Commit ·
9f380a2
1
Parent(s): ec54929
Trained on larger batch size
Browse files- trainer_state.json +388 -1279
- training_args.bin +3 -0
trainer_state.json
CHANGED
|
@@ -1,1516 +1,625 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "OutModelPolicy/checkpoint-50000",
|
| 4 |
-
"epoch": 0.
|
| 5 |
"global_step": 50000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
-
{
|
| 11 |
-
"epoch": 0.0,
|
| 12 |
-
"learning_rate": 1.9952746826949432e-05,
|
| 13 |
-
"loss": 0.4762,
|
| 14 |
-
"step": 500
|
| 15 |
-
},
|
| 16 |
-
{
|
| 17 |
-
"epoch": 0.0,
|
| 18 |
-
"eval_accuracy": 0.8208744710860366,
|
| 19 |
-
"eval_loss": 0.4056582748889923,
|
| 20 |
-
"eval_runtime": 129.4288,
|
| 21 |
-
"eval_samples_per_second": 82.169,
|
| 22 |
-
"eval_steps_per_second": 8.221,
|
| 23 |
-
"step": 500
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.0,
|
| 27 |
-
"learning_rate": 1.9905493653898863e-05,
|
| 28 |
-
"loss": 0.4222,
|
| 29 |
-
"step": 1000
|
| 30 |
-
},
|
| 31 |
-
{
|
| 32 |
-
"epoch": 0.0,
|
| 33 |
-
"eval_accuracy": 0.8309355900329102,
|
| 34 |
-
"eval_loss": 0.39738503098487854,
|
| 35 |
-
"eval_runtime": 129.5771,
|
| 36 |
-
"eval_samples_per_second": 82.075,
|
| 37 |
-
"eval_steps_per_second": 8.211,
|
| 38 |
-
"step": 1000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.01,
|
| 42 |
-
"learning_rate": 1.985824048084829e-05,
|
| 43 |
-
"loss": 0.4226,
|
| 44 |
-
"step": 1500
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.01,
|
| 48 |
-
"eval_accuracy": 0.8228490832157969,
|
| 49 |
-
"eval_loss": 0.4244444668292999,
|
| 50 |
-
"eval_runtime": 129.6288,
|
| 51 |
-
"eval_samples_per_second": 82.042,
|
| 52 |
-
"eval_steps_per_second": 8.208,
|
| 53 |
-
"step": 1500
|
| 54 |
-
},
|
| 55 |
-
{
|
| 56 |
-
"epoch": 0.01,
|
| 57 |
-
"learning_rate": 1.981098730779772e-05,
|
| 58 |
-
"loss": 0.4262,
|
| 59 |
-
"step": 2000
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.01,
|
| 63 |
-
"eval_accuracy": 0.7945463093559003,
|
| 64 |
-
"eval_loss": 0.4486662447452545,
|
| 65 |
-
"eval_runtime": 129.6271,
|
| 66 |
-
"eval_samples_per_second": 82.043,
|
| 67 |
-
"eval_steps_per_second": 8.208,
|
| 68 |
-
"step": 2000
|
| 69 |
-
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 0.01,
|
| 72 |
-
"learning_rate": 1.976373413474715e-05,
|
| 73 |
-
"loss": 0.3976,
|
| 74 |
-
"step": 2500
|
| 75 |
-
},
|
| 76 |
{
|
| 77 |
"epoch": 0.01,
|
| 78 |
-
"
|
| 79 |
-
"
|
| 80 |
-
"
|
| 81 |
-
"eval_samples_per_second": 82.055,
|
| 82 |
-
"eval_steps_per_second": 8.209,
|
| 83 |
-
"step": 2500
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 0.01,
|
| 87 |
-
"learning_rate": 1.971648096169658e-05,
|
| 88 |
-
"loss": 0.4094,
|
| 89 |
-
"step": 3000
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 0.01,
|
| 93 |
-
"
|
| 94 |
-
"
|
| 95 |
-
"
|
| 96 |
-
"eval_samples_per_second": 82.077,
|
| 97 |
-
"eval_steps_per_second": 8.212,
|
| 98 |
-
"step": 3000
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 0.02,
|
| 102 |
-
"learning_rate": 1.966922778864601e-05,
|
| 103 |
-
"loss": 0.4122,
|
| 104 |
-
"step": 3500
|
| 105 |
-
},
|
| 106 |
-
{
|
| 107 |
-
"epoch": 0.02,
|
| 108 |
-
"eval_accuracy": 0.8160789844851905,
|
| 109 |
-
"eval_loss": 0.4006648659706116,
|
| 110 |
-
"eval_runtime": 129.6541,
|
| 111 |
-
"eval_samples_per_second": 82.026,
|
| 112 |
-
"eval_steps_per_second": 8.206,
|
| 113 |
-
"step": 3500
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"epoch": 0.02,
|
| 117 |
-
"learning_rate": 1.962197461559544e-05,
|
| 118 |
-
"loss": 0.4086,
|
| 119 |
-
"step": 4000
|
| 120 |
-
},
|
| 121 |
-
{
|
| 122 |
-
"epoch": 0.02,
|
| 123 |
-
"eval_accuracy": 0.7988716502115656,
|
| 124 |
-
"eval_loss": 0.44481590390205383,
|
| 125 |
-
"eval_runtime": 129.6035,
|
| 126 |
-
"eval_samples_per_second": 82.058,
|
| 127 |
-
"eval_steps_per_second": 8.21,
|
| 128 |
-
"step": 4000
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.02,
|
| 132 |
-
"learning_rate": 1.9574721442544867e-05,
|
| 133 |
-
"loss": 0.3977,
|
| 134 |
-
"step": 4500
|
| 135 |
-
},
|
| 136 |
-
{
|
| 137 |
-
"epoch": 0.02,
|
| 138 |
-
"eval_accuracy": 0.8329102021626704,
|
| 139 |
-
"eval_loss": 0.38500234484672546,
|
| 140 |
-
"eval_runtime": 130.5943,
|
| 141 |
-
"eval_samples_per_second": 81.435,
|
| 142 |
-
"eval_steps_per_second": 8.147,
|
| 143 |
-
"step": 4500
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.02,
|
| 147 |
-
"learning_rate": 1.9527468269494298e-05,
|
| 148 |
-
"loss": 0.4121,
|
| 149 |
-
"step": 5000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.02,
|
| 153 |
-
"
|
| 154 |
-
"
|
| 155 |
-
"
|
| 156 |
-
"eval_samples_per_second": 81.835,
|
| 157 |
-
"eval_steps_per_second": 8.187,
|
| 158 |
-
"step": 5000
|
| 159 |
-
},
|
| 160 |
-
{
|
| 161 |
-
"epoch": 0.03,
|
| 162 |
-
"learning_rate": 1.948021509644373e-05,
|
| 163 |
-
"loss": 0.3913,
|
| 164 |
-
"step": 5500
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.03,
|
| 168 |
-
"eval_accuracy": 0.8207804419370005,
|
| 169 |
-
"eval_loss": 0.4078373908996582,
|
| 170 |
-
"eval_runtime": 129.7026,
|
| 171 |
-
"eval_samples_per_second": 81.995,
|
| 172 |
-
"eval_steps_per_second": 8.203,
|
| 173 |
-
"step": 5500
|
| 174 |
-
},
|
| 175 |
-
{
|
| 176 |
-
"epoch": 0.03,
|
| 177 |
-
"learning_rate": 1.9432961923393156e-05,
|
| 178 |
-
"loss": 0.4043,
|
| 179 |
-
"step": 6000
|
| 180 |
-
},
|
| 181 |
-
{
|
| 182 |
-
"epoch": 0.03,
|
| 183 |
-
"eval_accuracy": 0.8400564174894217,
|
| 184 |
-
"eval_loss": 0.40346187353134155,
|
| 185 |
-
"eval_runtime": 129.7498,
|
| 186 |
-
"eval_samples_per_second": 81.965,
|
| 187 |
-
"eval_steps_per_second": 8.2,
|
| 188 |
-
"step": 6000
|
| 189 |
-
},
|
| 190 |
-
{
|
| 191 |
-
"epoch": 0.03,
|
| 192 |
-
"learning_rate": 1.9385708750342586e-05,
|
| 193 |
-
"loss": 0.3943,
|
| 194 |
-
"step": 6500
|
| 195 |
-
},
|
| 196 |
-
{
|
| 197 |
-
"epoch": 0.03,
|
| 198 |
-
"eval_accuracy": 0.8372355430183357,
|
| 199 |
-
"eval_loss": 0.4044436514377594,
|
| 200 |
-
"eval_runtime": 129.5695,
|
| 201 |
-
"eval_samples_per_second": 82.079,
|
| 202 |
-
"eval_steps_per_second": 8.212,
|
| 203 |
-
"step": 6500
|
| 204 |
},
|
| 205 |
{
|
| 206 |
"epoch": 0.03,
|
| 207 |
-
"learning_rate": 1.
|
| 208 |
-
"loss": 0.
|
| 209 |
-
"step":
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 0.03,
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"
|
| 216 |
-
"eval_samples_per_second": 81.991,
|
| 217 |
-
"eval_steps_per_second": 8.203,
|
| 218 |
-
"step": 7000
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"epoch": 0.04,
|
| 222 |
-
"learning_rate": 1.9291202404241447e-05,
|
| 223 |
-
"loss": 0.3962,
|
| 224 |
-
"step": 7500
|
| 225 |
-
},
|
| 226 |
-
{
|
| 227 |
-
"epoch": 0.04,
|
| 228 |
-
"eval_accuracy": 0.8284908321579689,
|
| 229 |
-
"eval_loss": 0.41320571303367615,
|
| 230 |
-
"eval_runtime": 129.8631,
|
| 231 |
-
"eval_samples_per_second": 81.894,
|
| 232 |
-
"eval_steps_per_second": 8.193,
|
| 233 |
-
"step": 7500
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.04,
|
| 237 |
-
"learning_rate": 1.9243949231190878e-05,
|
| 238 |
-
"loss": 0.3776,
|
| 239 |
-
"step": 8000
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"epoch": 0.04,
|
| 243 |
-
"eval_accuracy": 0.8282087447108604,
|
| 244 |
-
"eval_loss": 0.4154951274394989,
|
| 245 |
-
"eval_runtime": 129.934,
|
| 246 |
-
"eval_samples_per_second": 81.849,
|
| 247 |
-
"eval_steps_per_second": 8.189,
|
| 248 |
-
"step": 8000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.04,
|
| 252 |
-
"learning_rate": 1.9196696058140305e-05,
|
| 253 |
-
"loss": 0.4021,
|
| 254 |
-
"step": 8500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.04,
|
| 258 |
-
"eval_accuracy": 0.8206864127879643,
|
| 259 |
-
"eval_loss": 0.4651769697666168,
|
| 260 |
-
"eval_runtime": 129.6522,
|
| 261 |
-
"eval_samples_per_second": 82.027,
|
| 262 |
-
"eval_steps_per_second": 8.207,
|
| 263 |
-
"step": 8500
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 0.04,
|
| 267 |
-
"learning_rate": 1.9149442885089736e-05,
|
| 268 |
-
"loss": 0.4031,
|
| 269 |
-
"step": 9000
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.04,
|
| 273 |
-
"eval_accuracy": 0.8377056887635167,
|
| 274 |
-
"eval_loss": 0.3797547519207001,
|
| 275 |
-
"eval_runtime": 129.7227,
|
| 276 |
-
"eval_samples_per_second": 81.983,
|
| 277 |
-
"eval_steps_per_second": 8.202,
|
| 278 |
-
"step": 9000
|
| 279 |
-
},
|
| 280 |
-
{
|
| 281 |
-
"epoch": 0.04,
|
| 282 |
-
"learning_rate": 1.9102189712039167e-05,
|
| 283 |
-
"loss": 0.3971,
|
| 284 |
-
"step": 9500
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.04,
|
| 288 |
-
"
|
| 289 |
-
"
|
| 290 |
-
"
|
| 291 |
-
"eval_samples_per_second": 81.944,
|
| 292 |
-
"eval_steps_per_second": 8.198,
|
| 293 |
-
"step": 9500
|
| 294 |
-
},
|
| 295 |
-
{
|
| 296 |
-
"epoch": 0.05,
|
| 297 |
-
"learning_rate": 1.9054936538988594e-05,
|
| 298 |
-
"loss": 0.4018,
|
| 299 |
-
"step": 10000
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 0.05,
|
| 303 |
-
"eval_accuracy": 0.8419370004701457,
|
| 304 |
-
"eval_loss": 0.3951157331466675,
|
| 305 |
-
"eval_runtime": 129.8232,
|
| 306 |
-
"eval_samples_per_second": 81.919,
|
| 307 |
-
"eval_steps_per_second": 8.196,
|
| 308 |
-
"step": 10000
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"epoch": 0.05,
|
| 312 |
-
"learning_rate": 1.9007683365938024e-05,
|
| 313 |
-
"loss": 0.3894,
|
| 314 |
-
"step": 10500
|
| 315 |
-
},
|
| 316 |
-
{
|
| 317 |
-
"epoch": 0.05,
|
| 318 |
-
"eval_accuracy": 0.8360131640808651,
|
| 319 |
-
"eval_loss": 0.37815341353416443,
|
| 320 |
-
"eval_runtime": 130.1142,
|
| 321 |
-
"eval_samples_per_second": 81.736,
|
| 322 |
-
"eval_steps_per_second": 8.177,
|
| 323 |
-
"step": 10500
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 0.05,
|
| 327 |
-
"learning_rate": 1.8960430192887455e-05,
|
| 328 |
-
"loss": 0.3947,
|
| 329 |
-
"step": 11000
|
| 330 |
-
},
|
| 331 |
-
{
|
| 332 |
-
"epoch": 0.05,
|
| 333 |
-
"eval_accuracy": 0.8376116596144805,
|
| 334 |
-
"eval_loss": 0.3747114837169647,
|
| 335 |
-
"eval_runtime": 129.8229,
|
| 336 |
-
"eval_samples_per_second": 81.919,
|
| 337 |
-
"eval_steps_per_second": 8.196,
|
| 338 |
-
"step": 11000
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.05,
|
| 342 |
-
"learning_rate": 1.8913177019836882e-05,
|
| 343 |
-
"loss": 0.3968,
|
| 344 |
-
"step": 11500
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.05,
|
| 348 |
-
"
|
| 349 |
-
"
|
| 350 |
-
"
|
| 351 |
-
"eval_samples_per_second": 82.0,
|
| 352 |
-
"eval_steps_per_second": 8.204,
|
| 353 |
-
"step": 11500
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.06,
|
| 357 |
-
"learning_rate": 1.8865923846786313e-05,
|
| 358 |
-
"loss": 0.3823,
|
| 359 |
-
"step": 12000
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.06,
|
| 363 |
-
"eval_accuracy": 0.8388340385519512,
|
| 364 |
-
"eval_loss": 0.39501070976257324,
|
| 365 |
-
"eval_runtime": 129.9377,
|
| 366 |
-
"eval_samples_per_second": 81.847,
|
| 367 |
-
"eval_steps_per_second": 8.189,
|
| 368 |
-
"step": 12000
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 0.06,
|
| 372 |
-
"learning_rate": 1.8818670673735744e-05,
|
| 373 |
-
"loss": 0.3865,
|
| 374 |
-
"step": 12500
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.06,
|
| 378 |
-
"eval_accuracy": 0.8396803008932769,
|
| 379 |
-
"eval_loss": 0.37278103828430176,
|
| 380 |
-
"eval_runtime": 129.7336,
|
| 381 |
-
"eval_samples_per_second": 81.976,
|
| 382 |
-
"eval_steps_per_second": 8.201,
|
| 383 |
-
"step": 12500
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.06,
|
| 387 |
-
"learning_rate": 1.877141750068517e-05,
|
| 388 |
-
"loss": 0.3895,
|
| 389 |
-
"step": 13000
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 0.06,
|
| 393 |
-
"eval_accuracy": 0.8393041842971322,
|
| 394 |
-
"eval_loss": 0.3665693402290344,
|
| 395 |
-
"eval_runtime": 129.2422,
|
| 396 |
-
"eval_samples_per_second": 82.287,
|
| 397 |
-
"eval_steps_per_second": 8.233,
|
| 398 |
-
"step": 13000
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.06,
|
| 402 |
-
"learning_rate": 1.
|
| 403 |
-
"loss": 0.
|
| 404 |
-
"step":
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 0.06,
|
| 408 |
-
"
|
| 409 |
-
"
|
| 410 |
-
"
|
| 411 |
-
"eval_samples_per_second": 82.355,
|
| 412 |
-
"eval_steps_per_second": 8.239,
|
| 413 |
-
"step": 13500
|
| 414 |
-
},
|
| 415 |
-
{
|
| 416 |
-
"epoch": 0.07,
|
| 417 |
-
"learning_rate": 1.8676911154584032e-05,
|
| 418 |
-
"loss": 0.3924,
|
| 419 |
-
"step": 14000
|
| 420 |
-
},
|
| 421 |
-
{
|
| 422 |
-
"epoch": 0.07,
|
| 423 |
-
"eval_accuracy": 0.8404325340855665,
|
| 424 |
-
"eval_loss": 0.37671658396720886,
|
| 425 |
-
"eval_runtime": 129.0223,
|
| 426 |
-
"eval_samples_per_second": 82.428,
|
| 427 |
-
"eval_steps_per_second": 8.247,
|
| 428 |
-
"step": 14000
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 0.07,
|
| 432 |
-
"learning_rate": 1.8629657981533463e-05,
|
| 433 |
-
"loss": 0.4017,
|
| 434 |
-
"step": 14500
|
| 435 |
-
},
|
| 436 |
-
{
|
| 437 |
-
"epoch": 0.07,
|
| 438 |
-
"eval_accuracy": 0.8437235543018335,
|
| 439 |
-
"eval_loss": 0.36841943860054016,
|
| 440 |
-
"eval_runtime": 130.2328,
|
| 441 |
-
"eval_samples_per_second": 81.661,
|
| 442 |
-
"eval_steps_per_second": 8.17,
|
| 443 |
-
"step": 14500
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.07,
|
| 447 |
-
"learning_rate": 1.858240480848289e-05,
|
| 448 |
-
"loss": 0.4186,
|
| 449 |
-
"step": 15000
|
| 450 |
-
},
|
| 451 |
-
{
|
| 452 |
-
"epoch": 0.07,
|
| 453 |
-
"eval_accuracy": 0.8300893276915844,
|
| 454 |
-
"eval_loss": 0.39305663108825684,
|
| 455 |
-
"eval_runtime": 130.4473,
|
| 456 |
-
"eval_samples_per_second": 81.527,
|
| 457 |
-
"eval_steps_per_second": 8.157,
|
| 458 |
-
"step": 15000
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.07,
|
| 462 |
-
"learning_rate": 1.853515163543232e-05,
|
| 463 |
-
"loss": 0.3766,
|
| 464 |
-
"step": 15500
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.07,
|
| 468 |
-
"
|
| 469 |
-
"
|
| 470 |
-
"
|
| 471 |
-
"eval_samples_per_second": 81.746,
|
| 472 |
-
"eval_steps_per_second": 8.178,
|
| 473 |
-
"step": 15500
|
| 474 |
-
},
|
| 475 |
-
{
|
| 476 |
-
"epoch": 0.08,
|
| 477 |
-
"learning_rate": 1.848789846238175e-05,
|
| 478 |
-
"loss": 0.376,
|
| 479 |
-
"step": 16000
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.08,
|
| 483 |
-
"eval_accuracy": 0.8409026798307475,
|
| 484 |
-
"eval_loss": 0.3675770163536072,
|
| 485 |
-
"eval_runtime": 130.1191,
|
| 486 |
-
"eval_samples_per_second": 81.733,
|
| 487 |
-
"eval_steps_per_second": 8.177,
|
| 488 |
-
"step": 16000
|
| 489 |
-
},
|
| 490 |
-
{
|
| 491 |
-
"epoch": 0.08,
|
| 492 |
-
"learning_rate": 1.8440645289331182e-05,
|
| 493 |
-
"loss": 0.3981,
|
| 494 |
-
"step": 16500
|
| 495 |
-
},
|
| 496 |
-
{
|
| 497 |
-
"epoch": 0.08,
|
| 498 |
-
"eval_accuracy": 0.8461683121767748,
|
| 499 |
-
"eval_loss": 0.36740317940711975,
|
| 500 |
-
"eval_runtime": 130.531,
|
| 501 |
-
"eval_samples_per_second": 81.475,
|
| 502 |
-
"eval_steps_per_second": 8.151,
|
| 503 |
-
"step": 16500
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.08,
|
| 507 |
-
"learning_rate": 1.839339211628061e-05,
|
| 508 |
-
"loss": 0.3805,
|
| 509 |
-
"step": 17000
|
| 510 |
-
},
|
| 511 |
-
{
|
| 512 |
-
"epoch": 0.08,
|
| 513 |
-
"eval_accuracy": 0.8456981664315938,
|
| 514 |
-
"eval_loss": 0.3746638894081116,
|
| 515 |
-
"eval_runtime": 130.414,
|
| 516 |
-
"eval_samples_per_second": 81.548,
|
| 517 |
-
"eval_steps_per_second": 8.159,
|
| 518 |
-
"step": 17000
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 0.08,
|
| 522 |
-
"learning_rate": 1.
|
| 523 |
-
"loss": 0.
|
| 524 |
-
"step":
|
| 525 |
},
|
| 526 |
{
|
| 527 |
"epoch": 0.08,
|
| 528 |
-
"
|
| 529 |
-
"
|
| 530 |
-
"
|
| 531 |
-
"eval_samples_per_second": 81.687,
|
| 532 |
-
"eval_steps_per_second": 8.173,
|
| 533 |
-
"step": 17500
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.09,
|
| 537 |
-
"learning_rate": 1.829888577017947e-05,
|
| 538 |
-
"loss": 0.3694,
|
| 539 |
-
"step": 18000
|
| 540 |
-
},
|
| 541 |
-
{
|
| 542 |
-
"epoch": 0.09,
|
| 543 |
-
"eval_accuracy": 0.8435354960037612,
|
| 544 |
-
"eval_loss": 0.36394280195236206,
|
| 545 |
-
"eval_runtime": 130.5033,
|
| 546 |
-
"eval_samples_per_second": 81.492,
|
| 547 |
-
"eval_steps_per_second": 8.153,
|
| 548 |
-
"step": 18000
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.09,
|
| 552 |
-
"learning_rate": 1.8251632597128898e-05,
|
| 553 |
-
"loss": 0.3937,
|
| 554 |
-
"step": 18500
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"epoch": 0.09,
|
| 558 |
-
"eval_accuracy": 0.8424071462153268,
|
| 559 |
-
"eval_loss": 0.36728155612945557,
|
| 560 |
-
"eval_runtime": 130.4243,
|
| 561 |
-
"eval_samples_per_second": 81.542,
|
| 562 |
-
"eval_steps_per_second": 8.158,
|
| 563 |
-
"step": 18500
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.09,
|
| 567 |
-
"learning_rate": 1.8204379424078328e-05,
|
| 568 |
-
"loss": 0.3839,
|
| 569 |
-
"step": 19000
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.09,
|
| 573 |
-
"eval_accuracy": 0.847202632816173,
|
| 574 |
-
"eval_loss": 0.3707529306411743,
|
| 575 |
-
"eval_runtime": 130.4582,
|
| 576 |
-
"eval_samples_per_second": 81.52,
|
| 577 |
-
"eval_steps_per_second": 8.156,
|
| 578 |
-
"step": 19000
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 0.09,
|
| 582 |
-
"learning_rate": 1.815712625102776e-05,
|
| 583 |
-
"loss": 0.3912,
|
| 584 |
-
"step": 19500
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.09,
|
| 588 |
-
"eval_accuracy": 0.8412787964268923,
|
| 589 |
-
"eval_loss": 0.37555885314941406,
|
| 590 |
-
"eval_runtime": 130.2205,
|
| 591 |
-
"eval_samples_per_second": 81.669,
|
| 592 |
-
"eval_steps_per_second": 8.171,
|
| 593 |
-
"step": 19500
|
| 594 |
-
},
|
| 595 |
-
{
|
| 596 |
-
"epoch": 0.09,
|
| 597 |
-
"learning_rate": 1.8109873077977186e-05,
|
| 598 |
-
"loss": 0.3863,
|
| 599 |
-
"step": 20000
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 0.09,
|
| 603 |
-
"
|
| 604 |
-
"
|
| 605 |
-
"
|
| 606 |
-
"eval_samples_per_second": 82.035,
|
| 607 |
-
"eval_steps_per_second": 8.207,
|
| 608 |
-
"step": 20000
|
| 609 |
-
},
|
| 610 |
-
{
|
| 611 |
-
"epoch": 0.1,
|
| 612 |
-
"learning_rate": 1.8062619904926617e-05,
|
| 613 |
-
"loss": 0.3961,
|
| 614 |
-
"step": 20500
|
| 615 |
-
},
|
| 616 |
-
{
|
| 617 |
-
"epoch": 0.1,
|
| 618 |
-
"eval_accuracy": 0.8417489421720733,
|
| 619 |
-
"eval_loss": 0.3776857852935791,
|
| 620 |
-
"eval_runtime": 129.4387,
|
| 621 |
-
"eval_samples_per_second": 82.162,
|
| 622 |
-
"eval_steps_per_second": 8.22,
|
| 623 |
-
"step": 20500
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 0.1,
|
| 627 |
-
"learning_rate": 1.8015366731876047e-05,
|
| 628 |
-
"loss": 0.3933,
|
| 629 |
-
"step": 21000
|
| 630 |
-
},
|
| 631 |
-
{
|
| 632 |
-
"epoch": 0.1,
|
| 633 |
-
"eval_accuracy": 0.8368594264221909,
|
| 634 |
-
"eval_loss": 0.3933545649051666,
|
| 635 |
-
"eval_runtime": 129.8078,
|
| 636 |
-
"eval_samples_per_second": 81.929,
|
| 637 |
-
"eval_steps_per_second": 8.197,
|
| 638 |
-
"step": 21000
|
| 639 |
-
},
|
| 640 |
-
{
|
| 641 |
-
"epoch": 0.1,
|
| 642 |
-
"learning_rate": 1.7968113558825478e-05,
|
| 643 |
-
"loss": 0.3902,
|
| 644 |
-
"step": 21500
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 0.1,
|
| 648 |
-
"eval_accuracy": 0.8445698166431593,
|
| 649 |
-
"eval_loss": 0.3732445240020752,
|
| 650 |
-
"eval_runtime": 129.549,
|
| 651 |
-
"eval_samples_per_second": 82.092,
|
| 652 |
-
"eval_steps_per_second": 8.213,
|
| 653 |
-
"step": 21500
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.1,
|
| 657 |
-
"learning_rate": 1.
|
| 658 |
-
"loss": 0.
|
| 659 |
-
"step":
|
| 660 |
},
|
| 661 |
{
|
| 662 |
"epoch": 0.1,
|
| 663 |
-
"
|
| 664 |
-
"
|
| 665 |
-
"
|
| 666 |
-
"eval_samples_per_second": 82.087,
|
| 667 |
-
"eval_steps_per_second": 8.213,
|
| 668 |
-
"step": 22000
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.11,
|
| 672 |
-
"learning_rate": 1.7873607212724336e-05,
|
| 673 |
-
"loss": 0.3938,
|
| 674 |
-
"step": 22500
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 0.11,
|
| 678 |
-
"eval_accuracy": 0.8378937470615891,
|
| 679 |
-
"eval_loss": 0.3706592619419098,
|
| 680 |
-
"eval_runtime": 129.5532,
|
| 681 |
-
"eval_samples_per_second": 82.09,
|
| 682 |
-
"eval_steps_per_second": 8.213,
|
| 683 |
-
"step": 22500
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 0.11,
|
| 687 |
-
"learning_rate": 1.7826354039673766e-05,
|
| 688 |
-
"loss": 0.3879,
|
| 689 |
-
"step": 23000
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 0.11,
|
| 693 |
-
"eval_accuracy": 0.8348848142924307,
|
| 694 |
-
"eval_loss": 0.4062308967113495,
|
| 695 |
-
"eval_runtime": 129.4397,
|
| 696 |
-
"eval_samples_per_second": 82.162,
|
| 697 |
-
"eval_steps_per_second": 8.22,
|
| 698 |
-
"step": 23000
|
| 699 |
-
},
|
| 700 |
-
{
|
| 701 |
-
"epoch": 0.11,
|
| 702 |
-
"learning_rate": 1.7779100866623197e-05,
|
| 703 |
-
"loss": 0.3958,
|
| 704 |
-
"step": 23500
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 0.11,
|
| 708 |
-
"eval_accuracy": 0.8449459332393042,
|
| 709 |
-
"eval_loss": 0.3733549416065216,
|
| 710 |
-
"eval_runtime": 129.5035,
|
| 711 |
-
"eval_samples_per_second": 82.121,
|
| 712 |
-
"eval_steps_per_second": 8.216,
|
| 713 |
-
"step": 23500
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"epoch": 0.11,
|
| 717 |
-
"learning_rate": 1.7731847693572624e-05,
|
| 718 |
-
"loss": 0.4004,
|
| 719 |
-
"step": 24000
|
| 720 |
},
|
| 721 |
{
|
| 722 |
"epoch": 0.11,
|
| 723 |
-
"
|
| 724 |
-
"
|
| 725 |
-
"
|
| 726 |
-
"eval_samples_per_second": 81.765,
|
| 727 |
-
"eval_steps_per_second": 8.18,
|
| 728 |
-
"step": 24000
|
| 729 |
-
},
|
| 730 |
-
{
|
| 731 |
-
"epoch": 0.12,
|
| 732 |
-
"learning_rate": 1.7684594520522055e-05,
|
| 733 |
-
"loss": 0.3989,
|
| 734 |
-
"step": 24500
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 0.12,
|
| 738 |
-
"eval_accuracy": 0.8387400094029149,
|
| 739 |
-
"eval_loss": 0.3675605356693268,
|
| 740 |
-
"eval_runtime": 129.789,
|
| 741 |
-
"eval_samples_per_second": 81.941,
|
| 742 |
-
"eval_steps_per_second": 8.198,
|
| 743 |
-
"step": 24500
|
| 744 |
-
},
|
| 745 |
-
{
|
| 746 |
-
"epoch": 0.12,
|
| 747 |
-
"learning_rate": 1.7637341347471485e-05,
|
| 748 |
-
"loss": 0.3993,
|
| 749 |
-
"step": 25000
|
| 750 |
-
},
|
| 751 |
-
{
|
| 752 |
-
"epoch": 0.12,
|
| 753 |
-
"eval_accuracy": 0.840620592383639,
|
| 754 |
-
"eval_loss": 0.3818625509738922,
|
| 755 |
-
"eval_runtime": 129.7415,
|
| 756 |
-
"eval_samples_per_second": 81.971,
|
| 757 |
-
"eval_steps_per_second": 8.201,
|
| 758 |
-
"step": 25000
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.12,
|
| 762 |
-
"learning_rate": 1.7590088174420913e-05,
|
| 763 |
-
"loss": 0.4096,
|
| 764 |
-
"step": 25500
|
| 765 |
-
},
|
| 766 |
-
{
|
| 767 |
-
"epoch": 0.12,
|
| 768 |
-
"eval_accuracy": 0.8440996708979783,
|
| 769 |
-
"eval_loss": 0.3813258409500122,
|
| 770 |
-
"eval_runtime": 129.9826,
|
| 771 |
-
"eval_samples_per_second": 81.819,
|
| 772 |
-
"eval_steps_per_second": 8.186,
|
| 773 |
-
"step": 25500
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.12,
|
| 777 |
-
"learning_rate": 1.7542835001370343e-05,
|
| 778 |
-
"loss": 0.3963,
|
| 779 |
-
"step": 26000
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.12,
|
| 783 |
-
"
|
| 784 |
-
"
|
| 785 |
-
"
|
| 786 |
-
"eval_samples_per_second": 82.013,
|
| 787 |
-
"eval_steps_per_second": 8.205,
|
| 788 |
-
"step": 26000
|
| 789 |
-
},
|
| 790 |
-
{
|
| 791 |
-
"epoch": 0.13,
|
| 792 |
-
"learning_rate": 1.749558182831977e-05,
|
| 793 |
-
"loss": 0.3901,
|
| 794 |
-
"step": 26500
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 0.13,
|
| 798 |
-
"eval_accuracy": 0.8412787964268923,
|
| 799 |
-
"eval_loss": 0.3806789815425873,
|
| 800 |
-
"eval_runtime": 129.8481,
|
| 801 |
-
"eval_samples_per_second": 81.903,
|
| 802 |
-
"eval_steps_per_second": 8.194,
|
| 803 |
-
"step": 26500
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 0.13,
|
| 807 |
-
"learning_rate": 1.74483286552692e-05,
|
| 808 |
-
"loss": 0.3953,
|
| 809 |
-
"step": 27000
|
| 810 |
-
},
|
| 811 |
-
{
|
| 812 |
-
"epoch": 0.13,
|
| 813 |
-
"eval_accuracy": 0.8321579689703809,
|
| 814 |
-
"eval_loss": 0.40024057030677795,
|
| 815 |
-
"eval_runtime": 130.2566,
|
| 816 |
-
"eval_samples_per_second": 81.647,
|
| 817 |
-
"eval_steps_per_second": 8.168,
|
| 818 |
-
"step": 27000
|
| 819 |
-
},
|
| 820 |
-
{
|
| 821 |
-
"epoch": 0.13,
|
| 822 |
-
"learning_rate": 1.7401075482218632e-05,
|
| 823 |
-
"loss": 0.3813,
|
| 824 |
-
"step": 27500
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 0.13,
|
| 828 |
-
"eval_accuracy": 0.8381758345086977,
|
| 829 |
-
"eval_loss": 0.41771912574768066,
|
| 830 |
-
"eval_runtime": 130.3445,
|
| 831 |
-
"eval_samples_per_second": 81.591,
|
| 832 |
-
"eval_steps_per_second": 8.163,
|
| 833 |
-
"step": 27500
|
| 834 |
-
},
|
| 835 |
-
{
|
| 836 |
-
"epoch": 0.13,
|
| 837 |
-
"learning_rate": 1.7353822309168062e-05,
|
| 838 |
-
"loss": 0.4054,
|
| 839 |
-
"step": 28000
|
| 840 |
-
},
|
| 841 |
-
{
|
| 842 |
-
"epoch": 0.13,
|
| 843 |
-
"eval_accuracy": 0.8449459332393042,
|
| 844 |
-
"eval_loss": 0.382539838552475,
|
| 845 |
-
"eval_runtime": 131.7654,
|
| 846 |
-
"eval_samples_per_second": 80.712,
|
| 847 |
-
"eval_steps_per_second": 8.075,
|
| 848 |
-
"step": 28000
|
| 849 |
},
|
| 850 |
{
|
| 851 |
"epoch": 0.13,
|
| 852 |
-
"learning_rate": 1.
|
| 853 |
-
"loss": 0.
|
| 854 |
-
"step":
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.13,
|
| 858 |
-
"
|
| 859 |
-
"
|
| 860 |
-
"
|
| 861 |
-
"eval_samples_per_second": 80.84,
|
| 862 |
-
"eval_steps_per_second": 8.088,
|
| 863 |
-
"step": 28500
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.14,
|
| 867 |
-
"learning_rate": 1.725931596306692e-05,
|
| 868 |
-
"loss": 0.3866,
|
| 869 |
-
"step": 29000
|
| 870 |
-
},
|
| 871 |
-
{
|
| 872 |
-
"epoch": 0.14,
|
| 873 |
-
"eval_accuracy": 0.8458862247296662,
|
| 874 |
-
"eval_loss": 0.37791889905929565,
|
| 875 |
-
"eval_runtime": 131.2917,
|
| 876 |
-
"eval_samples_per_second": 81.003,
|
| 877 |
-
"eval_steps_per_second": 8.104,
|
| 878 |
-
"step": 29000
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 0.14,
|
| 882 |
-
"learning_rate": 1.721206279001635e-05,
|
| 883 |
-
"loss": 0.4034,
|
| 884 |
-
"step": 29500
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 0.14,
|
| 888 |
-
"eval_accuracy": 0.841560883874001,
|
| 889 |
-
"eval_loss": 0.3799465000629425,
|
| 890 |
-
"eval_runtime": 131.3633,
|
| 891 |
-
"eval_samples_per_second": 80.959,
|
| 892 |
-
"eval_steps_per_second": 8.1,
|
| 893 |
-
"step": 29500
|
| 894 |
},
|
| 895 |
{
|
| 896 |
"epoch": 0.14,
|
| 897 |
-
"learning_rate": 1.
|
| 898 |
-
"loss": 0.
|
| 899 |
-
"step":
|
| 900 |
},
|
| 901 |
{
|
| 902 |
-
"epoch": 0.
|
| 903 |
-
"
|
| 904 |
-
"
|
| 905 |
-
"
|
| 906 |
-
"eval_samples_per_second": 80.988,
|
| 907 |
-
"eval_steps_per_second": 8.103,
|
| 908 |
-
"step": 30000
|
| 909 |
},
|
| 910 |
{
|
| 911 |
-
"epoch": 0.
|
| 912 |
-
"learning_rate": 1.
|
| 913 |
-
"loss": 0.
|
| 914 |
-
"step":
|
| 915 |
},
|
| 916 |
{
|
| 917 |
-
"epoch": 0.
|
| 918 |
-
"
|
| 919 |
-
"
|
| 920 |
-
"
|
| 921 |
-
"eval_samples_per_second": 80.931,
|
| 922 |
-
"eval_steps_per_second": 8.097,
|
| 923 |
-
"step": 30500
|
| 924 |
},
|
| 925 |
{
|
| 926 |
-
"epoch": 0.
|
| 927 |
-
"learning_rate": 1.
|
| 928 |
-
"loss": 0.
|
| 929 |
-
"step":
|
| 930 |
},
|
| 931 |
{
|
| 932 |
-
"epoch": 0.
|
| 933 |
-
"
|
| 934 |
-
"
|
| 935 |
-
"
|
| 936 |
-
"eval_samples_per_second": 80.807,
|
| 937 |
-
"eval_steps_per_second": 8.085,
|
| 938 |
-
"step": 31000
|
| 939 |
},
|
| 940 |
{
|
| 941 |
-
"epoch": 0.
|
| 942 |
-
"learning_rate": 1.
|
| 943 |
-
"loss": 0.
|
| 944 |
-
"step":
|
| 945 |
},
|
| 946 |
{
|
| 947 |
-
"epoch": 0.
|
| 948 |
-
"
|
| 949 |
-
"
|
| 950 |
-
"
|
| 951 |
-
"eval_samples_per_second": 80.931,
|
| 952 |
-
"eval_steps_per_second": 8.097,
|
| 953 |
-
"step": 31500
|
| 954 |
},
|
| 955 |
{
|
| 956 |
-
"epoch": 0.
|
| 957 |
-
"learning_rate": 1.
|
| 958 |
-
"loss": 0.
|
| 959 |
-
"step":
|
| 960 |
},
|
| 961 |
{
|
| 962 |
-
"epoch": 0.
|
| 963 |
-
"
|
| 964 |
-
"
|
| 965 |
-
"
|
| 966 |
-
"eval_samples_per_second": 80.8,
|
| 967 |
-
"eval_steps_per_second": 8.084,
|
| 968 |
-
"step": 32000
|
| 969 |
},
|
| 970 |
{
|
| 971 |
-
"epoch": 0.
|
| 972 |
-
"learning_rate": 1.
|
| 973 |
-
"loss": 0.
|
| 974 |
-
"step":
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"epoch": 0.
|
| 978 |
-
"
|
| 979 |
-
"
|
| 980 |
-
"
|
| 981 |
-
"eval_samples_per_second": 80.911,
|
| 982 |
-
"eval_steps_per_second": 8.095,
|
| 983 |
-
"step": 32500
|
| 984 |
},
|
| 985 |
{
|
| 986 |
-
"epoch": 0.
|
| 987 |
-
"learning_rate": 1.
|
| 988 |
-
"loss": 0.
|
| 989 |
-
"step":
|
| 990 |
},
|
| 991 |
{
|
| 992 |
-
"epoch": 0.
|
| 993 |
-
"
|
| 994 |
-
"
|
| 995 |
-
"
|
| 996 |
-
"eval_samples_per_second": 80.769,
|
| 997 |
-
"eval_steps_per_second": 8.081,
|
| 998 |
-
"step": 33000
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
-
"epoch": 0.
|
| 1002 |
-
"learning_rate": 1.
|
| 1003 |
-
"loss": 0.
|
| 1004 |
-
"step":
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"epoch": 0.
|
| 1008 |
-
"
|
| 1009 |
-
"
|
| 1010 |
-
"
|
| 1011 |
-
"eval_samples_per_second": 80.711,
|
| 1012 |
-
"eval_steps_per_second": 8.075,
|
| 1013 |
-
"step": 33500
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
-
"epoch": 0.
|
| 1017 |
-
"learning_rate": 1.
|
| 1018 |
-
"loss": 0.
|
| 1019 |
-
"step":
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
-
"epoch": 0.
|
| 1023 |
-
"
|
| 1024 |
-
"
|
| 1025 |
-
"
|
| 1026 |
-
"eval_samples_per_second": 80.59,
|
| 1027 |
-
"eval_steps_per_second": 8.063,
|
| 1028 |
-
"step": 34000
|
| 1029 |
},
|
| 1030 |
{
|
| 1031 |
-
"epoch": 0.
|
| 1032 |
-
"learning_rate": 1.
|
| 1033 |
-
"loss": 0.
|
| 1034 |
-
"step":
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
-
"epoch": 0.
|
| 1038 |
-
"
|
| 1039 |
-
"
|
| 1040 |
-
"
|
| 1041 |
-
"eval_samples_per_second": 80.824,
|
| 1042 |
-
"eval_steps_per_second": 8.086,
|
| 1043 |
-
"step": 34500
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
-
"epoch": 0.
|
| 1047 |
-
"learning_rate": 1.
|
| 1048 |
-
"loss": 0.
|
| 1049 |
-
"step":
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
-
"epoch": 0.
|
| 1053 |
-
"
|
| 1054 |
-
"
|
| 1055 |
-
"
|
| 1056 |
-
"eval_samples_per_second": 80.758,
|
| 1057 |
-
"eval_steps_per_second": 8.08,
|
| 1058 |
-
"step": 35000
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
-
"epoch": 0.
|
| 1062 |
-
"learning_rate": 1.
|
| 1063 |
-
"loss": 0.
|
| 1064 |
-
"step":
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
-
"epoch": 0.
|
| 1068 |
-
"
|
| 1069 |
-
"
|
| 1070 |
-
"
|
| 1071 |
-
"eval_samples_per_second": 80.676,
|
| 1072 |
-
"eval_steps_per_second": 8.071,
|
| 1073 |
-
"step": 35500
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch": 0.
|
| 1077 |
-
"learning_rate": 1.
|
| 1078 |
-
"loss": 0.
|
| 1079 |
-
"step":
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
-
"epoch": 0.
|
| 1083 |
-
"
|
| 1084 |
-
"
|
| 1085 |
-
"
|
| 1086 |
-
"eval_samples_per_second": 80.739,
|
| 1087 |
-
"eval_steps_per_second": 8.078,
|
| 1088 |
-
"step": 36000
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
-
"epoch": 0.
|
| 1092 |
-
"learning_rate": 1.
|
| 1093 |
-
"loss": 0.
|
| 1094 |
-
"step":
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch": 0.
|
| 1098 |
-
"
|
| 1099 |
-
"
|
| 1100 |
-
"
|
| 1101 |
-
"eval_samples_per_second": 80.675,
|
| 1102 |
-
"eval_steps_per_second": 8.071,
|
| 1103 |
-
"step": 36500
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
-
"epoch": 0.
|
| 1107 |
-
"learning_rate": 1.
|
| 1108 |
-
"loss": 0.
|
| 1109 |
-
"step":
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch": 0.
|
| 1113 |
-
"
|
| 1114 |
-
"
|
| 1115 |
-
"
|
| 1116 |
-
"eval_samples_per_second": 80.69,
|
| 1117 |
-
"eval_steps_per_second": 8.073,
|
| 1118 |
-
"step": 37000
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
-
"epoch": 0.
|
| 1122 |
-
"learning_rate": 1.
|
| 1123 |
-
"loss": 0.
|
| 1124 |
-
"step":
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
-
"epoch": 0.
|
| 1128 |
-
"
|
| 1129 |
-
"
|
| 1130 |
-
"
|
| 1131 |
-
"eval_samples_per_second": 80.713,
|
| 1132 |
-
"eval_steps_per_second": 8.075,
|
| 1133 |
-
"step": 37500
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
-
"epoch": 0.
|
| 1137 |
-
"learning_rate": 1.
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"step":
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
-
"epoch": 0.
|
| 1143 |
-
"
|
| 1144 |
-
"
|
| 1145 |
-
"
|
| 1146 |
-
"eval_samples_per_second": 80.531,
|
| 1147 |
-
"eval_steps_per_second": 8.057,
|
| 1148 |
-
"step": 38000
|
| 1149 |
},
|
| 1150 |
{
|
| 1151 |
-
"epoch": 0.
|
| 1152 |
-
"learning_rate": 1.
|
| 1153 |
-
"loss": 0.
|
| 1154 |
-
"step":
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
-
"epoch": 0.
|
| 1158 |
-
"
|
| 1159 |
-
"
|
| 1160 |
-
"
|
| 1161 |
-
"eval_samples_per_second": 80.779,
|
| 1162 |
-
"eval_steps_per_second": 8.082,
|
| 1163 |
-
"step": 38500
|
| 1164 |
},
|
| 1165 |
{
|
| 1166 |
-
"epoch": 0.
|
| 1167 |
-
"learning_rate": 1.
|
| 1168 |
-
"loss": 0.
|
| 1169 |
-
"step":
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
-
"epoch": 0.
|
| 1173 |
-
"
|
| 1174 |
-
"
|
| 1175 |
-
"
|
| 1176 |
-
"eval_samples_per_second": 80.619,
|
| 1177 |
-
"eval_steps_per_second": 8.066,
|
| 1178 |
-
"step": 39000
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch": 0.
|
| 1182 |
-
"learning_rate": 1.
|
| 1183 |
-
"loss": 0.
|
| 1184 |
-
"step":
|
| 1185 |
},
|
| 1186 |
{
|
| 1187 |
-
"epoch": 0.
|
| 1188 |
-
"
|
| 1189 |
-
"
|
| 1190 |
-
"
|
| 1191 |
-
"eval_samples_per_second": 80.732,
|
| 1192 |
-
"eval_steps_per_second": 8.077,
|
| 1193 |
-
"step": 39500
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch": 0.
|
| 1197 |
-
"learning_rate": 1.
|
| 1198 |
-
"loss": 0.
|
| 1199 |
-
"step":
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch": 0.
|
| 1203 |
-
"
|
| 1204 |
-
"
|
| 1205 |
-
"
|
| 1206 |
-
"eval_samples_per_second": 80.713,
|
| 1207 |
-
"eval_steps_per_second": 8.075,
|
| 1208 |
-
"step": 40000
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
-
"epoch": 0.
|
| 1212 |
-
"learning_rate": 1.
|
| 1213 |
-
"loss": 0.
|
| 1214 |
-
"step":
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
-
"epoch": 0.
|
| 1218 |
-
"
|
| 1219 |
-
"
|
| 1220 |
-
"
|
| 1221 |
-
"eval_samples_per_second": 80.663,
|
| 1222 |
-
"eval_steps_per_second": 8.07,
|
| 1223 |
-
"step": 40500
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
-
"epoch": 0.
|
| 1227 |
-
"learning_rate": 1.
|
| 1228 |
-
"loss": 0.
|
| 1229 |
-
"step":
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
-
"epoch": 0.
|
| 1233 |
-
"
|
| 1234 |
-
"
|
| 1235 |
-
"
|
| 1236 |
-
"eval_samples_per_second": 80.654,
|
| 1237 |
-
"eval_steps_per_second": 8.069,
|
| 1238 |
-
"step": 41000
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
-
"epoch": 0.
|
| 1242 |
-
"learning_rate": 1.
|
| 1243 |
-
"loss": 0.
|
| 1244 |
-
"step":
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
-
"epoch": 0.
|
| 1248 |
-
"
|
| 1249 |
-
"
|
| 1250 |
-
"
|
| 1251 |
-
"eval_samples_per_second": 80.477,
|
| 1252 |
-
"eval_steps_per_second": 8.051,
|
| 1253 |
-
"step": 41500
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
-
"epoch": 0.
|
| 1257 |
-
"learning_rate": 1.
|
| 1258 |
-
"loss": 0.
|
| 1259 |
-
"step":
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
-
"epoch": 0.
|
| 1263 |
-
"
|
| 1264 |
-
"
|
| 1265 |
-
"
|
| 1266 |
-
"eval_samples_per_second": 80.39,
|
| 1267 |
-
"eval_steps_per_second": 8.043,
|
| 1268 |
-
"step": 42000
|
| 1269 |
},
|
| 1270 |
{
|
| 1271 |
-
"epoch": 0.
|
| 1272 |
-
"learning_rate": 1.
|
| 1273 |
-
"loss": 0.
|
| 1274 |
-
"step":
|
| 1275 |
},
|
| 1276 |
{
|
| 1277 |
-
"epoch": 0.
|
| 1278 |
-
"
|
| 1279 |
-
"
|
| 1280 |
-
"
|
| 1281 |
-
"eval_samples_per_second": 80.481,
|
| 1282 |
-
"eval_steps_per_second": 8.052,
|
| 1283 |
-
"step": 42500
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
-
"epoch": 0.
|
| 1287 |
-
"learning_rate":
|
| 1288 |
-
"loss": 0.
|
| 1289 |
-
"step":
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
-
"epoch": 0.
|
| 1293 |
-
"
|
| 1294 |
-
"
|
| 1295 |
-
"
|
| 1296 |
-
"eval_samples_per_second": 80.259,
|
| 1297 |
-
"eval_steps_per_second": 8.03,
|
| 1298 |
-
"step": 43000
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
-
"epoch": 0.
|
| 1302 |
-
"learning_rate":
|
| 1303 |
-
"loss": 0.
|
| 1304 |
-
"step":
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
-
"epoch": 0.
|
| 1308 |
-
"
|
| 1309 |
-
"
|
| 1310 |
-
"
|
| 1311 |
-
"eval_samples_per_second": 80.36,
|
| 1312 |
-
"eval_steps_per_second": 8.04,
|
| 1313 |
-
"step": 43500
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
-
"epoch": 0.
|
| 1317 |
-
"learning_rate":
|
| 1318 |
-
"loss": 0.
|
| 1319 |
-
"step":
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
-
"epoch": 0.
|
| 1323 |
-
"
|
| 1324 |
-
"
|
| 1325 |
-
"
|
| 1326 |
-
"eval_samples_per_second": 80.272,
|
| 1327 |
-
"eval_steps_per_second": 8.031,
|
| 1328 |
-
"step": 44000
|
| 1329 |
},
|
| 1330 |
{
|
| 1331 |
-
"epoch": 0.
|
| 1332 |
-
"learning_rate":
|
| 1333 |
-
"loss": 0.
|
| 1334 |
-
"step":
|
| 1335 |
},
|
| 1336 |
{
|
| 1337 |
-
"epoch": 0.
|
| 1338 |
-
"
|
| 1339 |
-
"
|
| 1340 |
-
"
|
| 1341 |
-
"eval_samples_per_second": 80.348,
|
| 1342 |
-
"eval_steps_per_second": 8.039,
|
| 1343 |
-
"step": 44500
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
-
"epoch": 0.
|
| 1347 |
-
"learning_rate":
|
| 1348 |
-
"loss": 0.
|
| 1349 |
-
"step":
|
| 1350 |
},
|
| 1351 |
{
|
| 1352 |
-
"epoch": 0.
|
| 1353 |
-
"
|
| 1354 |
-
"
|
| 1355 |
-
"
|
| 1356 |
-
"eval_samples_per_second": 80.259,
|
| 1357 |
-
"eval_steps_per_second": 8.03,
|
| 1358 |
-
"step": 45000
|
| 1359 |
},
|
| 1360 |
{
|
| 1361 |
-
"epoch": 0.
|
| 1362 |
-
"learning_rate":
|
| 1363 |
-
"loss": 0.
|
| 1364 |
-
"step":
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
-
"epoch": 0.
|
| 1368 |
-
"
|
| 1369 |
-
"
|
| 1370 |
-
"
|
| 1371 |
-
"eval_samples_per_second": 80.39,
|
| 1372 |
-
"eval_steps_per_second": 8.043,
|
| 1373 |
-
"step": 45500
|
| 1374 |
},
|
| 1375 |
{
|
| 1376 |
-
"epoch": 0.
|
| 1377 |
-
"learning_rate":
|
| 1378 |
-
"loss": 0.
|
| 1379 |
-
"step":
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
-
"epoch": 0.
|
| 1383 |
-
"
|
| 1384 |
-
"
|
| 1385 |
-
"
|
| 1386 |
-
"eval_samples_per_second": 80.31,
|
| 1387 |
-
"eval_steps_per_second": 8.035,
|
| 1388 |
-
"step": 46000
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
-
"epoch": 0.
|
| 1392 |
-
"learning_rate":
|
| 1393 |
-
"loss": 0.
|
| 1394 |
-
"step":
|
| 1395 |
},
|
| 1396 |
{
|
| 1397 |
-
"epoch": 0.
|
| 1398 |
-
"
|
| 1399 |
-
"
|
| 1400 |
-
"
|
| 1401 |
-
"eval_samples_per_second": 80.249,
|
| 1402 |
-
"eval_steps_per_second": 8.029,
|
| 1403 |
-
"step": 46500
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
-
"epoch": 0.
|
| 1407 |
-
"learning_rate":
|
| 1408 |
-
"loss": 0.
|
| 1409 |
-
"step":
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
-
"epoch": 0.
|
| 1413 |
-
"
|
| 1414 |
-
"
|
| 1415 |
-
"
|
| 1416 |
-
"eval_samples_per_second": 80.344,
|
| 1417 |
-
"eval_steps_per_second": 8.038,
|
| 1418 |
-
"step": 47000
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
-
"epoch": 0.
|
| 1422 |
-
"learning_rate":
|
| 1423 |
-
"loss": 0.
|
| 1424 |
-
"step":
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
-
"epoch": 0.
|
| 1428 |
-
"
|
| 1429 |
-
"
|
| 1430 |
-
"
|
| 1431 |
-
"eval_samples_per_second": 80.301,
|
| 1432 |
-
"eval_steps_per_second": 8.034,
|
| 1433 |
-
"step": 47500
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
-
"epoch": 0.
|
| 1437 |
-
"learning_rate":
|
| 1438 |
-
"loss": 0.
|
| 1439 |
-
"step":
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
-
"epoch": 0.
|
| 1443 |
-
"
|
| 1444 |
-
"
|
| 1445 |
-
"
|
| 1446 |
-
"eval_samples_per_second": 80.242,
|
| 1447 |
-
"eval_steps_per_second": 8.028,
|
| 1448 |
-
"step": 48000
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
-
"epoch": 0.
|
| 1452 |
-
"learning_rate":
|
| 1453 |
-
"loss": 0.
|
| 1454 |
-
"step":
|
| 1455 |
},
|
| 1456 |
{
|
| 1457 |
-
"epoch": 0.
|
| 1458 |
-
"
|
| 1459 |
-
"
|
| 1460 |
-
"
|
| 1461 |
-
"eval_samples_per_second": 80.146,
|
| 1462 |
-
"eval_steps_per_second": 8.018,
|
| 1463 |
-
"step": 48500
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
-
"epoch": 0.
|
| 1467 |
-
"learning_rate":
|
| 1468 |
-
"loss": 0.
|
| 1469 |
-
"step":
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
-
"epoch": 0.
|
| 1473 |
-
"
|
| 1474 |
-
"
|
| 1475 |
-
"
|
| 1476 |
-
"eval_samples_per_second": 80.162,
|
| 1477 |
-
"eval_steps_per_second": 8.02,
|
| 1478 |
-
"step": 49000
|
| 1479 |
},
|
| 1480 |
{
|
| 1481 |
-
"epoch": 0.
|
| 1482 |
-
"learning_rate":
|
| 1483 |
-
"loss": 0.
|
| 1484 |
-
"step":
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
-
"epoch": 0.
|
| 1488 |
-
"
|
| 1489 |
-
"
|
| 1490 |
-
"eval_runtime": 132.561,
|
| 1491 |
-
"eval_samples_per_second": 80.227,
|
| 1492 |
-
"eval_steps_per_second": 8.026,
|
| 1493 |
"step": 49500
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
-
"epoch": 0.
|
| 1497 |
-
"learning_rate":
|
| 1498 |
-
"loss": 0.
|
| 1499 |
"step": 50000
|
| 1500 |
},
|
| 1501 |
{
|
| 1502 |
-
"epoch": 0.
|
| 1503 |
-
"eval_accuracy": 0.
|
| 1504 |
-
"eval_loss": 0.
|
| 1505 |
-
"eval_runtime":
|
| 1506 |
-
"eval_samples_per_second":
|
| 1507 |
-
"eval_steps_per_second":
|
| 1508 |
"step": 50000
|
| 1509 |
}
|
| 1510 |
],
|
| 1511 |
-
"max_steps":
|
| 1512 |
"num_train_epochs": 1,
|
| 1513 |
-
"total_flos":
|
| 1514 |
"trial_name": null,
|
| 1515 |
"trial_params": null
|
| 1516 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.4246249496936798,
|
| 3 |
"best_model_checkpoint": "OutModelPolicy/checkpoint-50000",
|
| 4 |
+
"epoch": 0.699623602501854,
|
| 5 |
"global_step": 50000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
{
|
| 11 |
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 1.986035512894063e-05,
|
| 13 |
+
"loss": 0.5152,
|
| 14 |
+
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 1.972043040844026e-05,
|
| 19 |
+
"loss": 0.4835,
|
| 20 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.02,
|
| 24 |
+
"learning_rate": 1.958050568793989e-05,
|
| 25 |
+
"loss": 0.4742,
|
| 26 |
+
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.03,
|
| 30 |
+
"learning_rate": 1.944058096743952e-05,
|
| 31 |
+
"loss": 0.4763,
|
| 32 |
+
"step": 2000
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.03,
|
| 36 |
+
"learning_rate": 1.930065624693915e-05,
|
| 37 |
+
"loss": 0.464,
|
| 38 |
+
"step": 2500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.04,
|
| 42 |
+
"learning_rate": 1.916101137587978e-05,
|
| 43 |
+
"loss": 0.4635,
|
| 44 |
+
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.05,
|
| 48 |
+
"learning_rate": 1.902108665537941e-05,
|
| 49 |
+
"loss": 0.4621,
|
| 50 |
+
"step": 3500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.06,
|
| 54 |
+
"learning_rate": 1.8881161934879037e-05,
|
| 55 |
+
"loss": 0.4525,
|
| 56 |
+
"step": 4000
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.06,
|
| 60 |
+
"learning_rate": 1.8741237214378666e-05,
|
| 61 |
+
"loss": 0.4636,
|
| 62 |
+
"step": 4500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.07,
|
| 66 |
+
"learning_rate": 1.8601872192760296e-05,
|
| 67 |
+
"loss": 0.4491,
|
| 68 |
+
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.08,
|
| 72 |
+
"learning_rate": 1.8461947472259925e-05,
|
| 73 |
+
"loss": 0.4502,
|
| 74 |
+
"step": 5500
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.08,
|
| 78 |
+
"learning_rate": 1.8322022751759554e-05,
|
| 79 |
+
"loss": 0.4512,
|
| 80 |
+
"step": 6000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.09,
|
| 84 |
+
"learning_rate": 1.8182377880700184e-05,
|
| 85 |
+
"loss": 0.4546,
|
| 86 |
+
"step": 6500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.1,
|
| 90 |
+
"learning_rate": 1.8042453160199812e-05,
|
| 91 |
+
"loss": 0.4477,
|
| 92 |
+
"step": 7000
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.1,
|
| 96 |
+
"learning_rate": 1.790252843969944e-05,
|
| 97 |
+
"loss": 0.4582,
|
| 98 |
+
"step": 7500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.11,
|
| 102 |
+
"learning_rate": 1.7762603719199074e-05,
|
| 103 |
+
"loss": 0.4452,
|
| 104 |
+
"step": 8000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.12,
|
| 108 |
+
"learning_rate": 1.7622678998698702e-05,
|
| 109 |
+
"loss": 0.4491,
|
| 110 |
+
"step": 8500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.13,
|
| 114 |
+
"learning_rate": 1.748275427819833e-05,
|
| 115 |
+
"loss": 0.4511,
|
| 116 |
+
"step": 9000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.13,
|
| 120 |
+
"learning_rate": 1.734282955769796e-05,
|
| 121 |
+
"loss": 0.442,
|
| 122 |
+
"step": 9500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.14,
|
| 126 |
+
"learning_rate": 1.720290483719759e-05,
|
| 127 |
+
"loss": 0.4403,
|
| 128 |
+
"step": 10000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.15,
|
| 132 |
+
"learning_rate": 1.7062980116697218e-05,
|
| 133 |
+
"loss": 0.4475,
|
| 134 |
+
"step": 10500
|
|
|
|
|
|
|
|
|
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.15,
|
| 138 |
+
"learning_rate": 1.6923335245637847e-05,
|
| 139 |
+
"loss": 0.4409,
|
| 140 |
+
"step": 11000
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 0.16,
|
| 144 |
+
"learning_rate": 1.678341052513748e-05,
|
| 145 |
+
"loss": 0.4427,
|
| 146 |
+
"step": 11500
|
|
|
|
|
|
|
|
|
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 0.17,
|
| 150 |
+
"learning_rate": 1.6643485804637108e-05,
|
| 151 |
+
"loss": 0.4335,
|
| 152 |
+
"step": 12000
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.17,
|
| 156 |
+
"learning_rate": 1.6503561084136737e-05,
|
| 157 |
+
"loss": 0.4474,
|
| 158 |
+
"step": 12500
|
|
|
|
|
|
|
|
|
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.18,
|
| 162 |
+
"learning_rate": 1.6363636363636366e-05,
|
| 163 |
+
"loss": 0.4552,
|
| 164 |
+
"step": 13000
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.19,
|
| 168 |
+
"learning_rate": 1.6223711643135995e-05,
|
| 169 |
+
"loss": 0.4405,
|
| 170 |
+
"step": 13500
|
|
|
|
|
|
|
|
|
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.2,
|
| 174 |
+
"learning_rate": 1.6083786922635624e-05,
|
| 175 |
+
"loss": 0.4331,
|
| 176 |
+
"step": 14000
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.2,
|
| 180 |
+
"learning_rate": 1.5943862202135252e-05,
|
| 181 |
+
"loss": 0.4338,
|
| 182 |
+
"step": 14500
|
|
|
|
|
|
|
|
|
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.21,
|
| 186 |
+
"learning_rate": 1.5804217331075882e-05,
|
| 187 |
+
"loss": 0.4339,
|
| 188 |
+
"step": 15000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 0.22,
|
| 192 |
+
"learning_rate": 1.566429261057551e-05,
|
| 193 |
+
"loss": 0.4355,
|
| 194 |
+
"step": 15500
|
|
|
|
|
|
|
|
|
|
| 195 |
},
|
| 196 |
{
|
| 197 |
+
"epoch": 0.22,
|
| 198 |
+
"learning_rate": 1.552436789007514e-05,
|
| 199 |
+
"loss": 0.4466,
|
| 200 |
+
"step": 16000
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 0.23,
|
| 204 |
+
"learning_rate": 1.538444316957477e-05,
|
| 205 |
+
"loss": 0.4368,
|
| 206 |
+
"step": 16500
|
|
|
|
|
|
|
|
|
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.24,
|
| 210 |
+
"learning_rate": 1.5244518449074399e-05,
|
| 211 |
+
"loss": 0.4343,
|
| 212 |
+
"step": 17000
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.24,
|
| 216 |
+
"learning_rate": 1.5104873578015029e-05,
|
| 217 |
+
"loss": 0.4404,
|
| 218 |
+
"step": 17500
|
|
|
|
|
|
|
|
|
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.25,
|
| 222 |
+
"learning_rate": 1.4964948857514659e-05,
|
| 223 |
+
"loss": 0.4367,
|
| 224 |
+
"step": 18000
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.26,
|
| 228 |
+
"learning_rate": 1.4825024137014288e-05,
|
| 229 |
+
"loss": 0.4389,
|
| 230 |
+
"step": 18500
|
|
|
|
|
|
|
|
|
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 0.27,
|
| 234 |
+
"learning_rate": 1.4685099416513917e-05,
|
| 235 |
+
"loss": 0.4396,
|
| 236 |
+
"step": 19000
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 0.27,
|
| 240 |
+
"learning_rate": 1.4545454545454546e-05,
|
| 241 |
+
"loss": 0.4364,
|
| 242 |
+
"step": 19500
|
|
|
|
|
|
|
|
|
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 0.28,
|
| 246 |
+
"learning_rate": 1.4405529824954177e-05,
|
| 247 |
+
"loss": 0.4434,
|
| 248 |
+
"step": 20000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.29,
|
| 252 |
+
"learning_rate": 1.4265605104453806e-05,
|
| 253 |
+
"loss": 0.4432,
|
| 254 |
+
"step": 20500
|
|
|
|
|
|
|
|
|
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.29,
|
| 258 |
+
"learning_rate": 1.4125680383953434e-05,
|
| 259 |
+
"loss": 0.4459,
|
| 260 |
+
"step": 21000
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"epoch": 0.3,
|
| 264 |
+
"learning_rate": 1.3986035512894066e-05,
|
| 265 |
+
"loss": 0.4415,
|
| 266 |
+
"step": 21500
|
|
|
|
|
|
|
|
|
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 0.31,
|
| 270 |
+
"learning_rate": 1.3846110792393695e-05,
|
| 271 |
+
"loss": 0.4373,
|
| 272 |
+
"step": 22000
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 0.31,
|
| 276 |
+
"learning_rate": 1.3706465921334324e-05,
|
| 277 |
+
"loss": 0.439,
|
| 278 |
+
"step": 22500
|
|
|
|
|
|
|
|
|
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 0.32,
|
| 282 |
+
"learning_rate": 1.3566541200833951e-05,
|
| 283 |
+
"loss": 0.4331,
|
| 284 |
+
"step": 23000
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 0.33,
|
| 288 |
+
"learning_rate": 1.342661648033358e-05,
|
| 289 |
+
"loss": 0.4361,
|
| 290 |
+
"step": 23500
|
|
|
|
|
|
|
|
|
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.34,
|
| 294 |
+
"learning_rate": 1.328669175983321e-05,
|
| 295 |
+
"loss": 0.4347,
|
| 296 |
+
"step": 24000
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.34,
|
| 300 |
+
"learning_rate": 1.314676703933284e-05,
|
| 301 |
+
"loss": 0.4319,
|
| 302 |
+
"step": 24500
|
|
|
|
|
|
|
|
|
|
| 303 |
},
|
| 304 |
{
|
| 305 |
+
"epoch": 0.35,
|
| 306 |
+
"learning_rate": 1.3006842318832468e-05,
|
| 307 |
+
"loss": 0.4321,
|
| 308 |
+
"step": 25000
|
| 309 |
},
|
| 310 |
{
|
| 311 |
+
"epoch": 0.36,
|
| 312 |
+
"learning_rate": 1.2866917598332099e-05,
|
| 313 |
+
"loss": 0.4208,
|
| 314 |
+
"step": 25500
|
|
|
|
|
|
|
|
|
|
| 315 |
},
|
| 316 |
{
|
| 317 |
+
"epoch": 0.36,
|
| 318 |
+
"learning_rate": 1.2727272727272728e-05,
|
| 319 |
+
"loss": 0.4404,
|
| 320 |
+
"step": 26000
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 0.37,
|
| 324 |
+
"learning_rate": 1.2587348006772357e-05,
|
| 325 |
+
"loss": 0.4248,
|
| 326 |
+
"step": 26500
|
|
|
|
|
|
|
|
|
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 0.38,
|
| 330 |
+
"learning_rate": 1.2447423286271986e-05,
|
| 331 |
+
"loss": 0.4351,
|
| 332 |
+
"step": 27000
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.38,
|
| 336 |
+
"learning_rate": 1.2307498565771617e-05,
|
| 337 |
+
"loss": 0.4302,
|
| 338 |
+
"step": 27500
|
|
|
|
|
|
|
|
|
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.39,
|
| 342 |
+
"learning_rate": 1.2167573845271245e-05,
|
| 343 |
+
"loss": 0.4335,
|
| 344 |
+
"step": 28000
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"epoch": 0.4,
|
| 348 |
+
"learning_rate": 1.2027649124770874e-05,
|
| 349 |
+
"loss": 0.4286,
|
| 350 |
+
"step": 28500
|
|
|
|
|
|
|
|
|
|
| 351 |
},
|
| 352 |
{
|
| 353 |
+
"epoch": 0.41,
|
| 354 |
+
"learning_rate": 1.1887724404270505e-05,
|
| 355 |
+
"loss": 0.4228,
|
| 356 |
+
"step": 29000
|
| 357 |
},
|
| 358 |
{
|
| 359 |
+
"epoch": 0.41,
|
| 360 |
+
"learning_rate": 1.1748079533211134e-05,
|
| 361 |
+
"loss": 0.4336,
|
| 362 |
+
"step": 29500
|
|
|
|
|
|
|
|
|
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 0.42,
|
| 366 |
+
"learning_rate": 1.1608154812710763e-05,
|
| 367 |
+
"loss": 0.4313,
|
| 368 |
+
"step": 30000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 0.43,
|
| 372 |
+
"learning_rate": 1.1468230092210392e-05,
|
| 373 |
+
"loss": 0.4322,
|
| 374 |
+
"step": 30500
|
|
|
|
|
|
|
|
|
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 0.43,
|
| 378 |
+
"learning_rate": 1.1328305371710019e-05,
|
| 379 |
+
"loss": 0.4439,
|
| 380 |
+
"step": 31000
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.44,
|
| 384 |
+
"learning_rate": 1.118866050065065e-05,
|
| 385 |
+
"loss": 0.4354,
|
| 386 |
+
"step": 31500
|
|
|
|
|
|
|
|
|
|
| 387 |
},
|
| 388 |
{
|
| 389 |
+
"epoch": 0.45,
|
| 390 |
+
"learning_rate": 1.104873578015028e-05,
|
| 391 |
+
"loss": 0.4268,
|
| 392 |
+
"step": 32000
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.45,
|
| 396 |
+
"learning_rate": 1.0908811059649908e-05,
|
| 397 |
+
"loss": 0.4281,
|
| 398 |
+
"step": 32500
|
|
|
|
|
|
|
|
|
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 0.46,
|
| 402 |
+
"learning_rate": 1.0768886339149539e-05,
|
| 403 |
+
"loss": 0.4261,
|
| 404 |
+
"step": 33000
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 0.47,
|
| 408 |
+
"learning_rate": 1.0628961618649167e-05,
|
| 409 |
+
"loss": 0.4272,
|
| 410 |
+
"step": 33500
|
|
|
|
|
|
|
|
|
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 0.48,
|
| 414 |
+
"learning_rate": 1.0489316747589797e-05,
|
| 415 |
+
"loss": 0.4261,
|
| 416 |
+
"step": 34000
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.48,
|
| 420 |
+
"learning_rate": 1.0349392027089426e-05,
|
| 421 |
+
"loss": 0.4345,
|
| 422 |
+
"step": 34500
|
|
|
|
|
|
|
|
|
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.49,
|
| 426 |
+
"learning_rate": 1.0209467306589056e-05,
|
| 427 |
+
"loss": 0.4292,
|
| 428 |
+
"step": 35000
|
| 429 |
},
|
| 430 |
{
|
| 431 |
+
"epoch": 0.5,
|
| 432 |
+
"learning_rate": 1.0069542586088685e-05,
|
| 433 |
+
"loss": 0.4323,
|
| 434 |
+
"step": 35500
|
|
|
|
|
|
|
|
|
|
| 435 |
},
|
| 436 |
{
|
| 437 |
+
"epoch": 0.5,
|
| 438 |
+
"learning_rate": 9.929617865588314e-06,
|
| 439 |
+
"loss": 0.4303,
|
| 440 |
+
"step": 36000
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 0.51,
|
| 444 |
+
"learning_rate": 9.789693145087945e-06,
|
| 445 |
+
"loss": 0.425,
|
| 446 |
+
"step": 36500
|
|
|
|
|
|
|
|
|
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 0.52,
|
| 450 |
+
"learning_rate": 9.649768424587572e-06,
|
| 451 |
+
"loss": 0.4297,
|
| 452 |
+
"step": 37000
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.52,
|
| 456 |
+
"learning_rate": 9.509843704087202e-06,
|
| 457 |
+
"loss": 0.4262,
|
| 458 |
+
"step": 37500
|
|
|
|
|
|
|
|
|
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.53,
|
| 462 |
+
"learning_rate": 9.370198833027832e-06,
|
| 463 |
+
"loss": 0.431,
|
| 464 |
+
"step": 38000
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.54,
|
| 468 |
+
"learning_rate": 9.230553961968461e-06,
|
| 469 |
+
"loss": 0.4225,
|
| 470 |
+
"step": 38500
|
|
|
|
|
|
|
|
|
|
| 471 |
},
|
| 472 |
{
|
| 473 |
+
"epoch": 0.55,
|
| 474 |
+
"learning_rate": 9.09062924146809e-06,
|
| 475 |
+
"loss": 0.44,
|
| 476 |
+
"step": 39000
|
| 477 |
},
|
| 478 |
{
|
| 479 |
+
"epoch": 0.55,
|
| 480 |
+
"learning_rate": 8.95070452096772e-06,
|
| 481 |
+
"loss": 0.4229,
|
| 482 |
+
"step": 39500
|
|
|
|
|
|
|
|
|
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 0.56,
|
| 486 |
+
"learning_rate": 8.81077980046735e-06,
|
| 487 |
+
"loss": 0.421,
|
| 488 |
+
"step": 40000
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 0.57,
|
| 492 |
+
"learning_rate": 8.670855079966978e-06,
|
| 493 |
+
"loss": 0.4366,
|
| 494 |
+
"step": 40500
|
|
|
|
|
|
|
|
|
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 0.57,
|
| 498 |
+
"learning_rate": 8.530930359466609e-06,
|
| 499 |
+
"loss": 0.424,
|
| 500 |
+
"step": 41000
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.58,
|
| 504 |
+
"learning_rate": 8.391005638966236e-06,
|
| 505 |
+
"loss": 0.4236,
|
| 506 |
+
"step": 41500
|
|
|
|
|
|
|
|
|
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.59,
|
| 510 |
+
"learning_rate": 8.251080918465867e-06,
|
| 511 |
+
"loss": 0.4222,
|
| 512 |
+
"step": 42000
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.59,
|
| 516 |
+
"learning_rate": 8.111436047406496e-06,
|
| 517 |
+
"loss": 0.4187,
|
| 518 |
+
"step": 42500
|
|
|
|
|
|
|
|
|
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.6,
|
| 522 |
+
"learning_rate": 7.971511326906125e-06,
|
| 523 |
+
"loss": 0.4444,
|
| 524 |
+
"step": 43000
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.61,
|
| 528 |
+
"learning_rate": 7.831586606405754e-06,
|
| 529 |
+
"loss": 0.4278,
|
| 530 |
+
"step": 43500
|
|
|
|
|
|
|
|
|
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.62,
|
| 534 |
+
"learning_rate": 7.691661885905384e-06,
|
| 535 |
+
"loss": 0.431,
|
| 536 |
+
"step": 44000
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 0.62,
|
| 540 |
+
"learning_rate": 7.552017014846014e-06,
|
| 541 |
+
"loss": 0.4258,
|
| 542 |
+
"step": 44500
|
|
|
|
|
|
|
|
|
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.63,
|
| 546 |
+
"learning_rate": 7.4123721437866435e-06,
|
| 547 |
+
"loss": 0.4252,
|
| 548 |
+
"step": 45000
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.64,
|
| 552 |
+
"learning_rate": 7.272447423286273e-06,
|
| 553 |
+
"loss": 0.4279,
|
| 554 |
+
"step": 45500
|
|
|
|
|
|
|
|
|
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 0.64,
|
| 558 |
+
"learning_rate": 7.132522702785902e-06,
|
| 559 |
+
"loss": 0.4279,
|
| 560 |
+
"step": 46000
|
| 561 |
},
|
| 562 |
{
|
| 563 |
+
"epoch": 0.65,
|
| 564 |
+
"learning_rate": 6.992597982285531e-06,
|
| 565 |
+
"loss": 0.4178,
|
| 566 |
+
"step": 46500
|
|
|
|
|
|
|
|
|
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 0.66,
|
| 570 |
+
"learning_rate": 6.85267326178516e-06,
|
| 571 |
+
"loss": 0.4201,
|
| 572 |
+
"step": 47000
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 0.66,
|
| 576 |
+
"learning_rate": 6.71302839072579e-06,
|
| 577 |
+
"loss": 0.4244,
|
| 578 |
+
"step": 47500
|
|
|
|
|
|
|
|
|
|
| 579 |
},
|
| 580 |
{
|
| 581 |
+
"epoch": 0.67,
|
| 582 |
+
"learning_rate": 6.573103670225419e-06,
|
| 583 |
+
"loss": 0.4169,
|
| 584 |
+
"step": 48000
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 0.68,
|
| 588 |
+
"learning_rate": 6.433178949725049e-06,
|
| 589 |
+
"loss": 0.427,
|
| 590 |
+
"step": 48500
|
|
|
|
|
|
|
|
|
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.69,
|
| 594 |
+
"learning_rate": 6.293254229224678e-06,
|
| 595 |
+
"loss": 0.4325,
|
| 596 |
+
"step": 49000
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 0.69,
|
| 600 |
+
"learning_rate": 6.153329508724307e-06,
|
| 601 |
+
"loss": 0.4206,
|
|
|
|
|
|
|
|
|
|
| 602 |
"step": 49500
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 0.7,
|
| 606 |
+
"learning_rate": 6.013404788223935e-06,
|
| 607 |
+
"loss": 0.4342,
|
| 608 |
"step": 50000
|
| 609 |
},
|
| 610 |
{
|
| 611 |
+
"epoch": 0.7,
|
| 612 |
+
"eval_accuracy": 0.807313903842584,
|
| 613 |
+
"eval_loss": 0.4246249496936798,
|
| 614 |
+
"eval_runtime": 59.7232,
|
| 615 |
+
"eval_samples_per_second": 90.199,
|
| 616 |
+
"eval_steps_per_second": 6.028,
|
| 617 |
"step": 50000
|
| 618 |
}
|
| 619 |
],
|
| 620 |
+
"max_steps": 71467,
|
| 621 |
"num_train_epochs": 1,
|
| 622 |
+
"total_flos": 9.090578558220155e+16,
|
| 623 |
"trial_name": null,
|
| 624 |
"trial_params": null
|
| 625 |
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9828a2737d623495fefef08d84cec0096f77c2371e80203d4f24c0e7da7cea9a
|
| 3 |
+
size 3311
|