Upload folder using huggingface_hub
Browse files- config.json +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1503 -443
- training_args.bin +2 -2
config.json
CHANGED
|
@@ -51,7 +51,7 @@
|
|
| 51 |
"ssm_cfg": {},
|
| 52 |
"torch_dtype": "float32",
|
| 53 |
"transformers_version": "4.17.0",
|
| 54 |
-
"use_auth_token": "redacted
|
| 55 |
"use_fast_path": true,
|
| 56 |
"vocab_size": 50277
|
| 57 |
}
|
|
|
|
| 51 |
"ssm_cfg": {},
|
| 52 |
"torch_dtype": "float32",
|
| 53 |
"transformers_version": "4.17.0",
|
| 54 |
+
"use_auth_token": "redacted",
|
| 55 |
"use_fast_path": true,
|
| 56 |
"vocab_size": 50277
|
| 57 |
}
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 190607546
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2105b8ff85b1b1cc88d293c74bdfc38fd244c9e9aff2573913cad0428f1af944
|
| 3 |
size 190607546
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 98566808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30b8bfe8890d332c49b8a0fed9423208d88581ac38c0c529dd2b20502f3ae46b
|
| 3 |
size 98566808
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f264f6864cd00354ec50b255292ba487c4cf42cbcbd037946b4559cba635d0c9
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aa6e6e0247647aad7e38a4ff2af871fe185ff494bc05375f7f54d1c1da31fc8
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,756 +1,1816 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.
|
| 5 |
-
"global_step":
|
| 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":
|
| 13 |
-
"loss":
|
| 14 |
"step": 1000
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 0.
|
| 20 |
"step": 2000
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss": 0.
|
| 26 |
"step": 3000
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss": 0.
|
| 32 |
"step": 4000
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 0.
|
| 38 |
-
"step": 5000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.06,
|
| 42 |
-
"eval_runtime": 104.0589,
|
| 43 |
-
"eval_samples_per_second": 1100.002,
|
| 44 |
-
"eval_steps_per_second": 137.509,
|
| 45 |
"step": 5000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 6000
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 7000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 8000
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 9000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 0.
|
| 75 |
-
"step": 10000
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.11,
|
| 79 |
-
"eval_runtime": 104.0864,
|
| 80 |
-
"eval_samples_per_second": 1099.711,
|
| 81 |
-
"eval_steps_per_second": 137.472,
|
| 82 |
"step": 10000
|
| 83 |
},
|
| 84 |
{
|
| 85 |
-
"epoch": 0.
|
| 86 |
-
"learning_rate":
|
| 87 |
-
"loss": 0.
|
| 88 |
"step": 11000
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"epoch": 0.
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 12000
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 13000
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"learning_rate":
|
| 105 |
-
"loss": 0.
|
| 106 |
"step": 14000
|
| 107 |
},
|
| 108 |
{
|
| 109 |
-
"epoch": 0.
|
| 110 |
-
"learning_rate":
|
| 111 |
-
"loss": 0.
|
| 112 |
-
"step": 15000
|
| 113 |
-
},
|
| 114 |
-
{
|
| 115 |
-
"epoch": 0.17,
|
| 116 |
-
"eval_runtime": 103.2729,
|
| 117 |
-
"eval_samples_per_second": 1108.374,
|
| 118 |
-
"eval_steps_per_second": 138.555,
|
| 119 |
"step": 15000
|
| 120 |
},
|
| 121 |
{
|
| 122 |
-
"epoch": 0.
|
| 123 |
-
"learning_rate":
|
| 124 |
-
"loss": 0.
|
| 125 |
"step": 16000
|
| 126 |
},
|
| 127 |
{
|
| 128 |
-
"epoch": 0.
|
| 129 |
-
"learning_rate":
|
| 130 |
-
"loss": 0.
|
| 131 |
"step": 17000
|
| 132 |
},
|
| 133 |
{
|
| 134 |
-
"epoch": 0.
|
| 135 |
-
"learning_rate":
|
| 136 |
-
"loss": 0.
|
| 137 |
"step": 18000
|
| 138 |
},
|
| 139 |
{
|
| 140 |
-
"epoch": 0.
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 19000
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 0.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 0.
|
| 149 |
-
"step": 20000
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.23,
|
| 153 |
-
"eval_runtime": 104.205,
|
| 154 |
-
"eval_samples_per_second": 1098.46,
|
| 155 |
-
"eval_steps_per_second": 137.316,
|
| 156 |
"step": 20000
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"learning_rate":
|
| 161 |
-
"loss": 0.
|
| 162 |
"step": 21000
|
| 163 |
},
|
| 164 |
{
|
| 165 |
-
"epoch": 0.
|
| 166 |
-
"learning_rate":
|
| 167 |
-
"loss": 0.
|
| 168 |
"step": 22000
|
| 169 |
},
|
| 170 |
{
|
| 171 |
-
"epoch": 0.
|
| 172 |
-
"learning_rate":
|
| 173 |
-
"loss": 0.
|
| 174 |
"step": 23000
|
| 175 |
},
|
| 176 |
{
|
| 177 |
-
"epoch": 0.
|
| 178 |
-
"learning_rate":
|
| 179 |
-
"loss": 0.
|
| 180 |
"step": 24000
|
| 181 |
},
|
| 182 |
{
|
| 183 |
-
"epoch": 0.
|
| 184 |
-
"learning_rate":
|
| 185 |
-
"loss": 0.
|
| 186 |
-
"step": 25000
|
| 187 |
-
},
|
| 188 |
-
{
|
| 189 |
-
"epoch": 0.28,
|
| 190 |
-
"eval_runtime": 104.294,
|
| 191 |
-
"eval_samples_per_second": 1097.522,
|
| 192 |
-
"eval_steps_per_second": 137.199,
|
| 193 |
"step": 25000
|
| 194 |
},
|
| 195 |
{
|
| 196 |
-
"epoch": 0.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 26000
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch": 0.
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 27000
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"learning_rate":
|
| 210 |
-
"loss": 0.
|
| 211 |
"step": 28000
|
| 212 |
},
|
| 213 |
{
|
| 214 |
-
"epoch": 0.
|
| 215 |
-
"learning_rate":
|
| 216 |
-
"loss": 0.
|
| 217 |
"step": 29000
|
| 218 |
},
|
| 219 |
{
|
| 220 |
-
"epoch": 0.
|
| 221 |
-
"learning_rate":
|
| 222 |
-
"loss": 0.
|
| 223 |
-
"step": 30000
|
| 224 |
-
},
|
| 225 |
-
{
|
| 226 |
-
"epoch": 0.34,
|
| 227 |
-
"eval_runtime": 103.1778,
|
| 228 |
-
"eval_samples_per_second": 1109.395,
|
| 229 |
-
"eval_steps_per_second": 138.683,
|
| 230 |
"step": 30000
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 0.
|
| 234 |
-
"learning_rate": 0.
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 31000
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch": 0.
|
| 240 |
-
"learning_rate": 0.
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 32000
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch": 0.
|
| 246 |
-
"learning_rate": 0.
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 33000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"learning_rate": 0.
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 34000
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"learning_rate": 0.
|
| 259 |
-
"loss": 0.
|
| 260 |
-
"step": 35000
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"epoch": 0.39,
|
| 264 |
-
"eval_runtime": 104.2139,
|
| 265 |
-
"eval_samples_per_second": 1098.366,
|
| 266 |
-
"eval_steps_per_second": 137.304,
|
| 267 |
"step": 35000
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch": 0.
|
| 271 |
-
"learning_rate": 0.
|
| 272 |
-
"loss": 0.
|
| 273 |
"step": 36000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch": 0.
|
| 277 |
-
"learning_rate": 0.
|
| 278 |
-
"loss": 0.
|
| 279 |
"step": 37000
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch": 0.
|
| 283 |
-
"learning_rate": 0.
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 38000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch": 0.
|
| 289 |
-
"learning_rate": 0.
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 39000
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 0.
|
| 295 |
-
"learning_rate": 0.
|
| 296 |
-
"loss": 0.
|
| 297 |
-
"step": 40000
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.45,
|
| 301 |
-
"eval_runtime": 103.2048,
|
| 302 |
-
"eval_samples_per_second": 1109.105,
|
| 303 |
-
"eval_steps_per_second": 138.647,
|
| 304 |
"step": 40000
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"epoch": 0.
|
| 308 |
-
"learning_rate": 0.
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 41000
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch": 0.
|
| 314 |
-
"learning_rate": 0.
|
| 315 |
-
"loss": 0.
|
| 316 |
"step": 42000
|
| 317 |
},
|
| 318 |
{
|
| 319 |
-
"epoch": 0.
|
| 320 |
-
"learning_rate": 0.
|
| 321 |
-
"loss": 0.
|
| 322 |
"step": 43000
|
| 323 |
},
|
| 324 |
{
|
| 325 |
-
"epoch": 0.
|
| 326 |
-
"learning_rate": 0.
|
| 327 |
-
"loss": 0.
|
| 328 |
"step": 44000
|
| 329 |
},
|
| 330 |
{
|
| 331 |
-
"epoch": 0.
|
| 332 |
-
"learning_rate":
|
| 333 |
-
"loss": 0.
|
| 334 |
-
"step": 45000
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"epoch": 0.51,
|
| 338 |
-
"eval_runtime": 103.2517,
|
| 339 |
-
"eval_samples_per_second": 1108.601,
|
| 340 |
-
"eval_steps_per_second": 138.584,
|
| 341 |
"step": 45000
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch": 0.
|
| 345 |
-
"learning_rate":
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 46000
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 47000
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"learning_rate":
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 48000
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"learning_rate":
|
| 364 |
-
"loss": 0.
|
| 365 |
"step": 49000
|
| 366 |
},
|
| 367 |
{
|
| 368 |
-
"epoch": 0.
|
| 369 |
-
"learning_rate":
|
| 370 |
-
"loss": 0.
|
| 371 |
-
"step": 50000
|
| 372 |
-
},
|
| 373 |
-
{
|
| 374 |
-
"epoch": 0.56,
|
| 375 |
-
"eval_runtime": 103.4614,
|
| 376 |
-
"eval_samples_per_second": 1106.355,
|
| 377 |
-
"eval_steps_per_second": 138.303,
|
| 378 |
"step": 50000
|
| 379 |
},
|
| 380 |
{
|
| 381 |
-
"epoch": 0.
|
| 382 |
-
"learning_rate":
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 51000
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch": 0.
|
| 388 |
-
"learning_rate":
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 52000
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch": 0.
|
| 394 |
-
"learning_rate":
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 53000
|
| 397 |
},
|
| 398 |
{
|
| 399 |
-
"epoch": 0.
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 54000
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch": 0.
|
| 406 |
-
"learning_rate":
|
| 407 |
-
"loss": 0.
|
| 408 |
-
"step": 55000
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.62,
|
| 412 |
-
"eval_runtime": 103.3222,
|
| 413 |
-
"eval_samples_per_second": 1107.845,
|
| 414 |
-
"eval_steps_per_second": 138.489,
|
| 415 |
"step": 55000
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch": 0.
|
| 419 |
-
"learning_rate":
|
| 420 |
-
"loss": 0.
|
| 421 |
"step": 56000
|
| 422 |
},
|
| 423 |
{
|
| 424 |
-
"epoch": 0.
|
| 425 |
-
"learning_rate":
|
| 426 |
-
"loss": 0.
|
| 427 |
"step": 57000
|
| 428 |
},
|
| 429 |
{
|
| 430 |
-
"epoch": 0.
|
| 431 |
-
"learning_rate":
|
| 432 |
-
"loss": 0.
|
| 433 |
"step": 58000
|
| 434 |
},
|
| 435 |
{
|
| 436 |
-
"epoch": 0.
|
| 437 |
-
"learning_rate":
|
| 438 |
-
"loss": 0.
|
| 439 |
"step": 59000
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch": 0.
|
| 443 |
-
"learning_rate":
|
| 444 |
-
"loss": 0.
|
| 445 |
-
"step": 60000
|
| 446 |
-
},
|
| 447 |
-
{
|
| 448 |
-
"epoch": 0.68,
|
| 449 |
-
"eval_runtime": 104.558,
|
| 450 |
-
"eval_samples_per_second": 1094.752,
|
| 451 |
-
"eval_steps_per_second": 136.852,
|
| 452 |
"step": 60000
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 61000
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 62000
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"learning_rate":
|
| 469 |
-
"loss": 0.
|
| 470 |
"step": 63000
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch": 0.
|
| 474 |
-
"learning_rate":
|
| 475 |
-
"loss": 0.
|
| 476 |
"step": 64000
|
| 477 |
},
|
| 478 |
{
|
| 479 |
-
"epoch": 0.
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 0.
|
| 482 |
-
"step": 65000
|
| 483 |
-
},
|
| 484 |
-
{
|
| 485 |
-
"epoch": 0.73,
|
| 486 |
-
"eval_runtime": 106.3469,
|
| 487 |
-
"eval_samples_per_second": 1076.336,
|
| 488 |
-
"eval_steps_per_second": 134.55,
|
| 489 |
"step": 65000
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch": 0.
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 66000
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 0.
|
| 499 |
-
"learning_rate":
|
| 500 |
-
"loss": 0.
|
| 501 |
"step": 67000
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch": 0.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 68000
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 0.
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 69000
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 0.
|
| 517 |
-
"learning_rate":
|
| 518 |
-
"loss": 0.
|
| 519 |
-
"step": 70000
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.79,
|
| 523 |
-
"eval_runtime": 133.9365,
|
| 524 |
-
"eval_samples_per_second": 854.621,
|
| 525 |
-
"eval_steps_per_second": 106.834,
|
| 526 |
"step": 70000
|
| 527 |
},
|
| 528 |
{
|
| 529 |
-
"epoch": 0.
|
| 530 |
-
"learning_rate":
|
| 531 |
-
"loss": 0.
|
| 532 |
"step": 71000
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"epoch": 0.
|
| 536 |
-
"learning_rate":
|
| 537 |
-
"loss": 0.
|
| 538 |
"step": 72000
|
| 539 |
},
|
| 540 |
{
|
| 541 |
-
"epoch": 0.
|
| 542 |
-
"learning_rate":
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 73000
|
| 545 |
},
|
| 546 |
{
|
| 547 |
-
"epoch": 0.
|
| 548 |
-
"learning_rate":
|
| 549 |
-
"loss": 0.
|
| 550 |
"step": 74000
|
| 551 |
},
|
| 552 |
{
|
| 553 |
-
"epoch": 0.
|
| 554 |
-
"learning_rate":
|
| 555 |
-
"loss": 0.
|
| 556 |
-
"step": 75000
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.84,
|
| 560 |
-
"eval_runtime": 104.105,
|
| 561 |
-
"eval_samples_per_second": 1099.515,
|
| 562 |
-
"eval_steps_per_second": 137.448,
|
| 563 |
"step": 75000
|
| 564 |
},
|
| 565 |
{
|
| 566 |
-
"epoch": 0.
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 76000
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch": 0.
|
| 573 |
-
"learning_rate":
|
| 574 |
-
"loss": 0.
|
| 575 |
"step": 77000
|
| 576 |
},
|
| 577 |
{
|
| 578 |
-
"epoch": 0.
|
| 579 |
-
"learning_rate":
|
| 580 |
-
"loss": 0.
|
| 581 |
"step": 78000
|
| 582 |
},
|
| 583 |
{
|
| 584 |
-
"epoch": 0.
|
| 585 |
-
"learning_rate":
|
| 586 |
-
"loss": 0.
|
| 587 |
"step": 79000
|
| 588 |
},
|
| 589 |
{
|
| 590 |
-
"epoch": 0.
|
| 591 |
-
"learning_rate":
|
| 592 |
-
"loss": 0.
|
| 593 |
-
"step": 80000
|
| 594 |
-
},
|
| 595 |
-
{
|
| 596 |
-
"epoch": 0.9,
|
| 597 |
-
"eval_runtime": 104.5533,
|
| 598 |
-
"eval_samples_per_second": 1094.801,
|
| 599 |
-
"eval_steps_per_second": 136.858,
|
| 600 |
"step": 80000
|
| 601 |
},
|
| 602 |
{
|
| 603 |
-
"epoch": 0.
|
| 604 |
-
"learning_rate":
|
| 605 |
-
"loss": 0.
|
| 606 |
"step": 81000
|
| 607 |
},
|
| 608 |
{
|
| 609 |
-
"epoch": 0.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 82000
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch": 0.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 83000
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch": 0.
|
| 622 |
-
"learning_rate":
|
| 623 |
-
"loss": 0.
|
| 624 |
"step": 84000
|
| 625 |
},
|
| 626 |
{
|
| 627 |
-
"epoch": 0.
|
| 628 |
-
"learning_rate":
|
| 629 |
-
"loss": 0.
|
| 630 |
"step": 85000
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch": 0.
|
| 634 |
-
"
|
| 635 |
-
"
|
| 636 |
-
"
|
| 637 |
-
"step": 85000
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch": 0.
|
| 641 |
-
"learning_rate":
|
| 642 |
-
"loss": 0.
|
| 643 |
-
"step": 86000
|
| 644 |
-
},
|
| 645 |
-
{
|
| 646 |
-
"epoch": 0.98,
|
| 647 |
-
"learning_rate": 1.6642125030122378e-05,
|
| 648 |
-
"loss": 0.2836,
|
| 649 |
"step": 87000
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch": 0.
|
| 653 |
-
"learning_rate":
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 88000
|
| 656 |
},
|
| 657 |
{
|
| 658 |
-
"epoch":
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 89000
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"learning_rate":
|
| 666 |
-
"loss": 0.
|
| 667 |
-
"step": 90000
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 1.01,
|
| 671 |
-
"eval_runtime": 102.1628,
|
| 672 |
-
"eval_samples_per_second": 1120.417,
|
| 673 |
-
"eval_steps_per_second": 140.061,
|
| 674 |
"step": 90000
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch":
|
| 678 |
-
"learning_rate":
|
| 679 |
-
"loss": 0.
|
| 680 |
"step": 91000
|
| 681 |
},
|
| 682 |
{
|
| 683 |
-
"epoch":
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss": 0.
|
| 686 |
"step": 92000
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch":
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 93000
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"epoch":
|
| 696 |
-
"learning_rate":
|
| 697 |
-
"loss": 0.
|
| 698 |
"step": 94000
|
| 699 |
},
|
| 700 |
{
|
| 701 |
-
"epoch":
|
| 702 |
-
"learning_rate":
|
| 703 |
-
"loss": 0.
|
| 704 |
-
"step": 95000
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 1.07,
|
| 708 |
-
"eval_runtime": 103.598,
|
| 709 |
-
"eval_samples_per_second": 1104.895,
|
| 710 |
-
"eval_steps_per_second": 138.12,
|
| 711 |
"step": 95000
|
| 712 |
},
|
| 713 |
{
|
| 714 |
-
"epoch":
|
| 715 |
-
"learning_rate":
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 96000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
-
"learning_rate":
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 97000
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch":
|
| 727 |
-
"learning_rate":
|
| 728 |
-
"loss": 0.
|
| 729 |
"step": 98000
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
-
"learning_rate":
|
| 734 |
-
"loss": 0.
|
| 735 |
"step": 99000
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
-
"learning_rate":
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 100000
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch":
|
| 745 |
-
"
|
| 746 |
-
"
|
| 747 |
-
"
|
| 748 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 749 |
}
|
| 750 |
],
|
| 751 |
-
"max_steps":
|
| 752 |
-
"num_train_epochs":
|
| 753 |
-
"total_flos":
|
| 754 |
"trial_name": null,
|
| 755 |
"trial_params": null
|
| 756 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.7776935019376858,
|
| 5 |
+
"global_step": 300000,
|
| 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": 3.3333333333333333e-06,
|
| 13 |
+
"loss": 1.243,
|
| 14 |
"step": 1000
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 6.666666666666667e-06,
|
| 19 |
+
"loss": 0.6562,
|
| 20 |
"step": 2000
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 0.02,
|
| 24 |
+
"learning_rate": 9.999999999999999e-06,
|
| 25 |
+
"loss": 0.4946,
|
| 26 |
"step": 3000
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 31 |
+
"loss": 0.46,
|
| 32 |
"step": 4000
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.03,
|
| 36 |
+
"learning_rate": 1.6666666666666664e-05,
|
| 37 |
+
"loss": 0.4355,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
"step": 5000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.04,
|
| 42 |
+
"learning_rate": 1.9999999999999998e-05,
|
| 43 |
+
"loss": 0.4203,
|
| 44 |
"step": 6000
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.04,
|
| 48 |
+
"learning_rate": 2.3333333333333332e-05,
|
| 49 |
+
"loss": 0.4075,
|
| 50 |
"step": 7000
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.05,
|
| 54 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 55 |
+
"loss": 0.3986,
|
| 56 |
"step": 8000
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.05,
|
| 60 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 61 |
+
"loss": 0.3922,
|
| 62 |
"step": 9000
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 0.06,
|
| 66 |
+
"learning_rate": 3.333333333333333e-05,
|
| 67 |
+
"loss": 0.3856,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
"step": 10000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 0.07,
|
| 72 |
+
"learning_rate": 3.666666666666666e-05,
|
| 73 |
+
"loss": 0.3801,
|
| 74 |
"step": 11000
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.07,
|
| 78 |
+
"learning_rate": 3.9999999999999996e-05,
|
| 79 |
+
"loss": 0.3758,
|
| 80 |
"step": 12000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.08,
|
| 84 |
+
"learning_rate": 4.333333333333333e-05,
|
| 85 |
+
"loss": 0.3712,
|
| 86 |
"step": 13000
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.08,
|
| 90 |
+
"learning_rate": 4.6666666666666665e-05,
|
| 91 |
+
"loss": 0.3671,
|
| 92 |
"step": 14000
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.09,
|
| 96 |
+
"learning_rate": 4.9999999999999996e-05,
|
| 97 |
+
"loss": 0.364,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
"step": 15000
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 0.09,
|
| 102 |
+
"learning_rate": 5.333333333333333e-05,
|
| 103 |
+
"loss": 0.3605,
|
| 104 |
"step": 16000
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 0.1,
|
| 108 |
+
"learning_rate": 5.666666666666666e-05,
|
| 109 |
+
"loss": 0.3574,
|
| 110 |
"step": 17000
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 0.11,
|
| 114 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 115 |
+
"loss": 0.3553,
|
| 116 |
"step": 18000
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 0.11,
|
| 120 |
+
"learning_rate": 6.333333333333333e-05,
|
| 121 |
+
"loss": 0.3516,
|
| 122 |
"step": 19000
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.12,
|
| 126 |
+
"learning_rate": 6.666666666666666e-05,
|
| 127 |
+
"loss": 0.3494,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 128 |
"step": 20000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.12,
|
| 132 |
+
"learning_rate": 7e-05,
|
| 133 |
+
"loss": 0.346,
|
| 134 |
"step": 21000
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.13,
|
| 138 |
+
"learning_rate": 7.333333333333332e-05,
|
| 139 |
+
"loss": 0.3432,
|
| 140 |
"step": 22000
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 0.14,
|
| 144 |
+
"learning_rate": 7.666666666666666e-05,
|
| 145 |
+
"loss": 0.3415,
|
| 146 |
"step": 23000
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 0.14,
|
| 150 |
+
"learning_rate": 7.999999999999999e-05,
|
| 151 |
+
"loss": 0.3385,
|
| 152 |
"step": 24000
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.15,
|
| 156 |
+
"learning_rate": 8.333333333333333e-05,
|
| 157 |
+
"loss": 0.3367,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
"step": 25000
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.15,
|
| 162 |
+
"learning_rate": 8.666666666666665e-05,
|
| 163 |
+
"loss": 0.3344,
|
| 164 |
"step": 26000
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.16,
|
| 168 |
+
"learning_rate": 8.999999999999999e-05,
|
| 169 |
+
"loss": 0.3322,
|
| 170 |
"step": 27000
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.17,
|
| 174 |
+
"learning_rate": 9.333333333333333e-05,
|
| 175 |
+
"loss": 0.331,
|
| 176 |
"step": 28000
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.17,
|
| 180 |
+
"learning_rate": 9.666666666666667e-05,
|
| 181 |
+
"loss": 0.3287,
|
| 182 |
"step": 29000
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.18,
|
| 186 |
+
"learning_rate": 9.999999999999999e-05,
|
| 187 |
+
"loss": 0.3266,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 188 |
"step": 30000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 0.18,
|
| 192 |
+
"learning_rate": 0.00010333333333333333,
|
| 193 |
+
"loss": 0.3252,
|
| 194 |
"step": 31000
|
| 195 |
},
|
| 196 |
{
|
| 197 |
+
"epoch": 0.19,
|
| 198 |
+
"learning_rate": 0.00010666666666666667,
|
| 199 |
+
"loss": 0.3237,
|
| 200 |
"step": 32000
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 0.2,
|
| 204 |
+
"learning_rate": 0.00010999999999999998,
|
| 205 |
+
"loss": 0.3215,
|
| 206 |
"step": 33000
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.2,
|
| 210 |
+
"learning_rate": 0.00011333333333333331,
|
| 211 |
+
"loss": 0.3207,
|
| 212 |
"step": 34000
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.21,
|
| 216 |
+
"learning_rate": 0.00011666666666666665,
|
| 217 |
+
"loss": 0.3197,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 218 |
"step": 35000
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.21,
|
| 222 |
+
"learning_rate": 0.00011999999999999999,
|
| 223 |
+
"loss": 0.3181,
|
| 224 |
"step": 36000
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.22,
|
| 228 |
+
"learning_rate": 0.0001233333333333333,
|
| 229 |
+
"loss": 0.3178,
|
| 230 |
"step": 37000
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 0.23,
|
| 234 |
+
"learning_rate": 0.00012666666666666666,
|
| 235 |
+
"loss": 0.3163,
|
| 236 |
"step": 38000
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 0.23,
|
| 240 |
+
"learning_rate": 0.00013,
|
| 241 |
+
"loss": 0.3154,
|
| 242 |
"step": 39000
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 0.24,
|
| 246 |
+
"learning_rate": 0.0001333333333333333,
|
| 247 |
+
"loss": 0.3146,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 248 |
"step": 40000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.24,
|
| 252 |
+
"learning_rate": 0.00013666666666666666,
|
| 253 |
+
"loss": 0.313,
|
| 254 |
"step": 41000
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.25,
|
| 258 |
+
"learning_rate": 0.00014,
|
| 259 |
+
"loss": 0.3131,
|
| 260 |
"step": 42000
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"epoch": 0.25,
|
| 264 |
+
"learning_rate": 0.00014333333333333334,
|
| 265 |
+
"loss": 0.3115,
|
| 266 |
"step": 43000
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 0.26,
|
| 270 |
+
"learning_rate": 0.00014666666666666664,
|
| 271 |
+
"loss": 0.3111,
|
| 272 |
"step": 44000
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 0.27,
|
| 276 |
+
"learning_rate": 0.00015,
|
| 277 |
+
"loss": 0.3102,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 278 |
"step": 45000
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 0.27,
|
| 282 |
+
"learning_rate": 0.00014999833143445625,
|
| 283 |
+
"loss": 0.3094,
|
| 284 |
"step": 46000
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 0.28,
|
| 288 |
+
"learning_rate": 0.000149993325817371,
|
| 289 |
+
"loss": 0.3091,
|
| 290 |
"step": 47000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.28,
|
| 294 |
+
"learning_rate": 0.00014998498338737861,
|
| 295 |
+
"loss": 0.3077,
|
| 296 |
"step": 48000
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.29,
|
| 300 |
+
"learning_rate": 0.00014997330454219024,
|
| 301 |
+
"loss": 0.3075,
|
| 302 |
"step": 49000
|
| 303 |
},
|
| 304 |
{
|
| 305 |
+
"epoch": 0.3,
|
| 306 |
+
"learning_rate": 0.0001499582898385749,
|
| 307 |
+
"loss": 0.3067,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 308 |
"step": 50000
|
| 309 |
},
|
| 310 |
{
|
| 311 |
+
"epoch": 0.3,
|
| 312 |
+
"learning_rate": 0.00014993993999233321,
|
| 313 |
+
"loss": 0.3054,
|
| 314 |
"step": 51000
|
| 315 |
},
|
| 316 |
{
|
| 317 |
+
"epoch": 0.31,
|
| 318 |
+
"learning_rate": 0.00014991825587826286,
|
| 319 |
+
"loss": 0.3048,
|
| 320 |
"step": 52000
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 0.31,
|
| 324 |
+
"learning_rate": 0.00014989323853011714,
|
| 325 |
+
"loss": 0.3043,
|
| 326 |
"step": 53000
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 0.32,
|
| 330 |
+
"learning_rate": 0.00014986488914055563,
|
| 331 |
+
"loss": 0.3036,
|
| 332 |
"step": 54000
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.33,
|
| 336 |
+
"learning_rate": 0.00014983320906108733,
|
| 337 |
+
"loss": 0.3025,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 338 |
"step": 55000
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.33,
|
| 342 |
+
"learning_rate": 0.0001497981998020062,
|
| 343 |
+
"loss": 0.3025,
|
| 344 |
"step": 56000
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"epoch": 0.34,
|
| 348 |
+
"learning_rate": 0.0001497598630323192,
|
| 349 |
+
"loss": 0.3022,
|
| 350 |
"step": 57000
|
| 351 |
},
|
| 352 |
{
|
| 353 |
+
"epoch": 0.34,
|
| 354 |
+
"learning_rate": 0.00014971820057966672,
|
| 355 |
+
"loss": 0.3016,
|
| 356 |
"step": 58000
|
| 357 |
},
|
| 358 |
{
|
| 359 |
+
"epoch": 0.35,
|
| 360 |
+
"learning_rate": 0.00014967321443023538,
|
| 361 |
+
"loss": 0.3012,
|
| 362 |
"step": 59000
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 0.36,
|
| 366 |
+
"learning_rate": 0.00014962490672866348,
|
| 367 |
+
"loss": 0.3006,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 368 |
"step": 60000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 0.36,
|
| 372 |
+
"learning_rate": 0.00014957327977793857,
|
| 373 |
+
"loss": 0.3003,
|
| 374 |
"step": 61000
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 0.37,
|
| 378 |
+
"learning_rate": 0.00014951833603928788,
|
| 379 |
+
"loss": 0.3008,
|
| 380 |
"step": 62000
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.37,
|
| 384 |
+
"learning_rate": 0.0001494600781320608,
|
| 385 |
+
"loss": 0.2995,
|
| 386 |
"step": 63000
|
| 387 |
},
|
| 388 |
{
|
| 389 |
+
"epoch": 0.38,
|
| 390 |
+
"learning_rate": 0.00014939850883360407,
|
| 391 |
+
"loss": 0.2988,
|
| 392 |
"step": 64000
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.39,
|
| 396 |
+
"learning_rate": 0.0001493336310791294,
|
| 397 |
+
"loss": 0.2988,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 398 |
"step": 65000
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 0.39,
|
| 402 |
+
"learning_rate": 0.00014926544796157356,
|
| 403 |
+
"loss": 0.2985,
|
| 404 |
"step": 66000
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 0.4,
|
| 408 |
+
"learning_rate": 0.00014919396273145084,
|
| 409 |
+
"loss": 0.2979,
|
| 410 |
"step": 67000
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 0.4,
|
| 414 |
+
"learning_rate": 0.00014911917879669812,
|
| 415 |
+
"loss": 0.2967,
|
| 416 |
"step": 68000
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.41,
|
| 420 |
+
"learning_rate": 0.00014904109972251249,
|
| 421 |
+
"loss": 0.2976,
|
| 422 |
"step": 69000
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.41,
|
| 426 |
+
"learning_rate": 0.00014895972923118116,
|
| 427 |
+
"loss": 0.2972,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 428 |
"step": 70000
|
| 429 |
},
|
| 430 |
{
|
| 431 |
+
"epoch": 0.42,
|
| 432 |
+
"learning_rate": 0.00014887507120190407,
|
| 433 |
+
"loss": 0.2965,
|
| 434 |
"step": 71000
|
| 435 |
},
|
| 436 |
{
|
| 437 |
+
"epoch": 0.43,
|
| 438 |
+
"learning_rate": 0.00014878712967060896,
|
| 439 |
+
"loss": 0.2956,
|
| 440 |
"step": 72000
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 0.43,
|
| 444 |
+
"learning_rate": 0.00014869590882975894,
|
| 445 |
+
"loss": 0.2954,
|
| 446 |
"step": 73000
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 0.44,
|
| 450 |
+
"learning_rate": 0.00014860141302815264,
|
| 451 |
+
"loss": 0.2952,
|
| 452 |
"step": 74000
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.44,
|
| 456 |
+
"learning_rate": 0.00014850364677071684,
|
| 457 |
+
"loss": 0.2955,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 458 |
"step": 75000
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.45,
|
| 462 |
+
"learning_rate": 0.00014840261471829187,
|
| 463 |
+
"loss": 0.2948,
|
| 464 |
"step": 76000
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.46,
|
| 468 |
+
"learning_rate": 0.00014829832168740913,
|
| 469 |
+
"loss": 0.2945,
|
| 470 |
"step": 77000
|
| 471 |
},
|
| 472 |
{
|
| 473 |
+
"epoch": 0.46,
|
| 474 |
+
"learning_rate": 0.0001481907726500617,
|
| 475 |
+
"loss": 0.2938,
|
| 476 |
"step": 78000
|
| 477 |
},
|
| 478 |
{
|
| 479 |
+
"epoch": 0.47,
|
| 480 |
+
"learning_rate": 0.00014807997273346728,
|
| 481 |
+
"loss": 0.2936,
|
| 482 |
"step": 79000
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 0.47,
|
| 486 |
+
"learning_rate": 0.00014796592721982364,
|
| 487 |
+
"loss": 0.2932,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 488 |
"step": 80000
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 0.48,
|
| 492 |
+
"learning_rate": 0.0001478486415460569,
|
| 493 |
+
"loss": 0.2938,
|
| 494 |
"step": 81000
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 0.49,
|
| 498 |
+
"learning_rate": 0.00014772812130356235,
|
| 499 |
+
"loss": 0.2931,
|
| 500 |
"step": 82000
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.49,
|
| 504 |
+
"learning_rate": 0.00014760437223793778,
|
| 505 |
+
"loss": 0.2933,
|
| 506 |
"step": 83000
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.5,
|
| 510 |
+
"learning_rate": 0.00014747740024870972,
|
| 511 |
+
"loss": 0.2924,
|
| 512 |
"step": 84000
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.5,
|
| 516 |
+
"learning_rate": 0.00014734721138905203,
|
| 517 |
+
"loss": 0.2915,
|
| 518 |
"step": 85000
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.51,
|
| 522 |
+
"learning_rate": 0.00014721381186549743,
|
| 523 |
+
"loss": 0.2916,
|
| 524 |
+
"step": 86000
|
|
|
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.52,
|
| 528 |
+
"learning_rate": 0.00014707720803764163,
|
| 529 |
+
"loss": 0.2921,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 530 |
"step": 87000
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.52,
|
| 534 |
+
"learning_rate": 0.00014693740641784,
|
| 535 |
+
"loss": 0.2914,
|
| 536 |
"step": 88000
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 0.53,
|
| 540 |
+
"learning_rate": 0.00014679441367089737,
|
| 541 |
+
"loss": 0.292,
|
| 542 |
"step": 89000
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.53,
|
| 546 |
+
"learning_rate": 0.00014664823661374998,
|
| 547 |
+
"loss": 0.2909,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 548 |
"step": 90000
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.54,
|
| 552 |
+
"learning_rate": 0.0001464988822151407,
|
| 553 |
+
"loss": 0.2913,
|
| 554 |
"step": 91000
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 0.55,
|
| 558 |
+
"learning_rate": 0.00014634635759528682,
|
| 559 |
+
"loss": 0.2905,
|
| 560 |
"step": 92000
|
| 561 |
},
|
| 562 |
{
|
| 563 |
+
"epoch": 0.55,
|
| 564 |
+
"learning_rate": 0.00014619067002554046,
|
| 565 |
+
"loss": 0.2911,
|
| 566 |
"step": 93000
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 0.56,
|
| 570 |
+
"learning_rate": 0.000146031826928042,
|
| 571 |
+
"loss": 0.2901,
|
| 572 |
"step": 94000
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 0.56,
|
| 576 |
+
"learning_rate": 0.00014586983587536634,
|
| 577 |
+
"loss": 0.2902,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 578 |
"step": 95000
|
| 579 |
},
|
| 580 |
{
|
| 581 |
+
"epoch": 0.57,
|
| 582 |
+
"learning_rate": 0.0001457047045901617,
|
| 583 |
+
"loss": 0.2897,
|
| 584 |
"step": 96000
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 0.57,
|
| 588 |
+
"learning_rate": 0.0001455364409447816,
|
| 589 |
+
"loss": 0.29,
|
| 590 |
"step": 97000
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.58,
|
| 594 |
+
"learning_rate": 0.00014536505296090948,
|
| 595 |
+
"loss": 0.2891,
|
| 596 |
"step": 98000
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 0.59,
|
| 600 |
+
"learning_rate": 0.00014519054880917634,
|
| 601 |
+
"loss": 0.2896,
|
| 602 |
"step": 99000
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 0.59,
|
| 606 |
+
"learning_rate": 0.00014501293680877115,
|
| 607 |
+
"loss": 0.2887,
|
| 608 |
"step": 100000
|
| 609 |
},
|
| 610 |
{
|
| 611 |
+
"epoch": 0.6,
|
| 612 |
+
"learning_rate": 0.00014483222542704434,
|
| 613 |
+
"loss": 0.2891,
|
| 614 |
+
"step": 101000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.6,
|
| 618 |
+
"learning_rate": 0.00014464842327910406,
|
| 619 |
+
"loss": 0.2892,
|
| 620 |
+
"step": 102000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.61,
|
| 624 |
+
"learning_rate": 0.00014446153912740542,
|
| 625 |
+
"loss": 0.2885,
|
| 626 |
+
"step": 103000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.62,
|
| 630 |
+
"learning_rate": 0.00014427158188133295,
|
| 631 |
+
"loss": 0.2888,
|
| 632 |
+
"step": 104000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.62,
|
| 636 |
+
"learning_rate": 0.00014407856059677564,
|
| 637 |
+
"loss": 0.288,
|
| 638 |
+
"step": 105000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.63,
|
| 642 |
+
"learning_rate": 0.0001438824844756953,
|
| 643 |
+
"loss": 0.2877,
|
| 644 |
+
"step": 106000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.63,
|
| 648 |
+
"learning_rate": 0.00014368336286568802,
|
| 649 |
+
"loss": 0.2875,
|
| 650 |
+
"step": 107000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.64,
|
| 654 |
+
"learning_rate": 0.00014348120525953817,
|
| 655 |
+
"loss": 0.2879,
|
| 656 |
+
"step": 108000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.65,
|
| 660 |
+
"learning_rate": 0.0001432760212947663,
|
| 661 |
+
"loss": 0.287,
|
| 662 |
+
"step": 109000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.65,
|
| 666 |
+
"learning_rate": 0.00014306782075316932,
|
| 667 |
+
"loss": 0.2875,
|
| 668 |
+
"step": 110000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.66,
|
| 672 |
+
"learning_rate": 0.00014285661356035442,
|
| 673 |
+
"loss": 0.287,
|
| 674 |
+
"step": 111000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.66,
|
| 678 |
+
"learning_rate": 0.0001426424097852657,
|
| 679 |
+
"loss": 0.2868,
|
| 680 |
+
"step": 112000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.67,
|
| 684 |
+
"learning_rate": 0.00014242521963970427,
|
| 685 |
+
"loss": 0.2868,
|
| 686 |
+
"step": 113000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.68,
|
| 690 |
+
"learning_rate": 0.00014220505347784133,
|
| 691 |
+
"loss": 0.287,
|
| 692 |
+
"step": 114000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.68,
|
| 696 |
+
"learning_rate": 0.00014198192179572467,
|
| 697 |
+
"loss": 0.2863,
|
| 698 |
+
"step": 115000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.69,
|
| 702 |
+
"learning_rate": 0.00014175583523077823,
|
| 703 |
+
"loss": 0.2867,
|
| 704 |
+
"step": 116000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.69,
|
| 708 |
+
"learning_rate": 0.00014152680456129484,
|
| 709 |
+
"loss": 0.2864,
|
| 710 |
+
"step": 117000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.7,
|
| 714 |
+
"learning_rate": 0.0001412948407059227,
|
| 715 |
+
"loss": 0.2861,
|
| 716 |
+
"step": 118000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.71,
|
| 720 |
+
"learning_rate": 0.00014105995472314447,
|
| 721 |
+
"loss": 0.2865,
|
| 722 |
+
"step": 119000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.71,
|
| 726 |
+
"learning_rate": 0.00014082215781075044,
|
| 727 |
+
"loss": 0.2857,
|
| 728 |
+
"step": 120000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.72,
|
| 732 |
+
"learning_rate": 0.00014058146130530442,
|
| 733 |
+
"loss": 0.2857,
|
| 734 |
+
"step": 121000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.72,
|
| 738 |
+
"learning_rate": 0.00014033787668160343,
|
| 739 |
+
"loss": 0.2852,
|
| 740 |
+
"step": 122000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.73,
|
| 744 |
+
"learning_rate": 0.0001400914155521306,
|
| 745 |
+
"loss": 0.2851,
|
| 746 |
+
"step": 123000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.73,
|
| 750 |
+
"learning_rate": 0.00013984208966650165,
|
| 751 |
+
"loss": 0.2856,
|
| 752 |
+
"step": 124000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.74,
|
| 756 |
+
"learning_rate": 0.00013958991091090456,
|
| 757 |
+
"loss": 0.2849,
|
| 758 |
+
"step": 125000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.75,
|
| 762 |
+
"learning_rate": 0.00013933489130753317,
|
| 763 |
+
"loss": 0.2845,
|
| 764 |
+
"step": 126000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.75,
|
| 768 |
+
"learning_rate": 0.00013907704301401386,
|
| 769 |
+
"loss": 0.2844,
|
| 770 |
+
"step": 127000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.76,
|
| 774 |
+
"learning_rate": 0.00013881637832282597,
|
| 775 |
+
"loss": 0.2852,
|
| 776 |
+
"step": 128000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.76,
|
| 780 |
+
"learning_rate": 0.00013855290966071585,
|
| 781 |
+
"loss": 0.2841,
|
| 782 |
+
"step": 129000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.77,
|
| 786 |
+
"learning_rate": 0.00013828664958810441,
|
| 787 |
+
"loss": 0.2843,
|
| 788 |
+
"step": 130000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.78,
|
| 792 |
+
"learning_rate": 0.00013801761079848836,
|
| 793 |
+
"loss": 0.2834,
|
| 794 |
+
"step": 131000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.78,
|
| 798 |
+
"learning_rate": 0.00013774580611783485,
|
| 799 |
+
"loss": 0.2842,
|
| 800 |
+
"step": 132000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.79,
|
| 804 |
+
"learning_rate": 0.00013747124850397037,
|
| 805 |
+
"loss": 0.2832,
|
| 806 |
+
"step": 133000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.79,
|
| 810 |
+
"learning_rate": 0.00013719395104596277,
|
| 811 |
+
"loss": 0.2836,
|
| 812 |
+
"step": 134000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.8,
|
| 816 |
+
"learning_rate": 0.00013691392696349726,
|
| 817 |
+
"loss": 0.2838,
|
| 818 |
+
"step": 135000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.81,
|
| 822 |
+
"learning_rate": 0.0001366311896062463,
|
| 823 |
+
"loss": 0.2835,
|
| 824 |
+
"step": 136000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.81,
|
| 828 |
+
"learning_rate": 0.00013634575245323311,
|
| 829 |
+
"loss": 0.2835,
|
| 830 |
+
"step": 137000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.82,
|
| 834 |
+
"learning_rate": 0.00013605762911218905,
|
| 835 |
+
"loss": 0.2831,
|
| 836 |
+
"step": 138000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.82,
|
| 840 |
+
"learning_rate": 0.00013576683331890498,
|
| 841 |
+
"loss": 0.2829,
|
| 842 |
+
"step": 139000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.83,
|
| 846 |
+
"learning_rate": 0.0001354733789365763,
|
| 847 |
+
"loss": 0.2825,
|
| 848 |
+
"step": 140000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.84,
|
| 852 |
+
"learning_rate": 0.00013517727995514226,
|
| 853 |
+
"loss": 0.2828,
|
| 854 |
+
"step": 141000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.84,
|
| 858 |
+
"learning_rate": 0.00013487855049061875,
|
| 859 |
+
"loss": 0.2826,
|
| 860 |
+
"step": 142000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.85,
|
| 864 |
+
"learning_rate": 0.00013457720478442554,
|
| 865 |
+
"loss": 0.2826,
|
| 866 |
+
"step": 143000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.85,
|
| 870 |
+
"learning_rate": 0.00013427325720270729,
|
| 871 |
+
"loss": 0.2824,
|
| 872 |
+
"step": 144000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.86,
|
| 876 |
+
"learning_rate": 0.00013396672223564855,
|
| 877 |
+
"loss": 0.2822,
|
| 878 |
+
"step": 145000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.87,
|
| 882 |
+
"learning_rate": 0.00013365761449678324,
|
| 883 |
+
"loss": 0.2819,
|
| 884 |
+
"step": 146000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.87,
|
| 888 |
+
"learning_rate": 0.00013334594872229763,
|
| 889 |
+
"loss": 0.282,
|
| 890 |
+
"step": 147000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.88,
|
| 894 |
+
"learning_rate": 0.00013303173977032815,
|
| 895 |
+
"loss": 0.2815,
|
| 896 |
+
"step": 148000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.88,
|
| 900 |
+
"learning_rate": 0.00013271500262025278,
|
| 901 |
+
"loss": 0.2815,
|
| 902 |
+
"step": 149000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.89,
|
| 906 |
+
"learning_rate": 0.00013239575237197705,
|
| 907 |
+
"loss": 0.2818,
|
| 908 |
+
"step": 150000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.89,
|
| 912 |
+
"learning_rate": 0.00013207400424521432,
|
| 913 |
+
"loss": 0.2818,
|
| 914 |
+
"step": 151000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.9,
|
| 918 |
+
"learning_rate": 0.0001317497735787599,
|
| 919 |
+
"loss": 0.2817,
|
| 920 |
+
"step": 152000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.91,
|
| 924 |
+
"learning_rate": 0.0001314230758297601,
|
| 925 |
+
"loss": 0.2817,
|
| 926 |
+
"step": 153000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.91,
|
| 930 |
+
"learning_rate": 0.000131093926572975,
|
| 931 |
+
"loss": 0.2809,
|
| 932 |
+
"step": 154000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.92,
|
| 936 |
+
"learning_rate": 0.00013076234150003634,
|
| 937 |
+
"loss": 0.281,
|
| 938 |
+
"step": 155000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.92,
|
| 942 |
+
"learning_rate": 0.00013042833641869922,
|
| 943 |
+
"loss": 0.2816,
|
| 944 |
+
"step": 156000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.93,
|
| 948 |
+
"learning_rate": 0.00013009192725208846,
|
| 949 |
+
"loss": 0.2812,
|
| 950 |
+
"step": 157000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.94,
|
| 954 |
+
"learning_rate": 0.00012975313003793962,
|
| 955 |
+
"loss": 0.2814,
|
| 956 |
+
"step": 158000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.94,
|
| 960 |
+
"learning_rate": 0.00012941196092783439,
|
| 961 |
+
"loss": 0.2803,
|
| 962 |
+
"step": 159000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.95,
|
| 966 |
+
"learning_rate": 0.00012906843618643046,
|
| 967 |
+
"loss": 0.2813,
|
| 968 |
+
"step": 160000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.95,
|
| 972 |
+
"learning_rate": 0.00012872257219068633,
|
| 973 |
+
"loss": 0.2809,
|
| 974 |
+
"step": 161000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.96,
|
| 978 |
+
"learning_rate": 0.00012837438542908053,
|
| 979 |
+
"loss": 0.2802,
|
| 980 |
+
"step": 162000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.97,
|
| 984 |
+
"learning_rate": 0.00012802389250082536,
|
| 985 |
+
"loss": 0.2806,
|
| 986 |
+
"step": 163000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.97,
|
| 990 |
+
"learning_rate": 0.0001276711101150758,
|
| 991 |
+
"loss": 0.2806,
|
| 992 |
+
"step": 164000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.98,
|
| 996 |
+
"learning_rate": 0.0001273160550901328,
|
| 997 |
+
"loss": 0.2801,
|
| 998 |
+
"step": 165000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.98,
|
| 1002 |
+
"learning_rate": 0.00012695874435264155,
|
| 1003 |
+
"loss": 0.2797,
|
| 1004 |
+
"step": 166000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.99,
|
| 1008 |
+
"learning_rate": 0.00012659919493678443,
|
| 1009 |
+
"loss": 0.2801,
|
| 1010 |
+
"step": 167000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.0,
|
| 1014 |
+
"learning_rate": 0.0001262374239834691,
|
| 1015 |
+
"loss": 0.28,
|
| 1016 |
+
"step": 168000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.0,
|
| 1020 |
+
"learning_rate": 0.00012587344873951116,
|
| 1021 |
+
"loss": 0.2797,
|
| 1022 |
+
"step": 169000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.01,
|
| 1026 |
+
"learning_rate": 0.0001255072865568121,
|
| 1027 |
+
"loss": 0.2798,
|
| 1028 |
+
"step": 170000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.01,
|
| 1032 |
+
"learning_rate": 0.00012513895489153193,
|
| 1033 |
+
"loss": 0.2798,
|
| 1034 |
+
"step": 171000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.02,
|
| 1038 |
+
"learning_rate": 0.00012476847130325712,
|
| 1039 |
+
"loss": 0.2792,
|
| 1040 |
+
"step": 172000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.03,
|
| 1044 |
+
"learning_rate": 0.0001243958534541634,
|
| 1045 |
+
"loss": 0.2795,
|
| 1046 |
+
"step": 173000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.03,
|
| 1050 |
+
"learning_rate": 0.00012402111910817367,
|
| 1051 |
+
"loss": 0.2796,
|
| 1052 |
+
"step": 174000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.04,
|
| 1056 |
+
"learning_rate": 0.00012364428613011134,
|
| 1057 |
+
"loss": 0.2785,
|
| 1058 |
+
"step": 175000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.04,
|
| 1062 |
+
"learning_rate": 0.0001232653724848485,
|
| 1063 |
+
"loss": 0.2793,
|
| 1064 |
+
"step": 176000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.05,
|
| 1068 |
+
"learning_rate": 0.0001228843962364495,
|
| 1069 |
+
"loss": 0.2787,
|
| 1070 |
+
"step": 177000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.05,
|
| 1074 |
+
"learning_rate": 0.0001225013755473098,
|
| 1075 |
+
"loss": 0.2797,
|
| 1076 |
+
"step": 178000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.06,
|
| 1080 |
+
"learning_rate": 0.00012211632867729008,
|
| 1081 |
+
"loss": 0.2791,
|
| 1082 |
+
"step": 179000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.07,
|
| 1086 |
+
"learning_rate": 0.00012172927398284578,
|
| 1087 |
+
"loss": 0.2787,
|
| 1088 |
+
"step": 180000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.07,
|
| 1092 |
+
"learning_rate": 0.00012134022991615195,
|
| 1093 |
+
"loss": 0.279,
|
| 1094 |
+
"step": 181000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.08,
|
| 1098 |
+
"learning_rate": 0.00012094921502422359,
|
| 1099 |
+
"loss": 0.2788,
|
| 1100 |
+
"step": 182000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 1.08,
|
| 1104 |
+
"learning_rate": 0.00012055624794803143,
|
| 1105 |
+
"loss": 0.2787,
|
| 1106 |
+
"step": 183000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.09,
|
| 1110 |
+
"learning_rate": 0.00012016134742161328,
|
| 1111 |
+
"loss": 0.2789,
|
| 1112 |
+
"step": 184000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.1,
|
| 1116 |
+
"learning_rate": 0.0001197645322711809,
|
| 1117 |
+
"loss": 0.2783,
|
| 1118 |
+
"step": 185000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.1,
|
| 1122 |
+
"learning_rate": 0.00011936582141422247,
|
| 1123 |
+
"loss": 0.2782,
|
| 1124 |
+
"step": 186000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.11,
|
| 1128 |
+
"learning_rate": 0.00011896523385860084,
|
| 1129 |
+
"loss": 0.2783,
|
| 1130 |
+
"step": 187000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.11,
|
| 1134 |
+
"learning_rate": 0.00011856278870164717,
|
| 1135 |
+
"loss": 0.2775,
|
| 1136 |
+
"step": 188000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.12,
|
| 1140 |
+
"learning_rate": 0.00011815850512925069,
|
| 1141 |
+
"loss": 0.2779,
|
| 1142 |
+
"step": 189000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 1.13,
|
| 1146 |
+
"learning_rate": 0.00011775240241494394,
|
| 1147 |
+
"loss": 0.2777,
|
| 1148 |
+
"step": 190000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 1.13,
|
| 1152 |
+
"learning_rate": 0.00011734449991898394,
|
| 1153 |
+
"loss": 0.2772,
|
| 1154 |
+
"step": 191000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.14,
|
| 1158 |
+
"learning_rate": 0.00011693481708742928,
|
| 1159 |
+
"loss": 0.2778,
|
| 1160 |
+
"step": 192000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.14,
|
| 1164 |
+
"learning_rate": 0.000116523373451213,
|
| 1165 |
+
"loss": 0.2773,
|
| 1166 |
+
"step": 193000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.15,
|
| 1170 |
+
"learning_rate": 0.0001161101886252115,
|
| 1171 |
+
"loss": 0.2779,
|
| 1172 |
+
"step": 194000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.16,
|
| 1176 |
+
"learning_rate": 0.00011569528230730951,
|
| 1177 |
+
"loss": 0.2778,
|
| 1178 |
+
"step": 195000
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.16,
|
| 1182 |
+
"learning_rate": 0.00011527867427746085,
|
| 1183 |
+
"loss": 0.2777,
|
| 1184 |
+
"step": 196000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 1.17,
|
| 1188 |
+
"learning_rate": 0.00011486038439674568,
|
| 1189 |
+
"loss": 0.2772,
|
| 1190 |
+
"step": 197000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 1.17,
|
| 1194 |
+
"learning_rate": 0.00011444043260642344,
|
| 1195 |
+
"loss": 0.2773,
|
| 1196 |
+
"step": 198000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.18,
|
| 1200 |
+
"learning_rate": 0.00011401883892698241,
|
| 1201 |
+
"loss": 0.2775,
|
| 1202 |
+
"step": 199000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 1.19,
|
| 1206 |
+
"learning_rate": 0.00011359562345718505,
|
| 1207 |
+
"loss": 0.2766,
|
| 1208 |
+
"step": 200000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.19,
|
| 1212 |
+
"learning_rate": 0.00011317080637310989,
|
| 1213 |
+
"loss": 0.2763,
|
| 1214 |
+
"step": 201000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.2,
|
| 1218 |
+
"learning_rate": 0.00011274440792718983,
|
| 1219 |
+
"loss": 0.2765,
|
| 1220 |
+
"step": 202000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.2,
|
| 1224 |
+
"learning_rate": 0.0001123164484472464,
|
| 1225 |
+
"loss": 0.2768,
|
| 1226 |
+
"step": 203000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.21,
|
| 1230 |
+
"learning_rate": 0.00011188694833552081,
|
| 1231 |
+
"loss": 0.276,
|
| 1232 |
+
"step": 204000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1.21,
|
| 1236 |
+
"learning_rate": 0.00011145592806770127,
|
| 1237 |
+
"loss": 0.2765,
|
| 1238 |
+
"step": 205000
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.22,
|
| 1242 |
+
"learning_rate": 0.00011102340819194689,
|
| 1243 |
+
"loss": 0.2763,
|
| 1244 |
+
"step": 206000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 1.23,
|
| 1248 |
+
"learning_rate": 0.00011058940932790795,
|
| 1249 |
+
"loss": 0.2763,
|
| 1250 |
+
"step": 207000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.23,
|
| 1254 |
+
"learning_rate": 0.00011015395216574308,
|
| 1255 |
+
"loss": 0.2764,
|
| 1256 |
+
"step": 208000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.24,
|
| 1260 |
+
"learning_rate": 0.00010971705746513272,
|
| 1261 |
+
"loss": 0.276,
|
| 1262 |
+
"step": 209000
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.24,
|
| 1266 |
+
"learning_rate": 0.00010927874605428955,
|
| 1267 |
+
"loss": 0.2759,
|
| 1268 |
+
"step": 210000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 1.25,
|
| 1272 |
+
"learning_rate": 0.00010883903882896555,
|
| 1273 |
+
"loss": 0.276,
|
| 1274 |
+
"step": 211000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.26,
|
| 1278 |
+
"learning_rate": 0.00010839795675145566,
|
| 1279 |
+
"loss": 0.2756,
|
| 1280 |
+
"step": 212000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.26,
|
| 1284 |
+
"learning_rate": 0.00010795552084959867,
|
| 1285 |
+
"loss": 0.2756,
|
| 1286 |
+
"step": 213000
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1.27,
|
| 1290 |
+
"learning_rate": 0.00010751175221577465,
|
| 1291 |
+
"loss": 0.2753,
|
| 1292 |
+
"step": 214000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.27,
|
| 1296 |
+
"learning_rate": 0.00010706667200589931,
|
| 1297 |
+
"loss": 0.2762,
|
| 1298 |
+
"step": 215000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.28,
|
| 1302 |
+
"learning_rate": 0.00010662030143841562,
|
| 1303 |
+
"loss": 0.2754,
|
| 1304 |
+
"step": 216000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.29,
|
| 1308 |
+
"learning_rate": 0.00010617266179328208,
|
| 1309 |
+
"loss": 0.2755,
|
| 1310 |
+
"step": 217000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 1.29,
|
| 1314 |
+
"learning_rate": 0.00010572377441095834,
|
| 1315 |
+
"loss": 0.2759,
|
| 1316 |
+
"step": 218000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 1.3,
|
| 1320 |
+
"learning_rate": 0.00010527366069138784,
|
| 1321 |
+
"loss": 0.2754,
|
| 1322 |
+
"step": 219000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.3,
|
| 1326 |
+
"learning_rate": 0.00010482234209297747,
|
| 1327 |
+
"loss": 0.2756,
|
| 1328 |
+
"step": 220000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 1.31,
|
| 1332 |
+
"learning_rate": 0.0001043698401315748,
|
| 1333 |
+
"loss": 0.2752,
|
| 1334 |
+
"step": 221000
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.32,
|
| 1338 |
+
"learning_rate": 0.00010391617637944215,
|
| 1339 |
+
"loss": 0.2751,
|
| 1340 |
+
"step": 222000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.32,
|
| 1344 |
+
"learning_rate": 0.00010346137246422819,
|
| 1345 |
+
"loss": 0.2745,
|
| 1346 |
+
"step": 223000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.33,
|
| 1350 |
+
"learning_rate": 0.00010300545006793699,
|
| 1351 |
+
"loss": 0.2747,
|
| 1352 |
+
"step": 224000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 1.33,
|
| 1356 |
+
"learning_rate": 0.00010254843092589434,
|
| 1357 |
+
"loss": 0.2746,
|
| 1358 |
+
"step": 225000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 1.34,
|
| 1362 |
+
"learning_rate": 0.00010209033682571144,
|
| 1363 |
+
"loss": 0.2749,
|
| 1364 |
+
"step": 226000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.35,
|
| 1368 |
+
"learning_rate": 0.00010163118960624632,
|
| 1369 |
+
"loss": 0.2739,
|
| 1370 |
+
"step": 227000
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1.35,
|
| 1374 |
+
"learning_rate": 0.0001011710111565627,
|
| 1375 |
+
"loss": 0.2746,
|
| 1376 |
+
"step": 228000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.36,
|
| 1380 |
+
"learning_rate": 0.00010070982341488645,
|
| 1381 |
+
"loss": 0.2744,
|
| 1382 |
+
"step": 229000
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.36,
|
| 1386 |
+
"learning_rate": 0.00010024764836755973,
|
| 1387 |
+
"loss": 0.2742,
|
| 1388 |
+
"step": 230000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.37,
|
| 1392 |
+
"learning_rate": 9.978450804799277e-05,
|
| 1393 |
+
"loss": 0.2737,
|
| 1394 |
+
"step": 231000
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 1.37,
|
| 1398 |
+
"learning_rate": 9.932042453561354e-05,
|
| 1399 |
+
"loss": 0.2747,
|
| 1400 |
+
"step": 232000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 1.38,
|
| 1404 |
+
"learning_rate": 9.885541995481517e-05,
|
| 1405 |
+
"loss": 0.2741,
|
| 1406 |
+
"step": 233000
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.39,
|
| 1410 |
+
"learning_rate": 9.838951647390106e-05,
|
| 1411 |
+
"loss": 0.2747,
|
| 1412 |
+
"step": 234000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.39,
|
| 1416 |
+
"learning_rate": 9.792273630402829e-05,
|
| 1417 |
+
"loss": 0.2739,
|
| 1418 |
+
"step": 235000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.4,
|
| 1422 |
+
"learning_rate": 9.745510169814845e-05,
|
| 1423 |
+
"loss": 0.2748,
|
| 1424 |
+
"step": 236000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.4,
|
| 1428 |
+
"learning_rate": 9.698663494994703e-05,
|
| 1429 |
+
"loss": 0.2736,
|
| 1430 |
+
"step": 237000
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.41,
|
| 1434 |
+
"learning_rate": 9.65173583927804e-05,
|
| 1435 |
+
"loss": 0.2732,
|
| 1436 |
+
"step": 238000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 1.42,
|
| 1440 |
+
"learning_rate": 9.604729439861125e-05,
|
| 1441 |
+
"loss": 0.274,
|
| 1442 |
+
"step": 239000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 1.42,
|
| 1446 |
+
"learning_rate": 9.5576465376942e-05,
|
| 1447 |
+
"loss": 0.2739,
|
| 1448 |
+
"step": 240000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.43,
|
| 1452 |
+
"learning_rate": 9.510489377374642e-05,
|
| 1453 |
+
"loss": 0.2723,
|
| 1454 |
+
"step": 241000
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 1.43,
|
| 1458 |
+
"learning_rate": 9.463260207039961e-05,
|
| 1459 |
+
"loss": 0.2732,
|
| 1460 |
+
"step": 242000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.44,
|
| 1464 |
+
"learning_rate": 9.415961278260625e-05,
|
| 1465 |
+
"loss": 0.2742,
|
| 1466 |
+
"step": 243000
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.45,
|
| 1470 |
+
"learning_rate": 9.36859484593271e-05,
|
| 1471 |
+
"loss": 0.2729,
|
| 1472 |
+
"step": 244000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.45,
|
| 1476 |
+
"learning_rate": 9.321163168170416e-05,
|
| 1477 |
+
"loss": 0.2732,
|
| 1478 |
+
"step": 245000
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 1.46,
|
| 1482 |
+
"learning_rate": 9.273668506198407e-05,
|
| 1483 |
+
"loss": 0.2728,
|
| 1484 |
+
"step": 246000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 1.46,
|
| 1488 |
+
"learning_rate": 9.226113124244006e-05,
|
| 1489 |
+
"loss": 0.2726,
|
| 1490 |
+
"step": 247000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.47,
|
| 1494 |
+
"learning_rate": 9.178499289429266e-05,
|
| 1495 |
+
"loss": 0.273,
|
| 1496 |
+
"step": 248000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.48,
|
| 1500 |
+
"learning_rate": 9.130829271662884e-05,
|
| 1501 |
+
"loss": 0.2731,
|
| 1502 |
+
"step": 249000
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.48,
|
| 1506 |
+
"learning_rate": 9.083105343531967e-05,
|
| 1507 |
+
"loss": 0.2726,
|
| 1508 |
+
"step": 250000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.49,
|
| 1512 |
+
"learning_rate": 9.035329780193725e-05,
|
| 1513 |
+
"loss": 0.2728,
|
| 1514 |
+
"step": 251000
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.49,
|
| 1518 |
+
"learning_rate": 8.987504859266978e-05,
|
| 1519 |
+
"loss": 0.2727,
|
| 1520 |
+
"step": 252000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 1.5,
|
| 1524 |
+
"learning_rate": 8.939632860723587e-05,
|
| 1525 |
+
"loss": 0.272,
|
| 1526 |
+
"step": 253000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1.51,
|
| 1530 |
+
"learning_rate": 8.891716066779766e-05,
|
| 1531 |
+
"loss": 0.2726,
|
| 1532 |
+
"step": 254000
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.51,
|
| 1536 |
+
"learning_rate": 8.843756761787259e-05,
|
| 1537 |
+
"loss": 0.272,
|
| 1538 |
+
"step": 255000
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.52,
|
| 1542 |
+
"learning_rate": 8.79575723212447e-05,
|
| 1543 |
+
"loss": 0.2716,
|
| 1544 |
+
"step": 256000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.52,
|
| 1548 |
+
"learning_rate": 8.747719766087431e-05,
|
| 1549 |
+
"loss": 0.272,
|
| 1550 |
+
"step": 257000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.53,
|
| 1554 |
+
"learning_rate": 8.69964665378073e-05,
|
| 1555 |
+
"loss": 0.2723,
|
| 1556 |
+
"step": 258000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.53,
|
| 1560 |
+
"learning_rate": 8.651540187008339e-05,
|
| 1561 |
+
"loss": 0.2719,
|
| 1562 |
+
"step": 259000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 1.54,
|
| 1566 |
+
"learning_rate": 8.603402659164335e-05,
|
| 1567 |
+
"loss": 0.2723,
|
| 1568 |
+
"step": 260000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 1.55,
|
| 1572 |
+
"learning_rate": 8.555236365123586e-05,
|
| 1573 |
+
"loss": 0.2714,
|
| 1574 |
+
"step": 261000
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.55,
|
| 1578 |
+
"learning_rate": 8.507043601132336e-05,
|
| 1579 |
+
"loss": 0.2719,
|
| 1580 |
+
"step": 262000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.56,
|
| 1584 |
+
"learning_rate": 8.458826664698743e-05,
|
| 1585 |
+
"loss": 0.2717,
|
| 1586 |
+
"step": 263000
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.56,
|
| 1590 |
+
"learning_rate": 8.410587854483341e-05,
|
| 1591 |
+
"loss": 0.2719,
|
| 1592 |
+
"step": 264000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.57,
|
| 1596 |
+
"learning_rate": 8.362329470189466e-05,
|
| 1597 |
+
"loss": 0.2715,
|
| 1598 |
+
"step": 265000
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.58,
|
| 1602 |
+
"learning_rate": 8.314053812453604e-05,
|
| 1603 |
+
"loss": 0.272,
|
| 1604 |
+
"step": 266000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 1.58,
|
| 1608 |
+
"learning_rate": 8.265763182735732e-05,
|
| 1609 |
+
"loss": 0.271,
|
| 1610 |
+
"step": 267000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.59,
|
| 1614 |
+
"learning_rate": 8.217459883209592e-05,
|
| 1615 |
+
"loss": 0.2715,
|
| 1616 |
+
"step": 268000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.59,
|
| 1620 |
+
"learning_rate": 8.169146216652924e-05,
|
| 1621 |
+
"loss": 0.2713,
|
| 1622 |
+
"step": 269000
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.6,
|
| 1626 |
+
"learning_rate": 8.12082448633772e-05,
|
| 1627 |
+
"loss": 0.2708,
|
| 1628 |
+
"step": 270000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 1.61,
|
| 1632 |
+
"learning_rate": 8.072496995920378e-05,
|
| 1633 |
+
"loss": 0.2708,
|
| 1634 |
+
"step": 271000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.61,
|
| 1638 |
+
"learning_rate": 8.024166049331909e-05,
|
| 1639 |
+
"loss": 0.2706,
|
| 1640 |
+
"step": 272000
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.62,
|
| 1644 |
+
"learning_rate": 7.975833950668091e-05,
|
| 1645 |
+
"loss": 0.2705,
|
| 1646 |
+
"step": 273000
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 1.62,
|
| 1650 |
+
"learning_rate": 7.927503004079623e-05,
|
| 1651 |
+
"loss": 0.2707,
|
| 1652 |
+
"step": 274000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 1.63,
|
| 1656 |
+
"learning_rate": 7.87917551366228e-05,
|
| 1657 |
+
"loss": 0.2705,
|
| 1658 |
+
"step": 275000
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 1.64,
|
| 1662 |
+
"learning_rate": 7.830853783347074e-05,
|
| 1663 |
+
"loss": 0.2703,
|
| 1664 |
+
"step": 276000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 1.64,
|
| 1668 |
+
"learning_rate": 7.782540116790411e-05,
|
| 1669 |
+
"loss": 0.2703,
|
| 1670 |
+
"step": 277000
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 1.65,
|
| 1674 |
+
"learning_rate": 7.734236817264267e-05,
|
| 1675 |
+
"loss": 0.2697,
|
| 1676 |
+
"step": 278000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.65,
|
| 1680 |
+
"learning_rate": 7.685946187546397e-05,
|
| 1681 |
+
"loss": 0.2699,
|
| 1682 |
+
"step": 279000
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.66,
|
| 1686 |
+
"learning_rate": 7.637670529810532e-05,
|
| 1687 |
+
"loss": 0.2703,
|
| 1688 |
+
"step": 280000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 1.67,
|
| 1692 |
+
"learning_rate": 7.589412145516656e-05,
|
| 1693 |
+
"loss": 0.2701,
|
| 1694 |
+
"step": 281000
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 1.67,
|
| 1698 |
+
"learning_rate": 7.541173335301255e-05,
|
| 1699 |
+
"loss": 0.2704,
|
| 1700 |
+
"step": 282000
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.68,
|
| 1704 |
+
"learning_rate": 7.492956398867663e-05,
|
| 1705 |
+
"loss": 0.2701,
|
| 1706 |
+
"step": 283000
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 1.68,
|
| 1710 |
+
"learning_rate": 7.444763634876413e-05,
|
| 1711 |
+
"loss": 0.2703,
|
| 1712 |
+
"step": 284000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 1.69,
|
| 1716 |
+
"learning_rate": 7.396597340835663e-05,
|
| 1717 |
+
"loss": 0.2701,
|
| 1718 |
+
"step": 285000
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.69,
|
| 1722 |
+
"learning_rate": 7.34845981299166e-05,
|
| 1723 |
+
"loss": 0.2703,
|
| 1724 |
+
"step": 286000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.7,
|
| 1728 |
+
"learning_rate": 7.300353346219268e-05,
|
| 1729 |
+
"loss": 0.2694,
|
| 1730 |
+
"step": 287000
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 1.71,
|
| 1734 |
+
"learning_rate": 7.252280233912569e-05,
|
| 1735 |
+
"loss": 0.2695,
|
| 1736 |
+
"step": 288000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 1.71,
|
| 1740 |
+
"learning_rate": 7.20424276787553e-05,
|
| 1741 |
+
"loss": 0.2694,
|
| 1742 |
+
"step": 289000
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.72,
|
| 1746 |
+
"learning_rate": 7.156243238212739e-05,
|
| 1747 |
+
"loss": 0.2701,
|
| 1748 |
+
"step": 290000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 1.72,
|
| 1752 |
+
"learning_rate": 7.108283933220234e-05,
|
| 1753 |
+
"loss": 0.2696,
|
| 1754 |
+
"step": 291000
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 1.73,
|
| 1758 |
+
"learning_rate": 7.060367139276414e-05,
|
| 1759 |
+
"loss": 0.2696,
|
| 1760 |
+
"step": 292000
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.74,
|
| 1764 |
+
"learning_rate": 7.012495140733023e-05,
|
| 1765 |
+
"loss": 0.2694,
|
| 1766 |
+
"step": 293000
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.74,
|
| 1770 |
+
"learning_rate": 6.964670219806273e-05,
|
| 1771 |
+
"loss": 0.2691,
|
| 1772 |
+
"step": 294000
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 1.75,
|
| 1776 |
+
"learning_rate": 6.91689465646803e-05,
|
| 1777 |
+
"loss": 0.2686,
|
| 1778 |
+
"step": 295000
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 1.75,
|
| 1782 |
+
"learning_rate": 6.869170728337114e-05,
|
| 1783 |
+
"loss": 0.2692,
|
| 1784 |
+
"step": 296000
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 1.76,
|
| 1788 |
+
"learning_rate": 6.82150071057073e-05,
|
| 1789 |
+
"loss": 0.2691,
|
| 1790 |
+
"step": 297000
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 1.77,
|
| 1794 |
+
"learning_rate": 6.773886875755992e-05,
|
| 1795 |
+
"loss": 0.2684,
|
| 1796 |
+
"step": 298000
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 1.77,
|
| 1800 |
+
"learning_rate": 6.726331493801593e-05,
|
| 1801 |
+
"loss": 0.2688,
|
| 1802 |
+
"step": 299000
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.78,
|
| 1806 |
+
"learning_rate": 6.678836831829584e-05,
|
| 1807 |
+
"loss": 0.2693,
|
| 1808 |
+
"step": 300000
|
| 1809 |
}
|
| 1810 |
],
|
| 1811 |
+
"max_steps": 500000,
|
| 1812 |
+
"num_train_epochs": 3,
|
| 1813 |
+
"total_flos": 1.1529473128203587e+21,
|
| 1814 |
"trial_name": null,
|
| 1815 |
"trial_params": null
|
| 1816 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba0013f502b335ff444eb9b6bf98acb3aff61765009c13640af7975297997821
|
| 3 |
+
size 3640
|