End of training
Browse files- all_results.json +5 -5
- test_results.json +5 -5
- trainer_state.json +825 -825
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.2,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 1.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.2,
|
| 3 |
+
"eval_accuracy": 0.8456375838926175,
|
| 4 |
+
"eval_loss": 0.7402271032333374,
|
| 5 |
+
"eval_runtime": 328.0839,
|
| 6 |
+
"eval_samples_per_second": 1.817,
|
| 7 |
+
"eval_steps_per_second": 1.817
|
| 8 |
}
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.2,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second": 1.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.2,
|
| 3 |
+
"eval_accuracy": 0.8456375838926175,
|
| 4 |
+
"eval_loss": 0.7402271032333374,
|
| 5 |
+
"eval_runtime": 328.0839,
|
| 6 |
+
"eval_samples_per_second": 1.817,
|
| 7 |
+
"eval_steps_per_second": 1.817
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "ctsinov1/checkpoint-
|
| 4 |
"epoch": 4.2,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 3935,
|
|
@@ -10,2825 +10,2825 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 2.538071065989848e-07,
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.01,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 5.076142131979696e-07,
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
-
"grad_norm": 6.
|
| 28 |
"learning_rate": 7.614213197969544e-07,
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01,
|
| 34 |
-
"grad_norm":
|
| 35 |
"learning_rate": 1.0152284263959392e-06,
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.01,
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 1.2690355329949238e-06,
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 1.5228426395939088e-06,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02,
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 1.7766497461928936e-06,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.02,
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 2.0304568527918785e-06,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.02,
|
| 69 |
-
"grad_norm":
|
| 70 |
"learning_rate": 2.284263959390863e-06,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03,
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 2.5380710659898476e-06,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.03,
|
| 83 |
-
"grad_norm":
|
| 84 |
"learning_rate": 2.7918781725888327e-06,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.03,
|
| 90 |
-
"grad_norm":
|
| 91 |
"learning_rate": 3.0456852791878177e-06,
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.03,
|
| 97 |
-
"grad_norm":
|
| 98 |
"learning_rate": 3.2994923857868023e-06,
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.04,
|
| 104 |
-
"grad_norm":
|
| 105 |
"learning_rate": 3.5532994923857873e-06,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.04,
|
| 111 |
-
"grad_norm":
|
| 112 |
"learning_rate": 3.8071065989847715e-06,
|
| 113 |
-
"loss":
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.04,
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 4.060913705583757e-06,
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.04,
|
| 125 |
-
"grad_norm":
|
| 126 |
"learning_rate": 4.3147208121827415e-06,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.05,
|
| 132 |
-
"grad_norm":
|
| 133 |
"learning_rate": 4.568527918781726e-06,
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.05,
|
| 139 |
-
"grad_norm":
|
| 140 |
"learning_rate": 4.822335025380711e-06,
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.05,
|
| 146 |
-
"grad_norm":
|
| 147 |
"learning_rate": 5.076142131979695e-06,
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.05,
|
| 153 |
-
"grad_norm":
|
| 154 |
"learning_rate": 5.329949238578681e-06,
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.06,
|
| 160 |
-
"grad_norm":
|
| 161 |
"learning_rate": 5.583756345177665e-06,
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.06,
|
| 167 |
-
"grad_norm":
|
| 168 |
"learning_rate": 5.83756345177665e-06,
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.06,
|
| 174 |
-
"grad_norm":
|
| 175 |
"learning_rate": 6.091370558375635e-06,
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.06,
|
| 181 |
-
"grad_norm":
|
| 182 |
"learning_rate": 6.34517766497462e-06,
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.07,
|
| 188 |
-
"grad_norm":
|
| 189 |
"learning_rate": 6.5989847715736045e-06,
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.07,
|
| 195 |
-
"grad_norm":
|
| 196 |
"learning_rate": 6.852791878172589e-06,
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.07,
|
| 202 |
-
"grad_norm":
|
| 203 |
"learning_rate": 7.106598984771575e-06,
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.07,
|
| 209 |
-
"grad_norm":
|
| 210 |
"learning_rate": 7.360406091370559e-06,
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.08,
|
| 216 |
-
"grad_norm":
|
| 217 |
"learning_rate": 7.614213197969543e-06,
|
| 218 |
-
"loss":
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.08,
|
| 223 |
-
"grad_norm":
|
| 224 |
"learning_rate": 7.86802030456853e-06,
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.08,
|
| 230 |
-
"grad_norm":
|
| 231 |
"learning_rate": 8.121827411167514e-06,
|
| 232 |
-
"loss":
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.08,
|
| 237 |
-
"grad_norm":
|
| 238 |
"learning_rate": 8.375634517766498e-06,
|
| 239 |
-
"loss":
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.09,
|
| 244 |
-
"grad_norm":
|
| 245 |
"learning_rate": 8.629441624365483e-06,
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.09,
|
| 251 |
-
"grad_norm":
|
| 252 |
"learning_rate": 8.883248730964468e-06,
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.09,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 9.137055837563452e-06,
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.09,
|
| 265 |
-
"grad_norm":
|
| 266 |
"learning_rate": 9.390862944162438e-06,
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1,
|
| 272 |
-
"grad_norm":
|
| 273 |
"learning_rate": 9.644670050761421e-06,
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.1,
|
| 279 |
-
"grad_norm":
|
| 280 |
"learning_rate": 9.898477157360406e-06,
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.1,
|
| 286 |
-
"grad_norm":
|
| 287 |
"learning_rate": 9.983055634001694e-06,
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.1,
|
| 293 |
-
"grad_norm":
|
| 294 |
"learning_rate": 9.95481502400452e-06,
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.11,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
"learning_rate": 9.926574414007344e-06,
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.11,
|
| 307 |
-
"grad_norm":
|
| 308 |
"learning_rate": 9.898333804010167e-06,
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.11,
|
| 314 |
-
"grad_norm":
|
| 315 |
"learning_rate": 9.870093194012991e-06,
|
| 316 |
-
"loss":
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.11,
|
| 321 |
-
"grad_norm":
|
| 322 |
"learning_rate": 9.841852584015815e-06,
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.12,
|
| 328 |
-
"grad_norm":
|
| 329 |
"learning_rate": 9.81361197401864e-06,
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.12,
|
| 335 |
-
"grad_norm":
|
| 336 |
"learning_rate": 9.785371364021462e-06,
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.12,
|
| 342 |
-
"grad_norm":
|
| 343 |
"learning_rate": 9.757130754024288e-06,
|
| 344 |
-
"loss":
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.12,
|
| 349 |
-
"grad_norm":
|
| 350 |
"learning_rate": 9.728890144027113e-06,
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.13,
|
| 356 |
-
"grad_norm":
|
| 357 |
"learning_rate": 9.700649534029935e-06,
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.13,
|
| 363 |
-
"grad_norm":
|
| 364 |
"learning_rate": 9.67240892403276e-06,
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.13,
|
| 370 |
-
"grad_norm":
|
| 371 |
"learning_rate": 9.644168314035584e-06,
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.13,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
"learning_rate": 9.615927704038408e-06,
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.14,
|
| 384 |
-
"grad_norm":
|
| 385 |
"learning_rate": 9.587687094041232e-06,
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.14,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
"learning_rate": 9.559446484044057e-06,
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.14,
|
| 398 |
-
"grad_norm":
|
| 399 |
"learning_rate": 9.531205874046881e-06,
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.14,
|
| 405 |
-
"grad_norm":
|
| 406 |
"learning_rate": 9.502965264049704e-06,
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.15,
|
| 412 |
-
"grad_norm":
|
| 413 |
"learning_rate": 9.474724654052528e-06,
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.15,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
"learning_rate": 9.446484044055352e-06,
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.15,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
"learning_rate": 9.418243434058176e-06,
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.16,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
"learning_rate": 9.390002824061e-06,
|
| 435 |
-
"loss": 2.
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.16,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
"learning_rate": 9.361762214063825e-06,
|
| 442 |
-
"loss": 1.
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.16,
|
| 447 |
-
"grad_norm":
|
| 448 |
"learning_rate": 9.33352160406665e-06,
|
| 449 |
-
"loss":
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.16,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
"learning_rate": 9.305280994069472e-06,
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.17,
|
| 461 |
-
"grad_norm":
|
| 462 |
"learning_rate": 9.277040384072296e-06,
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.17,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
"learning_rate": 9.24879977407512e-06,
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.17,
|
| 475 |
-
"grad_norm":
|
| 476 |
"learning_rate": 9.220559164077945e-06,
|
| 477 |
-
"loss": 1.
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.17,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
"learning_rate": 9.192318554080769e-06,
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.18,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 9.164077944083593e-06,
|
| 491 |
-
"loss": 0.
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.18,
|
| 496 |
-
"grad_norm":
|
| 497 |
"learning_rate": 9.135837334086418e-06,
|
| 498 |
-
"loss": 1.
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.18,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
"learning_rate": 9.107596724089242e-06,
|
| 505 |
-
"loss": 1.
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.18,
|
| 510 |
-
"grad_norm":
|
| 511 |
"learning_rate": 9.079356114092065e-06,
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.19,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 9.051115504094889e-06,
|
| 519 |
-
"loss": 1.
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.19,
|
| 524 |
-
"grad_norm":
|
| 525 |
"learning_rate": 9.022874894097713e-06,
|
| 526 |
-
"loss":
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.19,
|
| 531 |
-
"grad_norm":
|
| 532 |
"learning_rate": 8.994634284100537e-06,
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.19,
|
| 538 |
-
"grad_norm":
|
| 539 |
"learning_rate": 8.96639367410336e-06,
|
| 540 |
-
"loss":
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.2,
|
| 545 |
-
"grad_norm":
|
| 546 |
"learning_rate": 8.938153064106186e-06,
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.2,
|
| 552 |
-
"grad_norm":
|
| 553 |
"learning_rate": 8.90991245410901e-06,
|
| 554 |
-
"loss": 1.
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.2,
|
| 559 |
-
"eval_accuracy": 0.
|
| 560 |
-
"eval_loss": 1.
|
| 561 |
-
"eval_runtime":
|
| 562 |
-
"eval_samples_per_second": 1.
|
| 563 |
-
"eval_steps_per_second": 1.
|
| 564 |
"step": 787
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 1.0,
|
| 568 |
-
"grad_norm":
|
| 569 |
"learning_rate": 8.881671844111833e-06,
|
| 570 |
-
"loss": 1.
|
| 571 |
"step": 790
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 1.0,
|
| 575 |
-
"grad_norm":
|
| 576 |
"learning_rate": 8.853431234114657e-06,
|
| 577 |
-
"loss": 0.
|
| 578 |
"step": 800
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 1.01,
|
| 582 |
-
"grad_norm":
|
| 583 |
"learning_rate": 8.825190624117481e-06,
|
| 584 |
-
"loss": 0.
|
| 585 |
"step": 810
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 1.01,
|
| 589 |
-
"grad_norm":
|
| 590 |
"learning_rate": 8.796950014120306e-06,
|
| 591 |
-
"loss": 0.
|
| 592 |
"step": 820
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 1.01,
|
| 596 |
-
"grad_norm":
|
| 597 |
"learning_rate": 8.76870940412313e-06,
|
| 598 |
-
"loss": 1.
|
| 599 |
"step": 830
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 1.01,
|
| 603 |
-
"grad_norm": 0.
|
| 604 |
"learning_rate": 8.740468794125954e-06,
|
| 605 |
-
"loss":
|
| 606 |
"step": 840
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 1.02,
|
| 610 |
-
"grad_norm": 0.
|
| 611 |
"learning_rate": 8.712228184128779e-06,
|
| 612 |
-
"loss": 0.
|
| 613 |
"step": 850
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 1.02,
|
| 617 |
-
"grad_norm":
|
| 618 |
"learning_rate": 8.683987574131601e-06,
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 860
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 1.02,
|
| 624 |
-
"grad_norm":
|
| 625 |
"learning_rate": 8.655746964134425e-06,
|
| 626 |
-
"loss":
|
| 627 |
"step": 870
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 1.02,
|
| 631 |
-
"grad_norm": 0.
|
| 632 |
"learning_rate": 8.62750635413725e-06,
|
| 633 |
-
"loss": 1.
|
| 634 |
"step": 880
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 1.03,
|
| 638 |
-
"grad_norm":
|
| 639 |
"learning_rate": 8.599265744140074e-06,
|
| 640 |
-
"loss": 1.
|
| 641 |
"step": 890
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 1.03,
|
| 645 |
-
"grad_norm":
|
| 646 |
"learning_rate": 8.571025134142898e-06,
|
| 647 |
-
"loss": 1.
|
| 648 |
"step": 900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 1.03,
|
| 652 |
-
"grad_norm":
|
| 653 |
"learning_rate": 8.542784524145723e-06,
|
| 654 |
-
"loss":
|
| 655 |
"step": 910
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 1.03,
|
| 659 |
-
"grad_norm": 0.
|
| 660 |
"learning_rate": 8.514543914148547e-06,
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 920
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 1.04,
|
| 666 |
-
"grad_norm": 0.
|
| 667 |
"learning_rate": 8.48630330415137e-06,
|
| 668 |
-
"loss": 0.
|
| 669 |
"step": 930
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 1.04,
|
| 673 |
-
"grad_norm": 0.
|
| 674 |
"learning_rate": 8.458062694154194e-06,
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 940
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 1.04,
|
| 680 |
-
"grad_norm":
|
| 681 |
"learning_rate": 8.429822084157018e-06,
|
| 682 |
-
"loss": 1.
|
| 683 |
"step": 950
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 1.04,
|
| 687 |
-
"grad_norm":
|
| 688 |
"learning_rate": 8.401581474159842e-06,
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 960
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 1.05,
|
| 694 |
-
"grad_norm":
|
| 695 |
"learning_rate": 8.373340864162667e-06,
|
| 696 |
-
"loss":
|
| 697 |
"step": 970
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 1.05,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
"learning_rate": 8.345100254165491e-06,
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 980
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 1.05,
|
| 708 |
-
"grad_norm": 0.
|
| 709 |
"learning_rate": 8.316859644168315e-06,
|
| 710 |
-
"loss":
|
| 711 |
"step": 990
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 1.05,
|
| 715 |
-
"grad_norm":
|
| 716 |
"learning_rate": 8.28861903417114e-06,
|
| 717 |
-
"loss":
|
| 718 |
"step": 1000
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 1.06,
|
| 722 |
-
"grad_norm": 0.
|
| 723 |
"learning_rate": 8.260378424173962e-06,
|
| 724 |
-
"loss": 0.
|
| 725 |
"step": 1010
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 1.06,
|
| 729 |
-
"grad_norm": 0.
|
| 730 |
"learning_rate": 8.232137814176786e-06,
|
| 731 |
-
"loss": 0.
|
| 732 |
"step": 1020
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 1.06,
|
| 736 |
-
"grad_norm": 0.
|
| 737 |
"learning_rate": 8.20389720417961e-06,
|
| 738 |
-
"loss":
|
| 739 |
"step": 1030
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 1.06,
|
| 743 |
-
"grad_norm": 0.
|
| 744 |
"learning_rate": 8.175656594182435e-06,
|
| 745 |
-
"loss": 0.
|
| 746 |
"step": 1040
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 1.07,
|
| 750 |
-
"grad_norm":
|
| 751 |
"learning_rate": 8.14741598418526e-06,
|
| 752 |
-
"loss":
|
| 753 |
"step": 1050
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 1.07,
|
| 757 |
-
"grad_norm": 0.
|
| 758 |
"learning_rate": 8.119175374188084e-06,
|
| 759 |
-
"loss": 0.
|
| 760 |
"step": 1060
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 1.07,
|
| 764 |
-
"grad_norm": 0.
|
| 765 |
"learning_rate": 8.090934764190908e-06,
|
| 766 |
-
"loss": 0.
|
| 767 |
"step": 1070
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 1.07,
|
| 771 |
-
"grad_norm": 0.
|
| 772 |
"learning_rate": 8.06269415419373e-06,
|
| 773 |
-
"loss": 1.
|
| 774 |
"step": 1080
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 1.08,
|
| 778 |
-
"grad_norm": 0.
|
| 779 |
"learning_rate": 8.034453544196555e-06,
|
| 780 |
-
"loss": 0.
|
| 781 |
"step": 1090
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 1.08,
|
| 785 |
-
"grad_norm":
|
| 786 |
"learning_rate": 8.006212934199379e-06,
|
| 787 |
-
"loss": 0.
|
| 788 |
"step": 1100
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 1.08,
|
| 792 |
-
"grad_norm":
|
| 793 |
"learning_rate": 7.977972324202203e-06,
|
| 794 |
-
"loss": 1.
|
| 795 |
"step": 1110
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 1.08,
|
| 799 |
-
"grad_norm":
|
| 800 |
"learning_rate": 7.949731714205028e-06,
|
| 801 |
-
"loss": 0.
|
| 802 |
"step": 1120
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 1.09,
|
| 806 |
-
"grad_norm":
|
| 807 |
"learning_rate": 7.921491104207852e-06,
|
| 808 |
-
"loss":
|
| 809 |
"step": 1130
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 1.09,
|
| 813 |
-
"grad_norm":
|
| 814 |
"learning_rate": 7.893250494210676e-06,
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 1140
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 1.09,
|
| 820 |
-
"grad_norm": 0.
|
| 821 |
"learning_rate": 7.865009884213499e-06,
|
| 822 |
-
"loss":
|
| 823 |
"step": 1150
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 1.09,
|
| 827 |
-
"grad_norm":
|
| 828 |
"learning_rate": 7.836769274216323e-06,
|
| 829 |
-
"loss": 0.
|
| 830 |
"step": 1160
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 1.1,
|
| 834 |
-
"grad_norm": 0.
|
| 835 |
"learning_rate": 7.808528664219147e-06,
|
| 836 |
-
"loss": 1.
|
| 837 |
"step": 1170
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 1.1,
|
| 841 |
-
"grad_norm": 0.
|
| 842 |
"learning_rate": 7.780288054221972e-06,
|
| 843 |
-
"loss": 0.
|
| 844 |
"step": 1180
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 1.1,
|
| 848 |
-
"grad_norm": 0.
|
| 849 |
"learning_rate": 7.752047444224796e-06,
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 1190
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 1.1,
|
| 855 |
-
"grad_norm": 0.
|
| 856 |
"learning_rate": 7.72380683422762e-06,
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 1200
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 1.11,
|
| 862 |
-
"grad_norm": 0.
|
| 863 |
"learning_rate": 7.695566224230445e-06,
|
| 864 |
-
"loss": 0.
|
| 865 |
"step": 1210
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 1.11,
|
| 869 |
-
"grad_norm": 0.
|
| 870 |
"learning_rate": 7.667325614233269e-06,
|
| 871 |
-
"loss": 0.
|
| 872 |
"step": 1220
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 1.11,
|
| 876 |
-
"grad_norm": 0.
|
| 877 |
"learning_rate": 7.639085004236091e-06,
|
| 878 |
-
"loss": 2.
|
| 879 |
"step": 1230
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 1.12,
|
| 883 |
-
"grad_norm": 0.
|
| 884 |
"learning_rate": 7.610844394238917e-06,
|
| 885 |
-
"loss": 1.
|
| 886 |
"step": 1240
|
| 887 |
},
|
| 888 |
{
|
| 889 |
"epoch": 1.12,
|
| 890 |
-
"grad_norm": 0.
|
| 891 |
"learning_rate": 7.58260378424174e-06,
|
| 892 |
"loss": 0.0049,
|
| 893 |
"step": 1250
|
| 894 |
},
|
| 895 |
{
|
| 896 |
"epoch": 1.12,
|
| 897 |
-
"grad_norm": 0.
|
| 898 |
"learning_rate": 7.554363174244564e-06,
|
| 899 |
-
"loss": 0.
|
| 900 |
"step": 1260
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 1.12,
|
| 904 |
-
"grad_norm": 0.
|
| 905 |
"learning_rate": 7.526122564247388e-06,
|
| 906 |
-
"loss": 0.
|
| 907 |
"step": 1270
|
| 908 |
},
|
| 909 |
{
|
| 910 |
"epoch": 1.13,
|
| 911 |
-
"grad_norm":
|
| 912 |
"learning_rate": 7.497881954250212e-06,
|
| 913 |
-
"loss": 0.
|
| 914 |
"step": 1280
|
| 915 |
},
|
| 916 |
{
|
| 917 |
"epoch": 1.13,
|
| 918 |
-
"grad_norm":
|
| 919 |
"learning_rate": 7.469641344253037e-06,
|
| 920 |
-
"loss":
|
| 921 |
"step": 1290
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"epoch": 1.13,
|
| 925 |
-
"grad_norm": 0.
|
| 926 |
"learning_rate": 7.441400734255861e-06,
|
| 927 |
-
"loss": 0.
|
| 928 |
"step": 1300
|
| 929 |
},
|
| 930 |
{
|
| 931 |
"epoch": 1.13,
|
| 932 |
-
"grad_norm":
|
| 933 |
"learning_rate": 7.413160124258685e-06,
|
| 934 |
-
"loss": 1.
|
| 935 |
"step": 1310
|
| 936 |
},
|
| 937 |
{
|
| 938 |
"epoch": 1.14,
|
| 939 |
-
"grad_norm":
|
| 940 |
"learning_rate": 7.384919514261508e-06,
|
| 941 |
-
"loss":
|
| 942 |
"step": 1320
|
| 943 |
},
|
| 944 |
{
|
| 945 |
"epoch": 1.14,
|
| 946 |
-
"grad_norm": 0.
|
| 947 |
"learning_rate": 7.356678904264333e-06,
|
| 948 |
-
"loss": 0.
|
| 949 |
"step": 1330
|
| 950 |
},
|
| 951 |
{
|
| 952 |
"epoch": 1.14,
|
| 953 |
-
"grad_norm": 0.
|
| 954 |
"learning_rate": 7.328438294267157e-06,
|
| 955 |
-
"loss": 0.
|
| 956 |
"step": 1340
|
| 957 |
},
|
| 958 |
{
|
| 959 |
"epoch": 1.14,
|
| 960 |
-
"grad_norm": 0.
|
| 961 |
"learning_rate": 7.30019768426998e-06,
|
| 962 |
-
"loss": 0.
|
| 963 |
"step": 1350
|
| 964 |
},
|
| 965 |
{
|
| 966 |
"epoch": 1.15,
|
| 967 |
-
"grad_norm": 0.
|
| 968 |
"learning_rate": 7.2719570742728055e-06,
|
| 969 |
-
"loss": 1.
|
| 970 |
"step": 1360
|
| 971 |
},
|
| 972 |
{
|
| 973 |
"epoch": 1.15,
|
| 974 |
-
"grad_norm": 0.
|
| 975 |
"learning_rate": 7.243716464275629e-06,
|
| 976 |
-
"loss": 0.
|
| 977 |
"step": 1370
|
| 978 |
},
|
| 979 |
{
|
| 980 |
"epoch": 1.15,
|
| 981 |
-
"grad_norm": 0.
|
| 982 |
"learning_rate": 7.215475854278453e-06,
|
| 983 |
-
"loss":
|
| 984 |
"step": 1380
|
| 985 |
},
|
| 986 |
{
|
| 987 |
"epoch": 1.15,
|
| 988 |
-
"grad_norm": 0.
|
| 989 |
"learning_rate": 7.1872352442812775e-06,
|
| 990 |
-
"loss": 1.
|
| 991 |
"step": 1390
|
| 992 |
},
|
| 993 |
{
|
| 994 |
"epoch": 1.16,
|
| 995 |
-
"grad_norm": 0.
|
| 996 |
"learning_rate": 7.158994634284101e-06,
|
| 997 |
-
"loss": 1.
|
| 998 |
"step": 1400
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
"epoch": 1.16,
|
| 1002 |
-
"grad_norm": 0.
|
| 1003 |
"learning_rate": 7.130754024286925e-06,
|
| 1004 |
-
"loss": 1.
|
| 1005 |
"step": 1410
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
"epoch": 1.16,
|
| 1009 |
-
"grad_norm": 0.
|
| 1010 |
"learning_rate": 7.102513414289749e-06,
|
| 1011 |
-
"loss": 0.
|
| 1012 |
"step": 1420
|
| 1013 |
},
|
| 1014 |
{
|
| 1015 |
"epoch": 1.16,
|
| 1016 |
-
"grad_norm":
|
| 1017 |
"learning_rate": 7.074272804292574e-06,
|
| 1018 |
-
"loss": 0.
|
| 1019 |
"step": 1430
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
"epoch": 1.17,
|
| 1023 |
-
"grad_norm": 0.
|
| 1024 |
"learning_rate": 7.0460321942953965e-06,
|
| 1025 |
-
"loss": 0.
|
| 1026 |
"step": 1440
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
"epoch": 1.17,
|
| 1030 |
-
"grad_norm": 0.
|
| 1031 |
"learning_rate": 7.017791584298222e-06,
|
| 1032 |
-
"loss":
|
| 1033 |
"step": 1450
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 1.17,
|
| 1037 |
-
"grad_norm": 0.
|
| 1038 |
"learning_rate": 6.989550974301046e-06,
|
| 1039 |
-
"loss": 1.
|
| 1040 |
"step": 1460
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
"epoch": 1.17,
|
| 1044 |
-
"grad_norm": 0.
|
| 1045 |
"learning_rate": 6.961310364303869e-06,
|
| 1046 |
-
"loss":
|
| 1047 |
"step": 1470
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
"epoch": 1.18,
|
| 1051 |
-
"grad_norm": 0.
|
| 1052 |
"learning_rate": 6.933069754306694e-06,
|
| 1053 |
-
"loss":
|
| 1054 |
"step": 1480
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
"epoch": 1.18,
|
| 1058 |
-
"grad_norm":
|
| 1059 |
"learning_rate": 6.904829144309517e-06,
|
| 1060 |
-
"loss": 1.
|
| 1061 |
"step": 1490
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
"epoch": 1.18,
|
| 1065 |
-
"grad_norm": 0.
|
| 1066 |
"learning_rate": 6.876588534312341e-06,
|
| 1067 |
-
"loss": 0.
|
| 1068 |
"step": 1500
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 1.18,
|
| 1072 |
-
"grad_norm":
|
| 1073 |
"learning_rate": 6.8483479243151665e-06,
|
| 1074 |
-
"loss": 0.
|
| 1075 |
"step": 1510
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.19,
|
| 1079 |
-
"grad_norm":
|
| 1080 |
"learning_rate": 6.82010731431799e-06,
|
| 1081 |
-
"loss": 0.
|
| 1082 |
"step": 1520
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
"epoch": 1.19,
|
| 1086 |
-
"grad_norm":
|
| 1087 |
"learning_rate": 6.791866704320814e-06,
|
| 1088 |
-
"loss": 0.
|
| 1089 |
"step": 1530
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 1.19,
|
| 1093 |
-
"grad_norm": 0.
|
| 1094 |
"learning_rate": 6.763626094323638e-06,
|
| 1095 |
-
"loss":
|
| 1096 |
"step": 1540
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
"epoch": 1.19,
|
| 1100 |
-
"grad_norm": 0.
|
| 1101 |
"learning_rate": 6.735385484326462e-06,
|
| 1102 |
-
"loss": 0.
|
| 1103 |
"step": 1550
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
"epoch": 1.2,
|
| 1107 |
-
"grad_norm": 0.
|
| 1108 |
"learning_rate": 6.707144874329285e-06,
|
| 1109 |
-
"loss": 0.
|
| 1110 |
"step": 1560
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
"epoch": 1.2,
|
| 1114 |
-
"grad_norm": 0.
|
| 1115 |
"learning_rate": 6.67890426433211e-06,
|
| 1116 |
-
"loss": 0.
|
| 1117 |
"step": 1570
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
"epoch": 1.2,
|
| 1121 |
-
"eval_accuracy": 0.
|
| 1122 |
-
"eval_loss": 0.
|
| 1123 |
-
"eval_runtime":
|
| 1124 |
-
"eval_samples_per_second": 1.
|
| 1125 |
-
"eval_steps_per_second": 1.
|
| 1126 |
"step": 1574
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
"epoch": 2.0,
|
| 1130 |
-
"grad_norm":
|
| 1131 |
"learning_rate": 6.650663654334935e-06,
|
| 1132 |
-
"loss": 1.
|
| 1133 |
"step": 1580
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
"epoch": 2.0,
|
| 1137 |
-
"grad_norm": 0.
|
| 1138 |
"learning_rate": 6.622423044337758e-06,
|
| 1139 |
-
"loss": 0.
|
| 1140 |
"step": 1590
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 2.01,
|
| 1144 |
-
"grad_norm": 0.
|
| 1145 |
"learning_rate": 6.5941824343405826e-06,
|
| 1146 |
-
"loss": 1.
|
| 1147 |
"step": 1600
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
"epoch": 2.01,
|
| 1151 |
-
"grad_norm": 0.
|
| 1152 |
"learning_rate": 6.565941824343406e-06,
|
| 1153 |
-
"loss": 0.
|
| 1154 |
"step": 1610
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
"epoch": 2.01,
|
| 1158 |
-
"grad_norm": 0.
|
| 1159 |
"learning_rate": 6.53770121434623e-06,
|
| 1160 |
-
"loss":
|
| 1161 |
"step": 1620
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 2.01,
|
| 1165 |
-
"grad_norm": 0.
|
| 1166 |
"learning_rate": 6.5094606043490546e-06,
|
| 1167 |
-
"loss": 0.
|
| 1168 |
"step": 1630
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 2.02,
|
| 1172 |
-
"grad_norm": 0.
|
| 1173 |
"learning_rate": 6.481219994351878e-06,
|
| 1174 |
-
"loss": 0.
|
| 1175 |
"step": 1640
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 2.02,
|
| 1179 |
-
"grad_norm": 0.
|
| 1180 |
"learning_rate": 6.452979384354703e-06,
|
| 1181 |
-
"loss": 0.
|
| 1182 |
"step": 1650
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 2.02,
|
| 1186 |
-
"grad_norm": 0.
|
| 1187 |
"learning_rate": 6.424738774357527e-06,
|
| 1188 |
-
"loss": 1.
|
| 1189 |
"step": 1660
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 2.02,
|
| 1193 |
-
"grad_norm": 0.
|
| 1194 |
"learning_rate": 6.396498164360351e-06,
|
| 1195 |
-
"loss":
|
| 1196 |
"step": 1670
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 2.03,
|
| 1200 |
-
"grad_norm":
|
| 1201 |
"learning_rate": 6.368257554363175e-06,
|
| 1202 |
-
"loss": 0.
|
| 1203 |
"step": 1680
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 2.03,
|
| 1207 |
-
"grad_norm": 0.
|
| 1208 |
"learning_rate": 6.340016944365999e-06,
|
| 1209 |
-
"loss": 0.
|
| 1210 |
"step": 1690
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 2.03,
|
| 1214 |
-
"grad_norm":
|
| 1215 |
"learning_rate": 6.311776334368823e-06,
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 1700
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 2.03,
|
| 1221 |
-
"grad_norm":
|
| 1222 |
"learning_rate": 6.283535724371646e-06,
|
| 1223 |
-
"loss": 1.
|
| 1224 |
"step": 1710
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 2.04,
|
| 1228 |
-
"grad_norm": 0.
|
| 1229 |
"learning_rate": 6.2552951143744715e-06,
|
| 1230 |
-
"loss":
|
| 1231 |
"step": 1720
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 2.04,
|
| 1235 |
-
"grad_norm":
|
| 1236 |
"learning_rate": 6.227054504377295e-06,
|
| 1237 |
-
"loss": 0.
|
| 1238 |
"step": 1730
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 2.04,
|
| 1242 |
-
"grad_norm": 0.
|
| 1243 |
"learning_rate": 6.198813894380119e-06,
|
| 1244 |
-
"loss": 0.
|
| 1245 |
"step": 1740
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 2.04,
|
| 1249 |
-
"grad_norm":
|
| 1250 |
"learning_rate": 6.1705732843829435e-06,
|
| 1251 |
-
"loss": 1.
|
| 1252 |
"step": 1750
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 2.05,
|
| 1256 |
-
"grad_norm":
|
| 1257 |
"learning_rate": 6.142332674385767e-06,
|
| 1258 |
-
"loss": 1.
|
| 1259 |
"step": 1760
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 2.05,
|
| 1263 |
-
"grad_norm": 0.
|
| 1264 |
"learning_rate": 6.114092064388591e-06,
|
| 1265 |
-
"loss": 0.
|
| 1266 |
"step": 1770
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 2.05,
|
| 1270 |
-
"grad_norm": 0.
|
| 1271 |
"learning_rate": 6.085851454391415e-06,
|
| 1272 |
-
"loss": 0.
|
| 1273 |
"step": 1780
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 2.05,
|
| 1277 |
-
"grad_norm": 0.
|
| 1278 |
"learning_rate": 6.05761084439424e-06,
|
| 1279 |
-
"loss": 0.
|
| 1280 |
"step": 1790
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 2.06,
|
| 1284 |
-
"grad_norm": 0.
|
| 1285 |
"learning_rate": 6.029370234397064e-06,
|
| 1286 |
-
"loss":
|
| 1287 |
"step": 1800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 2.06,
|
| 1291 |
-
"grad_norm": 0.
|
| 1292 |
"learning_rate": 6.0011296243998876e-06,
|
| 1293 |
-
"loss": 0.
|
| 1294 |
"step": 1810
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 2.06,
|
| 1298 |
-
"grad_norm": 0.
|
| 1299 |
"learning_rate": 5.972889014402712e-06,
|
| 1300 |
-
"loss": 0.
|
| 1301 |
"step": 1820
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 2.07,
|
| 1305 |
-
"grad_norm": 0.
|
| 1306 |
"learning_rate": 5.944648404405535e-06,
|
| 1307 |
-
"loss":
|
| 1308 |
"step": 1830
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 2.07,
|
| 1312 |
-
"grad_norm": 0.
|
| 1313 |
"learning_rate": 5.9164077944083596e-06,
|
| 1314 |
-
"loss": 1.
|
| 1315 |
"step": 1840
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 2.07,
|
| 1319 |
-
"grad_norm": 0.
|
| 1320 |
"learning_rate": 5.888167184411185e-06,
|
| 1321 |
-
"loss": 0.
|
| 1322 |
"step": 1850
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 2.07,
|
| 1326 |
-
"grad_norm":
|
| 1327 |
"learning_rate": 5.859926574414007e-06,
|
| 1328 |
-
"loss":
|
| 1329 |
"step": 1860
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 2.08,
|
| 1333 |
-
"grad_norm": 0.
|
| 1334 |
"learning_rate": 5.8316859644168324e-06,
|
| 1335 |
-
"loss": 0.
|
| 1336 |
"step": 1870
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 2.08,
|
| 1340 |
-
"grad_norm": 0.
|
| 1341 |
"learning_rate": 5.803445354419656e-06,
|
| 1342 |
-
"loss": 0.
|
| 1343 |
"step": 1880
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 2.08,
|
| 1347 |
-
"grad_norm": 0.
|
| 1348 |
"learning_rate": 5.77520474442248e-06,
|
| 1349 |
-
"loss": 0.
|
| 1350 |
"step": 1890
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 2.08,
|
| 1354 |
-
"grad_norm": 0.
|
| 1355 |
"learning_rate": 5.746964134425304e-06,
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 1900
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 2.09,
|
| 1361 |
-
"grad_norm":
|
| 1362 |
"learning_rate": 5.718723524428128e-06,
|
| 1363 |
-
"loss": 0.
|
| 1364 |
"step": 1910
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 2.09,
|
| 1368 |
-
"grad_norm": 0.
|
| 1369 |
"learning_rate": 5.690482914430952e-06,
|
| 1370 |
-
"loss": 0.
|
| 1371 |
"step": 1920
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 2.09,
|
| 1375 |
-
"grad_norm": 0.
|
| 1376 |
"learning_rate": 5.662242304433776e-06,
|
| 1377 |
-
"loss": 0.
|
| 1378 |
"step": 1930
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 2.09,
|
| 1382 |
-
"grad_norm": 0.
|
| 1383 |
"learning_rate": 5.634001694436601e-06,
|
| 1384 |
-
"loss": 0.
|
| 1385 |
"step": 1940
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 2.1,
|
| 1389 |
-
"grad_norm":
|
| 1390 |
"learning_rate": 5.605761084439424e-06,
|
| 1391 |
-
"loss": 1.
|
| 1392 |
"step": 1950
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 2.1,
|
| 1396 |
-
"grad_norm": 0.
|
| 1397 |
"learning_rate": 5.5775204744422485e-06,
|
| 1398 |
-
"loss":
|
| 1399 |
"step": 1960
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 2.1,
|
| 1403 |
-
"grad_norm": 0.
|
| 1404 |
"learning_rate": 5.549279864445073e-06,
|
| 1405 |
-
"loss": 0.
|
| 1406 |
"step": 1970
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 2.1,
|
| 1410 |
-
"grad_norm": 0.
|
| 1411 |
"learning_rate": 5.521039254447896e-06,
|
| 1412 |
-
"loss":
|
| 1413 |
"step": 1980
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 2.11,
|
| 1417 |
-
"grad_norm": 0.
|
| 1418 |
"learning_rate": 5.4927986444507205e-06,
|
| 1419 |
-
"loss": 0.
|
| 1420 |
"step": 1990
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 2.11,
|
| 1424 |
-
"grad_norm": 0.
|
| 1425 |
"learning_rate": 5.464558034453544e-06,
|
| 1426 |
-
"loss": 0.
|
| 1427 |
"step": 2000
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 2.11,
|
| 1431 |
-
"grad_norm": 0.
|
| 1432 |
"learning_rate": 5.436317424456369e-06,
|
| 1433 |
-
"loss": 0.
|
| 1434 |
"step": 2010
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 2.11,
|
| 1438 |
-
"grad_norm":
|
| 1439 |
"learning_rate": 5.408076814459193e-06,
|
| 1440 |
-
"loss": 0.
|
| 1441 |
"step": 2020
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 2.12,
|
| 1445 |
-
"grad_norm":
|
| 1446 |
"learning_rate": 5.379836204462017e-06,
|
| 1447 |
-
"loss": 1.
|
| 1448 |
"step": 2030
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
"epoch": 2.12,
|
| 1452 |
-
"grad_norm":
|
| 1453 |
"learning_rate": 5.351595594464841e-06,
|
| 1454 |
-
"loss": 2.
|
| 1455 |
"step": 2040
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
"epoch": 2.12,
|
| 1459 |
-
"grad_norm": 0.
|
| 1460 |
"learning_rate": 5.323354984467665e-06,
|
| 1461 |
-
"loss": 1.
|
| 1462 |
"step": 2050
|
| 1463 |
},
|
| 1464 |
{
|
| 1465 |
"epoch": 2.12,
|
| 1466 |
-
"grad_norm": 0.
|
| 1467 |
"learning_rate": 5.295114374470489e-06,
|
| 1468 |
-
"loss": 0.
|
| 1469 |
"step": 2060
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
"epoch": 2.13,
|
| 1473 |
-
"grad_norm": 0.
|
| 1474 |
"learning_rate": 5.266873764473312e-06,
|
| 1475 |
-
"loss": 0.
|
| 1476 |
"step": 2070
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
"epoch": 2.13,
|
| 1480 |
-
"grad_norm": 0.
|
| 1481 |
"learning_rate": 5.2386331544761374e-06,
|
| 1482 |
-
"loss": 0.
|
| 1483 |
"step": 2080
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
"epoch": 2.13,
|
| 1487 |
-
"grad_norm": 0.
|
| 1488 |
"learning_rate": 5.210392544478962e-06,
|
| 1489 |
-
"loss": 0.
|
| 1490 |
"step": 2090
|
| 1491 |
},
|
| 1492 |
{
|
| 1493 |
"epoch": 2.13,
|
| 1494 |
-
"grad_norm":
|
| 1495 |
"learning_rate": 5.182151934481785e-06,
|
| 1496 |
-
"loss": 0.
|
| 1497 |
"step": 2100
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 2.14,
|
| 1501 |
-
"grad_norm": 0.
|
| 1502 |
"learning_rate": 5.1539113244846095e-06,
|
| 1503 |
-
"loss":
|
| 1504 |
"step": 2110
|
| 1505 |
},
|
| 1506 |
{
|
| 1507 |
"epoch": 2.14,
|
| 1508 |
-
"grad_norm": 0.
|
| 1509 |
"learning_rate": 5.125670714487433e-06,
|
| 1510 |
-
"loss": 0.
|
| 1511 |
"step": 2120
|
| 1512 |
},
|
| 1513 |
{
|
| 1514 |
"epoch": 2.14,
|
| 1515 |
-
"grad_norm":
|
| 1516 |
"learning_rate": 5.097430104490257e-06,
|
| 1517 |
-
"loss": 0.
|
| 1518 |
"step": 2130
|
| 1519 |
},
|
| 1520 |
{
|
| 1521 |
"epoch": 2.14,
|
| 1522 |
-
"grad_norm": 0.
|
| 1523 |
"learning_rate": 5.069189494493082e-06,
|
| 1524 |
-
"loss": 1.
|
| 1525 |
"step": 2140
|
| 1526 |
},
|
| 1527 |
{
|
| 1528 |
"epoch": 2.15,
|
| 1529 |
-
"grad_norm": 0.
|
| 1530 |
"learning_rate": 5.040948884495906e-06,
|
| 1531 |
-
"loss": 0.
|
| 1532 |
"step": 2150
|
| 1533 |
},
|
| 1534 |
{
|
| 1535 |
"epoch": 2.15,
|
| 1536 |
-
"grad_norm":
|
| 1537 |
"learning_rate": 5.01270827449873e-06,
|
| 1538 |
-
"loss": 0.
|
| 1539 |
"step": 2160
|
| 1540 |
},
|
| 1541 |
{
|
| 1542 |
"epoch": 2.15,
|
| 1543 |
-
"grad_norm": 0.
|
| 1544 |
"learning_rate": 4.9844676645015535e-06,
|
| 1545 |
-
"loss": 0.
|
| 1546 |
"step": 2170
|
| 1547 |
},
|
| 1548 |
{
|
| 1549 |
"epoch": 2.15,
|
| 1550 |
-
"grad_norm": 0.
|
| 1551 |
"learning_rate": 4.956227054504378e-06,
|
| 1552 |
-
"loss": 0.
|
| 1553 |
"step": 2180
|
| 1554 |
},
|
| 1555 |
{
|
| 1556 |
"epoch": 2.16,
|
| 1557 |
-
"grad_norm":
|
| 1558 |
"learning_rate": 4.927986444507202e-06,
|
| 1559 |
-
"loss": 1.
|
| 1560 |
"step": 2190
|
| 1561 |
},
|
| 1562 |
{
|
| 1563 |
"epoch": 2.16,
|
| 1564 |
-
"grad_norm": 0.
|
| 1565 |
"learning_rate": 4.8997458345100255e-06,
|
| 1566 |
-
"loss": 1.
|
| 1567 |
"step": 2200
|
| 1568 |
},
|
| 1569 |
{
|
| 1570 |
"epoch": 2.16,
|
| 1571 |
-
"grad_norm": 0.
|
| 1572 |
"learning_rate": 4.87150522451285e-06,
|
| 1573 |
-
"loss": 1.
|
| 1574 |
"step": 2210
|
| 1575 |
},
|
| 1576 |
{
|
| 1577 |
"epoch": 2.16,
|
| 1578 |
-
"grad_norm":
|
| 1579 |
"learning_rate": 4.843264614515674e-06,
|
| 1580 |
-
"loss":
|
| 1581 |
"step": 2220
|
| 1582 |
},
|
| 1583 |
{
|
| 1584 |
"epoch": 2.17,
|
| 1585 |
-
"grad_norm":
|
| 1586 |
"learning_rate": 4.815024004518498e-06,
|
| 1587 |
-
"loss": 1.
|
| 1588 |
"step": 2230
|
| 1589 |
},
|
| 1590 |
{
|
| 1591 |
"epoch": 2.17,
|
| 1592 |
-
"grad_norm":
|
| 1593 |
"learning_rate": 4.786783394521322e-06,
|
| 1594 |
-
"loss": 1.
|
| 1595 |
"step": 2240
|
| 1596 |
},
|
| 1597 |
{
|
| 1598 |
"epoch": 2.17,
|
| 1599 |
-
"grad_norm": 0.
|
| 1600 |
"learning_rate": 4.758542784524146e-06,
|
| 1601 |
-
"loss": 0.
|
| 1602 |
"step": 2250
|
| 1603 |
},
|
| 1604 |
{
|
| 1605 |
"epoch": 2.17,
|
| 1606 |
-
"grad_norm": 0.
|
| 1607 |
"learning_rate": 4.7303021745269704e-06,
|
| 1608 |
-
"loss": 0.
|
| 1609 |
"step": 2260
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 2.18,
|
| 1613 |
-
"grad_norm": 0.
|
| 1614 |
"learning_rate": 4.702061564529794e-06,
|
| 1615 |
-
"loss": 0.
|
| 1616 |
"step": 2270
|
| 1617 |
},
|
| 1618 |
{
|
| 1619 |
"epoch": 2.18,
|
| 1620 |
-
"grad_norm": 0.
|
| 1621 |
"learning_rate": 4.673820954532618e-06,
|
| 1622 |
-
"loss": 0.
|
| 1623 |
"step": 2280
|
| 1624 |
},
|
| 1625 |
{
|
| 1626 |
"epoch": 2.18,
|
| 1627 |
-
"grad_norm":
|
| 1628 |
"learning_rate": 4.6455803445354425e-06,
|
| 1629 |
-
"loss": 0.
|
| 1630 |
"step": 2290
|
| 1631 |
},
|
| 1632 |
{
|
| 1633 |
"epoch": 2.18,
|
| 1634 |
-
"grad_norm":
|
| 1635 |
"learning_rate": 4.617339734538267e-06,
|
| 1636 |
-
"loss": 0.
|
| 1637 |
"step": 2300
|
| 1638 |
},
|
| 1639 |
{
|
| 1640 |
"epoch": 2.19,
|
| 1641 |
-
"grad_norm": 0.
|
| 1642 |
"learning_rate": 4.58909912454109e-06,
|
| 1643 |
-
"loss": 0.
|
| 1644 |
"step": 2310
|
| 1645 |
},
|
| 1646 |
{
|
| 1647 |
"epoch": 2.19,
|
| 1648 |
-
"grad_norm":
|
| 1649 |
"learning_rate": 4.5608585145439145e-06,
|
| 1650 |
-
"loss": 0.
|
| 1651 |
"step": 2320
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
"epoch": 2.19,
|
| 1655 |
-
"grad_norm": 0.
|
| 1656 |
"learning_rate": 4.532617904546738e-06,
|
| 1657 |
-
"loss": 1.
|
| 1658 |
"step": 2330
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
"epoch": 2.19,
|
| 1662 |
-
"grad_norm": 0.
|
| 1663 |
"learning_rate": 4.504377294549563e-06,
|
| 1664 |
-
"loss": 0.
|
| 1665 |
"step": 2340
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 2.2,
|
| 1669 |
-
"grad_norm": 0.
|
| 1670 |
"learning_rate": 4.4761366845523865e-06,
|
| 1671 |
-
"loss": 0.
|
| 1672 |
"step": 2350
|
| 1673 |
},
|
| 1674 |
{
|
| 1675 |
"epoch": 2.2,
|
| 1676 |
-
"grad_norm": 0.
|
| 1677 |
"learning_rate": 4.447896074555211e-06,
|
| 1678 |
-
"loss": 0.
|
| 1679 |
"step": 2360
|
| 1680 |
},
|
| 1681 |
{
|
| 1682 |
"epoch": 2.2,
|
| 1683 |
-
"eval_accuracy": 0.
|
| 1684 |
-
"eval_loss": 0.
|
| 1685 |
-
"eval_runtime":
|
| 1686 |
-
"eval_samples_per_second": 1.
|
| 1687 |
-
"eval_steps_per_second": 1.
|
| 1688 |
"step": 2361
|
| 1689 |
},
|
| 1690 |
{
|
| 1691 |
"epoch": 3.0,
|
| 1692 |
-
"grad_norm": 0.
|
| 1693 |
"learning_rate": 4.419655464558035e-06,
|
| 1694 |
-
"loss": 0.
|
| 1695 |
"step": 2370
|
| 1696 |
},
|
| 1697 |
{
|
| 1698 |
"epoch": 3.0,
|
| 1699 |
-
"grad_norm":
|
| 1700 |
"learning_rate": 4.3914148545608585e-06,
|
| 1701 |
-
"loss": 0.
|
| 1702 |
"step": 2380
|
| 1703 |
},
|
| 1704 |
{
|
| 1705 |
"epoch": 3.01,
|
| 1706 |
-
"grad_norm":
|
| 1707 |
"learning_rate": 4.363174244563683e-06,
|
| 1708 |
-
"loss": 0.
|
| 1709 |
"step": 2390
|
| 1710 |
},
|
| 1711 |
{
|
| 1712 |
"epoch": 3.01,
|
| 1713 |
-
"grad_norm": 0.
|
| 1714 |
"learning_rate": 4.334933634566507e-06,
|
| 1715 |
-
"loss": 0.
|
| 1716 |
"step": 2400
|
| 1717 |
},
|
| 1718 |
{
|
| 1719 |
"epoch": 3.01,
|
| 1720 |
-
"grad_norm": 0.
|
| 1721 |
"learning_rate": 4.306693024569331e-06,
|
| 1722 |
-
"loss": 0.
|
| 1723 |
"step": 2410
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
"epoch": 3.01,
|
| 1727 |
-
"grad_norm": 0.
|
| 1728 |
"learning_rate": 4.278452414572155e-06,
|
| 1729 |
-
"loss": 1.
|
| 1730 |
"step": 2420
|
| 1731 |
},
|
| 1732 |
{
|
| 1733 |
"epoch": 3.02,
|
| 1734 |
-
"grad_norm": 0.
|
| 1735 |
"learning_rate": 4.250211804574979e-06,
|
| 1736 |
-
"loss": 0.
|
| 1737 |
"step": 2430
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 3.02,
|
| 1741 |
-
"grad_norm": 0.
|
| 1742 |
"learning_rate": 4.221971194577803e-06,
|
| 1743 |
-
"loss": 0.
|
| 1744 |
"step": 2440
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 3.02,
|
| 1748 |
-
"grad_norm": 0.
|
| 1749 |
"learning_rate": 4.193730584580628e-06,
|
| 1750 |
-
"loss": 0.
|
| 1751 |
"step": 2450
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 3.03,
|
| 1755 |
-
"grad_norm": 0.
|
| 1756 |
"learning_rate": 4.165489974583451e-06,
|
| 1757 |
-
"loss": 0.
|
| 1758 |
"step": 2460
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 3.03,
|
| 1762 |
-
"grad_norm":
|
| 1763 |
"learning_rate": 4.1372493645862754e-06,
|
| 1764 |
-
"loss": 0.
|
| 1765 |
"step": 2470
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 3.03,
|
| 1769 |
-
"grad_norm": 0.
|
| 1770 |
"learning_rate": 4.1090087545891e-06,
|
| 1771 |
-
"loss": 0.
|
| 1772 |
"step": 2480
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 3.03,
|
| 1776 |
-
"grad_norm": 0.
|
| 1777 |
"learning_rate": 4.080768144591923e-06,
|
| 1778 |
-
"loss": 1.
|
| 1779 |
"step": 2490
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 3.04,
|
| 1783 |
-
"grad_norm":
|
| 1784 |
"learning_rate": 4.0525275345947475e-06,
|
| 1785 |
-
"loss": 1.
|
| 1786 |
"step": 2500
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 3.04,
|
| 1790 |
-
"grad_norm": 0.
|
| 1791 |
"learning_rate": 4.024286924597572e-06,
|
| 1792 |
-
"loss": 2.
|
| 1793 |
"step": 2510
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 3.04,
|
| 1797 |
-
"grad_norm": 0.
|
| 1798 |
"learning_rate": 3.996046314600396e-06,
|
| 1799 |
-
"loss": 0.
|
| 1800 |
"step": 2520
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 3.04,
|
| 1804 |
-
"grad_norm": 0.
|
| 1805 |
"learning_rate": 3.9678057046032195e-06,
|
| 1806 |
-
"loss": 0.
|
| 1807 |
"step": 2530
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 3.05,
|
| 1811 |
-
"grad_norm": 0.
|
| 1812 |
"learning_rate": 3.939565094606044e-06,
|
| 1813 |
-
"loss": 0.
|
| 1814 |
"step": 2540
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 3.05,
|
| 1818 |
-
"grad_norm": 0.
|
| 1819 |
"learning_rate": 3.911324484608868e-06,
|
| 1820 |
-
"loss": 0.
|
| 1821 |
"step": 2550
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 3.05,
|
| 1825 |
-
"grad_norm": 0.
|
| 1826 |
"learning_rate": 3.8830838746116915e-06,
|
| 1827 |
-
"loss": 0.
|
| 1828 |
"step": 2560
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 3.05,
|
| 1832 |
-
"grad_norm": 0.
|
| 1833 |
"learning_rate": 3.854843264614516e-06,
|
| 1834 |
-
"loss": 1.
|
| 1835 |
"step": 2570
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 3.06,
|
| 1839 |
-
"grad_norm": 0.
|
| 1840 |
"learning_rate": 3.82660265461734e-06,
|
| 1841 |
-
"loss":
|
| 1842 |
"step": 2580
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 3.06,
|
| 1846 |
-
"grad_norm": 0.
|
| 1847 |
"learning_rate": 3.798362044620164e-06,
|
| 1848 |
-
"loss": 1.
|
| 1849 |
"step": 2590
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 3.06,
|
| 1853 |
-
"grad_norm": 0.
|
| 1854 |
"learning_rate": 3.7701214346229882e-06,
|
| 1855 |
-
"loss": 0.
|
| 1856 |
"step": 2600
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 3.06,
|
| 1860 |
-
"grad_norm":
|
| 1861 |
"learning_rate": 3.741880824625812e-06,
|
| 1862 |
-
"loss": 1.
|
| 1863 |
"step": 2610
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 3.07,
|
| 1867 |
-
"grad_norm": 0.
|
| 1868 |
"learning_rate": 3.7136402146286364e-06,
|
| 1869 |
-
"loss": 0.
|
| 1870 |
"step": 2620
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 3.07,
|
| 1874 |
-
"grad_norm": 0.
|
| 1875 |
"learning_rate": 3.6853996046314607e-06,
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 2630
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 3.07,
|
| 1881 |
-
"grad_norm": 0.
|
| 1882 |
"learning_rate": 3.6571589946342845e-06,
|
| 1883 |
-
"loss": 0.
|
| 1884 |
"step": 2640
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 3.07,
|
| 1888 |
-
"grad_norm": 0.
|
| 1889 |
"learning_rate": 3.6289183846371084e-06,
|
| 1890 |
-
"loss": 0.
|
| 1891 |
"step": 2650
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 3.08,
|
| 1895 |
-
"grad_norm": 0.
|
| 1896 |
"learning_rate": 3.6006777746399323e-06,
|
| 1897 |
-
"loss": 0.
|
| 1898 |
"step": 2660
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 3.08,
|
| 1902 |
-
"grad_norm":
|
| 1903 |
"learning_rate": 3.5724371646427566e-06,
|
| 1904 |
-
"loss":
|
| 1905 |
"step": 2670
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 3.08,
|
| 1909 |
-
"grad_norm": 0.
|
| 1910 |
"learning_rate": 3.544196554645581e-06,
|
| 1911 |
-
"loss":
|
| 1912 |
"step": 2680
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 3.08,
|
| 1916 |
-
"grad_norm": 0.
|
| 1917 |
"learning_rate": 3.5159559446484047e-06,
|
| 1918 |
-
"loss":
|
| 1919 |
"step": 2690
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 3.09,
|
| 1923 |
-
"grad_norm": 0.
|
| 1924 |
"learning_rate": 3.4877153346512286e-06,
|
| 1925 |
-
"loss": 0.
|
| 1926 |
"step": 2700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 3.09,
|
| 1930 |
-
"grad_norm": 0.
|
| 1931 |
"learning_rate": 3.459474724654053e-06,
|
| 1932 |
-
"loss": 0.
|
| 1933 |
"step": 2710
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 3.09,
|
| 1937 |
-
"grad_norm": 0.
|
| 1938 |
"learning_rate": 3.4312341146568767e-06,
|
| 1939 |
-
"loss": 0.
|
| 1940 |
"step": 2720
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 3.09,
|
| 1944 |
-
"grad_norm": 0.
|
| 1945 |
"learning_rate": 3.4029935046597006e-06,
|
| 1946 |
-
"loss": 0.
|
| 1947 |
"step": 2730
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 3.1,
|
| 1951 |
-
"grad_norm": 0.
|
| 1952 |
"learning_rate": 3.3747528946625253e-06,
|
| 1953 |
-
"loss": 0.
|
| 1954 |
"step": 2740
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 3.1,
|
| 1958 |
-
"grad_norm": 0.
|
| 1959 |
"learning_rate": 3.346512284665349e-06,
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 2750
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 3.1,
|
| 1965 |
-
"grad_norm": 0.
|
| 1966 |
"learning_rate": 3.318271674668173e-06,
|
| 1967 |
-
"loss": 0.
|
| 1968 |
"step": 2760
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 3.1,
|
| 1972 |
-
"grad_norm": 0.
|
| 1973 |
"learning_rate": 3.290031064670997e-06,
|
| 1974 |
-
"loss": 0.
|
| 1975 |
"step": 2770
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 3.11,
|
| 1979 |
-
"grad_norm": 0.
|
| 1980 |
"learning_rate": 3.2617904546738212e-06,
|
| 1981 |
-
"loss": 0.
|
| 1982 |
"step": 2780
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 3.11,
|
| 1986 |
-
"grad_norm": 0.
|
| 1987 |
"learning_rate": 3.233549844676645e-06,
|
| 1988 |
-
"loss": 0.
|
| 1989 |
"step": 2790
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 3.11,
|
| 1993 |
-
"grad_norm": 0.
|
| 1994 |
"learning_rate": 3.2053092346794694e-06,
|
| 1995 |
-
"loss": 0.
|
| 1996 |
"step": 2800
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 3.11,
|
| 2000 |
-
"grad_norm": 0.
|
| 2001 |
"learning_rate": 3.1770686246822937e-06,
|
| 2002 |
-
"loss": 0.
|
| 2003 |
"step": 2810
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 3.12,
|
| 2007 |
-
"grad_norm":
|
| 2008 |
"learning_rate": 3.1488280146851175e-06,
|
| 2009 |
-
"loss": 0.
|
| 2010 |
"step": 2820
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 3.12,
|
| 2014 |
-
"grad_norm":
|
| 2015 |
"learning_rate": 3.1205874046879414e-06,
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 2830
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 3.12,
|
| 2021 |
-
"grad_norm": 0.
|
| 2022 |
"learning_rate": 3.0923467946907653e-06,
|
| 2023 |
-
"loss":
|
| 2024 |
"step": 2840
|
| 2025 |
},
|
| 2026 |
{
|
| 2027 |
"epoch": 3.12,
|
| 2028 |
-
"grad_norm":
|
| 2029 |
"learning_rate": 3.06410618469359e-06,
|
| 2030 |
-
"loss": 0.
|
| 2031 |
"step": 2850
|
| 2032 |
},
|
| 2033 |
{
|
| 2034 |
"epoch": 3.13,
|
| 2035 |
-
"grad_norm": 0.
|
| 2036 |
"learning_rate": 3.035865574696414e-06,
|
| 2037 |
-
"loss": 0.
|
| 2038 |
"step": 2860
|
| 2039 |
},
|
| 2040 |
{
|
| 2041 |
"epoch": 3.13,
|
| 2042 |
-
"grad_norm": 0.
|
| 2043 |
"learning_rate": 3.0076249646992377e-06,
|
| 2044 |
-
"loss": 1.
|
| 2045 |
"step": 2870
|
| 2046 |
},
|
| 2047 |
{
|
| 2048 |
"epoch": 3.13,
|
| 2049 |
-
"grad_norm": 0.
|
| 2050 |
"learning_rate": 2.9793843547020616e-06,
|
| 2051 |
-
"loss": 0.
|
| 2052 |
"step": 2880
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 3.13,
|
| 2056 |
-
"grad_norm": 0.
|
| 2057 |
"learning_rate": 2.951143744704886e-06,
|
| 2058 |
-
"loss": 0.
|
| 2059 |
"step": 2890
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
"epoch": 3.14,
|
| 2063 |
-
"grad_norm":
|
| 2064 |
"learning_rate": 2.9229031347077097e-06,
|
| 2065 |
-
"loss": 1.
|
| 2066 |
"step": 2900
|
| 2067 |
},
|
| 2068 |
{
|
| 2069 |
"epoch": 3.14,
|
| 2070 |
-
"grad_norm":
|
| 2071 |
"learning_rate": 2.894662524710534e-06,
|
| 2072 |
-
"loss": 0.
|
| 2073 |
"step": 2910
|
| 2074 |
},
|
| 2075 |
{
|
| 2076 |
"epoch": 3.14,
|
| 2077 |
-
"grad_norm": 0.
|
| 2078 |
"learning_rate": 2.8664219147133583e-06,
|
| 2079 |
-
"loss": 1.
|
| 2080 |
"step": 2920
|
| 2081 |
},
|
| 2082 |
{
|
| 2083 |
"epoch": 3.14,
|
| 2084 |
-
"grad_norm": 0.
|
| 2085 |
"learning_rate": 2.838181304716182e-06,
|
| 2086 |
-
"loss": 0.
|
| 2087 |
"step": 2930
|
| 2088 |
},
|
| 2089 |
{
|
| 2090 |
"epoch": 3.15,
|
| 2091 |
-
"grad_norm":
|
| 2092 |
"learning_rate": 2.809940694719006e-06,
|
| 2093 |
-
"loss": 1.
|
| 2094 |
"step": 2940
|
| 2095 |
},
|
| 2096 |
{
|
| 2097 |
"epoch": 3.15,
|
| 2098 |
-
"grad_norm":
|
| 2099 |
"learning_rate": 2.78170008472183e-06,
|
| 2100 |
-
"loss": 0.
|
| 2101 |
"step": 2950
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
"epoch": 3.15,
|
| 2105 |
-
"grad_norm": 0.
|
| 2106 |
"learning_rate": 2.753459474724654e-06,
|
| 2107 |
-
"loss": 0.
|
| 2108 |
"step": 2960
|
| 2109 |
},
|
| 2110 |
{
|
| 2111 |
"epoch": 3.15,
|
| 2112 |
-
"grad_norm":
|
| 2113 |
"learning_rate": 2.7252188647274785e-06,
|
| 2114 |
-
"loss": 0.
|
| 2115 |
"step": 2970
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 3.16,
|
| 2119 |
-
"grad_norm": 0.
|
| 2120 |
"learning_rate": 2.6969782547303024e-06,
|
| 2121 |
-
"loss": 0.
|
| 2122 |
"step": 2980
|
| 2123 |
},
|
| 2124 |
{
|
| 2125 |
"epoch": 3.16,
|
| 2126 |
-
"grad_norm": 0.
|
| 2127 |
"learning_rate": 2.6687376447331266e-06,
|
| 2128 |
-
"loss": 0.
|
| 2129 |
"step": 2990
|
| 2130 |
},
|
| 2131 |
{
|
| 2132 |
"epoch": 3.16,
|
| 2133 |
-
"grad_norm": 0.
|
| 2134 |
"learning_rate": 2.6404970347359505e-06,
|
| 2135 |
-
"loss": 0.
|
| 2136 |
"step": 3000
|
| 2137 |
},
|
| 2138 |
{
|
| 2139 |
"epoch": 3.16,
|
| 2140 |
-
"grad_norm": 0.
|
| 2141 |
"learning_rate": 2.6122564247387744e-06,
|
| 2142 |
-
"loss": 1.
|
| 2143 |
"step": 3010
|
| 2144 |
},
|
| 2145 |
{
|
| 2146 |
"epoch": 3.17,
|
| 2147 |
-
"grad_norm": 0.
|
| 2148 |
"learning_rate": 2.584015814741599e-06,
|
| 2149 |
-
"loss": 0.
|
| 2150 |
"step": 3020
|
| 2151 |
},
|
| 2152 |
{
|
| 2153 |
"epoch": 3.17,
|
| 2154 |
-
"grad_norm": 0.
|
| 2155 |
"learning_rate": 2.555775204744423e-06,
|
| 2156 |
-
"loss": 0.
|
| 2157 |
"step": 3030
|
| 2158 |
},
|
| 2159 |
{
|
| 2160 |
"epoch": 3.17,
|
| 2161 |
-
"grad_norm": 0.
|
| 2162 |
"learning_rate": 2.527534594747247e-06,
|
| 2163 |
-
"loss": 0.
|
| 2164 |
"step": 3040
|
| 2165 |
},
|
| 2166 |
{
|
| 2167 |
"epoch": 3.18,
|
| 2168 |
-
"grad_norm": 0.
|
| 2169 |
"learning_rate": 2.4992939847500707e-06,
|
| 2170 |
-
"loss": 0.
|
| 2171 |
"step": 3050
|
| 2172 |
},
|
| 2173 |
{
|
| 2174 |
"epoch": 3.18,
|
| 2175 |
-
"grad_norm":
|
| 2176 |
"learning_rate": 2.4710533747528946e-06,
|
| 2177 |
-
"loss": 0.
|
| 2178 |
"step": 3060
|
| 2179 |
},
|
| 2180 |
{
|
| 2181 |
"epoch": 3.18,
|
| 2182 |
-
"grad_norm": 0.
|
| 2183 |
"learning_rate": 2.442812764755719e-06,
|
| 2184 |
-
"loss": 0.
|
| 2185 |
"step": 3070
|
| 2186 |
},
|
| 2187 |
{
|
| 2188 |
"epoch": 3.18,
|
| 2189 |
-
"grad_norm": 0.
|
| 2190 |
"learning_rate": 2.414572154758543e-06,
|
| 2191 |
-
"loss": 0.
|
| 2192 |
"step": 3080
|
| 2193 |
},
|
| 2194 |
{
|
| 2195 |
"epoch": 3.19,
|
| 2196 |
-
"grad_norm":
|
| 2197 |
"learning_rate": 2.386331544761367e-06,
|
| 2198 |
-
"loss": 0.
|
| 2199 |
"step": 3090
|
| 2200 |
},
|
| 2201 |
{
|
| 2202 |
"epoch": 3.19,
|
| 2203 |
-
"grad_norm": 0.
|
| 2204 |
"learning_rate": 2.3580909347641913e-06,
|
| 2205 |
-
"loss": 0.
|
| 2206 |
"step": 3100
|
| 2207 |
},
|
| 2208 |
{
|
| 2209 |
"epoch": 3.19,
|
| 2210 |
-
"grad_norm": 0.
|
| 2211 |
"learning_rate": 2.329850324767015e-06,
|
| 2212 |
-
"loss": 0.
|
| 2213 |
"step": 3110
|
| 2214 |
},
|
| 2215 |
{
|
| 2216 |
"epoch": 3.19,
|
| 2217 |
-
"grad_norm": 0.
|
| 2218 |
"learning_rate": 2.3016097147698394e-06,
|
| 2219 |
-
"loss":
|
| 2220 |
"step": 3120
|
| 2221 |
},
|
| 2222 |
{
|
| 2223 |
"epoch": 3.2,
|
| 2224 |
-
"grad_norm": 0.
|
| 2225 |
"learning_rate": 2.2733691047726633e-06,
|
| 2226 |
-
"loss": 1.
|
| 2227 |
"step": 3130
|
| 2228 |
},
|
| 2229 |
{
|
| 2230 |
"epoch": 3.2,
|
| 2231 |
-
"grad_norm": 0.
|
| 2232 |
"learning_rate": 2.245128494775487e-06,
|
| 2233 |
-
"loss": 0.
|
| 2234 |
"step": 3140
|
| 2235 |
},
|
| 2236 |
{
|
| 2237 |
"epoch": 3.2,
|
| 2238 |
-
"eval_accuracy": 0.
|
| 2239 |
-
"eval_loss": 0.
|
| 2240 |
-
"eval_runtime":
|
| 2241 |
-
"eval_samples_per_second": 1.
|
| 2242 |
-
"eval_steps_per_second": 1.
|
| 2243 |
"step": 3148
|
| 2244 |
},
|
| 2245 |
{
|
| 2246 |
"epoch": 4.0,
|
| 2247 |
-
"grad_norm": 0.
|
| 2248 |
"learning_rate": 2.2168878847783115e-06,
|
| 2249 |
-
"loss": 0.
|
| 2250 |
"step": 3150
|
| 2251 |
},
|
| 2252 |
{
|
| 2253 |
"epoch": 4.0,
|
| 2254 |
-
"grad_norm": 0.
|
| 2255 |
"learning_rate": 2.1886472747811353e-06,
|
| 2256 |
-
"loss": 0.
|
| 2257 |
"step": 3160
|
| 2258 |
},
|
| 2259 |
{
|
| 2260 |
"epoch": 4.01,
|
| 2261 |
-
"grad_norm": 0.
|
| 2262 |
"learning_rate": 2.1604066647839596e-06,
|
| 2263 |
-
"loss": 1.
|
| 2264 |
"step": 3170
|
| 2265 |
},
|
| 2266 |
{
|
| 2267 |
"epoch": 4.01,
|
| 2268 |
-
"grad_norm":
|
| 2269 |
"learning_rate": 2.1321660547867835e-06,
|
| 2270 |
-
"loss": 1.
|
| 2271 |
"step": 3180
|
| 2272 |
},
|
| 2273 |
{
|
| 2274 |
"epoch": 4.01,
|
| 2275 |
-
"grad_norm": 0.
|
| 2276 |
"learning_rate": 2.1039254447896078e-06,
|
| 2277 |
"loss": 0.0045,
|
| 2278 |
"step": 3190
|
| 2279 |
},
|
| 2280 |
{
|
| 2281 |
"epoch": 4.01,
|
| 2282 |
-
"grad_norm": 0.
|
| 2283 |
"learning_rate": 2.0756848347924316e-06,
|
| 2284 |
-
"loss": 0.
|
| 2285 |
"step": 3200
|
| 2286 |
},
|
| 2287 |
{
|
| 2288 |
"epoch": 4.02,
|
| 2289 |
-
"grad_norm": 0.
|
| 2290 |
"learning_rate": 2.047444224795256e-06,
|
| 2291 |
-
"loss": 0.
|
| 2292 |
"step": 3210
|
| 2293 |
},
|
| 2294 |
{
|
| 2295 |
"epoch": 4.02,
|
| 2296 |
-
"grad_norm": 0.
|
| 2297 |
"learning_rate": 2.01920361479808e-06,
|
| 2298 |
-
"loss": 0.
|
| 2299 |
"step": 3220
|
| 2300 |
},
|
| 2301 |
{
|
| 2302 |
"epoch": 4.02,
|
| 2303 |
-
"grad_norm": 0.
|
| 2304 |
"learning_rate": 1.9909630048009037e-06,
|
| 2305 |
-
"loss": 0.
|
| 2306 |
"step": 3230
|
| 2307 |
},
|
| 2308 |
{
|
| 2309 |
"epoch": 4.02,
|
| 2310 |
-
"grad_norm": 0.
|
| 2311 |
"learning_rate": 1.962722394803728e-06,
|
| 2312 |
-
"loss": 1.
|
| 2313 |
"step": 3240
|
| 2314 |
},
|
| 2315 |
{
|
| 2316 |
"epoch": 4.03,
|
| 2317 |
-
"grad_norm":
|
| 2318 |
"learning_rate": 1.934481784806552e-06,
|
| 2319 |
-
"loss": 0.
|
| 2320 |
"step": 3250
|
| 2321 |
},
|
| 2322 |
{
|
| 2323 |
"epoch": 4.03,
|
| 2324 |
-
"grad_norm": 0.
|
| 2325 |
"learning_rate": 1.906241174809376e-06,
|
| 2326 |
-
"loss": 0.
|
| 2327 |
"step": 3260
|
| 2328 |
},
|
| 2329 |
{
|
| 2330 |
"epoch": 4.03,
|
| 2331 |
-
"grad_norm": 0.
|
| 2332 |
"learning_rate": 1.8780005648122002e-06,
|
| 2333 |
-
"loss": 0.
|
| 2334 |
"step": 3270
|
| 2335 |
},
|
| 2336 |
{
|
| 2337 |
"epoch": 4.03,
|
| 2338 |
-
"grad_norm": 0.
|
| 2339 |
"learning_rate": 1.849759954815024e-06,
|
| 2340 |
-
"loss": 0.
|
| 2341 |
"step": 3280
|
| 2342 |
},
|
| 2343 |
{
|
| 2344 |
"epoch": 4.04,
|
| 2345 |
-
"grad_norm": 0.
|
| 2346 |
"learning_rate": 1.8215193448178481e-06,
|
| 2347 |
-
"loss": 1.
|
| 2348 |
"step": 3290
|
| 2349 |
},
|
| 2350 |
{
|
| 2351 |
"epoch": 4.04,
|
| 2352 |
-
"grad_norm": 0.
|
| 2353 |
"learning_rate": 1.7932787348206724e-06,
|
| 2354 |
-
"loss": 0.
|
| 2355 |
"step": 3300
|
| 2356 |
},
|
| 2357 |
{
|
| 2358 |
"epoch": 4.04,
|
| 2359 |
-
"grad_norm":
|
| 2360 |
"learning_rate": 1.7650381248234963e-06,
|
| 2361 |
-
"loss":
|
| 2362 |
"step": 3310
|
| 2363 |
},
|
| 2364 |
{
|
| 2365 |
"epoch": 4.04,
|
| 2366 |
-
"grad_norm": 0.
|
| 2367 |
"learning_rate": 1.7367975148263202e-06,
|
| 2368 |
-
"loss": 0.
|
| 2369 |
"step": 3320
|
| 2370 |
},
|
| 2371 |
{
|
| 2372 |
"epoch": 4.05,
|
| 2373 |
-
"grad_norm": 0.
|
| 2374 |
"learning_rate": 1.7085569048291444e-06,
|
| 2375 |
-
"loss": 0.
|
| 2376 |
"step": 3330
|
| 2377 |
},
|
| 2378 |
{
|
| 2379 |
"epoch": 4.05,
|
| 2380 |
-
"grad_norm": 0.
|
| 2381 |
"learning_rate": 1.6803162948319685e-06,
|
| 2382 |
-
"loss": 1.
|
| 2383 |
"step": 3340
|
| 2384 |
},
|
| 2385 |
{
|
| 2386 |
"epoch": 4.05,
|
| 2387 |
-
"grad_norm": 0.
|
| 2388 |
"learning_rate": 1.6520756848347926e-06,
|
| 2389 |
-
"loss": 0.
|
| 2390 |
"step": 3350
|
| 2391 |
},
|
| 2392 |
{
|
| 2393 |
"epoch": 4.05,
|
| 2394 |
-
"grad_norm": 0.
|
| 2395 |
"learning_rate": 1.6238350748376167e-06,
|
| 2396 |
-
"loss": 0.
|
| 2397 |
"step": 3360
|
| 2398 |
},
|
| 2399 |
{
|
| 2400 |
"epoch": 4.06,
|
| 2401 |
-
"grad_norm": 0.
|
| 2402 |
"learning_rate": 1.5955944648404405e-06,
|
| 2403 |
-
"loss": 0.
|
| 2404 |
"step": 3370
|
| 2405 |
},
|
| 2406 |
{
|
| 2407 |
"epoch": 4.06,
|
| 2408 |
-
"grad_norm": 0.
|
| 2409 |
"learning_rate": 1.5673538548432648e-06,
|
| 2410 |
-
"loss": 0.
|
| 2411 |
"step": 3380
|
| 2412 |
},
|
| 2413 |
{
|
| 2414 |
"epoch": 4.06,
|
| 2415 |
-
"grad_norm": 0.
|
| 2416 |
"learning_rate": 1.539113244846089e-06,
|
| 2417 |
-
"loss": 0.
|
| 2418 |
"step": 3390
|
| 2419 |
},
|
| 2420 |
{
|
| 2421 |
"epoch": 4.06,
|
| 2422 |
-
"grad_norm": 23.
|
| 2423 |
"learning_rate": 1.5108726348489128e-06,
|
| 2424 |
-
"loss": 1.
|
| 2425 |
"step": 3400
|
| 2426 |
},
|
| 2427 |
{
|
| 2428 |
"epoch": 4.07,
|
| 2429 |
-
"grad_norm":
|
| 2430 |
"learning_rate": 1.482632024851737e-06,
|
| 2431 |
-
"loss": 1.
|
| 2432 |
"step": 3410
|
| 2433 |
},
|
| 2434 |
{
|
| 2435 |
"epoch": 4.07,
|
| 2436 |
-
"grad_norm": 0.
|
| 2437 |
"learning_rate": 1.454391414854561e-06,
|
| 2438 |
-
"loss": 0.
|
| 2439 |
"step": 3420
|
| 2440 |
},
|
| 2441 |
{
|
| 2442 |
"epoch": 4.07,
|
| 2443 |
-
"grad_norm": 0.
|
| 2444 |
"learning_rate": 1.426150804857385e-06,
|
| 2445 |
-
"loss": 0.
|
| 2446 |
"step": 3430
|
| 2447 |
},
|
| 2448 |
{
|
| 2449 |
"epoch": 4.07,
|
| 2450 |
-
"grad_norm": 0.
|
| 2451 |
"learning_rate": 1.397910194860209e-06,
|
| 2452 |
-
"loss": 0.
|
| 2453 |
"step": 3440
|
| 2454 |
},
|
| 2455 |
{
|
| 2456 |
"epoch": 4.08,
|
| 2457 |
-
"grad_norm": 0.
|
| 2458 |
"learning_rate": 1.3696695848630332e-06,
|
| 2459 |
-
"loss": 0.
|
| 2460 |
"step": 3450
|
| 2461 |
},
|
| 2462 |
{
|
| 2463 |
"epoch": 4.08,
|
| 2464 |
-
"grad_norm":
|
| 2465 |
"learning_rate": 1.341428974865857e-06,
|
| 2466 |
-
"loss": 0.
|
| 2467 |
"step": 3460
|
| 2468 |
},
|
| 2469 |
{
|
| 2470 |
"epoch": 4.08,
|
| 2471 |
-
"grad_norm": 0.
|
| 2472 |
"learning_rate": 1.3131883648686813e-06,
|
| 2473 |
-
"loss": 0.
|
| 2474 |
"step": 3470
|
| 2475 |
},
|
| 2476 |
{
|
| 2477 |
"epoch": 4.08,
|
| 2478 |
-
"grad_norm":
|
| 2479 |
"learning_rate": 1.2849477548715054e-06,
|
| 2480 |
-
"loss": 1.
|
| 2481 |
"step": 3480
|
| 2482 |
},
|
| 2483 |
{
|
| 2484 |
"epoch": 4.09,
|
| 2485 |
-
"grad_norm": 0.
|
| 2486 |
"learning_rate": 1.2567071448743293e-06,
|
| 2487 |
-
"loss": 0.
|
| 2488 |
"step": 3490
|
| 2489 |
},
|
| 2490 |
{
|
| 2491 |
"epoch": 4.09,
|
| 2492 |
-
"grad_norm": 23.
|
| 2493 |
"learning_rate": 1.2284665348771536e-06,
|
| 2494 |
-
"loss": 1.
|
| 2495 |
"step": 3500
|
| 2496 |
},
|
| 2497 |
{
|
| 2498 |
"epoch": 4.09,
|
| 2499 |
-
"grad_norm": 0.
|
| 2500 |
"learning_rate": 1.2002259248799774e-06,
|
| 2501 |
-
"loss": 0.
|
| 2502 |
"step": 3510
|
| 2503 |
},
|
| 2504 |
{
|
| 2505 |
"epoch": 4.09,
|
| 2506 |
-
"grad_norm": 0.
|
| 2507 |
"learning_rate": 1.1719853148828015e-06,
|
| 2508 |
-
"loss": 0.
|
| 2509 |
"step": 3520
|
| 2510 |
},
|
| 2511 |
{
|
| 2512 |
"epoch": 4.1,
|
| 2513 |
-
"grad_norm":
|
| 2514 |
"learning_rate": 1.1437447048856256e-06,
|
| 2515 |
-
"loss": 0.
|
| 2516 |
"step": 3530
|
| 2517 |
},
|
| 2518 |
{
|
| 2519 |
"epoch": 4.1,
|
| 2520 |
-
"grad_norm": 0.
|
| 2521 |
"learning_rate": 1.1155040948884497e-06,
|
| 2522 |
-
"loss": 0.
|
| 2523 |
"step": 3540
|
| 2524 |
},
|
| 2525 |
{
|
| 2526 |
"epoch": 4.1,
|
| 2527 |
-
"grad_norm":
|
| 2528 |
"learning_rate": 1.0872634848912737e-06,
|
| 2529 |
-
"loss": 0.
|
| 2530 |
"step": 3550
|
| 2531 |
},
|
| 2532 |
{
|
| 2533 |
"epoch": 4.1,
|
| 2534 |
-
"grad_norm": 0.
|
| 2535 |
"learning_rate": 1.0590228748940978e-06,
|
| 2536 |
-
"loss": 0.
|
| 2537 |
"step": 3560
|
| 2538 |
},
|
| 2539 |
{
|
| 2540 |
"epoch": 4.11,
|
| 2541 |
-
"grad_norm": 0.
|
| 2542 |
"learning_rate": 1.0307822648969219e-06,
|
| 2543 |
-
"loss": 0.
|
| 2544 |
"step": 3570
|
| 2545 |
},
|
| 2546 |
{
|
| 2547 |
"epoch": 4.11,
|
| 2548 |
-
"grad_norm":
|
| 2549 |
"learning_rate": 1.002541654899746e-06,
|
| 2550 |
-
"loss": 0.
|
| 2551 |
"step": 3580
|
| 2552 |
},
|
| 2553 |
{
|
| 2554 |
"epoch": 4.11,
|
| 2555 |
-
"grad_norm": 0.
|
| 2556 |
"learning_rate": 9.7430104490257e-07,
|
| 2557 |
-
"loss": 0.
|
| 2558 |
"step": 3590
|
| 2559 |
},
|
| 2560 |
{
|
| 2561 |
"epoch": 4.11,
|
| 2562 |
-
"grad_norm":
|
| 2563 |
"learning_rate": 9.460604349053941e-07,
|
| 2564 |
-
"loss": 1.
|
| 2565 |
"step": 3600
|
| 2566 |
},
|
| 2567 |
{
|
| 2568 |
"epoch": 4.12,
|
| 2569 |
-
"grad_norm": 0.
|
| 2570 |
"learning_rate": 9.178198249082181e-07,
|
| 2571 |
-
"loss": 1.
|
| 2572 |
"step": 3610
|
| 2573 |
},
|
| 2574 |
{
|
| 2575 |
"epoch": 4.12,
|
| 2576 |
-
"grad_norm":
|
| 2577 |
"learning_rate": 8.895792149110422e-07,
|
| 2578 |
-
"loss": 0.
|
| 2579 |
"step": 3620
|
| 2580 |
},
|
| 2581 |
{
|
| 2582 |
"epoch": 4.12,
|
| 2583 |
-
"grad_norm":
|
| 2584 |
"learning_rate": 8.613386049138663e-07,
|
| 2585 |
-
"loss":
|
| 2586 |
"step": 3630
|
| 2587 |
},
|
| 2588 |
{
|
| 2589 |
"epoch": 4.13,
|
| 2590 |
-
"grad_norm":
|
| 2591 |
"learning_rate": 8.330979949166902e-07,
|
| 2592 |
-
"loss": 1.
|
| 2593 |
"step": 3640
|
| 2594 |
},
|
| 2595 |
{
|
| 2596 |
"epoch": 4.13,
|
| 2597 |
-
"grad_norm": 0.
|
| 2598 |
"learning_rate": 8.048573849195143e-07,
|
| 2599 |
-
"loss": 0.
|
| 2600 |
"step": 3650
|
| 2601 |
},
|
| 2602 |
{
|
| 2603 |
"epoch": 4.13,
|
| 2604 |
-
"grad_norm": 0.
|
| 2605 |
"learning_rate": 7.766167749223384e-07,
|
| 2606 |
-
"loss": 0.
|
| 2607 |
"step": 3660
|
| 2608 |
},
|
| 2609 |
{
|
| 2610 |
"epoch": 4.13,
|
| 2611 |
-
"grad_norm": 0.
|
| 2612 |
"learning_rate": 7.483761649251624e-07,
|
| 2613 |
-
"loss": 0.
|
| 2614 |
"step": 3670
|
| 2615 |
},
|
| 2616 |
{
|
| 2617 |
"epoch": 4.14,
|
| 2618 |
-
"grad_norm": 0.
|
| 2619 |
"learning_rate": 7.201355549279864e-07,
|
| 2620 |
-
"loss": 0.
|
| 2621 |
"step": 3680
|
| 2622 |
},
|
| 2623 |
{
|
| 2624 |
"epoch": 4.14,
|
| 2625 |
-
"grad_norm": 0.
|
| 2626 |
"learning_rate": 6.918949449308106e-07,
|
| 2627 |
-
"loss":
|
| 2628 |
"step": 3690
|
| 2629 |
},
|
| 2630 |
{
|
| 2631 |
"epoch": 4.14,
|
| 2632 |
-
"grad_norm":
|
| 2633 |
"learning_rate": 6.636543349336347e-07,
|
| 2634 |
-
"loss": 2.
|
| 2635 |
"step": 3700
|
| 2636 |
},
|
| 2637 |
{
|
| 2638 |
"epoch": 4.14,
|
| 2639 |
-
"grad_norm": 0.
|
| 2640 |
"learning_rate": 6.354137249364587e-07,
|
| 2641 |
-
"loss": 0.
|
| 2642 |
"step": 3710
|
| 2643 |
},
|
| 2644 |
{
|
| 2645 |
"epoch": 4.15,
|
| 2646 |
-
"grad_norm": 0.
|
| 2647 |
"learning_rate": 6.071731149392827e-07,
|
| 2648 |
-
"loss": 0.
|
| 2649 |
"step": 3720
|
| 2650 |
},
|
| 2651 |
{
|
| 2652 |
"epoch": 4.15,
|
| 2653 |
-
"grad_norm": 0.
|
| 2654 |
"learning_rate": 5.789325049421068e-07,
|
| 2655 |
-
"loss": 0.
|
| 2656 |
"step": 3730
|
| 2657 |
},
|
| 2658 |
{
|
| 2659 |
"epoch": 4.15,
|
| 2660 |
-
"grad_norm": 0.
|
| 2661 |
"learning_rate": 5.506918949449309e-07,
|
| 2662 |
-
"loss": 0.
|
| 2663 |
"step": 3740
|
| 2664 |
},
|
| 2665 |
{
|
| 2666 |
"epoch": 4.15,
|
| 2667 |
-
"grad_norm":
|
| 2668 |
"learning_rate": 5.224512849477549e-07,
|
| 2669 |
-
"loss":
|
| 2670 |
"step": 3750
|
| 2671 |
},
|
| 2672 |
{
|
| 2673 |
"epoch": 4.16,
|
| 2674 |
-
"grad_norm": 0.
|
| 2675 |
"learning_rate": 4.942106749505791e-07,
|
| 2676 |
-
"loss": 0.
|
| 2677 |
"step": 3760
|
| 2678 |
},
|
| 2679 |
{
|
| 2680 |
"epoch": 4.16,
|
| 2681 |
-
"grad_norm": 0.
|
| 2682 |
"learning_rate": 4.6597006495340303e-07,
|
| 2683 |
-
"loss": 0.
|
| 2684 |
"step": 3770
|
| 2685 |
},
|
| 2686 |
{
|
| 2687 |
"epoch": 4.16,
|
| 2688 |
-
"grad_norm":
|
| 2689 |
"learning_rate": 4.3772945495622706e-07,
|
| 2690 |
-
"loss": 1.
|
| 2691 |
"step": 3780
|
| 2692 |
},
|
| 2693 |
{
|
| 2694 |
"epoch": 4.16,
|
| 2695 |
-
"grad_norm": 0.
|
| 2696 |
"learning_rate": 4.094888449590512e-07,
|
| 2697 |
-
"loss": 0.
|
| 2698 |
"step": 3790
|
| 2699 |
},
|
| 2700 |
{
|
| 2701 |
"epoch": 4.17,
|
| 2702 |
-
"grad_norm": 0.
|
| 2703 |
"learning_rate": 3.812482349618752e-07,
|
| 2704 |
-
"loss": 0.
|
| 2705 |
"step": 3800
|
| 2706 |
},
|
| 2707 |
{
|
| 2708 |
"epoch": 4.17,
|
| 2709 |
-
"grad_norm": 0.
|
| 2710 |
"learning_rate": 3.530076249646993e-07,
|
| 2711 |
-
"loss": 0.
|
| 2712 |
"step": 3810
|
| 2713 |
},
|
| 2714 |
{
|
| 2715 |
"epoch": 4.17,
|
| 2716 |
-
"grad_norm": 0.
|
| 2717 |
"learning_rate": 3.247670149675233e-07,
|
| 2718 |
-
"loss": 0.
|
| 2719 |
"step": 3820
|
| 2720 |
},
|
| 2721 |
{
|
| 2722 |
"epoch": 4.17,
|
| 2723 |
-
"grad_norm": 0.
|
| 2724 |
"learning_rate": 2.965264049703474e-07,
|
| 2725 |
-
"loss": 0.
|
| 2726 |
"step": 3830
|
| 2727 |
},
|
| 2728 |
{
|
| 2729 |
"epoch": 4.18,
|
| 2730 |
-
"grad_norm": 0.
|
| 2731 |
"learning_rate": 2.682857949731714e-07,
|
| 2732 |
-
"loss": 0.
|
| 2733 |
"step": 3840
|
| 2734 |
},
|
| 2735 |
{
|
| 2736 |
"epoch": 4.18,
|
| 2737 |
-
"grad_norm": 0.
|
| 2738 |
"learning_rate": 2.400451849759955e-07,
|
| 2739 |
-
"loss": 0.
|
| 2740 |
"step": 3850
|
| 2741 |
},
|
| 2742 |
{
|
| 2743 |
"epoch": 4.18,
|
| 2744 |
-
"grad_norm": 0.
|
| 2745 |
"learning_rate": 2.1180457497881955e-07,
|
| 2746 |
-
"loss": 0.
|
| 2747 |
"step": 3860
|
| 2748 |
},
|
| 2749 |
{
|
| 2750 |
"epoch": 4.18,
|
| 2751 |
-
"grad_norm":
|
| 2752 |
"learning_rate": 1.8356396498164363e-07,
|
| 2753 |
-
"loss": 1.
|
| 2754 |
"step": 3870
|
| 2755 |
},
|
| 2756 |
{
|
| 2757 |
"epoch": 4.19,
|
| 2758 |
-
"grad_norm": 0.
|
| 2759 |
"learning_rate": 1.5532335498446768e-07,
|
| 2760 |
-
"loss": 1.
|
| 2761 |
"step": 3880
|
| 2762 |
},
|
| 2763 |
{
|
| 2764 |
"epoch": 4.19,
|
| 2765 |
-
"grad_norm": 0.
|
| 2766 |
"learning_rate": 1.2708274498729173e-07,
|
| 2767 |
-
"loss": 1.
|
| 2768 |
"step": 3890
|
| 2769 |
},
|
| 2770 |
{
|
| 2771 |
"epoch": 4.19,
|
| 2772 |
-
"grad_norm": 0.
|
| 2773 |
"learning_rate": 9.884213499011579e-08,
|
| 2774 |
-
"loss": 0.
|
| 2775 |
"step": 3900
|
| 2776 |
},
|
| 2777 |
{
|
| 2778 |
"epoch": 4.19,
|
| 2779 |
-
"grad_norm":
|
| 2780 |
"learning_rate": 7.060152499293986e-08,
|
| 2781 |
-
"loss": 1.
|
| 2782 |
"step": 3910
|
| 2783 |
},
|
| 2784 |
{
|
| 2785 |
"epoch": 4.2,
|
| 2786 |
-
"grad_norm": 0.
|
| 2787 |
"learning_rate": 4.236091499576391e-08,
|
| 2788 |
-
"loss": 0.
|
| 2789 |
"step": 3920
|
| 2790 |
},
|
| 2791 |
{
|
| 2792 |
"epoch": 4.2,
|
| 2793 |
-
"grad_norm": 0.
|
| 2794 |
"learning_rate": 1.4120304998587971e-08,
|
| 2795 |
-
"loss": 0.
|
| 2796 |
"step": 3930
|
| 2797 |
},
|
| 2798 |
{
|
| 2799 |
"epoch": 4.2,
|
| 2800 |
-
"eval_accuracy": 0.
|
| 2801 |
-
"eval_loss": 0.
|
| 2802 |
-
"eval_runtime": 327.
|
| 2803 |
-
"eval_samples_per_second": 1.
|
| 2804 |
-
"eval_steps_per_second": 1.
|
| 2805 |
"step": 3935
|
| 2806 |
},
|
| 2807 |
{
|
| 2808 |
"epoch": 4.2,
|
| 2809 |
"step": 3935,
|
| 2810 |
"total_flos": 1.7278688550692782e+19,
|
| 2811 |
-
"train_loss": 0.
|
| 2812 |
-
"train_runtime":
|
| 2813 |
-
"train_samples_per_second": 0.
|
| 2814 |
-
"train_steps_per_second": 0.
|
| 2815 |
},
|
| 2816 |
{
|
| 2817 |
"epoch": 4.2,
|
| 2818 |
-
"eval_accuracy": 0.
|
| 2819 |
-
"eval_loss": 0.
|
| 2820 |
-
"eval_runtime":
|
| 2821 |
-
"eval_samples_per_second": 1.
|
| 2822 |
-
"eval_steps_per_second": 1.
|
| 2823 |
"step": 3935
|
| 2824 |
},
|
| 2825 |
{
|
| 2826 |
"epoch": 4.2,
|
| 2827 |
-
"eval_accuracy": 0.
|
| 2828 |
-
"eval_loss": 0.
|
| 2829 |
-
"eval_runtime":
|
| 2830 |
-
"eval_samples_per_second": 1.
|
| 2831 |
-
"eval_steps_per_second": 1.
|
| 2832 |
"step": 3935
|
| 2833 |
}
|
| 2834 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8456375838926175,
|
| 3 |
+
"best_model_checkpoint": "ctsinov1/checkpoint-3148",
|
| 4 |
"epoch": 4.2,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 3935,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
+
"grad_norm": 6.577972888946533,
|
| 14 |
"learning_rate": 2.538071065989848e-07,
|
| 15 |
+
"loss": 0.7076,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 6.098381519317627,
|
| 21 |
"learning_rate": 5.076142131979696e-07,
|
| 22 |
+
"loss": 0.7072,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01,
|
| 27 |
+
"grad_norm": 6.236210346221924,
|
| 28 |
"learning_rate": 7.614213197969544e-07,
|
| 29 |
+
"loss": 0.7278,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01,
|
| 34 |
+
"grad_norm": 7.610434532165527,
|
| 35 |
"learning_rate": 1.0152284263959392e-06,
|
| 36 |
+
"loss": 0.6914,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.01,
|
| 41 |
+
"grad_norm": 5.927430152893066,
|
| 42 |
"learning_rate": 1.2690355329949238e-06,
|
| 43 |
+
"loss": 0.7105,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02,
|
| 48 |
+
"grad_norm": 22.82501220703125,
|
| 49 |
"learning_rate": 1.5228426395939088e-06,
|
| 50 |
+
"loss": 0.694,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02,
|
| 55 |
+
"grad_norm": 24.948528289794922,
|
| 56 |
"learning_rate": 1.7766497461928936e-06,
|
| 57 |
+
"loss": 0.7014,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.02,
|
| 62 |
+
"grad_norm": 12.509531021118164,
|
| 63 |
"learning_rate": 2.0304568527918785e-06,
|
| 64 |
+
"loss": 0.6915,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.02,
|
| 69 |
+
"grad_norm": 10.74423885345459,
|
| 70 |
"learning_rate": 2.284263959390863e-06,
|
| 71 |
+
"loss": 0.6788,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03,
|
| 76 |
+
"grad_norm": 16.888263702392578,
|
| 77 |
"learning_rate": 2.5380710659898476e-06,
|
| 78 |
+
"loss": 0.6601,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.03,
|
| 83 |
+
"grad_norm": 10.467493057250977,
|
| 84 |
"learning_rate": 2.7918781725888327e-06,
|
| 85 |
+
"loss": 0.6162,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.03,
|
| 90 |
+
"grad_norm": 13.966683387756348,
|
| 91 |
"learning_rate": 3.0456852791878177e-06,
|
| 92 |
+
"loss": 0.5912,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.03,
|
| 97 |
+
"grad_norm": 5.591104030609131,
|
| 98 |
"learning_rate": 3.2994923857868023e-06,
|
| 99 |
+
"loss": 0.4726,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.04,
|
| 104 |
+
"grad_norm": 8.37885856628418,
|
| 105 |
"learning_rate": 3.5532994923857873e-06,
|
| 106 |
+
"loss": 1.0922,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.04,
|
| 111 |
+
"grad_norm": 24.369611740112305,
|
| 112 |
"learning_rate": 3.8071065989847715e-06,
|
| 113 |
+
"loss": 1.0916,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.04,
|
| 118 |
+
"grad_norm": 12.621369361877441,
|
| 119 |
"learning_rate": 4.060913705583757e-06,
|
| 120 |
+
"loss": 0.6395,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.04,
|
| 125 |
+
"grad_norm": 7.74204683303833,
|
| 126 |
"learning_rate": 4.3147208121827415e-06,
|
| 127 |
+
"loss": 0.6146,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.05,
|
| 132 |
+
"grad_norm": 29.357898712158203,
|
| 133 |
"learning_rate": 4.568527918781726e-06,
|
| 134 |
+
"loss": 0.6861,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.05,
|
| 139 |
+
"grad_norm": 14.954898834228516,
|
| 140 |
"learning_rate": 4.822335025380711e-06,
|
| 141 |
+
"loss": 0.7809,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.05,
|
| 146 |
+
"grad_norm": 12.208268165588379,
|
| 147 |
"learning_rate": 5.076142131979695e-06,
|
| 148 |
+
"loss": 0.5812,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.05,
|
| 153 |
+
"grad_norm": 75.62841033935547,
|
| 154 |
"learning_rate": 5.329949238578681e-06,
|
| 155 |
+
"loss": 0.7408,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.06,
|
| 160 |
+
"grad_norm": 17.70330047607422,
|
| 161 |
"learning_rate": 5.583756345177665e-06,
|
| 162 |
+
"loss": 0.8077,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.06,
|
| 167 |
+
"grad_norm": 26.325668334960938,
|
| 168 |
"learning_rate": 5.83756345177665e-06,
|
| 169 |
+
"loss": 0.6845,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.06,
|
| 174 |
+
"grad_norm": 30.602115631103516,
|
| 175 |
"learning_rate": 6.091370558375635e-06,
|
| 176 |
+
"loss": 0.7371,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.06,
|
| 181 |
+
"grad_norm": 17.599937438964844,
|
| 182 |
"learning_rate": 6.34517766497462e-06,
|
| 183 |
+
"loss": 0.6351,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.07,
|
| 188 |
+
"grad_norm": 27.942121505737305,
|
| 189 |
"learning_rate": 6.5989847715736045e-06,
|
| 190 |
+
"loss": 0.6677,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.07,
|
| 195 |
+
"grad_norm": 22.718555450439453,
|
| 196 |
"learning_rate": 6.852791878172589e-06,
|
| 197 |
+
"loss": 0.6294,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.07,
|
| 202 |
+
"grad_norm": 15.279131889343262,
|
| 203 |
"learning_rate": 7.106598984771575e-06,
|
| 204 |
+
"loss": 0.7051,
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.07,
|
| 209 |
+
"grad_norm": 2.718256950378418,
|
| 210 |
"learning_rate": 7.360406091370559e-06,
|
| 211 |
+
"loss": 0.2624,
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.08,
|
| 216 |
+
"grad_norm": 7.366211414337158,
|
| 217 |
"learning_rate": 7.614213197969543e-06,
|
| 218 |
+
"loss": 1.0217,
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.08,
|
| 223 |
+
"grad_norm": 88.110595703125,
|
| 224 |
"learning_rate": 7.86802030456853e-06,
|
| 225 |
+
"loss": 0.97,
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.08,
|
| 230 |
+
"grad_norm": 8.04466438293457,
|
| 231 |
"learning_rate": 8.121827411167514e-06,
|
| 232 |
+
"loss": 0.7516,
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.08,
|
| 237 |
+
"grad_norm": 9.804755210876465,
|
| 238 |
"learning_rate": 8.375634517766498e-06,
|
| 239 |
+
"loss": 0.6807,
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.09,
|
| 244 |
+
"grad_norm": 8.710782051086426,
|
| 245 |
"learning_rate": 8.629441624365483e-06,
|
| 246 |
+
"loss": 0.431,
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.09,
|
| 251 |
+
"grad_norm": 6.989062786102295,
|
| 252 |
"learning_rate": 8.883248730964468e-06,
|
| 253 |
+
"loss": 0.4437,
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.09,
|
| 258 |
+
"grad_norm": 0.23465396463871002,
|
| 259 |
"learning_rate": 9.137055837563452e-06,
|
| 260 |
+
"loss": 0.6021,
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.09,
|
| 265 |
+
"grad_norm": 4.12635612487793,
|
| 266 |
"learning_rate": 9.390862944162438e-06,
|
| 267 |
+
"loss": 1.1584,
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1,
|
| 272 |
+
"grad_norm": 67.80364990234375,
|
| 273 |
"learning_rate": 9.644670050761421e-06,
|
| 274 |
+
"loss": 1.3129,
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.1,
|
| 279 |
+
"grad_norm": 1.1369630098342896,
|
| 280 |
"learning_rate": 9.898477157360406e-06,
|
| 281 |
+
"loss": 0.4667,
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.1,
|
| 286 |
+
"grad_norm": 0.777511477470398,
|
| 287 |
"learning_rate": 9.983055634001694e-06,
|
| 288 |
+
"loss": 1.0065,
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.1,
|
| 293 |
+
"grad_norm": 5.83087158203125,
|
| 294 |
"learning_rate": 9.95481502400452e-06,
|
| 295 |
+
"loss": 1.6977,
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.11,
|
| 300 |
+
"grad_norm": 0.4953118562698364,
|
| 301 |
"learning_rate": 9.926574414007344e-06,
|
| 302 |
+
"loss": 0.7726,
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.11,
|
| 307 |
+
"grad_norm": 15.068619728088379,
|
| 308 |
"learning_rate": 9.898333804010167e-06,
|
| 309 |
+
"loss": 0.6931,
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.11,
|
| 314 |
+
"grad_norm": 9.647561073303223,
|
| 315 |
"learning_rate": 9.870093194012991e-06,
|
| 316 |
+
"loss": 0.6504,
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.11,
|
| 321 |
+
"grad_norm": 5.142507076263428,
|
| 322 |
"learning_rate": 9.841852584015815e-06,
|
| 323 |
+
"loss": 0.4246,
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.12,
|
| 328 |
+
"grad_norm": 44.15083694458008,
|
| 329 |
"learning_rate": 9.81361197401864e-06,
|
| 330 |
+
"loss": 1.5019,
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.12,
|
| 335 |
+
"grad_norm": 73.28787231445312,
|
| 336 |
"learning_rate": 9.785371364021462e-06,
|
| 337 |
+
"loss": 0.9851,
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.12,
|
| 342 |
+
"grad_norm": 5.464163780212402,
|
| 343 |
"learning_rate": 9.757130754024288e-06,
|
| 344 |
+
"loss": 0.8729,
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.12,
|
| 349 |
+
"grad_norm": 0.4673428535461426,
|
| 350 |
"learning_rate": 9.728890144027113e-06,
|
| 351 |
+
"loss": 1.3793,
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.13,
|
| 356 |
+
"grad_norm": 68.62584686279297,
|
| 357 |
"learning_rate": 9.700649534029935e-06,
|
| 358 |
+
"loss": 1.7333,
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.13,
|
| 363 |
+
"grad_norm": 43.09763717651367,
|
| 364 |
"learning_rate": 9.67240892403276e-06,
|
| 365 |
+
"loss": 0.9818,
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.13,
|
| 370 |
+
"grad_norm": 1.3706504106521606,
|
| 371 |
"learning_rate": 9.644168314035584e-06,
|
| 372 |
+
"loss": 0.6076,
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.13,
|
| 377 |
+
"grad_norm": 0.3608866333961487,
|
| 378 |
"learning_rate": 9.615927704038408e-06,
|
| 379 |
+
"loss": 0.011,
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.14,
|
| 384 |
+
"grad_norm": 2.5606987476348877,
|
| 385 |
"learning_rate": 9.587687094041232e-06,
|
| 386 |
+
"loss": 1.1129,
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.14,
|
| 391 |
+
"grad_norm": 0.4630231559276581,
|
| 392 |
"learning_rate": 9.559446484044057e-06,
|
| 393 |
+
"loss": 0.5555,
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.14,
|
| 398 |
+
"grad_norm": 2.497171401977539,
|
| 399 |
"learning_rate": 9.531205874046881e-06,
|
| 400 |
+
"loss": 0.5851,
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.14,
|
| 405 |
+
"grad_norm": 72.00778198242188,
|
| 406 |
"learning_rate": 9.502965264049704e-06,
|
| 407 |
+
"loss": 1.1757,
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.15,
|
| 412 |
+
"grad_norm": 51.60250473022461,
|
| 413 |
"learning_rate": 9.474724654052528e-06,
|
| 414 |
+
"loss": 0.5766,
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.15,
|
| 419 |
+
"grad_norm": 0.111382856965065,
|
| 420 |
"learning_rate": 9.446484044055352e-06,
|
| 421 |
+
"loss": 0.0039,
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.15,
|
| 426 |
+
"grad_norm": 0.24558886885643005,
|
| 427 |
"learning_rate": 9.418243434058176e-06,
|
| 428 |
+
"loss": 0.0022,
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.16,
|
| 433 |
+
"grad_norm": 0.46412792801856995,
|
| 434 |
"learning_rate": 9.390002824061e-06,
|
| 435 |
+
"loss": 2.3804,
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.16,
|
| 440 |
+
"grad_norm": 0.6763355731964111,
|
| 441 |
"learning_rate": 9.361762214063825e-06,
|
| 442 |
+
"loss": 1.9181,
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.16,
|
| 447 |
+
"grad_norm": 0.18327564001083374,
|
| 448 |
"learning_rate": 9.33352160406665e-06,
|
| 449 |
+
"loss": 1.68,
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.16,
|
| 454 |
+
"grad_norm": 0.0858134999871254,
|
| 455 |
"learning_rate": 9.305280994069472e-06,
|
| 456 |
+
"loss": 0.4881,
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.17,
|
| 461 |
+
"grad_norm": 43.90891647338867,
|
| 462 |
"learning_rate": 9.277040384072296e-06,
|
| 463 |
+
"loss": 0.9335,
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.17,
|
| 468 |
+
"grad_norm": 0.740903913974762,
|
| 469 |
"learning_rate": 9.24879977407512e-06,
|
| 470 |
+
"loss": 1.8727,
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.17,
|
| 475 |
+
"grad_norm": 41.42824172973633,
|
| 476 |
"learning_rate": 9.220559164077945e-06,
|
| 477 |
+
"loss": 1.0477,
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.17,
|
| 482 |
+
"grad_norm": 0.6841148138046265,
|
| 483 |
"learning_rate": 9.192318554080769e-06,
|
| 484 |
+
"loss": 0.0093,
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.18,
|
| 489 |
+
"grad_norm": 0.15574102103710175,
|
| 490 |
"learning_rate": 9.164077944083593e-06,
|
| 491 |
+
"loss": 0.0061,
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.18,
|
| 496 |
+
"grad_norm": 30.2320613861084,
|
| 497 |
"learning_rate": 9.135837334086418e-06,
|
| 498 |
+
"loss": 1.2614,
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.18,
|
| 503 |
+
"grad_norm": 0.1653001755475998,
|
| 504 |
"learning_rate": 9.107596724089242e-06,
|
| 505 |
+
"loss": 1.4186,
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.18,
|
| 510 |
+
"grad_norm": 5.986731052398682,
|
| 511 |
"learning_rate": 9.079356114092065e-06,
|
| 512 |
+
"loss": 0.5271,
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.19,
|
| 517 |
+
"grad_norm": 0.5438697338104248,
|
| 518 |
"learning_rate": 9.051115504094889e-06,
|
| 519 |
+
"loss": 1.6235,
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.19,
|
| 524 |
+
"grad_norm": 132.91490173339844,
|
| 525 |
"learning_rate": 9.022874894097713e-06,
|
| 526 |
+
"loss": 0.8793,
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.19,
|
| 531 |
+
"grad_norm": 222.04197692871094,
|
| 532 |
"learning_rate": 8.994634284100537e-06,
|
| 533 |
+
"loss": 0.0887,
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.19,
|
| 538 |
+
"grad_norm": 55.007293701171875,
|
| 539 |
"learning_rate": 8.96639367410336e-06,
|
| 540 |
+
"loss": 3.5521,
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.2,
|
| 545 |
+
"grad_norm": 43.21747589111328,
|
| 546 |
"learning_rate": 8.938153064106186e-06,
|
| 547 |
+
"loss": 0.8727,
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.2,
|
| 552 |
+
"grad_norm": 52.716983795166016,
|
| 553 |
"learning_rate": 8.90991245410901e-06,
|
| 554 |
+
"loss": 1.8305,
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.2,
|
| 559 |
+
"eval_accuracy": 0.6023489932885906,
|
| 560 |
+
"eval_loss": 1.2154779434204102,
|
| 561 |
+
"eval_runtime": 323.3556,
|
| 562 |
+
"eval_samples_per_second": 1.843,
|
| 563 |
+
"eval_steps_per_second": 1.843,
|
| 564 |
"step": 787
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 1.0,
|
| 568 |
+
"grad_norm": 130.62637329101562,
|
| 569 |
"learning_rate": 8.881671844111833e-06,
|
| 570 |
+
"loss": 1.5259,
|
| 571 |
"step": 790
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 1.0,
|
| 575 |
+
"grad_norm": 28.14907455444336,
|
| 576 |
"learning_rate": 8.853431234114657e-06,
|
| 577 |
+
"loss": 0.9381,
|
| 578 |
"step": 800
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 1.01,
|
| 582 |
+
"grad_norm": 4.458512783050537,
|
| 583 |
"learning_rate": 8.825190624117481e-06,
|
| 584 |
+
"loss": 0.6209,
|
| 585 |
"step": 810
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 1.01,
|
| 589 |
+
"grad_norm": 2.9990813732147217,
|
| 590 |
"learning_rate": 8.796950014120306e-06,
|
| 591 |
+
"loss": 0.3972,
|
| 592 |
"step": 820
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 1.01,
|
| 596 |
+
"grad_norm": 317.2392272949219,
|
| 597 |
"learning_rate": 8.76870940412313e-06,
|
| 598 |
+
"loss": 1.7995,
|
| 599 |
"step": 830
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 1.01,
|
| 603 |
+
"grad_norm": 0.08616624772548676,
|
| 604 |
"learning_rate": 8.740468794125954e-06,
|
| 605 |
+
"loss": 0.94,
|
| 606 |
"step": 840
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 1.02,
|
| 610 |
+
"grad_norm": 0.08287107199430466,
|
| 611 |
"learning_rate": 8.712228184128779e-06,
|
| 612 |
+
"loss": 0.0513,
|
| 613 |
"step": 850
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 1.02,
|
| 617 |
+
"grad_norm": 177.66172790527344,
|
| 618 |
"learning_rate": 8.683987574131601e-06,
|
| 619 |
+
"loss": 0.7323,
|
| 620 |
"step": 860
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 1.02,
|
| 624 |
+
"grad_norm": 79.82263946533203,
|
| 625 |
"learning_rate": 8.655746964134425e-06,
|
| 626 |
+
"loss": 1.6152,
|
| 627 |
"step": 870
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 1.02,
|
| 631 |
+
"grad_norm": 0.3191066086292267,
|
| 632 |
"learning_rate": 8.62750635413725e-06,
|
| 633 |
+
"loss": 1.4359,
|
| 634 |
"step": 880
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 1.03,
|
| 638 |
+
"grad_norm": 172.41537475585938,
|
| 639 |
"learning_rate": 8.599265744140074e-06,
|
| 640 |
+
"loss": 1.473,
|
| 641 |
"step": 890
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 1.03,
|
| 645 |
+
"grad_norm": 9.612916946411133,
|
| 646 |
"learning_rate": 8.571025134142898e-06,
|
| 647 |
+
"loss": 1.0049,
|
| 648 |
"step": 900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 1.03,
|
| 652 |
+
"grad_norm": 27.31446647644043,
|
| 653 |
"learning_rate": 8.542784524145723e-06,
|
| 654 |
+
"loss": 0.7613,
|
| 655 |
"step": 910
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 1.03,
|
| 659 |
+
"grad_norm": 0.6966761350631714,
|
| 660 |
"learning_rate": 8.514543914148547e-06,
|
| 661 |
+
"loss": 0.3961,
|
| 662 |
"step": 920
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 1.04,
|
| 666 |
+
"grad_norm": 0.1365099400281906,
|
| 667 |
"learning_rate": 8.48630330415137e-06,
|
| 668 |
+
"loss": 0.1293,
|
| 669 |
"step": 930
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 1.04,
|
| 673 |
+
"grad_norm": 0.0884900763630867,
|
| 674 |
"learning_rate": 8.458062694154194e-06,
|
| 675 |
+
"loss": 0.5977,
|
| 676 |
"step": 940
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 1.04,
|
| 680 |
+
"grad_norm": 2.8027291297912598,
|
| 681 |
"learning_rate": 8.429822084157018e-06,
|
| 682 |
+
"loss": 1.2093,
|
| 683 |
"step": 950
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 1.04,
|
| 687 |
+
"grad_norm": 75.59962463378906,
|
| 688 |
"learning_rate": 8.401581474159842e-06,
|
| 689 |
+
"loss": 0.4025,
|
| 690 |
"step": 960
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 1.05,
|
| 694 |
+
"grad_norm": 2.279308795928955,
|
| 695 |
"learning_rate": 8.373340864162667e-06,
|
| 696 |
+
"loss": 2.1111,
|
| 697 |
"step": 970
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 1.05,
|
| 701 |
+
"grad_norm": 0.3451221287250519,
|
| 702 |
"learning_rate": 8.345100254165491e-06,
|
| 703 |
+
"loss": 0.8048,
|
| 704 |
"step": 980
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 1.05,
|
| 708 |
+
"grad_norm": 0.8533955812454224,
|
| 709 |
"learning_rate": 8.316859644168315e-06,
|
| 710 |
+
"loss": 0.9632,
|
| 711 |
"step": 990
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 1.05,
|
| 715 |
+
"grad_norm": 55.09623718261719,
|
| 716 |
"learning_rate": 8.28861903417114e-06,
|
| 717 |
+
"loss": 0.5962,
|
| 718 |
"step": 1000
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 1.06,
|
| 722 |
+
"grad_norm": 0.5793609023094177,
|
| 723 |
"learning_rate": 8.260378424173962e-06,
|
| 724 |
+
"loss": 0.8425,
|
| 725 |
"step": 1010
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 1.06,
|
| 729 |
+
"grad_norm": 0.020316464826464653,
|
| 730 |
"learning_rate": 8.232137814176786e-06,
|
| 731 |
+
"loss": 0.4848,
|
| 732 |
"step": 1020
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 1.06,
|
| 736 |
+
"grad_norm": 0.025828547775745392,
|
| 737 |
"learning_rate": 8.20389720417961e-06,
|
| 738 |
+
"loss": 0.4477,
|
| 739 |
"step": 1030
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 1.06,
|
| 743 |
+
"grad_norm": 0.02087996155023575,
|
| 744 |
"learning_rate": 8.175656594182435e-06,
|
| 745 |
+
"loss": 0.391,
|
| 746 |
"step": 1040
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 1.07,
|
| 750 |
+
"grad_norm": 172.80055236816406,
|
| 751 |
"learning_rate": 8.14741598418526e-06,
|
| 752 |
+
"loss": 1.9783,
|
| 753 |
"step": 1050
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 1.07,
|
| 757 |
+
"grad_norm": 0.15524566173553467,
|
| 758 |
"learning_rate": 8.119175374188084e-06,
|
| 759 |
+
"loss": 0.3193,
|
| 760 |
"step": 1060
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 1.07,
|
| 764 |
+
"grad_norm": 0.34669268131256104,
|
| 765 |
"learning_rate": 8.090934764190908e-06,
|
| 766 |
+
"loss": 0.5599,
|
| 767 |
"step": 1070
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 1.07,
|
| 771 |
+
"grad_norm": 0.14205870032310486,
|
| 772 |
"learning_rate": 8.06269415419373e-06,
|
| 773 |
+
"loss": 1.6577,
|
| 774 |
"step": 1080
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 1.08,
|
| 778 |
+
"grad_norm": 0.0540863536298275,
|
| 779 |
"learning_rate": 8.034453544196555e-06,
|
| 780 |
+
"loss": 0.4903,
|
| 781 |
"step": 1090
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 1.08,
|
| 785 |
+
"grad_norm": 3.833791494369507,
|
| 786 |
"learning_rate": 8.006212934199379e-06,
|
| 787 |
+
"loss": 0.3037,
|
| 788 |
"step": 1100
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 1.08,
|
| 792 |
+
"grad_norm": 37.181514739990234,
|
| 793 |
"learning_rate": 7.977972324202203e-06,
|
| 794 |
+
"loss": 1.1881,
|
| 795 |
"step": 1110
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 1.08,
|
| 799 |
+
"grad_norm": 66.44300079345703,
|
| 800 |
"learning_rate": 7.949731714205028e-06,
|
| 801 |
+
"loss": 0.5832,
|
| 802 |
"step": 1120
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 1.09,
|
| 806 |
+
"grad_norm": 154.6601104736328,
|
| 807 |
"learning_rate": 7.921491104207852e-06,
|
| 808 |
+
"loss": 2.519,
|
| 809 |
"step": 1130
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 1.09,
|
| 813 |
+
"grad_norm": 30.22860336303711,
|
| 814 |
"learning_rate": 7.893250494210676e-06,
|
| 815 |
+
"loss": 0.9667,
|
| 816 |
"step": 1140
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 1.09,
|
| 820 |
+
"grad_norm": 0.13177183270454407,
|
| 821 |
"learning_rate": 7.865009884213499e-06,
|
| 822 |
+
"loss": 1.0848,
|
| 823 |
"step": 1150
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 1.09,
|
| 827 |
+
"grad_norm": 39.34396743774414,
|
| 828 |
"learning_rate": 7.836769274216323e-06,
|
| 829 |
+
"loss": 0.5417,
|
| 830 |
"step": 1160
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 1.1,
|
| 834 |
+
"grad_norm": 0.5694972276687622,
|
| 835 |
"learning_rate": 7.808528664219147e-06,
|
| 836 |
+
"loss": 1.0765,
|
| 837 |
"step": 1170
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 1.1,
|
| 841 |
+
"grad_norm": 0.16328032314777374,
|
| 842 |
"learning_rate": 7.780288054221972e-06,
|
| 843 |
+
"loss": 0.01,
|
| 844 |
"step": 1180
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 1.1,
|
| 848 |
+
"grad_norm": 0.15245263278484344,
|
| 849 |
"learning_rate": 7.752047444224796e-06,
|
| 850 |
+
"loss": 0.6003,
|
| 851 |
"step": 1190
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 1.1,
|
| 855 |
+
"grad_norm": 0.11677064746618271,
|
| 856 |
"learning_rate": 7.72380683422762e-06,
|
| 857 |
+
"loss": 0.0037,
|
| 858 |
"step": 1200
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 1.11,
|
| 862 |
+
"grad_norm": 0.06681229919195175,
|
| 863 |
"learning_rate": 7.695566224230445e-06,
|
| 864 |
+
"loss": 0.0045,
|
| 865 |
"step": 1210
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 1.11,
|
| 869 |
+
"grad_norm": 0.03700481355190277,
|
| 870 |
"learning_rate": 7.667325614233269e-06,
|
| 871 |
+
"loss": 0.6897,
|
| 872 |
"step": 1220
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 1.11,
|
| 876 |
+
"grad_norm": 0.4864102005958557,
|
| 877 |
"learning_rate": 7.639085004236091e-06,
|
| 878 |
+
"loss": 2.1175,
|
| 879 |
"step": 1230
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 1.12,
|
| 883 |
+
"grad_norm": 0.16693036258220673,
|
| 884 |
"learning_rate": 7.610844394238917e-06,
|
| 885 |
+
"loss": 1.1404,
|
| 886 |
"step": 1240
|
| 887 |
},
|
| 888 |
{
|
| 889 |
"epoch": 1.12,
|
| 890 |
+
"grad_norm": 0.8580232858657837,
|
| 891 |
"learning_rate": 7.58260378424174e-06,
|
| 892 |
"loss": 0.0049,
|
| 893 |
"step": 1250
|
| 894 |
},
|
| 895 |
{
|
| 896 |
"epoch": 1.12,
|
| 897 |
+
"grad_norm": 0.25582146644592285,
|
| 898 |
"learning_rate": 7.554363174244564e-06,
|
| 899 |
+
"loss": 0.0043,
|
| 900 |
"step": 1260
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 1.12,
|
| 904 |
+
"grad_norm": 0.11917036771774292,
|
| 905 |
"learning_rate": 7.526122564247388e-06,
|
| 906 |
+
"loss": 0.2731,
|
| 907 |
"step": 1270
|
| 908 |
},
|
| 909 |
{
|
| 910 |
"epoch": 1.13,
|
| 911 |
+
"grad_norm": 46.90564727783203,
|
| 912 |
"learning_rate": 7.497881954250212e-06,
|
| 913 |
+
"loss": 0.7342,
|
| 914 |
"step": 1280
|
| 915 |
},
|
| 916 |
{
|
| 917 |
"epoch": 1.13,
|
| 918 |
+
"grad_norm": 24.272871017456055,
|
| 919 |
"learning_rate": 7.469641344253037e-06,
|
| 920 |
+
"loss": 1.1628,
|
| 921 |
"step": 1290
|
| 922 |
},
|
| 923 |
{
|
| 924 |
"epoch": 1.13,
|
| 925 |
+
"grad_norm": 0.12642551958560944,
|
| 926 |
"learning_rate": 7.441400734255861e-06,
|
| 927 |
+
"loss": 0.4991,
|
| 928 |
"step": 1300
|
| 929 |
},
|
| 930 |
{
|
| 931 |
"epoch": 1.13,
|
| 932 |
+
"grad_norm": 26.394733428955078,
|
| 933 |
"learning_rate": 7.413160124258685e-06,
|
| 934 |
+
"loss": 1.6511,
|
| 935 |
"step": 1310
|
| 936 |
},
|
| 937 |
{
|
| 938 |
"epoch": 1.14,
|
| 939 |
+
"grad_norm": 0.33993953466415405,
|
| 940 |
"learning_rate": 7.384919514261508e-06,
|
| 941 |
+
"loss": 0.4879,
|
| 942 |
"step": 1320
|
| 943 |
},
|
| 944 |
{
|
| 945 |
"epoch": 1.14,
|
| 946 |
+
"grad_norm": 0.2504374086856842,
|
| 947 |
"learning_rate": 7.356678904264333e-06,
|
| 948 |
+
"loss": 0.4509,
|
| 949 |
"step": 1330
|
| 950 |
},
|
| 951 |
{
|
| 952 |
"epoch": 1.14,
|
| 953 |
+
"grad_norm": 0.05574103072285652,
|
| 954 |
"learning_rate": 7.328438294267157e-06,
|
| 955 |
+
"loss": 0.4486,
|
| 956 |
"step": 1340
|
| 957 |
},
|
| 958 |
{
|
| 959 |
"epoch": 1.14,
|
| 960 |
+
"grad_norm": 0.23328031599521637,
|
| 961 |
"learning_rate": 7.30019768426998e-06,
|
| 962 |
+
"loss": 0.5054,
|
| 963 |
"step": 1350
|
| 964 |
},
|
| 965 |
{
|
| 966 |
"epoch": 1.15,
|
| 967 |
+
"grad_norm": 0.010512278415262699,
|
| 968 |
"learning_rate": 7.2719570742728055e-06,
|
| 969 |
+
"loss": 1.1599,
|
| 970 |
"step": 1360
|
| 971 |
},
|
| 972 |
{
|
| 973 |
"epoch": 1.15,
|
| 974 |
+
"grad_norm": 0.042285043746232986,
|
| 975 |
"learning_rate": 7.243716464275629e-06,
|
| 976 |
+
"loss": 0.4782,
|
| 977 |
"step": 1370
|
| 978 |
},
|
| 979 |
{
|
| 980 |
"epoch": 1.15,
|
| 981 |
+
"grad_norm": 0.012322720140218735,
|
| 982 |
"learning_rate": 7.215475854278453e-06,
|
| 983 |
+
"loss": 1.0997,
|
| 984 |
"step": 1380
|
| 985 |
},
|
| 986 |
{
|
| 987 |
"epoch": 1.15,
|
| 988 |
+
"grad_norm": 0.014798709191381931,
|
| 989 |
"learning_rate": 7.1872352442812775e-06,
|
| 990 |
+
"loss": 1.2561,
|
| 991 |
"step": 1390
|
| 992 |
},
|
| 993 |
{
|
| 994 |
"epoch": 1.16,
|
| 995 |
+
"grad_norm": 0.45606276392936707,
|
| 996 |
"learning_rate": 7.158994634284101e-06,
|
| 997 |
+
"loss": 1.2503,
|
| 998 |
"step": 1400
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
"epoch": 1.16,
|
| 1002 |
+
"grad_norm": 0.04647226259112358,
|
| 1003 |
"learning_rate": 7.130754024286925e-06,
|
| 1004 |
+
"loss": 1.2027,
|
| 1005 |
"step": 1410
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
"epoch": 1.16,
|
| 1009 |
+
"grad_norm": 0.34346136450767517,
|
| 1010 |
"learning_rate": 7.102513414289749e-06,
|
| 1011 |
+
"loss": 0.0049,
|
| 1012 |
"step": 1420
|
| 1013 |
},
|
| 1014 |
{
|
| 1015 |
"epoch": 1.16,
|
| 1016 |
+
"grad_norm": 23.320772171020508,
|
| 1017 |
"learning_rate": 7.074272804292574e-06,
|
| 1018 |
+
"loss": 0.8748,
|
| 1019 |
"step": 1430
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
"epoch": 1.17,
|
| 1023 |
+
"grad_norm": 0.4166504144668579,
|
| 1024 |
"learning_rate": 7.0460321942953965e-06,
|
| 1025 |
+
"loss": 0.8503,
|
| 1026 |
"step": 1440
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
"epoch": 1.17,
|
| 1030 |
+
"grad_norm": 0.028790142387151718,
|
| 1031 |
"learning_rate": 7.017791584298222e-06,
|
| 1032 |
+
"loss": 0.755,
|
| 1033 |
"step": 1450
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 1.17,
|
| 1037 |
+
"grad_norm": 0.10875601321458817,
|
| 1038 |
"learning_rate": 6.989550974301046e-06,
|
| 1039 |
+
"loss": 1.2674,
|
| 1040 |
"step": 1460
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
"epoch": 1.17,
|
| 1044 |
+
"grad_norm": 0.24065102636814117,
|
| 1045 |
"learning_rate": 6.961310364303869e-06,
|
| 1046 |
+
"loss": 1.0944,
|
| 1047 |
"step": 1470
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
"epoch": 1.18,
|
| 1051 |
+
"grad_norm": 0.17569658160209656,
|
| 1052 |
"learning_rate": 6.933069754306694e-06,
|
| 1053 |
+
"loss": 1.0381,
|
| 1054 |
"step": 1480
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
"epoch": 1.18,
|
| 1058 |
+
"grad_norm": 130.54307556152344,
|
| 1059 |
"learning_rate": 6.904829144309517e-06,
|
| 1060 |
+
"loss": 1.0751,
|
| 1061 |
"step": 1490
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
"epoch": 1.18,
|
| 1065 |
+
"grad_norm": 0.1745329350233078,
|
| 1066 |
"learning_rate": 6.876588534312341e-06,
|
| 1067 |
+
"loss": 0.97,
|
| 1068 |
"step": 1500
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 1.18,
|
| 1072 |
+
"grad_norm": 24.403528213500977,
|
| 1073 |
"learning_rate": 6.8483479243151665e-06,
|
| 1074 |
+
"loss": 0.8853,
|
| 1075 |
"step": 1510
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.19,
|
| 1079 |
+
"grad_norm": 3.1027746200561523,
|
| 1080 |
"learning_rate": 6.82010731431799e-06,
|
| 1081 |
+
"loss": 0.6651,
|
| 1082 |
"step": 1520
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
"epoch": 1.19,
|
| 1086 |
+
"grad_norm": 5.846127033233643,
|
| 1087 |
"learning_rate": 6.791866704320814e-06,
|
| 1088 |
+
"loss": 0.4536,
|
| 1089 |
"step": 1530
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 1.19,
|
| 1093 |
+
"grad_norm": 0.017441660165786743,
|
| 1094 |
"learning_rate": 6.763626094323638e-06,
|
| 1095 |
+
"loss": 1.0821,
|
| 1096 |
"step": 1540
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
"epoch": 1.19,
|
| 1100 |
+
"grad_norm": 0.20883318781852722,
|
| 1101 |
"learning_rate": 6.735385484326462e-06,
|
| 1102 |
+
"loss": 0.0086,
|
| 1103 |
"step": 1550
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
"epoch": 1.2,
|
| 1107 |
+
"grad_norm": 0.15525928139686584,
|
| 1108 |
"learning_rate": 6.707144874329285e-06,
|
| 1109 |
+
"loss": 0.5036,
|
| 1110 |
"step": 1560
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
"epoch": 1.2,
|
| 1114 |
+
"grad_norm": 0.15142841637134552,
|
| 1115 |
"learning_rate": 6.67890426433211e-06,
|
| 1116 |
+
"loss": 0.7465,
|
| 1117 |
"step": 1570
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
"epoch": 1.2,
|
| 1121 |
+
"eval_accuracy": 0.8338926174496645,
|
| 1122 |
+
"eval_loss": 0.7484979629516602,
|
| 1123 |
+
"eval_runtime": 324.9358,
|
| 1124 |
+
"eval_samples_per_second": 1.834,
|
| 1125 |
+
"eval_steps_per_second": 1.834,
|
| 1126 |
"step": 1574
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
"epoch": 2.0,
|
| 1130 |
+
"grad_norm": 0.13457632064819336,
|
| 1131 |
"learning_rate": 6.650663654334935e-06,
|
| 1132 |
+
"loss": 1.7798,
|
| 1133 |
"step": 1580
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
"epoch": 2.0,
|
| 1137 |
+
"grad_norm": 0.16015511751174927,
|
| 1138 |
"learning_rate": 6.622423044337758e-06,
|
| 1139 |
+
"loss": 0.3209,
|
| 1140 |
"step": 1590
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 2.01,
|
| 1144 |
+
"grad_norm": 0.16651424765586853,
|
| 1145 |
"learning_rate": 6.5941824343405826e-06,
|
| 1146 |
+
"loss": 1.5177,
|
| 1147 |
"step": 1600
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
"epoch": 2.01,
|
| 1151 |
+
"grad_norm": 0.35417357087135315,
|
| 1152 |
"learning_rate": 6.565941824343406e-06,
|
| 1153 |
+
"loss": 0.515,
|
| 1154 |
"step": 1610
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
"epoch": 2.01,
|
| 1158 |
+
"grad_norm": 0.05276577174663544,
|
| 1159 |
"learning_rate": 6.53770121434623e-06,
|
| 1160 |
+
"loss": 1.9682,
|
| 1161 |
"step": 1620
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 2.01,
|
| 1165 |
+
"grad_norm": 0.23119059205055237,
|
| 1166 |
"learning_rate": 6.5094606043490546e-06,
|
| 1167 |
+
"loss": 0.4395,
|
| 1168 |
"step": 1630
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 2.02,
|
| 1172 |
+
"grad_norm": 0.4055440425872803,
|
| 1173 |
"learning_rate": 6.481219994351878e-06,
|
| 1174 |
+
"loss": 0.0106,
|
| 1175 |
"step": 1640
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 2.02,
|
| 1179 |
+
"grad_norm": 0.15100878477096558,
|
| 1180 |
"learning_rate": 6.452979384354703e-06,
|
| 1181 |
+
"loss": 0.0059,
|
| 1182 |
"step": 1650
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 2.02,
|
| 1186 |
+
"grad_norm": 0.14264227449893951,
|
| 1187 |
"learning_rate": 6.424738774357527e-06,
|
| 1188 |
+
"loss": 1.2383,
|
| 1189 |
"step": 1660
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 2.02,
|
| 1193 |
+
"grad_norm": 0.16001996397972107,
|
| 1194 |
"learning_rate": 6.396498164360351e-06,
|
| 1195 |
+
"loss": 1.02,
|
| 1196 |
"step": 1670
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 2.03,
|
| 1200 |
+
"grad_norm": 0.15914620459079742,
|
| 1201 |
"learning_rate": 6.368257554363175e-06,
|
| 1202 |
+
"loss": 0.005,
|
| 1203 |
"step": 1680
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 2.03,
|
| 1207 |
+
"grad_norm": 0.11949492990970612,
|
| 1208 |
"learning_rate": 6.340016944365999e-06,
|
| 1209 |
+
"loss": 0.5224,
|
| 1210 |
"step": 1690
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 2.03,
|
| 1214 |
+
"grad_norm": 68.48541259765625,
|
| 1215 |
"learning_rate": 6.311776334368823e-06,
|
| 1216 |
+
"loss": 0.7585,
|
| 1217 |
"step": 1700
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 2.03,
|
| 1221 |
+
"grad_norm": 78.09130859375,
|
| 1222 |
"learning_rate": 6.283535724371646e-06,
|
| 1223 |
+
"loss": 1.5627,
|
| 1224 |
"step": 1710
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 2.04,
|
| 1228 |
+
"grad_norm": 0.06310901045799255,
|
| 1229 |
"learning_rate": 6.2552951143744715e-06,
|
| 1230 |
+
"loss": 1.0456,
|
| 1231 |
"step": 1720
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 2.04,
|
| 1235 |
+
"grad_norm": 0.7593610882759094,
|
| 1236 |
"learning_rate": 6.227054504377295e-06,
|
| 1237 |
+
"loss": 0.6762,
|
| 1238 |
"step": 1730
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 2.04,
|
| 1242 |
+
"grad_norm": 0.1845823973417282,
|
| 1243 |
"learning_rate": 6.198813894380119e-06,
|
| 1244 |
+
"loss": 0.9624,
|
| 1245 |
"step": 1740
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 2.04,
|
| 1249 |
+
"grad_norm": 0.49727919697761536,
|
| 1250 |
"learning_rate": 6.1705732843829435e-06,
|
| 1251 |
+
"loss": 1.1605,
|
| 1252 |
"step": 1750
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 2.05,
|
| 1256 |
+
"grad_norm": 25.680139541625977,
|
| 1257 |
"learning_rate": 6.142332674385767e-06,
|
| 1258 |
+
"loss": 1.5904,
|
| 1259 |
"step": 1760
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 2.05,
|
| 1263 |
+
"grad_norm": 0.20278450846672058,
|
| 1264 |
"learning_rate": 6.114092064388591e-06,
|
| 1265 |
+
"loss": 0.3835,
|
| 1266 |
"step": 1770
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 2.05,
|
| 1270 |
+
"grad_norm": 0.2322162538766861,
|
| 1271 |
"learning_rate": 6.085851454391415e-06,
|
| 1272 |
+
"loss": 0.9161,
|
| 1273 |
"step": 1780
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 2.05,
|
| 1277 |
+
"grad_norm": 0.2889300286769867,
|
| 1278 |
"learning_rate": 6.05761084439424e-06,
|
| 1279 |
+
"loss": 0.2446,
|
| 1280 |
"step": 1790
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 2.06,
|
| 1284 |
+
"grad_norm": 0.21073293685913086,
|
| 1285 |
"learning_rate": 6.029370234397064e-06,
|
| 1286 |
+
"loss": 0.6002,
|
| 1287 |
"step": 1800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 2.06,
|
| 1291 |
+
"grad_norm": 0.2775285243988037,
|
| 1292 |
"learning_rate": 6.0011296243998876e-06,
|
| 1293 |
+
"loss": 0.0199,
|
| 1294 |
"step": 1810
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 2.06,
|
| 1298 |
+
"grad_norm": 0.13404442369937897,
|
| 1299 |
"learning_rate": 5.972889014402712e-06,
|
| 1300 |
+
"loss": 0.0098,
|
| 1301 |
"step": 1820
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 2.07,
|
| 1305 |
+
"grad_norm": 0.19267532229423523,
|
| 1306 |
"learning_rate": 5.944648404405535e-06,
|
| 1307 |
+
"loss": 2.1965,
|
| 1308 |
"step": 1830
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 2.07,
|
| 1312 |
+
"grad_norm": 0.314384788274765,
|
| 1313 |
"learning_rate": 5.9164077944083596e-06,
|
| 1314 |
+
"loss": 1.875,
|
| 1315 |
"step": 1840
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 2.07,
|
| 1319 |
+
"grad_norm": 0.410067081451416,
|
| 1320 |
"learning_rate": 5.888167184411185e-06,
|
| 1321 |
+
"loss": 0.01,
|
| 1322 |
"step": 1850
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 2.07,
|
| 1326 |
+
"grad_norm": 26.344202041625977,
|
| 1327 |
"learning_rate": 5.859926574414007e-06,
|
| 1328 |
+
"loss": 0.8549,
|
| 1329 |
"step": 1860
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 2.08,
|
| 1333 |
+
"grad_norm": 0.2887156307697296,
|
| 1334 |
"learning_rate": 5.8316859644168324e-06,
|
| 1335 |
+
"loss": 0.0071,
|
| 1336 |
"step": 1870
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 2.08,
|
| 1340 |
+
"grad_norm": 0.2524619698524475,
|
| 1341 |
"learning_rate": 5.803445354419656e-06,
|
| 1342 |
+
"loss": 0.4629,
|
| 1343 |
"step": 1880
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 2.08,
|
| 1347 |
+
"grad_norm": 0.22347721457481384,
|
| 1348 |
"learning_rate": 5.77520474442248e-06,
|
| 1349 |
+
"loss": 0.454,
|
| 1350 |
"step": 1890
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 2.08,
|
| 1354 |
+
"grad_norm": 0.19489675760269165,
|
| 1355 |
"learning_rate": 5.746964134425304e-06,
|
| 1356 |
+
"loss": 0.0037,
|
| 1357 |
"step": 1900
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 2.09,
|
| 1361 |
+
"grad_norm": 25.353939056396484,
|
| 1362 |
"learning_rate": 5.718723524428128e-06,
|
| 1363 |
+
"loss": 0.5369,
|
| 1364 |
"step": 1910
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 2.09,
|
| 1368 |
+
"grad_norm": 0.01046634092926979,
|
| 1369 |
"learning_rate": 5.690482914430952e-06,
|
| 1370 |
+
"loss": 0.0017,
|
| 1371 |
"step": 1920
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 2.09,
|
| 1375 |
+
"grad_norm": 0.016024693846702576,
|
| 1376 |
"learning_rate": 5.662242304433776e-06,
|
| 1377 |
+
"loss": 0.6874,
|
| 1378 |
"step": 1930
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 2.09,
|
| 1382 |
+
"grad_norm": 0.13052043318748474,
|
| 1383 |
"learning_rate": 5.634001694436601e-06,
|
| 1384 |
+
"loss": 0.5459,
|
| 1385 |
"step": 1940
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 2.1,
|
| 1389 |
+
"grad_norm": 22.755247116088867,
|
| 1390 |
"learning_rate": 5.605761084439424e-06,
|
| 1391 |
+
"loss": 1.063,
|
| 1392 |
"step": 1950
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 2.1,
|
| 1396 |
+
"grad_norm": 0.017313385382294655,
|
| 1397 |
"learning_rate": 5.5775204744422485e-06,
|
| 1398 |
+
"loss": 1.3419,
|
| 1399 |
"step": 1960
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 2.1,
|
| 1403 |
+
"grad_norm": 0.27627086639404297,
|
| 1404 |
"learning_rate": 5.549279864445073e-06,
|
| 1405 |
+
"loss": 0.7342,
|
| 1406 |
"step": 1970
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 2.1,
|
| 1410 |
+
"grad_norm": 0.02726130001246929,
|
| 1411 |
"learning_rate": 5.521039254447896e-06,
|
| 1412 |
+
"loss": 1.1784,
|
| 1413 |
"step": 1980
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 2.11,
|
| 1417 |
+
"grad_norm": 0.3983820378780365,
|
| 1418 |
"learning_rate": 5.4927986444507205e-06,
|
| 1419 |
+
"loss": 0.018,
|
| 1420 |
"step": 1990
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 2.11,
|
| 1424 |
+
"grad_norm": 0.13953474164009094,
|
| 1425 |
"learning_rate": 5.464558034453544e-06,
|
| 1426 |
+
"loss": 0.0044,
|
| 1427 |
"step": 2000
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 2.11,
|
| 1431 |
+
"grad_norm": 0.04043055698275566,
|
| 1432 |
"learning_rate": 5.436317424456369e-06,
|
| 1433 |
+
"loss": 0.5193,
|
| 1434 |
"step": 2010
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 2.11,
|
| 1438 |
+
"grad_norm": 22.795866012573242,
|
| 1439 |
"learning_rate": 5.408076814459193e-06,
|
| 1440 |
+
"loss": 0.823,
|
| 1441 |
"step": 2020
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 2.12,
|
| 1445 |
+
"grad_norm": 23.257352828979492,
|
| 1446 |
"learning_rate": 5.379836204462017e-06,
|
| 1447 |
+
"loss": 1.0949,
|
| 1448 |
"step": 2030
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
"epoch": 2.12,
|
| 1452 |
+
"grad_norm": 22.406156539916992,
|
| 1453 |
"learning_rate": 5.351595594464841e-06,
|
| 1454 |
+
"loss": 2.1405,
|
| 1455 |
"step": 2040
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
"epoch": 2.12,
|
| 1459 |
+
"grad_norm": 0.4050818979740143,
|
| 1460 |
"learning_rate": 5.323354984467665e-06,
|
| 1461 |
+
"loss": 1.4621,
|
| 1462 |
"step": 2050
|
| 1463 |
},
|
| 1464 |
{
|
| 1465 |
"epoch": 2.12,
|
| 1466 |
+
"grad_norm": 0.4241192638874054,
|
| 1467 |
"learning_rate": 5.295114374470489e-06,
|
| 1468 |
+
"loss": 0.4477,
|
| 1469 |
"step": 2060
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
"epoch": 2.13,
|
| 1473 |
+
"grad_norm": 0.30280259251594543,
|
| 1474 |
"learning_rate": 5.266873764473312e-06,
|
| 1475 |
+
"loss": 0.0106,
|
| 1476 |
"step": 2070
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
"epoch": 2.13,
|
| 1480 |
+
"grad_norm": 0.18719686567783356,
|
| 1481 |
"learning_rate": 5.2386331544761374e-06,
|
| 1482 |
+
"loss": 0.4244,
|
| 1483 |
"step": 2080
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
"epoch": 2.13,
|
| 1487 |
+
"grad_norm": 0.04479646682739258,
|
| 1488 |
"learning_rate": 5.210392544478962e-06,
|
| 1489 |
+
"loss": 0.6732,
|
| 1490 |
"step": 2090
|
| 1491 |
},
|
| 1492 |
{
|
| 1493 |
"epoch": 2.13,
|
| 1494 |
+
"grad_norm": 22.563093185424805,
|
| 1495 |
"learning_rate": 5.182151934481785e-06,
|
| 1496 |
+
"loss": 0.5233,
|
| 1497 |
"step": 2100
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 2.14,
|
| 1501 |
+
"grad_norm": 0.0441102497279644,
|
| 1502 |
"learning_rate": 5.1539113244846095e-06,
|
| 1503 |
+
"loss": 1.7924,
|
| 1504 |
"step": 2110
|
| 1505 |
},
|
| 1506 |
{
|
| 1507 |
"epoch": 2.14,
|
| 1508 |
+
"grad_norm": 0.046370625495910645,
|
| 1509 |
"learning_rate": 5.125670714487433e-06,
|
| 1510 |
+
"loss": 0.0049,
|
| 1511 |
"step": 2120
|
| 1512 |
},
|
| 1513 |
{
|
| 1514 |
"epoch": 2.14,
|
| 1515 |
+
"grad_norm": 1.073113203048706,
|
| 1516 |
"learning_rate": 5.097430104490257e-06,
|
| 1517 |
+
"loss": 0.5907,
|
| 1518 |
"step": 2130
|
| 1519 |
},
|
| 1520 |
{
|
| 1521 |
"epoch": 2.14,
|
| 1522 |
+
"grad_norm": 0.08737409859895706,
|
| 1523 |
"learning_rate": 5.069189494493082e-06,
|
| 1524 |
+
"loss": 1.5069,
|
| 1525 |
"step": 2140
|
| 1526 |
},
|
| 1527 |
{
|
| 1528 |
"epoch": 2.15,
|
| 1529 |
+
"grad_norm": 0.058251939713954926,
|
| 1530 |
"learning_rate": 5.040948884495906e-06,
|
| 1531 |
+
"loss": 0.0395,
|
| 1532 |
"step": 2150
|
| 1533 |
},
|
| 1534 |
{
|
| 1535 |
"epoch": 2.15,
|
| 1536 |
+
"grad_norm": 39.82133102416992,
|
| 1537 |
"learning_rate": 5.01270827449873e-06,
|
| 1538 |
+
"loss": 0.01,
|
| 1539 |
"step": 2160
|
| 1540 |
},
|
| 1541 |
{
|
| 1542 |
"epoch": 2.15,
|
| 1543 |
+
"grad_norm": 0.18180860579013824,
|
| 1544 |
"learning_rate": 4.9844676645015535e-06,
|
| 1545 |
+
"loss": 0.0042,
|
| 1546 |
"step": 2170
|
| 1547 |
},
|
| 1548 |
{
|
| 1549 |
"epoch": 2.15,
|
| 1550 |
+
"grad_norm": 0.10672131180763245,
|
| 1551 |
"learning_rate": 4.956227054504378e-06,
|
| 1552 |
+
"loss": 0.01,
|
| 1553 |
"step": 2180
|
| 1554 |
},
|
| 1555 |
{
|
| 1556 |
"epoch": 2.16,
|
| 1557 |
+
"grad_norm": 28.28722381591797,
|
| 1558 |
"learning_rate": 4.927986444507202e-06,
|
| 1559 |
+
"loss": 1.221,
|
| 1560 |
"step": 2190
|
| 1561 |
},
|
| 1562 |
{
|
| 1563 |
"epoch": 2.16,
|
| 1564 |
+
"grad_norm": 0.12183073163032532,
|
| 1565 |
"learning_rate": 4.8997458345100255e-06,
|
| 1566 |
+
"loss": 1.1702,
|
| 1567 |
"step": 2200
|
| 1568 |
},
|
| 1569 |
{
|
| 1570 |
"epoch": 2.16,
|
| 1571 |
+
"grad_norm": 0.17091098427772522,
|
| 1572 |
"learning_rate": 4.87150522451285e-06,
|
| 1573 |
+
"loss": 1.9466,
|
| 1574 |
"step": 2210
|
| 1575 |
},
|
| 1576 |
{
|
| 1577 |
"epoch": 2.16,
|
| 1578 |
+
"grad_norm": 24.53551483154297,
|
| 1579 |
"learning_rate": 4.843264614515674e-06,
|
| 1580 |
+
"loss": 1.0011,
|
| 1581 |
"step": 2220
|
| 1582 |
},
|
| 1583 |
{
|
| 1584 |
"epoch": 2.17,
|
| 1585 |
+
"grad_norm": 33.58413314819336,
|
| 1586 |
"learning_rate": 4.815024004518498e-06,
|
| 1587 |
+
"loss": 1.1844,
|
| 1588 |
"step": 2230
|
| 1589 |
},
|
| 1590 |
{
|
| 1591 |
"epoch": 2.17,
|
| 1592 |
+
"grad_norm": 24.298912048339844,
|
| 1593 |
"learning_rate": 4.786783394521322e-06,
|
| 1594 |
+
"loss": 1.477,
|
| 1595 |
"step": 2240
|
| 1596 |
},
|
| 1597 |
{
|
| 1598 |
"epoch": 2.17,
|
| 1599 |
+
"grad_norm": 0.2964635193347931,
|
| 1600 |
"learning_rate": 4.758542784524146e-06,
|
| 1601 |
+
"loss": 0.5349,
|
| 1602 |
"step": 2250
|
| 1603 |
},
|
| 1604 |
{
|
| 1605 |
"epoch": 2.17,
|
| 1606 |
+
"grad_norm": 0.5078187584877014,
|
| 1607 |
"learning_rate": 4.7303021745269704e-06,
|
| 1608 |
+
"loss": 0.9149,
|
| 1609 |
"step": 2260
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 2.18,
|
| 1613 |
+
"grad_norm": 0.2942535877227783,
|
| 1614 |
"learning_rate": 4.702061564529794e-06,
|
| 1615 |
+
"loss": 0.9105,
|
| 1616 |
"step": 2270
|
| 1617 |
},
|
| 1618 |
{
|
| 1619 |
"epoch": 2.18,
|
| 1620 |
+
"grad_norm": 0.2372201532125473,
|
| 1621 |
"learning_rate": 4.673820954532618e-06,
|
| 1622 |
+
"loss": 0.4805,
|
| 1623 |
"step": 2280
|
| 1624 |
},
|
| 1625 |
{
|
| 1626 |
"epoch": 2.18,
|
| 1627 |
+
"grad_norm": 1.4744486808776855,
|
| 1628 |
"learning_rate": 4.6455803445354425e-06,
|
| 1629 |
+
"loss": 0.0077,
|
| 1630 |
"step": 2290
|
| 1631 |
},
|
| 1632 |
{
|
| 1633 |
"epoch": 2.18,
|
| 1634 |
+
"grad_norm": 3.4638776779174805,
|
| 1635 |
"learning_rate": 4.617339734538267e-06,
|
| 1636 |
+
"loss": 0.8804,
|
| 1637 |
"step": 2300
|
| 1638 |
},
|
| 1639 |
{
|
| 1640 |
"epoch": 2.19,
|
| 1641 |
+
"grad_norm": 0.03575224056839943,
|
| 1642 |
"learning_rate": 4.58909912454109e-06,
|
| 1643 |
+
"loss": 0.4857,
|
| 1644 |
"step": 2310
|
| 1645 |
},
|
| 1646 |
{
|
| 1647 |
"epoch": 2.19,
|
| 1648 |
+
"grad_norm": 595.8480834960938,
|
| 1649 |
"learning_rate": 4.5608585145439145e-06,
|
| 1650 |
+
"loss": 0.2092,
|
| 1651 |
"step": 2320
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
"epoch": 2.19,
|
| 1655 |
+
"grad_norm": 0.17693261802196503,
|
| 1656 |
"learning_rate": 4.532617904546738e-06,
|
| 1657 |
+
"loss": 1.4817,
|
| 1658 |
"step": 2330
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
"epoch": 2.19,
|
| 1662 |
+
"grad_norm": 0.1713973432779312,
|
| 1663 |
"learning_rate": 4.504377294549563e-06,
|
| 1664 |
+
"loss": 0.4704,
|
| 1665 |
"step": 2340
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 2.2,
|
| 1669 |
+
"grad_norm": 0.17562945187091827,
|
| 1670 |
"learning_rate": 4.4761366845523865e-06,
|
| 1671 |
+
"loss": 0.4608,
|
| 1672 |
"step": 2350
|
| 1673 |
},
|
| 1674 |
{
|
| 1675 |
"epoch": 2.2,
|
| 1676 |
+
"grad_norm": 0.04520571604371071,
|
| 1677 |
"learning_rate": 4.447896074555211e-06,
|
| 1678 |
+
"loss": 0.5066,
|
| 1679 |
"step": 2360
|
| 1680 |
},
|
| 1681 |
{
|
| 1682 |
"epoch": 2.2,
|
| 1683 |
+
"eval_accuracy": 0.8104026845637584,
|
| 1684 |
+
"eval_loss": 0.8862009644508362,
|
| 1685 |
+
"eval_runtime": 319.8982,
|
| 1686 |
+
"eval_samples_per_second": 1.863,
|
| 1687 |
+
"eval_steps_per_second": 1.863,
|
| 1688 |
"step": 2361
|
| 1689 |
},
|
| 1690 |
{
|
| 1691 |
"epoch": 3.0,
|
| 1692 |
+
"grad_norm": 0.06622195243835449,
|
| 1693 |
"learning_rate": 4.419655464558035e-06,
|
| 1694 |
+
"loss": 0.4614,
|
| 1695 |
"step": 2370
|
| 1696 |
},
|
| 1697 |
{
|
| 1698 |
"epoch": 3.0,
|
| 1699 |
+
"grad_norm": 24.007558822631836,
|
| 1700 |
"learning_rate": 4.3914148545608585e-06,
|
| 1701 |
+
"loss": 0.463,
|
| 1702 |
"step": 2380
|
| 1703 |
},
|
| 1704 |
{
|
| 1705 |
"epoch": 3.01,
|
| 1706 |
+
"grad_norm": 74.15673065185547,
|
| 1707 |
"learning_rate": 4.363174244563683e-06,
|
| 1708 |
+
"loss": 0.4436,
|
| 1709 |
"step": 2390
|
| 1710 |
},
|
| 1711 |
{
|
| 1712 |
"epoch": 3.01,
|
| 1713 |
+
"grad_norm": 0.021963810548186302,
|
| 1714 |
"learning_rate": 4.334933634566507e-06,
|
| 1715 |
+
"loss": 0.0038,
|
| 1716 |
"step": 2400
|
| 1717 |
},
|
| 1718 |
{
|
| 1719 |
"epoch": 3.01,
|
| 1720 |
+
"grad_norm": 0.015897750854492188,
|
| 1721 |
"learning_rate": 4.306693024569331e-06,
|
| 1722 |
+
"loss": 0.0044,
|
| 1723 |
"step": 2410
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
"epoch": 3.01,
|
| 1727 |
+
"grad_norm": 0.19425620138645172,
|
| 1728 |
"learning_rate": 4.278452414572155e-06,
|
| 1729 |
+
"loss": 1.5548,
|
| 1730 |
"step": 2420
|
| 1731 |
},
|
| 1732 |
{
|
| 1733 |
"epoch": 3.02,
|
| 1734 |
+
"grad_norm": 0.2513945400714874,
|
| 1735 |
"learning_rate": 4.250211804574979e-06,
|
| 1736 |
+
"loss": 0.5131,
|
| 1737 |
"step": 2430
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 3.02,
|
| 1741 |
+
"grad_norm": 0.010541065596044064,
|
| 1742 |
"learning_rate": 4.221971194577803e-06,
|
| 1743 |
+
"loss": 0.0067,
|
| 1744 |
"step": 2440
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 3.02,
|
| 1748 |
+
"grad_norm": 0.016293305903673172,
|
| 1749 |
"learning_rate": 4.193730584580628e-06,
|
| 1750 |
+
"loss": 0.9886,
|
| 1751 |
"step": 2450
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 3.03,
|
| 1755 |
+
"grad_norm": 0.009771335870027542,
|
| 1756 |
"learning_rate": 4.165489974583451e-06,
|
| 1757 |
+
"loss": 0.0052,
|
| 1758 |
"step": 2460
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 3.03,
|
| 1762 |
+
"grad_norm": 22.732620239257812,
|
| 1763 |
"learning_rate": 4.1372493645862754e-06,
|
| 1764 |
+
"loss": 0.9754,
|
| 1765 |
"step": 2470
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 3.03,
|
| 1769 |
+
"grad_norm": 0.01741587370634079,
|
| 1770 |
"learning_rate": 4.1090087545891e-06,
|
| 1771 |
+
"loss": 0.007,
|
| 1772 |
"step": 2480
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 3.03,
|
| 1776 |
+
"grad_norm": 0.011250239796936512,
|
| 1777 |
"learning_rate": 4.080768144591923e-06,
|
| 1778 |
+
"loss": 1.2522,
|
| 1779 |
"step": 2490
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 3.04,
|
| 1783 |
+
"grad_norm": 35.29149627685547,
|
| 1784 |
"learning_rate": 4.0525275345947475e-06,
|
| 1785 |
+
"loss": 1.0829,
|
| 1786 |
"step": 2500
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 3.04,
|
| 1790 |
+
"grad_norm": 0.2253457009792328,
|
| 1791 |
"learning_rate": 4.024286924597572e-06,
|
| 1792 |
+
"loss": 2.0893,
|
| 1793 |
"step": 2510
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 3.04,
|
| 1797 |
+
"grad_norm": 0.07997630536556244,
|
| 1798 |
"learning_rate": 3.996046314600396e-06,
|
| 1799 |
+
"loss": 0.5229,
|
| 1800 |
"step": 2520
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 3.04,
|
| 1804 |
+
"grad_norm": 0.24757777154445648,
|
| 1805 |
"learning_rate": 3.9678057046032195e-06,
|
| 1806 |
+
"loss": 0.0078,
|
| 1807 |
"step": 2530
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 3.05,
|
| 1811 |
+
"grad_norm": 0.20268729329109192,
|
| 1812 |
"learning_rate": 3.939565094606044e-06,
|
| 1813 |
+
"loss": 0.007,
|
| 1814 |
"step": 2540
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 3.05,
|
| 1818 |
+
"grad_norm": 0.18007797002792358,
|
| 1819 |
"learning_rate": 3.911324484608868e-06,
|
| 1820 |
+
"loss": 0.0057,
|
| 1821 |
"step": 2550
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 3.05,
|
| 1825 |
+
"grad_norm": 0.04412233456969261,
|
| 1826 |
"learning_rate": 3.8830838746116915e-06,
|
| 1827 |
+
"loss": 0.6502,
|
| 1828 |
"step": 2560
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 3.05,
|
| 1832 |
+
"grad_norm": 0.2677147686481476,
|
| 1833 |
"learning_rate": 3.854843264614516e-06,
|
| 1834 |
+
"loss": 1.0556,
|
| 1835 |
"step": 2570
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 3.06,
|
| 1839 |
+
"grad_norm": 0.13710559904575348,
|
| 1840 |
"learning_rate": 3.82660265461734e-06,
|
| 1841 |
+
"loss": 0.6048,
|
| 1842 |
"step": 2580
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 3.06,
|
| 1846 |
+
"grad_norm": 0.14919081330299377,
|
| 1847 |
"learning_rate": 3.798362044620164e-06,
|
| 1848 |
+
"loss": 1.0146,
|
| 1849 |
"step": 2590
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 3.06,
|
| 1853 |
+
"grad_norm": 0.06460587680339813,
|
| 1854 |
"learning_rate": 3.7701214346229882e-06,
|
| 1855 |
+
"loss": 0.0037,
|
| 1856 |
"step": 2600
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 3.06,
|
| 1860 |
+
"grad_norm": 22.712013244628906,
|
| 1861 |
"learning_rate": 3.741880824625812e-06,
|
| 1862 |
+
"loss": 1.1229,
|
| 1863 |
"step": 2610
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 3.07,
|
| 1867 |
+
"grad_norm": 0.03939145430922508,
|
| 1868 |
"learning_rate": 3.7136402146286364e-06,
|
| 1869 |
+
"loss": 0.5016,
|
| 1870 |
"step": 2620
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 3.07,
|
| 1874 |
+
"grad_norm": 0.179279625415802,
|
| 1875 |
"learning_rate": 3.6853996046314607e-06,
|
| 1876 |
+
"loss": 0.991,
|
| 1877 |
"step": 2630
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 3.07,
|
| 1881 |
+
"grad_norm": 0.16804170608520508,
|
| 1882 |
"learning_rate": 3.6571589946342845e-06,
|
| 1883 |
+
"loss": 0.0079,
|
| 1884 |
"step": 2640
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 3.07,
|
| 1888 |
+
"grad_norm": 0.20420299470424652,
|
| 1889 |
"learning_rate": 3.6289183846371084e-06,
|
| 1890 |
+
"loss": 0.9568,
|
| 1891 |
"step": 2650
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 3.08,
|
| 1895 |
+
"grad_norm": 0.22263668477535248,
|
| 1896 |
"learning_rate": 3.6006777746399323e-06,
|
| 1897 |
+
"loss": 0.4772,
|
| 1898 |
"step": 2660
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 3.08,
|
| 1902 |
+
"grad_norm": 22.77140998840332,
|
| 1903 |
"learning_rate": 3.5724371646427566e-06,
|
| 1904 |
+
"loss": 1.9533,
|
| 1905 |
"step": 2670
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 3.08,
|
| 1909 |
+
"grad_norm": 0.27024608850479126,
|
| 1910 |
"learning_rate": 3.544196554645581e-06,
|
| 1911 |
+
"loss": 2.2664,
|
| 1912 |
"step": 2680
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 3.08,
|
| 1916 |
+
"grad_norm": 0.4369601905345917,
|
| 1917 |
"learning_rate": 3.5159559446484047e-06,
|
| 1918 |
+
"loss": 1.1402,
|
| 1919 |
"step": 2690
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 3.09,
|
| 1923 |
+
"grad_norm": 0.055966880172491074,
|
| 1924 |
"learning_rate": 3.4877153346512286e-06,
|
| 1925 |
+
"loss": 0.0093,
|
| 1926 |
"step": 2700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 3.09,
|
| 1930 |
+
"grad_norm": 0.31162434816360474,
|
| 1931 |
"learning_rate": 3.459474724654053e-06,
|
| 1932 |
+
"loss": 0.0104,
|
| 1933 |
"step": 2710
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 3.09,
|
| 1937 |
+
"grad_norm": 0.09564479440450668,
|
| 1938 |
"learning_rate": 3.4312341146568767e-06,
|
| 1939 |
+
"loss": 0.4793,
|
| 1940 |
"step": 2720
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 3.09,
|
| 1944 |
+
"grad_norm": 0.20361103117465973,
|
| 1945 |
"learning_rate": 3.4029935046597006e-06,
|
| 1946 |
+
"loss": 0.9354,
|
| 1947 |
"step": 2730
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 3.1,
|
| 1951 |
+
"grad_norm": 0.03742867708206177,
|
| 1952 |
"learning_rate": 3.3747528946625253e-06,
|
| 1953 |
+
"loss": 0.0062,
|
| 1954 |
"step": 2740
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 3.1,
|
| 1958 |
+
"grad_norm": 0.17758674919605255,
|
| 1959 |
"learning_rate": 3.346512284665349e-06,
|
| 1960 |
+
"loss": 0.9682,
|
| 1961 |
"step": 2750
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 3.1,
|
| 1965 |
+
"grad_norm": 0.04006683826446533,
|
| 1966 |
"learning_rate": 3.318271674668173e-06,
|
| 1967 |
+
"loss": 0.6539,
|
| 1968 |
"step": 2760
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 3.1,
|
| 1972 |
+
"grad_norm": 0.2262236326932907,
|
| 1973 |
"learning_rate": 3.290031064670997e-06,
|
| 1974 |
+
"loss": 0.472,
|
| 1975 |
"step": 2770
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 3.11,
|
| 1979 |
+
"grad_norm": 0.021157417446374893,
|
| 1980 |
"learning_rate": 3.2617904546738212e-06,
|
| 1981 |
+
"loss": 0.4522,
|
| 1982 |
"step": 2780
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 3.11,
|
| 1986 |
+
"grad_norm": 0.20334932208061218,
|
| 1987 |
"learning_rate": 3.233549844676645e-06,
|
| 1988 |
+
"loss": 0.0057,
|
| 1989 |
"step": 2790
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 3.11,
|
| 1993 |
+
"grad_norm": 0.18717893958091736,
|
| 1994 |
"learning_rate": 3.2053092346794694e-06,
|
| 1995 |
+
"loss": 0.0058,
|
| 1996 |
"step": 2800
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 3.11,
|
| 2000 |
+
"grad_norm": 0.015843726694583893,
|
| 2001 |
"learning_rate": 3.1770686246822937e-06,
|
| 2002 |
+
"loss": 0.9782,
|
| 2003 |
"step": 2810
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 3.12,
|
| 2007 |
+
"grad_norm": 304.4051208496094,
|
| 2008 |
"learning_rate": 3.1488280146851175e-06,
|
| 2009 |
+
"loss": 0.3817,
|
| 2010 |
"step": 2820
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 3.12,
|
| 2014 |
+
"grad_norm": 22.5703067779541,
|
| 2015 |
"learning_rate": 3.1205874046879414e-06,
|
| 2016 |
+
"loss": 0.9501,
|
| 2017 |
"step": 2830
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 3.12,
|
| 2021 |
+
"grad_norm": 0.015223432332277298,
|
| 2022 |
"learning_rate": 3.0923467946907653e-06,
|
| 2023 |
+
"loss": 1.3262,
|
| 2024 |
"step": 2840
|
| 2025 |
},
|
| 2026 |
{
|
| 2027 |
"epoch": 3.12,
|
| 2028 |
+
"grad_norm": 22.17841339111328,
|
| 2029 |
"learning_rate": 3.06410618469359e-06,
|
| 2030 |
+
"loss": 0.4495,
|
| 2031 |
"step": 2850
|
| 2032 |
},
|
| 2033 |
{
|
| 2034 |
"epoch": 3.13,
|
| 2035 |
+
"grad_norm": 0.2771930694580078,
|
| 2036 |
"learning_rate": 3.035865574696414e-06,
|
| 2037 |
+
"loss": 0.005,
|
| 2038 |
"step": 2860
|
| 2039 |
},
|
| 2040 |
{
|
| 2041 |
"epoch": 3.13,
|
| 2042 |
+
"grad_norm": 0.24659463763237,
|
| 2043 |
"learning_rate": 3.0076249646992377e-06,
|
| 2044 |
+
"loss": 1.0779,
|
| 2045 |
"step": 2870
|
| 2046 |
},
|
| 2047 |
{
|
| 2048 |
"epoch": 3.13,
|
| 2049 |
+
"grad_norm": 0.017306260764598846,
|
| 2050 |
"learning_rate": 2.9793843547020616e-06,
|
| 2051 |
+
"loss": 0.0052,
|
| 2052 |
"step": 2880
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 3.13,
|
| 2056 |
+
"grad_norm": 0.16755558550357819,
|
| 2057 |
"learning_rate": 2.951143744704886e-06,
|
| 2058 |
+
"loss": 0.4679,
|
| 2059 |
"step": 2890
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
"epoch": 3.14,
|
| 2063 |
+
"grad_norm": 22.35555076599121,
|
| 2064 |
"learning_rate": 2.9229031347077097e-06,
|
| 2065 |
+
"loss": 1.6778,
|
| 2066 |
"step": 2900
|
| 2067 |
},
|
| 2068 |
{
|
| 2069 |
"epoch": 3.14,
|
| 2070 |
+
"grad_norm": 848.5310668945312,
|
| 2071 |
"learning_rate": 2.894662524710534e-06,
|
| 2072 |
+
"loss": 0.368,
|
| 2073 |
"step": 2910
|
| 2074 |
},
|
| 2075 |
{
|
| 2076 |
"epoch": 3.14,
|
| 2077 |
+
"grad_norm": 0.12664149701595306,
|
| 2078 |
"learning_rate": 2.8664219147133583e-06,
|
| 2079 |
+
"loss": 1.8596,
|
| 2080 |
"step": 2920
|
| 2081 |
},
|
| 2082 |
{
|
| 2083 |
"epoch": 3.14,
|
| 2084 |
+
"grad_norm": 0.15273146331310272,
|
| 2085 |
"learning_rate": 2.838181304716182e-06,
|
| 2086 |
+
"loss": 0.0056,
|
| 2087 |
"step": 2930
|
| 2088 |
},
|
| 2089 |
{
|
| 2090 |
"epoch": 3.15,
|
| 2091 |
+
"grad_norm": 22.398298263549805,
|
| 2092 |
"learning_rate": 2.809940694719006e-06,
|
| 2093 |
+
"loss": 1.4325,
|
| 2094 |
"step": 2940
|
| 2095 |
},
|
| 2096 |
{
|
| 2097 |
"epoch": 3.15,
|
| 2098 |
+
"grad_norm": 413.4339294433594,
|
| 2099 |
"learning_rate": 2.78170008472183e-06,
|
| 2100 |
+
"loss": 0.3167,
|
| 2101 |
"step": 2950
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
"epoch": 3.15,
|
| 2105 |
+
"grad_norm": 0.19188903272151947,
|
| 2106 |
"learning_rate": 2.753459474724654e-06,
|
| 2107 |
+
"loss": 0.904,
|
| 2108 |
"step": 2960
|
| 2109 |
},
|
| 2110 |
{
|
| 2111 |
"epoch": 3.15,
|
| 2112 |
+
"grad_norm": 3.69321608543396,
|
| 2113 |
"learning_rate": 2.7252188647274785e-06,
|
| 2114 |
+
"loss": 0.007,
|
| 2115 |
"step": 2970
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 3.16,
|
| 2119 |
+
"grad_norm": 0.2725941240787506,
|
| 2120 |
"learning_rate": 2.6969782547303024e-06,
|
| 2121 |
+
"loss": 0.9046,
|
| 2122 |
"step": 2980
|
| 2123 |
},
|
| 2124 |
{
|
| 2125 |
"epoch": 3.16,
|
| 2126 |
+
"grad_norm": 0.2501017451286316,
|
| 2127 |
"learning_rate": 2.6687376447331266e-06,
|
| 2128 |
+
"loss": 0.0085,
|
| 2129 |
"step": 2990
|
| 2130 |
},
|
| 2131 |
{
|
| 2132 |
"epoch": 3.16,
|
| 2133 |
+
"grad_norm": 0.08356527239084244,
|
| 2134 |
"learning_rate": 2.6404970347359505e-06,
|
| 2135 |
+
"loss": 0.6352,
|
| 2136 |
"step": 3000
|
| 2137 |
},
|
| 2138 |
{
|
| 2139 |
"epoch": 3.16,
|
| 2140 |
+
"grad_norm": 0.05544191226363182,
|
| 2141 |
"learning_rate": 2.6122564247387744e-06,
|
| 2142 |
+
"loss": 1.566,
|
| 2143 |
"step": 3010
|
| 2144 |
},
|
| 2145 |
{
|
| 2146 |
"epoch": 3.17,
|
| 2147 |
+
"grad_norm": 0.2636459469795227,
|
| 2148 |
"learning_rate": 2.584015814741599e-06,
|
| 2149 |
+
"loss": 0.9192,
|
| 2150 |
"step": 3020
|
| 2151 |
},
|
| 2152 |
{
|
| 2153 |
"epoch": 3.17,
|
| 2154 |
+
"grad_norm": 0.05203928053379059,
|
| 2155 |
"learning_rate": 2.555775204744423e-06,
|
| 2156 |
+
"loss": 0.4324,
|
| 2157 |
"step": 3030
|
| 2158 |
},
|
| 2159 |
{
|
| 2160 |
"epoch": 3.17,
|
| 2161 |
+
"grad_norm": 0.16593405604362488,
|
| 2162 |
"learning_rate": 2.527534594747247e-06,
|
| 2163 |
+
"loss": 0.8602,
|
| 2164 |
"step": 3040
|
| 2165 |
},
|
| 2166 |
{
|
| 2167 |
"epoch": 3.18,
|
| 2168 |
+
"grad_norm": 0.02371574193239212,
|
| 2169 |
"learning_rate": 2.4992939847500707e-06,
|
| 2170 |
+
"loss": 0.5104,
|
| 2171 |
"step": 3050
|
| 2172 |
},
|
| 2173 |
{
|
| 2174 |
"epoch": 3.18,
|
| 2175 |
+
"grad_norm": 4.215577125549316,
|
| 2176 |
"learning_rate": 2.4710533747528946e-06,
|
| 2177 |
+
"loss": 0.6802,
|
| 2178 |
"step": 3060
|
| 2179 |
},
|
| 2180 |
{
|
| 2181 |
"epoch": 3.18,
|
| 2182 |
+
"grad_norm": 0.1291200816631317,
|
| 2183 |
"learning_rate": 2.442812764755719e-06,
|
| 2184 |
+
"loss": 0.0066,
|
| 2185 |
"step": 3070
|
| 2186 |
},
|
| 2187 |
{
|
| 2188 |
"epoch": 3.18,
|
| 2189 |
+
"grad_norm": 0.31712397933006287,
|
| 2190 |
"learning_rate": 2.414572154758543e-06,
|
| 2191 |
+
"loss": 0.876,
|
| 2192 |
"step": 3080
|
| 2193 |
},
|
| 2194 |
{
|
| 2195 |
"epoch": 3.19,
|
| 2196 |
+
"grad_norm": 320.600341796875,
|
| 2197 |
"learning_rate": 2.386331544761367e-06,
|
| 2198 |
+
"loss": 0.9575,
|
| 2199 |
"step": 3090
|
| 2200 |
},
|
| 2201 |
{
|
| 2202 |
"epoch": 3.19,
|
| 2203 |
+
"grad_norm": 0.16467230021953583,
|
| 2204 |
"learning_rate": 2.3580909347641913e-06,
|
| 2205 |
+
"loss": 0.884,
|
| 2206 |
"step": 3100
|
| 2207 |
},
|
| 2208 |
{
|
| 2209 |
"epoch": 3.19,
|
| 2210 |
+
"grad_norm": 0.028689516708254814,
|
| 2211 |
"learning_rate": 2.329850324767015e-06,
|
| 2212 |
+
"loss": 0.0067,
|
| 2213 |
"step": 3110
|
| 2214 |
},
|
| 2215 |
{
|
| 2216 |
"epoch": 3.19,
|
| 2217 |
+
"grad_norm": 0.3429773151874542,
|
| 2218 |
"learning_rate": 2.3016097147698394e-06,
|
| 2219 |
+
"loss": 1.0181,
|
| 2220 |
"step": 3120
|
| 2221 |
},
|
| 2222 |
{
|
| 2223 |
"epoch": 3.2,
|
| 2224 |
+
"grad_norm": 0.039928894490003586,
|
| 2225 |
"learning_rate": 2.2733691047726633e-06,
|
| 2226 |
+
"loss": 1.1309,
|
| 2227 |
"step": 3130
|
| 2228 |
},
|
| 2229 |
{
|
| 2230 |
"epoch": 3.2,
|
| 2231 |
+
"grad_norm": 0.0716266855597496,
|
| 2232 |
"learning_rate": 2.245128494775487e-06,
|
| 2233 |
+
"loss": 0.0055,
|
| 2234 |
"step": 3140
|
| 2235 |
},
|
| 2236 |
{
|
| 2237 |
"epoch": 3.2,
|
| 2238 |
+
"eval_accuracy": 0.8456375838926175,
|
| 2239 |
+
"eval_loss": 0.7402271628379822,
|
| 2240 |
+
"eval_runtime": 327.8476,
|
| 2241 |
+
"eval_samples_per_second": 1.818,
|
| 2242 |
+
"eval_steps_per_second": 1.818,
|
| 2243 |
"step": 3148
|
| 2244 |
},
|
| 2245 |
{
|
| 2246 |
"epoch": 4.0,
|
| 2247 |
+
"grad_norm": 0.2870597243309021,
|
| 2248 |
"learning_rate": 2.2168878847783115e-06,
|
| 2249 |
+
"loss": 0.9632,
|
| 2250 |
"step": 3150
|
| 2251 |
},
|
| 2252 |
{
|
| 2253 |
"epoch": 4.0,
|
| 2254 |
+
"grad_norm": 0.29839763045310974,
|
| 2255 |
"learning_rate": 2.1886472747811353e-06,
|
| 2256 |
+
"loss": 0.5113,
|
| 2257 |
"step": 3160
|
| 2258 |
},
|
| 2259 |
{
|
| 2260 |
"epoch": 4.01,
|
| 2261 |
+
"grad_norm": 0.14275802671909332,
|
| 2262 |
"learning_rate": 2.1604066647839596e-06,
|
| 2263 |
+
"loss": 1.0579,
|
| 2264 |
"step": 3170
|
| 2265 |
},
|
| 2266 |
{
|
| 2267 |
"epoch": 4.01,
|
| 2268 |
+
"grad_norm": 25.018110275268555,
|
| 2269 |
"learning_rate": 2.1321660547867835e-06,
|
| 2270 |
+
"loss": 1.0512,
|
| 2271 |
"step": 3180
|
| 2272 |
},
|
| 2273 |
{
|
| 2274 |
"epoch": 4.01,
|
| 2275 |
+
"grad_norm": 0.07139433175325394,
|
| 2276 |
"learning_rate": 2.1039254447896078e-06,
|
| 2277 |
"loss": 0.0045,
|
| 2278 |
"step": 3190
|
| 2279 |
},
|
| 2280 |
{
|
| 2281 |
"epoch": 4.01,
|
| 2282 |
+
"grad_norm": 0.07226335257291794,
|
| 2283 |
"learning_rate": 2.0756848347924316e-06,
|
| 2284 |
+
"loss": 0.1377,
|
| 2285 |
"step": 3200
|
| 2286 |
},
|
| 2287 |
{
|
| 2288 |
"epoch": 4.02,
|
| 2289 |
+
"grad_norm": 0.05169367790222168,
|
| 2290 |
"learning_rate": 2.047444224795256e-06,
|
| 2291 |
+
"loss": 0.0114,
|
| 2292 |
"step": 3210
|
| 2293 |
},
|
| 2294 |
{
|
| 2295 |
"epoch": 4.02,
|
| 2296 |
+
"grad_norm": 0.2018321454524994,
|
| 2297 |
"learning_rate": 2.01920361479808e-06,
|
| 2298 |
+
"loss": 0.5378,
|
| 2299 |
"step": 3220
|
| 2300 |
},
|
| 2301 |
{
|
| 2302 |
"epoch": 4.02,
|
| 2303 |
+
"grad_norm": 0.023068850859999657,
|
| 2304 |
"learning_rate": 1.9909630048009037e-06,
|
| 2305 |
+
"loss": 0.5436,
|
| 2306 |
"step": 3230
|
| 2307 |
},
|
| 2308 |
{
|
| 2309 |
"epoch": 4.02,
|
| 2310 |
+
"grad_norm": 0.12592564523220062,
|
| 2311 |
"learning_rate": 1.962722394803728e-06,
|
| 2312 |
+
"loss": 1.2217,
|
| 2313 |
"step": 3240
|
| 2314 |
},
|
| 2315 |
{
|
| 2316 |
"epoch": 4.03,
|
| 2317 |
+
"grad_norm": 81.48845672607422,
|
| 2318 |
"learning_rate": 1.934481784806552e-06,
|
| 2319 |
+
"loss": 0.9742,
|
| 2320 |
"step": 3250
|
| 2321 |
},
|
| 2322 |
{
|
| 2323 |
"epoch": 4.03,
|
| 2324 |
+
"grad_norm": 0.19448934495449066,
|
| 2325 |
"learning_rate": 1.906241174809376e-06,
|
| 2326 |
+
"loss": 0.4926,
|
| 2327 |
"step": 3260
|
| 2328 |
},
|
| 2329 |
{
|
| 2330 |
"epoch": 4.03,
|
| 2331 |
+
"grad_norm": 0.3121252655982971,
|
| 2332 |
"learning_rate": 1.8780005648122002e-06,
|
| 2333 |
+
"loss": 0.4974,
|
| 2334 |
"step": 3270
|
| 2335 |
},
|
| 2336 |
{
|
| 2337 |
"epoch": 4.03,
|
| 2338 |
+
"grad_norm": 0.3977165222167969,
|
| 2339 |
"learning_rate": 1.849759954815024e-06,
|
| 2340 |
+
"loss": 0.5231,
|
| 2341 |
"step": 3280
|
| 2342 |
},
|
| 2343 |
{
|
| 2344 |
"epoch": 4.04,
|
| 2345 |
+
"grad_norm": 0.19376535713672638,
|
| 2346 |
"learning_rate": 1.8215193448178481e-06,
|
| 2347 |
+
"loss": 1.344,
|
| 2348 |
"step": 3290
|
| 2349 |
},
|
| 2350 |
{
|
| 2351 |
"epoch": 4.04,
|
| 2352 |
+
"grad_norm": 0.12233041226863861,
|
| 2353 |
"learning_rate": 1.7932787348206724e-06,
|
| 2354 |
+
"loss": 0.0042,
|
| 2355 |
"step": 3300
|
| 2356 |
},
|
| 2357 |
{
|
| 2358 |
"epoch": 4.04,
|
| 2359 |
+
"grad_norm": 1.1810427904129028,
|
| 2360 |
"learning_rate": 1.7650381248234963e-06,
|
| 2361 |
+
"loss": 1.0738,
|
| 2362 |
"step": 3310
|
| 2363 |
},
|
| 2364 |
{
|
| 2365 |
"epoch": 4.04,
|
| 2366 |
+
"grad_norm": 0.048179544508457184,
|
| 2367 |
"learning_rate": 1.7367975148263202e-06,
|
| 2368 |
+
"loss": 0.4614,
|
| 2369 |
"step": 3320
|
| 2370 |
},
|
| 2371 |
{
|
| 2372 |
"epoch": 4.05,
|
| 2373 |
+
"grad_norm": 0.058793701231479645,
|
| 2374 |
"learning_rate": 1.7085569048291444e-06,
|
| 2375 |
+
"loss": 0.4894,
|
| 2376 |
"step": 3330
|
| 2377 |
},
|
| 2378 |
{
|
| 2379 |
"epoch": 4.05,
|
| 2380 |
+
"grad_norm": 0.033707864582538605,
|
| 2381 |
"learning_rate": 1.6803162948319685e-06,
|
| 2382 |
+
"loss": 1.4924,
|
| 2383 |
"step": 3340
|
| 2384 |
},
|
| 2385 |
{
|
| 2386 |
"epoch": 4.05,
|
| 2387 |
+
"grad_norm": 0.12816235423088074,
|
| 2388 |
"learning_rate": 1.6520756848347926e-06,
|
| 2389 |
+
"loss": 0.0034,
|
| 2390 |
"step": 3350
|
| 2391 |
},
|
| 2392 |
{
|
| 2393 |
"epoch": 4.05,
|
| 2394 |
+
"grad_norm": 0.07232780009508133,
|
| 2395 |
"learning_rate": 1.6238350748376167e-06,
|
| 2396 |
+
"loss": 0.4661,
|
| 2397 |
"step": 3360
|
| 2398 |
},
|
| 2399 |
{
|
| 2400 |
"epoch": 4.06,
|
| 2401 |
+
"grad_norm": 0.027900898829102516,
|
| 2402 |
"learning_rate": 1.5955944648404405e-06,
|
| 2403 |
+
"loss": 0.671,
|
| 2404 |
"step": 3370
|
| 2405 |
},
|
| 2406 |
{
|
| 2407 |
"epoch": 4.06,
|
| 2408 |
+
"grad_norm": 0.1532682329416275,
|
| 2409 |
"learning_rate": 1.5673538548432648e-06,
|
| 2410 |
+
"loss": 0.4657,
|
| 2411 |
"step": 3380
|
| 2412 |
},
|
| 2413 |
{
|
| 2414 |
"epoch": 4.06,
|
| 2415 |
+
"grad_norm": 0.3767637014389038,
|
| 2416 |
"learning_rate": 1.539113244846089e-06,
|
| 2417 |
+
"loss": 0.0093,
|
| 2418 |
"step": 3390
|
| 2419 |
},
|
| 2420 |
{
|
| 2421 |
"epoch": 4.06,
|
| 2422 |
+
"grad_norm": 23.31383514404297,
|
| 2423 |
"learning_rate": 1.5108726348489128e-06,
|
| 2424 |
+
"loss": 1.4704,
|
| 2425 |
"step": 3400
|
| 2426 |
},
|
| 2427 |
{
|
| 2428 |
"epoch": 4.07,
|
| 2429 |
+
"grad_norm": 113.44857788085938,
|
| 2430 |
"learning_rate": 1.482632024851737e-06,
|
| 2431 |
+
"loss": 1.4151,
|
| 2432 |
"step": 3410
|
| 2433 |
},
|
| 2434 |
{
|
| 2435 |
"epoch": 4.07,
|
| 2436 |
+
"grad_norm": 0.1415228694677353,
|
| 2437 |
"learning_rate": 1.454391414854561e-06,
|
| 2438 |
+
"loss": 0.0069,
|
| 2439 |
"step": 3420
|
| 2440 |
},
|
| 2441 |
{
|
| 2442 |
"epoch": 4.07,
|
| 2443 |
+
"grad_norm": 0.12466636300086975,
|
| 2444 |
"learning_rate": 1.426150804857385e-06,
|
| 2445 |
+
"loss": 0.4532,
|
| 2446 |
"step": 3430
|
| 2447 |
},
|
| 2448 |
{
|
| 2449 |
"epoch": 4.07,
|
| 2450 |
+
"grad_norm": 0.2385646253824234,
|
| 2451 |
"learning_rate": 1.397910194860209e-06,
|
| 2452 |
+
"loss": 0.0064,
|
| 2453 |
"step": 3440
|
| 2454 |
},
|
| 2455 |
{
|
| 2456 |
"epoch": 4.08,
|
| 2457 |
+
"grad_norm": 0.0742715522646904,
|
| 2458 |
"learning_rate": 1.3696695848630332e-06,
|
| 2459 |
+
"loss": 0.0047,
|
| 2460 |
"step": 3450
|
| 2461 |
},
|
| 2462 |
{
|
| 2463 |
"epoch": 4.08,
|
| 2464 |
+
"grad_norm": 1417.8148193359375,
|
| 2465 |
"learning_rate": 1.341428974865857e-06,
|
| 2466 |
+
"loss": 0.2286,
|
| 2467 |
"step": 3460
|
| 2468 |
},
|
| 2469 |
{
|
| 2470 |
"epoch": 4.08,
|
| 2471 |
+
"grad_norm": 0.026701288297772408,
|
| 2472 |
"learning_rate": 1.3131883648686813e-06,
|
| 2473 |
+
"loss": 0.4781,
|
| 2474 |
"step": 3470
|
| 2475 |
},
|
| 2476 |
{
|
| 2477 |
"epoch": 4.08,
|
| 2478 |
+
"grad_norm": 24.302400588989258,
|
| 2479 |
"learning_rate": 1.2849477548715054e-06,
|
| 2480 |
+
"loss": 1.4506,
|
| 2481 |
"step": 3480
|
| 2482 |
},
|
| 2483 |
{
|
| 2484 |
"epoch": 4.09,
|
| 2485 |
+
"grad_norm": 0.20589368045330048,
|
| 2486 |
"learning_rate": 1.2567071448743293e-06,
|
| 2487 |
+
"loss": 0.5483,
|
| 2488 |
"step": 3490
|
| 2489 |
},
|
| 2490 |
{
|
| 2491 |
"epoch": 4.09,
|
| 2492 |
+
"grad_norm": 23.55211067199707,
|
| 2493 |
"learning_rate": 1.2284665348771536e-06,
|
| 2494 |
+
"loss": 1.104,
|
| 2495 |
"step": 3500
|
| 2496 |
},
|
| 2497 |
{
|
| 2498 |
"epoch": 4.09,
|
| 2499 |
+
"grad_norm": 0.11621620506048203,
|
| 2500 |
"learning_rate": 1.2002259248799774e-06,
|
| 2501 |
+
"loss": 0.6127,
|
| 2502 |
"step": 3510
|
| 2503 |
},
|
| 2504 |
{
|
| 2505 |
"epoch": 4.09,
|
| 2506 |
+
"grad_norm": 0.12657718360424042,
|
| 2507 |
"learning_rate": 1.1719853148828015e-06,
|
| 2508 |
+
"loss": 0.4753,
|
| 2509 |
"step": 3520
|
| 2510 |
},
|
| 2511 |
{
|
| 2512 |
"epoch": 4.1,
|
| 2513 |
+
"grad_norm": 62.24040222167969,
|
| 2514 |
"learning_rate": 1.1437447048856256e-06,
|
| 2515 |
+
"loss": 0.9395,
|
| 2516 |
"step": 3530
|
| 2517 |
},
|
| 2518 |
{
|
| 2519 |
"epoch": 4.1,
|
| 2520 |
+
"grad_norm": 0.11141040176153183,
|
| 2521 |
"learning_rate": 1.1155040948884497e-06,
|
| 2522 |
+
"loss": 0.4753,
|
| 2523 |
"step": 3540
|
| 2524 |
},
|
| 2525 |
{
|
| 2526 |
"epoch": 4.1,
|
| 2527 |
+
"grad_norm": 23.662633895874023,
|
| 2528 |
"learning_rate": 1.0872634848912737e-06,
|
| 2529 |
+
"loss": 0.472,
|
| 2530 |
"step": 3550
|
| 2531 |
},
|
| 2532 |
{
|
| 2533 |
"epoch": 4.1,
|
| 2534 |
+
"grad_norm": 0.029395530000329018,
|
| 2535 |
"learning_rate": 1.0590228748940978e-06,
|
| 2536 |
+
"loss": 0.0064,
|
| 2537 |
"step": 3560
|
| 2538 |
},
|
| 2539 |
{
|
| 2540 |
"epoch": 4.11,
|
| 2541 |
+
"grad_norm": 0.03742952272295952,
|
| 2542 |
"learning_rate": 1.0307822648969219e-06,
|
| 2543 |
+
"loss": 0.469,
|
| 2544 |
"step": 3570
|
| 2545 |
},
|
| 2546 |
{
|
| 2547 |
"epoch": 4.11,
|
| 2548 |
+
"grad_norm": 0.08150259405374527,
|
| 2549 |
"learning_rate": 1.002541654899746e-06,
|
| 2550 |
+
"loss": 0.4674,
|
| 2551 |
"step": 3580
|
| 2552 |
},
|
| 2553 |
{
|
| 2554 |
"epoch": 4.11,
|
| 2555 |
+
"grad_norm": 0.05578187480568886,
|
| 2556 |
"learning_rate": 9.7430104490257e-07,
|
| 2557 |
+
"loss": 0.5241,
|
| 2558 |
"step": 3590
|
| 2559 |
},
|
| 2560 |
{
|
| 2561 |
"epoch": 4.11,
|
| 2562 |
+
"grad_norm": 56.26320266723633,
|
| 2563 |
"learning_rate": 9.460604349053941e-07,
|
| 2564 |
+
"loss": 1.0151,
|
| 2565 |
"step": 3600
|
| 2566 |
},
|
| 2567 |
{
|
| 2568 |
"epoch": 4.12,
|
| 2569 |
+
"grad_norm": 0.1536935567855835,
|
| 2570 |
"learning_rate": 9.178198249082181e-07,
|
| 2571 |
+
"loss": 1.9666,
|
| 2572 |
"step": 3610
|
| 2573 |
},
|
| 2574 |
{
|
| 2575 |
"epoch": 4.12,
|
| 2576 |
+
"grad_norm": 2.32802677154541,
|
| 2577 |
"learning_rate": 8.895792149110422e-07,
|
| 2578 |
+
"loss": 0.4655,
|
| 2579 |
"step": 3620
|
| 2580 |
},
|
| 2581 |
{
|
| 2582 |
"epoch": 4.12,
|
| 2583 |
+
"grad_norm": 619.9100952148438,
|
| 2584 |
"learning_rate": 8.613386049138663e-07,
|
| 2585 |
+
"loss": 0.788,
|
| 2586 |
"step": 3630
|
| 2587 |
},
|
| 2588 |
{
|
| 2589 |
"epoch": 4.13,
|
| 2590 |
+
"grad_norm": 34.0890007019043,
|
| 2591 |
"learning_rate": 8.330979949166902e-07,
|
| 2592 |
+
"loss": 1.3822,
|
| 2593 |
"step": 3640
|
| 2594 |
},
|
| 2595 |
{
|
| 2596 |
"epoch": 4.13,
|
| 2597 |
+
"grad_norm": 0.2547508180141449,
|
| 2598 |
"learning_rate": 8.048573849195143e-07,
|
| 2599 |
+
"loss": 0.8994,
|
| 2600 |
"step": 3650
|
| 2601 |
},
|
| 2602 |
{
|
| 2603 |
"epoch": 4.13,
|
| 2604 |
+
"grad_norm": 0.27960023283958435,
|
| 2605 |
"learning_rate": 7.766167749223384e-07,
|
| 2606 |
+
"loss": 0.5559,
|
| 2607 |
"step": 3660
|
| 2608 |
},
|
| 2609 |
{
|
| 2610 |
"epoch": 4.13,
|
| 2611 |
+
"grad_norm": 0.05300937965512276,
|
| 2612 |
"learning_rate": 7.483761649251624e-07,
|
| 2613 |
+
"loss": 0.4556,
|
| 2614 |
"step": 3670
|
| 2615 |
},
|
| 2616 |
{
|
| 2617 |
"epoch": 4.14,
|
| 2618 |
+
"grad_norm": 0.024076635017991066,
|
| 2619 |
"learning_rate": 7.201355549279864e-07,
|
| 2620 |
+
"loss": 0.0023,
|
| 2621 |
"step": 3680
|
| 2622 |
},
|
| 2623 |
{
|
| 2624 |
"epoch": 4.14,
|
| 2625 |
+
"grad_norm": 0.27708545327186584,
|
| 2626 |
"learning_rate": 6.918949449308106e-07,
|
| 2627 |
+
"loss": 1.2558,
|
| 2628 |
"step": 3690
|
| 2629 |
},
|
| 2630 |
{
|
| 2631 |
"epoch": 4.14,
|
| 2632 |
+
"grad_norm": 25.573650360107422,
|
| 2633 |
"learning_rate": 6.636543349336347e-07,
|
| 2634 |
+
"loss": 2.7475,
|
| 2635 |
"step": 3700
|
| 2636 |
},
|
| 2637 |
{
|
| 2638 |
"epoch": 4.14,
|
| 2639 |
+
"grad_norm": 0.28637179732322693,
|
| 2640 |
"learning_rate": 6.354137249364587e-07,
|
| 2641 |
+
"loss": 0.8946,
|
| 2642 |
"step": 3710
|
| 2643 |
},
|
| 2644 |
{
|
| 2645 |
"epoch": 4.15,
|
| 2646 |
+
"grad_norm": 0.1498030722141266,
|
| 2647 |
"learning_rate": 6.071731149392827e-07,
|
| 2648 |
+
"loss": 0.0068,
|
| 2649 |
"step": 3720
|
| 2650 |
},
|
| 2651 |
{
|
| 2652 |
"epoch": 4.15,
|
| 2653 |
+
"grad_norm": 0.03544044494628906,
|
| 2654 |
"learning_rate": 5.789325049421068e-07,
|
| 2655 |
+
"loss": 0.1509,
|
| 2656 |
"step": 3730
|
| 2657 |
},
|
| 2658 |
{
|
| 2659 |
"epoch": 4.15,
|
| 2660 |
+
"grad_norm": 0.2865348756313324,
|
| 2661 |
"learning_rate": 5.506918949449309e-07,
|
| 2662 |
+
"loss": 0.0056,
|
| 2663 |
"step": 3740
|
| 2664 |
},
|
| 2665 |
{
|
| 2666 |
"epoch": 4.15,
|
| 2667 |
+
"grad_norm": 0.6191418170928955,
|
| 2668 |
"learning_rate": 5.224512849477549e-07,
|
| 2669 |
+
"loss": 0.987,
|
| 2670 |
"step": 3750
|
| 2671 |
},
|
| 2672 |
{
|
| 2673 |
"epoch": 4.16,
|
| 2674 |
+
"grad_norm": 0.03451582416892052,
|
| 2675 |
"learning_rate": 4.942106749505791e-07,
|
| 2676 |
+
"loss": 0.5444,
|
| 2677 |
"step": 3760
|
| 2678 |
},
|
| 2679 |
{
|
| 2680 |
"epoch": 4.16,
|
| 2681 |
+
"grad_norm": 0.3244670331478119,
|
| 2682 |
"learning_rate": 4.6597006495340303e-07,
|
| 2683 |
+
"loss": 0.6433,
|
| 2684 |
"step": 3770
|
| 2685 |
},
|
| 2686 |
{
|
| 2687 |
"epoch": 4.16,
|
| 2688 |
+
"grad_norm": 23.457420349121094,
|
| 2689 |
"learning_rate": 4.3772945495622706e-07,
|
| 2690 |
+
"loss": 1.4575,
|
| 2691 |
"step": 3780
|
| 2692 |
},
|
| 2693 |
{
|
| 2694 |
"epoch": 4.16,
|
| 2695 |
+
"grad_norm": 0.2671697735786438,
|
| 2696 |
"learning_rate": 4.094888449590512e-07,
|
| 2697 |
+
"loss": 0.0069,
|
| 2698 |
"step": 3790
|
| 2699 |
},
|
| 2700 |
{
|
| 2701 |
"epoch": 4.17,
|
| 2702 |
+
"grad_norm": 0.0375777892768383,
|
| 2703 |
"learning_rate": 3.812482349618752e-07,
|
| 2704 |
+
"loss": 0.5398,
|
| 2705 |
"step": 3800
|
| 2706 |
},
|
| 2707 |
{
|
| 2708 |
"epoch": 4.17,
|
| 2709 |
+
"grad_norm": 0.7031461596488953,
|
| 2710 |
"learning_rate": 3.530076249646993e-07,
|
| 2711 |
+
"loss": 0.0067,
|
| 2712 |
"step": 3810
|
| 2713 |
},
|
| 2714 |
{
|
| 2715 |
"epoch": 4.17,
|
| 2716 |
+
"grad_norm": 0.027200641110539436,
|
| 2717 |
"learning_rate": 3.247670149675233e-07,
|
| 2718 |
+
"loss": 0.0063,
|
| 2719 |
"step": 3820
|
| 2720 |
},
|
| 2721 |
{
|
| 2722 |
"epoch": 4.17,
|
| 2723 |
+
"grad_norm": 0.1259397715330124,
|
| 2724 |
"learning_rate": 2.965264049703474e-07,
|
| 2725 |
+
"loss": 0.619,
|
| 2726 |
"step": 3830
|
| 2727 |
},
|
| 2728 |
{
|
| 2729 |
"epoch": 4.18,
|
| 2730 |
+
"grad_norm": 0.12452813237905502,
|
| 2731 |
"learning_rate": 2.682857949731714e-07,
|
| 2732 |
+
"loss": 0.0046,
|
| 2733 |
"step": 3840
|
| 2734 |
},
|
| 2735 |
{
|
| 2736 |
"epoch": 4.18,
|
| 2737 |
+
"grad_norm": 0.25807440280914307,
|
| 2738 |
"learning_rate": 2.400451849759955e-07,
|
| 2739 |
+
"loss": 0.6994,
|
| 2740 |
"step": 3850
|
| 2741 |
},
|
| 2742 |
{
|
| 2743 |
"epoch": 4.18,
|
| 2744 |
+
"grad_norm": 0.13341209292411804,
|
| 2745 |
"learning_rate": 2.1180457497881955e-07,
|
| 2746 |
+
"loss": 0.5316,
|
| 2747 |
"step": 3860
|
| 2748 |
},
|
| 2749 |
{
|
| 2750 |
"epoch": 4.18,
|
| 2751 |
+
"grad_norm": 22.768707275390625,
|
| 2752 |
"learning_rate": 1.8356396498164363e-07,
|
| 2753 |
+
"loss": 1.5915,
|
| 2754 |
"step": 3870
|
| 2755 |
},
|
| 2756 |
{
|
| 2757 |
"epoch": 4.19,
|
| 2758 |
+
"grad_norm": 0.050129234790802,
|
| 2759 |
"learning_rate": 1.5532335498446768e-07,
|
| 2760 |
+
"loss": 1.065,
|
| 2761 |
"step": 3880
|
| 2762 |
},
|
| 2763 |
{
|
| 2764 |
"epoch": 4.19,
|
| 2765 |
+
"grad_norm": 0.10654575377702713,
|
| 2766 |
"learning_rate": 1.2708274498729173e-07,
|
| 2767 |
+
"loss": 1.1427,
|
| 2768 |
"step": 3890
|
| 2769 |
},
|
| 2770 |
{
|
| 2771 |
"epoch": 4.19,
|
| 2772 |
+
"grad_norm": 0.04145923629403114,
|
| 2773 |
"learning_rate": 9.884213499011579e-08,
|
| 2774 |
+
"loss": 0.0048,
|
| 2775 |
"step": 3900
|
| 2776 |
},
|
| 2777 |
{
|
| 2778 |
"epoch": 4.19,
|
| 2779 |
+
"grad_norm": 35.09569549560547,
|
| 2780 |
"learning_rate": 7.060152499293986e-08,
|
| 2781 |
+
"loss": 1.6677,
|
| 2782 |
"step": 3910
|
| 2783 |
},
|
| 2784 |
{
|
| 2785 |
"epoch": 4.2,
|
| 2786 |
+
"grad_norm": 0.2241126149892807,
|
| 2787 |
"learning_rate": 4.236091499576391e-08,
|
| 2788 |
+
"loss": 0.4727,
|
| 2789 |
"step": 3920
|
| 2790 |
},
|
| 2791 |
{
|
| 2792 |
"epoch": 4.2,
|
| 2793 |
+
"grad_norm": 0.043600454926490784,
|
| 2794 |
"learning_rate": 1.4120304998587971e-08,
|
| 2795 |
+
"loss": 0.0042,
|
| 2796 |
"step": 3930
|
| 2797 |
},
|
| 2798 |
{
|
| 2799 |
"epoch": 4.2,
|
| 2800 |
+
"eval_accuracy": 0.8104026845637584,
|
| 2801 |
+
"eval_loss": 0.8931716680526733,
|
| 2802 |
+
"eval_runtime": 327.1353,
|
| 2803 |
+
"eval_samples_per_second": 1.822,
|
| 2804 |
+
"eval_steps_per_second": 1.822,
|
| 2805 |
"step": 3935
|
| 2806 |
},
|
| 2807 |
{
|
| 2808 |
"epoch": 4.2,
|
| 2809 |
"step": 3935,
|
| 2810 |
"total_flos": 1.7278688550692782e+19,
|
| 2811 |
+
"train_loss": 0.7462392025486373,
|
| 2812 |
+
"train_runtime": 7487.9132,
|
| 2813 |
+
"train_samples_per_second": 0.526,
|
| 2814 |
+
"train_steps_per_second": 0.526
|
| 2815 |
},
|
| 2816 |
{
|
| 2817 |
"epoch": 4.2,
|
| 2818 |
+
"eval_accuracy": 0.8456375838926175,
|
| 2819 |
+
"eval_loss": 0.7402271628379822,
|
| 2820 |
+
"eval_runtime": 321.7128,
|
| 2821 |
+
"eval_samples_per_second": 1.853,
|
| 2822 |
+
"eval_steps_per_second": 1.853,
|
| 2823 |
"step": 3935
|
| 2824 |
},
|
| 2825 |
{
|
| 2826 |
"epoch": 4.2,
|
| 2827 |
+
"eval_accuracy": 0.8456375838926175,
|
| 2828 |
+
"eval_loss": 0.7402271032333374,
|
| 2829 |
+
"eval_runtime": 328.0839,
|
| 2830 |
+
"eval_samples_per_second": 1.817,
|
| 2831 |
+
"eval_steps_per_second": 1.817,
|
| 2832 |
"step": 3935
|
| 2833 |
}
|
| 2834 |
],
|