{ "best_metric": null, "best_model_checkpoint": null, "epoch": 20.0, "eval_steps": 500, "global_step": 320, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.06, "learning_rate": 2e-05, "loss": 2.3458, "step": 1 }, { "epoch": 0.12, "learning_rate": 4e-05, "loss": 2.1438, "step": 2 }, { "epoch": 0.19, "learning_rate": 6e-05, "loss": 2.1195, "step": 3 }, { "epoch": 0.25, "learning_rate": 8e-05, "loss": 2.5992, "step": 4 }, { "epoch": 0.31, "learning_rate": 0.0001, "loss": 2.2263, "step": 5 }, { "epoch": 0.38, "learning_rate": 9.968253968253969e-05, "loss": 2.3482, "step": 6 }, { "epoch": 0.44, "learning_rate": 9.936507936507938e-05, "loss": 1.9802, "step": 7 }, { "epoch": 0.5, "learning_rate": 9.904761904761905e-05, "loss": 1.9156, "step": 8 }, { "epoch": 0.56, "learning_rate": 9.873015873015874e-05, "loss": 1.7503, "step": 9 }, { "epoch": 0.62, "learning_rate": 9.841269841269841e-05, "loss": 2.059, "step": 10 }, { "epoch": 0.69, "learning_rate": 9.80952380952381e-05, "loss": 1.7051, "step": 11 }, { "epoch": 0.75, "learning_rate": 9.777777777777778e-05, "loss": 2.1344, "step": 12 }, { "epoch": 0.81, "learning_rate": 9.746031746031747e-05, "loss": 1.6857, "step": 13 }, { "epoch": 0.88, "learning_rate": 9.714285714285715e-05, "loss": 1.6324, "step": 14 }, { "epoch": 0.94, "learning_rate": 9.682539682539682e-05, "loss": 1.8047, "step": 15 }, { "epoch": 1.0, "learning_rate": 9.650793650793651e-05, "loss": 1.5296, "step": 16 }, { "epoch": 1.06, "learning_rate": 9.61904761904762e-05, "loss": 1.4797, "step": 17 }, { "epoch": 1.12, "learning_rate": 9.587301587301588e-05, "loss": 1.2823, "step": 18 }, { "epoch": 1.19, "learning_rate": 9.555555555555557e-05, "loss": 1.4775, "step": 19 }, { "epoch": 1.25, "learning_rate": 9.523809523809524e-05, "loss": 1.3844, "step": 20 }, { "epoch": 1.31, "learning_rate": 9.492063492063493e-05, "loss": 1.4312, "step": 21 }, { "epoch": 1.38, "learning_rate": 9.46031746031746e-05, "loss": 1.3182, "step": 22 }, { "epoch": 1.44, "learning_rate": 9.428571428571429e-05, "loss": 1.7689, "step": 23 }, { "epoch": 1.5, "learning_rate": 9.396825396825397e-05, "loss": 1.2934, "step": 24 }, { "epoch": 1.56, "learning_rate": 9.365079365079366e-05, "loss": 1.3469, "step": 25 }, { "epoch": 1.62, "learning_rate": 9.333333333333334e-05, "loss": 1.2163, "step": 26 }, { "epoch": 1.69, "learning_rate": 9.301587301587303e-05, "loss": 1.2185, "step": 27 }, { "epoch": 1.75, "learning_rate": 9.26984126984127e-05, "loss": 1.5183, "step": 28 }, { "epoch": 1.81, "learning_rate": 9.238095238095239e-05, "loss": 1.4769, "step": 29 }, { "epoch": 1.88, "learning_rate": 9.206349206349206e-05, "loss": 1.4717, "step": 30 }, { "epoch": 1.94, "learning_rate": 9.174603174603175e-05, "loss": 1.5168, "step": 31 }, { "epoch": 2.0, "learning_rate": 9.142857142857143e-05, "loss": 1.4267, "step": 32 }, { "epoch": 2.06, "learning_rate": 9.111111111111112e-05, "loss": 1.2049, "step": 33 }, { "epoch": 2.12, "learning_rate": 9.079365079365079e-05, "loss": 1.3333, "step": 34 }, { "epoch": 2.19, "learning_rate": 9.047619047619048e-05, "loss": 1.0416, "step": 35 }, { "epoch": 2.25, "learning_rate": 9.015873015873016e-05, "loss": 1.2593, "step": 36 }, { "epoch": 2.31, "learning_rate": 8.984126984126985e-05, "loss": 1.1917, "step": 37 }, { "epoch": 2.38, "learning_rate": 8.952380952380953e-05, "loss": 1.1144, "step": 38 }, { "epoch": 2.44, "learning_rate": 8.920634920634922e-05, "loss": 1.2757, "step": 39 }, { "epoch": 2.5, "learning_rate": 8.888888888888889e-05, "loss": 1.2067, "step": 40 }, { "epoch": 2.56, "learning_rate": 8.857142857142857e-05, "loss": 1.0615, "step": 41 }, { "epoch": 2.62, "learning_rate": 8.825396825396825e-05, "loss": 1.0798, "step": 42 }, { "epoch": 2.69, "learning_rate": 8.793650793650794e-05, "loss": 0.9969, "step": 43 }, { "epoch": 2.75, "learning_rate": 8.761904761904762e-05, "loss": 0.9464, "step": 44 }, { "epoch": 2.81, "learning_rate": 8.730158730158731e-05, "loss": 0.932, "step": 45 }, { "epoch": 2.88, "learning_rate": 8.6984126984127e-05, "loss": 1.1196, "step": 46 }, { "epoch": 2.94, "learning_rate": 8.666666666666667e-05, "loss": 1.2208, "step": 47 }, { "epoch": 3.0, "learning_rate": 8.634920634920635e-05, "loss": 1.1299, "step": 48 }, { "epoch": 3.06, "learning_rate": 8.603174603174604e-05, "loss": 0.7649, "step": 49 }, { "epoch": 3.12, "learning_rate": 8.571428571428571e-05, "loss": 0.8451, "step": 50 }, { "epoch": 3.19, "learning_rate": 8.53968253968254e-05, "loss": 0.8531, "step": 51 }, { "epoch": 3.25, "learning_rate": 8.507936507936508e-05, "loss": 0.703, "step": 52 }, { "epoch": 3.31, "learning_rate": 8.476190476190477e-05, "loss": 0.6589, "step": 53 }, { "epoch": 3.38, "learning_rate": 8.444444444444444e-05, "loss": 0.8864, "step": 54 }, { "epoch": 3.44, "learning_rate": 8.412698412698413e-05, "loss": 0.5939, "step": 55 }, { "epoch": 3.5, "learning_rate": 8.380952380952382e-05, "loss": 0.7822, "step": 56 }, { "epoch": 3.56, "learning_rate": 8.34920634920635e-05, "loss": 0.7617, "step": 57 }, { "epoch": 3.62, "learning_rate": 8.317460317460319e-05, "loss": 0.8168, "step": 58 }, { "epoch": 3.69, "learning_rate": 8.285714285714287e-05, "loss": 0.7862, "step": 59 }, { "epoch": 3.75, "learning_rate": 8.253968253968255e-05, "loss": 0.8804, "step": 60 }, { "epoch": 3.81, "learning_rate": 8.222222222222222e-05, "loss": 0.7369, "step": 61 }, { "epoch": 3.88, "learning_rate": 8.19047619047619e-05, "loss": 0.7369, "step": 62 }, { "epoch": 3.94, "learning_rate": 8.158730158730159e-05, "loss": 0.7458, "step": 63 }, { "epoch": 4.0, "learning_rate": 8.126984126984128e-05, "loss": 0.6612, "step": 64 }, { "epoch": 4.06, "learning_rate": 8.095238095238096e-05, "loss": 0.4904, "step": 65 }, { "epoch": 4.12, "learning_rate": 8.063492063492063e-05, "loss": 0.5237, "step": 66 }, { "epoch": 4.19, "learning_rate": 8.031746031746032e-05, "loss": 0.4462, "step": 67 }, { "epoch": 4.25, "learning_rate": 8e-05, "loss": 0.4381, "step": 68 }, { "epoch": 4.31, "learning_rate": 7.968253968253968e-05, "loss": 0.3626, "step": 69 }, { "epoch": 4.38, "learning_rate": 7.936507936507937e-05, "loss": 0.3554, "step": 70 }, { "epoch": 4.44, "learning_rate": 7.904761904761905e-05, "loss": 0.3706, "step": 71 }, { "epoch": 4.5, "learning_rate": 7.873015873015874e-05, "loss": 0.4913, "step": 72 }, { "epoch": 4.56, "learning_rate": 7.841269841269841e-05, "loss": 0.4625, "step": 73 }, { "epoch": 4.62, "learning_rate": 7.80952380952381e-05, "loss": 0.3817, "step": 74 }, { "epoch": 4.69, "learning_rate": 7.777777777777778e-05, "loss": 0.3561, "step": 75 }, { "epoch": 4.75, "learning_rate": 7.746031746031747e-05, "loss": 0.3431, "step": 76 }, { "epoch": 4.81, "learning_rate": 7.714285714285715e-05, "loss": 0.4003, "step": 77 }, { "epoch": 4.88, "learning_rate": 7.682539682539684e-05, "loss": 0.3712, "step": 78 }, { "epoch": 4.94, "learning_rate": 7.650793650793651e-05, "loss": 0.4937, "step": 79 }, { "epoch": 5.0, "learning_rate": 7.619047619047618e-05, "loss": 0.427, "step": 80 }, { "epoch": 5.06, "learning_rate": 7.587301587301587e-05, "loss": 0.2432, "step": 81 }, { "epoch": 5.12, "learning_rate": 7.555555555555556e-05, "loss": 0.1917, "step": 82 }, { "epoch": 5.19, "learning_rate": 7.523809523809524e-05, "loss": 0.247, "step": 83 }, { "epoch": 5.25, "learning_rate": 7.492063492063493e-05, "loss": 0.1778, "step": 84 }, { "epoch": 5.31, "learning_rate": 7.460317460317461e-05, "loss": 0.1838, "step": 85 }, { "epoch": 5.38, "learning_rate": 7.428571428571429e-05, "loss": 0.151, "step": 86 }, { "epoch": 5.44, "learning_rate": 7.396825396825397e-05, "loss": 0.1876, "step": 87 }, { "epoch": 5.5, "learning_rate": 7.365079365079366e-05, "loss": 0.18, "step": 88 }, { "epoch": 5.56, "learning_rate": 7.333333333333333e-05, "loss": 0.2786, "step": 89 }, { "epoch": 5.62, "learning_rate": 7.301587301587302e-05, "loss": 0.2065, "step": 90 }, { "epoch": 5.69, "learning_rate": 7.26984126984127e-05, "loss": 0.1887, "step": 91 }, { "epoch": 5.75, "learning_rate": 7.238095238095238e-05, "loss": 0.1627, "step": 92 }, { "epoch": 5.81, "learning_rate": 7.206349206349206e-05, "loss": 0.1663, "step": 93 }, { "epoch": 5.88, "learning_rate": 7.174603174603175e-05, "loss": 0.1822, "step": 94 }, { "epoch": 5.94, "learning_rate": 7.142857142857143e-05, "loss": 0.2143, "step": 95 }, { "epoch": 6.0, "learning_rate": 7.111111111111112e-05, "loss": 0.2168, "step": 96 }, { "epoch": 6.06, "learning_rate": 7.07936507936508e-05, "loss": 0.0953, "step": 97 }, { "epoch": 6.12, "learning_rate": 7.047619047619048e-05, "loss": 0.0841, "step": 98 }, { "epoch": 6.19, "learning_rate": 7.015873015873015e-05, "loss": 0.109, "step": 99 }, { "epoch": 6.25, "learning_rate": 6.984126984126984e-05, "loss": 0.0731, "step": 100 }, { "epoch": 6.31, "learning_rate": 6.952380952380952e-05, "loss": 0.0955, "step": 101 }, { "epoch": 6.38, "learning_rate": 6.920634920634921e-05, "loss": 0.0942, "step": 102 }, { "epoch": 6.44, "learning_rate": 6.88888888888889e-05, "loss": 0.094, "step": 103 }, { "epoch": 6.5, "learning_rate": 6.857142857142858e-05, "loss": 0.0996, "step": 104 }, { "epoch": 6.56, "learning_rate": 6.825396825396825e-05, "loss": 0.1177, "step": 105 }, { "epoch": 6.62, "learning_rate": 6.793650793650794e-05, "loss": 0.0872, "step": 106 }, { "epoch": 6.69, "learning_rate": 6.761904761904763e-05, "loss": 0.1114, "step": 107 }, { "epoch": 6.75, "learning_rate": 6.730158730158731e-05, "loss": 0.1503, "step": 108 }, { "epoch": 6.81, "learning_rate": 6.698412698412698e-05, "loss": 0.1005, "step": 109 }, { "epoch": 6.88, "learning_rate": 6.666666666666667e-05, "loss": 0.1266, "step": 110 }, { "epoch": 6.94, "learning_rate": 6.634920634920636e-05, "loss": 0.0862, "step": 111 }, { "epoch": 7.0, "learning_rate": 6.603174603174603e-05, "loss": 0.1063, "step": 112 }, { "epoch": 7.06, "learning_rate": 6.571428571428571e-05, "loss": 0.0481, "step": 113 }, { "epoch": 7.12, "learning_rate": 6.53968253968254e-05, "loss": 0.0563, "step": 114 }, { "epoch": 7.19, "learning_rate": 6.507936507936509e-05, "loss": 0.0927, "step": 115 }, { "epoch": 7.25, "learning_rate": 6.476190476190477e-05, "loss": 0.0577, "step": 116 }, { "epoch": 7.31, "learning_rate": 6.444444444444446e-05, "loss": 0.0741, "step": 117 }, { "epoch": 7.38, "learning_rate": 6.412698412698413e-05, "loss": 0.0493, "step": 118 }, { "epoch": 7.44, "learning_rate": 6.38095238095238e-05, "loss": 0.0566, "step": 119 }, { "epoch": 7.5, "learning_rate": 6.349206349206349e-05, "loss": 0.0687, "step": 120 }, { "epoch": 7.56, "learning_rate": 6.317460317460318e-05, "loss": 0.0646, "step": 121 }, { "epoch": 7.62, "learning_rate": 6.285714285714286e-05, "loss": 0.0537, "step": 122 }, { "epoch": 7.69, "learning_rate": 6.253968253968255e-05, "loss": 0.0893, "step": 123 }, { "epoch": 7.75, "learning_rate": 6.222222222222222e-05, "loss": 0.07, "step": 124 }, { "epoch": 7.81, "learning_rate": 6.19047619047619e-05, "loss": 0.06, "step": 125 }, { "epoch": 7.88, "learning_rate": 6.158730158730159e-05, "loss": 0.0595, "step": 126 }, { "epoch": 7.94, "learning_rate": 6.126984126984128e-05, "loss": 0.0508, "step": 127 }, { "epoch": 8.0, "learning_rate": 6.0952380952380964e-05, "loss": 0.055, "step": 128 }, { "epoch": 8.06, "learning_rate": 6.0634920634920636e-05, "loss": 0.0508, "step": 129 }, { "epoch": 8.12, "learning_rate": 6.0317460317460316e-05, "loss": 0.0379, "step": 130 }, { "epoch": 8.19, "learning_rate": 6e-05, "loss": 0.04, "step": 131 }, { "epoch": 8.25, "learning_rate": 5.968253968253969e-05, "loss": 0.0493, "step": 132 }, { "epoch": 8.31, "learning_rate": 5.936507936507937e-05, "loss": 0.0478, "step": 133 }, { "epoch": 8.38, "learning_rate": 5.904761904761905e-05, "loss": 0.0604, "step": 134 }, { "epoch": 8.44, "learning_rate": 5.873015873015873e-05, "loss": 0.0383, "step": 135 }, { "epoch": 8.5, "learning_rate": 5.841269841269842e-05, "loss": 0.0361, "step": 136 }, { "epoch": 8.56, "learning_rate": 5.8095238095238104e-05, "loss": 0.0588, "step": 137 }, { "epoch": 8.62, "learning_rate": 5.7777777777777776e-05, "loss": 0.0331, "step": 138 }, { "epoch": 8.69, "learning_rate": 5.7460317460317456e-05, "loss": 0.0446, "step": 139 }, { "epoch": 8.75, "learning_rate": 5.714285714285714e-05, "loss": 0.138, "step": 140 }, { "epoch": 8.81, "learning_rate": 5.682539682539683e-05, "loss": 0.0365, "step": 141 }, { "epoch": 8.88, "learning_rate": 5.650793650793651e-05, "loss": 0.0576, "step": 142 }, { "epoch": 8.94, "learning_rate": 5.619047619047619e-05, "loss": 0.0577, "step": 143 }, { "epoch": 9.0, "learning_rate": 5.587301587301588e-05, "loss": 0.0621, "step": 144 }, { "epoch": 9.06, "learning_rate": 5.555555555555556e-05, "loss": 0.0364, "step": 145 }, { "epoch": 9.12, "learning_rate": 5.5238095238095244e-05, "loss": 0.0427, "step": 146 }, { "epoch": 9.19, "learning_rate": 5.492063492063493e-05, "loss": 0.064, "step": 147 }, { "epoch": 9.25, "learning_rate": 5.46031746031746e-05, "loss": 0.0492, "step": 148 }, { "epoch": 9.31, "learning_rate": 5.428571428571428e-05, "loss": 0.0266, "step": 149 }, { "epoch": 9.38, "learning_rate": 5.396825396825397e-05, "loss": 0.0344, "step": 150 }, { "epoch": 9.44, "learning_rate": 5.3650793650793654e-05, "loss": 0.0341, "step": 151 }, { "epoch": 9.5, "learning_rate": 5.333333333333333e-05, "loss": 0.0338, "step": 152 }, { "epoch": 9.56, "learning_rate": 5.301587301587302e-05, "loss": 0.0398, "step": 153 }, { "epoch": 9.62, "learning_rate": 5.2698412698412705e-05, "loss": 0.053, "step": 154 }, { "epoch": 9.69, "learning_rate": 5.2380952380952384e-05, "loss": 0.0342, "step": 155 }, { "epoch": 9.75, "learning_rate": 5.206349206349207e-05, "loss": 0.0613, "step": 156 }, { "epoch": 9.81, "learning_rate": 5.1746031746031756e-05, "loss": 0.0347, "step": 157 }, { "epoch": 9.88, "learning_rate": 5.142857142857143e-05, "loss": 0.0454, "step": 158 }, { "epoch": 9.94, "learning_rate": 5.111111111111111e-05, "loss": 0.0378, "step": 159 }, { "epoch": 10.0, "learning_rate": 5.0793650793650794e-05, "loss": 0.0463, "step": 160 }, { "epoch": 10.06, "learning_rate": 5.047619047619048e-05, "loss": 0.0356, "step": 161 }, { "epoch": 10.12, "learning_rate": 5.015873015873016e-05, "loss": 0.0301, "step": 162 }, { "epoch": 10.19, "learning_rate": 4.9841269841269845e-05, "loss": 0.0281, "step": 163 }, { "epoch": 10.25, "learning_rate": 4.9523809523809525e-05, "loss": 0.0311, "step": 164 }, { "epoch": 10.31, "learning_rate": 4.9206349206349204e-05, "loss": 0.0324, "step": 165 }, { "epoch": 10.38, "learning_rate": 4.888888888888889e-05, "loss": 0.0397, "step": 166 }, { "epoch": 10.44, "learning_rate": 4.8571428571428576e-05, "loss": 0.0311, "step": 167 }, { "epoch": 10.5, "learning_rate": 4.8253968253968255e-05, "loss": 0.0489, "step": 168 }, { "epoch": 10.56, "learning_rate": 4.793650793650794e-05, "loss": 0.0333, "step": 169 }, { "epoch": 10.62, "learning_rate": 4.761904761904762e-05, "loss": 0.05, "step": 170 }, { "epoch": 10.69, "learning_rate": 4.73015873015873e-05, "loss": 0.0577, "step": 171 }, { "epoch": 10.75, "learning_rate": 4.6984126984126986e-05, "loss": 0.0561, "step": 172 }, { "epoch": 10.81, "learning_rate": 4.666666666666667e-05, "loss": 0.038, "step": 173 }, { "epoch": 10.88, "learning_rate": 4.634920634920635e-05, "loss": 0.0403, "step": 174 }, { "epoch": 10.94, "learning_rate": 4.603174603174603e-05, "loss": 0.0361, "step": 175 }, { "epoch": 11.0, "learning_rate": 4.5714285714285716e-05, "loss": 0.037, "step": 176 }, { "epoch": 11.06, "learning_rate": 4.5396825396825395e-05, "loss": 0.0263, "step": 177 }, { "epoch": 11.12, "learning_rate": 4.507936507936508e-05, "loss": 0.0275, "step": 178 }, { "epoch": 11.19, "learning_rate": 4.476190476190477e-05, "loss": 0.0355, "step": 179 }, { "epoch": 11.25, "learning_rate": 4.4444444444444447e-05, "loss": 0.0385, "step": 180 }, { "epoch": 11.31, "learning_rate": 4.4126984126984126e-05, "loss": 0.0302, "step": 181 }, { "epoch": 11.38, "learning_rate": 4.380952380952381e-05, "loss": 0.0243, "step": 182 }, { "epoch": 11.44, "learning_rate": 4.34920634920635e-05, "loss": 0.0333, "step": 183 }, { "epoch": 11.5, "learning_rate": 4.317460317460318e-05, "loss": 0.0485, "step": 184 }, { "epoch": 11.56, "learning_rate": 4.2857142857142856e-05, "loss": 0.03, "step": 185 }, { "epoch": 11.62, "learning_rate": 4.253968253968254e-05, "loss": 0.0344, "step": 186 }, { "epoch": 11.69, "learning_rate": 4.222222222222222e-05, "loss": 0.051, "step": 187 }, { "epoch": 11.75, "learning_rate": 4.190476190476191e-05, "loss": 0.0309, "step": 188 }, { "epoch": 11.81, "learning_rate": 4.1587301587301594e-05, "loss": 0.0277, "step": 189 }, { "epoch": 11.88, "learning_rate": 4.126984126984127e-05, "loss": 0.0355, "step": 190 }, { "epoch": 11.94, "learning_rate": 4.095238095238095e-05, "loss": 0.0364, "step": 191 }, { "epoch": 12.0, "learning_rate": 4.063492063492064e-05, "loss": 0.0246, "step": 192 }, { "epoch": 12.06, "learning_rate": 4.031746031746032e-05, "loss": 0.0583, "step": 193 }, { "epoch": 12.12, "learning_rate": 4e-05, "loss": 0.0291, "step": 194 }, { "epoch": 12.19, "learning_rate": 3.968253968253968e-05, "loss": 0.0263, "step": 195 }, { "epoch": 12.25, "learning_rate": 3.936507936507937e-05, "loss": 0.0269, "step": 196 }, { "epoch": 12.31, "learning_rate": 3.904761904761905e-05, "loss": 0.0238, "step": 197 }, { "epoch": 12.38, "learning_rate": 3.8730158730158734e-05, "loss": 0.0272, "step": 198 }, { "epoch": 12.44, "learning_rate": 3.841269841269842e-05, "loss": 0.028, "step": 199 }, { "epoch": 12.5, "learning_rate": 3.809523809523809e-05, "loss": 0.0251, "step": 200 }, { "epoch": 12.56, "learning_rate": 3.777777777777778e-05, "loss": 0.0327, "step": 201 }, { "epoch": 12.62, "learning_rate": 3.7460317460317464e-05, "loss": 0.0313, "step": 202 }, { "epoch": 12.69, "learning_rate": 3.7142857142857143e-05, "loss": 0.0346, "step": 203 }, { "epoch": 12.75, "learning_rate": 3.682539682539683e-05, "loss": 0.0266, "step": 204 }, { "epoch": 12.81, "learning_rate": 3.650793650793651e-05, "loss": 0.0352, "step": 205 }, { "epoch": 12.88, "learning_rate": 3.619047619047619e-05, "loss": 0.024, "step": 206 }, { "epoch": 12.94, "learning_rate": 3.5873015873015874e-05, "loss": 0.03, "step": 207 }, { "epoch": 13.0, "learning_rate": 3.555555555555556e-05, "loss": 0.0302, "step": 208 }, { "epoch": 13.06, "learning_rate": 3.523809523809524e-05, "loss": 0.0294, "step": 209 }, { "epoch": 13.12, "learning_rate": 3.492063492063492e-05, "loss": 0.0223, "step": 210 }, { "epoch": 13.19, "learning_rate": 3.4603174603174604e-05, "loss": 0.0202, "step": 211 }, { "epoch": 13.25, "learning_rate": 3.428571428571429e-05, "loss": 0.0259, "step": 212 }, { "epoch": 13.31, "learning_rate": 3.396825396825397e-05, "loss": 0.0204, "step": 213 }, { "epoch": 13.38, "learning_rate": 3.3650793650793656e-05, "loss": 0.0279, "step": 214 }, { "epoch": 13.44, "learning_rate": 3.3333333333333335e-05, "loss": 0.0214, "step": 215 }, { "epoch": 13.5, "learning_rate": 3.3015873015873014e-05, "loss": 0.0222, "step": 216 }, { "epoch": 13.56, "learning_rate": 3.26984126984127e-05, "loss": 0.0294, "step": 217 }, { "epoch": 13.62, "learning_rate": 3.2380952380952386e-05, "loss": 0.0269, "step": 218 }, { "epoch": 13.69, "learning_rate": 3.2063492063492065e-05, "loss": 0.0297, "step": 219 }, { "epoch": 13.75, "learning_rate": 3.1746031746031745e-05, "loss": 0.0319, "step": 220 }, { "epoch": 13.81, "learning_rate": 3.142857142857143e-05, "loss": 0.0309, "step": 221 }, { "epoch": 13.88, "learning_rate": 3.111111111111111e-05, "loss": 0.0323, "step": 222 }, { "epoch": 13.94, "learning_rate": 3.0793650793650796e-05, "loss": 0.0252, "step": 223 }, { "epoch": 14.0, "learning_rate": 3.0476190476190482e-05, "loss": 0.03, "step": 224 }, { "epoch": 14.06, "learning_rate": 3.0158730158730158e-05, "loss": 0.0244, "step": 225 }, { "epoch": 14.12, "learning_rate": 2.9841269841269844e-05, "loss": 0.0207, "step": 226 }, { "epoch": 14.19, "learning_rate": 2.9523809523809526e-05, "loss": 0.0258, "step": 227 }, { "epoch": 14.25, "learning_rate": 2.920634920634921e-05, "loss": 0.027, "step": 228 }, { "epoch": 14.31, "learning_rate": 2.8888888888888888e-05, "loss": 0.0236, "step": 229 }, { "epoch": 14.38, "learning_rate": 2.857142857142857e-05, "loss": 0.0234, "step": 230 }, { "epoch": 14.44, "learning_rate": 2.8253968253968253e-05, "loss": 0.0225, "step": 231 }, { "epoch": 14.5, "learning_rate": 2.793650793650794e-05, "loss": 0.0227, "step": 232 }, { "epoch": 14.56, "learning_rate": 2.7619047619047622e-05, "loss": 0.0265, "step": 233 }, { "epoch": 14.62, "learning_rate": 2.73015873015873e-05, "loss": 0.0219, "step": 234 }, { "epoch": 14.69, "learning_rate": 2.6984126984126984e-05, "loss": 0.0274, "step": 235 }, { "epoch": 14.75, "learning_rate": 2.6666666666666667e-05, "loss": 0.027, "step": 236 }, { "epoch": 14.81, "learning_rate": 2.6349206349206353e-05, "loss": 0.0257, "step": 237 }, { "epoch": 14.88, "learning_rate": 2.6031746031746035e-05, "loss": 0.0309, "step": 238 }, { "epoch": 14.94, "learning_rate": 2.5714285714285714e-05, "loss": 0.0268, "step": 239 }, { "epoch": 15.0, "learning_rate": 2.5396825396825397e-05, "loss": 0.023, "step": 240 }, { "epoch": 15.06, "learning_rate": 2.507936507936508e-05, "loss": 0.0239, "step": 241 }, { "epoch": 15.12, "learning_rate": 2.4761904761904762e-05, "loss": 0.0255, "step": 242 }, { "epoch": 15.19, "learning_rate": 2.4444444444444445e-05, "loss": 0.0246, "step": 243 }, { "epoch": 15.25, "learning_rate": 2.4126984126984128e-05, "loss": 0.0211, "step": 244 }, { "epoch": 15.31, "learning_rate": 2.380952380952381e-05, "loss": 0.0253, "step": 245 }, { "epoch": 15.38, "learning_rate": 2.3492063492063493e-05, "loss": 0.0217, "step": 246 }, { "epoch": 15.44, "learning_rate": 2.3174603174603175e-05, "loss": 0.0277, "step": 247 }, { "epoch": 15.5, "learning_rate": 2.2857142857142858e-05, "loss": 0.0235, "step": 248 }, { "epoch": 15.56, "learning_rate": 2.253968253968254e-05, "loss": 0.0235, "step": 249 }, { "epoch": 15.62, "learning_rate": 2.2222222222222223e-05, "loss": 0.0222, "step": 250 }, { "epoch": 15.69, "learning_rate": 2.1904761904761906e-05, "loss": 0.0226, "step": 251 }, { "epoch": 15.75, "learning_rate": 2.158730158730159e-05, "loss": 0.0236, "step": 252 }, { "epoch": 15.81, "learning_rate": 2.126984126984127e-05, "loss": 0.0278, "step": 253 }, { "epoch": 15.88, "learning_rate": 2.0952380952380954e-05, "loss": 0.0281, "step": 254 }, { "epoch": 15.94, "learning_rate": 2.0634920634920636e-05, "loss": 0.0252, "step": 255 }, { "epoch": 16.0, "learning_rate": 2.031746031746032e-05, "loss": 0.0284, "step": 256 }, { "epoch": 16.06, "learning_rate": 2e-05, "loss": 0.02, "step": 257 }, { "epoch": 16.12, "learning_rate": 1.9682539682539684e-05, "loss": 0.0222, "step": 258 }, { "epoch": 16.19, "learning_rate": 1.9365079365079367e-05, "loss": 0.0232, "step": 259 }, { "epoch": 16.25, "learning_rate": 1.9047619047619046e-05, "loss": 0.0223, "step": 260 }, { "epoch": 16.31, "learning_rate": 1.8730158730158732e-05, "loss": 0.0238, "step": 261 }, { "epoch": 16.38, "learning_rate": 1.8412698412698415e-05, "loss": 0.0249, "step": 262 }, { "epoch": 16.44, "learning_rate": 1.8095238095238094e-05, "loss": 0.0214, "step": 263 }, { "epoch": 16.5, "learning_rate": 1.777777777777778e-05, "loss": 0.0289, "step": 264 }, { "epoch": 16.56, "learning_rate": 1.746031746031746e-05, "loss": 0.025, "step": 265 }, { "epoch": 16.62, "learning_rate": 1.7142857142857145e-05, "loss": 0.0233, "step": 266 }, { "epoch": 16.69, "learning_rate": 1.6825396825396828e-05, "loss": 0.0255, "step": 267 }, { "epoch": 16.75, "learning_rate": 1.6507936507936507e-05, "loss": 0.0251, "step": 268 }, { "epoch": 16.81, "learning_rate": 1.6190476190476193e-05, "loss": 0.0229, "step": 269 }, { "epoch": 16.88, "learning_rate": 1.5873015873015872e-05, "loss": 0.0256, "step": 270 }, { "epoch": 16.94, "learning_rate": 1.5555555555555555e-05, "loss": 0.0288, "step": 271 }, { "epoch": 17.0, "learning_rate": 1.5238095238095241e-05, "loss": 0.0235, "step": 272 }, { "epoch": 17.06, "learning_rate": 1.4920634920634922e-05, "loss": 0.0213, "step": 273 }, { "epoch": 17.12, "learning_rate": 1.4603174603174605e-05, "loss": 0.0207, "step": 274 }, { "epoch": 17.19, "learning_rate": 1.4285714285714285e-05, "loss": 0.0246, "step": 275 }, { "epoch": 17.25, "learning_rate": 1.396825396825397e-05, "loss": 0.0227, "step": 276 }, { "epoch": 17.31, "learning_rate": 1.365079365079365e-05, "loss": 0.0234, "step": 277 }, { "epoch": 17.38, "learning_rate": 1.3333333333333333e-05, "loss": 0.0218, "step": 278 }, { "epoch": 17.44, "learning_rate": 1.3015873015873018e-05, "loss": 0.029, "step": 279 }, { "epoch": 17.5, "learning_rate": 1.2698412698412699e-05, "loss": 0.0236, "step": 280 }, { "epoch": 17.56, "learning_rate": 1.2380952380952381e-05, "loss": 0.0229, "step": 281 }, { "epoch": 17.62, "learning_rate": 1.2063492063492064e-05, "loss": 0.0229, "step": 282 }, { "epoch": 17.69, "learning_rate": 1.1746031746031746e-05, "loss": 0.0232, "step": 283 }, { "epoch": 17.75, "learning_rate": 1.1428571428571429e-05, "loss": 0.0288, "step": 284 }, { "epoch": 17.81, "learning_rate": 1.1111111111111112e-05, "loss": 0.0211, "step": 285 }, { "epoch": 17.88, "learning_rate": 1.0793650793650794e-05, "loss": 0.0198, "step": 286 }, { "epoch": 17.94, "learning_rate": 1.0476190476190477e-05, "loss": 0.0229, "step": 287 }, { "epoch": 18.0, "learning_rate": 1.015873015873016e-05, "loss": 0.0292, "step": 288 }, { "epoch": 18.06, "learning_rate": 9.841269841269842e-06, "loss": 0.0214, "step": 289 }, { "epoch": 18.12, "learning_rate": 9.523809523809523e-06, "loss": 0.0213, "step": 290 }, { "epoch": 18.19, "learning_rate": 9.206349206349207e-06, "loss": 0.022, "step": 291 }, { "epoch": 18.25, "learning_rate": 8.88888888888889e-06, "loss": 0.0202, "step": 292 }, { "epoch": 18.31, "learning_rate": 8.571428571428573e-06, "loss": 0.0217, "step": 293 }, { "epoch": 18.38, "learning_rate": 8.253968253968254e-06, "loss": 0.023, "step": 294 }, { "epoch": 18.44, "learning_rate": 7.936507936507936e-06, "loss": 0.0245, "step": 295 }, { "epoch": 18.5, "learning_rate": 7.6190476190476205e-06, "loss": 0.0284, "step": 296 }, { "epoch": 18.56, "learning_rate": 7.301587301587302e-06, "loss": 0.0266, "step": 297 }, { "epoch": 18.62, "learning_rate": 6.984126984126985e-06, "loss": 0.0222, "step": 298 }, { "epoch": 18.69, "learning_rate": 6.666666666666667e-06, "loss": 0.0222, "step": 299 }, { "epoch": 18.75, "learning_rate": 6.349206349206349e-06, "loss": 0.0207, "step": 300 }, { "epoch": 18.81, "learning_rate": 6.031746031746032e-06, "loss": 0.0262, "step": 301 }, { "epoch": 18.88, "learning_rate": 5.7142857142857145e-06, "loss": 0.0233, "step": 302 }, { "epoch": 18.94, "learning_rate": 5.396825396825397e-06, "loss": 0.0256, "step": 303 }, { "epoch": 19.0, "learning_rate": 5.07936507936508e-06, "loss": 0.0233, "step": 304 }, { "epoch": 19.06, "learning_rate": 4.7619047619047615e-06, "loss": 0.0225, "step": 305 }, { "epoch": 19.12, "learning_rate": 4.444444444444445e-06, "loss": 0.0236, "step": 306 }, { "epoch": 19.19, "learning_rate": 4.126984126984127e-06, "loss": 0.0229, "step": 307 }, { "epoch": 19.25, "learning_rate": 3.8095238095238102e-06, "loss": 0.029, "step": 308 }, { "epoch": 19.31, "learning_rate": 3.4920634920634924e-06, "loss": 0.0171, "step": 309 }, { "epoch": 19.38, "learning_rate": 3.1746031746031746e-06, "loss": 0.0252, "step": 310 }, { "epoch": 19.44, "learning_rate": 2.8571428571428573e-06, "loss": 0.0248, "step": 311 }, { "epoch": 19.5, "learning_rate": 2.53968253968254e-06, "loss": 0.0234, "step": 312 }, { "epoch": 19.56, "learning_rate": 2.2222222222222225e-06, "loss": 0.0206, "step": 313 }, { "epoch": 19.62, "learning_rate": 1.9047619047619051e-06, "loss": 0.0222, "step": 314 }, { "epoch": 19.69, "learning_rate": 1.5873015873015873e-06, "loss": 0.0231, "step": 315 }, { "epoch": 19.75, "learning_rate": 1.26984126984127e-06, "loss": 0.0221, "step": 316 }, { "epoch": 19.81, "learning_rate": 9.523809523809526e-07, "loss": 0.0209, "step": 317 }, { "epoch": 19.88, "learning_rate": 6.34920634920635e-07, "loss": 0.0247, "step": 318 }, { "epoch": 19.94, "learning_rate": 3.174603174603175e-07, "loss": 0.0219, "step": 319 }, { "epoch": 20.0, "learning_rate": 0.0, "loss": 0.0246, "step": 320 }, { "epoch": 20.0, "step": 320, "total_flos": 2.419837403922432e+16, "train_loss": 0.3230070192774292, "train_runtime": 746.9607, "train_samples_per_second": 1.714, "train_steps_per_second": 0.428 } ], "logging_steps": 1, "max_steps": 320, "num_train_epochs": 20, "save_steps": 500, "total_flos": 2.419837403922432e+16, "trial_name": null, "trial_params": null }