| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 295, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.003395585738539898, |
| "grad_norm": 6.320343624704885, |
| "learning_rate": 0.0, |
| "loss": 0.9899, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.006791171477079796, |
| "grad_norm": 16.545756923443665, |
| "learning_rate": 3.3333333333333335e-07, |
| "loss": 0.9683, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.010186757215619695, |
| "grad_norm": 6.124626290117518, |
| "learning_rate": 6.666666666666667e-07, |
| "loss": 0.9722, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.013582342954159592, |
| "grad_norm": 6.907471369837716, |
| "learning_rate": 1.0000000000000002e-06, |
| "loss": 0.9667, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.01697792869269949, |
| "grad_norm": 6.275431523928528, |
| "learning_rate": 1.3333333333333334e-06, |
| "loss": 0.9814, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.02037351443123939, |
| "grad_norm": 6.074623009583982, |
| "learning_rate": 1.6666666666666667e-06, |
| "loss": 0.9502, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.023769100169779286, |
| "grad_norm": 89.99670872123082, |
| "learning_rate": 2.0000000000000003e-06, |
| "loss": 0.9411, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.027164685908319185, |
| "grad_norm": 6.224698090768632, |
| "learning_rate": 2.3333333333333336e-06, |
| "loss": 0.9075, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.030560271646859084, |
| "grad_norm": 5.913600661145363, |
| "learning_rate": 2.666666666666667e-06, |
| "loss": 0.8979, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.03395585738539898, |
| "grad_norm": 7.267004601688997, |
| "learning_rate": 3e-06, |
| "loss": 0.8589, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.03735144312393888, |
| "grad_norm": 4.920293108168373, |
| "learning_rate": 3.3333333333333333e-06, |
| "loss": 0.8797, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.04074702886247878, |
| "grad_norm": 7.0251120254603325, |
| "learning_rate": 3.6666666666666666e-06, |
| "loss": 0.785, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.044142614601018676, |
| "grad_norm": 4.620531235688851, |
| "learning_rate": 4.000000000000001e-06, |
| "loss": 0.7932, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.04753820033955857, |
| "grad_norm": 8.017042756220283, |
| "learning_rate": 4.333333333333334e-06, |
| "loss": 0.7878, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.050933786078098474, |
| "grad_norm": 7.168335735527789, |
| "learning_rate": 4.666666666666667e-06, |
| "loss": 0.7854, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.05432937181663837, |
| "grad_norm": 2.6102783506270137, |
| "learning_rate": 5e-06, |
| "loss": 0.7194, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.057724957555178265, |
| "grad_norm": 2.8281275726026074, |
| "learning_rate": 5.333333333333334e-06, |
| "loss": 0.7059, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.06112054329371817, |
| "grad_norm": 2.1330350735122505, |
| "learning_rate": 5.666666666666667e-06, |
| "loss": 0.7236, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.06451612903225806, |
| "grad_norm": 3.3828007198200694, |
| "learning_rate": 6e-06, |
| "loss": 0.7133, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.06791171477079797, |
| "grad_norm": 1.733407506222324, |
| "learning_rate": 6.333333333333333e-06, |
| "loss": 0.6807, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.07130730050933787, |
| "grad_norm": 1.5732888077219966, |
| "learning_rate": 6.666666666666667e-06, |
| "loss": 0.6717, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.07470288624787776, |
| "grad_norm": 2.869294784618442, |
| "learning_rate": 7e-06, |
| "loss": 0.6836, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.07809847198641766, |
| "grad_norm": 1.539403446338188, |
| "learning_rate": 7.333333333333333e-06, |
| "loss": 0.656, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.08149405772495756, |
| "grad_norm": 1.390711875015812, |
| "learning_rate": 7.666666666666667e-06, |
| "loss": 0.652, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.08488964346349745, |
| "grad_norm": 1.4133171062619616, |
| "learning_rate": 8.000000000000001e-06, |
| "loss": 0.6237, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.08828522920203735, |
| "grad_norm": 1.2675100082185478, |
| "learning_rate": 8.333333333333334e-06, |
| "loss": 0.5999, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.09168081494057725, |
| "grad_norm": 1.4010718655700412, |
| "learning_rate": 8.666666666666668e-06, |
| "loss": 0.6229, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.09507640067911714, |
| "grad_norm": 1.4302934299507057, |
| "learning_rate": 9e-06, |
| "loss": 0.5909, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.09847198641765705, |
| "grad_norm": 1.3417528783040296, |
| "learning_rate": 9.333333333333334e-06, |
| "loss": 0.6236, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.10186757215619695, |
| "grad_norm": 1.2550326498175997, |
| "learning_rate": 9.666666666666667e-06, |
| "loss": 0.5978, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.10526315789473684, |
| "grad_norm": 1.2949203936642646, |
| "learning_rate": 1e-05, |
| "loss": 0.6206, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.10865874363327674, |
| "grad_norm": 1.3404316606873072, |
| "learning_rate": 9.999648647603774e-06, |
| "loss": 0.6103, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.11205432937181664, |
| "grad_norm": 1.2464543142979883, |
| "learning_rate": 9.998594639794502e-06, |
| "loss": 0.5823, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.11544991511035653, |
| "grad_norm": 1.28881560865773, |
| "learning_rate": 9.996838124703448e-06, |
| "loss": 0.5724, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.11884550084889643, |
| "grad_norm": 1.236715524722473, |
| "learning_rate": 9.994379349192927e-06, |
| "loss": 0.6006, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.12224108658743633, |
| "grad_norm": 1.1748122882399419, |
| "learning_rate": 9.991218658821609e-06, |
| "loss": 0.5834, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.12563667232597622, |
| "grad_norm": 1.3335055567165142, |
| "learning_rate": 9.987356497795944e-06, |
| "loss": 0.5781, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.12903225806451613, |
| "grad_norm": 1.2573177388462913, |
| "learning_rate": 9.982793408907747e-06, |
| "loss": 0.5613, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.13242784380305603, |
| "grad_norm": 614.0974853168025, |
| "learning_rate": 9.977530033457906e-06, |
| "loss": 0.5518, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.13582342954159593, |
| "grad_norm": 1.8130801394796245, |
| "learning_rate": 9.971567111166246e-06, |
| "loss": 0.5875, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.13921901528013583, |
| "grad_norm": 3.021540974319598, |
| "learning_rate": 9.964905480067585e-06, |
| "loss": 0.5666, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.14261460101867574, |
| "grad_norm": 1.4307732415515269, |
| "learning_rate": 9.957546076393944e-06, |
| "loss": 0.5829, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.1460101867572156, |
| "grad_norm": 1.2948953962351142, |
| "learning_rate": 9.949489934442966e-06, |
| "loss": 0.592, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.1494057724957555, |
| "grad_norm": 1.28013431192408, |
| "learning_rate": 9.940738186432565e-06, |
| "loss": 0.554, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.15280135823429541, |
| "grad_norm": 1.364283730780204, |
| "learning_rate": 9.931292062341793e-06, |
| "loss": 0.5563, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.15619694397283532, |
| "grad_norm": 1.1789812327801241, |
| "learning_rate": 9.921152889737985e-06, |
| "loss": 0.5349, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.15959252971137522, |
| "grad_norm": 1.272995264216437, |
| "learning_rate": 9.910322093590177e-06, |
| "loss": 0.5885, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.16298811544991512, |
| "grad_norm": 1.2071797315958241, |
| "learning_rate": 9.898801196068839e-06, |
| "loss": 0.555, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.166383701188455, |
| "grad_norm": 1.2466300266575705, |
| "learning_rate": 9.886591816331953e-06, |
| "loss": 0.5412, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.1697792869269949, |
| "grad_norm": 1.1951025261533827, |
| "learning_rate": 9.87369567029745e-06, |
| "loss": 0.5266, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.1731748726655348, |
| "grad_norm": 1.1409578758956818, |
| "learning_rate": 9.860114570402055e-06, |
| "loss": 0.5364, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.1765704584040747, |
| "grad_norm": 1.194968225105999, |
| "learning_rate": 9.845850425346563e-06, |
| "loss": 0.544, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.1799660441426146, |
| "grad_norm": 1.105032206844106, |
| "learning_rate": 9.830905239827592e-06, |
| "loss": 0.526, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.1833616298811545, |
| "grad_norm": 1.0951143452700236, |
| "learning_rate": 9.815281114255841e-06, |
| "loss": 0.5384, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.1867572156196944, |
| "grad_norm": 1.237628011550819, |
| "learning_rate": 9.798980244460892e-06, |
| "loss": 0.5216, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.19015280135823429, |
| "grad_norm": 1.2049941993061364, |
| "learning_rate": 9.782004921382612e-06, |
| "loss": 0.5602, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.1935483870967742, |
| "grad_norm": 1.0963401467603866, |
| "learning_rate": 9.764357530749178e-06, |
| "loss": 0.5211, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.1969439728353141, |
| "grad_norm": 1.1153693044651702, |
| "learning_rate": 9.74604055274178e-06, |
| "loss": 0.5408, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.200339558573854, |
| "grad_norm": 1.1446372149192978, |
| "learning_rate": 9.727056561646067e-06, |
| "loss": 0.5297, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.2037351443123939, |
| "grad_norm": 1.1379654178567067, |
| "learning_rate": 9.707408225490343e-06, |
| "loss": 0.5241, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.2071307300509338, |
| "grad_norm": 1.1133344090884694, |
| "learning_rate": 9.687098305670606e-06, |
| "loss": 0.5373, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.21052631578947367, |
| "grad_norm": 1.14642700050245, |
| "learning_rate": 9.66612965656245e-06, |
| "loss": 0.5397, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.21392190152801357, |
| "grad_norm": 1.0251599768158084, |
| "learning_rate": 9.644505225119922e-06, |
| "loss": 0.5444, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.21731748726655348, |
| "grad_norm": 0.9890854380699226, |
| "learning_rate": 9.622228050461345e-06, |
| "loss": 0.4984, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.22071307300509338, |
| "grad_norm": 1.1117068387014335, |
| "learning_rate": 9.599301263442194e-06, |
| "loss": 0.5214, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.22410865874363328, |
| "grad_norm": 1.1785090789686485, |
| "learning_rate": 9.575728086215093e-06, |
| "loss": 0.5265, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.22750424448217318, |
| "grad_norm": 1.0565597242070186, |
| "learning_rate": 9.551511831776966e-06, |
| "loss": 0.5155, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.23089983022071306, |
| "grad_norm": 1.2278748412937952, |
| "learning_rate": 9.526655903503423e-06, |
| "loss": 0.5277, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.23429541595925296, |
| "grad_norm": 1.1213279864335683, |
| "learning_rate": 9.501163794670445e-06, |
| "loss": 0.5307, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.23769100169779286, |
| "grad_norm": 1.1237007845490015, |
| "learning_rate": 9.475039087963443e-06, |
| "loss": 0.51, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.24108658743633277, |
| "grad_norm": 1.0615774210564253, |
| "learning_rate": 9.448285454973739e-06, |
| "loss": 0.517, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.24448217317487267, |
| "grad_norm": 1.2104405507373044, |
| "learning_rate": 9.420906655682553e-06, |
| "loss": 0.5223, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.24787775891341257, |
| "grad_norm": 1.0985265075768098, |
| "learning_rate": 9.392906537932582e-06, |
| "loss": 0.5257, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.25127334465195245, |
| "grad_norm": 1.0626836434651132, |
| "learning_rate": 9.364289036887214e-06, |
| "loss": 0.5187, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.2546689303904924, |
| "grad_norm": 1.0234412989238388, |
| "learning_rate": 9.335058174477472e-06, |
| "loss": 0.5127, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.25806451612903225, |
| "grad_norm": 1.0702433549745685, |
| "learning_rate": 9.305218058836778e-06, |
| "loss": 0.4966, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.2614601018675722, |
| "grad_norm": 1.1228773632449267, |
| "learning_rate": 9.274772883723587e-06, |
| "loss": 0.5265, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.26485568760611206, |
| "grad_norm": 1.0076068781933833, |
| "learning_rate": 9.24372692793199e-06, |
| "loss": 0.5003, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.26825127334465193, |
| "grad_norm": 1.0987680518642855, |
| "learning_rate": 9.21208455469037e-06, |
| "loss": 0.4971, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.27164685908319186, |
| "grad_norm": 1.1051172285171345, |
| "learning_rate": 9.179850211048193e-06, |
| "loss": 0.491, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.27504244482173174, |
| "grad_norm": 1.0886761114445709, |
| "learning_rate": 9.14702842725101e-06, |
| "loss": 0.5033, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.27843803056027167, |
| "grad_norm": 1.101989710769037, |
| "learning_rate": 9.113623816103775e-06, |
| "loss": 0.5157, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.28183361629881154, |
| "grad_norm": 1.0846044794179348, |
| "learning_rate": 9.079641072322555e-06, |
| "loss": 0.5103, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.28522920203735147, |
| "grad_norm": 1.0994243701550208, |
| "learning_rate": 9.045084971874738e-06, |
| "loss": 0.5118, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.28862478777589134, |
| "grad_norm": 0.982679989364588, |
| "learning_rate": 9.009960371307798e-06, |
| "loss": 0.4753, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.2920203735144312, |
| "grad_norm": 1.0974557383394132, |
| "learning_rate": 8.974272207066767e-06, |
| "loss": 0.5065, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.29541595925297115, |
| "grad_norm": 1.191767026119928, |
| "learning_rate": 8.938025494800454e-06, |
| "loss": 0.496, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.298811544991511, |
| "grad_norm": 1.0863558052080193, |
| "learning_rate": 8.901225328656543e-06, |
| "loss": 0.4768, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.30220713073005095, |
| "grad_norm": 1.0292339490206497, |
| "learning_rate": 8.863876880565656e-06, |
| "loss": 0.4982, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.30560271646859083, |
| "grad_norm": 1.0263488228624327, |
| "learning_rate": 8.825985399514488e-06, |
| "loss": 0.49, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.3089983022071307, |
| "grad_norm": 1.1083934275101859, |
| "learning_rate": 8.787556210808101e-06, |
| "loss": 0.508, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.31239388794567063, |
| "grad_norm": 1.0759718773481644, |
| "learning_rate": 8.748594715321512e-06, |
| "loss": 0.493, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.3157894736842105, |
| "grad_norm": 1.1275176872041444, |
| "learning_rate": 8.70910638874064e-06, |
| "loss": 0.5063, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.31918505942275044, |
| "grad_norm": 1.0268180450930244, |
| "learning_rate": 8.669096780792754e-06, |
| "loss": 0.4953, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.3225806451612903, |
| "grad_norm": 1.0957503547012442, |
| "learning_rate": 8.628571514466502e-06, |
| "loss": 0.5062, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.32597623089983024, |
| "grad_norm": 1.0476160480506713, |
| "learning_rate": 8.587536285221656e-06, |
| "loss": 0.5254, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.3293718166383701, |
| "grad_norm": 0.973467408933069, |
| "learning_rate": 8.545996860188668e-06, |
| "loss": 0.5107, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.33276740237691, |
| "grad_norm": 0.9690640110532758, |
| "learning_rate": 8.503959077358143e-06, |
| "loss": 0.4767, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.3361629881154499, |
| "grad_norm": 1.1272513638936046, |
| "learning_rate": 8.46142884476038e-06, |
| "loss": 0.4991, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.3395585738539898, |
| "grad_norm": 1.089949180068211, |
| "learning_rate": 8.418412139635026e-06, |
| "loss": 0.5202, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.34295415959252973, |
| "grad_norm": 1.0904976650398113, |
| "learning_rate": 8.374915007591053e-06, |
| "loss": 0.5019, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.3463497453310696, |
| "grad_norm": 1.0132973869482662, |
| "learning_rate": 8.330943561757092e-06, |
| "loss": 0.4981, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.34974533106960953, |
| "grad_norm": 1.0144345934272716, |
| "learning_rate": 8.286503981922284e-06, |
| "loss": 0.4828, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.3531409168081494, |
| "grad_norm": 1.0708364693456955, |
| "learning_rate": 8.241602513667775e-06, |
| "loss": 0.4677, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.3565365025466893, |
| "grad_norm": 1.0734641865863697, |
| "learning_rate": 8.19624546748895e-06, |
| "loss": 0.4991, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.3599320882852292, |
| "grad_norm": 1.206730960165882, |
| "learning_rate": 8.150439217908557e-06, |
| "loss": 0.4874, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.3633276740237691, |
| "grad_norm": 0.990112721589732, |
| "learning_rate": 8.104190202580811e-06, |
| "loss": 0.4613, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.366723259762309, |
| "grad_norm": 1.135017537331381, |
| "learning_rate": 8.057504921386661e-06, |
| "loss": 0.4981, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.3701188455008489, |
| "grad_norm": 1.1168906124405507, |
| "learning_rate": 8.010389935520269e-06, |
| "loss": 0.4812, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.3735144312393888, |
| "grad_norm": 1.109365518761395, |
| "learning_rate": 7.962851866566912e-06, |
| "loss": 0.4901, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.3769100169779287, |
| "grad_norm": 1.1121750297406146, |
| "learning_rate": 7.914897395572362e-06, |
| "loss": 0.4841, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.38030560271646857, |
| "grad_norm": 1.10407244769018, |
| "learning_rate": 7.866533262103937e-06, |
| "loss": 0.4813, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.3837011884550085, |
| "grad_norm": 1.109502100982312, |
| "learning_rate": 7.817766263303312e-06, |
| "loss": 0.4731, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.3870967741935484, |
| "grad_norm": 1.0709239172717544, |
| "learning_rate": 7.768603252931243e-06, |
| "loss": 0.4988, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.3904923599320883, |
| "grad_norm": 1.0915174076138725, |
| "learning_rate": 7.719051140404327e-06, |
| "loss": 0.4825, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.3938879456706282, |
| "grad_norm": 1.0965896575304275, |
| "learning_rate": 7.669116889823955e-06, |
| "loss": 0.4742, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.39728353140916806, |
| "grad_norm": 1.0241307032424805, |
| "learning_rate": 7.6188075189975644e-06, |
| "loss": 0.4743, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.400679117147708, |
| "grad_norm": 0.9435780818732129, |
| "learning_rate": 7.568130098452352e-06, |
| "loss": 0.4739, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.40407470288624786, |
| "grad_norm": 1.0248298007190784, |
| "learning_rate": 7.517091750441576e-06, |
| "loss": 0.4814, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.4074702886247878, |
| "grad_norm": 1.0315243401737155, |
| "learning_rate": 7.465699647943586e-06, |
| "loss": 0.4805, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.41086587436332767, |
| "grad_norm": 1.091443614674752, |
| "learning_rate": 7.413961013653725e-06, |
| "loss": 0.4953, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.4142614601018676, |
| "grad_norm": 1.0680561484041602, |
| "learning_rate": 7.361883118969248e-06, |
| "loss": 0.4588, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.41765704584040747, |
| "grad_norm": 0.9761759048656751, |
| "learning_rate": 7.309473282967387e-06, |
| "loss": 0.469, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.42105263157894735, |
| "grad_norm": 1.0119186737160737, |
| "learning_rate": 7.256738871376733e-06, |
| "loss": 0.4812, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.4244482173174873, |
| "grad_norm": 0.9670302415957511, |
| "learning_rate": 7.203687295542032e-06, |
| "loss": 0.456, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.42784380305602715, |
| "grad_norm": 1.001107021910696, |
| "learning_rate": 7.1503260113826035e-06, |
| "loss": 0.4597, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.4312393887945671, |
| "grad_norm": 1.024857257795839, |
| "learning_rate": 7.09666251834447e-06, |
| "loss": 0.4884, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.43463497453310695, |
| "grad_norm": 1.0331435407774852, |
| "learning_rate": 7.042704358346375e-06, |
| "loss": 0.4581, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.4380305602716469, |
| "grad_norm": 1.00793619623988, |
| "learning_rate": 6.988459114719849e-06, |
| "loss": 0.4672, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.44142614601018676, |
| "grad_norm": 0.9145035341259203, |
| "learning_rate": 6.933934411143419e-06, |
| "loss": 0.4524, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.44482173174872663, |
| "grad_norm": 1.1118101624400085, |
| "learning_rate": 6.879137910571191e-06, |
| "loss": 0.4738, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.44821731748726656, |
| "grad_norm": 1.0044078413498474, |
| "learning_rate": 6.824077314155877e-06, |
| "loss": 0.4679, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.45161290322580644, |
| "grad_norm": 1.0568591065810937, |
| "learning_rate": 6.768760360166471e-06, |
| "loss": 0.4909, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.45500848896434637, |
| "grad_norm": 0.9937372079053678, |
| "learning_rate": 6.713194822900707e-06, |
| "loss": 0.4734, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.45840407470288624, |
| "grad_norm": 1.0380172124065203, |
| "learning_rate": 6.657388511592453e-06, |
| "loss": 0.4979, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.4617996604414261, |
| "grad_norm": 0.9710945454163669, |
| "learning_rate": 6.601349269314188e-06, |
| "loss": 0.4826, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.46519524617996605, |
| "grad_norm": 1.1195229655523649, |
| "learning_rate": 6.545084971874738e-06, |
| "loss": 0.4685, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.4685908319185059, |
| "grad_norm": 1.079871197310449, |
| "learning_rate": 6.488603526712391e-06, |
| "loss": 0.483, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.47198641765704585, |
| "grad_norm": 1.0089317127122743, |
| "learning_rate": 6.431912871783587e-06, |
| "loss": 0.5028, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.47538200339558573, |
| "grad_norm": 0.9671431137430636, |
| "learning_rate": 6.3750209744473105e-06, |
| "loss": 0.4669, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.47877758913412566, |
| "grad_norm": 1.0520557175009426, |
| "learning_rate": 6.3179358303453386e-06, |
| "loss": 0.4735, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.48217317487266553, |
| "grad_norm": 1.0538462837406155, |
| "learning_rate": 6.260665462278544e-06, |
| "loss": 0.5051, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.4855687606112054, |
| "grad_norm": 1.037965055623141, |
| "learning_rate": 6.203217919079343e-06, |
| "loss": 0.4762, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.48896434634974534, |
| "grad_norm": 1.0176275927771237, |
| "learning_rate": 6.145601274480521e-06, |
| "loss": 0.4869, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.4923599320882852, |
| "grad_norm": 1.0431340329887582, |
| "learning_rate": 6.08782362598054e-06, |
| "loss": 0.4611, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.49575551782682514, |
| "grad_norm": 1.0308571103490511, |
| "learning_rate": 6.029893093705492e-06, |
| "loss": 0.4763, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.499151103565365, |
| "grad_norm": 1.0355702334516321, |
| "learning_rate": 5.971817819267914e-06, |
| "loss": 0.4654, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.5025466893039049, |
| "grad_norm": 0.9893529623332218, |
| "learning_rate": 5.9136059646225375e-06, |
| "loss": 0.5044, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.5059422750424448, |
| "grad_norm": 0.960195438000855, |
| "learning_rate": 5.855265710919211e-06, |
| "loss": 0.4435, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.5093378607809848, |
| "grad_norm": 0.9646163806867624, |
| "learning_rate": 5.796805257353109e-06, |
| "loss": 0.4834, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.5127334465195246, |
| "grad_norm": 0.9931874900764169, |
| "learning_rate": 5.738232820012407e-06, |
| "loss": 0.461, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.5161290322580645, |
| "grad_norm": 1.0149629817842267, |
| "learning_rate": 5.679556630723592e-06, |
| "loss": 0.4611, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.5195246179966044, |
| "grad_norm": 1.0022805808479243, |
| "learning_rate": 5.620784935894548e-06, |
| "loss": 0.4712, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.5229202037351444, |
| "grad_norm": 0.9091219788824919, |
| "learning_rate": 5.561925995355595e-06, |
| "loss": 0.4906, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.5263157894736842, |
| "grad_norm": 0.9508691466403413, |
| "learning_rate": 5.5029880811986546e-06, |
| "loss": 0.4836, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.5297113752122241, |
| "grad_norm": 0.9816091460679524, |
| "learning_rate": 5.443979476614674e-06, |
| "loss": 0.4568, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.533106960950764, |
| "grad_norm": 0.9890714751673821, |
| "learning_rate": 5.384908474729501e-06, |
| "loss": 0.4669, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.5365025466893039, |
| "grad_norm": 0.9852345937629808, |
| "learning_rate": 5.325783377438357e-06, |
| "loss": 0.4631, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.5398981324278438, |
| "grad_norm": 0.9992258942689997, |
| "learning_rate": 5.266612494239088e-06, |
| "loss": 0.4824, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.5432937181663837, |
| "grad_norm": 1.0010167382867965, |
| "learning_rate": 5.207404141064334e-06, |
| "loss": 0.4619, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.5466893039049237, |
| "grad_norm": 1.117194827109753, |
| "learning_rate": 5.148166639112799e-06, |
| "loss": 0.518, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.5500848896434635, |
| "grad_norm": 0.9423156223035444, |
| "learning_rate": 5.088908313679788e-06, |
| "loss": 0.4567, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.5534804753820034, |
| "grad_norm": 0.919109798014551, |
| "learning_rate": 5.029637492987153e-06, |
| "loss": 0.4649, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.5568760611205433, |
| "grad_norm": 1.1232411363329386, |
| "learning_rate": 4.970362507012848e-06, |
| "loss": 0.4682, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.5602716468590832, |
| "grad_norm": 0.9560703825281516, |
| "learning_rate": 4.911091686320213e-06, |
| "loss": 0.4594, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.5636672325976231, |
| "grad_norm": 0.9927617945956294, |
| "learning_rate": 4.8518333608872015e-06, |
| "loss": 0.4266, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.567062818336163, |
| "grad_norm": 1.0053279255521879, |
| "learning_rate": 4.792595858935668e-06, |
| "loss": 0.4745, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.5704584040747029, |
| "grad_norm": 1.1247627665383018, |
| "learning_rate": 4.733387505760913e-06, |
| "loss": 0.4863, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.5738539898132428, |
| "grad_norm": 0.9506838521386066, |
| "learning_rate": 4.674216622561645e-06, |
| "loss": 0.4459, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.5772495755517827, |
| "grad_norm": 0.8918545360346245, |
| "learning_rate": 4.6150915252705005e-06, |
| "loss": 0.4571, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.5806451612903226, |
| "grad_norm": 0.9704497436094051, |
| "learning_rate": 4.556020523385326e-06, |
| "loss": 0.4701, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.5840407470288624, |
| "grad_norm": 0.9830084007827714, |
| "learning_rate": 4.497011918801347e-06, |
| "loss": 0.477, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.5874363327674024, |
| "grad_norm": 1.0446285595393363, |
| "learning_rate": 4.438074004644407e-06, |
| "loss": 0.4963, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.5908319185059423, |
| "grad_norm": 0.9430261659822053, |
| "learning_rate": 4.379215064105454e-06, |
| "loss": 0.4502, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.5942275042444821, |
| "grad_norm": 0.9528761706191039, |
| "learning_rate": 4.32044336927641e-06, |
| "loss": 0.4667, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.597623089983022, |
| "grad_norm": 0.9712945630496029, |
| "learning_rate": 4.261767179987595e-06, |
| "loss": 0.469, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.601018675721562, |
| "grad_norm": 0.9203712027671123, |
| "learning_rate": 4.203194742646893e-06, |
| "loss": 0.4549, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.6044142614601019, |
| "grad_norm": 0.9335750991253596, |
| "learning_rate": 4.1447342890807905e-06, |
| "loss": 0.4589, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.6078098471986417, |
| "grad_norm": 0.9302840956981335, |
| "learning_rate": 4.086394035377463e-06, |
| "loss": 0.4525, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.6112054329371817, |
| "grad_norm": 0.9205763667360544, |
| "learning_rate": 4.028182180732088e-06, |
| "loss": 0.4652, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.6146010186757216, |
| "grad_norm": 0.9053159533350853, |
| "learning_rate": 3.970106906294509e-06, |
| "loss": 0.4887, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.6179966044142614, |
| "grad_norm": 0.9639072424421392, |
| "learning_rate": 3.912176374019462e-06, |
| "loss": 0.4928, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.6213921901528013, |
| "grad_norm": 0.8934882136410526, |
| "learning_rate": 3.85439872551948e-06, |
| "loss": 0.4613, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.6247877758913413, |
| "grad_norm": 0.9142153731395757, |
| "learning_rate": 3.796782080920659e-06, |
| "loss": 0.4489, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.6281833616298812, |
| "grad_norm": 0.9197655133200423, |
| "learning_rate": 3.7393345377214584e-06, |
| "loss": 0.4433, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.631578947368421, |
| "grad_norm": 0.9569197839619695, |
| "learning_rate": 3.682064169654663e-06, |
| "loss": 0.4665, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.634974533106961, |
| "grad_norm": 0.9362494462879061, |
| "learning_rate": 3.6249790255526916e-06, |
| "loss": 0.4363, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.6383701188455009, |
| "grad_norm": 0.8671267136185908, |
| "learning_rate": 3.568087128216414e-06, |
| "loss": 0.4646, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.6417657045840407, |
| "grad_norm": 0.9203019838984255, |
| "learning_rate": 3.511396473287611e-06, |
| "loss": 0.4571, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.6451612903225806, |
| "grad_norm": 0.9768723974984408, |
| "learning_rate": 3.4549150281252635e-06, |
| "loss": 0.4484, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.6485568760611206, |
| "grad_norm": 0.9396504959111149, |
| "learning_rate": 3.398650730685813e-06, |
| "loss": 0.457, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.6519524617996605, |
| "grad_norm": 0.9336335579059931, |
| "learning_rate": 3.3426114884075488e-06, |
| "loss": 0.4467, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.6553480475382003, |
| "grad_norm": 0.8976643217426382, |
| "learning_rate": 3.2868051770992935e-06, |
| "loss": 0.4572, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.6587436332767402, |
| "grad_norm": 0.9373033603378723, |
| "learning_rate": 3.2312396398335312e-06, |
| "loss": 0.4687, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.6621392190152802, |
| "grad_norm": 0.9707294952319975, |
| "learning_rate": 3.175922685844125e-06, |
| "loss": 0.4336, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.66553480475382, |
| "grad_norm": 0.9583493306758707, |
| "learning_rate": 3.1208620894288105e-06, |
| "loss": 0.4486, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.6689303904923599, |
| "grad_norm": 0.9820021819293268, |
| "learning_rate": 3.0660655888565827e-06, |
| "loss": 0.4588, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.6723259762308998, |
| "grad_norm": 0.9400174461623624, |
| "learning_rate": 3.0115408852801535e-06, |
| "loss": 0.4423, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.6757215619694398, |
| "grad_norm": 0.9603934801371697, |
| "learning_rate": 2.9572956416536267e-06, |
| "loss": 0.444, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.6791171477079796, |
| "grad_norm": 0.9537870436181752, |
| "learning_rate": 2.9033374816555338e-06, |
| "loss": 0.4556, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.6825127334465195, |
| "grad_norm": 0.9149135094841083, |
| "learning_rate": 2.8496739886173994e-06, |
| "loss": 0.4727, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.6859083191850595, |
| "grad_norm": 0.9486316397589806, |
| "learning_rate": 2.7963127044579697e-06, |
| "loss": 0.432, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.6893039049235993, |
| "grad_norm": 0.9240748880967119, |
| "learning_rate": 2.743261128623269e-06, |
| "loss": 0.4596, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.6926994906621392, |
| "grad_norm": 0.9573847569611018, |
| "learning_rate": 2.6905267170326143e-06, |
| "loss": 0.4679, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.6960950764006791, |
| "grad_norm": 0.8845992056166951, |
| "learning_rate": 2.6381168810307536e-06, |
| "loss": 0.4631, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.6994906621392191, |
| "grad_norm": 1.004797015683845, |
| "learning_rate": 2.5860389863462765e-06, |
| "loss": 0.4475, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.7028862478777589, |
| "grad_norm": 0.9408081397458115, |
| "learning_rate": 2.534300352056416e-06, |
| "loss": 0.4529, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.7062818336162988, |
| "grad_norm": 0.8830371296945094, |
| "learning_rate": 2.4829082495584244e-06, |
| "loss": 0.4431, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.7096774193548387, |
| "grad_norm": 0.918591969840702, |
| "learning_rate": 2.4318699015476495e-06, |
| "loss": 0.4295, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.7130730050933786, |
| "grad_norm": 0.9126233293580099, |
| "learning_rate": 2.3811924810024385e-06, |
| "loss": 0.4462, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.7164685908319185, |
| "grad_norm": 0.944049132172904, |
| "learning_rate": 2.330883110176049e-06, |
| "loss": 0.4764, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.7198641765704584, |
| "grad_norm": 0.8594916774059891, |
| "learning_rate": 2.2809488595956746e-06, |
| "loss": 0.4245, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.7232597623089984, |
| "grad_norm": 0.9004237430035651, |
| "learning_rate": 2.2313967470687593e-06, |
| "loss": 0.4618, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.7266553480475382, |
| "grad_norm": 0.9087973453481795, |
| "learning_rate": 2.18223373669669e-06, |
| "loss": 0.4614, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.7300509337860781, |
| "grad_norm": 0.8919978483526617, |
| "learning_rate": 2.1334667378960642e-06, |
| "loss": 0.462, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.733446519524618, |
| "grad_norm": 0.9099211098465206, |
| "learning_rate": 2.0851026044276405e-06, |
| "loss": 0.4634, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.7368421052631579, |
| "grad_norm": 0.8896204484265512, |
| "learning_rate": 2.0371481334330913e-06, |
| "loss": 0.463, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.7402376910016978, |
| "grad_norm": 0.9519257493559129, |
| "learning_rate": 1.9896100644797316e-06, |
| "loss": 0.4788, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.7436332767402377, |
| "grad_norm": 0.9116718849117885, |
| "learning_rate": 1.9424950786133414e-06, |
| "loss": 0.4634, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.7470288624787776, |
| "grad_norm": 0.9263410885293661, |
| "learning_rate": 1.8958097974191909e-06, |
| "loss": 0.484, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.7504244482173175, |
| "grad_norm": 0.9023645875611066, |
| "learning_rate": 1.8495607820914451e-06, |
| "loss": 0.4442, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.7538200339558574, |
| "grad_norm": 0.8875288134755263, |
| "learning_rate": 1.8037545325110506e-06, |
| "loss": 0.4241, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.7572156196943973, |
| "grad_norm": 0.9432820849930534, |
| "learning_rate": 1.7583974863322272e-06, |
| "loss": 0.4837, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.7606112054329371, |
| "grad_norm": 0.9658199086993665, |
| "learning_rate": 1.7134960180777171e-06, |
| "loss": 0.4728, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.7640067911714771, |
| "grad_norm": 0.8457866400541412, |
| "learning_rate": 1.6690564382429104e-06, |
| "loss": 0.4301, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.767402376910017, |
| "grad_norm": 0.8989697214841547, |
| "learning_rate": 1.6250849924089485e-06, |
| "loss": 0.4527, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.7707979626485568, |
| "grad_norm": 0.9022700419018949, |
| "learning_rate": 1.581587860364977e-06, |
| "loss": 0.4543, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.7741935483870968, |
| "grad_norm": 0.9173241233268217, |
| "learning_rate": 1.5385711552396227e-06, |
| "loss": 0.4466, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.7775891341256367, |
| "grad_norm": 0.8617337826425147, |
| "learning_rate": 1.4960409226418576e-06, |
| "loss": 0.4386, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.7809847198641766, |
| "grad_norm": 0.8625616538339923, |
| "learning_rate": 1.4540031398113335e-06, |
| "loss": 0.4533, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.7843803056027164, |
| "grad_norm": 0.894048701617508, |
| "learning_rate": 1.4124637147783431e-06, |
| "loss": 0.4447, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.7877758913412564, |
| "grad_norm": 0.9208844935282854, |
| "learning_rate": 1.371428485533498e-06, |
| "loss": 0.4826, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.7911714770797963, |
| "grad_norm": 0.8883039726822113, |
| "learning_rate": 1.3309032192072463e-06, |
| "loss": 0.4486, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.7945670628183361, |
| "grad_norm": 0.899703690304073, |
| "learning_rate": 1.29089361125936e-06, |
| "loss": 0.4396, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.797962648556876, |
| "grad_norm": 0.942126628969688, |
| "learning_rate": 1.251405284678488e-06, |
| "loss": 0.4666, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.801358234295416, |
| "grad_norm": 0.876574882692073, |
| "learning_rate": 1.2124437891918995e-06, |
| "loss": 0.4338, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.8047538200339559, |
| "grad_norm": 0.8738541998828565, |
| "learning_rate": 1.1740146004855141e-06, |
| "loss": 0.4609, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.8081494057724957, |
| "grad_norm": 0.8732256896169818, |
| "learning_rate": 1.1361231194343436e-06, |
| "loss": 0.4505, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.8115449915110357, |
| "grad_norm": 0.8634374293174691, |
| "learning_rate": 1.0987746713434578e-06, |
| "loss": 0.4626, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.8149405772495756, |
| "grad_norm": 0.8709360064531477, |
| "learning_rate": 1.0619745051995473e-06, |
| "loss": 0.465, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.8183361629881154, |
| "grad_norm": 0.8612129831209381, |
| "learning_rate": 1.0257277929332332e-06, |
| "loss": 0.4543, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.8217317487266553, |
| "grad_norm": 0.9089176413771565, |
| "learning_rate": 9.900396286922025e-07, |
| "loss": 0.4676, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.8251273344651953, |
| "grad_norm": 0.8856710278778406, |
| "learning_rate": 9.549150281252633e-07, |
| "loss": 0.4366, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.8285229202037352, |
| "grad_norm": 0.8188623584178454, |
| "learning_rate": 9.203589276774438e-07, |
| "loss": 0.4404, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.831918505942275, |
| "grad_norm": 0.9252917982729971, |
| "learning_rate": 8.86376183896226e-07, |
| "loss": 0.4447, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.8353140916808149, |
| "grad_norm": 0.8784423429363009, |
| "learning_rate": 8.529715727489912e-07, |
| "loss": 0.4488, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.8387096774193549, |
| "grad_norm": 0.9207047553546756, |
| "learning_rate": 8.201497889518073e-07, |
| "loss": 0.4561, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.8421052631578947, |
| "grad_norm": 0.886908726068121, |
| "learning_rate": 7.879154453096305e-07, |
| "loss": 0.4556, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.8455008488964346, |
| "grad_norm": 0.8585705366949494, |
| "learning_rate": 7.562730720680111e-07, |
| "loss": 0.4333, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.8488964346349746, |
| "grad_norm": 0.8504815527006262, |
| "learning_rate": 7.25227116276413e-07, |
| "loss": 0.4512, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.8522920203735145, |
| "grad_norm": 0.8201572218259242, |
| "learning_rate": 6.947819411632223e-07, |
| "loss": 0.4492, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.8556876061120543, |
| "grad_norm": 0.8579475897548395, |
| "learning_rate": 6.649418255225298e-07, |
| "loss": 0.4633, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.8590831918505942, |
| "grad_norm": 0.8140416489339838, |
| "learning_rate": 6.357109631127889e-07, |
| "loss": 0.4418, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.8624787775891342, |
| "grad_norm": 0.8552881051968678, |
| "learning_rate": 6.07093462067419e-07, |
| "loss": 0.4572, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.865874363327674, |
| "grad_norm": 0.8971540169404577, |
| "learning_rate": 5.79093344317449e-07, |
| "loss": 0.4775, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.8692699490662139, |
| "grad_norm": 0.8058254068183305, |
| "learning_rate": 5.517145450262639e-07, |
| "loss": 0.4313, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.8726655348047538, |
| "grad_norm": 0.8530487589073922, |
| "learning_rate": 5.249609120365579e-07, |
| "loss": 0.4397, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.8760611205432938, |
| "grad_norm": 0.9429999718413201, |
| "learning_rate": 4.988362053295564e-07, |
| "loss": 0.4496, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.8794567062818336, |
| "grad_norm": 0.8375356078246441, |
| "learning_rate": 4.733440964965791e-07, |
| "loss": 0.4303, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.8828522920203735, |
| "grad_norm": 0.8256703834683652, |
| "learning_rate": 4.484881682230341e-07, |
| "loss": 0.4268, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.8862478777589134, |
| "grad_norm": 0.8543078561007811, |
| "learning_rate": 4.242719137849077e-07, |
| "loss": 0.4514, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.8896434634974533, |
| "grad_norm": 0.905414358204414, |
| "learning_rate": 4.00698736557808e-07, |
| "loss": 0.4475, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.8930390492359932, |
| "grad_norm": 0.8479836834072867, |
| "learning_rate": 3.777719495386567e-07, |
| "loss": 0.4479, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.8964346349745331, |
| "grad_norm": 0.8646327130622585, |
| "learning_rate": 3.5549477488007853e-07, |
| "loss": 0.4563, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.8998302207130731, |
| "grad_norm": 0.8363483102479007, |
| "learning_rate": 3.3387034343755063e-07, |
| "loss": 0.4447, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.9032258064516129, |
| "grad_norm": 0.8772923486186023, |
| "learning_rate": 3.1290169432939556e-07, |
| "loss": 0.4416, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.9066213921901528, |
| "grad_norm": 0.8320996733730202, |
| "learning_rate": 2.925917745096568e-07, |
| "loss": 0.4263, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.9100169779286927, |
| "grad_norm": 0.8735391402072182, |
| "learning_rate": 2.7294343835393366e-07, |
| "loss": 0.4635, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.9134125636672326, |
| "grad_norm": 0.8476560390812548, |
| "learning_rate": 2.539594472582213e-07, |
| "loss": 0.4387, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.9168081494057725, |
| "grad_norm": 0.8265472830061668, |
| "learning_rate": 2.3564246925082358e-07, |
| "loss": 0.4409, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.9202037351443124, |
| "grad_norm": 0.8275085794146296, |
| "learning_rate": 2.179950786173879e-07, |
| "loss": 0.422, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.9235993208828522, |
| "grad_norm": 0.8378368477541204, |
| "learning_rate": 2.01019755539108e-07, |
| "loss": 0.4533, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.9269949066213922, |
| "grad_norm": 0.8225280576542648, |
| "learning_rate": 1.8471888574415953e-07, |
| "loss": 0.4504, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.9303904923599321, |
| "grad_norm": 0.8689138287376853, |
| "learning_rate": 1.690947601724091e-07, |
| "loss": 0.4616, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.933786078098472, |
| "grad_norm": 0.8153963652578823, |
| "learning_rate": 1.5414957465343883e-07, |
| "loss": 0.4362, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.9371816638370118, |
| "grad_norm": 0.7970051219455544, |
| "learning_rate": 1.3988542959794627e-07, |
| "loss": 0.4476, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.9405772495755518, |
| "grad_norm": 0.8286263852975556, |
| "learning_rate": 1.2630432970255014e-07, |
| "loss": 0.4317, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.9439728353140917, |
| "grad_norm": 0.819703548732724, |
| "learning_rate": 1.1340818366804728e-07, |
| "loss": 0.4374, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.9473684210526315, |
| "grad_norm": 0.8630335285550543, |
| "learning_rate": 1.0119880393116177e-07, |
| "loss": 0.4463, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.9507640067911715, |
| "grad_norm": 0.8905416016998572, |
| "learning_rate": 8.967790640982466e-08, |
| "loss": 0.4534, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.9541595925297114, |
| "grad_norm": 0.869553726844365, |
| "learning_rate": 7.884711026201586e-08, |
| "loss": 0.4474, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.9575551782682513, |
| "grad_norm": 0.8516101851144577, |
| "learning_rate": 6.870793765820783e-08, |
| "loss": 0.4735, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.9609507640067911, |
| "grad_norm": 0.834166916764426, |
| "learning_rate": 5.92618135674361e-08, |
| "loss": 0.4506, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.9643463497453311, |
| "grad_norm": 0.8500281237762543, |
| "learning_rate": 5.0510065557034526e-08, |
| "loss": 0.4677, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.967741935483871, |
| "grad_norm": 0.833367943844974, |
| "learning_rate": 4.245392360605727e-08, |
| "loss": 0.4582, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.9711375212224108, |
| "grad_norm": 0.8458107328647104, |
| "learning_rate": 3.5094519932415417e-08, |
| "loss": 0.4505, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.9745331069609507, |
| "grad_norm": 0.8443895579435003, |
| "learning_rate": 2.843288883375539e-08, |
| "loss": 0.4538, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.9779286926994907, |
| "grad_norm": 0.8631958241337926, |
| "learning_rate": 2.2469966542096323e-08, |
| "loss": 0.4432, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.9813242784380306, |
| "grad_norm": 0.886260366562854, |
| "learning_rate": 1.7206591092253642e-08, |
| "loss": 0.4567, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.9847198641765704, |
| "grad_norm": 0.7994144470255627, |
| "learning_rate": 1.264350220405719e-08, |
| "loss": 0.4317, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.9881154499151104, |
| "grad_norm": 0.8700681537853684, |
| "learning_rate": 8.781341178393244e-09, |
| "loss": 0.4555, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.9915110356536503, |
| "grad_norm": 0.8575898812488901, |
| "learning_rate": 5.620650807073857e-09, |
| "loss": 0.4606, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.9949066213921901, |
| "grad_norm": 0.8616135380601608, |
| "learning_rate": 3.1618752965534295e-09, |
| "loss": 0.4408, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.99830220713073, |
| "grad_norm": 0.7784658740487929, |
| "learning_rate": 1.4053602054991954e-09, |
| "loss": 0.4157, |
| "step": 294 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 1.2063313976605035, |
| "learning_rate": 3.513523962256349e-10, |
| "loss": 0.4174, |
| "step": 295 |
| }, |
| { |
| "epoch": 1.0, |
| "step": 295, |
| "total_flos": 44242544295936.0, |
| "train_loss": 0.5097209928399425, |
| "train_runtime": 1256.4644, |
| "train_samples_per_second": 14.995, |
| "train_steps_per_second": 0.235 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 295, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 2500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 44242544295936.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
