "grad_norm": 38.9715690612793, "learning_rate": 6.7927631578947375e-06, "loss": 0.438, "step": 1560 }, { "epoch": 0.3209292763157895, "grad_norm": 141.52691650390625, "learning_rate": 6.790707236842105e-06, "loss": 2.2812, "step": 1561 }, { "epoch": 0.32113486842105265, "grad_norm": 61.1801872253418, "learning_rate": 6.788651315789474e-06, "loss": 0.7432, "step": 1562 }, { "epoch": 0.32134046052631576, "grad_norm": 83.11520385742188, "learning_rate": 6.786595394736842e-06, "loss": 1.5547, "step": 1563 }, { "epoch": 0.3215460526315789, "grad_norm": 67.85233306884766, "learning_rate": 6.784539473684211e-06, "loss": 1.6367, "step": 1564 }, { "epoch": 0.3217516447368421, "grad_norm": 82.47225189208984, "learning_rate": 6.7824835526315794e-06, "loss": 1.4766, "step": 1565 }, { "epoch": 0.32195723684210525, "grad_norm": 60.8392333984375, "learning_rate": 6.780427631578947e-06, "loss": 0.7227, "step": 1566 }, { "epoch": 0.3221628289473684, "grad_norm": 287.39013671875, "learning_rate": 6.778371710526316e-06, "loss": 4.3359, "step": 1567 }, { "epoch": 0.3223684210526316, "grad_norm": 61.76559066772461, "learning_rate": 6.776315789473686e-06, "loss": 0.5957, "step": 1568 }, { "epoch": 0.32257401315789475, "grad_norm": 46.684661865234375, "learning_rate": 6.7742598684210535e-06, "loss": 0.3866, "step": 1569 }, { "epoch": 0.3227796052631579, "grad_norm": 50.911888122558594, "learning_rate": 6.772203947368422e-06, "loss": 0.7529, "step": 1570 }, { "epoch": 0.3229851973684211, "grad_norm": 22.99010467529297, "learning_rate": 6.77014802631579e-06, "loss": 0.4189, "step": 1571 }, { "epoch": 0.32319078947368424, "grad_norm": 39.66630172729492, "learning_rate": 6.768092105263159e-06, "loss": 0.5474, "step": 1572 }, { "epoch": 0.32339638157894735, "grad_norm": 46.29843521118164, "learning_rate": 6.766036184210527e-06, "loss": 0.792, "step": 1573 }, { "epoch": 0.3236019736842105, "grad_norm": 26.287357330322266, "learning_rate": 6.7639802631578954e-06, "loss": 0.1729, "step": 1574 }, { "epoch": 0.3238075657894737, "grad_norm": 115.89899444580078, "learning_rate": 6.761924342105264e-06, "loss": 0.6521, "step": 1575 }, { "epoch": 0.32401315789473684, "grad_norm": 26.13722038269043, "learning_rate": 6.759868421052632e-06, "loss": 0.2168, "step": 1576 }, { "epoch": 0.32421875, "grad_norm": 31.734689712524414, "learning_rate": 6.757812500000001e-06, "loss": 0.3623, "step": 1577 }, { "epoch": 0.32442434210526316, "grad_norm": 44.2773551940918, "learning_rate": 6.755756578947369e-06, "loss": 0.6831, "step": 1578 }, { "epoch": 0.3246299342105263, "grad_norm": 45.666770935058594, "learning_rate": 6.753700657894737e-06, "loss": 0.3667, "step": 1579 }, { "epoch": 0.3248355263157895, "grad_norm": 69.68789672851562, "learning_rate": 6.751644736842105e-06, "loss": 1.0859, "step": 1580 }, { "epoch": 0.32504111842105265, "grad_norm": 84.81520080566406, "learning_rate": 6.749588815789474e-06, "loss": 0.4397, "step": 1581 }, { "epoch": 0.32524671052631576, "grad_norm": 15.952046394348145, "learning_rate": 6.747532894736843e-06, "loss": 0.1301, "step": 1582 }, { "epoch": 0.3254523026315789, "grad_norm": 139.7264862060547, "learning_rate": 6.745476973684211e-06, "loss": 1.9316, "step": 1583 }, { "epoch": 0.3256578947368421, "grad_norm": 146.51995849609375, "learning_rate": 6.743421052631579e-06, "loss": 1.2148, "step": 1584 }, { "epoch": 0.32586348684210525, "grad_norm": 94.00788116455078, "learning_rate": 6.741365131578947e-06, "loss": 2.2266, "step": 1585 }, { "epoch": 0.3260690789473684, "grad_norm": 54.67285919189453, "learning_rate": 6.739309210526316e-06, "loss": 0.4824, "step": 1586 }, { "epoch": 0.3262746710526316, "grad_norm": 138.4453582763672, "learning_rate": 6.7372532894736855e-06, "loss": 2.3071, "step": 1587 }, { "epoch": 0.32648026315789475, "grad_norm": 25.677104949951172, "learning_rate": 6.735197368421053e-06, "loss": 0.1226, "step": 1588 }, { "epoch": 0.3266858552631579, "grad_norm": 48.207275390625, "learning_rate": 6.733141447368422e-06, "loss": 0.4492, "step": 1589 }, { "epoch": 0.3268914473684211, "grad_norm": 56.43464279174805, "learning_rate": 6.73108552631579e-06, "loss": 0.6318, "step": 1590 }, { "epoch": 0.32709703947368424, "grad_norm": 131.25244140625, "learning_rate": 6.729029605263159e-06, "loss": 1.793, "step": 1591 }, { "epoch": 0.32730263157894735, "grad_norm": 77.39800262451172, "learning_rate": 6.7269736842105275e-06, "loss": 1.2057, "step": 1592 }, { "epoch": 0.3275082236842105, "grad_norm": 34.00719451904297, "learning_rate": 6.724917763157895e-06, "loss": 0.3875, "step": 1593 }, { "epoch": 0.3277138157894737, "grad_norm": 135.5121612548828, "learning_rate": 6.722861842105264e-06, "loss": 1.168, "step": 1594 }, { "epoch": 0.32791940789473684, "grad_norm": 24.21680450439453, "learning_rate": 6.720805921052632e-06, "loss": 0.2263, "step": 1595 }, { "epoch": 0.328125, "grad_norm": 75.1781234741211, "learning_rate": 6.718750000000001e-06, "loss": 1.1082, "step": 1596 }, { "epoch": 0.32833059210526316, "grad_norm": 86.79224395751953, "learning_rate": 6.7166940789473686e-06, "loss": 0.8555, "step": 1597 }, { "epoch": 0.3285361842105263, "grad_norm": 100.21611022949219, "learning_rate": 6.714638157894737e-06, "loss": 0.7988, "step": 1598 }, { "epoch": 0.3287417763157895, "grad_norm": 57.391578674316406, "learning_rate": 6.712582236842106e-06, "loss": 0.7148, "step": 1599 }, { "epoch": 0.32894736842105265, "grad_norm": 33.76616287231445, "learning_rate": 6.710526315789474e-06, "loss": 0.3975, "step": 1600 }, { "epoch": 0.32915296052631576, "grad_norm": 41.550994873046875, "learning_rate": 6.708470394736843e-06, "loss": 0.3071, "step": 1601 }, { "epoch": 0.3293585526315789, "grad_norm": 9.753896713256836, "learning_rate": 6.7064144736842105e-06, "loss": 0.0771, "step": 1602 }, { "epoch": 0.3295641447368421, "grad_norm": 1.7913440465927124, "learning_rate": 6.704358552631579e-06, "loss": 0.0022, "step": 1603 }, { "epoch": 0.32976973684210525, "grad_norm": 18.202478408813477, "learning_rate": 6.702302631578947e-06, "loss": 0.0962, "step": 1604 }, { "epoch": 0.3299753289473684, "grad_norm": 52.62664794921875, "learning_rate": 6.700246710526316e-06, "loss": 0.2869, "step": 1605 }, { "epoch": 0.3301809210526316, "grad_norm": 205.4130401611328, "learning_rate": 6.698190789473685e-06, "loss": 1.8027, "step": 1606 }, { "epoch": 0.33038651315789475, "grad_norm": 53.28230285644531, "learning_rate": 6.696134868421053e-06, "loss": 0.4104, "step": 1607 }, { "epoch": 0.3305921052631579, "grad_norm": 23.392269134521484, "learning_rate": 6.694078947368422e-06, "loss": 0.1394, "step": 1608 }, { "epoch": 0.3307976973684211, "grad_norm": 247.58700561523438, "learning_rate": 6.692023026315791e-06, "loss": 2.2988, "step": 1609 }, { "epoch": 0.33100328947368424, "grad_norm": 296.11639404296875, "learning_rate": 6.689967105263159e-06, "loss": 2.3086, "step": 1610 }, { "epoch": 0.33120888157894735, "grad_norm": 271.97100830078125, "learning_rate": 6.687911184210527e-06, "loss": 2.1953, "step": 1611 }, { "epoch": 0.3314144736842105, "grad_norm": 93.68125915527344, "learning_rate": 6.685855263157895e-06, "loss": 1.248, "step": 1612 }, { "epoch": 0.3316200657894737, "grad_norm": 32.75151443481445, "learning_rate": 6.683799342105264e-06, "loss": 0.2422, "step": 1613 }, { "epoch": 0.33182565789473684, "grad_norm": 12.092679977416992, "learning_rate": 6.681743421052632e-06, "loss": 0.0779, "step": 1614 }, { "epoch": 0.33203125, "grad_norm": 80.91167449951172, "learning_rate": 6.679687500000001e-06, "loss": 1.3032, "step": 1615 }, { "epoch": 0.33223684210526316, "grad_norm": 112.66715240478516, "learning_rate": 6.677631578947369e-06, "loss": 1.2202, "step": 1616 }, { "epoch": 0.3324424342105263, "grad_norm": 207.0120391845703, "learning_rate": 6.675575657894737e-06, "loss": 1.6182, "step": 1617 }, { "epoch": 0.3326480263157895, "grad_norm": 57.881134033203125, "learning_rate": 6.673519736842106e-06, "loss": 1.2034, "step": 1618 }, { "epoch": 0.33285361842105265, "grad_norm": 9.7418851852417, "learning_rate": 6.671463815789474e-06, "loss": 0.0224, "step": 1619 }, { "epoch": 0.33305921052631576, "grad_norm": 48.86962890625, "learning_rate": 6.6694078947368425e-06, "loss": 0.6523, "step": 1620 }, { "epoch": 0.3332648026315789, "grad_norm": 52.335968017578125, "learning_rate": 6.66735197368421e-06, "loss": 0.4707, "step": 1621 }, { "epoch": 0.3334703947368421, "grad_norm": 126.75320434570312, "learning_rate": 6.665296052631579e-06, "loss": 1.838, "step": 1622 }, { "epoch": 0.33367598684210525, "grad_norm": 110.4897689819336, "learning_rate": 6.663240131578948e-06, "loss": 1.2705, "step": 1623 }, { "epoch": 0.3338815789473684, "grad_norm": 199.62620544433594, "learning_rate": 6.661184210526316e-06, "loss": 1.7676, "step": 1624 }, { "epoch": 0.3340871710526316, "grad_norm": 90.06951141357422, "learning_rate": 6.659128289473685e-06, "loss": 1.2944, "step": 1625 }, { "epoch": 0.33429276315789475, "grad_norm": 21.003843307495117, "learning_rate": 6.657072368421054e-06, "loss": 0.0663, "step": 1626 }, { "epoch": 0.3344983552631579, "grad_norm": 157.55374145507812, "learning_rate": 6.655016447368422e-06, "loss": 1.792, "step": 1627 }, { "epoch": 0.3347039473684211, "grad_norm": 59.0074348449707, "learning_rate": 6.652960526315791e-06, "loss": 0.6753, "step": 1628 }, { "epoch": 0.33490953947368424, "grad_norm": 99.99365234375, "learning_rate": 6.6509046052631585e-06, "loss": 0.5476, "step": 1629 }, { "epoch": 0.33511513157894735, "grad_norm": 212.07275390625, "learning_rate": 6.648848684210527e-06, "loss": 4.6719, "step": 1630 }, { "epoch": 0.3353207236842105, "grad_norm": 21.646448135375977, "learning_rate": 6.646792763157895e-06, "loss": 0.0729, "step": 1631 }, { "epoch": 0.3355263157894737, "grad_norm": 87.55914306640625, "learning_rate": 6.644736842105264e-06, "loss": 0.7078, "step": 1632 }, { "epoch": 0.33573190789473684, "grad_norm": 22.133098602294922, "learning_rate": 6.642680921052633e-06, "loss": 0.261, "step": 1633 }, { "epoch": 0.3359375, "grad_norm": 26.535722732543945, "learning_rate": 6.6406250000000005e-06, "loss": 0.1455, "step": 1634 }, { "epoch": 0.33614309210526316, "grad_norm": 48.21051025390625, "learning_rate": 6.638569078947369e-06, "loss": 0.772, "step": 1635 }, { "epoch": 0.3363486842105263, "grad_norm": 16.59963607788086, "learning_rate": 6.636513157894737e-06, "loss": 0.1313, "step": 1636 }, { "epoch": 0.3365542763157895, "grad_norm": 84.16139221191406, "learning_rate": 6.634457236842106e-06, "loss": 0.5967, "step": 1637 }, { "epoch": 0.33675986842105265, "grad_norm": 36.99968338012695, "learning_rate": 6.632401315789474e-06, "loss": 0.6934, "step": 1638 }, { "epoch": 0.33696546052631576, "grad_norm": 20.38227653503418, "learning_rate": 6.630345394736842e-06, "loss": 0.1072, "step": 1639 }, { "epoch": 0.3371710526315789, "grad_norm": 26.25441551208496, "learning_rate": 6.62828947368421e-06, "loss": 0.1139, "step": 1640 }, { "epoch": 0.3373766447368421, "grad_norm": 26.166664123535156, "learning_rate": 6.626233552631579e-06, "loss": 0.1011, "step": 1641 }, { "epoch": 0.33758223684210525, "grad_norm": 179.46163940429688, "learning_rate": 6.624177631578948e-06, "loss": 2.1641, "step": 1642 }, { "epoch": 0.3377878289473684, "grad_norm": 177.61485290527344, "learning_rate": 6.622121710526316e-06, "loss": 3.336, "step": 1643 }, { "epoch": 0.3379934210526316, "grad_norm": 113.41539764404297, "learning_rate": 6.620065789473685e-06, "loss": 1.1582, "step": 1644 }, { "epoch": 0.33819901315789475, "grad_norm": 54.61103439331055, "learning_rate": 6.618009868421054e-06, "loss": 0.386, "step": 1645 }, { "epoch": 0.3384046052631579, "grad_norm": 39.232337951660156, "learning_rate": 6.615953947368422e-06, "loss": 0.291, "step": 1646 }, { "epoch": 0.3386101973684211, "grad_norm": 52.451446533203125, "learning_rate": 6.6138980263157905e-06, "loss": 0.3535, "step": 1647 }, { "epoch": 0.33881578947368424, "grad_norm": 19.472457885742188, "learning_rate": 6.6118421052631584e-06, "loss": 0.1929, "step": 1648 }, { "epoch": 0.33902138157894735, "grad_norm": 138.72142028808594, "learning_rate": 6.609786184210527e-06, "loss": 1.1562, "step": 1649 }, { "epoch": 0.3392269736842105, "grad_norm": 10.99063777923584, "learning_rate": 6.607730263157895e-06, "loss": 0.0252, "step": 1650 }, { "epoch": 0.3394325657894737, "grad_norm": 50.59587097167969, "learning_rate": 6.605674342105264e-06, "loss": 0.8057, "step": 1651 }, { "epoch": 0.33963815789473684, "grad_norm": 23.45220184326172, "learning_rate": 6.6036184210526325e-06, "loss": 0.0836, "step": 1652 }, { "epoch": 0.33984375, "grad_norm": 173.83029174804688, "learning_rate": 6.6015625e-06, "loss": 2.7094, "step": 1653 }, { "epoch": 0.34004934210526316, "grad_norm": 182.44964599609375, "learning_rate": 6.599506578947369e-06, "loss": 4.1719, "step": 1654 }, { "epoch": 0.3402549342105263, "grad_norm": 77.18601989746094, "learning_rate": 6.597450657894737e-06, "loss": 0.6504, "step": 1655 }, { "epoch": 0.3404605263157895, "grad_norm": 46.5672607421875, "learning_rate": 6.595394736842106e-06, "loss": 0.269, "step": 1656 }, { "epoch": 0.34066611842105265, "grad_norm": 216.75355529785156, "learning_rate": 6.593338815789474e-06, "loss": 1.4766, "step": 1657 }, { "epoch": 0.34087171052631576, "grad_norm": 11.793583869934082, "learning_rate": 6.591282894736842e-06, "loss": 0.0682, "step": 1658 }, { "epoch": 0.3410773026315789, "grad_norm": 125.91575622558594, "learning_rate": 6.589226973684211e-06, "loss": 1.2637, "step": 1659 }, { "epoch": 0.3412828947368421, "grad_norm": 80.02279663085938, "learning_rate": 6.587171052631579e-06, "loss": 0.8042, "step": 1660 }, { "epoch": 0.34148848684210525, "grad_norm": 57.481842041015625, "learning_rate": 6.585115131578948e-06, "loss": 0.4849, "step": 1661 }, { "epoch": 0.3416940789473684, "grad_norm": 60.27647399902344, "learning_rate": 6.5830592105263155e-06, "loss": 1.0543, "step": 1662 }, { "epoch": 0.3418996710526316, "grad_norm": 31.531932830810547, "learning_rate": 6.581003289473685e-06, "loss": 0.1899, "step": 1663 }, { "epoch": 0.34210526315789475, "grad_norm": 184.09835815429688, "learning_rate": 6.578947368421054e-06, "loss": 2.0332, "step": 1664 }, { "epoch": 0.3423108552631579, "grad_norm": 126.99698638916016, "learning_rate": 6.576891447368422e-06, "loss": 1.5098, "step": 1665 }, { "epoch": 0.3425164473684211, "grad_norm": 39.306480407714844, "learning_rate": 6.5748355263157904e-06, "loss": 0.3203, "step": 1666 }, { "epoch": 0.34272203947368424, "grad_norm": 52.3098258972168, "learning_rate": 6.572779605263158e-06, "loss": 0.3506, "step": 1667 }, { "epoch": 0.34292763157894735, "grad_norm": 4.892480850219727, "learning_rate": 6.570723684210527e-06, "loss": 0.0325, "step": 1668 }, { "epoch": 0.3431332236842105, "grad_norm": 52.222232818603516, "learning_rate": 6.568667763157896e-06, "loss": 0.668, "step": 1669 }, { "epoch": 0.3433388157894737, "grad_norm": 12.470879554748535, "learning_rate": 6.566611842105264e-06, "loss": 0.0701, "step": 1670 }, { "epoch": 0.34354440789473684, "grad_norm": 43.45999526977539, "learning_rate": 6.564555921052632e-06, "loss": 0.3112, "step": 1671 }, { "epoch": 0.34375, "grad_norm": 130.84783935546875, "learning_rate": 6.5625e-06, "loss": 1.9883, "step": 1672 }, { "epoch": 0.34395559210526316, "grad_norm": 55.82377243041992, "learning_rate": 6.560444078947369e-06, "loss": 0.4922, "step": 1673 }, { "epoch": 0.3441611842105263, "grad_norm": 75.8546142578125, "learning_rate": 6.558388157894737e-06, "loss": 0.6406, "step": 1674 }, { "epoch": 0.3443667763157895, "grad_norm": 57.31818389892578, "learning_rate": 6.556332236842106e-06, "loss": 0.2913, "step": 1675 }, { "epoch": 0.34457236842105265, "grad_norm": 54.459434509277344, "learning_rate": 6.554276315789474e-06, "loss": 0.4014, "step": 1676 }, { "epoch": 0.34477796052631576, "grad_norm": 49.913360595703125, "learning_rate": 6.552220394736842e-06, "loss": 0.5918, "step": 1677 }, { "epoch": 0.3449835526315789, "grad_norm": 46.59246826171875, "learning_rate": 6.550164473684211e-06, "loss": 0.2876, "step": 1678 }, { "epoch": 0.3451891447368421, "grad_norm": 40.4019889831543, "learning_rate": 6.548108552631579e-06, "loss": 0.3513, "step": 1679 }, { "epoch": 0.34539473684210525, "grad_norm": 26.40839958190918, "learning_rate": 6.5460526315789476e-06, "loss": 0.2183, "step": 1680 }, { "epoch": 0.3456003289473684, "grad_norm": 35.27826690673828, "learning_rate": 6.5439967105263154e-06, "loss": 0.1675, "step": 1681 }, { "epoch": 0.3458059210526316, "grad_norm": 141.83111572265625, "learning_rate": 6.541940789473685e-06, "loss": 2.3145, "step": 1682 }, { "epoch": 0.34601151315789475, "grad_norm": 36.3303337097168, "learning_rate": 6.539884868421054e-06, "loss": 0.2358, "step": 1683 }, { "epoch": 0.3462171052631579, "grad_norm": 78.06477355957031, "learning_rate": 6.537828947368422e-06, "loss": 0.5742, "step": 1684 }, { "epoch": 0.3464226973684211, "grad_norm": 39.224143981933594, "learning_rate": 6.53577302631579e-06, "loss": 0.2776, "step": 1685 }, { "epoch": 0.34662828947368424, "grad_norm": 190.8227081298828, "learning_rate": 6.533717105263159e-06, "loss": 4.3008, "step": 1686 }, { "epoch": 0.34683388157894735, "grad_norm": 79.70282745361328, "learning_rate": 6.531661184210527e-06, "loss": 0.2761, "step": 1687 }, { "epoch": 0.3470394736842105, "grad_norm": 34.339900970458984, "learning_rate": 6.529605263157896e-06, "loss": 0.2271, "step": 1688 }, { "epoch": 0.3472450657894737, "grad_norm": 177.7330322265625, "learning_rate": 6.5275493421052636e-06, "loss": 2.0581, "step": 1689 }, { "epoch": 0.34745065789473684, "grad_norm": 28.12782859802246, "learning_rate": 6.525493421052632e-06, "loss": 0.1412, "step": 1690 }, { "epoch": 0.34765625, "grad_norm": 53.67451477050781, "learning_rate": 6.5234375e-06, "loss": 0.7168, "step": 1691 }, { "epoch": 0.34786184210526316, "grad_norm": 11.377715110778809, "learning_rate": 6.521381578947369e-06, "loss": 0.0501, "step": 1692 }, { "epoch": 0.3480674342105263, "grad_norm": 34.107635498046875, "learning_rate": 6.519325657894738e-06, "loss": 0.207, "step": 1693 }, { "epoch": 0.3482730263157895, "grad_norm": 45.93149185180664, "learning_rate": 6.5172697368421055e-06, "loss": 0.5094, "step": 1694 }, { "epoch": 0.34847861842105265, "grad_norm": 27.23862648010254, "learning_rate": 6.515213815789474e-06, "loss": 0.1854, "step": 1695 }, { "epoch": 0.34868421052631576, "grad_norm": 218.95082092285156, "learning_rate": 6.513157894736842e-06, "loss": 1.9395, "step": 1696 }, { "epoch": 0.3488898026315789, "grad_norm": 11.125772476196289, "learning_rate": 6.511101973684211e-06, "loss": 0.1282, "step": 1697 }, { "epoch": 0.3490953947368421, "grad_norm": 85.58750915527344, "learning_rate": 6.509046052631579e-06, "loss": 0.7787, "step": 1698 }, { "epoch": 0.34930098684210525, "grad_norm": 60.1148567199707, "learning_rate": 6.5069901315789475e-06, "loss": 0.4756, "step": 1699 }, { "epoch": 0.3495065789473684, "grad_norm": 81.92547607421875, "learning_rate": 6.504934210526316e-06, "loss": 0.8184, "step": 1700 }, { "epoch": 0.3497121710526316, "grad_norm": 186.79617309570312, "learning_rate": 6.502878289473685e-06, "loss": 1.6699, "step": 1701 }, { "epoch": 0.34991776315789475, "grad_norm": 53.26115798950195, "learning_rate": 6.500822368421054e-06, "loss": 0.3662, "step": 1702 }, { "epoch": 0.3501233552631579, "grad_norm": 122.37206268310547, "learning_rate": 6.498766447368422e-06, "loss": 1.8887, "step": 1703 }, { "epoch": 0.3503289473684211, "grad_norm": 67.96122741699219, "learning_rate": 6.49671052631579e-06, "loss": 0.9043, "step": 1704 }, { "epoch": 0.35053453947368424, "grad_norm": 111.12496185302734, "learning_rate": 6.494654605263159e-06, "loss": 0.7047, "step": 1705 }, { "epoch": 0.35074013157894735, "grad_norm": 172.60826110839844, "learning_rate": 6.492598684210527e-06, "loss": 2.2266, "step": 1706 }, { "epoch": 0.3509457236842105, "grad_norm": 31.974088668823242, "learning_rate": 6.490542763157896e-06, "loss": 0.5891, "step": 1707 }, { "epoch": 0.3511513157894737, "grad_norm": 102.7982406616211, "learning_rate": 6.4884868421052635e-06, "loss": 1.4922, "step": 1708 }, { "epoch": 0.35135690789473684, "grad_norm": 20.008371353149414, "learning_rate": 6.486430921052632e-06, "loss": 0.062, "step": 1709 }, { "epoch": 0.3515625, "grad_norm": 33.4422721862793, "learning_rate": 6.484375000000001e-06, "loss": 0.1763, "step": 1710 }, { "epoch": 0.35176809210526316, "grad_norm": 30.91750717163086, "learning_rate": 6.482319078947369e-06, "loss": 0.1699, "step": 1711 }, { "epoch": 0.3519736842105263, "grad_norm": 77.92295837402344, "learning_rate": 6.4802631578947375e-06, "loss": 1.2065, "step": 1712 }, { "epoch": 0.3521792763157895, "grad_norm": 52.81715393066406, "learning_rate": 6.478207236842105e-06, "loss": 0.3887, "step": 1713 }, { "epoch": 0.35238486842105265, "grad_norm": 49.343788146972656, "learning_rate": 6.476151315789474e-06, "loss": 0.325, "step": 1714 }, { "epoch": 0.35259046052631576, "grad_norm": 69.04796600341797, "learning_rate": 6.474095394736842e-06, "loss": 0.7786, "step": 1715 }, { "epoch": 0.3527960526315789, "grad_norm": 217.13165283203125, "learning_rate": 6.472039473684211e-06, "loss": 1.7158, "step": 1716 }, { "epoch": 0.3530016447368421, "grad_norm": 58.680999755859375, "learning_rate": 6.4699835526315795e-06, "loss": 0.5039, "step": 1717 }, { "epoch": 0.35320723684210525, "grad_norm": 49.3232536315918, "learning_rate": 6.467927631578947e-06, "loss": 0.418, "step": 1718 }, { "epoch": 0.3534128289473684, "grad_norm": 44.0789794921875, "learning_rate": 6.465871710526316e-06, "loss": 0.5368, "step": 1719 }, { "epoch": 0.3536184210526316, "grad_norm": 141.9054412841797, "learning_rate": 6.463815789473686e-06, "loss": 1.3586, "step": 1720 }, { "epoch": 0.35382401315789475, "grad_norm": 75.50473022460938, "learning_rate": 6.4617598684210535e-06, "loss": 0.8438, "step": 1721 }, { "epoch": 0.3540296052631579, "grad_norm": 116.03173065185547, "learning_rate": 6.459703947368422e-06, "loss": 1.4355, "step": 1722 }, { "epoch": 0.3542351973684211, "grad_norm": 36.138824462890625, "learning_rate": 6.45764802631579e-06, "loss": 0.3533, "step": 1723 }, { "epoch": 0.35444078947368424, "grad_norm": 35.17411804199219, "learning_rate": 6.455592105263159e-06, "loss": 0.7256, "step": 1724 }, { "epoch": 0.35464638157894735, "grad_norm": 24.29652214050293, "learning_rate": 6.453536184210527e-06, "loss": 0.1865, "step": 1725 }, { "epoch": 0.3548519736842105, "grad_norm": 84.5289077758789, "learning_rate": 6.4514802631578955e-06, "loss": 1.0801, "step": 1726 }, { "epoch": 0.3550575657894737, "grad_norm": 6.6488037109375, "learning_rate": 6.449424342105264e-06, "loss": 0.0602, "step": 1727 }, { "epoch": 0.35526315789473684, "grad_norm": 29.400081634521484, "learning_rate": 6.447368421052632e-06, "loss": 0.226, "step": 1728 }, { "epoch": 0.35546875, "grad_norm": 162.30868530273438, "learning_rate": 6.445312500000001e-06, "loss": 2.5, "step": 1729 }, { "epoch": 0.35567434210526316, "grad_norm": 140.79823303222656, "learning_rate": 6.443256578947369e-06, "loss": 2.5884, "step": 1730 }, { "epoch": 0.3558799342105263, "grad_norm": 75.21633911132812, "learning_rate": 6.441200657894737e-06, "loss": 0.7, "step": 1731 }, { "epoch": 0.3560855263157895, "grad_norm": 28.657859802246094, "learning_rate": 6.439144736842105e-06, "loss": 0.2149, "step": 1732 }, { "epoch": 0.35629111842105265, "grad_norm": 55.08251190185547, "learning_rate": 6.437088815789474e-06, "loss": 0.377, "step": 1733 }, { "epoch": 0.35649671052631576, "grad_norm": 12.479059219360352, "learning_rate": 6.435032894736843e-06, "loss": 0.0758, "step": 1734 }, { "epoch": 0.3567023026315789, "grad_norm": 58.31557846069336, "learning_rate": 6.432976973684211e-06, "loss": 0.7559, "step": 1735 }, { "epoch": 0.3569078947368421, "grad_norm": 13.980241775512695, "learning_rate": 6.430921052631579e-06, "loss": 0.0913, "step": 1736 }, { "epoch": 0.35711348684210525, "grad_norm": 97.7601089477539, "learning_rate": 6.428865131578947e-06, "loss": 1.6602, "step": 1737 }, { "epoch": 0.3573190789473684, "grad_norm": 18.744369506835938, "learning_rate": 6.426809210526316e-06, "loss": 0.1926, "step": 1738 }, { "epoch": 0.3575246710526316, "grad_norm": 85.79634094238281, "learning_rate": 6.4247532894736855e-06, "loss": 0.9452, "step": 1739 }, { "epoch": 0.35773026315789475, "grad_norm": 26.32024383544922, "learning_rate": 6.4226973684210534e-06, "loss": 0.1287, "step": 1740 }, { "epoch": 0.3579358552631579, "grad_norm": 111.11727142333984, "learning_rate": 6.420641447368422e-06, "loss": 0.7161, "step": 1741 }, { "epoch": 0.3581414473684211, "grad_norm": 144.812744140625, "learning_rate": 6.41858552631579e-06, "loss": 3.1437, "step": 1742 }, { "epoch": 0.35834703947368424, "grad_norm": 91.51591491699219, "learning_rate": 6.416529605263159e-06, "loss": 1.6152, "step": 1743 }, { "epoch": 0.35855263157894735, "grad_norm": 201.9343719482422, "learning_rate": 6.4144736842105275e-06, "loss": 2.4585, "step": 1744 }, { "epoch": 0.3587582236842105, "grad_norm": 22.975847244262695, "learning_rate": 6.412417763157895e-06, "loss": 0.257, "step": 1745 }, { "epoch": 0.3589638157894737, "grad_norm": 72.0746841430664, "learning_rate": 6.410361842105264e-06, "loss": 0.9771, "step": 1746 }, { "epoch": 0.35916940789473684, "grad_norm": 5.967100143432617, "learning_rate": 6.408305921052632e-06, "loss": 0.0318, "step": 1747 }, { "epoch": 0.359375, "grad_norm": 77.31904602050781, "learning_rate": 6.406250000000001e-06, "loss": 1.0747, "step": 1748 }, { "epoch": 0.35958059210526316, "grad_norm": 28.64374351501465, "learning_rate": 6.404194078947369e-06, "loss": 0.1892, "step": 1749 }, { "epoch": 0.3597861842105263, "grad_norm": 55.15675735473633, "learning_rate": 6.402138157894737e-06, "loss": 0.5684, "step": 1750 }, { "epoch": 0.3599917763157895, "grad_norm": 117.6438217163086, "learning_rate": 6.400082236842105e-06, "loss": 0.7769, "step": 1751 }, { "epoch": 0.36019736842105265, "grad_norm": 150.47210693359375, "learning_rate": 6.398026315789474e-06, "loss": 1.1196, "step": 1752 }, { "epoch": 0.36040296052631576, "grad_norm": 12.370079040527344, "learning_rate": 6.395970394736843e-06, "loss": 0.0824, "step": 1753 }, { "epoch": 0.3606085526315789, "grad_norm": 86.35796356201172, "learning_rate": 6.3939144736842105e-06, "loss": 0.9204, "step": 1754 }, { "epoch": 0.3608141447368421, "grad_norm": 93.94051361083984, "learning_rate": 6.391858552631579e-06, "loss": 0.9036, "step": 1755 }, { "epoch": 0.36101973684210525, "grad_norm": 32.59375, "learning_rate": 6.389802631578947e-06, "loss": 0.1733, "step": 1756 }, { "epoch": 0.3612253289473684, "grad_norm": 95.61215209960938, "learning_rate": 6.387746710526316e-06, "loss": 0.5554, "step": 1757 }, { "epoch": 0.3614309210526316, "grad_norm": 31.372770309448242, "learning_rate": 6.3856907894736854e-06, "loss": 0.2593, "step": 1758 }, { "epoch": 0.36163651315789475, "grad_norm": 216.5350341796875, "learning_rate": 6.383634868421053e-06, "loss": 1.9961, "step": 1759 }, { "epoch": 0.3618421052631579, "grad_norm": 23.404321670532227, "learning_rate": 6.381578947368422e-06, "loss": 0.1998, "step": 1760 }, { "epoch": 0.3620476973684211, "grad_norm": 51.661376953125, "learning_rate": 6.37952302631579e-06, "loss": 0.4146, "step": 1761 }, { "epoch": 0.36225328947368424, "grad_norm": 109.75849151611328, "learning_rate": 6.377467105263159e-06, "loss": 1.8098, "step": 1762 }, { "epoch": 0.36245888157894735, "grad_norm": 65.20433807373047, "learning_rate": 6.375411184210527e-06, "loss": 0.8457, "step": 1763 }, { "epoch": 0.3626644736842105, "grad_norm": 183.99314880371094, "learning_rate": 6.373355263157895e-06, "loss": 1.6211, "step": 1764 }, { "epoch": 0.3628700657894737, "grad_norm": 88.8577880859375, "learning_rate": 6.371299342105264e-06, "loss": 0.5469, "step": 1765 }, { "epoch": 0.36307565789473684, "grad_norm": 28.96632957458496, "learning_rate": 6.369243421052632e-06, "loss": 0.1472, "step": 1766 }, { "epoch": 0.36328125, "grad_norm": 72.58882904052734, "learning_rate": 6.367187500000001e-06, "loss": 0.725, "step": 1767 }, { "epoch": 0.36348684210526316, "grad_norm": 64.22361755371094, "learning_rate": 6.3651315789473685e-06, "loss": 0.5289, "step": 1768 }, { "epoch": 0.3636924342105263, "grad_norm": 192.05043029785156, "learning_rate": 6.363075657894737e-06, "loss": 1.375, "step": 1769 }, { "epoch": 0.3638980263157895, "grad_norm": 30.603837966918945, "learning_rate": 6.361019736842106e-06, "loss": 0.3916, "step": 1770 }, { "epoch": 0.36410361842105265, "grad_norm": 50.626399993896484, "learning_rate": 6.358963815789474e-06, "loss": 0.2783, "step": 1771 }, { "epoch": 0.36430921052631576, "grad_norm": 27.65091323852539, "learning_rate": 6.3569078947368426e-06, "loss": 0.3047, "step": 1772 }, { "epoch": 0.3645148026315789, "grad_norm": 14.762500762939453, "learning_rate": 6.3548519736842104e-06, "loss": 0.088, "step": 1773 }, { "epoch": 0.3647203947368421, "grad_norm": 86.77203369140625, "learning_rate": 6.352796052631579e-06, "loss": 0.6995, "step": 1774 }, { "epoch": 0.36492598684210525, "grad_norm": 27.237089157104492, "learning_rate": 6.350740131578947e-06, "loss": 0.1476, "step": 1775 }, { "epoch": 0.3651315789473684, "grad_norm": 150.95716857910156, "learning_rate": 6.348684210526316e-06, "loss": 1.8398, "step": 1776 }, { "epoch": 0.3653371710526316, "grad_norm": 31.87674903869629, "learning_rate": 6.346628289473685e-06, "loss": 0.2457, "step": 1777 }, { "epoch": 0.36554276315789475, "grad_norm": 46.588409423828125, "learning_rate": 6.344572368421053e-06, "loss": 0.5117, "step": 1778 }, { "epoch": 0.3657483552631579, "grad_norm": 24.15743064880371, "learning_rate": 6.342516447368422e-06, "loss": 0.0944, "step": 1779 }, { "epoch": 0.3659539473684211, "grad_norm": 85.32207489013672, "learning_rate": 6.340460526315791e-06, "loss": 1.3813, "step": 1780 }, { "epoch": 0.36615953947368424, "grad_norm": 88.76026916503906, "learning_rate": 6.3384046052631586e-06, "loss": 0.657, "step": 1781 }, { "epoch": 0.36636513157894735, "grad_norm": 79.16852569580078, "learning_rate": 6.336348684210527e-06, "loss": 0.7158, "step": 1782 }, { "epoch": 0.3665707236842105, "grad_norm": 43.840240478515625, "learning_rate": 6.334292763157895e-06, "loss": 0.2588, "step": 1783 }, { "epoch": 0.3667763157894737, "grad_norm": 18.523073196411133, "learning_rate": 6.332236842105264e-06, "loss": 0.1208, "step": 1784 }, { "epoch": 0.36698190789473684, "grad_norm": 37.69044876098633, "learning_rate": 6.330180921052632e-06, "loss": 0.2358, "step": 1785 }, { "epoch": 0.3671875, "grad_norm": 60.980987548828125, "learning_rate": 6.3281250000000005e-06, "loss": 0.7688, "step": 1786 }, { "epoch": 0.36739309210526316, "grad_norm": 99.3213119506836, "learning_rate": 6.326069078947369e-06, "loss": 1.5493, "step": 1787 }, { "epoch": 0.3675986842105263, "grad_norm": 72.27870178222656, "learning_rate": 6.324013157894737e-06, "loss": 0.7617, "step": 1788 }, { "epoch": 0.3678042763157895, "grad_norm": 22.702686309814453, "learning_rate": 6.321957236842106e-06, "loss": 0.1338, "step": 1789 }, { "epoch": 0.36800986842105265, "grad_norm": 143.1873321533203, "learning_rate": 6.319901315789474e-06, "loss": 3.1211, "step": 1790 }, { "epoch": 0.36821546052631576, "grad_norm": 48.38536834716797, "learning_rate": 6.3178453947368424e-06, "loss": 0.5859, "step": 1791 }, { "epoch": 0.3684210526315789, "grad_norm": 58.08936309814453, "learning_rate": 6.31578947368421e-06, "loss": 1.0996, "step": 1792 }, { "epoch": 0.3686266447368421, "grad_norm": 20.161375045776367, "learning_rate": 6.313733552631579e-06, "loss": 0.0801, "step": 1793 }, { "epoch": 0.36883223684210525, "grad_norm": 51.91925048828125, "learning_rate": 6.311677631578948e-06, "loss": 0.3253, "step": 1794 }, { "epoch": 0.3690378289473684, "grad_norm": 70.86725616455078, "learning_rate": 6.309621710526316e-06, "loss": 0.6151, "step": 1795 }, { "epoch": 0.3692434210526316, "grad_norm": 69.88233947753906, "learning_rate": 6.307565789473685e-06, "loss": 0.6445, "step": 1796 }, { "epoch": 0.36944901315789475, "grad_norm": 229.4588623046875, "learning_rate": 6.305509868421054e-06, "loss": 1.8192, "step": 1797 }, { "epoch": 0.3696546052631579, "grad_norm": 35.728153228759766, "learning_rate": 6.303453947368422e-06, "loss": 0.3911, "step": 1798 }, { "epoch": 0.3698601973684211, "grad_norm": 31.045211791992188, "learning_rate": 6.301398026315791e-06, "loss": 0.2296, "step": 1799 }, { "epoch": 0.37006578947368424, "grad_norm": 95.45970916748047, "learning_rate": 6.2993421052631585e-06, "loss": 1.1602, "step": 1800 }, { "epoch": 0.37027138157894735, "grad_norm": 49.38419723510742, "learning_rate": 6.297286184210527e-06, "loss": 0.2441, "step": 1801 }, { "epoch": 0.3704769736842105, "grad_norm": 81.38131713867188, "learning_rate": 6.295230263157895e-06, "loss": 0.3673, "step": 1802 }, { "epoch": 0.3706825657894737, "grad_norm": 15.714239120483398, "learning_rate": 6.293174342105264e-06, "loss": 0.1248, "step": 1803 }, { "epoch": 0.37088815789473684, "grad_norm": 35.70469665527344, "learning_rate": 6.2911184210526325e-06, "loss": 0.1621, "step": 1804 }, { "epoch": 0.37109375, "grad_norm": 145.09588623046875, "learning_rate": 6.2890625e-06, "loss": 1.5352, "step": 1805 }, { "epoch": 0.37129934210526316, "grad_norm": 48.971248626708984, "learning_rate": 6.287006578947369e-06, "loss": 0.6816, "step": 1806 }, { "epoch": 0.3715049342105263, "grad_norm": 16.729246139526367, "learning_rate": 6.284950657894737e-06, "loss": 0.0689, "step": 1807 }, { "epoch": 0.3717105263157895, "grad_norm": 139.58921813964844, "learning_rate": 6.282894736842106e-06, "loss": 1.8398, "step": 1808 }, { "epoch": 0.37191611842105265, "grad_norm": 51.11547088623047, "learning_rate": 6.280838815789474e-06, "loss": 0.499, "step": 1809 }, { "epoch": 0.37212171052631576, "grad_norm": 79.49119567871094, "learning_rate": 6.278782894736842e-06, "loss": 1.0273, "step": 1810 }, { "epoch": 0.3723273026315789, "grad_norm": 60.45927810668945, "learning_rate": 6.276726973684211e-06, "loss": 0.5835, "step": 1811 }, { "epoch": 0.3725328947368421, "grad_norm": 27.305150985717773, "learning_rate": 6.274671052631579e-06, "loss": 0.2656, "step": 1812 }, { "epoch": 0.37273848684210525, "grad_norm": 63.95374298095703, "learning_rate": 6.272615131578948e-06, "loss": 1.1719, "step": 1813 }, { "epoch": 0.3729440789473684, "grad_norm": 82.82866668701172, "learning_rate": 6.2705592105263156e-06, "loss": 0.5488, "step": 1814 }, { "epoch": 0.3731496710526316, "grad_norm": 203.31976318359375, "learning_rate": 6.268503289473685e-06, "loss": 2.6445, "step": 1815 }, { "epoch": 0.37335526315789475, "grad_norm": 104.02290344238281, "learning_rate": 6.266447368421054e-06, "loss": 1.2656, "step": 1816 }, { "epoch": 0.3735608552631579, "grad_norm": 88.02436828613281, "learning_rate": 6.264391447368422e-06, "loss": 0.9902, "step": 1817 }, { "epoch": 0.3737664473684211, "grad_norm": 76.1810073852539, "learning_rate": 6.2623355263157905e-06, "loss": 1.3359, "step": 1818 }, { "epoch": 0.37397203947368424, "grad_norm": 63.08906173706055, "learning_rate": 6.260279605263158e-06, "loss": 0.7949, "step": 1819 }, { "epoch": 0.37417763157894735, "grad_norm": 39.2204704284668, "learning_rate": 6.258223684210527e-06, "loss": 0.2129, "step": 1820 }, { "epoch": 0.3743832236842105, "grad_norm": 52.655025482177734, "learning_rate": 6.256167763157896e-06, "loss": 0.4175, "step": 1821 }, { "epoch": 0.3745888157894737, "grad_norm": 217.12217712402344, "learning_rate": 6.254111842105264e-06, "loss": 1.4023, "step": 1822 }, { "epoch": 0.37479440789473684, "grad_norm": 112.09642791748047, "learning_rate": 6.252055921052632e-06, "loss": 1.4531, "step": 1823 }, { "epoch": 0.375, "grad_norm": 89.83623504638672, "learning_rate": 6.25e-06, "loss": 1.2708, "step": 1824 }, { "epoch": 0.37520559210526316, "grad_norm": 90.86077117919922, "learning_rate": 6.247944078947369e-06, "loss": 1.1592, "step": 1825 }, { "epoch": 0.3754111842105263, "grad_norm": 28.48163414001465, "learning_rate": 6.245888157894737e-06, "loss": 0.0988, "step": 1826 }, { "epoch": 0.3756167763157895, "grad_norm": 132.5896759033203, "learning_rate": 6.243832236842106e-06, "loss": 1.375, "step": 1827 }, { "epoch": 0.37582236842105265, "grad_norm": 71.10856628417969, "learning_rate": 6.241776315789474e-06, "loss": 0.5311, "step": 1828 }, { "epoch": 0.37602796052631576, "grad_norm": 28.545549392700195, "learning_rate": 6.239720394736842e-06, "loss": 0.1583, "step": 1829 }, { "epoch": 0.3762335526315789, "grad_norm": 108.81592559814453, "learning_rate": 6.237664473684211e-06, "loss": 2.1895, "step": 1830 }, { "epoch": 0.3764391447368421, "grad_norm": 233.5877685546875, "learning_rate": 6.235608552631579e-06, "loss": 1.6299, "step": 1831 }, { "epoch": 0.37664473684210525, "grad_norm": 83.05500030517578, "learning_rate": 6.233552631578948e-06, "loss": 0.8086, "step": 1832 }, { "epoch": 0.3768503289473684, "grad_norm": 90.20846557617188, "learning_rate": 6.2314967105263155e-06, "loss": 0.8848, "step": 1833 }, { "epoch": 0.3770559210526316, "grad_norm": 53.613834381103516, "learning_rate": 6.229440789473685e-06, "loss": 0.4634, "step": 1834 }, { "epoch": 0.37726151315789475, "grad_norm": 53.57508850097656, "learning_rate": 6.227384868421054e-06, "loss": 1.0859, "step": 1835 }, { "epoch": 0.3774671052631579, "grad_norm": 84.80597686767578, "learning_rate": 6.225328947368422e-06, "loss": 0.8967, "step": 1836 }, { "epoch": 0.3776726973684211, "grad_norm": 33.116477966308594, "learning_rate": 6.22327302631579e-06, "loss": 0.3918, "step": 1837 }, { "epoch": 0.37787828947368424, "grad_norm": 71.1299057006836, "learning_rate": 6.221217105263159e-06, "loss": 1.2852, "step": 1838 }, { "epoch": 0.37808388157894735, "grad_norm": 93.61790466308594, "learning_rate": 6.219161184210527e-06, "loss": 1.0244, "step": 1839 }, { "epoch": 0.3782894736842105, "grad_norm": 21.581235885620117, "learning_rate": 6.217105263157896e-06, "loss": 0.0948, "step": 1840 }, { "epoch": 0.3784950657894737, "grad_norm": 26.668346405029297, "learning_rate": 6.215049342105264e-06, "loss": 0.231, "step": 1841 }, { "epoch": 0.37870065789473684, "grad_norm": 211.19798278808594, "learning_rate": 6.212993421052632e-06, "loss": 2.5859, "step": 1842 }, { "epoch": 0.37890625, "grad_norm": 67.9140853881836, "learning_rate": 6.2109375e-06, "loss": 0.9961, "step": 1843 }, { "epoch": 0.37911184210526316, "grad_norm": 43.64830780029297, "learning_rate": 6.208881578947369e-06, "loss": 0.4122, "step": 1844 }, { "epoch": 0.3793174342105263, "grad_norm": 28.0200252532959, "learning_rate": 6.206825657894738e-06, "loss": 0.248, "step": 1845 }, { "epoch": 0.3795230263157895, "grad_norm": 78.75090026855469, "learning_rate": 6.2047697368421055e-06, "loss": 0.8481, "step": 1846 }, { "epoch": 0.37972861842105265, "grad_norm": 44.75204086303711, "learning_rate": 6.202713815789474e-06, "loss": 0.7378, "step": 1847 }, { "epoch": 0.37993421052631576, "grad_norm": 42.55432891845703, "learning_rate": 6.200657894736842e-06, "loss": 0.7227, "step": 1848 }, { "epoch": 0.3801398026315789, "grad_norm": 56.92592239379883, "learning_rate": 6.198601973684211e-06, "loss": 0.7776, "step": 1849 }, { "epoch": 0.3803453947368421, "grad_norm": 154.01870727539062, "learning_rate": 6.196546052631579e-06, "loss": 1.1094, "step": 1850 }, { "epoch": 0.38055098684210525, "grad_norm": 27.501487731933594, "learning_rate": 6.1944901315789475e-06, "loss": 0.3135, "step": 1851 }, { "epoch": 0.3807565789473684, "grad_norm": 98.59254455566406, "learning_rate": 6.192434210526315e-06, "loss": 1.6797, "step": 1852 }, { "epoch": 0.3809621710526316, "grad_norm": 67.48822021484375, "learning_rate": 6.190378289473685e-06, "loss": 0.4707, "step": 1853 }, { "epoch": 0.38116776315789475, "grad_norm": 124.2984390258789, "learning_rate": 6.188322368421054e-06, "loss": 1.0812, "step": 1854 }, { "epoch": 0.3813733552631579, "grad_norm": 41.53175354003906, "learning_rate": 6.186266447368422e-06, "loss": 0.3497, "step": 1855 }, { "epoch": 0.3815789473684211, "grad_norm": 49.95635223388672, "learning_rate": 6.18421052631579e-06, "loss": 0.9142, "step": 1856 }, { "epoch": 0.38178453947368424, "grad_norm": 21.334421157836914, "learning_rate": 6.182154605263159e-06, "loss": 0.1091, "step": 1857 }, { "epoch": 0.38199013157894735, "grad_norm": 29.693836212158203, "learning_rate": 6.180098684210527e-06, "loss": 0.2119, "step": 1858 }, { "epoch": 0.3821957236842105, "grad_norm": 14.961471557617188, "learning_rate": 6.178042763157896e-06, "loss": 0.0987, "step": 1859 }, { "epoch": 0.3824013157894737, "grad_norm": 94.25920104980469, "learning_rate": 6.1759868421052635e-06, "loss": 1.3106, "step": 1860 }, { "epoch": 0.38260690789473684, "grad_norm": 88.9892807006836, "learning_rate": 6.173930921052632e-06, "loss": 1.3242, "step": 1861 }, { "epoch": 0.3828125, "grad_norm": 206.4331817626953, "learning_rate": 6.171875e-06, "loss": 2.477, "step": 1862 }, { "epoch": 0.38301809210526316, "grad_norm": 25.662851333618164, "learning_rate": 6.169819078947369e-06, "loss": 0.3589, "step": 1863 }, { "epoch": 0.3832236842105263, "grad_norm": 95.57720184326172, "learning_rate": 6.1677631578947376e-06, "loss": 0.8271, "step": 1864 }, { "epoch": 0.3834292763157895, "grad_norm": 102.79107666015625, "learning_rate": 6.1657072368421054e-06, "loss": 0.9844, "step": 1865 }, { "epoch": 0.38363486842105265, "grad_norm": 32.5120849609375, "learning_rate": 6.163651315789474e-06, "loss": 0.1938, "step": 1866 }, { "epoch": 0.38384046052631576, "grad_norm": 8.862659454345703, "learning_rate": 6.161595394736842e-06, "loss": 0.0477, "step": 1867 }, { "epoch": 0.3840460526315789, "grad_norm": 7.8384690284729, "learning_rate": 6.159539473684211e-06, "loss": 0.0324, "step": 1868 }, { "epoch": 0.3842516447368421, "grad_norm": 22.055673599243164, "learning_rate": 6.157483552631579e-06, "loss": 0.2002, "step": 1869 }, { "epoch": 0.38445723684210525, "grad_norm": 37.05948257446289, "learning_rate": 6.155427631578947e-06, "loss": 0.3982, "step": 1870 }, { "epoch": 0.3846628289473684, "grad_norm": 67.2182846069336, "learning_rate": 6.153371710526316e-06, "loss": 0.6445, "step": 1871 }, { "epoch": 0.3848684210526316, "grad_norm": 100.71815490722656, "learning_rate": 6.151315789473685e-06, "loss": 1.5059, "step": 1872 }, { "epoch": 0.38507401315789475, "grad_norm": 78.71395874023438, "learning_rate": 6.1492598684210536e-06, "loss": 0.7715, "step": 1873 }, { "epoch": 0.3852796052631579, "grad_norm": 85.35462188720703, "learning_rate": 6.147203947368422e-06, "loss": 0.8641, "step": 1874 }, { "epoch": 0.3854851973684211, "grad_norm": 3.827369451522827, "learning_rate": 6.14514802631579e-06, "loss": 0.0329, "step": 1875 }, { "epoch": 0.38569078947368424, "grad_norm": 56.63518524169922, "learning_rate": 6.143092105263159e-06, "loss": 0.4741, "step": 1876 }, { "epoch": 0.38589638157894735, "grad_norm": 26.294586181640625, "learning_rate": 6.141036184210527e-06, "loss": 0.4614, "step": 1877 }, { "epoch": 0.3861019736842105, "grad_norm": 26.115215301513672, "learning_rate": 6.1389802631578955e-06, "loss": 0.1035, "step": 1878 }, { "epoch": 0.3863075657894737, "grad_norm": 53.036685943603516, "learning_rate": 6.136924342105263e-06, "loss": 0.5359, "step": 1879 }, { "epoch": 0.38651315789473684, "grad_norm": 28.333843231201172, "learning_rate": 6.134868421052632e-06, "loss": 0.1652, "step": 1880 }, { "epoch": 0.38671875, "grad_norm": 57.40380096435547, "learning_rate": 6.132812500000001e-06, "loss": 0.2822, "step": 1881 }, { "epoch": 0.38692434210526316, "grad_norm": 67.03347778320312, "learning_rate": 6.130756578947369e-06, "loss": 0.7043, "step": 1882 }, { "epoch": 0.3871299342105263, "grad_norm": 67.39081573486328, "learning_rate": 6.1287006578947374e-06, "loss": 0.7852, "step": 1883 }, { "epoch": 0.3873355263157895, "grad_norm": 14.187596321105957, "learning_rate": 6.126644736842105e-06, "loss": 0.1204, "step": 1884 }, { "epoch": 0.38754111842105265, "grad_norm": 89.09370422363281, "learning_rate": 6.124588815789474e-06, "loss": 1.5273, "step": 1885 }, { "epoch": 0.38774671052631576, "grad_norm": 66.48914337158203, "learning_rate": 6.122532894736842e-06, "loss": 1.1141, "step": 1886 }, { "epoch": 0.3879523026315789, "grad_norm": 12.095267295837402, "learning_rate": 6.120476973684211e-06, "loss": 0.0216, "step": 1887 }, { "epoch": 0.3881578947368421, "grad_norm": 71.29183197021484, "learning_rate": 6.118421052631579e-06, "loss": 0.2191, "step": 1888 }, { "epoch": 0.38836348684210525, "grad_norm": 40.22163772583008, "learning_rate": 6.116365131578947e-06, "loss": 0.2339, "step": 1889 }, { "epoch": 0.3885690789473684, "grad_norm": 34.40623474121094, "learning_rate": 6.114309210526316e-06, "loss": 0.1698, "step": 1890 }, { "epoch": 0.3887746710526316, "grad_norm": 101.75137329101562, "learning_rate": 6.1122532894736856e-06, "loss": 0.6465, "step": 1891 }, { "epoch": 0.38898026315789475, "grad_norm": 3.374587297439575, "learning_rate": 6.1101973684210535e-06, "loss": 0.0134, "step": 1892 }, { "epoch": 0.3891858552631579, "grad_norm": 204.31594848632812, "learning_rate": 6.108141447368422e-06, "loss": 2.2266, "step": 1893 }, { "epoch": 0.3893914473684211, "grad_norm": 24.310970306396484, "learning_rate": 6.10608552631579e-06, "loss": 0.3916, "step": 1894 }, { "epoch": 0.38959703947368424, "grad_norm": 94.7091064453125, "learning_rate": 6.104029605263159e-06, "loss": 2.0391, "step": 1895 }, { "epoch": 0.38980263157894735, "grad_norm": 308.0036926269531, "learning_rate": 6.101973684210527e-06, "loss": 3.8633, "step": 1896 }, { "epoch": 0.3900082236842105, "grad_norm": 33.45579147338867, "learning_rate": 6.099917763157895e-06, "loss": 0.3293, "step": 1897 }, { "epoch": 0.3902138157894737, "grad_norm": 184.68736267089844, "learning_rate": 6.097861842105264e-06, "loss": 2.3984, "step": 1898 }, { "epoch": 0.39041940789473684, "grad_norm": 26.24018096923828, "learning_rate": 6.095805921052632e-06, "loss": 0.1538, "step": 1899 }, { "epoch": 0.390625, "grad_norm": 97.77106475830078, "learning_rate": 6.093750000000001e-06, "loss": 0.8643, "step": 1900 }, { "epoch": 0.39083059210526316, "grad_norm": 39.89152145385742, "learning_rate": 6.091694078947369e-06, "loss": 0.7749, "step": 1901 }, { "epoch": 0.3910361842105263, "grad_norm": 16.946640014648438, "learning_rate": 6.089638157894737e-06, "loss": 0.1553, "step": 1902 }, { "epoch": 0.3912417763157895, "grad_norm": 93.54100036621094, "learning_rate": 6.087582236842105e-06, "loss": 1.4165, "step": 1903 }, { "epoch": 0.39144736842105265, "grad_norm": 194.84523010253906, "learning_rate": 6.085526315789474e-06, "loss": 4.6719, "step": 1904 }, { "epoch": 0.39165296052631576, "grad_norm": 13.809347152709961, "learning_rate": 6.083470394736843e-06, "loss": 0.1638, "step": 1905 }, { "epoch": 0.3918585526315789, "grad_norm": 273.7816162109375, "learning_rate": 6.0814144736842106e-06, "loss": 3.6719, "step": 1906 }, { "epoch": 0.3920641447368421, "grad_norm": 18.039457321166992, "learning_rate": 6.079358552631579e-06, "loss": 0.0795, "step": 1907 }, { "epoch": 0.39226973684210525, "grad_norm": 143.18167114257812, "learning_rate": 6.077302631578947e-06, "loss": 2.8125, "step": 1908 }, { "epoch": 0.3924753289473684, "grad_norm": 18.255138397216797, "learning_rate": 6.075246710526316e-06, "loss": 0.1539, "step": 1909 }, { "epoch": 0.3926809210526316, "grad_norm": 57.78819274902344, "learning_rate": 6.0731907894736855e-06, "loss": 0.5156, "step": 1910 }, { "epoch": 0.39288651315789475, "grad_norm": 108.34468841552734, "learning_rate": 6.071134868421053e-06, "loss": 0.759, "step": 1911 }, { "epoch": 0.3930921052631579, "grad_norm": 89.17839050292969, "learning_rate": 6.069078947368422e-06, "loss": 1.168, "step": 1912 }, { "epoch": 0.3932976973684211, "grad_norm": 130.67453002929688, "learning_rate": 6.06702302631579e-06, "loss": 1.7808, "step": 1913 }, { "epoch": 0.39350328947368424, "grad_norm": 24.981653213500977, "learning_rate": 6.064967105263159e-06, "loss": 0.1627, "step": 1914 }, { "epoch": 0.39370888157894735, "grad_norm": 19.34363555908203, "learning_rate": 6.062911184210527e-06, "loss": 0.0989, "step": 1915 }, { "epoch": 0.3939144736842105, "grad_norm": 55.64130401611328, "learning_rate": 6.060855263157895e-06, "loss": 1.0596, "step": 1916 }, { "epoch": 0.3941200657894737, "grad_norm": 51.68553924560547, "learning_rate": 6.058799342105264e-06, "loss": 0.6162, "step": 1917 }, { "epoch": 0.39432565789473684, "grad_norm": 27.38245964050293, "learning_rate": 6.056743421052632e-06, "loss": 0.1667, "step": 1918 }, { "epoch": 0.39453125, "grad_norm": 29.091188430786133, "learning_rate": 6.054687500000001e-06, "loss": 0.1492, "step": 1919 }, { "epoch": 0.39473684210526316, "grad_norm": 41.413543701171875, "learning_rate": 6.0526315789473685e-06, "loss": 0.3979, "step": 1920 }, { "epoch": 0.3949424342105263, "grad_norm": 129.63726806640625, "learning_rate": 6.050575657894737e-06, "loss": 0.8589, "step": 1921 }, { "epoch": 0.3951480263157895, "grad_norm": 97.42372131347656, "learning_rate": 6.048519736842106e-06, "loss": 1.3789, "step": 1922 }, { "epoch": 0.39535361842105265, "grad_norm": 162.7395782470703, "learning_rate": 6.046463815789474e-06, "loss": 2.5391, "step": 1923 }, { "epoch": 0.39555921052631576, "grad_norm": 127.71554565429688, "learning_rate": 6.044407894736843e-06, "loss": 1.2246, "step": 1924 }, { "epoch": 0.3957648026315789, "grad_norm": 114.12483215332031, "learning_rate": 6.0423519736842105e-06, "loss": 1.5645, "step": 1925 }, { "epoch": 0.3959703947368421, "grad_norm": 324.4521789550781, "learning_rate": 6.040296052631579e-06, "loss": 2.1768, "step": 1926 }, { "epoch": 0.39617598684210525, "grad_norm": 111.8772201538086, "learning_rate": 6.038240131578947e-06, "loss": 1.6094, "step": 1927 }, { "epoch": 0.3963815789473684, "grad_norm": 36.31013107299805, "learning_rate": 6.036184210526316e-06, "loss": 0.2285, "step": 1928 }, { "epoch": 0.3965871710526316, "grad_norm": 38.53400421142578, "learning_rate": 6.034128289473685e-06, "loss": 0.3877, "step": 1929 }, { "epoch": 0.39679276315789475, "grad_norm": 23.10186004638672, "learning_rate": 6.032072368421053e-06, "loss": 0.1375, "step": 1930 }, { "epoch": 0.3969983552631579, "grad_norm": 46.2637939453125, "learning_rate": 6.030016447368422e-06, "loss": 0.4019, "step": 1931 }, { "epoch": 0.3972039473684211, "grad_norm": 6.784730434417725, "learning_rate": 6.027960526315791e-06, "loss": 0.0369, "step": 1932 }, { "epoch": 0.39740953947368424, "grad_norm": 115.88247680664062, "learning_rate": 6.025904605263159e-06, "loss": 1.4473, "step": 1933 }, { "epoch": 0.39761513157894735, "grad_norm": 33.31977844238281, "learning_rate": 6.023848684210527e-06, "loss": 0.2158, "step": 1934 }, { "epoch": 0.3978207236842105, "grad_norm": 72.49617767333984, "learning_rate": 6.021792763157895e-06, "loss": 0.8049, "step": 1935 }, { "epoch": 0.3980263157894737, "grad_norm": 91.0794677734375, "learning_rate": 6.019736842105264e-06, "loss": 1.0527, "step": 1936 }, { "epoch": 0.39823190789473684, "grad_norm": 54.061458587646484, "learning_rate": 6.017680921052632e-06, "loss": 0.3125, "step": 1937 }, { "epoch": 0.3984375, "grad_norm": 71.1333236694336, "learning_rate": 6.0156250000000005e-06, "loss": 0.5972, "step": 1938 }, { "epoch": 0.39864309210526316, "grad_norm": 39.52738571166992, "learning_rate": 6.013569078947369e-06, "loss": 0.9377, "step": 1939 }, { "epoch": 0.3988486842105263, "grad_norm": 56.291866302490234, "learning_rate": 6.011513157894737e-06, "loss": 0.6895, "step": 1940 }, { "epoch": 0.3990542763157895, "grad_norm": 114.96957397460938, "learning_rate": 6.009457236842106e-06, "loss": 1.2559, "step": 1941 }, { "epoch": 0.39925986842105265, "grad_norm": 23.858776092529297, "learning_rate": 6.007401315789474e-06, "loss": 0.2632, "step": 1942 }, { "epoch": 0.39946546052631576, "grad_norm": 64.88748168945312, "learning_rate": 6.0053453947368425e-06, "loss": 0.4805, "step": 1943 }, { "epoch": 0.3996710526315789, "grad_norm": 24.509737014770508, "learning_rate": 6.00328947368421e-06, "loss": 0.2065, "step": 1944 }, { "epoch": 0.3998766447368421, "grad_norm": 15.895304679870605, "learning_rate": 6.001233552631579e-06, "loss": 0.0484, "step": 1945 }, { "epoch": 0.40008223684210525, "grad_norm": 68.4259033203125, "learning_rate": 5.999177631578948e-06, "loss": 0.5746, "step": 1946 }, { "epoch": 0.4002878289473684, "grad_norm": 53.81132888793945, "learning_rate": 5.997121710526316e-06, "loss": 0.6599, "step": 1947 }, { "epoch": 0.4004934210526316, "grad_norm": 60.51774215698242, "learning_rate": 5.995065789473685e-06, "loss": 1.1738, "step": 1948 }, { "epoch": 0.40069901315789475, "grad_norm": 12.07656192779541, "learning_rate": 5.993009868421054e-06, "loss": 0.0807, "step": 1949 }, { "epoch": 0.4009046052631579, "grad_norm": 76.57506561279297, "learning_rate": 5.990953947368422e-06, "loss": 0.8555, "step": 1950 }, { "epoch": 0.4011101973684211, "grad_norm": 250.23104858398438, "learning_rate": 5.988898026315791e-06, "loss": 2.6658, "step": 1951 }, { "epoch": 0.40131578947368424, "grad_norm": 80.83528137207031, "learning_rate": 5.9868421052631585e-06, "loss": 0.9906, "step": 1952 }, { "epoch": 0.40152138157894735, "grad_norm": 94.2220687866211, "learning_rate": 5.984786184210527e-06, "loss": 1.8049, "step": 1953 }, { "epoch": 0.4017269736842105, "grad_norm": 5.878194808959961, "learning_rate": 5.982730263157895e-06, "loss": 0.0316, "step": 1954 }, { "epoch": 0.4019325657894737, "grad_norm": 56.587562561035156, "learning_rate": 5.980674342105264e-06, "loss": 0.7676, "step": 1955 }, { "epoch": 0.40213815789473684, "grad_norm": 62.51740646362305, "learning_rate": 5.9786184210526325e-06, "loss": 0.6133, "step": 1956 }, { "epoch": 0.40234375, "grad_norm": 41.449371337890625, "learning_rate": 5.9765625000000004e-06, "loss": 0.4033, "step": 1957 }, { "epoch": 0.40254934210526316, "grad_norm": 31.63766098022461, "learning_rate": 5.974506578947369e-06, "loss": 0.1582, "step": 1958 }, { "epoch": 0.4027549342105263, "grad_norm": 24.43867301940918, "learning_rate": 5.972450657894737e-06, "loss": 0.2021, "step": 1959 }, { "epoch": 0.4029605263157895, "grad_norm": 64.72654724121094, "learning_rate": 5.970394736842106e-06, "loss": 0.5664, "step": 1960 }, { "epoch": 0.40316611842105265, "grad_norm": 54.762882232666016, "learning_rate": 5.968338815789474e-06, "loss": 0.4645, "step": 1961 }, { "epoch": 0.40337171052631576, "grad_norm": 78.1889877319336, "learning_rate": 5.966282894736842e-06, "loss": 0.9893, "step": 1962 }, { "epoch": 0.4035773026315789, "grad_norm": 122.76702880859375, "learning_rate": 5.96422697368421e-06, "loss": 1.5645, "step": 1963 }, { "epoch": 0.4037828947368421, "grad_norm": 130.9840545654297, "learning_rate": 5.962171052631579e-06, "loss": 1.3723, "step": 1964 }, { "epoch": 0.40398848684210525, "grad_norm": 33.82376480102539, "learning_rate": 5.960115131578948e-06, "loss": 0.3169, "step": 1965 }, { "epoch": 0.4041940789473684, "grad_norm": 117.71057891845703, "learning_rate": 5.958059210526316e-06, "loss": 1.9615, "step": 1966 }, { "epoch": 0.4043996710526316, "grad_norm": 23.733036041259766, "learning_rate": 5.956003289473685e-06, "loss": 0.0857, "step": 1967 }, { "epoch": 0.40460526315789475, "grad_norm": 236.36920166015625, "learning_rate": 5.953947368421054e-06, "loss": 1.1855, "step": 1968 }, { "epoch": 0.4048108552631579, "grad_norm": 27.331249237060547, "learning_rate": 5.951891447368422e-06, "loss": 0.2297, "step": 1969 }, { "epoch": 0.4050164473684211, "grad_norm": 42.75347137451172, "learning_rate": 5.9498355263157905e-06, "loss": 0.4163, "step": 1970 }, { "epoch": 0.40522203947368424, "grad_norm": 96.48021697998047, "learning_rate": 5.947779605263158e-06, "loss": 0.8516, "step": 1971 }, { "epoch": 0.40542763157894735, "grad_norm": 37.10918426513672, "learning_rate": 5.945723684210527e-06, "loss": 0.1375, "step": 1972 }, { "epoch": 0.4056332236842105, "grad_norm": 140.38882446289062, "learning_rate": 5.943667763157895e-06, "loss": 1.3975, "step": 1973 }, { "epoch": 0.4058388157894737, "grad_norm": 46.91209030151367, "learning_rate": 5.941611842105264e-06, "loss": 0.3213, "step": 1974 }, { "epoch": 0.40604440789473684, "grad_norm": 140.69073486328125, "learning_rate": 5.9395559210526324e-06, "loss": 1.7305, "step": 1975 }, { "epoch": 0.40625, "grad_norm": 108.95596313476562, "learning_rate": 5.9375e-06, "loss": 0.8838, "step": 1976 }, { "epoch": 0.40645559210526316, "grad_norm": 126.52888488769531, "learning_rate": 5.935444078947369e-06, "loss": 1.4375, "step": 1977 }, { "epoch": 0.4066611842105263, "grad_norm": 26.320322036743164, "learning_rate": 5.933388157894737e-06, "loss": 0.1606, "step": 1978 }, { "epoch": 0.4068667763157895, "grad_norm": 45.55905532836914, "learning_rate": 5.931332236842106e-06, "loss": 0.2952, "step": 1979 }, { "epoch": 0.40707236842105265, "grad_norm": 143.04432678222656, "learning_rate": 5.9292763157894735e-06, "loss": 1.0508, "step": 1980 }, { "epoch": 0.40727796052631576, "grad_norm": 35.580169677734375, "learning_rate": 5.927220394736842e-06, "loss": 0.148, "step": 1981 }, { "epoch": 0.4074835526315789, "grad_norm": 128.48410034179688, "learning_rate": 5.925164473684211e-06, "loss": 1.4258, "step": 1982 }, { "epoch": 0.4076891447368421, "grad_norm": 48.066165924072266, "learning_rate": 5.923108552631579e-06, "loss": 0.2597, "step": 1983 }, { "epoch": 0.40789473684210525, "grad_norm": 126.2625961303711, "learning_rate": 5.921052631578948e-06, "loss": 1.8047, "step": 1984 }, { "epoch": 0.4081003289473684, "grad_norm": 17.726171493530273, "learning_rate": 5.9189967105263155e-06, "loss": 0.0842, "step": 1985 }, { "epoch": 0.4083059210526316, "grad_norm": 12.546233177185059, "learning_rate": 5.916940789473685e-06, "loss": 0.0542, "step": 1986 }, { "epoch": 0.40851151315789475, "grad_norm": 78.86286926269531, "learning_rate": 5.914884868421054e-06, "loss": 0.4822, "step": 1987 }, { "epoch": 0.4087171052631579, "grad_norm": 67.45187377929688, "learning_rate": 5.912828947368422e-06, "loss": 0.6777, "step": 1988 }, { "epoch": 0.4089226973684211, "grad_norm": 38.1794319152832, "learning_rate": 5.91077302631579e-06, "loss": 0.3433, "step": 1989 }, { "epoch": 0.40912828947368424, "grad_norm": 18.428997039794922, "learning_rate": 5.908717105263158e-06, "loss": 0.1299, "step": 1990 }, { "epoch": 0.40933388157894735, "grad_norm": 60.26972579956055, "learning_rate": 5.906661184210527e-06, "loss": 0.4585, "step": 1991 }, { "epoch": 0.4095394736842105, "grad_norm": 26.902416229248047, "learning_rate": 5.904605263157896e-06, "loss": 0.1309, "step": 1992 }, { "epoch": 0.4097450657894737, "grad_norm": 56.15910339355469, "learning_rate": 5.902549342105264e-06, "loss": 0.3525, "step": 1993 }, { "epoch": 0.40995065789473684, "grad_norm": 36.257266998291016, "learning_rate": 5.900493421052632e-06, "loss": 0.1438, "step": 1994 }, { "epoch": 0.41015625, "grad_norm": 74.58038330078125, "learning_rate": 5.8984375e-06, "loss": 1.3447, "step": 1995 }, { "epoch": 0.41036184210526316, "grad_norm": 115.14432525634766, "learning_rate": 5.896381578947369e-06, "loss": 0.7256, "step": 1996 }, { "epoch": 0.4105674342105263, "grad_norm": 88.9087905883789, "learning_rate": 5.894325657894737e-06, "loss": 0.6221, "step": 1997 }, { "epoch": 0.4107730263157895, "grad_norm": 9.664305686950684, "learning_rate": 5.8922697368421056e-06, "loss": 0.0607, "step": 1998 }, { "epoch": 0.41097861842105265, "grad_norm": 55.12565231323242, "learning_rate": 5.890213815789474e-06, "loss": 0.4435, "step": 1999 }, { "epoch": 0.41118421052631576, "grad_norm": 74.54291534423828, "learning_rate": 5.888157894736842e-06, "loss": 0.4945, "step": 2000 }, { "epoch": 0.4113898026315789, "grad_norm": 94.62518310546875, "learning_rate": 5.886101973684211e-06, "loss": 1.0234, "step": 2001 }, { "epoch": 0.4115953947368421, "grad_norm": 284.4317932128906, "learning_rate": 5.884046052631579e-06, "loss": 1.259, "step": 2002 }, { "epoch": 0.41180098684210525, "grad_norm": 61.806522369384766, "learning_rate": 5.8819901315789475e-06, "loss": 0.3986, "step": 2003 }, { "epoch": 0.4120065789473684, "grad_norm": 26.955337524414062, "learning_rate": 5.879934210526315e-06, "loss": 0.2015, "step": 2004 }, { "epoch": 0.4122121710526316, "grad_norm": 131.27210998535156, "learning_rate": 5.877878289473685e-06, "loss": 1.3611, "step": 2005 }, { "epoch": 0.41241776315789475, "grad_norm": 171.38475036621094, "learning_rate": 5.875822368421054e-06, "loss": 1.5447, "step": 2006 }, { "epoch": 0.4126233552631579, "grad_norm": 111.6689453125, "learning_rate": 5.8737664473684216e-06, "loss": 0.8379, "step": 2007 }, { "epoch": 0.4128289473684211, "grad_norm": 97.95816802978516, "learning_rate": 5.87171052631579e-06, "loss": 1.4375, "step": 2008 }, { "epoch": 0.41303453947368424, "grad_norm": 73.59220886230469, "learning_rate": 5.869654605263159e-06, "loss": 0.3438, "step": 2009 }, { "epoch": 0.41324013157894735, "grad_norm": 190.30267333984375, "learning_rate": 5.867598684210527e-06, "loss": 1.7915, "step": 2010 }, { "epoch": 0.4134457236842105, "grad_norm": 67.83707427978516, "learning_rate": 5.865542763157896e-06, "loss": 0.7688, "step": 2011 }, { "epoch": 0.4136513157894737, "grad_norm": 43.900177001953125, "learning_rate": 5.8634868421052635e-06, "loss": 0.3066, "step": 2012 }, { "epoch": 0.41385690789473684, "grad_norm": 46.86185073852539, "learning_rate": 5.861430921052632e-06, "loss": 0.7266, "step": 2013 }, { "epoch": 0.4140625, "grad_norm": 57.41328811645508, "learning_rate": 5.859375e-06, "loss": 0.7686, "step": 2014 }, { "epoch": 0.41426809210526316, "grad_norm": 29.013418197631836, "learning_rate": 5.857319078947369e-06, "loss": 0.2715, "step": 2015 }, { "epoch": 0.4144736842105263, "grad_norm": 44.03178024291992, "learning_rate": 5.855263157894738e-06, "loss": 0.2642, "step": 2016 }, { "epoch": 0.4146792763157895, "grad_norm": 154.146240234375, "learning_rate": 5.8532072368421055e-06, "loss": 2.1016, "step": 2017 }, { "epoch": 0.41488486842105265, "grad_norm": 87.9669418334961, "learning_rate": 5.851151315789474e-06, "loss": 0.8447, "step": 2018 }, { "epoch": 0.41509046052631576, "grad_norm": 41.30827331542969, "learning_rate": 5.849095394736842e-06, "loss": 0.368, "step": 2019 }, { "epoch": 0.4152960526315789, "grad_norm": 10.340839385986328, "learning_rate": 5.847039473684211e-06, "loss": 0.0267, "step": 2020 }, { "epoch": 0.4155016447368421, "grad_norm": 90.07345581054688, "learning_rate": 5.844983552631579e-06, "loss": 0.9292, "step": 2021 }, { "epoch": 0.41570723684210525, "grad_norm": 184.1570281982422, "learning_rate": 5.842927631578947e-06, "loss": 3.1406, "step": 2022 }, { "epoch": 0.4159128289473684, "grad_norm": 13.752165794372559, "learning_rate": 5.840871710526316e-06, "loss": 0.0776, "step": 2023 }, { "epoch": 0.4161184210526316, "grad_norm": 29.639036178588867, "learning_rate": 5.838815789473685e-06, "loss": 0.3931, "step": 2024 }, { "epoch": 0.41632401315789475, "grad_norm": 146.849853515625, "learning_rate": 5.836759868421054e-06, "loss": 3.0781, "step": 2025 }, { "epoch": 0.4165296052631579, "grad_norm": 101.57930755615234, "learning_rate": 5.834703947368422e-06, "loss": 1.0273, "step": 2026 }, { "epoch": 0.4167351973684211, "grad_norm": 28.573368072509766, "learning_rate": 5.83264802631579e-06, "loss": 0.2109, "step": 2027 }, { "epoch": 0.41694078947368424, "grad_norm": 30.397846221923828, "learning_rate": 5.830592105263159e-06, "loss": 0.2025, "step": 2028 }, { "epoch": 0.41714638157894735, "grad_norm": 48.90084457397461, "learning_rate": 5.828536184210527e-06, "loss": 0.4346, "step": 2029 }, { "epoch": 0.4173519736842105, "grad_norm": 152.56224060058594, "learning_rate": 5.8264802631578955e-06, "loss": 1.457, "step": 2030 }, { "epoch": 0.4175575657894737, "grad_norm": 125.71556091308594, "learning_rate": 5.824424342105263e-06, "loss": 2.168, "step": 2031 }, { "epoch": 0.41776315789473684, "grad_norm": 35.15667724609375, "learning_rate": 5.822368421052632e-06, "loss": 0.6621, "step": 2032 }, { "epoch": 0.41796875, "grad_norm": 38.99718475341797, "learning_rate": 5.820312500000001e-06, "loss": 0.2925, "step": 2033 }, { "epoch": 0.41817434210526316, "grad_norm": 47.8563117980957, "learning_rate": 5.818256578947369e-06, "loss": 0.3867, "step": 2034 }, { "epoch": 0.4183799342105263, "grad_norm": 144.9488067626953, "learning_rate": 5.8162006578947375e-06, "loss": 2.875, "step": 2035 }, { "epoch": 0.4185855263157895, "grad_norm": 85.16053771972656, "learning_rate": 5.814144736842105e-06, "loss": 1.5957, "step": 2036 }, { "epoch": 0.41879111842105265, "grad_norm": 215.34422302246094, "learning_rate": 5.812088815789474e-06, "loss": 1.4199, "step": 2037 }, { "epoch": 0.41899671052631576, "grad_norm": 47.046939849853516, "learning_rate": 5.810032894736842e-06, "loss": 0.2812, "step": 2038 }, { "epoch": 0.4192023026315789, "grad_norm": 20.7152042388916, "learning_rate": 5.807976973684211e-06, "loss": 0.1475, "step": 2039 }, { "epoch": 0.4194078947368421, "grad_norm": 93.45299530029297, "learning_rate": 5.805921052631579e-06, "loss": 0.8125, "step": 2040 }, { "epoch": 0.41961348684210525, "grad_norm": 74.42697143554688, "learning_rate": 5.803865131578947e-06, "loss": 1.23, "step": 2041 }, { "epoch": 0.4198190789473684, "grad_norm": 62.83835983276367, "learning_rate": 5.801809210526316e-06, "loss": 0.6035, "step": 2042 }, { "epoch": 0.4200246710526316, "grad_norm": 76.1596450805664, "learning_rate": 5.799753289473686e-06, "loss": 1.0654, "step": 2043 }, { "epoch": 0.42023026315789475, "grad_norm": 151.00640869140625, "learning_rate": 5.7976973684210535e-06, "loss": 1.8281, "step": 2044 }, { "epoch": 0.4204358552631579, "grad_norm": 64.20552825927734, "learning_rate": 5.795641447368422e-06, "loss": 0.959, "step": 2045 }, { "epoch": 0.4206414473684211, "grad_norm": 97.44773864746094, "learning_rate": 5.79358552631579e-06, "loss": 1.3345, "step": 2046 }, { "epoch": 0.42084703947368424, "grad_norm": 22.996286392211914, "learning_rate": 5.791529605263159e-06, "loss": 0.2241, "step": 2047 }, { "epoch": 0.42105263157894735, "grad_norm": 74.41117095947266, "learning_rate": 5.789473684210527e-06, "loss": 1.1602, "step": 2048 }, { "epoch": 0.4212582236842105, "grad_norm": 26.629070281982422, "learning_rate": 5.7874177631578954e-06, "loss": 0.1179, "step": 2049 }, { "epoch": 0.4214638157894737, "grad_norm": 129.08856201171875, "learning_rate": 5.785361842105264e-06, "loss": 2.6562, "step": 2050 }, { "epoch": 0.42166940789473684, "grad_norm": 89.20227813720703, "learning_rate": 5.783305921052632e-06, "loss": 0.6709, "step": 2051 }, { "epoch": 0.421875, "grad_norm": 12.063786506652832, "learning_rate": 5.781250000000001e-06, "loss": 0.0508, "step": 2052 }, { "epoch": 0.42208059210526316, "grad_norm": 138.15719604492188, "learning_rate": 5.779194078947369e-06, "loss": 2.3906, "step": 2053 }, { "epoch": 0.4222861842105263, "grad_norm": 23.493499755859375, "learning_rate": 5.777138157894737e-06, "loss": 0.1447, "step": 2054 }, { "epoch": 0.4224917763157895, "grad_norm": 95.48187255859375, "learning_rate": 5.775082236842105e-06, "loss": 1.2715, "step": 2055 }, { "epoch": 0.42269736842105265, "grad_norm": 34.968849182128906, "learning_rate": 5.773026315789474e-06, "loss": 0.2063, "step": 2056 }, { "epoch": 0.42290296052631576, "grad_norm": 109.55110931396484, "learning_rate": 5.770970394736843e-06, "loss": 0.9297, "step": 2057 }, { "epoch": 0.4231085526315789, "grad_norm": 16.702089309692383, "learning_rate": 5.768914473684211e-06, "loss": 0.0675, "step": 2058 }, { "epoch": 0.4233141447368421, "grad_norm": 22.763612747192383, "learning_rate": 5.766858552631579e-06, "loss": 0.1033, "step": 2059 }, { "epoch": 0.42351973684210525, "grad_norm": 18.945098876953125, "learning_rate": 5.764802631578947e-06, "loss": 0.1561, "step": 2060 }, { "epoch": 0.4237253289473684, "grad_norm": 24.070188522338867, "learning_rate": 5.762746710526316e-06, "loss": 0.1616, "step": 2061 }, { "epoch": 0.4239309210526316, "grad_norm": 66.78330993652344, "learning_rate": 5.7606907894736855e-06, "loss": 0.7656, "step": 2062 }, { "epoch": 0.42413651315789475, "grad_norm": 11.192353248596191, "learning_rate": 5.758634868421053e-06, "loss": 0.0465, "step": 2063 }, { "epoch": 0.4243421052631579, "grad_norm": 54.64042282104492, "learning_rate": 5.756578947368422e-06, "loss": 0.4873, "step": 2064 }, { "epoch": 0.4245476973684211, "grad_norm": 44.8065299987793, "learning_rate": 5.75452302631579e-06, "loss": 0.2368, "step": 2065 }, { "epoch": 0.42475328947368424, "grad_norm": 53.94464111328125, "learning_rate": 5.752467105263159e-06, "loss": 0.2361, "step": 2066 }, { "epoch": 0.42495888157894735, "grad_norm": 98.46588134765625, "learning_rate": 5.750411184210527e-06, "loss": 0.7773, "step": 2067 }, { "epoch": 0.4251644736842105, "grad_norm": 131.31756591796875, "learning_rate": 5.748355263157895e-06, "loss": 2.248, "step": 2068 }, { "epoch": 0.4253700657894737, "grad_norm": 33.413787841796875, "learning_rate": 5.746299342105264e-06, "loss": 0.1449, "step": 2069 }, { "epoch": 0.42557565789473684, "grad_norm": 27.16778564453125, "learning_rate": 5.744243421052632e-06, "loss": 0.3179, "step": 2070 }, { "epoch": 0.42578125, "grad_norm": 41.874267578125, "learning_rate": 5.742187500000001e-06, "loss": 0.2378, "step": 2071 }, { "epoch": 0.42598684210526316, "grad_norm": 49.253047943115234, "learning_rate": 5.7401315789473685e-06, "loss": 0.3064, "step": 2072 }, { "epoch": 0.4261924342105263, "grad_norm": 92.310546875, "learning_rate": 5.738075657894737e-06, "loss": 1.2427, "step": 2073 }, { "epoch": 0.4263980263157895, "grad_norm": 43.624366760253906, "learning_rate": 5.736019736842105e-06, "loss": 0.3506, "step": 2074 }, { "epoch": 0.42660361842105265, "grad_norm": 191.864013671875, "learning_rate": 5.733963815789474e-06, "loss": 1.5508, "step": 2075 }, { "epoch": 0.42680921052631576, "grad_norm": 58.576297760009766, "learning_rate": 5.731907894736843e-06, "loss": 0.3945, "step": 2076 }, { "epoch": 0.4270148026315789, "grad_norm": 69.74302673339844, "learning_rate": 5.7298519736842105e-06, "loss": 0.7012, "step": 2077 }, { "epoch": 0.4272203947368421, "grad_norm": 66.51490783691406, "learning_rate": 5.727796052631579e-06, "loss": 0.5413, "step": 2078 }, { "epoch": 0.42742598684210525, "grad_norm": 50.662269592285156, "learning_rate": 5.725740131578947e-06, "loss": 0.3467, "step": 2079 }, { "epoch": 0.4276315789473684, "grad_norm": 181.9344940185547, "learning_rate": 5.723684210526316e-06, "loss": 1.453, "step": 2080 }, { "epoch": 0.4278371710526316, "grad_norm": 73.70962524414062, "learning_rate": 5.721628289473685e-06, "loss": 0.6318, "step": 2081 }, { "epoch": 0.42804276315789475, "grad_norm": 23.34363555908203, "learning_rate": 5.719572368421053e-06, "loss": 0.1616, "step": 2082 }, { "epoch": 0.4282483552631579, "grad_norm": 65.03227233886719, "learning_rate": 5.717516447368422e-06, "loss": 0.6301, "step": 2083 }, { "epoch": 0.4284539473684211, "grad_norm": 71.48735809326172, "learning_rate": 5.71546052631579e-06, "loss": 0.3336, "step": 2084 }, { "epoch": 0.42865953947368424, "grad_norm": 104.86648559570312, "learning_rate": 5.713404605263159e-06, "loss": 0.8085, "step": 2085 }, { "epoch": 0.42886513157894735, "grad_norm": 34.47964859008789, "learning_rate": 5.711348684210527e-06, "loss": 0.5479, "step": 2086 }, { "epoch": 0.4290707236842105, "grad_norm": 5.01019811630249, "learning_rate": 5.709292763157895e-06, "loss": 0.0143, "step": 2087 }, { "epoch": 0.4292763157894737, "grad_norm": 39.47612762451172, "learning_rate": 5.707236842105264e-06, "loss": 0.2468, "step": 2088 }, { "epoch": 0.42948190789473684, "grad_norm": 49.73085403442383, "learning_rate": 5.705180921052632e-06, "loss": 0.3071, "step": 2089 }, { "epoch": 0.4296875, "grad_norm": 43.12952423095703, "learning_rate": 5.7031250000000006e-06, "loss": 0.3721, "step": 2090 }, { "epoch": 0.42989309210526316, "grad_norm": 145.12277221679688, "learning_rate": 5.7010690789473684e-06, "loss": 1.6328, "step": 2091 }, { "epoch": 0.4300986842105263, "grad_norm": 36.11959457397461, "learning_rate": 5.699013157894737e-06, "loss": 0.2529, "step": 2092 }, { "epoch": 0.4303042763157895, "grad_norm": 104.68338775634766, "learning_rate": 5.696957236842106e-06, "loss": 1.8594, "step": 2093 }, { "epoch": 0.43050986842105265, "grad_norm": 14.242650032043457, "learning_rate": 5.694901315789474e-06, "loss": 0.0427, "step": 2094 }, { "epoch": 0.43071546052631576, "grad_norm": 17.324167251586914, "learning_rate": 5.6928453947368425e-06, "loss": 0.0894, "step": 2095 }, { "epoch": 0.4309210526315789, "grad_norm": 200.8926239013672, "learning_rate": 5.69078947368421e-06, "loss": 1.5093, "step": 2096 }, { "epoch": 0.4311266447368421, "grad_norm": 70.89171600341797, "learning_rate": 5.688733552631579e-06, "loss": 1.4785, "step": 2097 }, { "epoch": 0.43133223684210525, "grad_norm": 26.85736083984375, "learning_rate": 5.686677631578947e-06, "loss": 0.1709, "step": 2098 }, { "epoch": 0.4315378289473684, "grad_norm": 160.5074462890625, "learning_rate": 5.684621710526316e-06, "loss": 2.0295, "step": 2099 }, { "epoch": 0.4317434210526316, "grad_norm": 94.33827209472656, "learning_rate": 5.682565789473685e-06, "loss": 1.0605, "step": 2100 }, { "epoch": 0.43194901315789475, "grad_norm": 98.50626373291016, "learning_rate": 5.680509868421053e-06, "loss": 1.918, "step": 2101 }, { "epoch": 0.4321546052631579, "grad_norm": 111.65919494628906, "learning_rate": 5.678453947368422e-06, "loss": 1.4553, "step": 2102 }, { "epoch": 0.4323601973684211, "grad_norm": 21.47045135498047, "learning_rate": 5.676398026315791e-06, "loss": 0.2742, "step": 2103 }, { "epoch": 0.43256578947368424, "grad_norm": 47.1291618347168, "learning_rate": 5.6743421052631585e-06, "loss": 0.7529, "step": 2104 }, { "epoch": 0.43277138157894735, "grad_norm": 27.98121452331543, "learning_rate": 5.672286184210527e-06, "loss": 0.1782, "step": 2105 }, { "epoch": 0.4329769736842105, "grad_norm": 31.919300079345703, "learning_rate": 5.670230263157895e-06, "loss": 0.4547, "step": 2106 }, { "epoch": 0.4331825657894737, "grad_norm": 55.177059173583984, "learning_rate": 5.668174342105264e-06, "loss": 0.641, "step": 2107 }, { "epoch": 0.43338815789473684, "grad_norm": 43.884613037109375, "learning_rate": 5.666118421052632e-06, "loss": 0.6021, "step": 2108 }, { "epoch": 0.43359375, "grad_norm": 102.81952667236328, "learning_rate": 5.6640625000000005e-06, "loss": 1.4248, "step": 2109 }, { "epoch": 0.43379934210526316, "grad_norm": 42.11323928833008, "learning_rate": 5.662006578947369e-06, "loss": 0.2637, "step": 2110 }, { "epoch": 0.4340049342105263, "grad_norm": 55.570472717285156, "learning_rate": 5.659950657894737e-06, "loss": 0.4409, "step": 2111 }, { "epoch": 0.4342105263157895, "grad_norm": 69.68585968017578, "learning_rate": 5.657894736842106e-06, "loss": 0.7285, "step": 2112 }, { "epoch": 0.43441611842105265, "grad_norm": 120.17338562011719, "learning_rate": 5.655838815789474e-06, "loss": 1.6504, "step": 2113 }, { "epoch": 0.43462171052631576, "grad_norm": 74.65156555175781, "learning_rate": 5.653782894736842e-06, "loss": 0.9336, "step": 2114 }, { "epoch": 0.4348273026315789, "grad_norm": 52.08017349243164, "learning_rate": 5.65172697368421e-06, "loss": 0.3447, "step": 2115 }, { "epoch": 0.4350328947368421, "grad_norm": 57.68596649169922, "learning_rate": 5.649671052631579e-06, "loss": 0.4504, "step": 2116 }, { "epoch": 0.43523848684210525, "grad_norm": 59.85456848144531, "learning_rate": 5.647615131578948e-06, "loss": 0.8633, "step": 2117 }, { "epoch": 0.4354440789473684, "grad_norm": 73.73190307617188, "learning_rate": 5.645559210526316e-06, "loss": 0.7178, "step": 2118 }, { "epoch": 0.4356496710526316, "grad_norm": 24.267845153808594, "learning_rate": 5.643503289473685e-06, "loss": 0.2529, "step": 2119 }, { "epoch": 0.43585526315789475, "grad_norm": 47.07035827636719, "learning_rate": 5.641447368421054e-06, "loss": 0.5186, "step": 2120 }, { "epoch": 0.4360608552631579, "grad_norm": 66.62100219726562, "learning_rate": 5.639391447368422e-06, "loss": 0.9095, "step": 2121 }, { "epoch": 0.4362664473684211, "grad_norm": 169.6749725341797, "learning_rate": 5.6373355263157905e-06, "loss": 2.2656, "step": 2122 }, { "epoch": 0.43647203947368424, "grad_norm": 29.210668563842773, "learning_rate": 5.635279605263158e-06, "loss": 0.1976, "step": 2123 }, { "epoch": 0.43667763157894735, "grad_norm": 69.50837707519531, "learning_rate": 5.633223684210527e-06, "loss": 0.6526, "step": 2124 }, { "epoch": 0.4368832236842105, "grad_norm": 76.24977111816406, "learning_rate": 5.631167763157895e-06, "loss": 0.7385, "step": 2125 }, { "epoch": 0.4370888157894737, "grad_norm": 65.69280242919922, "learning_rate": 5.629111842105264e-06, "loss": 0.9214, "step": 2126 }, { "epoch": 0.43729440789473684, "grad_norm": 61.347896575927734, "learning_rate": 5.6270559210526325e-06, "loss": 0.3179, "step": 2127 }, { "epoch": 0.4375, "grad_norm": 84.578369140625, "learning_rate": 5.625e-06, "loss": 1.6699, "step": 2128 }, { "epoch": 0.43770559210526316, "grad_norm": 13.07283878326416, "learning_rate": 5.622944078947369e-06, "loss": 0.0533, "step": 2129 }, { "epoch": 0.4379111842105263, "grad_norm": 50.13556671142578, "learning_rate": 5.620888157894737e-06, "loss": 0.7407, "step": 2130 }, { "epoch": 0.4381167763157895, "grad_norm": 118.19499969482422, "learning_rate": 5.618832236842106e-06, "loss": 1.3125, "step": 2131 }, { "epoch": 0.43832236842105265, "grad_norm": 14.575752258300781, "learning_rate": 5.6167763157894736e-06, "loss": 0.0833, "step": 2132 }, { "epoch": 0.43852796052631576, "grad_norm": 9.551566123962402, "learning_rate": 5.614720394736842e-06, "loss": 0.0421, "step": 2133 }, { "epoch": 0.4387335526315789, "grad_norm": 33.860904693603516, "learning_rate": 5.612664473684211e-06, "loss": 0.1727, "step": 2134 }, { "epoch": 0.4389391447368421, "grad_norm": 16.320707321166992, "learning_rate": 5.610608552631579e-06, "loss": 0.1987, "step": 2135 }, { "epoch": 0.43914473684210525, "grad_norm": 114.37794494628906, "learning_rate": 5.608552631578948e-06, "loss": 2.1523, "step": 2136 }, { "epoch": 0.4393503289473684, "grad_norm": 171.62841796875, "learning_rate": 5.6064967105263155e-06, "loss": 3.168, "step": 2137 }, { "epoch": 0.4395559210526316, "grad_norm": 40.550201416015625, "learning_rate": 5.604440789473685e-06, "loss": 0.3657, "step": 2138 }, { "epoch": 0.43976151315789475, "grad_norm": 45.98616027832031, "learning_rate": 5.602384868421054e-06, "loss": 0.3867, "step": 2139 }, { "epoch": 0.4399671052631579, "grad_norm": 36.116458892822266, "learning_rate": 5.600328947368422e-06, "loss": 0.2275, "step": 2140 }, { "epoch": 0.4401726973684211, "grad_norm": 97.54595947265625, "learning_rate": 5.5982730263157904e-06, "loss": 1.8945, "step": 2141 }, { "epoch": 0.44037828947368424, "grad_norm": 113.62591552734375, "learning_rate": 5.596217105263158e-06, "loss": 1.3169, "step": 2142 }, { "epoch": 0.44058388157894735, "grad_norm": 52.98300552368164, "learning_rate": 5.594161184210527e-06, "loss": 0.5127, "step": 2143 }, { "epoch": 0.4407894736842105, "grad_norm": 113.43074798583984, "learning_rate": 5.592105263157896e-06, "loss": 0.8179, "step": 2144 }, { "epoch": 0.4409950657894737, "grad_norm": 28.55093765258789, "learning_rate": 5.590049342105264e-06, "loss": 0.1543, "step": 2145 }, { "epoch": 0.44120065789473684, "grad_norm": 32.1033821105957, "learning_rate": 5.587993421052632e-06, "loss": 0.4609, "step": 2146 }, { "epoch": 0.44140625, "grad_norm": 27.91405487060547, "learning_rate": 5.5859375e-06, "loss": 0.1367, "step": 2147 }, { "epoch": 0.44161184210526316, "grad_norm": 38.06175994873047, "learning_rate": 5.583881578947369e-06, "loss": 0.0857, "step": 2148 }, { "epoch": 0.4418174342105263, "grad_norm": 7.461126804351807, "learning_rate": 5.581825657894737e-06, "loss": 0.0492, "step": 2149 }, { "epoch": 0.4420230263157895, "grad_norm": 34.90956497192383, "learning_rate": 5.579769736842106e-06, "loss": 0.4473, "step": 2150 }, { "epoch": 0.44222861842105265, "grad_norm": 88.27356719970703, "learning_rate": 5.577713815789474e-06, "loss": 1.04, "step": 2151 }, { "epoch": 0.44243421052631576, "grad_norm": 13.723530769348145, "learning_rate": 5.575657894736842e-06, "loss": 0.0876, "step": 2152 }, { "epoch": 0.4426398026315789, "grad_norm": 89.1964340209961, "learning_rate": 5.573601973684211e-06, "loss": 1.1704, "step": 2153 }, { "epoch": 0.4428453947368421, "grad_norm": 128.08401489257812, "learning_rate": 5.571546052631579e-06, "loss": 1.6406, "step": 2154 }, { "epoch": 0.44305098684210525, "grad_norm": 48.338600158691406, "learning_rate": 5.5694901315789475e-06, "loss": 0.2745, "step": 2155 }, { "epoch": 0.4432565789473684, "grad_norm": 5.757163047790527, "learning_rate": 5.567434210526315e-06, "loss": 0.0142, "step": 2156 }, { "epoch": 0.4434621710526316, "grad_norm": 46.891929626464844, "learning_rate": 5.565378289473685e-06, "loss": 0.2593, "step": 2157 }, { "epoch": 0.44366776315789475, "grad_norm": 35.7086181640625, "learning_rate": 5.563322368421054e-06, "loss": 0.3052, "step": 2158 }, { "epoch": 0.4438733552631579, "grad_norm": 14.258553504943848, "learning_rate": 5.561266447368422e-06, "loss": 0.0471, "step": 2159 }, { "epoch": 0.4440789473684211, "grad_norm": 87.7942886352539, "learning_rate": 5.55921052631579e-06, "loss": 1.3801, "step": 2160 }, { "epoch": 0.44428453947368424, "grad_norm": 100.42745971679688, "learning_rate": 5.557154605263159e-06, "loss": 0.9141, "step": 2161 }, { "epoch": 0.44449013157894735, "grad_norm": 109.86695861816406, "learning_rate": 5.555098684210527e-06, "loss": 1.1738, "step": 2162 }, { "epoch": 0.4446957236842105, "grad_norm": 102.30524444580078, "learning_rate": 5.553042763157896e-06, "loss": 1.3633, "step": 2163 }, { "epoch": 0.4449013157894737, "grad_norm": 10.261860847473145, "learning_rate": 5.5509868421052635e-06, "loss": 0.0372, "step": 2164 }, { "epoch": 0.44510690789473684, "grad_norm": 48.87498092651367, "learning_rate": 5.548930921052632e-06, "loss": 0.3608, "step": 2165 }, { "epoch": 0.4453125, "grad_norm": 32.25232696533203, "learning_rate": 5.546875e-06, "loss": 0.1599, "step": 2166 }, { "epoch": 0.44551809210526316, "grad_norm": 39.74791717529297, "learning_rate": 5.544819078947369e-06, "loss": 0.2173, "step": 2167 }, { "epoch": 0.4457236842105263, "grad_norm": 56.46585464477539, "learning_rate": 5.542763157894737e-06, "loss": 0.623, "step": 2168 }, { "epoch": 0.4459292763157895, "grad_norm": 224.99737548828125, "learning_rate": 5.5407072368421055e-06, "loss": 1.4102, "step": 2169 }, { "epoch": 0.44613486842105265, "grad_norm": 36.65513610839844, "learning_rate": 5.538651315789474e-06, "loss": 0.3379, "step": 2170 }, { "epoch": 0.44634046052631576, "grad_norm": 160.5054168701172, "learning_rate": 5.536595394736842e-06, "loss": 1.7891, "step": 2171 }, { "epoch": 0.4465460526315789, "grad_norm": 59.382041931152344, "learning_rate": 5.534539473684211e-06, "loss": 0.7051, "step": 2172 }, { "epoch": 0.4467516447368421, "grad_norm": 74.1430435180664, "learning_rate": 5.532483552631579e-06, "loss": 0.6282, "step": 2173 }, { "epoch": 0.44695723684210525, "grad_norm": 78.62765502929688, "learning_rate": 5.5304276315789474e-06, "loss": 0.7441, "step": 2174 }, { "epoch": 0.4471628289473684, "grad_norm": 94.7033920288086, "learning_rate": 5.528371710526315e-06, "loss": 0.9468, "step": 2175 }, { "epoch": 0.4473684210526316, "grad_norm": 63.69320297241211, "learning_rate": 5.526315789473685e-06, "loss": 0.5469, "step": 2176 }, { "epoch": 0.44757401315789475, "grad_norm": 116.96182250976562, "learning_rate": 5.524259868421054e-06, "loss": 1.9805, "step": 2177 }, { "epoch": 0.4477796052631579, "grad_norm": 93.27742004394531, "learning_rate": 5.5222039473684215e-06, "loss": 0.7065, "step": 2178 }, { "epoch": 0.4479851973684211, "grad_norm": 151.447998046875, "learning_rate": 5.52014802631579e-06, "loss": 1.123, "step": 2179 }, { "epoch": 0.44819078947368424, "grad_norm": 26.769309997558594, "learning_rate": 5.518092105263159e-06, "loss": 0.0818, "step": 2180 }, { "epoch": 0.44839638157894735, "grad_norm": 156.28541564941406, "learning_rate": 5.516036184210527e-06, "loss": 4.0732, "step": 2181 }, { "epoch": 0.4486019736842105, "grad_norm": 22.551742553710938, "learning_rate": 5.5139802631578956e-06, "loss": 0.1956, "step": 2182 }, { "epoch": 0.4488075657894737, "grad_norm": 76.9942398071289, "learning_rate": 5.5119243421052634e-06, "loss": 1.2871, "step": 2183 }, { "epoch": 0.44901315789473684, "grad_norm": 12.154285430908203, "learning_rate": 5.509868421052632e-06, "loss": 0.0525, "step": 2184 }, { "epoch": 0.44921875, "grad_norm": 65.95537567138672, "learning_rate": 5.5078125e-06, "loss": 0.8918, "step": 2185 }, { "epoch": 0.44942434210526316, "grad_norm": 124.6831283569336, "learning_rate": 5.505756578947369e-06, "loss": 1.748, "step": 2186 }, { "epoch": 0.4496299342105263, "grad_norm": 120.58658599853516, "learning_rate": 5.5037006578947375e-06, "loss": 1.1818, "step": 2187 }, { "epoch": 0.4498355263157895, "grad_norm": 22.78190040588379, "learning_rate": 5.501644736842105e-06, "loss": 0.1604, "step": 2188 }, { "epoch": 0.45004111842105265, "grad_norm": 45.05170822143555, "learning_rate": 5.499588815789474e-06, "loss": 0.2812, "step": 2189 }, { "epoch": 0.45024671052631576, "grad_norm": 60.48125457763672, "learning_rate": 5.497532894736842e-06, "loss": 0.4301, "step": 2190 }, { "epoch": 0.4504523026315789, "grad_norm": 7.772931098937988, "learning_rate": 5.495476973684211e-06, "loss": 0.026, "step": 2191 }, { "epoch": 0.4506578947368421, "grad_norm": 71.32646179199219, "learning_rate": 5.493421052631579e-06, "loss": 0.8782, "step": 2192 }, { "epoch": 0.45086348684210525, "grad_norm": 39.65742111206055, "learning_rate": 5.491365131578947e-06, "loss": 0.4307, "step": 2193 }, { "epoch": 0.4510690789473684, "grad_norm": 435.2488098144531, "learning_rate": 5.489309210526316e-06, "loss": 2.1289, "step": 2194 }, { "epoch": 0.4512746710526316, "grad_norm": 37.907169342041016, "learning_rate": 5.487253289473685e-06, "loss": 0.2407, "step": 2195 }, { "epoch": 0.45148026315789475, "grad_norm": 71.29874420166016, "learning_rate": 5.4851973684210535e-06, "loss": 0.8477, "step": 2196 }, { "epoch": 0.4516858552631579, "grad_norm": 37.618919372558594, "learning_rate": 5.483141447368422e-06, "loss": 0.2866, "step": 2197 }, { "epoch": 0.4518914473684211, "grad_norm": 95.83308410644531, "learning_rate": 5.48108552631579e-06, "loss": 1.332, "step": 2198 }, { "epoch": 0.45209703947368424, "grad_norm": 92.70743560791016, "learning_rate": 5.479029605263159e-06, "loss": 0.7526, "step": 2199 }, { "epoch": 0.45230263157894735, "grad_norm": 81.58545684814453, "learning_rate": 5.476973684210527e-06, "loss": 1.043, "step": 2200 }, { "epoch": 0.4525082236842105, "grad_norm": 203.60203552246094, "learning_rate": 5.4749177631578955e-06, "loss": 1.4453, "step": 2201 }, { "epoch": 0.4527138157894737, "grad_norm": 93.03060150146484, "learning_rate": 5.472861842105263e-06, "loss": 0.6484, "step": 2202 }, { "epoch": 0.45291940789473684, "grad_norm": 47.20277786254883, "learning_rate": 5.470805921052632e-06, "loss": 0.5131, "step": 2203 }, { "epoch": 0.453125, "grad_norm": 20.870620727539062, "learning_rate": 5.468750000000001e-06, "loss": 0.1063, "step": 2204 }, { "epoch": 0.45333059210526316, "grad_norm": 19.70764923095703, "learning_rate": 5.466694078947369e-06, "loss": 0.083, "step": 2205 }, { "epoch": 0.4535361842105263, "grad_norm": 97.82512664794922, "learning_rate": 5.464638157894737e-06, "loss": 0.7344, "step": 2206 }, { "epoch": 0.4537417763157895, "grad_norm": 1.880107045173645, "learning_rate": 5.462582236842105e-06, "loss": 0.0088, "step": 2207 }, { "epoch": 0.45394736842105265, "grad_norm": 229.75209045410156, "learning_rate": 5.460526315789474e-06, "loss": 3.8633, "step": 2208 }, { "epoch": 0.45415296052631576, "grad_norm": 60.317665100097656, "learning_rate": 5.458470394736842e-06, "loss": 0.7461, "step": 2209 }, { "epoch": 0.4543585526315789, "grad_norm": 159.6494140625, "learning_rate": 5.456414473684211e-06, "loss": 2.4248, "step": 2210 }, { "epoch": 0.4545641447368421, "grad_norm": 61.16450500488281, "learning_rate": 5.454358552631579e-06, "loss": 0.6055, "step": 2211 }, { "epoch": 0.45476973684210525, "grad_norm": 87.54611206054688, "learning_rate": 5.452302631578947e-06, "loss": 0.7026, "step": 2212 }, { "epoch": 0.4549753289473684, "grad_norm": 45.78079605102539, "learning_rate": 5.450246710526316e-06, "loss": 0.3223, "step": 2213 }, { "epoch": 0.4551809210526316, "grad_norm": 83.92278289794922, "learning_rate": 5.4481907894736855e-06, "loss": 0.5469, "step": 2214 }, { "epoch": 0.45538651315789475, "grad_norm": 67.95943450927734, "learning_rate": 5.446134868421053e-06, "loss": 1.4521, "step": 2215 }, { "epoch": 0.4555921052631579, "grad_norm": 176.879638671875, "learning_rate": 5.444078947368422e-06, "loss": 3.6128, "step": 2216 }, { "epoch": 0.4557976973684211, "grad_norm": 67.3059310913086, "learning_rate": 5.44202302631579e-06, "loss": 0.5156, "step": 2217 }, { "epoch": 0.45600328947368424, "grad_norm": 140.7138671875, "learning_rate": 5.439967105263159e-06, "loss": 2.1211, "step": 2218 }, { "epoch": 0.45620888157894735, "grad_norm": 33.47517395019531, "learning_rate": 5.437911184210527e-06, "loss": 0.2344, "step": 2219 }, { "epoch": 0.4564144736842105, "grad_norm": 20.869125366210938, "learning_rate": 5.435855263157895e-06, "loss": 0.0994, "step": 2220 }, { "epoch": 0.4566200657894737, "grad_norm": 82.64616394042969, "learning_rate": 5.433799342105264e-06, "loss": 1.29, "step": 2221 }, { "epoch": 0.45682565789473684, "grad_norm": 73.86136627197266, "learning_rate": 5.431743421052632e-06, "loss": 0.7854, "step": 2222 }, { "epoch": 0.45703125, "grad_norm": 79.1838150024414, "learning_rate": 5.429687500000001e-06, "loss": 0.9116, "step": 2223 }, { "epoch": 0.45723684210526316, "grad_norm": 138.31045532226562, "learning_rate": 5.4276315789473686e-06, "loss": 0.9907, "step": 2224 }, { "epoch": 0.4574424342105263, "grad_norm": 122.5740737915039, "learning_rate": 5.425575657894737e-06, "loss": 0.5977, "step": 2225 }, { "epoch": 0.4576480263157895, "grad_norm": 63.34665298461914, "learning_rate": 5.423519736842105e-06, "loss": 0.8511, "step": 2226 }, { "epoch": 0.45785361842105265, "grad_norm": 19.32025146484375, "learning_rate": 5.421463815789474e-06, "loss": 0.098, "step": 2227 }, { "epoch": 0.45805921052631576, "grad_norm": 22.811641693115234, "learning_rate": 5.419407894736843e-06, "loss": 0.0852, "step": 2228 }, { "epoch": 0.4582648026315789, "grad_norm": 86.62748718261719, "learning_rate": 5.4173519736842105e-06, "loss": 1.1201, "step": 2229 }, { "epoch": 0.4584703947368421, "grad_norm": 10.291800498962402, "learning_rate": 5.415296052631579e-06, "loss": 0.0191, "step": 2230 }, { "epoch": 0.45867598684210525, "grad_norm": 131.99404907226562, "learning_rate": 5.413240131578947e-06, "loss": 0.5022, "step": 2231 }, { "epoch": 0.4588815789473684, "grad_norm": 129.52847290039062, "learning_rate": 5.411184210526316e-06, "loss": 1.2949, "step": 2232 }, { "epoch": 0.4590871710526316, "grad_norm": 19.865447998046875, "learning_rate": 5.4091282894736854e-06, "loss": 0.1553, "step": 2233 }, { "epoch": 0.45929276315789475, "grad_norm": 52.17932891845703, "learning_rate": 5.407072368421053e-06, "loss": 0.3167, "step": 2234 }, { "epoch": 0.4594983552631579, "grad_norm": 25.268348693847656, "learning_rate": 5.405016447368422e-06, "loss": 0.0857, "step": 2235 }, { "epoch": 0.4597039473684211, "grad_norm": 30.30575942993164, "learning_rate": 5.40296052631579e-06, "loss": 0.1968, "step": 2236 }, { "epoch": 0.45990953947368424, "grad_norm": 42.582115173339844, "learning_rate": 5.400904605263159e-06, "loss": 0.3456, "step": 2237 }, { "epoch": 0.46011513157894735, "grad_norm": 21.693111419677734, "learning_rate": 5.398848684210527e-06, "loss": 0.2393, "step": 2238 }, { "epoch": 0.4603207236842105, "grad_norm": 34.41257858276367, "learning_rate": 5.396792763157895e-06, "loss": 0.2153, "step": 2239 }, { "epoch": 0.4605263157894737, "grad_norm": 241.35472106933594, "learning_rate": 5.394736842105264e-06, "loss": 2.134, "step": 2240 }, { "epoch": 0.46073190789473684, "grad_norm": 39.204898834228516, "learning_rate": 5.392680921052632e-06, "loss": 0.4008, "step": 2241 }, { "epoch": 0.4609375, "grad_norm": 27.47697639465332, "learning_rate": 5.390625000000001e-06, "loss": 0.1587, "step": 2242 }, { "epoch": 0.46114309210526316, "grad_norm": 28.24747657775879, "learning_rate": 5.3885690789473685e-06, "loss": 0.1287, "step": 2243 }, { "epoch": 0.4613486842105263, "grad_norm": 139.3961944580078, "learning_rate": 5.386513157894737e-06, "loss": 0.5884, "step": 2244 }, { "epoch": 0.4615542763157895, "grad_norm": 31.27547836303711, "learning_rate": 5.384457236842106e-06, "loss": 0.2919, "step": 2245 }, { "epoch": 0.46175986842105265, "grad_norm": 84.11787414550781, "learning_rate": 5.382401315789474e-06, "loss": 0.9023, "step": 2246 }, { "epoch": 0.46196546052631576, "grad_norm": 93.190185546875, "learning_rate": 5.3803453947368425e-06, "loss": 0.8428, "step": 2247 }, { "epoch": 0.4621710526315789, "grad_norm": 18.921268463134766, "learning_rate": 5.37828947368421e-06, "loss": 0.0782, "step": 2248 }, { "epoch": 0.4623766447368421, "grad_norm": 72.8770980834961, "learning_rate": 5.376233552631579e-06, "loss": 1.1406, "step": 2249 }, { "epoch": 0.46258223684210525, "grad_norm": 58.10158157348633, "learning_rate": 5.374177631578947e-06, "loss": 0.543, "step": 2250 }, { "epoch": 0.4627878289473684, "grad_norm": 66.5830078125, "learning_rate": 5.372121710526316e-06, "loss": 0.7156, "step": 2251 }, { "epoch": 0.4629934210526316, "grad_norm": 83.36282348632812, "learning_rate": 5.370065789473685e-06, "loss": 1.1055, "step": 2252 }, { "epoch": 0.46319901315789475, "grad_norm": 63.09612274169922, "learning_rate": 5.368009868421053e-06, "loss": 0.3208, "step": 2253 }, { "epoch": 0.4634046052631579, "grad_norm": 51.19950485229492, "learning_rate": 5.365953947368422e-06, "loss": 0.3571, "step": 2254 }, { "epoch": 0.4636101973684211, "grad_norm": 29.22952651977539, "learning_rate": 5.363898026315791e-06, "loss": 0.2354, "step": 2255 }, { "epoch": 0.46381578947368424, "grad_norm": 107.38790130615234, "learning_rate": 5.3618421052631585e-06, "loss": 0.7827, "step": 2256 }, { "epoch": 0.46402138157894735, "grad_norm": 60.90756607055664, "learning_rate": 5.359786184210527e-06, "loss": 0.4515, "step": 2257 }, { "epoch": 0.4642269736842105, "grad_norm": 137.5308837890625, "learning_rate": 5.357730263157895e-06, "loss": 1.1719, "step": 2258 }, { "epoch": 0.4644325657894737, "grad_norm": 62.70950698852539, "learning_rate": 5.355674342105264e-06, "loss": 0.7051, "step": 2259 }, { "epoch": 0.46463815789473684, "grad_norm": 169.47792053222656, "learning_rate": 5.353618421052632e-06, "loss": 1.7305, "step": 2260 }, { "epoch": 0.46484375, "grad_norm": 185.45407104492188, "learning_rate": 5.3515625000000005e-06, "loss": 2.8828, "step": 2261 }, { "epoch": 0.46504934210526316, "grad_norm": 16.225032806396484, "learning_rate": 5.349506578947369e-06, "loss": 0.0598, "step": 2262 }, { "epoch": 0.4652549342105263, "grad_norm": 31.832530975341797, "learning_rate": 5.347450657894737e-06, "loss": 0.1453, "step": 2263 }, { "epoch": 0.4654605263157895, "grad_norm": 15.255589485168457, "learning_rate": 5.345394736842106e-06, "loss": 0.0806, "step": 2264 }, { "epoch": 0.46566611842105265, "grad_norm": 22.511999130249023, "learning_rate": 5.343338815789474e-06, "loss": 0.1416, "step": 2265 }, { "epoch": 0.46587171052631576, "grad_norm": 32.66456985473633, "learning_rate": 5.3412828947368424e-06, "loss": 0.2307, "step": 2266 }, { "epoch": 0.4660773026315789, "grad_norm": 36.65477752685547, "learning_rate": 5.33922697368421e-06, "loss": 0.209, "step": 2267 }, { "epoch": 0.4662828947368421, "grad_norm": 69.11346435546875, "learning_rate": 5.337171052631579e-06, "loss": 0.4807, "step": 2268 }, { "epoch": 0.46648848684210525, "grad_norm": 119.99474334716797, "learning_rate": 5.335115131578947e-06, "loss": 1.0938, "step": 2269 }, { "epoch": 0.4666940789473684, "grad_norm": 72.48067474365234, "learning_rate": 5.333059210526316e-06, "loss": 0.34, "step": 2270 }, { "epoch": 0.4668996710526316, "grad_norm": 71.48184967041016, "learning_rate": 5.331003289473685e-06, "loss": 0.4158, "step": 2271 }, { "epoch": 0.46710526315789475, "grad_norm": 106.27686309814453, "learning_rate": 5.328947368421054e-06, "loss": 1.0554, "step": 2272 }, { "epoch": 0.4673108552631579, "grad_norm": 87.34750366210938, "learning_rate": 5.326891447368422e-06, "loss": 0.4487, "step": 2273 }, { "epoch": 0.4675164473684211, "grad_norm": 99.95732116699219, "learning_rate": 5.3248355263157906e-06, "loss": 1.4258, "step": 2274 }, { "epoch": 0.46772203947368424, "grad_norm": 56.87318801879883, "learning_rate": 5.3227796052631584e-06, "loss": 0.8048, "step": 2275 }, { "epoch": 0.46792763157894735, "grad_norm": 66.48043823242188, "learning_rate": 5.320723684210527e-06, "loss": 0.458, "step": 2276 }, { "epoch": 0.4681332236842105, "grad_norm": 13.604301452636719, "learning_rate": 5.318667763157895e-06, "loss": 0.0381, "step": 2277 }, { "epoch": 0.4683388157894737, "grad_norm": 122.62925720214844, "learning_rate": 5.316611842105264e-06, "loss": 0.8413, "step": 2278 }, { "epoch": 0.46854440789473684, "grad_norm": 23.963611602783203, "learning_rate": 5.314555921052632e-06, "loss": 0.2256, "step": 2279 }, { "epoch": 0.46875, "grad_norm": 21.11497688293457, "learning_rate": 5.3125e-06, "loss": 0.145, "step": 2280 }, { "epoch": 0.46895559210526316, "grad_norm": 13.243685722351074, "learning_rate": 5.310444078947369e-06, "loss": 0.0714, "step": 2281 }, { "epoch": 0.4691611842105263, "grad_norm": 38.76900863647461, "learning_rate": 5.308388157894737e-06, "loss": 0.3559, "step": 2282 }, { "epoch": 0.4693667763157895, "grad_norm": 69.9073257446289, "learning_rate": 5.306332236842106e-06, "loss": 0.377, "step": 2283 }, { "epoch": 0.46957236842105265, "grad_norm": 141.59156799316406, "learning_rate": 5.304276315789474e-06, "loss": 1.3906, "step": 2284 }, { "epoch": 0.46977796052631576, "grad_norm": 14.897806167602539, "learning_rate": 5.302220394736842e-06, "loss": 0.0471, "step": 2285 }, { "epoch": 0.4699835526315789, "grad_norm": 137.9097137451172, "learning_rate": 5.30016447368421e-06, "loss": 0.729, "step": 2286 }, { "epoch": 0.4701891447368421, "grad_norm": 43.31760787963867, "learning_rate": 5.298108552631579e-06, "loss": 0.2778, "step": 2287 }, { "epoch": 0.47039473684210525, "grad_norm": 22.03982925415039, "learning_rate": 5.296052631578948e-06, "loss": 0.1541, "step": 2288 }, { "epoch": 0.4706003289473684, "grad_norm": 76.19063568115234, "learning_rate": 5.2939967105263155e-06, "loss": 0.8672, "step": 2289 }, { "epoch": 0.4708059210526316, "grad_norm": 5.71510124206543, "learning_rate": 5.291940789473685e-06, "loss": 0.0186, "step": 2290 }, { "epoch": 0.47101151315789475, "grad_norm": 121.10185241699219, "learning_rate": 5.289884868421054e-06, "loss": 1.1992, "step": 2291 }, { "epoch": 0.4712171052631579, "grad_norm": 41.63251876831055, "learning_rate": 5.287828947368422e-06, "loss": 0.3047, "step": 2292 }, { "epoch": 0.4714226973684211, "grad_norm": 102.78498840332031, "learning_rate": 5.2857730263157905e-06, "loss": 0.8223, "step": 2293 }, { "epoch": 0.47162828947368424, "grad_norm": 95.88848876953125, "learning_rate": 5.283717105263158e-06, "loss": 2.3281, "step": 2294 }, { "epoch": 0.47183388157894735, "grad_norm": 50.10077667236328, "learning_rate": 5.281661184210527e-06, "loss": 0.7148, "step": 2295 }, { "epoch": 0.4720394736842105, "grad_norm": 15.069315910339355, "learning_rate": 5.279605263157895e-06, "loss": 0.0681, "step": 2296 }, { "epoch": 0.4722450657894737, "grad_norm": 96.7359390258789, "learning_rate": 5.277549342105264e-06, "loss": 1.1682, "step": 2297 }, { "epoch": 0.47245065789473684, "grad_norm": 19.88226318359375, "learning_rate": 5.275493421052632e-06, "loss": 0.0565, "step": 2298 }, { "epoch": 0.47265625, "grad_norm": 42.63949966430664, "learning_rate": 5.2734375e-06, "loss": 0.3271, "step": 2299 }, { "epoch": 0.47286184210526316, "grad_norm": 215.39588928222656, "learning_rate": 5.271381578947369e-06, "loss": 2.582, "step": 2300 }, { "epoch": 0.4730674342105263, "grad_norm": 13.266740798950195, "learning_rate": 5.269325657894737e-06, "loss": 0.0513, "step": 2301 }, { "epoch": 0.4732730263157895, "grad_norm": 89.06000518798828, "learning_rate": 5.267269736842106e-06, "loss": 0.6824, "step": 2302 }, { "epoch": 0.47347861842105265, "grad_norm": 241.9901123046875, "learning_rate": 5.2652138157894735e-06, "loss": 1.8633, "step": 2303 }, { "epoch": 0.47368421052631576, "grad_norm": 16.954750061035156, "learning_rate": 5.263157894736842e-06, "loss": 0.1123, "step": 2304 }, { "epoch": 0.4738898026315789, "grad_norm": 12.952689170837402, "learning_rate": 5.261101973684211e-06, "loss": 0.0613, "step": 2305 }, { "epoch": 0.4740953947368421, "grad_norm": 84.71417236328125, "learning_rate": 5.259046052631579e-06, "loss": 1.5547, "step": 2306 }, { "epoch": 0.47430098684210525, "grad_norm": 39.38886642456055, "learning_rate": 5.2569901315789476e-06, "loss": 0.4922, "step": 2307 }, { "epoch": 0.4745065789473684, "grad_norm": 80.70586395263672, "learning_rate": 5.2549342105263154e-06, "loss": 1.0566, "step": 2308 }, { "epoch": 0.4747121710526316, "grad_norm": 36.7666130065918, "learning_rate": 5.252878289473685e-06, "loss": 0.2173, "step": 2309 }, { "epoch": 0.47491776315789475, "grad_norm": 33.24839401245117, "learning_rate": 5.250822368421054e-06, "loss": 0.1782, "step": 2310 }, { "epoch": 0.4751233552631579, "grad_norm": 111.84273529052734, "learning_rate": 5.248766447368422e-06, "loss": 0.8055, "step": 2311 }, { "epoch": 0.4753289473684211, "grad_norm": 25.041423797607422, "learning_rate": 5.24671052631579e-06, "loss": 0.1373, "step": 2312 }, { "epoch": 0.47553453947368424, "grad_norm": 116.9102783203125, "learning_rate": 5.244654605263158e-06, "loss": 1.6934, "step": 2313 }, { "epoch": 0.47574013157894735, "grad_norm": 16.74965476989746, "learning_rate": 5.242598684210527e-06, "loss": 0.0776, "step": 2314 }, { "epoch": 0.4759457236842105, "grad_norm": 53.60133743286133, "learning_rate": 5.240542763157896e-06, "loss": 0.3252, "step": 2315 }, { "epoch": 0.4761513157894737, "grad_norm": 28.9056339263916, "learning_rate": 5.2384868421052636e-06, "loss": 0.2166, "step": 2316 }, { "epoch": 0.47635690789473684, "grad_norm": 26.974119186401367, "learning_rate": 5.236430921052632e-06, "loss": 0.2114, "step": 2317 }, { "epoch": 0.4765625, "grad_norm": 8.609535217285156, "learning_rate": 5.234375e-06, "loss": 0.0311, "step": 2318 }, { "epoch": 0.47676809210526316, "grad_norm": 104.93476104736328, "learning_rate": 5.232319078947369e-06, "loss": 0.8059, "step": 2319 }, { "epoch": 0.4769736842105263, "grad_norm": 61.51924514770508, "learning_rate": 5.230263157894737e-06, "loss": 0.5674, "step": 2320 }, { "epoch": 0.4771792763157895, "grad_norm": 26.482332229614258, "learning_rate": 5.2282072368421055e-06, "loss": 0.1268, "step": 2321 }, { "epoch": 0.47738486842105265, "grad_norm": 38.34437561035156, "learning_rate": 5.226151315789474e-06, "loss": 0.292, "step": 2322 }, { "epoch": 0.47759046052631576, "grad_norm": 45.16140365600586, "learning_rate": 5.224095394736842e-06, "loss": 0.2729, "step": 2323 }, { "epoch": 0.4777960526315789, "grad_norm": 193.146240234375, "learning_rate": 5.222039473684211e-06, "loss": 1.9785, "step": 2324 }, { "epoch": 0.4780016447368421, "grad_norm": 150.05584716796875, "learning_rate": 5.219983552631579e-06, "loss": 0.8359, "step": 2325 }, { "epoch": 0.47820723684210525, "grad_norm": 59.697750091552734, "learning_rate": 5.2179276315789475e-06, "loss": 0.4547, "step": 2326 }, { "epoch": 0.4784128289473684, "grad_norm": 92.0274658203125, "learning_rate": 5.215871710526315e-06, "loss": 0.8662, "step": 2327 }, { "epoch": 0.4786184210526316, "grad_norm": 8.391749382019043, "learning_rate": 5.213815789473685e-06, "loss": 0.0155, "step": 2328 }, { "epoch": 0.47882401315789475, "grad_norm": 231.23291015625, "learning_rate": 5.211759868421054e-06, "loss": 3.45, "step": 2329 }, { "epoch": 0.4790296052631579, "grad_norm": 363.240478515625, "learning_rate": 5.2097039473684215e-06, "loss": 3.5408, "step": 2330 }, { "epoch": 0.4792351973684211, "grad_norm": 7.54001522064209, "learning_rate": 5.20764802631579e-06, "loss": 0.0275, "step": 2331 }, { "epoch": 0.47944078947368424, "grad_norm": 6.956185817718506, "learning_rate": 5.205592105263159e-06, "loss": 0.0183, "step": 2332 }, { "epoch": 0.47964638157894735, "grad_norm": 226.07701110839844, "learning_rate": 5.203536184210527e-06, "loss": 1.6484, "step": 2333 }, { "epoch": 0.4798519736842105, "grad_norm": 45.584224700927734, "learning_rate": 5.201480263157896e-06, "loss": 0.2781, "step": 2334 }, { "epoch": 0.4800575657894737, "grad_norm": 15.4505615234375, "learning_rate": 5.1994243421052635e-06, "loss": 0.0911, "step": 2335 }, { "epoch": 0.48026315789473684, "grad_norm": 62.67617416381836, "learning_rate": 5.197368421052632e-06, "loss": 0.4326, "step": 2336 }, { "epoch": 0.48046875, "grad_norm": 47.94852066040039, "learning_rate": 5.1953125e-06, "loss": 0.6045, "step": 2337 }, { "epoch": 0.48067434210526316, "grad_norm": 83.68470764160156, "learning_rate": 5.193256578947369e-06, "loss": 0.6196, "step": 2338 }, { "epoch": 0.4808799342105263, "grad_norm": 66.00609588623047, "learning_rate": 5.1912006578947375e-06, "loss": 1.371, "step": 2339 }, { "epoch": 0.4810855263157895, "grad_norm": 10.02939224243164, "learning_rate": 5.189144736842105e-06, "loss": 0.0183, "step": 2340 }, { "epoch": 0.48129111842105265, "grad_norm": 15.998161315917969, "learning_rate": 5.187088815789474e-06, "loss": 0.0746, "step": 2341 }, { "epoch": 0.48149671052631576, "grad_norm": 81.58523559570312, "learning_rate": 5.185032894736842e-06, "loss": 0.7832, "step": 2342 }, { "epoch": 0.4817023026315789, "grad_norm": 99.93974304199219, "learning_rate": 5.182976973684211e-06, "loss": 1.4531, "step": 2343 }, { "epoch": 0.4819078947368421, "grad_norm": 55.56233215332031, "learning_rate": 5.180921052631579e-06, "loss": 0.5557, "step": 2344 }, { "epoch": 0.48211348684210525, "grad_norm": 145.71836853027344, "learning_rate": 5.178865131578947e-06, "loss": 0.9912, "step": 2345 }, { "epoch": 0.4823190789473684, "grad_norm": 34.319034576416016, "learning_rate": 5.176809210526316e-06, "loss": 0.1837, "step": 2346 }, { "epoch": 0.4825246710526316, "grad_norm": 109.79023742675781, "learning_rate": 5.174753289473685e-06, "loss": 1.104, "step": 2347 }, { "epoch": 0.48273026315789475, "grad_norm": 138.68930053710938, "learning_rate": 5.1726973684210535e-06, "loss": 2.017, "step": 2348 }, { "epoch": 0.4829358552631579, "grad_norm": 85.45179748535156, "learning_rate": 5.170641447368422e-06, "loss": 0.582, "step": 2349 }, { "epoch": 0.4831414473684211, "grad_norm": 9.717490196228027, "learning_rate": 5.16858552631579e-06, "loss": 0.0504, "step": 2350 }, { "epoch": 0.48334703947368424, "grad_norm": 49.215152740478516, "learning_rate": 5.166529605263159e-06, "loss": 0.3711, "step": 2351 }, { "epoch": 0.48355263157894735, "grad_norm": 152.89309692382812, "learning_rate": 5.164473684210527e-06, "loss": 1.0684, "step": 2352 }, { "epoch": 0.4837582236842105, "grad_norm": 75.16875457763672, "learning_rate": 5.1624177631578955e-06, "loss": 0.7837, "step": 2353 }, { "epoch": 0.4839638157894737, "grad_norm": 82.52830505371094, "learning_rate": 5.160361842105263e-06, "loss": 0.5527, "step": 2354 }, { "epoch": 0.48416940789473684, "grad_norm": 123.05012512207031, "learning_rate": 5.158305921052632e-06, "loss": 1.2747, "step": 2355 }, { "epoch": 0.484375, "grad_norm": 63.32767868041992, "learning_rate": 5.156250000000001e-06, "loss": 0.1128, "step": 2356 }, { "epoch": 0.48458059210526316, "grad_norm": 53.867210388183594, "learning_rate": 5.154194078947369e-06, "loss": 0.9141, "step": 2357 }, { "epoch": 0.4847861842105263, "grad_norm": 33.818336486816406, "learning_rate": 5.1521381578947374e-06, "loss": 0.1421, "step": 2358 }, { "epoch": 0.4849917763157895, "grad_norm": 35.07719421386719, "learning_rate": 5.150082236842105e-06, "loss": 0.227, "step": 2359 }, { "epoch": 0.48519736842105265, "grad_norm": 42.34528350830078, "learning_rate": 5.148026315789474e-06, "loss": 0.2546, "step": 2360 }, { "epoch": 0.48540296052631576, "grad_norm": 198.51248168945312, "learning_rate": 5.145970394736842e-06, "loss": 4.0273, "step": 2361 }, { "epoch": 0.4856085526315789, "grad_norm": 31.751073837280273, "learning_rate": 5.143914473684211e-06, "loss": 0.3184, "step": 2362 }, { "epoch": 0.4858141447368421, "grad_norm": 225.57908630371094, "learning_rate": 5.141858552631579e-06, "loss": 0.8488, "step": 2363 }, { "epoch": 0.48601973684210525, "grad_norm": 35.79840850830078, "learning_rate": 5.139802631578947e-06, "loss": 0.3049, "step": 2364 }, { "epoch": 0.4862253289473684, "grad_norm": 67.66960144042969, "learning_rate": 5.137746710526316e-06, "loss": 0.4663, "step": 2365 }, { "epoch": 0.4864309210526316, "grad_norm": 148.8115997314453, "learning_rate": 5.1356907894736856e-06, "loss": 1.0156, "step": 2366 }, { "epoch": 0.48663651315789475, "grad_norm": 20.171966552734375, "learning_rate": 5.1336348684210534e-06, "loss": 0.1182, "step": 2367 }, { "epoch": 0.4868421052631579, "grad_norm": 112.57107543945312, "learning_rate": 5.131578947368422e-06, "loss": 1.1729, "step": 2368 }, { "epoch": 0.4870476973684211, "grad_norm": 20.01609992980957, "learning_rate": 5.12952302631579e-06, "loss": 0.0859, "step": 2369 }, { "epoch": 0.48725328947368424, "grad_norm": 65.70594787597656, "learning_rate": 5.127467105263159e-06, "loss": 0.7056, "step": 2370 }, { "epoch": 0.48745888157894735, "grad_norm": 20.96782112121582, "learning_rate": 5.125411184210527e-06, "loss": 0.0894, "step": 2371 }, { "epoch": 0.4876644736842105, "grad_norm": 33.21198654174805, "learning_rate": 5.123355263157895e-06, "loss": 0.1615, "step": 2372 }, { "epoch": 0.4878700657894737, "grad_norm": 68.09440612792969, "learning_rate": 5.121299342105264e-06, "loss": 0.5986, "step": 2373 }, { "epoch": 0.48807565789473684, "grad_norm": 55.00299072265625, "learning_rate": 5.119243421052632e-06, "loss": 0.6497, "step": 2374 }, { "epoch": 0.48828125, "grad_norm": 12.726222038269043, "learning_rate": 5.117187500000001e-06, "loss": 0.078, "step": 2375 }, { "epoch": 0.48848684210526316, "grad_norm": 181.8579864501953, "learning_rate": 5.115131578947369e-06, "loss": 2.7266, "step": 2376 }, { "epoch": 0.4886924342105263, "grad_norm": 188.7910919189453, "learning_rate": 5.113075657894737e-06, "loss": 2.6093, "step": 2377 }, { "epoch": 0.4888980263157895, "grad_norm": 111.1697006225586, "learning_rate": 5.111019736842105e-06, "loss": 1.767, "step": 2378 }, { "epoch": 0.48910361842105265, "grad_norm": 50.43949890136719, "learning_rate": 5.108963815789474e-06, "loss": 0.3813, "step": 2379 }, { "epoch": 0.48930921052631576, "grad_norm": 38.4139404296875, "learning_rate": 5.106907894736842e-06, "loss": 0.1885, "step": 2380 }, { "epoch": 0.4895148026315789, "grad_norm": 49.90441131591797, "learning_rate": 5.1048519736842105e-06, "loss": 0.4541, "step": 2381 }, { "epoch": 0.4897203947368421, "grad_norm": 62.533138275146484, "learning_rate": 5.102796052631579e-06, "loss": 1.1665, "step": 2382 }, { "epoch": 0.48992598684210525, "grad_norm": 92.21707153320312, "learning_rate": 5.100740131578947e-06, "loss": 0.7545, "step": 2383 }, { "epoch": 0.4901315789473684, "grad_norm": 17.241853713989258, "learning_rate": 5.098684210526316e-06, "loss": 0.0911, "step": 2384 }, { "epoch": 0.4903371710526316, "grad_norm": 103.47820281982422, "learning_rate": 5.0966282894736855e-06, "loss": 1.1785, "step": 2385 }, { "epoch": 0.49054276315789475, "grad_norm": 86.26962280273438, "learning_rate": 5.094572368421053e-06, "loss": 0.8059, "step": 2386 }, { "epoch": 0.4907483552631579, "grad_norm": 104.66565704345703, "learning_rate": 5.092516447368422e-06, "loss": 0.6422, "step": 2387 }, { "epoch": 0.4909539473684211, "grad_norm": 13.138910293579102, "learning_rate": 5.09046052631579e-06, "loss": 0.0382, "step": 2388 }, { "epoch": 0.49115953947368424, "grad_norm": 45.18273162841797, "learning_rate": 5.088404605263159e-06, "loss": 0.4023, "step": 2389 }, { "epoch": 0.49136513157894735, "grad_norm": 27.048444747924805, "learning_rate": 5.0863486842105266e-06, "loss": 0.2715, "step": 2390 }, { "epoch": 0.4915707236842105, "grad_norm": 165.1260223388672, "learning_rate": 5.084292763157895e-06, "loss": 1.2993, "step": 2391 }, { "epoch": 0.4917763157894737, "grad_norm": 201.4365234375, "learning_rate": 5.082236842105264e-06, "loss": 2.0488, "step": 2392 }, { "epoch": 0.49198190789473684, "grad_norm": 31.879009246826172, "learning_rate": 5.080180921052632e-06, "loss": 0.271, "step": 2393 }, { "epoch": 0.4921875, "grad_norm": 35.6368408203125, "learning_rate": 5.078125000000001e-06, "loss": 0.3538, "step": 2394 }, { "epoch": 0.49239309210526316, "grad_norm": 152.65185546875, "learning_rate": 5.0760690789473685e-06, "loss": 1.0898, "step": 2395 }, { "epoch": 0.4925986842105263, "grad_norm": 13.73837661743164, "learning_rate": 5.074013157894737e-06, "loss": 0.0728, "step": 2396 }, { "epoch": 0.4928042763157895, "grad_norm": 57.46505355834961, "learning_rate": 5.071957236842105e-06, "loss": 0.4158, "step": 2397 }, { "epoch": 0.49300986842105265, "grad_norm": 6.069149971008301, "learning_rate": 5.069901315789474e-06, "loss": 0.0411, "step": 2398 }, { "epoch": 0.49321546052631576, "grad_norm": 11.433404922485352, "learning_rate": 5.0678453947368426e-06, "loss": 0.0471, "step": 2399 }, { "epoch": 0.4934210526315789, "grad_norm": 88.08452606201172, "learning_rate": 5.0657894736842104e-06, "loss": 0.8516, "step": 2400 }, { "epoch": 0.4936266447368421, "grad_norm": 148.7735595703125, "learning_rate": 5.063733552631579e-06, "loss": 2.7656, "step": 2401 }, { "epoch": 0.49383223684210525, "grad_norm": 17.38428497314453, "learning_rate": 5.061677631578947e-06, "loss": 0.1194, "step": 2402 }, { "epoch": 0.4940378289473684, "grad_norm": 118.82417297363281, "learning_rate": 5.059621710526316e-06, "loss": 1.9561, "step": 2403 }, { "epoch": 0.4942434210526316, "grad_norm": 58.89101028442383, "learning_rate": 5.057565789473685e-06, "loss": 0.5449, "step": 2404 }, { "epoch": 0.49444901315789475, "grad_norm": 80.85092163085938, "learning_rate": 5.055509868421053e-06, "loss": 0.6758, "step": 2405 }, { "epoch": 0.4946546052631579, "grad_norm": 19.51936912536621, "learning_rate": 5.053453947368422e-06, "loss": 0.1163, "step": 2406 }, { "epoch": 0.4948601973684211, "grad_norm": 6.5008158683776855, "learning_rate": 5.05139802631579e-06, "loss": 0.0477, "step": 2407 }, { "epoch": 0.49506578947368424, "grad_norm": 135.5983428955078, "learning_rate": 5.0493421052631586e-06, "loss": 0.5946, "step": 2408 }, { "epoch": 0.49527138157894735, "grad_norm": 151.26031494140625, "learning_rate": 5.047286184210527e-06, "loss": 4.7061, "step": 2409 }, { "epoch": 0.4954769736842105, "grad_norm": 61.932369232177734, "learning_rate": 5.045230263157895e-06, "loss": 0.8652, "step": 2410 }, { "epoch": 0.4956825657894737, "grad_norm": 79.2599105834961, "learning_rate": 5.043174342105264e-06, "loss": 0.3647, "step": 2411 }, { "epoch": 0.49588815789473684, "grad_norm": 300.2742614746094, "learning_rate": 5.041118421052632e-06, "loss": 1.729, "step": 2412 }, { "epoch": 0.49609375, "grad_norm": 61.15687942504883, "learning_rate": 5.0390625000000005e-06, "loss": 0.5557, "step": 2413 }, { "epoch": 0.49629934210526316, "grad_norm": 29.7255916595459, "learning_rate": 5.037006578947368e-06, "loss": 0.1777, "step": 2414 }, { "epoch": 0.4965049342105263, "grad_norm": 88.48554229736328, "learning_rate": 5.034950657894737e-06, "loss": 1.134, "step": 2415 }, { "epoch": 0.4967105263157895, "grad_norm": 59.43230056762695, "learning_rate": 5.032894736842106e-06, "loss": 0.3348, "step": 2416 }, { "epoch": 0.49691611842105265, "grad_norm": 56.83453369140625, "learning_rate": 5.030838815789474e-06, "loss": 0.4639, "step": 2417 }, { "epoch": 0.49712171052631576, "grad_norm": 16.888874053955078, "learning_rate": 5.0287828947368425e-06, "loss": 0.1338, "step": 2418 }, { "epoch": 0.4973273026315789, "grad_norm": 61.937767028808594, "learning_rate": 5.02672697368421e-06, "loss": 0.4561, "step": 2419 }, { "epoch": 0.4975328947368421, "grad_norm": 31.236202239990234, "learning_rate": 5.024671052631579e-06, "loss": 0.269, "step": 2420 }, { "epoch": 0.49773848684210525, "grad_norm": 51.99871063232422, "learning_rate": 5.022615131578947e-06, "loss": 0.6875, "step": 2421 }, { "epoch": 0.4979440789473684, "grad_norm": 78.95699310302734, "learning_rate": 5.020559210526316e-06, "loss": 0.4141, "step": 2422 }, { "epoch": 0.4981496710526316, "grad_norm": 16.806411743164062, "learning_rate": 5.018503289473685e-06, "loss": 0.0671, "step": 2423 }, { "epoch": 0.49835526315789475, "grad_norm": 175.56675720214844, "learning_rate": 5.016447368421053e-06, "loss": 2.7578, "step": 2424 }, { "epoch": 0.4985608552631579, "grad_norm": 76.85417175292969, "learning_rate": 5.014391447368422e-06, "loss": 0.5977, "step": 2425 }, { "epoch": 0.4987664473684211, "grad_norm": 24.0203800201416, "learning_rate": 5.012335526315791e-06, "loss": 0.1812, "step": 2426 }, { "epoch": 0.49897203947368424, "grad_norm": 73.02824401855469, "learning_rate": 5.0102796052631585e-06, "loss": 0.5627, "step": 2427 }, { "epoch": 0.49917763157894735, "grad_norm": 13.548409461975098, "learning_rate": 5.008223684210527e-06, "loss": 0.0713, "step": 2428 }, { "epoch": 0.4993832236842105, "grad_norm": 120.8748779296875, "learning_rate": 5.006167763157895e-06, "loss": 0.4026, "step": 2429 }, { "epoch": 0.4995888157894737, "grad_norm": 158.2575225830078, "learning_rate": 5.004111842105264e-06, "loss": 0.6602, "step": 2430 }, { "epoch": 0.49979440789473684, "grad_norm": 99.43636322021484, "learning_rate": 5.002055921052632e-06, "loss": 1.1572, "step": 2431 }, { "epoch": 0.5, "grad_norm": 68.66667175292969, "learning_rate": 5e-06, "loss": 0.6055, "step": 2432 }, { "epoch": 0.5002055921052632, "grad_norm": 190.86886596679688, "learning_rate": 4.997944078947369e-06, "loss": 0.5612, "step": 2433 }, { "epoch": 0.5004111842105263, "grad_norm": 27.264219284057617, "learning_rate": 4.995888157894737e-06, "loss": 0.1187, "step": 2434 }, { "epoch": 0.5006167763157895, "grad_norm": 76.575439453125, "learning_rate": 4.993832236842106e-06, "loss": 0.9707, "step": 2435 }, { "epoch": 0.5008223684210527, "grad_norm": 113.88542938232422, "learning_rate": 4.991776315789474e-06, "loss": 1.6562, "step": 2436 }, { "epoch": 0.5010279605263158, "grad_norm": 97.75337219238281, "learning_rate": 4.989720394736842e-06, "loss": 0.4353, "step": 2437 }, { "epoch": 0.501233552631579, "grad_norm": 21.515779495239258, "learning_rate": 4.987664473684211e-06, "loss": 0.0739, "step": 2438 }, { "epoch": 0.5014391447368421, "grad_norm": 157.55288696289062, "learning_rate": 4.98560855263158e-06, "loss": 2.4375, "step": 2439 }, { "epoch": 0.5016447368421053, "grad_norm": 175.17376708984375, "learning_rate": 4.983552631578948e-06, "loss": 2.7441, "step": 2440 }, { "epoch": 0.5018503289473685, "grad_norm": 158.07101440429688, "learning_rate": 4.981496710526316e-06, "loss": 0.9824, "step": 2441 }, { "epoch": 0.5020559210526315, "grad_norm": 213.67237854003906, "learning_rate": 4.979440789473684e-06, "loss": 2.7461, "step": 2442 }, { "epoch": 0.5022615131578947, "grad_norm": 48.697757720947266, "learning_rate": 4.977384868421053e-06, "loss": 0.3191, "step": 2443 }, { "epoch": 0.5024671052631579, "grad_norm": 103.99913787841797, "learning_rate": 4.975328947368421e-06, "loss": 0.8212, "step": 2444 }, { "epoch": 0.502672697368421, "grad_norm": 42.268310546875, "learning_rate": 4.97327302631579e-06, "loss": 0.3008, "step": 2445 }, { "epoch": 0.5028782894736842, "grad_norm": 80.66464233398438, "learning_rate": 4.971217105263158e-06, "loss": 0.7888, "step": 2446 }, { "epoch": 0.5030838815789473, "grad_norm": 89.53196716308594, "learning_rate": 4.969161184210527e-06, "loss": 0.3257, "step": 2447 }, { "epoch": 0.5032894736842105, "grad_norm": 11.7713041305542, "learning_rate": 4.967105263157895e-06, "loss": 0.0879, "step": 2448 }, { "epoch": 0.5034950657894737, "grad_norm": 11.453790664672852, "learning_rate": 4.965049342105264e-06, "loss": 0.095, "step": 2449 }, { "epoch": 0.5037006578947368, "grad_norm": 38.11654281616211, "learning_rate": 4.9629934210526324e-06, "loss": 0.7471, "step": 2450 }, { "epoch": 0.50390625, "grad_norm": 42.06119155883789, "learning_rate": 4.9609375e-06, "loss": 0.2471, "step": 2451 }, { "epoch": 0.5041118421052632, "grad_norm": 83.2532958984375, "learning_rate": 4.958881578947369e-06, "loss": 0.7339, "step": 2452 }, { "epoch": 0.5043174342105263, "grad_norm": 150.11978149414062, "learning_rate": 4.956825657894737e-06, "loss": 1.4468, "step": 2453 }, { "epoch": 0.5045230263157895, "grad_norm": 34.25699234008789, "learning_rate": 4.954769736842106e-06, "loss": 0.2602, "step": 2454 }, { "epoch": 0.5047286184210527, "grad_norm": 18.424821853637695, "learning_rate": 4.9527138157894735e-06, "loss": 0.0991, "step": 2455 }, { "epoch": 0.5049342105263158, "grad_norm": 10.870779037475586, "learning_rate": 4.950657894736843e-06, "loss": 0.0351, "step": 2456 }, { "epoch": 0.505139802631579, "grad_norm": 230.38046264648438, "learning_rate": 4.948601973684211e-06, "loss": 2.5869, "step": 2457 }, { "epoch": 0.5053453947368421, "grad_norm": 14.71656608581543, "learning_rate": 4.94654605263158e-06, "loss": 0.0801, "step": 2458 }, { "epoch": 0.5055509868421053, "grad_norm": 66.83483123779297, "learning_rate": 4.944490131578948e-06, "loss": 0.311, "step": 2459 }, { "epoch": 0.5057565789473685, "grad_norm": 33.84559631347656, "learning_rate": 4.942434210526316e-06, "loss": 0.168, "step": 2460 }, { "epoch": 0.5059621710526315, "grad_norm": 20.594324111938477, "learning_rate": 4.940378289473684e-06, "loss": 0.1276, "step": 2461 }, { "epoch": 0.5061677631578947, "grad_norm": 28.21623420715332, "learning_rate": 4.938322368421053e-06, "loss": 0.1069, "step": 2462 }, { "epoch": 0.5063733552631579, "grad_norm": 148.474609375, "learning_rate": 4.936266447368422e-06, "loss": 1.3882, "step": 2463 }, { "epoch": 0.506578947368421, "grad_norm": 11.587239265441895, "learning_rate": 4.9342105263157895e-06, "loss": 0.0669, "step": 2464 }, { "epoch": 0.5067845394736842, "grad_norm": 102.733154296875, "learning_rate": 4.932154605263158e-06, "loss": 0.814, "step": 2465 }, { "epoch": 0.5069901315789473, "grad_norm": 17.05055809020996, "learning_rate": 4.930098684210527e-06, "loss": 0.0459, "step": 2466 }, { "epoch": 0.5071957236842105, "grad_norm": 76.01818084716797, "learning_rate": 4.928042763157896e-06, "loss": 0.8174, "step": 2467 }, { "epoch": 0.5074013157894737, "grad_norm": 64.72319793701172, "learning_rate": 4.925986842105264e-06, "loss": 0.3451, "step": 2468 }, { "epoch": 0.5076069078947368, "grad_norm": 114.83609008789062, "learning_rate": 4.923930921052632e-06, "loss": 0.8406, "step": 2469 }, { "epoch": 0.5078125, "grad_norm": 94.45932006835938, "learning_rate": 4.921875e-06, "loss": 0.71, "step": 2470 }, { "epoch": 0.5080180921052632, "grad_norm": 130.278076171875, "learning_rate": 4.919819078947369e-06, "loss": 0.8184, "step": 2471 }, { "epoch": 0.5082236842105263, "grad_norm": 28.123783111572266, "learning_rate": 4.917763157894737e-06, "loss": 0.1702, "step": 2472 }, { "epoch": 0.5084292763157895, "grad_norm": 41.86946105957031, "learning_rate": 4.9157072368421055e-06, "loss": 0.2423, "step": 2473 }, { "epoch": 0.5086348684210527, "grad_norm": 164.53985595703125, "learning_rate": 4.913651315789474e-06, "loss": 1.0459, "step": 2474 }, { "epoch": 0.5088404605263158, "grad_norm": 36.96900177001953, "learning_rate": 4.911595394736843e-06, "loss": 0.2462, "step": 2475 }, { "epoch": 0.509046052631579, "grad_norm": 58.50993347167969, "learning_rate": 4.909539473684211e-06, "loss": 0.2241, "step": 2476 }, { "epoch": 0.5092516447368421, "grad_norm": 37.82704544067383, "learning_rate": 4.90748355263158e-06, "loss": 0.4746, "step": 2477 }, { "epoch": 0.5094572368421053, "grad_norm": 145.01361083984375, "learning_rate": 4.9054276315789475e-06, "loss": 0.8702, "step": 2478 }, { "epoch": 0.5096628289473685, "grad_norm": 57.858489990234375, "learning_rate": 4.903371710526316e-06, "loss": 0.1865, "step": 2479 }, { "epoch": 0.5098684210526315, "grad_norm": 60.046180725097656, "learning_rate": 4.901315789473685e-06, "loss": 0.297, "step": 2480 }, { "epoch": 0.5100740131578947, "grad_norm": 169.00167846679688, "learning_rate": 4.899259868421053e-06, "loss": 3.3857, "step": 2481 }, { "epoch": 0.5102796052631579, "grad_norm": 195.74803161621094, "learning_rate": 4.8972039473684215e-06, "loss": 3.2656, "step": 2482 }, { "epoch": 0.510485197368421, "grad_norm": 16.38349723815918, "learning_rate": 4.8951480263157894e-06, "loss": 0.1058, "step": 2483 }, { "epoch": 0.5106907894736842, "grad_norm": 88.71133422851562, "learning_rate": 4.893092105263158e-06, "loss": 0.1722, "step": 2484 }, { "epoch": 0.5108963815789473, "grad_norm": 8.97817611694336, "learning_rate": 4.891036184210527e-06, "loss": 0.0413, "step": 2485 }, { "epoch": 0.5111019736842105, "grad_norm": 90.18560028076172, "learning_rate": 4.888980263157896e-06, "loss": 0.8252, "step": 2486 }, { "epoch": 0.5113075657894737, "grad_norm": 22.789478302001953, "learning_rate": 4.8869243421052635e-06, "loss": 0.1294, "step": 2487 }, { "epoch": 0.5115131578947368, "grad_norm": 64.30099487304688, "learning_rate": 4.884868421052632e-06, "loss": 0.4834, "step": 2488 }, { "epoch": 0.51171875, "grad_norm": 5.649136066436768, "learning_rate": 4.8828125e-06, "loss": 0.0333, "step": 2489 }, { "epoch": 0.5119243421052632, "grad_norm": 178.56556701660156, "learning_rate": 4.880756578947369e-06, "loss": 1.3203, "step": 2490 }, { "epoch": 0.5121299342105263, "grad_norm": 119.26478576660156, "learning_rate": 4.878700657894737e-06, "loss": 0.7107, "step": 2491 }, { "epoch": 0.5123355263157895, "grad_norm": 52.980403900146484, "learning_rate": 4.8766447368421054e-06, "loss": 0.5356, "step": 2492 }, { "epoch": 0.5125411184210527, "grad_norm": 5.514158725738525, "learning_rate": 4.874588815789474e-06, "loss": 0.0193, "step": 2493 }, { "epoch": 0.5127467105263158, "grad_norm": 89.0917739868164, "learning_rate": 4.872532894736843e-06, "loss": 1.2561, "step": 2494 }, { "epoch": 0.512952302631579, "grad_norm": 32.74583053588867, "learning_rate": 4.870476973684211e-06, "loss": 0.2626, "step": 2495 }, { "epoch": 0.5131578947368421, "grad_norm": 266.98541259765625, "learning_rate": 4.8684210526315795e-06, "loss": 1.0469, "step": 2496 }, { "epoch": 0.5133634868421053, "grad_norm": 126.97945404052734, "learning_rate": 4.866365131578948e-06, "loss": 0.998, "step": 2497 }, { "epoch": 0.5135690789473685, "grad_norm": 51.909183502197266, "learning_rate": 4.864309210526316e-06, "loss": 0.3016, "step": 2498 }, { "epoch": 0.5137746710526315, "grad_norm": 4.098408222198486, "learning_rate": 4.862253289473685e-06, "loss": 0.0233, "step": 2499 }, { "epoch": 0.5139802631578947, "grad_norm": 63.16706848144531, "learning_rate": 4.860197368421053e-06, "loss": 0.375, "step": 2500 }, { "epoch": 0.5141858552631579, "grad_norm": 80.20988464355469, "learning_rate": 4.8581414473684214e-06, "loss": 0.3498, "step": 2501 }, { "epoch": 0.514391447368421, "grad_norm": 31.7918758392334, "learning_rate": 4.856085526315789e-06, "loss": 0.1831, "step": 2502 }, { "epoch": 0.5145970394736842, "grad_norm": 14.639384269714355, "learning_rate": 4.854029605263158e-06, "loss": 0.1311, "step": 2503 }, { "epoch": 0.5148026315789473, "grad_norm": 86.41026306152344, "learning_rate": 4.851973684210527e-06, "loss": 0.621, "step": 2504 }, { "epoch": 0.5150082236842105, "grad_norm": 126.66027069091797, "learning_rate": 4.8499177631578955e-06, "loss": 0.8497, "step": 2505 }, { "epoch": 0.5152138157894737, "grad_norm": 113.05310821533203, "learning_rate": 4.847861842105263e-06, "loss": 2.1055, "step": 2506 }, { "epoch": 0.5154194078947368, "grad_norm": 35.39124298095703, "learning_rate": 4.845805921052632e-06, "loss": 0.1929, "step": 2507 }, { "epoch": 0.515625, "grad_norm": 14.753957748413086, "learning_rate": 4.84375e-06, "loss": 0.0356, "step": 2508 }, { "epoch": 0.5158305921052632, "grad_norm": 81.07274627685547, "learning_rate": 4.841694078947369e-06, "loss": 0.9121, "step": 2509 }, { "epoch": 0.5160361842105263, "grad_norm": 15.900760650634766, "learning_rate": 4.8396381578947375e-06, "loss": 0.0706, "step": 2510 }, { "epoch": 0.5162417763157895, "grad_norm": 185.40921020507812, "learning_rate": 4.837582236842105e-06, "loss": 2.0825, "step": 2511 }, { "epoch": 0.5164473684210527, "grad_norm": 88.94738006591797, "learning_rate": 4.835526315789474e-06, "loss": 0.4604, "step": 2512 }, { "epoch": 0.5166529605263158, "grad_norm": 226.10386657714844, "learning_rate": 4.833470394736843e-06, "loss": 1.4297, "step": 2513 }, { "epoch": 0.516858552631579, "grad_norm": 74.75138092041016, "learning_rate": 4.8314144736842115e-06, "loss": 0.9883, "step": 2514 }, { "epoch": 0.5170641447368421, "grad_norm": 28.804298400878906, "learning_rate": 4.829358552631579e-06, "loss": 0.1664, "step": 2515 }, { "epoch": 0.5172697368421053, "grad_norm": 48.378421783447266, "learning_rate": 4.827302631578948e-06, "loss": 0.4824, "step": 2516 }, { "epoch": 0.5174753289473685, "grad_norm": 261.0107727050781, "learning_rate": 4.825246710526316e-06, "loss": 3.4766, "step": 2517 }, { "epoch": 0.5176809210526315, "grad_norm": 110.87188720703125, "learning_rate": 4.823190789473685e-06, "loss": 0.7026, "step": 2518 }, { "epoch": 0.5178865131578947, "grad_norm": 19.381305694580078, "learning_rate": 4.821134868421053e-06, "loss": 0.0925, "step": 2519 }, { "epoch": 0.5180921052631579, "grad_norm": 114.70738220214844, "learning_rate": 4.819078947368421e-06, "loss": 1.4888, "step": 2520 }, { "epoch": 0.518297697368421, "grad_norm": 25.266878128051758, "learning_rate": 4.81702302631579e-06, "loss": 0.1368, "step": 2521 }, { "epoch": 0.5185032894736842, "grad_norm": 8.754734992980957, "learning_rate": 4.814967105263158e-06, "loss": 0.038, "step": 2522 }, { "epoch": 0.5187088815789473, "grad_norm": 125.33676147460938, "learning_rate": 4.812911184210527e-06, "loss": 1.0527, "step": 2523 }, { "epoch": 0.5189144736842105, "grad_norm": 32.01841354370117, "learning_rate": 4.810855263157895e-06, "loss": 0.131, "step": 2524 }, { "epoch": 0.5191200657894737, "grad_norm": 39.360740661621094, "learning_rate": 4.808799342105263e-06, "loss": 0.1851, "step": 2525 }, { "epoch": 0.5193256578947368, "grad_norm": 13.109325408935547, "learning_rate": 4.806743421052632e-06, "loss": 0.0582, "step": 2526 }, { "epoch": 0.51953125, "grad_norm": 210.92990112304688, "learning_rate": 4.804687500000001e-06, "loss": 1.8555, "step": 2527 }, { "epoch": 0.5197368421052632, "grad_norm": 88.42770385742188, "learning_rate": 4.802631578947369e-06, "loss": 0.793, "step": 2528 }, { "epoch": 0.5199424342105263, "grad_norm": 90.5772705078125, "learning_rate": 4.800575657894737e-06, "loss": 0.7052, "step": 2529 }, { "epoch": 0.5201480263157895, "grad_norm": 64.58061981201172, "learning_rate": 4.798519736842105e-06, "loss": 0.2249, "step": 2530 }, { "epoch": 0.5203536184210527, "grad_norm": 217.70079040527344, "learning_rate": 4.796463815789474e-06, "loss": 1.2891, "step": 2531 }, { "epoch": 0.5205592105263158, "grad_norm": 21.65707015991211, "learning_rate": 4.794407894736843e-06, "loss": 0.0902, "step": 2532 }, { "epoch": 0.520764802631579, "grad_norm": 227.15826416015625, "learning_rate": 4.792351973684211e-06, "loss": 2.9844, "step": 2533 }, { "epoch": 0.5209703947368421, "grad_norm": 55.53754425048828, "learning_rate": 4.790296052631579e-06, "loss": 0.5224, "step": 2534 }, { "epoch": 0.5211759868421053, "grad_norm": 83.62336730957031, "learning_rate": 4.788240131578948e-06, "loss": 0.6519, "step": 2535 }, { "epoch": 0.5213815789473685, "grad_norm": 52.10239028930664, "learning_rate": 4.786184210526316e-06, "loss": 0.1831, "step": 2536 }, { "epoch": 0.5215871710526315, "grad_norm": 29.586956024169922, "learning_rate": 4.784128289473685e-06, "loss": 0.1636, "step": 2537 }, { "epoch": 0.5217927631578947, "grad_norm": 48.24326705932617, "learning_rate": 4.7820723684210525e-06, "loss": 0.3887, "step": 2538 }, { "epoch": 0.5219983552631579, "grad_norm": 28.781389236450195, "learning_rate": 4.780016447368421e-06, "loss": 0.1705, "step": 2539 }, { "epoch": 0.522203947368421, "grad_norm": 269.209228515625, "learning_rate": 4.77796052631579e-06, "loss": 3.9609, "step": 2540 }, { "epoch": 0.5224095394736842, "grad_norm": 164.52513122558594, "learning_rate": 4.775904605263158e-06, "loss": 2.0, "step": 2541 }, { "epoch": 0.5226151315789473, "grad_norm": 41.78136444091797, "learning_rate": 4.773848684210527e-06, "loss": 0.2198, "step": 2542 }, { "epoch": 0.5228207236842105, "grad_norm": 51.07315444946289, "learning_rate": 4.771792763157895e-06, "loss": 0.5039, "step": 2543 }, { "epoch": 0.5230263157894737, "grad_norm": 87.6606216430664, "learning_rate": 4.769736842105264e-06, "loss": 1.0142, "step": 2544 }, { "epoch": 0.5232319078947368, "grad_norm": 70.41867065429688, "learning_rate": 4.767680921052632e-06, "loss": 0.5748, "step": 2545 }, { "epoch": 0.5234375, "grad_norm": 11.13808822631836, "learning_rate": 4.765625000000001e-06, "loss": 0.0602, "step": 2546 }, { "epoch": 0.5236430921052632, "grad_norm": 128.92909240722656, "learning_rate": 4.7635690789473685e-06, "loss": 1.4883, "step": 2547 }, { "epoch": 0.5238486842105263, "grad_norm": 113.47508239746094, "learning_rate": 4.761513157894737e-06, "loss": 1.8252, "step": 2548 }, { "epoch": 0.5240542763157895, "grad_norm": 55.2276725769043, "learning_rate": 4.759457236842105e-06, "loss": 0.6553, "step": 2549 }, { "epoch": 0.5242598684210527, "grad_norm": 17.871379852294922, "learning_rate": 4.757401315789474e-06, "loss": 0.0554, "step": 2550 }, { "epoch": 0.5244654605263158, "grad_norm": 74.57561492919922, "learning_rate": 4.755345394736843e-06, "loss": 0.5054, "step": 2551 }, { "epoch": 0.524671052631579, "grad_norm": 151.5752716064453, "learning_rate": 4.753289473684211e-06, "loss": 1.0371, "step": 2552 }, { "epoch": 0.5248766447368421, "grad_norm": 15.03533935546875, "learning_rate": 4.751233552631579e-06, "loss": 0.0574, "step": 2553 }, { "epoch": 0.5250822368421053, "grad_norm": 34.21398162841797, "learning_rate": 4.749177631578948e-06, "loss": 0.1926, "step": 2554 }, { "epoch": 0.5252878289473685, "grad_norm": 17.548542022705078, "learning_rate": 4.747121710526316e-06, "loss": 0.1387, "step": 2555 }, { "epoch": 0.5254934210526315, "grad_norm": 33.55874252319336, "learning_rate": 4.7450657894736845e-06, "loss": 0.1487, "step": 2556 }, { "epoch": 0.5256990131578947, "grad_norm": 21.576313018798828, "learning_rate": 4.743009868421053e-06, "loss": 0.0577, "step": 2557 }, { "epoch": 0.5259046052631579, "grad_norm": 51.996978759765625, "learning_rate": 4.740953947368421e-06, "loss": 0.3311, "step": 2558 }, { "epoch": 0.526110197368421, "grad_norm": 97.5577163696289, "learning_rate": 4.73889802631579e-06, "loss": 0.6719, "step": 2559 }, { "epoch": 0.5263157894736842, "grad_norm": 33.2539176940918, "learning_rate": 4.736842105263158e-06, "loss": 0.208, "step": 2560 }, { "epoch": 0.5265213815789473, "grad_norm": 30.054655075073242, "learning_rate": 4.734786184210527e-06, "loss": 0.137, "step": 2561 }, { "epoch": 0.5267269736842105, "grad_norm": 31.449615478515625, "learning_rate": 4.732730263157895e-06, "loss": 0.0977, "step": 2562 }, { "epoch": 0.5269325657894737, "grad_norm": 38.69242858886719, "learning_rate": 4.730674342105264e-06, "loss": 0.3672, "step": 2563 }, { "epoch": 0.5271381578947368, "grad_norm": 75.14065551757812, "learning_rate": 4.728618421052632e-06, "loss": 1.2335, "step": 2564 }, { "epoch": 0.52734375, "grad_norm": 64.29286193847656, "learning_rate": 4.7265625000000005e-06, "loss": 1.0312, "step": 2565 }, { "epoch": 0.5275493421052632, "grad_norm": 34.886619567871094, "learning_rate": 4.724506578947368e-06, "loss": 0.2476, "step": 2566 }, { "epoch": 0.5277549342105263, "grad_norm": 24.065324783325195, "learning_rate": 4.722450657894737e-06, "loss": 0.1277, "step": 2567 }, { "epoch": 0.5279605263157895, "grad_norm": 32.50429153442383, "learning_rate": 4.720394736842106e-06, "loss": 0.2283, "step": 2568 }, { "epoch": 0.5281661184210527, "grad_norm": 121.2823486328125, "learning_rate": 4.718338815789474e-06, "loss": 1.1367, "step": 2569 }, { "epoch": 0.5283717105263158, "grad_norm": 269.4390563964844, "learning_rate": 4.7162828947368425e-06, "loss": 1.7266, "step": 2570 }, { "epoch": 0.528577302631579, "grad_norm": 15.788678169250488, "learning_rate": 4.714226973684211e-06, "loss": 0.1401, "step": 2571 }, { "epoch": 0.5287828947368421, "grad_norm": 135.06661987304688, "learning_rate": 4.712171052631579e-06, "loss": 1.0312, "step": 2572 }, { "epoch": 0.5289884868421053, "grad_norm": 10.906783103942871, "learning_rate": 4.710115131578948e-06, "loss": 0.0406, "step": 2573 }, { "epoch": 0.5291940789473685, "grad_norm": 4.752174377441406, "learning_rate": 4.7080592105263165e-06, "loss": 0.0237, "step": 2574 }, { "epoch": 0.5293996710526315, "grad_norm": 47.52561569213867, "learning_rate": 4.7060032894736844e-06, "loss": 0.1641, "step": 2575 }, { "epoch": 0.5296052631578947, "grad_norm": 224.59402465820312, "learning_rate": 4.703947368421053e-06, "loss": 1.1088, "step": 2576 }, { "epoch": 0.5298108552631579, "grad_norm": 27.41977882385254, "learning_rate": 4.701891447368421e-06, "loss": 0.2131, "step": 2577 }, { "epoch": 0.530016447368421, "grad_norm": 81.95917510986328, "learning_rate": 4.69983552631579e-06, "loss": 0.936, "step": 2578 }, { "epoch": 0.5302220394736842, "grad_norm": 33.77156066894531, "learning_rate": 4.697779605263158e-06, "loss": 0.2764, "step": 2579 }, { "epoch": 0.5304276315789473, "grad_norm": 143.51617431640625, "learning_rate": 4.695723684210527e-06, "loss": 0.9658, "step": 2580 }, { "epoch": 0.5306332236842105, "grad_norm": 62.45726776123047, "learning_rate": 4.693667763157895e-06, "loss": 0.3497, "step": 2581 }, { "epoch": 0.5308388157894737, "grad_norm": 74.44328308105469, "learning_rate": 4.691611842105264e-06, "loss": 0.5, "step": 2582 }, { "epoch": 0.5310444078947368, "grad_norm": 124.4732894897461, "learning_rate": 4.689555921052632e-06, "loss": 1.1714, "step": 2583 }, { "epoch": 0.53125, "grad_norm": 73.89456176757812, "learning_rate": 4.6875000000000004e-06, "loss": 0.3511, "step": 2584 }, { "epoch": 0.5314555921052632, "grad_norm": 47.344295501708984, "learning_rate": 4.685444078947369e-06, "loss": 1.1094, "step": 2585 }, { "epoch": 0.5316611842105263, "grad_norm": 136.5515594482422, "learning_rate": 4.683388157894737e-06, "loss": 1.1104, "step": 2586 }, { "epoch": 0.5318667763157895, "grad_norm": 150.0205841064453, "learning_rate": 4.681332236842106e-06, "loss": 0.751, "step": 2587 }, { "epoch": 0.5320723684210527, "grad_norm": 29.98991584777832, "learning_rate": 4.679276315789474e-06, "loss": 0.1039, "step": 2588 }, { "epoch": 0.5322779605263158, "grad_norm": 28.51000213623047, "learning_rate": 4.677220394736842e-06, "loss": 0.075, "step": 2589 }, { "epoch": 0.532483552631579, "grad_norm": 336.19171142578125, "learning_rate": 4.675164473684211e-06, "loss": 1.3699, "step": 2590 }, { "epoch": 0.5326891447368421, "grad_norm": 53.924991607666016, "learning_rate": 4.67310855263158e-06, "loss": 0.3208, "step": 2591 }, { "epoch": 0.5328947368421053, "grad_norm": 36.063358306884766, "learning_rate": 4.671052631578948e-06, "loss": 0.2317, "step": 2592 }, { "epoch": 0.5331003289473685, "grad_norm": 33.8284912109375, "learning_rate": 4.6689967105263164e-06, "loss": 0.2104, "step": 2593 }, { "epoch": 0.5333059210526315, "grad_norm": 44.55500793457031, "learning_rate": 4.666940789473684e-06, "loss": 0.4697, "step": 2594 }, { "epoch": 0.5335115131578947, "grad_norm": 101.83476257324219, "learning_rate": 4.664884868421053e-06, "loss": 1.8081, "step": 2595 }, { "epoch": 0.5337171052631579, "grad_norm": 126.4298095703125, "learning_rate": 4.662828947368421e-06, "loss": 0.6367, "step": 2596 }, { "epoch": 0.533922697368421, "grad_norm": 27.634214401245117, "learning_rate": 4.66077302631579e-06, "loss": 0.1829, "step": 2597 }, { "epoch": 0.5341282894736842, "grad_norm": 300.606201171875, "learning_rate": 4.658717105263158e-06, "loss": 3.042, "step": 2598 }, { "epoch": 0.5343338815789473, "grad_norm": 31.236316680908203, "learning_rate": 4.656661184210527e-06, "loss": 0.1221, "step": 2599 }, { "epoch": 0.5345394736842105, "grad_norm": 15.824540138244629, "learning_rate": 4.654605263157895e-06, "loss": 0.0536, "step": 2600 }, { "epoch": 0.5347450657894737, "grad_norm": 7.855496883392334, "learning_rate": 4.652549342105264e-06, "loss": 0.0406, "step": 2601 }, { "epoch": 0.5349506578947368, "grad_norm": 103.66116333007812, "learning_rate": 4.650493421052632e-06, "loss": 0.5293, "step": 2602 }, { "epoch": 0.53515625, "grad_norm": 22.62796974182129, "learning_rate": 4.6484375e-06, "loss": 0.0854, "step": 2603 }, { "epoch": 0.5353618421052632, "grad_norm": 35.07988739013672, "learning_rate": 4.646381578947369e-06, "loss": 0.199, "step": 2604 }, { "epoch": 0.5355674342105263, "grad_norm": 152.89393615722656, "learning_rate": 4.644325657894737e-06, "loss": 0.8489, "step": 2605 }, { "epoch": 0.5357730263157895, "grad_norm": 29.131595611572266, "learning_rate": 4.642269736842106e-06, "loss": 0.1335, "step": 2606 }, { "epoch": 0.5359786184210527, "grad_norm": 79.40202331542969, "learning_rate": 4.6402138157894736e-06, "loss": 0.4951, "step": 2607 }, { "epoch": 0.5361842105263158, "grad_norm": 66.31497192382812, "learning_rate": 4.638157894736843e-06, "loss": 0.9355, "step": 2608 }, { "epoch": 0.536389802631579, "grad_norm": 6.714367866516113, "learning_rate": 4.636101973684211e-06, "loss": 0.0143, "step": 2609 }, { "epoch": 0.5365953947368421, "grad_norm": 158.49237060546875, "learning_rate": 4.63404605263158e-06, "loss": 1.1316, "step": 2610 }, { "epoch": 0.5368009868421053, "grad_norm": 23.168806076049805, "learning_rate": 4.631990131578948e-06, "loss": 0.1494, "step": 2611 }, { "epoch": 0.5370065789473685, "grad_norm": 88.49691772460938, "learning_rate": 4.629934210526316e-06, "loss": 0.3879, "step": 2612 }, { "epoch": 0.5372121710526315, "grad_norm": 278.0403747558594, "learning_rate": 4.627878289473684e-06, "loss": 1.158, "step": 2613 }, { "epoch": 0.5374177631578947, "grad_norm": 42.53784942626953, "learning_rate": 4.625822368421053e-06, "loss": 0.302, "step": 2614 }, { "epoch": 0.5376233552631579, "grad_norm": 82.35794067382812, "learning_rate": 4.623766447368422e-06, "loss": 0.3833, "step": 2615 }, { "epoch": 0.537828947368421, "grad_norm": 33.796714782714844, "learning_rate": 4.6217105263157896e-06, "loss": 0.1133, "step": 2616 }, { "epoch": 0.5380345394736842, "grad_norm": 75.96158599853516, "learning_rate": 4.619654605263158e-06, "loss": 0.4426, "step": 2617 }, { "epoch": 0.5382401315789473, "grad_norm": 34.498374938964844, "learning_rate": 4.617598684210527e-06, "loss": 0.1151, "step": 2618 }, { "epoch": 0.5384457236842105, "grad_norm": 86.24179077148438, "learning_rate": 4.615542763157895e-06, "loss": 0.4444, "step": 2619 }, { "epoch": 0.5386513157894737, "grad_norm": 39.68503952026367, "learning_rate": 4.613486842105264e-06, "loss": 0.1675, "step": 2620 }, { "epoch": 0.5388569078947368, "grad_norm": 45.76935577392578, "learning_rate": 4.611430921052632e-06, "loss": 0.3411, "step": 2621 }, { "epoch": 0.5390625, "grad_norm": 34.04155349731445, "learning_rate": 4.609375e-06, "loss": 0.2408, "step": 2622 }, { "epoch": 0.5392680921052632, "grad_norm": 37.11834716796875, "learning_rate": 4.607319078947369e-06, "loss": 0.2015, "step": 2623 }, { "epoch": 0.5394736842105263, "grad_norm": 132.9386444091797, "learning_rate": 4.605263157894737e-06, "loss": 1.2227, "step": 2624 }, { "epoch": 0.5396792763157895, "grad_norm": 83.42247772216797, "learning_rate": 4.6032072368421056e-06, "loss": 0.4072, "step": 2625 }, { "epoch": 0.5398848684210527, "grad_norm": 70.1539535522461, "learning_rate": 4.6011513157894734e-06, "loss": 0.4878, "step": 2626 }, { "epoch": 0.5400904605263158, "grad_norm": 43.383544921875, "learning_rate": 4.599095394736843e-06, "loss": 0.1278, "step": 2627 }, { "epoch": 0.540296052631579, "grad_norm": 168.23631286621094, "learning_rate": 4.597039473684211e-06, "loss": 2.1328, "step": 2628 }, { "epoch": 0.5405016447368421, "grad_norm": 54.730037689208984, "learning_rate": 4.59498355263158e-06, "loss": 0.5114, "step": 2629 }, { "epoch": 0.5407072368421053, "grad_norm": 120.12215423583984, "learning_rate": 4.5929276315789475e-06, "loss": 1.1143, "step": 2630 }, { "epoch": 0.5409128289473685, "grad_norm": 123.1876220703125, "learning_rate": 4.590871710526316e-06, "loss": 1.1813, "step": 2631 }, { "epoch": 0.5411184210526315, "grad_norm": 24.11412811279297, "learning_rate": 4.588815789473685e-06, "loss": 0.0869, "step": 2632 }, { "epoch": 0.5413240131578947, "grad_norm": 34.266021728515625, "learning_rate": 4.586759868421053e-06, "loss": 0.1436, "step": 2633 }, { "epoch": 0.5415296052631579, "grad_norm": 201.4383087158203, "learning_rate": 4.5847039473684216e-06, "loss": 1.3488, "step": 2634 }, { "epoch": 0.541735197368421, "grad_norm": 233.6190185546875, "learning_rate": 4.5826480263157895e-06, "loss": 1.9512, "step": 2635 }, { "epoch": 0.5419407894736842, "grad_norm": 11.55430793762207, "learning_rate": 4.580592105263158e-06, "loss": 0.04, "step": 2636 }, { "epoch": 0.5421463815789473, "grad_norm": 55.1565055847168, "learning_rate": 4.578536184210527e-06, "loss": 0.4258, "step": 2637 }, { "epoch": 0.5423519736842105, "grad_norm": 452.1308288574219, "learning_rate": 4.576480263157896e-06, "loss": 1.918, "step": 2638 }, { "epoch": 0.5425575657894737, "grad_norm": 243.11659240722656, "learning_rate": 4.5744243421052635e-06, "loss": 2.043, "step": 2639 }, { "epoch": 0.5427631578947368, "grad_norm": 52.04668426513672, "learning_rate": 4.572368421052632e-06, "loss": 0.5666, "step": 2640 }, { "epoch": 0.54296875, "grad_norm": 17.783679962158203, "learning_rate": 4.5703125e-06, "loss": 0.0774, "step": 2641 }, { "epoch": 0.5431743421052632, "grad_norm": 66.38726806640625, "learning_rate": 4.568256578947369e-06, "loss": 0.3896, "step": 2642 }, { "epoch": 0.5433799342105263, "grad_norm": 75.26175689697266, "learning_rate": 4.566200657894737e-06, "loss": 0.5781, "step": 2643 }, { "epoch": 0.5435855263157895, "grad_norm": 79.85877990722656, "learning_rate": 4.5641447368421055e-06, "loss": 1.0072, "step": 2644 }, { "epoch": 0.5437911184210527, "grad_norm": 20.63709259033203, "learning_rate": 4.562088815789474e-06, "loss": 0.1982, "step": 2645 }, { "epoch": 0.5439967105263158, "grad_norm": 13.35194206237793, "learning_rate": 4.560032894736843e-06, "loss": 0.0681, "step": 2646 }, { "epoch": 0.544202302631579, "grad_norm": 243.44720458984375, "learning_rate": 4.557976973684211e-06, "loss": 0.876, "step": 2647 }, { "epoch": 0.5444078947368421, "grad_norm": 12.634410858154297, "learning_rate": 4.5559210526315795e-06, "loss": 0.0283, "step": 2648 }, { "epoch": 0.5446134868421053, "grad_norm": 153.57223510742188, "learning_rate": 4.553865131578947e-06, "loss": 0.8646, "step": 2649 }, { "epoch": 0.5448190789473685, "grad_norm": 76.71031188964844, "learning_rate": 4.551809210526316e-06, "loss": 1.1831, "step": 2650 }, { "epoch": 0.5450246710526315, "grad_norm": 17.772626876831055, "learning_rate": 4.549753289473685e-06, "loss": 0.1416, "step": 2651 }, { "epoch": 0.5452302631578947, "grad_norm": 100.31713104248047, "learning_rate": 4.547697368421053e-06, "loss": 1.4844, "step": 2652 }, { "epoch": 0.5454358552631579, "grad_norm": 48.75006866455078, "learning_rate": 4.5456414473684215e-06, "loss": 0.759, "step": 2653 }, { "epoch": 0.545641447368421, "grad_norm": 23.441190719604492, "learning_rate": 4.543585526315789e-06, "loss": 0.1055, "step": 2654 }, { "epoch": 0.5458470394736842, "grad_norm": 175.57774353027344, "learning_rate": 4.541529605263158e-06, "loss": 2.6953, "step": 2655 }, { "epoch": 0.5460526315789473, "grad_norm": 173.57566833496094, "learning_rate": 4.539473684210527e-06, "loss": 2.2031, "step": 2656 }, { "epoch": 0.5462582236842105, "grad_norm": 52.3631591796875, "learning_rate": 4.5374177631578955e-06, "loss": 0.2465, "step": 2657 }, { "epoch": 0.5464638157894737, "grad_norm": 39.41755294799805, "learning_rate": 4.535361842105263e-06, "loss": 0.4223, "step": 2658 }, { "epoch": 0.5466694078947368, "grad_norm": 59.34963607788086, "learning_rate": 4.533305921052632e-06, "loss": 0.5115, "step": 2659 }, { "epoch": 0.546875, "grad_norm": 17.513065338134766, "learning_rate": 4.53125e-06, "loss": 0.1299, "step": 2660 }, { "epoch": 0.5470805921052632, "grad_norm": 40.43659210205078, "learning_rate": 4.529194078947369e-06, "loss": 0.2134, "step": 2661 }, { "epoch": 0.5472861842105263, "grad_norm": 26.503286361694336, "learning_rate": 4.5271381578947375e-06, "loss": 0.121, "step": 2662 }, { "epoch": 0.5474917763157895, "grad_norm": 7.216920852661133, "learning_rate": 4.525082236842105e-06, "loss": 0.0525, "step": 2663 }, { "epoch": 0.5476973684210527, "grad_norm": 23.461132049560547, "learning_rate": 4.523026315789474e-06, "loss": 0.1443, "step": 2664 }, { "epoch": 0.5479029605263158, "grad_norm": 85.2921142578125, "learning_rate": 4.520970394736843e-06, "loss": 1.002, "step": 2665 }, { "epoch": 0.548108552631579, "grad_norm": 26.27740478515625, "learning_rate": 4.518914473684211e-06, "loss": 0.1095, "step": 2666 }, { "epoch": 0.5483141447368421, "grad_norm": 41.473655700683594, "learning_rate": 4.5168585526315794e-06, "loss": 0.3234, "step": 2667 }, { "epoch": 0.5485197368421053, "grad_norm": 172.3384246826172, "learning_rate": 4.514802631578948e-06, "loss": 1.4394, "step": 2668 }, { "epoch": 0.5487253289473685, "grad_norm": 56.134098052978516, "learning_rate": 4.512746710526316e-06, "loss": 0.7883, "step": 2669 }, { "epoch": 0.5489309210526315, "grad_norm": 28.01708984375, "learning_rate": 4.510690789473685e-06, "loss": 0.119, "step": 2670 }, { "epoch": 0.5491365131578947, "grad_norm": 34.41838455200195, "learning_rate": 4.508634868421053e-06, "loss": 0.1855, "step": 2671 }, { "epoch": 0.5493421052631579, "grad_norm": 34.75790786743164, "learning_rate": 4.506578947368421e-06, "loss": 0.1359, "step": 2672 }, { "epoch": 0.549547697368421, "grad_norm": 223.4787139892578, "learning_rate": 4.504523026315789e-06, "loss": 1.6509, "step": 2673 }, { "epoch": 0.5497532894736842, "grad_norm": 89.535400390625, "learning_rate": 4.502467105263158e-06, "loss": 1.0352, "step": 2674 }, { "epoch": 0.5499588815789473, "grad_norm": 37.20682144165039, "learning_rate": 4.500411184210527e-06, "loss": 0.3037, "step": 2675 }, { "epoch": 0.5501644736842105, "grad_norm": 19.969158172607422, "learning_rate": 4.4983552631578954e-06, "loss": 0.1006, "step": 2676 }, { "epoch": 0.5503700657894737, "grad_norm": 77.7493896484375, "learning_rate": 4.496299342105263e-06, "loss": 0.6914, "step": 2677 }, { "epoch": 0.5505756578947368, "grad_norm": 71.66079711914062, "learning_rate": 4.494243421052632e-06, "loss": 0.2342, "step": 2678 }, { "epoch": 0.55078125, "grad_norm": 47.1254997253418, "learning_rate": 4.492187500000001e-06, "loss": 0.4133, "step": 2679 }, { "epoch": 0.5509868421052632, "grad_norm": 97.34684753417969, "learning_rate": 4.490131578947369e-06, "loss": 0.4795, "step": 2680 }, { "epoch": 0.5511924342105263, "grad_norm": 109.09290313720703, "learning_rate": 4.488075657894737e-06, "loss": 0.8379, "step": 2681 }, { "epoch": 0.5513980263157895, "grad_norm": 51.29004669189453, "learning_rate": 4.486019736842105e-06, "loss": 0.9189, "step": 2682 }, { "epoch": 0.5516036184210527, "grad_norm": 186.6429443359375, "learning_rate": 4.483963815789474e-06, "loss": 0.604, "step": 2683 }, { "epoch": 0.5518092105263158, "grad_norm": 75.94380950927734, "learning_rate": 4.481907894736843e-06, "loss": 0.6553, "step": 2684 }, { "epoch": 0.552014802631579, "grad_norm": 24.751834869384766, "learning_rate": 4.4798519736842114e-06, "loss": 0.0961, "step": 2685 }, { "epoch": 0.5522203947368421, "grad_norm": 132.3941192626953, "learning_rate": 4.477796052631579e-06, "loss": 1.4531, "step": 2686 }, { "epoch": 0.5524259868421053, "grad_norm": 34.10093688964844, "learning_rate": 4.475740131578948e-06, "loss": 0.1973, "step": 2687 }, { "epoch": 0.5526315789473685, "grad_norm": 23.60625457763672, "learning_rate": 4.473684210526316e-06, "loss": 0.092, "step": 2688 }, { "epoch": 0.5528371710526315, "grad_norm": 95.33992767333984, "learning_rate": 4.471628289473685e-06, "loss": 0.8408, "step": 2689 }, { "epoch": 0.5530427631578947, "grad_norm": 47.256439208984375, "learning_rate": 4.4695723684210525e-06, "loss": 0.282, "step": 2690 }, { "epoch": 0.5532483552631579, "grad_norm": 114.31657409667969, "learning_rate": 4.467516447368421e-06, "loss": 0.5103, "step": 2691 }, { "epoch": 0.553453947368421, "grad_norm": 8.245072364807129, "learning_rate": 4.46546052631579e-06, "loss": 0.0701, "step": 2692 }, { "epoch": 0.5536595394736842, "grad_norm": 122.31348419189453, "learning_rate": 4.463404605263158e-06, "loss": 2.3281, "step": 2693 }, { "epoch": 0.5538651315789473, "grad_norm": 177.15191650390625, "learning_rate": 4.461348684210527e-06, "loss": 0.7432, "step": 2694 }, { "epoch": 0.5540707236842105, "grad_norm": 12.106295585632324, "learning_rate": 4.459292763157895e-06, "loss": 0.0335, "step": 2695 }, { "epoch": 0.5542763157894737, "grad_norm": 21.985013961791992, "learning_rate": 4.457236842105264e-06, "loss": 0.1591, "step": 2696 }, { "epoch": 0.5544819078947368, "grad_norm": 2.595909833908081, "learning_rate": 4.455180921052632e-06, "loss": 0.0163, "step": 2697 }, { "epoch": 0.5546875, "grad_norm": 220.0554656982422, "learning_rate": 4.453125000000001e-06, "loss": 2.0312, "step": 2698 }, { "epoch": 0.5548930921052632, "grad_norm": 199.4898223876953, "learning_rate": 4.4510690789473685e-06, "loss": 2.5537, "step": 2699 }, { "epoch": 0.5550986842105263, "grad_norm": 81.89306640625, "learning_rate": 4.449013157894737e-06, "loss": 0.4905, "step": 2700 }, { "epoch": 0.5553042763157895, "grad_norm": 35.72357940673828, "learning_rate": 4.446957236842105e-06, "loss": 0.1494, "step": 2701 }, { "epoch": 0.5555098684210527, "grad_norm": 50.251991271972656, "learning_rate": 4.444901315789474e-06, "loss": 0.248, "step": 2702 }, { "epoch": 0.5557154605263158, "grad_norm": 16.8245906829834, "learning_rate": 4.442845394736843e-06, "loss": 0.0454, "step": 2703 }, { "epoch": 0.555921052631579, "grad_norm": 24.887706756591797, "learning_rate": 4.440789473684211e-06, "loss": 0.084, "step": 2704 }, { "epoch": 0.5561266447368421, "grad_norm": 49.74919509887695, "learning_rate": 4.438733552631579e-06, "loss": 0.2339, "step": 2705 }, { "epoch": 0.5563322368421053, "grad_norm": 125.84746551513672, "learning_rate": 4.436677631578948e-06, "loss": 1.5947, "step": 2706 }, { "epoch": 0.5565378289473685, "grad_norm": 36.671817779541016, "learning_rate": 4.434621710526316e-06, "loss": 0.1626, "step": 2707 }, { "epoch": 0.5567434210526315, "grad_norm": 149.85043334960938, "learning_rate": 4.4325657894736846e-06, "loss": 1.8608, "step": 2708 }, { "epoch": 0.5569490131578947, "grad_norm": 122.0292739868164, "learning_rate": 4.430509868421053e-06, "loss": 1.0938, "step": 2709 }, { "epoch": 0.5571546052631579, "grad_norm": 49.988624572753906, "learning_rate": 4.428453947368421e-06, "loss": 0.1201, "step": 2710 }, { "epoch": 0.557360197368421, "grad_norm": 117.28536987304688, "learning_rate": 4.42639802631579e-06, "loss": 0.7944, "step": 2711 }, { "epoch": 0.5575657894736842, "grad_norm": 241.70684814453125, "learning_rate": 4.424342105263158e-06, "loss": 2.0234, "step": 2712 }, { "epoch": 0.5577713815789473, "grad_norm": 236.93154907226562, "learning_rate": 4.4222861842105265e-06, "loss": 3.3125, "step": 2713 }, { "epoch": 0.5579769736842105, "grad_norm": 20.91286849975586, "learning_rate": 4.420230263157895e-06, "loss": 0.0756, "step": 2714 }, { "epoch": 0.5581825657894737, "grad_norm": 121.6884994506836, "learning_rate": 4.418174342105264e-06, "loss": 0.9763, "step": 2715 }, { "epoch": 0.5583881578947368, "grad_norm": 33.16695785522461, "learning_rate": 4.416118421052632e-06, "loss": 0.1353, "step": 2716 }, { "epoch": 0.55859375, "grad_norm": 31.354463577270508, "learning_rate": 4.4140625000000006e-06, "loss": 0.137, "step": 2717 }, { "epoch": 0.5587993421052632, "grad_norm": 93.02320098876953, "learning_rate": 4.4120065789473684e-06, "loss": 0.7642, "step": 2718 }, { "epoch": 0.5590049342105263, "grad_norm": 77.95773315429688, "learning_rate": 4.409950657894737e-06, "loss": 0.5547, "step": 2719 }, { "epoch": 0.5592105263157895, "grad_norm": 4.327332496643066, "learning_rate": 4.407894736842105e-06, "loss": 0.0108, "step": 2720 }, { "epoch": 0.5594161184210527, "grad_norm": 285.7721862792969, "learning_rate": 4.405838815789474e-06, "loss": 1.3945, "step": 2721 }, { "epoch": 0.5596217105263158, "grad_norm": 79.60211944580078, "learning_rate": 4.4037828947368425e-06, "loss": 0.9247, "step": 2722 }, { "epoch": 0.559827302631579, "grad_norm": 19.76434898376465, "learning_rate": 4.401726973684211e-06, "loss": 0.0861, "step": 2723 }, { "epoch": 0.5600328947368421, "grad_norm": 141.31329345703125, "learning_rate": 4.399671052631579e-06, "loss": 1.4634, "step": 2724 }, { "epoch": 0.5602384868421053, "grad_norm": 8.661529541015625, "learning_rate": 4.397615131578948e-06, "loss": 0.0158, "step": 2725 }, { "epoch": 0.5604440789473685, "grad_norm": 95.23927307128906, "learning_rate": 4.3955592105263166e-06, "loss": 0.9648, "step": 2726 }, { "epoch": 0.5606496710526315, "grad_norm": 69.27098846435547, "learning_rate": 4.3935032894736845e-06, "loss": 0.7246, "step": 2727 }, { "epoch": 0.5608552631578947, "grad_norm": 197.3812713623047, "learning_rate": 4.391447368421053e-06, "loss": 1.5782, "step": 2728 }, { "epoch": 0.5610608552631579, "grad_norm": 100.77295684814453, "learning_rate": 4.389391447368421e-06, "loss": 0.9658, "step": 2729 }, { "epoch": 0.561266447368421, "grad_norm": 257.58526611328125, "learning_rate": 4.38733552631579e-06, "loss": 1.7191, "step": 2730 }, { "epoch": 0.5614720394736842, "grad_norm": 49.661678314208984, "learning_rate": 4.385279605263158e-06, "loss": 0.3145, "step": 2731 }, { "epoch": 0.5616776315789473, "grad_norm": 25.084274291992188, "learning_rate": 4.383223684210527e-06, "loss": 0.0789, "step": 2732 }, { "epoch": 0.5618832236842105, "grad_norm": 15.972902297973633, "learning_rate": 4.381167763157895e-06, "loss": 0.0466, "step": 2733 }, { "epoch": 0.5620888157894737, "grad_norm": 46.127830505371094, "learning_rate": 4.379111842105264e-06, "loss": 0.4346, "step": 2734 }, { "epoch": 0.5622944078947368, "grad_norm": 47.916648864746094, "learning_rate": 4.377055921052632e-06, "loss": 0.4231, "step": 2735 }, { "epoch": 0.5625, "grad_norm": 80.90521240234375, "learning_rate": 4.3750000000000005e-06, "loss": 1.3281, "step": 2736 }, { "epoch": 0.5627055921052632, "grad_norm": 65.42305755615234, "learning_rate": 4.372944078947368e-06, "loss": 0.7051, "step": 2737 }, { "epoch": 0.5629111842105263, "grad_norm": 83.08053588867188, "learning_rate": 4.370888157894737e-06, "loss": 0.2654, "step": 2738 }, { "epoch": 0.5631167763157895, "grad_norm": 21.157258987426758, "learning_rate": 4.368832236842106e-06, "loss": 0.0948, "step": 2739 }, { "epoch": 0.5633223684210527, "grad_norm": 16.655014038085938, "learning_rate": 4.366776315789474e-06, "loss": 0.0432, "step": 2740 }, { "epoch": 0.5635279605263158, "grad_norm": 52.747596740722656, "learning_rate": 4.364720394736842e-06, "loss": 0.236, "step": 2741 }, { "epoch": 0.563733552631579, "grad_norm": 68.42676544189453, "learning_rate": 4.362664473684211e-06, "loss": 0.6278, "step": 2742 }, { "epoch": 0.5639391447368421, "grad_norm": 103.55390930175781, "learning_rate": 4.36060855263158e-06, "loss": 0.5923, "step": 2743 }, { "epoch": 0.5641447368421053, "grad_norm": 68.04568481445312, "learning_rate": 4.358552631578948e-06, "loss": 0.5127, "step": 2744 }, { "epoch": 0.5643503289473685, "grad_norm": 96.2052230834961, "learning_rate": 4.3564967105263165e-06, "loss": 0.7749, "step": 2745 }, { "epoch": 0.5645559210526315, "grad_norm": 44.11578369140625, "learning_rate": 4.354440789473684e-06, "loss": 0.3784, "step": 2746 }, { "epoch": 0.5647615131578947, "grad_norm": 111.66246032714844, "learning_rate": 4.352384868421053e-06, "loss": 0.6152, "step": 2747 }, { "epoch": 0.5649671052631579, "grad_norm": 68.40763854980469, "learning_rate": 4.350328947368421e-06, "loss": 0.6387, "step": 2748 }, { "epoch": 0.565172697368421, "grad_norm": 35.87992858886719, "learning_rate": 4.34827302631579e-06, "loss": 0.1799, "step": 2749 }, { "epoch": 0.5653782894736842, "grad_norm": 69.8268051147461, "learning_rate": 4.3462171052631576e-06, "loss": 0.179, "step": 2750 }, { "epoch": 0.5655838815789473, "grad_norm": 127.11815643310547, "learning_rate": 4.344161184210527e-06, "loss": 1.207, "step": 2751 }, { "epoch": 0.5657894736842105, "grad_norm": 51.13294219970703, "learning_rate": 4.342105263157895e-06, "loss": 0.332, "step": 2752 }, { "epoch": 0.5659950657894737, "grad_norm": 233.89508056640625, "learning_rate": 4.340049342105264e-06, "loss": 5.1406, "step": 2753 }, { "epoch": 0.5662006578947368, "grad_norm": 8.17660140991211, "learning_rate": 4.337993421052632e-06, "loss": 0.0179, "step": 2754 }, { "epoch": 0.56640625, "grad_norm": 17.871286392211914, "learning_rate": 4.3359375e-06, "loss": 0.053, "step": 2755 }, { "epoch": 0.5666118421052632, "grad_norm": 23.299301147460938, "learning_rate": 4.333881578947369e-06, "loss": 0.0684, "step": 2756 }, { "epoch": 0.5668174342105263, "grad_norm": 170.5945587158203, "learning_rate": 4.331825657894737e-06, "loss": 0.8555, "step": 2757 }, { "epoch": 0.5670230263157895, "grad_norm": 65.7887191772461, "learning_rate": 4.329769736842106e-06, "loss": 0.5059, "step": 2758 }, { "epoch": 0.5672286184210527, "grad_norm": 72.46701049804688, "learning_rate": 4.327713815789474e-06, "loss": 0.5168, "step": 2759 }, { "epoch": 0.5674342105263158, "grad_norm": 112.4258804321289, "learning_rate": 4.325657894736842e-06, "loss": 0.9224, "step": 2760 }, { "epoch": 0.567639802631579, "grad_norm": 73.72272491455078, "learning_rate": 4.323601973684211e-06, "loss": 0.6709, "step": 2761 }, { "epoch": 0.5678453947368421, "grad_norm": 27.65538215637207, "learning_rate": 4.32154605263158e-06, "loss": 0.0728, "step": 2762 }, { "epoch": 0.5680509868421053, "grad_norm": 124.26715850830078, "learning_rate": 4.319490131578948e-06, "loss": 1.2988, "step": 2763 }, { "epoch": 0.5682565789473685, "grad_norm": 43.70951843261719, "learning_rate": 4.317434210526316e-06, "loss": 0.2019, "step": 2764 }, { "epoch": 0.5684621710526315, "grad_norm": 210.58128356933594, "learning_rate": 4.315378289473684e-06, "loss": 1.5156, "step": 2765 }, { "epoch": 0.5686677631578947, "grad_norm": 122.78688049316406, "learning_rate": 4.313322368421053e-06, "loss": 0.8857, "step": 2766 }, { "epoch": 0.5688733552631579, "grad_norm": 104.1529769897461, "learning_rate": 4.311266447368421e-06, "loss": 0.6511, "step": 2767 }, { "epoch": 0.569078947368421, "grad_norm": 148.3342742919922, "learning_rate": 4.30921052631579e-06, "loss": 1.9595, "step": 2768 }, { "epoch": 0.5692845394736842, "grad_norm": 30.386573791503906, "learning_rate": 4.307154605263158e-06, "loss": 0.1658, "step": 2769 }, { "epoch": 0.5694901315789473, "grad_norm": 25.265195846557617, "learning_rate": 4.305098684210527e-06, "loss": 0.1775, "step": 2770 }, { "epoch": 0.5696957236842105, "grad_norm": 39.02897644042969, "learning_rate": 4.303042763157895e-06, "loss": 0.2715, "step": 2771 }, { "epoch": 0.5699013157894737, "grad_norm": 73.61375427246094, "learning_rate": 4.300986842105264e-06, "loss": 0.4292, "step": 2772 }, { "epoch": 0.5701069078947368, "grad_norm": 124.55946350097656, "learning_rate": 4.298930921052632e-06, "loss": 0.5846, "step": 2773 }, { "epoch": 0.5703125, "grad_norm": 184.02439880371094, "learning_rate": 4.296875e-06, "loss": 1.5325, "step": 2774 }, { "epoch": 0.5705180921052632, "grad_norm": 58.998878479003906, "learning_rate": 4.294819078947369e-06, "loss": 0.4029, "step": 2775 }, { "epoch": 0.5707236842105263, "grad_norm": 128.88482666015625, "learning_rate": 4.292763157894737e-06, "loss": 1.4595, "step": 2776 }, { "epoch": 0.5709292763157895, "grad_norm": 6.547189235687256, "learning_rate": 4.290707236842106e-06, "loss": 0.0374, "step": 2777 }, { "epoch": 0.5711348684210527, "grad_norm": 82.46482849121094, "learning_rate": 4.2886513157894735e-06, "loss": 0.3979, "step": 2778 }, { "epoch": 0.5713404605263158, "grad_norm": 51.881500244140625, "learning_rate": 4.286595394736843e-06, "loss": 0.2997, "step": 2779 }, { "epoch": 0.571546052631579, "grad_norm": 42.46624755859375, "learning_rate": 4.284539473684211e-06, "loss": 0.2825, "step": 2780 }, { "epoch": 0.5717516447368421, "grad_norm": 67.74058532714844, "learning_rate": 4.28248355263158e-06, "loss": 0.5507, "step": 2781 }, { "epoch": 0.5719572368421053, "grad_norm": 53.088661193847656, "learning_rate": 4.2804276315789475e-06, "loss": 0.7632, "step": 2782 }, { "epoch": 0.5721628289473685, "grad_norm": 14.167082786560059, "learning_rate": 4.278371710526316e-06, "loss": 0.0651, "step": 2783 }, { "epoch": 0.5723684210526315, "grad_norm": 99.54460144042969, "learning_rate": 4.276315789473684e-06, "loss": 0.9082, "step": 2784 }, { "epoch": 0.5725740131578947, "grad_norm": 65.67488861083984, "learning_rate": 4.274259868421053e-06, "loss": 0.3953, "step": 2785 }, { "epoch": 0.5727796052631579, "grad_norm": 26.701501846313477, "learning_rate": 4.272203947368422e-06, "loss": 0.0955, "step": 2786 }, { "epoch": 0.572985197368421, "grad_norm": 4.945610523223877, "learning_rate": 4.2701480263157895e-06, "loss": 0.0255, "step": 2787 }, { "epoch": 0.5731907894736842, "grad_norm": 21.781038284301758, "learning_rate": 4.268092105263158e-06, "loss": 0.0911, "step": 2788 }, { "epoch": 0.5733963815789473, "grad_norm": 3.349013328552246, "learning_rate": 4.266036184210527e-06, "loss": 0.0091, "step": 2789 }, { "epoch": 0.5736019736842105, "grad_norm": 85.16702270507812, "learning_rate": 4.263980263157896e-06, "loss": 0.3777, "step": 2790 }, { "epoch": 0.5738075657894737, "grad_norm": 12.389396667480469, "learning_rate": 4.2619243421052635e-06, "loss": 0.0433, "step": 2791 }, { "epoch": 0.5740131578947368, "grad_norm": 25.869338989257812, "learning_rate": 4.259868421052632e-06, "loss": 0.109, "step": 2792 }, { "epoch": 0.57421875, "grad_norm": 42.43351364135742, "learning_rate": 4.2578125e-06, "loss": 0.1829, "step": 2793 }, { "epoch": 0.5744243421052632, "grad_norm": 107.66553497314453, "learning_rate": 4.255756578947369e-06, "loss": 0.7168, "step": 2794 }, { "epoch": 0.5746299342105263, "grad_norm": 32.066062927246094, "learning_rate": 4.253700657894737e-06, "loss": 0.269, "step": 2795 }, { "epoch": 0.5748355263157895, "grad_norm": 21.0797119140625, "learning_rate": 4.2516447368421055e-06, "loss": 0.1322, "step": 2796 }, { "epoch": 0.5750411184210527, "grad_norm": 116.98875427246094, "learning_rate": 4.249588815789474e-06, "loss": 0.8389, "step": 2797 }, { "epoch": 0.5752467105263158, "grad_norm": 92.56160736083984, "learning_rate": 4.247532894736843e-06, "loss": 0.6543, "step": 2798 }, { "epoch": 0.575452302631579, "grad_norm": 38.756324768066406, "learning_rate": 4.245476973684211e-06, "loss": 0.2876, "step": 2799 }, { "epoch": 0.5756578947368421, "grad_norm": 63.9505729675293, "learning_rate": 4.2434210526315796e-06, "loss": 0.5527, "step": 2800 }, { "epoch": 0.5758634868421053, "grad_norm": 9.893265724182129, "learning_rate": 4.2413651315789474e-06, "loss": 0.0402, "step": 2801 }, { "epoch": 0.5760690789473685, "grad_norm": 53.67276382446289, "learning_rate": 4.239309210526316e-06, "loss": 0.3082, "step": 2802 }, { "epoch": 0.5762746710526315, "grad_norm": 2.3762123584747314, "learning_rate": 4.237253289473685e-06, "loss": 0.0096, "step": 2803 }, { "epoch": 0.5764802631578947, "grad_norm": 39.123878479003906, "learning_rate": 4.235197368421053e-06, "loss": 0.2837, "step": 2804 }, { "epoch": 0.5766858552631579, "grad_norm": 38.92578887939453, "learning_rate": 4.2331414473684215e-06, "loss": 0.2749, "step": 2805 }, { "epoch": 0.576891447368421, "grad_norm": 7.114015102386475, "learning_rate": 4.231085526315789e-06, "loss": 0.0162, "step": 2806 }, { "epoch": 0.5770970394736842, "grad_norm": 106.4642562866211, "learning_rate": 4.229029605263158e-06, "loss": 0.7944, "step": 2807 }, { "epoch": 0.5773026315789473, "grad_norm": 71.68002319335938, "learning_rate": 4.226973684210527e-06, "loss": 0.4423, "step": 2808 }, { "epoch": 0.5775082236842105, "grad_norm": 143.30130004882812, "learning_rate": 4.2249177631578956e-06, "loss": 0.9586, "step": 2809 }, { "epoch": 0.5777138157894737, "grad_norm": 72.24253845214844, "learning_rate": 4.2228618421052634e-06, "loss": 0.9746, "step": 2810 }, { "epoch": 0.5779194078947368, "grad_norm": 19.58362579345703, "learning_rate": 4.220805921052632e-06, "loss": 0.0767, "step": 2811 }, { "epoch": 0.578125, "grad_norm": 201.89117431640625, "learning_rate": 4.21875e-06, "loss": 2.2266, "step": 2812 }, { "epoch": 0.5783305921052632, "grad_norm": 139.77154541015625, "learning_rate": 4.216694078947369e-06, "loss": 0.9033, "step": 2813 }, { "epoch": 0.5785361842105263, "grad_norm": 88.25430297851562, "learning_rate": 4.214638157894737e-06, "loss": 1.2559, "step": 2814 }, { "epoch": 0.5787417763157895, "grad_norm": 52.31370544433594, "learning_rate": 4.212582236842105e-06, "loss": 0.2922, "step": 2815 }, { "epoch": 0.5789473684210527, "grad_norm": 26.358150482177734, "learning_rate": 4.210526315789474e-06, "loss": 0.1284, "step": 2816 }, { "epoch": 0.5791529605263158, "grad_norm": 69.80047607421875, "learning_rate": 4.208470394736843e-06, "loss": 0.3413, "step": 2817 }, { "epoch": 0.579358552631579, "grad_norm": 75.73021697998047, "learning_rate": 4.206414473684211e-06, "loss": 0.6523, "step": 2818 }, { "epoch": 0.5795641447368421, "grad_norm": 10.736067771911621, "learning_rate": 4.2043585526315795e-06, "loss": 0.0555, "step": 2819 }, { "epoch": 0.5797697368421053, "grad_norm": 112.51097869873047, "learning_rate": 4.202302631578948e-06, "loss": 0.8994, "step": 2820 }, { "epoch": 0.5799753289473685, "grad_norm": 48.393104553222656, "learning_rate": 4.200246710526316e-06, "loss": 0.2375, "step": 2821 }, { "epoch": 0.5801809210526315, "grad_norm": 100.97250366210938, "learning_rate": 4.198190789473685e-06, "loss": 0.8691, "step": 2822 }, { "epoch": 0.5803865131578947, "grad_norm": 11.87575912475586, "learning_rate": 4.196134868421053e-06, "loss": 0.0212, "step": 2823 }, { "epoch": 0.5805921052631579, "grad_norm": 26.949445724487305, "learning_rate": 4.194078947368421e-06, "loss": 0.1421, "step": 2824 }, { "epoch": 0.580797697368421, "grad_norm": 45.144256591796875, "learning_rate": 4.192023026315789e-06, "loss": 0.4137, "step": 2825 }, { "epoch": 0.5810032894736842, "grad_norm": 85.61089324951172, "learning_rate": 4.189967105263158e-06, "loss": 0.4707, "step": 2826 }, { "epoch": 0.5812088815789473, "grad_norm": 129.4268035888672, "learning_rate": 4.187911184210527e-06, "loss": 1.1895, "step": 2827 }, { "epoch": 0.5814144736842105, "grad_norm": 61.05864715576172, "learning_rate": 4.1858552631578955e-06, "loss": 0.4199, "step": 2828 }, { "epoch": 0.5816200657894737, "grad_norm": 59.24404525756836, "learning_rate": 4.183799342105263e-06, "loss": 0.3482, "step": 2829 }, { "epoch": 0.5818256578947368, "grad_norm": 182.6363983154297, "learning_rate": 4.181743421052632e-06, "loss": 0.9424, "step": 2830 }, { "epoch": 0.58203125, "grad_norm": 171.6084747314453, "learning_rate": 4.1796875e-06, "loss": 1.1758, "step": 2831 }, { "epoch": 0.5822368421052632, "grad_norm": 46.322540283203125, "learning_rate": 4.177631578947369e-06, "loss": 0.2237, "step": 2832 }, { "epoch": 0.5824424342105263, "grad_norm": 201.1350555419922, "learning_rate": 4.175575657894737e-06, "loss": 0.8567, "step": 2833 }, { "epoch": 0.5826480263157895, "grad_norm": 209.67359924316406, "learning_rate": 4.173519736842105e-06, "loss": 1.7703, "step": 2834 }, { "epoch": 0.5828536184210527, "grad_norm": 32.55963897705078, "learning_rate": 4.171463815789474e-06, "loss": 0.1113, "step": 2835 }, { "epoch": 0.5830592105263158, "grad_norm": 129.57720947265625, "learning_rate": 4.169407894736843e-06, "loss": 1.2587, "step": 2836 }, { "epoch": 0.583264802631579, "grad_norm": 13.087209701538086, "learning_rate": 4.1673519736842115e-06, "loss": 0.0352, "step": 2837 }, { "epoch": 0.5834703947368421, "grad_norm": 71.87480926513672, "learning_rate": 4.165296052631579e-06, "loss": 0.8472, "step": 2838 }, { "epoch": 0.5836759868421053, "grad_norm": 53.57924270629883, "learning_rate": 4.163240131578948e-06, "loss": 0.4384, "step": 2839 }, { "epoch": 0.5838815789473685, "grad_norm": 38.236324310302734, "learning_rate": 4.161184210526316e-06, "loss": 0.3574, "step": 2840 }, { "epoch": 0.5840871710526315, "grad_norm": 170.32594299316406, "learning_rate": 4.159128289473685e-06, "loss": 2.5371, "step": 2841 }, { "epoch": 0.5842927631578947, "grad_norm": 53.66725158691406, "learning_rate": 4.1570723684210526e-06, "loss": 0.248, "step": 2842 }, { "epoch": 0.5844983552631579, "grad_norm": 262.5569152832031, "learning_rate": 4.155016447368421e-06, "loss": 0.9435, "step": 2843 }, { "epoch": 0.584703947368421, "grad_norm": 25.49895477294922, "learning_rate": 4.15296052631579e-06, "loss": 0.1084, "step": 2844 }, { "epoch": 0.5849095394736842, "grad_norm": 15.197224617004395, "learning_rate": 4.150904605263158e-06, "loss": 0.0397, "step": 2845 }, { "epoch": 0.5851151315789473, "grad_norm": 207.52467346191406, "learning_rate": 4.148848684210527e-06, "loss": 1.0859, "step": 2846 }, { "epoch": 0.5853207236842105, "grad_norm": 64.08685302734375, "learning_rate": 4.146792763157895e-06, "loss": 0.5117, "step": 2847 }, { "epoch": 0.5855263157894737, "grad_norm": 120.55329132080078, "learning_rate": 4.144736842105263e-06, "loss": 1.1289, "step": 2848 }, { "epoch": 0.5857319078947368, "grad_norm": 120.426513671875, "learning_rate": 4.142680921052632e-06, "loss": 0.9102, "step": 2849 }, { "epoch": 0.5859375, "grad_norm": 216.6763916015625, "learning_rate": 4.140625000000001e-06, "loss": 4.8931, "step": 2850 }, { "epoch": 0.5861430921052632, "grad_norm": 57.13277816772461, "learning_rate": 4.138569078947369e-06, "loss": 0.2817, "step": 2851 }, { "epoch": 0.5863486842105263, "grad_norm": 48.03614044189453, "learning_rate": 4.136513157894737e-06, "loss": 0.6113, "step": 2852 }, { "epoch": 0.5865542763157895, "grad_norm": 116.650634765625, "learning_rate": 4.134457236842105e-06, "loss": 0.499, "step": 2853 }, { "epoch": 0.5867598684210527, "grad_norm": 41.174903869628906, "learning_rate": 4.132401315789474e-06, "loss": 0.2004, "step": 2854 }, { "epoch": 0.5869654605263158, "grad_norm": 72.22600555419922, "learning_rate": 4.130345394736843e-06, "loss": 0.4492, "step": 2855 }, { "epoch": 0.587171052631579, "grad_norm": 60.389766693115234, "learning_rate": 4.128289473684211e-06, "loss": 0.4551, "step": 2856 }, { "epoch": 0.5873766447368421, "grad_norm": 13.2418794631958, "learning_rate": 4.126233552631579e-06, "loss": 0.1414, "step": 2857 }, { "epoch": 0.5875822368421053, "grad_norm": 92.44645690917969, "learning_rate": 4.124177631578948e-06, "loss": 0.5195, "step": 2858 }, { "epoch": 0.5877878289473685, "grad_norm": 20.594921112060547, "learning_rate": 4.122121710526316e-06, "loss": 0.0547, "step": 2859 }, { "epoch": 0.5879934210526315, "grad_norm": 79.06034088134766, "learning_rate": 4.120065789473685e-06, "loss": 0.6592, "step": 2860 }, { "epoch": 0.5881990131578947, "grad_norm": 127.6043472290039, "learning_rate": 4.1180098684210525e-06, "loss": 0.458, "step": 2861 }, { "epoch": 0.5884046052631579, "grad_norm": 49.56166076660156, "learning_rate": 4.115953947368421e-06, "loss": 0.2429, "step": 2862 }, { "epoch": 0.588610197368421, "grad_norm": 47.4669303894043, "learning_rate": 4.11389802631579e-06, "loss": 0.6425, "step": 2863 }, { "epoch": 0.5888157894736842, "grad_norm": 18.772640228271484, "learning_rate": 4.111842105263158e-06, "loss": 0.1091, "step": 2864 }, { "epoch": 0.5890213815789473, "grad_norm": 156.7685546875, "learning_rate": 4.1097861842105265e-06, "loss": 1.3672, "step": 2865 }, { "epoch": 0.5892269736842105, "grad_norm": 135.0237274169922, "learning_rate": 4.107730263157895e-06, "loss": 1.25, "step": 2866 }, { "epoch": 0.5894325657894737, "grad_norm": 7.376319408416748, "learning_rate": 4.105674342105264e-06, "loss": 0.0205, "step": 2867 }, { "epoch": 0.5896381578947368, "grad_norm": 103.17843627929688, "learning_rate": 4.103618421052632e-06, "loss": 0.8711, "step": 2868 }, { "epoch": 0.58984375, "grad_norm": 46.26371765136719, "learning_rate": 4.101562500000001e-06, "loss": 0.2603, "step": 2869 }, { "epoch": 0.5900493421052632, "grad_norm": 9.074191093444824, "learning_rate": 4.0995065789473685e-06, "loss": 0.0392, "step": 2870 }, { "epoch": 0.5902549342105263, "grad_norm": 140.7956085205078, "learning_rate": 4.097450657894737e-06, "loss": 0.7031, "step": 2871 }, { "epoch": 0.5904605263157895, "grad_norm": 28.07443618774414, "learning_rate": 4.095394736842105e-06, "loss": 0.1043, "step": 2872 }, { "epoch": 0.5906661184210527, "grad_norm": 99.35338592529297, "learning_rate": 4.093338815789474e-06, "loss": 0.8338, "step": 2873 }, { "epoch": 0.5908717105263158, "grad_norm": 131.19456481933594, "learning_rate": 4.0912828947368425e-06, "loss": 0.8477, "step": 2874 }, { "epoch": 0.591077302631579, "grad_norm": 32.440040588378906, "learning_rate": 4.089226973684211e-06, "loss": 0.1289, "step": 2875 }, { "epoch": 0.5912828947368421, "grad_norm": 14.89501667022705, "learning_rate": 4.087171052631579e-06, "loss": 0.0449, "step": 2876 }, { "epoch": 0.5914884868421053, "grad_norm": 29.309009552001953, "learning_rate": 4.085115131578948e-06, "loss": 0.1843, "step": 2877 }, { "epoch": 0.5916940789473685, "grad_norm": 67.90098571777344, "learning_rate": 4.083059210526316e-06, "loss": 0.3086, "step": 2878 }, { "epoch": 0.5918996710526315, "grad_norm": 347.5747985839844, "learning_rate": 4.0810032894736845e-06, "loss": 1.1416, "step": 2879 }, { "epoch": 0.5921052631578947, "grad_norm": 6.873557090759277, "learning_rate": 4.078947368421053e-06, "loss": 0.0141, "step": 2880 }, { "epoch": 0.5923108552631579, "grad_norm": 28.233097076416016, "learning_rate": 4.076891447368421e-06, "loss": 0.0877, "step": 2881 }, { "epoch": 0.592516447368421, "grad_norm": 141.81524658203125, "learning_rate": 4.07483552631579e-06, "loss": 0.9531, "step": 2882 }, { "epoch": 0.5927220394736842, "grad_norm": 83.35848236083984, "learning_rate": 4.072779605263158e-06, "loss": 0.4985, "step": 2883 }, { "epoch": 0.5929276315789473, "grad_norm": 184.9790496826172, "learning_rate": 4.070723684210527e-06, "loss": 1.2734, "step": 2884 }, { "epoch": 0.5931332236842105, "grad_norm": 179.2019805908203, "learning_rate": 4.068667763157895e-06, "loss": 0.5044, "step": 2885 }, { "epoch": 0.5933388157894737, "grad_norm": 209.57765197753906, "learning_rate": 4.066611842105264e-06, "loss": 0.8688, "step": 2886 }, { "epoch": 0.5935444078947368, "grad_norm": 104.82598114013672, "learning_rate": 4.064555921052632e-06, "loss": 1.4658, "step": 2887 }, { "epoch": 0.59375, "grad_norm": 339.33221435546875, "learning_rate": 4.0625000000000005e-06, "loss": 1.3418, "step": 2888 }, { "epoch": 0.5939555921052632, "grad_norm": 25.761659622192383, "learning_rate": 4.060444078947368e-06, "loss": 0.1124, "step": 2889 }, { "epoch": 0.5941611842105263, "grad_norm": 445.25958251953125, "learning_rate": 4.058388157894737e-06, "loss": 1.8512, "step": 2890 }, { "epoch": 0.5943667763157895, "grad_norm": 39.95951843261719, "learning_rate": 4.056332236842106e-06, "loss": 0.2351, "step": 2891 }, { "epoch": 0.5945723684210527, "grad_norm": 43.80483627319336, "learning_rate": 4.054276315789474e-06, "loss": 0.1785, "step": 2892 }, { "epoch": 0.5947779605263158, "grad_norm": 104.96612548828125, "learning_rate": 4.0522203947368424e-06, "loss": 0.4827, "step": 2893 }, { "epoch": 0.594983552631579, "grad_norm": 184.87445068359375, "learning_rate": 4.050164473684211e-06, "loss": 2.7588, "step": 2894 }, { "epoch": 0.5951891447368421, "grad_norm": 5.93311882019043, "learning_rate": 4.048108552631579e-06, "loss": 0.0259, "step": 2895 }, { "epoch": 0.5953947368421053, "grad_norm": 75.1473617553711, "learning_rate": 4.046052631578948e-06, "loss": 0.5781, "step": 2896 }, { "epoch": 0.5956003289473685, "grad_norm": 90.8294906616211, "learning_rate": 4.0439967105263165e-06, "loss": 1.3574, "step": 2897 }, { "epoch": 0.5958059210526315, "grad_norm": 15.337647438049316, "learning_rate": 4.041940789473684e-06, "loss": 0.0423, "step": 2898 }, { "epoch": 0.5960115131578947, "grad_norm": 133.85731506347656, "learning_rate": 4.039884868421053e-06, "loss": 1.6108, "step": 2899 }, { "epoch": 0.5962171052631579, "grad_norm": 22.446016311645508, "learning_rate": 4.037828947368421e-06, "loss": 0.0952, "step": 2900 }, { "epoch": 0.596422697368421, "grad_norm": 70.18352508544922, "learning_rate": 4.03577302631579e-06, "loss": 0.9083, "step": 2901 }, { "epoch": 0.5966282894736842, "grad_norm": 69.46656036376953, "learning_rate": 4.033717105263158e-06, "loss": 0.6279, "step": 2902 }, { "epoch": 0.5968338815789473, "grad_norm": 66.83455657958984, "learning_rate": 4.031661184210527e-06, "loss": 0.543, "step": 2903 }, { "epoch": 0.5970394736842105, "grad_norm": 39.249610900878906, "learning_rate": 4.029605263157895e-06, "loss": 0.2534, "step": 2904 }, { "epoch": 0.5972450657894737, "grad_norm": 60.42684555053711, "learning_rate": 4.027549342105264e-06, "loss": 0.4082, "step": 2905 }, { "epoch": 0.5974506578947368, "grad_norm": 31.018754959106445, "learning_rate": 4.025493421052632e-06, "loss": 0.1552, "step": 2906 }, { "epoch": 0.59765625, "grad_norm": 293.05133056640625, "learning_rate": 4.0234375e-06, "loss": 1.231, "step": 2907 }, { "epoch": 0.5978618421052632, "grad_norm": 17.670955657958984, "learning_rate": 4.021381578947368e-06, "loss": 0.0996, "step": 2908 }, { "epoch": 0.5980674342105263, "grad_norm": 38.85425567626953, "learning_rate": 4.019325657894737e-06, "loss": 0.3267, "step": 2909 }, { "epoch": 0.5982730263157895, "grad_norm": 78.5801773071289, "learning_rate": 4.017269736842106e-06, "loss": 0.5293, "step": 2910 }, { "epoch": 0.5984786184210527, "grad_norm": 53.42984390258789, "learning_rate": 4.015213815789474e-06, "loss": 0.4448, "step": 2911 }, { "epoch": 0.5986842105263158, "grad_norm": 51.28424835205078, "learning_rate": 4.013157894736842e-06, "loss": 0.3384, "step": 2912 }, { "epoch": 0.598889802631579, "grad_norm": 43.42974853515625, "learning_rate": 4.011101973684211e-06, "loss": 0.7812, "step": 2913 }, { "epoch": 0.5990953947368421, "grad_norm": 26.587881088256836, "learning_rate": 4.00904605263158e-06, "loss": 0.2407, "step": 2914 }, { "epoch": 0.5993009868421053, "grad_norm": 113.32221221923828, "learning_rate": 4.006990131578948e-06, "loss": 1.0693, "step": 2915 }, { "epoch": 0.5995065789473685, "grad_norm": 47.83821487426758, "learning_rate": 4.004934210526316e-06, "loss": 0.373, "step": 2916 }, { "epoch": 0.5997121710526315, "grad_norm": 97.62679290771484, "learning_rate": 4.002878289473684e-06, "loss": 0.7432, "step": 2917 }, { "epoch": 0.5999177631578947, "grad_norm": 85.13323211669922, "learning_rate": 4.000822368421053e-06, "loss": 1.2323, "step": 2918 }, { "epoch": 0.6001233552631579, "grad_norm": 23.263538360595703, "learning_rate": 3.998766447368421e-06, "loss": 0.0842, "step": 2919 }, { "epoch": 0.600328947368421, "grad_norm": 22.549163818359375, "learning_rate": 3.99671052631579e-06, "loss": 0.0706, "step": 2920 }, { "epoch": 0.6005345394736842, "grad_norm": 7.245571136474609, "learning_rate": 3.994654605263158e-06, "loss": 0.0166, "step": 2921 }, { "epoch": 0.6007401315789473, "grad_norm": 62.116641998291016, "learning_rate": 3.992598684210527e-06, "loss": 0.6138, "step": 2922 }, { "epoch": 0.6009457236842105, "grad_norm": 100.08612823486328, "learning_rate": 3.990542763157895e-06, "loss": 0.5051, "step": 2923 }, { "epoch": 0.6011513157894737, "grad_norm": 125.55828094482422, "learning_rate": 3.988486842105264e-06, "loss": 1.8601, "step": 2924 }, { "epoch": 0.6013569078947368, "grad_norm": 70.15196990966797, "learning_rate": 3.9864309210526316e-06, "loss": 0.2285, "step": 2925 }, { "epoch": 0.6015625, "grad_norm": 186.0143585205078, "learning_rate": 3.984375e-06, "loss": 1.4399, "step": 2926 }, { "epoch": 0.6017680921052632, "grad_norm": 109.79124450683594, "learning_rate": 3.982319078947369e-06, "loss": 0.9097, "step": 2927 }, { "epoch": 0.6019736842105263, "grad_norm": 66.0390853881836, "learning_rate": 3.980263157894737e-06, "loss": 1.0957, "step": 2928 }, { "epoch": 0.6021792763157895, "grad_norm": 39.72788619995117, "learning_rate": 3.978207236842106e-06, "loss": 0.2866, "step": 2929 }, { "epoch": 0.6023848684210527, "grad_norm": 65.73970794677734, "learning_rate": 3.9761513157894735e-06, "loss": 0.7109, "step": 2930 }, { "epoch": 0.6025904605263158, "grad_norm": 4.820986747741699, "learning_rate": 3.974095394736843e-06, "loss": 0.0172, "step": 2931 }, { "epoch": 0.602796052631579, "grad_norm": 25.651464462280273, "learning_rate": 3.972039473684211e-06, "loss": 0.072, "step": 2932 }, { "epoch": 0.6030016447368421, "grad_norm": 66.51493835449219, "learning_rate": 3.96998355263158e-06, "loss": 0.8574, "step": 2933 }, { "epoch": 0.6032072368421053, "grad_norm": 49.10223388671875, "learning_rate": 3.9679276315789476e-06, "loss": 0.2513, "step": 2934 }, { "epoch": 0.6034128289473685, "grad_norm": 3.329286575317383, "learning_rate": 3.965871710526316e-06, "loss": 0.0142, "step": 2935 }, { "epoch": 0.6036184210526315, "grad_norm": 126.18141174316406, "learning_rate": 3.963815789473684e-06, "loss": 1.332, "step": 2936 }, { "epoch": 0.6038240131578947, "grad_norm": 24.758825302124023, "learning_rate": 3.961759868421053e-06, "loss": 0.1685, "step": 2937 }, { "epoch": 0.6040296052631579, "grad_norm": 36.22583770751953, "learning_rate": 3.959703947368422e-06, "loss": 0.1357, "step": 2938 }, { "epoch": 0.604235197368421, "grad_norm": 45.00383377075195, "learning_rate": 3.9576480263157895e-06, "loss": 0.2386, "step": 2939 }, { "epoch": 0.6044407894736842, "grad_norm": 61.163150787353516, "learning_rate": 3.955592105263158e-06, "loss": 0.4824, "step": 2940 }, { "epoch": 0.6046463815789473, "grad_norm": 33.523536682128906, "learning_rate": 3.953536184210527e-06, "loss": 0.1658, "step": 2941 }, { "epoch": 0.6048519736842105, "grad_norm": 213.0513458251953, "learning_rate": 3.951480263157895e-06, "loss": 1.7344, "step": 2942 }, { "epoch": 0.6050575657894737, "grad_norm": 154.14242553710938, "learning_rate": 3.9494243421052636e-06, "loss": 1.1641, "step": 2943 }, { "epoch": 0.6052631578947368, "grad_norm": 24.982690811157227, "learning_rate": 3.947368421052632e-06, "loss": 0.1437, "step": 2944 }, { "epoch": 0.60546875, "grad_norm": 33.158409118652344, "learning_rate": 3.9453125e-06, "loss": 0.1807, "step": 2945 }, { "epoch": 0.6056743421052632, "grad_norm": 32.535552978515625, "learning_rate": 3.943256578947369e-06, "loss": 0.2043, "step": 2946 }, { "epoch": 0.6058799342105263, "grad_norm": 82.36363983154297, "learning_rate": 3.941200657894737e-06, "loss": 1.1348, "step": 2947 }, { "epoch": 0.6060855263157895, "grad_norm": 27.639474868774414, "learning_rate": 3.9391447368421055e-06, "loss": 0.1666, "step": 2948 }, { "epoch": 0.6062911184210527, "grad_norm": 7.795935153961182, "learning_rate": 3.937088815789473e-06, "loss": 0.0232, "step": 2949 }, { "epoch": 0.6064967105263158, "grad_norm": 48.76676559448242, "learning_rate": 3.935032894736843e-06, "loss": 0.2646, "step": 2950 }, { "epoch": 0.606702302631579, "grad_norm": 301.8919982910156, "learning_rate": 3.932976973684211e-06, "loss": 3.4805, "step": 2951 }, { "epoch": 0.6069078947368421, "grad_norm": 96.42731475830078, "learning_rate": 3.93092105263158e-06, "loss": 0.3589, "step": 2952 }, { "epoch": 0.6071134868421053, "grad_norm": 28.83100700378418, "learning_rate": 3.9288651315789475e-06, "loss": 0.1103, "step": 2953 }, { "epoch": 0.6073190789473685, "grad_norm": 48.73336410522461, "learning_rate": 3.926809210526316e-06, "loss": 0.1463, "step": 2954 }, { "epoch": 0.6075246710526315, "grad_norm": 86.17424774169922, "learning_rate": 3.924753289473685e-06, "loss": 0.7109, "step": 2955 }, { "epoch": 0.6077302631578947, "grad_norm": 163.81114196777344, "learning_rate": 3.922697368421053e-06, "loss": 0.9512, "step": 2956 }, { "epoch": 0.6079358552631579, "grad_norm": 58.4301643371582, "learning_rate": 3.9206414473684215e-06, "loss": 0.686, "step": 2957 }, { "epoch": 0.608141447368421, "grad_norm": 27.873249053955078, "learning_rate": 3.918585526315789e-06, "loss": 0.0857, "step": 2958 }, { "epoch": 0.6083470394736842, "grad_norm": 214.70391845703125, "learning_rate": 3.916529605263158e-06, "loss": 1.2114, "step": 2959 }, { "epoch": 0.6085526315789473, "grad_norm": 48.34069061279297, "learning_rate": 3.914473684210527e-06, "loss": 1.0767, "step": 2960 }, { "epoch": 0.6087582236842105, "grad_norm": 143.89100646972656, "learning_rate": 3.912417763157896e-06, "loss": 0.2601, "step": 2961 }, { "epoch": 0.6089638157894737, "grad_norm": 214.7021484375, "learning_rate": 3.9103618421052635e-06, "loss": 1.3516, "step": 2962 }, { "epoch": 0.6091694078947368, "grad_norm": 94.79835510253906, "learning_rate": 3.908305921052632e-06, "loss": 0.6562, "step": 2963 }, { "epoch": 0.609375, "grad_norm": 145.60464477539062, "learning_rate": 3.90625e-06, "loss": 1.582, "step": 2964 }, { "epoch": 0.6095805921052632, "grad_norm": 185.40184020996094, "learning_rate": 3.904194078947369e-06, "loss": 2.375, "step": 2965 }, { "epoch": 0.6097861842105263, "grad_norm": 58.59943389892578, "learning_rate": 3.902138157894737e-06, "loss": 0.416, "step": 2966 }, { "epoch": 0.6099917763157895, "grad_norm": 23.911781311035156, "learning_rate": 3.900082236842105e-06, "loss": 0.1401, "step": 2967 }, { "epoch": 0.6101973684210527, "grad_norm": 72.9454574584961, "learning_rate": 3.898026315789474e-06, "loss": 0.4939, "step": 2968 }, { "epoch": 0.6104029605263158, "grad_norm": 87.19686889648438, "learning_rate": 3.895970394736843e-06, "loss": 0.5747, "step": 2969 }, { "epoch": 0.610608552631579, "grad_norm": 14.566104888916016, "learning_rate": 3.893914473684211e-06, "loss": 0.0856, "step": 2970 }, { "epoch": 0.6108141447368421, "grad_norm": 30.912155151367188, "learning_rate": 3.8918585526315795e-06, "loss": 0.3553, "step": 2971 }, { "epoch": 0.6110197368421053, "grad_norm": 38.03947448730469, "learning_rate": 3.889802631578947e-06, "loss": 0.0714, "step": 2972 }, { "epoch": 0.6112253289473685, "grad_norm": 18.726146697998047, "learning_rate": 3.887746710526316e-06, "loss": 0.1164, "step": 2973 }, { "epoch": 0.6114309210526315, "grad_norm": 13.745087623596191, "learning_rate": 3.885690789473685e-06, "loss": 0.0674, "step": 2974 }, { "epoch": 0.6116365131578947, "grad_norm": 18.427637100219727, "learning_rate": 3.883634868421053e-06, "loss": 0.0439, "step": 2975 }, { "epoch": 0.6118421052631579, "grad_norm": 1.9210829734802246, "learning_rate": 3.8815789473684214e-06, "loss": 0.0071, "step": 2976 }, { "epoch": 0.612047697368421, "grad_norm": 58.49398422241211, "learning_rate": 3.879523026315789e-06, "loss": 0.4298, "step": 2977 }, { "epoch": 0.6122532894736842, "grad_norm": 87.32428741455078, "learning_rate": 3.877467105263158e-06, "loss": 0.4519, "step": 2978 }, { "epoch": 0.6124588815789473, "grad_norm": 68.74420166015625, "learning_rate": 3.875411184210527e-06, "loss": 0.7339, "step": 2979 }, { "epoch": 0.6126644736842105, "grad_norm": 218.61643981933594, "learning_rate": 3.8733552631578955e-06, "loss": 1.7524, "step": 2980 }, { "epoch": 0.6128700657894737, "grad_norm": 110.84031677246094, "learning_rate": 3.871299342105263e-06, "loss": 1.3379, "step": 2981 }, { "epoch": 0.6130756578947368, "grad_norm": 14.353666305541992, "learning_rate": 3.869243421052632e-06, "loss": 0.0664, "step": 2982 }, { "epoch": 0.61328125, "grad_norm": 111.86526489257812, "learning_rate": 3.8671875e-06, "loss": 0.5103, "step": 2983 }, { "epoch": 0.6134868421052632, "grad_norm": 81.48924255371094, "learning_rate": 3.865131578947369e-06, "loss": 0.8076, "step": 2984 }, { "epoch": 0.6136924342105263, "grad_norm": 56.58262252807617, "learning_rate": 3.8630756578947374e-06, "loss": 1.1459, "step": 2985 }, { "epoch": 0.6138980263157895, "grad_norm": 67.2794418334961, "learning_rate": 3.861019736842105e-06, "loss": 0.4785, "step": 2986 }, { "epoch": 0.6141036184210527, "grad_norm": 380.1355895996094, "learning_rate": 3.858963815789474e-06, "loss": 0.4165, "step": 2987 }, { "epoch": 0.6143092105263158, "grad_norm": 24.88655662536621, "learning_rate": 3.856907894736843e-06, "loss": 0.165, "step": 2988 }, { "epoch": 0.614514802631579, "grad_norm": 7.305649280548096, "learning_rate": 3.854851973684211e-06, "loss": 0.0214, "step": 2989 }, { "epoch": 0.6147203947368421, "grad_norm": 48.309288024902344, "learning_rate": 3.852796052631579e-06, "loss": 0.4727, "step": 2990 }, { "epoch": 0.6149259868421053, "grad_norm": 6.971612930297852, "learning_rate": 3.850740131578948e-06, "loss": 0.0163, "step": 2991 }, { "epoch": 0.6151315789473685, "grad_norm": 11.797289848327637, "learning_rate": 3.848684210526316e-06, "loss": 0.025, "step": 2992 }, { "epoch": 0.6153371710526315, "grad_norm": 14.574793815612793, "learning_rate": 3.846628289473685e-06, "loss": 0.0544, "step": 2993 }, { "epoch": 0.6155427631578947, "grad_norm": 41.31720733642578, "learning_rate": 3.844572368421053e-06, "loss": 0.1635, "step": 2994 }, { "epoch": 0.6157483552631579, "grad_norm": 114.87956237792969, "learning_rate": 3.842516447368421e-06, "loss": 0.7246, "step": 2995 }, { "epoch": 0.615953947368421, "grad_norm": 15.706768035888672, "learning_rate": 3.840460526315789e-06, "loss": 0.0469, "step": 2996 }, { "epoch": 0.6161595394736842, "grad_norm": 126.6058578491211, "learning_rate": 3.838404605263158e-06, "loss": 0.7329, "step": 2997 }, { "epoch": 0.6163651315789473, "grad_norm": 73.98974609375, "learning_rate": 3.836348684210527e-06, "loss": 0.3649, "step": 2998 }, { "epoch": 0.6165707236842105, "grad_norm": 68.81523895263672, "learning_rate": 3.834292763157895e-06, "loss": 0.6123, "step": 2999 }, { "epoch": 0.6167763157894737, "grad_norm": 78.69081115722656, "learning_rate": 3.832236842105263e-06, "loss": 0.6562, "step": 3000 }, { "epoch": 0.6169819078947368, "grad_norm": 159.8917999267578, "learning_rate": 3.830180921052632e-06, "loss": 1.3, "step": 3001 }, { "epoch": 0.6171875, "grad_norm": 10.381299018859863, "learning_rate": 3.828125000000001e-06, "loss": 0.0371, "step": 3002 }, { "epoch": 0.6173930921052632, "grad_norm": 14.345075607299805, "learning_rate": 3.826069078947369e-06, "loss": 0.0769, "step": 3003 }, { "epoch": 0.6175986842105263, "grad_norm": 50.893611907958984, "learning_rate": 3.824013157894737e-06, "loss": 0.2246, "step": 3004 }, { "epoch": 0.6178042763157895, "grad_norm": 102.69839477539062, "learning_rate": 3.821957236842105e-06, "loss": 1.4531, "step": 3005 }, { "epoch": 0.6180098684210527, "grad_norm": 23.60352325439453, "learning_rate": 3.819901315789474e-06, "loss": 0.1658, "step": 3006 }, { "epoch": 0.6182154605263158, "grad_norm": 111.5946044921875, "learning_rate": 3.817845394736843e-06, "loss": 1.0547, "step": 3007 }, { "epoch": 0.618421052631579, "grad_norm": 45.55596160888672, "learning_rate": 3.815789473684211e-06, "loss": 0.1719, "step": 3008 }, { "epoch": 0.6186266447368421, "grad_norm": 72.77814483642578, "learning_rate": 3.8137335526315793e-06, "loss": 0.6543, "step": 3009 }, { "epoch": 0.6188322368421053, "grad_norm": 6.910340309143066, "learning_rate": 3.8116776315789476e-06, "loss": 0.0386, "step": 3010 }, { "epoch": 0.6190378289473685, "grad_norm": 9.22628116607666, "learning_rate": 3.8096217105263163e-06, "loss": 0.0263, "step": 3011 }, { "epoch": 0.6192434210526315, "grad_norm": 12.718021392822266, "learning_rate": 3.8075657894736846e-06, "loss": 0.0328, "step": 3012 }, { "epoch": 0.6194490131578947, "grad_norm": 306.0678405761719, "learning_rate": 3.805509868421053e-06, "loss": 1.8391, "step": 3013 }, { "epoch": 0.6196546052631579, "grad_norm": 16.660133361816406, "learning_rate": 3.8034539473684212e-06, "loss": 0.0919, "step": 3014 }, { "epoch": 0.619860197368421, "grad_norm": 11.382872581481934, "learning_rate": 3.8013980263157895e-06, "loss": 0.0356, "step": 3015 }, { "epoch": 0.6200657894736842, "grad_norm": 130.11062622070312, "learning_rate": 3.799342105263158e-06, "loss": 1.5994, "step": 3016 }, { "epoch": 0.6202713815789473, "grad_norm": 5.162786483764648, "learning_rate": 3.797286184210527e-06, "loss": 0.0276, "step": 3017 }, { "epoch": 0.6204769736842105, "grad_norm": 48.869354248046875, "learning_rate": 3.7952302631578953e-06, "loss": 0.3009, "step": 3018 }, { "epoch": 0.6206825657894737, "grad_norm": 28.228675842285156, "learning_rate": 3.7931743421052636e-06, "loss": 0.1079, "step": 3019 }, { "epoch": 0.6208881578947368, "grad_norm": 35.6295280456543, "learning_rate": 3.791118421052632e-06, "loss": 0.1815, "step": 3020 }, { "epoch": 0.62109375, "grad_norm": 46.84367752075195, "learning_rate": 3.7890625e-06, "loss": 1.1396, "step": 3021 }, { "epoch": 0.6212993421052632, "grad_norm": 78.19721984863281, "learning_rate": 3.7870065789473685e-06, "loss": 0.408, "step": 3022 }, { "epoch": 0.6215049342105263, "grad_norm": 34.55738830566406, "learning_rate": 3.7849506578947372e-06, "loss": 0.1225, "step": 3023 }, { "epoch": 0.6217105263157895, "grad_norm": 37.923274993896484, "learning_rate": 3.7828947368421055e-06, "loss": 0.1769, "step": 3024 }, { "epoch": 0.6219161184210527, "grad_norm": 252.22579956054688, "learning_rate": 3.780838815789474e-06, "loss": 1.5234, "step": 3025 }, { "epoch": 0.6221217105263158, "grad_norm": 106.46589660644531, "learning_rate": 3.7787828947368426e-06, "loss": 0.6422, "step": 3026 }, { "epoch": 0.622327302631579, "grad_norm": 23.139705657958984, "learning_rate": 3.776726973684211e-06, "loss": 0.1002, "step": 3027 }, { "epoch": 0.6225328947368421, "grad_norm": 274.6846618652344, "learning_rate": 3.7746710526315796e-06, "loss": 2.9341, "step": 3028 }, { "epoch": 0.6227384868421053, "grad_norm": 15.81564712524414, "learning_rate": 3.772615131578948e-06, "loss": 0.0327, "step": 3029 }, { "epoch": 0.6229440789473685, "grad_norm": 85.33782958984375, "learning_rate": 3.770559210526316e-06, "loss": 0.374, "step": 3030 }, { "epoch": 0.6231496710526315, "grad_norm": 8.911646842956543, "learning_rate": 3.7685032894736845e-06, "loss": 0.0349, "step": 3031 }, { "epoch": 0.6233552631578947, "grad_norm": 273.3668212890625, "learning_rate": 3.766447368421053e-06, "loss": 1.6562, "step": 3032 }, { "epoch": 0.6235608552631579, "grad_norm": 78.16008758544922, "learning_rate": 3.764391447368421e-06, "loss": 0.6855, "step": 3033 }, { "epoch": 0.623766447368421, "grad_norm": 272.2970886230469, "learning_rate": 3.7623355263157894e-06, "loss": 1.418, "step": 3034 }, { "epoch": 0.6239720394736842, "grad_norm": 86.03559875488281, "learning_rate": 3.760279605263158e-06, "loss": 0.3335, "step": 3035 }, { "epoch": 0.6241776315789473, "grad_norm": 63.09440231323242, "learning_rate": 3.758223684210527e-06, "loss": 0.4609, "step": 3036 }, { "epoch": 0.6243832236842105, "grad_norm": 12.845219612121582, "learning_rate": 3.756167763157895e-06, "loss": 0.0556, "step": 3037 }, { "epoch": 0.6245888157894737, "grad_norm": 73.8133773803711, "learning_rate": 3.7541118421052635e-06, "loss": 1.0171, "step": 3038 }, { "epoch": 0.6247944078947368, "grad_norm": 31.156518936157227, "learning_rate": 3.752055921052632e-06, "loss": 0.0969, "step": 3039 }, { "epoch": 0.625, "grad_norm": 195.18698120117188, "learning_rate": 3.7500000000000005e-06, "loss": 1.0049, "step": 3040 }, { "epoch": 0.6252055921052632, "grad_norm": 130.98939514160156, "learning_rate": 3.747944078947369e-06, "loss": 1.2461, "step": 3041 }, { "epoch": 0.6254111842105263, "grad_norm": 120.5953598022461, "learning_rate": 3.745888157894737e-06, "loss": 0.9956, "step": 3042 }, { "epoch": 0.6256167763157895, "grad_norm": 26.116661071777344, "learning_rate": 3.7438322368421054e-06, "loss": 0.1619, "step": 3043 }, { "epoch": 0.6258223684210527, "grad_norm": 51.90599822998047, "learning_rate": 3.7417763157894737e-06, "loss": 0.2637, "step": 3044 }, { "epoch": 0.6260279605263158, "grad_norm": 56.899166107177734, "learning_rate": 3.7397203947368425e-06, "loss": 0.5507, "step": 3045 }, { "epoch": 0.626233552631579, "grad_norm": 188.2094268798828, "learning_rate": 3.737664473684211e-06, "loss": 1.4003, "step": 3046 }, { "epoch": 0.6264391447368421, "grad_norm": 10.64781379699707, "learning_rate": 3.7356085526315795e-06, "loss": 0.0209, "step": 3047 }, { "epoch": 0.6266447368421053, "grad_norm": 92.44720458984375, "learning_rate": 3.733552631578948e-06, "loss": 0.9292, "step": 3048 }, { "epoch": 0.6268503289473685, "grad_norm": 81.34446716308594, "learning_rate": 3.731496710526316e-06, "loss": 0.2773, "step": 3049 }, { "epoch": 0.6270559210526315, "grad_norm": 6.181501388549805, "learning_rate": 3.7294407894736844e-06, "loss": 0.0216, "step": 3050 }, { "epoch": 0.6272615131578947, "grad_norm": 93.75605010986328, "learning_rate": 3.7273848684210527e-06, "loss": 0.5339, "step": 3051 }, { "epoch": 0.6274671052631579, "grad_norm": 25.27839469909668, "learning_rate": 3.725328947368421e-06, "loss": 0.1904, "step": 3052 }, { "epoch": 0.627672697368421, "grad_norm": 290.1649169921875, "learning_rate": 3.7232730263157897e-06, "loss": 6.9922, "step": 3053 }, { "epoch": 0.6278782894736842, "grad_norm": 58.1987419128418, "learning_rate": 3.721217105263158e-06, "loss": 0.4951, "step": 3054 }, { "epoch": 0.6280838815789473, "grad_norm": 50.64646911621094, "learning_rate": 3.7191611842105268e-06, "loss": 0.4727, "step": 3055 }, { "epoch": 0.6282894736842105, "grad_norm": 281.0575866699219, "learning_rate": 3.717105263157895e-06, "loss": 1.5088, "step": 3056 }, { "epoch": 0.6284950657894737, "grad_norm": 210.42665100097656, "learning_rate": 3.7150493421052634e-06, "loss": 1.8594, "step": 3057 }, { "epoch": 0.6287006578947368, "grad_norm": 220.4858856201172, "learning_rate": 3.712993421052632e-06, "loss": 1.8516, "step": 3058 }, { "epoch": 0.62890625, "grad_norm": 19.82638931274414, "learning_rate": 3.7109375000000004e-06, "loss": 0.1426, "step": 3059 }, { "epoch": 0.6291118421052632, "grad_norm": 38.84360885620117, "learning_rate": 3.7088815789473687e-06, "loss": 0.1567, "step": 3060 }, { "epoch": 0.6293174342105263, "grad_norm": 90.0738296508789, "learning_rate": 3.706825657894737e-06, "loss": 0.6924, "step": 3061 }, { "epoch": 0.6295230263157895, "grad_norm": 156.3818359375, "learning_rate": 3.7047697368421053e-06, "loss": 4.0312, "step": 3062 }, { "epoch": 0.6297286184210527, "grad_norm": 13.9113130569458, "learning_rate": 3.7027138157894736e-06, "loss": 0.0436, "step": 3063 }, { "epoch": 0.6299342105263158, "grad_norm": 13.95801830291748, "learning_rate": 3.7006578947368428e-06, "loss": 0.0598, "step": 3064 }, { "epoch": 0.630139802631579, "grad_norm": 80.27362823486328, "learning_rate": 3.698601973684211e-06, "loss": 0.7644, "step": 3065 }, { "epoch": 0.6303453947368421, "grad_norm": 25.933847427368164, "learning_rate": 3.6965460526315794e-06, "loss": 0.1202, "step": 3066 }, { "epoch": 0.6305509868421053, "grad_norm": 1.7056341171264648, "learning_rate": 3.6944901315789477e-06, "loss": 0.0149, "step": 3067 }, { "epoch": 0.6307565789473685, "grad_norm": 71.80921936035156, "learning_rate": 3.692434210526316e-06, "loss": 0.3212, "step": 3068 }, { "epoch": 0.6309621710526315, "grad_norm": 136.79132080078125, "learning_rate": 3.6903782894736843e-06, "loss": 1.6016, "step": 3069 }, { "epoch": 0.6311677631578947, "grad_norm": 29.179485321044922, "learning_rate": 3.688322368421053e-06, "loss": 0.249, "step": 3070 }, { "epoch": 0.6313733552631579, "grad_norm": 7.887320041656494, "learning_rate": 3.6862664473684213e-06, "loss": 0.0408, "step": 3071 }, { "epoch": 0.631578947368421, "grad_norm": 192.6964569091797, "learning_rate": 3.6842105263157896e-06, "loss": 3.0781, "step": 3072 }, { "epoch": 0.6317845394736842, "grad_norm": 24.15157699584961, "learning_rate": 3.682154605263158e-06, "loss": 0.0643, "step": 3073 }, { "epoch": 0.6319901315789473, "grad_norm": 63.423065185546875, "learning_rate": 3.6800986842105267e-06, "loss": 0.8096, "step": 3074 }, { "epoch": 0.6321957236842105, "grad_norm": 17.51265525817871, "learning_rate": 3.6780427631578954e-06, "loss": 0.0548, "step": 3075 }, { "epoch": 0.6324013157894737, "grad_norm": 43.094146728515625, "learning_rate": 3.6759868421052637e-06, "loss": 0.4062, "step": 3076 }, { "epoch": 0.6326069078947368, "grad_norm": 75.5141830444336, "learning_rate": 3.673930921052632e-06, "loss": 0.4463, "step": 3077 }, { "epoch": 0.6328125, "grad_norm": 177.7023162841797, "learning_rate": 3.6718750000000003e-06, "loss": 0.5419, "step": 3078 }, { "epoch": 0.6330180921052632, "grad_norm": 41.45880889892578, "learning_rate": 3.6698190789473686e-06, "loss": 0.2769, "step": 3079 }, { "epoch": 0.6332236842105263, "grad_norm": 39.05188751220703, "learning_rate": 3.667763157894737e-06, "loss": 0.1777, "step": 3080 }, { "epoch": 0.6334292763157895, "grad_norm": 25.432228088378906, "learning_rate": 3.6657072368421052e-06, "loss": 0.1069, "step": 3081 }, { "epoch": 0.6336348684210527, "grad_norm": 52.14925765991211, "learning_rate": 3.663651315789474e-06, "loss": 0.4355, "step": 3082 }, { "epoch": 0.6338404605263158, "grad_norm": 48.919376373291016, "learning_rate": 3.6615953947368427e-06, "loss": 0.3457, "step": 3083 }, { "epoch": 0.634046052631579, "grad_norm": 76.8521728515625, "learning_rate": 3.659539473684211e-06, "loss": 0.5571, "step": 3084 }, { "epoch": 0.6342516447368421, "grad_norm": 56.81825637817383, "learning_rate": 3.6574835526315793e-06, "loss": 0.4053, "step": 3085 }, { "epoch": 0.6344572368421053, "grad_norm": 23.865190505981445, "learning_rate": 3.6554276315789476e-06, "loss": 0.0938, "step": 3086 }, { "epoch": 0.6346628289473685, "grad_norm": 48.03897476196289, "learning_rate": 3.6533717105263163e-06, "loss": 0.2354, "step": 3087 }, { "epoch": 0.6348684210526315, "grad_norm": 16.575376510620117, "learning_rate": 3.6513157894736846e-06, "loss": 0.1307, "step": 3088 }, { "epoch": 0.6350740131578947, "grad_norm": 15.570993423461914, "learning_rate": 3.649259868421053e-06, "loss": 0.0776, "step": 3089 }, { "epoch": 0.6352796052631579, "grad_norm": 104.2299575805664, "learning_rate": 3.6472039473684212e-06, "loss": 0.6108, "step": 3090 }, { "epoch": 0.635485197368421, "grad_norm": 30.886699676513672, "learning_rate": 3.6451480263157895e-06, "loss": 0.1184, "step": 3091 }, { "epoch": 0.6356907894736842, "grad_norm": 6.679627895355225, "learning_rate": 3.643092105263158e-06, "loss": 0.0188, "step": 3092 }, { "epoch": 0.6358963815789473, "grad_norm": 41.379486083984375, "learning_rate": 3.641036184210527e-06, "loss": 0.2657, "step": 3093 }, { "epoch": 0.6361019736842105, "grad_norm": 227.99899291992188, "learning_rate": 3.6389802631578953e-06, "loss": 3.2051, "step": 3094 }, { "epoch": 0.6363075657894737, "grad_norm": 66.64334106445312, "learning_rate": 3.6369243421052636e-06, "loss": 0.3052, "step": 3095 }, { "epoch": 0.6365131578947368, "grad_norm": 154.73300170898438, "learning_rate": 3.634868421052632e-06, "loss": 1.2402, "step": 3096 }, { "epoch": 0.63671875, "grad_norm": 22.066747665405273, "learning_rate": 3.6328125e-06, "loss": 0.1602, "step": 3097 }, { "epoch": 0.6369243421052632, "grad_norm": 24.434141159057617, "learning_rate": 3.6307565789473685e-06, "loss": 0.1138, "step": 3098 }, { "epoch": 0.6371299342105263, "grad_norm": 13.209729194641113, "learning_rate": 3.628700657894737e-06, "loss": 0.0374, "step": 3099 }, { "epoch": 0.6373355263157895, "grad_norm": 33.27187728881836, "learning_rate": 3.6266447368421055e-06, "loss": 0.1598, "step": 3100 }, { "epoch": 0.6375411184210527, "grad_norm": 181.0970001220703, "learning_rate": 3.624588815789474e-06, "loss": 0.7554, "step": 3101 }, { "epoch": 0.6377467105263158, "grad_norm": 14.186434745788574, "learning_rate": 3.6225328947368426e-06, "loss": 0.0403, "step": 3102 }, { "epoch": 0.637952302631579, "grad_norm": 151.62107849121094, "learning_rate": 3.620476973684211e-06, "loss": 0.7837, "step": 3103 }, { "epoch": 0.6381578947368421, "grad_norm": 66.51253509521484, "learning_rate": 3.618421052631579e-06, "loss": 0.4541, "step": 3104 }, { "epoch": 0.6383634868421053, "grad_norm": 78.89129638671875, "learning_rate": 3.616365131578948e-06, "loss": 0.3008, "step": 3105 }, { "epoch": 0.6385690789473685, "grad_norm": 83.75529479980469, "learning_rate": 3.6143092105263162e-06, "loss": 0.8242, "step": 3106 }, { "epoch": 0.6387746710526315, "grad_norm": 55.27632522583008, "learning_rate": 3.6122532894736845e-06, "loss": 0.7607, "step": 3107 }, { "epoch": 0.6389802631578947, "grad_norm": 269.60205078125, "learning_rate": 3.610197368421053e-06, "loss": 3.5859, "step": 3108 }, { "epoch": 0.6391858552631579, "grad_norm": 30.80267906188965, "learning_rate": 3.608141447368421e-06, "loss": 0.186, "step": 3109 }, { "epoch": 0.639391447368421, "grad_norm": 286.6560974121094, "learning_rate": 3.6060855263157894e-06, "loss": 2.557, "step": 3110 }, { "epoch": 0.6395970394736842, "grad_norm": 55.31343078613281, "learning_rate": 3.6040296052631577e-06, "loss": 0.2019, "step": 3111 }, { "epoch": 0.6398026315789473, "grad_norm": 91.99237060546875, "learning_rate": 3.601973684210527e-06, "loss": 1.0671, "step": 3112 }, { "epoch": 0.6400082236842105, "grad_norm": 77.53325653076172, "learning_rate": 3.599917763157895e-06, "loss": 0.4028, "step": 3113 }, { "epoch": 0.6402138157894737, "grad_norm": 121.19267272949219, "learning_rate": 3.5978618421052635e-06, "loss": 1.4805, "step": 3114 }, { "epoch": 0.6404194078947368, "grad_norm": 74.09706115722656, "learning_rate": 3.595805921052632e-06, "loss": 0.3799, "step": 3115 }, { "epoch": 0.640625, "grad_norm": 7.16273307800293, "learning_rate": 3.59375e-06, "loss": 0.0334, "step": 3116 }, { "epoch": 0.6408305921052632, "grad_norm": 82.85372161865234, "learning_rate": 3.591694078947369e-06, "loss": 0.3466, "step": 3117 }, { "epoch": 0.6410361842105263, "grad_norm": 157.82568359375, "learning_rate": 3.589638157894737e-06, "loss": 2.0801, "step": 3118 }, { "epoch": 0.6412417763157895, "grad_norm": 21.15732765197754, "learning_rate": 3.5875822368421054e-06, "loss": 0.1625, "step": 3119 }, { "epoch": 0.6414473684210527, "grad_norm": 92.05467224121094, "learning_rate": 3.5855263157894737e-06, "loss": 0.7227, "step": 3120 }, { "epoch": 0.6416529605263158, "grad_norm": 59.44108200073242, "learning_rate": 3.5834703947368425e-06, "loss": 0.4119, "step": 3121 }, { "epoch": 0.641858552631579, "grad_norm": 77.30799865722656, "learning_rate": 3.581414473684211e-06, "loss": 0.5615, "step": 3122 }, { "epoch": 0.6420641447368421, "grad_norm": 39.51390075683594, "learning_rate": 3.5793585526315795e-06, "loss": 0.208, "step": 3123 }, { "epoch": 0.6422697368421053, "grad_norm": 10.58827018737793, "learning_rate": 3.577302631578948e-06, "loss": 0.0664, "step": 3124 }, { "epoch": 0.6424753289473685, "grad_norm": 51.237483978271484, "learning_rate": 3.575246710526316e-06, "loss": 0.2739, "step": 3125 }, { "epoch": 0.6426809210526315, "grad_norm": 128.9248809814453, "learning_rate": 3.5731907894736844e-06, "loss": 1.5781, "step": 3126 }, { "epoch": 0.6428865131578947, "grad_norm": 124.37478637695312, "learning_rate": 3.5711348684210527e-06, "loss": 0.7417, "step": 3127 }, { "epoch": 0.6430921052631579, "grad_norm": 21.678897857666016, "learning_rate": 3.569078947368421e-06, "loss": 0.077, "step": 3128 }, { "epoch": 0.643297697368421, "grad_norm": 82.77493286132812, "learning_rate": 3.5670230263157898e-06, "loss": 0.8571, "step": 3129 }, { "epoch": 0.6435032894736842, "grad_norm": 4.961030006408691, "learning_rate": 3.564967105263158e-06, "loss": 0.0242, "step": 3130 }, { "epoch": 0.6437088815789473, "grad_norm": 7.824987411499023, "learning_rate": 3.562911184210527e-06, "loss": 0.0383, "step": 3131 }, { "epoch": 0.6439144736842105, "grad_norm": 44.31489181518555, "learning_rate": 3.560855263157895e-06, "loss": 0.1769, "step": 3132 }, { "epoch": 0.6441200657894737, "grad_norm": 7.785831451416016, "learning_rate": 3.5587993421052634e-06, "loss": 0.0197, "step": 3133 }, { "epoch": 0.6443256578947368, "grad_norm": 108.23649597167969, "learning_rate": 3.556743421052632e-06, "loss": 1.5122, "step": 3134 }, { "epoch": 0.64453125, "grad_norm": 249.96385192871094, "learning_rate": 3.5546875000000004e-06, "loss": 2.1001, "step": 3135 }, { "epoch": 0.6447368421052632, "grad_norm": 44.426605224609375, "learning_rate": 3.5526315789473687e-06, "loss": 0.3362, "step": 3136 }, { "epoch": 0.6449424342105263, "grad_norm": 75.90795135498047, "learning_rate": 3.550575657894737e-06, "loss": 0.4028, "step": 3137 }, { "epoch": 0.6451480263157895, "grad_norm": 26.48517417907715, "learning_rate": 3.5485197368421053e-06, "loss": 0.201, "step": 3138 }, { "epoch": 0.6453536184210527, "grad_norm": 350.2469787597656, "learning_rate": 3.5464638157894736e-06, "loss": 2.1758, "step": 3139 }, { "epoch": 0.6455592105263158, "grad_norm": 215.62979125976562, "learning_rate": 3.544407894736843e-06, "loss": 1.0637, "step": 3140 }, { "epoch": 0.645764802631579, "grad_norm": 85.76861572265625, "learning_rate": 3.542351973684211e-06, "loss": 0.6162, "step": 3141 }, { "epoch": 0.6459703947368421, "grad_norm": 47.0927848815918, "learning_rate": 3.5402960526315794e-06, "loss": 0.1606, "step": 3142 }, { "epoch": 0.6461759868421053, "grad_norm": 11.67752742767334, "learning_rate": 3.5382401315789477e-06, "loss": 0.0507, "step": 3143 }, { "epoch": 0.6463815789473685, "grad_norm": 210.50845336914062, "learning_rate": 3.536184210526316e-06, "loss": 3.2803, "step": 3144 }, { "epoch": 0.6465871710526315, "grad_norm": 64.89917755126953, "learning_rate": 3.5341282894736843e-06, "loss": 0.4854, "step": 3145 }, { "epoch": 0.6467927631578947, "grad_norm": 44.814552307128906, "learning_rate": 3.5320723684210526e-06, "loss": 0.2405, "step": 3146 }, { "epoch": 0.6469983552631579, "grad_norm": 139.78465270996094, "learning_rate": 3.5300164473684214e-06, "loss": 1.7299, "step": 3147 }, { "epoch": 0.647203947368421, "grad_norm": 42.38409423828125, "learning_rate": 3.5279605263157897e-06, "loss": 0.1758, "step": 3148 }, { "epoch": 0.6474095394736842, "grad_norm": 52.895355224609375, "learning_rate": 3.525904605263158e-06, "loss": 0.3369, "step": 3149 }, { "epoch": 0.6476151315789473, "grad_norm": 37.557254791259766, "learning_rate": 3.5238486842105267e-06, "loss": 0.2969, "step": 3150 }, { "epoch": 0.6478207236842105, "grad_norm": 44.674861907958984, "learning_rate": 3.521792763157895e-06, "loss": 0.3528, "step": 3151 }, { "epoch": 0.6480263157894737, "grad_norm": 56.13893508911133, "learning_rate": 3.5197368421052637e-06, "loss": 0.2539, "step": 3152 }, { "epoch": 0.6482319078947368, "grad_norm": 81.45880889892578, "learning_rate": 3.517680921052632e-06, "loss": 0.4458, "step": 3153 }, { "epoch": 0.6484375, "grad_norm": 16.000099182128906, "learning_rate": 3.5156250000000003e-06, "loss": 0.0565, "step": 3154 }, { "epoch": 0.6486430921052632, "grad_norm": 27.399072647094727, "learning_rate": 3.5135690789473686e-06, "loss": 0.1685, "step": 3155 }, { "epoch": 0.6488486842105263, "grad_norm": 97.06348419189453, "learning_rate": 3.511513157894737e-06, "loss": 1.0283, "step": 3156 }, { "epoch": 0.6490542763157895, "grad_norm": 75.66485595703125, "learning_rate": 3.5094572368421052e-06, "loss": 0.5708, "step": 3157 }, { "epoch": 0.6492598684210527, "grad_norm": 30.852996826171875, "learning_rate": 3.5074013157894735e-06, "loss": 0.2207, "step": 3158 }, { "epoch": 0.6494654605263158, "grad_norm": 159.2292022705078, "learning_rate": 3.5053453947368427e-06, "loss": 1.0933, "step": 3159 }, { "epoch": 0.649671052631579, "grad_norm": 193.49783325195312, "learning_rate": 3.503289473684211e-06, "loss": 1.0957, "step": 3160 }, { "epoch": 0.6498766447368421, "grad_norm": 14.792075157165527, "learning_rate": 3.5012335526315793e-06, "loss": 0.04, "step": 3161 }, { "epoch": 0.6500822368421053, "grad_norm": 4.340132236480713, "learning_rate": 3.4991776315789476e-06, "loss": 0.0182, "step": 3162 }, { "epoch": 0.6502878289473685, "grad_norm": 164.7305145263672, "learning_rate": 3.497121710526316e-06, "loss": 0.5986, "step": 3163 }, { "epoch": 0.6504934210526315, "grad_norm": 10.377251625061035, "learning_rate": 3.4950657894736846e-06, "loss": 0.0913, "step": 3164 }, { "epoch": 0.6506990131578947, "grad_norm": 104.15735626220703, "learning_rate": 3.493009868421053e-06, "loss": 0.5234, "step": 3165 }, { "epoch": 0.6509046052631579, "grad_norm": 108.16358947753906, "learning_rate": 3.4909539473684212e-06, "loss": 0.5708, "step": 3166 }, { "epoch": 0.651110197368421, "grad_norm": 179.53842163085938, "learning_rate": 3.4888980263157896e-06, "loss": 2.0121, "step": 3167 }, { "epoch": 0.6513157894736842, "grad_norm": 34.50899124145508, "learning_rate": 3.486842105263158e-06, "loss": 0.2715, "step": 3168 }, { "epoch": 0.6515213815789473, "grad_norm": 96.20963287353516, "learning_rate": 3.484786184210527e-06, "loss": 0.5691, "step": 3169 }, { "epoch": 0.6517269736842105, "grad_norm": 53.6518669128418, "learning_rate": 3.4827302631578953e-06, "loss": 1.0603, "step": 3170 }, { "epoch": 0.6519325657894737, "grad_norm": 46.49071502685547, "learning_rate": 3.4806743421052636e-06, "loss": 0.2417, "step": 3171 }, { "epoch": 0.6521381578947368, "grad_norm": 212.54161071777344, "learning_rate": 3.478618421052632e-06, "loss": 3.9355, "step": 3172 }, { "epoch": 0.65234375, "grad_norm": 149.0517578125, "learning_rate": 3.4765625000000002e-06, "loss": 2.4963, "step": 3173 }, { "epoch": 0.6525493421052632, "grad_norm": 50.21551513671875, "learning_rate": 3.4745065789473685e-06, "loss": 0.8823, "step": 3174 }, { "epoch": 0.6527549342105263, "grad_norm": 16.226369857788086, "learning_rate": 3.472450657894737e-06, "loss": 0.1119, "step": 3175 }, { "epoch": 0.6529605263157895, "grad_norm": 87.44296264648438, "learning_rate": 3.4703947368421056e-06, "loss": 0.9211, "step": 3176 }, { "epoch": 0.6531661184210527, "grad_norm": 341.7458190917969, "learning_rate": 3.468338815789474e-06, "loss": 2.0176, "step": 3177 }, { "epoch": 0.6533717105263158, "grad_norm": 45.576454162597656, "learning_rate": 3.4662828947368426e-06, "loss": 0.5179, "step": 3178 }, { "epoch": 0.653577302631579, "grad_norm": 2.6023755073547363, "learning_rate": 3.464226973684211e-06, "loss": 0.0182, "step": 3179 }, { "epoch": 0.6537828947368421, "grad_norm": 37.29802322387695, "learning_rate": 3.462171052631579e-06, "loss": 0.1499, "step": 3180 }, { "epoch": 0.6539884868421053, "grad_norm": 80.45775604248047, "learning_rate": 3.460115131578948e-06, "loss": 0.7656, "step": 3181 }, { "epoch": 0.6541940789473685, "grad_norm": 94.13098907470703, "learning_rate": 3.4580592105263162e-06, "loss": 0.6172, "step": 3182 }, { "epoch": 0.6543996710526315, "grad_norm": 49.114540100097656, "learning_rate": 3.4560032894736845e-06, "loss": 0.5891, "step": 3183 }, { "epoch": 0.6546052631578947, "grad_norm": 5.352883338928223, "learning_rate": 3.453947368421053e-06, "loss": 0.0316, "step": 3184 }, { "epoch": 0.6548108552631579, "grad_norm": 212.9857940673828, "learning_rate": 3.451891447368421e-06, "loss": 1.8887, "step": 3185 }, { "epoch": 0.655016447368421, "grad_norm": 156.31256103515625, "learning_rate": 3.4498355263157895e-06, "loss": 0.7271, "step": 3186 }, { "epoch": 0.6552220394736842, "grad_norm": 153.89443969726562, "learning_rate": 3.4477796052631578e-06, "loss": 0.3936, "step": 3187 }, { "epoch": 0.6554276315789473, "grad_norm": 71.2162857055664, "learning_rate": 3.445723684210527e-06, "loss": 0.6831, "step": 3188 }, { "epoch": 0.6556332236842105, "grad_norm": 70.92512512207031, "learning_rate": 3.443667763157895e-06, "loss": 1.2344, "step": 3189 }, { "epoch": 0.6558388157894737, "grad_norm": 42.88156509399414, "learning_rate": 3.4416118421052635e-06, "loss": 0.5996, "step": 3190 }, { "epoch": 0.6560444078947368, "grad_norm": 29.331872940063477, "learning_rate": 3.439555921052632e-06, "loss": 0.178, "step": 3191 }, { "epoch": 0.65625, "grad_norm": 10.978585243225098, "learning_rate": 3.4375e-06, "loss": 0.0507, "step": 3192 }, { "epoch": 0.6564555921052632, "grad_norm": 109.93766784667969, "learning_rate": 3.435444078947369e-06, "loss": 0.4141, "step": 3193 }, { "epoch": 0.6566611842105263, "grad_norm": 62.97780227661133, "learning_rate": 3.433388157894737e-06, "loss": 0.3738, "step": 3194 }, { "epoch": 0.6568667763157895, "grad_norm": 101.40888214111328, "learning_rate": 3.4313322368421055e-06, "loss": 0.4717, "step": 3195 }, { "epoch": 0.6570723684210527, "grad_norm": 85.25467681884766, "learning_rate": 3.4292763157894738e-06, "loss": 0.5078, "step": 3196 }, { "epoch": 0.6572779605263158, "grad_norm": 311.12896728515625, "learning_rate": 3.4272203947368425e-06, "loss": 2.0234, "step": 3197 }, { "epoch": 0.657483552631579, "grad_norm": 134.25634765625, "learning_rate": 3.4251644736842112e-06, "loss": 0.624, "step": 3198 }, { "epoch": 0.6576891447368421, "grad_norm": 136.97348022460938, "learning_rate": 3.4231085526315795e-06, "loss": 0.6392, "step": 3199 }, { "epoch": 0.6578947368421053, "grad_norm": 57.581295013427734, "learning_rate": 3.421052631578948e-06, "loss": 0.2651, "step": 3200 }, { "epoch": 0.6581003289473685, "grad_norm": 116.44409942626953, "learning_rate": 3.418996710526316e-06, "loss": 1.2163, "step": 3201 }, { "epoch": 0.6583059210526315, "grad_norm": 46.1597785949707, "learning_rate": 3.4169407894736844e-06, "loss": 0.363, "step": 3202 }, { "epoch": 0.6585115131578947, "grad_norm": 115.06037902832031, "learning_rate": 3.4148848684210527e-06, "loss": 0.6846, "step": 3203 }, { "epoch": 0.6587171052631579, "grad_norm": 51.736854553222656, "learning_rate": 3.412828947368421e-06, "loss": 0.38, "step": 3204 }, { "epoch": 0.658922697368421, "grad_norm": 84.20677947998047, "learning_rate": 3.4107730263157893e-06, "loss": 0.4746, "step": 3205 }, { "epoch": 0.6591282894736842, "grad_norm": 322.1570739746094, "learning_rate": 3.408717105263158e-06, "loss": 1.7383, "step": 3206 }, { "epoch": 0.6593338815789473, "grad_norm": 32.89204406738281, "learning_rate": 3.406661184210527e-06, "loss": 0.1348, "step": 3207 }, { "epoch": 0.6595394736842105, "grad_norm": 64.17630004882812, "learning_rate": 3.404605263157895e-06, "loss": 0.3184, "step": 3208 }, { "epoch": 0.6597450657894737, "grad_norm": 66.1485595703125, "learning_rate": 3.4025493421052634e-06, "loss": 0.4746, "step": 3209 }, { "epoch": 0.6599506578947368, "grad_norm": 45.919761657714844, "learning_rate": 3.4004934210526317e-06, "loss": 0.2534, "step": 3210 }, { "epoch": 0.66015625, "grad_norm": 10.345285415649414, "learning_rate": 3.3984375000000004e-06, "loss": 0.0558, "step": 3211 }, { "epoch": 0.6603618421052632, "grad_norm": 4.789392948150635, "learning_rate": 3.3963815789473687e-06, "loss": 0.0346, "step": 3212 }, { "epoch": 0.6605674342105263, "grad_norm": 55.93247985839844, "learning_rate": 3.394325657894737e-06, "loss": 0.5265, "step": 3213 }, { "epoch": 0.6607730263157895, "grad_norm": 157.648681640625, "learning_rate": 3.3922697368421054e-06, "loss": 0.9993, "step": 3214 }, { "epoch": 0.6609786184210527, "grad_norm": 63.11244201660156, "learning_rate": 3.3902138157894737e-06, "loss": 0.4219, "step": 3215 }, { "epoch": 0.6611842105263158, "grad_norm": 58.97265625, "learning_rate": 3.388157894736843e-06, "loss": 0.2764, "step": 3216 }, { "epoch": 0.661389802631579, "grad_norm": 16.413705825805664, "learning_rate": 3.386101973684211e-06, "loss": 0.0874, "step": 3217 }, { "epoch": 0.6615953947368421, "grad_norm": 28.57497215270996, "learning_rate": 3.3840460526315794e-06, "loss": 0.1262, "step": 3218 }, { "epoch": 0.6618009868421053, "grad_norm": 12.752662658691406, "learning_rate": 3.3819901315789477e-06, "loss": 0.031, "step": 3219 }, { "epoch": 0.6620065789473685, "grad_norm": 40.67525863647461, "learning_rate": 3.379934210526316e-06, "loss": 0.1836, "step": 3220 }, { "epoch": 0.6622121710526315, "grad_norm": 127.39356994628906, "learning_rate": 3.3778782894736843e-06, "loss": 1.2589, "step": 3221 }, { "epoch": 0.6624177631578947, "grad_norm": 29.218990325927734, "learning_rate": 3.3758223684210526e-06, "loss": 0.1157, "step": 3222 }, { "epoch": 0.6626233552631579, "grad_norm": 79.3394546508789, "learning_rate": 3.3737664473684214e-06, "loss": 0.5171, "step": 3223 }, { "epoch": 0.662828947368421, "grad_norm": 22.72614288330078, "learning_rate": 3.3717105263157897e-06, "loss": 0.0734, "step": 3224 }, { "epoch": 0.6630345394736842, "grad_norm": 100.64647674560547, "learning_rate": 3.369654605263158e-06, "loss": 0.2825, "step": 3225 }, { "epoch": 0.6632401315789473, "grad_norm": 111.43272399902344, "learning_rate": 3.3675986842105267e-06, "loss": 0.9766, "step": 3226 }, { "epoch": 0.6634457236842105, "grad_norm": 11.597419738769531, "learning_rate": 3.365542763157895e-06, "loss": 0.0285, "step": 3227 }, { "epoch": 0.6636513157894737, "grad_norm": 53.736785888671875, "learning_rate": 3.3634868421052637e-06, "loss": 0.4497, "step": 3228 }, { "epoch": 0.6638569078947368, "grad_norm": 12.969392776489258, "learning_rate": 3.361430921052632e-06, "loss": 0.0267, "step": 3229 }, { "epoch": 0.6640625, "grad_norm": 12.707687377929688, "learning_rate": 3.3593750000000003e-06, "loss": 0.0576, "step": 3230 }, { "epoch": 0.6642680921052632, "grad_norm": 67.7114028930664, "learning_rate": 3.3573190789473686e-06, "loss": 0.8477, "step": 3231 }, { "epoch": 0.6644736842105263, "grad_norm": 136.9271697998047, "learning_rate": 3.355263157894737e-06, "loss": 1.0117, "step": 3232 }, { "epoch": 0.6646792763157895, "grad_norm": 19.096012115478516, "learning_rate": 3.3532072368421053e-06, "loss": 0.0375, "step": 3233 }, { "epoch": 0.6648848684210527, "grad_norm": 17.807567596435547, "learning_rate": 3.3511513157894736e-06, "loss": 0.1042, "step": 3234 }, { "epoch": 0.6650904605263158, "grad_norm": 193.1005401611328, "learning_rate": 3.3490953947368427e-06, "loss": 1.6834, "step": 3235 }, { "epoch": 0.665296052631579, "grad_norm": 24.10215187072754, "learning_rate": 3.347039473684211e-06, "loss": 0.067, "step": 3236 }, { "epoch": 0.6655016447368421, "grad_norm": 20.50545883178711, "learning_rate": 3.3449835526315793e-06, "loss": 0.1064, "step": 3237 }, { "epoch": 0.6657072368421053, "grad_norm": 237.63134765625, "learning_rate": 3.3429276315789476e-06, "loss": 1.8896, "step": 3238 }, { "epoch": 0.6659128289473685, "grad_norm": 157.13015747070312, "learning_rate": 3.340871710526316e-06, "loss": 1.4141, "step": 3239 }, { "epoch": 0.6661184210526315, "grad_norm": 50.243473052978516, "learning_rate": 3.3388157894736847e-06, "loss": 0.1548, "step": 3240 }, { "epoch": 0.6663240131578947, "grad_norm": 40.15450668334961, "learning_rate": 3.336759868421053e-06, "loss": 0.158, "step": 3241 }, { "epoch": 0.6665296052631579, "grad_norm": 32.83635330200195, "learning_rate": 3.3347039473684213e-06, "loss": 0.1926, "step": 3242 }, { "epoch": 0.666735197368421, "grad_norm": 194.82899475097656, "learning_rate": 3.3326480263157896e-06, "loss": 1.0696, "step": 3243 }, { "epoch": 0.6669407894736842, "grad_norm": 97.34325408935547, "learning_rate": 3.330592105263158e-06, "loss": 0.3147, "step": 3244 }, { "epoch": 0.6671463815789473, "grad_norm": 116.61618041992188, "learning_rate": 3.328536184210527e-06, "loss": 0.8232, "step": 3245 }, { "epoch": 0.6673519736842105, "grad_norm": 74.43878936767578, "learning_rate": 3.3264802631578953e-06, "loss": 0.4126, "step": 3246 }, { "epoch": 0.6675575657894737, "grad_norm": 79.01127624511719, "learning_rate": 3.3244243421052636e-06, "loss": 0.896, "step": 3247 }, { "epoch": 0.6677631578947368, "grad_norm": 105.87078857421875, "learning_rate": 3.322368421052632e-06, "loss": 0.6873, "step": 3248 }, { "epoch": 0.66796875, "grad_norm": 214.1470947265625, "learning_rate": 3.3203125000000002e-06, "loss": 1.1866, "step": 3249 }, { "epoch": 0.6681743421052632, "grad_norm": 48.77829360961914, "learning_rate": 3.3182565789473685e-06, "loss": 0.1844, "step": 3250 }, { "epoch": 0.6683799342105263, "grad_norm": 63.23047637939453, "learning_rate": 3.316200657894737e-06, "loss": 0.4084, "step": 3251 }, { "epoch": 0.6685855263157895, "grad_norm": 22.30518341064453, "learning_rate": 3.314144736842105e-06, "loss": 0.1091, "step": 3252 }, { "epoch": 0.6687911184210527, "grad_norm": 59.115047454833984, "learning_rate": 3.312088815789474e-06, "loss": 1.0452, "step": 3253 }, { "epoch": 0.6689967105263158, "grad_norm": 62.17098617553711, "learning_rate": 3.3100328947368426e-06, "loss": 0.2827, "step": 3254 }, { "epoch": 0.669202302631579, "grad_norm": 80.80205535888672, "learning_rate": 3.307976973684211e-06, "loss": 0.3718, "step": 3255 }, { "epoch": 0.6694078947368421, "grad_norm": 25.268495559692383, "learning_rate": 3.3059210526315792e-06, "loss": 0.0938, "step": 3256 }, { "epoch": 0.6696134868421053, "grad_norm": 70.50684356689453, "learning_rate": 3.3038651315789475e-06, "loss": 0.7617, "step": 3257 }, { "epoch": 0.6698190789473685, "grad_norm": 75.90278625488281, "learning_rate": 3.3018092105263162e-06, "loss": 0.3823, "step": 3258 }, { "epoch": 0.6700246710526315, "grad_norm": 26.843549728393555, "learning_rate": 3.2997532894736846e-06, "loss": 0.2371, "step": 3259 }, { "epoch": 0.6702302631578947, "grad_norm": 17.75262451171875, "learning_rate": 3.297697368421053e-06, "loss": 0.0475, "step": 3260 }, { "epoch": 0.6704358552631579, "grad_norm": 13.3474702835083, "learning_rate": 3.295641447368421e-06, "loss": 0.0367, "step": 3261 }, { "epoch": 0.670641447368421, "grad_norm": 74.675048828125, "learning_rate": 3.2935855263157895e-06, "loss": 0.4546, "step": 3262 }, { "epoch": 0.6708470394736842, "grad_norm": 72.75067901611328, "learning_rate": 3.2915296052631578e-06, "loss": 0.2834, "step": 3263 }, { "epoch": 0.6710526315789473, "grad_norm": 99.70536804199219, "learning_rate": 3.289473684210527e-06, "loss": 0.283, "step": 3264 }, { "epoch": 0.6712582236842105, "grad_norm": 85.30469512939453, "learning_rate": 3.2874177631578952e-06, "loss": 0.6484, "step": 3265 }, { "epoch": 0.6714638157894737, "grad_norm": 26.27799415588379, "learning_rate": 3.2853618421052635e-06, "loss": 0.0748, "step": 3266 }, { "epoch": 0.6716694078947368, "grad_norm": 36.755226135253906, "learning_rate": 3.283305921052632e-06, "loss": 0.2861, "step": 3267 }, { "epoch": 0.671875, "grad_norm": 41.10977554321289, "learning_rate": 3.28125e-06, "loss": 0.1633, "step": 3268 }, { "epoch": 0.6720805921052632, "grad_norm": 5.010176658630371, "learning_rate": 3.2791940789473684e-06, "loss": 0.0148, "step": 3269 }, { "epoch": 0.6722861842105263, "grad_norm": 206.39895629882812, "learning_rate": 3.277138157894737e-06, "loss": 1.395, "step": 3270 }, { "epoch": 0.6724917763157895, "grad_norm": 35.15998840332031, "learning_rate": 3.2750822368421055e-06, "loss": 0.1631, "step": 3271 }, { "epoch": 0.6726973684210527, "grad_norm": 88.14653778076172, "learning_rate": 3.2730263157894738e-06, "loss": 1.4863, "step": 3272 }, { "epoch": 0.6729029605263158, "grad_norm": 10.832938194274902, "learning_rate": 3.2709703947368425e-06, "loss": 0.0224, "step": 3273 }, { "epoch": 0.673108552631579, "grad_norm": 4.167391300201416, "learning_rate": 3.268914473684211e-06, "loss": 0.01, "step": 3274 }, { "epoch": 0.6733141447368421, "grad_norm": 78.18887329101562, "learning_rate": 3.2668585526315795e-06, "loss": 0.3818, "step": 3275 }, { "epoch": 0.6735197368421053, "grad_norm": 30.035018920898438, "learning_rate": 3.264802631578948e-06, "loss": 0.1101, "step": 3276 }, { "epoch": 0.6737253289473685, "grad_norm": 67.00541687011719, "learning_rate": 3.262746710526316e-06, "loss": 0.7168, "step": 3277 }, { "epoch": 0.6739309210526315, "grad_norm": 20.658916473388672, "learning_rate": 3.2606907894736844e-06, "loss": 0.0575, "step": 3278 }, { "epoch": 0.6741365131578947, "grad_norm": 84.07476806640625, "learning_rate": 3.2586348684210528e-06, "loss": 0.2954, "step": 3279 }, { "epoch": 0.6743421052631579, "grad_norm": 16.64449691772461, "learning_rate": 3.256578947368421e-06, "loss": 0.049, "step": 3280 }, { "epoch": 0.674547697368421, "grad_norm": 1.9693042039871216, "learning_rate": 3.2545230263157894e-06, "loss": 0.0096, "step": 3281 }, { "epoch": 0.6747532894736842, "grad_norm": 7.381502628326416, "learning_rate": 3.252467105263158e-06, "loss": 0.0294, "step": 3282 }, { "epoch": 0.6749588815789473, "grad_norm": 8.049131393432617, "learning_rate": 3.250411184210527e-06, "loss": 0.0359, "step": 3283 }, { "epoch": 0.6751644736842105, "grad_norm": 292.46478271484375, "learning_rate": 3.248355263157895e-06, "loss": 2.5938, "step": 3284 }, { "epoch": 0.6753700657894737, "grad_norm": 55.00822067260742, "learning_rate": 3.2462993421052634e-06, "loss": 0.3257, "step": 3285 }, { "epoch": 0.6755756578947368, "grad_norm": 85.55534362792969, "learning_rate": 3.2442434210526317e-06, "loss": 0.7344, "step": 3286 }, { "epoch": 0.67578125, "grad_norm": 13.499138832092285, "learning_rate": 3.2421875000000005e-06, "loss": 0.056, "step": 3287 }, { "epoch": 0.6759868421052632, "grad_norm": 8.864892959594727, "learning_rate": 3.2401315789473688e-06, "loss": 0.0306, "step": 3288 }, { "epoch": 0.6761924342105263, "grad_norm": 15.075209617614746, "learning_rate": 3.238075657894737e-06, "loss": 0.0374, "step": 3289 }, { "epoch": 0.6763980263157895, "grad_norm": 100.03958129882812, "learning_rate": 3.2360197368421054e-06, "loss": 0.6396, "step": 3290 }, { "epoch": 0.6766036184210527, "grad_norm": 17.477874755859375, "learning_rate": 3.2339638157894737e-06, "loss": 0.0568, "step": 3291 }, { "epoch": 0.6768092105263158, "grad_norm": 165.34417724609375, "learning_rate": 3.231907894736843e-06, "loss": 0.8341, "step": 3292 }, { "epoch": 0.677014802631579, "grad_norm": 99.70112609863281, "learning_rate": 3.229851973684211e-06, "loss": 0.7082, "step": 3293 }, { "epoch": 0.6772203947368421, "grad_norm": 20.990474700927734, "learning_rate": 3.2277960526315794e-06, "loss": 0.0511, "step": 3294 }, { "epoch": 0.6774259868421053, "grad_norm": 75.85401916503906, "learning_rate": 3.2257401315789477e-06, "loss": 0.2578, "step": 3295 }, { "epoch": 0.6776315789473685, "grad_norm": 122.92005920410156, "learning_rate": 3.223684210526316e-06, "loss": 0.9365, "step": 3296 }, { "epoch": 0.6778371710526315, "grad_norm": 24.13018035888672, "learning_rate": 3.2216282894736843e-06, "loss": 0.084, "step": 3297 }, { "epoch": 0.6780427631578947, "grad_norm": 43.882537841796875, "learning_rate": 3.2195723684210527e-06, "loss": 0.275, "step": 3298 }, { "epoch": 0.6782483552631579, "grad_norm": 331.625732421875, "learning_rate": 3.2175164473684214e-06, "loss": 1.9658, "step": 3299 }, { "epoch": 0.678453947368421, "grad_norm": 18.516796112060547, "learning_rate": 3.2154605263157897e-06, "loss": 0.0343, "step": 3300 }, { "epoch": 0.6786595394736842, "grad_norm": 63.71052169799805, "learning_rate": 3.213404605263158e-06, "loss": 0.3242, "step": 3301 }, { "epoch": 0.6788651315789473, "grad_norm": 27.304218292236328, "learning_rate": 3.2113486842105267e-06, "loss": 0.0901, "step": 3302 }, { "epoch": 0.6790707236842105, "grad_norm": 10.372776985168457, "learning_rate": 3.209292763157895e-06, "loss": 0.0406, "step": 3303 }, { "epoch": 0.6792763157894737, "grad_norm": 370.8432922363281, "learning_rate": 3.2072368421052637e-06, "loss": 3.5913, "step": 3304 }, { "epoch": 0.6794819078947368, "grad_norm": 444.7209777832031, "learning_rate": 3.205180921052632e-06, "loss": 2.3018, "step": 3305 }, { "epoch": 0.6796875, "grad_norm": 40.198917388916016, "learning_rate": 3.2031250000000004e-06, "loss": 0.1348, "step": 3306 }, { "epoch": 0.6798930921052632, "grad_norm": 100.21803283691406, "learning_rate": 3.2010690789473687e-06, "loss": 0.48, "step": 3307 }, { "epoch": 0.6800986842105263, "grad_norm": 176.85357666015625, "learning_rate": 3.199013157894737e-06, "loss": 2.0938, "step": 3308 }, { "epoch": 0.6803042763157895, "grad_norm": 48.267669677734375, "learning_rate": 3.1969572368421053e-06, "loss": 0.2378, "step": 3309 }, { "epoch": 0.6805098684210527, "grad_norm": 182.934814453125, "learning_rate": 3.1949013157894736e-06, "loss": 2.1943, "step": 3310 }, { "epoch": 0.6807154605263158, "grad_norm": 11.440282821655273, "learning_rate": 3.1928453947368427e-06, "loss": 0.0456, "step": 3311 }, { "epoch": 0.680921052631579, "grad_norm": 46.359275817871094, "learning_rate": 3.190789473684211e-06, "loss": 0.3066, "step": 3312 }, { "epoch": 0.6811266447368421, "grad_norm": 208.1730194091797, "learning_rate": 3.1887335526315793e-06, "loss": 1.2603, "step": 3313 }, { "epoch": 0.6813322368421053, "grad_norm": 83.39657592773438, "learning_rate": 3.1866776315789476e-06, "loss": 0.4561, "step": 3314 }, { "epoch": 0.6815378289473685, "grad_norm": 23.404781341552734, "learning_rate": 3.184621710526316e-06, "loss": 0.0781, "step": 3315 }, { "epoch": 0.6817434210526315, "grad_norm": 11.594120025634766, "learning_rate": 3.1825657894736842e-06, "loss": 0.0537, "step": 3316 }, { "epoch": 0.6819490131578947, "grad_norm": 21.154258728027344, "learning_rate": 3.180509868421053e-06, "loss": 0.1012, "step": 3317 }, { "epoch": 0.6821546052631579, "grad_norm": 151.18084716796875, "learning_rate": 3.1784539473684213e-06, "loss": 0.3872, "step": 3318 }, { "epoch": 0.682360197368421, "grad_norm": 78.29886627197266, "learning_rate": 3.1763980263157896e-06, "loss": 0.5068, "step": 3319 }, { "epoch": 0.6825657894736842, "grad_norm": 18.019350051879883, "learning_rate": 3.174342105263158e-06, "loss": 0.0747, "step": 3320 }, { "epoch": 0.6827713815789473, "grad_norm": 51.904693603515625, "learning_rate": 3.1722861842105266e-06, "loss": 0.153, "step": 3321 }, { "epoch": 0.6829769736842105, "grad_norm": 4.642718315124512, "learning_rate": 3.1702302631578953e-06, "loss": 0.0173, "step": 3322 }, { "epoch": 0.6831825657894737, "grad_norm": 262.5654602050781, "learning_rate": 3.1681743421052636e-06, "loss": 0.9023, "step": 3323 }, { "epoch": 0.6833881578947368, "grad_norm": 446.20977783203125, "learning_rate": 3.166118421052632e-06, "loss": 3.8042, "step": 3324 }, { "epoch": 0.68359375, "grad_norm": 6.6159491539001465, "learning_rate": 3.1640625000000003e-06, "loss": 0.0229, "step": 3325 }, { "epoch": 0.6837993421052632, "grad_norm": 69.24520111083984, "learning_rate": 3.1620065789473686e-06, "loss": 0.8218, "step": 3326 }, { "epoch": 0.6840049342105263, "grad_norm": 87.0827865600586, "learning_rate": 3.159950657894737e-06, "loss": 0.4414, "step": 3327 }, { "epoch": 0.6842105263157895, "grad_norm": 29.62224769592285, "learning_rate": 3.157894736842105e-06, "loss": 0.1229, "step": 3328 }, { "epoch": 0.6844161184210527, "grad_norm": 120.4643325805664, "learning_rate": 3.155838815789474e-06, "loss": 1.1523, "step": 3329 }, { "epoch": 0.6846217105263158, "grad_norm": 41.537147521972656, "learning_rate": 3.1537828947368426e-06, "loss": 0.115, "step": 3330 }, { "epoch": 0.684827302631579, "grad_norm": 13.852596282958984, "learning_rate": 3.151726973684211e-06, "loss": 0.0294, "step": 3331 }, { "epoch": 0.6850328947368421, "grad_norm": 392.4903259277344, "learning_rate": 3.1496710526315792e-06, "loss": 1.192, "step": 3332 }, { "epoch": 0.6852384868421053, "grad_norm": 50.01044464111328, "learning_rate": 3.1476151315789475e-06, "loss": 0.4821, "step": 3333 }, { "epoch": 0.6854440789473685, "grad_norm": 44.1708869934082, "learning_rate": 3.1455592105263163e-06, "loss": 0.1864, "step": 3334 }, { "epoch": 0.6856496710526315, "grad_norm": 21.9021053314209, "learning_rate": 3.1435032894736846e-06, "loss": 0.1038, "step": 3335 }, { "epoch": 0.6858552631578947, "grad_norm": 26.434432983398438, "learning_rate": 3.141447368421053e-06, "loss": 0.0614, "step": 3336 }, { "epoch": 0.6860608552631579, "grad_norm": 117.45870208740234, "learning_rate": 3.139391447368421e-06, "loss": 0.4876, "step": 3337 }, { "epoch": 0.686266447368421, "grad_norm": 28.194772720336914, "learning_rate": 3.1373355263157895e-06, "loss": 0.1292, "step": 3338 }, { "epoch": 0.6864720394736842, "grad_norm": 113.220947265625, "learning_rate": 3.1352796052631578e-06, "loss": 0.928, "step": 3339 }, { "epoch": 0.6866776315789473, "grad_norm": 28.999126434326172, "learning_rate": 3.133223684210527e-06, "loss": 0.1147, "step": 3340 }, { "epoch": 0.6868832236842105, "grad_norm": 33.28362274169922, "learning_rate": 3.1311677631578952e-06, "loss": 0.1189, "step": 3341 }, { "epoch": 0.6870888157894737, "grad_norm": 17.352548599243164, "learning_rate": 3.1291118421052635e-06, "loss": 0.0612, "step": 3342 }, { "epoch": 0.6872944078947368, "grad_norm": 23.192705154418945, "learning_rate": 3.127055921052632e-06, "loss": 0.0581, "step": 3343 }, { "epoch": 0.6875, "grad_norm": 153.1804962158203, "learning_rate": 3.125e-06, "loss": 1.7793, "step": 3344 }, { "epoch": 0.6877055921052632, "grad_norm": 12.044906616210938, "learning_rate": 3.1229440789473685e-06, "loss": 0.0305, "step": 3345 }, { "epoch": 0.6879111842105263, "grad_norm": 85.57589721679688, "learning_rate": 3.120888157894737e-06, "loss": 0.4304, "step": 3346 }, { "epoch": 0.6881167763157895, "grad_norm": 25.999723434448242, "learning_rate": 3.1188322368421055e-06, "loss": 0.0721, "step": 3347 }, { "epoch": 0.6883223684210527, "grad_norm": 79.24471282958984, "learning_rate": 3.116776315789474e-06, "loss": 0.2808, "step": 3348 }, { "epoch": 0.6885279605263158, "grad_norm": 123.840576171875, "learning_rate": 3.1147203947368425e-06, "loss": 0.8135, "step": 3349 }, { "epoch": 0.688733552631579, "grad_norm": 9.612730979919434, "learning_rate": 3.112664473684211e-06, "loss": 0.0342, "step": 3350 }, { "epoch": 0.6889391447368421, "grad_norm": 8.449972152709961, "learning_rate": 3.1106085526315795e-06, "loss": 0.0251, "step": 3351 }, { "epoch": 0.6891447368421053, "grad_norm": 58.4804573059082, "learning_rate": 3.108552631578948e-06, "loss": 0.3596, "step": 3352 }, { "epoch": 0.6893503289473685, "grad_norm": 10.939740180969238, "learning_rate": 3.106496710526316e-06, "loss": 0.0265, "step": 3353 }, { "epoch": 0.6895559210526315, "grad_norm": 11.97055435180664, "learning_rate": 3.1044407894736845e-06, "loss": 0.04, "step": 3354 }, { "epoch": 0.6897615131578947, "grad_norm": 42.67375946044922, "learning_rate": 3.1023848684210528e-06, "loss": 0.3606, "step": 3355 }, { "epoch": 0.6899671052631579, "grad_norm": 138.6296844482422, "learning_rate": 3.100328947368421e-06, "loss": 0.7573, "step": 3356 }, { "epoch": 0.690172697368421, "grad_norm": 163.70718383789062, "learning_rate": 3.0982730263157894e-06, "loss": 0.6306, "step": 3357 }, { "epoch": 0.6903782894736842, "grad_norm": 27.16557502746582, "learning_rate": 3.0962171052631577e-06, "loss": 0.0928, "step": 3358 }, { "epoch": 0.6905838815789473, "grad_norm": 16.146018981933594, "learning_rate": 3.094161184210527e-06, "loss": 0.0303, "step": 3359 }, { "epoch": 0.6907894736842105, "grad_norm": 205.02914428710938, "learning_rate": 3.092105263157895e-06, "loss": 0.9766, "step": 3360 }, { "epoch": 0.6909950657894737, "grad_norm": 275.0433349609375, "learning_rate": 3.0900493421052634e-06, "loss": 1.0273, "step": 3361 }, { "epoch": 0.6912006578947368, "grad_norm": 84.11676025390625, "learning_rate": 3.0879934210526317e-06, "loss": 0.4956, "step": 3362 }, { "epoch": 0.69140625, "grad_norm": 78.59513854980469, "learning_rate": 3.0859375e-06, "loss": 0.5105, "step": 3363 }, { "epoch": 0.6916118421052632, "grad_norm": 62.19889450073242, "learning_rate": 3.0838815789473688e-06, "loss": 0.297, "step": 3364 }, { "epoch": 0.6918174342105263, "grad_norm": 74.00753021240234, "learning_rate": 3.081825657894737e-06, "loss": 0.6774, "step": 3365 }, { "epoch": 0.6920230263157895, "grad_norm": 65.46403503417969, "learning_rate": 3.0797697368421054e-06, "loss": 0.1715, "step": 3366 }, { "epoch": 0.6922286184210527, "grad_norm": 5.273571014404297, "learning_rate": 3.0777138157894737e-06, "loss": 0.0094, "step": 3367 }, { "epoch": 0.6924342105263158, "grad_norm": 36.14723587036133, "learning_rate": 3.0756578947368424e-06, "loss": 0.2231, "step": 3368 }, { "epoch": 0.692639802631579, "grad_norm": 58.71042251586914, "learning_rate": 3.073601973684211e-06, "loss": 0.1807, "step": 3369 }, { "epoch": 0.6928453947368421, "grad_norm": 112.91867065429688, "learning_rate": 3.0715460526315794e-06, "loss": 0.6636, "step": 3370 }, { "epoch": 0.6930509868421053, "grad_norm": 16.58724021911621, "learning_rate": 3.0694901315789478e-06, "loss": 0.0818, "step": 3371 }, { "epoch": 0.6932565789473685, "grad_norm": 7.87601900100708, "learning_rate": 3.067434210526316e-06, "loss": 0.0184, "step": 3372 }, { "epoch": 0.6934621710526315, "grad_norm": 99.63256072998047, "learning_rate": 3.0653782894736844e-06, "loss": 0.4766, "step": 3373 }, { "epoch": 0.6936677631578947, "grad_norm": 54.426666259765625, "learning_rate": 3.0633223684210527e-06, "loss": 0.1294, "step": 3374 }, { "epoch": 0.6938733552631579, "grad_norm": 102.09659576416016, "learning_rate": 3.061266447368421e-06, "loss": 0.5432, "step": 3375 }, { "epoch": 0.694078947368421, "grad_norm": 7.308078765869141, "learning_rate": 3.0592105263157897e-06, "loss": 0.0258, "step": 3376 }, { "epoch": 0.6942845394736842, "grad_norm": 93.13860321044922, "learning_rate": 3.057154605263158e-06, "loss": 0.3588, "step": 3377 }, { "epoch": 0.6944901315789473, "grad_norm": 201.73904418945312, "learning_rate": 3.0550986842105267e-06, "loss": 2.1309, "step": 3378 }, { "epoch": 0.6946957236842105, "grad_norm": 17.66130256652832, "learning_rate": 3.053042763157895e-06, "loss": 0.0568, "step": 3379 }, { "epoch": 0.6949013157894737, "grad_norm": 36.20933151245117, "learning_rate": 3.0509868421052633e-06, "loss": 0.08, "step": 3380 }, { "epoch": 0.6951069078947368, "grad_norm": 149.1884765625, "learning_rate": 3.048930921052632e-06, "loss": 0.8496, "step": 3381 }, { "epoch": 0.6953125, "grad_norm": 287.61090087890625, "learning_rate": 3.0468750000000004e-06, "loss": 2.5789, "step": 3382 }, { "epoch": 0.6955180921052632, "grad_norm": 17.56841278076172, "learning_rate": 3.0448190789473687e-06, "loss": 0.0774, "step": 3383 }, { "epoch": 0.6957236842105263, "grad_norm": 9.729691505432129, "learning_rate": 3.042763157894737e-06, "loss": 0.0404, "step": 3384 }, { "epoch": 0.6959292763157895, "grad_norm": 198.21624755859375, "learning_rate": 3.0407072368421053e-06, "loss": 1.6572, "step": 3385 }, { "epoch": 0.6961348684210527, "grad_norm": 112.7850112915039, "learning_rate": 3.0386513157894736e-06, "loss": 1.4648, "step": 3386 }, { "epoch": 0.6963404605263158, "grad_norm": 7.375352382659912, "learning_rate": 3.0365953947368427e-06, "loss": 0.0125, "step": 3387 }, { "epoch": 0.696546052631579, "grad_norm": 371.6708984375, "learning_rate": 3.034539473684211e-06, "loss": 4.4824, "step": 3388 }, { "epoch": 0.6967516447368421, "grad_norm": 59.32506561279297, "learning_rate": 3.0324835526315793e-06, "loss": 0.3149, "step": 3389 }, { "epoch": 0.6969572368421053, "grad_norm": 130.39996337890625, "learning_rate": 3.0304276315789476e-06, "loss": 0.7954, "step": 3390 }, { "epoch": 0.6971628289473685, "grad_norm": 187.2244415283203, "learning_rate": 3.028371710526316e-06, "loss": 1.3621, "step": 3391 }, { "epoch": 0.6973684210526315, "grad_norm": 78.69903564453125, "learning_rate": 3.0263157894736843e-06, "loss": 0.2903, "step": 3392 }, { "epoch": 0.6975740131578947, "grad_norm": 10.035125732421875, "learning_rate": 3.024259868421053e-06, "loss": 0.045, "step": 3393 }, { "epoch": 0.6977796052631579, "grad_norm": 9.696579933166504, "learning_rate": 3.0222039473684213e-06, "loss": 0.0282, "step": 3394 }, { "epoch": 0.697985197368421, "grad_norm": 56.52014923095703, "learning_rate": 3.0201480263157896e-06, "loss": 0.2593, "step": 3395 }, { "epoch": 0.6981907894736842, "grad_norm": 71.46183776855469, "learning_rate": 3.018092105263158e-06, "loss": 0.7671, "step": 3396 }, { "epoch": 0.6983963815789473, "grad_norm": 208.21092224121094, "learning_rate": 3.0160361842105266e-06, "loss": 1.2686, "step": 3397 }, { "epoch": 0.6986019736842105, "grad_norm": 71.5319595336914, "learning_rate": 3.0139802631578954e-06, "loss": 0.2834, "step": 3398 }, { "epoch": 0.6988075657894737, "grad_norm": 30.85727882385254, "learning_rate": 3.0119243421052637e-06, "loss": 0.1133, "step": 3399 }, { "epoch": 0.6990131578947368, "grad_norm": 118.65003967285156, "learning_rate": 3.009868421052632e-06, "loss": 0.3848, "step": 3400 }, { "epoch": 0.69921875, "grad_norm": 35.51658248901367, "learning_rate": 3.0078125000000003e-06, "loss": 0.098, "step": 3401 }, { "epoch": 0.6994243421052632, "grad_norm": 166.58946228027344, "learning_rate": 3.0057565789473686e-06, "loss": 1.2227, "step": 3402 }, { "epoch": 0.6996299342105263, "grad_norm": 269.583984375, "learning_rate": 3.003700657894737e-06, "loss": 1.7461, "step": 3403 }, { "epoch": 0.6998355263157895, "grad_norm": 2.891037702560425, "learning_rate": 3.001644736842105e-06, "loss": 0.0057, "step": 3404 }, { "epoch": 0.7000411184210527, "grad_norm": 19.326215744018555, "learning_rate": 2.999588815789474e-06, "loss": 0.0612, "step": 3405 }, { "epoch": 0.7002467105263158, "grad_norm": 167.4220733642578, "learning_rate": 2.9975328947368426e-06, "loss": 1.5229, "step": 3406 }, { "epoch": 0.700452302631579, "grad_norm": 32.15286636352539, "learning_rate": 2.995476973684211e-06, "loss": 0.1584, "step": 3407 }, { "epoch": 0.7006578947368421, "grad_norm": 27.928058624267578, "learning_rate": 2.9934210526315792e-06, "loss": 0.147, "step": 3408 }, { "epoch": 0.7008634868421053, "grad_norm": 9.766959190368652, "learning_rate": 2.9913651315789475e-06, "loss": 0.0411, "step": 3409 }, { "epoch": 0.7010690789473685, "grad_norm": 17.770694732666016, "learning_rate": 2.9893092105263163e-06, "loss": 0.0542, "step": 3410 }, { "epoch": 0.7012746710526315, "grad_norm": 216.3892822265625, "learning_rate": 2.9872532894736846e-06, "loss": 0.9492, "step": 3411 }, { "epoch": 0.7014802631578947, "grad_norm": 110.46758270263672, "learning_rate": 2.985197368421053e-06, "loss": 0.2543, "step": 3412 }, { "epoch": 0.7016858552631579, "grad_norm": 24.411813735961914, "learning_rate": 2.983141447368421e-06, "loss": 0.1053, "step": 3413 }, { "epoch": 0.701891447368421, "grad_norm": 85.42145538330078, "learning_rate": 2.9810855263157895e-06, "loss": 0.4009, "step": 3414 }, { "epoch": 0.7020970394736842, "grad_norm": 64.26917266845703, "learning_rate": 2.979029605263158e-06, "loss": 0.2238, "step": 3415 }, { "epoch": 0.7023026315789473, "grad_norm": 47.64011001586914, "learning_rate": 2.976973684210527e-06, "loss": 0.207, "step": 3416 }, { "epoch": 0.7025082236842105, "grad_norm": 11.225168228149414, "learning_rate": 2.9749177631578953e-06, "loss": 0.035, "step": 3417 }, { "epoch": 0.7027138157894737, "grad_norm": 18.76313018798828, "learning_rate": 2.9728618421052636e-06, "loss": 0.0425, "step": 3418 }, { "epoch": 0.7029194078947368, "grad_norm": 329.18212890625, "learning_rate": 2.970805921052632e-06, "loss": 2.1602, "step": 3419 }, { "epoch": 0.703125, "grad_norm": 72.75701904296875, "learning_rate": 2.96875e-06, "loss": 0.2817, "step": 3420 }, { "epoch": 0.7033305921052632, "grad_norm": 60.14959716796875, "learning_rate": 2.9666940789473685e-06, "loss": 0.542, "step": 3421 }, { "epoch": 0.7035361842105263, "grad_norm": 5.1692047119140625, "learning_rate": 2.9646381578947368e-06, "loss": 0.009, "step": 3422 }, { "epoch": 0.7037417763157895, "grad_norm": 54.832244873046875, "learning_rate": 2.9625822368421055e-06, "loss": 0.216, "step": 3423 }, { "epoch": 0.7039473684210527, "grad_norm": 44.158138275146484, "learning_rate": 2.960526315789474e-06, "loss": 0.2217, "step": 3424 }, { "epoch": 0.7041529605263158, "grad_norm": 37.87010955810547, "learning_rate": 2.9584703947368425e-06, "loss": 0.2466, "step": 3425 }, { "epoch": 0.704358552631579, "grad_norm": 90.8692626953125, "learning_rate": 2.956414473684211e-06, "loss": 0.3464, "step": 3426 }, { "epoch": 0.7045641447368421, "grad_norm": 176.05650329589844, "learning_rate": 2.954358552631579e-06, "loss": 1.3535, "step": 3427 }, { "epoch": 0.7047697368421053, "grad_norm": 29.216745376586914, "learning_rate": 2.952302631578948e-06, "loss": 0.1331, "step": 3428 }, { "epoch": 0.7049753289473685, "grad_norm": 54.78789520263672, "learning_rate": 2.950246710526316e-06, "loss": 0.2773, "step": 3429 }, { "epoch": 0.7051809210526315, "grad_norm": 35.18437576293945, "learning_rate": 2.9481907894736845e-06, "loss": 0.2085, "step": 3430 }, { "epoch": 0.7053865131578947, "grad_norm": 12.263779640197754, "learning_rate": 2.9461348684210528e-06, "loss": 0.0287, "step": 3431 }, { "epoch": 0.7055921052631579, "grad_norm": 7.749385833740234, "learning_rate": 2.944078947368421e-06, "loss": 0.0132, "step": 3432 }, { "epoch": 0.705797697368421, "grad_norm": 129.71316528320312, "learning_rate": 2.9420230263157894e-06, "loss": 1.0078, "step": 3433 }, { "epoch": 0.7060032894736842, "grad_norm": 36.13776397705078, "learning_rate": 2.9399671052631577e-06, "loss": 0.1083, "step": 3434 }, { "epoch": 0.7062088815789473, "grad_norm": 18.773517608642578, "learning_rate": 2.937911184210527e-06, "loss": 0.0686, "step": 3435 }, { "epoch": 0.7064144736842105, "grad_norm": 39.79354476928711, "learning_rate": 2.935855263157895e-06, "loss": 0.0767, "step": 3436 }, { "epoch": 0.7066200657894737, "grad_norm": 218.01246643066406, "learning_rate": 2.9337993421052635e-06, "loss": 1.1748, "step": 3437 }, { "epoch": 0.7068256578947368, "grad_norm": 88.22172546386719, "learning_rate": 2.9317434210526318e-06, "loss": 0.2578, "step": 3438 }, { "epoch": 0.70703125, "grad_norm": 74.5235824584961, "learning_rate": 2.9296875e-06, "loss": 0.2168, "step": 3439 }, { "epoch": 0.7072368421052632, "grad_norm": 96.4437484741211, "learning_rate": 2.927631578947369e-06, "loss": 1.356, "step": 3440 }, { "epoch": 0.7074424342105263, "grad_norm": 25.573415756225586, "learning_rate": 2.925575657894737e-06, "loss": 0.0914, "step": 3441 }, { "epoch": 0.7076480263157895, "grad_norm": 46.694000244140625, "learning_rate": 2.9235197368421054e-06, "loss": 0.2917, "step": 3442 }, { "epoch": 0.7078536184210527, "grad_norm": 15.460578918457031, "learning_rate": 2.9214638157894737e-06, "loss": 0.0303, "step": 3443 }, { "epoch": 0.7080592105263158, "grad_norm": 201.07655334472656, "learning_rate": 2.9194078947368424e-06, "loss": 0.8213, "step": 3444 }, { "epoch": 0.708264802631579, "grad_norm": 18.71917724609375, "learning_rate": 2.917351973684211e-06, "loss": 0.0297, "step": 3445 }, { "epoch": 0.7084703947368421, "grad_norm": 4.431929111480713, "learning_rate": 2.9152960526315795e-06, "loss": 0.0168, "step": 3446 }, { "epoch": 0.7086759868421053, "grad_norm": 39.46223831176758, "learning_rate": 2.9132401315789478e-06, "loss": 0.2019, "step": 3447 }, { "epoch": 0.7088815789473685, "grad_norm": 503.3543701171875, "learning_rate": 2.911184210526316e-06, "loss": 3.4336, "step": 3448 }, { "epoch": 0.7090871710526315, "grad_norm": 17.33144187927246, "learning_rate": 2.9091282894736844e-06, "loss": 0.0331, "step": 3449 }, { "epoch": 0.7092927631578947, "grad_norm": 100.51371765136719, "learning_rate": 2.9070723684210527e-06, "loss": 0.15, "step": 3450 }, { "epoch": 0.7094983552631579, "grad_norm": 119.1025619506836, "learning_rate": 2.905016447368421e-06, "loss": 0.5027, "step": 3451 }, { "epoch": 0.709703947368421, "grad_norm": 46.90568161010742, "learning_rate": 2.9029605263157897e-06, "loss": 0.2983, "step": 3452 }, { "epoch": 0.7099095394736842, "grad_norm": 52.71277618408203, "learning_rate": 2.900904605263158e-06, "loss": 0.2012, "step": 3453 }, { "epoch": 0.7101151315789473, "grad_norm": 4.968891620635986, "learning_rate": 2.8988486842105267e-06, "loss": 0.0059, "step": 3454 }, { "epoch": 0.7103207236842105, "grad_norm": 62.884796142578125, "learning_rate": 2.896792763157895e-06, "loss": 0.1716, "step": 3455 }, { "epoch": 0.7105263157894737, "grad_norm": 211.1700439453125, "learning_rate": 2.8947368421052634e-06, "loss": 1.1748, "step": 3456 }, { "epoch": 0.7107319078947368, "grad_norm": 23.24755859375, "learning_rate": 2.892680921052632e-06, "loss": 0.0555, "step": 3457 }, { "epoch": 0.7109375, "grad_norm": 71.75526428222656, "learning_rate": 2.8906250000000004e-06, "loss": 0.2969, "step": 3458 }, { "epoch": 0.7111430921052632, "grad_norm": 34.74869918823242, "learning_rate": 2.8885690789473687e-06, "loss": 0.1101, "step": 3459 }, { "epoch": 0.7113486842105263, "grad_norm": 191.77093505859375, "learning_rate": 2.886513157894737e-06, "loss": 0.9531, "step": 3460 }, { "epoch": 0.7115542763157895, "grad_norm": 19.585506439208984, "learning_rate": 2.8844572368421053e-06, "loss": 0.0979, "step": 3461 }, { "epoch": 0.7117598684210527, "grad_norm": 423.5087585449219, "learning_rate": 2.8824013157894736e-06, "loss": 1.9216, "step": 3462 }, { "epoch": 0.7119654605263158, "grad_norm": 36.591835021972656, "learning_rate": 2.8803453947368428e-06, "loss": 0.2584, "step": 3463 }, { "epoch": 0.712171052631579, "grad_norm": 21.901615142822266, "learning_rate": 2.878289473684211e-06, "loss": 0.049, "step": 3464 }, { "epoch": 0.7123766447368421, "grad_norm": 21.537960052490234, "learning_rate": 2.8762335526315794e-06, "loss": 0.1189, "step": 3465 }, { "epoch": 0.7125822368421053, "grad_norm": 45.33368682861328, "learning_rate": 2.8741776315789477e-06, "loss": 0.2041, "step": 3466 }, { "epoch": 0.7127878289473685, "grad_norm": 5.358014106750488, "learning_rate": 2.872121710526316e-06, "loss": 0.0101, "step": 3467 }, { "epoch": 0.7129934210526315, "grad_norm": 263.60577392578125, "learning_rate": 2.8700657894736843e-06, "loss": 4.8862, "step": 3468 }, { "epoch": 0.7131990131578947, "grad_norm": 8.156244277954102, "learning_rate": 2.8680098684210526e-06, "loss": 0.045, "step": 3469 }, { "epoch": 0.7134046052631579, "grad_norm": 78.88241577148438, "learning_rate": 2.8659539473684213e-06, "loss": 0.9043, "step": 3470 }, { "epoch": 0.713610197368421, "grad_norm": 143.46307373046875, "learning_rate": 2.8638980263157896e-06, "loss": 1.1904, "step": 3471 }, { "epoch": 0.7138157894736842, "grad_norm": 59.15509033203125, "learning_rate": 2.861842105263158e-06, "loss": 0.3609, "step": 3472 }, { "epoch": 0.7140213815789473, "grad_norm": 93.00202941894531, "learning_rate": 2.8597861842105266e-06, "loss": 0.5493, "step": 3473 }, { "epoch": 0.7142269736842105, "grad_norm": 337.72308349609375, "learning_rate": 2.857730263157895e-06, "loss": 1.1934, "step": 3474 }, { "epoch": 0.7144325657894737, "grad_norm": 24.73811149597168, "learning_rate": 2.8556743421052637e-06, "loss": 0.0562, "step": 3475 }, { "epoch": 0.7146381578947368, "grad_norm": 43.676544189453125, "learning_rate": 2.853618421052632e-06, "loss": 0.094, "step": 3476 }, { "epoch": 0.71484375, "grad_norm": 129.22911071777344, "learning_rate": 2.8515625000000003e-06, "loss": 0.6348, "step": 3477 }, { "epoch": 0.7150493421052632, "grad_norm": 440.61846923828125, "learning_rate": 2.8495065789473686e-06, "loss": 2.6523, "step": 3478 }, { "epoch": 0.7152549342105263, "grad_norm": 55.753334045410156, "learning_rate": 2.847450657894737e-06, "loss": 0.8281, "step": 3479 }, { "epoch": 0.7154605263157895, "grad_norm": 38.434024810791016, "learning_rate": 2.845394736842105e-06, "loss": 0.1221, "step": 3480 }, { "epoch": 0.7156661184210527, "grad_norm": 40.777008056640625, "learning_rate": 2.8433388157894735e-06, "loss": 0.1248, "step": 3481 }, { "epoch": 0.7158717105263158, "grad_norm": 17.7447509765625, "learning_rate": 2.8412828947368426e-06, "loss": 0.0593, "step": 3482 }, { "epoch": 0.716077302631579, "grad_norm": 104.90145874023438, "learning_rate": 2.839226973684211e-06, "loss": 0.4014, "step": 3483 }, { "epoch": 0.7162828947368421, "grad_norm": 297.40972900390625, "learning_rate": 2.8371710526315793e-06, "loss": 1.6514, "step": 3484 }, { "epoch": 0.7164884868421053, "grad_norm": 71.3394775390625, "learning_rate": 2.8351151315789476e-06, "loss": 0.3811, "step": 3485 }, { "epoch": 0.7166940789473685, "grad_norm": 33.056732177734375, "learning_rate": 2.833059210526316e-06, "loss": 0.1337, "step": 3486 }, { "epoch": 0.7168996710526315, "grad_norm": 55.672523498535156, "learning_rate": 2.8310032894736846e-06, "loss": 0.3164, "step": 3487 }, { "epoch": 0.7171052631578947, "grad_norm": 212.36622619628906, "learning_rate": 2.828947368421053e-06, "loss": 1.0544, "step": 3488 }, { "epoch": 0.7173108552631579, "grad_norm": 54.09208297729492, "learning_rate": 2.826891447368421e-06, "loss": 0.2546, "step": 3489 }, { "epoch": 0.717516447368421, "grad_norm": 16.458982467651367, "learning_rate": 2.8248355263157895e-06, "loss": 0.04, "step": 3490 }, { "epoch": 0.7177220394736842, "grad_norm": 100.16616821289062, "learning_rate": 2.822779605263158e-06, "loss": 0.7324, "step": 3491 }, { "epoch": 0.7179276315789473, "grad_norm": 37.36361312866211, "learning_rate": 2.820723684210527e-06, "loss": 0.1404, "step": 3492 }, { "epoch": 0.7181332236842105, "grad_norm": 121.71073150634766, "learning_rate": 2.8186677631578953e-06, "loss": 1.692, "step": 3493 }, { "epoch": 0.7183388157894737, "grad_norm": 13.033686637878418, "learning_rate": 2.8166118421052636e-06, "loss": 0.0334, "step": 3494 }, { "epoch": 0.7185444078947368, "grad_norm": 101.7873306274414, "learning_rate": 2.814555921052632e-06, "loss": 0.4283, "step": 3495 }, { "epoch": 0.71875, "grad_norm": 9.65357780456543, "learning_rate": 2.8125e-06, "loss": 0.0135, "step": 3496 }, { "epoch": 0.7189555921052632, "grad_norm": 99.15284729003906, "learning_rate": 2.8104440789473685e-06, "loss": 1.0348, "step": 3497 }, { "epoch": 0.7191611842105263, "grad_norm": 88.98593139648438, "learning_rate": 2.8083881578947368e-06, "loss": 0.4866, "step": 3498 }, { "epoch": 0.7193667763157895, "grad_norm": 27.259063720703125, "learning_rate": 2.8063322368421055e-06, "loss": 0.0575, "step": 3499 }, { "epoch": 0.7195723684210527, "grad_norm": 21.776153564453125, "learning_rate": 2.804276315789474e-06, "loss": 0.0797, "step": 3500 }, { "epoch": 0.7197779605263158, "grad_norm": 63.541080474853516, "learning_rate": 2.8022203947368425e-06, "loss": 0.2539, "step": 3501 }, { "epoch": 0.719983552631579, "grad_norm": 3.156515598297119, "learning_rate": 2.800164473684211e-06, "loss": 0.0134, "step": 3502 }, { "epoch": 0.7201891447368421, "grad_norm": 38.587730407714844, "learning_rate": 2.798108552631579e-06, "loss": 0.249, "step": 3503 }, { "epoch": 0.7203947368421053, "grad_norm": 7.554232597351074, "learning_rate": 2.796052631578948e-06, "loss": 0.0236, "step": 3504 }, { "epoch": 0.7206003289473685, "grad_norm": 58.962772369384766, "learning_rate": 2.793996710526316e-06, "loss": 0.2776, "step": 3505 }, { "epoch": 0.7208059210526315, "grad_norm": 200.5242156982422, "learning_rate": 2.7919407894736845e-06, "loss": 1.7852, "step": 3506 }, { "epoch": 0.7210115131578947, "grad_norm": 56.429718017578125, "learning_rate": 2.789884868421053e-06, "loss": 0.2196, "step": 3507 }, { "epoch": 0.7212171052631579, "grad_norm": 103.68309783935547, "learning_rate": 2.787828947368421e-06, "loss": 0.7329, "step": 3508 }, { "epoch": 0.721422697368421, "grad_norm": 210.6591796875, "learning_rate": 2.7857730263157894e-06, "loss": 1.2261, "step": 3509 }, { "epoch": 0.7216282894736842, "grad_norm": 20.61272430419922, "learning_rate": 2.7837171052631577e-06, "loss": 0.0481, "step": 3510 }, { "epoch": 0.7218338815789473, "grad_norm": 24.65221405029297, "learning_rate": 2.781661184210527e-06, "loss": 0.1348, "step": 3511 }, { "epoch": 0.7220394736842105, "grad_norm": 189.09934997558594, "learning_rate": 2.779605263157895e-06, "loss": 0.3909, "step": 3512 }, { "epoch": 0.7222450657894737, "grad_norm": 96.51495361328125, "learning_rate": 2.7775493421052635e-06, "loss": 0.5232, "step": 3513 }, { "epoch": 0.7224506578947368, "grad_norm": 50.474552154541016, "learning_rate": 2.7754934210526318e-06, "loss": 0.1285, "step": 3514 }, { "epoch": 0.72265625, "grad_norm": 175.14212036132812, "learning_rate": 2.7734375e-06, "loss": 0.4199, "step": 3515 }, { "epoch": 0.7228618421052632, "grad_norm": 13.405654907226562, "learning_rate": 2.7713815789473684e-06, "loss": 0.0316, "step": 3516 }, { "epoch": 0.7230674342105263, "grad_norm": 144.1034393310547, "learning_rate": 2.769325657894737e-06, "loss": 0.5535, "step": 3517 }, { "epoch": 0.7232730263157895, "grad_norm": 36.74195098876953, "learning_rate": 2.7672697368421054e-06, "loss": 0.1167, "step": 3518 }, { "epoch": 0.7234786184210527, "grad_norm": 97.91667938232422, "learning_rate": 2.7652138157894737e-06, "loss": 0.4596, "step": 3519 }, { "epoch": 0.7236842105263158, "grad_norm": 74.33655548095703, "learning_rate": 2.7631578947368424e-06, "loss": 1.418, "step": 3520 }, { "epoch": 0.723889802631579, "grad_norm": 174.85791015625, "learning_rate": 2.7611019736842107e-06, "loss": 1.4727, "step": 3521 }, { "epoch": 0.7240953947368421, "grad_norm": 240.42657470703125, "learning_rate": 2.7590460526315795e-06, "loss": 1.23, "step": 3522 }, { "epoch": 0.7243009868421053, "grad_norm": 6.222539901733398, "learning_rate": 2.7569901315789478e-06, "loss": 0.0088, "step": 3523 }, { "epoch": 0.7245065789473685, "grad_norm": 106.07415008544922, "learning_rate": 2.754934210526316e-06, "loss": 0.603, "step": 3524 }, { "epoch": 0.7247121710526315, "grad_norm": 31.83077621459961, "learning_rate": 2.7528782894736844e-06, "loss": 0.1223, "step": 3525 }, { "epoch": 0.7249177631578947, "grad_norm": 82.22484588623047, "learning_rate": 2.7508223684210527e-06, "loss": 0.8291, "step": 3526 }, { "epoch": 0.7251233552631579, "grad_norm": 88.62938690185547, "learning_rate": 2.748766447368421e-06, "loss": 0.3682, "step": 3527 }, { "epoch": 0.725328947368421, "grad_norm": 204.2471923828125, "learning_rate": 2.7467105263157893e-06, "loss": 1.5898, "step": 3528 }, { "epoch": 0.7255345394736842, "grad_norm": 272.6456298828125, "learning_rate": 2.744654605263158e-06, "loss": 2.1332, "step": 3529 }, { "epoch": 0.7257401315789473, "grad_norm": 33.14569854736328, "learning_rate": 2.7425986842105268e-06, "loss": 0.1508, "step": 3530 }, { "epoch": 0.7259457236842105, "grad_norm": 15.737799644470215, "learning_rate": 2.740542763157895e-06, "loss": 0.0587, "step": 3531 }, { "epoch": 0.7261513157894737, "grad_norm": 56.50210952758789, "learning_rate": 2.7384868421052634e-06, "loss": 0.2654, "step": 3532 }, { "epoch": 0.7263569078947368, "grad_norm": 27.312339782714844, "learning_rate": 2.7364309210526317e-06, "loss": 0.1163, "step": 3533 }, { "epoch": 0.7265625, "grad_norm": 107.76058197021484, "learning_rate": 2.7343750000000004e-06, "loss": 0.4219, "step": 3534 }, { "epoch": 0.7267680921052632, "grad_norm": 83.24580383300781, "learning_rate": 2.7323190789473687e-06, "loss": 1.1077, "step": 3535 }, { "epoch": 0.7269736842105263, "grad_norm": 10.951513290405273, "learning_rate": 2.730263157894737e-06, "loss": 0.0202, "step": 3536 }, { "epoch": 0.7271792763157895, "grad_norm": 15.630464553833008, "learning_rate": 2.7282072368421053e-06, "loss": 0.0377, "step": 3537 }, { "epoch": 0.7273848684210527, "grad_norm": 3.94260311126709, "learning_rate": 2.7261513157894736e-06, "loss": 0.0124, "step": 3538 }, { "epoch": 0.7275904605263158, "grad_norm": 26.023653030395508, "learning_rate": 2.7240953947368428e-06, "loss": 0.0831, "step": 3539 }, { "epoch": 0.727796052631579, "grad_norm": 65.11114501953125, "learning_rate": 2.722039473684211e-06, "loss": 0.1526, "step": 3540 }, { "epoch": 0.7280016447368421, "grad_norm": 54.11731719970703, "learning_rate": 2.7199835526315794e-06, "loss": 0.123, "step": 3541 }, { "epoch": 0.7282072368421053, "grad_norm": 71.61018371582031, "learning_rate": 2.7179276315789477e-06, "loss": 0.5086, "step": 3542 }, { "epoch": 0.7284128289473685, "grad_norm": 13.918490409851074, "learning_rate": 2.715871710526316e-06, "loss": 0.0324, "step": 3543 }, { "epoch": 0.7286184210526315, "grad_norm": 31.91781997680664, "learning_rate": 2.7138157894736843e-06, "loss": 0.1582, "step": 3544 }, { "epoch": 0.7288240131578947, "grad_norm": 98.08060455322266, "learning_rate": 2.7117598684210526e-06, "loss": 0.8813, "step": 3545 }, { "epoch": 0.7290296052631579, "grad_norm": 5.447727680206299, "learning_rate": 2.7097039473684213e-06, "loss": 0.0359, "step": 3546 }, { "epoch": 0.729235197368421, "grad_norm": 76.70152282714844, "learning_rate": 2.7076480263157896e-06, "loss": 0.3291, "step": 3547 }, { "epoch": 0.7294407894736842, "grad_norm": 121.20935821533203, "learning_rate": 2.705592105263158e-06, "loss": 0.6357, "step": 3548 }, { "epoch": 0.7296463815789473, "grad_norm": 111.73406982421875, "learning_rate": 2.7035361842105267e-06, "loss": 0.8867, "step": 3549 }, { "epoch": 0.7298519736842105, "grad_norm": 50.18510818481445, "learning_rate": 2.701480263157895e-06, "loss": 0.6136, "step": 3550 }, { "epoch": 0.7300575657894737, "grad_norm": 10.090394973754883, "learning_rate": 2.6994243421052637e-06, "loss": 0.0231, "step": 3551 }, { "epoch": 0.7302631578947368, "grad_norm": 34.66122817993164, "learning_rate": 2.697368421052632e-06, "loss": 0.082, "step": 3552 }, { "epoch": 0.73046875, "grad_norm": 53.39572525024414, "learning_rate": 2.6953125000000003e-06, "loss": 0.2875, "step": 3553 }, { "epoch": 0.7306743421052632, "grad_norm": 22.112354278564453, "learning_rate": 2.6932565789473686e-06, "loss": 0.0464, "step": 3554 }, { "epoch": 0.7308799342105263, "grad_norm": 72.07640075683594, "learning_rate": 2.691200657894737e-06, "loss": 0.8735, "step": 3555 }, { "epoch": 0.7310855263157895, "grad_norm": 329.8201599121094, "learning_rate": 2.689144736842105e-06, "loss": 0.752, "step": 3556 }, { "epoch": 0.7312911184210527, "grad_norm": 74.36724853515625, "learning_rate": 2.6870888157894735e-06, "loss": 0.2097, "step": 3557 }, { "epoch": 0.7314967105263158, "grad_norm": 160.5743408203125, "learning_rate": 2.6850328947368427e-06, "loss": 1.0132, "step": 3558 }, { "epoch": 0.731702302631579, "grad_norm": 204.90911865234375, "learning_rate": 2.682976973684211e-06, "loss": 2.168, "step": 3559 }, { "epoch": 0.7319078947368421, "grad_norm": 10.122312545776367, "learning_rate": 2.6809210526315793e-06, "loss": 0.0176, "step": 3560 }, { "epoch": 0.7321134868421053, "grad_norm": 71.72776794433594, "learning_rate": 2.6788651315789476e-06, "loss": 0.1299, "step": 3561 }, { "epoch": 0.7323190789473685, "grad_norm": 107.17243194580078, "learning_rate": 2.676809210526316e-06, "loss": 0.5098, "step": 3562 }, { "epoch": 0.7325246710526315, "grad_norm": 49.035675048828125, "learning_rate": 2.6747532894736846e-06, "loss": 0.1643, "step": 3563 }, { "epoch": 0.7327302631578947, "grad_norm": 58.070762634277344, "learning_rate": 2.672697368421053e-06, "loss": 0.1233, "step": 3564 }, { "epoch": 0.7329358552631579, "grad_norm": 64.98745727539062, "learning_rate": 2.6706414473684212e-06, "loss": 0.1855, "step": 3565 }, { "epoch": 0.733141447368421, "grad_norm": 43.357181549072266, "learning_rate": 2.6685855263157895e-06, "loss": 0.2246, "step": 3566 }, { "epoch": 0.7333470394736842, "grad_norm": 91.11510467529297, "learning_rate": 2.666529605263158e-06, "loss": 0.3369, "step": 3567 }, { "epoch": 0.7335526315789473, "grad_norm": 78.26539611816406, "learning_rate": 2.664473684210527e-06, "loss": 1.4167, "step": 3568 }, { "epoch": 0.7337582236842105, "grad_norm": 73.10590362548828, "learning_rate": 2.6624177631578953e-06, "loss": 0.4468, "step": 3569 }, { "epoch": 0.7339638157894737, "grad_norm": 115.01590728759766, "learning_rate": 2.6603618421052636e-06, "loss": 0.5938, "step": 3570 }, { "epoch": 0.7341694078947368, "grad_norm": 35.57275390625, "learning_rate": 2.658305921052632e-06, "loss": 0.1718, "step": 3571 }, { "epoch": 0.734375, "grad_norm": 99.73725128173828, "learning_rate": 2.65625e-06, "loss": 0.6006, "step": 3572 }, { "epoch": 0.7345805921052632, "grad_norm": 314.8186950683594, "learning_rate": 2.6541940789473685e-06, "loss": 2.3789, "step": 3573 }, { "epoch": 0.7347861842105263, "grad_norm": 186.60052490234375, "learning_rate": 2.652138157894737e-06, "loss": 1.02, "step": 3574 }, { "epoch": 0.7349917763157895, "grad_norm": 22.820024490356445, "learning_rate": 2.650082236842105e-06, "loss": 0.0712, "step": 3575 }, { "epoch": 0.7351973684210527, "grad_norm": 29.0562744140625, "learning_rate": 2.648026315789474e-06, "loss": 0.0754, "step": 3576 }, { "epoch": 0.7354029605263158, "grad_norm": 88.75892639160156, "learning_rate": 2.6459703947368426e-06, "loss": 0.5079, "step": 3577 }, { "epoch": 0.735608552631579, "grad_norm": 35.25148391723633, "learning_rate": 2.643914473684211e-06, "loss": 0.1662, "step": 3578 }, { "epoch": 0.7358141447368421, "grad_norm": 8.62734603881836, "learning_rate": 2.641858552631579e-06, "loss": 0.0155, "step": 3579 }, { "epoch": 0.7360197368421053, "grad_norm": 6.588563919067383, "learning_rate": 2.6398026315789475e-06, "loss": 0.0162, "step": 3580 }, { "epoch": 0.7362253289473685, "grad_norm": 121.97249603271484, "learning_rate": 2.637746710526316e-06, "loss": 0.3867, "step": 3581 }, { "epoch": 0.7364309210526315, "grad_norm": 76.69316864013672, "learning_rate": 2.6356907894736845e-06, "loss": 1.4048, "step": 3582 }, { "epoch": 0.7366365131578947, "grad_norm": 148.37229919433594, "learning_rate": 2.633634868421053e-06, "loss": 0.9023, "step": 3583 }, { "epoch": 0.7368421052631579, "grad_norm": 3.2815656661987305, "learning_rate": 2.631578947368421e-06, "loss": 0.0172, "step": 3584 }, { "epoch": 0.737047697368421, "grad_norm": 13.676870346069336, "learning_rate": 2.6295230263157894e-06, "loss": 0.0385, "step": 3585 }, { "epoch": 0.7372532894736842, "grad_norm": 148.768798828125, "learning_rate": 2.6274671052631577e-06, "loss": 0.6285, "step": 3586 }, { "epoch": 0.7374588815789473, "grad_norm": 45.271514892578125, "learning_rate": 2.625411184210527e-06, "loss": 0.1931, "step": 3587 }, { "epoch": 0.7376644736842105, "grad_norm": 7.405820369720459, "learning_rate": 2.623355263157895e-06, "loss": 0.0169, "step": 3588 }, { "epoch": 0.7378700657894737, "grad_norm": 30.360132217407227, "learning_rate": 2.6212993421052635e-06, "loss": 0.1371, "step": 3589 }, { "epoch": 0.7380756578947368, "grad_norm": 4.113637447357178, "learning_rate": 2.6192434210526318e-06, "loss": 0.0205, "step": 3590 }, { "epoch": 0.73828125, "grad_norm": 9.163910865783691, "learning_rate": 2.6171875e-06, "loss": 0.0345, "step": 3591 }, { "epoch": 0.7384868421052632, "grad_norm": 14.367008209228516, "learning_rate": 2.6151315789473684e-06, "loss": 0.0514, "step": 3592 }, { "epoch": 0.7386924342105263, "grad_norm": 57.50127410888672, "learning_rate": 2.613075657894737e-06, "loss": 0.1506, "step": 3593 }, { "epoch": 0.7388980263157895, "grad_norm": 3.6257567405700684, "learning_rate": 2.6110197368421054e-06, "loss": 0.0102, "step": 3594 }, { "epoch": 0.7391036184210527, "grad_norm": 80.42253875732422, "learning_rate": 2.6089638157894737e-06, "loss": 0.3645, "step": 3595 }, { "epoch": 0.7393092105263158, "grad_norm": 5.917730331420898, "learning_rate": 2.6069078947368425e-06, "loss": 0.0254, "step": 3596 }, { "epoch": 0.739514802631579, "grad_norm": 34.62810516357422, "learning_rate": 2.6048519736842108e-06, "loss": 0.1876, "step": 3597 }, { "epoch": 0.7397203947368421, "grad_norm": 70.89485168457031, "learning_rate": 2.6027960526315795e-06, "loss": 1.1038, "step": 3598 }, { "epoch": 0.7399259868421053, "grad_norm": 114.81167602539062, "learning_rate": 2.600740131578948e-06, "loss": 0.4531, "step": 3599 }, { "epoch": 0.7401315789473685, "grad_norm": 23.8623046875, "learning_rate": 2.598684210526316e-06, "loss": 0.0274, "step": 3600 }, { "epoch": 0.7403371710526315, "grad_norm": 81.47771453857422, "learning_rate": 2.5966282894736844e-06, "loss": 0.2928, "step": 3601 }, { "epoch": 0.7405427631578947, "grad_norm": 53.876609802246094, "learning_rate": 2.5945723684210527e-06, "loss": 0.2363, "step": 3602 }, { "epoch": 0.7407483552631579, "grad_norm": 29.992258071899414, "learning_rate": 2.592516447368421e-06, "loss": 0.0566, "step": 3603 }, { "epoch": 0.740953947368421, "grad_norm": 25.008996963500977, "learning_rate": 2.5904605263157893e-06, "loss": 0.0909, "step": 3604 }, { "epoch": 0.7411595394736842, "grad_norm": 79.22197723388672, "learning_rate": 2.588404605263158e-06, "loss": 0.7461, "step": 3605 }, { "epoch": 0.7413651315789473, "grad_norm": 24.1767635345459, "learning_rate": 2.5863486842105268e-06, "loss": 0.0635, "step": 3606 }, { "epoch": 0.7415707236842105, "grad_norm": 77.07569885253906, "learning_rate": 2.584292763157895e-06, "loss": 0.7744, "step": 3607 }, { "epoch": 0.7417763157894737, "grad_norm": 122.12882995605469, "learning_rate": 2.5822368421052634e-06, "loss": 0.9987, "step": 3608 }, { "epoch": 0.7419819078947368, "grad_norm": 47.41072463989258, "learning_rate": 2.5801809210526317e-06, "loss": 0.0942, "step": 3609 }, { "epoch": 0.7421875, "grad_norm": 24.635000228881836, "learning_rate": 2.5781250000000004e-06, "loss": 0.1145, "step": 3610 }, { "epoch": 0.7423930921052632, "grad_norm": 23.507518768310547, "learning_rate": 2.5760690789473687e-06, "loss": 0.1455, "step": 3611 }, { "epoch": 0.7425986842105263, "grad_norm": 45.55577087402344, "learning_rate": 2.574013157894737e-06, "loss": 0.1083, "step": 3612 }, { "epoch": 0.7428042763157895, "grad_norm": 40.297611236572266, "learning_rate": 2.5719572368421053e-06, "loss": 0.2139, "step": 3613 }, { "epoch": 0.7430098684210527, "grad_norm": 194.46791076660156, "learning_rate": 2.5699013157894736e-06, "loss": 1.9496, "step": 3614 }, { "epoch": 0.7432154605263158, "grad_norm": 29.334068298339844, "learning_rate": 2.5678453947368428e-06, "loss": 0.1233, "step": 3615 }, { "epoch": 0.743421052631579, "grad_norm": 16.704179763793945, "learning_rate": 2.565789473684211e-06, "loss": 0.0341, "step": 3616 }, { "epoch": 0.7436266447368421, "grad_norm": 22.17168426513672, "learning_rate": 2.5637335526315794e-06, "loss": 0.0551, "step": 3617 }, { "epoch": 0.7438322368421053, "grad_norm": 61.32476043701172, "learning_rate": 2.5616776315789477e-06, "loss": 0.699, "step": 3618 }, { "epoch": 0.7440378289473685, "grad_norm": 26.660017013549805, "learning_rate": 2.559621710526316e-06, "loss": 0.0849, "step": 3619 }, { "epoch": 0.7442434210526315, "grad_norm": 52.67859649658203, "learning_rate": 2.5575657894736843e-06, "loss": 0.1962, "step": 3620 }, { "epoch": 0.7444490131578947, "grad_norm": 107.78015899658203, "learning_rate": 2.5555098684210526e-06, "loss": 0.4426, "step": 3621 }, { "epoch": 0.7446546052631579, "grad_norm": 32.806365966796875, "learning_rate": 2.553453947368421e-06, "loss": 0.1072, "step": 3622 }, { "epoch": 0.744860197368421, "grad_norm": 115.70319366455078, "learning_rate": 2.5513980263157896e-06, "loss": 0.5883, "step": 3623 }, { "epoch": 0.7450657894736842, "grad_norm": 30.400981903076172, "learning_rate": 2.549342105263158e-06, "loss": 0.0898, "step": 3624 }, { "epoch": 0.7452713815789473, "grad_norm": 111.27259063720703, "learning_rate": 2.5472861842105267e-06, "loss": 0.6089, "step": 3625 }, { "epoch": 0.7454769736842105, "grad_norm": 346.755859375, "learning_rate": 2.545230263157895e-06, "loss": 4.0625, "step": 3626 }, { "epoch": 0.7456825657894737, "grad_norm": 5.8636155128479, "learning_rate": 2.5431743421052633e-06, "loss": 0.0067, "step": 3627 }, { "epoch": 0.7458881578947368, "grad_norm": 82.44305419921875, "learning_rate": 2.541118421052632e-06, "loss": 0.7715, "step": 3628 }, { "epoch": 0.74609375, "grad_norm": 84.36956024169922, "learning_rate": 2.5390625000000003e-06, "loss": 0.2352, "step": 3629 }, { "epoch": 0.7462993421052632, "grad_norm": 11.801691055297852, "learning_rate": 2.5370065789473686e-06, "loss": 0.0466, "step": 3630 }, { "epoch": 0.7465049342105263, "grad_norm": 25.151893615722656, "learning_rate": 2.534950657894737e-06, "loss": 0.0833, "step": 3631 }, { "epoch": 0.7467105263157895, "grad_norm": 3.3215866088867188, "learning_rate": 2.5328947368421052e-06, "loss": 0.005, "step": 3632 }, { "epoch": 0.7469161184210527, "grad_norm": 80.93511199951172, "learning_rate": 2.5308388157894735e-06, "loss": 0.325, "step": 3633 }, { "epoch": 0.7471217105263158, "grad_norm": 355.3158874511719, "learning_rate": 2.5287828947368427e-06, "loss": 1.8682, "step": 3634 }, { "epoch": 0.747327302631579, "grad_norm": 129.37911987304688, "learning_rate": 2.526726973684211e-06, "loss": 0.4479, "step": 3635 }, { "epoch": 0.7475328947368421, "grad_norm": 46.987953186035156, "learning_rate": 2.5246710526315793e-06, "loss": 0.2087, "step": 3636 }, { "epoch": 0.7477384868421053, "grad_norm": 261.3052062988281, "learning_rate": 2.5226151315789476e-06, "loss": 2.0547, "step": 3637 }, { "epoch": 0.7479440789473685, "grad_norm": 211.39614868164062, "learning_rate": 2.520559210526316e-06, "loss": 1.6131, "step": 3638 }, { "epoch": 0.7481496710526315, "grad_norm": 523.9588623046875, "learning_rate": 2.518503289473684e-06, "loss": 1.7373, "step": 3639 }, { "epoch": 0.7483552631578947, "grad_norm": 12.58764362335205, "learning_rate": 2.516447368421053e-06, "loss": 0.0293, "step": 3640 }, { "epoch": 0.7485608552631579, "grad_norm": 102.91284942626953, "learning_rate": 2.5143914473684212e-06, "loss": 0.6245, "step": 3641 }, { "epoch": 0.748766447368421, "grad_norm": 25.132051467895508, "learning_rate": 2.5123355263157895e-06, "loss": 0.1309, "step": 3642 }, { "epoch": 0.7489720394736842, "grad_norm": 19.492740631103516, "learning_rate": 2.510279605263158e-06, "loss": 0.095, "step": 3643 }, { "epoch": 0.7491776315789473, "grad_norm": 262.0729064941406, "learning_rate": 2.5082236842105266e-06, "loss": 1.5469, "step": 3644 }, { "epoch": 0.7493832236842105, "grad_norm": 16.4080753326416, "learning_rate": 2.5061677631578953e-06, "loss": 0.0612, "step": 3645 }, { "epoch": 0.7495888157894737, "grad_norm": 89.43165588378906, "learning_rate": 2.5041118421052636e-06, "loss": 0.2809, "step": 3646 }, { "epoch": 0.7497944078947368, "grad_norm": 142.6821746826172, "learning_rate": 2.502055921052632e-06, "loss": 0.5645, "step": 3647 }, { "epoch": 0.75, "grad_norm": 23.560609817504883, "learning_rate": 2.5e-06, "loss": 0.1032, "step": 3648 }, { "epoch": 0.7502055921052632, "grad_norm": 4.308425426483154, "learning_rate": 2.4979440789473685e-06, "loss": 0.0202, "step": 3649 }, { "epoch": 0.7504111842105263, "grad_norm": 39.25981903076172, "learning_rate": 2.495888157894737e-06, "loss": 0.182, "step": 3650 }, { "epoch": 0.7506167763157895, "grad_norm": 15.598639488220215, "learning_rate": 2.4938322368421055e-06, "loss": 0.0731, "step": 3651 }, { "epoch": 0.7508223684210527, "grad_norm": 27.238496780395508, "learning_rate": 2.491776315789474e-06, "loss": 0.0891, "step": 3652 }, { "epoch": 0.7510279605263158, "grad_norm": 46.11358642578125, "learning_rate": 2.489720394736842e-06, "loss": 0.0813, "step": 3653 }, { "epoch": 0.751233552631579, "grad_norm": 71.89110565185547, "learning_rate": 2.4876644736842105e-06, "loss": 0.5137, "step": 3654 }, { "epoch": 0.7514391447368421, "grad_norm": 28.998519897460938, "learning_rate": 2.485608552631579e-06, "loss": 0.1768, "step": 3655 }, { "epoch": 0.7516447368421053, "grad_norm": 9.567795753479004, "learning_rate": 2.4835526315789475e-06, "loss": 0.0169, "step": 3656 }, { "epoch": 0.7518503289473685, "grad_norm": 154.23226928710938, "learning_rate": 2.4814967105263162e-06, "loss": 0.5511, "step": 3657 }, { "epoch": 0.7520559210526315, "grad_norm": 27.442045211791992, "learning_rate": 2.4794407894736845e-06, "loss": 0.116, "step": 3658 }, { "epoch": 0.7522615131578947, "grad_norm": 7.110412120819092, "learning_rate": 2.477384868421053e-06, "loss": 0.0276, "step": 3659 }, { "epoch": 0.7524671052631579, "grad_norm": 27.284408569335938, "learning_rate": 2.4753289473684215e-06, "loss": 0.1061, "step": 3660 }, { "epoch": 0.752672697368421, "grad_norm": 45.332191467285156, "learning_rate": 2.47327302631579e-06, "loss": 0.2397, "step": 3661 }, { "epoch": 0.7528782894736842, "grad_norm": 7.499504089355469, "learning_rate": 2.471217105263158e-06, "loss": 0.0068, "step": 3662 }, { "epoch": 0.7530838815789473, "grad_norm": 15.542688369750977, "learning_rate": 2.4691611842105265e-06, "loss": 0.075, "step": 3663 }, { "epoch": 0.7532894736842105, "grad_norm": 116.59285736083984, "learning_rate": 2.4671052631578948e-06, "loss": 0.2621, "step": 3664 }, { "epoch": 0.7534950657894737, "grad_norm": 37.049068450927734, "learning_rate": 2.4650493421052635e-06, "loss": 0.2651, "step": 3665 }, { "epoch": 0.7537006578947368, "grad_norm": 77.22190856933594, "learning_rate": 2.462993421052632e-06, "loss": 0.4657, "step": 3666 }, { "epoch": 0.75390625, "grad_norm": 22.6605167388916, "learning_rate": 2.4609375e-06, "loss": 0.1021, "step": 3667 }, { "epoch": 0.7541118421052632, "grad_norm": 61.8411979675293, "learning_rate": 2.4588815789473684e-06, "loss": 0.2627, "step": 3668 }, { "epoch": 0.7543174342105263, "grad_norm": 55.85305404663086, "learning_rate": 2.456825657894737e-06, "loss": 0.8221, "step": 3669 }, { "epoch": 0.7545230263157895, "grad_norm": 7.834507942199707, "learning_rate": 2.4547697368421054e-06, "loss": 0.0167, "step": 3670 }, { "epoch": 0.7547286184210527, "grad_norm": 126.93580627441406, "learning_rate": 2.4527138157894737e-06, "loss": 0.9307, "step": 3671 }, { "epoch": 0.7549342105263158, "grad_norm": 30.897869110107422, "learning_rate": 2.4506578947368425e-06, "loss": 0.0989, "step": 3672 }, { "epoch": 0.755139802631579, "grad_norm": 15.484679222106934, "learning_rate": 2.4486019736842108e-06, "loss": 0.0291, "step": 3673 }, { "epoch": 0.7553453947368421, "grad_norm": 113.77031707763672, "learning_rate": 2.446546052631579e-06, "loss": 0.668, "step": 3674 }, { "epoch": 0.7555509868421053, "grad_norm": 122.7127914428711, "learning_rate": 2.444490131578948e-06, "loss": 0.4673, "step": 3675 }, { "epoch": 0.7557565789473685, "grad_norm": 68.85379791259766, "learning_rate": 2.442434210526316e-06, "loss": 0.8042, "step": 3676 }, { "epoch": 0.7559621710526315, "grad_norm": 32.25138473510742, "learning_rate": 2.4403782894736844e-06, "loss": 0.144, "step": 3677 }, { "epoch": 0.7561677631578947, "grad_norm": 41.640869140625, "learning_rate": 2.4383223684210527e-06, "loss": 0.1362, "step": 3678 }, { "epoch": 0.7563733552631579, "grad_norm": 121.50898742675781, "learning_rate": 2.4362664473684214e-06, "loss": 0.627, "step": 3679 }, { "epoch": 0.756578947368421, "grad_norm": 49.86701965332031, "learning_rate": 2.4342105263157898e-06, "loss": 0.1652, "step": 3680 }, { "epoch": 0.7567845394736842, "grad_norm": 49.7274055480957, "learning_rate": 2.432154605263158e-06, "loss": 0.2446, "step": 3681 }, { "epoch": 0.7569901315789473, "grad_norm": 126.41793060302734, "learning_rate": 2.4300986842105264e-06, "loss": 0.3716, "step": 3682 }, { "epoch": 0.7571957236842105, "grad_norm": 26.694507598876953, "learning_rate": 2.4280427631578947e-06, "loss": 0.0629, "step": 3683 }, { "epoch": 0.7574013157894737, "grad_norm": 84.23800659179688, "learning_rate": 2.4259868421052634e-06, "loss": 0.5459, "step": 3684 }, { "epoch": 0.7576069078947368, "grad_norm": 59.0726432800293, "learning_rate": 2.4239309210526317e-06, "loss": 0.3193, "step": 3685 }, { "epoch": 0.7578125, "grad_norm": 105.40119934082031, "learning_rate": 2.421875e-06, "loss": 1.2497, "step": 3686 }, { "epoch": 0.7580180921052632, "grad_norm": 48.55855178833008, "learning_rate": 2.4198190789473687e-06, "loss": 0.1587, "step": 3687 }, { "epoch": 0.7582236842105263, "grad_norm": 50.2239875793457, "learning_rate": 2.417763157894737e-06, "loss": 0.3448, "step": 3688 }, { "epoch": 0.7584292763157895, "grad_norm": 106.05863952636719, "learning_rate": 2.4157072368421058e-06, "loss": 0.5498, "step": 3689 }, { "epoch": 0.7586348684210527, "grad_norm": 91.13721466064453, "learning_rate": 2.413651315789474e-06, "loss": 0.4648, "step": 3690 }, { "epoch": 0.7588404605263158, "grad_norm": 6.716926574707031, "learning_rate": 2.4115953947368424e-06, "loss": 0.0113, "step": 3691 }, { "epoch": 0.759046052631579, "grad_norm": 11.761219024658203, "learning_rate": 2.4095394736842107e-06, "loss": 0.0239, "step": 3692 }, { "epoch": 0.7592516447368421, "grad_norm": 16.428436279296875, "learning_rate": 2.407483552631579e-06, "loss": 0.0658, "step": 3693 }, { "epoch": 0.7594572368421053, "grad_norm": 444.9305725097656, "learning_rate": 2.4054276315789477e-06, "loss": 3.1964, "step": 3694 }, { "epoch": 0.7596628289473685, "grad_norm": 64.56708526611328, "learning_rate": 2.403371710526316e-06, "loss": 0.1855, "step": 3695 }, { "epoch": 0.7598684210526315, "grad_norm": 29.644567489624023, "learning_rate": 2.4013157894736843e-06, "loss": 0.0872, "step": 3696 }, { "epoch": 0.7600740131578947, "grad_norm": 36.71732711791992, "learning_rate": 2.3992598684210526e-06, "loss": 0.2646, "step": 3697 }, { "epoch": 0.7602796052631579, "grad_norm": 16.534637451171875, "learning_rate": 2.3972039473684213e-06, "loss": 0.0408, "step": 3698 }, { "epoch": 0.760485197368421, "grad_norm": 60.0391731262207, "learning_rate": 2.3951480263157896e-06, "loss": 0.3158, "step": 3699 }, { "epoch": 0.7606907894736842, "grad_norm": 205.5224151611328, "learning_rate": 2.393092105263158e-06, "loss": 0.7146, "step": 3700 }, { "epoch": 0.7608963815789473, "grad_norm": 3.0523593425750732, "learning_rate": 2.3910361842105263e-06, "loss": 0.0089, "step": 3701 }, { "epoch": 0.7611019736842105, "grad_norm": 181.88380432128906, "learning_rate": 2.388980263157895e-06, "loss": 0.7127, "step": 3702 }, { "epoch": 0.7613075657894737, "grad_norm": 5.415604114532471, "learning_rate": 2.3869243421052633e-06, "loss": 0.0217, "step": 3703 }, { "epoch": 0.7615131578947368, "grad_norm": 43.689781188964844, "learning_rate": 2.384868421052632e-06, "loss": 0.2739, "step": 3704 }, { "epoch": 0.76171875, "grad_norm": 11.874030113220215, "learning_rate": 2.3828125000000003e-06, "loss": 0.0468, "step": 3705 }, { "epoch": 0.7619243421052632, "grad_norm": 68.78587341308594, "learning_rate": 2.3807565789473686e-06, "loss": 0.2422, "step": 3706 }, { "epoch": 0.7621299342105263, "grad_norm": 17.78615379333496, "learning_rate": 2.378700657894737e-06, "loss": 0.0675, "step": 3707 }, { "epoch": 0.7623355263157895, "grad_norm": 94.36660766601562, "learning_rate": 2.3766447368421057e-06, "loss": 1.2983, "step": 3708 }, { "epoch": 0.7625411184210527, "grad_norm": 364.6266784667969, "learning_rate": 2.374588815789474e-06, "loss": 2.1748, "step": 3709 }, { "epoch": 0.7627467105263158, "grad_norm": 255.99229431152344, "learning_rate": 2.3725328947368423e-06, "loss": 3.0508, "step": 3710 }, { "epoch": 0.762952302631579, "grad_norm": 3.877934455871582, "learning_rate": 2.3704769736842106e-06, "loss": 0.0114, "step": 3711 }, { "epoch": 0.7631578947368421, "grad_norm": 1.7142843008041382, "learning_rate": 2.368421052631579e-06, "loss": 0.0037, "step": 3712 }, { "epoch": 0.7633634868421053, "grad_norm": 33.12556838989258, "learning_rate": 2.3663651315789476e-06, "loss": 0.0951, "step": 3713 }, { "epoch": 0.7635690789473685, "grad_norm": 298.64532470703125, "learning_rate": 2.364309210526316e-06, "loss": 1.3832, "step": 3714 }, { "epoch": 0.7637746710526315, "grad_norm": 198.05995178222656, "learning_rate": 2.362253289473684e-06, "loss": 0.9893, "step": 3715 }, { "epoch": 0.7639802631578947, "grad_norm": 14.813396453857422, "learning_rate": 2.360197368421053e-06, "loss": 0.059, "step": 3716 }, { "epoch": 0.7641858552631579, "grad_norm": 86.44794464111328, "learning_rate": 2.3581414473684212e-06, "loss": 0.6055, "step": 3717 }, { "epoch": 0.764391447368421, "grad_norm": 55.64730453491211, "learning_rate": 2.3560855263157895e-06, "loss": 0.2361, "step": 3718 }, { "epoch": 0.7645970394736842, "grad_norm": 27.401554107666016, "learning_rate": 2.3540296052631583e-06, "loss": 0.0885, "step": 3719 }, { "epoch": 0.7648026315789473, "grad_norm": 49.44086456298828, "learning_rate": 2.3519736842105266e-06, "loss": 0.3418, "step": 3720 }, { "epoch": 0.7650082236842105, "grad_norm": 28.689714431762695, "learning_rate": 2.349917763157895e-06, "loss": 0.1038, "step": 3721 }, { "epoch": 0.7652138157894737, "grad_norm": 77.34204864501953, "learning_rate": 2.3478618421052636e-06, "loss": 0.3382, "step": 3722 }, { "epoch": 0.7654194078947368, "grad_norm": 93.18431091308594, "learning_rate": 2.345805921052632e-06, "loss": 0.4719, "step": 3723 }, { "epoch": 0.765625, "grad_norm": 24.969600677490234, "learning_rate": 2.3437500000000002e-06, "loss": 0.1738, "step": 3724 }, { "epoch": 0.7658305921052632, "grad_norm": 18.33623695373535, "learning_rate": 2.3416940789473685e-06, "loss": 0.0715, "step": 3725 }, { "epoch": 0.7660361842105263, "grad_norm": 62.83742904663086, "learning_rate": 2.339638157894737e-06, "loss": 0.4365, "step": 3726 }, { "epoch": 0.7662417763157895, "grad_norm": 104.71324157714844, "learning_rate": 2.3375822368421056e-06, "loss": 0.5405, "step": 3727 }, { "epoch": 0.7664473684210527, "grad_norm": 208.43687438964844, "learning_rate": 2.335526315789474e-06, "loss": 1.5664, "step": 3728 }, { "epoch": 0.7666529605263158, "grad_norm": 80.03944396972656, "learning_rate": 2.333470394736842e-06, "loss": 0.3813, "step": 3729 }, { "epoch": 0.766858552631579, "grad_norm": 215.00926208496094, "learning_rate": 2.3314144736842105e-06, "loss": 0.7446, "step": 3730 }, { "epoch": 0.7670641447368421, "grad_norm": 160.57859802246094, "learning_rate": 2.329358552631579e-06, "loss": 0.6553, "step": 3731 }, { "epoch": 0.7672697368421053, "grad_norm": 90.25740814208984, "learning_rate": 2.3273026315789475e-06, "loss": 0.5532, "step": 3732 }, { "epoch": 0.7674753289473685, "grad_norm": 41.61115264892578, "learning_rate": 2.325246710526316e-06, "loss": 0.0955, "step": 3733 }, { "epoch": 0.7676809210526315, "grad_norm": 19.293258666992188, "learning_rate": 2.3231907894736845e-06, "loss": 0.0681, "step": 3734 }, { "epoch": 0.7678865131578947, "grad_norm": 80.14214324951172, "learning_rate": 2.321134868421053e-06, "loss": 0.751, "step": 3735 }, { "epoch": 0.7680921052631579, "grad_norm": 185.5877685546875, "learning_rate": 2.3190789473684216e-06, "loss": 0.6794, "step": 3736 }, { "epoch": 0.768297697368421, "grad_norm": 27.078731536865234, "learning_rate": 2.31702302631579e-06, "loss": 0.116, "step": 3737 }, { "epoch": 0.7685032894736842, "grad_norm": 171.78616333007812, "learning_rate": 2.314967105263158e-06, "loss": 0.6105, "step": 3738 }, { "epoch": 0.7687088815789473, "grad_norm": 33.198143005371094, "learning_rate": 2.3129111842105265e-06, "loss": 0.0867, "step": 3739 }, { "epoch": 0.7689144736842105, "grad_norm": 106.36891174316406, "learning_rate": 2.3108552631578948e-06, "loss": 0.9873, "step": 3740 }, { "epoch": 0.7691200657894737, "grad_norm": 53.12094497680664, "learning_rate": 2.3087993421052635e-06, "loss": 0.1764, "step": 3741 }, { "epoch": 0.7693256578947368, "grad_norm": 34.8367805480957, "learning_rate": 2.306743421052632e-06, "loss": 0.1685, "step": 3742 }, { "epoch": 0.76953125, "grad_norm": 35.026458740234375, "learning_rate": 2.3046875e-06, "loss": 0.2058, "step": 3743 }, { "epoch": 0.7697368421052632, "grad_norm": 60.58132553100586, "learning_rate": 2.3026315789473684e-06, "loss": 0.2646, "step": 3744 }, { "epoch": 0.7699424342105263, "grad_norm": 35.54399490356445, "learning_rate": 2.3005756578947367e-06, "loss": 0.125, "step": 3745 }, { "epoch": 0.7701480263157895, "grad_norm": 1.980395793914795, "learning_rate": 2.2985197368421055e-06, "loss": 0.0096, "step": 3746 }, { "epoch": 0.7703536184210527, "grad_norm": 135.00328063964844, "learning_rate": 2.2964638157894738e-06, "loss": 0.9508, "step": 3747 }, { "epoch": 0.7705592105263158, "grad_norm": 9.887651443481445, "learning_rate": 2.2944078947368425e-06, "loss": 0.0166, "step": 3748 }, { "epoch": 0.770764802631579, "grad_norm": 27.322999954223633, "learning_rate": 2.2923519736842108e-06, "loss": 0.2107, "step": 3749 }, { "epoch": 0.7709703947368421, "grad_norm": 13.181171417236328, "learning_rate": 2.290296052631579e-06, "loss": 0.0443, "step": 3750 }, { "epoch": 0.7711759868421053, "grad_norm": 20.91106605529785, "learning_rate": 2.288240131578948e-06, "loss": 0.0901, "step": 3751 }, { "epoch": 0.7713815789473685, "grad_norm": 51.37228775024414, "learning_rate": 2.286184210526316e-06, "loss": 0.1686, "step": 3752 }, { "epoch": 0.7715871710526315, "grad_norm": 25.480112075805664, "learning_rate": 2.2841282894736844e-06, "loss": 0.11, "step": 3753 }, { "epoch": 0.7717927631578947, "grad_norm": 115.10031127929688, "learning_rate": 2.2820723684210527e-06, "loss": 0.4253, "step": 3754 }, { "epoch": 0.7719983552631579, "grad_norm": 12.002701759338379, "learning_rate": 2.2800164473684215e-06, "loss": 0.0511, "step": 3755 }, { "epoch": 0.772203947368421, "grad_norm": 23.657726287841797, "learning_rate": 2.2779605263157898e-06, "loss": 0.1013, "step": 3756 }, { "epoch": 0.7724095394736842, "grad_norm": 15.579713821411133, "learning_rate": 2.275904605263158e-06, "loss": 0.0743, "step": 3757 }, { "epoch": 0.7726151315789473, "grad_norm": 34.54164505004883, "learning_rate": 2.2738486842105264e-06, "loss": 0.2134, "step": 3758 }, { "epoch": 0.7728207236842105, "grad_norm": 189.53782653808594, "learning_rate": 2.2717927631578947e-06, "loss": 0.7695, "step": 3759 }, { "epoch": 0.7730263157894737, "grad_norm": 178.03546142578125, "learning_rate": 2.2697368421052634e-06, "loss": 0.7202, "step": 3760 }, { "epoch": 0.7732319078947368, "grad_norm": 8.814651489257812, "learning_rate": 2.2676809210526317e-06, "loss": 0.0151, "step": 3761 }, { "epoch": 0.7734375, "grad_norm": 75.83627319335938, "learning_rate": 2.265625e-06, "loss": 0.1396, "step": 3762 }, { "epoch": 0.7736430921052632, "grad_norm": 241.36666870117188, "learning_rate": 2.2635690789473687e-06, "loss": 0.3087, "step": 3763 }, { "epoch": 0.7738486842105263, "grad_norm": 16.926546096801758, "learning_rate": 2.261513157894737e-06, "loss": 0.0847, "step": 3764 }, { "epoch": 0.7740542763157895, "grad_norm": 83.80651092529297, "learning_rate": 2.2594572368421053e-06, "loss": 0.3557, "step": 3765 }, { "epoch": 0.7742598684210527, "grad_norm": 16.096050262451172, "learning_rate": 2.257401315789474e-06, "loss": 0.0759, "step": 3766 }, { "epoch": 0.7744654605263158, "grad_norm": 7.9707112312316895, "learning_rate": 2.2553453947368424e-06, "loss": 0.0319, "step": 3767 }, { "epoch": 0.774671052631579, "grad_norm": 385.8392333984375, "learning_rate": 2.2532894736842107e-06, "loss": 2.6719, "step": 3768 }, { "epoch": 0.7748766447368421, "grad_norm": 248.45449829101562, "learning_rate": 2.251233552631579e-06, "loss": 1.6956, "step": 3769 }, { "epoch": 0.7750822368421053, "grad_norm": 15.314764976501465, "learning_rate": 2.2491776315789477e-06, "loss": 0.0272, "step": 3770 }, { "epoch": 0.7752878289473685, "grad_norm": 77.65638732910156, "learning_rate": 2.247121710526316e-06, "loss": 0.6865, "step": 3771 }, { "epoch": 0.7754934210526315, "grad_norm": 6.631051063537598, "learning_rate": 2.2450657894736843e-06, "loss": 0.038, "step": 3772 }, { "epoch": 0.7756990131578947, "grad_norm": 32.96369934082031, "learning_rate": 2.2430098684210526e-06, "loss": 0.1348, "step": 3773 }, { "epoch": 0.7759046052631579, "grad_norm": 25.293153762817383, "learning_rate": 2.2409539473684214e-06, "loss": 0.0406, "step": 3774 }, { "epoch": 0.776110197368421, "grad_norm": 52.52391052246094, "learning_rate": 2.2388980263157897e-06, "loss": 0.2428, "step": 3775 }, { "epoch": 0.7763157894736842, "grad_norm": 17.602645874023438, "learning_rate": 2.236842105263158e-06, "loss": 0.0275, "step": 3776 }, { "epoch": 0.7765213815789473, "grad_norm": 94.35929870605469, "learning_rate": 2.2347861842105263e-06, "loss": 0.5879, "step": 3777 }, { "epoch": 0.7767269736842105, "grad_norm": 7.947362422943115, "learning_rate": 2.232730263157895e-06, "loss": 0.0201, "step": 3778 }, { "epoch": 0.7769325657894737, "grad_norm": 312.4162292480469, "learning_rate": 2.2306743421052633e-06, "loss": 1.4023, "step": 3779 }, { "epoch": 0.7771381578947368, "grad_norm": 412.363037109375, "learning_rate": 2.228618421052632e-06, "loss": 3.3242, "step": 3780 }, { "epoch": 0.77734375, "grad_norm": 37.336544036865234, "learning_rate": 2.2265625000000003e-06, "loss": 0.1106, "step": 3781 }, { "epoch": 0.7775493421052632, "grad_norm": 86.6886978149414, "learning_rate": 2.2245065789473686e-06, "loss": 0.2217, "step": 3782 }, { "epoch": 0.7777549342105263, "grad_norm": 128.10740661621094, "learning_rate": 2.222450657894737e-06, "loss": 0.615, "step": 3783 }, { "epoch": 0.7779605263157895, "grad_norm": 37.54725646972656, "learning_rate": 2.2203947368421057e-06, "loss": 0.1621, "step": 3784 }, { "epoch": 0.7781661184210527, "grad_norm": 33.4779052734375, "learning_rate": 2.218338815789474e-06, "loss": 0.0962, "step": 3785 }, { "epoch": 0.7783717105263158, "grad_norm": 93.78205108642578, "learning_rate": 2.2162828947368423e-06, "loss": 0.4774, "step": 3786 }, { "epoch": 0.778577302631579, "grad_norm": 218.20208740234375, "learning_rate": 2.2142269736842106e-06, "loss": 1.0195, "step": 3787 }, { "epoch": 0.7787828947368421, "grad_norm": 20.705589294433594, "learning_rate": 2.212171052631579e-06, "loss": 0.0522, "step": 3788 }, { "epoch": 0.7789884868421053, "grad_norm": 38.38737487792969, "learning_rate": 2.2101151315789476e-06, "loss": 0.1792, "step": 3789 }, { "epoch": 0.7791940789473685, "grad_norm": 41.742671966552734, "learning_rate": 2.208059210526316e-06, "loss": 0.1303, "step": 3790 }, { "epoch": 0.7793996710526315, "grad_norm": 443.13507080078125, "learning_rate": 2.2060032894736842e-06, "loss": 2.1784, "step": 3791 }, { "epoch": 0.7796052631578947, "grad_norm": 13.260154724121094, "learning_rate": 2.2039473684210525e-06, "loss": 0.0294, "step": 3792 }, { "epoch": 0.7798108552631579, "grad_norm": 133.23472595214844, "learning_rate": 2.2018914473684213e-06, "loss": 0.4644, "step": 3793 }, { "epoch": 0.780016447368421, "grad_norm": 112.51393127441406, "learning_rate": 2.1998355263157896e-06, "loss": 0.2937, "step": 3794 }, { "epoch": 0.7802220394736842, "grad_norm": 33.73923873901367, "learning_rate": 2.1977796052631583e-06, "loss": 0.0886, "step": 3795 }, { "epoch": 0.7804276315789473, "grad_norm": 80.2051010131836, "learning_rate": 2.1957236842105266e-06, "loss": 0.9492, "step": 3796 }, { "epoch": 0.7806332236842105, "grad_norm": 31.72256088256836, "learning_rate": 2.193667763157895e-06, "loss": 0.0944, "step": 3797 }, { "epoch": 0.7808388157894737, "grad_norm": 24.854541778564453, "learning_rate": 2.1916118421052636e-06, "loss": 0.0588, "step": 3798 }, { "epoch": 0.7810444078947368, "grad_norm": 9.49222183227539, "learning_rate": 2.189555921052632e-06, "loss": 0.0191, "step": 3799 }, { "epoch": 0.78125, "grad_norm": 23.034263610839844, "learning_rate": 2.1875000000000002e-06, "loss": 0.0916, "step": 3800 }, { "epoch": 0.7814555921052632, "grad_norm": 201.7903289794922, "learning_rate": 2.1854440789473685e-06, "loss": 0.4031, "step": 3801 }, { "epoch": 0.7816611842105263, "grad_norm": 41.62652587890625, "learning_rate": 2.183388157894737e-06, "loss": 0.1238, "step": 3802 }, { "epoch": 0.7818667763157895, "grad_norm": 11.899552345275879, "learning_rate": 2.1813322368421056e-06, "loss": 0.0308, "step": 3803 }, { "epoch": 0.7820723684210527, "grad_norm": 3.6177420616149902, "learning_rate": 2.179276315789474e-06, "loss": 0.0164, "step": 3804 }, { "epoch": 0.7822779605263158, "grad_norm": 192.50665283203125, "learning_rate": 2.177220394736842e-06, "loss": 1.0811, "step": 3805 }, { "epoch": 0.782483552631579, "grad_norm": 70.19947814941406, "learning_rate": 2.1751644736842105e-06, "loss": 0.6286, "step": 3806 }, { "epoch": 0.7826891447368421, "grad_norm": 166.2666778564453, "learning_rate": 2.1731085526315788e-06, "loss": 0.4912, "step": 3807 }, { "epoch": 0.7828947368421053, "grad_norm": 28.982505798339844, "learning_rate": 2.1710526315789475e-06, "loss": 0.1099, "step": 3808 }, { "epoch": 0.7831003289473685, "grad_norm": 0.9836202263832092, "learning_rate": 2.168996710526316e-06, "loss": 0.0014, "step": 3809 }, { "epoch": 0.7833059210526315, "grad_norm": 7.341226577758789, "learning_rate": 2.1669407894736845e-06, "loss": 0.0189, "step": 3810 }, { "epoch": 0.7835115131578947, "grad_norm": 25.73600959777832, "learning_rate": 2.164884868421053e-06, "loss": 0.0498, "step": 3811 }, { "epoch": 0.7837171052631579, "grad_norm": 125.24524688720703, "learning_rate": 2.162828947368421e-06, "loss": 0.7371, "step": 3812 }, { "epoch": 0.783922697368421, "grad_norm": 13.499073028564453, "learning_rate": 2.16077302631579e-06, "loss": 0.028, "step": 3813 }, { "epoch": 0.7841282894736842, "grad_norm": 16.697525024414062, "learning_rate": 2.158717105263158e-06, "loss": 0.0446, "step": 3814 }, { "epoch": 0.7843338815789473, "grad_norm": 22.00243377685547, "learning_rate": 2.1566611842105265e-06, "loss": 0.104, "step": 3815 }, { "epoch": 0.7845394736842105, "grad_norm": 150.87286376953125, "learning_rate": 2.154605263157895e-06, "loss": 0.5583, "step": 3816 }, { "epoch": 0.7847450657894737, "grad_norm": 138.15945434570312, "learning_rate": 2.1525493421052635e-06, "loss": 0.2913, "step": 3817 }, { "epoch": 0.7849506578947368, "grad_norm": 18.78952980041504, "learning_rate": 2.150493421052632e-06, "loss": 0.04, "step": 3818 }, { "epoch": 0.78515625, "grad_norm": 180.27259826660156, "learning_rate": 2.1484375e-06, "loss": 1.6204, "step": 3819 }, { "epoch": 0.7853618421052632, "grad_norm": 55.05657196044922, "learning_rate": 2.1463815789473684e-06, "loss": 0.4414, "step": 3820 }, { "epoch": 0.7855674342105263, "grad_norm": 7.231216907501221, "learning_rate": 2.1443256578947367e-06, "loss": 0.0161, "step": 3821 }, { "epoch": 0.7857730263157895, "grad_norm": 70.269287109375, "learning_rate": 2.1422697368421055e-06, "loss": 0.46, "step": 3822 }, { "epoch": 0.7859786184210527, "grad_norm": 45.443721771240234, "learning_rate": 2.1402138157894738e-06, "loss": 0.0746, "step": 3823 }, { "epoch": 0.7861842105263158, "grad_norm": 31.893457412719727, "learning_rate": 2.138157894736842e-06, "loss": 0.1077, "step": 3824 }, { "epoch": 0.786389802631579, "grad_norm": 91.42501068115234, "learning_rate": 2.136101973684211e-06, "loss": 0.3116, "step": 3825 }, { "epoch": 0.7865953947368421, "grad_norm": 239.64727783203125, "learning_rate": 2.134046052631579e-06, "loss": 3.3484, "step": 3826 }, { "epoch": 0.7868009868421053, "grad_norm": 73.92472076416016, "learning_rate": 2.131990131578948e-06, "loss": 0.2047, "step": 3827 }, { "epoch": 0.7870065789473685, "grad_norm": 312.7259826660156, "learning_rate": 2.129934210526316e-06, "loss": 1.3984, "step": 3828 }, { "epoch": 0.7872121710526315, "grad_norm": 12.883113861083984, "learning_rate": 2.1278782894736844e-06, "loss": 0.0481, "step": 3829 }, { "epoch": 0.7874177631578947, "grad_norm": 12.285533905029297, "learning_rate": 2.1258223684210527e-06, "loss": 0.0203, "step": 3830 }, { "epoch": 0.7876233552631579, "grad_norm": 7.4538164138793945, "learning_rate": 2.1237664473684215e-06, "loss": 0.0332, "step": 3831 }, { "epoch": 0.787828947368421, "grad_norm": 35.55167770385742, "learning_rate": 2.1217105263157898e-06, "loss": 0.1055, "step": 3832 }, { "epoch": 0.7880345394736842, "grad_norm": 102.41555786132812, "learning_rate": 2.119654605263158e-06, "loss": 0.3574, "step": 3833 }, { "epoch": 0.7882401315789473, "grad_norm": 5.677938938140869, "learning_rate": 2.1175986842105264e-06, "loss": 0.0167, "step": 3834 }, { "epoch": 0.7884457236842105, "grad_norm": 6.571420669555664, "learning_rate": 2.1155427631578947e-06, "loss": 0.0121, "step": 3835 }, { "epoch": 0.7886513157894737, "grad_norm": 20.72281265258789, "learning_rate": 2.1134868421052634e-06, "loss": 0.0933, "step": 3836 }, { "epoch": 0.7888569078947368, "grad_norm": 33.99110794067383, "learning_rate": 2.1114309210526317e-06, "loss": 0.1341, "step": 3837 }, { "epoch": 0.7890625, "grad_norm": 23.34837532043457, "learning_rate": 2.109375e-06, "loss": 0.0693, "step": 3838 }, { "epoch": 0.7892680921052632, "grad_norm": 113.03446197509766, "learning_rate": 2.1073190789473683e-06, "loss": 0.4207, "step": 3839 }, { "epoch": 0.7894736842105263, "grad_norm": 109.5128173828125, "learning_rate": 2.105263157894737e-06, "loss": 0.329, "step": 3840 }, { "epoch": 0.7896792763157895, "grad_norm": 113.12918853759766, "learning_rate": 2.1032072368421054e-06, "loss": 0.6601, "step": 3841 }, { "epoch": 0.7898848684210527, "grad_norm": 218.21221923828125, "learning_rate": 2.101151315789474e-06, "loss": 0.8223, "step": 3842 }, { "epoch": 0.7900904605263158, "grad_norm": 15.603633880615234, "learning_rate": 2.0990953947368424e-06, "loss": 0.0576, "step": 3843 }, { "epoch": 0.790296052631579, "grad_norm": 63.489906311035156, "learning_rate": 2.0970394736842107e-06, "loss": 0.2314, "step": 3844 }, { "epoch": 0.7905016447368421, "grad_norm": 174.59524536132812, "learning_rate": 2.094983552631579e-06, "loss": 1.4258, "step": 3845 }, { "epoch": 0.7907072368421053, "grad_norm": 17.928386688232422, "learning_rate": 2.0929276315789477e-06, "loss": 0.0819, "step": 3846 }, { "epoch": 0.7909128289473685, "grad_norm": 74.6920166015625, "learning_rate": 2.090871710526316e-06, "loss": 0.2605, "step": 3847 }, { "epoch": 0.7911184210526315, "grad_norm": 6.284909725189209, "learning_rate": 2.0888157894736843e-06, "loss": 0.0103, "step": 3848 }, { "epoch": 0.7913240131578947, "grad_norm": 25.968393325805664, "learning_rate": 2.0867598684210526e-06, "loss": 0.1345, "step": 3849 }, { "epoch": 0.7915296052631579, "grad_norm": 81.11013793945312, "learning_rate": 2.0847039473684214e-06, "loss": 1.3712, "step": 3850 }, { "epoch": 0.791735197368421, "grad_norm": 393.8890380859375, "learning_rate": 2.0826480263157897e-06, "loss": 1.0347, "step": 3851 }, { "epoch": 0.7919407894736842, "grad_norm": 10.472250938415527, "learning_rate": 2.080592105263158e-06, "loss": 0.0565, "step": 3852 }, { "epoch": 0.7921463815789473, "grad_norm": 41.06206130981445, "learning_rate": 2.0785361842105263e-06, "loss": 0.1507, "step": 3853 }, { "epoch": 0.7923519736842105, "grad_norm": 3.9689884185791016, "learning_rate": 2.076480263157895e-06, "loss": 0.0169, "step": 3854 }, { "epoch": 0.7925575657894737, "grad_norm": 32.49616622924805, "learning_rate": 2.0744243421052633e-06, "loss": 0.0793, "step": 3855 }, { "epoch": 0.7927631578947368, "grad_norm": 33.872066497802734, "learning_rate": 2.0723684210526316e-06, "loss": 0.0898, "step": 3856 }, { "epoch": 0.79296875, "grad_norm": 534.7158813476562, "learning_rate": 2.0703125000000003e-06, "loss": 2.9697, "step": 3857 }, { "epoch": 0.7931743421052632, "grad_norm": 16.135204315185547, "learning_rate": 2.0682565789473687e-06, "loss": 0.0399, "step": 3858 }, { "epoch": 0.7933799342105263, "grad_norm": 21.315214157104492, "learning_rate": 2.066200657894737e-06, "loss": 0.0582, "step": 3859 }, { "epoch": 0.7935855263157895, "grad_norm": 165.37342834472656, "learning_rate": 2.0641447368421057e-06, "loss": 1.0566, "step": 3860 }, { "epoch": 0.7937911184210527, "grad_norm": 307.6373291015625, "learning_rate": 2.062088815789474e-06, "loss": 1.1724, "step": 3861 }, { "epoch": 0.7939967105263158, "grad_norm": 96.17504119873047, "learning_rate": 2.0600328947368423e-06, "loss": 0.4624, "step": 3862 }, { "epoch": 0.794202302631579, "grad_norm": 203.75088500976562, "learning_rate": 2.0579769736842106e-06, "loss": 2.1875, "step": 3863 }, { "epoch": 0.7944078947368421, "grad_norm": 8.284014701843262, "learning_rate": 2.055921052631579e-06, "loss": 0.009, "step": 3864 }, { "epoch": 0.7946134868421053, "grad_norm": 36.181640625, "learning_rate": 2.0538651315789476e-06, "loss": 0.0474, "step": 3865 }, { "epoch": 0.7948190789473685, "grad_norm": 52.94401931762695, "learning_rate": 2.051809210526316e-06, "loss": 0.121, "step": 3866 }, { "epoch": 0.7950246710526315, "grad_norm": 69.76756286621094, "learning_rate": 2.0497532894736842e-06, "loss": 0.283, "step": 3867 }, { "epoch": 0.7952302631578947, "grad_norm": 96.04601287841797, "learning_rate": 2.0476973684210525e-06, "loss": 0.2805, "step": 3868 }, { "epoch": 0.7954358552631579, "grad_norm": 199.02012634277344, "learning_rate": 2.0456414473684213e-06, "loss": 1.1104, "step": 3869 }, { "epoch": 0.795641447368421, "grad_norm": 27.621618270874023, "learning_rate": 2.0435855263157896e-06, "loss": 0.0903, "step": 3870 }, { "epoch": 0.7958470394736842, "grad_norm": 10.333816528320312, "learning_rate": 2.041529605263158e-06, "loss": 0.041, "step": 3871 }, { "epoch": 0.7960526315789473, "grad_norm": 35.06134796142578, "learning_rate": 2.0394736842105266e-06, "loss": 0.1909, "step": 3872 }, { "epoch": 0.7962582236842105, "grad_norm": 3.6540350914001465, "learning_rate": 2.037417763157895e-06, "loss": 0.0132, "step": 3873 }, { "epoch": 0.7964638157894737, "grad_norm": 25.931354522705078, "learning_rate": 2.0353618421052636e-06, "loss": 0.076, "step": 3874 }, { "epoch": 0.7966694078947368, "grad_norm": 139.2017364501953, "learning_rate": 2.033305921052632e-06, "loss": 1.0602, "step": 3875 }, { "epoch": 0.796875, "grad_norm": 16.32374382019043, "learning_rate": 2.0312500000000002e-06, "loss": 0.0386, "step": 3876 }, { "epoch": 0.7970805921052632, "grad_norm": 112.57413482666016, "learning_rate": 2.0291940789473685e-06, "loss": 0.8457, "step": 3877 }, { "epoch": 0.7972861842105263, "grad_norm": 68.8168716430664, "learning_rate": 2.027138157894737e-06, "loss": 0.3185, "step": 3878 }, { "epoch": 0.7974917763157895, "grad_norm": 12.425107955932617, "learning_rate": 2.0250822368421056e-06, "loss": 0.0374, "step": 3879 }, { "epoch": 0.7976973684210527, "grad_norm": 53.923641204833984, "learning_rate": 2.023026315789474e-06, "loss": 0.105, "step": 3880 }, { "epoch": 0.7979029605263158, "grad_norm": 74.71512603759766, "learning_rate": 2.020970394736842e-06, "loss": 0.2343, "step": 3881 }, { "epoch": 0.798108552631579, "grad_norm": 40.24238204956055, "learning_rate": 2.0189144736842105e-06, "loss": 0.2007, "step": 3882 }, { "epoch": 0.7983141447368421, "grad_norm": 418.4888916015625, "learning_rate": 2.016858552631579e-06, "loss": 0.9316, "step": 3883 }, { "epoch": 0.7985197368421053, "grad_norm": 43.61506652832031, "learning_rate": 2.0148026315789475e-06, "loss": 0.0691, "step": 3884 }, { "epoch": 0.7987253289473685, "grad_norm": 86.6198501586914, "learning_rate": 2.012746710526316e-06, "loss": 0.4543, "step": 3885 }, { "epoch": 0.7989309210526315, "grad_norm": 214.4209747314453, "learning_rate": 2.010690789473684e-06, "loss": 0.8447, "step": 3886 }, { "epoch": 0.7991365131578947, "grad_norm": 169.50303649902344, "learning_rate": 2.008634868421053e-06, "loss": 0.4905, "step": 3887 }, { "epoch": 0.7993421052631579, "grad_norm": 90.98804473876953, "learning_rate": 2.006578947368421e-06, "loss": 1.1353, "step": 3888 }, { "epoch": 0.799547697368421, "grad_norm": 2.0262467861175537, "learning_rate": 2.00452302631579e-06, "loss": 0.0092, "step": 3889 }, { "epoch": 0.7997532894736842, "grad_norm": 17.600923538208008, "learning_rate": 2.002467105263158e-06, "loss": 0.0429, "step": 3890 }, { "epoch": 0.7999588815789473, "grad_norm": 12.102437973022461, "learning_rate": 2.0004111842105265e-06, "loss": 0.0435, "step": 3891 }, { "epoch": 0.8001644736842105, "grad_norm": 2.635213613510132, "learning_rate": 1.998355263157895e-06, "loss": 0.0125, "step": 3892 }, { "epoch": 0.8003700657894737, "grad_norm": 65.650390625, "learning_rate": 1.9962993421052635e-06, "loss": 0.2725, "step": 3893 }, { "epoch": 0.8005756578947368, "grad_norm": 58.52676773071289, "learning_rate": 1.994243421052632e-06, "loss": 0.3003, "step": 3894 }, { "epoch": 0.80078125, "grad_norm": 299.8064880371094, "learning_rate": 1.9921875e-06, "loss": 1.9062, "step": 3895 }, { "epoch": 0.8009868421052632, "grad_norm": 122.035888671875, "learning_rate": 1.9901315789473684e-06, "loss": 0.834, "step": 3896 }, { "epoch": 0.8011924342105263, "grad_norm": 59.61692810058594, "learning_rate": 1.9880756578947368e-06, "loss": 0.8999, "step": 3897 }, { "epoch": 0.8013980263157895, "grad_norm": 8.036462783813477, "learning_rate": 1.9860197368421055e-06, "loss": 0.0224, "step": 3898 }, { "epoch": 0.8016036184210527, "grad_norm": 20.421615600585938, "learning_rate": 1.9839638157894738e-06, "loss": 0.0625, "step": 3899 }, { "epoch": 0.8018092105263158, "grad_norm": 57.253055572509766, "learning_rate": 1.981907894736842e-06, "loss": 0.3594, "step": 3900 }, { "epoch": 0.802014802631579, "grad_norm": 28.97466278076172, "learning_rate": 1.979851973684211e-06, "loss": 0.0925, "step": 3901 }, { "epoch": 0.8022203947368421, "grad_norm": 19.303133010864258, "learning_rate": 1.977796052631579e-06, "loss": 0.0508, "step": 3902 }, { "epoch": 0.8024259868421053, "grad_norm": 68.71725463867188, "learning_rate": 1.9757401315789474e-06, "loss": 0.272, "step": 3903 }, { "epoch": 0.8026315789473685, "grad_norm": 27.4210147857666, "learning_rate": 1.973684210526316e-06, "loss": 0.1929, "step": 3904 }, { "epoch": 0.8028371710526315, "grad_norm": 4.489548206329346, "learning_rate": 1.9716282894736845e-06, "loss": 0.0143, "step": 3905 }, { "epoch": 0.8030427631578947, "grad_norm": 46.17130661010742, "learning_rate": 1.9695723684210528e-06, "loss": 0.1147, "step": 3906 }, { "epoch": 0.8032483552631579, "grad_norm": 10.51746940612793, "learning_rate": 1.9675164473684215e-06, "loss": 0.0417, "step": 3907 }, { "epoch": 0.803453947368421, "grad_norm": 65.13725280761719, "learning_rate": 1.96546052631579e-06, "loss": 0.3438, "step": 3908 }, { "epoch": 0.8036595394736842, "grad_norm": 33.85992431640625, "learning_rate": 1.963404605263158e-06, "loss": 0.0987, "step": 3909 }, { "epoch": 0.8038651315789473, "grad_norm": 185.3708038330078, "learning_rate": 1.9613486842105264e-06, "loss": 1.2791, "step": 3910 }, { "epoch": 0.8040707236842105, "grad_norm": 51.490516662597656, "learning_rate": 1.9592927631578947e-06, "loss": 0.1846, "step": 3911 }, { "epoch": 0.8042763157894737, "grad_norm": 29.1131534576416, "learning_rate": 1.9572368421052634e-06, "loss": 0.0508, "step": 3912 }, { "epoch": 0.8044819078947368, "grad_norm": 20.14287757873535, "learning_rate": 1.9551809210526317e-06, "loss": 0.08, "step": 3913 }, { "epoch": 0.8046875, "grad_norm": 34.440643310546875, "learning_rate": 1.953125e-06, "loss": 0.1143, "step": 3914 }, { "epoch": 0.8048930921052632, "grad_norm": 2.955683708190918, "learning_rate": 1.9510690789473683e-06, "loss": 0.0088, "step": 3915 }, { "epoch": 0.8050986842105263, "grad_norm": 20.757091522216797, "learning_rate": 1.949013157894737e-06, "loss": 0.1163, "step": 3916 }, { "epoch": 0.8053042763157895, "grad_norm": 139.1640625, "learning_rate": 1.9469572368421054e-06, "loss": 0.8559, "step": 3917 }, { "epoch": 0.8055098684210527, "grad_norm": 29.755611419677734, "learning_rate": 1.9449013157894737e-06, "loss": 0.0728, "step": 3918 }, { "epoch": 0.8057154605263158, "grad_norm": 50.335140228271484, "learning_rate": 1.9428453947368424e-06, "loss": 0.106, "step": 3919 }, { "epoch": 0.805921052631579, "grad_norm": 58.08919906616211, "learning_rate": 1.9407894736842107e-06, "loss": 0.3773, "step": 3920 }, { "epoch": 0.8061266447368421, "grad_norm": 5.0370707511901855, "learning_rate": 1.938733552631579e-06, "loss": 0.0097, "step": 3921 }, { "epoch": 0.8063322368421053, "grad_norm": 210.83750915527344, "learning_rate": 1.9366776315789477e-06, "loss": 0.9666, "step": 3922 }, { "epoch": 0.8065378289473685, "grad_norm": 2.402164936065674, "learning_rate": 1.934621710526316e-06, "loss": 0.0081, "step": 3923 }, { "epoch": 0.8067434210526315, "grad_norm": 55.88710021972656, "learning_rate": 1.9325657894736844e-06, "loss": 0.9595, "step": 3924 }, { "epoch": 0.8069490131578947, "grad_norm": 96.56846618652344, "learning_rate": 1.9305098684210527e-06, "loss": 0.4684, "step": 3925 }, { "epoch": 0.8071546052631579, "grad_norm": 78.36083984375, "learning_rate": 1.9284539473684214e-06, "loss": 0.9506, "step": 3926 }, { "epoch": 0.807360197368421, "grad_norm": 33.46701431274414, "learning_rate": 1.9263980263157897e-06, "loss": 0.0392, "step": 3927 }, { "epoch": 0.8075657894736842, "grad_norm": 344.85003662109375, "learning_rate": 1.924342105263158e-06, "loss": 1.0977, "step": 3928 }, { "epoch": 0.8077713815789473, "grad_norm": 42.40590286254883, "learning_rate": 1.9222861842105263e-06, "loss": 0.2134, "step": 3929 }, { "epoch": 0.8079769736842105, "grad_norm": 7.331238746643066, "learning_rate": 1.9202302631578946e-06, "loss": 0.0207, "step": 3930 }, { "epoch": 0.8081825657894737, "grad_norm": 98.03681182861328, "learning_rate": 1.9181743421052633e-06, "loss": 0.3138, "step": 3931 }, { "epoch": 0.8083881578947368, "grad_norm": 97.69049072265625, "learning_rate": 1.9161184210526316e-06, "loss": 0.7695, "step": 3932 }, { "epoch": 0.80859375, "grad_norm": 44.255863189697266, "learning_rate": 1.9140625000000004e-06, "loss": 0.1119, "step": 3933 }, { "epoch": 0.8087993421052632, "grad_norm": 81.73925018310547, "learning_rate": 1.9120065789473687e-06, "loss": 0.9009, "step": 3934 }, { "epoch": 0.8090049342105263, "grad_norm": 44.4053840637207, "learning_rate": 1.909950657894737e-06, "loss": 0.205, "step": 3935 }, { "epoch": 0.8092105263157895, "grad_norm": 61.93326950073242, "learning_rate": 1.9078947368421057e-06, "loss": 0.895, "step": 3936 }, { "epoch": 0.8094161184210527, "grad_norm": 69.32760620117188, "learning_rate": 1.9058388157894738e-06, "loss": 0.2425, "step": 3937 }, { "epoch": 0.8096217105263158, "grad_norm": 55.087554931640625, "learning_rate": 1.9037828947368423e-06, "loss": 1.0127, "step": 3938 }, { "epoch": 0.809827302631579, "grad_norm": 22.24225425720215, "learning_rate": 1.9017269736842106e-06, "loss": 0.0647, "step": 3939 }, { "epoch": 0.8100328947368421, "grad_norm": 7.063749313354492, "learning_rate": 1.899671052631579e-06, "loss": 0.0119, "step": 3940 }, { "epoch": 0.8102384868421053, "grad_norm": 9.39652156829834, "learning_rate": 1.8976151315789476e-06, "loss": 0.0201, "step": 3941 }, { "epoch": 0.8104440789473685, "grad_norm": 37.59337615966797, "learning_rate": 1.895559210526316e-06, "loss": 0.1262, "step": 3942 }, { "epoch": 0.8106496710526315, "grad_norm": 66.30458068847656, "learning_rate": 1.8935032894736843e-06, "loss": 1.3337, "step": 3943 }, { "epoch": 0.8108552631578947, "grad_norm": 72.58637237548828, "learning_rate": 1.8914473684210528e-06, "loss": 0.2495, "step": 3944 }, { "epoch": 0.8110608552631579, "grad_norm": 77.12279510498047, "learning_rate": 1.8893914473684213e-06, "loss": 0.9492, "step": 3945 }, { "epoch": 0.811266447368421, "grad_norm": 195.95127868652344, "learning_rate": 1.8873355263157898e-06, "loss": 0.6533, "step": 3946 }, { "epoch": 0.8114720394736842, "grad_norm": 15.382251739501953, "learning_rate": 1.885279605263158e-06, "loss": 0.0293, "step": 3947 }, { "epoch": 0.8116776315789473, "grad_norm": 3.812593698501587, "learning_rate": 1.8832236842105264e-06, "loss": 0.0048, "step": 3948 }, { "epoch": 0.8118832236842105, "grad_norm": 34.97233963012695, "learning_rate": 1.8811677631578947e-06, "loss": 0.1129, "step": 3949 }, { "epoch": 0.8120888157894737, "grad_norm": 517.6947021484375, "learning_rate": 1.8791118421052634e-06, "loss": 2.6313, "step": 3950 }, { "epoch": 0.8122944078947368, "grad_norm": 79.08204650878906, "learning_rate": 1.8770559210526317e-06, "loss": 0.2104, "step": 3951 }, { "epoch": 0.8125, "grad_norm": 7.056222438812256, "learning_rate": 1.8750000000000003e-06, "loss": 0.0127, "step": 3952 }, { "epoch": 0.8127055921052632, "grad_norm": 8.743961334228516, "learning_rate": 1.8729440789473686e-06, "loss": 0.0206, "step": 3953 }, { "epoch": 0.8129111842105263, "grad_norm": 9.861189842224121, "learning_rate": 1.8708881578947369e-06, "loss": 0.0385, "step": 3954 }, { "epoch": 0.8131167763157895, "grad_norm": 86.80718231201172, "learning_rate": 1.8688322368421056e-06, "loss": 0.284, "step": 3955 }, { "epoch": 0.8133223684210527, "grad_norm": 11.540085792541504, "learning_rate": 1.866776315789474e-06, "loss": 0.0183, "step": 3956 }, { "epoch": 0.8135279605263158, "grad_norm": 79.40608978271484, "learning_rate": 1.8647203947368422e-06, "loss": 0.6057, "step": 3957 }, { "epoch": 0.813733552631579, "grad_norm": 56.84041976928711, "learning_rate": 1.8626644736842105e-06, "loss": 0.2855, "step": 3958 }, { "epoch": 0.8139391447368421, "grad_norm": 111.82420349121094, "learning_rate": 1.860608552631579e-06, "loss": 0.5713, "step": 3959 }, { "epoch": 0.8141447368421053, "grad_norm": 329.2220458984375, "learning_rate": 1.8585526315789475e-06, "loss": 1.6035, "step": 3960 }, { "epoch": 0.8143503289473685, "grad_norm": 77.03360748291016, "learning_rate": 1.856496710526316e-06, "loss": 0.4819, "step": 3961 }, { "epoch": 0.8145559210526315, "grad_norm": 49.8905143737793, "learning_rate": 1.8544407894736844e-06, "loss": 0.2593, "step": 3962 }, { "epoch": 0.8147615131578947, "grad_norm": 9.080778121948242, "learning_rate": 1.8523848684210527e-06, "loss": 0.0229, "step": 3963 }, { "epoch": 0.8149671052631579, "grad_norm": 63.80656051635742, "learning_rate": 1.8503289473684214e-06, "loss": 0.4026, "step": 3964 }, { "epoch": 0.815172697368421, "grad_norm": 165.91708374023438, "learning_rate": 1.8482730263157897e-06, "loss": 1.2188, "step": 3965 }, { "epoch": 0.8153782894736842, "grad_norm": 59.11594009399414, "learning_rate": 1.846217105263158e-06, "loss": 0.5547, "step": 3966 }, { "epoch": 0.8155838815789473, "grad_norm": 84.25189208984375, "learning_rate": 1.8441611842105265e-06, "loss": 1.0254, "step": 3967 }, { "epoch": 0.8157894736842105, "grad_norm": 85.0768814086914, "learning_rate": 1.8421052631578948e-06, "loss": 0.3634, "step": 3968 }, { "epoch": 0.8159950657894737, "grad_norm": 28.940053939819336, "learning_rate": 1.8400493421052633e-06, "loss": 0.1089, "step": 3969 }, { "epoch": 0.8162006578947368, "grad_norm": 18.801612854003906, "learning_rate": 1.8379934210526319e-06, "loss": 0.0905, "step": 3970 }, { "epoch": 0.81640625, "grad_norm": 43.2366828918457, "learning_rate": 1.8359375000000002e-06, "loss": 0.1317, "step": 3971 }, { "epoch": 0.8166118421052632, "grad_norm": 83.37356567382812, "learning_rate": 1.8338815789473685e-06, "loss": 0.6995, "step": 3972 }, { "epoch": 0.8168174342105263, "grad_norm": 244.939453125, "learning_rate": 1.831825657894737e-06, "loss": 1.9102, "step": 3973 }, { "epoch": 0.8170230263157895, "grad_norm": 191.13916015625, "learning_rate": 1.8297697368421055e-06, "loss": 1.7678, "step": 3974 }, { "epoch": 0.8172286184210527, "grad_norm": 12.521944999694824, "learning_rate": 1.8277138157894738e-06, "loss": 0.0299, "step": 3975 }, { "epoch": 0.8174342105263158, "grad_norm": 13.898992538452148, "learning_rate": 1.8256578947368423e-06, "loss": 0.0371, "step": 3976 }, { "epoch": 0.817639802631579, "grad_norm": 58.913761138916016, "learning_rate": 1.8236019736842106e-06, "loss": 1.0017, "step": 3977 }, { "epoch": 0.8178453947368421, "grad_norm": 47.197052001953125, "learning_rate": 1.821546052631579e-06, "loss": 0.1376, "step": 3978 }, { "epoch": 0.8180509868421053, "grad_norm": 106.80194091796875, "learning_rate": 1.8194901315789476e-06, "loss": 0.5308, "step": 3979 }, { "epoch": 0.8182565789473685, "grad_norm": 7.436956882476807, "learning_rate": 1.817434210526316e-06, "loss": 0.0159, "step": 3980 }, { "epoch": 0.8184621710526315, "grad_norm": 19.525615692138672, "learning_rate": 1.8153782894736843e-06, "loss": 0.0566, "step": 3981 }, { "epoch": 0.8186677631578947, "grad_norm": 96.94981384277344, "learning_rate": 1.8133223684210528e-06, "loss": 0.5923, "step": 3982 }, { "epoch": 0.8188733552631579, "grad_norm": 29.251623153686523, "learning_rate": 1.8112664473684213e-06, "loss": 0.0604, "step": 3983 }, { "epoch": 0.819078947368421, "grad_norm": 23.1419734954834, "learning_rate": 1.8092105263157896e-06, "loss": 0.0681, "step": 3984 }, { "epoch": 0.8192845394736842, "grad_norm": 52.284950256347656, "learning_rate": 1.8071546052631581e-06, "loss": 0.2302, "step": 3985 }, { "epoch": 0.8194901315789473, "grad_norm": 58.63282775878906, "learning_rate": 1.8050986842105264e-06, "loss": 0.167, "step": 3986 }, { "epoch": 0.8196957236842105, "grad_norm": 170.68580627441406, "learning_rate": 1.8030427631578947e-06, "loss": 0.6817, "step": 3987 }, { "epoch": 0.8199013157894737, "grad_norm": 50.17905044555664, "learning_rate": 1.8009868421052634e-06, "loss": 0.2446, "step": 3988 }, { "epoch": 0.8201069078947368, "grad_norm": 37.944549560546875, "learning_rate": 1.7989309210526318e-06, "loss": 0.1066, "step": 3989 }, { "epoch": 0.8203125, "grad_norm": 250.40460205078125, "learning_rate": 1.796875e-06, "loss": 0.5045, "step": 3990 }, { "epoch": 0.8205180921052632, "grad_norm": 14.02924919128418, "learning_rate": 1.7948190789473686e-06, "loss": 0.0537, "step": 3991 }, { "epoch": 0.8207236842105263, "grad_norm": 62.72391891479492, "learning_rate": 1.7927631578947369e-06, "loss": 0.2109, "step": 3992 }, { "epoch": 0.8209292763157895, "grad_norm": 9.317177772521973, "learning_rate": 1.7907072368421056e-06, "loss": 0.0231, "step": 3993 }, { "epoch": 0.8211348684210527, "grad_norm": 56.20705032348633, "learning_rate": 1.788651315789474e-06, "loss": 0.7743, "step": 3994 }, { "epoch": 0.8213404605263158, "grad_norm": 54.82453918457031, "learning_rate": 1.7865953947368422e-06, "loss": 0.2066, "step": 3995 }, { "epoch": 0.821546052631579, "grad_norm": 10.374639511108398, "learning_rate": 1.7845394736842105e-06, "loss": 0.0369, "step": 3996 }, { "epoch": 0.8217516447368421, "grad_norm": 146.23143005371094, "learning_rate": 1.782483552631579e-06, "loss": 0.6777, "step": 3997 }, { "epoch": 0.8219572368421053, "grad_norm": 22.99365997314453, "learning_rate": 1.7804276315789475e-06, "loss": 0.0819, "step": 3998 }, { "epoch": 0.8221628289473685, "grad_norm": 84.8017807006836, "learning_rate": 1.778371710526316e-06, "loss": 0.275, "step": 3999 }, { "epoch": 0.8223684210526315, "grad_norm": 6.311519622802734, "learning_rate": 1.7763157894736844e-06, "loss": 0.0205, "step": 4000 }, { "epoch": 0.8225740131578947, "grad_norm": 26.686317443847656, "learning_rate": 1.7742598684210527e-06, "loss": 0.0773, "step": 4001 }, { "epoch": 0.8227796052631579, "grad_norm": 29.168804168701172, "learning_rate": 1.7722039473684214e-06, "loss": 0.0663, "step": 4002 }, { "epoch": 0.822985197368421, "grad_norm": 19.5196533203125, "learning_rate": 1.7701480263157897e-06, "loss": 0.0551, "step": 4003 }, { "epoch": 0.8231907894736842, "grad_norm": 12.392281532287598, "learning_rate": 1.768092105263158e-06, "loss": 0.0302, "step": 4004 }, { "epoch": 0.8233963815789473, "grad_norm": 103.83707427978516, "learning_rate": 1.7660361842105263e-06, "loss": 0.2415, "step": 4005 }, { "epoch": 0.8236019736842105, "grad_norm": 12.509015083312988, "learning_rate": 1.7639802631578948e-06, "loss": 0.0276, "step": 4006 }, { "epoch": 0.8238075657894737, "grad_norm": 73.88495635986328, "learning_rate": 1.7619243421052633e-06, "loss": 0.179, "step": 4007 }, { "epoch": 0.8240131578947368, "grad_norm": 79.0031509399414, "learning_rate": 1.7598684210526319e-06, "loss": 0.8685, "step": 4008 }, { "epoch": 0.82421875, "grad_norm": 62.9752311706543, "learning_rate": 1.7578125000000002e-06, "loss": 0.2719, "step": 4009 }, { "epoch": 0.8244243421052632, "grad_norm": 4.0483527183532715, "learning_rate": 1.7557565789473685e-06, "loss": 0.0102, "step": 4010 }, { "epoch": 0.8246299342105263, "grad_norm": 22.953166961669922, "learning_rate": 1.7537006578947368e-06, "loss": 0.068, "step": 4011 }, { "epoch": 0.8248355263157895, "grad_norm": 17.339942932128906, "learning_rate": 1.7516447368421055e-06, "loss": 0.0479, "step": 4012 }, { "epoch": 0.8250411184210527, "grad_norm": 12.252467155456543, "learning_rate": 1.7495888157894738e-06, "loss": 0.0217, "step": 4013 }, { "epoch": 0.8252467105263158, "grad_norm": 21.704545974731445, "learning_rate": 1.7475328947368423e-06, "loss": 0.0654, "step": 4014 }, { "epoch": 0.825452302631579, "grad_norm": 142.79515075683594, "learning_rate": 1.7454769736842106e-06, "loss": 0.1207, "step": 4015 }, { "epoch": 0.8256578947368421, "grad_norm": 16.85356330871582, "learning_rate": 1.743421052631579e-06, "loss": 0.0402, "step": 4016 }, { "epoch": 0.8258634868421053, "grad_norm": 63.45553207397461, "learning_rate": 1.7413651315789477e-06, "loss": 0.5842, "step": 4017 }, { "epoch": 0.8260690789473685, "grad_norm": 5.508555889129639, "learning_rate": 1.739309210526316e-06, "loss": 0.0189, "step": 4018 }, { "epoch": 0.8262746710526315, "grad_norm": 11.171289443969727, "learning_rate": 1.7372532894736843e-06, "loss": 0.0238, "step": 4019 }, { "epoch": 0.8264802631578947, "grad_norm": 69.70010375976562, "learning_rate": 1.7351973684210528e-06, "loss": 0.2471, "step": 4020 }, { "epoch": 0.8266858552631579, "grad_norm": 83.8728256225586, "learning_rate": 1.7331414473684213e-06, "loss": 0.2598, "step": 4021 }, { "epoch": 0.826891447368421, "grad_norm": 7.154050350189209, "learning_rate": 1.7310855263157896e-06, "loss": 0.0147, "step": 4022 }, { "epoch": 0.8270970394736842, "grad_norm": 57.844696044921875, "learning_rate": 1.7290296052631581e-06, "loss": 0.4654, "step": 4023 }, { "epoch": 0.8273026315789473, "grad_norm": 17.29763412475586, "learning_rate": 1.7269736842105264e-06, "loss": 0.03, "step": 4024 }, { "epoch": 0.8275082236842105, "grad_norm": 25.293874740600586, "learning_rate": 1.7249177631578947e-06, "loss": 0.0839, "step": 4025 }, { "epoch": 0.8277138157894737, "grad_norm": 3.8781557083129883, "learning_rate": 1.7228618421052635e-06, "loss": 0.0123, "step": 4026 }, { "epoch": 0.8279194078947368, "grad_norm": 121.98361206054688, "learning_rate": 1.7208059210526318e-06, "loss": 0.3855, "step": 4027 }, { "epoch": 0.828125, "grad_norm": 6.353709697723389, "learning_rate": 1.71875e-06, "loss": 0.011, "step": 4028 }, { "epoch": 0.8283305921052632, "grad_norm": 105.37330627441406, "learning_rate": 1.7166940789473686e-06, "loss": 0.344, "step": 4029 }, { "epoch": 0.8285361842105263, "grad_norm": 31.296815872192383, "learning_rate": 1.7146381578947369e-06, "loss": 0.1003, "step": 4030 }, { "epoch": 0.8287417763157895, "grad_norm": 28.793991088867188, "learning_rate": 1.7125822368421056e-06, "loss": 0.1437, "step": 4031 }, { "epoch": 0.8289473684210527, "grad_norm": 158.04751586914062, "learning_rate": 1.710526315789474e-06, "loss": 0.8164, "step": 4032 }, { "epoch": 0.8291529605263158, "grad_norm": 7.542282581329346, "learning_rate": 1.7084703947368422e-06, "loss": 0.0199, "step": 4033 }, { "epoch": 0.829358552631579, "grad_norm": 73.50212097167969, "learning_rate": 1.7064144736842105e-06, "loss": 1.0215, "step": 4034 }, { "epoch": 0.8295641447368421, "grad_norm": 41.158267974853516, "learning_rate": 1.704358552631579e-06, "loss": 0.0702, "step": 4035 }, { "epoch": 0.8297697368421053, "grad_norm": 87.10340881347656, "learning_rate": 1.7023026315789476e-06, "loss": 0.595, "step": 4036 }, { "epoch": 0.8299753289473685, "grad_norm": 159.43724060058594, "learning_rate": 1.7002467105263159e-06, "loss": 0.5625, "step": 4037 }, { "epoch": 0.8301809210526315, "grad_norm": 11.530341148376465, "learning_rate": 1.6981907894736844e-06, "loss": 0.0214, "step": 4038 }, { "epoch": 0.8303865131578947, "grad_norm": 21.226879119873047, "learning_rate": 1.6961348684210527e-06, "loss": 0.0547, "step": 4039 }, { "epoch": 0.8305921052631579, "grad_norm": 79.26226043701172, "learning_rate": 1.6940789473684214e-06, "loss": 0.4199, "step": 4040 }, { "epoch": 0.830797697368421, "grad_norm": 39.07068634033203, "learning_rate": 1.6920230263157897e-06, "loss": 0.0757, "step": 4041 }, { "epoch": 0.8310032894736842, "grad_norm": 18.871803283691406, "learning_rate": 1.689967105263158e-06, "loss": 0.0322, "step": 4042 }, { "epoch": 0.8312088815789473, "grad_norm": 60.12132263183594, "learning_rate": 1.6879111842105263e-06, "loss": 0.6777, "step": 4043 }, { "epoch": 0.8314144736842105, "grad_norm": 61.333744049072266, "learning_rate": 1.6858552631578948e-06, "loss": 0.4463, "step": 4044 }, { "epoch": 0.8316200657894737, "grad_norm": 416.3518371582031, "learning_rate": 1.6837993421052634e-06, "loss": 0.8887, "step": 4045 }, { "epoch": 0.8318256578947368, "grad_norm": 31.48638343811035, "learning_rate": 1.6817434210526319e-06, "loss": 0.0722, "step": 4046 }, { "epoch": 0.83203125, "grad_norm": 25.908260345458984, "learning_rate": 1.6796875000000002e-06, "loss": 0.0714, "step": 4047 }, { "epoch": 0.8322368421052632, "grad_norm": 24.494220733642578, "learning_rate": 1.6776315789473685e-06, "loss": 0.0609, "step": 4048 }, { "epoch": 0.8324424342105263, "grad_norm": 61.56914138793945, "learning_rate": 1.6755756578947368e-06, "loss": 0.2632, "step": 4049 }, { "epoch": 0.8326480263157895, "grad_norm": 8.864385604858398, "learning_rate": 1.6735197368421055e-06, "loss": 0.0215, "step": 4050 }, { "epoch": 0.8328536184210527, "grad_norm": 2.1148531436920166, "learning_rate": 1.6714638157894738e-06, "loss": 0.0091, "step": 4051 }, { "epoch": 0.8330592105263158, "grad_norm": 68.75609588623047, "learning_rate": 1.6694078947368423e-06, "loss": 0.2083, "step": 4052 }, { "epoch": 0.833264802631579, "grad_norm": 109.47864532470703, "learning_rate": 1.6673519736842106e-06, "loss": 0.3594, "step": 4053 }, { "epoch": 0.8334703947368421, "grad_norm": 42.31819534301758, "learning_rate": 1.665296052631579e-06, "loss": 0.1768, "step": 4054 }, { "epoch": 0.8336759868421053, "grad_norm": 11.108480453491211, "learning_rate": 1.6632401315789477e-06, "loss": 0.0275, "step": 4055 }, { "epoch": 0.8338815789473685, "grad_norm": 35.67988204956055, "learning_rate": 1.661184210526316e-06, "loss": 0.1196, "step": 4056 }, { "epoch": 0.8340871710526315, "grad_norm": 12.68108081817627, "learning_rate": 1.6591282894736843e-06, "loss": 0.0316, "step": 4057 }, { "epoch": 0.8342927631578947, "grad_norm": 37.77415084838867, "learning_rate": 1.6570723684210526e-06, "loss": 0.1603, "step": 4058 }, { "epoch": 0.8344983552631579, "grad_norm": 3.254408359527588, "learning_rate": 1.6550164473684213e-06, "loss": 0.0075, "step": 4059 }, { "epoch": 0.834703947368421, "grad_norm": 99.93093872070312, "learning_rate": 1.6529605263157896e-06, "loss": 0.5693, "step": 4060 }, { "epoch": 0.8349095394736842, "grad_norm": 157.5975341796875, "learning_rate": 1.6509046052631581e-06, "loss": 0.8687, "step": 4061 }, { "epoch": 0.8351151315789473, "grad_norm": 6.743834495544434, "learning_rate": 1.6488486842105264e-06, "loss": 0.0162, "step": 4062 }, { "epoch": 0.8353207236842105, "grad_norm": 8.610640525817871, "learning_rate": 1.6467927631578947e-06, "loss": 0.0323, "step": 4063 }, { "epoch": 0.8355263157894737, "grad_norm": 63.82880401611328, "learning_rate": 1.6447368421052635e-06, "loss": 1.1787, "step": 4064 }, { "epoch": 0.8357319078947368, "grad_norm": 42.89739227294922, "learning_rate": 1.6426809210526318e-06, "loss": 0.2148, "step": 4065 }, { "epoch": 0.8359375, "grad_norm": 13.920499801635742, "learning_rate": 1.640625e-06, "loss": 0.0326, "step": 4066 }, { "epoch": 0.8361430921052632, "grad_norm": 105.82884216308594, "learning_rate": 1.6385690789473686e-06, "loss": 1.1973, "step": 4067 }, { "epoch": 0.8363486842105263, "grad_norm": 4.592624664306641, "learning_rate": 1.6365131578947369e-06, "loss": 0.0091, "step": 4068 }, { "epoch": 0.8365542763157895, "grad_norm": 17.4531307220459, "learning_rate": 1.6344572368421054e-06, "loss": 0.0397, "step": 4069 }, { "epoch": 0.8367598684210527, "grad_norm": 34.21488571166992, "learning_rate": 1.632401315789474e-06, "loss": 0.0967, "step": 4070 }, { "epoch": 0.8369654605263158, "grad_norm": 138.1781768798828, "learning_rate": 1.6303453947368422e-06, "loss": 0.2537, "step": 4071 }, { "epoch": 0.837171052631579, "grad_norm": 50.95158386230469, "learning_rate": 1.6282894736842105e-06, "loss": 0.3037, "step": 4072 }, { "epoch": 0.8373766447368421, "grad_norm": 29.49737548828125, "learning_rate": 1.626233552631579e-06, "loss": 0.1685, "step": 4073 }, { "epoch": 0.8375822368421053, "grad_norm": 13.69301700592041, "learning_rate": 1.6241776315789476e-06, "loss": 0.0485, "step": 4074 }, { "epoch": 0.8377878289473685, "grad_norm": 260.7752990722656, "learning_rate": 1.6221217105263159e-06, "loss": 4.2969, "step": 4075 }, { "epoch": 0.8379934210526315, "grad_norm": 6.30063533782959, "learning_rate": 1.6200657894736844e-06, "loss": 0.018, "step": 4076 }, { "epoch": 0.8381990131578947, "grad_norm": 46.152244567871094, "learning_rate": 1.6180098684210527e-06, "loss": 0.1531, "step": 4077 }, { "epoch": 0.8384046052631579, "grad_norm": 52.986759185791016, "learning_rate": 1.6159539473684214e-06, "loss": 0.8225, "step": 4078 }, { "epoch": 0.838610197368421, "grad_norm": 89.96046447753906, "learning_rate": 1.6138980263157897e-06, "loss": 1.0232, "step": 4079 }, { "epoch": 0.8388157894736842, "grad_norm": 14.562856674194336, "learning_rate": 1.611842105263158e-06, "loss": 0.0604, "step": 4080 }, { "epoch": 0.8390213815789473, "grad_norm": 12.723231315612793, "learning_rate": 1.6097861842105263e-06, "loss": 0.0176, "step": 4081 }, { "epoch": 0.8392269736842105, "grad_norm": 24.029449462890625, "learning_rate": 1.6077302631578948e-06, "loss": 0.0558, "step": 4082 }, { "epoch": 0.8394325657894737, "grad_norm": 17.460296630859375, "learning_rate": 1.6056743421052634e-06, "loss": 0.1028, "step": 4083 }, { "epoch": 0.8396381578947368, "grad_norm": 113.18022155761719, "learning_rate": 1.6036184210526319e-06, "loss": 0.7773, "step": 4084 }, { "epoch": 0.83984375, "grad_norm": 33.49111557006836, "learning_rate": 1.6015625000000002e-06, "loss": 0.1227, "step": 4085 }, { "epoch": 0.8400493421052632, "grad_norm": 6.417932510375977, "learning_rate": 1.5995065789473685e-06, "loss": 0.0123, "step": 4086 }, { "epoch": 0.8402549342105263, "grad_norm": 85.36563110351562, "learning_rate": 1.5974506578947368e-06, "loss": 0.7494, "step": 4087 }, { "epoch": 0.8404605263157895, "grad_norm": 19.224332809448242, "learning_rate": 1.5953947368421055e-06, "loss": 0.0448, "step": 4088 }, { "epoch": 0.8406661184210527, "grad_norm": 88.58084106445312, "learning_rate": 1.5933388157894738e-06, "loss": 0.2217, "step": 4089 }, { "epoch": 0.8408717105263158, "grad_norm": 56.91676330566406, "learning_rate": 1.5912828947368421e-06, "loss": 0.1746, "step": 4090 }, { "epoch": 0.841077302631579, "grad_norm": 150.23553466796875, "learning_rate": 1.5892269736842106e-06, "loss": 0.4966, "step": 4091 }, { "epoch": 0.8412828947368421, "grad_norm": 8.366230010986328, "learning_rate": 1.587171052631579e-06, "loss": 0.0176, "step": 4092 }, { "epoch": 0.8414884868421053, "grad_norm": 7.15810489654541, "learning_rate": 1.5851151315789477e-06, "loss": 0.0161, "step": 4093 }, { "epoch": 0.8416940789473685, "grad_norm": 150.18899536132812, "learning_rate": 1.583059210526316e-06, "loss": 1.3362, "step": 4094 }, { "epoch": 0.8418996710526315, "grad_norm": 83.8276138305664, "learning_rate": 1.5810032894736843e-06, "loss": 1.2619, "step": 4095 }, { "epoch": 0.8421052631578947, "grad_norm": 155.04173278808594, "learning_rate": 1.5789473684210526e-06, "loss": 0.5869, "step": 4096 }, { "epoch": 0.8423108552631579, "grad_norm": 21.906641006469727, "learning_rate": 1.5768914473684213e-06, "loss": 0.0566, "step": 4097 }, { "epoch": 0.842516447368421, "grad_norm": 9.05708122253418, "learning_rate": 1.5748355263157896e-06, "loss": 0.0179, "step": 4098 }, { "epoch": 0.8427220394736842, "grad_norm": 279.4018859863281, "learning_rate": 1.5727796052631581e-06, "loss": 1.1006, "step": 4099 }, { "epoch": 0.8429276315789473, "grad_norm": 14.029984474182129, "learning_rate": 1.5707236842105264e-06, "loss": 0.0386, "step": 4100 }, { "epoch": 0.8431332236842105, "grad_norm": 10.055534362792969, "learning_rate": 1.5686677631578947e-06, "loss": 0.0172, "step": 4101 }, { "epoch": 0.8433388157894737, "grad_norm": 20.803895950317383, "learning_rate": 1.5666118421052635e-06, "loss": 0.0896, "step": 4102 }, { "epoch": 0.8435444078947368, "grad_norm": 244.06739807128906, "learning_rate": 1.5645559210526318e-06, "loss": 0.6862, "step": 4103 }, { "epoch": 0.84375, "grad_norm": 21.585704803466797, "learning_rate": 1.5625e-06, "loss": 0.0707, "step": 4104 }, { "epoch": 0.8439555921052632, "grad_norm": 20.54097557067871, "learning_rate": 1.5604440789473686e-06, "loss": 0.0671, "step": 4105 }, { "epoch": 0.8441611842105263, "grad_norm": 16.797290802001953, "learning_rate": 1.558388157894737e-06, "loss": 0.0554, "step": 4106 }, { "epoch": 0.8443667763157895, "grad_norm": 20.5579833984375, "learning_rate": 1.5563322368421054e-06, "loss": 0.071, "step": 4107 }, { "epoch": 0.8445723684210527, "grad_norm": 25.208255767822266, "learning_rate": 1.554276315789474e-06, "loss": 0.0714, "step": 4108 }, { "epoch": 0.8447779605263158, "grad_norm": 78.24785614013672, "learning_rate": 1.5522203947368422e-06, "loss": 0.3481, "step": 4109 }, { "epoch": 0.844983552631579, "grad_norm": 34.088260650634766, "learning_rate": 1.5501644736842105e-06, "loss": 0.1379, "step": 4110 }, { "epoch": 0.8451891447368421, "grad_norm": 19.50751495361328, "learning_rate": 1.5481085526315788e-06, "loss": 0.033, "step": 4111 }, { "epoch": 0.8453947368421053, "grad_norm": 25.20381736755371, "learning_rate": 1.5460526315789476e-06, "loss": 0.0996, "step": 4112 }, { "epoch": 0.8456003289473685, "grad_norm": 7.294246673583984, "learning_rate": 1.5439967105263159e-06, "loss": 0.0291, "step": 4113 }, { "epoch": 0.8458059210526315, "grad_norm": 8.84616470336914, "learning_rate": 1.5419407894736844e-06, "loss": 0.0224, "step": 4114 }, { "epoch": 0.8460115131578947, "grad_norm": 303.89990234375, "learning_rate": 1.5398848684210527e-06, "loss": 1.7757, "step": 4115 }, { "epoch": 0.8462171052631579, "grad_norm": 7.176075458526611, "learning_rate": 1.5378289473684212e-06, "loss": 0.0304, "step": 4116 }, { "epoch": 0.846422697368421, "grad_norm": 52.0678825378418, "learning_rate": 1.5357730263157897e-06, "loss": 0.1864, "step": 4117 }, { "epoch": 0.8466282894736842, "grad_norm": 22.09749984741211, "learning_rate": 1.533717105263158e-06, "loss": 0.0712, "step": 4118 }, { "epoch": 0.8468338815789473, "grad_norm": 16.80439567565918, "learning_rate": 1.5316611842105263e-06, "loss": 0.0261, "step": 4119 }, { "epoch": 0.8470394736842105, "grad_norm": 64.1692123413086, "learning_rate": 1.5296052631578948e-06, "loss": 0.0726, "step": 4120 }, { "epoch": 0.8472450657894737, "grad_norm": 258.8418884277344, "learning_rate": 1.5275493421052634e-06, "loss": 0.4061, "step": 4121 }, { "epoch": 0.8474506578947368, "grad_norm": 22.010547637939453, "learning_rate": 1.5254934210526317e-06, "loss": 0.0801, "step": 4122 }, { "epoch": 0.84765625, "grad_norm": 11.040350914001465, "learning_rate": 1.5234375000000002e-06, "loss": 0.0285, "step": 4123 }, { "epoch": 0.8478618421052632, "grad_norm": 2.402625322341919, "learning_rate": 1.5213815789473685e-06, "loss": 0.007, "step": 4124 }, { "epoch": 0.8480674342105263, "grad_norm": 12.526683807373047, "learning_rate": 1.5193256578947368e-06, "loss": 0.0391, "step": 4125 }, { "epoch": 0.8482730263157895, "grad_norm": 96.041259765625, "learning_rate": 1.5172697368421055e-06, "loss": 0.5618, "step": 4126 }, { "epoch": 0.8484786184210527, "grad_norm": 154.77809143066406, "learning_rate": 1.5152138157894738e-06, "loss": 1.2793, "step": 4127 }, { "epoch": 0.8486842105263158, "grad_norm": 269.17822265625, "learning_rate": 1.5131578947368421e-06, "loss": 1.5001, "step": 4128 }, { "epoch": 0.848889802631579, "grad_norm": 29.333026885986328, "learning_rate": 1.5111019736842106e-06, "loss": 0.0692, "step": 4129 }, { "epoch": 0.8490953947368421, "grad_norm": 202.80734252929688, "learning_rate": 1.509046052631579e-06, "loss": 1.6445, "step": 4130 }, { "epoch": 0.8493009868421053, "grad_norm": 8.473822593688965, "learning_rate": 1.5069901315789477e-06, "loss": 0.0193, "step": 4131 }, { "epoch": 0.8495065789473685, "grad_norm": 47.95869064331055, "learning_rate": 1.504934210526316e-06, "loss": 0.0945, "step": 4132 }, { "epoch": 0.8497121710526315, "grad_norm": 6.150716781616211, "learning_rate": 1.5028782894736843e-06, "loss": 0.0182, "step": 4133 }, { "epoch": 0.8499177631578947, "grad_norm": 25.313199996948242, "learning_rate": 1.5008223684210526e-06, "loss": 0.0713, "step": 4134 }, { "epoch": 0.8501233552631579, "grad_norm": 28.03720474243164, "learning_rate": 1.4987664473684213e-06, "loss": 0.0686, "step": 4135 }, { "epoch": 0.850328947368421, "grad_norm": 36.54502868652344, "learning_rate": 1.4967105263157896e-06, "loss": 0.123, "step": 4136 }, { "epoch": 0.8505345394736842, "grad_norm": 18.765226364135742, "learning_rate": 1.4946546052631581e-06, "loss": 0.0325, "step": 4137 }, { "epoch": 0.8507401315789473, "grad_norm": 437.4544982910156, "learning_rate": 1.4925986842105264e-06, "loss": 1.457, "step": 4138 }, { "epoch": 0.8509457236842105, "grad_norm": 77.14716339111328, "learning_rate": 1.4905427631578947e-06, "loss": 0.6646, "step": 4139 }, { "epoch": 0.8511513157894737, "grad_norm": 8.818377494812012, "learning_rate": 1.4884868421052635e-06, "loss": 0.0214, "step": 4140 }, { "epoch": 0.8513569078947368, "grad_norm": 94.06212615966797, "learning_rate": 1.4864309210526318e-06, "loss": 0.7498, "step": 4141 }, { "epoch": 0.8515625, "grad_norm": 27.569242477416992, "learning_rate": 1.484375e-06, "loss": 0.1152, "step": 4142 }, { "epoch": 0.8517680921052632, "grad_norm": 39.29465866088867, "learning_rate": 1.4823190789473684e-06, "loss": 0.1458, "step": 4143 }, { "epoch": 0.8519736842105263, "grad_norm": 198.1977081298828, "learning_rate": 1.480263157894737e-06, "loss": 0.8022, "step": 4144 }, { "epoch": 0.8521792763157895, "grad_norm": 7.826861381530762, "learning_rate": 1.4782072368421054e-06, "loss": 0.0344, "step": 4145 }, { "epoch": 0.8523848684210527, "grad_norm": 2.399611711502075, "learning_rate": 1.476151315789474e-06, "loss": 0.0093, "step": 4146 }, { "epoch": 0.8525904605263158, "grad_norm": 61.404335021972656, "learning_rate": 1.4740953947368422e-06, "loss": 0.2711, "step": 4147 }, { "epoch": 0.852796052631579, "grad_norm": 48.419620513916016, "learning_rate": 1.4720394736842105e-06, "loss": 0.1913, "step": 4148 }, { "epoch": 0.8530016447368421, "grad_norm": 40.5113525390625, "learning_rate": 1.4699835526315788e-06, "loss": 0.1154, "step": 4149 }, { "epoch": 0.8532072368421053, "grad_norm": 18.529836654663086, "learning_rate": 1.4679276315789476e-06, "loss": 0.0706, "step": 4150 }, { "epoch": 0.8534128289473685, "grad_norm": 14.456820487976074, "learning_rate": 1.4658717105263159e-06, "loss": 0.0409, "step": 4151 }, { "epoch": 0.8536184210526315, "grad_norm": 110.6467056274414, "learning_rate": 1.4638157894736844e-06, "loss": 0.2871, "step": 4152 }, { "epoch": 0.8538240131578947, "grad_norm": 26.654563903808594, "learning_rate": 1.4617598684210527e-06, "loss": 0.1351, "step": 4153 }, { "epoch": 0.8540296052631579, "grad_norm": 348.35321044921875, "learning_rate": 1.4597039473684212e-06, "loss": 1.4951, "step": 4154 }, { "epoch": 0.854235197368421, "grad_norm": 56.60215377807617, "learning_rate": 1.4576480263157897e-06, "loss": 0.5842, "step": 4155 }, { "epoch": 0.8544407894736842, "grad_norm": 26.040132522583008, "learning_rate": 1.455592105263158e-06, "loss": 0.0489, "step": 4156 }, { "epoch": 0.8546463815789473, "grad_norm": 2.1207237243652344, "learning_rate": 1.4535361842105263e-06, "loss": 0.0085, "step": 4157 }, { "epoch": 0.8548519736842105, "grad_norm": 124.91786193847656, "learning_rate": 1.4514802631578949e-06, "loss": 0.3625, "step": 4158 }, { "epoch": 0.8550575657894737, "grad_norm": 69.20238494873047, "learning_rate": 1.4494243421052634e-06, "loss": 0.273, "step": 4159 }, { "epoch": 0.8552631578947368, "grad_norm": 12.30941390991211, "learning_rate": 1.4473684210526317e-06, "loss": 0.0231, "step": 4160 }, { "epoch": 0.85546875, "grad_norm": 4.121579647064209, "learning_rate": 1.4453125000000002e-06, "loss": 0.0132, "step": 4161 }, { "epoch": 0.8556743421052632, "grad_norm": 35.328392028808594, "learning_rate": 1.4432565789473685e-06, "loss": 0.0885, "step": 4162 }, { "epoch": 0.8558799342105263, "grad_norm": 91.65717315673828, "learning_rate": 1.4412006578947368e-06, "loss": 0.4304, "step": 4163 }, { "epoch": 0.8560855263157895, "grad_norm": 11.597101211547852, "learning_rate": 1.4391447368421055e-06, "loss": 0.0354, "step": 4164 }, { "epoch": 0.8562911184210527, "grad_norm": 84.08939361572266, "learning_rate": 1.4370888157894738e-06, "loss": 0.8141, "step": 4165 }, { "epoch": 0.8564967105263158, "grad_norm": 68.38386535644531, "learning_rate": 1.4350328947368421e-06, "loss": 0.4761, "step": 4166 }, { "epoch": 0.856702302631579, "grad_norm": 16.56690216064453, "learning_rate": 1.4329769736842107e-06, "loss": 0.0506, "step": 4167 }, { "epoch": 0.8569078947368421, "grad_norm": 39.40007400512695, "learning_rate": 1.430921052631579e-06, "loss": 0.0775, "step": 4168 }, { "epoch": 0.8571134868421053, "grad_norm": 37.20402908325195, "learning_rate": 1.4288651315789475e-06, "loss": 0.1089, "step": 4169 }, { "epoch": 0.8573190789473685, "grad_norm": 12.300044059753418, "learning_rate": 1.426809210526316e-06, "loss": 0.0292, "step": 4170 }, { "epoch": 0.8575246710526315, "grad_norm": 64.2070541381836, "learning_rate": 1.4247532894736843e-06, "loss": 0.1083, "step": 4171 }, { "epoch": 0.8577302631578947, "grad_norm": 13.745223045349121, "learning_rate": 1.4226973684210526e-06, "loss": 0.05, "step": 4172 }, { "epoch": 0.8579358552631579, "grad_norm": 143.2691192626953, "learning_rate": 1.4206414473684213e-06, "loss": 0.7048, "step": 4173 }, { "epoch": 0.858141447368421, "grad_norm": 152.4798583984375, "learning_rate": 1.4185855263157896e-06, "loss": 1.6641, "step": 4174 }, { "epoch": 0.8583470394736842, "grad_norm": 263.1554870605469, "learning_rate": 1.416529605263158e-06, "loss": 0.619, "step": 4175 }, { "epoch": 0.8585526315789473, "grad_norm": 186.8441925048828, "learning_rate": 1.4144736842105264e-06, "loss": 1.061, "step": 4176 }, { "epoch": 0.8587582236842105, "grad_norm": 3.3064169883728027, "learning_rate": 1.4124177631578948e-06, "loss": 0.0093, "step": 4177 }, { "epoch": 0.8589638157894737, "grad_norm": 172.69606018066406, "learning_rate": 1.4103618421052635e-06, "loss": 0.5527, "step": 4178 }, { "epoch": 0.8591694078947368, "grad_norm": 13.34320068359375, "learning_rate": 1.4083059210526318e-06, "loss": 0.0211, "step": 4179 }, { "epoch": 0.859375, "grad_norm": 8.231344223022461, "learning_rate": 1.40625e-06, "loss": 0.0153, "step": 4180 }, { "epoch": 0.8595805921052632, "grad_norm": 423.04339599609375, "learning_rate": 1.4041940789473684e-06, "loss": 0.8205, "step": 4181 }, { "epoch": 0.8597861842105263, "grad_norm": 18.691150665283203, "learning_rate": 1.402138157894737e-06, "loss": 0.106, "step": 4182 }, { "epoch": 0.8599917763157895, "grad_norm": 23.95819664001465, "learning_rate": 1.4000822368421054e-06, "loss": 0.0213, "step": 4183 }, { "epoch": 0.8601973684210527, "grad_norm": 66.67456817626953, "learning_rate": 1.398026315789474e-06, "loss": 0.3401, "step": 4184 }, { "epoch": 0.8604029605263158, "grad_norm": 274.5731506347656, "learning_rate": 1.3959703947368422e-06, "loss": 1.4774, "step": 4185 }, { "epoch": 0.860608552631579, "grad_norm": 24.713489532470703, "learning_rate": 1.3939144736842105e-06, "loss": 0.0275, "step": 4186 }, { "epoch": 0.8608141447368421, "grad_norm": 160.28378295898438, "learning_rate": 1.3918585526315789e-06, "loss": 0.8208, "step": 4187 }, { "epoch": 0.8610197368421053, "grad_norm": 20.087350845336914, "learning_rate": 1.3898026315789476e-06, "loss": 0.0424, "step": 4188 }, { "epoch": 0.8612253289473685, "grad_norm": 29.810941696166992, "learning_rate": 1.3877467105263159e-06, "loss": 0.0425, "step": 4189 }, { "epoch": 0.8614309210526315, "grad_norm": 6.778649806976318, "learning_rate": 1.3856907894736842e-06, "loss": 0.01, "step": 4190 }, { "epoch": 0.8616365131578947, "grad_norm": 5.800992488861084, "learning_rate": 1.3836348684210527e-06, "loss": 0.0076, "step": 4191 }, { "epoch": 0.8618421052631579, "grad_norm": 164.29014587402344, "learning_rate": 1.3815789473684212e-06, "loss": 0.5031, "step": 4192 }, { "epoch": 0.862047697368421, "grad_norm": 218.51132202148438, "learning_rate": 1.3795230263157897e-06, "loss": 0.3618, "step": 4193 }, { "epoch": 0.8622532894736842, "grad_norm": 86.2065658569336, "learning_rate": 1.377467105263158e-06, "loss": 0.4912, "step": 4194 }, { "epoch": 0.8624588815789473, "grad_norm": 69.60919952392578, "learning_rate": 1.3754111842105263e-06, "loss": 0.5269, "step": 4195 }, { "epoch": 0.8626644736842105, "grad_norm": 22.222612380981445, "learning_rate": 1.3733552631578946e-06, "loss": 0.0326, "step": 4196 }, { "epoch": 0.8628700657894737, "grad_norm": 370.60504150390625, "learning_rate": 1.3712993421052634e-06, "loss": 1.7422, "step": 4197 }, { "epoch": 0.8630756578947368, "grad_norm": 50.3830451965332, "learning_rate": 1.3692434210526317e-06, "loss": 0.3281, "step": 4198 }, { "epoch": 0.86328125, "grad_norm": 154.42898559570312, "learning_rate": 1.3671875000000002e-06, "loss": 0.5181, "step": 4199 }, { "epoch": 0.8634868421052632, "grad_norm": 32.83826446533203, "learning_rate": 1.3651315789473685e-06, "loss": 0.0839, "step": 4200 }, { "epoch": 0.8636924342105263, "grad_norm": 83.38299560546875, "learning_rate": 1.3630756578947368e-06, "loss": 0.3545, "step": 4201 }, { "epoch": 0.8638980263157895, "grad_norm": 114.11356353759766, "learning_rate": 1.3610197368421055e-06, "loss": 0.2734, "step": 4202 }, { "epoch": 0.8641036184210527, "grad_norm": 8.385477066040039, "learning_rate": 1.3589638157894738e-06, "loss": 0.021, "step": 4203 }, { "epoch": 0.8643092105263158, "grad_norm": 291.9954833984375, "learning_rate": 1.3569078947368421e-06, "loss": 1.3809, "step": 4204 }, { "epoch": 0.864514802631579, "grad_norm": 9.092930793762207, "learning_rate": 1.3548519736842107e-06, "loss": 0.03, "step": 4205 }, { "epoch": 0.8647203947368421, "grad_norm": 382.77154541015625, "learning_rate": 1.352796052631579e-06, "loss": 2.0157, "step": 4206 }, { "epoch": 0.8649259868421053, "grad_norm": 3.5622153282165527, "learning_rate": 1.3507401315789475e-06, "loss": 0.0147, "step": 4207 }, { "epoch": 0.8651315789473685, "grad_norm": 33.33682632446289, "learning_rate": 1.348684210526316e-06, "loss": 0.0467, "step": 4208 }, { "epoch": 0.8653371710526315, "grad_norm": 50.70518493652344, "learning_rate": 1.3466282894736843e-06, "loss": 0.1529, "step": 4209 }, { "epoch": 0.8655427631578947, "grad_norm": 284.1226501464844, "learning_rate": 1.3445723684210526e-06, "loss": 1.2246, "step": 4210 }, { "epoch": 0.8657483552631579, "grad_norm": 72.44793701171875, "learning_rate": 1.3425164473684213e-06, "loss": 0.6855, "step": 4211 }, { "epoch": 0.865953947368421, "grad_norm": 3.7982914447784424, "learning_rate": 1.3404605263157896e-06, "loss": 0.0075, "step": 4212 }, { "epoch": 0.8661595394736842, "grad_norm": 12.511398315429688, "learning_rate": 1.338404605263158e-06, "loss": 0.0325, "step": 4213 }, { "epoch": 0.8663651315789473, "grad_norm": 28.26056671142578, "learning_rate": 1.3363486842105265e-06, "loss": 0.0726, "step": 4214 }, { "epoch": 0.8665707236842105, "grad_norm": 7.149920463562012, "learning_rate": 1.3342927631578948e-06, "loss": 0.0245, "step": 4215 }, { "epoch": 0.8667763157894737, "grad_norm": 157.82989501953125, "learning_rate": 1.3322368421052635e-06, "loss": 0.3848, "step": 4216 }, { "epoch": 0.8669819078947368, "grad_norm": 138.7653350830078, "learning_rate": 1.3301809210526318e-06, "loss": 0.8708, "step": 4217 }, { "epoch": 0.8671875, "grad_norm": 17.69416046142578, "learning_rate": 1.328125e-06, "loss": 0.0342, "step": 4218 }, { "epoch": 0.8673930921052632, "grad_norm": 216.7586212158203, "learning_rate": 1.3260690789473684e-06, "loss": 0.6294, "step": 4219 }, { "epoch": 0.8675986842105263, "grad_norm": 193.25180053710938, "learning_rate": 1.324013157894737e-06, "loss": 1.3262, "step": 4220 }, { "epoch": 0.8678042763157895, "grad_norm": 20.667030334472656, "learning_rate": 1.3219572368421054e-06, "loss": 0.0554, "step": 4221 }, { "epoch": 0.8680098684210527, "grad_norm": 97.43656921386719, "learning_rate": 1.3199013157894737e-06, "loss": 0.2075, "step": 4222 }, { "epoch": 0.8682154605263158, "grad_norm": 117.93614196777344, "learning_rate": 1.3178453947368423e-06, "loss": 0.9126, "step": 4223 }, { "epoch": 0.868421052631579, "grad_norm": 8.40267276763916, "learning_rate": 1.3157894736842106e-06, "loss": 0.0231, "step": 4224 }, { "epoch": 0.8686266447368421, "grad_norm": 150.80422973632812, "learning_rate": 1.3137335526315789e-06, "loss": 1.0796, "step": 4225 }, { "epoch": 0.8688322368421053, "grad_norm": 8.248418807983398, "learning_rate": 1.3116776315789476e-06, "loss": 0.0344, "step": 4226 }, { "epoch": 0.8690378289473685, "grad_norm": 10.385671615600586, "learning_rate": 1.3096217105263159e-06, "loss": 0.0174, "step": 4227 }, { "epoch": 0.8692434210526315, "grad_norm": 68.62429809570312, "learning_rate": 1.3075657894736842e-06, "loss": 0.6073, "step": 4228 }, { "epoch": 0.8694490131578947, "grad_norm": 19.505733489990234, "learning_rate": 1.3055098684210527e-06, "loss": 0.0767, "step": 4229 }, { "epoch": 0.8696546052631579, "grad_norm": 130.40560913085938, "learning_rate": 1.3034539473684212e-06, "loss": 0.6367, "step": 4230 }, { "epoch": 0.869860197368421, "grad_norm": 202.49203491210938, "learning_rate": 1.3013980263157897e-06, "loss": 1.0117, "step": 4231 }, { "epoch": 0.8700657894736842, "grad_norm": 13.621561050415039, "learning_rate": 1.299342105263158e-06, "loss": 0.0309, "step": 4232 }, { "epoch": 0.8702713815789473, "grad_norm": 28.599430084228516, "learning_rate": 1.2972861842105264e-06, "loss": 0.0819, "step": 4233 }, { "epoch": 0.8704769736842105, "grad_norm": 381.72296142578125, "learning_rate": 1.2952302631578947e-06, "loss": 1.21, "step": 4234 }, { "epoch": 0.8706825657894737, "grad_norm": 4.924722671508789, "learning_rate": 1.2931743421052634e-06, "loss": 0.0087, "step": 4235 }, { "epoch": 0.8708881578947368, "grad_norm": 37.04420471191406, "learning_rate": 1.2911184210526317e-06, "loss": 0.3091, "step": 4236 }, { "epoch": 0.87109375, "grad_norm": 23.196121215820312, "learning_rate": 1.2890625000000002e-06, "loss": 0.0792, "step": 4237 }, { "epoch": 0.8712993421052632, "grad_norm": 16.660249710083008, "learning_rate": 1.2870065789473685e-06, "loss": 0.0407, "step": 4238 }, { "epoch": 0.8715049342105263, "grad_norm": 20.161922454833984, "learning_rate": 1.2849506578947368e-06, "loss": 0.0365, "step": 4239 }, { "epoch": 0.8717105263157895, "grad_norm": 98.5383529663086, "learning_rate": 1.2828947368421055e-06, "loss": 0.5342, "step": 4240 }, { "epoch": 0.8719161184210527, "grad_norm": 90.89311981201172, "learning_rate": 1.2808388157894738e-06, "loss": 0.3607, "step": 4241 }, { "epoch": 0.8721217105263158, "grad_norm": 20.391902923583984, "learning_rate": 1.2787828947368421e-06, "loss": 0.0471, "step": 4242 }, { "epoch": 0.872327302631579, "grad_norm": 329.98944091796875, "learning_rate": 1.2767269736842105e-06, "loss": 1.7228, "step": 4243 }, { "epoch": 0.8725328947368421, "grad_norm": 201.01641845703125, "learning_rate": 1.274671052631579e-06, "loss": 2.5293, "step": 4244 }, { "epoch": 0.8727384868421053, "grad_norm": 81.79631805419922, "learning_rate": 1.2726151315789475e-06, "loss": 1.2007, "step": 4245 }, { "epoch": 0.8729440789473685, "grad_norm": 9.782173156738281, "learning_rate": 1.270559210526316e-06, "loss": 0.0723, "step": 4246 }, { "epoch": 0.8731496710526315, "grad_norm": 50.44871520996094, "learning_rate": 1.2685032894736843e-06, "loss": 0.2817, "step": 4247 }, { "epoch": 0.8733552631578947, "grad_norm": 4.442914009094238, "learning_rate": 1.2664473684210526e-06, "loss": 0.0099, "step": 4248 }, { "epoch": 0.8735608552631579, "grad_norm": 76.2933349609375, "learning_rate": 1.2643914473684213e-06, "loss": 1.1714, "step": 4249 }, { "epoch": 0.873766447368421, "grad_norm": 140.9583740234375, "learning_rate": 1.2623355263157896e-06, "loss": 0.7622, "step": 4250 }, { "epoch": 0.8739720394736842, "grad_norm": 43.843902587890625, "learning_rate": 1.260279605263158e-06, "loss": 0.2473, "step": 4251 }, { "epoch": 0.8741776315789473, "grad_norm": 12.225296974182129, "learning_rate": 1.2582236842105265e-06, "loss": 0.0157, "step": 4252 }, { "epoch": 0.8743832236842105, "grad_norm": 208.9625244140625, "learning_rate": 1.2561677631578948e-06, "loss": 0.5834, "step": 4253 }, { "epoch": 0.8745888157894737, "grad_norm": 13.466287612915039, "learning_rate": 1.2541118421052633e-06, "loss": 0.0613, "step": 4254 }, { "epoch": 0.8747944078947368, "grad_norm": 37.58946990966797, "learning_rate": 1.2520559210526318e-06, "loss": 0.113, "step": 4255 }, { "epoch": 0.875, "grad_norm": 36.96464157104492, "learning_rate": 1.25e-06, "loss": 0.075, "step": 4256 }, { "epoch": 0.8752055921052632, "grad_norm": 43.28800964355469, "learning_rate": 1.2479440789473684e-06, "loss": 0.1849, "step": 4257 }, { "epoch": 0.8754111842105263, "grad_norm": 52.21931457519531, "learning_rate": 1.245888157894737e-06, "loss": 0.1562, "step": 4258 }, { "epoch": 0.8756167763157895, "grad_norm": 3.2959372997283936, "learning_rate": 1.2438322368421052e-06, "loss": 0.0133, "step": 4259 }, { "epoch": 0.8758223684210527, "grad_norm": 148.39540100097656, "learning_rate": 1.2417763157894737e-06, "loss": 0.8589, "step": 4260 }, { "epoch": 0.8760279605263158, "grad_norm": 47.599937438964844, "learning_rate": 1.2397203947368423e-06, "loss": 0.1492, "step": 4261 }, { "epoch": 0.876233552631579, "grad_norm": 3.685516357421875, "learning_rate": 1.2376644736842108e-06, "loss": 0.0143, "step": 4262 }, { "epoch": 0.8764391447368421, "grad_norm": 7.099997520446777, "learning_rate": 1.235608552631579e-06, "loss": 0.0177, "step": 4263 }, { "epoch": 0.8766447368421053, "grad_norm": 13.971193313598633, "learning_rate": 1.2335526315789474e-06, "loss": 0.0803, "step": 4264 }, { "epoch": 0.8768503289473685, "grad_norm": 37.36457061767578, "learning_rate": 1.231496710526316e-06, "loss": 0.1369, "step": 4265 }, { "epoch": 0.8770559210526315, "grad_norm": 36.587379455566406, "learning_rate": 1.2294407894736842e-06, "loss": 0.0733, "step": 4266 }, { "epoch": 0.8772615131578947, "grad_norm": 180.31039428710938, "learning_rate": 1.2273848684210527e-06, "loss": 0.6423, "step": 4267 }, { "epoch": 0.8774671052631579, "grad_norm": 28.166929244995117, "learning_rate": 1.2253289473684212e-06, "loss": 0.0914, "step": 4268 }, { "epoch": 0.877672697368421, "grad_norm": 10.138984680175781, "learning_rate": 1.2232730263157895e-06, "loss": 0.0396, "step": 4269 }, { "epoch": 0.8778782894736842, "grad_norm": 176.0272674560547, "learning_rate": 1.221217105263158e-06, "loss": 1.2939, "step": 4270 }, { "epoch": 0.8780838815789473, "grad_norm": 1.2634819746017456, "learning_rate": 1.2191611842105264e-06, "loss": 0.0061, "step": 4271 }, { "epoch": 0.8782894736842105, "grad_norm": 57.718299865722656, "learning_rate": 1.2171052631578949e-06, "loss": 0.2665, "step": 4272 }, { "epoch": 0.8784950657894737, "grad_norm": 92.94248962402344, "learning_rate": 1.2150493421052632e-06, "loss": 0.2877, "step": 4273 }, { "epoch": 0.8787006578947368, "grad_norm": 122.80567932128906, "learning_rate": 1.2129934210526317e-06, "loss": 1.1914, "step": 4274 }, { "epoch": 0.87890625, "grad_norm": 132.80438232421875, "learning_rate": 1.2109375e-06, "loss": 1.1644, "step": 4275 }, { "epoch": 0.8791118421052632, "grad_norm": 14.417152404785156, "learning_rate": 1.2088815789473685e-06, "loss": 0.0377, "step": 4276 }, { "epoch": 0.8793174342105263, "grad_norm": 34.748104095458984, "learning_rate": 1.206825657894737e-06, "loss": 0.1414, "step": 4277 }, { "epoch": 0.8795230263157895, "grad_norm": 45.461647033691406, "learning_rate": 1.2047697368421053e-06, "loss": 0.1604, "step": 4278 }, { "epoch": 0.8797286184210527, "grad_norm": 12.543782234191895, "learning_rate": 1.2027138157894739e-06, "loss": 0.045, "step": 4279 }, { "epoch": 0.8799342105263158, "grad_norm": 96.10321807861328, "learning_rate": 1.2006578947368422e-06, "loss": 0.5308, "step": 4280 }, { "epoch": 0.880139802631579, "grad_norm": 12.012904167175293, "learning_rate": 1.1986019736842107e-06, "loss": 0.033, "step": 4281 }, { "epoch": 0.8803453947368421, "grad_norm": 31.228557586669922, "learning_rate": 1.196546052631579e-06, "loss": 0.0774, "step": 4282 }, { "epoch": 0.8805509868421053, "grad_norm": 12.919451713562012, "learning_rate": 1.1944901315789475e-06, "loss": 0.0219, "step": 4283 }, { "epoch": 0.8807565789473685, "grad_norm": 132.05059814453125, "learning_rate": 1.192434210526316e-06, "loss": 0.5107, "step": 4284 }, { "epoch": 0.8809621710526315, "grad_norm": 5.8427886962890625, "learning_rate": 1.1903782894736843e-06, "loss": 0.0102, "step": 4285 }, { "epoch": 0.8811677631578947, "grad_norm": 7.174164295196533, "learning_rate": 1.1883223684210528e-06, "loss": 0.0234, "step": 4286 }, { "epoch": 0.8813733552631579, "grad_norm": 12.849000930786133, "learning_rate": 1.1862664473684211e-06, "loss": 0.0366, "step": 4287 }, { "epoch": 0.881578947368421, "grad_norm": 10.54331111907959, "learning_rate": 1.1842105263157894e-06, "loss": 0.0146, "step": 4288 }, { "epoch": 0.8817845394736842, "grad_norm": 60.1806640625, "learning_rate": 1.182154605263158e-06, "loss": 0.4894, "step": 4289 }, { "epoch": 0.8819901315789473, "grad_norm": 34.218143463134766, "learning_rate": 1.1800986842105265e-06, "loss": 0.104, "step": 4290 }, { "epoch": 0.8821957236842105, "grad_norm": 148.39346313476562, "learning_rate": 1.1780427631578948e-06, "loss": 0.7871, "step": 4291 }, { "epoch": 0.8824013157894737, "grad_norm": 25.80231475830078, "learning_rate": 1.1759868421052633e-06, "loss": 0.0865, "step": 4292 }, { "epoch": 0.8826069078947368, "grad_norm": 401.84515380859375, "learning_rate": 1.1739309210526318e-06, "loss": 4.0352, "step": 4293 }, { "epoch": 0.8828125, "grad_norm": 103.99893188476562, "learning_rate": 1.1718750000000001e-06, "loss": 0.5068, "step": 4294 }, { "epoch": 0.8830180921052632, "grad_norm": 17.161266326904297, "learning_rate": 1.1698190789473684e-06, "loss": 0.025, "step": 4295 }, { "epoch": 0.8832236842105263, "grad_norm": 78.78156280517578, "learning_rate": 1.167763157894737e-06, "loss": 0.2455, "step": 4296 }, { "epoch": 0.8834292763157895, "grad_norm": 72.79710388183594, "learning_rate": 1.1657072368421052e-06, "loss": 0.6194, "step": 4297 }, { "epoch": 0.8836348684210527, "grad_norm": 79.38926696777344, "learning_rate": 1.1636513157894737e-06, "loss": 0.8521, "step": 4298 }, { "epoch": 0.8838404605263158, "grad_norm": 191.69525146484375, "learning_rate": 1.1615953947368423e-06, "loss": 0.6074, "step": 4299 }, { "epoch": 0.884046052631579, "grad_norm": 110.6837387084961, "learning_rate": 1.1595394736842108e-06, "loss": 0.4014, "step": 4300 }, { "epoch": 0.8842516447368421, "grad_norm": 35.22356033325195, "learning_rate": 1.157483552631579e-06, "loss": 0.1155, "step": 4301 }, { "epoch": 0.8844572368421053, "grad_norm": 3.0745580196380615, "learning_rate": 1.1554276315789474e-06, "loss": 0.0048, "step": 4302 }, { "epoch": 0.8846628289473685, "grad_norm": 4.311916351318359, "learning_rate": 1.153371710526316e-06, "loss": 0.0137, "step": 4303 }, { "epoch": 0.8848684210526315, "grad_norm": 38.196868896484375, "learning_rate": 1.1513157894736842e-06, "loss": 0.0543, "step": 4304 }, { "epoch": 0.8850740131578947, "grad_norm": 9.1575927734375, "learning_rate": 1.1492598684210527e-06, "loss": 0.0219, "step": 4305 }, { "epoch": 0.8852796052631579, "grad_norm": 380.3833312988281, "learning_rate": 1.1472039473684212e-06, "loss": 1.3453, "step": 4306 }, { "epoch": 0.885485197368421, "grad_norm": 123.9330062866211, "learning_rate": 1.1451480263157895e-06, "loss": 0.2686, "step": 4307 }, { "epoch": 0.8856907894736842, "grad_norm": 3.111504554748535, "learning_rate": 1.143092105263158e-06, "loss": 0.0154, "step": 4308 }, { "epoch": 0.8858963815789473, "grad_norm": 133.60597229003906, "learning_rate": 1.1410361842105264e-06, "loss": 0.8477, "step": 4309 }, { "epoch": 0.8861019736842105, "grad_norm": 3.7995033264160156, "learning_rate": 1.1389802631578949e-06, "loss": 0.012, "step": 4310 }, { "epoch": 0.8863075657894737, "grad_norm": 172.6561279296875, "learning_rate": 1.1369243421052632e-06, "loss": 1.1367, "step": 4311 }, { "epoch": 0.8865131578947368, "grad_norm": 51.41292190551758, "learning_rate": 1.1348684210526317e-06, "loss": 0.1129, "step": 4312 }, { "epoch": 0.88671875, "grad_norm": 77.44418334960938, "learning_rate": 1.1328125e-06, "loss": 0.4517, "step": 4313 }, { "epoch": 0.8869243421052632, "grad_norm": 76.51758575439453, "learning_rate": 1.1307565789473685e-06, "loss": 0.403, "step": 4314 }, { "epoch": 0.8871299342105263, "grad_norm": 54.24419021606445, "learning_rate": 1.128700657894737e-06, "loss": 0.1904, "step": 4315 }, { "epoch": 0.8873355263157895, "grad_norm": 296.7976379394531, "learning_rate": 1.1266447368421053e-06, "loss": 1.1402, "step": 4316 }, { "epoch": 0.8875411184210527, "grad_norm": 4.682039260864258, "learning_rate": 1.1245888157894739e-06, "loss": 0.0087, "step": 4317 }, { "epoch": 0.8877467105263158, "grad_norm": 20.44976234436035, "learning_rate": 1.1225328947368422e-06, "loss": 0.0774, "step": 4318 }, { "epoch": 0.887952302631579, "grad_norm": 6.719164848327637, "learning_rate": 1.1204769736842107e-06, "loss": 0.011, "step": 4319 }, { "epoch": 0.8881578947368421, "grad_norm": 13.01517391204834, "learning_rate": 1.118421052631579e-06, "loss": 0.048, "step": 4320 }, { "epoch": 0.8883634868421053, "grad_norm": 45.82166290283203, "learning_rate": 1.1163651315789475e-06, "loss": 0.1289, "step": 4321 }, { "epoch": 0.8885690789473685, "grad_norm": 250.495849609375, "learning_rate": 1.114309210526316e-06, "loss": 2.6406, "step": 4322 }, { "epoch": 0.8887746710526315, "grad_norm": 9.358997344970703, "learning_rate": 1.1122532894736843e-06, "loss": 0.0274, "step": 4323 }, { "epoch": 0.8889802631578947, "grad_norm": 130.5025177001953, "learning_rate": 1.1101973684210528e-06, "loss": 0.215, "step": 4324 }, { "epoch": 0.8891858552631579, "grad_norm": 22.93352699279785, "learning_rate": 1.1081414473684211e-06, "loss": 0.0448, "step": 4325 }, { "epoch": 0.889391447368421, "grad_norm": 23.475440979003906, "learning_rate": 1.1060855263157894e-06, "loss": 0.0676, "step": 4326 }, { "epoch": 0.8895970394736842, "grad_norm": 36.97688674926758, "learning_rate": 1.104029605263158e-06, "loss": 0.1165, "step": 4327 }, { "epoch": 0.8898026315789473, "grad_norm": 9.141515731811523, "learning_rate": 1.1019736842105263e-06, "loss": 0.017, "step": 4328 }, { "epoch": 0.8900082236842105, "grad_norm": 18.561616897583008, "learning_rate": 1.0999177631578948e-06, "loss": 0.0612, "step": 4329 }, { "epoch": 0.8902138157894737, "grad_norm": 310.7505798339844, "learning_rate": 1.0978618421052633e-06, "loss": 3.0469, "step": 4330 }, { "epoch": 0.8904194078947368, "grad_norm": 18.67057228088379, "learning_rate": 1.0958059210526318e-06, "loss": 0.0662, "step": 4331 }, { "epoch": 0.890625, "grad_norm": 43.458587646484375, "learning_rate": 1.0937500000000001e-06, "loss": 0.2458, "step": 4332 }, { "epoch": 0.8908305921052632, "grad_norm": 57.703182220458984, "learning_rate": 1.0916940789473684e-06, "loss": 0.3494, "step": 4333 }, { "epoch": 0.8910361842105263, "grad_norm": 2.839923620223999, "learning_rate": 1.089638157894737e-06, "loss": 0.0083, "step": 4334 }, { "epoch": 0.8912417763157895, "grad_norm": 21.422710418701172, "learning_rate": 1.0875822368421052e-06, "loss": 0.0463, "step": 4335 }, { "epoch": 0.8914473684210527, "grad_norm": 66.96261596679688, "learning_rate": 1.0855263157894738e-06, "loss": 0.2203, "step": 4336 }, { "epoch": 0.8916529605263158, "grad_norm": 3.6795055866241455, "learning_rate": 1.0834703947368423e-06, "loss": 0.0084, "step": 4337 }, { "epoch": 0.891858552631579, "grad_norm": 45.17431640625, "learning_rate": 1.0814144736842106e-06, "loss": 0.113, "step": 4338 }, { "epoch": 0.8920641447368421, "grad_norm": 100.57147979736328, "learning_rate": 1.079358552631579e-06, "loss": 0.7645, "step": 4339 }, { "epoch": 0.8922697368421053, "grad_norm": 7.693138122558594, "learning_rate": 1.0773026315789474e-06, "loss": 0.0198, "step": 4340 }, { "epoch": 0.8924753289473685, "grad_norm": 207.00035095214844, "learning_rate": 1.075246710526316e-06, "loss": 1.2514, "step": 4341 }, { "epoch": 0.8926809210526315, "grad_norm": 22.846376419067383, "learning_rate": 1.0731907894736842e-06, "loss": 0.0668, "step": 4342 }, { "epoch": 0.8928865131578947, "grad_norm": 37.307472229003906, "learning_rate": 1.0711348684210527e-06, "loss": 0.1434, "step": 4343 }, { "epoch": 0.8930921052631579, "grad_norm": 156.57522583007812, "learning_rate": 1.069078947368421e-06, "loss": 0.8574, "step": 4344 }, { "epoch": 0.893297697368421, "grad_norm": 190.9065399169922, "learning_rate": 1.0670230263157896e-06, "loss": 0.5244, "step": 4345 }, { "epoch": 0.8935032894736842, "grad_norm": 108.37313842773438, "learning_rate": 1.064967105263158e-06, "loss": 0.2937, "step": 4346 }, { "epoch": 0.8937088815789473, "grad_norm": 101.62408447265625, "learning_rate": 1.0629111842105264e-06, "loss": 0.4199, "step": 4347 }, { "epoch": 0.8939144736842105, "grad_norm": 43.13327407836914, "learning_rate": 1.0608552631578949e-06, "loss": 0.145, "step": 4348 }, { "epoch": 0.8941200657894737, "grad_norm": 89.86854553222656, "learning_rate": 1.0587993421052632e-06, "loss": 0.5015, "step": 4349 }, { "epoch": 0.8943256578947368, "grad_norm": 15.610183715820312, "learning_rate": 1.0567434210526317e-06, "loss": 0.0354, "step": 4350 }, { "epoch": 0.89453125, "grad_norm": 36.630828857421875, "learning_rate": 1.0546875e-06, "loss": 0.1094, "step": 4351 }, { "epoch": 0.8947368421052632, "grad_norm": 83.89944458007812, "learning_rate": 1.0526315789473685e-06, "loss": 1.8203, "step": 4352 }, { "epoch": 0.8949424342105263, "grad_norm": 69.71206665039062, "learning_rate": 1.050575657894737e-06, "loss": 0.634, "step": 4353 }, { "epoch": 0.8951480263157895, "grad_norm": 31.017841339111328, "learning_rate": 1.0485197368421053e-06, "loss": 0.0597, "step": 4354 }, { "epoch": 0.8953536184210527, "grad_norm": 26.602537155151367, "learning_rate": 1.0464638157894739e-06, "loss": 0.0942, "step": 4355 }, { "epoch": 0.8955592105263158, "grad_norm": 77.94243621826172, "learning_rate": 1.0444078947368422e-06, "loss": 0.2227, "step": 4356 }, { "epoch": 0.895764802631579, "grad_norm": 3.2677054405212402, "learning_rate": 1.0423519736842107e-06, "loss": 0.009, "step": 4357 }, { "epoch": 0.8959703947368421, "grad_norm": 23.080514907836914, "learning_rate": 1.040296052631579e-06, "loss": 0.0406, "step": 4358 }, { "epoch": 0.8961759868421053, "grad_norm": 14.540717124938965, "learning_rate": 1.0382401315789475e-06, "loss": 0.0425, "step": 4359 }, { "epoch": 0.8963815789473685, "grad_norm": 52.73678970336914, "learning_rate": 1.0361842105263158e-06, "loss": 0.1562, "step": 4360 }, { "epoch": 0.8965871710526315, "grad_norm": 24.59278678894043, "learning_rate": 1.0341282894736843e-06, "loss": 0.0287, "step": 4361 }, { "epoch": 0.8967927631578947, "grad_norm": 99.65592193603516, "learning_rate": 1.0320723684210528e-06, "loss": 1.0942, "step": 4362 }, { "epoch": 0.8969983552631579, "grad_norm": 62.729591369628906, "learning_rate": 1.0300164473684211e-06, "loss": 0.7056, "step": 4363 }, { "epoch": 0.897203947368421, "grad_norm": 122.19451904296875, "learning_rate": 1.0279605263157895e-06, "loss": 0.5606, "step": 4364 }, { "epoch": 0.8974095394736842, "grad_norm": 181.2677764892578, "learning_rate": 1.025904605263158e-06, "loss": 0.7614, "step": 4365 }, { "epoch": 0.8976151315789473, "grad_norm": 343.4331970214844, "learning_rate": 1.0238486842105263e-06, "loss": 3.499, "step": 4366 }, { "epoch": 0.8978207236842105, "grad_norm": 34.850521087646484, "learning_rate": 1.0217927631578948e-06, "loss": 0.0775, "step": 4367 }, { "epoch": 0.8980263157894737, "grad_norm": 27.721126556396484, "learning_rate": 1.0197368421052633e-06, "loss": 0.1355, "step": 4368 }, { "epoch": 0.8982319078947368, "grad_norm": 67.82433319091797, "learning_rate": 1.0176809210526318e-06, "loss": 0.853, "step": 4369 }, { "epoch": 0.8984375, "grad_norm": 108.45392608642578, "learning_rate": 1.0156250000000001e-06, "loss": 0.4319, "step": 4370 }, { "epoch": 0.8986430921052632, "grad_norm": 171.71385192871094, "learning_rate": 1.0135690789473684e-06, "loss": 0.2576, "step": 4371 }, { "epoch": 0.8988486842105263, "grad_norm": 19.292911529541016, "learning_rate": 1.011513157894737e-06, "loss": 0.0476, "step": 4372 }, { "epoch": 0.8990542763157895, "grad_norm": 13.94814682006836, "learning_rate": 1.0094572368421052e-06, "loss": 0.0279, "step": 4373 }, { "epoch": 0.8992598684210527, "grad_norm": 91.97233581542969, "learning_rate": 1.0074013157894738e-06, "loss": 0.1406, "step": 4374 }, { "epoch": 0.8994654605263158, "grad_norm": 219.91867065429688, "learning_rate": 1.005345394736842e-06, "loss": 0.791, "step": 4375 }, { "epoch": 0.899671052631579, "grad_norm": 87.6978530883789, "learning_rate": 1.0032894736842106e-06, "loss": 0.4199, "step": 4376 }, { "epoch": 0.8998766447368421, "grad_norm": 13.618881225585938, "learning_rate": 1.001233552631579e-06, "loss": 0.0593, "step": 4377 }, { "epoch": 0.9000822368421053, "grad_norm": 34.98102569580078, "learning_rate": 9.991776315789474e-07, "loss": 0.167, "step": 4378 }, { "epoch": 0.9002878289473685, "grad_norm": 272.9820251464844, "learning_rate": 9.97121710526316e-07, "loss": 3.007, "step": 4379 }, { "epoch": 0.9004934210526315, "grad_norm": 264.06787109375, "learning_rate": 9.950657894736842e-07, "loss": 2.4276, "step": 4380 }, { "epoch": 0.9006990131578947, "grad_norm": 39.13681411743164, "learning_rate": 9.930098684210527e-07, "loss": 0.0981, "step": 4381 }, { "epoch": 0.9009046052631579, "grad_norm": 9.931355476379395, "learning_rate": 9.90953947368421e-07, "loss": 0.0145, "step": 4382 }, { "epoch": 0.901110197368421, "grad_norm": 87.60244750976562, "learning_rate": 9.888980263157896e-07, "loss": 0.8062, "step": 4383 }, { "epoch": 0.9013157894736842, "grad_norm": 35.329063415527344, "learning_rate": 9.86842105263158e-07, "loss": 0.0742, "step": 4384 }, { "epoch": 0.9015213815789473, "grad_norm": 19.36434555053711, "learning_rate": 9.847861842105264e-07, "loss": 0.0505, "step": 4385 }, { "epoch": 0.9017269736842105, "grad_norm": 59.76234436035156, "learning_rate": 9.82730263157895e-07, "loss": 0.245, "step": 4386 }, { "epoch": 0.9019325657894737, "grad_norm": 4.298952579498291, "learning_rate": 9.806743421052632e-07, "loss": 0.0127, "step": 4387 }, { "epoch": 0.9021381578947368, "grad_norm": 94.5030517578125, "learning_rate": 9.786184210526317e-07, "loss": 0.8564, "step": 4388 }, { "epoch": 0.90234375, "grad_norm": 36.15732955932617, "learning_rate": 9.765625e-07, "loss": 0.1385, "step": 4389 }, { "epoch": 0.9025493421052632, "grad_norm": 582.646484375, "learning_rate": 9.745065789473685e-07, "loss": 4.6052, "step": 4390 }, { "epoch": 0.9027549342105263, "grad_norm": 45.41581344604492, "learning_rate": 9.724506578947368e-07, "loss": 0.0825, "step": 4391 }, { "epoch": 0.9029605263157895, "grad_norm": 33.14326095581055, "learning_rate": 9.703947368421054e-07, "loss": 0.1317, "step": 4392 }, { "epoch": 0.9031661184210527, "grad_norm": 7.628839015960693, "learning_rate": 9.683388157894739e-07, "loss": 0.0117, "step": 4393 }, { "epoch": 0.9033717105263158, "grad_norm": 115.70303344726562, "learning_rate": 9.662828947368422e-07, "loss": 0.8208, "step": 4394 }, { "epoch": 0.903577302631579, "grad_norm": 6.43751859664917, "learning_rate": 9.642269736842107e-07, "loss": 0.012, "step": 4395 }, { "epoch": 0.9037828947368421, "grad_norm": 87.76516723632812, "learning_rate": 9.62171052631579e-07, "loss": 0.7846, "step": 4396 }, { "epoch": 0.9039884868421053, "grad_norm": 148.974365234375, "learning_rate": 9.601151315789473e-07, "loss": 1.4844, "step": 4397 }, { "epoch": 0.9041940789473685, "grad_norm": 64.55545806884766, "learning_rate": 9.580592105263158e-07, "loss": 0.9043, "step": 4398 }, { "epoch": 0.9043996710526315, "grad_norm": 50.73778533935547, "learning_rate": 9.560032894736843e-07, "loss": 0.4673, "step": 4399 }, { "epoch": 0.9046052631578947, "grad_norm": 55.30797576904297, "learning_rate": 9.539473684210528e-07, "loss": 0.2257, "step": 4400 }, { "epoch": 0.9048108552631579, "grad_norm": 287.81353759765625, "learning_rate": 9.518914473684212e-07, "loss": 0.4871, "step": 4401 }, { "epoch": 0.905016447368421, "grad_norm": 133.29666137695312, "learning_rate": 9.498355263157895e-07, "loss": 0.5555, "step": 4402 }, { "epoch": 0.9052220394736842, "grad_norm": 47.087833404541016, "learning_rate": 9.47779605263158e-07, "loss": 0.2186, "step": 4403 }, { "epoch": 0.9054276315789473, "grad_norm": 101.16303253173828, "learning_rate": 9.457236842105264e-07, "loss": 0.509, "step": 4404 }, { "epoch": 0.9056332236842105, "grad_norm": 416.8521423339844, "learning_rate": 9.436677631578949e-07, "loss": 0.5945, "step": 4405 }, { "epoch": 0.9058388157894737, "grad_norm": 40.327640533447266, "learning_rate": 9.416118421052632e-07, "loss": 0.1262, "step": 4406 }, { "epoch": 0.9060444078947368, "grad_norm": 8.920562744140625, "learning_rate": 9.395559210526317e-07, "loss": 0.0302, "step": 4407 }, { "epoch": 0.90625, "grad_norm": 80.50902557373047, "learning_rate": 9.375000000000001e-07, "loss": 0.4956, "step": 4408 }, { "epoch": 0.9064555921052632, "grad_norm": 21.152942657470703, "learning_rate": 9.354440789473684e-07, "loss": 0.1169, "step": 4409 }, { "epoch": 0.9066611842105263, "grad_norm": 8.35204029083252, "learning_rate": 9.33388157894737e-07, "loss": 0.0234, "step": 4410 }, { "epoch": 0.9068667763157895, "grad_norm": 213.04493713378906, "learning_rate": 9.313322368421053e-07, "loss": 1.3074, "step": 4411 }, { "epoch": 0.9070723684210527, "grad_norm": 3.3027291297912598, "learning_rate": 9.292763157894738e-07, "loss": 0.0099, "step": 4412 }, { "epoch": 0.9072779605263158, "grad_norm": 33.62758255004883, "learning_rate": 9.272203947368422e-07, "loss": 0.1331, "step": 4413 }, { "epoch": 0.907483552631579, "grad_norm": 783.193359375, "learning_rate": 9.251644736842107e-07, "loss": 1.4543, "step": 4414 }, { "epoch": 0.9076891447368421, "grad_norm": 85.72102355957031, "learning_rate": 9.23108552631579e-07, "loss": 0.535, "step": 4415 }, { "epoch": 0.9078947368421053, "grad_norm": 4.204531192779541, "learning_rate": 9.210526315789474e-07, "loss": 0.0114, "step": 4416 }, { "epoch": 0.9081003289473685, "grad_norm": 41.993038177490234, "learning_rate": 9.189967105263159e-07, "loss": 0.1123, "step": 4417 }, { "epoch": 0.9083059210526315, "grad_norm": 68.93370056152344, "learning_rate": 9.169407894736842e-07, "loss": 0.2375, "step": 4418 }, { "epoch": 0.9085115131578947, "grad_norm": 14.382918357849121, "learning_rate": 9.148848684210527e-07, "loss": 0.044, "step": 4419 }, { "epoch": 0.9087171052631579, "grad_norm": 104.46744537353516, "learning_rate": 9.128289473684212e-07, "loss": 0.4492, "step": 4420 }, { "epoch": 0.908922697368421, "grad_norm": 6.544371128082275, "learning_rate": 9.107730263157895e-07, "loss": 0.0173, "step": 4421 }, { "epoch": 0.9091282894736842, "grad_norm": 85.1509017944336, "learning_rate": 9.08717105263158e-07, "loss": 0.426, "step": 4422 }, { "epoch": 0.9093338815789473, "grad_norm": 36.06679153442383, "learning_rate": 9.066611842105264e-07, "loss": 0.1494, "step": 4423 }, { "epoch": 0.9095394736842105, "grad_norm": 154.7923583984375, "learning_rate": 9.046052631578948e-07, "loss": 0.782, "step": 4424 }, { "epoch": 0.9097450657894737, "grad_norm": 21.28101348876953, "learning_rate": 9.025493421052632e-07, "loss": 0.0715, "step": 4425 }, { "epoch": 0.9099506578947368, "grad_norm": 236.45045471191406, "learning_rate": 9.004934210526317e-07, "loss": 5.061, "step": 4426 }, { "epoch": 0.91015625, "grad_norm": 207.3253173828125, "learning_rate": 8.984375e-07, "loss": 4.0547, "step": 4427 }, { "epoch": 0.9103618421052632, "grad_norm": 53.237464904785156, "learning_rate": 8.963815789473684e-07, "loss": 0.1774, "step": 4428 }, { "epoch": 0.9105674342105263, "grad_norm": 57.36980056762695, "learning_rate": 8.94325657894737e-07, "loss": 0.2189, "step": 4429 }, { "epoch": 0.9107730263157895, "grad_norm": 23.437795639038086, "learning_rate": 8.922697368421053e-07, "loss": 0.1406, "step": 4430 }, { "epoch": 0.9109786184210527, "grad_norm": 13.446184158325195, "learning_rate": 8.902138157894738e-07, "loss": 0.0324, "step": 4431 }, { "epoch": 0.9111842105263158, "grad_norm": 23.58283042907715, "learning_rate": 8.881578947368422e-07, "loss": 0.0756, "step": 4432 }, { "epoch": 0.911389802631579, "grad_norm": 70.12969207763672, "learning_rate": 8.861019736842107e-07, "loss": 0.3862, "step": 4433 }, { "epoch": 0.9115953947368421, "grad_norm": 162.5630645751953, "learning_rate": 8.84046052631579e-07, "loss": 1.1636, "step": 4434 }, { "epoch": 0.9118009868421053, "grad_norm": 20.290063858032227, "learning_rate": 8.819901315789474e-07, "loss": 0.1049, "step": 4435 }, { "epoch": 0.9120065789473685, "grad_norm": 44.13910675048828, "learning_rate": 8.799342105263159e-07, "loss": 0.1121, "step": 4436 }, { "epoch": 0.9122121710526315, "grad_norm": 292.234375, "learning_rate": 8.778782894736842e-07, "loss": 1.3633, "step": 4437 }, { "epoch": 0.9124177631578947, "grad_norm": 23.627761840820312, "learning_rate": 8.758223684210527e-07, "loss": 0.0581, "step": 4438 }, { "epoch": 0.9126233552631579, "grad_norm": 22.44738006591797, "learning_rate": 8.737664473684212e-07, "loss": 0.0524, "step": 4439 }, { "epoch": 0.912828947368421, "grad_norm": 7.763708114624023, "learning_rate": 8.717105263157895e-07, "loss": 0.0169, "step": 4440 }, { "epoch": 0.9130345394736842, "grad_norm": 17.766826629638672, "learning_rate": 8.69654605263158e-07, "loss": 0.0778, "step": 4441 }, { "epoch": 0.9132401315789473, "grad_norm": 118.04015350341797, "learning_rate": 8.675986842105264e-07, "loss": 0.3544, "step": 4442 }, { "epoch": 0.9134457236842105, "grad_norm": 74.67247772216797, "learning_rate": 8.655427631578948e-07, "loss": 0.4997, "step": 4443 }, { "epoch": 0.9136513157894737, "grad_norm": 7.834837913513184, "learning_rate": 8.634868421052632e-07, "loss": 0.0153, "step": 4444 }, { "epoch": 0.9138569078947368, "grad_norm": 135.55667114257812, "learning_rate": 8.614309210526317e-07, "loss": 0.6562, "step": 4445 }, { "epoch": 0.9140625, "grad_norm": 22.02406883239746, "learning_rate": 8.59375e-07, "loss": 0.0607, "step": 4446 }, { "epoch": 0.9142680921052632, "grad_norm": 62.9029655456543, "learning_rate": 8.573190789473684e-07, "loss": 0.286, "step": 4447 }, { "epoch": 0.9144736842105263, "grad_norm": 96.89710998535156, "learning_rate": 8.55263157894737e-07, "loss": 0.6133, "step": 4448 }, { "epoch": 0.9146792763157895, "grad_norm": 53.88705825805664, "learning_rate": 8.532072368421053e-07, "loss": 0.1481, "step": 4449 }, { "epoch": 0.9148848684210527, "grad_norm": 72.58381652832031, "learning_rate": 8.511513157894738e-07, "loss": 0.4541, "step": 4450 }, { "epoch": 0.9150904605263158, "grad_norm": 21.501615524291992, "learning_rate": 8.490953947368422e-07, "loss": 0.0535, "step": 4451 }, { "epoch": 0.915296052631579, "grad_norm": 10.332111358642578, "learning_rate": 8.470394736842107e-07, "loss": 0.025, "step": 4452 }, { "epoch": 0.9155016447368421, "grad_norm": 58.20761489868164, "learning_rate": 8.44983552631579e-07, "loss": 0.2267, "step": 4453 }, { "epoch": 0.9157072368421053, "grad_norm": 40.035011291503906, "learning_rate": 8.429276315789474e-07, "loss": 0.1206, "step": 4454 }, { "epoch": 0.9159128289473685, "grad_norm": 95.60464477539062, "learning_rate": 8.408717105263159e-07, "loss": 0.8066, "step": 4455 }, { "epoch": 0.9161184210526315, "grad_norm": 18.93918800354004, "learning_rate": 8.388157894736842e-07, "loss": 0.0842, "step": 4456 }, { "epoch": 0.9163240131578947, "grad_norm": 20.194177627563477, "learning_rate": 8.367598684210528e-07, "loss": 0.045, "step": 4457 }, { "epoch": 0.9165296052631579, "grad_norm": 75.73439025878906, "learning_rate": 8.347039473684212e-07, "loss": 0.2402, "step": 4458 }, { "epoch": 0.916735197368421, "grad_norm": 16.22085189819336, "learning_rate": 8.326480263157895e-07, "loss": 0.05, "step": 4459 }, { "epoch": 0.9169407894736842, "grad_norm": 9.741081237792969, "learning_rate": 8.30592105263158e-07, "loss": 0.0165, "step": 4460 }, { "epoch": 0.9171463815789473, "grad_norm": 4.047877311706543, "learning_rate": 8.285361842105263e-07, "loss": 0.0085, "step": 4461 }, { "epoch": 0.9173519736842105, "grad_norm": 47.58340072631836, "learning_rate": 8.264802631578948e-07, "loss": 0.6436, "step": 4462 }, { "epoch": 0.9175575657894737, "grad_norm": 33.29838562011719, "learning_rate": 8.244243421052632e-07, "loss": 0.1489, "step": 4463 }, { "epoch": 0.9177631578947368, "grad_norm": 43.25688552856445, "learning_rate": 8.223684210526317e-07, "loss": 0.0669, "step": 4464 }, { "epoch": 0.91796875, "grad_norm": 25.282060623168945, "learning_rate": 8.203125e-07, "loss": 0.0496, "step": 4465 }, { "epoch": 0.9181743421052632, "grad_norm": 47.5530891418457, "learning_rate": 8.182565789473684e-07, "loss": 0.1252, "step": 4466 }, { "epoch": 0.9183799342105263, "grad_norm": 21.18844223022461, "learning_rate": 8.16200657894737e-07, "loss": 0.0553, "step": 4467 }, { "epoch": 0.9185855263157895, "grad_norm": 125.68174743652344, "learning_rate": 8.141447368421053e-07, "loss": 0.3687, "step": 4468 }, { "epoch": 0.9187911184210527, "grad_norm": 95.51394653320312, "learning_rate": 8.120888157894738e-07, "loss": 0.1954, "step": 4469 }, { "epoch": 0.9189967105263158, "grad_norm": 69.72273254394531, "learning_rate": 8.100328947368422e-07, "loss": 0.4263, "step": 4470 }, { "epoch": 0.919202302631579, "grad_norm": 75.29082489013672, "learning_rate": 8.079769736842107e-07, "loss": 0.251, "step": 4471 }, { "epoch": 0.9194078947368421, "grad_norm": 23.525150299072266, "learning_rate": 8.05921052631579e-07, "loss": 0.0366, "step": 4472 }, { "epoch": 0.9196134868421053, "grad_norm": 11.195196151733398, "learning_rate": 8.038651315789474e-07, "loss": 0.0217, "step": 4473 }, { "epoch": 0.9198190789473685, "grad_norm": 204.7476348876953, "learning_rate": 8.018092105263159e-07, "loss": 1.3718, "step": 4474 }, { "epoch": 0.9200246710526315, "grad_norm": 124.635498046875, "learning_rate": 7.997532894736842e-07, "loss": 0.3721, "step": 4475 }, { "epoch": 0.9202302631578947, "grad_norm": 3.011404514312744, "learning_rate": 7.976973684210528e-07, "loss": 0.0114, "step": 4476 }, { "epoch": 0.9204358552631579, "grad_norm": 102.2296142578125, "learning_rate": 7.956414473684211e-07, "loss": 0.561, "step": 4477 }, { "epoch": 0.920641447368421, "grad_norm": 62.67337417602539, "learning_rate": 7.935855263157895e-07, "loss": 0.2534, "step": 4478 }, { "epoch": 0.9208470394736842, "grad_norm": 15.02822208404541, "learning_rate": 7.91529605263158e-07, "loss": 0.0507, "step": 4479 }, { "epoch": 0.9210526315789473, "grad_norm": 207.06411743164062, "learning_rate": 7.894736842105263e-07, "loss": 0.519, "step": 4480 }, { "epoch": 0.9212582236842105, "grad_norm": 73.54297637939453, "learning_rate": 7.874177631578948e-07, "loss": 0.5928, "step": 4481 }, { "epoch": 0.9214638157894737, "grad_norm": 60.66908645629883, "learning_rate": 7.853618421052632e-07, "loss": 0.2476, "step": 4482 }, { "epoch": 0.9216694078947368, "grad_norm": 11.933110237121582, "learning_rate": 7.833059210526317e-07, "loss": 0.0233, "step": 4483 }, { "epoch": 0.921875, "grad_norm": 242.1448516845703, "learning_rate": 7.8125e-07, "loss": 0.5598, "step": 4484 }, { "epoch": 0.9220805921052632, "grad_norm": 15.101078033447266, "learning_rate": 7.791940789473684e-07, "loss": 0.0309, "step": 4485 }, { "epoch": 0.9222861842105263, "grad_norm": 8.824238777160645, "learning_rate": 7.77138157894737e-07, "loss": 0.0278, "step": 4486 }, { "epoch": 0.9224917763157895, "grad_norm": 2.2568905353546143, "learning_rate": 7.750822368421053e-07, "loss": 0.007, "step": 4487 }, { "epoch": 0.9226973684210527, "grad_norm": 68.97731018066406, "learning_rate": 7.730263157894738e-07, "loss": 0.3239, "step": 4488 }, { "epoch": 0.9229029605263158, "grad_norm": 194.66712951660156, "learning_rate": 7.709703947368422e-07, "loss": 0.4593, "step": 4489 }, { "epoch": 0.923108552631579, "grad_norm": 10.896127700805664, "learning_rate": 7.689144736842106e-07, "loss": 0.0213, "step": 4490 }, { "epoch": 0.9233141447368421, "grad_norm": 44.317317962646484, "learning_rate": 7.66858552631579e-07, "loss": 0.152, "step": 4491 }, { "epoch": 0.9235197368421053, "grad_norm": 27.13084602355957, "learning_rate": 7.648026315789474e-07, "loss": 0.1042, "step": 4492 }, { "epoch": 0.9237253289473685, "grad_norm": 74.9000473022461, "learning_rate": 7.627467105263158e-07, "loss": 0.1516, "step": 4493 }, { "epoch": 0.9239309210526315, "grad_norm": 31.392770767211914, "learning_rate": 7.606907894736842e-07, "loss": 0.062, "step": 4494 }, { "epoch": 0.9241365131578947, "grad_norm": 115.71273040771484, "learning_rate": 7.586348684210528e-07, "loss": 1.0893, "step": 4495 }, { "epoch": 0.9243421052631579, "grad_norm": 54.585147857666016, "learning_rate": 7.565789473684211e-07, "loss": 0.684, "step": 4496 }, { "epoch": 0.924547697368421, "grad_norm": 3.6683831214904785, "learning_rate": 7.545230263157895e-07, "loss": 0.0149, "step": 4497 }, { "epoch": 0.9247532894736842, "grad_norm": 158.70474243164062, "learning_rate": 7.52467105263158e-07, "loss": 0.8197, "step": 4498 }, { "epoch": 0.9249588815789473, "grad_norm": 139.88897705078125, "learning_rate": 7.504111842105263e-07, "loss": 0.3149, "step": 4499 }, { "epoch": 0.9251644736842105, "grad_norm": 22.58222198486328, "learning_rate": 7.483552631578948e-07, "loss": 0.0858, "step": 4500 }, { "epoch": 0.9253700657894737, "grad_norm": 9.90334701538086, "learning_rate": 7.462993421052632e-07, "loss": 0.0231, "step": 4501 }, { "epoch": 0.9255756578947368, "grad_norm": 34.38615417480469, "learning_rate": 7.442434210526317e-07, "loss": 0.1184, "step": 4502 }, { "epoch": 0.92578125, "grad_norm": 91.75637817382812, "learning_rate": 7.421875e-07, "loss": 0.829, "step": 4503 }, { "epoch": 0.9259868421052632, "grad_norm": 197.43699645996094, "learning_rate": 7.401315789473685e-07, "loss": 0.5521, "step": 4504 }, { "epoch": 0.9261924342105263, "grad_norm": 10.652177810668945, "learning_rate": 7.38075657894737e-07, "loss": 0.0277, "step": 4505 }, { "epoch": 0.9263980263157895, "grad_norm": 143.4932861328125, "learning_rate": 7.360197368421053e-07, "loss": 0.4848, "step": 4506 }, { "epoch": 0.9266036184210527, "grad_norm": 117.64117431640625, "learning_rate": 7.339638157894738e-07, "loss": 1.2754, "step": 4507 }, { "epoch": 0.9268092105263158, "grad_norm": 36.464481353759766, "learning_rate": 7.319078947368422e-07, "loss": 0.022, "step": 4508 }, { "epoch": 0.927014802631579, "grad_norm": 45.17190933227539, "learning_rate": 7.298519736842106e-07, "loss": 0.123, "step": 4509 }, { "epoch": 0.9272203947368421, "grad_norm": 122.24324035644531, "learning_rate": 7.27796052631579e-07, "loss": 1.4098, "step": 4510 }, { "epoch": 0.9274259868421053, "grad_norm": 204.5577392578125, "learning_rate": 7.257401315789474e-07, "loss": 0.5287, "step": 4511 }, { "epoch": 0.9276315789473685, "grad_norm": 45.29383087158203, "learning_rate": 7.236842105263158e-07, "loss": 0.1455, "step": 4512 }, { "epoch": 0.9278371710526315, "grad_norm": 103.43951416015625, "learning_rate": 7.216282894736842e-07, "loss": 0.6816, "step": 4513 }, { "epoch": 0.9280427631578947, "grad_norm": 7.848034381866455, "learning_rate": 7.195723684210528e-07, "loss": 0.0287, "step": 4514 }, { "epoch": 0.9282483552631579, "grad_norm": 211.5523223876953, "learning_rate": 7.175164473684211e-07, "loss": 0.7452, "step": 4515 }, { "epoch": 0.928453947368421, "grad_norm": 144.12710571289062, "learning_rate": 7.154605263157895e-07, "loss": 0.8514, "step": 4516 }, { "epoch": 0.9286595394736842, "grad_norm": 7.463077545166016, "learning_rate": 7.13404605263158e-07, "loss": 0.0247, "step": 4517 }, { "epoch": 0.9288651315789473, "grad_norm": 68.5098648071289, "learning_rate": 7.113486842105263e-07, "loss": 0.2953, "step": 4518 }, { "epoch": 0.9290707236842105, "grad_norm": 41.59684753417969, "learning_rate": 7.092927631578948e-07, "loss": 0.1108, "step": 4519 }, { "epoch": 0.9292763157894737, "grad_norm": 27.76818084716797, "learning_rate": 7.072368421052632e-07, "loss": 0.0903, "step": 4520 }, { "epoch": 0.9294819078947368, "grad_norm": 44.358131408691406, "learning_rate": 7.051809210526317e-07, "loss": 0.1807, "step": 4521 }, { "epoch": 0.9296875, "grad_norm": 22.79079246520996, "learning_rate": 7.03125e-07, "loss": 0.0809, "step": 4522 }, { "epoch": 0.9298930921052632, "grad_norm": 64.68053436279297, "learning_rate": 7.010690789473685e-07, "loss": 0.1677, "step": 4523 }, { "epoch": 0.9300986842105263, "grad_norm": 12.791939735412598, "learning_rate": 6.99013157894737e-07, "loss": 0.062, "step": 4524 }, { "epoch": 0.9303042763157895, "grad_norm": 3.731938123703003, "learning_rate": 6.969572368421053e-07, "loss": 0.0136, "step": 4525 }, { "epoch": 0.9305098684210527, "grad_norm": 6.715244770050049, "learning_rate": 6.949013157894738e-07, "loss": 0.0139, "step": 4526 }, { "epoch": 0.9307154605263158, "grad_norm": 8.270549774169922, "learning_rate": 6.928453947368421e-07, "loss": 0.0244, "step": 4527 }, { "epoch": 0.930921052631579, "grad_norm": 3.806056022644043, "learning_rate": 6.907894736842106e-07, "loss": 0.0166, "step": 4528 }, { "epoch": 0.9311266447368421, "grad_norm": 34.32426834106445, "learning_rate": 6.88733552631579e-07, "loss": 0.1084, "step": 4529 }, { "epoch": 0.9313322368421053, "grad_norm": 22.3934268951416, "learning_rate": 6.866776315789473e-07, "loss": 0.0696, "step": 4530 }, { "epoch": 0.9315378289473685, "grad_norm": 11.176623344421387, "learning_rate": 6.846217105263158e-07, "loss": 0.0345, "step": 4531 }, { "epoch": 0.9317434210526315, "grad_norm": 42.97905731201172, "learning_rate": 6.825657894736843e-07, "loss": 0.1295, "step": 4532 }, { "epoch": 0.9319490131578947, "grad_norm": 60.96287536621094, "learning_rate": 6.805098684210528e-07, "loss": 0.2842, "step": 4533 }, { "epoch": 0.9321546052631579, "grad_norm": 9.625274658203125, "learning_rate": 6.784539473684211e-07, "loss": 0.0139, "step": 4534 }, { "epoch": 0.932360197368421, "grad_norm": 62.00644302368164, "learning_rate": 6.763980263157895e-07, "loss": 0.2411, "step": 4535 }, { "epoch": 0.9325657894736842, "grad_norm": 165.27720642089844, "learning_rate": 6.74342105263158e-07, "loss": 1.1607, "step": 4536 }, { "epoch": 0.9327713815789473, "grad_norm": 42.505592346191406, "learning_rate": 6.722861842105263e-07, "loss": 0.0766, "step": 4537 }, { "epoch": 0.9329769736842105, "grad_norm": 18.154895782470703, "learning_rate": 6.702302631578948e-07, "loss": 0.0279, "step": 4538 }, { "epoch": 0.9331825657894737, "grad_norm": 63.94998550415039, "learning_rate": 6.681743421052632e-07, "loss": 0.4183, "step": 4539 }, { "epoch": 0.9333881578947368, "grad_norm": 19.329326629638672, "learning_rate": 6.661184210526317e-07, "loss": 0.0549, "step": 4540 }, { "epoch": 0.93359375, "grad_norm": 147.4353485107422, "learning_rate": 6.640625e-07, "loss": 0.3555, "step": 4541 }, { "epoch": 0.9337993421052632, "grad_norm": 8.80218505859375, "learning_rate": 6.620065789473685e-07, "loss": 0.0158, "step": 4542 }, { "epoch": 0.9340049342105263, "grad_norm": 64.10334014892578, "learning_rate": 6.599506578947369e-07, "loss": 0.356, "step": 4543 }, { "epoch": 0.9342105263157895, "grad_norm": 119.26411437988281, "learning_rate": 6.578947368421053e-07, "loss": 0.3674, "step": 4544 }, { "epoch": 0.9344161184210527, "grad_norm": 27.626998901367188, "learning_rate": 6.558388157894738e-07, "loss": 0.0529, "step": 4545 }, { "epoch": 0.9346217105263158, "grad_norm": 12.35934066772461, "learning_rate": 6.537828947368421e-07, "loss": 0.0281, "step": 4546 }, { "epoch": 0.934827302631579, "grad_norm": 11.069808959960938, "learning_rate": 6.517269736842106e-07, "loss": 0.0256, "step": 4547 }, { "epoch": 0.9350328947368421, "grad_norm": 50.782493591308594, "learning_rate": 6.49671052631579e-07, "loss": 0.2202, "step": 4548 }, { "epoch": 0.9352384868421053, "grad_norm": 27.55223846435547, "learning_rate": 6.476151315789473e-07, "loss": 0.0499, "step": 4549 }, { "epoch": 0.9354440789473685, "grad_norm": 13.641596794128418, "learning_rate": 6.455592105263158e-07, "loss": 0.0468, "step": 4550 }, { "epoch": 0.9356496710526315, "grad_norm": 52.53327178955078, "learning_rate": 6.435032894736843e-07, "loss": 0.0994, "step": 4551 }, { "epoch": 0.9358552631578947, "grad_norm": 61.97407531738281, "learning_rate": 6.414473684210528e-07, "loss": 0.1245, "step": 4552 }, { "epoch": 0.9360608552631579, "grad_norm": 12.17210578918457, "learning_rate": 6.393914473684211e-07, "loss": 0.0422, "step": 4553 }, { "epoch": 0.936266447368421, "grad_norm": 5.372280597686768, "learning_rate": 6.373355263157895e-07, "loss": 0.0191, "step": 4554 }, { "epoch": 0.9364720394736842, "grad_norm": 19.22481346130371, "learning_rate": 6.35279605263158e-07, "loss": 0.1057, "step": 4555 }, { "epoch": 0.9366776315789473, "grad_norm": 106.4769058227539, "learning_rate": 6.332236842105263e-07, "loss": 0.4238, "step": 4556 }, { "epoch": 0.9368832236842105, "grad_norm": 243.33168029785156, "learning_rate": 6.311677631578948e-07, "loss": 0.6973, "step": 4557 }, { "epoch": 0.9370888157894737, "grad_norm": 68.5351333618164, "learning_rate": 6.291118421052632e-07, "loss": 0.7007, "step": 4558 }, { "epoch": 0.9372944078947368, "grad_norm": 11.30962085723877, "learning_rate": 6.270559210526316e-07, "loss": 0.0154, "step": 4559 }, { "epoch": 0.9375, "grad_norm": 124.47985076904297, "learning_rate": 6.25e-07, "loss": 0.4525, "step": 4560 }, { "epoch": 0.9377055921052632, "grad_norm": 39.520851135253906, "learning_rate": 6.229440789473685e-07, "loss": 0.2546, "step": 4561 }, { "epoch": 0.9379111842105263, "grad_norm": 439.4424743652344, "learning_rate": 6.208881578947369e-07, "loss": 3.4263, "step": 4562 }, { "epoch": 0.9381167763157895, "grad_norm": 30.78619384765625, "learning_rate": 6.188322368421054e-07, "loss": 0.0883, "step": 4563 }, { "epoch": 0.9383223684210527, "grad_norm": 6.322934627532959, "learning_rate": 6.167763157894737e-07, "loss": 0.0091, "step": 4564 }, { "epoch": 0.9385279605263158, "grad_norm": 29.554258346557617, "learning_rate": 6.147203947368421e-07, "loss": 0.0809, "step": 4565 }, { "epoch": 0.938733552631579, "grad_norm": 76.10272979736328, "learning_rate": 6.126644736842106e-07, "loss": 0.916, "step": 4566 }, { "epoch": 0.9389391447368421, "grad_norm": 31.563325881958008, "learning_rate": 6.10608552631579e-07, "loss": 0.0736, "step": 4567 }, { "epoch": 0.9391447368421053, "grad_norm": 90.90599060058594, "learning_rate": 6.085526315789474e-07, "loss": 0.3257, "step": 4568 }, { "epoch": 0.9393503289473685, "grad_norm": 14.452227592468262, "learning_rate": 6.064967105263158e-07, "loss": 0.031, "step": 4569 }, { "epoch": 0.9395559210526315, "grad_norm": 7.625570297241211, "learning_rate": 6.044407894736843e-07, "loss": 0.0257, "step": 4570 }, { "epoch": 0.9397615131578947, "grad_norm": 9.00818920135498, "learning_rate": 6.023848684210527e-07, "loss": 0.0143, "step": 4571 }, { "epoch": 0.9399671052631579, "grad_norm": 8.5675048828125, "learning_rate": 6.003289473684211e-07, "loss": 0.0168, "step": 4572 }, { "epoch": 0.940172697368421, "grad_norm": 37.979759216308594, "learning_rate": 5.982730263157895e-07, "loss": 0.0655, "step": 4573 }, { "epoch": 0.9403782894736842, "grad_norm": 10.401434898376465, "learning_rate": 5.96217105263158e-07, "loss": 0.0124, "step": 4574 }, { "epoch": 0.9405838815789473, "grad_norm": 5.718742847442627, "learning_rate": 5.941611842105264e-07, "loss": 0.0119, "step": 4575 }, { "epoch": 0.9407894736842105, "grad_norm": 11.664346694946289, "learning_rate": 5.921052631578947e-07, "loss": 0.0314, "step": 4576 }, { "epoch": 0.9409950657894737, "grad_norm": 139.36434936523438, "learning_rate": 5.900493421052632e-07, "loss": 1.1565, "step": 4577 }, { "epoch": 0.9412006578947368, "grad_norm": 16.566926956176758, "learning_rate": 5.879934210526316e-07, "loss": 0.0314, "step": 4578 }, { "epoch": 0.94140625, "grad_norm": 68.66162109375, "learning_rate": 5.859375000000001e-07, "loss": 0.2402, "step": 4579 }, { "epoch": 0.9416118421052632, "grad_norm": 434.1451721191406, "learning_rate": 5.838815789473685e-07, "loss": 0.8839, "step": 4580 }, { "epoch": 0.9418174342105263, "grad_norm": 1.5334924459457397, "learning_rate": 5.818256578947369e-07, "loss": 0.0049, "step": 4581 }, { "epoch": 0.9420230263157895, "grad_norm": 4.364607334136963, "learning_rate": 5.797697368421054e-07, "loss": 0.0094, "step": 4582 }, { "epoch": 0.9422286184210527, "grad_norm": 23.318342208862305, "learning_rate": 5.777138157894737e-07, "loss": 0.0698, "step": 4583 }, { "epoch": 0.9424342105263158, "grad_norm": 30.156753540039062, "learning_rate": 5.756578947368421e-07, "loss": 0.0958, "step": 4584 }, { "epoch": 0.942639802631579, "grad_norm": 47.7461051940918, "learning_rate": 5.736019736842106e-07, "loss": 0.126, "step": 4585 }, { "epoch": 0.9428453947368421, "grad_norm": 48.5218620300293, "learning_rate": 5.71546052631579e-07, "loss": 0.1158, "step": 4586 }, { "epoch": 0.9430509868421053, "grad_norm": 102.95700073242188, "learning_rate": 5.694901315789474e-07, "loss": 0.286, "step": 4587 }, { "epoch": 0.9432565789473685, "grad_norm": 7.496748924255371, "learning_rate": 5.674342105263159e-07, "loss": 0.0206, "step": 4588 }, { "epoch": 0.9434621710526315, "grad_norm": 52.27516174316406, "learning_rate": 5.653782894736843e-07, "loss": 0.1022, "step": 4589 }, { "epoch": 0.9436677631578947, "grad_norm": 11.43175220489502, "learning_rate": 5.633223684210527e-07, "loss": 0.0166, "step": 4590 }, { "epoch": 0.9438733552631579, "grad_norm": 100.02024841308594, "learning_rate": 5.612664473684211e-07, "loss": 0.364, "step": 4591 }, { "epoch": 0.944078947368421, "grad_norm": 48.79420471191406, "learning_rate": 5.592105263157895e-07, "loss": 0.5208, "step": 4592 }, { "epoch": 0.9442845394736842, "grad_norm": 118.60394287109375, "learning_rate": 5.57154605263158e-07, "loss": 0.2568, "step": 4593 }, { "epoch": 0.9444901315789473, "grad_norm": 453.6911926269531, "learning_rate": 5.550986842105264e-07, "loss": 0.9282, "step": 4594 }, { "epoch": 0.9446957236842105, "grad_norm": 10.403020858764648, "learning_rate": 5.530427631578947e-07, "loss": 0.0515, "step": 4595 }, { "epoch": 0.9449013157894737, "grad_norm": 71.22209930419922, "learning_rate": 5.509868421052631e-07, "loss": 0.5492, "step": 4596 }, { "epoch": 0.9451069078947368, "grad_norm": 114.26655578613281, "learning_rate": 5.489309210526316e-07, "loss": 0.4626, "step": 4597 }, { "epoch": 0.9453125, "grad_norm": 92.29237365722656, "learning_rate": 5.468750000000001e-07, "loss": 0.5605, "step": 4598 }, { "epoch": 0.9455180921052632, "grad_norm": 91.638671875, "learning_rate": 5.448190789473685e-07, "loss": 0.0474, "step": 4599 }, { "epoch": 0.9457236842105263, "grad_norm": 62.312984466552734, "learning_rate": 5.427631578947369e-07, "loss": 0.9697, "step": 4600 }, { "epoch": 0.9459292763157895, "grad_norm": 43.65196990966797, "learning_rate": 5.407072368421053e-07, "loss": 0.1665, "step": 4601 }, { "epoch": 0.9461348684210527, "grad_norm": 123.58331298828125, "learning_rate": 5.386513157894737e-07, "loss": 0.8409, "step": 4602 }, { "epoch": 0.9463404605263158, "grad_norm": 14.401314735412598, "learning_rate": 5.365953947368421e-07, "loss": 0.0847, "step": 4603 }, { "epoch": 0.946546052631579, "grad_norm": 10.216249465942383, "learning_rate": 5.345394736842105e-07, "loss": 0.0226, "step": 4604 }, { "epoch": 0.9467516447368421, "grad_norm": 114.96063995361328, "learning_rate": 5.32483552631579e-07, "loss": 0.5118, "step": 4605 }, { "epoch": 0.9469572368421053, "grad_norm": 83.34912872314453, "learning_rate": 5.304276315789474e-07, "loss": 0.4302, "step": 4606 }, { "epoch": 0.9471628289473685, "grad_norm": 25.777864456176758, "learning_rate": 5.283717105263159e-07, "loss": 0.0723, "step": 4607 }, { "epoch": 0.9473684210526315, "grad_norm": 42.6572265625, "learning_rate": 5.263157894736843e-07, "loss": 0.0872, "step": 4608 }, { "epoch": 0.9475740131578947, "grad_norm": 173.82716369628906, "learning_rate": 5.242598684210527e-07, "loss": 0.8735, "step": 4609 }, { "epoch": 0.9477796052631579, "grad_norm": 26.666812896728516, "learning_rate": 5.222039473684211e-07, "loss": 0.0712, "step": 4610 }, { "epoch": 0.947985197368421, "grad_norm": 129.20741271972656, "learning_rate": 5.201480263157895e-07, "loss": 0.5869, "step": 4611 }, { "epoch": 0.9481907894736842, "grad_norm": 60.337318420410156, "learning_rate": 5.180921052631579e-07, "loss": 0.1371, "step": 4612 }, { "epoch": 0.9483963815789473, "grad_norm": 209.69854736328125, "learning_rate": 5.160361842105264e-07, "loss": 0.7295, "step": 4613 }, { "epoch": 0.9486019736842105, "grad_norm": 10.338408470153809, "learning_rate": 5.139802631578947e-07, "loss": 0.0156, "step": 4614 }, { "epoch": 0.9488075657894737, "grad_norm": 82.12993621826172, "learning_rate": 5.119243421052631e-07, "loss": 0.1699, "step": 4615 }, { "epoch": 0.9490131578947368, "grad_norm": 157.33653259277344, "learning_rate": 5.098684210526317e-07, "loss": 1.2925, "step": 4616 }, { "epoch": 0.94921875, "grad_norm": 86.37425231933594, "learning_rate": 5.078125000000001e-07, "loss": 0.1477, "step": 4617 }, { "epoch": 0.9494243421052632, "grad_norm": 19.84884262084961, "learning_rate": 5.057565789473685e-07, "loss": 0.0299, "step": 4618 }, { "epoch": 0.9496299342105263, "grad_norm": 8.57498550415039, "learning_rate": 5.037006578947369e-07, "loss": 0.0123, "step": 4619 }, { "epoch": 0.9498355263157895, "grad_norm": 83.99246978759766, "learning_rate": 5.016447368421053e-07, "loss": 0.4598, "step": 4620 }, { "epoch": 0.9500411184210527, "grad_norm": 171.5601806640625, "learning_rate": 4.995888157894737e-07, "loss": 0.4502, "step": 4621 }, { "epoch": 0.9502467105263158, "grad_norm": 26.62830924987793, "learning_rate": 4.975328947368421e-07, "loss": 0.0559, "step": 4622 }, { "epoch": 0.950452302631579, "grad_norm": 171.25729370117188, "learning_rate": 4.954769736842105e-07, "loss": 0.8157, "step": 4623 }, { "epoch": 0.9506578947368421, "grad_norm": 104.57876586914062, "learning_rate": 4.93421052631579e-07, "loss": 1.6875, "step": 4624 }, { "epoch": 0.9508634868421053, "grad_norm": 88.26397705078125, "learning_rate": 4.913651315789474e-07, "loss": 0.1996, "step": 4625 }, { "epoch": 0.9510690789473685, "grad_norm": 137.33229064941406, "learning_rate": 4.893092105263159e-07, "loss": 0.5598, "step": 4626 }, { "epoch": 0.9512746710526315, "grad_norm": 32.97045135498047, "learning_rate": 4.872532894736843e-07, "loss": 0.0896, "step": 4627 }, { "epoch": 0.9514802631578947, "grad_norm": 16.40233612060547, "learning_rate": 4.851973684210527e-07, "loss": 0.0457, "step": 4628 }, { "epoch": 0.9516858552631579, "grad_norm": 49.76991271972656, "learning_rate": 4.831414473684211e-07, "loss": 0.1655, "step": 4629 }, { "epoch": 0.951891447368421, "grad_norm": 20.12278175354004, "learning_rate": 4.810855263157895e-07, "loss": 0.0535, "step": 4630 }, { "epoch": 0.9520970394736842, "grad_norm": 21.503782272338867, "learning_rate": 4.790296052631579e-07, "loss": 0.0853, "step": 4631 }, { "epoch": 0.9523026315789473, "grad_norm": 34.61383056640625, "learning_rate": 4.769736842105264e-07, "loss": 0.089, "step": 4632 }, { "epoch": 0.9525082236842105, "grad_norm": 209.30076599121094, "learning_rate": 4.7491776315789473e-07, "loss": 1.0273, "step": 4633 }, { "epoch": 0.9527138157894737, "grad_norm": 33.34831237792969, "learning_rate": 4.728618421052632e-07, "loss": 0.1206, "step": 4634 }, { "epoch": 0.9529194078947368, "grad_norm": 96.42822265625, "learning_rate": 4.708059210526316e-07, "loss": 0.438, "step": 4635 }, { "epoch": 0.953125, "grad_norm": 71.66071319580078, "learning_rate": 4.6875000000000006e-07, "loss": 0.4932, "step": 4636 }, { "epoch": 0.9533305921052632, "grad_norm": 63.939674377441406, "learning_rate": 4.666940789473685e-07, "loss": 0.2651, "step": 4637 }, { "epoch": 0.9535361842105263, "grad_norm": 1.7261792421340942, "learning_rate": 4.646381578947369e-07, "loss": 0.0043, "step": 4638 }, { "epoch": 0.9537417763157895, "grad_norm": 4.389571666717529, "learning_rate": 4.6258223684210535e-07, "loss": 0.0078, "step": 4639 }, { "epoch": 0.9539473684210527, "grad_norm": 36.70496368408203, "learning_rate": 4.605263157894737e-07, "loss": 0.1491, "step": 4640 }, { "epoch": 0.9541529605263158, "grad_norm": 136.96302795410156, "learning_rate": 4.584703947368421e-07, "loss": 0.6978, "step": 4641 }, { "epoch": 0.954358552631579, "grad_norm": 5.369485378265381, "learning_rate": 4.564144736842106e-07, "loss": 0.0203, "step": 4642 }, { "epoch": 0.9545641447368421, "grad_norm": 57.853511810302734, "learning_rate": 4.54358552631579e-07, "loss": 0.2025, "step": 4643 }, { "epoch": 0.9547697368421053, "grad_norm": 21.28458023071289, "learning_rate": 4.523026315789474e-07, "loss": 0.0664, "step": 4644 }, { "epoch": 0.9549753289473685, "grad_norm": 48.48908615112305, "learning_rate": 4.5024671052631586e-07, "loss": 0.155, "step": 4645 }, { "epoch": 0.9551809210526315, "grad_norm": 208.00006103515625, "learning_rate": 4.481907894736842e-07, "loss": 0.4753, "step": 4646 }, { "epoch": 0.9553865131578947, "grad_norm": 331.43701171875, "learning_rate": 4.4613486842105263e-07, "loss": 0.908, "step": 4647 }, { "epoch": 0.9555921052631579, "grad_norm": 4.851216793060303, "learning_rate": 4.440789473684211e-07, "loss": 0.0089, "step": 4648 }, { "epoch": 0.955797697368421, "grad_norm": 18.92717170715332, "learning_rate": 4.420230263157895e-07, "loss": 0.0668, "step": 4649 }, { "epoch": 0.9560032894736842, "grad_norm": 150.63973999023438, "learning_rate": 4.3996710526315796e-07, "loss": 0.6934, "step": 4650 }, { "epoch": 0.9562088815789473, "grad_norm": 14.31980037689209, "learning_rate": 4.379111842105264e-07, "loss": 0.0195, "step": 4651 }, { "epoch": 0.9564144736842105, "grad_norm": 14.621371269226074, "learning_rate": 4.3585526315789473e-07, "loss": 0.0464, "step": 4652 }, { "epoch": 0.9566200657894737, "grad_norm": 82.20146179199219, "learning_rate": 4.337993421052632e-07, "loss": 0.5918, "step": 4653 }, { "epoch": 0.9568256578947368, "grad_norm": 79.6134033203125, "learning_rate": 4.317434210526316e-07, "loss": 0.2607, "step": 4654 }, { "epoch": 0.95703125, "grad_norm": 55.5378532409668, "learning_rate": 4.296875e-07, "loss": 0.209, "step": 4655 }, { "epoch": 0.9572368421052632, "grad_norm": 34.24761199951172, "learning_rate": 4.276315789473685e-07, "loss": 0.1172, "step": 4656 }, { "epoch": 0.9574424342105263, "grad_norm": 36.89188003540039, "learning_rate": 4.255756578947369e-07, "loss": 0.0887, "step": 4657 }, { "epoch": 0.9576480263157895, "grad_norm": 86.79782104492188, "learning_rate": 4.2351973684210535e-07, "loss": 0.3018, "step": 4658 }, { "epoch": 0.9578536184210527, "grad_norm": 10.522330284118652, "learning_rate": 4.214638157894737e-07, "loss": 0.0364, "step": 4659 }, { "epoch": 0.9580592105263158, "grad_norm": 52.457523345947266, "learning_rate": 4.194078947368421e-07, "loss": 0.0713, "step": 4660 }, { "epoch": 0.958264802631579, "grad_norm": 39.04718780517578, "learning_rate": 4.173519736842106e-07, "loss": 0.0402, "step": 4661 }, { "epoch": 0.9584703947368421, "grad_norm": 61.647682189941406, "learning_rate": 4.15296052631579e-07, "loss": 0.2281, "step": 4662 }, { "epoch": 0.9586759868421053, "grad_norm": 23.498077392578125, "learning_rate": 4.132401315789474e-07, "loss": 0.1029, "step": 4663 }, { "epoch": 0.9588815789473685, "grad_norm": 65.85363006591797, "learning_rate": 4.1118421052631586e-07, "loss": 0.2649, "step": 4664 }, { "epoch": 0.9590871710526315, "grad_norm": 5.784830570220947, "learning_rate": 4.091282894736842e-07, "loss": 0.0187, "step": 4665 }, { "epoch": 0.9592927631578947, "grad_norm": 129.40028381347656, "learning_rate": 4.0707236842105263e-07, "loss": 0.7681, "step": 4666 }, { "epoch": 0.9594983552631579, "grad_norm": 163.146484375, "learning_rate": 4.050164473684211e-07, "loss": 1.3848, "step": 4667 }, { "epoch": 0.959703947368421, "grad_norm": 10.879562377929688, "learning_rate": 4.029605263157895e-07, "loss": 0.0229, "step": 4668 }, { "epoch": 0.9599095394736842, "grad_norm": 6.843327522277832, "learning_rate": 4.0090460526315797e-07, "loss": 0.0192, "step": 4669 }, { "epoch": 0.9601151315789473, "grad_norm": 12.177515983581543, "learning_rate": 3.988486842105264e-07, "loss": 0.0137, "step": 4670 }, { "epoch": 0.9603207236842105, "grad_norm": 11.068267822265625, "learning_rate": 3.9679276315789474e-07, "loss": 0.0583, "step": 4671 }, { "epoch": 0.9605263157894737, "grad_norm": 141.68719482421875, "learning_rate": 3.9473684210526315e-07, "loss": 0.6411, "step": 4672 }, { "epoch": 0.9607319078947368, "grad_norm": 35.59389877319336, "learning_rate": 3.926809210526316e-07, "loss": 0.1596, "step": 4673 }, { "epoch": 0.9609375, "grad_norm": 37.93256759643555, "learning_rate": 3.90625e-07, "loss": 0.0407, "step": 4674 }, { "epoch": 0.9611430921052632, "grad_norm": 7.762557506561279, "learning_rate": 3.885690789473685e-07, "loss": 0.0314, "step": 4675 }, { "epoch": 0.9613486842105263, "grad_norm": 134.6548614501953, "learning_rate": 3.865131578947369e-07, "loss": 1.7422, "step": 4676 }, { "epoch": 0.9615542763157895, "grad_norm": 6.132359504699707, "learning_rate": 3.844572368421053e-07, "loss": 0.019, "step": 4677 }, { "epoch": 0.9617598684210527, "grad_norm": 9.010673522949219, "learning_rate": 3.824013157894737e-07, "loss": 0.0294, "step": 4678 }, { "epoch": 0.9619654605263158, "grad_norm": 24.748001098632812, "learning_rate": 3.803453947368421e-07, "loss": 0.1123, "step": 4679 }, { "epoch": 0.962171052631579, "grad_norm": 79.16499328613281, "learning_rate": 3.7828947368421053e-07, "loss": 0.9417, "step": 4680 }, { "epoch": 0.9623766447368421, "grad_norm": 20.076400756835938, "learning_rate": 3.76233552631579e-07, "loss": 0.0887, "step": 4681 }, { "epoch": 0.9625822368421053, "grad_norm": 43.26259231567383, "learning_rate": 3.741776315789474e-07, "loss": 0.086, "step": 4682 }, { "epoch": 0.9627878289473685, "grad_norm": 18.738826751708984, "learning_rate": 3.7212171052631587e-07, "loss": 0.0509, "step": 4683 }, { "epoch": 0.9629934210526315, "grad_norm": 31.465682983398438, "learning_rate": 3.700657894736842e-07, "loss": 0.1179, "step": 4684 }, { "epoch": 0.9631990131578947, "grad_norm": 12.779158592224121, "learning_rate": 3.6800986842105264e-07, "loss": 0.038, "step": 4685 }, { "epoch": 0.9634046052631579, "grad_norm": 67.32354736328125, "learning_rate": 3.659539473684211e-07, "loss": 0.1411, "step": 4686 }, { "epoch": 0.963610197368421, "grad_norm": 121.72602844238281, "learning_rate": 3.638980263157895e-07, "loss": 0.5635, "step": 4687 }, { "epoch": 0.9638157894736842, "grad_norm": 4.402018070220947, "learning_rate": 3.618421052631579e-07, "loss": 0.0105, "step": 4688 }, { "epoch": 0.9640213815789473, "grad_norm": 25.89415168762207, "learning_rate": 3.597861842105264e-07, "loss": 0.0629, "step": 4689 }, { "epoch": 0.9642269736842105, "grad_norm": 95.36943817138672, "learning_rate": 3.5773026315789474e-07, "loss": 0.2697, "step": 4690 }, { "epoch": 0.9644325657894737, "grad_norm": 55.32246017456055, "learning_rate": 3.5567434210526315e-07, "loss": 0.1478, "step": 4691 }, { "epoch": 0.9646381578947368, "grad_norm": 0.9094275832176208, "learning_rate": 3.536184210526316e-07, "loss": 0.0041, "step": 4692 }, { "epoch": 0.96484375, "grad_norm": 27.245052337646484, "learning_rate": 3.515625e-07, "loss": 0.0629, "step": 4693 }, { "epoch": 0.9650493421052632, "grad_norm": 222.42112731933594, "learning_rate": 3.495065789473685e-07, "loss": 0.597, "step": 4694 }, { "epoch": 0.9652549342105263, "grad_norm": 29.720718383789062, "learning_rate": 3.474506578947369e-07, "loss": 0.0715, "step": 4695 }, { "epoch": 0.9654605263157895, "grad_norm": 23.037633895874023, "learning_rate": 3.453947368421053e-07, "loss": 0.1011, "step": 4696 }, { "epoch": 0.9656661184210527, "grad_norm": 49.283443450927734, "learning_rate": 3.4333881578947366e-07, "loss": 0.6932, "step": 4697 }, { "epoch": 0.9658717105263158, "grad_norm": 5.258636951446533, "learning_rate": 3.412828947368421e-07, "loss": 0.0128, "step": 4698 }, { "epoch": 0.966077302631579, "grad_norm": 21.39716339111328, "learning_rate": 3.3922697368421054e-07, "loss": 0.0388, "step": 4699 }, { "epoch": 0.9662828947368421, "grad_norm": 26.50792121887207, "learning_rate": 3.37171052631579e-07, "loss": 0.1112, "step": 4700 }, { "epoch": 0.9664884868421053, "grad_norm": 13.205742835998535, "learning_rate": 3.351151315789474e-07, "loss": 0.0356, "step": 4701 }, { "epoch": 0.9666940789473685, "grad_norm": 76.14449310302734, "learning_rate": 3.3305921052631587e-07, "loss": 0.2935, "step": 4702 }, { "epoch": 0.9668996710526315, "grad_norm": 5.908542156219482, "learning_rate": 3.3100328947368423e-07, "loss": 0.0084, "step": 4703 }, { "epoch": 0.9671052631578947, "grad_norm": 28.620697021484375, "learning_rate": 3.2894736842105264e-07, "loss": 0.095, "step": 4704 }, { "epoch": 0.9673108552631579, "grad_norm": 10.756001472473145, "learning_rate": 3.2689144736842105e-07, "loss": 0.0201, "step": 4705 }, { "epoch": 0.967516447368421, "grad_norm": 210.53028869628906, "learning_rate": 3.248355263157895e-07, "loss": 1.7656, "step": 4706 }, { "epoch": 0.9677220394736842, "grad_norm": 6.024347305297852, "learning_rate": 3.227796052631579e-07, "loss": 0.0188, "step": 4707 }, { "epoch": 0.9679276315789473, "grad_norm": 281.3194885253906, "learning_rate": 3.207236842105264e-07, "loss": 2.0586, "step": 4708 }, { "epoch": 0.9681332236842105, "grad_norm": 7.977750778198242, "learning_rate": 3.1866776315789474e-07, "loss": 0.0149, "step": 4709 }, { "epoch": 0.9683388157894737, "grad_norm": 9.099468231201172, "learning_rate": 3.1661184210526315e-07, "loss": 0.0216, "step": 4710 }, { "epoch": 0.9685444078947368, "grad_norm": 67.63301849365234, "learning_rate": 3.145559210526316e-07, "loss": 1.0095, "step": 4711 }, { "epoch": 0.96875, "grad_norm": 104.10371398925781, "learning_rate": 3.125e-07, "loss": 0.6038, "step": 4712 }, { "epoch": 0.9689555921052632, "grad_norm": 241.93003845214844, "learning_rate": 3.1044407894736844e-07, "loss": 1.332, "step": 4713 }, { "epoch": 0.9691611842105263, "grad_norm": 114.56693267822266, "learning_rate": 3.0838815789473685e-07, "loss": 0.2117, "step": 4714 }, { "epoch": 0.9693667763157895, "grad_norm": 290.83526611328125, "learning_rate": 3.063322368421053e-07, "loss": 1.124, "step": 4715 }, { "epoch": 0.9695723684210527, "grad_norm": 31.565879821777344, "learning_rate": 3.042763157894737e-07, "loss": 0.1234, "step": 4716 }, { "epoch": 0.9697779605263158, "grad_norm": 225.61761474609375, "learning_rate": 3.0222039473684213e-07, "loss": 0.9907, "step": 4717 }, { "epoch": 0.969983552631579, "grad_norm": 82.72940826416016, "learning_rate": 3.0016447368421054e-07, "loss": 1.1713, "step": 4718 }, { "epoch": 0.9701891447368421, "grad_norm": 15.999428749084473, "learning_rate": 2.98108552631579e-07, "loss": 0.0492, "step": 4719 }, { "epoch": 0.9703947368421053, "grad_norm": 83.82850646972656, "learning_rate": 2.9605263157894736e-07, "loss": 0.3129, "step": 4720 }, { "epoch": 0.9706003289473685, "grad_norm": 32.4890251159668, "learning_rate": 2.939967105263158e-07, "loss": 0.0657, "step": 4721 }, { "epoch": 0.9708059210526315, "grad_norm": 19.138690948486328, "learning_rate": 2.9194078947368423e-07, "loss": 0.0437, "step": 4722 }, { "epoch": 0.9710115131578947, "grad_norm": 92.28924560546875, "learning_rate": 2.898848684210527e-07, "loss": 0.3959, "step": 4723 }, { "epoch": 0.9712171052631579, "grad_norm": 30.48360824584961, "learning_rate": 2.8782894736842105e-07, "loss": 0.0989, "step": 4724 }, { "epoch": 0.971422697368421, "grad_norm": 68.63191223144531, "learning_rate": 2.857730263157895e-07, "loss": 0.7229, "step": 4725 }, { "epoch": 0.9716282894736842, "grad_norm": 33.31428146362305, "learning_rate": 2.837171052631579e-07, "loss": 0.1262, "step": 4726 }, { "epoch": 0.9718338815789473, "grad_norm": 41.75996017456055, "learning_rate": 2.8166118421052634e-07, "loss": 0.1128, "step": 4727 }, { "epoch": 0.9720394736842105, "grad_norm": 92.09583282470703, "learning_rate": 2.7960526315789475e-07, "loss": 0.1307, "step": 4728 }, { "epoch": 0.9722450657894737, "grad_norm": 90.8582763671875, "learning_rate": 2.775493421052632e-07, "loss": 0.7681, "step": 4729 }, { "epoch": 0.9724506578947368, "grad_norm": 43.138938903808594, "learning_rate": 2.7549342105263157e-07, "loss": 0.0476, "step": 4730 }, { "epoch": 0.97265625, "grad_norm": 15.104092597961426, "learning_rate": 2.7343750000000003e-07, "loss": 0.0417, "step": 4731 }, { "epoch": 0.9728618421052632, "grad_norm": 110.40101623535156, "learning_rate": 2.7138157894736844e-07, "loss": 0.2842, "step": 4732 }, { "epoch": 0.9730674342105263, "grad_norm": 16.85199737548828, "learning_rate": 2.6932565789473685e-07, "loss": 0.0388, "step": 4733 }, { "epoch": 0.9732730263157895, "grad_norm": 3.235013961791992, "learning_rate": 2.6726973684210526e-07, "loss": 0.0138, "step": 4734 }, { "epoch": 0.9734786184210527, "grad_norm": 9.201597213745117, "learning_rate": 2.652138157894737e-07, "loss": 0.0334, "step": 4735 }, { "epoch": 0.9736842105263158, "grad_norm": 180.1560821533203, "learning_rate": 2.6315789473684213e-07, "loss": 1.6523, "step": 4736 }, { "epoch": 0.973889802631579, "grad_norm": 41.92189025878906, "learning_rate": 2.6110197368421054e-07, "loss": 0.1313, "step": 4737 }, { "epoch": 0.9740953947368421, "grad_norm": 7.917209148406982, "learning_rate": 2.5904605263157895e-07, "loss": 0.0187, "step": 4738 }, { "epoch": 0.9743009868421053, "grad_norm": 82.68880462646484, "learning_rate": 2.5699013157894736e-07, "loss": 0.1431, "step": 4739 }, { "epoch": 0.9745065789473685, "grad_norm": 22.69619369506836, "learning_rate": 2.549342105263158e-07, "loss": 0.0342, "step": 4740 }, { "epoch": 0.9747121710526315, "grad_norm": 16.08176040649414, "learning_rate": 2.5287828947368424e-07, "loss": 0.0609, "step": 4741 }, { "epoch": 0.9749177631578947, "grad_norm": 622.791748046875, "learning_rate": 2.5082236842105265e-07, "loss": 1.918, "step": 4742 }, { "epoch": 0.9751233552631579, "grad_norm": 19.065059661865234, "learning_rate": 2.4876644736842106e-07, "loss": 0.0413, "step": 4743 }, { "epoch": 0.975328947368421, "grad_norm": 117.22676086425781, "learning_rate": 2.467105263157895e-07, "loss": 0.4619, "step": 4744 }, { "epoch": 0.9755345394736842, "grad_norm": 164.26028442382812, "learning_rate": 2.4465460526315793e-07, "loss": 0.7886, "step": 4745 }, { "epoch": 0.9757401315789473, "grad_norm": 33.84805679321289, "learning_rate": 2.4259868421052634e-07, "loss": 0.1094, "step": 4746 }, { "epoch": 0.9759457236842105, "grad_norm": 5.445568561553955, "learning_rate": 2.4054276315789475e-07, "loss": 0.013, "step": 4747 }, { "epoch": 0.9761513157894737, "grad_norm": 96.22848510742188, "learning_rate": 2.384868421052632e-07, "loss": 0.7093, "step": 4748 }, { "epoch": 0.9763569078947368, "grad_norm": 92.61377716064453, "learning_rate": 2.364309210526316e-07, "loss": 0.6885, "step": 4749 }, { "epoch": 0.9765625, "grad_norm": 13.326744079589844, "learning_rate": 2.3437500000000003e-07, "loss": 0.0392, "step": 4750 }, { "epoch": 0.9767680921052632, "grad_norm": 165.1881866455078, "learning_rate": 2.3231907894736844e-07, "loss": 0.5442, "step": 4751 }, { "epoch": 0.9769736842105263, "grad_norm": 103.15234375, "learning_rate": 2.3026315789473685e-07, "loss": 0.427, "step": 4752 }, { "epoch": 0.9771792763157895, "grad_norm": 4.468114376068115, "learning_rate": 2.282072368421053e-07, "loss": 0.0076, "step": 4753 }, { "epoch": 0.9773848684210527, "grad_norm": 23.414913177490234, "learning_rate": 2.261513157894737e-07, "loss": 0.0599, "step": 4754 }, { "epoch": 0.9775904605263158, "grad_norm": 16.041675567626953, "learning_rate": 2.240953947368421e-07, "loss": 0.0571, "step": 4755 }, { "epoch": 0.977796052631579, "grad_norm": 177.9516143798828, "learning_rate": 2.2203947368421055e-07, "loss": 0.623, "step": 4756 }, { "epoch": 0.9780016447368421, "grad_norm": 41.40887451171875, "learning_rate": 2.1998355263157898e-07, "loss": 0.056, "step": 4757 }, { "epoch": 0.9782072368421053, "grad_norm": 44.10049057006836, "learning_rate": 2.1792763157894737e-07, "loss": 0.1871, "step": 4758 }, { "epoch": 0.9784128289473685, "grad_norm": 32.07536697387695, "learning_rate": 2.158717105263158e-07, "loss": 0.0598, "step": 4759 }, { "epoch": 0.9786184210526315, "grad_norm": 56.327030181884766, "learning_rate": 2.1381578947368424e-07, "loss": 0.659, "step": 4760 }, { "epoch": 0.9788240131578947, "grad_norm": 32.162506103515625, "learning_rate": 2.1175986842105268e-07, "loss": 0.0787, "step": 4761 }, { "epoch": 0.9790296052631579, "grad_norm": 14.568992614746094, "learning_rate": 2.0970394736842106e-07, "loss": 0.0304, "step": 4762 }, { "epoch": 0.979235197368421, "grad_norm": 27.114839553833008, "learning_rate": 2.076480263157895e-07, "loss": 0.0747, "step": 4763 }, { "epoch": 0.9794407894736842, "grad_norm": 94.70319366455078, "learning_rate": 2.0559210526315793e-07, "loss": 0.2246, "step": 4764 }, { "epoch": 0.9796463815789473, "grad_norm": 5.897781848907471, "learning_rate": 2.0353618421052632e-07, "loss": 0.0078, "step": 4765 }, { "epoch": 0.9798519736842105, "grad_norm": 27.81266975402832, "learning_rate": 2.0148026315789475e-07, "loss": 0.0593, "step": 4766 }, { "epoch": 0.9800575657894737, "grad_norm": 4.214711666107178, "learning_rate": 1.994243421052632e-07, "loss": 0.0112, "step": 4767 }, { "epoch": 0.9802631578947368, "grad_norm": 60.423187255859375, "learning_rate": 1.9736842105263157e-07, "loss": 0.1053, "step": 4768 }, { "epoch": 0.98046875, "grad_norm": 8.021951675415039, "learning_rate": 1.953125e-07, "loss": 0.0355, "step": 4769 }, { "epoch": 0.9806743421052632, "grad_norm": 164.4381561279297, "learning_rate": 1.9325657894736845e-07, "loss": 0.511, "step": 4770 }, { "epoch": 0.9808799342105263, "grad_norm": 6.89481258392334, "learning_rate": 1.9120065789473686e-07, "loss": 0.0168, "step": 4771 }, { "epoch": 0.9810855263157895, "grad_norm": 91.4432601928711, "learning_rate": 1.8914473684210527e-07, "loss": 0.4785, "step": 4772 }, { "epoch": 0.9812911184210527, "grad_norm": 91.46087646484375, "learning_rate": 1.870888157894737e-07, "loss": 0.5764, "step": 4773 }, { "epoch": 0.9814967105263158, "grad_norm": 73.45159912109375, "learning_rate": 1.850328947368421e-07, "loss": 0.2764, "step": 4774 }, { "epoch": 0.981702302631579, "grad_norm": 13.855812072753906, "learning_rate": 1.8297697368421055e-07, "loss": 0.0278, "step": 4775 }, { "epoch": 0.9819078947368421, "grad_norm": 2.264664649963379, "learning_rate": 1.8092105263157896e-07, "loss": 0.007, "step": 4776 }, { "epoch": 0.9821134868421053, "grad_norm": 59.14225387573242, "learning_rate": 1.7886513157894737e-07, "loss": 0.6026, "step": 4777 }, { "epoch": 0.9823190789473685, "grad_norm": 10.969922065734863, "learning_rate": 1.768092105263158e-07, "loss": 0.0224, "step": 4778 }, { "epoch": 0.9825246710526315, "grad_norm": 11.121930122375488, "learning_rate": 1.7475328947368424e-07, "loss": 0.0258, "step": 4779 }, { "epoch": 0.9827302631578947, "grad_norm": 17.650808334350586, "learning_rate": 1.7269736842105265e-07, "loss": 0.0592, "step": 4780 }, { "epoch": 0.9829358552631579, "grad_norm": 38.443546295166016, "learning_rate": 1.7064144736842106e-07, "loss": 0.1702, "step": 4781 }, { "epoch": 0.983141447368421, "grad_norm": 356.13861083984375, "learning_rate": 1.685855263157895e-07, "loss": 2.5713, "step": 4782 }, { "epoch": 0.9833470394736842, "grad_norm": 3.651233434677124, "learning_rate": 1.6652960526315794e-07, "loss": 0.0078, "step": 4783 }, { "epoch": 0.9835526315789473, "grad_norm": 222.65162658691406, "learning_rate": 1.6447368421052632e-07, "loss": 0.7316, "step": 4784 }, { "epoch": 0.9837582236842105, "grad_norm": 5.7314558029174805, "learning_rate": 1.6241776315789476e-07, "loss": 0.0108, "step": 4785 }, { "epoch": 0.9839638157894737, "grad_norm": 196.007568359375, "learning_rate": 1.603618421052632e-07, "loss": 2.5731, "step": 4786 }, { "epoch": 0.9841694078947368, "grad_norm": 8.562294006347656, "learning_rate": 1.5830592105263158e-07, "loss": 0.0116, "step": 4787 }, { "epoch": 0.984375, "grad_norm": 54.7445068359375, "learning_rate": 1.5625e-07, "loss": 0.1687, "step": 4788 }, { "epoch": 0.9845805921052632, "grad_norm": 6.222727298736572, "learning_rate": 1.5419407894736842e-07, "loss": 0.0115, "step": 4789 }, { "epoch": 0.9847861842105263, "grad_norm": 2.5321907997131348, "learning_rate": 1.5213815789473686e-07, "loss": 0.0081, "step": 4790 }, { "epoch": 0.9849917763157895, "grad_norm": 10.487621307373047, "learning_rate": 1.5008223684210527e-07, "loss": 0.043, "step": 4791 }, { "epoch": 0.9851973684210527, "grad_norm": 19.877349853515625, "learning_rate": 1.4802631578947368e-07, "loss": 0.0452, "step": 4792 }, { "epoch": 0.9854029605263158, "grad_norm": 18.825862884521484, "learning_rate": 1.4597039473684212e-07, "loss": 0.0845, "step": 4793 }, { "epoch": 0.985608552631579, "grad_norm": 4.3244309425354, "learning_rate": 1.4391447368421053e-07, "loss": 0.0053, "step": 4794 }, { "epoch": 0.9858141447368421, "grad_norm": 12.952360153198242, "learning_rate": 1.4185855263157896e-07, "loss": 0.0631, "step": 4795 }, { "epoch": 0.9860197368421053, "grad_norm": 80.86505889892578, "learning_rate": 1.3980263157894737e-07, "loss": 0.3563, "step": 4796 }, { "epoch": 0.9862253289473685, "grad_norm": 158.40733337402344, "learning_rate": 1.3774671052631578e-07, "loss": 0.6445, "step": 4797 }, { "epoch": 0.9864309210526315, "grad_norm": 61.26328659057617, "learning_rate": 1.3569078947368422e-07, "loss": 0.2651, "step": 4798 }, { "epoch": 0.9866365131578947, "grad_norm": 76.87362670898438, "learning_rate": 1.3363486842105263e-07, "loss": 1.418, "step": 4799 }, { "epoch": 0.9868421052631579, "grad_norm": 75.13018035888672, "learning_rate": 1.3157894736842107e-07, "loss": 0.1562, "step": 4800 }, { "epoch": 0.987047697368421, "grad_norm": 112.9920654296875, "learning_rate": 1.2952302631578948e-07, "loss": 0.2348, "step": 4801 }, { "epoch": 0.9872532894736842, "grad_norm": 304.11614990234375, "learning_rate": 1.274671052631579e-07, "loss": 1.2441, "step": 4802 }, { "epoch": 0.9874588815789473, "grad_norm": 22.641719818115234, "learning_rate": 1.2541118421052632e-07, "loss": 0.0544, "step": 4803 }, { "epoch": 0.9876644736842105, "grad_norm": 66.4428482055664, "learning_rate": 1.2335526315789476e-07, "loss": 0.4551, "step": 4804 }, { "epoch": 0.9878700657894737, "grad_norm": 6.83738374710083, "learning_rate": 1.2129934210526317e-07, "loss": 0.0144, "step": 4805 }, { "epoch": 0.9880756578947368, "grad_norm": 27.136829376220703, "learning_rate": 1.192434210526316e-07, "loss": 0.0614, "step": 4806 }, { "epoch": 0.98828125, "grad_norm": 49.123016357421875, "learning_rate": 1.1718750000000002e-07, "loss": 0.1265, "step": 4807 }, { "epoch": 0.9884868421052632, "grad_norm": 36.36695861816406, "learning_rate": 1.1513157894736843e-07, "loss": 0.0849, "step": 4808 }, { "epoch": 0.9886924342105263, "grad_norm": 21.206819534301758, "learning_rate": 1.1307565789473685e-07, "loss": 0.041, "step": 4809 }, { "epoch": 0.9888980263157895, "grad_norm": 3.6354928016662598, "learning_rate": 1.1101973684210527e-07, "loss": 0.0062, "step": 4810 }, { "epoch": 0.9891036184210527, "grad_norm": 128.33126831054688, "learning_rate": 1.0896381578947368e-07, "loss": 0.216, "step": 4811 }, { "epoch": 0.9893092105263158, "grad_norm": 180.9879608154297, "learning_rate": 1.0690789473684212e-07, "loss": 1.0859, "step": 4812 }, { "epoch": 0.989514802631579, "grad_norm": 291.8801574707031, "learning_rate": 1.0485197368421053e-07, "loss": 0.5141, "step": 4813 }, { "epoch": 0.9897203947368421, "grad_norm": 453.0140075683594, "learning_rate": 1.0279605263157897e-07, "loss": 1.6189, "step": 4814 }, { "epoch": 0.9899259868421053, "grad_norm": 17.189939498901367, "learning_rate": 1.0074013157894738e-07, "loss": 0.0304, "step": 4815 }, { "epoch": 0.9901315789473685, "grad_norm": 47.082847595214844, "learning_rate": 9.868421052631579e-08, "loss": 0.2402, "step": 4816 }, { "epoch": 0.9903371710526315, "grad_norm": 57.63482666015625, "learning_rate": 9.662828947368422e-08, "loss": 0.3623, "step": 4817 }, { "epoch": 0.9905427631578947, "grad_norm": 12.408506393432617, "learning_rate": 9.457236842105263e-08, "loss": 0.0209, "step": 4818 }, { "epoch": 0.9907483552631579, "grad_norm": 141.89816284179688, "learning_rate": 9.251644736842106e-08, "loss": 0.4062, "step": 4819 }, { "epoch": 0.990953947368421, "grad_norm": 53.703025817871094, "learning_rate": 9.046052631578948e-08, "loss": 0.1211, "step": 4820 }, { "epoch": 0.9911595394736842, "grad_norm": 311.29315185546875, "learning_rate": 8.84046052631579e-08, "loss": 0.6777, "step": 4821 }, { "epoch": 0.9913651315789473, "grad_norm": 51.08918380737305, "learning_rate": 8.634868421052633e-08, "loss": 0.1844, "step": 4822 }, { "epoch": 0.9915707236842105, "grad_norm": 126.20582580566406, "learning_rate": 8.429276315789475e-08, "loss": 1.29, "step": 4823 }, { "epoch": 0.9917763157894737, "grad_norm": 26.55607795715332, "learning_rate": 8.223684210526316e-08, "loss": 0.0652, "step": 4824 }, { "epoch": 0.9919819078947368, "grad_norm": 12.744282722473145, "learning_rate": 8.01809210526316e-08, "loss": 0.0183, "step": 4825 }, { "epoch": 0.9921875, "grad_norm": 124.55632019042969, "learning_rate": 7.8125e-08, "loss": 0.3918, "step": 4826 }, { "epoch": 0.9923930921052632, "grad_norm": 164.97576904296875, "learning_rate": 7.606907894736843e-08, "loss": 0.8774, "step": 4827 }, { "epoch": 0.9925986842105263, "grad_norm": 36.00751495361328, "learning_rate": 7.401315789473684e-08, "loss": 0.1042, "step": 4828 }, { "epoch": 0.9928042763157895, "grad_norm": 25.414155960083008, "learning_rate": 7.195723684210526e-08, "loss": 0.0453, "step": 4829 }, { "epoch": 0.9930098684210527, "grad_norm": 69.08782958984375, "learning_rate": 6.990131578947369e-08, "loss": 0.3642, "step": 4830 }, { "epoch": 0.9932154605263158, "grad_norm": 109.7258071899414, "learning_rate": 6.784539473684211e-08, "loss": 0.137, "step": 4831 }, { "epoch": 0.993421052631579, "grad_norm": 75.57425689697266, "learning_rate": 6.578947368421053e-08, "loss": 0.148, "step": 4832 }, { "epoch": 0.9936266447368421, "grad_norm": 178.1168670654297, "learning_rate": 6.373355263157896e-08, "loss": 1.0282, "step": 4833 }, { "epoch": 0.9938322368421053, "grad_norm": 3.7407517433166504, "learning_rate": 6.167763157894738e-08, "loss": 0.0221, "step": 4834 }, { "epoch": 0.9940378289473685, "grad_norm": 10.285094261169434, "learning_rate": 5.96217105263158e-08, "loss": 0.0173, "step": 4835 }, { "epoch": 0.9942434210526315, "grad_norm": 3.5968551635742188, "learning_rate": 5.756578947368421e-08, "loss": 0.0103, "step": 4836 }, { "epoch": 0.9944490131578947, "grad_norm": 159.630615234375, "learning_rate": 5.5509868421052636e-08, "loss": 0.9033, "step": 4837 }, { "epoch": 0.9946546052631579, "grad_norm": 2.3044841289520264, "learning_rate": 5.345394736842106e-08, "loss": 0.0139, "step": 4838 }, { "epoch": 0.994860197368421, "grad_norm": 89.9747543334961, "learning_rate": 5.139802631578948e-08, "loss": 0.1814, "step": 4839 }, { "epoch": 0.9950657894736842, "grad_norm": 56.441741943359375, "learning_rate": 4.934210526315789e-08, "loss": 0.2108, "step": 4840 }, { "epoch": 0.9952713815789473, "grad_norm": 192.45004272460938, "learning_rate": 4.7286184210526317e-08, "loss": 1.2238, "step": 4841 }, { "epoch": 0.9954769736842105, "grad_norm": 44.12889862060547, "learning_rate": 4.523026315789474e-08, "loss": 0.6824, "step": 4842 }, { "epoch": 0.9956825657894737, "grad_norm": 47.149818420410156, "learning_rate": 4.317434210526316e-08, "loss": 0.2163, "step": 4843 }, { "epoch": 0.9958881578947368, "grad_norm": 85.46209716796875, "learning_rate": 4.111842105263158e-08, "loss": 0.6719, "step": 4844 }, { "epoch": 0.99609375, "grad_norm": 14.93366527557373, "learning_rate": 3.90625e-08, "loss": 0.0503, "step": 4845 }, { "epoch": 0.9962993421052632, "grad_norm": 22.114988327026367, "learning_rate": 3.700657894736842e-08, "loss": 0.0351, "step": 4846 }, { "epoch": 0.9965049342105263, "grad_norm": 91.51122283935547, "learning_rate": 3.495065789473684e-08, "loss": 0.5808, "step": 4847 }, { "epoch": 0.9967105263157895, "grad_norm": 8.234692573547363, "learning_rate": 3.2894736842105267e-08, "loss": 0.0295, "step": 4848 }, { "epoch": 0.9969161184210527, "grad_norm": 21.95867919921875, "learning_rate": 3.083881578947369e-08, "loss": 0.0776, "step": 4849 }, { "epoch": 0.9971217105263158, "grad_norm": 35.53308868408203, "learning_rate": 2.8782894736842107e-08, "loss": 0.1031, "step": 4850 }, { "epoch": 0.997327302631579, "grad_norm": 25.555606842041016, "learning_rate": 2.672697368421053e-08, "loss": 0.064, "step": 4851 }, { "epoch": 0.9975328947368421, "grad_norm": 24.26386833190918, "learning_rate": 2.4671052631578947e-08, "loss": 0.0614, "step": 4852 }, { "epoch": 0.9977384868421053, "grad_norm": 17.68927001953125, "learning_rate": 2.261513157894737e-08, "loss": 0.0549, "step": 4853 }, { "epoch": 0.9979440789473685, "grad_norm": 14.734000205993652, "learning_rate": 2.055921052631579e-08, "loss": 0.0277, "step": 4854 }, { "epoch": 0.9981496710526315, "grad_norm": 40.01732635498047, "learning_rate": 1.850328947368421e-08, "loss": 0.1147, "step": 4855 }, { "epoch": 0.9983552631578947, "grad_norm": 238.25328063964844, "learning_rate": 1.6447368421052633e-08, "loss": 1.1533, "step": 4856 }, { "epoch": 0.9985608552631579, "grad_norm": 255.74339294433594, "learning_rate": 1.4391447368421053e-08, "loss": 1.6484, "step": 4857 }, { "epoch": 0.998766447368421, "grad_norm": 207.79933166503906, "learning_rate": 1.2335526315789473e-08, "loss": 0.5876, "step": 4858 }, { "epoch": 0.9989720394736842, "grad_norm": 10.222477912902832, "learning_rate": 1.0279605263157895e-08, "loss": 0.0177, "step": 4859 }, { "epoch": 0.9991776315789473, "grad_norm": 4.787688732147217, "learning_rate": 8.223684210526317e-09, "loss": 0.0085, "step": 4860 }, { "epoch": 0.9993832236842105, "grad_norm": 135.98751831054688, "learning_rate": 6.167763157894737e-09, "loss": 0.4688, "step": 4861 }, { "epoch": 0.9995888157894737, "grad_norm": 177.64784240722656, "learning_rate": 4.111842105263158e-09, "loss": 0.9193, "step": 4862 }, { "epoch": 0.9997944078947368, "grad_norm": 32.11542510986328, "learning_rate": 2.055921052631579e-09, "loss": 0.0718, "step": 4863 }, { "epoch": 1.0, "grad_norm": 12.889538764953613, "learning_rate": 0.0, "loss": 0.0419, "step": 4864 }, { "epoch": 1.0, "eval_accuracy": 0.7184284377923292, "eval_loss": 0.746451199054718, "eval_runtime": 172.4929, "eval_samples_per_second": 6.215, "eval_steps_per_second": 6.215, "step": 4864 } ], "logging_steps": 1, "max_steps": 4864, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }