Upload trainer_state.json
Browse files- trainer_state.json +721 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,721 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 58011,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 4.956904724965955e-05,
|
| 13 |
+
"loss": 1.7071,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"learning_rate": 4.91380944993191e-05,
|
| 19 |
+
"loss": 1.6086,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.08,
|
| 24 |
+
"learning_rate": 4.8707141748978643e-05,
|
| 25 |
+
"loss": 1.5292,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.1,
|
| 30 |
+
"learning_rate": 4.8276188998638195e-05,
|
| 31 |
+
"loss": 1.475,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.13,
|
| 36 |
+
"learning_rate": 4.784523624829774e-05,
|
| 37 |
+
"loss": 1.4358,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.16,
|
| 42 |
+
"learning_rate": 4.7414283497957284e-05,
|
| 43 |
+
"loss": 1.4158,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18,
|
| 48 |
+
"learning_rate": 4.6983330747616836e-05,
|
| 49 |
+
"loss": 1.3844,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.21,
|
| 54 |
+
"learning_rate": 4.655237799727638e-05,
|
| 55 |
+
"loss": 1.3696,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.23,
|
| 60 |
+
"learning_rate": 4.6121425246935925e-05,
|
| 61 |
+
"loss": 1.33,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.26,
|
| 66 |
+
"learning_rate": 4.569047249659548e-05,
|
| 67 |
+
"loss": 1.2824,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.28,
|
| 72 |
+
"learning_rate": 4.525951974625502e-05,
|
| 73 |
+
"loss": 1.2842,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.31,
|
| 78 |
+
"learning_rate": 4.482856699591457e-05,
|
| 79 |
+
"loss": 1.282,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.34,
|
| 84 |
+
"learning_rate": 4.439761424557412e-05,
|
| 85 |
+
"loss": 1.2593,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36,
|
| 90 |
+
"learning_rate": 4.396666149523366e-05,
|
| 91 |
+
"loss": 1.238,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.39,
|
| 96 |
+
"learning_rate": 4.3535708744893214e-05,
|
| 97 |
+
"loss": 1.2322,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.41,
|
| 102 |
+
"learning_rate": 4.310475599455276e-05,
|
| 103 |
+
"loss": 1.2107,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.44,
|
| 108 |
+
"learning_rate": 4.2673803244212304e-05,
|
| 109 |
+
"loss": 1.2007,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.47,
|
| 114 |
+
"learning_rate": 4.2242850493871855e-05,
|
| 115 |
+
"loss": 1.1947,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.49,
|
| 120 |
+
"learning_rate": 4.18118977435314e-05,
|
| 121 |
+
"loss": 1.1792,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.52,
|
| 126 |
+
"learning_rate": 4.138094499319095e-05,
|
| 127 |
+
"loss": 1.1771,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.54,
|
| 132 |
+
"learning_rate": 4.0949992242850496e-05,
|
| 133 |
+
"loss": 1.1634,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.57,
|
| 138 |
+
"learning_rate": 4.051903949251004e-05,
|
| 139 |
+
"loss": 1.1487,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.59,
|
| 144 |
+
"learning_rate": 4.008808674216959e-05,
|
| 145 |
+
"loss": 1.1607,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.62,
|
| 150 |
+
"learning_rate": 3.965713399182914e-05,
|
| 151 |
+
"loss": 1.1333,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.65,
|
| 156 |
+
"learning_rate": 3.922618124148869e-05,
|
| 157 |
+
"loss": 1.1344,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.67,
|
| 162 |
+
"learning_rate": 3.8795228491148233e-05,
|
| 163 |
+
"loss": 1.1319,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7,
|
| 168 |
+
"learning_rate": 3.836427574080778e-05,
|
| 169 |
+
"loss": 1.1236,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.72,
|
| 174 |
+
"learning_rate": 3.793332299046733e-05,
|
| 175 |
+
"loss": 1.1088,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.75,
|
| 180 |
+
"learning_rate": 3.7502370240126874e-05,
|
| 181 |
+
"loss": 1.1135,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.78,
|
| 186 |
+
"learning_rate": 3.707141748978642e-05,
|
| 187 |
+
"loss": 1.0985,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.8,
|
| 192 |
+
"learning_rate": 3.664046473944597e-05,
|
| 193 |
+
"loss": 1.0886,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.83,
|
| 198 |
+
"learning_rate": 3.6209511989105515e-05,
|
| 199 |
+
"loss": 1.0961,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.85,
|
| 204 |
+
"learning_rate": 3.577855923876507e-05,
|
| 205 |
+
"loss": 1.0846,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.88,
|
| 210 |
+
"learning_rate": 3.534760648842461e-05,
|
| 211 |
+
"loss": 1.0638,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.91,
|
| 216 |
+
"learning_rate": 3.4916653738084156e-05,
|
| 217 |
+
"loss": 1.0757,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.93,
|
| 222 |
+
"learning_rate": 3.448570098774371e-05,
|
| 223 |
+
"loss": 1.0676,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.96,
|
| 228 |
+
"learning_rate": 3.405474823740325e-05,
|
| 229 |
+
"loss": 1.0547,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.98,
|
| 234 |
+
"learning_rate": 3.36237954870628e-05,
|
| 235 |
+
"loss": 1.0657,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.01,
|
| 240 |
+
"learning_rate": 3.319284273672235e-05,
|
| 241 |
+
"loss": 1.0545,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.03,
|
| 246 |
+
"learning_rate": 3.2761889986381894e-05,
|
| 247 |
+
"loss": 1.0325,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.06,
|
| 252 |
+
"learning_rate": 3.2330937236041445e-05,
|
| 253 |
+
"loss": 1.0263,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.09,
|
| 258 |
+
"learning_rate": 3.189998448570099e-05,
|
| 259 |
+
"loss": 1.0273,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.11,
|
| 264 |
+
"learning_rate": 3.1469031735360535e-05,
|
| 265 |
+
"loss": 1.0176,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.14,
|
| 270 |
+
"learning_rate": 3.1038078985020086e-05,
|
| 271 |
+
"loss": 1.0119,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.16,
|
| 276 |
+
"learning_rate": 3.060712623467963e-05,
|
| 277 |
+
"loss": 1.0184,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.19,
|
| 282 |
+
"learning_rate": 3.0176173484339176e-05,
|
| 283 |
+
"loss": 1.0226,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.22,
|
| 288 |
+
"learning_rate": 2.9745220733998724e-05,
|
| 289 |
+
"loss": 0.9991,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.24,
|
| 294 |
+
"learning_rate": 2.9314267983658272e-05,
|
| 295 |
+
"loss": 0.9955,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.27,
|
| 300 |
+
"learning_rate": 2.888331523331782e-05,
|
| 301 |
+
"loss": 0.9834,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.29,
|
| 306 |
+
"learning_rate": 2.8452362482977368e-05,
|
| 307 |
+
"loss": 0.9906,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.32,
|
| 312 |
+
"learning_rate": 2.8021409732636916e-05,
|
| 313 |
+
"loss": 0.988,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.34,
|
| 318 |
+
"learning_rate": 2.7590456982296464e-05,
|
| 319 |
+
"loss": 0.9742,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.37,
|
| 324 |
+
"learning_rate": 2.7159504231956013e-05,
|
| 325 |
+
"loss": 0.9712,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.4,
|
| 330 |
+
"learning_rate": 2.6728551481615554e-05,
|
| 331 |
+
"loss": 0.9755,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.42,
|
| 336 |
+
"learning_rate": 2.6297598731275102e-05,
|
| 337 |
+
"loss": 0.9807,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.45,
|
| 342 |
+
"learning_rate": 2.586664598093465e-05,
|
| 343 |
+
"loss": 0.9752,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.47,
|
| 348 |
+
"learning_rate": 2.54356932305942e-05,
|
| 349 |
+
"loss": 0.9714,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.5,
|
| 354 |
+
"learning_rate": 2.5004740480253746e-05,
|
| 355 |
+
"loss": 0.9579,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.53,
|
| 360 |
+
"learning_rate": 2.4573787729913295e-05,
|
| 361 |
+
"loss": 0.9462,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.55,
|
| 366 |
+
"learning_rate": 2.414283497957284e-05,
|
| 367 |
+
"loss": 0.9435,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.58,
|
| 372 |
+
"learning_rate": 2.3711882229232387e-05,
|
| 373 |
+
"loss": 0.9429,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.6,
|
| 378 |
+
"learning_rate": 2.3280929478891936e-05,
|
| 379 |
+
"loss": 0.9521,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.63,
|
| 384 |
+
"learning_rate": 2.2849976728551484e-05,
|
| 385 |
+
"loss": 0.9409,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.65,
|
| 390 |
+
"learning_rate": 2.241902397821103e-05,
|
| 391 |
+
"loss": 0.9434,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.68,
|
| 396 |
+
"learning_rate": 2.1988071227870577e-05,
|
| 397 |
+
"loss": 0.9357,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.71,
|
| 402 |
+
"learning_rate": 2.1557118477530125e-05,
|
| 403 |
+
"loss": 0.9256,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.73,
|
| 408 |
+
"learning_rate": 2.1126165727189673e-05,
|
| 409 |
+
"loss": 0.9379,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.76,
|
| 414 |
+
"learning_rate": 2.0695212976849218e-05,
|
| 415 |
+
"loss": 0.9346,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.78,
|
| 420 |
+
"learning_rate": 2.0264260226508766e-05,
|
| 421 |
+
"loss": 0.9238,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.81,
|
| 426 |
+
"learning_rate": 1.9833307476168314e-05,
|
| 427 |
+
"loss": 0.9266,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.84,
|
| 432 |
+
"learning_rate": 1.9402354725827862e-05,
|
| 433 |
+
"loss": 0.924,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.86,
|
| 438 |
+
"learning_rate": 1.8971401975487407e-05,
|
| 439 |
+
"loss": 0.9082,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.89,
|
| 444 |
+
"learning_rate": 1.8540449225146955e-05,
|
| 445 |
+
"loss": 0.9124,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.91,
|
| 450 |
+
"learning_rate": 1.8109496474806503e-05,
|
| 451 |
+
"loss": 0.9265,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.94,
|
| 456 |
+
"learning_rate": 1.767854372446605e-05,
|
| 457 |
+
"loss": 0.9153,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.97,
|
| 462 |
+
"learning_rate": 1.7247590974125596e-05,
|
| 463 |
+
"loss": 0.9056,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.99,
|
| 468 |
+
"learning_rate": 1.6816638223785144e-05,
|
| 469 |
+
"loss": 0.8929,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.02,
|
| 474 |
+
"learning_rate": 1.6385685473444692e-05,
|
| 475 |
+
"loss": 0.8964,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.04,
|
| 480 |
+
"learning_rate": 1.595473272310424e-05,
|
| 481 |
+
"loss": 0.8852,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.07,
|
| 486 |
+
"learning_rate": 1.5523779972763785e-05,
|
| 487 |
+
"loss": 0.8845,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.09,
|
| 492 |
+
"learning_rate": 1.5092827222423333e-05,
|
| 493 |
+
"loss": 0.9005,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.12,
|
| 498 |
+
"learning_rate": 1.4661874472082881e-05,
|
| 499 |
+
"loss": 0.8803,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.15,
|
| 504 |
+
"learning_rate": 1.423092172174243e-05,
|
| 505 |
+
"loss": 0.8864,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.17,
|
| 510 |
+
"learning_rate": 1.3799968971401977e-05,
|
| 511 |
+
"loss": 0.8821,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.2,
|
| 516 |
+
"learning_rate": 1.3369016221061522e-05,
|
| 517 |
+
"loss": 0.887,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.22,
|
| 522 |
+
"learning_rate": 1.293806347072107e-05,
|
| 523 |
+
"loss": 0.8738,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.25,
|
| 528 |
+
"learning_rate": 1.2507110720380618e-05,
|
| 529 |
+
"loss": 0.8776,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.28,
|
| 534 |
+
"learning_rate": 1.2076157970040165e-05,
|
| 535 |
+
"loss": 0.8589,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.3,
|
| 540 |
+
"learning_rate": 1.1645205219699713e-05,
|
| 541 |
+
"loss": 0.8543,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.33,
|
| 546 |
+
"learning_rate": 1.121425246935926e-05,
|
| 547 |
+
"loss": 0.8584,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.35,
|
| 552 |
+
"learning_rate": 1.0783299719018808e-05,
|
| 553 |
+
"loss": 0.8614,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.38,
|
| 558 |
+
"learning_rate": 1.0352346968678354e-05,
|
| 559 |
+
"loss": 0.8537,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.4,
|
| 564 |
+
"learning_rate": 9.921394218337902e-06,
|
| 565 |
+
"loss": 0.8625,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.43,
|
| 570 |
+
"learning_rate": 9.490441467997449e-06,
|
| 571 |
+
"loss": 0.8532,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.46,
|
| 576 |
+
"learning_rate": 9.059488717656997e-06,
|
| 577 |
+
"loss": 0.8559,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.48,
|
| 582 |
+
"learning_rate": 8.628535967316543e-06,
|
| 583 |
+
"loss": 0.8424,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.51,
|
| 588 |
+
"learning_rate": 8.197583216976091e-06,
|
| 589 |
+
"loss": 0.8475,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.53,
|
| 594 |
+
"learning_rate": 7.766630466635638e-06,
|
| 595 |
+
"loss": 0.8525,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.56,
|
| 600 |
+
"learning_rate": 7.335677716295186e-06,
|
| 601 |
+
"loss": 0.8457,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.59,
|
| 606 |
+
"learning_rate": 6.904724965954733e-06,
|
| 607 |
+
"loss": 0.8503,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.61,
|
| 612 |
+
"learning_rate": 6.4737722156142804e-06,
|
| 613 |
+
"loss": 0.8469,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.64,
|
| 618 |
+
"learning_rate": 6.042819465273828e-06,
|
| 619 |
+
"loss": 0.8416,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.66,
|
| 624 |
+
"learning_rate": 5.611866714933375e-06,
|
| 625 |
+
"loss": 0.8463,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 2.69,
|
| 630 |
+
"learning_rate": 5.180913964592922e-06,
|
| 631 |
+
"loss": 0.843,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.72,
|
| 636 |
+
"learning_rate": 4.7499612142524696e-06,
|
| 637 |
+
"loss": 0.8298,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.74,
|
| 642 |
+
"learning_rate": 4.319008463912017e-06,
|
| 643 |
+
"loss": 0.8335,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.77,
|
| 648 |
+
"learning_rate": 3.888055713571564e-06,
|
| 649 |
+
"loss": 0.8363,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.79,
|
| 654 |
+
"learning_rate": 3.457102963231112e-06,
|
| 655 |
+
"loss": 0.834,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.82,
|
| 660 |
+
"learning_rate": 3.0261502128906587e-06,
|
| 661 |
+
"loss": 0.8289,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.84,
|
| 666 |
+
"learning_rate": 2.595197462550206e-06,
|
| 667 |
+
"loss": 0.8384,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.87,
|
| 672 |
+
"learning_rate": 2.1642447122097533e-06,
|
| 673 |
+
"loss": 0.8255,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9,
|
| 678 |
+
"learning_rate": 1.7332919618693005e-06,
|
| 679 |
+
"loss": 0.825,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.92,
|
| 684 |
+
"learning_rate": 1.302339211528848e-06,
|
| 685 |
+
"loss": 0.8398,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 2.95,
|
| 690 |
+
"learning_rate": 8.713864611883953e-07,
|
| 691 |
+
"loss": 0.8264,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.97,
|
| 696 |
+
"learning_rate": 4.4043371084794265e-07,
|
| 697 |
+
"loss": 0.8196,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 3.0,
|
| 702 |
+
"learning_rate": 9.48096050748996e-09,
|
| 703 |
+
"loss": 0.8307,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 3.0,
|
| 708 |
+
"step": 58011,
|
| 709 |
+
"total_flos": 5.4106824484952474e+17,
|
| 710 |
+
"train_loss": 1.0139114455800622,
|
| 711 |
+
"train_runtime": 59211.1926,
|
| 712 |
+
"train_samples_per_second": 9.797,
|
| 713 |
+
"train_steps_per_second": 0.98
|
| 714 |
+
}
|
| 715 |
+
],
|
| 716 |
+
"max_steps": 58011,
|
| 717 |
+
"num_train_epochs": 3,
|
| 718 |
+
"total_flos": 5.4106824484952474e+17,
|
| 719 |
+
"trial_name": null,
|
| 720 |
+
"trial_params": null
|
| 721 |
+
}
|