{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 1960, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005102040816326531, "grad_norm": 180.20428466796875, "learning_rate": 9.994897959183675e-06, "loss": 2.25, "step": 1 }, { "epoch": 0.0010204081632653062, "grad_norm": 149.1138153076172, "learning_rate": 9.989795918367348e-06, "loss": 1.4375, "step": 2 }, { "epoch": 0.001530612244897959, "grad_norm": 106.21714782714844, "learning_rate": 9.984693877551021e-06, "loss": 0.8242, "step": 3 }, { "epoch": 0.0020408163265306124, "grad_norm": 67.01580047607422, "learning_rate": 9.979591836734694e-06, "loss": 1.0645, "step": 4 }, { "epoch": 0.002551020408163265, "grad_norm": 174.26470947265625, "learning_rate": 9.974489795918369e-06, "loss": 1.5879, "step": 5 }, { "epoch": 0.003061224489795918, "grad_norm": 102.49396514892578, "learning_rate": 9.969387755102042e-06, "loss": 0.7559, "step": 6 }, { "epoch": 0.0035714285714285713, "grad_norm": 186.5989227294922, "learning_rate": 9.964285714285714e-06, "loss": 1.1484, "step": 7 }, { "epoch": 0.004081632653061225, "grad_norm": 99.42011260986328, "learning_rate": 9.959183673469387e-06, "loss": 0.7744, "step": 8 }, { "epoch": 0.004591836734693878, "grad_norm": 74.93321228027344, "learning_rate": 9.954081632653062e-06, "loss": 0.9199, "step": 9 }, { "epoch": 0.00510204081632653, "grad_norm": 24.61996841430664, "learning_rate": 9.948979591836737e-06, "loss": 0.5117, "step": 10 }, { "epoch": 0.005612244897959183, "grad_norm": 89.22261047363281, "learning_rate": 9.94387755102041e-06, "loss": 0.5977, "step": 11 }, { "epoch": 0.006122448979591836, "grad_norm": 77.13282775878906, "learning_rate": 9.938775510204082e-06, "loss": 0.8242, "step": 12 }, { "epoch": 0.0066326530612244895, "grad_norm": 35.671268463134766, "learning_rate": 9.933673469387755e-06, "loss": 0.6543, "step": 13 }, { "epoch": 0.007142857142857143, "grad_norm": 40.284549713134766, "learning_rate": 9.92857142857143e-06, "loss": 0.7598, "step": 14 }, { "epoch": 0.007653061224489796, "grad_norm": 37.70195770263672, "learning_rate": 9.923469387755103e-06, "loss": 0.4873, "step": 15 }, { "epoch": 0.00816326530612245, "grad_norm": 75.62923431396484, "learning_rate": 9.918367346938776e-06, "loss": 0.9355, "step": 16 }, { "epoch": 0.008673469387755102, "grad_norm": 27.662946701049805, "learning_rate": 9.913265306122449e-06, "loss": 0.5928, "step": 17 }, { "epoch": 0.009183673469387756, "grad_norm": 29.03522491455078, "learning_rate": 9.908163265306123e-06, "loss": 0.832, "step": 18 }, { "epoch": 0.009693877551020408, "grad_norm": 21.77155876159668, "learning_rate": 9.903061224489798e-06, "loss": 0.6777, "step": 19 }, { "epoch": 0.01020408163265306, "grad_norm": 28.85708999633789, "learning_rate": 9.89795918367347e-06, "loss": 0.6582, "step": 20 }, { "epoch": 0.010714285714285714, "grad_norm": 33.97568893432617, "learning_rate": 9.892857142857143e-06, "loss": 0.5811, "step": 21 }, { "epoch": 0.011224489795918367, "grad_norm": 32.343624114990234, "learning_rate": 9.887755102040816e-06, "loss": 0.4805, "step": 22 }, { "epoch": 0.01173469387755102, "grad_norm": 25.349021911621094, "learning_rate": 9.882653061224491e-06, "loss": 0.5635, "step": 23 }, { "epoch": 0.012244897959183673, "grad_norm": 27.375892639160156, "learning_rate": 9.877551020408164e-06, "loss": 0.5742, "step": 24 }, { "epoch": 0.012755102040816327, "grad_norm": 25.68079376220703, "learning_rate": 9.872448979591838e-06, "loss": 0.6357, "step": 25 }, { "epoch": 0.013265306122448979, "grad_norm": 28.40098762512207, "learning_rate": 9.867346938775511e-06, "loss": 0.7676, "step": 26 }, { "epoch": 0.013775510204081633, "grad_norm": 36.143836975097656, "learning_rate": 9.862244897959184e-06, "loss": 0.4937, "step": 27 }, { "epoch": 0.014285714285714285, "grad_norm": 28.520051956176758, "learning_rate": 9.857142857142859e-06, "loss": 0.2034, "step": 28 }, { "epoch": 0.01479591836734694, "grad_norm": 23.373916625976562, "learning_rate": 9.852040816326532e-06, "loss": 0.4355, "step": 29 }, { "epoch": 0.015306122448979591, "grad_norm": 34.331886291503906, "learning_rate": 9.846938775510205e-06, "loss": 0.7305, "step": 30 }, { "epoch": 0.015816326530612244, "grad_norm": 59.27937698364258, "learning_rate": 9.841836734693878e-06, "loss": 0.8945, "step": 31 }, { "epoch": 0.0163265306122449, "grad_norm": 47.941097259521484, "learning_rate": 9.836734693877552e-06, "loss": 0.8848, "step": 32 }, { "epoch": 0.01683673469387755, "grad_norm": 43.998870849609375, "learning_rate": 9.831632653061225e-06, "loss": 0.7773, "step": 33 }, { "epoch": 0.017346938775510204, "grad_norm": 36.20676803588867, "learning_rate": 9.8265306122449e-06, "loss": 0.4961, "step": 34 }, { "epoch": 0.017857142857142856, "grad_norm": 34.36786651611328, "learning_rate": 9.821428571428573e-06, "loss": 0.6562, "step": 35 }, { "epoch": 0.018367346938775512, "grad_norm": 54.18436050415039, "learning_rate": 9.816326530612245e-06, "loss": 1.168, "step": 36 }, { "epoch": 0.018877551020408164, "grad_norm": 25.10146713256836, "learning_rate": 9.81122448979592e-06, "loss": 0.3267, "step": 37 }, { "epoch": 0.019387755102040816, "grad_norm": 32.08784103393555, "learning_rate": 9.806122448979593e-06, "loss": 0.7422, "step": 38 }, { "epoch": 0.01989795918367347, "grad_norm": 44.92784881591797, "learning_rate": 9.801020408163266e-06, "loss": 0.6052, "step": 39 }, { "epoch": 0.02040816326530612, "grad_norm": 30.02330780029297, "learning_rate": 9.795918367346939e-06, "loss": 0.6367, "step": 40 }, { "epoch": 0.020918367346938777, "grad_norm": 26.374889373779297, "learning_rate": 9.790816326530613e-06, "loss": 0.6016, "step": 41 }, { "epoch": 0.02142857142857143, "grad_norm": 43.53873062133789, "learning_rate": 9.785714285714286e-06, "loss": 0.793, "step": 42 }, { "epoch": 0.02193877551020408, "grad_norm": 34.2724609375, "learning_rate": 9.78061224489796e-06, "loss": 0.4814, "step": 43 }, { "epoch": 0.022448979591836733, "grad_norm": 34.583919525146484, "learning_rate": 9.775510204081634e-06, "loss": 0.6797, "step": 44 }, { "epoch": 0.02295918367346939, "grad_norm": 18.27403450012207, "learning_rate": 9.770408163265307e-06, "loss": 0.2637, "step": 45 }, { "epoch": 0.02346938775510204, "grad_norm": 20.295339584350586, "learning_rate": 9.765306122448981e-06, "loss": 0.3481, "step": 46 }, { "epoch": 0.023979591836734693, "grad_norm": 40.48591995239258, "learning_rate": 9.760204081632654e-06, "loss": 0.7305, "step": 47 }, { "epoch": 0.024489795918367346, "grad_norm": 24.177547454833984, "learning_rate": 9.755102040816327e-06, "loss": 0.3242, "step": 48 }, { "epoch": 0.025, "grad_norm": 28.06923484802246, "learning_rate": 9.75e-06, "loss": 0.4741, "step": 49 }, { "epoch": 0.025510204081632654, "grad_norm": 33.15027618408203, "learning_rate": 9.744897959183674e-06, "loss": 0.8574, "step": 50 }, { "epoch": 0.026020408163265306, "grad_norm": 37.48566436767578, "learning_rate": 9.739795918367347e-06, "loss": 0.3125, "step": 51 }, { "epoch": 0.026530612244897958, "grad_norm": 30.875812530517578, "learning_rate": 9.734693877551022e-06, "loss": 0.5361, "step": 52 }, { "epoch": 0.027040816326530614, "grad_norm": 33.03073501586914, "learning_rate": 9.729591836734695e-06, "loss": 0.7598, "step": 53 }, { "epoch": 0.027551020408163266, "grad_norm": 18.41679573059082, "learning_rate": 9.724489795918368e-06, "loss": 0.6582, "step": 54 }, { "epoch": 0.02806122448979592, "grad_norm": 38.92633819580078, "learning_rate": 9.719387755102042e-06, "loss": 0.9336, "step": 55 }, { "epoch": 0.02857142857142857, "grad_norm": 40.54533767700195, "learning_rate": 9.714285714285715e-06, "loss": 0.7236, "step": 56 }, { "epoch": 0.029081632653061223, "grad_norm": 56.868778228759766, "learning_rate": 9.70918367346939e-06, "loss": 1.0371, "step": 57 }, { "epoch": 0.02959183673469388, "grad_norm": 23.465797424316406, "learning_rate": 9.704081632653061e-06, "loss": 0.3408, "step": 58 }, { "epoch": 0.03010204081632653, "grad_norm": 22.835206985473633, "learning_rate": 9.698979591836736e-06, "loss": 0.4004, "step": 59 }, { "epoch": 0.030612244897959183, "grad_norm": 46.72260284423828, "learning_rate": 9.693877551020408e-06, "loss": 0.8105, "step": 60 }, { "epoch": 0.031122448979591835, "grad_norm": 24.229944229125977, "learning_rate": 9.688775510204083e-06, "loss": 0.2803, "step": 61 }, { "epoch": 0.03163265306122449, "grad_norm": 18.874591827392578, "learning_rate": 9.683673469387756e-06, "loss": 0.2314, "step": 62 }, { "epoch": 0.03214285714285714, "grad_norm": 34.32584762573242, "learning_rate": 9.678571428571429e-06, "loss": 0.5273, "step": 63 }, { "epoch": 0.0326530612244898, "grad_norm": 25.608600616455078, "learning_rate": 9.673469387755103e-06, "loss": 0.3853, "step": 64 }, { "epoch": 0.03316326530612245, "grad_norm": 36.641815185546875, "learning_rate": 9.668367346938776e-06, "loss": 0.6602, "step": 65 }, { "epoch": 0.0336734693877551, "grad_norm": 26.1795654296875, "learning_rate": 9.663265306122451e-06, "loss": 0.3999, "step": 66 }, { "epoch": 0.034183673469387756, "grad_norm": 22.43852424621582, "learning_rate": 9.658163265306124e-06, "loss": 0.5112, "step": 67 }, { "epoch": 0.03469387755102041, "grad_norm": 42.580074310302734, "learning_rate": 9.653061224489797e-06, "loss": 0.9629, "step": 68 }, { "epoch": 0.03520408163265306, "grad_norm": 17.4409236907959, "learning_rate": 9.64795918367347e-06, "loss": 0.2938, "step": 69 }, { "epoch": 0.03571428571428571, "grad_norm": 60.15330123901367, "learning_rate": 9.642857142857144e-06, "loss": 0.7676, "step": 70 }, { "epoch": 0.036224489795918365, "grad_norm": 42.13832473754883, "learning_rate": 9.637755102040817e-06, "loss": 0.6738, "step": 71 }, { "epoch": 0.036734693877551024, "grad_norm": 23.257869720458984, "learning_rate": 9.63265306122449e-06, "loss": 0.2612, "step": 72 }, { "epoch": 0.037244897959183676, "grad_norm": 21.014530181884766, "learning_rate": 9.627551020408165e-06, "loss": 0.3354, "step": 73 }, { "epoch": 0.03775510204081633, "grad_norm": 31.90937042236328, "learning_rate": 9.622448979591837e-06, "loss": 0.374, "step": 74 }, { "epoch": 0.03826530612244898, "grad_norm": 31.11526107788086, "learning_rate": 9.617346938775512e-06, "loss": 0.4902, "step": 75 }, { "epoch": 0.03877551020408163, "grad_norm": 52.741756439208984, "learning_rate": 9.612244897959185e-06, "loss": 0.8027, "step": 76 }, { "epoch": 0.039285714285714285, "grad_norm": 36.732784271240234, "learning_rate": 9.607142857142858e-06, "loss": 0.5645, "step": 77 }, { "epoch": 0.03979591836734694, "grad_norm": 29.65674591064453, "learning_rate": 9.60204081632653e-06, "loss": 0.5674, "step": 78 }, { "epoch": 0.04030612244897959, "grad_norm": 26.993762969970703, "learning_rate": 9.596938775510205e-06, "loss": 0.5596, "step": 79 }, { "epoch": 0.04081632653061224, "grad_norm": 21.767559051513672, "learning_rate": 9.591836734693878e-06, "loss": 0.3237, "step": 80 }, { "epoch": 0.0413265306122449, "grad_norm": 38.7492790222168, "learning_rate": 9.586734693877551e-06, "loss": 0.3857, "step": 81 }, { "epoch": 0.04183673469387755, "grad_norm": 49.705142974853516, "learning_rate": 9.581632653061226e-06, "loss": 0.8877, "step": 82 }, { "epoch": 0.042346938775510205, "grad_norm": 42.470428466796875, "learning_rate": 9.576530612244899e-06, "loss": 0.7812, "step": 83 }, { "epoch": 0.04285714285714286, "grad_norm": 52.263572692871094, "learning_rate": 9.571428571428573e-06, "loss": 0.6875, "step": 84 }, { "epoch": 0.04336734693877551, "grad_norm": 54.38450622558594, "learning_rate": 9.566326530612246e-06, "loss": 0.9053, "step": 85 }, { "epoch": 0.04387755102040816, "grad_norm": 41.437015533447266, "learning_rate": 9.561224489795919e-06, "loss": 0.6113, "step": 86 }, { "epoch": 0.044387755102040814, "grad_norm": 43.69349670410156, "learning_rate": 9.556122448979592e-06, "loss": 0.4023, "step": 87 }, { "epoch": 0.044897959183673466, "grad_norm": 28.58114242553711, "learning_rate": 9.551020408163266e-06, "loss": 0.3247, "step": 88 }, { "epoch": 0.045408163265306126, "grad_norm": 76.00294494628906, "learning_rate": 9.54591836734694e-06, "loss": 1.002, "step": 89 }, { "epoch": 0.04591836734693878, "grad_norm": 37.57072830200195, "learning_rate": 9.540816326530612e-06, "loss": 0.4824, "step": 90 }, { "epoch": 0.04642857142857143, "grad_norm": 70.09291076660156, "learning_rate": 9.535714285714287e-06, "loss": 0.791, "step": 91 }, { "epoch": 0.04693877551020408, "grad_norm": 38.46149826049805, "learning_rate": 9.53061224489796e-06, "loss": 0.3252, "step": 92 }, { "epoch": 0.047448979591836735, "grad_norm": 24.384801864624023, "learning_rate": 9.525510204081634e-06, "loss": 0.1816, "step": 93 }, { "epoch": 0.04795918367346939, "grad_norm": 46.24349594116211, "learning_rate": 9.520408163265307e-06, "loss": 0.5439, "step": 94 }, { "epoch": 0.04846938775510204, "grad_norm": 33.58824157714844, "learning_rate": 9.51530612244898e-06, "loss": 0.3662, "step": 95 }, { "epoch": 0.04897959183673469, "grad_norm": 37.34817886352539, "learning_rate": 9.510204081632653e-06, "loss": 0.4263, "step": 96 }, { "epoch": 0.049489795918367344, "grad_norm": 38.73717498779297, "learning_rate": 9.505102040816328e-06, "loss": 0.3535, "step": 97 }, { "epoch": 0.05, "grad_norm": 42.282875061035156, "learning_rate": 9.5e-06, "loss": 0.7773, "step": 98 }, { "epoch": 0.050510204081632655, "grad_norm": 56.41849899291992, "learning_rate": 9.494897959183675e-06, "loss": 0.7148, "step": 99 }, { "epoch": 0.05102040816326531, "grad_norm": 50.924407958984375, "learning_rate": 9.489795918367348e-06, "loss": 0.9141, "step": 100 }, { "epoch": 0.05153061224489796, "grad_norm": 25.65245819091797, "learning_rate": 9.484693877551021e-06, "loss": 0.2117, "step": 101 }, { "epoch": 0.05204081632653061, "grad_norm": 53.109500885009766, "learning_rate": 9.479591836734695e-06, "loss": 0.4648, "step": 102 }, { "epoch": 0.052551020408163264, "grad_norm": 20.992956161499023, "learning_rate": 9.474489795918368e-06, "loss": 0.0345, "step": 103 }, { "epoch": 0.053061224489795916, "grad_norm": 56.162376403808594, "learning_rate": 9.469387755102041e-06, "loss": 0.1924, "step": 104 }, { "epoch": 0.05357142857142857, "grad_norm": 73.63543701171875, "learning_rate": 9.464285714285714e-06, "loss": 0.9707, "step": 105 }, { "epoch": 0.05408163265306123, "grad_norm": 83.5208740234375, "learning_rate": 9.459183673469389e-06, "loss": 0.6406, "step": 106 }, { "epoch": 0.05459183673469388, "grad_norm": 53.107818603515625, "learning_rate": 9.454081632653062e-06, "loss": 0.1807, "step": 107 }, { "epoch": 0.05510204081632653, "grad_norm": 94.28384399414062, "learning_rate": 9.448979591836736e-06, "loss": 1.3633, "step": 108 }, { "epoch": 0.055612244897959184, "grad_norm": 58.84535217285156, "learning_rate": 9.44387755102041e-06, "loss": 0.4478, "step": 109 }, { "epoch": 0.05612244897959184, "grad_norm": 43.12350082397461, "learning_rate": 9.438775510204082e-06, "loss": 0.1705, "step": 110 }, { "epoch": 0.05663265306122449, "grad_norm": 51.086463928222656, "learning_rate": 9.433673469387757e-06, "loss": 0.3604, "step": 111 }, { "epoch": 0.05714285714285714, "grad_norm": 39.57880783081055, "learning_rate": 9.42857142857143e-06, "loss": 0.2487, "step": 112 }, { "epoch": 0.05765306122448979, "grad_norm": 57.11735153198242, "learning_rate": 9.423469387755102e-06, "loss": 0.4033, "step": 113 }, { "epoch": 0.058163265306122446, "grad_norm": 80.10859680175781, "learning_rate": 9.418367346938775e-06, "loss": 0.5635, "step": 114 }, { "epoch": 0.058673469387755105, "grad_norm": 85.74658203125, "learning_rate": 9.41326530612245e-06, "loss": 1.1875, "step": 115 }, { "epoch": 0.05918367346938776, "grad_norm": 88.73854064941406, "learning_rate": 9.408163265306123e-06, "loss": 0.4883, "step": 116 }, { "epoch": 0.05969387755102041, "grad_norm": 25.52219009399414, "learning_rate": 9.403061224489797e-06, "loss": 0.4316, "step": 117 }, { "epoch": 0.06020408163265306, "grad_norm": 47.028541564941406, "learning_rate": 9.39795918367347e-06, "loss": 0.3401, "step": 118 }, { "epoch": 0.060714285714285714, "grad_norm": 65.34862518310547, "learning_rate": 9.392857142857143e-06, "loss": 0.2803, "step": 119 }, { "epoch": 0.061224489795918366, "grad_norm": 50.170082092285156, "learning_rate": 9.387755102040818e-06, "loss": 0.8887, "step": 120 }, { "epoch": 0.06173469387755102, "grad_norm": 64.69440460205078, "learning_rate": 9.38265306122449e-06, "loss": 0.6387, "step": 121 }, { "epoch": 0.06224489795918367, "grad_norm": 28.455402374267578, "learning_rate": 9.377551020408164e-06, "loss": 0.3779, "step": 122 }, { "epoch": 0.06275510204081633, "grad_norm": 56.851295471191406, "learning_rate": 9.372448979591836e-06, "loss": 0.458, "step": 123 }, { "epoch": 0.06326530612244897, "grad_norm": 37.103233337402344, "learning_rate": 9.367346938775511e-06, "loss": 0.3921, "step": 124 }, { "epoch": 0.06377551020408163, "grad_norm": 38.03989791870117, "learning_rate": 9.362244897959184e-06, "loss": 0.3174, "step": 125 }, { "epoch": 0.06428571428571428, "grad_norm": 29.40471076965332, "learning_rate": 9.357142857142859e-06, "loss": 0.3262, "step": 126 }, { "epoch": 0.06479591836734694, "grad_norm": 67.49293518066406, "learning_rate": 9.352040816326531e-06, "loss": 1.0171, "step": 127 }, { "epoch": 0.0653061224489796, "grad_norm": 28.81426429748535, "learning_rate": 9.346938775510204e-06, "loss": 0.3369, "step": 128 }, { "epoch": 0.06581632653061224, "grad_norm": 39.364646911621094, "learning_rate": 9.341836734693879e-06, "loss": 0.6484, "step": 129 }, { "epoch": 0.0663265306122449, "grad_norm": 35.88399887084961, "learning_rate": 9.336734693877552e-06, "loss": 0.5918, "step": 130 }, { "epoch": 0.06683673469387755, "grad_norm": 50.083805084228516, "learning_rate": 9.331632653061225e-06, "loss": 0.5059, "step": 131 }, { "epoch": 0.0673469387755102, "grad_norm": 15.41246223449707, "learning_rate": 9.326530612244898e-06, "loss": 0.1045, "step": 132 }, { "epoch": 0.06785714285714285, "grad_norm": 34.074527740478516, "learning_rate": 9.321428571428572e-06, "loss": 0.417, "step": 133 }, { "epoch": 0.06836734693877551, "grad_norm": 126.97840881347656, "learning_rate": 9.316326530612245e-06, "loss": 1.0015, "step": 134 }, { "epoch": 0.06887755102040816, "grad_norm": 38.16463851928711, "learning_rate": 9.31122448979592e-06, "loss": 0.5752, "step": 135 }, { "epoch": 0.06938775510204082, "grad_norm": 42.31215286254883, "learning_rate": 9.306122448979593e-06, "loss": 0.1948, "step": 136 }, { "epoch": 0.06989795918367347, "grad_norm": 66.4386215209961, "learning_rate": 9.301020408163265e-06, "loss": 0.5332, "step": 137 }, { "epoch": 0.07040816326530612, "grad_norm": 43.8679084777832, "learning_rate": 9.29591836734694e-06, "loss": 0.8613, "step": 138 }, { "epoch": 0.07091836734693878, "grad_norm": 46.80387878417969, "learning_rate": 9.290816326530613e-06, "loss": 0.407, "step": 139 }, { "epoch": 0.07142857142857142, "grad_norm": 27.621755599975586, "learning_rate": 9.285714285714288e-06, "loss": 0.1606, "step": 140 }, { "epoch": 0.07193877551020408, "grad_norm": 38.73698425292969, "learning_rate": 9.28061224489796e-06, "loss": 0.6013, "step": 141 }, { "epoch": 0.07244897959183673, "grad_norm": 82.09680938720703, "learning_rate": 9.275510204081633e-06, "loss": 1.4434, "step": 142 }, { "epoch": 0.07295918367346939, "grad_norm": 72.57801818847656, "learning_rate": 9.270408163265306e-06, "loss": 0.4014, "step": 143 }, { "epoch": 0.07346938775510205, "grad_norm": 74.1526107788086, "learning_rate": 9.26530612244898e-06, "loss": 0.624, "step": 144 }, { "epoch": 0.07397959183673469, "grad_norm": 59.2291374206543, "learning_rate": 9.260204081632654e-06, "loss": 0.7891, "step": 145 }, { "epoch": 0.07448979591836735, "grad_norm": 68.28843688964844, "learning_rate": 9.255102040816327e-06, "loss": 1.0869, "step": 146 }, { "epoch": 0.075, "grad_norm": 85.41793823242188, "learning_rate": 9.250000000000001e-06, "loss": 1.4824, "step": 147 }, { "epoch": 0.07551020408163266, "grad_norm": 47.234989166259766, "learning_rate": 9.244897959183674e-06, "loss": 0.3838, "step": 148 }, { "epoch": 0.0760204081632653, "grad_norm": 30.908872604370117, "learning_rate": 9.239795918367349e-06, "loss": 0.2627, "step": 149 }, { "epoch": 0.07653061224489796, "grad_norm": 34.30613708496094, "learning_rate": 9.234693877551022e-06, "loss": 0.3032, "step": 150 }, { "epoch": 0.0770408163265306, "grad_norm": 96.20804595947266, "learning_rate": 9.229591836734695e-06, "loss": 0.6504, "step": 151 }, { "epoch": 0.07755102040816327, "grad_norm": 45.695011138916016, "learning_rate": 9.224489795918367e-06, "loss": 0.3818, "step": 152 }, { "epoch": 0.07806122448979592, "grad_norm": 61.580448150634766, "learning_rate": 9.219387755102042e-06, "loss": 0.9773, "step": 153 }, { "epoch": 0.07857142857142857, "grad_norm": 90.28081512451172, "learning_rate": 9.214285714285715e-06, "loss": 0.8379, "step": 154 }, { "epoch": 0.07908163265306123, "grad_norm": 22.588993072509766, "learning_rate": 9.209183673469388e-06, "loss": 0.1917, "step": 155 }, { "epoch": 0.07959183673469387, "grad_norm": 27.31729507446289, "learning_rate": 9.204081632653062e-06, "loss": 0.3022, "step": 156 }, { "epoch": 0.08010204081632653, "grad_norm": 69.88858795166016, "learning_rate": 9.198979591836735e-06, "loss": 0.8125, "step": 157 }, { "epoch": 0.08061224489795918, "grad_norm": 25.693775177001953, "learning_rate": 9.19387755102041e-06, "loss": 0.1002, "step": 158 }, { "epoch": 0.08112244897959184, "grad_norm": 47.98956298828125, "learning_rate": 9.188775510204083e-06, "loss": 0.4863, "step": 159 }, { "epoch": 0.08163265306122448, "grad_norm": 45.434932708740234, "learning_rate": 9.183673469387756e-06, "loss": 0.5537, "step": 160 }, { "epoch": 0.08214285714285714, "grad_norm": 5.55327844619751, "learning_rate": 9.178571428571429e-06, "loss": 0.0277, "step": 161 }, { "epoch": 0.0826530612244898, "grad_norm": 28.61231803894043, "learning_rate": 9.173469387755103e-06, "loss": 0.5063, "step": 162 }, { "epoch": 0.08316326530612245, "grad_norm": 25.984447479248047, "learning_rate": 9.168367346938776e-06, "loss": 0.209, "step": 163 }, { "epoch": 0.0836734693877551, "grad_norm": 54.507667541503906, "learning_rate": 9.163265306122449e-06, "loss": 0.6504, "step": 164 }, { "epoch": 0.08418367346938775, "grad_norm": 46.38584518432617, "learning_rate": 9.158163265306124e-06, "loss": 0.7793, "step": 165 }, { "epoch": 0.08469387755102041, "grad_norm": 93.0276107788086, "learning_rate": 9.153061224489796e-06, "loss": 0.7607, "step": 166 }, { "epoch": 0.08520408163265306, "grad_norm": 60.44808578491211, "learning_rate": 9.147959183673471e-06, "loss": 1.0938, "step": 167 }, { "epoch": 0.08571428571428572, "grad_norm": 104.53335571289062, "learning_rate": 9.142857142857144e-06, "loss": 0.6191, "step": 168 }, { "epoch": 0.08622448979591837, "grad_norm": 45.57658386230469, "learning_rate": 9.137755102040817e-06, "loss": 0.6094, "step": 169 }, { "epoch": 0.08673469387755102, "grad_norm": 34.81111145019531, "learning_rate": 9.13265306122449e-06, "loss": 0.6953, "step": 170 }, { "epoch": 0.08724489795918368, "grad_norm": 26.484888076782227, "learning_rate": 9.127551020408164e-06, "loss": 0.2764, "step": 171 }, { "epoch": 0.08775510204081632, "grad_norm": 57.10886764526367, "learning_rate": 9.122448979591837e-06, "loss": 0.8979, "step": 172 }, { "epoch": 0.08826530612244898, "grad_norm": 37.437904357910156, "learning_rate": 9.11734693877551e-06, "loss": 0.585, "step": 173 }, { "epoch": 0.08877551020408163, "grad_norm": 61.384769439697266, "learning_rate": 9.112244897959185e-06, "loss": 0.6758, "step": 174 }, { "epoch": 0.08928571428571429, "grad_norm": 127.39149475097656, "learning_rate": 9.107142857142858e-06, "loss": 1.3096, "step": 175 }, { "epoch": 0.08979591836734693, "grad_norm": 27.312244415283203, "learning_rate": 9.102040816326532e-06, "loss": 0.707, "step": 176 }, { "epoch": 0.09030612244897959, "grad_norm": 53.83796310424805, "learning_rate": 9.096938775510205e-06, "loss": 0.5449, "step": 177 }, { "epoch": 0.09081632653061225, "grad_norm": 43.877750396728516, "learning_rate": 9.091836734693878e-06, "loss": 0.458, "step": 178 }, { "epoch": 0.0913265306122449, "grad_norm": 80.93883514404297, "learning_rate": 9.08673469387755e-06, "loss": 1.0059, "step": 179 }, { "epoch": 0.09183673469387756, "grad_norm": 59.49557113647461, "learning_rate": 9.081632653061225e-06, "loss": 0.7266, "step": 180 }, { "epoch": 0.0923469387755102, "grad_norm": 35.15586853027344, "learning_rate": 9.076530612244898e-06, "loss": 0.1199, "step": 181 }, { "epoch": 0.09285714285714286, "grad_norm": 41.139923095703125, "learning_rate": 9.071428571428573e-06, "loss": 0.4775, "step": 182 }, { "epoch": 0.0933673469387755, "grad_norm": 46.234535217285156, "learning_rate": 9.066326530612246e-06, "loss": 0.6602, "step": 183 }, { "epoch": 0.09387755102040816, "grad_norm": 49.565345764160156, "learning_rate": 9.061224489795919e-06, "loss": 0.8809, "step": 184 }, { "epoch": 0.09438775510204081, "grad_norm": 37.45954132080078, "learning_rate": 9.056122448979593e-06, "loss": 0.2021, "step": 185 }, { "epoch": 0.09489795918367347, "grad_norm": 64.1097183227539, "learning_rate": 9.051020408163266e-06, "loss": 0.957, "step": 186 }, { "epoch": 0.09540816326530613, "grad_norm": 27.192216873168945, "learning_rate": 9.045918367346939e-06, "loss": 0.2319, "step": 187 }, { "epoch": 0.09591836734693877, "grad_norm": 46.71800231933594, "learning_rate": 9.040816326530612e-06, "loss": 0.792, "step": 188 }, { "epoch": 0.09642857142857143, "grad_norm": 37.28218460083008, "learning_rate": 9.035714285714287e-06, "loss": 0.5078, "step": 189 }, { "epoch": 0.09693877551020408, "grad_norm": 29.099224090576172, "learning_rate": 9.03061224489796e-06, "loss": 0.3232, "step": 190 }, { "epoch": 0.09744897959183674, "grad_norm": 21.810150146484375, "learning_rate": 9.025510204081634e-06, "loss": 0.6426, "step": 191 }, { "epoch": 0.09795918367346938, "grad_norm": 71.40327453613281, "learning_rate": 9.020408163265307e-06, "loss": 0.6064, "step": 192 }, { "epoch": 0.09846938775510204, "grad_norm": 25.434703826904297, "learning_rate": 9.01530612244898e-06, "loss": 0.3604, "step": 193 }, { "epoch": 0.09897959183673469, "grad_norm": 26.399639129638672, "learning_rate": 9.010204081632654e-06, "loss": 0.3208, "step": 194 }, { "epoch": 0.09948979591836735, "grad_norm": 64.2388687133789, "learning_rate": 9.005102040816327e-06, "loss": 0.7422, "step": 195 }, { "epoch": 0.1, "grad_norm": 36.59234619140625, "learning_rate": 9e-06, "loss": 0.1924, "step": 196 }, { "epoch": 0.10051020408163265, "grad_norm": 33.895286560058594, "learning_rate": 8.994897959183673e-06, "loss": 0.4346, "step": 197 }, { "epoch": 0.10102040816326531, "grad_norm": 36.25426483154297, "learning_rate": 8.989795918367348e-06, "loss": 0.0918, "step": 198 }, { "epoch": 0.10153061224489796, "grad_norm": 56.27308654785156, "learning_rate": 8.98469387755102e-06, "loss": 0.8496, "step": 199 }, { "epoch": 0.10204081632653061, "grad_norm": 48.48802947998047, "learning_rate": 8.979591836734695e-06, "loss": 0.7109, "step": 200 }, { "epoch": 0.10255102040816326, "grad_norm": 54.04899978637695, "learning_rate": 8.974489795918368e-06, "loss": 0.8984, "step": 201 }, { "epoch": 0.10306122448979592, "grad_norm": 25.264175415039062, "learning_rate": 8.969387755102041e-06, "loss": 0.5039, "step": 202 }, { "epoch": 0.10357142857142858, "grad_norm": 35.44578170776367, "learning_rate": 8.964285714285716e-06, "loss": 0.3164, "step": 203 }, { "epoch": 0.10408163265306122, "grad_norm": 62.46015167236328, "learning_rate": 8.959183673469388e-06, "loss": 0.7227, "step": 204 }, { "epoch": 0.10459183673469388, "grad_norm": 24.94677734375, "learning_rate": 8.954081632653061e-06, "loss": 0.1694, "step": 205 }, { "epoch": 0.10510204081632653, "grad_norm": 74.01685333251953, "learning_rate": 8.948979591836734e-06, "loss": 0.7949, "step": 206 }, { "epoch": 0.10561224489795919, "grad_norm": 32.872379302978516, "learning_rate": 8.943877551020409e-06, "loss": 0.3557, "step": 207 }, { "epoch": 0.10612244897959183, "grad_norm": 27.903867721557617, "learning_rate": 8.938775510204082e-06, "loss": 0.2368, "step": 208 }, { "epoch": 0.10663265306122449, "grad_norm": 22.090301513671875, "learning_rate": 8.933673469387756e-06, "loss": 0.0959, "step": 209 }, { "epoch": 0.10714285714285714, "grad_norm": 27.679231643676758, "learning_rate": 8.92857142857143e-06, "loss": 0.2947, "step": 210 }, { "epoch": 0.1076530612244898, "grad_norm": 39.07429504394531, "learning_rate": 8.923469387755102e-06, "loss": 0.1277, "step": 211 }, { "epoch": 0.10816326530612246, "grad_norm": 33.66646194458008, "learning_rate": 8.918367346938777e-06, "loss": 0.3818, "step": 212 }, { "epoch": 0.1086734693877551, "grad_norm": 28.026960372924805, "learning_rate": 8.91326530612245e-06, "loss": 0.2286, "step": 213 }, { "epoch": 0.10918367346938776, "grad_norm": 33.26628494262695, "learning_rate": 8.908163265306124e-06, "loss": 0.4546, "step": 214 }, { "epoch": 0.1096938775510204, "grad_norm": 52.94529724121094, "learning_rate": 8.903061224489795e-06, "loss": 0.4824, "step": 215 }, { "epoch": 0.11020408163265306, "grad_norm": 15.159096717834473, "learning_rate": 8.89795918367347e-06, "loss": 0.1144, "step": 216 }, { "epoch": 0.11071428571428571, "grad_norm": 23.162355422973633, "learning_rate": 8.892857142857143e-06, "loss": 0.1323, "step": 217 }, { "epoch": 0.11122448979591837, "grad_norm": 49.617061614990234, "learning_rate": 8.887755102040817e-06, "loss": 0.2441, "step": 218 }, { "epoch": 0.11173469387755101, "grad_norm": 26.43107795715332, "learning_rate": 8.88265306122449e-06, "loss": 0.356, "step": 219 }, { "epoch": 0.11224489795918367, "grad_norm": 43.631553649902344, "learning_rate": 8.877551020408163e-06, "loss": 0.4194, "step": 220 }, { "epoch": 0.11275510204081633, "grad_norm": 71.73251342773438, "learning_rate": 8.872448979591838e-06, "loss": 0.3906, "step": 221 }, { "epoch": 0.11326530612244898, "grad_norm": 61.167144775390625, "learning_rate": 8.86734693877551e-06, "loss": 0.5012, "step": 222 }, { "epoch": 0.11377551020408164, "grad_norm": 60.311222076416016, "learning_rate": 8.862244897959185e-06, "loss": 0.6016, "step": 223 }, { "epoch": 0.11428571428571428, "grad_norm": 49.39049530029297, "learning_rate": 8.857142857142858e-06, "loss": 0.5391, "step": 224 }, { "epoch": 0.11479591836734694, "grad_norm": 28.716745376586914, "learning_rate": 8.852040816326531e-06, "loss": 0.2886, "step": 225 }, { "epoch": 0.11530612244897959, "grad_norm": 42.27504348754883, "learning_rate": 8.846938775510204e-06, "loss": 0.3074, "step": 226 }, { "epoch": 0.11581632653061225, "grad_norm": 36.80427551269531, "learning_rate": 8.841836734693879e-06, "loss": 0.2863, "step": 227 }, { "epoch": 0.11632653061224489, "grad_norm": 18.35335922241211, "learning_rate": 8.836734693877552e-06, "loss": 0.0645, "step": 228 }, { "epoch": 0.11683673469387755, "grad_norm": 55.21095275878906, "learning_rate": 8.831632653061224e-06, "loss": 0.2705, "step": 229 }, { "epoch": 0.11734693877551021, "grad_norm": 57.27177047729492, "learning_rate": 8.826530612244899e-06, "loss": 0.8574, "step": 230 }, { "epoch": 0.11785714285714285, "grad_norm": 48.23779296875, "learning_rate": 8.821428571428572e-06, "loss": 0.4484, "step": 231 }, { "epoch": 0.11836734693877551, "grad_norm": 33.41584014892578, "learning_rate": 8.816326530612247e-06, "loss": 0.1866, "step": 232 }, { "epoch": 0.11887755102040816, "grad_norm": 71.69505310058594, "learning_rate": 8.81122448979592e-06, "loss": 0.4823, "step": 233 }, { "epoch": 0.11938775510204082, "grad_norm": 37.181087493896484, "learning_rate": 8.806122448979592e-06, "loss": 0.2637, "step": 234 }, { "epoch": 0.11989795918367346, "grad_norm": 49.456912994384766, "learning_rate": 8.801020408163265e-06, "loss": 0.1661, "step": 235 }, { "epoch": 0.12040816326530612, "grad_norm": 34.55421829223633, "learning_rate": 8.79591836734694e-06, "loss": 0.3809, "step": 236 }, { "epoch": 0.12091836734693878, "grad_norm": 106.53833770751953, "learning_rate": 8.790816326530613e-06, "loss": 0.4915, "step": 237 }, { "epoch": 0.12142857142857143, "grad_norm": 82.4144058227539, "learning_rate": 8.785714285714286e-06, "loss": 0.9395, "step": 238 }, { "epoch": 0.12193877551020409, "grad_norm": 8.338484764099121, "learning_rate": 8.78061224489796e-06, "loss": 0.0359, "step": 239 }, { "epoch": 0.12244897959183673, "grad_norm": 27.24197769165039, "learning_rate": 8.775510204081633e-06, "loss": 0.0732, "step": 240 }, { "epoch": 0.12295918367346939, "grad_norm": 59.791656494140625, "learning_rate": 8.770408163265308e-06, "loss": 0.5913, "step": 241 }, { "epoch": 0.12346938775510204, "grad_norm": 88.53572845458984, "learning_rate": 8.76530612244898e-06, "loss": 0.332, "step": 242 }, { "epoch": 0.1239795918367347, "grad_norm": 77.56645202636719, "learning_rate": 8.760204081632653e-06, "loss": 0.2969, "step": 243 }, { "epoch": 0.12448979591836734, "grad_norm": 47.407623291015625, "learning_rate": 8.755102040816326e-06, "loss": 0.4927, "step": 244 }, { "epoch": 0.125, "grad_norm": 33.048248291015625, "learning_rate": 8.750000000000001e-06, "loss": 0.074, "step": 245 }, { "epoch": 0.12551020408163266, "grad_norm": 19.97512435913086, "learning_rate": 8.744897959183676e-06, "loss": 0.0337, "step": 246 }, { "epoch": 0.12602040816326532, "grad_norm": 50.18816375732422, "learning_rate": 8.739795918367347e-06, "loss": 0.2217, "step": 247 }, { "epoch": 0.12653061224489795, "grad_norm": 75.94056701660156, "learning_rate": 8.734693877551021e-06, "loss": 0.6309, "step": 248 }, { "epoch": 0.1270408163265306, "grad_norm": 104.46366882324219, "learning_rate": 8.729591836734694e-06, "loss": 0.5479, "step": 249 }, { "epoch": 0.12755102040816327, "grad_norm": 26.87546157836914, "learning_rate": 8.724489795918369e-06, "loss": 0.1982, "step": 250 }, { "epoch": 0.12806122448979593, "grad_norm": 16.216379165649414, "learning_rate": 8.719387755102042e-06, "loss": 0.0601, "step": 251 }, { "epoch": 0.12857142857142856, "grad_norm": 39.083560943603516, "learning_rate": 8.714285714285715e-06, "loss": 0.1228, "step": 252 }, { "epoch": 0.12908163265306122, "grad_norm": 75.17076873779297, "learning_rate": 8.709183673469387e-06, "loss": 0.9609, "step": 253 }, { "epoch": 0.12959183673469388, "grad_norm": 133.8837432861328, "learning_rate": 8.704081632653062e-06, "loss": 0.8516, "step": 254 }, { "epoch": 0.13010204081632654, "grad_norm": 61.974365234375, "learning_rate": 8.698979591836737e-06, "loss": 0.2451, "step": 255 }, { "epoch": 0.1306122448979592, "grad_norm": 34.44236755371094, "learning_rate": 8.69387755102041e-06, "loss": 0.2124, "step": 256 }, { "epoch": 0.13112244897959183, "grad_norm": 55.44428634643555, "learning_rate": 8.688775510204082e-06, "loss": 0.7656, "step": 257 }, { "epoch": 0.13163265306122449, "grad_norm": 116.1556167602539, "learning_rate": 8.683673469387755e-06, "loss": 1.4521, "step": 258 }, { "epoch": 0.13214285714285715, "grad_norm": 110.94064331054688, "learning_rate": 8.67857142857143e-06, "loss": 0.9434, "step": 259 }, { "epoch": 0.1326530612244898, "grad_norm": 100.89263916015625, "learning_rate": 8.673469387755103e-06, "loss": 1.1631, "step": 260 }, { "epoch": 0.13316326530612244, "grad_norm": 78.21894836425781, "learning_rate": 8.668367346938776e-06, "loss": 0.499, "step": 261 }, { "epoch": 0.1336734693877551, "grad_norm": 37.99648666381836, "learning_rate": 8.663265306122449e-06, "loss": 0.4854, "step": 262 }, { "epoch": 0.13418367346938775, "grad_norm": 86.06537628173828, "learning_rate": 8.658163265306123e-06, "loss": 0.4014, "step": 263 }, { "epoch": 0.1346938775510204, "grad_norm": 59.674598693847656, "learning_rate": 8.653061224489798e-06, "loss": 0.3076, "step": 264 }, { "epoch": 0.13520408163265307, "grad_norm": 63.46868896484375, "learning_rate": 8.64795918367347e-06, "loss": 0.6289, "step": 265 }, { "epoch": 0.1357142857142857, "grad_norm": 74.9300308227539, "learning_rate": 8.642857142857144e-06, "loss": 0.5176, "step": 266 }, { "epoch": 0.13622448979591836, "grad_norm": 57.741172790527344, "learning_rate": 8.637755102040816e-06, "loss": 0.4883, "step": 267 }, { "epoch": 0.13673469387755102, "grad_norm": 60.409149169921875, "learning_rate": 8.632653061224491e-06, "loss": 0.4551, "step": 268 }, { "epoch": 0.13724489795918368, "grad_norm": 79.2987060546875, "learning_rate": 8.627551020408164e-06, "loss": 0.5596, "step": 269 }, { "epoch": 0.1377551020408163, "grad_norm": 45.00891876220703, "learning_rate": 8.622448979591837e-06, "loss": 0.5879, "step": 270 }, { "epoch": 0.13826530612244897, "grad_norm": 40.72996139526367, "learning_rate": 8.61734693877551e-06, "loss": 0.3721, "step": 271 }, { "epoch": 0.13877551020408163, "grad_norm": 46.73809051513672, "learning_rate": 8.612244897959184e-06, "loss": 0.2412, "step": 272 }, { "epoch": 0.1392857142857143, "grad_norm": 26.15532112121582, "learning_rate": 8.607142857142859e-06, "loss": 0.1895, "step": 273 }, { "epoch": 0.13979591836734695, "grad_norm": 41.36906814575195, "learning_rate": 8.602040816326532e-06, "loss": 0.3081, "step": 274 }, { "epoch": 0.14030612244897958, "grad_norm": 104.63702392578125, "learning_rate": 8.596938775510205e-06, "loss": 1.0986, "step": 275 }, { "epoch": 0.14081632653061224, "grad_norm": 100.24798583984375, "learning_rate": 8.591836734693878e-06, "loss": 0.7061, "step": 276 }, { "epoch": 0.1413265306122449, "grad_norm": 43.33245086669922, "learning_rate": 8.586734693877552e-06, "loss": 0.4072, "step": 277 }, { "epoch": 0.14183673469387756, "grad_norm": 46.28691864013672, "learning_rate": 8.581632653061225e-06, "loss": 0.201, "step": 278 }, { "epoch": 0.14234693877551022, "grad_norm": 25.770387649536133, "learning_rate": 8.576530612244898e-06, "loss": 0.1011, "step": 279 }, { "epoch": 0.14285714285714285, "grad_norm": 42.79970932006836, "learning_rate": 8.571428571428571e-06, "loss": 0.2358, "step": 280 }, { "epoch": 0.1433673469387755, "grad_norm": 50.44066619873047, "learning_rate": 8.566326530612246e-06, "loss": 0.241, "step": 281 }, { "epoch": 0.14387755102040817, "grad_norm": 53.05989074707031, "learning_rate": 8.56122448979592e-06, "loss": 0.6592, "step": 282 }, { "epoch": 0.14438775510204083, "grad_norm": 63.34260559082031, "learning_rate": 8.556122448979593e-06, "loss": 0.4815, "step": 283 }, { "epoch": 0.14489795918367346, "grad_norm": 41.73698425292969, "learning_rate": 8.551020408163266e-06, "loss": 0.2329, "step": 284 }, { "epoch": 0.14540816326530612, "grad_norm": 6.303662300109863, "learning_rate": 8.545918367346939e-06, "loss": 0.0266, "step": 285 }, { "epoch": 0.14591836734693878, "grad_norm": 88.41893768310547, "learning_rate": 8.540816326530613e-06, "loss": 1.0347, "step": 286 }, { "epoch": 0.14642857142857144, "grad_norm": 31.252849578857422, "learning_rate": 8.535714285714286e-06, "loss": 0.1816, "step": 287 }, { "epoch": 0.1469387755102041, "grad_norm": 98.52784729003906, "learning_rate": 8.530612244897961e-06, "loss": 0.8613, "step": 288 }, { "epoch": 0.14744897959183673, "grad_norm": 49.3430290222168, "learning_rate": 8.525510204081632e-06, "loss": 0.2776, "step": 289 }, { "epoch": 0.14795918367346939, "grad_norm": 97.83133697509766, "learning_rate": 8.520408163265307e-06, "loss": 0.499, "step": 290 }, { "epoch": 0.14846938775510204, "grad_norm": 80.4065933227539, "learning_rate": 8.515306122448981e-06, "loss": 1.123, "step": 291 }, { "epoch": 0.1489795918367347, "grad_norm": 30.211336135864258, "learning_rate": 8.510204081632654e-06, "loss": 0.2061, "step": 292 }, { "epoch": 0.14948979591836734, "grad_norm": 45.938838958740234, "learning_rate": 8.505102040816327e-06, "loss": 0.2998, "step": 293 }, { "epoch": 0.15, "grad_norm": 57.71543884277344, "learning_rate": 8.5e-06, "loss": 0.4883, "step": 294 }, { "epoch": 0.15051020408163265, "grad_norm": 10.640128135681152, "learning_rate": 8.494897959183675e-06, "loss": 0.0518, "step": 295 }, { "epoch": 0.1510204081632653, "grad_norm": 76.93633270263672, "learning_rate": 8.489795918367347e-06, "loss": 0.7891, "step": 296 }, { "epoch": 0.15153061224489797, "grad_norm": 82.10570526123047, "learning_rate": 8.484693877551022e-06, "loss": 0.6436, "step": 297 }, { "epoch": 0.1520408163265306, "grad_norm": 67.86988830566406, "learning_rate": 8.479591836734695e-06, "loss": 0.5068, "step": 298 }, { "epoch": 0.15255102040816326, "grad_norm": 41.14324951171875, "learning_rate": 8.474489795918368e-06, "loss": 0.1956, "step": 299 }, { "epoch": 0.15306122448979592, "grad_norm": 60.73426818847656, "learning_rate": 8.469387755102042e-06, "loss": 0.7998, "step": 300 }, { "epoch": 0.15357142857142858, "grad_norm": 47.681617736816406, "learning_rate": 8.464285714285715e-06, "loss": 0.3171, "step": 301 }, { "epoch": 0.1540816326530612, "grad_norm": 43.50709533691406, "learning_rate": 8.459183673469388e-06, "loss": 0.5742, "step": 302 }, { "epoch": 0.15459183673469387, "grad_norm": 28.713869094848633, "learning_rate": 8.454081632653061e-06, "loss": 0.2756, "step": 303 }, { "epoch": 0.15510204081632653, "grad_norm": 55.816200256347656, "learning_rate": 8.448979591836736e-06, "loss": 0.3618, "step": 304 }, { "epoch": 0.1556122448979592, "grad_norm": 39.46714401245117, "learning_rate": 8.443877551020409e-06, "loss": 0.3535, "step": 305 }, { "epoch": 0.15612244897959185, "grad_norm": 84.02921295166016, "learning_rate": 8.438775510204083e-06, "loss": 0.748, "step": 306 }, { "epoch": 0.15663265306122448, "grad_norm": 52.73015213012695, "learning_rate": 8.433673469387756e-06, "loss": 0.2314, "step": 307 }, { "epoch": 0.15714285714285714, "grad_norm": 31.677162170410156, "learning_rate": 8.428571428571429e-06, "loss": 0.624, "step": 308 }, { "epoch": 0.1576530612244898, "grad_norm": 34.53183364868164, "learning_rate": 8.423469387755104e-06, "loss": 0.4263, "step": 309 }, { "epoch": 0.15816326530612246, "grad_norm": 40.19322967529297, "learning_rate": 8.418367346938776e-06, "loss": 0.1184, "step": 310 }, { "epoch": 0.1586734693877551, "grad_norm": 124.65576171875, "learning_rate": 8.41326530612245e-06, "loss": 1.3789, "step": 311 }, { "epoch": 0.15918367346938775, "grad_norm": 52.95748519897461, "learning_rate": 8.408163265306122e-06, "loss": 0.5312, "step": 312 }, { "epoch": 0.1596938775510204, "grad_norm": 66.03582000732422, "learning_rate": 8.403061224489797e-06, "loss": 0.5742, "step": 313 }, { "epoch": 0.16020408163265307, "grad_norm": 35.39268493652344, "learning_rate": 8.39795918367347e-06, "loss": 0.1003, "step": 314 }, { "epoch": 0.16071428571428573, "grad_norm": 28.167009353637695, "learning_rate": 8.392857142857144e-06, "loss": 0.1018, "step": 315 }, { "epoch": 0.16122448979591836, "grad_norm": 48.40702438354492, "learning_rate": 8.387755102040817e-06, "loss": 0.417, "step": 316 }, { "epoch": 0.16173469387755102, "grad_norm": 74.44703674316406, "learning_rate": 8.38265306122449e-06, "loss": 0.5732, "step": 317 }, { "epoch": 0.16224489795918368, "grad_norm": 55.36091232299805, "learning_rate": 8.377551020408165e-06, "loss": 0.2676, "step": 318 }, { "epoch": 0.16275510204081634, "grad_norm": 49.48396682739258, "learning_rate": 8.372448979591838e-06, "loss": 0.3145, "step": 319 }, { "epoch": 0.16326530612244897, "grad_norm": 39.45724868774414, "learning_rate": 8.36734693877551e-06, "loss": 0.2256, "step": 320 }, { "epoch": 0.16377551020408163, "grad_norm": 121.82907104492188, "learning_rate": 8.362244897959183e-06, "loss": 1.0195, "step": 321 }, { "epoch": 0.16428571428571428, "grad_norm": 37.260616302490234, "learning_rate": 8.357142857142858e-06, "loss": 0.0977, "step": 322 }, { "epoch": 0.16479591836734694, "grad_norm": 26.400325775146484, "learning_rate": 8.352040816326531e-06, "loss": 0.1295, "step": 323 }, { "epoch": 0.1653061224489796, "grad_norm": 44.40492630004883, "learning_rate": 8.346938775510205e-06, "loss": 0.1959, "step": 324 }, { "epoch": 0.16581632653061223, "grad_norm": 31.654462814331055, "learning_rate": 8.341836734693878e-06, "loss": 0.3108, "step": 325 }, { "epoch": 0.1663265306122449, "grad_norm": 35.24894714355469, "learning_rate": 8.336734693877551e-06, "loss": 0.2936, "step": 326 }, { "epoch": 0.16683673469387755, "grad_norm": 25.59202766418457, "learning_rate": 8.331632653061226e-06, "loss": 0.4714, "step": 327 }, { "epoch": 0.1673469387755102, "grad_norm": 33.37561798095703, "learning_rate": 8.326530612244899e-06, "loss": 0.584, "step": 328 }, { "epoch": 0.16785714285714284, "grad_norm": 123.29547882080078, "learning_rate": 8.321428571428573e-06, "loss": 0.9277, "step": 329 }, { "epoch": 0.1683673469387755, "grad_norm": 33.114742279052734, "learning_rate": 8.316326530612246e-06, "loss": 0.1851, "step": 330 }, { "epoch": 0.16887755102040816, "grad_norm": 119.26338195800781, "learning_rate": 8.311224489795919e-06, "loss": 1.168, "step": 331 }, { "epoch": 0.16938775510204082, "grad_norm": 97.96987915039062, "learning_rate": 8.306122448979592e-06, "loss": 0.7793, "step": 332 }, { "epoch": 0.16989795918367348, "grad_norm": 50.52949142456055, "learning_rate": 8.301020408163267e-06, "loss": 0.4727, "step": 333 }, { "epoch": 0.1704081632653061, "grad_norm": 130.82582092285156, "learning_rate": 8.29591836734694e-06, "loss": 1.0137, "step": 334 }, { "epoch": 0.17091836734693877, "grad_norm": 66.32474517822266, "learning_rate": 8.290816326530612e-06, "loss": 0.5317, "step": 335 }, { "epoch": 0.17142857142857143, "grad_norm": 40.856529235839844, "learning_rate": 8.285714285714287e-06, "loss": 0.2861, "step": 336 }, { "epoch": 0.1719387755102041, "grad_norm": 25.520469665527344, "learning_rate": 8.28061224489796e-06, "loss": 0.0726, "step": 337 }, { "epoch": 0.17244897959183675, "grad_norm": 30.47495460510254, "learning_rate": 8.275510204081634e-06, "loss": 0.0952, "step": 338 }, { "epoch": 0.17295918367346938, "grad_norm": 16.54120635986328, "learning_rate": 8.270408163265307e-06, "loss": 0.0461, "step": 339 }, { "epoch": 0.17346938775510204, "grad_norm": 82.70003509521484, "learning_rate": 8.26530612244898e-06, "loss": 0.6833, "step": 340 }, { "epoch": 0.1739795918367347, "grad_norm": 43.37309265136719, "learning_rate": 8.260204081632653e-06, "loss": 0.5439, "step": 341 }, { "epoch": 0.17448979591836736, "grad_norm": 23.585269927978516, "learning_rate": 8.255102040816328e-06, "loss": 0.1714, "step": 342 }, { "epoch": 0.175, "grad_norm": 44.23314666748047, "learning_rate": 8.25e-06, "loss": 0.433, "step": 343 }, { "epoch": 0.17551020408163265, "grad_norm": 38.68181228637695, "learning_rate": 8.244897959183674e-06, "loss": 0.5098, "step": 344 }, { "epoch": 0.1760204081632653, "grad_norm": 112.05716705322266, "learning_rate": 8.239795918367348e-06, "loss": 1.1885, "step": 345 }, { "epoch": 0.17653061224489797, "grad_norm": 63.49458694458008, "learning_rate": 8.234693877551021e-06, "loss": 0.2717, "step": 346 }, { "epoch": 0.17704081632653063, "grad_norm": 14.389328956604004, "learning_rate": 8.229591836734696e-06, "loss": 0.0342, "step": 347 }, { "epoch": 0.17755102040816326, "grad_norm": 64.75745391845703, "learning_rate": 8.224489795918369e-06, "loss": 0.6685, "step": 348 }, { "epoch": 0.17806122448979592, "grad_norm": 51.707847595214844, "learning_rate": 8.219387755102041e-06, "loss": 0.5898, "step": 349 }, { "epoch": 0.17857142857142858, "grad_norm": 28.19064712524414, "learning_rate": 8.214285714285714e-06, "loss": 0.207, "step": 350 }, { "epoch": 0.17908163265306123, "grad_norm": 62.81899642944336, "learning_rate": 8.209183673469389e-06, "loss": 0.4961, "step": 351 }, { "epoch": 0.17959183673469387, "grad_norm": 78.13274383544922, "learning_rate": 8.204081632653062e-06, "loss": 0.791, "step": 352 }, { "epoch": 0.18010204081632653, "grad_norm": 47.81070327758789, "learning_rate": 8.198979591836735e-06, "loss": 0.6602, "step": 353 }, { "epoch": 0.18061224489795918, "grad_norm": 35.674034118652344, "learning_rate": 8.19387755102041e-06, "loss": 0.1392, "step": 354 }, { "epoch": 0.18112244897959184, "grad_norm": 56.06858825683594, "learning_rate": 8.188775510204082e-06, "loss": 0.3277, "step": 355 }, { "epoch": 0.1816326530612245, "grad_norm": 51.35543441772461, "learning_rate": 8.183673469387757e-06, "loss": 0.9551, "step": 356 }, { "epoch": 0.18214285714285713, "grad_norm": 50.242496490478516, "learning_rate": 8.17857142857143e-06, "loss": 0.3262, "step": 357 }, { "epoch": 0.1826530612244898, "grad_norm": 73.2733383178711, "learning_rate": 8.173469387755103e-06, "loss": 0.5041, "step": 358 }, { "epoch": 0.18316326530612245, "grad_norm": 55.5291748046875, "learning_rate": 8.168367346938775e-06, "loss": 0.5957, "step": 359 }, { "epoch": 0.1836734693877551, "grad_norm": 36.09253692626953, "learning_rate": 8.16326530612245e-06, "loss": 0.1963, "step": 360 }, { "epoch": 0.18418367346938774, "grad_norm": 59.55515670776367, "learning_rate": 8.158163265306123e-06, "loss": 0.3503, "step": 361 }, { "epoch": 0.1846938775510204, "grad_norm": 37.886474609375, "learning_rate": 8.153061224489796e-06, "loss": 0.1621, "step": 362 }, { "epoch": 0.18520408163265306, "grad_norm": 102.13749694824219, "learning_rate": 8.14795918367347e-06, "loss": 1.0117, "step": 363 }, { "epoch": 0.18571428571428572, "grad_norm": 32.727352142333984, "learning_rate": 8.142857142857143e-06, "loss": 0.3516, "step": 364 }, { "epoch": 0.18622448979591838, "grad_norm": 36.9388313293457, "learning_rate": 8.137755102040818e-06, "loss": 0.4268, "step": 365 }, { "epoch": 0.186734693877551, "grad_norm": 38.68507766723633, "learning_rate": 8.13265306122449e-06, "loss": 0.3945, "step": 366 }, { "epoch": 0.18724489795918367, "grad_norm": 60.041080474853516, "learning_rate": 8.127551020408164e-06, "loss": 0.5752, "step": 367 }, { "epoch": 0.18775510204081633, "grad_norm": 33.511962890625, "learning_rate": 8.122448979591837e-06, "loss": 0.1885, "step": 368 }, { "epoch": 0.188265306122449, "grad_norm": 44.21847152709961, "learning_rate": 8.117346938775511e-06, "loss": 0.228, "step": 369 }, { "epoch": 0.18877551020408162, "grad_norm": 60.78368377685547, "learning_rate": 8.112244897959184e-06, "loss": 0.3486, "step": 370 }, { "epoch": 0.18928571428571428, "grad_norm": 48.3172607421875, "learning_rate": 8.107142857142859e-06, "loss": 0.2396, "step": 371 }, { "epoch": 0.18979591836734694, "grad_norm": 69.15843963623047, "learning_rate": 8.102040816326532e-06, "loss": 0.7051, "step": 372 }, { "epoch": 0.1903061224489796, "grad_norm": 56.71660232543945, "learning_rate": 8.096938775510204e-06, "loss": 0.4492, "step": 373 }, { "epoch": 0.19081632653061226, "grad_norm": 53.852088928222656, "learning_rate": 8.091836734693879e-06, "loss": 0.3027, "step": 374 }, { "epoch": 0.1913265306122449, "grad_norm": 105.52400970458984, "learning_rate": 8.086734693877552e-06, "loss": 0.6055, "step": 375 }, { "epoch": 0.19183673469387755, "grad_norm": 30.79682159423828, "learning_rate": 8.081632653061225e-06, "loss": 0.1924, "step": 376 }, { "epoch": 0.1923469387755102, "grad_norm": 60.20758056640625, "learning_rate": 8.076530612244898e-06, "loss": 0.3481, "step": 377 }, { "epoch": 0.19285714285714287, "grad_norm": 68.6085433959961, "learning_rate": 8.071428571428572e-06, "loss": 0.3486, "step": 378 }, { "epoch": 0.1933673469387755, "grad_norm": 36.71294021606445, "learning_rate": 8.066326530612245e-06, "loss": 0.157, "step": 379 }, { "epoch": 0.19387755102040816, "grad_norm": 66.08944702148438, "learning_rate": 8.06122448979592e-06, "loss": 0.418, "step": 380 }, { "epoch": 0.19438775510204082, "grad_norm": 28.767972946166992, "learning_rate": 8.056122448979593e-06, "loss": 0.0811, "step": 381 }, { "epoch": 0.19489795918367347, "grad_norm": 74.8750228881836, "learning_rate": 8.051020408163266e-06, "loss": 0.7305, "step": 382 }, { "epoch": 0.19540816326530613, "grad_norm": 126.6944808959961, "learning_rate": 8.04591836734694e-06, "loss": 0.8604, "step": 383 }, { "epoch": 0.19591836734693877, "grad_norm": 28.484832763671875, "learning_rate": 8.040816326530613e-06, "loss": 0.0796, "step": 384 }, { "epoch": 0.19642857142857142, "grad_norm": 99.94397735595703, "learning_rate": 8.035714285714286e-06, "loss": 0.9609, "step": 385 }, { "epoch": 0.19693877551020408, "grad_norm": 90.40021514892578, "learning_rate": 8.030612244897959e-06, "loss": 0.7686, "step": 386 }, { "epoch": 0.19744897959183674, "grad_norm": 49.889373779296875, "learning_rate": 8.025510204081633e-06, "loss": 0.3564, "step": 387 }, { "epoch": 0.19795918367346937, "grad_norm": 99.73383331298828, "learning_rate": 8.020408163265306e-06, "loss": 0.6348, "step": 388 }, { "epoch": 0.19846938775510203, "grad_norm": 22.81363868713379, "learning_rate": 8.015306122448981e-06, "loss": 0.1312, "step": 389 }, { "epoch": 0.1989795918367347, "grad_norm": 55.26231384277344, "learning_rate": 8.010204081632654e-06, "loss": 0.6108, "step": 390 }, { "epoch": 0.19948979591836735, "grad_norm": 22.79471206665039, "learning_rate": 8.005102040816327e-06, "loss": 0.087, "step": 391 }, { "epoch": 0.2, "grad_norm": 29.473344802856445, "learning_rate": 8.000000000000001e-06, "loss": 0.1812, "step": 392 }, { "epoch": 0.20051020408163264, "grad_norm": 32.93170928955078, "learning_rate": 7.994897959183674e-06, "loss": 0.4022, "step": 393 }, { "epoch": 0.2010204081632653, "grad_norm": 76.46302032470703, "learning_rate": 7.989795918367347e-06, "loss": 0.4575, "step": 394 }, { "epoch": 0.20153061224489796, "grad_norm": 61.31814956665039, "learning_rate": 7.98469387755102e-06, "loss": 0.5129, "step": 395 }, { "epoch": 0.20204081632653062, "grad_norm": 35.000831604003906, "learning_rate": 7.979591836734695e-06, "loss": 0.2104, "step": 396 }, { "epoch": 0.20255102040816325, "grad_norm": 44.25748062133789, "learning_rate": 7.974489795918368e-06, "loss": 0.1719, "step": 397 }, { "epoch": 0.2030612244897959, "grad_norm": 70.69973754882812, "learning_rate": 7.969387755102042e-06, "loss": 0.458, "step": 398 }, { "epoch": 0.20357142857142857, "grad_norm": 64.64582061767578, "learning_rate": 7.964285714285715e-06, "loss": 0.3066, "step": 399 }, { "epoch": 0.20408163265306123, "grad_norm": 46.0272216796875, "learning_rate": 7.959183673469388e-06, "loss": 0.3442, "step": 400 }, { "epoch": 0.2045918367346939, "grad_norm": 18.049333572387695, "learning_rate": 7.954081632653062e-06, "loss": 0.1038, "step": 401 }, { "epoch": 0.20510204081632652, "grad_norm": 60.02465057373047, "learning_rate": 7.948979591836735e-06, "loss": 0.2827, "step": 402 }, { "epoch": 0.20561224489795918, "grad_norm": 92.20713806152344, "learning_rate": 7.94387755102041e-06, "loss": 0.8533, "step": 403 }, { "epoch": 0.20612244897959184, "grad_norm": 26.00432777404785, "learning_rate": 7.938775510204081e-06, "loss": 0.1641, "step": 404 }, { "epoch": 0.2066326530612245, "grad_norm": 80.35779571533203, "learning_rate": 7.933673469387756e-06, "loss": 0.6309, "step": 405 }, { "epoch": 0.20714285714285716, "grad_norm": 67.13218688964844, "learning_rate": 7.928571428571429e-06, "loss": 0.4263, "step": 406 }, { "epoch": 0.2076530612244898, "grad_norm": 71.70327758789062, "learning_rate": 7.923469387755103e-06, "loss": 0.5332, "step": 407 }, { "epoch": 0.20816326530612245, "grad_norm": 16.56011962890625, "learning_rate": 7.918367346938776e-06, "loss": 0.0714, "step": 408 }, { "epoch": 0.2086734693877551, "grad_norm": 81.34461212158203, "learning_rate": 7.913265306122449e-06, "loss": 0.8555, "step": 409 }, { "epoch": 0.20918367346938777, "grad_norm": 98.36890411376953, "learning_rate": 7.908163265306124e-06, "loss": 1.0818, "step": 410 }, { "epoch": 0.2096938775510204, "grad_norm": 18.448646545410156, "learning_rate": 7.903061224489797e-06, "loss": 0.1075, "step": 411 }, { "epoch": 0.21020408163265306, "grad_norm": 49.35126876831055, "learning_rate": 7.897959183673471e-06, "loss": 0.4111, "step": 412 }, { "epoch": 0.21071428571428572, "grad_norm": 40.58681869506836, "learning_rate": 7.892857142857144e-06, "loss": 0.5557, "step": 413 }, { "epoch": 0.21122448979591837, "grad_norm": 66.8772964477539, "learning_rate": 7.887755102040817e-06, "loss": 0.665, "step": 414 }, { "epoch": 0.21173469387755103, "grad_norm": 80.0522689819336, "learning_rate": 7.88265306122449e-06, "loss": 0.4092, "step": 415 }, { "epoch": 0.21224489795918366, "grad_norm": 59.471046447753906, "learning_rate": 7.877551020408164e-06, "loss": 0.644, "step": 416 }, { "epoch": 0.21275510204081632, "grad_norm": 70.87850189208984, "learning_rate": 7.872448979591837e-06, "loss": 0.5942, "step": 417 }, { "epoch": 0.21326530612244898, "grad_norm": 45.05499267578125, "learning_rate": 7.86734693877551e-06, "loss": 0.2426, "step": 418 }, { "epoch": 0.21377551020408164, "grad_norm": 65.7256088256836, "learning_rate": 7.862244897959185e-06, "loss": 0.4143, "step": 419 }, { "epoch": 0.21428571428571427, "grad_norm": 34.18880081176758, "learning_rate": 7.857142857142858e-06, "loss": 0.3828, "step": 420 }, { "epoch": 0.21479591836734693, "grad_norm": 46.575557708740234, "learning_rate": 7.852040816326532e-06, "loss": 0.3311, "step": 421 }, { "epoch": 0.2153061224489796, "grad_norm": 44.27016067504883, "learning_rate": 7.846938775510205e-06, "loss": 0.2987, "step": 422 }, { "epoch": 0.21581632653061225, "grad_norm": 28.1955509185791, "learning_rate": 7.841836734693878e-06, "loss": 0.1346, "step": 423 }, { "epoch": 0.2163265306122449, "grad_norm": 62.506290435791016, "learning_rate": 7.836734693877551e-06, "loss": 0.6191, "step": 424 }, { "epoch": 0.21683673469387754, "grad_norm": 30.974912643432617, "learning_rate": 7.831632653061226e-06, "loss": 0.1741, "step": 425 }, { "epoch": 0.2173469387755102, "grad_norm": 27.254467010498047, "learning_rate": 7.826530612244898e-06, "loss": 0.1143, "step": 426 }, { "epoch": 0.21785714285714286, "grad_norm": 49.02326965332031, "learning_rate": 7.821428571428571e-06, "loss": 0.5283, "step": 427 }, { "epoch": 0.21836734693877552, "grad_norm": 40.55220031738281, "learning_rate": 7.816326530612246e-06, "loss": 0.228, "step": 428 }, { "epoch": 0.21887755102040815, "grad_norm": 60.710601806640625, "learning_rate": 7.811224489795919e-06, "loss": 0.4326, "step": 429 }, { "epoch": 0.2193877551020408, "grad_norm": 52.045169830322266, "learning_rate": 7.806122448979593e-06, "loss": 0.415, "step": 430 }, { "epoch": 0.21989795918367347, "grad_norm": 59.936588287353516, "learning_rate": 7.801020408163266e-06, "loss": 0.4504, "step": 431 }, { "epoch": 0.22040816326530613, "grad_norm": 68.2706069946289, "learning_rate": 7.79591836734694e-06, "loss": 0.3926, "step": 432 }, { "epoch": 0.2209183673469388, "grad_norm": 99.52059173583984, "learning_rate": 7.790816326530612e-06, "loss": 1.041, "step": 433 }, { "epoch": 0.22142857142857142, "grad_norm": 33.34510040283203, "learning_rate": 7.785714285714287e-06, "loss": 0.149, "step": 434 }, { "epoch": 0.22193877551020408, "grad_norm": 85.89479064941406, "learning_rate": 7.78061224489796e-06, "loss": 0.6328, "step": 435 }, { "epoch": 0.22244897959183674, "grad_norm": 55.58793640136719, "learning_rate": 7.775510204081632e-06, "loss": 0.3188, "step": 436 }, { "epoch": 0.2229591836734694, "grad_norm": 93.09330749511719, "learning_rate": 7.770408163265307e-06, "loss": 0.6992, "step": 437 }, { "epoch": 0.22346938775510203, "grad_norm": 57.54879379272461, "learning_rate": 7.76530612244898e-06, "loss": 0.5186, "step": 438 }, { "epoch": 0.2239795918367347, "grad_norm": 72.57948303222656, "learning_rate": 7.760204081632655e-06, "loss": 0.3975, "step": 439 }, { "epoch": 0.22448979591836735, "grad_norm": 74.01660919189453, "learning_rate": 7.755102040816327e-06, "loss": 0.4058, "step": 440 }, { "epoch": 0.225, "grad_norm": 28.245174407958984, "learning_rate": 7.75e-06, "loss": 0.2017, "step": 441 }, { "epoch": 0.22551020408163266, "grad_norm": 58.74325180053711, "learning_rate": 7.744897959183673e-06, "loss": 0.4375, "step": 442 }, { "epoch": 0.2260204081632653, "grad_norm": 17.515609741210938, "learning_rate": 7.739795918367348e-06, "loss": 0.0647, "step": 443 }, { "epoch": 0.22653061224489796, "grad_norm": 56.626060485839844, "learning_rate": 7.73469387755102e-06, "loss": 0.2925, "step": 444 }, { "epoch": 0.22704081632653061, "grad_norm": 38.42918395996094, "learning_rate": 7.729591836734695e-06, "loss": 0.2178, "step": 445 }, { "epoch": 0.22755102040816327, "grad_norm": 107.21630096435547, "learning_rate": 7.724489795918368e-06, "loss": 0.7979, "step": 446 }, { "epoch": 0.2280612244897959, "grad_norm": 54.65217208862305, "learning_rate": 7.719387755102041e-06, "loss": 0.626, "step": 447 }, { "epoch": 0.22857142857142856, "grad_norm": 22.198514938354492, "learning_rate": 7.714285714285716e-06, "loss": 0.1077, "step": 448 }, { "epoch": 0.22908163265306122, "grad_norm": 67.50344848632812, "learning_rate": 7.709183673469389e-06, "loss": 0.4111, "step": 449 }, { "epoch": 0.22959183673469388, "grad_norm": 21.145931243896484, "learning_rate": 7.704081632653061e-06, "loss": 0.0358, "step": 450 }, { "epoch": 0.23010204081632654, "grad_norm": 45.56168746948242, "learning_rate": 7.698979591836734e-06, "loss": 0.2812, "step": 451 }, { "epoch": 0.23061224489795917, "grad_norm": 18.347579956054688, "learning_rate": 7.693877551020409e-06, "loss": 0.0704, "step": 452 }, { "epoch": 0.23112244897959183, "grad_norm": 31.427236557006836, "learning_rate": 7.688775510204082e-06, "loss": 0.2996, "step": 453 }, { "epoch": 0.2316326530612245, "grad_norm": 40.55540084838867, "learning_rate": 7.683673469387756e-06, "loss": 0.0907, "step": 454 }, { "epoch": 0.23214285714285715, "grad_norm": 65.94026184082031, "learning_rate": 7.67857142857143e-06, "loss": 0.6895, "step": 455 }, { "epoch": 0.23265306122448978, "grad_norm": 56.37742233276367, "learning_rate": 7.673469387755102e-06, "loss": 0.3179, "step": 456 }, { "epoch": 0.23316326530612244, "grad_norm": 34.6407470703125, "learning_rate": 7.668367346938777e-06, "loss": 0.1238, "step": 457 }, { "epoch": 0.2336734693877551, "grad_norm": 8.487417221069336, "learning_rate": 7.66326530612245e-06, "loss": 0.0125, "step": 458 }, { "epoch": 0.23418367346938776, "grad_norm": 33.489078521728516, "learning_rate": 7.658163265306123e-06, "loss": 0.129, "step": 459 }, { "epoch": 0.23469387755102042, "grad_norm": 70.91468811035156, "learning_rate": 7.653061224489796e-06, "loss": 0.7324, "step": 460 }, { "epoch": 0.23520408163265305, "grad_norm": 47.71525192260742, "learning_rate": 7.64795918367347e-06, "loss": 0.248, "step": 461 }, { "epoch": 0.2357142857142857, "grad_norm": 60.962249755859375, "learning_rate": 7.642857142857143e-06, "loss": 0.2427, "step": 462 }, { "epoch": 0.23622448979591837, "grad_norm": 57.502925872802734, "learning_rate": 7.637755102040818e-06, "loss": 0.5176, "step": 463 }, { "epoch": 0.23673469387755103, "grad_norm": 26.173612594604492, "learning_rate": 7.63265306122449e-06, "loss": 0.1877, "step": 464 }, { "epoch": 0.2372448979591837, "grad_norm": 22.56398582458496, "learning_rate": 7.627551020408163e-06, "loss": 0.0405, "step": 465 }, { "epoch": 0.23775510204081632, "grad_norm": 47.31758117675781, "learning_rate": 7.622448979591838e-06, "loss": 0.3818, "step": 466 }, { "epoch": 0.23826530612244898, "grad_norm": 17.2362003326416, "learning_rate": 7.617346938775511e-06, "loss": 0.0488, "step": 467 }, { "epoch": 0.23877551020408164, "grad_norm": 95.24469757080078, "learning_rate": 7.612244897959185e-06, "loss": 1.3447, "step": 468 }, { "epoch": 0.2392857142857143, "grad_norm": 63.82474136352539, "learning_rate": 7.6071428571428575e-06, "loss": 0.2791, "step": 469 }, { "epoch": 0.23979591836734693, "grad_norm": 46.894325256347656, "learning_rate": 7.602040816326531e-06, "loss": 0.1836, "step": 470 }, { "epoch": 0.2403061224489796, "grad_norm": 84.83946990966797, "learning_rate": 7.596938775510204e-06, "loss": 0.5215, "step": 471 }, { "epoch": 0.24081632653061225, "grad_norm": 57.90958786010742, "learning_rate": 7.591836734693878e-06, "loss": 0.1667, "step": 472 }, { "epoch": 0.2413265306122449, "grad_norm": 88.54853057861328, "learning_rate": 7.5867346938775525e-06, "loss": 1.0625, "step": 473 }, { "epoch": 0.24183673469387756, "grad_norm": 119.522705078125, "learning_rate": 7.581632653061225e-06, "loss": 0.7262, "step": 474 }, { "epoch": 0.2423469387755102, "grad_norm": 17.98221206665039, "learning_rate": 7.576530612244899e-06, "loss": 0.0916, "step": 475 }, { "epoch": 0.24285714285714285, "grad_norm": 96.19312286376953, "learning_rate": 7.571428571428572e-06, "loss": 0.707, "step": 476 }, { "epoch": 0.24336734693877551, "grad_norm": 9.142897605895996, "learning_rate": 7.566326530612246e-06, "loss": 0.0359, "step": 477 }, { "epoch": 0.24387755102040817, "grad_norm": 164.375732421875, "learning_rate": 7.561224489795919e-06, "loss": 0.9707, "step": 478 }, { "epoch": 0.2443877551020408, "grad_norm": 60.962886810302734, "learning_rate": 7.556122448979592e-06, "loss": 0.4119, "step": 479 }, { "epoch": 0.24489795918367346, "grad_norm": 77.64954376220703, "learning_rate": 7.551020408163265e-06, "loss": 0.3125, "step": 480 }, { "epoch": 0.24540816326530612, "grad_norm": 72.75625610351562, "learning_rate": 7.545918367346939e-06, "loss": 0.5176, "step": 481 }, { "epoch": 0.24591836734693878, "grad_norm": 146.1666717529297, "learning_rate": 7.540816326530614e-06, "loss": 1.4927, "step": 482 }, { "epoch": 0.24642857142857144, "grad_norm": 80.99240112304688, "learning_rate": 7.5357142857142865e-06, "loss": 0.7744, "step": 483 }, { "epoch": 0.24693877551020407, "grad_norm": 103.67998504638672, "learning_rate": 7.53061224489796e-06, "loss": 0.8945, "step": 484 }, { "epoch": 0.24744897959183673, "grad_norm": 90.2336654663086, "learning_rate": 7.525510204081633e-06, "loss": 0.4443, "step": 485 }, { "epoch": 0.2479591836734694, "grad_norm": 44.69839096069336, "learning_rate": 7.520408163265307e-06, "loss": 0.161, "step": 486 }, { "epoch": 0.24846938775510205, "grad_norm": 121.79985809326172, "learning_rate": 7.51530612244898e-06, "loss": 0.8957, "step": 487 }, { "epoch": 0.24897959183673468, "grad_norm": 60.5068359375, "learning_rate": 7.5102040816326536e-06, "loss": 0.1909, "step": 488 }, { "epoch": 0.24948979591836734, "grad_norm": 30.39191436767578, "learning_rate": 7.5051020408163264e-06, "loss": 0.1595, "step": 489 }, { "epoch": 0.25, "grad_norm": 57.5645637512207, "learning_rate": 7.500000000000001e-06, "loss": 0.3745, "step": 490 }, { "epoch": 0.25051020408163266, "grad_norm": 136.97561645507812, "learning_rate": 7.494897959183675e-06, "loss": 0.9375, "step": 491 }, { "epoch": 0.2510204081632653, "grad_norm": 86.24913024902344, "learning_rate": 7.489795918367348e-06, "loss": 0.4902, "step": 492 }, { "epoch": 0.251530612244898, "grad_norm": 169.8101806640625, "learning_rate": 7.484693877551021e-06, "loss": 1.7891, "step": 493 }, { "epoch": 0.25204081632653064, "grad_norm": 30.859149932861328, "learning_rate": 7.479591836734694e-06, "loss": 0.0671, "step": 494 }, { "epoch": 0.25255102040816324, "grad_norm": 28.901206970214844, "learning_rate": 7.474489795918368e-06, "loss": 0.1367, "step": 495 }, { "epoch": 0.2530612244897959, "grad_norm": 47.944461822509766, "learning_rate": 7.469387755102041e-06, "loss": 0.3054, "step": 496 }, { "epoch": 0.25357142857142856, "grad_norm": 44.865684509277344, "learning_rate": 7.464285714285715e-06, "loss": 0.2778, "step": 497 }, { "epoch": 0.2540816326530612, "grad_norm": 29.194162368774414, "learning_rate": 7.459183673469388e-06, "loss": 0.1472, "step": 498 }, { "epoch": 0.2545918367346939, "grad_norm": 55.07899475097656, "learning_rate": 7.454081632653062e-06, "loss": 0.3232, "step": 499 }, { "epoch": 0.25510204081632654, "grad_norm": 78.74903106689453, "learning_rate": 7.448979591836736e-06, "loss": 0.6416, "step": 500 }, { "epoch": 0.2556122448979592, "grad_norm": 85.8030014038086, "learning_rate": 7.443877551020409e-06, "loss": 0.5293, "step": 501 }, { "epoch": 0.25612244897959185, "grad_norm": 49.712646484375, "learning_rate": 7.4387755102040826e-06, "loss": 0.1484, "step": 502 }, { "epoch": 0.2566326530612245, "grad_norm": 69.66159057617188, "learning_rate": 7.4336734693877555e-06, "loss": 0.4165, "step": 503 }, { "epoch": 0.2571428571428571, "grad_norm": 13.328307151794434, "learning_rate": 7.428571428571429e-06, "loss": 0.025, "step": 504 }, { "epoch": 0.2576530612244898, "grad_norm": 44.703025817871094, "learning_rate": 7.423469387755102e-06, "loss": 0.4515, "step": 505 }, { "epoch": 0.25816326530612244, "grad_norm": 68.42399597167969, "learning_rate": 7.418367346938776e-06, "loss": 0.2542, "step": 506 }, { "epoch": 0.2586734693877551, "grad_norm": 17.02503776550293, "learning_rate": 7.413265306122449e-06, "loss": 0.0648, "step": 507 }, { "epoch": 0.25918367346938775, "grad_norm": 68.38497161865234, "learning_rate": 7.408163265306123e-06, "loss": 0.3335, "step": 508 }, { "epoch": 0.2596938775510204, "grad_norm": 26.620222091674805, "learning_rate": 7.403061224489797e-06, "loss": 0.2466, "step": 509 }, { "epoch": 0.2602040816326531, "grad_norm": 19.28280258178711, "learning_rate": 7.39795918367347e-06, "loss": 0.1226, "step": 510 }, { "epoch": 0.26071428571428573, "grad_norm": 51.019100189208984, "learning_rate": 7.392857142857144e-06, "loss": 0.1372, "step": 511 }, { "epoch": 0.2612244897959184, "grad_norm": 35.12116241455078, "learning_rate": 7.387755102040817e-06, "loss": 0.1414, "step": 512 }, { "epoch": 0.261734693877551, "grad_norm": 163.62289428710938, "learning_rate": 7.38265306122449e-06, "loss": 1.4883, "step": 513 }, { "epoch": 0.26224489795918365, "grad_norm": 47.85900115966797, "learning_rate": 7.377551020408163e-06, "loss": 0.2164, "step": 514 }, { "epoch": 0.2627551020408163, "grad_norm": 92.01315307617188, "learning_rate": 7.372448979591838e-06, "loss": 0.5645, "step": 515 }, { "epoch": 0.26326530612244897, "grad_norm": 15.43292236328125, "learning_rate": 7.367346938775511e-06, "loss": 0.0324, "step": 516 }, { "epoch": 0.26377551020408163, "grad_norm": 58.146873474121094, "learning_rate": 7.3622448979591845e-06, "loss": 0.5654, "step": 517 }, { "epoch": 0.2642857142857143, "grad_norm": 52.54106521606445, "learning_rate": 7.357142857142858e-06, "loss": 0.7227, "step": 518 }, { "epoch": 0.26479591836734695, "grad_norm": 20.122568130493164, "learning_rate": 7.352040816326531e-06, "loss": 0.0303, "step": 519 }, { "epoch": 0.2653061224489796, "grad_norm": 136.90113830566406, "learning_rate": 7.346938775510205e-06, "loss": 1.9375, "step": 520 }, { "epoch": 0.26581632653061227, "grad_norm": 44.984230041503906, "learning_rate": 7.341836734693878e-06, "loss": 0.5605, "step": 521 }, { "epoch": 0.26632653061224487, "grad_norm": 35.16033172607422, "learning_rate": 7.3367346938775515e-06, "loss": 0.3604, "step": 522 }, { "epoch": 0.26683673469387753, "grad_norm": 29.710918426513672, "learning_rate": 7.331632653061224e-06, "loss": 0.1489, "step": 523 }, { "epoch": 0.2673469387755102, "grad_norm": 38.523651123046875, "learning_rate": 7.326530612244899e-06, "loss": 0.5332, "step": 524 }, { "epoch": 0.26785714285714285, "grad_norm": 34.49848556518555, "learning_rate": 7.321428571428572e-06, "loss": 0.0919, "step": 525 }, { "epoch": 0.2683673469387755, "grad_norm": 21.391084671020508, "learning_rate": 7.316326530612246e-06, "loss": 0.1531, "step": 526 }, { "epoch": 0.26887755102040817, "grad_norm": 68.62610626220703, "learning_rate": 7.311224489795919e-06, "loss": 0.6123, "step": 527 }, { "epoch": 0.2693877551020408, "grad_norm": 70.21046447753906, "learning_rate": 7.306122448979592e-06, "loss": 0.585, "step": 528 }, { "epoch": 0.2698979591836735, "grad_norm": 61.862369537353516, "learning_rate": 7.301020408163266e-06, "loss": 0.4712, "step": 529 }, { "epoch": 0.27040816326530615, "grad_norm": 76.4526596069336, "learning_rate": 7.295918367346939e-06, "loss": 0.8271, "step": 530 }, { "epoch": 0.27091836734693875, "grad_norm": 11.558497428894043, "learning_rate": 7.2908163265306135e-06, "loss": 0.0996, "step": 531 }, { "epoch": 0.2714285714285714, "grad_norm": 80.89340209960938, "learning_rate": 7.285714285714286e-06, "loss": 0.6611, "step": 532 }, { "epoch": 0.27193877551020407, "grad_norm": 41.57585906982422, "learning_rate": 7.28061224489796e-06, "loss": 0.2812, "step": 533 }, { "epoch": 0.2724489795918367, "grad_norm": 59.999488830566406, "learning_rate": 7.275510204081633e-06, "loss": 0.1909, "step": 534 }, { "epoch": 0.2729591836734694, "grad_norm": 59.19394302368164, "learning_rate": 7.270408163265307e-06, "loss": 0.3467, "step": 535 }, { "epoch": 0.27346938775510204, "grad_norm": 54.81898498535156, "learning_rate": 7.2653061224489805e-06, "loss": 0.3662, "step": 536 }, { "epoch": 0.2739795918367347, "grad_norm": 110.44430541992188, "learning_rate": 7.260204081632653e-06, "loss": 0.6921, "step": 537 }, { "epoch": 0.27448979591836736, "grad_norm": 19.55908966064453, "learning_rate": 7.255102040816327e-06, "loss": 0.105, "step": 538 }, { "epoch": 0.275, "grad_norm": 9.943583488464355, "learning_rate": 7.25e-06, "loss": 0.056, "step": 539 }, { "epoch": 0.2755102040816326, "grad_norm": 38.76340866088867, "learning_rate": 7.244897959183675e-06, "loss": 0.1365, "step": 540 }, { "epoch": 0.2760204081632653, "grad_norm": 10.29061508178711, "learning_rate": 7.2397959183673475e-06, "loss": 0.046, "step": 541 }, { "epoch": 0.27653061224489794, "grad_norm": 86.3921890258789, "learning_rate": 7.234693877551021e-06, "loss": 0.4463, "step": 542 }, { "epoch": 0.2770408163265306, "grad_norm": 6.460108280181885, "learning_rate": 7.229591836734694e-06, "loss": 0.0279, "step": 543 }, { "epoch": 0.27755102040816326, "grad_norm": 94.395751953125, "learning_rate": 7.224489795918368e-06, "loss": 0.7773, "step": 544 }, { "epoch": 0.2780612244897959, "grad_norm": 94.9793930053711, "learning_rate": 7.219387755102042e-06, "loss": 0.6055, "step": 545 }, { "epoch": 0.2785714285714286, "grad_norm": 60.07078170776367, "learning_rate": 7.2142857142857145e-06, "loss": 0.5176, "step": 546 }, { "epoch": 0.27908163265306124, "grad_norm": 57.19651412963867, "learning_rate": 7.209183673469389e-06, "loss": 0.1489, "step": 547 }, { "epoch": 0.2795918367346939, "grad_norm": 93.2054214477539, "learning_rate": 7.204081632653061e-06, "loss": 1.0586, "step": 548 }, { "epoch": 0.2801020408163265, "grad_norm": 76.24694061279297, "learning_rate": 7.198979591836736e-06, "loss": 0.5352, "step": 549 }, { "epoch": 0.28061224489795916, "grad_norm": 86.98812866210938, "learning_rate": 7.193877551020409e-06, "loss": 0.6018, "step": 550 }, { "epoch": 0.2811224489795918, "grad_norm": 15.802156448364258, "learning_rate": 7.188775510204082e-06, "loss": 0.0587, "step": 551 }, { "epoch": 0.2816326530612245, "grad_norm": 35.18263626098633, "learning_rate": 7.183673469387755e-06, "loss": 0.2646, "step": 552 }, { "epoch": 0.28214285714285714, "grad_norm": 39.418212890625, "learning_rate": 7.178571428571429e-06, "loss": 0.1509, "step": 553 }, { "epoch": 0.2826530612244898, "grad_norm": 86.09902954101562, "learning_rate": 7.173469387755103e-06, "loss": 0.6016, "step": 554 }, { "epoch": 0.28316326530612246, "grad_norm": 104.65943145751953, "learning_rate": 7.168367346938776e-06, "loss": 0.5039, "step": 555 }, { "epoch": 0.2836734693877551, "grad_norm": 23.996835708618164, "learning_rate": 7.16326530612245e-06, "loss": 0.1389, "step": 556 }, { "epoch": 0.2841836734693878, "grad_norm": 62.28258514404297, "learning_rate": 7.158163265306123e-06, "loss": 0.2954, "step": 557 }, { "epoch": 0.28469387755102044, "grad_norm": 54.2957763671875, "learning_rate": 7.153061224489797e-06, "loss": 0.3303, "step": 558 }, { "epoch": 0.28520408163265304, "grad_norm": 18.173784255981445, "learning_rate": 7.14795918367347e-06, "loss": 0.0736, "step": 559 }, { "epoch": 0.2857142857142857, "grad_norm": 87.5218734741211, "learning_rate": 7.1428571428571436e-06, "loss": 0.5242, "step": 560 }, { "epoch": 0.28622448979591836, "grad_norm": 46.485416412353516, "learning_rate": 7.1377551020408165e-06, "loss": 0.1783, "step": 561 }, { "epoch": 0.286734693877551, "grad_norm": 53.482994079589844, "learning_rate": 7.13265306122449e-06, "loss": 0.3632, "step": 562 }, { "epoch": 0.2872448979591837, "grad_norm": 83.8442153930664, "learning_rate": 7.127551020408165e-06, "loss": 0.5039, "step": 563 }, { "epoch": 0.28775510204081634, "grad_norm": 22.975019454956055, "learning_rate": 7.122448979591837e-06, "loss": 0.0403, "step": 564 }, { "epoch": 0.288265306122449, "grad_norm": 45.07960510253906, "learning_rate": 7.1173469387755114e-06, "loss": 0.1768, "step": 565 }, { "epoch": 0.28877551020408165, "grad_norm": 18.278635025024414, "learning_rate": 7.112244897959184e-06, "loss": 0.0886, "step": 566 }, { "epoch": 0.2892857142857143, "grad_norm": 24.451723098754883, "learning_rate": 7.107142857142858e-06, "loss": 0.0328, "step": 567 }, { "epoch": 0.2897959183673469, "grad_norm": 151.42164611816406, "learning_rate": 7.102040816326531e-06, "loss": 0.6089, "step": 568 }, { "epoch": 0.2903061224489796, "grad_norm": 43.341007232666016, "learning_rate": 7.096938775510205e-06, "loss": 0.1113, "step": 569 }, { "epoch": 0.29081632653061223, "grad_norm": 67.61224365234375, "learning_rate": 7.091836734693878e-06, "loss": 0.6084, "step": 570 }, { "epoch": 0.2913265306122449, "grad_norm": 34.41045379638672, "learning_rate": 7.086734693877551e-06, "loss": 0.2834, "step": 571 }, { "epoch": 0.29183673469387755, "grad_norm": 66.11260223388672, "learning_rate": 7.081632653061226e-06, "loss": 0.4607, "step": 572 }, { "epoch": 0.2923469387755102, "grad_norm": 74.0250244140625, "learning_rate": 7.076530612244899e-06, "loss": 0.7295, "step": 573 }, { "epoch": 0.29285714285714287, "grad_norm": 22.60026741027832, "learning_rate": 7.0714285714285726e-06, "loss": 0.1753, "step": 574 }, { "epoch": 0.29336734693877553, "grad_norm": 73.4660415649414, "learning_rate": 7.0663265306122455e-06, "loss": 0.2163, "step": 575 }, { "epoch": 0.2938775510204082, "grad_norm": 73.34139251708984, "learning_rate": 7.061224489795919e-06, "loss": 0.2651, "step": 576 }, { "epoch": 0.2943877551020408, "grad_norm": 169.72520446777344, "learning_rate": 7.056122448979592e-06, "loss": 0.7058, "step": 577 }, { "epoch": 0.29489795918367345, "grad_norm": 47.97903823852539, "learning_rate": 7.051020408163266e-06, "loss": 0.3447, "step": 578 }, { "epoch": 0.2954081632653061, "grad_norm": 51.342437744140625, "learning_rate": 7.045918367346939e-06, "loss": 0.1798, "step": 579 }, { "epoch": 0.29591836734693877, "grad_norm": 72.24365234375, "learning_rate": 7.0408163265306125e-06, "loss": 0.29, "step": 580 }, { "epoch": 0.29642857142857143, "grad_norm": 10.434687614440918, "learning_rate": 7.035714285714287e-06, "loss": 0.0147, "step": 581 }, { "epoch": 0.2969387755102041, "grad_norm": 73.64238739013672, "learning_rate": 7.03061224489796e-06, "loss": 0.4507, "step": 582 }, { "epoch": 0.29744897959183675, "grad_norm": 24.94285774230957, "learning_rate": 7.025510204081634e-06, "loss": 0.0677, "step": 583 }, { "epoch": 0.2979591836734694, "grad_norm": 50.386173248291016, "learning_rate": 7.020408163265307e-06, "loss": 0.3867, "step": 584 }, { "epoch": 0.29846938775510207, "grad_norm": 42.0314826965332, "learning_rate": 7.01530612244898e-06, "loss": 0.1572, "step": 585 }, { "epoch": 0.29897959183673467, "grad_norm": 103.0766830444336, "learning_rate": 7.010204081632653e-06, "loss": 0.4795, "step": 586 }, { "epoch": 0.29948979591836733, "grad_norm": 33.716793060302734, "learning_rate": 7.005102040816327e-06, "loss": 0.1272, "step": 587 }, { "epoch": 0.3, "grad_norm": 69.68559265136719, "learning_rate": 7e-06, "loss": 0.2988, "step": 588 }, { "epoch": 0.30051020408163265, "grad_norm": 85.51964569091797, "learning_rate": 6.9948979591836745e-06, "loss": 0.263, "step": 589 }, { "epoch": 0.3010204081632653, "grad_norm": 66.89107513427734, "learning_rate": 6.989795918367348e-06, "loss": 0.1748, "step": 590 }, { "epoch": 0.30153061224489797, "grad_norm": 179.29434204101562, "learning_rate": 6.984693877551021e-06, "loss": 1.4829, "step": 591 }, { "epoch": 0.3020408163265306, "grad_norm": 47.46409606933594, "learning_rate": 6.979591836734695e-06, "loss": 0.2734, "step": 592 }, { "epoch": 0.3025510204081633, "grad_norm": 96.49907684326172, "learning_rate": 6.974489795918368e-06, "loss": 0.9722, "step": 593 }, { "epoch": 0.30306122448979594, "grad_norm": 65.21955108642578, "learning_rate": 6.9693877551020415e-06, "loss": 0.2036, "step": 594 }, { "epoch": 0.30357142857142855, "grad_norm": 88.91761779785156, "learning_rate": 6.964285714285714e-06, "loss": 0.7012, "step": 595 }, { "epoch": 0.3040816326530612, "grad_norm": 14.813088417053223, "learning_rate": 6.959183673469388e-06, "loss": 0.0288, "step": 596 }, { "epoch": 0.30459183673469387, "grad_norm": 115.04486083984375, "learning_rate": 6.954081632653061e-06, "loss": 0.6367, "step": 597 }, { "epoch": 0.3051020408163265, "grad_norm": 57.65116882324219, "learning_rate": 6.948979591836736e-06, "loss": 0.1396, "step": 598 }, { "epoch": 0.3056122448979592, "grad_norm": 69.71482849121094, "learning_rate": 6.943877551020409e-06, "loss": 0.5265, "step": 599 }, { "epoch": 0.30612244897959184, "grad_norm": 95.42530059814453, "learning_rate": 6.938775510204082e-06, "loss": 0.4297, "step": 600 }, { "epoch": 0.3066326530612245, "grad_norm": 27.282943725585938, "learning_rate": 6.933673469387756e-06, "loss": 0.1362, "step": 601 }, { "epoch": 0.30714285714285716, "grad_norm": 62.66109848022461, "learning_rate": 6.928571428571429e-06, "loss": 0.356, "step": 602 }, { "epoch": 0.3076530612244898, "grad_norm": 85.93942260742188, "learning_rate": 6.923469387755103e-06, "loss": 0.5703, "step": 603 }, { "epoch": 0.3081632653061224, "grad_norm": 56.665740966796875, "learning_rate": 6.9183673469387755e-06, "loss": 0.2363, "step": 604 }, { "epoch": 0.3086734693877551, "grad_norm": 55.03900146484375, "learning_rate": 6.91326530612245e-06, "loss": 0.3955, "step": 605 }, { "epoch": 0.30918367346938774, "grad_norm": 84.49185943603516, "learning_rate": 6.908163265306122e-06, "loss": 0.6543, "step": 606 }, { "epoch": 0.3096938775510204, "grad_norm": 92.68756103515625, "learning_rate": 6.903061224489797e-06, "loss": 0.4712, "step": 607 }, { "epoch": 0.31020408163265306, "grad_norm": 33.713523864746094, "learning_rate": 6.8979591836734705e-06, "loss": 0.1246, "step": 608 }, { "epoch": 0.3107142857142857, "grad_norm": 10.993498802185059, "learning_rate": 6.892857142857143e-06, "loss": 0.0364, "step": 609 }, { "epoch": 0.3112244897959184, "grad_norm": 58.532447814941406, "learning_rate": 6.887755102040817e-06, "loss": 0.3183, "step": 610 }, { "epoch": 0.31173469387755104, "grad_norm": 84.3249740600586, "learning_rate": 6.88265306122449e-06, "loss": 0.5312, "step": 611 }, { "epoch": 0.3122448979591837, "grad_norm": 34.80390548706055, "learning_rate": 6.877551020408164e-06, "loss": 0.1831, "step": 612 }, { "epoch": 0.3127551020408163, "grad_norm": 36.095428466796875, "learning_rate": 6.872448979591837e-06, "loss": 0.1057, "step": 613 }, { "epoch": 0.31326530612244896, "grad_norm": 120.47953033447266, "learning_rate": 6.867346938775511e-06, "loss": 0.5972, "step": 614 }, { "epoch": 0.3137755102040816, "grad_norm": 103.96500396728516, "learning_rate": 6.862244897959184e-06, "loss": 0.9121, "step": 615 }, { "epoch": 0.3142857142857143, "grad_norm": 146.47897338867188, "learning_rate": 6.857142857142858e-06, "loss": 1.126, "step": 616 }, { "epoch": 0.31479591836734694, "grad_norm": 10.60655689239502, "learning_rate": 6.852040816326531e-06, "loss": 0.0089, "step": 617 }, { "epoch": 0.3153061224489796, "grad_norm": 18.95148277282715, "learning_rate": 6.8469387755102046e-06, "loss": 0.0493, "step": 618 }, { "epoch": 0.31581632653061226, "grad_norm": 59.76114273071289, "learning_rate": 6.841836734693878e-06, "loss": 0.3306, "step": 619 }, { "epoch": 0.3163265306122449, "grad_norm": 32.05243682861328, "learning_rate": 6.836734693877551e-06, "loss": 0.2437, "step": 620 }, { "epoch": 0.3168367346938776, "grad_norm": 6.291755199432373, "learning_rate": 6.831632653061225e-06, "loss": 0.0202, "step": 621 }, { "epoch": 0.3173469387755102, "grad_norm": 25.707189559936523, "learning_rate": 6.826530612244898e-06, "loss": 0.1176, "step": 622 }, { "epoch": 0.31785714285714284, "grad_norm": 112.7210464477539, "learning_rate": 6.8214285714285724e-06, "loss": 0.6221, "step": 623 }, { "epoch": 0.3183673469387755, "grad_norm": 104.612548828125, "learning_rate": 6.816326530612245e-06, "loss": 0.6216, "step": 624 }, { "epoch": 0.31887755102040816, "grad_norm": 63.23408126831055, "learning_rate": 6.811224489795919e-06, "loss": 0.399, "step": 625 }, { "epoch": 0.3193877551020408, "grad_norm": 58.56471252441406, "learning_rate": 6.806122448979592e-06, "loss": 0.5479, "step": 626 }, { "epoch": 0.3198979591836735, "grad_norm": 9.94200325012207, "learning_rate": 6.801020408163266e-06, "loss": 0.0253, "step": 627 }, { "epoch": 0.32040816326530613, "grad_norm": 16.286706924438477, "learning_rate": 6.7959183673469394e-06, "loss": 0.046, "step": 628 }, { "epoch": 0.3209183673469388, "grad_norm": 52.487613677978516, "learning_rate": 6.790816326530612e-06, "loss": 0.4473, "step": 629 }, { "epoch": 0.32142857142857145, "grad_norm": 29.131595611572266, "learning_rate": 6.785714285714287e-06, "loss": 0.0747, "step": 630 }, { "epoch": 0.32193877551020406, "grad_norm": 160.53924560546875, "learning_rate": 6.78061224489796e-06, "loss": 1.7327, "step": 631 }, { "epoch": 0.3224489795918367, "grad_norm": 122.8255615234375, "learning_rate": 6.7755102040816336e-06, "loss": 0.8127, "step": 632 }, { "epoch": 0.3229591836734694, "grad_norm": 8.849479675292969, "learning_rate": 6.7704081632653065e-06, "loss": 0.0299, "step": 633 }, { "epoch": 0.32346938775510203, "grad_norm": 54.16599655151367, "learning_rate": 6.76530612244898e-06, "loss": 0.3207, "step": 634 }, { "epoch": 0.3239795918367347, "grad_norm": 39.692989349365234, "learning_rate": 6.760204081632653e-06, "loss": 0.1351, "step": 635 }, { "epoch": 0.32448979591836735, "grad_norm": 33.28297424316406, "learning_rate": 6.755102040816327e-06, "loss": 0.198, "step": 636 }, { "epoch": 0.325, "grad_norm": 35.21436309814453, "learning_rate": 6.750000000000001e-06, "loss": 0.1343, "step": 637 }, { "epoch": 0.32551020408163267, "grad_norm": 17.411897659301758, "learning_rate": 6.7448979591836735e-06, "loss": 0.058, "step": 638 }, { "epoch": 0.32602040816326533, "grad_norm": 184.54757690429688, "learning_rate": 6.739795918367348e-06, "loss": 0.752, "step": 639 }, { "epoch": 0.32653061224489793, "grad_norm": 54.572505950927734, "learning_rate": 6.734693877551021e-06, "loss": 0.1548, "step": 640 }, { "epoch": 0.3270408163265306, "grad_norm": 124.35067749023438, "learning_rate": 6.729591836734695e-06, "loss": 1.0635, "step": 641 }, { "epoch": 0.32755102040816325, "grad_norm": 34.6632194519043, "learning_rate": 6.724489795918368e-06, "loss": 0.1105, "step": 642 }, { "epoch": 0.3280612244897959, "grad_norm": 34.782447814941406, "learning_rate": 6.719387755102041e-06, "loss": 0.0668, "step": 643 }, { "epoch": 0.32857142857142857, "grad_norm": 46.264686584472656, "learning_rate": 6.714285714285714e-06, "loss": 0.2549, "step": 644 }, { "epoch": 0.32908163265306123, "grad_norm": 36.21946334838867, "learning_rate": 6.709183673469388e-06, "loss": 0.123, "step": 645 }, { "epoch": 0.3295918367346939, "grad_norm": 117.5719223022461, "learning_rate": 6.704081632653063e-06, "loss": 0.5339, "step": 646 }, { "epoch": 0.33010204081632655, "grad_norm": 96.33863830566406, "learning_rate": 6.6989795918367355e-06, "loss": 0.4741, "step": 647 }, { "epoch": 0.3306122448979592, "grad_norm": 52.50869369506836, "learning_rate": 6.693877551020409e-06, "loss": 0.2856, "step": 648 }, { "epoch": 0.3311224489795918, "grad_norm": 127.28580474853516, "learning_rate": 6.688775510204082e-06, "loss": 0.9945, "step": 649 }, { "epoch": 0.33163265306122447, "grad_norm": 71.89524841308594, "learning_rate": 6.683673469387756e-06, "loss": 0.2942, "step": 650 }, { "epoch": 0.33214285714285713, "grad_norm": 51.458351135253906, "learning_rate": 6.678571428571429e-06, "loss": 0.1433, "step": 651 }, { "epoch": 0.3326530612244898, "grad_norm": 46.63521957397461, "learning_rate": 6.6734693877551025e-06, "loss": 0.1109, "step": 652 }, { "epoch": 0.33316326530612245, "grad_norm": 84.99858093261719, "learning_rate": 6.668367346938775e-06, "loss": 0.7637, "step": 653 }, { "epoch": 0.3336734693877551, "grad_norm": 69.51248168945312, "learning_rate": 6.663265306122449e-06, "loss": 0.1526, "step": 654 }, { "epoch": 0.33418367346938777, "grad_norm": 46.13043212890625, "learning_rate": 6.658163265306124e-06, "loss": 0.0925, "step": 655 }, { "epoch": 0.3346938775510204, "grad_norm": 56.35503387451172, "learning_rate": 6.653061224489797e-06, "loss": 0.3403, "step": 656 }, { "epoch": 0.3352040816326531, "grad_norm": 129.4577178955078, "learning_rate": 6.64795918367347e-06, "loss": 0.8574, "step": 657 }, { "epoch": 0.3357142857142857, "grad_norm": 33.84550476074219, "learning_rate": 6.642857142857143e-06, "loss": 0.106, "step": 658 }, { "epoch": 0.33622448979591835, "grad_norm": 108.67666625976562, "learning_rate": 6.637755102040817e-06, "loss": 0.812, "step": 659 }, { "epoch": 0.336734693877551, "grad_norm": 33.17964553833008, "learning_rate": 6.63265306122449e-06, "loss": 0.0979, "step": 660 }, { "epoch": 0.33724489795918366, "grad_norm": 105.8989028930664, "learning_rate": 6.627551020408164e-06, "loss": 0.4463, "step": 661 }, { "epoch": 0.3377551020408163, "grad_norm": 20.95460319519043, "learning_rate": 6.6224489795918365e-06, "loss": 0.0828, "step": 662 }, { "epoch": 0.338265306122449, "grad_norm": 13.823375701904297, "learning_rate": 6.61734693877551e-06, "loss": 0.0322, "step": 663 }, { "epoch": 0.33877551020408164, "grad_norm": 44.28261947631836, "learning_rate": 6.612244897959185e-06, "loss": 0.1475, "step": 664 }, { "epoch": 0.3392857142857143, "grad_norm": 53.08664321899414, "learning_rate": 6.607142857142858e-06, "loss": 0.2573, "step": 665 }, { "epoch": 0.33979591836734696, "grad_norm": 33.39915466308594, "learning_rate": 6.6020408163265315e-06, "loss": 0.1631, "step": 666 }, { "epoch": 0.34030612244897956, "grad_norm": 27.01349639892578, "learning_rate": 6.596938775510204e-06, "loss": 0.0981, "step": 667 }, { "epoch": 0.3408163265306122, "grad_norm": 178.18438720703125, "learning_rate": 6.591836734693878e-06, "loss": 1.293, "step": 668 }, { "epoch": 0.3413265306122449, "grad_norm": 52.705623626708984, "learning_rate": 6.586734693877551e-06, "loss": 0.4609, "step": 669 }, { "epoch": 0.34183673469387754, "grad_norm": 46.110130310058594, "learning_rate": 6.581632653061225e-06, "loss": 0.2939, "step": 670 }, { "epoch": 0.3423469387755102, "grad_norm": 6.2022881507873535, "learning_rate": 6.576530612244898e-06, "loss": 0.0181, "step": 671 }, { "epoch": 0.34285714285714286, "grad_norm": 21.12042999267578, "learning_rate": 6.571428571428572e-06, "loss": 0.0717, "step": 672 }, { "epoch": 0.3433673469387755, "grad_norm": 58.40286636352539, "learning_rate": 6.566326530612246e-06, "loss": 0.1954, "step": 673 }, { "epoch": 0.3438775510204082, "grad_norm": 101.62349700927734, "learning_rate": 6.561224489795919e-06, "loss": 0.6377, "step": 674 }, { "epoch": 0.34438775510204084, "grad_norm": 58.80821228027344, "learning_rate": 6.556122448979593e-06, "loss": 0.3762, "step": 675 }, { "epoch": 0.3448979591836735, "grad_norm": 24.995363235473633, "learning_rate": 6.5510204081632656e-06, "loss": 0.0511, "step": 676 }, { "epoch": 0.3454081632653061, "grad_norm": 46.22563552856445, "learning_rate": 6.545918367346939e-06, "loss": 0.2354, "step": 677 }, { "epoch": 0.34591836734693876, "grad_norm": 204.7807159423828, "learning_rate": 6.540816326530612e-06, "loss": 1.2695, "step": 678 }, { "epoch": 0.3464285714285714, "grad_norm": 18.232885360717773, "learning_rate": 6.535714285714286e-06, "loss": 0.0884, "step": 679 }, { "epoch": 0.3469387755102041, "grad_norm": 50.47403335571289, "learning_rate": 6.530612244897959e-06, "loss": 0.4492, "step": 680 }, { "epoch": 0.34744897959183674, "grad_norm": 48.998600006103516, "learning_rate": 6.5255102040816334e-06, "loss": 0.2241, "step": 681 }, { "epoch": 0.3479591836734694, "grad_norm": 77.34185028076172, "learning_rate": 6.520408163265307e-06, "loss": 0.5176, "step": 682 }, { "epoch": 0.34846938775510206, "grad_norm": 42.773014068603516, "learning_rate": 6.51530612244898e-06, "loss": 0.5508, "step": 683 }, { "epoch": 0.3489795918367347, "grad_norm": 89.96984100341797, "learning_rate": 6.510204081632654e-06, "loss": 0.4714, "step": 684 }, { "epoch": 0.3494897959183674, "grad_norm": 99.75232696533203, "learning_rate": 6.505102040816327e-06, "loss": 0.9199, "step": 685 }, { "epoch": 0.35, "grad_norm": 127.10189056396484, "learning_rate": 6.5000000000000004e-06, "loss": 1.1113, "step": 686 }, { "epoch": 0.35051020408163264, "grad_norm": 55.310176849365234, "learning_rate": 6.494897959183673e-06, "loss": 0.2316, "step": 687 }, { "epoch": 0.3510204081632653, "grad_norm": 168.09078979492188, "learning_rate": 6.489795918367348e-06, "loss": 1.5391, "step": 688 }, { "epoch": 0.35153061224489796, "grad_norm": 45.228919982910156, "learning_rate": 6.484693877551021e-06, "loss": 0.1755, "step": 689 }, { "epoch": 0.3520408163265306, "grad_norm": 29.311098098754883, "learning_rate": 6.4795918367346946e-06, "loss": 0.172, "step": 690 }, { "epoch": 0.3525510204081633, "grad_norm": 35.478515625, "learning_rate": 6.474489795918368e-06, "loss": 0.0848, "step": 691 }, { "epoch": 0.35306122448979593, "grad_norm": 78.30326080322266, "learning_rate": 6.469387755102041e-06, "loss": 0.5016, "step": 692 }, { "epoch": 0.3535714285714286, "grad_norm": 33.30475616455078, "learning_rate": 6.464285714285715e-06, "loss": 0.1831, "step": 693 }, { "epoch": 0.35408163265306125, "grad_norm": 18.659116744995117, "learning_rate": 6.459183673469388e-06, "loss": 0.0482, "step": 694 }, { "epoch": 0.35459183673469385, "grad_norm": 95.95103454589844, "learning_rate": 6.454081632653062e-06, "loss": 0.6113, "step": 695 }, { "epoch": 0.3551020408163265, "grad_norm": 109.00550842285156, "learning_rate": 6.4489795918367345e-06, "loss": 1.0957, "step": 696 }, { "epoch": 0.3556122448979592, "grad_norm": 31.346492767333984, "learning_rate": 6.443877551020409e-06, "loss": 0.2964, "step": 697 }, { "epoch": 0.35612244897959183, "grad_norm": 16.00472640991211, "learning_rate": 6.438775510204082e-06, "loss": 0.0874, "step": 698 }, { "epoch": 0.3566326530612245, "grad_norm": 27.065603256225586, "learning_rate": 6.433673469387756e-06, "loss": 0.1139, "step": 699 }, { "epoch": 0.35714285714285715, "grad_norm": 40.208343505859375, "learning_rate": 6.4285714285714295e-06, "loss": 0.3691, "step": 700 }, { "epoch": 0.3576530612244898, "grad_norm": 41.742271423339844, "learning_rate": 6.423469387755102e-06, "loss": 0.2476, "step": 701 }, { "epoch": 0.35816326530612247, "grad_norm": 34.36666488647461, "learning_rate": 6.418367346938776e-06, "loss": 0.125, "step": 702 }, { "epoch": 0.35867346938775513, "grad_norm": 76.45577239990234, "learning_rate": 6.413265306122449e-06, "loss": 0.6396, "step": 703 }, { "epoch": 0.35918367346938773, "grad_norm": 78.31637573242188, "learning_rate": 6.408163265306124e-06, "loss": 0.6113, "step": 704 }, { "epoch": 0.3596938775510204, "grad_norm": 57.81171798706055, "learning_rate": 6.403061224489796e-06, "loss": 0.1982, "step": 705 }, { "epoch": 0.36020408163265305, "grad_norm": 113.83602142333984, "learning_rate": 6.39795918367347e-06, "loss": 0.7129, "step": 706 }, { "epoch": 0.3607142857142857, "grad_norm": 9.619640350341797, "learning_rate": 6.392857142857143e-06, "loss": 0.0369, "step": 707 }, { "epoch": 0.36122448979591837, "grad_norm": 43.44652557373047, "learning_rate": 6.387755102040817e-06, "loss": 0.3083, "step": 708 }, { "epoch": 0.36173469387755103, "grad_norm": 75.51934051513672, "learning_rate": 6.382653061224491e-06, "loss": 0.6865, "step": 709 }, { "epoch": 0.3622448979591837, "grad_norm": 69.7166976928711, "learning_rate": 6.3775510204081635e-06, "loss": 0.748, "step": 710 }, { "epoch": 0.36275510204081635, "grad_norm": 26.1529483795166, "learning_rate": 6.372448979591837e-06, "loss": 0.1191, "step": 711 }, { "epoch": 0.363265306122449, "grad_norm": 49.0225944519043, "learning_rate": 6.36734693877551e-06, "loss": 0.2146, "step": 712 }, { "epoch": 0.3637755102040816, "grad_norm": 55.459983825683594, "learning_rate": 6.362244897959185e-06, "loss": 0.2958, "step": 713 }, { "epoch": 0.36428571428571427, "grad_norm": 70.99153137207031, "learning_rate": 6.357142857142858e-06, "loss": 0.469, "step": 714 }, { "epoch": 0.3647959183673469, "grad_norm": 72.18878936767578, "learning_rate": 6.352040816326531e-06, "loss": 0.6104, "step": 715 }, { "epoch": 0.3653061224489796, "grad_norm": 62.39371109008789, "learning_rate": 6.346938775510204e-06, "loss": 0.3428, "step": 716 }, { "epoch": 0.36581632653061225, "grad_norm": 37.23216247558594, "learning_rate": 6.341836734693878e-06, "loss": 0.3311, "step": 717 }, { "epoch": 0.3663265306122449, "grad_norm": 57.2979736328125, "learning_rate": 6.336734693877552e-06, "loss": 0.4099, "step": 718 }, { "epoch": 0.36683673469387756, "grad_norm": 52.950531005859375, "learning_rate": 6.331632653061225e-06, "loss": 0.4111, "step": 719 }, { "epoch": 0.3673469387755102, "grad_norm": 51.09238815307617, "learning_rate": 6.326530612244899e-06, "loss": 0.5452, "step": 720 }, { "epoch": 0.3678571428571429, "grad_norm": 63.770957946777344, "learning_rate": 6.321428571428571e-06, "loss": 0.4248, "step": 721 }, { "epoch": 0.3683673469387755, "grad_norm": 87.24472045898438, "learning_rate": 6.316326530612246e-06, "loss": 0.7324, "step": 722 }, { "epoch": 0.36887755102040815, "grad_norm": 50.441463470458984, "learning_rate": 6.311224489795919e-06, "loss": 0.233, "step": 723 }, { "epoch": 0.3693877551020408, "grad_norm": 17.58578872680664, "learning_rate": 6.3061224489795925e-06, "loss": 0.0942, "step": 724 }, { "epoch": 0.36989795918367346, "grad_norm": 68.0986328125, "learning_rate": 6.301020408163265e-06, "loss": 0.7158, "step": 725 }, { "epoch": 0.3704081632653061, "grad_norm": 162.9749298095703, "learning_rate": 6.295918367346939e-06, "loss": 0.6909, "step": 726 }, { "epoch": 0.3709183673469388, "grad_norm": 12.698287963867188, "learning_rate": 6.290816326530613e-06, "loss": 0.0308, "step": 727 }, { "epoch": 0.37142857142857144, "grad_norm": 36.611209869384766, "learning_rate": 6.285714285714286e-06, "loss": 0.0959, "step": 728 }, { "epoch": 0.3719387755102041, "grad_norm": 32.74394226074219, "learning_rate": 6.28061224489796e-06, "loss": 0.2021, "step": 729 }, { "epoch": 0.37244897959183676, "grad_norm": 47.79265213012695, "learning_rate": 6.275510204081633e-06, "loss": 0.4531, "step": 730 }, { "epoch": 0.37295918367346936, "grad_norm": 58.80137252807617, "learning_rate": 6.270408163265307e-06, "loss": 0.3057, "step": 731 }, { "epoch": 0.373469387755102, "grad_norm": 32.15926742553711, "learning_rate": 6.26530612244898e-06, "loss": 0.1187, "step": 732 }, { "epoch": 0.3739795918367347, "grad_norm": 118.35709381103516, "learning_rate": 6.260204081632654e-06, "loss": 0.3237, "step": 733 }, { "epoch": 0.37448979591836734, "grad_norm": 85.90968322753906, "learning_rate": 6.2551020408163266e-06, "loss": 0.4209, "step": 734 }, { "epoch": 0.375, "grad_norm": 36.53342819213867, "learning_rate": 6.25e-06, "loss": 0.208, "step": 735 }, { "epoch": 0.37551020408163266, "grad_norm": 23.086210250854492, "learning_rate": 6.244897959183675e-06, "loss": 0.1121, "step": 736 }, { "epoch": 0.3760204081632653, "grad_norm": 15.672575950622559, "learning_rate": 6.239795918367347e-06, "loss": 0.0588, "step": 737 }, { "epoch": 0.376530612244898, "grad_norm": 84.16727447509766, "learning_rate": 6.2346938775510215e-06, "loss": 0.3035, "step": 738 }, { "epoch": 0.37704081632653064, "grad_norm": 54.97834777832031, "learning_rate": 6.2295918367346944e-06, "loss": 0.2094, "step": 739 }, { "epoch": 0.37755102040816324, "grad_norm": 54.41300582885742, "learning_rate": 6.224489795918368e-06, "loss": 0.3986, "step": 740 }, { "epoch": 0.3780612244897959, "grad_norm": 67.71231079101562, "learning_rate": 6.219387755102041e-06, "loss": 0.3794, "step": 741 }, { "epoch": 0.37857142857142856, "grad_norm": 20.777193069458008, "learning_rate": 6.214285714285715e-06, "loss": 0.0916, "step": 742 }, { "epoch": 0.3790816326530612, "grad_norm": 28.154146194458008, "learning_rate": 6.209183673469388e-06, "loss": 0.1233, "step": 743 }, { "epoch": 0.3795918367346939, "grad_norm": 24.241737365722656, "learning_rate": 6.2040816326530614e-06, "loss": 0.1451, "step": 744 }, { "epoch": 0.38010204081632654, "grad_norm": 22.251419067382812, "learning_rate": 6.198979591836736e-06, "loss": 0.0942, "step": 745 }, { "epoch": 0.3806122448979592, "grad_norm": 35.25226974487305, "learning_rate": 6.193877551020409e-06, "loss": 0.0835, "step": 746 }, { "epoch": 0.38112244897959185, "grad_norm": 29.26425552368164, "learning_rate": 6.188775510204083e-06, "loss": 0.2065, "step": 747 }, { "epoch": 0.3816326530612245, "grad_norm": 120.84806823730469, "learning_rate": 6.1836734693877556e-06, "loss": 0.8281, "step": 748 }, { "epoch": 0.3821428571428571, "grad_norm": 25.176809310913086, "learning_rate": 6.178571428571429e-06, "loss": 0.0527, "step": 749 }, { "epoch": 0.3826530612244898, "grad_norm": 17.097774505615234, "learning_rate": 6.173469387755102e-06, "loss": 0.0523, "step": 750 }, { "epoch": 0.38316326530612244, "grad_norm": 27.714588165283203, "learning_rate": 6.168367346938776e-06, "loss": 0.0781, "step": 751 }, { "epoch": 0.3836734693877551, "grad_norm": 78.85135650634766, "learning_rate": 6.163265306122449e-06, "loss": 0.4395, "step": 752 }, { "epoch": 0.38418367346938775, "grad_norm": 25.059621810913086, "learning_rate": 6.158163265306123e-06, "loss": 0.1636, "step": 753 }, { "epoch": 0.3846938775510204, "grad_norm": 118.79293823242188, "learning_rate": 6.153061224489797e-06, "loss": 0.3486, "step": 754 }, { "epoch": 0.3852040816326531, "grad_norm": 13.154561042785645, "learning_rate": 6.14795918367347e-06, "loss": 0.0419, "step": 755 }, { "epoch": 0.38571428571428573, "grad_norm": 15.614875793457031, "learning_rate": 6.142857142857144e-06, "loss": 0.0492, "step": 756 }, { "epoch": 0.3862244897959184, "grad_norm": 99.27397918701172, "learning_rate": 6.137755102040817e-06, "loss": 0.4492, "step": 757 }, { "epoch": 0.386734693877551, "grad_norm": 34.83305740356445, "learning_rate": 6.1326530612244905e-06, "loss": 0.072, "step": 758 }, { "epoch": 0.38724489795918365, "grad_norm": 45.87968444824219, "learning_rate": 6.127551020408163e-06, "loss": 0.0995, "step": 759 }, { "epoch": 0.3877551020408163, "grad_norm": 17.09336280822754, "learning_rate": 6.122448979591837e-06, "loss": 0.0571, "step": 760 }, { "epoch": 0.38826530612244897, "grad_norm": 54.45036315917969, "learning_rate": 6.11734693877551e-06, "loss": 0.2256, "step": 761 }, { "epoch": 0.38877551020408163, "grad_norm": 30.03205108642578, "learning_rate": 6.112244897959185e-06, "loss": 0.0969, "step": 762 }, { "epoch": 0.3892857142857143, "grad_norm": 134.72689819335938, "learning_rate": 6.107142857142858e-06, "loss": 0.709, "step": 763 }, { "epoch": 0.38979591836734695, "grad_norm": 40.75043869018555, "learning_rate": 6.102040816326531e-06, "loss": 0.1127, "step": 764 }, { "epoch": 0.3903061224489796, "grad_norm": 17.195371627807617, "learning_rate": 6.096938775510205e-06, "loss": 0.0338, "step": 765 }, { "epoch": 0.39081632653061227, "grad_norm": 42.76658248901367, "learning_rate": 6.091836734693878e-06, "loss": 0.0624, "step": 766 }, { "epoch": 0.39132653061224487, "grad_norm": 56.503509521484375, "learning_rate": 6.086734693877552e-06, "loss": 0.161, "step": 767 }, { "epoch": 0.39183673469387753, "grad_norm": 32.44635009765625, "learning_rate": 6.0816326530612245e-06, "loss": 0.0788, "step": 768 }, { "epoch": 0.3923469387755102, "grad_norm": 107.02574920654297, "learning_rate": 6.076530612244898e-06, "loss": 0.6797, "step": 769 }, { "epoch": 0.39285714285714285, "grad_norm": 12.567275047302246, "learning_rate": 6.071428571428571e-06, "loss": 0.0435, "step": 770 }, { "epoch": 0.3933673469387755, "grad_norm": 53.879886627197266, "learning_rate": 6.066326530612246e-06, "loss": 0.3179, "step": 771 }, { "epoch": 0.39387755102040817, "grad_norm": 26.226062774658203, "learning_rate": 6.0612244897959195e-06, "loss": 0.092, "step": 772 }, { "epoch": 0.3943877551020408, "grad_norm": 80.61046600341797, "learning_rate": 6.056122448979592e-06, "loss": 0.3716, "step": 773 }, { "epoch": 0.3948979591836735, "grad_norm": 153.5399169921875, "learning_rate": 6.051020408163266e-06, "loss": 0.9648, "step": 774 }, { "epoch": 0.39540816326530615, "grad_norm": 59.661495208740234, "learning_rate": 6.045918367346939e-06, "loss": 0.1382, "step": 775 }, { "epoch": 0.39591836734693875, "grad_norm": 71.9591064453125, "learning_rate": 6.040816326530613e-06, "loss": 0.2725, "step": 776 }, { "epoch": 0.3964285714285714, "grad_norm": 50.36683654785156, "learning_rate": 6.035714285714286e-06, "loss": 0.1346, "step": 777 }, { "epoch": 0.39693877551020407, "grad_norm": 9.422607421875, "learning_rate": 6.03061224489796e-06, "loss": 0.0145, "step": 778 }, { "epoch": 0.3974489795918367, "grad_norm": 29.006954193115234, "learning_rate": 6.025510204081632e-06, "loss": 0.0822, "step": 779 }, { "epoch": 0.3979591836734694, "grad_norm": 66.5331039428711, "learning_rate": 6.020408163265307e-06, "loss": 0.168, "step": 780 }, { "epoch": 0.39846938775510204, "grad_norm": 54.40912628173828, "learning_rate": 6.015306122448981e-06, "loss": 0.3719, "step": 781 }, { "epoch": 0.3989795918367347, "grad_norm": 112.97367858886719, "learning_rate": 6.0102040816326535e-06, "loss": 0.4385, "step": 782 }, { "epoch": 0.39948979591836736, "grad_norm": 58.616695404052734, "learning_rate": 6.005102040816327e-06, "loss": 0.2078, "step": 783 }, { "epoch": 0.4, "grad_norm": 19.790332794189453, "learning_rate": 6e-06, "loss": 0.0471, "step": 784 }, { "epoch": 0.4005102040816326, "grad_norm": 224.37953186035156, "learning_rate": 5.994897959183674e-06, "loss": 1.5293, "step": 785 }, { "epoch": 0.4010204081632653, "grad_norm": 52.49542999267578, "learning_rate": 5.989795918367347e-06, "loss": 0.1899, "step": 786 }, { "epoch": 0.40153061224489794, "grad_norm": 193.25123596191406, "learning_rate": 5.984693877551021e-06, "loss": 1.4043, "step": 787 }, { "epoch": 0.4020408163265306, "grad_norm": 34.90439224243164, "learning_rate": 5.979591836734694e-06, "loss": 0.0681, "step": 788 }, { "epoch": 0.40255102040816326, "grad_norm": 65.14236450195312, "learning_rate": 5.974489795918368e-06, "loss": 0.2324, "step": 789 }, { "epoch": 0.4030612244897959, "grad_norm": 116.25560760498047, "learning_rate": 5.969387755102042e-06, "loss": 0.479, "step": 790 }, { "epoch": 0.4035714285714286, "grad_norm": 75.42031860351562, "learning_rate": 5.964285714285715e-06, "loss": 0.3206, "step": 791 }, { "epoch": 0.40408163265306124, "grad_norm": 58.61258316040039, "learning_rate": 5.959183673469388e-06, "loss": 0.1588, "step": 792 }, { "epoch": 0.4045918367346939, "grad_norm": 141.23422241210938, "learning_rate": 5.954081632653061e-06, "loss": 0.4495, "step": 793 }, { "epoch": 0.4051020408163265, "grad_norm": 187.69253540039062, "learning_rate": 5.948979591836735e-06, "loss": 1.0176, "step": 794 }, { "epoch": 0.40561224489795916, "grad_norm": 93.17790985107422, "learning_rate": 5.943877551020408e-06, "loss": 0.3057, "step": 795 }, { "epoch": 0.4061224489795918, "grad_norm": 26.17630386352539, "learning_rate": 5.9387755102040825e-06, "loss": 0.0745, "step": 796 }, { "epoch": 0.4066326530612245, "grad_norm": 45.29244613647461, "learning_rate": 5.933673469387755e-06, "loss": 0.1208, "step": 797 }, { "epoch": 0.40714285714285714, "grad_norm": 105.22510528564453, "learning_rate": 5.928571428571429e-06, "loss": 0.3031, "step": 798 }, { "epoch": 0.4076530612244898, "grad_norm": 69.72262573242188, "learning_rate": 5.923469387755103e-06, "loss": 0.1753, "step": 799 }, { "epoch": 0.40816326530612246, "grad_norm": 46.3089485168457, "learning_rate": 5.918367346938776e-06, "loss": 0.0952, "step": 800 }, { "epoch": 0.4086734693877551, "grad_norm": 82.36885833740234, "learning_rate": 5.9132653061224495e-06, "loss": 0.4172, "step": 801 }, { "epoch": 0.4091836734693878, "grad_norm": 82.40156555175781, "learning_rate": 5.9081632653061224e-06, "loss": 0.2668, "step": 802 }, { "epoch": 0.40969387755102044, "grad_norm": 39.34758377075195, "learning_rate": 5.903061224489797e-06, "loss": 0.1405, "step": 803 }, { "epoch": 0.41020408163265304, "grad_norm": 32.440101623535156, "learning_rate": 5.89795918367347e-06, "loss": 0.1064, "step": 804 }, { "epoch": 0.4107142857142857, "grad_norm": 131.41925048828125, "learning_rate": 5.892857142857144e-06, "loss": 0.7061, "step": 805 }, { "epoch": 0.41122448979591836, "grad_norm": 55.66576385498047, "learning_rate": 5.8877551020408166e-06, "loss": 0.1423, "step": 806 }, { "epoch": 0.411734693877551, "grad_norm": 32.686405181884766, "learning_rate": 5.88265306122449e-06, "loss": 0.0608, "step": 807 }, { "epoch": 0.4122448979591837, "grad_norm": 76.84168243408203, "learning_rate": 5.877551020408164e-06, "loss": 0.2812, "step": 808 }, { "epoch": 0.41275510204081634, "grad_norm": 37.915809631347656, "learning_rate": 5.872448979591837e-06, "loss": 0.0464, "step": 809 }, { "epoch": 0.413265306122449, "grad_norm": 160.66697692871094, "learning_rate": 5.867346938775511e-06, "loss": 0.9922, "step": 810 }, { "epoch": 0.41377551020408165, "grad_norm": 89.00968933105469, "learning_rate": 5.862244897959184e-06, "loss": 0.4907, "step": 811 }, { "epoch": 0.4142857142857143, "grad_norm": 119.16693115234375, "learning_rate": 5.857142857142858e-06, "loss": 0.5338, "step": 812 }, { "epoch": 0.4147959183673469, "grad_norm": 64.09293365478516, "learning_rate": 5.852040816326531e-06, "loss": 0.1733, "step": 813 }, { "epoch": 0.4153061224489796, "grad_norm": 9.424681663513184, "learning_rate": 5.846938775510205e-06, "loss": 0.0207, "step": 814 }, { "epoch": 0.41581632653061223, "grad_norm": 50.30925750732422, "learning_rate": 5.841836734693878e-06, "loss": 0.1084, "step": 815 }, { "epoch": 0.4163265306122449, "grad_norm": 3.1643331050872803, "learning_rate": 5.8367346938775515e-06, "loss": 0.0069, "step": 816 }, { "epoch": 0.41683673469387755, "grad_norm": 56.3040771484375, "learning_rate": 5.831632653061225e-06, "loss": 0.1697, "step": 817 }, { "epoch": 0.4173469387755102, "grad_norm": 87.26913452148438, "learning_rate": 5.826530612244898e-06, "loss": 0.3428, "step": 818 }, { "epoch": 0.41785714285714287, "grad_norm": 66.46612548828125, "learning_rate": 5.821428571428573e-06, "loss": 0.1997, "step": 819 }, { "epoch": 0.41836734693877553, "grad_norm": 121.48152923583984, "learning_rate": 5.816326530612246e-06, "loss": 0.582, "step": 820 }, { "epoch": 0.4188775510204082, "grad_norm": 152.61952209472656, "learning_rate": 5.811224489795919e-06, "loss": 0.9365, "step": 821 }, { "epoch": 0.4193877551020408, "grad_norm": 93.44112396240234, "learning_rate": 5.806122448979592e-06, "loss": 0.3889, "step": 822 }, { "epoch": 0.41989795918367345, "grad_norm": 113.58998107910156, "learning_rate": 5.801020408163266e-06, "loss": 0.7451, "step": 823 }, { "epoch": 0.4204081632653061, "grad_norm": 54.79607009887695, "learning_rate": 5.795918367346939e-06, "loss": 0.1489, "step": 824 }, { "epoch": 0.42091836734693877, "grad_norm": 87.33096313476562, "learning_rate": 5.790816326530613e-06, "loss": 0.5739, "step": 825 }, { "epoch": 0.42142857142857143, "grad_norm": 78.1507797241211, "learning_rate": 5.785714285714286e-06, "loss": 0.365, "step": 826 }, { "epoch": 0.4219387755102041, "grad_norm": 185.05941772460938, "learning_rate": 5.780612244897959e-06, "loss": 1.1938, "step": 827 }, { "epoch": 0.42244897959183675, "grad_norm": 191.7788848876953, "learning_rate": 5.775510204081634e-06, "loss": 1.4102, "step": 828 }, { "epoch": 0.4229591836734694, "grad_norm": 26.269508361816406, "learning_rate": 5.770408163265307e-06, "loss": 0.0681, "step": 829 }, { "epoch": 0.42346938775510207, "grad_norm": 240.48480224609375, "learning_rate": 5.7653061224489805e-06, "loss": 1.6182, "step": 830 }, { "epoch": 0.42397959183673467, "grad_norm": 90.67774963378906, "learning_rate": 5.760204081632653e-06, "loss": 0.5566, "step": 831 }, { "epoch": 0.42448979591836733, "grad_norm": 115.1650619506836, "learning_rate": 5.755102040816327e-06, "loss": 1.043, "step": 832 }, { "epoch": 0.425, "grad_norm": 17.023971557617188, "learning_rate": 5.75e-06, "loss": 0.0371, "step": 833 }, { "epoch": 0.42551020408163265, "grad_norm": 38.90401077270508, "learning_rate": 5.744897959183674e-06, "loss": 0.0923, "step": 834 }, { "epoch": 0.4260204081632653, "grad_norm": 13.465786933898926, "learning_rate": 5.739795918367348e-06, "loss": 0.0277, "step": 835 }, { "epoch": 0.42653061224489797, "grad_norm": 82.19554138183594, "learning_rate": 5.73469387755102e-06, "loss": 0.2559, "step": 836 }, { "epoch": 0.4270408163265306, "grad_norm": 37.804439544677734, "learning_rate": 5.729591836734695e-06, "loss": 0.0656, "step": 837 }, { "epoch": 0.4275510204081633, "grad_norm": 114.35718536376953, "learning_rate": 5.724489795918368e-06, "loss": 0.7441, "step": 838 }, { "epoch": 0.42806122448979594, "grad_norm": 27.299636840820312, "learning_rate": 5.719387755102042e-06, "loss": 0.0717, "step": 839 }, { "epoch": 0.42857142857142855, "grad_norm": 19.31078338623047, "learning_rate": 5.7142857142857145e-06, "loss": 0.0237, "step": 840 }, { "epoch": 0.4290816326530612, "grad_norm": 84.52933502197266, "learning_rate": 5.709183673469388e-06, "loss": 0.3677, "step": 841 }, { "epoch": 0.42959183673469387, "grad_norm": 42.0095100402832, "learning_rate": 5.704081632653061e-06, "loss": 0.1655, "step": 842 }, { "epoch": 0.4301020408163265, "grad_norm": 49.2426872253418, "learning_rate": 5.698979591836735e-06, "loss": 0.2953, "step": 843 }, { "epoch": 0.4306122448979592, "grad_norm": 54.21651077270508, "learning_rate": 5.6938775510204095e-06, "loss": 0.1747, "step": 844 }, { "epoch": 0.43112244897959184, "grad_norm": 98.43576049804688, "learning_rate": 5.688775510204082e-06, "loss": 0.4727, "step": 845 }, { "epoch": 0.4316326530612245, "grad_norm": 117.73087310791016, "learning_rate": 5.683673469387756e-06, "loss": 0.5642, "step": 846 }, { "epoch": 0.43214285714285716, "grad_norm": 125.51811218261719, "learning_rate": 5.678571428571429e-06, "loss": 0.3872, "step": 847 }, { "epoch": 0.4326530612244898, "grad_norm": 66.57725524902344, "learning_rate": 5.673469387755103e-06, "loss": 0.2705, "step": 848 }, { "epoch": 0.4331632653061224, "grad_norm": 84.17050170898438, "learning_rate": 5.668367346938776e-06, "loss": 0.2695, "step": 849 }, { "epoch": 0.4336734693877551, "grad_norm": 38.085330963134766, "learning_rate": 5.663265306122449e-06, "loss": 0.1846, "step": 850 }, { "epoch": 0.43418367346938774, "grad_norm": 35.52751922607422, "learning_rate": 5.658163265306122e-06, "loss": 0.1014, "step": 851 }, { "epoch": 0.4346938775510204, "grad_norm": 37.8881950378418, "learning_rate": 5.653061224489796e-06, "loss": 0.0724, "step": 852 }, { "epoch": 0.43520408163265306, "grad_norm": 91.28995513916016, "learning_rate": 5.647959183673471e-06, "loss": 0.3076, "step": 853 }, { "epoch": 0.4357142857142857, "grad_norm": 30.705759048461914, "learning_rate": 5.6428571428571435e-06, "loss": 0.0875, "step": 854 }, { "epoch": 0.4362244897959184, "grad_norm": 150.85446166992188, "learning_rate": 5.637755102040817e-06, "loss": 1.0117, "step": 855 }, { "epoch": 0.43673469387755104, "grad_norm": 29.93207550048828, "learning_rate": 5.63265306122449e-06, "loss": 0.0981, "step": 856 }, { "epoch": 0.4372448979591837, "grad_norm": 66.7497329711914, "learning_rate": 5.627551020408164e-06, "loss": 0.2161, "step": 857 }, { "epoch": 0.4377551020408163, "grad_norm": 124.7976303100586, "learning_rate": 5.622448979591837e-06, "loss": 0.5911, "step": 858 }, { "epoch": 0.43826530612244896, "grad_norm": 72.10818481445312, "learning_rate": 5.6173469387755105e-06, "loss": 0.4004, "step": 859 }, { "epoch": 0.4387755102040816, "grad_norm": 20.81977081298828, "learning_rate": 5.6122448979591834e-06, "loss": 0.0468, "step": 860 }, { "epoch": 0.4392857142857143, "grad_norm": 33.23506546020508, "learning_rate": 5.607142857142858e-06, "loss": 0.1274, "step": 861 }, { "epoch": 0.43979591836734694, "grad_norm": 29.233243942260742, "learning_rate": 5.602040816326531e-06, "loss": 0.1123, "step": 862 }, { "epoch": 0.4403061224489796, "grad_norm": 84.1875228881836, "learning_rate": 5.596938775510205e-06, "loss": 0.418, "step": 863 }, { "epoch": 0.44081632653061226, "grad_norm": 95.02442169189453, "learning_rate": 5.591836734693878e-06, "loss": 0.6426, "step": 864 }, { "epoch": 0.4413265306122449, "grad_norm": 14.344368934631348, "learning_rate": 5.586734693877551e-06, "loss": 0.0311, "step": 865 }, { "epoch": 0.4418367346938776, "grad_norm": 27.272762298583984, "learning_rate": 5.581632653061225e-06, "loss": 0.1038, "step": 866 }, { "epoch": 0.4423469387755102, "grad_norm": 116.32434844970703, "learning_rate": 5.576530612244898e-06, "loss": 0.5784, "step": 867 }, { "epoch": 0.44285714285714284, "grad_norm": 9.394695281982422, "learning_rate": 5.571428571428572e-06, "loss": 0.0227, "step": 868 }, { "epoch": 0.4433673469387755, "grad_norm": 54.688724517822266, "learning_rate": 5.566326530612245e-06, "loss": 0.1864, "step": 869 }, { "epoch": 0.44387755102040816, "grad_norm": 108.8348159790039, "learning_rate": 5.561224489795919e-06, "loss": 0.7016, "step": 870 }, { "epoch": 0.4443877551020408, "grad_norm": 109.46830749511719, "learning_rate": 5.556122448979592e-06, "loss": 0.5322, "step": 871 }, { "epoch": 0.4448979591836735, "grad_norm": 88.74755096435547, "learning_rate": 5.551020408163266e-06, "loss": 0.3572, "step": 872 }, { "epoch": 0.44540816326530613, "grad_norm": 11.095545768737793, "learning_rate": 5.5459183673469396e-06, "loss": 0.0179, "step": 873 }, { "epoch": 0.4459183673469388, "grad_norm": 45.457515716552734, "learning_rate": 5.5408163265306125e-06, "loss": 0.1835, "step": 874 }, { "epoch": 0.44642857142857145, "grad_norm": 61.88081741333008, "learning_rate": 5.535714285714286e-06, "loss": 0.392, "step": 875 }, { "epoch": 0.44693877551020406, "grad_norm": 15.803656578063965, "learning_rate": 5.530612244897959e-06, "loss": 0.0154, "step": 876 }, { "epoch": 0.4474489795918367, "grad_norm": 37.131622314453125, "learning_rate": 5.525510204081634e-06, "loss": 0.1788, "step": 877 }, { "epoch": 0.4479591836734694, "grad_norm": 7.437021732330322, "learning_rate": 5.520408163265306e-06, "loss": 0.0113, "step": 878 }, { "epoch": 0.44846938775510203, "grad_norm": 191.79010009765625, "learning_rate": 5.51530612244898e-06, "loss": 0.8258, "step": 879 }, { "epoch": 0.4489795918367347, "grad_norm": 48.643402099609375, "learning_rate": 5.510204081632653e-06, "loss": 0.2729, "step": 880 }, { "epoch": 0.44948979591836735, "grad_norm": 52.35139083862305, "learning_rate": 5.505102040816327e-06, "loss": 0.1257, "step": 881 }, { "epoch": 0.45, "grad_norm": 47.52025604248047, "learning_rate": 5.500000000000001e-06, "loss": 0.1797, "step": 882 }, { "epoch": 0.45051020408163267, "grad_norm": 118.15834045410156, "learning_rate": 5.494897959183674e-06, "loss": 0.5996, "step": 883 }, { "epoch": 0.45102040816326533, "grad_norm": 81.12723541259766, "learning_rate": 5.489795918367347e-06, "loss": 0.2935, "step": 884 }, { "epoch": 0.45153061224489793, "grad_norm": 2.23016357421875, "learning_rate": 5.48469387755102e-06, "loss": 0.0049, "step": 885 }, { "epoch": 0.4520408163265306, "grad_norm": 29.23896026611328, "learning_rate": 5.479591836734695e-06, "loss": 0.0527, "step": 886 }, { "epoch": 0.45255102040816325, "grad_norm": 192.8160400390625, "learning_rate": 5.474489795918368e-06, "loss": 1.8956, "step": 887 }, { "epoch": 0.4530612244897959, "grad_norm": 16.27593421936035, "learning_rate": 5.4693877551020415e-06, "loss": 0.0397, "step": 888 }, { "epoch": 0.45357142857142857, "grad_norm": 23.059371948242188, "learning_rate": 5.464285714285714e-06, "loss": 0.0492, "step": 889 }, { "epoch": 0.45408163265306123, "grad_norm": 165.79234313964844, "learning_rate": 5.459183673469388e-06, "loss": 0.7269, "step": 890 }, { "epoch": 0.4545918367346939, "grad_norm": 78.9997329711914, "learning_rate": 5.454081632653062e-06, "loss": 0.3845, "step": 891 }, { "epoch": 0.45510204081632655, "grad_norm": 57.4855842590332, "learning_rate": 5.448979591836735e-06, "loss": 0.2894, "step": 892 }, { "epoch": 0.4556122448979592, "grad_norm": 78.73365783691406, "learning_rate": 5.443877551020409e-06, "loss": 0.3291, "step": 893 }, { "epoch": 0.4561224489795918, "grad_norm": 57.83993911743164, "learning_rate": 5.438775510204081e-06, "loss": 0.1387, "step": 894 }, { "epoch": 0.45663265306122447, "grad_norm": 64.99560546875, "learning_rate": 5.433673469387756e-06, "loss": 0.3547, "step": 895 }, { "epoch": 0.45714285714285713, "grad_norm": 49.238773345947266, "learning_rate": 5.428571428571429e-06, "loss": 0.2412, "step": 896 }, { "epoch": 0.4576530612244898, "grad_norm": 168.53558349609375, "learning_rate": 5.423469387755103e-06, "loss": 1.1951, "step": 897 }, { "epoch": 0.45816326530612245, "grad_norm": 16.0331974029541, "learning_rate": 5.4183673469387755e-06, "loss": 0.0322, "step": 898 }, { "epoch": 0.4586734693877551, "grad_norm": 91.76226043701172, "learning_rate": 5.413265306122449e-06, "loss": 0.4209, "step": 899 }, { "epoch": 0.45918367346938777, "grad_norm": 37.28987121582031, "learning_rate": 5.408163265306123e-06, "loss": 0.1172, "step": 900 }, { "epoch": 0.4596938775510204, "grad_norm": 121.7083511352539, "learning_rate": 5.403061224489796e-06, "loss": 0.7656, "step": 901 }, { "epoch": 0.4602040816326531, "grad_norm": 70.30885314941406, "learning_rate": 5.3979591836734705e-06, "loss": 0.1694, "step": 902 }, { "epoch": 0.4607142857142857, "grad_norm": 48.03559494018555, "learning_rate": 5.392857142857143e-06, "loss": 0.0986, "step": 903 }, { "epoch": 0.46122448979591835, "grad_norm": 30.67242431640625, "learning_rate": 5.387755102040817e-06, "loss": 0.0631, "step": 904 }, { "epoch": 0.461734693877551, "grad_norm": 39.45822525024414, "learning_rate": 5.38265306122449e-06, "loss": 0.2061, "step": 905 }, { "epoch": 0.46224489795918366, "grad_norm": 75.01351928710938, "learning_rate": 5.377551020408164e-06, "loss": 0.3905, "step": 906 }, { "epoch": 0.4627551020408163, "grad_norm": 25.438642501831055, "learning_rate": 5.372448979591837e-06, "loss": 0.1475, "step": 907 }, { "epoch": 0.463265306122449, "grad_norm": 49.262054443359375, "learning_rate": 5.36734693877551e-06, "loss": 0.1945, "step": 908 }, { "epoch": 0.46377551020408164, "grad_norm": 40.99546813964844, "learning_rate": 5.362244897959185e-06, "loss": 0.1846, "step": 909 }, { "epoch": 0.4642857142857143, "grad_norm": 66.46600341796875, "learning_rate": 5.357142857142857e-06, "loss": 0.2046, "step": 910 }, { "epoch": 0.46479591836734696, "grad_norm": 8.338855743408203, "learning_rate": 5.352040816326532e-06, "loss": 0.0084, "step": 911 }, { "epoch": 0.46530612244897956, "grad_norm": 18.362627029418945, "learning_rate": 5.3469387755102045e-06, "loss": 0.0364, "step": 912 }, { "epoch": 0.4658163265306122, "grad_norm": 99.8839111328125, "learning_rate": 5.341836734693878e-06, "loss": 0.5039, "step": 913 }, { "epoch": 0.4663265306122449, "grad_norm": 22.357343673706055, "learning_rate": 5.336734693877551e-06, "loss": 0.0894, "step": 914 }, { "epoch": 0.46683673469387754, "grad_norm": 75.112548828125, "learning_rate": 5.331632653061225e-06, "loss": 0.2759, "step": 915 }, { "epoch": 0.4673469387755102, "grad_norm": 22.462825775146484, "learning_rate": 5.326530612244898e-06, "loss": 0.0388, "step": 916 }, { "epoch": 0.46785714285714286, "grad_norm": 247.50254821777344, "learning_rate": 5.3214285714285715e-06, "loss": 1.7466, "step": 917 }, { "epoch": 0.4683673469387755, "grad_norm": 14.881189346313477, "learning_rate": 5.316326530612246e-06, "loss": 0.0258, "step": 918 }, { "epoch": 0.4688775510204082, "grad_norm": 166.11410522460938, "learning_rate": 5.311224489795919e-06, "loss": 1.5039, "step": 919 }, { "epoch": 0.46938775510204084, "grad_norm": 28.569332122802734, "learning_rate": 5.306122448979593e-06, "loss": 0.0516, "step": 920 }, { "epoch": 0.4698979591836735, "grad_norm": 81.61792755126953, "learning_rate": 5.301020408163266e-06, "loss": 0.5024, "step": 921 }, { "epoch": 0.4704081632653061, "grad_norm": 38.372100830078125, "learning_rate": 5.295918367346939e-06, "loss": 0.0957, "step": 922 }, { "epoch": 0.47091836734693876, "grad_norm": 55.27342224121094, "learning_rate": 5.290816326530612e-06, "loss": 0.2822, "step": 923 }, { "epoch": 0.4714285714285714, "grad_norm": 41.712120056152344, "learning_rate": 5.285714285714286e-06, "loss": 0.174, "step": 924 }, { "epoch": 0.4719387755102041, "grad_norm": 32.05448913574219, "learning_rate": 5.280612244897959e-06, "loss": 0.114, "step": 925 }, { "epoch": 0.47244897959183674, "grad_norm": 24.655641555786133, "learning_rate": 5.275510204081633e-06, "loss": 0.0673, "step": 926 }, { "epoch": 0.4729591836734694, "grad_norm": 17.1375789642334, "learning_rate": 5.270408163265307e-06, "loss": 0.0634, "step": 927 }, { "epoch": 0.47346938775510206, "grad_norm": 17.208967208862305, "learning_rate": 5.26530612244898e-06, "loss": 0.0232, "step": 928 }, { "epoch": 0.4739795918367347, "grad_norm": 37.40412521362305, "learning_rate": 5.260204081632654e-06, "loss": 0.1777, "step": 929 }, { "epoch": 0.4744897959183674, "grad_norm": 23.040136337280273, "learning_rate": 5.255102040816327e-06, "loss": 0.033, "step": 930 }, { "epoch": 0.475, "grad_norm": 24.881595611572266, "learning_rate": 5.2500000000000006e-06, "loss": 0.0466, "step": 931 }, { "epoch": 0.47551020408163264, "grad_norm": 101.77995300292969, "learning_rate": 5.2448979591836735e-06, "loss": 0.416, "step": 932 }, { "epoch": 0.4760204081632653, "grad_norm": 96.30113983154297, "learning_rate": 5.239795918367347e-06, "loss": 0.6919, "step": 933 }, { "epoch": 0.47653061224489796, "grad_norm": 34.56840133666992, "learning_rate": 5.23469387755102e-06, "loss": 0.0771, "step": 934 }, { "epoch": 0.4770408163265306, "grad_norm": 61.37550354003906, "learning_rate": 5.229591836734695e-06, "loss": 0.2388, "step": 935 }, { "epoch": 0.4775510204081633, "grad_norm": 13.98636531829834, "learning_rate": 5.2244897959183684e-06, "loss": 0.0554, "step": 936 }, { "epoch": 0.47806122448979593, "grad_norm": 42.99785232543945, "learning_rate": 5.219387755102041e-06, "loss": 0.0825, "step": 937 }, { "epoch": 0.4785714285714286, "grad_norm": 33.697265625, "learning_rate": 5.214285714285715e-06, "loss": 0.0588, "step": 938 }, { "epoch": 0.47908163265306125, "grad_norm": 152.74156188964844, "learning_rate": 5.209183673469388e-06, "loss": 0.7107, "step": 939 }, { "epoch": 0.47959183673469385, "grad_norm": 52.04684066772461, "learning_rate": 5.204081632653062e-06, "loss": 0.1452, "step": 940 }, { "epoch": 0.4801020408163265, "grad_norm": 63.641693115234375, "learning_rate": 5.198979591836735e-06, "loss": 0.2371, "step": 941 }, { "epoch": 0.4806122448979592, "grad_norm": 142.92893981933594, "learning_rate": 5.193877551020408e-06, "loss": 0.6191, "step": 942 }, { "epoch": 0.48112244897959183, "grad_norm": 75.175537109375, "learning_rate": 5.188775510204081e-06, "loss": 0.2529, "step": 943 }, { "epoch": 0.4816326530612245, "grad_norm": 135.66404724121094, "learning_rate": 5.183673469387756e-06, "loss": 1.0938, "step": 944 }, { "epoch": 0.48214285714285715, "grad_norm": 75.39822387695312, "learning_rate": 5.1785714285714296e-06, "loss": 0.4828, "step": 945 }, { "epoch": 0.4826530612244898, "grad_norm": 99.71654510498047, "learning_rate": 5.1734693877551025e-06, "loss": 0.4419, "step": 946 }, { "epoch": 0.48316326530612247, "grad_norm": 27.568586349487305, "learning_rate": 5.168367346938776e-06, "loss": 0.0454, "step": 947 }, { "epoch": 0.48367346938775513, "grad_norm": 54.348358154296875, "learning_rate": 5.163265306122449e-06, "loss": 0.2104, "step": 948 }, { "epoch": 0.48418367346938773, "grad_norm": 26.25860595703125, "learning_rate": 5.158163265306123e-06, "loss": 0.0721, "step": 949 }, { "epoch": 0.4846938775510204, "grad_norm": 38.578060150146484, "learning_rate": 5.153061224489796e-06, "loss": 0.1436, "step": 950 }, { "epoch": 0.48520408163265305, "grad_norm": 37.945186614990234, "learning_rate": 5.14795918367347e-06, "loss": 0.0752, "step": 951 }, { "epoch": 0.4857142857142857, "grad_norm": 90.31257629394531, "learning_rate": 5.142857142857142e-06, "loss": 0.5676, "step": 952 }, { "epoch": 0.48622448979591837, "grad_norm": 42.869964599609375, "learning_rate": 5.137755102040817e-06, "loss": 0.1455, "step": 953 }, { "epoch": 0.48673469387755103, "grad_norm": 78.92963409423828, "learning_rate": 5.132653061224491e-06, "loss": 0.4077, "step": 954 }, { "epoch": 0.4872448979591837, "grad_norm": 146.7748260498047, "learning_rate": 5.127551020408164e-06, "loss": 0.884, "step": 955 }, { "epoch": 0.48775510204081635, "grad_norm": 49.687557220458984, "learning_rate": 5.122448979591837e-06, "loss": 0.1445, "step": 956 }, { "epoch": 0.488265306122449, "grad_norm": 38.43147277832031, "learning_rate": 5.11734693877551e-06, "loss": 0.1981, "step": 957 }, { "epoch": 0.4887755102040816, "grad_norm": 37.470951080322266, "learning_rate": 5.112244897959184e-06, "loss": 0.1487, "step": 958 }, { "epoch": 0.48928571428571427, "grad_norm": 69.49468231201172, "learning_rate": 5.107142857142857e-06, "loss": 0.3364, "step": 959 }, { "epoch": 0.4897959183673469, "grad_norm": 30.60089111328125, "learning_rate": 5.1020408163265315e-06, "loss": 0.0789, "step": 960 }, { "epoch": 0.4903061224489796, "grad_norm": 60.42655944824219, "learning_rate": 5.096938775510204e-06, "loss": 0.1979, "step": 961 }, { "epoch": 0.49081632653061225, "grad_norm": 62.25179672241211, "learning_rate": 5.091836734693878e-06, "loss": 0.2146, "step": 962 }, { "epoch": 0.4913265306122449, "grad_norm": 47.38945007324219, "learning_rate": 5.086734693877552e-06, "loss": 0.1206, "step": 963 }, { "epoch": 0.49183673469387756, "grad_norm": 30.31964874267578, "learning_rate": 5.081632653061225e-06, "loss": 0.1284, "step": 964 }, { "epoch": 0.4923469387755102, "grad_norm": 128.3580322265625, "learning_rate": 5.0765306122448985e-06, "loss": 0.5518, "step": 965 }, { "epoch": 0.4928571428571429, "grad_norm": 40.96641540527344, "learning_rate": 5.071428571428571e-06, "loss": 0.0874, "step": 966 }, { "epoch": 0.4933673469387755, "grad_norm": 34.32910919189453, "learning_rate": 5.066326530612245e-06, "loss": 0.0822, "step": 967 }, { "epoch": 0.49387755102040815, "grad_norm": 35.15664291381836, "learning_rate": 5.061224489795918e-06, "loss": 0.1233, "step": 968 }, { "epoch": 0.4943877551020408, "grad_norm": 69.57056427001953, "learning_rate": 5.056122448979593e-06, "loss": 0.4414, "step": 969 }, { "epoch": 0.49489795918367346, "grad_norm": 59.03541564941406, "learning_rate": 5.0510204081632655e-06, "loss": 0.1851, "step": 970 }, { "epoch": 0.4954081632653061, "grad_norm": 73.79888916015625, "learning_rate": 5.045918367346939e-06, "loss": 0.5439, "step": 971 }, { "epoch": 0.4959183673469388, "grad_norm": 91.31954956054688, "learning_rate": 5.040816326530613e-06, "loss": 0.3561, "step": 972 }, { "epoch": 0.49642857142857144, "grad_norm": 47.51620864868164, "learning_rate": 5.035714285714286e-06, "loss": 0.2305, "step": 973 }, { "epoch": 0.4969387755102041, "grad_norm": 19.094947814941406, "learning_rate": 5.03061224489796e-06, "loss": 0.0503, "step": 974 }, { "epoch": 0.49744897959183676, "grad_norm": 40.33220672607422, "learning_rate": 5.0255102040816325e-06, "loss": 0.1042, "step": 975 }, { "epoch": 0.49795918367346936, "grad_norm": 30.916818618774414, "learning_rate": 5.020408163265307e-06, "loss": 0.1012, "step": 976 }, { "epoch": 0.498469387755102, "grad_norm": 40.604496002197266, "learning_rate": 5.01530612244898e-06, "loss": 0.144, "step": 977 }, { "epoch": 0.4989795918367347, "grad_norm": 28.227510452270508, "learning_rate": 5.010204081632654e-06, "loss": 0.1111, "step": 978 }, { "epoch": 0.49948979591836734, "grad_norm": 29.846145629882812, "learning_rate": 5.005102040816327e-06, "loss": 0.0419, "step": 979 }, { "epoch": 0.5, "grad_norm": 91.87031555175781, "learning_rate": 5e-06, "loss": 0.4995, "step": 980 }, { "epoch": 0.5005102040816326, "grad_norm": 98.25852966308594, "learning_rate": 4.994897959183674e-06, "loss": 0.4189, "step": 981 }, { "epoch": 0.5010204081632653, "grad_norm": 70.63806915283203, "learning_rate": 4.989795918367347e-06, "loss": 0.2546, "step": 982 }, { "epoch": 0.5015306122448979, "grad_norm": 36.56729507446289, "learning_rate": 4.984693877551021e-06, "loss": 0.1519, "step": 983 }, { "epoch": 0.5020408163265306, "grad_norm": 26.349708557128906, "learning_rate": 4.979591836734694e-06, "loss": 0.037, "step": 984 }, { "epoch": 0.5025510204081632, "grad_norm": 134.0643768310547, "learning_rate": 4.974489795918368e-06, "loss": 0.6069, "step": 985 }, { "epoch": 0.503061224489796, "grad_norm": 122.48017120361328, "learning_rate": 4.969387755102041e-06, "loss": 0.7402, "step": 986 }, { "epoch": 0.5035714285714286, "grad_norm": 40.79277420043945, "learning_rate": 4.964285714285715e-06, "loss": 0.1929, "step": 987 }, { "epoch": 0.5040816326530613, "grad_norm": 60.614173889160156, "learning_rate": 4.959183673469388e-06, "loss": 0.2266, "step": 988 }, { "epoch": 0.5045918367346939, "grad_norm": 57.49152755737305, "learning_rate": 4.9540816326530616e-06, "loss": 0.1545, "step": 989 }, { "epoch": 0.5051020408163265, "grad_norm": 28.051353454589844, "learning_rate": 4.948979591836735e-06, "loss": 0.0817, "step": 990 }, { "epoch": 0.5056122448979592, "grad_norm": 185.3383026123047, "learning_rate": 4.943877551020408e-06, "loss": 1.1863, "step": 991 }, { "epoch": 0.5061224489795918, "grad_norm": 42.14871597290039, "learning_rate": 4.938775510204082e-06, "loss": 0.0894, "step": 992 }, { "epoch": 0.5066326530612245, "grad_norm": 99.87521362304688, "learning_rate": 4.933673469387756e-06, "loss": 0.7246, "step": 993 }, { "epoch": 0.5071428571428571, "grad_norm": 108.5139389038086, "learning_rate": 4.928571428571429e-06, "loss": 0.3535, "step": 994 }, { "epoch": 0.5076530612244898, "grad_norm": 29.921340942382812, "learning_rate": 4.923469387755102e-06, "loss": 0.0505, "step": 995 }, { "epoch": 0.5081632653061224, "grad_norm": 38.35554885864258, "learning_rate": 4.918367346938776e-06, "loss": 0.1611, "step": 996 }, { "epoch": 0.5086734693877552, "grad_norm": 23.304006576538086, "learning_rate": 4.91326530612245e-06, "loss": 0.0864, "step": 997 }, { "epoch": 0.5091836734693878, "grad_norm": 9.254980087280273, "learning_rate": 4.908163265306123e-06, "loss": 0.0209, "step": 998 }, { "epoch": 0.5096938775510204, "grad_norm": 90.1374740600586, "learning_rate": 4.9030612244897964e-06, "loss": 0.2637, "step": 999 }, { "epoch": 0.5102040816326531, "grad_norm": 10.114912033081055, "learning_rate": 4.897959183673469e-06, "loss": 0.0303, "step": 1000 }, { "epoch": 0.5107142857142857, "grad_norm": 48.52086639404297, "learning_rate": 4.892857142857143e-06, "loss": 0.2069, "step": 1001 }, { "epoch": 0.5112244897959184, "grad_norm": 7.190910339355469, "learning_rate": 4.887755102040817e-06, "loss": 0.0174, "step": 1002 }, { "epoch": 0.511734693877551, "grad_norm": 32.72443389892578, "learning_rate": 4.8826530612244906e-06, "loss": 0.1934, "step": 1003 }, { "epoch": 0.5122448979591837, "grad_norm": 29.396038055419922, "learning_rate": 4.8775510204081635e-06, "loss": 0.0801, "step": 1004 }, { "epoch": 0.5127551020408163, "grad_norm": 67.30844116210938, "learning_rate": 4.872448979591837e-06, "loss": 0.2212, "step": 1005 }, { "epoch": 0.513265306122449, "grad_norm": 54.95892333984375, "learning_rate": 4.867346938775511e-06, "loss": 0.1591, "step": 1006 }, { "epoch": 0.5137755102040816, "grad_norm": 23.679336547851562, "learning_rate": 4.862244897959184e-06, "loss": 0.0727, "step": 1007 }, { "epoch": 0.5142857142857142, "grad_norm": 33.733333587646484, "learning_rate": 4.857142857142858e-06, "loss": 0.1157, "step": 1008 }, { "epoch": 0.514795918367347, "grad_norm": 76.26932525634766, "learning_rate": 4.8520408163265305e-06, "loss": 0.5298, "step": 1009 }, { "epoch": 0.5153061224489796, "grad_norm": 78.60010528564453, "learning_rate": 4.846938775510204e-06, "loss": 0.3364, "step": 1010 }, { "epoch": 0.5158163265306123, "grad_norm": 131.48106384277344, "learning_rate": 4.841836734693878e-06, "loss": 0.6903, "step": 1011 }, { "epoch": 0.5163265306122449, "grad_norm": 12.410789489746094, "learning_rate": 4.836734693877552e-06, "loss": 0.0382, "step": 1012 }, { "epoch": 0.5168367346938776, "grad_norm": 38.29351043701172, "learning_rate": 4.8316326530612255e-06, "loss": 0.1652, "step": 1013 }, { "epoch": 0.5173469387755102, "grad_norm": 54.62494659423828, "learning_rate": 4.826530612244898e-06, "loss": 0.2568, "step": 1014 }, { "epoch": 0.5178571428571429, "grad_norm": 34.04876708984375, "learning_rate": 4.821428571428572e-06, "loss": 0.1274, "step": 1015 }, { "epoch": 0.5183673469387755, "grad_norm": 17.40114402770996, "learning_rate": 4.816326530612245e-06, "loss": 0.0432, "step": 1016 }, { "epoch": 0.5188775510204081, "grad_norm": 12.665262222290039, "learning_rate": 4.811224489795919e-06, "loss": 0.0102, "step": 1017 }, { "epoch": 0.5193877551020408, "grad_norm": 26.640396118164062, "learning_rate": 4.8061224489795925e-06, "loss": 0.0427, "step": 1018 }, { "epoch": 0.5198979591836734, "grad_norm": 9.97052001953125, "learning_rate": 4.801020408163265e-06, "loss": 0.0221, "step": 1019 }, { "epoch": 0.5204081632653061, "grad_norm": 124.32201385498047, "learning_rate": 4.795918367346939e-06, "loss": 0.499, "step": 1020 }, { "epoch": 0.5209183673469387, "grad_norm": 96.84408569335938, "learning_rate": 4.790816326530613e-06, "loss": 0.377, "step": 1021 }, { "epoch": 0.5214285714285715, "grad_norm": 80.85256958007812, "learning_rate": 4.785714285714287e-06, "loss": 0.3945, "step": 1022 }, { "epoch": 0.5219387755102041, "grad_norm": 34.850059509277344, "learning_rate": 4.7806122448979595e-06, "loss": 0.1747, "step": 1023 }, { "epoch": 0.5224489795918368, "grad_norm": 22.001426696777344, "learning_rate": 4.775510204081633e-06, "loss": 0.0903, "step": 1024 }, { "epoch": 0.5229591836734694, "grad_norm": 59.71054458618164, "learning_rate": 4.770408163265306e-06, "loss": 0.1785, "step": 1025 }, { "epoch": 0.523469387755102, "grad_norm": 23.64380645751953, "learning_rate": 4.76530612244898e-06, "loss": 0.046, "step": 1026 }, { "epoch": 0.5239795918367347, "grad_norm": 34.67354965209961, "learning_rate": 4.760204081632654e-06, "loss": 0.155, "step": 1027 }, { "epoch": 0.5244897959183673, "grad_norm": 30.6621150970459, "learning_rate": 4.7551020408163265e-06, "loss": 0.0651, "step": 1028 }, { "epoch": 0.525, "grad_norm": 70.2184829711914, "learning_rate": 4.75e-06, "loss": 0.2809, "step": 1029 }, { "epoch": 0.5255102040816326, "grad_norm": 111.02610778808594, "learning_rate": 4.744897959183674e-06, "loss": 0.4553, "step": 1030 }, { "epoch": 0.5260204081632653, "grad_norm": 75.04460144042969, "learning_rate": 4.739795918367348e-06, "loss": 0.2954, "step": 1031 }, { "epoch": 0.5265306122448979, "grad_norm": 53.26376724243164, "learning_rate": 4.734693877551021e-06, "loss": 0.1485, "step": 1032 }, { "epoch": 0.5270408163265307, "grad_norm": 41.53818130493164, "learning_rate": 4.729591836734694e-06, "loss": 0.1733, "step": 1033 }, { "epoch": 0.5275510204081633, "grad_norm": 41.35334777832031, "learning_rate": 4.724489795918368e-06, "loss": 0.0819, "step": 1034 }, { "epoch": 0.5280612244897959, "grad_norm": 66.74555206298828, "learning_rate": 4.719387755102041e-06, "loss": 0.292, "step": 1035 }, { "epoch": 0.5285714285714286, "grad_norm": 47.67375183105469, "learning_rate": 4.714285714285715e-06, "loss": 0.3088, "step": 1036 }, { "epoch": 0.5290816326530612, "grad_norm": 20.527833938598633, "learning_rate": 4.709183673469388e-06, "loss": 0.0375, "step": 1037 }, { "epoch": 0.5295918367346939, "grad_norm": 35.292083740234375, "learning_rate": 4.704081632653061e-06, "loss": 0.0758, "step": 1038 }, { "epoch": 0.5301020408163265, "grad_norm": 65.31835174560547, "learning_rate": 4.698979591836735e-06, "loss": 0.1677, "step": 1039 }, { "epoch": 0.5306122448979592, "grad_norm": 66.1888198852539, "learning_rate": 4.693877551020409e-06, "loss": 0.3184, "step": 1040 }, { "epoch": 0.5311224489795918, "grad_norm": 57.02951431274414, "learning_rate": 4.688775510204082e-06, "loss": 0.1871, "step": 1041 }, { "epoch": 0.5316326530612245, "grad_norm": 65.1299057006836, "learning_rate": 4.6836734693877555e-06, "loss": 0.2379, "step": 1042 }, { "epoch": 0.5321428571428571, "grad_norm": 99.16394805908203, "learning_rate": 4.678571428571429e-06, "loss": 0.7177, "step": 1043 }, { "epoch": 0.5326530612244897, "grad_norm": 15.95533561706543, "learning_rate": 4.673469387755102e-06, "loss": 0.0355, "step": 1044 }, { "epoch": 0.5331632653061225, "grad_norm": 12.035353660583496, "learning_rate": 4.668367346938776e-06, "loss": 0.0198, "step": 1045 }, { "epoch": 0.5336734693877551, "grad_norm": 8.484695434570312, "learning_rate": 4.663265306122449e-06, "loss": 0.0118, "step": 1046 }, { "epoch": 0.5341836734693878, "grad_norm": 33.43607711791992, "learning_rate": 4.6581632653061226e-06, "loss": 0.1002, "step": 1047 }, { "epoch": 0.5346938775510204, "grad_norm": 20.48702621459961, "learning_rate": 4.653061224489796e-06, "loss": 0.0517, "step": 1048 }, { "epoch": 0.5352040816326531, "grad_norm": 26.089405059814453, "learning_rate": 4.64795918367347e-06, "loss": 0.0678, "step": 1049 }, { "epoch": 0.5357142857142857, "grad_norm": 182.61375427246094, "learning_rate": 4.642857142857144e-06, "loss": 0.9756, "step": 1050 }, { "epoch": 0.5362244897959184, "grad_norm": 12.384181022644043, "learning_rate": 4.637755102040817e-06, "loss": 0.027, "step": 1051 }, { "epoch": 0.536734693877551, "grad_norm": 150.1430206298828, "learning_rate": 4.63265306122449e-06, "loss": 1.12, "step": 1052 }, { "epoch": 0.5372448979591836, "grad_norm": 34.49362564086914, "learning_rate": 4.627551020408163e-06, "loss": 0.0596, "step": 1053 }, { "epoch": 0.5377551020408163, "grad_norm": 189.46942138671875, "learning_rate": 4.622448979591837e-06, "loss": 1.0016, "step": 1054 }, { "epoch": 0.5382653061224489, "grad_norm": 41.58380126953125, "learning_rate": 4.617346938775511e-06, "loss": 0.0896, "step": 1055 }, { "epoch": 0.5387755102040817, "grad_norm": 56.49480438232422, "learning_rate": 4.612244897959184e-06, "loss": 0.1696, "step": 1056 }, { "epoch": 0.5392857142857143, "grad_norm": 9.867877006530762, "learning_rate": 4.6071428571428574e-06, "loss": 0.0195, "step": 1057 }, { "epoch": 0.539795918367347, "grad_norm": 85.77963256835938, "learning_rate": 4.602040816326531e-06, "loss": 0.3362, "step": 1058 }, { "epoch": 0.5403061224489796, "grad_norm": 22.966876983642578, "learning_rate": 4.596938775510205e-06, "loss": 0.0436, "step": 1059 }, { "epoch": 0.5408163265306123, "grad_norm": 38.66813278198242, "learning_rate": 4.591836734693878e-06, "loss": 0.0979, "step": 1060 }, { "epoch": 0.5413265306122449, "grad_norm": 51.63981246948242, "learning_rate": 4.5867346938775516e-06, "loss": 0.1248, "step": 1061 }, { "epoch": 0.5418367346938775, "grad_norm": 130.71229553222656, "learning_rate": 4.5816326530612245e-06, "loss": 0.5098, "step": 1062 }, { "epoch": 0.5423469387755102, "grad_norm": 42.43650436401367, "learning_rate": 4.576530612244898e-06, "loss": 0.0862, "step": 1063 }, { "epoch": 0.5428571428571428, "grad_norm": 204.41929626464844, "learning_rate": 4.571428571428572e-06, "loss": 0.9092, "step": 1064 }, { "epoch": 0.5433673469387755, "grad_norm": 148.25765991210938, "learning_rate": 4.566326530612245e-06, "loss": 0.874, "step": 1065 }, { "epoch": 0.5438775510204081, "grad_norm": 18.85063934326172, "learning_rate": 4.561224489795919e-06, "loss": 0.037, "step": 1066 }, { "epoch": 0.5443877551020408, "grad_norm": 65.95673370361328, "learning_rate": 4.556122448979592e-06, "loss": 0.1582, "step": 1067 }, { "epoch": 0.5448979591836735, "grad_norm": 25.030733108520508, "learning_rate": 4.551020408163266e-06, "loss": 0.0599, "step": 1068 }, { "epoch": 0.5454081632653062, "grad_norm": 103.979248046875, "learning_rate": 4.545918367346939e-06, "loss": 0.4747, "step": 1069 }, { "epoch": 0.5459183673469388, "grad_norm": 37.320556640625, "learning_rate": 4.540816326530613e-06, "loss": 0.0808, "step": 1070 }, { "epoch": 0.5464285714285714, "grad_norm": 47.32221221923828, "learning_rate": 4.5357142857142865e-06, "loss": 0.1147, "step": 1071 }, { "epoch": 0.5469387755102041, "grad_norm": 34.7042350769043, "learning_rate": 4.530612244897959e-06, "loss": 0.1137, "step": 1072 }, { "epoch": 0.5474489795918367, "grad_norm": 22.95425796508789, "learning_rate": 4.525510204081633e-06, "loss": 0.085, "step": 1073 }, { "epoch": 0.5479591836734694, "grad_norm": 21.26638412475586, "learning_rate": 4.520408163265306e-06, "loss": 0.0706, "step": 1074 }, { "epoch": 0.548469387755102, "grad_norm": 71.8822250366211, "learning_rate": 4.51530612244898e-06, "loss": 0.178, "step": 1075 }, { "epoch": 0.5489795918367347, "grad_norm": 56.172264099121094, "learning_rate": 4.5102040816326535e-06, "loss": 0.1572, "step": 1076 }, { "epoch": 0.5494897959183673, "grad_norm": 120.4759292602539, "learning_rate": 4.505102040816327e-06, "loss": 0.3064, "step": 1077 }, { "epoch": 0.55, "grad_norm": 56.84187698364258, "learning_rate": 4.5e-06, "loss": 0.189, "step": 1078 }, { "epoch": 0.5505102040816326, "grad_norm": 41.006309509277344, "learning_rate": 4.494897959183674e-06, "loss": 0.1346, "step": 1079 }, { "epoch": 0.5510204081632653, "grad_norm": 137.96702575683594, "learning_rate": 4.489795918367348e-06, "loss": 0.8018, "step": 1080 }, { "epoch": 0.551530612244898, "grad_norm": 40.97859191894531, "learning_rate": 4.4846938775510205e-06, "loss": 0.1034, "step": 1081 }, { "epoch": 0.5520408163265306, "grad_norm": 94.38196563720703, "learning_rate": 4.479591836734694e-06, "loss": 0.291, "step": 1082 }, { "epoch": 0.5525510204081633, "grad_norm": 73.9909896850586, "learning_rate": 4.474489795918367e-06, "loss": 0.2222, "step": 1083 }, { "epoch": 0.5530612244897959, "grad_norm": 45.52335739135742, "learning_rate": 4.469387755102041e-06, "loss": 0.1763, "step": 1084 }, { "epoch": 0.5535714285714286, "grad_norm": 32.92928695678711, "learning_rate": 4.464285714285715e-06, "loss": 0.1438, "step": 1085 }, { "epoch": 0.5540816326530612, "grad_norm": 52.818668365478516, "learning_rate": 4.459183673469388e-06, "loss": 0.1612, "step": 1086 }, { "epoch": 0.5545918367346939, "grad_norm": 32.944515228271484, "learning_rate": 4.454081632653062e-06, "loss": 0.106, "step": 1087 }, { "epoch": 0.5551020408163265, "grad_norm": 47.35269546508789, "learning_rate": 4.448979591836735e-06, "loss": 0.167, "step": 1088 }, { "epoch": 0.5556122448979591, "grad_norm": 84.26695251464844, "learning_rate": 4.443877551020409e-06, "loss": 0.2997, "step": 1089 }, { "epoch": 0.5561224489795918, "grad_norm": 184.76657104492188, "learning_rate": 4.438775510204082e-06, "loss": 0.7729, "step": 1090 }, { "epoch": 0.5566326530612244, "grad_norm": 25.757083892822266, "learning_rate": 4.433673469387755e-06, "loss": 0.0889, "step": 1091 }, { "epoch": 0.5571428571428572, "grad_norm": 8.70840072631836, "learning_rate": 4.428571428571429e-06, "loss": 0.0171, "step": 1092 }, { "epoch": 0.5576530612244898, "grad_norm": 11.402214050292969, "learning_rate": 4.423469387755102e-06, "loss": 0.015, "step": 1093 }, { "epoch": 0.5581632653061225, "grad_norm": 39.23138427734375, "learning_rate": 4.418367346938776e-06, "loss": 0.0931, "step": 1094 }, { "epoch": 0.5586734693877551, "grad_norm": 78.50774383544922, "learning_rate": 4.4132653061224495e-06, "loss": 0.2761, "step": 1095 }, { "epoch": 0.5591836734693878, "grad_norm": 58.12727355957031, "learning_rate": 4.408163265306123e-06, "loss": 0.2581, "step": 1096 }, { "epoch": 0.5596938775510204, "grad_norm": 28.589599609375, "learning_rate": 4.403061224489796e-06, "loss": 0.0675, "step": 1097 }, { "epoch": 0.560204081632653, "grad_norm": 11.528457641601562, "learning_rate": 4.39795918367347e-06, "loss": 0.0194, "step": 1098 }, { "epoch": 0.5607142857142857, "grad_norm": 25.837549209594727, "learning_rate": 4.392857142857143e-06, "loss": 0.0448, "step": 1099 }, { "epoch": 0.5612244897959183, "grad_norm": 14.447965621948242, "learning_rate": 4.3877551020408165e-06, "loss": 0.0214, "step": 1100 }, { "epoch": 0.561734693877551, "grad_norm": 11.722237586975098, "learning_rate": 4.38265306122449e-06, "loss": 0.0113, "step": 1101 }, { "epoch": 0.5622448979591836, "grad_norm": 12.302326202392578, "learning_rate": 4.377551020408163e-06, "loss": 0.0077, "step": 1102 }, { "epoch": 0.5627551020408164, "grad_norm": 175.99769592285156, "learning_rate": 4.372448979591838e-06, "loss": 0.7064, "step": 1103 }, { "epoch": 0.563265306122449, "grad_norm": 19.69425392150879, "learning_rate": 4.367346938775511e-06, "loss": 0.0299, "step": 1104 }, { "epoch": 0.5637755102040817, "grad_norm": 26.145238876342773, "learning_rate": 4.362244897959184e-06, "loss": 0.0364, "step": 1105 }, { "epoch": 0.5642857142857143, "grad_norm": 34.93511199951172, "learning_rate": 4.357142857142857e-06, "loss": 0.0979, "step": 1106 }, { "epoch": 0.564795918367347, "grad_norm": 26.582805633544922, "learning_rate": 4.352040816326531e-06, "loss": 0.0677, "step": 1107 }, { "epoch": 0.5653061224489796, "grad_norm": 102.3194580078125, "learning_rate": 4.346938775510205e-06, "loss": 0.3367, "step": 1108 }, { "epoch": 0.5658163265306122, "grad_norm": 27.020679473876953, "learning_rate": 4.341836734693878e-06, "loss": 0.0345, "step": 1109 }, { "epoch": 0.5663265306122449, "grad_norm": 84.52767944335938, "learning_rate": 4.336734693877551e-06, "loss": 0.5693, "step": 1110 }, { "epoch": 0.5668367346938775, "grad_norm": 75.42728424072266, "learning_rate": 4.331632653061224e-06, "loss": 0.3511, "step": 1111 }, { "epoch": 0.5673469387755102, "grad_norm": 63.397884368896484, "learning_rate": 4.326530612244899e-06, "loss": 0.2388, "step": 1112 }, { "epoch": 0.5678571428571428, "grad_norm": 107.91397094726562, "learning_rate": 4.321428571428572e-06, "loss": 0.2758, "step": 1113 }, { "epoch": 0.5683673469387756, "grad_norm": 52.15452575683594, "learning_rate": 4.3163265306122455e-06, "loss": 0.1657, "step": 1114 }, { "epoch": 0.5688775510204082, "grad_norm": 25.304603576660156, "learning_rate": 4.3112244897959184e-06, "loss": 0.0825, "step": 1115 }, { "epoch": 0.5693877551020409, "grad_norm": 36.77111053466797, "learning_rate": 4.306122448979592e-06, "loss": 0.1145, "step": 1116 }, { "epoch": 0.5698979591836735, "grad_norm": 30.951614379882812, "learning_rate": 4.301020408163266e-06, "loss": 0.0869, "step": 1117 }, { "epoch": 0.5704081632653061, "grad_norm": 187.9287872314453, "learning_rate": 4.295918367346939e-06, "loss": 0.6621, "step": 1118 }, { "epoch": 0.5709183673469388, "grad_norm": 15.91569995880127, "learning_rate": 4.2908163265306126e-06, "loss": 0.0126, "step": 1119 }, { "epoch": 0.5714285714285714, "grad_norm": 52.37792205810547, "learning_rate": 4.2857142857142855e-06, "loss": 0.1482, "step": 1120 }, { "epoch": 0.5719387755102041, "grad_norm": 42.59999465942383, "learning_rate": 4.28061224489796e-06, "loss": 0.0935, "step": 1121 }, { "epoch": 0.5724489795918367, "grad_norm": 10.453607559204102, "learning_rate": 4.275510204081633e-06, "loss": 0.0231, "step": 1122 }, { "epoch": 0.5729591836734694, "grad_norm": 196.1283721923828, "learning_rate": 4.270408163265307e-06, "loss": 0.6196, "step": 1123 }, { "epoch": 0.573469387755102, "grad_norm": 58.997623443603516, "learning_rate": 4.2653061224489804e-06, "loss": 0.146, "step": 1124 }, { "epoch": 0.5739795918367347, "grad_norm": 78.53055572509766, "learning_rate": 4.260204081632653e-06, "loss": 0.4459, "step": 1125 }, { "epoch": 0.5744897959183674, "grad_norm": 21.07754135131836, "learning_rate": 4.255102040816327e-06, "loss": 0.0402, "step": 1126 }, { "epoch": 0.575, "grad_norm": 14.86988639831543, "learning_rate": 4.25e-06, "loss": 0.0193, "step": 1127 }, { "epoch": 0.5755102040816327, "grad_norm": 209.4464569091797, "learning_rate": 4.244897959183674e-06, "loss": 1.2529, "step": 1128 }, { "epoch": 0.5760204081632653, "grad_norm": 77.81645202636719, "learning_rate": 4.2397959183673475e-06, "loss": 0.3547, "step": 1129 }, { "epoch": 0.576530612244898, "grad_norm": 52.24641418457031, "learning_rate": 4.234693877551021e-06, "loss": 0.3387, "step": 1130 }, { "epoch": 0.5770408163265306, "grad_norm": 100.74240112304688, "learning_rate": 4.229591836734694e-06, "loss": 0.25, "step": 1131 }, { "epoch": 0.5775510204081633, "grad_norm": 52.4826774597168, "learning_rate": 4.224489795918368e-06, "loss": 0.1359, "step": 1132 }, { "epoch": 0.5780612244897959, "grad_norm": 80.50521087646484, "learning_rate": 4.219387755102042e-06, "loss": 0.2856, "step": 1133 }, { "epoch": 0.5785714285714286, "grad_norm": 18.94988250732422, "learning_rate": 4.2142857142857145e-06, "loss": 0.0332, "step": 1134 }, { "epoch": 0.5790816326530612, "grad_norm": 195.42677307128906, "learning_rate": 4.209183673469388e-06, "loss": 0.6992, "step": 1135 }, { "epoch": 0.5795918367346938, "grad_norm": 74.5130615234375, "learning_rate": 4.204081632653061e-06, "loss": 0.1548, "step": 1136 }, { "epoch": 0.5801020408163265, "grad_norm": 120.46251678466797, "learning_rate": 4.198979591836735e-06, "loss": 0.5366, "step": 1137 }, { "epoch": 0.5806122448979592, "grad_norm": 34.89018249511719, "learning_rate": 4.193877551020409e-06, "loss": 0.0616, "step": 1138 }, { "epoch": 0.5811224489795919, "grad_norm": 12.079283714294434, "learning_rate": 4.188775510204082e-06, "loss": 0.0233, "step": 1139 }, { "epoch": 0.5816326530612245, "grad_norm": 42.169036865234375, "learning_rate": 4.183673469387755e-06, "loss": 0.1248, "step": 1140 }, { "epoch": 0.5821428571428572, "grad_norm": 26.696977615356445, "learning_rate": 4.178571428571429e-06, "loss": 0.0823, "step": 1141 }, { "epoch": 0.5826530612244898, "grad_norm": 26.527971267700195, "learning_rate": 4.173469387755103e-06, "loss": 0.0232, "step": 1142 }, { "epoch": 0.5831632653061225, "grad_norm": 88.48262023925781, "learning_rate": 4.168367346938776e-06, "loss": 0.4366, "step": 1143 }, { "epoch": 0.5836734693877551, "grad_norm": 30.39879035949707, "learning_rate": 4.163265306122449e-06, "loss": 0.0889, "step": 1144 }, { "epoch": 0.5841836734693877, "grad_norm": 35.27324295043945, "learning_rate": 4.158163265306123e-06, "loss": 0.155, "step": 1145 }, { "epoch": 0.5846938775510204, "grad_norm": 35.58400344848633, "learning_rate": 4.153061224489796e-06, "loss": 0.098, "step": 1146 }, { "epoch": 0.585204081632653, "grad_norm": 50.426151275634766, "learning_rate": 4.14795918367347e-06, "loss": 0.2729, "step": 1147 }, { "epoch": 0.5857142857142857, "grad_norm": 15.953079223632812, "learning_rate": 4.1428571428571435e-06, "loss": 0.0475, "step": 1148 }, { "epoch": 0.5862244897959183, "grad_norm": 114.388671875, "learning_rate": 4.137755102040817e-06, "loss": 0.4231, "step": 1149 }, { "epoch": 0.5867346938775511, "grad_norm": 73.48621368408203, "learning_rate": 4.13265306122449e-06, "loss": 0.28, "step": 1150 }, { "epoch": 0.5872448979591837, "grad_norm": 27.3116455078125, "learning_rate": 4.127551020408164e-06, "loss": 0.0632, "step": 1151 }, { "epoch": 0.5877551020408164, "grad_norm": 23.819259643554688, "learning_rate": 4.122448979591837e-06, "loss": 0.085, "step": 1152 }, { "epoch": 0.588265306122449, "grad_norm": 16.269594192504883, "learning_rate": 4.1173469387755105e-06, "loss": 0.0487, "step": 1153 }, { "epoch": 0.5887755102040816, "grad_norm": 98.25843811035156, "learning_rate": 4.112244897959184e-06, "loss": 0.5981, "step": 1154 }, { "epoch": 0.5892857142857143, "grad_norm": 40.647159576416016, "learning_rate": 4.107142857142857e-06, "loss": 0.0819, "step": 1155 }, { "epoch": 0.5897959183673469, "grad_norm": 88.21232604980469, "learning_rate": 4.102040816326531e-06, "loss": 0.2599, "step": 1156 }, { "epoch": 0.5903061224489796, "grad_norm": 22.803922653198242, "learning_rate": 4.096938775510205e-06, "loss": 0.0442, "step": 1157 }, { "epoch": 0.5908163265306122, "grad_norm": 191.99766540527344, "learning_rate": 4.091836734693878e-06, "loss": 0.7539, "step": 1158 }, { "epoch": 0.5913265306122449, "grad_norm": 33.14800262451172, "learning_rate": 4.086734693877551e-06, "loss": 0.1942, "step": 1159 }, { "epoch": 0.5918367346938775, "grad_norm": 16.813562393188477, "learning_rate": 4.081632653061225e-06, "loss": 0.0172, "step": 1160 }, { "epoch": 0.5923469387755103, "grad_norm": 22.4349422454834, "learning_rate": 4.076530612244898e-06, "loss": 0.094, "step": 1161 }, { "epoch": 0.5928571428571429, "grad_norm": 27.382585525512695, "learning_rate": 4.071428571428572e-06, "loss": 0.0443, "step": 1162 }, { "epoch": 0.5933673469387755, "grad_norm": 31.349393844604492, "learning_rate": 4.066326530612245e-06, "loss": 0.0402, "step": 1163 }, { "epoch": 0.5938775510204082, "grad_norm": 16.04868507385254, "learning_rate": 4.061224489795918e-06, "loss": 0.0311, "step": 1164 }, { "epoch": 0.5943877551020408, "grad_norm": 70.97560119628906, "learning_rate": 4.056122448979592e-06, "loss": 0.4067, "step": 1165 }, { "epoch": 0.5948979591836735, "grad_norm": 59.251224517822266, "learning_rate": 4.051020408163266e-06, "loss": 0.1772, "step": 1166 }, { "epoch": 0.5954081632653061, "grad_norm": 117.09127807617188, "learning_rate": 4.0459183673469395e-06, "loss": 0.395, "step": 1167 }, { "epoch": 0.5959183673469388, "grad_norm": 15.023975372314453, "learning_rate": 4.040816326530612e-06, "loss": 0.028, "step": 1168 }, { "epoch": 0.5964285714285714, "grad_norm": 18.057344436645508, "learning_rate": 4.035714285714286e-06, "loss": 0.0771, "step": 1169 }, { "epoch": 0.5969387755102041, "grad_norm": 77.79132080078125, "learning_rate": 4.03061224489796e-06, "loss": 0.4722, "step": 1170 }, { "epoch": 0.5974489795918367, "grad_norm": 35.494808197021484, "learning_rate": 4.025510204081633e-06, "loss": 0.0722, "step": 1171 }, { "epoch": 0.5979591836734693, "grad_norm": 48.964759826660156, "learning_rate": 4.0204081632653065e-06, "loss": 0.1946, "step": 1172 }, { "epoch": 0.5984693877551021, "grad_norm": 23.44445037841797, "learning_rate": 4.0153061224489794e-06, "loss": 0.0221, "step": 1173 }, { "epoch": 0.5989795918367347, "grad_norm": 34.20009231567383, "learning_rate": 4.010204081632653e-06, "loss": 0.0698, "step": 1174 }, { "epoch": 0.5994897959183674, "grad_norm": 4.541125297546387, "learning_rate": 4.005102040816327e-06, "loss": 0.0059, "step": 1175 }, { "epoch": 0.6, "grad_norm": 43.639251708984375, "learning_rate": 4.000000000000001e-06, "loss": 0.0898, "step": 1176 }, { "epoch": 0.6005102040816327, "grad_norm": 6.789349555969238, "learning_rate": 3.9948979591836736e-06, "loss": 0.0062, "step": 1177 }, { "epoch": 0.6010204081632653, "grad_norm": 33.63608932495117, "learning_rate": 3.989795918367347e-06, "loss": 0.0439, "step": 1178 }, { "epoch": 0.601530612244898, "grad_norm": 72.63557434082031, "learning_rate": 3.984693877551021e-06, "loss": 0.3079, "step": 1179 }, { "epoch": 0.6020408163265306, "grad_norm": 47.210086822509766, "learning_rate": 3.979591836734694e-06, "loss": 0.1156, "step": 1180 }, { "epoch": 0.6025510204081632, "grad_norm": 8.225129127502441, "learning_rate": 3.974489795918368e-06, "loss": 0.0068, "step": 1181 }, { "epoch": 0.6030612244897959, "grad_norm": 135.2066192626953, "learning_rate": 3.969387755102041e-06, "loss": 0.4619, "step": 1182 }, { "epoch": 0.6035714285714285, "grad_norm": 28.120323181152344, "learning_rate": 3.964285714285714e-06, "loss": 0.125, "step": 1183 }, { "epoch": 0.6040816326530613, "grad_norm": 22.527610778808594, "learning_rate": 3.959183673469388e-06, "loss": 0.066, "step": 1184 }, { "epoch": 0.6045918367346939, "grad_norm": 35.383697509765625, "learning_rate": 3.954081632653062e-06, "loss": 0.1388, "step": 1185 }, { "epoch": 0.6051020408163266, "grad_norm": 32.38810729980469, "learning_rate": 3.9489795918367356e-06, "loss": 0.0779, "step": 1186 }, { "epoch": 0.6056122448979592, "grad_norm": 57.720272064208984, "learning_rate": 3.9438775510204085e-06, "loss": 0.1393, "step": 1187 }, { "epoch": 0.6061224489795919, "grad_norm": 47.67719268798828, "learning_rate": 3.938775510204082e-06, "loss": 0.1383, "step": 1188 }, { "epoch": 0.6066326530612245, "grad_norm": 28.851638793945312, "learning_rate": 3.933673469387755e-06, "loss": 0.0825, "step": 1189 }, { "epoch": 0.6071428571428571, "grad_norm": 112.25385284423828, "learning_rate": 3.928571428571429e-06, "loss": 0.282, "step": 1190 }, { "epoch": 0.6076530612244898, "grad_norm": 23.00905990600586, "learning_rate": 3.9234693877551026e-06, "loss": 0.0285, "step": 1191 }, { "epoch": 0.6081632653061224, "grad_norm": 43.42043685913086, "learning_rate": 3.9183673469387755e-06, "loss": 0.1978, "step": 1192 }, { "epoch": 0.6086734693877551, "grad_norm": 121.05421447753906, "learning_rate": 3.913265306122449e-06, "loss": 0.5898, "step": 1193 }, { "epoch": 0.6091836734693877, "grad_norm": 28.14826202392578, "learning_rate": 3.908163265306123e-06, "loss": 0.0433, "step": 1194 }, { "epoch": 0.6096938775510204, "grad_norm": 16.82575798034668, "learning_rate": 3.903061224489797e-06, "loss": 0.041, "step": 1195 }, { "epoch": 0.610204081632653, "grad_norm": 45.92105484008789, "learning_rate": 3.89795918367347e-06, "loss": 0.0614, "step": 1196 }, { "epoch": 0.6107142857142858, "grad_norm": 47.72837448120117, "learning_rate": 3.892857142857143e-06, "loss": 0.0915, "step": 1197 }, { "epoch": 0.6112244897959184, "grad_norm": 13.878188133239746, "learning_rate": 3.887755102040816e-06, "loss": 0.0302, "step": 1198 }, { "epoch": 0.611734693877551, "grad_norm": 80.20044708251953, "learning_rate": 3.88265306122449e-06, "loss": 0.3169, "step": 1199 }, { "epoch": 0.6122448979591837, "grad_norm": 7.011779308319092, "learning_rate": 3.877551020408164e-06, "loss": 0.0105, "step": 1200 }, { "epoch": 0.6127551020408163, "grad_norm": 26.344812393188477, "learning_rate": 3.872448979591837e-06, "loss": 0.0793, "step": 1201 }, { "epoch": 0.613265306122449, "grad_norm": 36.299407958984375, "learning_rate": 3.86734693877551e-06, "loss": 0.1262, "step": 1202 }, { "epoch": 0.6137755102040816, "grad_norm": 37.133602142333984, "learning_rate": 3.862244897959184e-06, "loss": 0.063, "step": 1203 }, { "epoch": 0.6142857142857143, "grad_norm": 23.012828826904297, "learning_rate": 3.857142857142858e-06, "loss": 0.0596, "step": 1204 }, { "epoch": 0.6147959183673469, "grad_norm": 22.260913848876953, "learning_rate": 3.852040816326531e-06, "loss": 0.0598, "step": 1205 }, { "epoch": 0.6153061224489796, "grad_norm": 27.0489559173584, "learning_rate": 3.8469387755102045e-06, "loss": 0.0735, "step": 1206 }, { "epoch": 0.6158163265306122, "grad_norm": 137.62973022460938, "learning_rate": 3.841836734693878e-06, "loss": 0.4932, "step": 1207 }, { "epoch": 0.6163265306122448, "grad_norm": 48.01408767700195, "learning_rate": 3.836734693877551e-06, "loss": 0.1592, "step": 1208 }, { "epoch": 0.6168367346938776, "grad_norm": 26.560874938964844, "learning_rate": 3.831632653061225e-06, "loss": 0.0417, "step": 1209 }, { "epoch": 0.6173469387755102, "grad_norm": 62.42783737182617, "learning_rate": 3.826530612244898e-06, "loss": 0.2222, "step": 1210 }, { "epoch": 0.6178571428571429, "grad_norm": 8.874138832092285, "learning_rate": 3.8214285714285715e-06, "loss": 0.0278, "step": 1211 }, { "epoch": 0.6183673469387755, "grad_norm": 101.31809997558594, "learning_rate": 3.816326530612245e-06, "loss": 0.454, "step": 1212 }, { "epoch": 0.6188775510204082, "grad_norm": 5.149127006530762, "learning_rate": 3.811224489795919e-06, "loss": 0.0089, "step": 1213 }, { "epoch": 0.6193877551020408, "grad_norm": 262.3978271484375, "learning_rate": 3.8061224489795923e-06, "loss": 2.0168, "step": 1214 }, { "epoch": 0.6198979591836735, "grad_norm": 29.437698364257812, "learning_rate": 3.8010204081632656e-06, "loss": 0.0883, "step": 1215 }, { "epoch": 0.6204081632653061, "grad_norm": 16.41016387939453, "learning_rate": 3.795918367346939e-06, "loss": 0.0207, "step": 1216 }, { "epoch": 0.6209183673469387, "grad_norm": 89.62061309814453, "learning_rate": 3.7908163265306127e-06, "loss": 0.3007, "step": 1217 }, { "epoch": 0.6214285714285714, "grad_norm": 41.82023620605469, "learning_rate": 3.785714285714286e-06, "loss": 0.1101, "step": 1218 }, { "epoch": 0.621938775510204, "grad_norm": 62.925296783447266, "learning_rate": 3.7806122448979593e-06, "loss": 0.272, "step": 1219 }, { "epoch": 0.6224489795918368, "grad_norm": 10.880748748779297, "learning_rate": 3.7755102040816327e-06, "loss": 0.009, "step": 1220 }, { "epoch": 0.6229591836734694, "grad_norm": 29.278751373291016, "learning_rate": 3.770408163265307e-06, "loss": 0.0733, "step": 1221 }, { "epoch": 0.6234693877551021, "grad_norm": 149.2815704345703, "learning_rate": 3.76530612244898e-06, "loss": 0.4274, "step": 1222 }, { "epoch": 0.6239795918367347, "grad_norm": 78.3371353149414, "learning_rate": 3.7602040816326535e-06, "loss": 0.2695, "step": 1223 }, { "epoch": 0.6244897959183674, "grad_norm": 37.416473388671875, "learning_rate": 3.7551020408163268e-06, "loss": 0.0591, "step": 1224 }, { "epoch": 0.625, "grad_norm": 60.67212677001953, "learning_rate": 3.7500000000000005e-06, "loss": 0.1754, "step": 1225 }, { "epoch": 0.6255102040816326, "grad_norm": 31.832717895507812, "learning_rate": 3.744897959183674e-06, "loss": 0.0608, "step": 1226 }, { "epoch": 0.6260204081632653, "grad_norm": 68.5280532836914, "learning_rate": 3.739795918367347e-06, "loss": 0.2205, "step": 1227 }, { "epoch": 0.6265306122448979, "grad_norm": 59.511043548583984, "learning_rate": 3.7346938775510205e-06, "loss": 0.1914, "step": 1228 }, { "epoch": 0.6270408163265306, "grad_norm": 14.966217041015625, "learning_rate": 3.729591836734694e-06, "loss": 0.0391, "step": 1229 }, { "epoch": 0.6275510204081632, "grad_norm": 74.5845718383789, "learning_rate": 3.724489795918368e-06, "loss": 0.1821, "step": 1230 }, { "epoch": 0.628061224489796, "grad_norm": 90.99407196044922, "learning_rate": 3.7193877551020413e-06, "loss": 0.377, "step": 1231 }, { "epoch": 0.6285714285714286, "grad_norm": 93.59234619140625, "learning_rate": 3.7142857142857146e-06, "loss": 0.418, "step": 1232 }, { "epoch": 0.6290816326530613, "grad_norm": 30.9075984954834, "learning_rate": 3.709183673469388e-06, "loss": 0.0938, "step": 1233 }, { "epoch": 0.6295918367346939, "grad_norm": 45.57841873168945, "learning_rate": 3.7040816326530617e-06, "loss": 0.1659, "step": 1234 }, { "epoch": 0.6301020408163265, "grad_norm": 21.35421371459961, "learning_rate": 3.698979591836735e-06, "loss": 0.0295, "step": 1235 }, { "epoch": 0.6306122448979592, "grad_norm": 48.690364837646484, "learning_rate": 3.6938775510204083e-06, "loss": 0.1776, "step": 1236 }, { "epoch": 0.6311224489795918, "grad_norm": 40.7950325012207, "learning_rate": 3.6887755102040816e-06, "loss": 0.1833, "step": 1237 }, { "epoch": 0.6316326530612245, "grad_norm": 30.980669021606445, "learning_rate": 3.6836734693877554e-06, "loss": 0.0631, "step": 1238 }, { "epoch": 0.6321428571428571, "grad_norm": 23.059757232666016, "learning_rate": 3.678571428571429e-06, "loss": 0.0487, "step": 1239 }, { "epoch": 0.6326530612244898, "grad_norm": 46.648170471191406, "learning_rate": 3.6734693877551024e-06, "loss": 0.0938, "step": 1240 }, { "epoch": 0.6331632653061224, "grad_norm": 137.18975830078125, "learning_rate": 3.6683673469387757e-06, "loss": 0.4893, "step": 1241 }, { "epoch": 0.6336734693877552, "grad_norm": 37.0784912109375, "learning_rate": 3.6632653061224495e-06, "loss": 0.0968, "step": 1242 }, { "epoch": 0.6341836734693878, "grad_norm": 58.62586975097656, "learning_rate": 3.658163265306123e-06, "loss": 0.1553, "step": 1243 }, { "epoch": 0.6346938775510204, "grad_norm": 29.37779998779297, "learning_rate": 3.653061224489796e-06, "loss": 0.0913, "step": 1244 }, { "epoch": 0.6352040816326531, "grad_norm": 55.759151458740234, "learning_rate": 3.6479591836734694e-06, "loss": 0.2056, "step": 1245 }, { "epoch": 0.6357142857142857, "grad_norm": 37.2816047668457, "learning_rate": 3.642857142857143e-06, "loss": 0.1276, "step": 1246 }, { "epoch": 0.6362244897959184, "grad_norm": 5.7401299476623535, "learning_rate": 3.6377551020408165e-06, "loss": 0.0073, "step": 1247 }, { "epoch": 0.636734693877551, "grad_norm": 227.3195343017578, "learning_rate": 3.6326530612244903e-06, "loss": 0.7695, "step": 1248 }, { "epoch": 0.6372448979591837, "grad_norm": 154.99246215820312, "learning_rate": 3.6275510204081636e-06, "loss": 0.5535, "step": 1249 }, { "epoch": 0.6377551020408163, "grad_norm": 57.26854705810547, "learning_rate": 3.6224489795918373e-06, "loss": 0.135, "step": 1250 }, { "epoch": 0.638265306122449, "grad_norm": 188.82696533203125, "learning_rate": 3.6173469387755106e-06, "loss": 0.6754, "step": 1251 }, { "epoch": 0.6387755102040816, "grad_norm": 87.55656433105469, "learning_rate": 3.612244897959184e-06, "loss": 0.3618, "step": 1252 }, { "epoch": 0.6392857142857142, "grad_norm": 161.7622833251953, "learning_rate": 3.6071428571428573e-06, "loss": 0.7307, "step": 1253 }, { "epoch": 0.639795918367347, "grad_norm": 6.065882205963135, "learning_rate": 3.6020408163265306e-06, "loss": 0.0096, "step": 1254 }, { "epoch": 0.6403061224489796, "grad_norm": 22.998043060302734, "learning_rate": 3.5969387755102043e-06, "loss": 0.0453, "step": 1255 }, { "epoch": 0.6408163265306123, "grad_norm": 30.687366485595703, "learning_rate": 3.5918367346938777e-06, "loss": 0.0825, "step": 1256 }, { "epoch": 0.6413265306122449, "grad_norm": 41.66588592529297, "learning_rate": 3.5867346938775514e-06, "loss": 0.2372, "step": 1257 }, { "epoch": 0.6418367346938776, "grad_norm": 152.4623260498047, "learning_rate": 3.581632653061225e-06, "loss": 0.52, "step": 1258 }, { "epoch": 0.6423469387755102, "grad_norm": 42.19049072265625, "learning_rate": 3.5765306122448985e-06, "loss": 0.0579, "step": 1259 }, { "epoch": 0.6428571428571429, "grad_norm": 17.78597640991211, "learning_rate": 3.5714285714285718e-06, "loss": 0.0189, "step": 1260 }, { "epoch": 0.6433673469387755, "grad_norm": 90.68384552001953, "learning_rate": 3.566326530612245e-06, "loss": 0.3799, "step": 1261 }, { "epoch": 0.6438775510204081, "grad_norm": 48.90946960449219, "learning_rate": 3.5612244897959184e-06, "loss": 0.1118, "step": 1262 }, { "epoch": 0.6443877551020408, "grad_norm": 29.387908935546875, "learning_rate": 3.556122448979592e-06, "loss": 0.0964, "step": 1263 }, { "epoch": 0.6448979591836734, "grad_norm": 26.874492645263672, "learning_rate": 3.5510204081632655e-06, "loss": 0.0604, "step": 1264 }, { "epoch": 0.6454081632653061, "grad_norm": 4.693141460418701, "learning_rate": 3.545918367346939e-06, "loss": 0.0069, "step": 1265 }, { "epoch": 0.6459183673469387, "grad_norm": 69.35267639160156, "learning_rate": 3.540816326530613e-06, "loss": 0.3027, "step": 1266 }, { "epoch": 0.6464285714285715, "grad_norm": 138.06796264648438, "learning_rate": 3.5357142857142863e-06, "loss": 0.7339, "step": 1267 }, { "epoch": 0.6469387755102041, "grad_norm": 21.743228912353516, "learning_rate": 3.5306122448979596e-06, "loss": 0.0252, "step": 1268 }, { "epoch": 0.6474489795918368, "grad_norm": 17.22263526916504, "learning_rate": 3.525510204081633e-06, "loss": 0.0371, "step": 1269 }, { "epoch": 0.6479591836734694, "grad_norm": 41.038551330566406, "learning_rate": 3.5204081632653062e-06, "loss": 0.0771, "step": 1270 }, { "epoch": 0.648469387755102, "grad_norm": 32.074947357177734, "learning_rate": 3.51530612244898e-06, "loss": 0.0784, "step": 1271 }, { "epoch": 0.6489795918367347, "grad_norm": 9.528373718261719, "learning_rate": 3.5102040816326533e-06, "loss": 0.0234, "step": 1272 }, { "epoch": 0.6494897959183673, "grad_norm": 60.76799011230469, "learning_rate": 3.5051020408163266e-06, "loss": 0.2524, "step": 1273 }, { "epoch": 0.65, "grad_norm": 38.63211441040039, "learning_rate": 3.5e-06, "loss": 0.1427, "step": 1274 }, { "epoch": 0.6505102040816326, "grad_norm": 7.10417366027832, "learning_rate": 3.494897959183674e-06, "loss": 0.0093, "step": 1275 }, { "epoch": 0.6510204081632653, "grad_norm": 50.85823059082031, "learning_rate": 3.4897959183673474e-06, "loss": 0.3553, "step": 1276 }, { "epoch": 0.6515306122448979, "grad_norm": 24.623882293701172, "learning_rate": 3.4846938775510208e-06, "loss": 0.0433, "step": 1277 }, { "epoch": 0.6520408163265307, "grad_norm": 80.55793762207031, "learning_rate": 3.479591836734694e-06, "loss": 0.2173, "step": 1278 }, { "epoch": 0.6525510204081633, "grad_norm": 42.476985931396484, "learning_rate": 3.474489795918368e-06, "loss": 0.1538, "step": 1279 }, { "epoch": 0.6530612244897959, "grad_norm": 37.45490646362305, "learning_rate": 3.469387755102041e-06, "loss": 0.0751, "step": 1280 }, { "epoch": 0.6535714285714286, "grad_norm": 128.91746520996094, "learning_rate": 3.4642857142857145e-06, "loss": 0.3445, "step": 1281 }, { "epoch": 0.6540816326530612, "grad_norm": 61.338314056396484, "learning_rate": 3.4591836734693878e-06, "loss": 0.166, "step": 1282 }, { "epoch": 0.6545918367346939, "grad_norm": 74.79781341552734, "learning_rate": 3.454081632653061e-06, "loss": 0.2139, "step": 1283 }, { "epoch": 0.6551020408163265, "grad_norm": 199.9740447998047, "learning_rate": 3.4489795918367353e-06, "loss": 0.7637, "step": 1284 }, { "epoch": 0.6556122448979592, "grad_norm": 71.30680847167969, "learning_rate": 3.4438775510204086e-06, "loss": 0.2422, "step": 1285 }, { "epoch": 0.6561224489795918, "grad_norm": 204.0145721435547, "learning_rate": 3.438775510204082e-06, "loss": 0.5482, "step": 1286 }, { "epoch": 0.6566326530612245, "grad_norm": 150.47691345214844, "learning_rate": 3.4336734693877556e-06, "loss": 0.3059, "step": 1287 }, { "epoch": 0.6571428571428571, "grad_norm": 66.11913299560547, "learning_rate": 3.428571428571429e-06, "loss": 0.167, "step": 1288 }, { "epoch": 0.6576530612244897, "grad_norm": 34.886905670166016, "learning_rate": 3.4234693877551023e-06, "loss": 0.0716, "step": 1289 }, { "epoch": 0.6581632653061225, "grad_norm": 12.028800964355469, "learning_rate": 3.4183673469387756e-06, "loss": 0.0144, "step": 1290 }, { "epoch": 0.6586734693877551, "grad_norm": 51.986209869384766, "learning_rate": 3.413265306122449e-06, "loss": 0.1106, "step": 1291 }, { "epoch": 0.6591836734693878, "grad_norm": 168.58013916015625, "learning_rate": 3.4081632653061227e-06, "loss": 0.3833, "step": 1292 }, { "epoch": 0.6596938775510204, "grad_norm": 45.02932357788086, "learning_rate": 3.403061224489796e-06, "loss": 0.0609, "step": 1293 }, { "epoch": 0.6602040816326531, "grad_norm": 113.24119567871094, "learning_rate": 3.3979591836734697e-06, "loss": 0.2992, "step": 1294 }, { "epoch": 0.6607142857142857, "grad_norm": 50.9926643371582, "learning_rate": 3.3928571428571435e-06, "loss": 0.1458, "step": 1295 }, { "epoch": 0.6612244897959184, "grad_norm": 23.126304626464844, "learning_rate": 3.3877551020408168e-06, "loss": 0.0398, "step": 1296 }, { "epoch": 0.661734693877551, "grad_norm": 117.82582092285156, "learning_rate": 3.38265306122449e-06, "loss": 0.3174, "step": 1297 }, { "epoch": 0.6622448979591836, "grad_norm": 159.70225524902344, "learning_rate": 3.3775510204081634e-06, "loss": 0.7715, "step": 1298 }, { "epoch": 0.6627551020408163, "grad_norm": 47.05710983276367, "learning_rate": 3.3724489795918367e-06, "loss": 0.0938, "step": 1299 }, { "epoch": 0.6632653061224489, "grad_norm": 105.86807250976562, "learning_rate": 3.3673469387755105e-06, "loss": 0.3846, "step": 1300 }, { "epoch": 0.6637755102040817, "grad_norm": 149.07904052734375, "learning_rate": 3.362244897959184e-06, "loss": 0.5735, "step": 1301 }, { "epoch": 0.6642857142857143, "grad_norm": 28.682085037231445, "learning_rate": 3.357142857142857e-06, "loss": 0.0663, "step": 1302 }, { "epoch": 0.664795918367347, "grad_norm": 151.19749450683594, "learning_rate": 3.3520408163265313e-06, "loss": 0.5215, "step": 1303 }, { "epoch": 0.6653061224489796, "grad_norm": 237.76632690429688, "learning_rate": 3.3469387755102046e-06, "loss": 0.5919, "step": 1304 }, { "epoch": 0.6658163265306123, "grad_norm": 82.95072937011719, "learning_rate": 3.341836734693878e-06, "loss": 0.2051, "step": 1305 }, { "epoch": 0.6663265306122449, "grad_norm": 8.847975730895996, "learning_rate": 3.3367346938775513e-06, "loss": 0.0116, "step": 1306 }, { "epoch": 0.6668367346938775, "grad_norm": 98.38676452636719, "learning_rate": 3.3316326530612246e-06, "loss": 0.2585, "step": 1307 }, { "epoch": 0.6673469387755102, "grad_norm": 65.65143585205078, "learning_rate": 3.3265306122448983e-06, "loss": 0.1846, "step": 1308 }, { "epoch": 0.6678571428571428, "grad_norm": 23.375133514404297, "learning_rate": 3.3214285714285716e-06, "loss": 0.0394, "step": 1309 }, { "epoch": 0.6683673469387755, "grad_norm": 14.622981071472168, "learning_rate": 3.316326530612245e-06, "loss": 0.0184, "step": 1310 }, { "epoch": 0.6688775510204081, "grad_norm": 55.967010498046875, "learning_rate": 3.3112244897959183e-06, "loss": 0.1106, "step": 1311 }, { "epoch": 0.6693877551020408, "grad_norm": 44.46063232421875, "learning_rate": 3.3061224489795924e-06, "loss": 0.1154, "step": 1312 }, { "epoch": 0.6698979591836735, "grad_norm": 30.630435943603516, "learning_rate": 3.3010204081632658e-06, "loss": 0.0485, "step": 1313 }, { "epoch": 0.6704081632653062, "grad_norm": 46.796878814697266, "learning_rate": 3.295918367346939e-06, "loss": 0.1836, "step": 1314 }, { "epoch": 0.6709183673469388, "grad_norm": 112.92406463623047, "learning_rate": 3.2908163265306124e-06, "loss": 0.5488, "step": 1315 }, { "epoch": 0.6714285714285714, "grad_norm": 6.440260410308838, "learning_rate": 3.285714285714286e-06, "loss": 0.0032, "step": 1316 }, { "epoch": 0.6719387755102041, "grad_norm": 53.364845275878906, "learning_rate": 3.2806122448979595e-06, "loss": 0.189, "step": 1317 }, { "epoch": 0.6724489795918367, "grad_norm": 63.86994171142578, "learning_rate": 3.2755102040816328e-06, "loss": 0.1868, "step": 1318 }, { "epoch": 0.6729591836734694, "grad_norm": 75.62535858154297, "learning_rate": 3.270408163265306e-06, "loss": 0.1777, "step": 1319 }, { "epoch": 0.673469387755102, "grad_norm": 53.42654037475586, "learning_rate": 3.2653061224489794e-06, "loss": 0.1637, "step": 1320 }, { "epoch": 0.6739795918367347, "grad_norm": 105.72643280029297, "learning_rate": 3.2602040816326536e-06, "loss": 0.4586, "step": 1321 }, { "epoch": 0.6744897959183673, "grad_norm": 16.240333557128906, "learning_rate": 3.255102040816327e-06, "loss": 0.039, "step": 1322 }, { "epoch": 0.675, "grad_norm": 53.56648254394531, "learning_rate": 3.2500000000000002e-06, "loss": 0.1201, "step": 1323 }, { "epoch": 0.6755102040816326, "grad_norm": 34.67395782470703, "learning_rate": 3.244897959183674e-06, "loss": 0.0732, "step": 1324 }, { "epoch": 0.6760204081632653, "grad_norm": 33.72609329223633, "learning_rate": 3.2397959183673473e-06, "loss": 0.0486, "step": 1325 }, { "epoch": 0.676530612244898, "grad_norm": 304.20111083984375, "learning_rate": 3.2346938775510206e-06, "loss": 2.5117, "step": 1326 }, { "epoch": 0.6770408163265306, "grad_norm": 101.3328628540039, "learning_rate": 3.229591836734694e-06, "loss": 0.4333, "step": 1327 }, { "epoch": 0.6775510204081633, "grad_norm": 161.29364013671875, "learning_rate": 3.2244897959183672e-06, "loss": 0.7537, "step": 1328 }, { "epoch": 0.6780612244897959, "grad_norm": 273.8958435058594, "learning_rate": 3.219387755102041e-06, "loss": 0.6699, "step": 1329 }, { "epoch": 0.6785714285714286, "grad_norm": 52.420066833496094, "learning_rate": 3.2142857142857147e-06, "loss": 0.0795, "step": 1330 }, { "epoch": 0.6790816326530612, "grad_norm": 24.306180953979492, "learning_rate": 3.209183673469388e-06, "loss": 0.0459, "step": 1331 }, { "epoch": 0.6795918367346939, "grad_norm": 18.17367935180664, "learning_rate": 3.204081632653062e-06, "loss": 0.0314, "step": 1332 }, { "epoch": 0.6801020408163265, "grad_norm": 92.50967407226562, "learning_rate": 3.198979591836735e-06, "loss": 0.3462, "step": 1333 }, { "epoch": 0.6806122448979591, "grad_norm": 109.51399230957031, "learning_rate": 3.1938775510204084e-06, "loss": 0.2937, "step": 1334 }, { "epoch": 0.6811224489795918, "grad_norm": 62.40757369995117, "learning_rate": 3.1887755102040818e-06, "loss": 0.1665, "step": 1335 }, { "epoch": 0.6816326530612244, "grad_norm": 50.535804748535156, "learning_rate": 3.183673469387755e-06, "loss": 0.1323, "step": 1336 }, { "epoch": 0.6821428571428572, "grad_norm": 46.113521575927734, "learning_rate": 3.178571428571429e-06, "loss": 0.1304, "step": 1337 }, { "epoch": 0.6826530612244898, "grad_norm": 16.513460159301758, "learning_rate": 3.173469387755102e-06, "loss": 0.0516, "step": 1338 }, { "epoch": 0.6831632653061225, "grad_norm": 20.424901962280273, "learning_rate": 3.168367346938776e-06, "loss": 0.036, "step": 1339 }, { "epoch": 0.6836734693877551, "grad_norm": 227.9959259033203, "learning_rate": 3.1632653061224496e-06, "loss": 2.1481, "step": 1340 }, { "epoch": 0.6841836734693878, "grad_norm": 36.656368255615234, "learning_rate": 3.158163265306123e-06, "loss": 0.0788, "step": 1341 }, { "epoch": 0.6846938775510204, "grad_norm": 16.0291748046875, "learning_rate": 3.1530612244897963e-06, "loss": 0.0269, "step": 1342 }, { "epoch": 0.685204081632653, "grad_norm": 41.7044792175293, "learning_rate": 3.1479591836734696e-06, "loss": 0.1091, "step": 1343 }, { "epoch": 0.6857142857142857, "grad_norm": 35.44289016723633, "learning_rate": 3.142857142857143e-06, "loss": 0.045, "step": 1344 }, { "epoch": 0.6862244897959183, "grad_norm": 13.630613327026367, "learning_rate": 3.1377551020408166e-06, "loss": 0.0193, "step": 1345 }, { "epoch": 0.686734693877551, "grad_norm": 54.319122314453125, "learning_rate": 3.13265306122449e-06, "loss": 0.125, "step": 1346 }, { "epoch": 0.6872448979591836, "grad_norm": 14.014261245727539, "learning_rate": 3.1275510204081633e-06, "loss": 0.0289, "step": 1347 }, { "epoch": 0.6877551020408164, "grad_norm": 30.22063446044922, "learning_rate": 3.1224489795918374e-06, "loss": 0.0482, "step": 1348 }, { "epoch": 0.688265306122449, "grad_norm": 12.280858993530273, "learning_rate": 3.1173469387755108e-06, "loss": 0.0269, "step": 1349 }, { "epoch": 0.6887755102040817, "grad_norm": 36.94214630126953, "learning_rate": 3.112244897959184e-06, "loss": 0.0555, "step": 1350 }, { "epoch": 0.6892857142857143, "grad_norm": 55.85078048706055, "learning_rate": 3.1071428571428574e-06, "loss": 0.1272, "step": 1351 }, { "epoch": 0.689795918367347, "grad_norm": 169.54934692382812, "learning_rate": 3.1020408163265307e-06, "loss": 0.5809, "step": 1352 }, { "epoch": 0.6903061224489796, "grad_norm": 44.607322692871094, "learning_rate": 3.0969387755102045e-06, "loss": 0.1147, "step": 1353 }, { "epoch": 0.6908163265306122, "grad_norm": 24.514175415039062, "learning_rate": 3.0918367346938778e-06, "loss": 0.0347, "step": 1354 }, { "epoch": 0.6913265306122449, "grad_norm": 24.046022415161133, "learning_rate": 3.086734693877551e-06, "loss": 0.0312, "step": 1355 }, { "epoch": 0.6918367346938775, "grad_norm": 38.210418701171875, "learning_rate": 3.0816326530612244e-06, "loss": 0.0592, "step": 1356 }, { "epoch": 0.6923469387755102, "grad_norm": 17.068584442138672, "learning_rate": 3.0765306122448986e-06, "loss": 0.0195, "step": 1357 }, { "epoch": 0.6928571428571428, "grad_norm": 129.2494659423828, "learning_rate": 3.071428571428572e-06, "loss": 0.5249, "step": 1358 }, { "epoch": 0.6933673469387756, "grad_norm": 55.43430709838867, "learning_rate": 3.0663265306122452e-06, "loss": 0.2256, "step": 1359 }, { "epoch": 0.6938775510204082, "grad_norm": 252.0205535888672, "learning_rate": 3.0612244897959185e-06, "loss": 1.033, "step": 1360 }, { "epoch": 0.6943877551020409, "grad_norm": 130.76499938964844, "learning_rate": 3.0561224489795923e-06, "loss": 0.2821, "step": 1361 }, { "epoch": 0.6948979591836735, "grad_norm": 29.244552612304688, "learning_rate": 3.0510204081632656e-06, "loss": 0.1099, "step": 1362 }, { "epoch": 0.6954081632653061, "grad_norm": 27.399456024169922, "learning_rate": 3.045918367346939e-06, "loss": 0.061, "step": 1363 }, { "epoch": 0.6959183673469388, "grad_norm": 19.995752334594727, "learning_rate": 3.0408163265306122e-06, "loss": 0.0263, "step": 1364 }, { "epoch": 0.6964285714285714, "grad_norm": 107.93071746826172, "learning_rate": 3.0357142857142856e-06, "loss": 0.2656, "step": 1365 }, { "epoch": 0.6969387755102041, "grad_norm": 35.49892044067383, "learning_rate": 3.0306122448979597e-06, "loss": 0.0819, "step": 1366 }, { "epoch": 0.6974489795918367, "grad_norm": 92.10172271728516, "learning_rate": 3.025510204081633e-06, "loss": 0.561, "step": 1367 }, { "epoch": 0.6979591836734694, "grad_norm": 52.84951400756836, "learning_rate": 3.0204081632653064e-06, "loss": 0.1683, "step": 1368 }, { "epoch": 0.698469387755102, "grad_norm": 18.06717872619629, "learning_rate": 3.01530612244898e-06, "loss": 0.0329, "step": 1369 }, { "epoch": 0.6989795918367347, "grad_norm": 86.25357055664062, "learning_rate": 3.0102040816326534e-06, "loss": 0.2227, "step": 1370 }, { "epoch": 0.6994897959183674, "grad_norm": 40.815616607666016, "learning_rate": 3.0051020408163268e-06, "loss": 0.103, "step": 1371 }, { "epoch": 0.7, "grad_norm": 78.3099594116211, "learning_rate": 3e-06, "loss": 0.2516, "step": 1372 }, { "epoch": 0.7005102040816327, "grad_norm": 47.50718307495117, "learning_rate": 2.9948979591836734e-06, "loss": 0.1595, "step": 1373 }, { "epoch": 0.7010204081632653, "grad_norm": 124.28340148925781, "learning_rate": 2.989795918367347e-06, "loss": 0.6194, "step": 1374 }, { "epoch": 0.701530612244898, "grad_norm": 49.12411880493164, "learning_rate": 2.984693877551021e-06, "loss": 0.2718, "step": 1375 }, { "epoch": 0.7020408163265306, "grad_norm": 45.886680603027344, "learning_rate": 2.979591836734694e-06, "loss": 0.1262, "step": 1376 }, { "epoch": 0.7025510204081633, "grad_norm": 128.42434692382812, "learning_rate": 2.9744897959183675e-06, "loss": 0.5167, "step": 1377 }, { "epoch": 0.7030612244897959, "grad_norm": 86.9940414428711, "learning_rate": 2.9693877551020413e-06, "loss": 0.4423, "step": 1378 }, { "epoch": 0.7035714285714286, "grad_norm": 206.6812286376953, "learning_rate": 2.9642857142857146e-06, "loss": 0.5991, "step": 1379 }, { "epoch": 0.7040816326530612, "grad_norm": 18.718168258666992, "learning_rate": 2.959183673469388e-06, "loss": 0.0409, "step": 1380 }, { "epoch": 0.7045918367346938, "grad_norm": 77.02133178710938, "learning_rate": 2.9540816326530612e-06, "loss": 0.2314, "step": 1381 }, { "epoch": 0.7051020408163265, "grad_norm": 4.018664836883545, "learning_rate": 2.948979591836735e-06, "loss": 0.0038, "step": 1382 }, { "epoch": 0.7056122448979592, "grad_norm": 157.15811157226562, "learning_rate": 2.9438775510204083e-06, "loss": 0.998, "step": 1383 }, { "epoch": 0.7061224489795919, "grad_norm": 39.757713317871094, "learning_rate": 2.938775510204082e-06, "loss": 0.0664, "step": 1384 }, { "epoch": 0.7066326530612245, "grad_norm": 15.020140647888184, "learning_rate": 2.9336734693877553e-06, "loss": 0.0144, "step": 1385 }, { "epoch": 0.7071428571428572, "grad_norm": 35.86440658569336, "learning_rate": 2.928571428571429e-06, "loss": 0.0775, "step": 1386 }, { "epoch": 0.7076530612244898, "grad_norm": 123.66632843017578, "learning_rate": 2.9234693877551024e-06, "loss": 0.5956, "step": 1387 }, { "epoch": 0.7081632653061225, "grad_norm": 40.352012634277344, "learning_rate": 2.9183673469387757e-06, "loss": 0.1117, "step": 1388 }, { "epoch": 0.7086734693877551, "grad_norm": 11.866608619689941, "learning_rate": 2.913265306122449e-06, "loss": 0.0253, "step": 1389 }, { "epoch": 0.7091836734693877, "grad_norm": 33.4807014465332, "learning_rate": 2.908163265306123e-06, "loss": 0.0828, "step": 1390 }, { "epoch": 0.7096938775510204, "grad_norm": 31.16853904724121, "learning_rate": 2.903061224489796e-06, "loss": 0.0594, "step": 1391 }, { "epoch": 0.710204081632653, "grad_norm": 10.228103637695312, "learning_rate": 2.8979591836734694e-06, "loss": 0.0137, "step": 1392 }, { "epoch": 0.7107142857142857, "grad_norm": 42.001739501953125, "learning_rate": 2.892857142857143e-06, "loss": 0.0728, "step": 1393 }, { "epoch": 0.7112244897959183, "grad_norm": 30.58222770690918, "learning_rate": 2.887755102040817e-06, "loss": 0.0439, "step": 1394 }, { "epoch": 0.7117346938775511, "grad_norm": 91.97080993652344, "learning_rate": 2.8826530612244902e-06, "loss": 0.2506, "step": 1395 }, { "epoch": 0.7122448979591837, "grad_norm": 65.85636138916016, "learning_rate": 2.8775510204081636e-06, "loss": 0.1409, "step": 1396 }, { "epoch": 0.7127551020408164, "grad_norm": 81.19803619384766, "learning_rate": 2.872448979591837e-06, "loss": 0.2297, "step": 1397 }, { "epoch": 0.713265306122449, "grad_norm": 17.057456970214844, "learning_rate": 2.86734693877551e-06, "loss": 0.0237, "step": 1398 }, { "epoch": 0.7137755102040816, "grad_norm": 13.33524227142334, "learning_rate": 2.862244897959184e-06, "loss": 0.0296, "step": 1399 }, { "epoch": 0.7142857142857143, "grad_norm": 115.03265380859375, "learning_rate": 2.8571428571428573e-06, "loss": 0.5859, "step": 1400 }, { "epoch": 0.7147959183673469, "grad_norm": 10.494661331176758, "learning_rate": 2.8520408163265306e-06, "loss": 0.0133, "step": 1401 }, { "epoch": 0.7153061224489796, "grad_norm": 164.53602600097656, "learning_rate": 2.8469387755102047e-06, "loss": 0.5603, "step": 1402 }, { "epoch": 0.7158163265306122, "grad_norm": 96.7363052368164, "learning_rate": 2.841836734693878e-06, "loss": 0.4878, "step": 1403 }, { "epoch": 0.7163265306122449, "grad_norm": 11.817408561706543, "learning_rate": 2.8367346938775514e-06, "loss": 0.0202, "step": 1404 }, { "epoch": 0.7168367346938775, "grad_norm": 105.869873046875, "learning_rate": 2.8316326530612247e-06, "loss": 0.4158, "step": 1405 }, { "epoch": 0.7173469387755103, "grad_norm": 51.042598724365234, "learning_rate": 2.826530612244898e-06, "loss": 0.1283, "step": 1406 }, { "epoch": 0.7178571428571429, "grad_norm": 23.02140235900879, "learning_rate": 2.8214285714285718e-06, "loss": 0.0337, "step": 1407 }, { "epoch": 0.7183673469387755, "grad_norm": 16.88048553466797, "learning_rate": 2.816326530612245e-06, "loss": 0.0134, "step": 1408 }, { "epoch": 0.7188775510204082, "grad_norm": 21.138795852661133, "learning_rate": 2.8112244897959184e-06, "loss": 0.0353, "step": 1409 }, { "epoch": 0.7193877551020408, "grad_norm": 24.64508628845215, "learning_rate": 2.8061224489795917e-06, "loss": 0.0325, "step": 1410 }, { "epoch": 0.7198979591836735, "grad_norm": 118.86836242675781, "learning_rate": 2.8010204081632655e-06, "loss": 0.5259, "step": 1411 }, { "epoch": 0.7204081632653061, "grad_norm": 33.078041076660156, "learning_rate": 2.795918367346939e-06, "loss": 0.0337, "step": 1412 }, { "epoch": 0.7209183673469388, "grad_norm": 213.21453857421875, "learning_rate": 2.7908163265306125e-06, "loss": 0.8176, "step": 1413 }, { "epoch": 0.7214285714285714, "grad_norm": 66.0126953125, "learning_rate": 2.785714285714286e-06, "loss": 0.2976, "step": 1414 }, { "epoch": 0.7219387755102041, "grad_norm": 44.177181243896484, "learning_rate": 2.7806122448979596e-06, "loss": 0.1295, "step": 1415 }, { "epoch": 0.7224489795918367, "grad_norm": 46.14854431152344, "learning_rate": 2.775510204081633e-06, "loss": 0.1444, "step": 1416 }, { "epoch": 0.7229591836734693, "grad_norm": 16.20336151123047, "learning_rate": 2.7704081632653062e-06, "loss": 0.033, "step": 1417 }, { "epoch": 0.7234693877551021, "grad_norm": 23.427196502685547, "learning_rate": 2.7653061224489795e-06, "loss": 0.0427, "step": 1418 }, { "epoch": 0.7239795918367347, "grad_norm": 65.11282348632812, "learning_rate": 2.760204081632653e-06, "loss": 0.1342, "step": 1419 }, { "epoch": 0.7244897959183674, "grad_norm": 184.5493621826172, "learning_rate": 2.7551020408163266e-06, "loss": 1.2261, "step": 1420 }, { "epoch": 0.725, "grad_norm": 79.66617584228516, "learning_rate": 2.7500000000000004e-06, "loss": 0.2528, "step": 1421 }, { "epoch": 0.7255102040816327, "grad_norm": 56.8701286315918, "learning_rate": 2.7448979591836737e-06, "loss": 0.1462, "step": 1422 }, { "epoch": 0.7260204081632653, "grad_norm": 39.70444869995117, "learning_rate": 2.7397959183673474e-06, "loss": 0.0862, "step": 1423 }, { "epoch": 0.726530612244898, "grad_norm": 46.36750793457031, "learning_rate": 2.7346938775510207e-06, "loss": 0.109, "step": 1424 }, { "epoch": 0.7270408163265306, "grad_norm": 47.32767105102539, "learning_rate": 2.729591836734694e-06, "loss": 0.1262, "step": 1425 }, { "epoch": 0.7275510204081632, "grad_norm": 85.91315460205078, "learning_rate": 2.7244897959183674e-06, "loss": 0.2844, "step": 1426 }, { "epoch": 0.7280612244897959, "grad_norm": 15.44376277923584, "learning_rate": 2.7193877551020407e-06, "loss": 0.0181, "step": 1427 }, { "epoch": 0.7285714285714285, "grad_norm": 27.30631446838379, "learning_rate": 2.7142857142857144e-06, "loss": 0.0598, "step": 1428 }, { "epoch": 0.7290816326530613, "grad_norm": 19.146581649780273, "learning_rate": 2.7091836734693878e-06, "loss": 0.0376, "step": 1429 }, { "epoch": 0.7295918367346939, "grad_norm": 83.55378723144531, "learning_rate": 2.7040816326530615e-06, "loss": 0.2251, "step": 1430 }, { "epoch": 0.7301020408163266, "grad_norm": 36.272315979003906, "learning_rate": 2.6989795918367352e-06, "loss": 0.0794, "step": 1431 }, { "epoch": 0.7306122448979592, "grad_norm": 26.295085906982422, "learning_rate": 2.6938775510204086e-06, "loss": 0.0358, "step": 1432 }, { "epoch": 0.7311224489795919, "grad_norm": 109.32255554199219, "learning_rate": 2.688775510204082e-06, "loss": 0.4917, "step": 1433 }, { "epoch": 0.7316326530612245, "grad_norm": 21.722349166870117, "learning_rate": 2.683673469387755e-06, "loss": 0.0471, "step": 1434 }, { "epoch": 0.7321428571428571, "grad_norm": 42.616268157958984, "learning_rate": 2.6785714285714285e-06, "loss": 0.066, "step": 1435 }, { "epoch": 0.7326530612244898, "grad_norm": 5.597631931304932, "learning_rate": 2.6734693877551023e-06, "loss": 0.0052, "step": 1436 }, { "epoch": 0.7331632653061224, "grad_norm": 75.387939453125, "learning_rate": 2.6683673469387756e-06, "loss": 0.3193, "step": 1437 }, { "epoch": 0.7336734693877551, "grad_norm": 52.64888381958008, "learning_rate": 2.663265306122449e-06, "loss": 0.082, "step": 1438 }, { "epoch": 0.7341836734693877, "grad_norm": 157.04159545898438, "learning_rate": 2.658163265306123e-06, "loss": 0.6252, "step": 1439 }, { "epoch": 0.7346938775510204, "grad_norm": 10.403306007385254, "learning_rate": 2.6530612244897964e-06, "loss": 0.0144, "step": 1440 }, { "epoch": 0.735204081632653, "grad_norm": 55.93538284301758, "learning_rate": 2.6479591836734697e-06, "loss": 0.1375, "step": 1441 }, { "epoch": 0.7357142857142858, "grad_norm": 17.284523010253906, "learning_rate": 2.642857142857143e-06, "loss": 0.0208, "step": 1442 }, { "epoch": 0.7362244897959184, "grad_norm": 129.656005859375, "learning_rate": 2.6377551020408163e-06, "loss": 0.7139, "step": 1443 }, { "epoch": 0.736734693877551, "grad_norm": 69.57071685791016, "learning_rate": 2.63265306122449e-06, "loss": 0.2533, "step": 1444 }, { "epoch": 0.7372448979591837, "grad_norm": 92.33845520019531, "learning_rate": 2.6275510204081634e-06, "loss": 0.3643, "step": 1445 }, { "epoch": 0.7377551020408163, "grad_norm": 10.969841003417969, "learning_rate": 2.6224489795918367e-06, "loss": 0.0193, "step": 1446 }, { "epoch": 0.738265306122449, "grad_norm": 52.245994567871094, "learning_rate": 2.61734693877551e-06, "loss": 0.1133, "step": 1447 }, { "epoch": 0.7387755102040816, "grad_norm": 37.03947067260742, "learning_rate": 2.6122448979591842e-06, "loss": 0.0509, "step": 1448 }, { "epoch": 0.7392857142857143, "grad_norm": 180.9546661376953, "learning_rate": 2.6071428571428575e-06, "loss": 0.6484, "step": 1449 }, { "epoch": 0.7397959183673469, "grad_norm": 117.41265106201172, "learning_rate": 2.602040816326531e-06, "loss": 0.3016, "step": 1450 }, { "epoch": 0.7403061224489796, "grad_norm": 26.961854934692383, "learning_rate": 2.596938775510204e-06, "loss": 0.0442, "step": 1451 }, { "epoch": 0.7408163265306122, "grad_norm": 15.344857215881348, "learning_rate": 2.591836734693878e-06, "loss": 0.0204, "step": 1452 }, { "epoch": 0.7413265306122448, "grad_norm": 20.366559982299805, "learning_rate": 2.5867346938775512e-06, "loss": 0.0146, "step": 1453 }, { "epoch": 0.7418367346938776, "grad_norm": 72.08346557617188, "learning_rate": 2.5816326530612246e-06, "loss": 0.2332, "step": 1454 }, { "epoch": 0.7423469387755102, "grad_norm": 43.80073547363281, "learning_rate": 2.576530612244898e-06, "loss": 0.1533, "step": 1455 }, { "epoch": 0.7428571428571429, "grad_norm": 24.11647605895996, "learning_rate": 2.571428571428571e-06, "loss": 0.0473, "step": 1456 }, { "epoch": 0.7433673469387755, "grad_norm": 12.898087501525879, "learning_rate": 2.5663265306122454e-06, "loss": 0.0235, "step": 1457 }, { "epoch": 0.7438775510204082, "grad_norm": 29.00116729736328, "learning_rate": 2.5612244897959187e-06, "loss": 0.0449, "step": 1458 }, { "epoch": 0.7443877551020408, "grad_norm": 20.561870574951172, "learning_rate": 2.556122448979592e-06, "loss": 0.0278, "step": 1459 }, { "epoch": 0.7448979591836735, "grad_norm": 86.57313537597656, "learning_rate": 2.5510204081632657e-06, "loss": 0.2163, "step": 1460 }, { "epoch": 0.7454081632653061, "grad_norm": 252.12301635742188, "learning_rate": 2.545918367346939e-06, "loss": 2.3477, "step": 1461 }, { "epoch": 0.7459183673469387, "grad_norm": 17.29354476928711, "learning_rate": 2.5408163265306124e-06, "loss": 0.0283, "step": 1462 }, { "epoch": 0.7464285714285714, "grad_norm": 10.567573547363281, "learning_rate": 2.5357142857142857e-06, "loss": 0.0135, "step": 1463 }, { "epoch": 0.746938775510204, "grad_norm": 83.74082946777344, "learning_rate": 2.530612244897959e-06, "loss": 0.2482, "step": 1464 }, { "epoch": 0.7474489795918368, "grad_norm": 248.4337615966797, "learning_rate": 2.5255102040816328e-06, "loss": 1.9395, "step": 1465 }, { "epoch": 0.7479591836734694, "grad_norm": 83.67340087890625, "learning_rate": 2.5204081632653065e-06, "loss": 0.3276, "step": 1466 }, { "epoch": 0.7484693877551021, "grad_norm": 51.42272186279297, "learning_rate": 2.51530612244898e-06, "loss": 0.1858, "step": 1467 }, { "epoch": 0.7489795918367347, "grad_norm": 79.38951873779297, "learning_rate": 2.5102040816326536e-06, "loss": 0.213, "step": 1468 }, { "epoch": 0.7494897959183674, "grad_norm": 82.63340759277344, "learning_rate": 2.505102040816327e-06, "loss": 0.2478, "step": 1469 }, { "epoch": 0.75, "grad_norm": 76.12025451660156, "learning_rate": 2.5e-06, "loss": 0.1787, "step": 1470 }, { "epoch": 0.7505102040816326, "grad_norm": 16.220476150512695, "learning_rate": 2.4948979591836735e-06, "loss": 0.0112, "step": 1471 }, { "epoch": 0.7510204081632653, "grad_norm": 87.3941879272461, "learning_rate": 2.489795918367347e-06, "loss": 0.2866, "step": 1472 }, { "epoch": 0.7515306122448979, "grad_norm": 34.68866729736328, "learning_rate": 2.4846938775510206e-06, "loss": 0.045, "step": 1473 }, { "epoch": 0.7520408163265306, "grad_norm": 38.04828643798828, "learning_rate": 2.479591836734694e-06, "loss": 0.0378, "step": 1474 }, { "epoch": 0.7525510204081632, "grad_norm": 35.3651123046875, "learning_rate": 2.4744897959183676e-06, "loss": 0.0681, "step": 1475 }, { "epoch": 0.753061224489796, "grad_norm": 49.94139862060547, "learning_rate": 2.469387755102041e-06, "loss": 0.0964, "step": 1476 }, { "epoch": 0.7535714285714286, "grad_norm": 8.493317604064941, "learning_rate": 2.4642857142857147e-06, "loss": 0.0075, "step": 1477 }, { "epoch": 0.7540816326530613, "grad_norm": 42.9823112487793, "learning_rate": 2.459183673469388e-06, "loss": 0.1685, "step": 1478 }, { "epoch": 0.7545918367346939, "grad_norm": 59.73786544799805, "learning_rate": 2.4540816326530614e-06, "loss": 0.1969, "step": 1479 }, { "epoch": 0.7551020408163265, "grad_norm": 12.715458869934082, "learning_rate": 2.4489795918367347e-06, "loss": 0.0144, "step": 1480 }, { "epoch": 0.7556122448979592, "grad_norm": 24.145008087158203, "learning_rate": 2.4438775510204084e-06, "loss": 0.0468, "step": 1481 }, { "epoch": 0.7561224489795918, "grad_norm": 102.89299011230469, "learning_rate": 2.4387755102040817e-06, "loss": 0.4302, "step": 1482 }, { "epoch": 0.7566326530612245, "grad_norm": 34.67300033569336, "learning_rate": 2.4336734693877555e-06, "loss": 0.1124, "step": 1483 }, { "epoch": 0.7571428571428571, "grad_norm": 52.589637756347656, "learning_rate": 2.428571428571429e-06, "loss": 0.1419, "step": 1484 }, { "epoch": 0.7576530612244898, "grad_norm": 35.514198303222656, "learning_rate": 2.423469387755102e-06, "loss": 0.127, "step": 1485 }, { "epoch": 0.7581632653061224, "grad_norm": 25.508012771606445, "learning_rate": 2.418367346938776e-06, "loss": 0.0286, "step": 1486 }, { "epoch": 0.7586734693877552, "grad_norm": 65.10322570800781, "learning_rate": 2.413265306122449e-06, "loss": 0.2701, "step": 1487 }, { "epoch": 0.7591836734693878, "grad_norm": 19.466571807861328, "learning_rate": 2.4081632653061225e-06, "loss": 0.0359, "step": 1488 }, { "epoch": 0.7596938775510204, "grad_norm": 25.860706329345703, "learning_rate": 2.4030612244897962e-06, "loss": 0.0631, "step": 1489 }, { "epoch": 0.7602040816326531, "grad_norm": 33.01938247680664, "learning_rate": 2.3979591836734696e-06, "loss": 0.113, "step": 1490 }, { "epoch": 0.7607142857142857, "grad_norm": 57.13579559326172, "learning_rate": 2.3928571428571433e-06, "loss": 0.1775, "step": 1491 }, { "epoch": 0.7612244897959184, "grad_norm": 39.65348815917969, "learning_rate": 2.3877551020408166e-06, "loss": 0.1484, "step": 1492 }, { "epoch": 0.761734693877551, "grad_norm": 8.661036491394043, "learning_rate": 2.38265306122449e-06, "loss": 0.0102, "step": 1493 }, { "epoch": 0.7622448979591837, "grad_norm": 28.961196899414062, "learning_rate": 2.3775510204081633e-06, "loss": 0.0388, "step": 1494 }, { "epoch": 0.7627551020408163, "grad_norm": 5.049112319946289, "learning_rate": 2.372448979591837e-06, "loss": 0.0071, "step": 1495 }, { "epoch": 0.763265306122449, "grad_norm": 31.569419860839844, "learning_rate": 2.3673469387755103e-06, "loss": 0.0995, "step": 1496 }, { "epoch": 0.7637755102040816, "grad_norm": 14.157614707946777, "learning_rate": 2.362244897959184e-06, "loss": 0.0322, "step": 1497 }, { "epoch": 0.7642857142857142, "grad_norm": 235.01722717285156, "learning_rate": 2.3571428571428574e-06, "loss": 2.0586, "step": 1498 }, { "epoch": 0.764795918367347, "grad_norm": 126.89030456542969, "learning_rate": 2.3520408163265307e-06, "loss": 0.5945, "step": 1499 }, { "epoch": 0.7653061224489796, "grad_norm": 14.448871612548828, "learning_rate": 2.3469387755102044e-06, "loss": 0.0212, "step": 1500 }, { "epoch": 0.7658163265306123, "grad_norm": 20.922523498535156, "learning_rate": 2.3418367346938778e-06, "loss": 0.0349, "step": 1501 }, { "epoch": 0.7663265306122449, "grad_norm": 44.30598831176758, "learning_rate": 2.336734693877551e-06, "loss": 0.092, "step": 1502 }, { "epoch": 0.7668367346938776, "grad_norm": 10.125328063964844, "learning_rate": 2.3316326530612244e-06, "loss": 0.0184, "step": 1503 }, { "epoch": 0.7673469387755102, "grad_norm": 33.082603454589844, "learning_rate": 2.326530612244898e-06, "loss": 0.0479, "step": 1504 }, { "epoch": 0.7678571428571429, "grad_norm": 132.82418823242188, "learning_rate": 2.321428571428572e-06, "loss": 0.6182, "step": 1505 }, { "epoch": 0.7683673469387755, "grad_norm": 43.8972282409668, "learning_rate": 2.316326530612245e-06, "loss": 0.1675, "step": 1506 }, { "epoch": 0.7688775510204081, "grad_norm": 54.11611557006836, "learning_rate": 2.3112244897959185e-06, "loss": 0.2362, "step": 1507 }, { "epoch": 0.7693877551020408, "grad_norm": 114.667236328125, "learning_rate": 2.306122448979592e-06, "loss": 0.4849, "step": 1508 }, { "epoch": 0.7698979591836734, "grad_norm": 188.29730224609375, "learning_rate": 2.3010204081632656e-06, "loss": 0.6564, "step": 1509 }, { "epoch": 0.7704081632653061, "grad_norm": 15.461508750915527, "learning_rate": 2.295918367346939e-06, "loss": 0.0394, "step": 1510 }, { "epoch": 0.7709183673469387, "grad_norm": 9.587504386901855, "learning_rate": 2.2908163265306122e-06, "loss": 0.0192, "step": 1511 }, { "epoch": 0.7714285714285715, "grad_norm": 13.499245643615723, "learning_rate": 2.285714285714286e-06, "loss": 0.0182, "step": 1512 }, { "epoch": 0.7719387755102041, "grad_norm": 82.3281021118164, "learning_rate": 2.2806122448979593e-06, "loss": 0.1887, "step": 1513 }, { "epoch": 0.7724489795918368, "grad_norm": 33.15078353881836, "learning_rate": 2.275510204081633e-06, "loss": 0.0699, "step": 1514 }, { "epoch": 0.7729591836734694, "grad_norm": 43.90126419067383, "learning_rate": 2.2704081632653064e-06, "loss": 0.0753, "step": 1515 }, { "epoch": 0.773469387755102, "grad_norm": 31.114566802978516, "learning_rate": 2.2653061224489797e-06, "loss": 0.0935, "step": 1516 }, { "epoch": 0.7739795918367347, "grad_norm": 26.375986099243164, "learning_rate": 2.260204081632653e-06, "loss": 0.0256, "step": 1517 }, { "epoch": 0.7744897959183673, "grad_norm": 3.2220516204833984, "learning_rate": 2.2551020408163267e-06, "loss": 0.005, "step": 1518 }, { "epoch": 0.775, "grad_norm": 31.91994285583496, "learning_rate": 2.25e-06, "loss": 0.0411, "step": 1519 }, { "epoch": 0.7755102040816326, "grad_norm": 12.998215675354004, "learning_rate": 2.244897959183674e-06, "loss": 0.0174, "step": 1520 }, { "epoch": 0.7760204081632653, "grad_norm": 98.06134033203125, "learning_rate": 2.239795918367347e-06, "loss": 0.4042, "step": 1521 }, { "epoch": 0.7765306122448979, "grad_norm": 28.76230812072754, "learning_rate": 2.2346938775510204e-06, "loss": 0.0563, "step": 1522 }, { "epoch": 0.7770408163265307, "grad_norm": 11.840611457824707, "learning_rate": 2.229591836734694e-06, "loss": 0.0116, "step": 1523 }, { "epoch": 0.7775510204081633, "grad_norm": 58.363094329833984, "learning_rate": 2.2244897959183675e-06, "loss": 0.1617, "step": 1524 }, { "epoch": 0.7780612244897959, "grad_norm": 34.03742599487305, "learning_rate": 2.219387755102041e-06, "loss": 0.0935, "step": 1525 }, { "epoch": 0.7785714285714286, "grad_norm": 32.51577377319336, "learning_rate": 2.2142857142857146e-06, "loss": 0.071, "step": 1526 }, { "epoch": 0.7790816326530612, "grad_norm": 82.36705780029297, "learning_rate": 2.209183673469388e-06, "loss": 0.2271, "step": 1527 }, { "epoch": 0.7795918367346939, "grad_norm": 39.26930236816406, "learning_rate": 2.2040816326530616e-06, "loss": 0.1157, "step": 1528 }, { "epoch": 0.7801020408163265, "grad_norm": 8.136474609375, "learning_rate": 2.198979591836735e-06, "loss": 0.015, "step": 1529 }, { "epoch": 0.7806122448979592, "grad_norm": 137.43228149414062, "learning_rate": 2.1938775510204083e-06, "loss": 0.3573, "step": 1530 }, { "epoch": 0.7811224489795918, "grad_norm": 15.095215797424316, "learning_rate": 2.1887755102040816e-06, "loss": 0.013, "step": 1531 }, { "epoch": 0.7816326530612245, "grad_norm": 75.42620086669922, "learning_rate": 2.1836734693877553e-06, "loss": 0.2595, "step": 1532 }, { "epoch": 0.7821428571428571, "grad_norm": 43.69291687011719, "learning_rate": 2.1785714285714286e-06, "loss": 0.1562, "step": 1533 }, { "epoch": 0.7826530612244897, "grad_norm": 41.985633850097656, "learning_rate": 2.1734693877551024e-06, "loss": 0.0989, "step": 1534 }, { "epoch": 0.7831632653061225, "grad_norm": 32.59589767456055, "learning_rate": 2.1683673469387757e-06, "loss": 0.0952, "step": 1535 }, { "epoch": 0.7836734693877551, "grad_norm": 16.01462745666504, "learning_rate": 2.1632653061224495e-06, "loss": 0.0279, "step": 1536 }, { "epoch": 0.7841836734693878, "grad_norm": 47.78495788574219, "learning_rate": 2.1581632653061228e-06, "loss": 0.1318, "step": 1537 }, { "epoch": 0.7846938775510204, "grad_norm": 75.25386047363281, "learning_rate": 2.153061224489796e-06, "loss": 0.6448, "step": 1538 }, { "epoch": 0.7852040816326531, "grad_norm": 20.506999969482422, "learning_rate": 2.1479591836734694e-06, "loss": 0.0289, "step": 1539 }, { "epoch": 0.7857142857142857, "grad_norm": 27.270687103271484, "learning_rate": 2.1428571428571427e-06, "loss": 0.0262, "step": 1540 }, { "epoch": 0.7862244897959184, "grad_norm": 151.60926818847656, "learning_rate": 2.1377551020408165e-06, "loss": 0.6218, "step": 1541 }, { "epoch": 0.786734693877551, "grad_norm": 46.76213455200195, "learning_rate": 2.1326530612244902e-06, "loss": 0.165, "step": 1542 }, { "epoch": 0.7872448979591836, "grad_norm": 38.31576919555664, "learning_rate": 2.1275510204081635e-06, "loss": 0.0481, "step": 1543 }, { "epoch": 0.7877551020408163, "grad_norm": 174.6278839111328, "learning_rate": 2.122448979591837e-06, "loss": 0.4326, "step": 1544 }, { "epoch": 0.7882653061224489, "grad_norm": 86.84017944335938, "learning_rate": 2.1173469387755106e-06, "loss": 0.4597, "step": 1545 }, { "epoch": 0.7887755102040817, "grad_norm": 6.915668487548828, "learning_rate": 2.112244897959184e-06, "loss": 0.0057, "step": 1546 }, { "epoch": 0.7892857142857143, "grad_norm": 95.63287353515625, "learning_rate": 2.1071428571428572e-06, "loss": 0.4859, "step": 1547 }, { "epoch": 0.789795918367347, "grad_norm": 43.60197067260742, "learning_rate": 2.1020408163265306e-06, "loss": 0.0722, "step": 1548 }, { "epoch": 0.7903061224489796, "grad_norm": 34.78852844238281, "learning_rate": 2.0969387755102043e-06, "loss": 0.1172, "step": 1549 }, { "epoch": 0.7908163265306123, "grad_norm": 38.102638244628906, "learning_rate": 2.0918367346938776e-06, "loss": 0.0667, "step": 1550 }, { "epoch": 0.7913265306122449, "grad_norm": 60.205810546875, "learning_rate": 2.0867346938775514e-06, "loss": 0.1895, "step": 1551 }, { "epoch": 0.7918367346938775, "grad_norm": 34.19519805908203, "learning_rate": 2.0816326530612247e-06, "loss": 0.0515, "step": 1552 }, { "epoch": 0.7923469387755102, "grad_norm": 48.727352142333984, "learning_rate": 2.076530612244898e-06, "loss": 0.1292, "step": 1553 }, { "epoch": 0.7928571428571428, "grad_norm": 43.371429443359375, "learning_rate": 2.0714285714285717e-06, "loss": 0.0734, "step": 1554 }, { "epoch": 0.7933673469387755, "grad_norm": 26.997167587280273, "learning_rate": 2.066326530612245e-06, "loss": 0.0354, "step": 1555 }, { "epoch": 0.7938775510204081, "grad_norm": 66.93327331542969, "learning_rate": 2.0612244897959184e-06, "loss": 0.2485, "step": 1556 }, { "epoch": 0.7943877551020408, "grad_norm": 52.11791229248047, "learning_rate": 2.056122448979592e-06, "loss": 0.1613, "step": 1557 }, { "epoch": 0.7948979591836735, "grad_norm": 214.78028869628906, "learning_rate": 2.0510204081632654e-06, "loss": 0.6616, "step": 1558 }, { "epoch": 0.7954081632653062, "grad_norm": 9.581907272338867, "learning_rate": 2.045918367346939e-06, "loss": 0.0141, "step": 1559 }, { "epoch": 0.7959183673469388, "grad_norm": 10.442397117614746, "learning_rate": 2.0408163265306125e-06, "loss": 0.0226, "step": 1560 }, { "epoch": 0.7964285714285714, "grad_norm": 13.964073181152344, "learning_rate": 2.035714285714286e-06, "loss": 0.0178, "step": 1561 }, { "epoch": 0.7969387755102041, "grad_norm": 20.8344669342041, "learning_rate": 2.030612244897959e-06, "loss": 0.0184, "step": 1562 }, { "epoch": 0.7974489795918367, "grad_norm": 35.57036590576172, "learning_rate": 2.025510204081633e-06, "loss": 0.0567, "step": 1563 }, { "epoch": 0.7979591836734694, "grad_norm": 75.13511657714844, "learning_rate": 2.020408163265306e-06, "loss": 0.2987, "step": 1564 }, { "epoch": 0.798469387755102, "grad_norm": 82.64997100830078, "learning_rate": 2.01530612244898e-06, "loss": 0.3052, "step": 1565 }, { "epoch": 0.7989795918367347, "grad_norm": 18.5529842376709, "learning_rate": 2.0102040816326533e-06, "loss": 0.0369, "step": 1566 }, { "epoch": 0.7994897959183673, "grad_norm": 34.405487060546875, "learning_rate": 2.0051020408163266e-06, "loss": 0.0523, "step": 1567 }, { "epoch": 0.8, "grad_norm": 35.6617546081543, "learning_rate": 2.0000000000000003e-06, "loss": 0.0818, "step": 1568 }, { "epoch": 0.8005102040816326, "grad_norm": 22.76129913330078, "learning_rate": 1.9948979591836737e-06, "loss": 0.0314, "step": 1569 }, { "epoch": 0.8010204081632653, "grad_norm": 39.97718048095703, "learning_rate": 1.989795918367347e-06, "loss": 0.0826, "step": 1570 }, { "epoch": 0.801530612244898, "grad_norm": 49.059940338134766, "learning_rate": 1.9846938775510203e-06, "loss": 0.2014, "step": 1571 }, { "epoch": 0.8020408163265306, "grad_norm": 22.63507080078125, "learning_rate": 1.979591836734694e-06, "loss": 0.0408, "step": 1572 }, { "epoch": 0.8025510204081633, "grad_norm": 29.96210289001465, "learning_rate": 1.9744897959183678e-06, "loss": 0.1052, "step": 1573 }, { "epoch": 0.8030612244897959, "grad_norm": 212.38417053222656, "learning_rate": 1.969387755102041e-06, "loss": 0.4572, "step": 1574 }, { "epoch": 0.8035714285714286, "grad_norm": 97.4085922241211, "learning_rate": 1.9642857142857144e-06, "loss": 0.2998, "step": 1575 }, { "epoch": 0.8040816326530612, "grad_norm": 15.412939071655273, "learning_rate": 1.9591836734693877e-06, "loss": 0.0192, "step": 1576 }, { "epoch": 0.8045918367346939, "grad_norm": 58.076297760009766, "learning_rate": 1.9540816326530615e-06, "loss": 0.1426, "step": 1577 }, { "epoch": 0.8051020408163265, "grad_norm": 31.622852325439453, "learning_rate": 1.948979591836735e-06, "loss": 0.0502, "step": 1578 }, { "epoch": 0.8056122448979591, "grad_norm": 7.737627983093262, "learning_rate": 1.943877551020408e-06, "loss": 0.0096, "step": 1579 }, { "epoch": 0.8061224489795918, "grad_norm": 41.37223434448242, "learning_rate": 1.938775510204082e-06, "loss": 0.109, "step": 1580 }, { "epoch": 0.8066326530612244, "grad_norm": 27.352214813232422, "learning_rate": 1.933673469387755e-06, "loss": 0.0342, "step": 1581 }, { "epoch": 0.8071428571428572, "grad_norm": 200.47552490234375, "learning_rate": 1.928571428571429e-06, "loss": 1.1396, "step": 1582 }, { "epoch": 0.8076530612244898, "grad_norm": 39.45965576171875, "learning_rate": 1.9234693877551022e-06, "loss": 0.1143, "step": 1583 }, { "epoch": 0.8081632653061225, "grad_norm": 9.9906587600708, "learning_rate": 1.9183673469387756e-06, "loss": 0.0087, "step": 1584 }, { "epoch": 0.8086734693877551, "grad_norm": 13.158907890319824, "learning_rate": 1.913265306122449e-06, "loss": 0.0148, "step": 1585 }, { "epoch": 0.8091836734693878, "grad_norm": 25.028213500976562, "learning_rate": 1.9081632653061226e-06, "loss": 0.0789, "step": 1586 }, { "epoch": 0.8096938775510204, "grad_norm": 48.3575439453125, "learning_rate": 1.9030612244897962e-06, "loss": 0.1019, "step": 1587 }, { "epoch": 0.810204081632653, "grad_norm": 16.3282527923584, "learning_rate": 1.8979591836734695e-06, "loss": 0.0132, "step": 1588 }, { "epoch": 0.8107142857142857, "grad_norm": 41.098114013671875, "learning_rate": 1.892857142857143e-06, "loss": 0.0571, "step": 1589 }, { "epoch": 0.8112244897959183, "grad_norm": 67.68418884277344, "learning_rate": 1.8877551020408163e-06, "loss": 0.1499, "step": 1590 }, { "epoch": 0.811734693877551, "grad_norm": 19.04847526550293, "learning_rate": 1.88265306122449e-06, "loss": 0.0416, "step": 1591 }, { "epoch": 0.8122448979591836, "grad_norm": 136.28330993652344, "learning_rate": 1.8775510204081634e-06, "loss": 0.5708, "step": 1592 }, { "epoch": 0.8127551020408164, "grad_norm": 34.64924621582031, "learning_rate": 1.872448979591837e-06, "loss": 0.083, "step": 1593 }, { "epoch": 0.813265306122449, "grad_norm": 62.08856201171875, "learning_rate": 1.8673469387755102e-06, "loss": 0.1073, "step": 1594 }, { "epoch": 0.8137755102040817, "grad_norm": 28.567176818847656, "learning_rate": 1.862244897959184e-06, "loss": 0.0827, "step": 1595 }, { "epoch": 0.8142857142857143, "grad_norm": 101.51065826416016, "learning_rate": 1.8571428571428573e-06, "loss": 0.3472, "step": 1596 }, { "epoch": 0.814795918367347, "grad_norm": 138.28138732910156, "learning_rate": 1.8520408163265308e-06, "loss": 0.4261, "step": 1597 }, { "epoch": 0.8153061224489796, "grad_norm": 68.98278045654297, "learning_rate": 1.8469387755102042e-06, "loss": 0.234, "step": 1598 }, { "epoch": 0.8158163265306122, "grad_norm": 13.879040718078613, "learning_rate": 1.8418367346938777e-06, "loss": 0.0104, "step": 1599 }, { "epoch": 0.8163265306122449, "grad_norm": 68.20362854003906, "learning_rate": 1.8367346938775512e-06, "loss": 0.1636, "step": 1600 }, { "epoch": 0.8168367346938775, "grad_norm": 42.17952346801758, "learning_rate": 1.8316326530612247e-06, "loss": 0.0691, "step": 1601 }, { "epoch": 0.8173469387755102, "grad_norm": 63.54762268066406, "learning_rate": 1.826530612244898e-06, "loss": 0.2168, "step": 1602 }, { "epoch": 0.8178571428571428, "grad_norm": 81.17512512207031, "learning_rate": 1.8214285714285716e-06, "loss": 0.2646, "step": 1603 }, { "epoch": 0.8183673469387756, "grad_norm": 16.984235763549805, "learning_rate": 1.8163265306122451e-06, "loss": 0.0165, "step": 1604 }, { "epoch": 0.8188775510204082, "grad_norm": 45.44636917114258, "learning_rate": 1.8112244897959187e-06, "loss": 0.2047, "step": 1605 }, { "epoch": 0.8193877551020409, "grad_norm": 48.716861724853516, "learning_rate": 1.806122448979592e-06, "loss": 0.1794, "step": 1606 }, { "epoch": 0.8198979591836735, "grad_norm": 77.2822036743164, "learning_rate": 1.8010204081632653e-06, "loss": 0.1934, "step": 1607 }, { "epoch": 0.8204081632653061, "grad_norm": 45.385250091552734, "learning_rate": 1.7959183673469388e-06, "loss": 0.0789, "step": 1608 }, { "epoch": 0.8209183673469388, "grad_norm": 33.82747268676758, "learning_rate": 1.7908163265306126e-06, "loss": 0.0542, "step": 1609 }, { "epoch": 0.8214285714285714, "grad_norm": 74.98731231689453, "learning_rate": 1.7857142857142859e-06, "loss": 0.1061, "step": 1610 }, { "epoch": 0.8219387755102041, "grad_norm": 106.78945922851562, "learning_rate": 1.7806122448979592e-06, "loss": 0.2755, "step": 1611 }, { "epoch": 0.8224489795918367, "grad_norm": 132.12538146972656, "learning_rate": 1.7755102040816327e-06, "loss": 0.5573, "step": 1612 }, { "epoch": 0.8229591836734694, "grad_norm": 60.1646842956543, "learning_rate": 1.7704081632653065e-06, "loss": 0.1338, "step": 1613 }, { "epoch": 0.823469387755102, "grad_norm": 50.866024017333984, "learning_rate": 1.7653061224489798e-06, "loss": 0.1045, "step": 1614 }, { "epoch": 0.8239795918367347, "grad_norm": 21.68210220336914, "learning_rate": 1.7602040816326531e-06, "loss": 0.0378, "step": 1615 }, { "epoch": 0.8244897959183674, "grad_norm": 203.8197021484375, "learning_rate": 1.7551020408163267e-06, "loss": 0.7637, "step": 1616 }, { "epoch": 0.825, "grad_norm": 17.579792022705078, "learning_rate": 1.75e-06, "loss": 0.0311, "step": 1617 }, { "epoch": 0.8255102040816327, "grad_norm": 17.39439582824707, "learning_rate": 1.7448979591836737e-06, "loss": 0.0403, "step": 1618 }, { "epoch": 0.8260204081632653, "grad_norm": 35.72249984741211, "learning_rate": 1.739795918367347e-06, "loss": 0.0408, "step": 1619 }, { "epoch": 0.826530612244898, "grad_norm": 256.79876708984375, "learning_rate": 1.7346938775510206e-06, "loss": 1.6188, "step": 1620 }, { "epoch": 0.8270408163265306, "grad_norm": 95.08283996582031, "learning_rate": 1.7295918367346939e-06, "loss": 0.4097, "step": 1621 }, { "epoch": 0.8275510204081633, "grad_norm": 96.16004943847656, "learning_rate": 1.7244897959183676e-06, "loss": 0.3232, "step": 1622 }, { "epoch": 0.8280612244897959, "grad_norm": 39.543331146240234, "learning_rate": 1.719387755102041e-06, "loss": 0.0818, "step": 1623 }, { "epoch": 0.8285714285714286, "grad_norm": 107.75029754638672, "learning_rate": 1.7142857142857145e-06, "loss": 0.4755, "step": 1624 }, { "epoch": 0.8290816326530612, "grad_norm": 3.273580312728882, "learning_rate": 1.7091836734693878e-06, "loss": 0.0026, "step": 1625 }, { "epoch": 0.8295918367346938, "grad_norm": 53.13883590698242, "learning_rate": 1.7040816326530613e-06, "loss": 0.1068, "step": 1626 }, { "epoch": 0.8301020408163265, "grad_norm": 52.06472396850586, "learning_rate": 1.6989795918367349e-06, "loss": 0.1123, "step": 1627 }, { "epoch": 0.8306122448979592, "grad_norm": 21.84333610534668, "learning_rate": 1.6938775510204084e-06, "loss": 0.0303, "step": 1628 }, { "epoch": 0.8311224489795919, "grad_norm": 25.925434112548828, "learning_rate": 1.6887755102040817e-06, "loss": 0.0725, "step": 1629 }, { "epoch": 0.8316326530612245, "grad_norm": 14.395247459411621, "learning_rate": 1.6836734693877552e-06, "loss": 0.0095, "step": 1630 }, { "epoch": 0.8321428571428572, "grad_norm": 58.66457748413086, "learning_rate": 1.6785714285714286e-06, "loss": 0.2073, "step": 1631 }, { "epoch": 0.8326530612244898, "grad_norm": 45.20904541015625, "learning_rate": 1.6734693877551023e-06, "loss": 0.1613, "step": 1632 }, { "epoch": 0.8331632653061225, "grad_norm": 47.974578857421875, "learning_rate": 1.6683673469387756e-06, "loss": 0.1714, "step": 1633 }, { "epoch": 0.8336734693877551, "grad_norm": 18.06430435180664, "learning_rate": 1.6632653061224492e-06, "loss": 0.034, "step": 1634 }, { "epoch": 0.8341836734693877, "grad_norm": 22.524972915649414, "learning_rate": 1.6581632653061225e-06, "loss": 0.0286, "step": 1635 }, { "epoch": 0.8346938775510204, "grad_norm": 25.046804428100586, "learning_rate": 1.6530612244897962e-06, "loss": 0.0542, "step": 1636 }, { "epoch": 0.835204081632653, "grad_norm": 79.65563201904297, "learning_rate": 1.6479591836734695e-06, "loss": 0.2951, "step": 1637 }, { "epoch": 0.8357142857142857, "grad_norm": 10.093883514404297, "learning_rate": 1.642857142857143e-06, "loss": 0.015, "step": 1638 }, { "epoch": 0.8362244897959183, "grad_norm": 10.402588844299316, "learning_rate": 1.6377551020408164e-06, "loss": 0.0085, "step": 1639 }, { "epoch": 0.8367346938775511, "grad_norm": 71.53190612792969, "learning_rate": 1.6326530612244897e-06, "loss": 0.261, "step": 1640 }, { "epoch": 0.8372448979591837, "grad_norm": 74.29647064208984, "learning_rate": 1.6275510204081635e-06, "loss": 0.2866, "step": 1641 }, { "epoch": 0.8377551020408164, "grad_norm": 47.65853500366211, "learning_rate": 1.622448979591837e-06, "loss": 0.0896, "step": 1642 }, { "epoch": 0.838265306122449, "grad_norm": 14.142234802246094, "learning_rate": 1.6173469387755103e-06, "loss": 0.0456, "step": 1643 }, { "epoch": 0.8387755102040816, "grad_norm": 92.6864242553711, "learning_rate": 1.6122448979591836e-06, "loss": 0.4136, "step": 1644 }, { "epoch": 0.8392857142857143, "grad_norm": 19.498517990112305, "learning_rate": 1.6071428571428574e-06, "loss": 0.0261, "step": 1645 }, { "epoch": 0.8397959183673469, "grad_norm": 9.267862319946289, "learning_rate": 1.602040816326531e-06, "loss": 0.0175, "step": 1646 }, { "epoch": 0.8403061224489796, "grad_norm": 122.37619018554688, "learning_rate": 1.5969387755102042e-06, "loss": 0.7329, "step": 1647 }, { "epoch": 0.8408163265306122, "grad_norm": 80.23882293701172, "learning_rate": 1.5918367346938775e-06, "loss": 0.1524, "step": 1648 }, { "epoch": 0.8413265306122449, "grad_norm": 105.64079284667969, "learning_rate": 1.586734693877551e-06, "loss": 0.6533, "step": 1649 }, { "epoch": 0.8418367346938775, "grad_norm": 58.492767333984375, "learning_rate": 1.5816326530612248e-06, "loss": 0.2861, "step": 1650 }, { "epoch": 0.8423469387755103, "grad_norm": 146.7424774169922, "learning_rate": 1.5765306122448981e-06, "loss": 0.6602, "step": 1651 }, { "epoch": 0.8428571428571429, "grad_norm": 23.90764045715332, "learning_rate": 1.5714285714285714e-06, "loss": 0.0586, "step": 1652 }, { "epoch": 0.8433673469387755, "grad_norm": 5.658835411071777, "learning_rate": 1.566326530612245e-06, "loss": 0.0073, "step": 1653 }, { "epoch": 0.8438775510204082, "grad_norm": 171.6650848388672, "learning_rate": 1.5612244897959187e-06, "loss": 0.7512, "step": 1654 }, { "epoch": 0.8443877551020408, "grad_norm": 11.547285079956055, "learning_rate": 1.556122448979592e-06, "loss": 0.0193, "step": 1655 }, { "epoch": 0.8448979591836735, "grad_norm": 14.197469711303711, "learning_rate": 1.5510204081632654e-06, "loss": 0.0139, "step": 1656 }, { "epoch": 0.8454081632653061, "grad_norm": 19.023624420166016, "learning_rate": 1.5459183673469389e-06, "loss": 0.0408, "step": 1657 }, { "epoch": 0.8459183673469388, "grad_norm": 30.725902557373047, "learning_rate": 1.5408163265306122e-06, "loss": 0.0764, "step": 1658 }, { "epoch": 0.8464285714285714, "grad_norm": 35.312191009521484, "learning_rate": 1.535714285714286e-06, "loss": 0.1343, "step": 1659 }, { "epoch": 0.8469387755102041, "grad_norm": 89.01182556152344, "learning_rate": 1.5306122448979593e-06, "loss": 0.2317, "step": 1660 }, { "epoch": 0.8474489795918367, "grad_norm": 27.79344367980957, "learning_rate": 1.5255102040816328e-06, "loss": 0.0651, "step": 1661 }, { "epoch": 0.8479591836734693, "grad_norm": 45.35776901245117, "learning_rate": 1.5204081632653061e-06, "loss": 0.1357, "step": 1662 }, { "epoch": 0.8484693877551021, "grad_norm": 12.834630012512207, "learning_rate": 1.5153061224489799e-06, "loss": 0.0224, "step": 1663 }, { "epoch": 0.8489795918367347, "grad_norm": 157.52676391601562, "learning_rate": 1.5102040816326532e-06, "loss": 0.485, "step": 1664 }, { "epoch": 0.8494897959183674, "grad_norm": 226.44723510742188, "learning_rate": 1.5051020408163267e-06, "loss": 1.293, "step": 1665 }, { "epoch": 0.85, "grad_norm": 12.708098411560059, "learning_rate": 1.5e-06, "loss": 0.0205, "step": 1666 }, { "epoch": 0.8505102040816327, "grad_norm": 51.63559341430664, "learning_rate": 1.4948979591836736e-06, "loss": 0.106, "step": 1667 }, { "epoch": 0.8510204081632653, "grad_norm": 32.47344970703125, "learning_rate": 1.489795918367347e-06, "loss": 0.0616, "step": 1668 }, { "epoch": 0.851530612244898, "grad_norm": 35.2677116394043, "learning_rate": 1.4846938775510206e-06, "loss": 0.0819, "step": 1669 }, { "epoch": 0.8520408163265306, "grad_norm": 44.31740188598633, "learning_rate": 1.479591836734694e-06, "loss": 0.0737, "step": 1670 }, { "epoch": 0.8525510204081632, "grad_norm": 193.04901123046875, "learning_rate": 1.4744897959183675e-06, "loss": 1.0457, "step": 1671 }, { "epoch": 0.8530612244897959, "grad_norm": 224.84246826171875, "learning_rate": 1.469387755102041e-06, "loss": 0.9897, "step": 1672 }, { "epoch": 0.8535714285714285, "grad_norm": 22.06763458251953, "learning_rate": 1.4642857142857145e-06, "loss": 0.0498, "step": 1673 }, { "epoch": 0.8540816326530613, "grad_norm": 168.8067626953125, "learning_rate": 1.4591836734693879e-06, "loss": 0.6157, "step": 1674 }, { "epoch": 0.8545918367346939, "grad_norm": 19.34302520751953, "learning_rate": 1.4540816326530614e-06, "loss": 0.0326, "step": 1675 }, { "epoch": 0.8551020408163266, "grad_norm": 172.10694885253906, "learning_rate": 1.4489795918367347e-06, "loss": 0.6113, "step": 1676 }, { "epoch": 0.8556122448979592, "grad_norm": 12.380440711975098, "learning_rate": 1.4438775510204085e-06, "loss": 0.0105, "step": 1677 }, { "epoch": 0.8561224489795919, "grad_norm": 106.87449645996094, "learning_rate": 1.4387755102040818e-06, "loss": 0.2786, "step": 1678 }, { "epoch": 0.8566326530612245, "grad_norm": 16.902877807617188, "learning_rate": 1.433673469387755e-06, "loss": 0.0468, "step": 1679 }, { "epoch": 0.8571428571428571, "grad_norm": 77.20829772949219, "learning_rate": 1.4285714285714286e-06, "loss": 0.2466, "step": 1680 }, { "epoch": 0.8576530612244898, "grad_norm": 111.4642105102539, "learning_rate": 1.4234693877551024e-06, "loss": 0.3018, "step": 1681 }, { "epoch": 0.8581632653061224, "grad_norm": 30.602102279663086, "learning_rate": 1.4183673469387757e-06, "loss": 0.0646, "step": 1682 }, { "epoch": 0.8586734693877551, "grad_norm": 109.96634674072266, "learning_rate": 1.413265306122449e-06, "loss": 0.2781, "step": 1683 }, { "epoch": 0.8591836734693877, "grad_norm": 44.29183578491211, "learning_rate": 1.4081632653061225e-06, "loss": 0.1163, "step": 1684 }, { "epoch": 0.8596938775510204, "grad_norm": 24.841474533081055, "learning_rate": 1.4030612244897959e-06, "loss": 0.043, "step": 1685 }, { "epoch": 0.860204081632653, "grad_norm": 45.232059478759766, "learning_rate": 1.3979591836734696e-06, "loss": 0.1067, "step": 1686 }, { "epoch": 0.8607142857142858, "grad_norm": 30.001625061035156, "learning_rate": 1.392857142857143e-06, "loss": 0.0852, "step": 1687 }, { "epoch": 0.8612244897959184, "grad_norm": 36.59764099121094, "learning_rate": 1.3877551020408165e-06, "loss": 0.0557, "step": 1688 }, { "epoch": 0.861734693877551, "grad_norm": 37.628963470458984, "learning_rate": 1.3826530612244898e-06, "loss": 0.1101, "step": 1689 }, { "epoch": 0.8622448979591837, "grad_norm": 45.95731735229492, "learning_rate": 1.3775510204081633e-06, "loss": 0.0906, "step": 1690 }, { "epoch": 0.8627551020408163, "grad_norm": 16.36394691467285, "learning_rate": 1.3724489795918368e-06, "loss": 0.0178, "step": 1691 }, { "epoch": 0.863265306122449, "grad_norm": 78.62411499023438, "learning_rate": 1.3673469387755104e-06, "loss": 0.172, "step": 1692 }, { "epoch": 0.8637755102040816, "grad_norm": 30.521854400634766, "learning_rate": 1.3622448979591837e-06, "loss": 0.0566, "step": 1693 }, { "epoch": 0.8642857142857143, "grad_norm": 88.35821533203125, "learning_rate": 1.3571428571428572e-06, "loss": 0.3584, "step": 1694 }, { "epoch": 0.8647959183673469, "grad_norm": 8.724749565124512, "learning_rate": 1.3520408163265307e-06, "loss": 0.0085, "step": 1695 }, { "epoch": 0.8653061224489796, "grad_norm": 12.359358787536621, "learning_rate": 1.3469387755102043e-06, "loss": 0.0263, "step": 1696 }, { "epoch": 0.8658163265306122, "grad_norm": 23.792844772338867, "learning_rate": 1.3418367346938776e-06, "loss": 0.0547, "step": 1697 }, { "epoch": 0.8663265306122448, "grad_norm": 5.009494304656982, "learning_rate": 1.3367346938775511e-06, "loss": 0.0058, "step": 1698 }, { "epoch": 0.8668367346938776, "grad_norm": 58.78827667236328, "learning_rate": 1.3316326530612245e-06, "loss": 0.3142, "step": 1699 }, { "epoch": 0.8673469387755102, "grad_norm": 20.8699951171875, "learning_rate": 1.3265306122448982e-06, "loss": 0.032, "step": 1700 }, { "epoch": 0.8678571428571429, "grad_norm": 59.02088928222656, "learning_rate": 1.3214285714285715e-06, "loss": 0.0678, "step": 1701 }, { "epoch": 0.8683673469387755, "grad_norm": 50.45330047607422, "learning_rate": 1.316326530612245e-06, "loss": 0.1433, "step": 1702 }, { "epoch": 0.8688775510204082, "grad_norm": 18.355016708374023, "learning_rate": 1.3112244897959184e-06, "loss": 0.0277, "step": 1703 }, { "epoch": 0.8693877551020408, "grad_norm": 108.52497100830078, "learning_rate": 1.3061224489795921e-06, "loss": 0.3748, "step": 1704 }, { "epoch": 0.8698979591836735, "grad_norm": 60.76481628417969, "learning_rate": 1.3010204081632654e-06, "loss": 0.1509, "step": 1705 }, { "epoch": 0.8704081632653061, "grad_norm": 7.561622619628906, "learning_rate": 1.295918367346939e-06, "loss": 0.0123, "step": 1706 }, { "epoch": 0.8709183673469387, "grad_norm": 85.80319213867188, "learning_rate": 1.2908163265306123e-06, "loss": 0.2803, "step": 1707 }, { "epoch": 0.8714285714285714, "grad_norm": 72.44151306152344, "learning_rate": 1.2857142857142856e-06, "loss": 0.2851, "step": 1708 }, { "epoch": 0.871938775510204, "grad_norm": 20.17258644104004, "learning_rate": 1.2806122448979593e-06, "loss": 0.0449, "step": 1709 }, { "epoch": 0.8724489795918368, "grad_norm": 79.65473175048828, "learning_rate": 1.2755102040816329e-06, "loss": 0.2996, "step": 1710 }, { "epoch": 0.8729591836734694, "grad_norm": 81.82600402832031, "learning_rate": 1.2704081632653062e-06, "loss": 0.2842, "step": 1711 }, { "epoch": 0.8734693877551021, "grad_norm": 10.086222648620605, "learning_rate": 1.2653061224489795e-06, "loss": 0.0125, "step": 1712 }, { "epoch": 0.8739795918367347, "grad_norm": 129.26893615722656, "learning_rate": 1.2602040816326533e-06, "loss": 0.5156, "step": 1713 }, { "epoch": 0.8744897959183674, "grad_norm": 127.4863510131836, "learning_rate": 1.2551020408163268e-06, "loss": 0.4102, "step": 1714 }, { "epoch": 0.875, "grad_norm": 18.443086624145508, "learning_rate": 1.25e-06, "loss": 0.0417, "step": 1715 }, { "epoch": 0.8755102040816326, "grad_norm": 21.914278030395508, "learning_rate": 1.2448979591836734e-06, "loss": 0.0349, "step": 1716 }, { "epoch": 0.8760204081632653, "grad_norm": 88.07373046875, "learning_rate": 1.239795918367347e-06, "loss": 0.2705, "step": 1717 }, { "epoch": 0.8765306122448979, "grad_norm": 55.5157585144043, "learning_rate": 1.2346938775510205e-06, "loss": 0.1167, "step": 1718 }, { "epoch": 0.8770408163265306, "grad_norm": 17.817052841186523, "learning_rate": 1.229591836734694e-06, "loss": 0.029, "step": 1719 }, { "epoch": 0.8775510204081632, "grad_norm": 39.46003341674805, "learning_rate": 1.2244897959183673e-06, "loss": 0.0839, "step": 1720 }, { "epoch": 0.878061224489796, "grad_norm": 40.47589874267578, "learning_rate": 1.2193877551020409e-06, "loss": 0.0696, "step": 1721 }, { "epoch": 0.8785714285714286, "grad_norm": 130.30999755859375, "learning_rate": 1.2142857142857144e-06, "loss": 0.6223, "step": 1722 }, { "epoch": 0.8790816326530613, "grad_norm": 69.1437759399414, "learning_rate": 1.209183673469388e-06, "loss": 0.167, "step": 1723 }, { "epoch": 0.8795918367346939, "grad_norm": 53.8173942565918, "learning_rate": 1.2040816326530612e-06, "loss": 0.1035, "step": 1724 }, { "epoch": 0.8801020408163265, "grad_norm": 6.767220497131348, "learning_rate": 1.1989795918367348e-06, "loss": 0.0088, "step": 1725 }, { "epoch": 0.8806122448979592, "grad_norm": 9.070991516113281, "learning_rate": 1.1938775510204083e-06, "loss": 0.0071, "step": 1726 }, { "epoch": 0.8811224489795918, "grad_norm": 48.50843048095703, "learning_rate": 1.1887755102040816e-06, "loss": 0.0635, "step": 1727 }, { "epoch": 0.8816326530612245, "grad_norm": 108.60492706298828, "learning_rate": 1.1836734693877552e-06, "loss": 0.25, "step": 1728 }, { "epoch": 0.8821428571428571, "grad_norm": 11.828723907470703, "learning_rate": 1.1785714285714287e-06, "loss": 0.0298, "step": 1729 }, { "epoch": 0.8826530612244898, "grad_norm": 21.794103622436523, "learning_rate": 1.1734693877551022e-06, "loss": 0.0212, "step": 1730 }, { "epoch": 0.8831632653061224, "grad_norm": 45.481414794921875, "learning_rate": 1.1683673469387755e-06, "loss": 0.1499, "step": 1731 }, { "epoch": 0.8836734693877552, "grad_norm": 11.526408195495605, "learning_rate": 1.163265306122449e-06, "loss": 0.0166, "step": 1732 }, { "epoch": 0.8841836734693878, "grad_norm": 50.062782287597656, "learning_rate": 1.1581632653061226e-06, "loss": 0.1347, "step": 1733 }, { "epoch": 0.8846938775510204, "grad_norm": 43.38827133178711, "learning_rate": 1.153061224489796e-06, "loss": 0.0709, "step": 1734 }, { "epoch": 0.8852040816326531, "grad_norm": 32.921302795410156, "learning_rate": 1.1479591836734695e-06, "loss": 0.0579, "step": 1735 }, { "epoch": 0.8857142857142857, "grad_norm": 34.36521530151367, "learning_rate": 1.142857142857143e-06, "loss": 0.0653, "step": 1736 }, { "epoch": 0.8862244897959184, "grad_norm": 7.297756671905518, "learning_rate": 1.1377551020408165e-06, "loss": 0.0089, "step": 1737 }, { "epoch": 0.886734693877551, "grad_norm": 14.430858612060547, "learning_rate": 1.1326530612244898e-06, "loss": 0.0117, "step": 1738 }, { "epoch": 0.8872448979591837, "grad_norm": 9.275567054748535, "learning_rate": 1.1275510204081634e-06, "loss": 0.0102, "step": 1739 }, { "epoch": 0.8877551020408163, "grad_norm": 27.939598083496094, "learning_rate": 1.122448979591837e-06, "loss": 0.0468, "step": 1740 }, { "epoch": 0.888265306122449, "grad_norm": 71.71142578125, "learning_rate": 1.1173469387755102e-06, "loss": 0.0961, "step": 1741 }, { "epoch": 0.8887755102040816, "grad_norm": 5.4155354499816895, "learning_rate": 1.1122448979591838e-06, "loss": 0.0057, "step": 1742 }, { "epoch": 0.8892857142857142, "grad_norm": 51.73462677001953, "learning_rate": 1.1071428571428573e-06, "loss": 0.1196, "step": 1743 }, { "epoch": 0.889795918367347, "grad_norm": 60.26512145996094, "learning_rate": 1.1020408163265308e-06, "loss": 0.1974, "step": 1744 }, { "epoch": 0.8903061224489796, "grad_norm": 15.46298599243164, "learning_rate": 1.0969387755102041e-06, "loss": 0.013, "step": 1745 }, { "epoch": 0.8908163265306123, "grad_norm": 64.82288360595703, "learning_rate": 1.0918367346938777e-06, "loss": 0.1934, "step": 1746 }, { "epoch": 0.8913265306122449, "grad_norm": 9.879206657409668, "learning_rate": 1.0867346938775512e-06, "loss": 0.0187, "step": 1747 }, { "epoch": 0.8918367346938776, "grad_norm": 77.44990539550781, "learning_rate": 1.0816326530612247e-06, "loss": 0.2288, "step": 1748 }, { "epoch": 0.8923469387755102, "grad_norm": 8.562651634216309, "learning_rate": 1.076530612244898e-06, "loss": 0.0053, "step": 1749 }, { "epoch": 0.8928571428571429, "grad_norm": 15.115448951721191, "learning_rate": 1.0714285714285714e-06, "loss": 0.0093, "step": 1750 }, { "epoch": 0.8933673469387755, "grad_norm": 27.10320281982422, "learning_rate": 1.0663265306122451e-06, "loss": 0.0556, "step": 1751 }, { "epoch": 0.8938775510204081, "grad_norm": 87.38886260986328, "learning_rate": 1.0612244897959184e-06, "loss": 0.3447, "step": 1752 }, { "epoch": 0.8943877551020408, "grad_norm": 51.67900848388672, "learning_rate": 1.056122448979592e-06, "loss": 0.1438, "step": 1753 }, { "epoch": 0.8948979591836734, "grad_norm": 27.0687255859375, "learning_rate": 1.0510204081632653e-06, "loss": 0.0336, "step": 1754 }, { "epoch": 0.8954081632653061, "grad_norm": 50.72376251220703, "learning_rate": 1.0459183673469388e-06, "loss": 0.1819, "step": 1755 }, { "epoch": 0.8959183673469387, "grad_norm": 168.94314575195312, "learning_rate": 1.0408163265306123e-06, "loss": 0.5978, "step": 1756 }, { "epoch": 0.8964285714285715, "grad_norm": 17.114953994750977, "learning_rate": 1.0357142857142859e-06, "loss": 0.0228, "step": 1757 }, { "epoch": 0.8969387755102041, "grad_norm": 37.79789733886719, "learning_rate": 1.0306122448979592e-06, "loss": 0.0859, "step": 1758 }, { "epoch": 0.8974489795918368, "grad_norm": 8.591856956481934, "learning_rate": 1.0255102040816327e-06, "loss": 0.0124, "step": 1759 }, { "epoch": 0.8979591836734694, "grad_norm": 52.08307647705078, "learning_rate": 1.0204081632653063e-06, "loss": 0.1575, "step": 1760 }, { "epoch": 0.898469387755102, "grad_norm": 23.553621292114258, "learning_rate": 1.0153061224489796e-06, "loss": 0.0389, "step": 1761 }, { "epoch": 0.8989795918367347, "grad_norm": 139.28387451171875, "learning_rate": 1.010204081632653e-06, "loss": 0.4758, "step": 1762 }, { "epoch": 0.8994897959183673, "grad_norm": 103.49319458007812, "learning_rate": 1.0051020408163266e-06, "loss": 0.5459, "step": 1763 }, { "epoch": 0.9, "grad_norm": 41.79087829589844, "learning_rate": 1.0000000000000002e-06, "loss": 0.0992, "step": 1764 }, { "epoch": 0.9005102040816326, "grad_norm": 53.32533645629883, "learning_rate": 9.948979591836735e-07, "loss": 0.1108, "step": 1765 }, { "epoch": 0.9010204081632653, "grad_norm": 54.43605422973633, "learning_rate": 9.89795918367347e-07, "loss": 0.1982, "step": 1766 }, { "epoch": 0.9015306122448979, "grad_norm": 64.04644012451172, "learning_rate": 9.846938775510205e-07, "loss": 0.1526, "step": 1767 }, { "epoch": 0.9020408163265307, "grad_norm": 27.412206649780273, "learning_rate": 9.795918367346939e-07, "loss": 0.0446, "step": 1768 }, { "epoch": 0.9025510204081633, "grad_norm": 116.86224365234375, "learning_rate": 9.744897959183674e-07, "loss": 0.3848, "step": 1769 }, { "epoch": 0.9030612244897959, "grad_norm": 32.51971435546875, "learning_rate": 9.69387755102041e-07, "loss": 0.0366, "step": 1770 }, { "epoch": 0.9035714285714286, "grad_norm": 65.04561614990234, "learning_rate": 9.642857142857145e-07, "loss": 0.1499, "step": 1771 }, { "epoch": 0.9040816326530612, "grad_norm": 34.000640869140625, "learning_rate": 9.591836734693878e-07, "loss": 0.0657, "step": 1772 }, { "epoch": 0.9045918367346939, "grad_norm": 22.766403198242188, "learning_rate": 9.540816326530613e-07, "loss": 0.0202, "step": 1773 }, { "epoch": 0.9051020408163265, "grad_norm": 10.108830451965332, "learning_rate": 9.489795918367347e-07, "loss": 0.0121, "step": 1774 }, { "epoch": 0.9056122448979592, "grad_norm": 42.229915618896484, "learning_rate": 9.438775510204082e-07, "loss": 0.0957, "step": 1775 }, { "epoch": 0.9061224489795918, "grad_norm": 33.78862380981445, "learning_rate": 9.387755102040817e-07, "loss": 0.0598, "step": 1776 }, { "epoch": 0.9066326530612245, "grad_norm": 43.12638473510742, "learning_rate": 9.336734693877551e-07, "loss": 0.1104, "step": 1777 }, { "epoch": 0.9071428571428571, "grad_norm": 30.806509017944336, "learning_rate": 9.285714285714287e-07, "loss": 0.0664, "step": 1778 }, { "epoch": 0.9076530612244897, "grad_norm": 64.92422485351562, "learning_rate": 9.234693877551021e-07, "loss": 0.1735, "step": 1779 }, { "epoch": 0.9081632653061225, "grad_norm": 52.63892364501953, "learning_rate": 9.183673469387756e-07, "loss": 0.1602, "step": 1780 }, { "epoch": 0.9086734693877551, "grad_norm": 102.01897430419922, "learning_rate": 9.13265306122449e-07, "loss": 0.7076, "step": 1781 }, { "epoch": 0.9091836734693878, "grad_norm": 107.39047241210938, "learning_rate": 9.081632653061226e-07, "loss": 0.5422, "step": 1782 }, { "epoch": 0.9096938775510204, "grad_norm": 69.64804077148438, "learning_rate": 9.03061224489796e-07, "loss": 0.2856, "step": 1783 }, { "epoch": 0.9102040816326531, "grad_norm": 24.721681594848633, "learning_rate": 8.979591836734694e-07, "loss": 0.0424, "step": 1784 }, { "epoch": 0.9107142857142857, "grad_norm": 44.75162124633789, "learning_rate": 8.928571428571429e-07, "loss": 0.0717, "step": 1785 }, { "epoch": 0.9112244897959184, "grad_norm": 61.8922119140625, "learning_rate": 8.877551020408164e-07, "loss": 0.2554, "step": 1786 }, { "epoch": 0.911734693877551, "grad_norm": 9.52384090423584, "learning_rate": 8.826530612244899e-07, "loss": 0.0127, "step": 1787 }, { "epoch": 0.9122448979591836, "grad_norm": 44.36495590209961, "learning_rate": 8.775510204081633e-07, "loss": 0.1001, "step": 1788 }, { "epoch": 0.9127551020408163, "grad_norm": 15.083593368530273, "learning_rate": 8.724489795918369e-07, "loss": 0.0253, "step": 1789 }, { "epoch": 0.9132653061224489, "grad_norm": 9.114136695861816, "learning_rate": 8.673469387755103e-07, "loss": 0.0095, "step": 1790 }, { "epoch": 0.9137755102040817, "grad_norm": 8.574913024902344, "learning_rate": 8.622448979591838e-07, "loss": 0.0029, "step": 1791 }, { "epoch": 0.9142857142857143, "grad_norm": 43.88521957397461, "learning_rate": 8.571428571428572e-07, "loss": 0.1135, "step": 1792 }, { "epoch": 0.914795918367347, "grad_norm": 128.2406005859375, "learning_rate": 8.520408163265307e-07, "loss": 0.3072, "step": 1793 }, { "epoch": 0.9153061224489796, "grad_norm": 42.145755767822266, "learning_rate": 8.469387755102042e-07, "loss": 0.072, "step": 1794 }, { "epoch": 0.9158163265306123, "grad_norm": 38.136085510253906, "learning_rate": 8.418367346938776e-07, "loss": 0.084, "step": 1795 }, { "epoch": 0.9163265306122449, "grad_norm": 61.06956100463867, "learning_rate": 8.367346938775512e-07, "loss": 0.2144, "step": 1796 }, { "epoch": 0.9168367346938775, "grad_norm": 28.127412796020508, "learning_rate": 8.316326530612246e-07, "loss": 0.0552, "step": 1797 }, { "epoch": 0.9173469387755102, "grad_norm": 45.883758544921875, "learning_rate": 8.265306122448981e-07, "loss": 0.1448, "step": 1798 }, { "epoch": 0.9178571428571428, "grad_norm": 19.285743713378906, "learning_rate": 8.214285714285715e-07, "loss": 0.0439, "step": 1799 }, { "epoch": 0.9183673469387755, "grad_norm": 13.482954025268555, "learning_rate": 8.163265306122449e-07, "loss": 0.0229, "step": 1800 }, { "epoch": 0.9188775510204081, "grad_norm": 63.35903549194336, "learning_rate": 8.112244897959185e-07, "loss": 0.3052, "step": 1801 }, { "epoch": 0.9193877551020408, "grad_norm": 42.6038703918457, "learning_rate": 8.061224489795918e-07, "loss": 0.1074, "step": 1802 }, { "epoch": 0.9198979591836735, "grad_norm": 20.829423904418945, "learning_rate": 8.010204081632654e-07, "loss": 0.0215, "step": 1803 }, { "epoch": 0.9204081632653062, "grad_norm": 57.93727111816406, "learning_rate": 7.959183673469388e-07, "loss": 0.1621, "step": 1804 }, { "epoch": 0.9209183673469388, "grad_norm": 25.5330753326416, "learning_rate": 7.908163265306124e-07, "loss": 0.0439, "step": 1805 }, { "epoch": 0.9214285714285714, "grad_norm": 11.569315910339355, "learning_rate": 7.857142857142857e-07, "loss": 0.0071, "step": 1806 }, { "epoch": 0.9219387755102041, "grad_norm": 19.363128662109375, "learning_rate": 7.806122448979594e-07, "loss": 0.0328, "step": 1807 }, { "epoch": 0.9224489795918367, "grad_norm": 39.65160369873047, "learning_rate": 7.755102040816327e-07, "loss": 0.1042, "step": 1808 }, { "epoch": 0.9229591836734694, "grad_norm": 83.85896301269531, "learning_rate": 7.704081632653061e-07, "loss": 0.3469, "step": 1809 }, { "epoch": 0.923469387755102, "grad_norm": 26.37398338317871, "learning_rate": 7.653061224489796e-07, "loss": 0.0373, "step": 1810 }, { "epoch": 0.9239795918367347, "grad_norm": 23.56862449645996, "learning_rate": 7.602040816326531e-07, "loss": 0.0669, "step": 1811 }, { "epoch": 0.9244897959183673, "grad_norm": 97.2119369506836, "learning_rate": 7.551020408163266e-07, "loss": 0.3887, "step": 1812 }, { "epoch": 0.925, "grad_norm": 19.859603881835938, "learning_rate": 7.5e-07, "loss": 0.0299, "step": 1813 }, { "epoch": 0.9255102040816326, "grad_norm": 87.05843353271484, "learning_rate": 7.448979591836736e-07, "loss": 0.2724, "step": 1814 }, { "epoch": 0.9260204081632653, "grad_norm": 134.05902099609375, "learning_rate": 7.39795918367347e-07, "loss": 0.4277, "step": 1815 }, { "epoch": 0.926530612244898, "grad_norm": 33.366668701171875, "learning_rate": 7.346938775510205e-07, "loss": 0.0521, "step": 1816 }, { "epoch": 0.9270408163265306, "grad_norm": 15.293240547180176, "learning_rate": 7.295918367346939e-07, "loss": 0.0296, "step": 1817 }, { "epoch": 0.9275510204081633, "grad_norm": 37.8026237487793, "learning_rate": 7.244897959183674e-07, "loss": 0.0482, "step": 1818 }, { "epoch": 0.9280612244897959, "grad_norm": 35.44809341430664, "learning_rate": 7.193877551020409e-07, "loss": 0.072, "step": 1819 }, { "epoch": 0.9285714285714286, "grad_norm": 49.77033615112305, "learning_rate": 7.142857142857143e-07, "loss": 0.0874, "step": 1820 }, { "epoch": 0.9290816326530612, "grad_norm": 25.838193893432617, "learning_rate": 7.091836734693878e-07, "loss": 0.0265, "step": 1821 }, { "epoch": 0.9295918367346939, "grad_norm": 84.27186584472656, "learning_rate": 7.040816326530613e-07, "loss": 0.2347, "step": 1822 }, { "epoch": 0.9301020408163265, "grad_norm": 23.853670120239258, "learning_rate": 6.989795918367348e-07, "loss": 0.0284, "step": 1823 }, { "epoch": 0.9306122448979591, "grad_norm": 27.78782844543457, "learning_rate": 6.938775510204082e-07, "loss": 0.0276, "step": 1824 }, { "epoch": 0.9311224489795918, "grad_norm": 10.703680992126465, "learning_rate": 6.887755102040817e-07, "loss": 0.0113, "step": 1825 }, { "epoch": 0.9316326530612244, "grad_norm": 39.53456115722656, "learning_rate": 6.836734693877552e-07, "loss": 0.0594, "step": 1826 }, { "epoch": 0.9321428571428572, "grad_norm": 52.501800537109375, "learning_rate": 6.785714285714286e-07, "loss": 0.119, "step": 1827 }, { "epoch": 0.9326530612244898, "grad_norm": 52.031211853027344, "learning_rate": 6.734693877551021e-07, "loss": 0.1611, "step": 1828 }, { "epoch": 0.9331632653061225, "grad_norm": 161.4055938720703, "learning_rate": 6.683673469387756e-07, "loss": 0.9683, "step": 1829 }, { "epoch": 0.9336734693877551, "grad_norm": 25.514795303344727, "learning_rate": 6.632653061224491e-07, "loss": 0.0328, "step": 1830 }, { "epoch": 0.9341836734693878, "grad_norm": 21.19870948791504, "learning_rate": 6.581632653061225e-07, "loss": 0.0298, "step": 1831 }, { "epoch": 0.9346938775510204, "grad_norm": 65.30322265625, "learning_rate": 6.530612244897961e-07, "loss": 0.1914, "step": 1832 }, { "epoch": 0.935204081632653, "grad_norm": 80.42169189453125, "learning_rate": 6.479591836734695e-07, "loss": 0.2172, "step": 1833 }, { "epoch": 0.9357142857142857, "grad_norm": 14.232757568359375, "learning_rate": 6.428571428571428e-07, "loss": 0.0112, "step": 1834 }, { "epoch": 0.9362244897959183, "grad_norm": 9.400615692138672, "learning_rate": 6.377551020408164e-07, "loss": 0.0125, "step": 1835 }, { "epoch": 0.936734693877551, "grad_norm": 15.705513954162598, "learning_rate": 6.326530612244898e-07, "loss": 0.0264, "step": 1836 }, { "epoch": 0.9372448979591836, "grad_norm": 25.730365753173828, "learning_rate": 6.275510204081634e-07, "loss": 0.0424, "step": 1837 }, { "epoch": 0.9377551020408164, "grad_norm": 19.419479370117188, "learning_rate": 6.224489795918367e-07, "loss": 0.0255, "step": 1838 }, { "epoch": 0.938265306122449, "grad_norm": 56.0516357421875, "learning_rate": 6.173469387755102e-07, "loss": 0.124, "step": 1839 }, { "epoch": 0.9387755102040817, "grad_norm": 28.035486221313477, "learning_rate": 6.122448979591837e-07, "loss": 0.0425, "step": 1840 }, { "epoch": 0.9392857142857143, "grad_norm": 22.393129348754883, "learning_rate": 6.071428571428572e-07, "loss": 0.0219, "step": 1841 }, { "epoch": 0.939795918367347, "grad_norm": 6.1879119873046875, "learning_rate": 6.020408163265306e-07, "loss": 0.0041, "step": 1842 }, { "epoch": 0.9403061224489796, "grad_norm": 192.1066131591797, "learning_rate": 5.969387755102042e-07, "loss": 0.792, "step": 1843 }, { "epoch": 0.9408163265306122, "grad_norm": 21.685039520263672, "learning_rate": 5.918367346938776e-07, "loss": 0.0493, "step": 1844 }, { "epoch": 0.9413265306122449, "grad_norm": 30.35733413696289, "learning_rate": 5.867346938775511e-07, "loss": 0.0739, "step": 1845 }, { "epoch": 0.9418367346938775, "grad_norm": 149.482421875, "learning_rate": 5.816326530612245e-07, "loss": 0.5303, "step": 1846 }, { "epoch": 0.9423469387755102, "grad_norm": 18.289812088012695, "learning_rate": 5.76530612244898e-07, "loss": 0.0159, "step": 1847 }, { "epoch": 0.9428571428571428, "grad_norm": 14.96845817565918, "learning_rate": 5.714285714285715e-07, "loss": 0.0124, "step": 1848 }, { "epoch": 0.9433673469387756, "grad_norm": 20.382551193237305, "learning_rate": 5.663265306122449e-07, "loss": 0.0355, "step": 1849 }, { "epoch": 0.9438775510204082, "grad_norm": 140.33241271972656, "learning_rate": 5.612244897959184e-07, "loss": 0.2527, "step": 1850 }, { "epoch": 0.9443877551020409, "grad_norm": 48.51268768310547, "learning_rate": 5.561224489795919e-07, "loss": 0.1185, "step": 1851 }, { "epoch": 0.9448979591836735, "grad_norm": 232.44107055664062, "learning_rate": 5.510204081632654e-07, "loss": 0.8978, "step": 1852 }, { "epoch": 0.9454081632653061, "grad_norm": 172.5355682373047, "learning_rate": 5.459183673469388e-07, "loss": 0.7322, "step": 1853 }, { "epoch": 0.9459183673469388, "grad_norm": 54.25844955444336, "learning_rate": 5.408163265306124e-07, "loss": 0.2705, "step": 1854 }, { "epoch": 0.9464285714285714, "grad_norm": 29.3337345123291, "learning_rate": 5.357142857142857e-07, "loss": 0.0863, "step": 1855 }, { "epoch": 0.9469387755102041, "grad_norm": 29.966856002807617, "learning_rate": 5.306122448979592e-07, "loss": 0.0671, "step": 1856 }, { "epoch": 0.9474489795918367, "grad_norm": 359.2778015136719, "learning_rate": 5.255102040816326e-07, "loss": 1.2808, "step": 1857 }, { "epoch": 0.9479591836734694, "grad_norm": 64.52867126464844, "learning_rate": 5.204081632653062e-07, "loss": 0.1335, "step": 1858 }, { "epoch": 0.948469387755102, "grad_norm": 16.693159103393555, "learning_rate": 5.153061224489796e-07, "loss": 0.0167, "step": 1859 }, { "epoch": 0.9489795918367347, "grad_norm": 4.164316177368164, "learning_rate": 5.102040816326531e-07, "loss": 0.0044, "step": 1860 }, { "epoch": 0.9494897959183674, "grad_norm": 50.277099609375, "learning_rate": 5.051020408163266e-07, "loss": 0.1725, "step": 1861 }, { "epoch": 0.95, "grad_norm": 80.37914276123047, "learning_rate": 5.000000000000001e-07, "loss": 0.137, "step": 1862 }, { "epoch": 0.9505102040816327, "grad_norm": 225.8207550048828, "learning_rate": 4.948979591836735e-07, "loss": 0.6211, "step": 1863 }, { "epoch": 0.9510204081632653, "grad_norm": 58.20606231689453, "learning_rate": 4.897959183673469e-07, "loss": 0.2397, "step": 1864 }, { "epoch": 0.951530612244898, "grad_norm": 40.430110931396484, "learning_rate": 4.846938775510205e-07, "loss": 0.0502, "step": 1865 }, { "epoch": 0.9520408163265306, "grad_norm": 51.28953170776367, "learning_rate": 4.795918367346939e-07, "loss": 0.2307, "step": 1866 }, { "epoch": 0.9525510204081633, "grad_norm": 45.20911407470703, "learning_rate": 4.7448979591836737e-07, "loss": 0.1075, "step": 1867 }, { "epoch": 0.9530612244897959, "grad_norm": 70.71084594726562, "learning_rate": 4.6938775510204085e-07, "loss": 0.3982, "step": 1868 }, { "epoch": 0.9535714285714286, "grad_norm": 55.147430419921875, "learning_rate": 4.642857142857143e-07, "loss": 0.0749, "step": 1869 }, { "epoch": 0.9540816326530612, "grad_norm": 23.354724884033203, "learning_rate": 4.591836734693878e-07, "loss": 0.0436, "step": 1870 }, { "epoch": 0.9545918367346938, "grad_norm": 14.571335792541504, "learning_rate": 4.540816326530613e-07, "loss": 0.0209, "step": 1871 }, { "epoch": 0.9551020408163265, "grad_norm": 39.840789794921875, "learning_rate": 4.489795918367347e-07, "loss": 0.0749, "step": 1872 }, { "epoch": 0.9556122448979592, "grad_norm": 18.900100708007812, "learning_rate": 4.438775510204082e-07, "loss": 0.0272, "step": 1873 }, { "epoch": 0.9561224489795919, "grad_norm": 31.33615493774414, "learning_rate": 4.3877551020408166e-07, "loss": 0.0519, "step": 1874 }, { "epoch": 0.9566326530612245, "grad_norm": 55.044769287109375, "learning_rate": 4.3367346938775514e-07, "loss": 0.1044, "step": 1875 }, { "epoch": 0.9571428571428572, "grad_norm": 44.02341842651367, "learning_rate": 4.285714285714286e-07, "loss": 0.1319, "step": 1876 }, { "epoch": 0.9576530612244898, "grad_norm": 19.920608520507812, "learning_rate": 4.234693877551021e-07, "loss": 0.0186, "step": 1877 }, { "epoch": 0.9581632653061225, "grad_norm": 21.928133010864258, "learning_rate": 4.183673469387756e-07, "loss": 0.02, "step": 1878 }, { "epoch": 0.9586734693877551, "grad_norm": 60.0349235534668, "learning_rate": 4.1326530612244905e-07, "loss": 0.1196, "step": 1879 }, { "epoch": 0.9591836734693877, "grad_norm": 175.4227752685547, "learning_rate": 4.0816326530612243e-07, "loss": 0.6348, "step": 1880 }, { "epoch": 0.9596938775510204, "grad_norm": 70.27619934082031, "learning_rate": 4.030612244897959e-07, "loss": 0.2231, "step": 1881 }, { "epoch": 0.960204081632653, "grad_norm": 19.09709358215332, "learning_rate": 3.979591836734694e-07, "loss": 0.0314, "step": 1882 }, { "epoch": 0.9607142857142857, "grad_norm": 57.044918060302734, "learning_rate": 3.9285714285714286e-07, "loss": 0.1221, "step": 1883 }, { "epoch": 0.9612244897959183, "grad_norm": 17.479978561401367, "learning_rate": 3.8775510204081634e-07, "loss": 0.0235, "step": 1884 }, { "epoch": 0.9617346938775511, "grad_norm": 19.57831382751465, "learning_rate": 3.826530612244898e-07, "loss": 0.0508, "step": 1885 }, { "epoch": 0.9622448979591837, "grad_norm": 18.676528930664062, "learning_rate": 3.775510204081633e-07, "loss": 0.0252, "step": 1886 }, { "epoch": 0.9627551020408164, "grad_norm": 29.298446655273438, "learning_rate": 3.724489795918368e-07, "loss": 0.0934, "step": 1887 }, { "epoch": 0.963265306122449, "grad_norm": 25.776002883911133, "learning_rate": 3.6734693877551025e-07, "loss": 0.0494, "step": 1888 }, { "epoch": 0.9637755102040816, "grad_norm": 27.09151268005371, "learning_rate": 3.622448979591837e-07, "loss": 0.0657, "step": 1889 }, { "epoch": 0.9642857142857143, "grad_norm": 38.82861328125, "learning_rate": 3.5714285714285716e-07, "loss": 0.0869, "step": 1890 }, { "epoch": 0.9647959183673469, "grad_norm": 10.882378578186035, "learning_rate": 3.5204081632653064e-07, "loss": 0.0146, "step": 1891 }, { "epoch": 0.9653061224489796, "grad_norm": 17.872644424438477, "learning_rate": 3.469387755102041e-07, "loss": 0.0277, "step": 1892 }, { "epoch": 0.9658163265306122, "grad_norm": 15.142598152160645, "learning_rate": 3.418367346938776e-07, "loss": 0.0306, "step": 1893 }, { "epoch": 0.9663265306122449, "grad_norm": 35.74921798706055, "learning_rate": 3.3673469387755107e-07, "loss": 0.1062, "step": 1894 }, { "epoch": 0.9668367346938775, "grad_norm": 21.885042190551758, "learning_rate": 3.3163265306122455e-07, "loss": 0.0334, "step": 1895 }, { "epoch": 0.9673469387755103, "grad_norm": 47.778167724609375, "learning_rate": 3.2653061224489803e-07, "loss": 0.1049, "step": 1896 }, { "epoch": 0.9678571428571429, "grad_norm": 36.559059143066406, "learning_rate": 3.214285714285714e-07, "loss": 0.0681, "step": 1897 }, { "epoch": 0.9683673469387755, "grad_norm": 31.44573402404785, "learning_rate": 3.163265306122449e-07, "loss": 0.0649, "step": 1898 }, { "epoch": 0.9688775510204082, "grad_norm": 32.41381072998047, "learning_rate": 3.1122448979591836e-07, "loss": 0.0637, "step": 1899 }, { "epoch": 0.9693877551020408, "grad_norm": 42.2923469543457, "learning_rate": 3.0612244897959183e-07, "loss": 0.0664, "step": 1900 }, { "epoch": 0.9698979591836735, "grad_norm": 37.621517181396484, "learning_rate": 3.010204081632653e-07, "loss": 0.0548, "step": 1901 }, { "epoch": 0.9704081632653061, "grad_norm": 29.556533813476562, "learning_rate": 2.959183673469388e-07, "loss": 0.0686, "step": 1902 }, { "epoch": 0.9709183673469388, "grad_norm": 17.86480712890625, "learning_rate": 2.9081632653061227e-07, "loss": 0.0183, "step": 1903 }, { "epoch": 0.9714285714285714, "grad_norm": 32.14585494995117, "learning_rate": 2.8571428571428575e-07, "loss": 0.0543, "step": 1904 }, { "epoch": 0.9719387755102041, "grad_norm": 21.268980026245117, "learning_rate": 2.806122448979592e-07, "loss": 0.024, "step": 1905 }, { "epoch": 0.9724489795918367, "grad_norm": 18.416324615478516, "learning_rate": 2.755102040816327e-07, "loss": 0.0186, "step": 1906 }, { "epoch": 0.9729591836734693, "grad_norm": 149.3097381591797, "learning_rate": 2.704081632653062e-07, "loss": 0.5932, "step": 1907 }, { "epoch": 0.9734693877551021, "grad_norm": 265.5702819824219, "learning_rate": 2.653061224489796e-07, "loss": 1.5978, "step": 1908 }, { "epoch": 0.9739795918367347, "grad_norm": 217.5521697998047, "learning_rate": 2.602040816326531e-07, "loss": 0.6175, "step": 1909 }, { "epoch": 0.9744897959183674, "grad_norm": 50.80287170410156, "learning_rate": 2.5510204081632656e-07, "loss": 0.0544, "step": 1910 }, { "epoch": 0.975, "grad_norm": 36.13534164428711, "learning_rate": 2.5000000000000004e-07, "loss": 0.0762, "step": 1911 }, { "epoch": 0.9755102040816327, "grad_norm": 13.718497276306152, "learning_rate": 2.4489795918367347e-07, "loss": 0.0157, "step": 1912 }, { "epoch": 0.9760204081632653, "grad_norm": 22.25408172607422, "learning_rate": 2.3979591836734695e-07, "loss": 0.0382, "step": 1913 }, { "epoch": 0.976530612244898, "grad_norm": 33.572242736816406, "learning_rate": 2.3469387755102042e-07, "loss": 0.0585, "step": 1914 }, { "epoch": 0.9770408163265306, "grad_norm": 17.347084045410156, "learning_rate": 2.295918367346939e-07, "loss": 0.0191, "step": 1915 }, { "epoch": 0.9775510204081632, "grad_norm": 36.30152130126953, "learning_rate": 2.2448979591836735e-07, "loss": 0.1267, "step": 1916 }, { "epoch": 0.9780612244897959, "grad_norm": 49.27014923095703, "learning_rate": 2.1938775510204083e-07, "loss": 0.0899, "step": 1917 }, { "epoch": 0.9785714285714285, "grad_norm": 22.828025817871094, "learning_rate": 2.142857142857143e-07, "loss": 0.0421, "step": 1918 }, { "epoch": 0.9790816326530613, "grad_norm": 4.881752014160156, "learning_rate": 2.091836734693878e-07, "loss": 0.0092, "step": 1919 }, { "epoch": 0.9795918367346939, "grad_norm": 54.7541618347168, "learning_rate": 2.0408163265306121e-07, "loss": 0.1208, "step": 1920 }, { "epoch": 0.9801020408163266, "grad_norm": 33.822330474853516, "learning_rate": 1.989795918367347e-07, "loss": 0.0377, "step": 1921 }, { "epoch": 0.9806122448979592, "grad_norm": 25.09183120727539, "learning_rate": 1.9387755102040817e-07, "loss": 0.037, "step": 1922 }, { "epoch": 0.9811224489795919, "grad_norm": 20.31435775756836, "learning_rate": 1.8877551020408165e-07, "loss": 0.05, "step": 1923 }, { "epoch": 0.9816326530612245, "grad_norm": 25.90929412841797, "learning_rate": 1.8367346938775513e-07, "loss": 0.0502, "step": 1924 }, { "epoch": 0.9821428571428571, "grad_norm": 12.506081581115723, "learning_rate": 1.7857142857142858e-07, "loss": 0.0383, "step": 1925 }, { "epoch": 0.9826530612244898, "grad_norm": 49.26741027832031, "learning_rate": 1.7346938775510206e-07, "loss": 0.0848, "step": 1926 }, { "epoch": 0.9831632653061224, "grad_norm": 51.34260177612305, "learning_rate": 1.6836734693877553e-07, "loss": 0.146, "step": 1927 }, { "epoch": 0.9836734693877551, "grad_norm": 17.97673988342285, "learning_rate": 1.6326530612244901e-07, "loss": 0.0389, "step": 1928 }, { "epoch": 0.9841836734693877, "grad_norm": 45.052276611328125, "learning_rate": 1.5816326530612244e-07, "loss": 0.0993, "step": 1929 }, { "epoch": 0.9846938775510204, "grad_norm": 28.089515686035156, "learning_rate": 1.5306122448979592e-07, "loss": 0.0604, "step": 1930 }, { "epoch": 0.985204081632653, "grad_norm": 184.30123901367188, "learning_rate": 1.479591836734694e-07, "loss": 1.3192, "step": 1931 }, { "epoch": 0.9857142857142858, "grad_norm": 18.166275024414062, "learning_rate": 1.4285714285714287e-07, "loss": 0.0571, "step": 1932 }, { "epoch": 0.9862244897959184, "grad_norm": 160.2290802001953, "learning_rate": 1.3775510204081635e-07, "loss": 0.3059, "step": 1933 }, { "epoch": 0.986734693877551, "grad_norm": 25.441471099853516, "learning_rate": 1.326530612244898e-07, "loss": 0.0533, "step": 1934 }, { "epoch": 0.9872448979591837, "grad_norm": 33.45953369140625, "learning_rate": 1.2755102040816328e-07, "loss": 0.0393, "step": 1935 }, { "epoch": 0.9877551020408163, "grad_norm": 14.509445190429688, "learning_rate": 1.2244897959183673e-07, "loss": 0.0146, "step": 1936 }, { "epoch": 0.988265306122449, "grad_norm": 44.613983154296875, "learning_rate": 1.1734693877551021e-07, "loss": 0.1602, "step": 1937 }, { "epoch": 0.9887755102040816, "grad_norm": 18.62423324584961, "learning_rate": 1.1224489795918368e-07, "loss": 0.0136, "step": 1938 }, { "epoch": 0.9892857142857143, "grad_norm": 12.593050003051758, "learning_rate": 1.0714285714285716e-07, "loss": 0.0116, "step": 1939 }, { "epoch": 0.9897959183673469, "grad_norm": 64.98340606689453, "learning_rate": 1.0204081632653061e-07, "loss": 0.1222, "step": 1940 }, { "epoch": 0.9903061224489796, "grad_norm": 44.38301086425781, "learning_rate": 9.693877551020409e-08, "loss": 0.0629, "step": 1941 }, { "epoch": 0.9908163265306122, "grad_norm": 10.974567413330078, "learning_rate": 9.183673469387756e-08, "loss": 0.0123, "step": 1942 }, { "epoch": 0.9913265306122448, "grad_norm": 63.343048095703125, "learning_rate": 8.673469387755103e-08, "loss": 0.1472, "step": 1943 }, { "epoch": 0.9918367346938776, "grad_norm": 8.707964897155762, "learning_rate": 8.163265306122451e-08, "loss": 0.0094, "step": 1944 }, { "epoch": 0.9923469387755102, "grad_norm": 58.531578063964844, "learning_rate": 7.653061224489796e-08, "loss": 0.1401, "step": 1945 }, { "epoch": 0.9928571428571429, "grad_norm": 18.82192611694336, "learning_rate": 7.142857142857144e-08, "loss": 0.0448, "step": 1946 }, { "epoch": 0.9933673469387755, "grad_norm": 41.70481872558594, "learning_rate": 6.63265306122449e-08, "loss": 0.0552, "step": 1947 }, { "epoch": 0.9938775510204082, "grad_norm": 12.247783660888672, "learning_rate": 6.122448979591837e-08, "loss": 0.0204, "step": 1948 }, { "epoch": 0.9943877551020408, "grad_norm": 72.17545318603516, "learning_rate": 5.612244897959184e-08, "loss": 0.3496, "step": 1949 }, { "epoch": 0.9948979591836735, "grad_norm": 11.649019241333008, "learning_rate": 5.1020408163265303e-08, "loss": 0.0132, "step": 1950 }, { "epoch": 0.9954081632653061, "grad_norm": 32.90751647949219, "learning_rate": 4.591836734693878e-08, "loss": 0.1179, "step": 1951 }, { "epoch": 0.9959183673469387, "grad_norm": 29.89322280883789, "learning_rate": 4.0816326530612253e-08, "loss": 0.0494, "step": 1952 }, { "epoch": 0.9964285714285714, "grad_norm": 152.76602172851562, "learning_rate": 3.571428571428572e-08, "loss": 0.462, "step": 1953 }, { "epoch": 0.996938775510204, "grad_norm": 25.69182777404785, "learning_rate": 3.0612244897959183e-08, "loss": 0.0352, "step": 1954 }, { "epoch": 0.9974489795918368, "grad_norm": 44.416603088378906, "learning_rate": 2.5510204081632652e-08, "loss": 0.1082, "step": 1955 }, { "epoch": 0.9979591836734694, "grad_norm": 37.72198486328125, "learning_rate": 2.0408163265306127e-08, "loss": 0.0705, "step": 1956 }, { "epoch": 0.9984693877551021, "grad_norm": 69.20677185058594, "learning_rate": 1.5306122448979592e-08, "loss": 0.223, "step": 1957 }, { "epoch": 0.9989795918367347, "grad_norm": 28.68073081970215, "learning_rate": 1.0204081632653063e-08, "loss": 0.0269, "step": 1958 }, { "epoch": 0.9994897959183674, "grad_norm": 22.890968322753906, "learning_rate": 5.102040816326532e-09, "loss": 0.0492, "step": 1959 }, { "epoch": 1.0, "grad_norm": 112.00133514404297, "learning_rate": 0.0, "loss": 0.2345, "step": 1960 }, { "epoch": 1.0, "eval_accuracy": 0.768361581920904, "eval_loss": 0.6275630593299866, "eval_runtime": 29.6136, "eval_samples_per_second": 11.954, "eval_steps_per_second": 11.954, "step": 1960 } ], "logging_steps": 1, "max_steps": 1960, "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 }