Upload trainer_state.json
Browse files- trainer_state.json +375 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,375 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2994,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"learning_rate": 0.00019989878688605225,
|
| 14 |
+
"loss": 1.6802,
|
| 15 |
+
"step": 50
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.2,
|
| 19 |
+
"learning_rate": 0.00019959535242609772,
|
| 20 |
+
"loss": 1.5884,
|
| 21 |
+
"step": 100
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.3,
|
| 25 |
+
"learning_rate": 0.0001990903108510678,
|
| 26 |
+
"loss": 1.5617,
|
| 27 |
+
"step": 150
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.4,
|
| 31 |
+
"learning_rate": 0.00019838468449757217,
|
| 32 |
+
"loss": 1.5139,
|
| 33 |
+
"step": 200
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.5,
|
| 37 |
+
"learning_rate": 0.00019747990173842118,
|
| 38 |
+
"loss": 1.5259,
|
| 39 |
+
"step": 250
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.6,
|
| 43 |
+
"learning_rate": 0.00019637779409122493,
|
| 44 |
+
"loss": 1.4783,
|
| 45 |
+
"step": 300
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.7,
|
| 49 |
+
"learning_rate": 0.00019508059251092087,
|
| 50 |
+
"loss": 1.5202,
|
| 51 |
+
"step": 350
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8,
|
| 55 |
+
"learning_rate": 0.0001935909228737363,
|
| 56 |
+
"loss": 1.7727,
|
| 57 |
+
"step": 400
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.9,
|
| 61 |
+
"learning_rate": 0.00019191180066172584,
|
| 62 |
+
"loss": 1.8033,
|
| 63 |
+
"step": 450
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0,
|
| 67 |
+
"learning_rate": 0.00019004662485864503,
|
| 68 |
+
"loss": 1.7674,
|
| 69 |
+
"step": 500
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.1,
|
| 73 |
+
"learning_rate": 0.00018799917106951565,
|
| 74 |
+
"loss": 1.3627,
|
| 75 |
+
"step": 550
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 1.2,
|
| 79 |
+
"learning_rate": 0.00018577358387781095,
|
| 80 |
+
"loss": 1.3747,
|
| 81 |
+
"step": 600
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.3,
|
| 85 |
+
"learning_rate": 0.0001833743684557316,
|
| 86 |
+
"loss": 1.3806,
|
| 87 |
+
"step": 650
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.4,
|
| 91 |
+
"learning_rate": 0.00018080638144455567,
|
| 92 |
+
"loss": 1.3878,
|
| 93 |
+
"step": 700
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.5,
|
| 97 |
+
"learning_rate": 0.00017807482112352263,
|
| 98 |
+
"loss": 1.3824,
|
| 99 |
+
"step": 750
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.6,
|
| 103 |
+
"learning_rate": 0.0001751852168871531,
|
| 104 |
+
"loss": 1.4058,
|
| 105 |
+
"step": 800
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7,
|
| 109 |
+
"learning_rate": 0.00017214341805230388,
|
| 110 |
+
"loss": 1.4329,
|
| 111 |
+
"step": 850
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.8,
|
| 115 |
+
"learning_rate": 0.0001689555820176165,
|
| 116 |
+
"loss": 1.4325,
|
| 117 |
+
"step": 900
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.9,
|
| 121 |
+
"learning_rate": 0.00016562816179932748,
|
| 122 |
+
"loss": 1.4175,
|
| 123 |
+
"step": 950
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.0,
|
| 127 |
+
"learning_rate": 0.0001621678929686709,
|
| 128 |
+
"loss": 1.4092,
|
| 129 |
+
"step": 1000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.1,
|
| 133 |
+
"learning_rate": 0.00015858178001731577,
|
| 134 |
+
"loss": 0.969,
|
| 135 |
+
"step": 1050
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.2,
|
| 139 |
+
"learning_rate": 0.00015487708217843755,
|
| 140 |
+
"loss": 0.9795,
|
| 141 |
+
"step": 1100
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 2.3,
|
| 145 |
+
"learning_rate": 0.0001510612987321264,
|
| 146 |
+
"loss": 0.991,
|
| 147 |
+
"step": 1150
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 2.4,
|
| 151 |
+
"learning_rate": 0.00014714215382487735,
|
| 152 |
+
"loss": 1.0001,
|
| 153 |
+
"step": 1200
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.51,
|
| 157 |
+
"learning_rate": 0.00014312758083389192,
|
| 158 |
+
"loss": 1.0092,
|
| 159 |
+
"step": 1250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.61,
|
| 163 |
+
"learning_rate": 0.00013902570630784184,
|
| 164 |
+
"loss": 1.0274,
|
| 165 |
+
"step": 1300
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.71,
|
| 169 |
+
"learning_rate": 0.00013484483351660324,
|
| 170 |
+
"loss": 1.0369,
|
| 171 |
+
"step": 1350
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.81,
|
| 175 |
+
"learning_rate": 0.00013059342564326052,
|
| 176 |
+
"loss": 1.043,
|
| 177 |
+
"step": 1400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.91,
|
| 181 |
+
"learning_rate": 0.00012628008865240412,
|
| 182 |
+
"loss": 1.0355,
|
| 183 |
+
"step": 1450
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.01,
|
| 187 |
+
"learning_rate": 0.00012191355386940109,
|
| 188 |
+
"loss": 1.0137,
|
| 189 |
+
"step": 1500
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 3.11,
|
| 193 |
+
"learning_rate": 0.0001175912552720942,
|
| 194 |
+
"loss": 0.5538,
|
| 195 |
+
"step": 1550
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.21,
|
| 199 |
+
"learning_rate": 0.00011314555233691659,
|
| 200 |
+
"loss": 0.5644,
|
| 201 |
+
"step": 1600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.31,
|
| 205 |
+
"learning_rate": 0.00010867323935600735,
|
| 206 |
+
"loss": 0.5862,
|
| 207 |
+
"step": 1650
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 3.41,
|
| 211 |
+
"learning_rate": 0.00010418336946383338,
|
| 212 |
+
"loss": 0.595,
|
| 213 |
+
"step": 1700
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 3.51,
|
| 217 |
+
"learning_rate": 9.968503133465485e-05,
|
| 218 |
+
"loss": 0.5974,
|
| 219 |
+
"step": 1750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.61,
|
| 223 |
+
"learning_rate": 9.518733078466463e-05,
|
| 224 |
+
"loss": 0.5951,
|
| 225 |
+
"step": 1800
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.71,
|
| 229 |
+
"learning_rate": 9.069937233942807e-05,
|
| 230 |
+
"loss": 0.5998,
|
| 231 |
+
"step": 1850
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 3.81,
|
| 235 |
+
"learning_rate": 8.623024080393546e-05,
|
| 236 |
+
"loss": 0.5986,
|
| 237 |
+
"step": 1900
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 3.91,
|
| 241 |
+
"learning_rate": 8.178898287257368e-05,
|
| 242 |
+
"loss": 0.5888,
|
| 243 |
+
"step": 1950
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.01,
|
| 247 |
+
"learning_rate": 7.738458881624441e-05,
|
| 248 |
+
"loss": 0.576,
|
| 249 |
+
"step": 2000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 4.11,
|
| 253 |
+
"learning_rate": 7.302597428369742e-05,
|
| 254 |
+
"loss": 0.271,
|
| 255 |
+
"step": 2050
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.21,
|
| 259 |
+
"learning_rate": 6.872196225391955e-05,
|
| 260 |
+
"loss": 0.2735,
|
| 261 |
+
"step": 2100
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.31,
|
| 265 |
+
"learning_rate": 6.448126517611075e-05,
|
| 266 |
+
"loss": 0.2818,
|
| 267 |
+
"step": 2150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.41,
|
| 271 |
+
"learning_rate": 6.031246733340219e-05,
|
| 272 |
+
"loss": 0.2814,
|
| 273 |
+
"step": 2200
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 4.51,
|
| 277 |
+
"learning_rate": 5.6224007466015405e-05,
|
| 278 |
+
"loss": 0.277,
|
| 279 |
+
"step": 2250
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.61,
|
| 283 |
+
"learning_rate": 5.2224161689038975e-05,
|
| 284 |
+
"loss": 0.275,
|
| 285 |
+
"step": 2300
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.71,
|
| 289 |
+
"learning_rate": 4.8321026739400875e-05,
|
| 290 |
+
"loss": 0.2763,
|
| 291 |
+
"step": 2350
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 4.81,
|
| 295 |
+
"learning_rate": 4.4522503585949327e-05,
|
| 296 |
+
"loss": 0.2766,
|
| 297 |
+
"step": 2400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.91,
|
| 301 |
+
"learning_rate": 4.083628143581959e-05,
|
| 302 |
+
"loss": 0.2775,
|
| 303 |
+
"step": 2450
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.01,
|
| 307 |
+
"learning_rate": 3.726982216946203e-05,
|
| 308 |
+
"loss": 0.264,
|
| 309 |
+
"step": 2500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.11,
|
| 313 |
+
"learning_rate": 3.383034523583898e-05,
|
| 314 |
+
"loss": 0.1316,
|
| 315 |
+
"step": 2550
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 5.21,
|
| 319 |
+
"learning_rate": 3.052481303836647e-05,
|
| 320 |
+
"loss": 0.1326,
|
| 321 |
+
"step": 2600
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 5.31,
|
| 325 |
+
"learning_rate": 2.7359916841183754e-05,
|
| 326 |
+
"loss": 0.1349,
|
| 327 |
+
"step": 2650
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.41,
|
| 331 |
+
"learning_rate": 2.4342063224279566e-05,
|
| 332 |
+
"loss": 0.1344,
|
| 333 |
+
"step": 2700
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 5.51,
|
| 337 |
+
"learning_rate": 2.1477361114894035e-05,
|
| 338 |
+
"loss": 0.1326,
|
| 339 |
+
"step": 2750
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.61,
|
| 343 |
+
"learning_rate": 1.8771609421447622e-05,
|
| 344 |
+
"loss": 0.1278,
|
| 345 |
+
"step": 2800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.71,
|
| 349 |
+
"learning_rate": 1.6230285295029556e-05,
|
| 350 |
+
"loss": 0.129,
|
| 351 |
+
"step": 2850
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 5.81,
|
| 355 |
+
"learning_rate": 1.3858533042207633e-05,
|
| 356 |
+
"loss": 0.1282,
|
| 357 |
+
"step": 2900
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 5.91,
|
| 361 |
+
"learning_rate": 1.166115371160218e-05,
|
| 362 |
+
"loss": 0.1285,
|
| 363 |
+
"step": 2950
|
| 364 |
+
}
|
| 365 |
+
],
|
| 366 |
+
"logging_steps": 50,
|
| 367 |
+
"max_steps": 3491,
|
| 368 |
+
"num_input_tokens_seen": 0,
|
| 369 |
+
"num_train_epochs": 7,
|
| 370 |
+
"save_steps": 500,
|
| 371 |
+
"total_flos": 2.316912930642985e+17,
|
| 372 |
+
"train_batch_size": 8,
|
| 373 |
+
"trial_name": null,
|
| 374 |
+
"trial_params": null
|
| 375 |
+
}
|