Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- checkpoint-200/adapter_model.safetensors +1 -1
- checkpoint-200/optimizer.pt +1 -1
- checkpoint-200/rng_state.pth +1 -1
- checkpoint-200/scheduler.pt +1 -1
- checkpoint-200/trainer_state.json +785 -785
- checkpoint-200/training_args.bin +1 -1
- training_args.bin +1 -1
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 34100216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c58239a67307473e58bbc09292e8bf457bccbf3d8f093413f2a9983160f1373
|
| 3 |
size 34100216
|
checkpoint-200/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 34100216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c58239a67307473e58bbc09292e8bf457bccbf3d8f093413f2a9983160f1373
|
| 3 |
size 34100216
|
checkpoint-200/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 68292346
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93a1f6aa5183962874fd62e0d358e0f0ea27e6a68a0535d07f990842e7013963
|
| 3 |
size 68292346
|
checkpoint-200/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 sha256:04745946c21cb6d9e7af44a9e436e4207c8437de1201d76c48a7771e0e0f3bad
|
| 3 |
size 14244
|
checkpoint-200/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:b5fcb2e0acbfc451125ca5cd950e0420d092eee3eeb6dd5b682202b7a14cd22e
|
| 3 |
size 1064
|
checkpoint-200/trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -9,1403 +9,1403 @@
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch":
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch":
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch":
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch":
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss":
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss":
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 1.
|
| 64 |
-
"loss":
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm": 1.
|
| 70 |
-
"learning_rate": 1.
|
| 71 |
-
"loss":
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 2.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 2.
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm": 1.
|
| 91 |
-
"learning_rate": 2.
|
| 92 |
-
"loss":
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate": 2.
|
| 99 |
-
"loss": 2.
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate": 2.
|
| 106 |
-
"loss":
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss":
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss":
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 3.
|
| 127 |
-
"loss":
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate": 3.
|
| 134 |
-
"loss": 2.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate": 3.
|
| 141 |
-
"loss": 2.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss":
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss":
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 4.
|
| 162 |
-
"loss": 2.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm": 1.
|
| 168 |
-
"learning_rate": 4.
|
| 169 |
-
"loss":
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm":
|
| 175 |
-
"learning_rate": 4.
|
| 176 |
-
"loss":
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm":
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss":
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm":
|
| 189 |
-
"learning_rate":
|
| 190 |
-
"loss": 2.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate": 5.
|
| 197 |
-
"loss":
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate": 5.
|
| 204 |
-
"loss":
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm":
|
| 210 |
-
"learning_rate": 5.
|
| 211 |
-
"loss": 2.
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm":
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss":
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate":
|
| 225 |
-
"loss":
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate": 6.
|
| 232 |
-
"loss":
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm":
|
| 238 |
-
"learning_rate": 6.
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate": 6.
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch": 0.
|
| 251 |
-
"grad_norm":
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate": 7.
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch": 0.
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 7.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"grad_norm":
|
| 280 |
-
"learning_rate": 7.
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 0.
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 0.
|
| 293 |
-
"grad_norm":
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate": 8.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0.
|
| 307 |
-
"grad_norm":
|
| 308 |
-
"learning_rate": 8.
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch": 0.
|
| 314 |
-
"grad_norm": 3.
|
| 315 |
-
"learning_rate": 8.
|
| 316 |
-
"loss":
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"grad_norm":
|
| 322 |
-
"learning_rate":
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch": 0.
|
| 328 |
-
"grad_norm": 2.
|
| 329 |
-
"learning_rate":
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch": 0.
|
| 335 |
-
"grad_norm":
|
| 336 |
-
"learning_rate": 9.
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"grad_norm":
|
| 343 |
-
"learning_rate": 9.
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 0.
|
| 349 |
-
"grad_norm":
|
| 350 |
-
"learning_rate": 9.
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch": 0.
|
| 356 |
-
"grad_norm":
|
| 357 |
-
"learning_rate":
|
| 358 |
-
"loss":
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"grad_norm":
|
| 364 |
-
"learning_rate": 0.
|
| 365 |
-
"loss":
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch": 0.
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate": 0.
|
| 372 |
-
"loss": 1.
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
-
"epoch": 0.
|
| 377 |
-
"grad_norm":
|
| 378 |
-
"learning_rate": 0.
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
-
"grad_norm":
|
| 385 |
-
"learning_rate": 0.
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 0.
|
| 391 |
-
"grad_norm":
|
| 392 |
-
"learning_rate": 0.
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch": 0.
|
| 398 |
-
"grad_norm":
|
| 399 |
-
"learning_rate": 0.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"grad_norm":
|
| 406 |
-
"learning_rate": 0.
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch": 0.
|
| 412 |
-
"grad_norm":
|
| 413 |
"learning_rate": 0.00011399999999999999,
|
| 414 |
-
"loss": 1.
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch": 0.
|
| 419 |
-
"grad_norm":
|
| 420 |
"learning_rate": 0.000116,
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch": 0.
|
| 426 |
-
"grad_norm":
|
| 427 |
"learning_rate": 0.000118,
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 0.
|
| 433 |
-
"grad_norm":
|
| 434 |
"learning_rate": 0.00012,
|
| 435 |
-
"loss": 1.
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
-
"epoch": 0.
|
| 440 |
-
"grad_norm":
|
| 441 |
"learning_rate": 0.000122,
|
| 442 |
-
"loss":
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch": 0.
|
| 447 |
-
"grad_norm":
|
| 448 |
-
"learning_rate": 0.
|
| 449 |
-
"loss":
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
-
"epoch": 0.
|
| 454 |
-
"grad_norm":
|
| 455 |
-
"learning_rate": 0.
|
| 456 |
-
"loss": 1.
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
-
"epoch": 0.
|
| 461 |
-
"grad_norm":
|
| 462 |
-
"learning_rate": 0.
|
| 463 |
-
"loss":
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"grad_norm":
|
| 469 |
-
"learning_rate": 0.
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch": 0.
|
| 475 |
-
"grad_norm":
|
| 476 |
-
"learning_rate": 0.
|
| 477 |
-
"loss":
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
-
"epoch": 0.
|
| 482 |
-
"grad_norm":
|
| 483 |
-
"learning_rate": 0.
|
| 484 |
-
"loss":
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
-
"epoch": 0.
|
| 489 |
-
"grad_norm":
|
| 490 |
-
"learning_rate": 0.
|
| 491 |
-
"loss":
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"epoch": 0.
|
| 496 |
-
"grad_norm":
|
| 497 |
-
"learning_rate": 0.
|
| 498 |
-
"loss":
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
-
"epoch": 0.
|
| 503 |
-
"grad_norm":
|
| 504 |
-
"learning_rate": 0.
|
| 505 |
-
"loss":
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
-
"grad_norm":
|
| 511 |
-
"learning_rate": 0.
|
| 512 |
-
"loss": 1.
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 0.
|
| 517 |
-
"grad_norm": 2.
|
| 518 |
-
"learning_rate": 0.
|
| 519 |
-
"loss":
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
-
"epoch": 0.
|
| 524 |
-
"grad_norm":
|
| 525 |
-
"learning_rate": 0.
|
| 526 |
-
"loss":
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
-
"epoch": 0.
|
| 531 |
-
"grad_norm":
|
| 532 |
-
"learning_rate": 0.
|
| 533 |
-
"loss":
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
-
"epoch": 0.
|
| 538 |
-
"grad_norm":
|
| 539 |
-
"learning_rate": 0.
|
| 540 |
-
"loss":
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch": 0.
|
| 545 |
-
"grad_norm":
|
| 546 |
-
"learning_rate": 0.
|
| 547 |
-
"loss":
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch": 0.
|
| 552 |
-
"grad_norm":
|
| 553 |
-
"learning_rate": 0.
|
| 554 |
-
"loss": 1.
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 0.
|
| 559 |
-
"grad_norm":
|
| 560 |
-
"learning_rate": 0.
|
| 561 |
-
"loss":
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch": 0.
|
| 566 |
-
"grad_norm": 3.
|
| 567 |
-
"learning_rate": 0.
|
| 568 |
-
"loss":
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch": 0.
|
| 573 |
-
"grad_norm":
|
| 574 |
"learning_rate": 0.00015800000000000002,
|
| 575 |
-
"loss":
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch": 0.
|
| 580 |
-
"grad_norm":
|
| 581 |
"learning_rate": 0.00016,
|
| 582 |
-
"loss": 1.
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch": 0.
|
| 587 |
-
"grad_norm":
|
| 588 |
"learning_rate": 0.000162,
|
| 589 |
-
"loss":
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch": 0.
|
| 594 |
-
"grad_norm":
|
| 595 |
"learning_rate": 0.000164,
|
| 596 |
-
"loss":
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch": 0.
|
| 601 |
-
"grad_norm":
|
| 602 |
"learning_rate": 0.000166,
|
| 603 |
-
"loss":
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch": 0.
|
| 608 |
-
"grad_norm":
|
| 609 |
"learning_rate": 0.000168,
|
| 610 |
-
"loss":
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
-
"epoch": 0.
|
| 615 |
-
"grad_norm":
|
| 616 |
"learning_rate": 0.00017,
|
| 617 |
-
"loss":
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch": 0.
|
| 622 |
-
"grad_norm":
|
| 623 |
"learning_rate": 0.000172,
|
| 624 |
-
"loss": 1.
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch": 0.
|
| 629 |
-
"grad_norm":
|
| 630 |
"learning_rate": 0.000174,
|
| 631 |
-
"loss":
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch": 0.
|
| 636 |
-
"grad_norm":
|
| 637 |
"learning_rate": 0.00017600000000000002,
|
| 638 |
-
"loss":
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch": 0.
|
| 643 |
-
"grad_norm":
|
| 644 |
"learning_rate": 0.00017800000000000002,
|
| 645 |
-
"loss": 1.
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
-
"epoch": 0.
|
| 650 |
-
"grad_norm":
|
| 651 |
"learning_rate": 0.00018,
|
| 652 |
-
"loss":
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
-
"epoch": 0.
|
| 657 |
-
"grad_norm":
|
| 658 |
-
"learning_rate": 0.
|
| 659 |
-
"loss": 1.
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch": 0.
|
| 664 |
-
"grad_norm":
|
| 665 |
-
"learning_rate": 0.
|
| 666 |
-
"loss": 1.
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
-
"epoch": 0.
|
| 671 |
-
"grad_norm":
|
| 672 |
-
"learning_rate": 0.
|
| 673 |
-
"loss": 1.
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch": 0.
|
| 678 |
-
"grad_norm":
|
| 679 |
-
"learning_rate": 0.
|
| 680 |
-
"loss":
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch": 0.
|
| 685 |
-
"grad_norm":
|
| 686 |
-
"learning_rate": 0.
|
| 687 |
-
"loss":
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
-
"epoch": 0.
|
| 692 |
-
"grad_norm":
|
| 693 |
-
"learning_rate": 0.
|
| 694 |
-
"loss":
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch": 0.
|
| 699 |
-
"grad_norm":
|
| 700 |
-
"learning_rate": 0.
|
| 701 |
-
"loss":
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch": 0.
|
| 706 |
-
"grad_norm":
|
| 707 |
-
"learning_rate": 0.
|
| 708 |
-
"loss":
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch": 0.
|
| 713 |
-
"grad_norm":
|
| 714 |
-
"learning_rate": 0.
|
| 715 |
-
"loss":
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch": 0.
|
| 720 |
-
"grad_norm": 2.
|
| 721 |
-
"learning_rate": 0.
|
| 722 |
-
"loss": 1.
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch": 0.
|
| 727 |
-
"grad_norm": 2.
|
| 728 |
-
"learning_rate": 0.
|
| 729 |
-
"loss": 1.
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
-
"epoch": 0.
|
| 734 |
-
"grad_norm":
|
| 735 |
-
"learning_rate": 0.
|
| 736 |
-
"loss":
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
-
"epoch": 0.
|
| 741 |
-
"grad_norm":
|
| 742 |
-
"learning_rate": 0.
|
| 743 |
-
"loss": 1.
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch": 0.
|
| 748 |
-
"grad_norm":
|
| 749 |
-
"learning_rate": 0.
|
| 750 |
-
"loss":
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
-
"epoch": 0.
|
| 755 |
-
"grad_norm":
|
| 756 |
-
"learning_rate": 0.
|
| 757 |
-
"loss":
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
-
"epoch": 0.
|
| 762 |
-
"grad_norm":
|
| 763 |
-
"learning_rate": 0.
|
| 764 |
-
"loss":
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch": 0.
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate": 0.
|
| 771 |
-
"loss":
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
-
"epoch": 0.
|
| 776 |
-
"grad_norm":
|
| 777 |
-
"learning_rate": 0.
|
| 778 |
-
"loss":
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
-
"epoch": 0.
|
| 783 |
-
"grad_norm":
|
| 784 |
-
"learning_rate": 0.
|
| 785 |
-
"loss": 1.
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
-
"epoch": 0.
|
| 790 |
-
"grad_norm":
|
| 791 |
-
"learning_rate": 0.
|
| 792 |
-
"loss":
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
-
"epoch": 0.
|
| 797 |
-
"grad_norm":
|
| 798 |
-
"learning_rate": 0.
|
| 799 |
-
"loss":
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
-
"epoch": 0.
|
| 804 |
-
"grad_norm":
|
| 805 |
-
"learning_rate": 0.
|
| 806 |
-
"loss": 1.
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch": 0.
|
| 811 |
-
"grad_norm":
|
| 812 |
-
"learning_rate": 0.
|
| 813 |
-
"loss":
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
-
"epoch": 0.
|
| 818 |
-
"grad_norm":
|
| 819 |
-
"learning_rate": 0.
|
| 820 |
-
"loss":
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
-
"epoch": 0.
|
| 825 |
-
"grad_norm":
|
| 826 |
-
"learning_rate": 0.
|
| 827 |
-
"loss": 1.
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
-
"epoch": 0.
|
| 832 |
-
"grad_norm": 1.
|
| 833 |
-
"learning_rate": 0.
|
| 834 |
-
"loss":
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
-
"epoch": 0.
|
| 839 |
-
"grad_norm":
|
| 840 |
-
"learning_rate": 0.
|
| 841 |
-
"loss": 1.
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
-
"epoch": 0.
|
| 846 |
-
"grad_norm":
|
| 847 |
-
"learning_rate": 0.
|
| 848 |
-
"loss":
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch": 0.
|
| 853 |
-
"grad_norm":
|
| 854 |
-
"learning_rate": 0.
|
| 855 |
-
"loss":
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
-
"epoch": 0.
|
| 860 |
-
"grad_norm": 1.
|
| 861 |
-
"learning_rate": 0.
|
| 862 |
-
"loss":
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
-
"epoch": 0.
|
| 867 |
-
"grad_norm":
|
| 868 |
-
"learning_rate": 0.
|
| 869 |
-
"loss":
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
-
"epoch": 0.
|
| 874 |
-
"grad_norm":
|
| 875 |
-
"learning_rate": 0.
|
| 876 |
-
"loss": 1.
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
-
"epoch": 0.
|
| 881 |
-
"grad_norm":
|
| 882 |
-
"learning_rate": 0.
|
| 883 |
-
"loss":
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
-
"epoch": 0.
|
| 888 |
-
"grad_norm":
|
| 889 |
-
"learning_rate": 0.
|
| 890 |
-
"loss": 1.
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch": 0.
|
| 895 |
-
"grad_norm":
|
| 896 |
-
"learning_rate": 0.
|
| 897 |
-
"loss": 1.
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
-
"epoch": 0.
|
| 902 |
-
"grad_norm":
|
| 903 |
-
"learning_rate": 0.
|
| 904 |
-
"loss":
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
-
"epoch": 0.
|
| 909 |
-
"grad_norm":
|
| 910 |
-
"learning_rate": 0.
|
| 911 |
-
"loss":
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
-
"epoch": 0.
|
| 916 |
-
"grad_norm": 1.
|
| 917 |
-
"learning_rate": 0.
|
| 918 |
-
"loss":
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
-
"epoch": 0.
|
| 923 |
-
"grad_norm":
|
| 924 |
-
"learning_rate": 0.
|
| 925 |
-
"loss": 1.
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
-
"epoch": 0.
|
| 930 |
-
"grad_norm":
|
| 931 |
-
"learning_rate": 0.
|
| 932 |
-
"loss":
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch": 0.
|
| 937 |
-
"grad_norm": 1.
|
| 938 |
-
"learning_rate": 0.
|
| 939 |
-
"loss":
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
-
"epoch": 0.
|
| 944 |
-
"grad_norm": 1.
|
| 945 |
-
"learning_rate": 0.
|
| 946 |
-
"loss":
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
-
"epoch": 0.
|
| 951 |
-
"grad_norm": 1.
|
| 952 |
-
"learning_rate": 0.
|
| 953 |
-
"loss":
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"epoch": 0.
|
| 958 |
-
"grad_norm": 1.
|
| 959 |
-
"learning_rate": 0.
|
| 960 |
-
"loss":
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
-
"epoch": 0.
|
| 965 |
-
"grad_norm":
|
| 966 |
-
"learning_rate": 0.
|
| 967 |
-
"loss":
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
-
"epoch": 0.
|
| 972 |
-
"grad_norm": 1.
|
| 973 |
-
"learning_rate": 0.
|
| 974 |
-
"loss":
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch": 0.
|
| 979 |
-
"grad_norm":
|
| 980 |
-
"learning_rate": 0.
|
| 981 |
-
"loss":
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
-
"epoch": 0.
|
| 986 |
-
"grad_norm":
|
| 987 |
-
"learning_rate": 0.
|
| 988 |
-
"loss":
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
-
"epoch": 0.
|
| 993 |
-
"grad_norm": 1.
|
| 994 |
-
"learning_rate": 0.
|
| 995 |
-
"loss":
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
-
"epoch": 0.
|
| 1000 |
-
"grad_norm": 2.
|
| 1001 |
-
"learning_rate": 0.
|
| 1002 |
-
"loss": 1.
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
-
"epoch": 0.
|
| 1007 |
-
"grad_norm":
|
| 1008 |
-
"learning_rate": 0.
|
| 1009 |
-
"loss":
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
-
"epoch": 0.
|
| 1014 |
-
"grad_norm":
|
| 1015 |
-
"learning_rate": 0.
|
| 1016 |
-
"loss":
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
-
"epoch": 0.
|
| 1021 |
-
"grad_norm":
|
| 1022 |
-
"learning_rate": 0.
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
-
"epoch": 0.
|
| 1028 |
-
"grad_norm":
|
| 1029 |
-
"learning_rate": 0.
|
| 1030 |
-
"loss":
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
-
"epoch": 0.
|
| 1035 |
-
"grad_norm": 1.
|
| 1036 |
-
"learning_rate": 0.
|
| 1037 |
-
"loss":
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
-
"epoch": 0.
|
| 1042 |
-
"grad_norm":
|
| 1043 |
-
"learning_rate": 0.
|
| 1044 |
-
"loss":
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
-
"epoch": 0.
|
| 1049 |
-
"grad_norm": 1.
|
| 1050 |
-
"learning_rate": 0.
|
| 1051 |
-
"loss":
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
-
"epoch": 0.
|
| 1056 |
-
"grad_norm":
|
| 1057 |
-
"learning_rate": 0.
|
| 1058 |
-
"loss": 1.
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch": 0.
|
| 1063 |
-
"grad_norm": 2.
|
| 1064 |
-
"learning_rate": 0.
|
| 1065 |
-
"loss":
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
-
"epoch": 0.
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate": 0.
|
| 1072 |
-
"loss":
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch": 0.
|
| 1077 |
-
"grad_norm":
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss":
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch": 0.
|
| 1084 |
-
"grad_norm":
|
| 1085 |
-
"learning_rate": 9.
|
| 1086 |
-
"loss":
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
-
"epoch": 0.
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate": 9.
|
| 1093 |
-
"loss":
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch": 0.
|
| 1098 |
-
"grad_norm":
|
| 1099 |
-
"learning_rate": 9.
|
| 1100 |
-
"loss":
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch": 0.
|
| 1105 |
-
"grad_norm":
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss":
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
-
"epoch": 0.
|
| 1112 |
-
"grad_norm":
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss": 1.
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch": 0.
|
| 1119 |
-
"grad_norm": 1.
|
| 1120 |
-
"learning_rate": 8.
|
| 1121 |
-
"loss":
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
-
"epoch": 0.
|
| 1126 |
-
"grad_norm":
|
| 1127 |
-
"learning_rate": 8.
|
| 1128 |
-
"loss":
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch": 0.
|
| 1133 |
-
"grad_norm":
|
| 1134 |
-
"learning_rate": 8.
|
| 1135 |
-
"loss": 1.
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch": 0.
|
| 1140 |
-
"grad_norm":
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss":
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch": 0.
|
| 1147 |
-
"grad_norm":
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss":
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
-
"epoch": 0.
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate": 7.
|
| 1156 |
-
"loss":
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch": 0.
|
| 1161 |
-
"grad_norm": 1.
|
| 1162 |
-
"learning_rate": 7.
|
| 1163 |
-
"loss":
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
-
"epoch": 0.
|
| 1168 |
-
"grad_norm":
|
| 1169 |
-
"learning_rate": 7.
|
| 1170 |
-
"loss": 1.
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
-
"epoch": 0.
|
| 1175 |
-
"grad_norm":
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss":
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch": 0.
|
| 1182 |
-
"grad_norm":
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss":
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch": 0.
|
| 1189 |
-
"grad_norm":
|
| 1190 |
-
"learning_rate": 6.
|
| 1191 |
-
"loss":
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
-
"epoch": 0.
|
| 1196 |
-
"grad_norm":
|
| 1197 |
-
"learning_rate": 6.
|
| 1198 |
-
"loss":
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch": 0.
|
| 1203 |
-
"grad_norm":
|
| 1204 |
-
"learning_rate": 6.
|
| 1205 |
-
"loss":
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
-
"epoch": 0.
|
| 1210 |
-
"grad_norm":
|
| 1211 |
-
"learning_rate":
|
| 1212 |
-
"loss":
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
-
"epoch": 0.
|
| 1217 |
-
"grad_norm": 1.
|
| 1218 |
-
"learning_rate":
|
| 1219 |
-
"loss":
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
-
"epoch": 0.
|
| 1224 |
-
"grad_norm":
|
| 1225 |
-
"learning_rate": 5.
|
| 1226 |
-
"loss":
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
-
"epoch": 0.
|
| 1231 |
-
"grad_norm":
|
| 1232 |
-
"learning_rate": 5.
|
| 1233 |
-
"loss":
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
-
"epoch": 0.
|
| 1238 |
-
"grad_norm":
|
| 1239 |
-
"learning_rate": 5.
|
| 1240 |
-
"loss":
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
-
"epoch": 0.
|
| 1245 |
-
"grad_norm": 1.
|
| 1246 |
-
"learning_rate":
|
| 1247 |
-
"loss":
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
-
"epoch": 0.
|
| 1252 |
-
"grad_norm":
|
| 1253 |
-
"learning_rate":
|
| 1254 |
-
"loss":
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
-
"epoch": 0.
|
| 1259 |
-
"grad_norm": 2.
|
| 1260 |
-
"learning_rate": 4.
|
| 1261 |
-
"loss":
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
-
"epoch": 0.
|
| 1266 |
-
"grad_norm":
|
| 1267 |
-
"learning_rate": 4.
|
| 1268 |
-
"loss":
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
-
"epoch": 0.
|
| 1273 |
-
"grad_norm":
|
| 1274 |
-
"learning_rate": 4.
|
| 1275 |
-
"loss": 1.
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
-
"epoch": 0.
|
| 1280 |
-
"grad_norm":
|
| 1281 |
-
"learning_rate":
|
| 1282 |
-
"loss":
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
-
"epoch": 0.
|
| 1287 |
-
"grad_norm": 1.
|
| 1288 |
-
"learning_rate":
|
| 1289 |
-
"loss":
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
-
"epoch": 0.
|
| 1294 |
-
"grad_norm":
|
| 1295 |
-
"learning_rate": 3.
|
| 1296 |
-
"loss":
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
-
"epoch": 0.
|
| 1301 |
-
"grad_norm":
|
| 1302 |
-
"learning_rate": 3.
|
| 1303 |
-
"loss":
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
-
"epoch": 0.
|
| 1308 |
-
"grad_norm":
|
| 1309 |
-
"learning_rate": 3.
|
| 1310 |
-
"loss":
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
-
"epoch": 0.
|
| 1315 |
-
"grad_norm":
|
| 1316 |
-
"learning_rate":
|
| 1317 |
-
"loss":
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
-
"epoch": 0.
|
| 1322 |
-
"grad_norm": 1.
|
| 1323 |
-
"learning_rate":
|
| 1324 |
-
"loss":
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
-
"epoch": 0.
|
| 1329 |
-
"grad_norm":
|
| 1330 |
-
"learning_rate": 2.
|
| 1331 |
-
"loss":
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
-
"epoch": 0.
|
| 1336 |
-
"grad_norm":
|
| 1337 |
-
"learning_rate": 2.
|
| 1338 |
-
"loss":
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
-
"epoch": 0.
|
| 1343 |
-
"grad_norm": 1.
|
| 1344 |
-
"learning_rate": 2.
|
| 1345 |
-
"loss":
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
-
"epoch": 0.
|
| 1350 |
-
"grad_norm": 1.
|
| 1351 |
-
"learning_rate":
|
| 1352 |
-
"loss":
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
-
"epoch": 0.
|
| 1357 |
-
"grad_norm": 1.
|
| 1358 |
-
"learning_rate":
|
| 1359 |
-
"loss":
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
-
"epoch": 0.
|
| 1364 |
-
"grad_norm": 2.
|
| 1365 |
-
"learning_rate": 1.
|
| 1366 |
-
"loss":
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
-
"epoch": 0.
|
| 1371 |
-
"grad_norm":
|
| 1372 |
-
"learning_rate": 1.
|
| 1373 |
-
"loss":
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
-
"epoch": 0.
|
| 1378 |
-
"grad_norm":
|
| 1379 |
-
"learning_rate": 1.
|
| 1380 |
-
"loss":
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
-
"epoch": 0.
|
| 1385 |
-
"grad_norm":
|
| 1386 |
-
"learning_rate":
|
| 1387 |
-
"loss":
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
-
"epoch": 0.
|
| 1392 |
-
"grad_norm": 1.
|
| 1393 |
-
"learning_rate":
|
| 1394 |
-
"loss":
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
-
"epoch": 0.
|
| 1399 |
-
"grad_norm": 2.
|
| 1400 |
-
"learning_rate":
|
| 1401 |
-
"loss": 1.
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
-
"epoch": 0.
|
| 1406 |
-
"grad_norm": 1.
|
| 1407 |
-
"learning_rate":
|
| 1408 |
-
"loss":
|
| 1409 |
"step": 200
|
| 1410 |
}
|
| 1411 |
],
|
|
@@ -1426,7 +1426,7 @@
|
|
| 1426 |
"attributes": {}
|
| 1427 |
}
|
| 1428 |
},
|
| 1429 |
-
"total_flos":
|
| 1430 |
"train_batch_size": 1,
|
| 1431 |
"trial_name": null,
|
| 1432 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.00485074853113271,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 2.4253742655663552e-05,
|
| 13 |
+
"grad_norm": 2.9237160682678223,
|
| 14 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 15 |
+
"loss": 1.4192,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 4.8507485311327104e-05,
|
| 20 |
+
"grad_norm": 8.286487579345703,
|
| 21 |
+
"learning_rate": 4.000000000000001e-06,
|
| 22 |
+
"loss": 1.7523,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 7.276122796699065e-05,
|
| 27 |
+
"grad_norm": 1.957672119140625,
|
| 28 |
+
"learning_rate": 6e-06,
|
| 29 |
+
"loss": 1.9875,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 9.701497062265421e-05,
|
| 34 |
+
"grad_norm": 0.5736268758773804,
|
| 35 |
+
"learning_rate": 8.000000000000001e-06,
|
| 36 |
+
"loss": 1.427,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.00012126871327831776,
|
| 41 |
+
"grad_norm": 0.7024741172790527,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 1.4541,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.0001455224559339813,
|
| 48 |
+
"grad_norm": 6.171888828277588,
|
| 49 |
+
"learning_rate": 1.2e-05,
|
| 50 |
+
"loss": 1.9617,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.00016977619858964487,
|
| 55 |
+
"grad_norm": 1.4335829019546509,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 57 |
+
"loss": 1.6549,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.00019402994124530841,
|
| 62 |
+
"grad_norm": 0.8052822947502136,
|
| 63 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 64 |
+
"loss": 1.681,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.00021828368390097196,
|
| 69 |
+
"grad_norm": 1.0054032802581787,
|
| 70 |
+
"learning_rate": 1.8e-05,
|
| 71 |
+
"loss": 1.7113,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.00024253742655663553,
|
| 76 |
+
"grad_norm": 2.378899097442627,
|
| 77 |
+
"learning_rate": 2e-05,
|
| 78 |
+
"loss": 2.0225,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.00026679116921229907,
|
| 83 |
+
"grad_norm": 12.526110649108887,
|
| 84 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 85 |
+
"loss": 2.3219,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.0002910449118679626,
|
| 90 |
+
"grad_norm": 1.658133864402771,
|
| 91 |
+
"learning_rate": 2.4e-05,
|
| 92 |
+
"loss": 1.5336,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.00031529865452362615,
|
| 97 |
+
"grad_norm": 18.127710342407227,
|
| 98 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 99 |
+
"loss": 2.1609,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.00033955239717928975,
|
| 104 |
+
"grad_norm": 0.7923838496208191,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 106 |
+
"loss": 1.7354,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.0003638061398349533,
|
| 111 |
+
"grad_norm": 1.6248362064361572,
|
| 112 |
+
"learning_rate": 3e-05,
|
| 113 |
+
"loss": 1.5774,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.00038805988249061683,
|
| 118 |
+
"grad_norm": 1.6371029615402222,
|
| 119 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 120 |
+
"loss": 2.2978,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.00041231362514628037,
|
| 125 |
+
"grad_norm": 1.409232497215271,
|
| 126 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 127 |
+
"loss": 1.1136,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.0004365673678019439,
|
| 132 |
+
"grad_norm": 5.593843936920166,
|
| 133 |
+
"learning_rate": 3.6e-05,
|
| 134 |
+
"loss": 2.3497,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.0004608211104576075,
|
| 139 |
+
"grad_norm": 1.9881709814071655,
|
| 140 |
+
"learning_rate": 3.8e-05,
|
| 141 |
+
"loss": 2.0849,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.00048507485311327105,
|
| 146 |
+
"grad_norm": 1.7124191522598267,
|
| 147 |
+
"learning_rate": 4e-05,
|
| 148 |
+
"loss": 1.565,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.0005093285957689346,
|
| 153 |
+
"grad_norm": 0.6472250819206238,
|
| 154 |
+
"learning_rate": 4.2e-05,
|
| 155 |
+
"loss": 1.0869,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.0005335823384245981,
|
| 160 |
+
"grad_norm": 4.816655158996582,
|
| 161 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 162 |
+
"loss": 2.2343,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.0005578360810802617,
|
| 167 |
+
"grad_norm": 1.3076063394546509,
|
| 168 |
+
"learning_rate": 4.600000000000001e-05,
|
| 169 |
+
"loss": 1.8767,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.0005820898237359252,
|
| 174 |
+
"grad_norm": 1.0358009338378906,
|
| 175 |
+
"learning_rate": 4.8e-05,
|
| 176 |
+
"loss": 1.7001,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.0006063435663915888,
|
| 181 |
+
"grad_norm": 0.8611243367195129,
|
| 182 |
+
"learning_rate": 5e-05,
|
| 183 |
+
"loss": 1.0703,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.0006305973090472523,
|
| 188 |
+
"grad_norm": 1.8082849979400635,
|
| 189 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 190 |
+
"loss": 2.409,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.000654851051702916,
|
| 195 |
+
"grad_norm": 2.4376935958862305,
|
| 196 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 197 |
+
"loss": 1.6188,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.0006791047943585795,
|
| 202 |
+
"grad_norm": 1.7561759948730469,
|
| 203 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 204 |
+
"loss": 1.8782,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.000703358537014243,
|
| 209 |
+
"grad_norm": 1.4189746379852295,
|
| 210 |
+
"learning_rate": 5.8e-05,
|
| 211 |
+
"loss": 2.1414,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.0007276122796699066,
|
| 216 |
+
"grad_norm": 1.186546802520752,
|
| 217 |
+
"learning_rate": 6e-05,
|
| 218 |
+
"loss": 1.5557,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.0007518660223255701,
|
| 223 |
+
"grad_norm": 1.365446925163269,
|
| 224 |
+
"learning_rate": 6.2e-05,
|
| 225 |
+
"loss": 1.9806,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.0007761197649812337,
|
| 230 |
+
"grad_norm": 5.6013641357421875,
|
| 231 |
+
"learning_rate": 6.400000000000001e-05,
|
| 232 |
+
"loss": 2.836,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.0008003735076368972,
|
| 237 |
+
"grad_norm": 1.5393822193145752,
|
| 238 |
+
"learning_rate": 6.6e-05,
|
| 239 |
+
"loss": 1.6833,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.0008246272502925607,
|
| 244 |
+
"grad_norm": 1.7934705018997192,
|
| 245 |
+
"learning_rate": 6.800000000000001e-05,
|
| 246 |
+
"loss": 1.6762,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 0.0008488809929482243,
|
| 251 |
+
"grad_norm": 1.1468597650527954,
|
| 252 |
+
"learning_rate": 7e-05,
|
| 253 |
+
"loss": 1.2819,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.0008731347356038878,
|
| 258 |
+
"grad_norm": 4.899059772491455,
|
| 259 |
+
"learning_rate": 7.2e-05,
|
| 260 |
+
"loss": 1.7406,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.0008973884782595515,
|
| 265 |
+
"grad_norm": 1.6008192300796509,
|
| 266 |
+
"learning_rate": 7.4e-05,
|
| 267 |
+
"loss": 1.5028,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 0.000921642220915215,
|
| 272 |
+
"grad_norm": 3.4235761165618896,
|
| 273 |
+
"learning_rate": 7.6e-05,
|
| 274 |
+
"loss": 1.6816,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.0009458959635708786,
|
| 279 |
+
"grad_norm": 1.1148329973220825,
|
| 280 |
+
"learning_rate": 7.800000000000001e-05,
|
| 281 |
+
"loss": 1.4805,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 0.0009701497062265421,
|
| 286 |
+
"grad_norm": 4.683089733123779,
|
| 287 |
+
"learning_rate": 8e-05,
|
| 288 |
+
"loss": 1.9679,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 0.0009944034488822055,
|
| 293 |
+
"grad_norm": 0.75764000415802,
|
| 294 |
+
"learning_rate": 8.2e-05,
|
| 295 |
+
"loss": 1.3406,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.0010186571915378692,
|
| 300 |
+
"grad_norm": 1.3041810989379883,
|
| 301 |
+
"learning_rate": 8.4e-05,
|
| 302 |
+
"loss": 1.4009,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.0010429109341935328,
|
| 307 |
+
"grad_norm": 1.4765511751174927,
|
| 308 |
+
"learning_rate": 8.6e-05,
|
| 309 |
+
"loss": 1.3446,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 0.0010671646768491963,
|
| 314 |
+
"grad_norm": 3.716846227645874,
|
| 315 |
+
"learning_rate": 8.800000000000001e-05,
|
| 316 |
+
"loss": 2.1417,
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 0.00109141841950486,
|
| 321 |
+
"grad_norm": 1.0331724882125854,
|
| 322 |
+
"learning_rate": 9e-05,
|
| 323 |
+
"loss": 1.396,
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 0.0011156721621605234,
|
| 328 |
+
"grad_norm": 2.426039934158325,
|
| 329 |
+
"learning_rate": 9.200000000000001e-05,
|
| 330 |
+
"loss": 1.466,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 0.001139925904816187,
|
| 335 |
+
"grad_norm": 1.3340977430343628,
|
| 336 |
+
"learning_rate": 9.4e-05,
|
| 337 |
+
"loss": 1.7891,
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.0011641796474718504,
|
| 342 |
+
"grad_norm": 3.5191874504089355,
|
| 343 |
+
"learning_rate": 9.6e-05,
|
| 344 |
+
"loss": 1.6633,
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.001188433390127514,
|
| 349 |
+
"grad_norm": 1.6775161027908325,
|
| 350 |
+
"learning_rate": 9.8e-05,
|
| 351 |
+
"loss": 1.4875,
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 0.0012126871327831775,
|
| 356 |
+
"grad_norm": 8.406580924987793,
|
| 357 |
+
"learning_rate": 0.0001,
|
| 358 |
+
"loss": 2.1048,
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 0.0012369408754388412,
|
| 363 |
+
"grad_norm": 6.012252330780029,
|
| 364 |
+
"learning_rate": 0.00010200000000000001,
|
| 365 |
+
"loss": 2.0742,
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 0.0012611946180945046,
|
| 370 |
+
"grad_norm": 12.059433937072754,
|
| 371 |
+
"learning_rate": 0.00010400000000000001,
|
| 372 |
+
"loss": 1.843,
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 0.0012854483607501683,
|
| 377 |
+
"grad_norm": 1.3695584535598755,
|
| 378 |
+
"learning_rate": 0.00010600000000000002,
|
| 379 |
+
"loss": 1.8865,
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.001309702103405832,
|
| 384 |
+
"grad_norm": 2.331925630569458,
|
| 385 |
+
"learning_rate": 0.00010800000000000001,
|
| 386 |
+
"loss": 1.8662,
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.0013339558460614953,
|
| 391 |
+
"grad_norm": 2.2023472785949707,
|
| 392 |
+
"learning_rate": 0.00011000000000000002,
|
| 393 |
+
"loss": 1.3941,
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 0.001358209588717159,
|
| 398 |
+
"grad_norm": 1.971096396446228,
|
| 399 |
+
"learning_rate": 0.00011200000000000001,
|
| 400 |
+
"loss": 1.9362,
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.0013824633313728224,
|
| 405 |
+
"grad_norm": 1.254345417022705,
|
| 406 |
+
"learning_rate": 0.00011399999999999999,
|
| 407 |
+
"loss": 1.5581,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 0.001406717074028486,
|
| 412 |
+
"grad_norm": Infinity,
|
| 413 |
"learning_rate": 0.00011399999999999999,
|
| 414 |
+
"loss": 1.5454,
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
+
"epoch": 0.0014309708166841495,
|
| 419 |
+
"grad_norm": 1.5801585912704468,
|
| 420 |
"learning_rate": 0.000116,
|
| 421 |
+
"loss": 1.553,
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.0014552245593398132,
|
| 426 |
+
"grad_norm": 0.9962314963340759,
|
| 427 |
"learning_rate": 0.000118,
|
| 428 |
+
"loss": 1.4924,
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 0.0014794783019954766,
|
| 433 |
+
"grad_norm": 5.291327476501465,
|
| 434 |
"learning_rate": 0.00012,
|
| 435 |
+
"loss": 1.8365,
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
+
"epoch": 0.0015037320446511402,
|
| 440 |
+
"grad_norm": 1.6945866346359253,
|
| 441 |
"learning_rate": 0.000122,
|
| 442 |
+
"loss": 1.4683,
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 0.0015279857873068039,
|
| 447 |
+
"grad_norm": 1.0400323867797852,
|
| 448 |
+
"learning_rate": 0.000124,
|
| 449 |
+
"loss": 0.8587,
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
+
"epoch": 0.0015522395299624673,
|
| 454 |
+
"grad_norm": 3.5120689868927,
|
| 455 |
+
"learning_rate": 0.000126,
|
| 456 |
+
"loss": 1.2127,
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
+
"epoch": 0.001576493272618131,
|
| 461 |
+
"grad_norm": 1.0469675064086914,
|
| 462 |
+
"learning_rate": 0.00012800000000000002,
|
| 463 |
+
"loss": 1.654,
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.0016007470152737944,
|
| 468 |
+
"grad_norm": 1.5673848390579224,
|
| 469 |
+
"learning_rate": 0.00013000000000000002,
|
| 470 |
+
"loss": 1.801,
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 0.001625000757929458,
|
| 475 |
+
"grad_norm": 2.0134799480438232,
|
| 476 |
+
"learning_rate": 0.000132,
|
| 477 |
+
"loss": 2.0432,
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
+
"epoch": 0.0016492545005851215,
|
| 482 |
+
"grad_norm": 7.747060775756836,
|
| 483 |
+
"learning_rate": 0.000134,
|
| 484 |
+
"loss": 1.4823,
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 0.0016735082432407851,
|
| 489 |
+
"grad_norm": 3.2787840366363525,
|
| 490 |
+
"learning_rate": 0.00013600000000000003,
|
| 491 |
+
"loss": 1.4865,
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"epoch": 0.0016977619858964486,
|
| 496 |
+
"grad_norm": 4.277163028717041,
|
| 497 |
+
"learning_rate": 0.000138,
|
| 498 |
+
"loss": 1.4665,
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
+
"epoch": 0.0017220157285521122,
|
| 503 |
+
"grad_norm": 11.562691688537598,
|
| 504 |
+
"learning_rate": 0.00014,
|
| 505 |
+
"loss": 2.0221,
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.0017462694712077757,
|
| 510 |
+
"grad_norm": 2.5653882026672363,
|
| 511 |
+
"learning_rate": 0.000142,
|
| 512 |
+
"loss": 1.4631,
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 0.0017705232138634393,
|
| 517 |
+
"grad_norm": 2.392688751220703,
|
| 518 |
+
"learning_rate": 0.000144,
|
| 519 |
+
"loss": 1.3514,
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
+
"epoch": 0.001794776956519103,
|
| 524 |
+
"grad_norm": 3.874115467071533,
|
| 525 |
+
"learning_rate": 0.000146,
|
| 526 |
+
"loss": 1.5983,
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
+
"epoch": 0.0018190306991747664,
|
| 531 |
+
"grad_norm": 3.085160732269287,
|
| 532 |
+
"learning_rate": 0.000148,
|
| 533 |
+
"loss": 1.321,
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
+
"epoch": 0.00184328444183043,
|
| 538 |
+
"grad_norm": 1.7499727010726929,
|
| 539 |
+
"learning_rate": 0.00015000000000000001,
|
| 540 |
+
"loss": 1.385,
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
+
"epoch": 0.0018675381844860935,
|
| 545 |
+
"grad_norm": 14.051904678344727,
|
| 546 |
+
"learning_rate": 0.000152,
|
| 547 |
+
"loss": 1.5748,
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.0018917919271417571,
|
| 552 |
+
"grad_norm": 4.689906597137451,
|
| 553 |
+
"learning_rate": 0.000154,
|
| 554 |
+
"loss": 1.2126,
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 0.0019160456697974206,
|
| 559 |
+
"grad_norm": 1.0368577241897583,
|
| 560 |
+
"learning_rate": 0.00015600000000000002,
|
| 561 |
+
"loss": 1.0542,
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"epoch": 0.0019402994124530842,
|
| 566 |
+
"grad_norm": 3.577094078063965,
|
| 567 |
+
"learning_rate": 0.00015800000000000002,
|
| 568 |
+
"loss": 1.9481,
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 0.001964553155108748,
|
| 573 |
+
"grad_norm": NaN,
|
| 574 |
"learning_rate": 0.00015800000000000002,
|
| 575 |
+
"loss": 1.2913,
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 0.001988806897764411,
|
| 580 |
+
"grad_norm": 4.969227313995361,
|
| 581 |
"learning_rate": 0.00016,
|
| 582 |
+
"loss": 1.5551,
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 0.0020130606404200747,
|
| 587 |
+
"grad_norm": 71.41371154785156,
|
| 588 |
"learning_rate": 0.000162,
|
| 589 |
+
"loss": 1.5847,
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.0020373143830757384,
|
| 594 |
+
"grad_norm": 0.9679685831069946,
|
| 595 |
"learning_rate": 0.000164,
|
| 596 |
+
"loss": 1.1516,
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 0.002061568125731402,
|
| 601 |
+
"grad_norm": 7.050138473510742,
|
| 602 |
"learning_rate": 0.000166,
|
| 603 |
+
"loss": 1.6904,
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 0.0020858218683870657,
|
| 608 |
+
"grad_norm": 1.4181028604507446,
|
| 609 |
"learning_rate": 0.000168,
|
| 610 |
+
"loss": 1.5385,
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 0.002110075611042729,
|
| 615 |
+
"grad_norm": 1.122887134552002,
|
| 616 |
"learning_rate": 0.00017,
|
| 617 |
+
"loss": 1.282,
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 0.0021343293536983925,
|
| 622 |
+
"grad_norm": 3.0532054901123047,
|
| 623 |
"learning_rate": 0.000172,
|
| 624 |
+
"loss": 1.7126,
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
+
"epoch": 0.002158583096354056,
|
| 629 |
+
"grad_norm": 1.1368091106414795,
|
| 630 |
"learning_rate": 0.000174,
|
| 631 |
+
"loss": 1.3792,
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 0.00218283683900972,
|
| 636 |
+
"grad_norm": 1.5556614398956299,
|
| 637 |
"learning_rate": 0.00017600000000000002,
|
| 638 |
+
"loss": 1.1266,
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 0.002207090581665383,
|
| 643 |
+
"grad_norm": 2.5969090461730957,
|
| 644 |
"learning_rate": 0.00017800000000000002,
|
| 645 |
+
"loss": 1.1916,
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
+
"epoch": 0.0022313443243210467,
|
| 650 |
+
"grad_norm": 7.9968719482421875,
|
| 651 |
"learning_rate": 0.00018,
|
| 652 |
+
"loss": 2.3005,
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
+
"epoch": 0.0022555980669767104,
|
| 657 |
+
"grad_norm": 1.0678150653839111,
|
| 658 |
+
"learning_rate": 0.000182,
|
| 659 |
+
"loss": 1.1991,
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
+
"epoch": 0.002279851809632374,
|
| 664 |
+
"grad_norm": 6.678997993469238,
|
| 665 |
+
"learning_rate": 0.00018400000000000003,
|
| 666 |
+
"loss": 1.9749,
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
+
"epoch": 0.0023041055522880377,
|
| 671 |
+
"grad_norm": 2.6500961780548096,
|
| 672 |
+
"learning_rate": 0.00018600000000000002,
|
| 673 |
+
"loss": 1.7569,
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
+
"epoch": 0.002328359294943701,
|
| 678 |
+
"grad_norm": 1.547878384590149,
|
| 679 |
+
"learning_rate": 0.000188,
|
| 680 |
+
"loss": 1.7286,
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 0.0023526130375993645,
|
| 685 |
+
"grad_norm": 1.910861849784851,
|
| 686 |
+
"learning_rate": 0.00019,
|
| 687 |
+
"loss": 2.0815,
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
+
"epoch": 0.002376866780255028,
|
| 692 |
+
"grad_norm": 3.678144693374634,
|
| 693 |
+
"learning_rate": 0.000192,
|
| 694 |
+
"loss": 1.7011,
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
+
"epoch": 0.002401120522910692,
|
| 699 |
+
"grad_norm": 1.6496407985687256,
|
| 700 |
+
"learning_rate": 0.000194,
|
| 701 |
+
"loss": 1.8013,
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"epoch": 0.002425374265566355,
|
| 706 |
+
"grad_norm": 1.3245445489883423,
|
| 707 |
+
"learning_rate": 0.000196,
|
| 708 |
+
"loss": 1.5302,
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
+
"epoch": 0.0024496280082220187,
|
| 713 |
+
"grad_norm": 1.609408974647522,
|
| 714 |
+
"learning_rate": 0.00019800000000000002,
|
| 715 |
+
"loss": 1.6082,
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
+
"epoch": 0.0024738817508776823,
|
| 720 |
+
"grad_norm": 2.7757482528686523,
|
| 721 |
+
"learning_rate": 0.0002,
|
| 722 |
+
"loss": 1.6157,
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 0.002498135493533346,
|
| 727 |
+
"grad_norm": 2.0618674755096436,
|
| 728 |
+
"learning_rate": 0.00019800000000000002,
|
| 729 |
+
"loss": 1.7212,
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
+
"epoch": 0.002522389236189009,
|
| 734 |
+
"grad_norm": 1.7433160543441772,
|
| 735 |
+
"learning_rate": 0.000196,
|
| 736 |
+
"loss": 1.9133,
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
+
"epoch": 0.002546642978844673,
|
| 741 |
+
"grad_norm": 1.4754176139831543,
|
| 742 |
+
"learning_rate": 0.000194,
|
| 743 |
+
"loss": 1.3311,
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"epoch": 0.0025708967215003365,
|
| 748 |
+
"grad_norm": 1.4527512788772583,
|
| 749 |
+
"learning_rate": 0.000192,
|
| 750 |
+
"loss": 1.6432,
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
+
"epoch": 0.002595150464156,
|
| 755 |
+
"grad_norm": 5.3099045753479,
|
| 756 |
+
"learning_rate": 0.00019,
|
| 757 |
+
"loss": 1.3233,
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
+
"epoch": 0.002619404206811664,
|
| 762 |
+
"grad_norm": 1.016135573387146,
|
| 763 |
+
"learning_rate": 0.000188,
|
| 764 |
+
"loss": 1.3291,
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 0.002643657949467327,
|
| 769 |
+
"grad_norm": 1.0468915700912476,
|
| 770 |
+
"learning_rate": 0.00018600000000000002,
|
| 771 |
+
"loss": 1.4454,
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
+
"epoch": 0.0026679116921229907,
|
| 776 |
+
"grad_norm": 3.245159387588501,
|
| 777 |
+
"learning_rate": 0.00018400000000000003,
|
| 778 |
+
"loss": 1.6501,
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 0.0026921654347786543,
|
| 783 |
+
"grad_norm": 1.1628005504608154,
|
| 784 |
+
"learning_rate": 0.000182,
|
| 785 |
+
"loss": 1.399,
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
+
"epoch": 0.002716419177434318,
|
| 790 |
+
"grad_norm": 3.3215761184692383,
|
| 791 |
+
"learning_rate": 0.00018,
|
| 792 |
+
"loss": 1.4829,
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
+
"epoch": 0.002740672920089981,
|
| 797 |
+
"grad_norm": 1.424054503440857,
|
| 798 |
+
"learning_rate": 0.00017800000000000002,
|
| 799 |
+
"loss": 1.7692,
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
+
"epoch": 0.002764926662745645,
|
| 804 |
+
"grad_norm": 1.2700444459915161,
|
| 805 |
+
"learning_rate": 0.00017600000000000002,
|
| 806 |
+
"loss": 1.6959,
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 0.0027891804054013085,
|
| 811 |
+
"grad_norm": 1.262736201286316,
|
| 812 |
+
"learning_rate": 0.000174,
|
| 813 |
+
"loss": 1.5445,
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"epoch": 0.002813434148056972,
|
| 818 |
+
"grad_norm": 4.976276397705078,
|
| 819 |
+
"learning_rate": 0.000172,
|
| 820 |
+
"loss": 1.4571,
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 0.002837687890712636,
|
| 825 |
+
"grad_norm": 1.2716983556747437,
|
| 826 |
+
"learning_rate": 0.00017,
|
| 827 |
+
"loss": 1.217,
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
+
"epoch": 0.002861941633368299,
|
| 832 |
+
"grad_norm": 1.5143672227859497,
|
| 833 |
+
"learning_rate": 0.000168,
|
| 834 |
+
"loss": 1.3059,
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
+
"epoch": 0.0028861953760239627,
|
| 839 |
+
"grad_norm": 0.9907928705215454,
|
| 840 |
+
"learning_rate": 0.000166,
|
| 841 |
+
"loss": 1.8035,
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
+
"epoch": 0.0029104491186796263,
|
| 846 |
+
"grad_norm": 1.7163773775100708,
|
| 847 |
+
"learning_rate": 0.000164,
|
| 848 |
+
"loss": 2.1635,
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 0.00293470286133529,
|
| 853 |
+
"grad_norm": 7.744179725646973,
|
| 854 |
+
"learning_rate": 0.000162,
|
| 855 |
+
"loss": 2.2964,
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
+
"epoch": 0.002958956603990953,
|
| 860 |
+
"grad_norm": 1.101208209991455,
|
| 861 |
+
"learning_rate": 0.00016,
|
| 862 |
+
"loss": 1.46,
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 0.002983210346646617,
|
| 867 |
+
"grad_norm": 0.7589418888092041,
|
| 868 |
+
"learning_rate": 0.00015800000000000002,
|
| 869 |
+
"loss": 1.1036,
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
+
"epoch": 0.0030074640893022805,
|
| 874 |
+
"grad_norm": 4.626345634460449,
|
| 875 |
+
"learning_rate": 0.00015600000000000002,
|
| 876 |
+
"loss": 1.5257,
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
+
"epoch": 0.003031717831957944,
|
| 881 |
+
"grad_norm": 1.5994218587875366,
|
| 882 |
+
"learning_rate": 0.000154,
|
| 883 |
+
"loss": 1.2856,
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
+
"epoch": 0.0030559715746136078,
|
| 888 |
+
"grad_norm": 1.3893097639083862,
|
| 889 |
+
"learning_rate": 0.000152,
|
| 890 |
+
"loss": 1.3359,
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 0.003080225317269271,
|
| 895 |
+
"grad_norm": 1.0536174774169922,
|
| 896 |
+
"learning_rate": 0.00015000000000000001,
|
| 897 |
+
"loss": 1.1718,
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
+
"epoch": 0.0031044790599249346,
|
| 902 |
+
"grad_norm": 1.2898694276809692,
|
| 903 |
+
"learning_rate": 0.000148,
|
| 904 |
+
"loss": 1.2908,
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
+
"epoch": 0.0031287328025805983,
|
| 909 |
+
"grad_norm": 1.6943238973617554,
|
| 910 |
+
"learning_rate": 0.000146,
|
| 911 |
+
"loss": 1.7723,
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
+
"epoch": 0.003152986545236262,
|
| 916 |
+
"grad_norm": 1.2562038898468018,
|
| 917 |
+
"learning_rate": 0.000144,
|
| 918 |
+
"loss": 1.4476,
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
+
"epoch": 0.003177240287891925,
|
| 923 |
+
"grad_norm": 1.0170366764068604,
|
| 924 |
+
"learning_rate": 0.000142,
|
| 925 |
+
"loss": 1.2967,
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
+
"epoch": 0.003201494030547589,
|
| 930 |
+
"grad_norm": 5.646616458892822,
|
| 931 |
+
"learning_rate": 0.00014,
|
| 932 |
+
"loss": 1.4366,
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 0.0032257477732032525,
|
| 937 |
+
"grad_norm": 1.29356849193573,
|
| 938 |
+
"learning_rate": 0.000138,
|
| 939 |
+
"loss": 1.4713,
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
+
"epoch": 0.003250001515858916,
|
| 944 |
+
"grad_norm": 1.3575730323791504,
|
| 945 |
+
"learning_rate": 0.00013600000000000003,
|
| 946 |
+
"loss": 1.2233,
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
+
"epoch": 0.0032742552585145793,
|
| 951 |
+
"grad_norm": 1.158686876296997,
|
| 952 |
+
"learning_rate": 0.000134,
|
| 953 |
+
"loss": 1.3622,
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"epoch": 0.003298509001170243,
|
| 958 |
+
"grad_norm": 1.4680081605911255,
|
| 959 |
+
"learning_rate": 0.000132,
|
| 960 |
+
"loss": 1.7191,
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
+
"epoch": 0.0033227627438259066,
|
| 965 |
+
"grad_norm": 1.2448982000350952,
|
| 966 |
+
"learning_rate": 0.00013000000000000002,
|
| 967 |
+
"loss": 1.9376,
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
+
"epoch": 0.0033470164864815703,
|
| 972 |
+
"grad_norm": 1.2404478788375854,
|
| 973 |
+
"learning_rate": 0.00012800000000000002,
|
| 974 |
+
"loss": 1.5334,
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 0.003371270229137234,
|
| 979 |
+
"grad_norm": 0.9862974882125854,
|
| 980 |
+
"learning_rate": 0.000126,
|
| 981 |
+
"loss": 1.2073,
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
+
"epoch": 0.003395523971792897,
|
| 986 |
+
"grad_norm": 7.36776876449585,
|
| 987 |
+
"learning_rate": 0.000124,
|
| 988 |
+
"loss": 1.5745,
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 0.003419777714448561,
|
| 993 |
+
"grad_norm": 1.1356145143508911,
|
| 994 |
+
"learning_rate": 0.000122,
|
| 995 |
+
"loss": 1.563,
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
+
"epoch": 0.0034440314571042244,
|
| 1000 |
+
"grad_norm": 2.073420286178589,
|
| 1001 |
+
"learning_rate": 0.00012,
|
| 1002 |
+
"loss": 1.7438,
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
+
"epoch": 0.003468285199759888,
|
| 1007 |
+
"grad_norm": 3.029547691345215,
|
| 1008 |
+
"learning_rate": 0.000118,
|
| 1009 |
+
"loss": 1.6755,
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
+
"epoch": 0.0034925389424155513,
|
| 1014 |
+
"grad_norm": 14.66256046295166,
|
| 1015 |
+
"learning_rate": 0.000116,
|
| 1016 |
+
"loss": 1.8375,
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 0.003516792685071215,
|
| 1021 |
+
"grad_norm": 0.8765383958816528,
|
| 1022 |
+
"learning_rate": 0.00011399999999999999,
|
| 1023 |
+
"loss": 0.6782,
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"epoch": 0.0035410464277268786,
|
| 1028 |
+
"grad_norm": 3.167731761932373,
|
| 1029 |
+
"learning_rate": 0.00011200000000000001,
|
| 1030 |
+
"loss": 1.4619,
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 0.0035653001703825423,
|
| 1035 |
+
"grad_norm": 1.1016606092453003,
|
| 1036 |
+
"learning_rate": 0.00011000000000000002,
|
| 1037 |
+
"loss": 1.5817,
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
+
"epoch": 0.003589553913038206,
|
| 1042 |
+
"grad_norm": 2.5535902976989746,
|
| 1043 |
+
"learning_rate": 0.00010800000000000001,
|
| 1044 |
+
"loss": 1.3491,
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
+
"epoch": 0.003613807655693869,
|
| 1049 |
+
"grad_norm": 1.1920926570892334,
|
| 1050 |
+
"learning_rate": 0.00010600000000000002,
|
| 1051 |
+
"loss": 1.5675,
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
+
"epoch": 0.0036380613983495328,
|
| 1056 |
+
"grad_norm": 1.1023300886154175,
|
| 1057 |
+
"learning_rate": 0.00010400000000000001,
|
| 1058 |
+
"loss": 1.2821,
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 0.0036623151410051964,
|
| 1063 |
+
"grad_norm": 2.1146063804626465,
|
| 1064 |
+
"learning_rate": 0.00010200000000000001,
|
| 1065 |
+
"loss": 1.5396,
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
+
"epoch": 0.00368656888366086,
|
| 1070 |
+
"grad_norm": 1.1838195323944092,
|
| 1071 |
+
"learning_rate": 0.0001,
|
| 1072 |
+
"loss": 1.2992,
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
+
"epoch": 0.0037108226263165233,
|
| 1077 |
+
"grad_norm": 1.414258360862732,
|
| 1078 |
+
"learning_rate": 9.8e-05,
|
| 1079 |
+
"loss": 1.6888,
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
+
"epoch": 0.003735076368972187,
|
| 1084 |
+
"grad_norm": 0.8630995154380798,
|
| 1085 |
+
"learning_rate": 9.6e-05,
|
| 1086 |
+
"loss": 1.141,
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
+
"epoch": 0.0037593301116278506,
|
| 1091 |
+
"grad_norm": 2.4763197898864746,
|
| 1092 |
+
"learning_rate": 9.4e-05,
|
| 1093 |
+
"loss": 1.5743,
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 0.0037835838542835142,
|
| 1098 |
+
"grad_norm": 0.9457703828811646,
|
| 1099 |
+
"learning_rate": 9.200000000000001e-05,
|
| 1100 |
+
"loss": 1.3879,
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 0.003807837596939178,
|
| 1105 |
+
"grad_norm": 1.308862328529358,
|
| 1106 |
+
"learning_rate": 9e-05,
|
| 1107 |
+
"loss": 1.1564,
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
+
"epoch": 0.003832091339594841,
|
| 1112 |
+
"grad_norm": 1.208833932876587,
|
| 1113 |
+
"learning_rate": 8.800000000000001e-05,
|
| 1114 |
+
"loss": 1.2059,
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
+
"epoch": 0.0038563450822505048,
|
| 1119 |
+
"grad_norm": 1.9145225286483765,
|
| 1120 |
+
"learning_rate": 8.6e-05,
|
| 1121 |
+
"loss": 1.4744,
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
+
"epoch": 0.0038805988249061684,
|
| 1126 |
+
"grad_norm": 1.0121599435806274,
|
| 1127 |
+
"learning_rate": 8.4e-05,
|
| 1128 |
+
"loss": 1.8676,
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
+
"epoch": 0.003904852567561832,
|
| 1133 |
+
"grad_norm": 1.9128226041793823,
|
| 1134 |
+
"learning_rate": 8.2e-05,
|
| 1135 |
+
"loss": 1.6948,
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"epoch": 0.003929106310217496,
|
| 1140 |
+
"grad_norm": 1.7783511877059937,
|
| 1141 |
+
"learning_rate": 8e-05,
|
| 1142 |
+
"loss": 1.4198,
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 0.003953360052873159,
|
| 1147 |
+
"grad_norm": 0.9799928665161133,
|
| 1148 |
+
"learning_rate": 7.800000000000001e-05,
|
| 1149 |
+
"loss": 1.185,
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
+
"epoch": 0.003977613795528822,
|
| 1154 |
+
"grad_norm": 1.5025819540023804,
|
| 1155 |
+
"learning_rate": 7.6e-05,
|
| 1156 |
+
"loss": 1.0638,
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
+
"epoch": 0.004001867538184486,
|
| 1161 |
+
"grad_norm": 1.2552540302276611,
|
| 1162 |
+
"learning_rate": 7.4e-05,
|
| 1163 |
+
"loss": 1.3463,
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
+
"epoch": 0.0040261212808401494,
|
| 1168 |
+
"grad_norm": 1.0616928339004517,
|
| 1169 |
+
"learning_rate": 7.2e-05,
|
| 1170 |
+
"loss": 1.9275,
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 0.004050375023495813,
|
| 1175 |
+
"grad_norm": 2.8277781009674072,
|
| 1176 |
+
"learning_rate": 7e-05,
|
| 1177 |
+
"loss": 1.9653,
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
+
"epoch": 0.004074628766151477,
|
| 1182 |
+
"grad_norm": 1.7752915620803833,
|
| 1183 |
+
"learning_rate": 6.800000000000001e-05,
|
| 1184 |
+
"loss": 1.2783,
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 0.00409888250880714,
|
| 1189 |
+
"grad_norm": 2.4863102436065674,
|
| 1190 |
+
"learning_rate": 6.6e-05,
|
| 1191 |
+
"loss": 1.2803,
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
+
"epoch": 0.004123136251462804,
|
| 1196 |
+
"grad_norm": 0.9784297943115234,
|
| 1197 |
+
"learning_rate": 6.400000000000001e-05,
|
| 1198 |
+
"loss": 1.0347,
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
+
"epoch": 0.004147389994118468,
|
| 1203 |
+
"grad_norm": 2.0077602863311768,
|
| 1204 |
+
"learning_rate": 6.2e-05,
|
| 1205 |
+
"loss": 1.5035,
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
+
"epoch": 0.004171643736774131,
|
| 1210 |
+
"grad_norm": 1.5943875312805176,
|
| 1211 |
+
"learning_rate": 6e-05,
|
| 1212 |
+
"loss": 1.7834,
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
+
"epoch": 0.004195897479429794,
|
| 1217 |
+
"grad_norm": 1.0898661613464355,
|
| 1218 |
+
"learning_rate": 5.8e-05,
|
| 1219 |
+
"loss": 1.5043,
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
+
"epoch": 0.004220151222085458,
|
| 1224 |
+
"grad_norm": 1.8181873559951782,
|
| 1225 |
+
"learning_rate": 5.6000000000000006e-05,
|
| 1226 |
+
"loss": 1.5026,
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
+
"epoch": 0.004244404964741121,
|
| 1231 |
+
"grad_norm": 0.9227128028869629,
|
| 1232 |
+
"learning_rate": 5.4000000000000005e-05,
|
| 1233 |
+
"loss": 1.3945,
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
+
"epoch": 0.004268658707396785,
|
| 1238 |
+
"grad_norm": 1.3131438493728638,
|
| 1239 |
+
"learning_rate": 5.2000000000000004e-05,
|
| 1240 |
+
"loss": 1.2629,
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
+
"epoch": 0.004292912450052449,
|
| 1245 |
+
"grad_norm": 1.388433814048767,
|
| 1246 |
+
"learning_rate": 5e-05,
|
| 1247 |
+
"loss": 1.607,
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
+
"epoch": 0.004317166192708112,
|
| 1252 |
+
"grad_norm": 1.5145083665847778,
|
| 1253 |
+
"learning_rate": 4.8e-05,
|
| 1254 |
+
"loss": 1.8671,
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
+
"epoch": 0.004341419935363776,
|
| 1259 |
+
"grad_norm": 2.0021004676818848,
|
| 1260 |
+
"learning_rate": 4.600000000000001e-05,
|
| 1261 |
+
"loss": 2.0679,
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
+
"epoch": 0.00436567367801944,
|
| 1266 |
+
"grad_norm": 1.8512483835220337,
|
| 1267 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 1268 |
+
"loss": 1.3095,
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
+
"epoch": 0.004389927420675103,
|
| 1273 |
+
"grad_norm": 1.0164090394973755,
|
| 1274 |
+
"learning_rate": 4.2e-05,
|
| 1275 |
+
"loss": 1.3749,
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
+
"epoch": 0.004414181163330766,
|
| 1280 |
+
"grad_norm": 30.730253219604492,
|
| 1281 |
+
"learning_rate": 4e-05,
|
| 1282 |
+
"loss": 2.3106,
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
+
"epoch": 0.00443843490598643,
|
| 1287 |
+
"grad_norm": 1.2311471700668335,
|
| 1288 |
+
"learning_rate": 3.8e-05,
|
| 1289 |
+
"loss": 1.1078,
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
+
"epoch": 0.004462688648642093,
|
| 1294 |
+
"grad_norm": 0.9555554389953613,
|
| 1295 |
+
"learning_rate": 3.6e-05,
|
| 1296 |
+
"loss": 1.4175,
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
+
"epoch": 0.004486942391297757,
|
| 1301 |
+
"grad_norm": 1.4511969089508057,
|
| 1302 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 1303 |
+
"loss": 1.4099,
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
+
"epoch": 0.004511196133953421,
|
| 1308 |
+
"grad_norm": 5.319721221923828,
|
| 1309 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 1310 |
+
"loss": 1.2749,
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
+
"epoch": 0.004535449876609084,
|
| 1315 |
+
"grad_norm": 1.8668482303619385,
|
| 1316 |
+
"learning_rate": 3e-05,
|
| 1317 |
+
"loss": 1.4658,
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
+
"epoch": 0.004559703619264748,
|
| 1322 |
+
"grad_norm": 1.737425684928894,
|
| 1323 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 1324 |
+
"loss": 1.8142,
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
+
"epoch": 0.004583957361920412,
|
| 1329 |
+
"grad_norm": 12.860699653625488,
|
| 1330 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 1331 |
+
"loss": 1.6338,
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
+
"epoch": 0.004608211104576075,
|
| 1336 |
+
"grad_norm": 0.6315305233001709,
|
| 1337 |
+
"learning_rate": 2.4e-05,
|
| 1338 |
+
"loss": 1.0426,
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
+
"epoch": 0.004632464847231738,
|
| 1343 |
+
"grad_norm": 1.0747138261795044,
|
| 1344 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 1345 |
+
"loss": 1.4042,
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
+
"epoch": 0.004656718589887402,
|
| 1350 |
+
"grad_norm": 1.1410670280456543,
|
| 1351 |
+
"learning_rate": 2e-05,
|
| 1352 |
+
"loss": 1.2148,
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
+
"epoch": 0.004680972332543065,
|
| 1357 |
+
"grad_norm": 1.4486732482910156,
|
| 1358 |
+
"learning_rate": 1.8e-05,
|
| 1359 |
+
"loss": 1.3208,
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
+
"epoch": 0.004705226075198729,
|
| 1364 |
+
"grad_norm": 2.5336716175079346,
|
| 1365 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 1366 |
+
"loss": 1.7136,
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
+
"epoch": 0.004729479817854393,
|
| 1371 |
+
"grad_norm": 3.4056637287139893,
|
| 1372 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 1373 |
+
"loss": 1.5443,
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
+
"epoch": 0.004753733560510056,
|
| 1378 |
+
"grad_norm": 1.1721996068954468,
|
| 1379 |
+
"learning_rate": 1.2e-05,
|
| 1380 |
+
"loss": 1.4518,
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
+
"epoch": 0.00477798730316572,
|
| 1385 |
+
"grad_norm": 1.9326874017715454,
|
| 1386 |
+
"learning_rate": 1e-05,
|
| 1387 |
+
"loss": 1.6381,
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
+
"epoch": 0.004802241045821384,
|
| 1392 |
+
"grad_norm": 1.0234136581420898,
|
| 1393 |
+
"learning_rate": 8.000000000000001e-06,
|
| 1394 |
+
"loss": 1.4116,
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
+
"epoch": 0.004826494788477047,
|
| 1399 |
+
"grad_norm": 2.1815526485443115,
|
| 1400 |
+
"learning_rate": 6e-06,
|
| 1401 |
+
"loss": 1.7442,
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
+
"epoch": 0.00485074853113271,
|
| 1406 |
+
"grad_norm": 1.9557406902313232,
|
| 1407 |
+
"learning_rate": 4.000000000000001e-06,
|
| 1408 |
+
"loss": 1.9568,
|
| 1409 |
"step": 200
|
| 1410 |
}
|
| 1411 |
],
|
|
|
|
| 1426 |
"attributes": {}
|
| 1427 |
}
|
| 1428 |
},
|
| 1429 |
+
"total_flos": 6034692870365184.0,
|
| 1430 |
"train_batch_size": 1,
|
| 1431 |
"trial_name": null,
|
| 1432 |
"trial_params": null
|
checkpoint-200/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02e9962458f016a1c07dc5280d7007cfd14653c05542b51edd26eefb09ce3f00
|
| 3 |
size 5176
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02e9962458f016a1c07dc5280d7007cfd14653c05542b51edd26eefb09ce3f00
|
| 3 |
size 5176
|