End of training
Browse files- README.md +2 -2
- all_results.json +10 -10
- eval_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +727 -1983
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -16,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# full_review
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 1.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 16 |
|
| 17 |
# full_review
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on the openreview_full_review dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 1.6258
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"eval_loss": 1.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second": 11.
|
| 7 |
-
"perplexity": 5.
|
| 8 |
-
"total_flos": 1.
|
| 9 |
-
"train_loss": 1.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.6257721185684204,
|
| 4 |
+
"eval_runtime": 4.2134,
|
| 5 |
+
"eval_samples_per_second": 90.426,
|
| 6 |
+
"eval_steps_per_second": 11.392,
|
| 7 |
+
"perplexity": 5.082341692999726,
|
| 8 |
+
"total_flos": 1.0291750314285466e+19,
|
| 9 |
+
"train_loss": 1.6438454094579664,
|
| 10 |
+
"train_runtime": 3711.4901,
|
| 11 |
+
"train_samples_per_second": 30.48,
|
| 12 |
+
"train_steps_per_second": 0.477
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"eval_loss": 1.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second": 11.
|
| 7 |
-
"perplexity": 5.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.6257721185684204,
|
| 4 |
+
"eval_runtime": 4.2134,
|
| 5 |
+
"eval_samples_per_second": 90.426,
|
| 6 |
+
"eval_steps_per_second": 11.392,
|
| 7 |
+
"perplexity": 5.082341692999726
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss": 1.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.0291750314285466e+19,
|
| 4 |
+
"train_loss": 1.6438454094579664,
|
| 5 |
+
"train_runtime": 3711.4901,
|
| 6 |
+
"train_samples_per_second": 30.48,
|
| 7 |
+
"train_steps_per_second": 0.477
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -3,2534 +3,1278 @@
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 3.0,
|
| 5 |
"eval_steps": 600,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss": 1.
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate":
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate":
|
| 197 |
-
"loss": 1.
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss": 1.
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 4.
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate": 4.
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"learning_rate": 4.
|
| 225 |
-
"loss": 1.
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
-
"learning_rate": 4.
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 4.
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate": 4.
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch": 0.
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate": 4.
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate": 4.
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate": 4.
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch": 0.
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate": 4.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
-
"learning_rate": 4.
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 0.
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
-
"learning_rate": 4.
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 0.
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
-
"learning_rate": 4.
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
-
"learning_rate": 4.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0.
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
-
"learning_rate": 4.
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch": 0.
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
-
"learning_rate": 4.
|
| 316 |
-
"loss": 1.
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate": 4.
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch": 0.
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate": 4.
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch": 0.
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 4.
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate": 4.
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 0.
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate": 4.
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch": 0.
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate": 4.
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
-
"learning_rate": 4.
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch": 0.
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
-
"learning_rate": 4.
|
| 372 |
-
"loss": 1.
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
-
"epoch": 0.
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate": 4.
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate": 4.
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 0.
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
-
"learning_rate": 4.
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch": 0.
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
-
"learning_rate": 4.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate": 4.
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch": 0.
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
-
"learning_rate": 4.
|
| 414 |
-
"loss": 1.
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch": 0
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
-
"learning_rate": 4.
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch":
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate": 4.
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"eval_loss": 1.
|
| 434 |
-
"eval_runtime":
|
| 435 |
-
"eval_samples_per_second":
|
| 436 |
-
"eval_steps_per_second":
|
| 437 |
"step": 600
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
-
"learning_rate": 4.
|
| 443 |
-
"loss": 1.
|
| 444 |
"step": 610
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch":
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate": 4.
|
| 450 |
-
"loss": 1.
|
| 451 |
"step": 620
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch":
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
-
"learning_rate": 4.
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 630
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
-
"learning_rate": 4.
|
| 464 |
-
"loss": 1.
|
| 465 |
"step": 640
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss": 1.
|
| 472 |
"step": 650
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 1.
|
| 479 |
"step": 660
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss": 1.
|
| 486 |
"step": 670
|
| 487 |
},
|
| 488 |
{
|
| 489 |
-
"epoch":
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss": 1.
|
| 493 |
"step": 680
|
| 494 |
},
|
| 495 |
{
|
| 496 |
-
"epoch":
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 690
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch":
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 700
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"learning_rate":
|
| 513 |
-
"loss": 1.
|
| 514 |
"step": 710
|
| 515 |
},
|
| 516 |
{
|
| 517 |
-
"epoch":
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
-
"learning_rate":
|
| 520 |
-
"loss": 1.
|
| 521 |
"step": 720
|
| 522 |
},
|
| 523 |
{
|
| 524 |
-
"epoch":
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss": 1.
|
| 528 |
"step": 730
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch":
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
-
"learning_rate":
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 740
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch":
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 750
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch":
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 760
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
-
"learning_rate":
|
| 555 |
-
"loss": 1.
|
| 556 |
"step": 770
|
| 557 |
},
|
| 558 |
{
|
| 559 |
-
"epoch":
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
-
"learning_rate":
|
| 562 |
-
"loss": 1.
|
| 563 |
"step": 780
|
| 564 |
},
|
| 565 |
{
|
| 566 |
-
"epoch":
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
-
"learning_rate":
|
| 569 |
-
"loss": 1.
|
| 570 |
"step": 790
|
| 571 |
},
|
| 572 |
{
|
| 573 |
-
"epoch":
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
-
"learning_rate":
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 800
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch":
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
-
"learning_rate":
|
| 583 |
-
"loss": 1.
|
| 584 |
"step": 810
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch":
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
-
"learning_rate":
|
| 590 |
-
"loss": 1.
|
| 591 |
"step": 820
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"grad_norm": 0.
|
| 596 |
-
"learning_rate":
|
| 597 |
-
"loss": 1.
|
| 598 |
"step": 830
|
| 599 |
},
|
| 600 |
{
|
| 601 |
-
"epoch":
|
| 602 |
-
"grad_norm": 0.
|
| 603 |
-
"learning_rate":
|
| 604 |
-
"loss": 1.
|
| 605 |
"step": 840
|
| 606 |
},
|
| 607 |
{
|
| 608 |
-
"epoch":
|
| 609 |
-
"grad_norm": 0.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss": 1.
|
| 612 |
"step": 850
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch":
|
| 616 |
-
"grad_norm": 0.
|
| 617 |
-
"learning_rate":
|
| 618 |
-
"loss": 1.
|
| 619 |
"step": 860
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch":
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss": 1.
|
| 626 |
"step": 870
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"grad_norm": 0.
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss": 1.
|
| 633 |
"step": 880
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
-
"learning_rate":
|
| 639 |
-
"loss": 1.
|
| 640 |
"step": 890
|
| 641 |
},
|
| 642 |
{
|
| 643 |
-
"epoch":
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
-
"learning_rate":
|
| 646 |
-
"loss": 1.
|
| 647 |
"step": 900
|
| 648 |
},
|
| 649 |
{
|
| 650 |
-
"epoch":
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss": 1.
|
| 654 |
"step": 910
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch":
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"loss": 1.
|
| 661 |
"step": 920
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss": 1.
|
| 668 |
"step": 930
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"grad_norm": 0.
|
| 673 |
-
"learning_rate":
|
| 674 |
-
"loss": 1.
|
| 675 |
"step": 940
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch":
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
-
"learning_rate":
|
| 681 |
-
"loss": 1.
|
| 682 |
"step": 950
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"epoch":
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
-
"learning_rate":
|
| 688 |
-
"loss": 1.
|
| 689 |
"step": 960
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss": 1.
|
| 696 |
"step": 970
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch":
|
| 700 |
-
"grad_norm": 0.
|
| 701 |
-
"learning_rate":
|
| 702 |
-
"loss": 1.
|
| 703 |
"step": 980
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch":
|
| 707 |
-
"grad_norm": 0.
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss": 1.
|
| 710 |
"step": 990
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"grad_norm": 0.
|
| 715 |
-
"learning_rate":
|
| 716 |
-
"loss": 1.
|
| 717 |
"step": 1000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
-
"grad_norm": 0.
|
| 722 |
-
"learning_rate":
|
| 723 |
-
"loss": 1.
|
| 724 |
"step": 1010
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch":
|
| 728 |
-
"grad_norm": 0.
|
| 729 |
-
"learning_rate":
|
| 730 |
-
"loss": 1.
|
| 731 |
"step": 1020
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch":
|
| 735 |
-
"grad_norm": 0.
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss": 1.
|
| 738 |
"step": 1030
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch":
|
| 742 |
-
"grad_norm": 0.
|
| 743 |
-
"learning_rate":
|
| 744 |
-
"loss": 1.
|
| 745 |
"step": 1040
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch":
|
| 749 |
-
"grad_norm": 0.
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss": 1.
|
| 752 |
"step": 1050
|
| 753 |
},
|
| 754 |
{
|
| 755 |
-
"epoch":
|
| 756 |
-
"grad_norm": 0.
|
| 757 |
-
"learning_rate":
|
| 758 |
-
"loss": 1.
|
| 759 |
"step": 1060
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
-
"grad_norm": 0.
|
| 764 |
-
"learning_rate":
|
| 765 |
-
"loss": 1.
|
| 766 |
"step": 1070
|
| 767 |
},
|
| 768 |
{
|
| 769 |
-
"epoch":
|
| 770 |
-
"grad_norm": 0.
|
| 771 |
-
"learning_rate":
|
| 772 |
-
"loss": 1.
|
| 773 |
"step": 1080
|
| 774 |
},
|
| 775 |
{
|
| 776 |
-
"epoch":
|
| 777 |
-
"grad_norm": 0.
|
| 778 |
-
"learning_rate":
|
| 779 |
-
"loss": 1.
|
| 780 |
"step": 1090
|
| 781 |
},
|
| 782 |
{
|
| 783 |
-
"epoch":
|
| 784 |
-
"grad_norm": 0.
|
| 785 |
-
"learning_rate":
|
| 786 |
-
"loss": 1.
|
| 787 |
"step": 1100
|
| 788 |
},
|
| 789 |
{
|
| 790 |
-
"epoch":
|
| 791 |
-
"grad_norm": 0.
|
| 792 |
-
"learning_rate":
|
| 793 |
-
"loss": 1.
|
| 794 |
"step": 1110
|
| 795 |
},
|
| 796 |
{
|
| 797 |
-
"epoch":
|
| 798 |
-
"grad_norm": 0.
|
| 799 |
-
"learning_rate":
|
| 800 |
-
"loss": 1.
|
| 801 |
"step": 1120
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch":
|
| 805 |
-
"grad_norm": 0.
|
| 806 |
-
"learning_rate":
|
| 807 |
-
"loss": 1.
|
| 808 |
"step": 1130
|
| 809 |
},
|
| 810 |
{
|
| 811 |
-
"epoch":
|
| 812 |
-
"grad_norm": 0.
|
| 813 |
-
"learning_rate":
|
| 814 |
-
"loss": 1.
|
| 815 |
"step": 1140
|
| 816 |
},
|
| 817 |
{
|
| 818 |
-
"epoch":
|
| 819 |
-
"grad_norm": 0.
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 1.
|
| 822 |
"step": 1150
|
| 823 |
},
|
| 824 |
{
|
| 825 |
-
"epoch":
|
| 826 |
-
"grad_norm": 0.
|
| 827 |
-
"learning_rate":
|
| 828 |
-
"loss": 1.
|
| 829 |
"step": 1160
|
| 830 |
},
|
| 831 |
{
|
| 832 |
-
"epoch":
|
| 833 |
-
"grad_norm": 0.
|
| 834 |
-
"learning_rate":
|
| 835 |
-
"loss": 1.
|
| 836 |
"step": 1170
|
| 837 |
},
|
| 838 |
{
|
| 839 |
-
"epoch":
|
| 840 |
-
"grad_norm": 0.
|
| 841 |
-
"learning_rate":
|
| 842 |
-
"loss": 1.
|
| 843 |
"step": 1180
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch":
|
| 847 |
-
"grad_norm": 0.
|
| 848 |
-
"learning_rate":
|
| 849 |
-
"loss": 1.
|
| 850 |
"step": 1190
|
| 851 |
},
|
| 852 |
{
|
| 853 |
-
"epoch":
|
| 854 |
-
"grad_norm": 0.
|
| 855 |
-
"learning_rate":
|
| 856 |
-
"loss": 1.
|
| 857 |
"step": 1200
|
| 858 |
},
|
| 859 |
{
|
| 860 |
-
"epoch":
|
| 861 |
-
"eval_loss": 1.
|
| 862 |
-
"eval_runtime":
|
| 863 |
-
"eval_samples_per_second":
|
| 864 |
-
"eval_steps_per_second": 11.
|
| 865 |
"step": 1200
|
| 866 |
},
|
| 867 |
{
|
| 868 |
-
"epoch":
|
| 869 |
-
"grad_norm": 0.
|
| 870 |
-
"learning_rate":
|
| 871 |
-
"loss": 1.
|
| 872 |
"step": 1210
|
| 873 |
},
|
| 874 |
{
|
| 875 |
-
"epoch":
|
| 876 |
-
"grad_norm": 0.
|
| 877 |
-
"learning_rate":
|
| 878 |
-
"loss": 1.
|
| 879 |
"step": 1220
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch":
|
| 883 |
-
"grad_norm": 0.
|
| 884 |
-
"learning_rate":
|
| 885 |
-
"loss": 1.
|
| 886 |
"step": 1230
|
| 887 |
},
|
| 888 |
{
|
| 889 |
-
"epoch":
|
| 890 |
-
"grad_norm": 0.
|
| 891 |
-
"learning_rate":
|
| 892 |
-
"loss": 1.
|
| 893 |
"step": 1240
|
| 894 |
},
|
| 895 |
{
|
| 896 |
-
"epoch":
|
| 897 |
-
"grad_norm": 0.
|
| 898 |
-
"learning_rate":
|
| 899 |
-
"loss": 1.
|
| 900 |
"step": 1250
|
| 901 |
},
|
| 902 |
{
|
| 903 |
-
"epoch":
|
| 904 |
-
"grad_norm": 0.
|
| 905 |
-
"learning_rate":
|
| 906 |
-
"loss": 1.
|
| 907 |
"step": 1260
|
| 908 |
},
|
| 909 |
{
|
| 910 |
-
"epoch":
|
| 911 |
-
"grad_norm": 0.
|
| 912 |
-
"learning_rate":
|
| 913 |
-
"loss": 1.
|
| 914 |
"step": 1270
|
| 915 |
},
|
| 916 |
{
|
| 917 |
-
"epoch":
|
| 918 |
-
"grad_norm": 0.
|
| 919 |
-
"learning_rate":
|
| 920 |
-
"loss": 1.
|
| 921 |
"step": 1280
|
| 922 |
},
|
| 923 |
{
|
| 924 |
-
"epoch":
|
| 925 |
-
"grad_norm": 0.
|
| 926 |
-
"learning_rate":
|
| 927 |
-
"loss": 1.
|
| 928 |
"step": 1290
|
| 929 |
},
|
| 930 |
{
|
| 931 |
-
"epoch":
|
| 932 |
-
"grad_norm": 0.
|
| 933 |
-
"learning_rate":
|
| 934 |
-
"loss": 1.
|
| 935 |
"step": 1300
|
| 936 |
},
|
| 937 |
{
|
| 938 |
-
"epoch":
|
| 939 |
-
"grad_norm": 0.
|
| 940 |
-
"learning_rate":
|
| 941 |
-
"loss": 1.
|
| 942 |
"step": 1310
|
| 943 |
},
|
| 944 |
{
|
| 945 |
-
"epoch":
|
| 946 |
-
"grad_norm": 0.
|
| 947 |
-
"learning_rate":
|
| 948 |
-
"loss": 1.
|
| 949 |
"step": 1320
|
| 950 |
},
|
| 951 |
{
|
| 952 |
-
"epoch":
|
| 953 |
-
"grad_norm": 0.
|
| 954 |
-
"learning_rate":
|
| 955 |
-
"loss": 1.
|
| 956 |
"step": 1330
|
| 957 |
},
|
| 958 |
{
|
| 959 |
-
"epoch":
|
| 960 |
-
"grad_norm": 0.
|
| 961 |
-
"learning_rate":
|
| 962 |
-
"loss": 1.
|
| 963 |
"step": 1340
|
| 964 |
},
|
| 965 |
{
|
| 966 |
-
"epoch":
|
| 967 |
-
"grad_norm": 0.
|
| 968 |
-
"learning_rate":
|
| 969 |
-
"loss": 1.
|
| 970 |
"step": 1350
|
| 971 |
},
|
| 972 |
{
|
| 973 |
-
"epoch":
|
| 974 |
-
"grad_norm": 0.
|
| 975 |
-
"learning_rate":
|
| 976 |
-
"loss": 1.
|
| 977 |
"step": 1360
|
| 978 |
},
|
| 979 |
{
|
| 980 |
-
"epoch":
|
| 981 |
-
"grad_norm": 0.
|
| 982 |
-
"learning_rate":
|
| 983 |
-
"loss": 1.
|
| 984 |
"step": 1370
|
| 985 |
},
|
| 986 |
{
|
| 987 |
-
"epoch":
|
| 988 |
-
"grad_norm": 0.
|
| 989 |
-
"learning_rate":
|
| 990 |
-
"loss": 1.
|
| 991 |
"step": 1380
|
| 992 |
},
|
| 993 |
{
|
| 994 |
-
"epoch":
|
| 995 |
-
"grad_norm": 0.
|
| 996 |
-
"learning_rate":
|
| 997 |
-
"loss": 1.
|
| 998 |
"step": 1390
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
-
"epoch":
|
| 1002 |
-
"grad_norm": 0.
|
| 1003 |
-
"learning_rate":
|
| 1004 |
-
"loss": 1.
|
| 1005 |
"step": 1400
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
-
"epoch":
|
| 1009 |
-
"grad_norm": 0.
|
| 1010 |
-
"learning_rate":
|
| 1011 |
-
"loss": 1.
|
| 1012 |
"step": 1410
|
| 1013 |
},
|
| 1014 |
{
|
| 1015 |
-
"epoch":
|
| 1016 |
-
"grad_norm": 0.
|
| 1017 |
-
"learning_rate":
|
| 1018 |
-
"loss": 1.
|
| 1019 |
"step": 1420
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
-
"epoch":
|
| 1023 |
-
"grad_norm": 0.
|
| 1024 |
-
"learning_rate":
|
| 1025 |
-
"loss": 1.
|
| 1026 |
"step": 1430
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
-
"epoch":
|
| 1030 |
-
"grad_norm": 0.
|
| 1031 |
-
"learning_rate":
|
| 1032 |
-
"loss": 1.
|
| 1033 |
"step": 1440
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
-
"epoch":
|
| 1037 |
-
"grad_norm": 0.
|
| 1038 |
-
"learning_rate":
|
| 1039 |
-
"loss": 1.
|
| 1040 |
"step": 1450
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
-
"epoch":
|
| 1044 |
-
"grad_norm": 0.
|
| 1045 |
-
"learning_rate":
|
| 1046 |
-
"loss": 1.
|
| 1047 |
"step": 1460
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
-
"epoch":
|
| 1051 |
-
"grad_norm": 0.
|
| 1052 |
-
"learning_rate":
|
| 1053 |
-
"loss": 1.
|
| 1054 |
"step": 1470
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
-
"epoch":
|
| 1058 |
-
"grad_norm": 0.
|
| 1059 |
-
"learning_rate": 3.
|
| 1060 |
-
"loss": 1.
|
| 1061 |
"step": 1480
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
-
"epoch":
|
| 1065 |
-
"grad_norm": 0.
|
| 1066 |
-
"learning_rate": 3.
|
| 1067 |
-
"loss": 1.
|
| 1068 |
"step": 1490
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
-
"epoch":
|
| 1072 |
-
"grad_norm": 0.
|
| 1073 |
-
"learning_rate": 3.
|
| 1074 |
-
"loss": 1.
|
| 1075 |
"step": 1500
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
-
"epoch":
|
| 1079 |
-
"grad_norm": 0.
|
| 1080 |
-
"learning_rate": 3.
|
| 1081 |
-
"loss": 1.
|
| 1082 |
"step": 1510
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
-
"epoch":
|
| 1086 |
-
"grad_norm": 0.
|
| 1087 |
-
"learning_rate":
|
| 1088 |
-
"loss": 1.
|
| 1089 |
"step": 1520
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
-
"epoch":
|
| 1093 |
-
"grad_norm": 0.
|
| 1094 |
-
"learning_rate":
|
| 1095 |
-
"loss": 1.
|
| 1096 |
"step": 1530
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
-
"epoch":
|
| 1100 |
-
"grad_norm": 0.
|
| 1101 |
-
"learning_rate":
|
| 1102 |
-
"loss": 1.
|
| 1103 |
"step": 1540
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
-
"epoch":
|
| 1107 |
-
"grad_norm": 0.
|
| 1108 |
-
"learning_rate":
|
| 1109 |
-
"loss": 1.
|
| 1110 |
"step": 1550
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
-
"epoch":
|
| 1114 |
-
"grad_norm": 0.
|
| 1115 |
-
"learning_rate":
|
| 1116 |
-
"loss": 1.
|
| 1117 |
"step": 1560
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
-
"epoch":
|
| 1121 |
-
"grad_norm": 0.
|
| 1122 |
-
"learning_rate":
|
| 1123 |
-
"loss": 1.
|
| 1124 |
"step": 1570
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
-
"epoch":
|
| 1128 |
-
"grad_norm": 0.
|
| 1129 |
-
"learning_rate":
|
| 1130 |
-
"loss": 1.
|
| 1131 |
"step": 1580
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
-
"epoch":
|
| 1135 |
-
"grad_norm": 0.
|
| 1136 |
-
"learning_rate":
|
| 1137 |
-
"loss": 1.
|
| 1138 |
"step": 1590
|
| 1139 |
},
|
| 1140 |
{
|
| 1141 |
-
"epoch":
|
| 1142 |
-
"grad_norm": 0.
|
| 1143 |
-
"learning_rate":
|
| 1144 |
-
"loss": 1.
|
| 1145 |
"step": 1600
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
-
"epoch":
|
| 1149 |
-
"grad_norm": 0.
|
| 1150 |
-
"learning_rate":
|
| 1151 |
-
"loss": 1.
|
| 1152 |
"step": 1610
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
-
"epoch":
|
| 1156 |
-
"grad_norm": 0.
|
| 1157 |
-
"learning_rate":
|
| 1158 |
-
"loss": 1.
|
| 1159 |
"step": 1620
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
-
"epoch":
|
| 1163 |
-
"grad_norm": 0.
|
| 1164 |
-
"learning_rate":
|
| 1165 |
-
"loss": 1.
|
| 1166 |
"step": 1630
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
-
"epoch":
|
| 1170 |
-
"grad_norm": 0.
|
| 1171 |
-
"learning_rate":
|
| 1172 |
-
"loss": 1.
|
| 1173 |
"step": 1640
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
-
"epoch":
|
| 1177 |
-
"grad_norm": 0.
|
| 1178 |
-
"learning_rate":
|
| 1179 |
-
"loss": 1.
|
| 1180 |
"step": 1650
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
-
"epoch":
|
| 1184 |
-
"grad_norm": 0.
|
| 1185 |
-
"learning_rate":
|
| 1186 |
-
"loss": 1.
|
| 1187 |
"step": 1660
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
-
"epoch":
|
| 1191 |
-
"grad_norm": 0.
|
| 1192 |
-
"learning_rate":
|
| 1193 |
-
"loss": 1.
|
| 1194 |
"step": 1670
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
-
"epoch":
|
| 1198 |
-
"grad_norm": 0.
|
| 1199 |
-
"learning_rate": 3.
|
| 1200 |
-
"loss": 1.
|
| 1201 |
"step": 1680
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
-
"epoch":
|
| 1205 |
-
"grad_norm": 0.
|
| 1206 |
-
"learning_rate": 3.
|
| 1207 |
-
"loss": 1.
|
| 1208 |
"step": 1690
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
-
"epoch":
|
| 1212 |
-
"grad_norm": 0.
|
| 1213 |
-
"learning_rate":
|
| 1214 |
-
"loss": 1.
|
| 1215 |
"step": 1700
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
-
"epoch":
|
| 1219 |
-
"grad_norm": 0.
|
| 1220 |
-
"learning_rate":
|
| 1221 |
-
"loss": 1.
|
| 1222 |
"step": 1710
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
-
"epoch":
|
| 1226 |
-
"grad_norm": 0.
|
| 1227 |
-
"learning_rate":
|
| 1228 |
-
"loss": 1.
|
| 1229 |
"step": 1720
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
-
"epoch":
|
| 1233 |
-
"grad_norm": 0.
|
| 1234 |
-
"learning_rate":
|
| 1235 |
-
"loss": 1.
|
| 1236 |
"step": 1730
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
-
"epoch":
|
| 1240 |
-
"grad_norm": 0.
|
| 1241 |
-
"learning_rate":
|
| 1242 |
-
"loss": 1.
|
| 1243 |
"step": 1740
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
-
"epoch":
|
| 1247 |
-
"grad_norm": 0.
|
| 1248 |
-
"learning_rate":
|
| 1249 |
-
"loss": 1.
|
| 1250 |
"step": 1750
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
-
"epoch":
|
| 1254 |
-
"grad_norm": 0.
|
| 1255 |
-
"learning_rate":
|
| 1256 |
-
"loss": 1.
|
| 1257 |
"step": 1760
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
-
"epoch":
|
| 1261 |
-
"grad_norm": 0.
|
| 1262 |
-
"learning_rate":
|
| 1263 |
-
"loss": 1.
|
| 1264 |
"step": 1770
|
| 1265 |
},
|
| 1266 |
-
{
|
| 1267 |
-
"epoch": 1.5097540288379983,
|
| 1268 |
-
"grad_norm": 0.21272026002407074,
|
| 1269 |
-
"learning_rate": 2.9065539589269147e-05,
|
| 1270 |
-
"loss": 1.6252,
|
| 1271 |
-
"step": 1780
|
| 1272 |
-
},
|
| 1273 |
-
{
|
| 1274 |
-
"epoch": 1.5182357930449535,
|
| 1275 |
-
"grad_norm": 0.20586243271827698,
|
| 1276 |
-
"learning_rate": 2.8821882327445594e-05,
|
| 1277 |
-
"loss": 1.6216,
|
| 1278 |
-
"step": 1790
|
| 1279 |
-
},
|
| 1280 |
-
{
|
| 1281 |
-
"epoch": 1.5267175572519083,
|
| 1282 |
-
"grad_norm": 0.20619966089725494,
|
| 1283 |
-
"learning_rate": 2.857785275943458e-05,
|
| 1284 |
-
"loss": 1.6182,
|
| 1285 |
-
"step": 1800
|
| 1286 |
-
},
|
| 1287 |
-
{
|
| 1288 |
-
"epoch": 1.5267175572519083,
|
| 1289 |
-
"eval_loss": 1.6240788698196411,
|
| 1290 |
-
"eval_runtime": 8.3704,
|
| 1291 |
-
"eval_samples_per_second": 45.518,
|
| 1292 |
-
"eval_steps_per_second": 11.469,
|
| 1293 |
-
"step": 1800
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 1.5351993214588635,
|
| 1297 |
-
"grad_norm": 0.2109595090150833,
|
| 1298 |
-
"learning_rate": 2.8333474657218174e-05,
|
| 1299 |
-
"loss": 1.6002,
|
| 1300 |
-
"step": 1810
|
| 1301 |
-
},
|
| 1302 |
-
{
|
| 1303 |
-
"epoch": 1.5436810856658185,
|
| 1304 |
-
"grad_norm": 0.2052215188741684,
|
| 1305 |
-
"learning_rate": 2.8088771826730643e-05,
|
| 1306 |
-
"loss": 1.5973,
|
| 1307 |
-
"step": 1820
|
| 1308 |
-
},
|
| 1309 |
-
{
|
| 1310 |
-
"epoch": 1.5521628498727735,
|
| 1311 |
-
"grad_norm": 0.21743585169315338,
|
| 1312 |
-
"learning_rate": 2.784376810553947e-05,
|
| 1313 |
-
"loss": 1.6262,
|
| 1314 |
-
"step": 1830
|
| 1315 |
-
},
|
| 1316 |
-
{
|
| 1317 |
-
"epoch": 1.5606446140797285,
|
| 1318 |
-
"grad_norm": 0.20770463347434998,
|
| 1319 |
-
"learning_rate": 2.759848736052321e-05,
|
| 1320 |
-
"loss": 1.5753,
|
| 1321 |
-
"step": 1840
|
| 1322 |
-
},
|
| 1323 |
-
{
|
| 1324 |
-
"epoch": 1.5691263782866836,
|
| 1325 |
-
"grad_norm": 0.21669910848140717,
|
| 1326 |
-
"learning_rate": 2.735295348554651e-05,
|
| 1327 |
-
"loss": 1.6209,
|
| 1328 |
-
"step": 1850
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 1.5776081424936388,
|
| 1332 |
-
"grad_norm": 0.21204906702041626,
|
| 1333 |
-
"learning_rate": 2.7107190399132526e-05,
|
| 1334 |
-
"loss": 1.6088,
|
| 1335 |
-
"step": 1860
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 1.5860899067005936,
|
| 1339 |
-
"grad_norm": 0.2132139503955841,
|
| 1340 |
-
"learning_rate": 2.686122204213285e-05,
|
| 1341 |
-
"loss": 1.6252,
|
| 1342 |
-
"step": 1870
|
| 1343 |
-
},
|
| 1344 |
-
{
|
| 1345 |
-
"epoch": 1.5945716709075488,
|
| 1346 |
-
"grad_norm": 0.22568291425704956,
|
| 1347 |
-
"learning_rate": 2.6615072375395423e-05,
|
| 1348 |
-
"loss": 1.6112,
|
| 1349 |
-
"step": 1880
|
| 1350 |
-
},
|
| 1351 |
-
{
|
| 1352 |
-
"epoch": 1.6030534351145038,
|
| 1353 |
-
"grad_norm": 0.20681500434875488,
|
| 1354 |
-
"learning_rate": 2.6368765377430344e-05,
|
| 1355 |
-
"loss": 1.6146,
|
| 1356 |
-
"step": 1890
|
| 1357 |
-
},
|
| 1358 |
-
{
|
| 1359 |
-
"epoch": 1.6115351993214588,
|
| 1360 |
-
"grad_norm": 0.20736300945281982,
|
| 1361 |
-
"learning_rate": 2.6122325042074014e-05,
|
| 1362 |
-
"loss": 1.5834,
|
| 1363 |
-
"step": 1900
|
| 1364 |
-
},
|
| 1365 |
-
{
|
| 1366 |
-
"epoch": 1.620016963528414,
|
| 1367 |
-
"grad_norm": 0.20751895010471344,
|
| 1368 |
-
"learning_rate": 2.5875775376151823e-05,
|
| 1369 |
-
"loss": 1.6151,
|
| 1370 |
-
"step": 1910
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 1.6284987277353689,
|
| 1374 |
-
"grad_norm": 0.22000038623809814,
|
| 1375 |
-
"learning_rate": 2.562914039713953e-05,
|
| 1376 |
-
"loss": 1.5838,
|
| 1377 |
-
"step": 1920
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 1.636980491942324,
|
| 1381 |
-
"grad_norm": 0.20806461572647095,
|
| 1382 |
-
"learning_rate": 2.5382444130823612e-05,
|
| 1383 |
-
"loss": 1.6254,
|
| 1384 |
-
"step": 1930
|
| 1385 |
-
},
|
| 1386 |
-
{
|
| 1387 |
-
"epoch": 1.645462256149279,
|
| 1388 |
-
"grad_norm": 0.21619880199432373,
|
| 1389 |
-
"learning_rate": 2.5135710608960807e-05,
|
| 1390 |
-
"loss": 1.5999,
|
| 1391 |
-
"step": 1940
|
| 1392 |
-
},
|
| 1393 |
-
{
|
| 1394 |
-
"epoch": 1.6539440203562341,
|
| 1395 |
-
"grad_norm": 0.22222678363323212,
|
| 1396 |
-
"learning_rate": 2.4888963866937067e-05,
|
| 1397 |
-
"loss": 1.581,
|
| 1398 |
-
"step": 1950
|
| 1399 |
-
},
|
| 1400 |
-
{
|
| 1401 |
-
"epoch": 1.6624257845631891,
|
| 1402 |
-
"grad_norm": 0.21882720291614532,
|
| 1403 |
-
"learning_rate": 2.464222794142619e-05,
|
| 1404 |
-
"loss": 1.6151,
|
| 1405 |
-
"step": 1960
|
| 1406 |
-
},
|
| 1407 |
-
{
|
| 1408 |
-
"epoch": 1.6709075487701441,
|
| 1409 |
-
"grad_norm": 0.22589364647865295,
|
| 1410 |
-
"learning_rate": 2.43955268680483e-05,
|
| 1411 |
-
"loss": 1.6062,
|
| 1412 |
-
"step": 1970
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 1.6793893129770994,
|
| 1416 |
-
"grad_norm": 0.21262851357460022,
|
| 1417 |
-
"learning_rate": 2.4148884679028402e-05,
|
| 1418 |
-
"loss": 1.6096,
|
| 1419 |
-
"step": 1980
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 1.6878710771840542,
|
| 1423 |
-
"grad_norm": 0.21310704946517944,
|
| 1424 |
-
"learning_rate": 2.390232540085533e-05,
|
| 1425 |
-
"loss": 1.5932,
|
| 1426 |
-
"step": 1990
|
| 1427 |
-
},
|
| 1428 |
-
{
|
| 1429 |
-
"epoch": 1.6963528413910094,
|
| 1430 |
-
"grad_norm": 0.21306438744068146,
|
| 1431 |
-
"learning_rate": 2.3655873051941203e-05,
|
| 1432 |
-
"loss": 1.5809,
|
| 1433 |
-
"step": 2000
|
| 1434 |
-
},
|
| 1435 |
-
{
|
| 1436 |
-
"epoch": 1.7048346055979644,
|
| 1437 |
-
"grad_norm": 0.20825529098510742,
|
| 1438 |
-
"learning_rate": 2.340955164028171e-05,
|
| 1439 |
-
"loss": 1.5818,
|
| 1440 |
-
"step": 2010
|
| 1441 |
-
},
|
| 1442 |
-
{
|
| 1443 |
-
"epoch": 1.7133163698049194,
|
| 1444 |
-
"grad_norm": 0.21667224168777466,
|
| 1445 |
-
"learning_rate": 2.3163385161117367e-05,
|
| 1446 |
-
"loss": 1.5911,
|
| 1447 |
-
"step": 2020
|
| 1448 |
-
},
|
| 1449 |
-
{
|
| 1450 |
-
"epoch": 1.7217981340118744,
|
| 1451 |
-
"grad_norm": 0.2139502763748169,
|
| 1452 |
-
"learning_rate": 2.291739759459605e-05,
|
| 1453 |
-
"loss": 1.5816,
|
| 1454 |
-
"step": 2030
|
| 1455 |
-
},
|
| 1456 |
-
{
|
| 1457 |
-
"epoch": 1.7302798982188294,
|
| 1458 |
-
"grad_norm": 0.2165052443742752,
|
| 1459 |
-
"learning_rate": 2.267161290343694e-05,
|
| 1460 |
-
"loss": 1.6033,
|
| 1461 |
-
"step": 2040
|
| 1462 |
-
},
|
| 1463 |
-
{
|
| 1464 |
-
"epoch": 1.7387616624257847,
|
| 1465 |
-
"grad_norm": 0.2145167738199234,
|
| 1466 |
-
"learning_rate": 2.2426055030596282e-05,
|
| 1467 |
-
"loss": 1.6122,
|
| 1468 |
-
"step": 2050
|
| 1469 |
-
},
|
| 1470 |
-
{
|
| 1471 |
-
"epoch": 1.7472434266327395,
|
| 1472 |
-
"grad_norm": 0.2100827991962433,
|
| 1473 |
-
"learning_rate": 2.2180747896934956e-05,
|
| 1474 |
-
"loss": 1.5995,
|
| 1475 |
-
"step": 2060
|
| 1476 |
-
},
|
| 1477 |
-
{
|
| 1478 |
-
"epoch": 1.7557251908396947,
|
| 1479 |
-
"grad_norm": 0.22637659311294556,
|
| 1480 |
-
"learning_rate": 2.193571539888824e-05,
|
| 1481 |
-
"loss": 1.5938,
|
| 1482 |
-
"step": 2070
|
| 1483 |
-
},
|
| 1484 |
-
{
|
| 1485 |
-
"epoch": 1.7642069550466497,
|
| 1486 |
-
"grad_norm": 0.2245241105556488,
|
| 1487 |
-
"learning_rate": 2.169098140613797e-05,
|
| 1488 |
-
"loss": 1.5792,
|
| 1489 |
-
"step": 2080
|
| 1490 |
-
},
|
| 1491 |
-
{
|
| 1492 |
-
"epoch": 1.7726887192536047,
|
| 1493 |
-
"grad_norm": 0.2259100377559662,
|
| 1494 |
-
"learning_rate": 2.1446569759287267e-05,
|
| 1495 |
-
"loss": 1.5856,
|
| 1496 |
-
"step": 2090
|
| 1497 |
-
},
|
| 1498 |
-
{
|
| 1499 |
-
"epoch": 1.78117048346056,
|
| 1500 |
-
"grad_norm": 0.22213774919509888,
|
| 1501 |
-
"learning_rate": 2.120250426753815e-05,
|
| 1502 |
-
"loss": 1.5875,
|
| 1503 |
-
"step": 2100
|
| 1504 |
-
},
|
| 1505 |
-
{
|
| 1506 |
-
"epoch": 1.7896522476675147,
|
| 1507 |
-
"grad_norm": 0.22346043586730957,
|
| 1508 |
-
"learning_rate": 2.095880870637216e-05,
|
| 1509 |
-
"loss": 1.5909,
|
| 1510 |
-
"step": 2110
|
| 1511 |
-
},
|
| 1512 |
-
{
|
| 1513 |
-
"epoch": 1.79813401187447,
|
| 1514 |
-
"grad_norm": 0.2293236404657364,
|
| 1515 |
-
"learning_rate": 2.071550681523429e-05,
|
| 1516 |
-
"loss": 1.5812,
|
| 1517 |
-
"step": 2120
|
| 1518 |
-
},
|
| 1519 |
-
{
|
| 1520 |
-
"epoch": 1.806615776081425,
|
| 1521 |
-
"grad_norm": 0.22071833908557892,
|
| 1522 |
-
"learning_rate": 2.0472622295220413e-05,
|
| 1523 |
-
"loss": 1.5909,
|
| 1524 |
-
"step": 2130
|
| 1525 |
-
},
|
| 1526 |
-
{
|
| 1527 |
-
"epoch": 1.81509754028838,
|
| 1528 |
-
"grad_norm": 0.22376208007335663,
|
| 1529 |
-
"learning_rate": 2.023017880676849e-05,
|
| 1530 |
-
"loss": 1.5852,
|
| 1531 |
-
"step": 2140
|
| 1532 |
-
},
|
| 1533 |
-
{
|
| 1534 |
-
"epoch": 1.823579304495335,
|
| 1535 |
-
"grad_norm": 0.23016902804374695,
|
| 1536 |
-
"learning_rate": 1.9988199967353642e-05,
|
| 1537 |
-
"loss": 1.5997,
|
| 1538 |
-
"step": 2150
|
| 1539 |
-
},
|
| 1540 |
-
{
|
| 1541 |
-
"epoch": 1.83206106870229,
|
| 1542 |
-
"grad_norm": 0.22549933195114136,
|
| 1543 |
-
"learning_rate": 1.974670934918753e-05,
|
| 1544 |
-
"loss": 1.6062,
|
| 1545 |
-
"step": 2160
|
| 1546 |
-
},
|
| 1547 |
-
{
|
| 1548 |
-
"epoch": 1.8405428329092453,
|
| 1549 |
-
"grad_norm": 0.2221376597881317,
|
| 1550 |
-
"learning_rate": 1.9505730476922064e-05,
|
| 1551 |
-
"loss": 1.6112,
|
| 1552 |
-
"step": 2170
|
| 1553 |
-
},
|
| 1554 |
-
{
|
| 1555 |
-
"epoch": 1.8490245971162,
|
| 1556 |
-
"grad_norm": 0.22362485527992249,
|
| 1557 |
-
"learning_rate": 1.9265286825357742e-05,
|
| 1558 |
-
"loss": 1.6001,
|
| 1559 |
-
"step": 2180
|
| 1560 |
-
},
|
| 1561 |
-
{
|
| 1562 |
-
"epoch": 1.8575063613231553,
|
| 1563 |
-
"grad_norm": 0.22703245282173157,
|
| 1564 |
-
"learning_rate": 1.9025401817156905e-05,
|
| 1565 |
-
"loss": 1.6122,
|
| 1566 |
-
"step": 2190
|
| 1567 |
-
},
|
| 1568 |
-
{
|
| 1569 |
-
"epoch": 1.8659881255301103,
|
| 1570 |
-
"grad_norm": 0.2253347635269165,
|
| 1571 |
-
"learning_rate": 1.8786098820561972e-05,
|
| 1572 |
-
"loss": 1.5837,
|
| 1573 |
-
"step": 2200
|
| 1574 |
-
},
|
| 1575 |
-
{
|
| 1576 |
-
"epoch": 1.8744698897370653,
|
| 1577 |
-
"grad_norm": 0.2247924655675888,
|
| 1578 |
-
"learning_rate": 1.8547401147119127e-05,
|
| 1579 |
-
"loss": 1.6056,
|
| 1580 |
-
"step": 2210
|
| 1581 |
-
},
|
| 1582 |
-
{
|
| 1583 |
-
"epoch": 1.8829516539440203,
|
| 1584 |
-
"grad_norm": 0.2206358015537262,
|
| 1585 |
-
"learning_rate": 1.8309332049407363e-05,
|
| 1586 |
-
"loss": 1.5898,
|
| 1587 |
-
"step": 2220
|
| 1588 |
-
},
|
| 1589 |
-
{
|
| 1590 |
-
"epoch": 1.8914334181509753,
|
| 1591 |
-
"grad_norm": 0.23167766630649567,
|
| 1592 |
-
"learning_rate": 1.8071914718773393e-05,
|
| 1593 |
-
"loss": 1.5803,
|
| 1594 |
-
"step": 2230
|
| 1595 |
-
},
|
| 1596 |
-
{
|
| 1597 |
-
"epoch": 1.8999151823579306,
|
| 1598 |
-
"grad_norm": 0.22543945908546448,
|
| 1599 |
-
"learning_rate": 1.783517228307245e-05,
|
| 1600 |
-
"loss": 1.5908,
|
| 1601 |
-
"step": 2240
|
| 1602 |
-
},
|
| 1603 |
-
{
|
| 1604 |
-
"epoch": 1.9083969465648853,
|
| 1605 |
-
"grad_norm": 0.22845198214054108,
|
| 1606 |
-
"learning_rate": 1.7599127804415317e-05,
|
| 1607 |
-
"loss": 1.6114,
|
| 1608 |
-
"step": 2250
|
| 1609 |
-
},
|
| 1610 |
-
{
|
| 1611 |
-
"epoch": 1.9168787107718406,
|
| 1612 |
-
"grad_norm": 0.22125060856342316,
|
| 1613 |
-
"learning_rate": 1.7363804276921748e-05,
|
| 1614 |
-
"loss": 1.5835,
|
| 1615 |
-
"step": 2260
|
| 1616 |
-
},
|
| 1617 |
-
{
|
| 1618 |
-
"epoch": 1.9253604749787956,
|
| 1619 |
-
"grad_norm": 0.22639022767543793,
|
| 1620 |
-
"learning_rate": 1.7129224624480506e-05,
|
| 1621 |
-
"loss": 1.5937,
|
| 1622 |
-
"step": 2270
|
| 1623 |
-
},
|
| 1624 |
-
{
|
| 1625 |
-
"epoch": 1.9338422391857506,
|
| 1626 |
-
"grad_norm": 0.23626133799552917,
|
| 1627 |
-
"learning_rate": 1.689541169851625e-05,
|
| 1628 |
-
"loss": 1.6138,
|
| 1629 |
-
"step": 2280
|
| 1630 |
-
},
|
| 1631 |
-
{
|
| 1632 |
-
"epoch": 1.9423240033927058,
|
| 1633 |
-
"grad_norm": 0.2249774932861328,
|
| 1634 |
-
"learning_rate": 1.6662388275763484e-05,
|
| 1635 |
-
"loss": 1.5972,
|
| 1636 |
-
"step": 2290
|
| 1637 |
-
},
|
| 1638 |
-
{
|
| 1639 |
-
"epoch": 1.9508057675996606,
|
| 1640 |
-
"grad_norm": 0.22697053849697113,
|
| 1641 |
-
"learning_rate": 1.643017705604774e-05,
|
| 1642 |
-
"loss": 1.6007,
|
| 1643 |
-
"step": 2300
|
| 1644 |
-
},
|
| 1645 |
-
{
|
| 1646 |
-
"epoch": 1.9592875318066159,
|
| 1647 |
-
"grad_norm": 0.23137332499027252,
|
| 1648 |
-
"learning_rate": 1.6198800660074353e-05,
|
| 1649 |
-
"loss": 1.5868,
|
| 1650 |
-
"step": 2310
|
| 1651 |
-
},
|
| 1652 |
-
{
|
| 1653 |
-
"epoch": 1.9677692960135709,
|
| 1654 |
-
"grad_norm": 0.2348482310771942,
|
| 1655 |
-
"learning_rate": 1.5968281627224823e-05,
|
| 1656 |
-
"loss": 1.5554,
|
| 1657 |
-
"step": 2320
|
| 1658 |
-
},
|
| 1659 |
-
{
|
| 1660 |
-
"epoch": 1.9762510602205259,
|
| 1661 |
-
"grad_norm": 0.2407807558774948,
|
| 1662 |
-
"learning_rate": 1.573864241336119e-05,
|
| 1663 |
-
"loss": 1.6222,
|
| 1664 |
-
"step": 2330
|
| 1665 |
-
},
|
| 1666 |
-
{
|
| 1667 |
-
"epoch": 1.984732824427481,
|
| 1668 |
-
"grad_norm": 0.22666969895362854,
|
| 1669 |
-
"learning_rate": 1.550990538863848e-05,
|
| 1670 |
-
"loss": 1.5684,
|
| 1671 |
-
"step": 2340
|
| 1672 |
-
},
|
| 1673 |
-
{
|
| 1674 |
-
"epoch": 1.993214588634436,
|
| 1675 |
-
"grad_norm": 0.23586216568946838,
|
| 1676 |
-
"learning_rate": 1.528209283532559e-05,
|
| 1677 |
-
"loss": 1.6103,
|
| 1678 |
-
"step": 2350
|
| 1679 |
-
},
|
| 1680 |
-
{
|
| 1681 |
-
"epoch": 2.001696352841391,
|
| 1682 |
-
"grad_norm": 0.2294708639383316,
|
| 1683 |
-
"learning_rate": 1.5055226945634571e-05,
|
| 1684 |
-
"loss": 1.5866,
|
| 1685 |
-
"step": 2360
|
| 1686 |
-
},
|
| 1687 |
-
{
|
| 1688 |
-
"epoch": 2.010178117048346,
|
| 1689 |
-
"grad_norm": 0.2632901072502136,
|
| 1690 |
-
"learning_rate": 1.4829329819558905e-05,
|
| 1691 |
-
"loss": 1.524,
|
| 1692 |
-
"step": 2370
|
| 1693 |
-
},
|
| 1694 |
-
{
|
| 1695 |
-
"epoch": 2.018659881255301,
|
| 1696 |
-
"grad_norm": 0.26669132709503174,
|
| 1697 |
-
"learning_rate": 1.4604423462720574e-05,
|
| 1698 |
-
"loss": 1.5222,
|
| 1699 |
-
"step": 2380
|
| 1700 |
-
},
|
| 1701 |
-
{
|
| 1702 |
-
"epoch": 2.027141645462256,
|
| 1703 |
-
"grad_norm": 0.245851069688797,
|
| 1704 |
-
"learning_rate": 1.438052978422641e-05,
|
| 1705 |
-
"loss": 1.538,
|
| 1706 |
-
"step": 2390
|
| 1707 |
-
},
|
| 1708 |
-
{
|
| 1709 |
-
"epoch": 2.035623409669211,
|
| 1710 |
-
"grad_norm": 0.2613518238067627,
|
| 1711 |
-
"learning_rate": 1.4157670594533879e-05,
|
| 1712 |
-
"loss": 1.5463,
|
| 1713 |
-
"step": 2400
|
| 1714 |
-
},
|
| 1715 |
-
{
|
| 1716 |
-
"epoch": 2.035623409669211,
|
| 1717 |
-
"eval_loss": 1.6182470321655273,
|
| 1718 |
-
"eval_runtime": 8.3631,
|
| 1719 |
-
"eval_samples_per_second": 45.557,
|
| 1720 |
-
"eval_steps_per_second": 11.479,
|
| 1721 |
-
"step": 2400
|
| 1722 |
-
},
|
| 1723 |
-
{
|
| 1724 |
-
"epoch": 2.0441051738761664,
|
| 1725 |
-
"grad_norm": 0.25607672333717346,
|
| 1726 |
-
"learning_rate": 1.393586760332633e-05,
|
| 1727 |
-
"loss": 1.5402,
|
| 1728 |
-
"step": 2410
|
| 1729 |
-
},
|
| 1730 |
-
{
|
| 1731 |
-
"epoch": 2.052586938083121,
|
| 1732 |
-
"grad_norm": 0.24335207045078278,
|
| 1733 |
-
"learning_rate": 1.3715142417398275e-05,
|
| 1734 |
-
"loss": 1.5312,
|
| 1735 |
-
"step": 2420
|
| 1736 |
-
},
|
| 1737 |
-
{
|
| 1738 |
-
"epoch": 2.0610687022900764,
|
| 1739 |
-
"grad_norm": 0.25492334365844727,
|
| 1740 |
-
"learning_rate": 1.3495516538550495e-05,
|
| 1741 |
-
"loss": 1.5251,
|
| 1742 |
-
"step": 2430
|
| 1743 |
-
},
|
| 1744 |
-
{
|
| 1745 |
-
"epoch": 2.0695504664970312,
|
| 1746 |
-
"grad_norm": 0.26885706186294556,
|
| 1747 |
-
"learning_rate": 1.3277011361495498e-05,
|
| 1748 |
-
"loss": 1.5536,
|
| 1749 |
-
"step": 2440
|
| 1750 |
-
},
|
| 1751 |
-
{
|
| 1752 |
-
"epoch": 2.0780322307039865,
|
| 1753 |
-
"grad_norm": 0.2573489844799042,
|
| 1754 |
-
"learning_rate": 1.3059648171773361e-05,
|
| 1755 |
-
"loss": 1.5608,
|
| 1756 |
-
"step": 2450
|
| 1757 |
-
},
|
| 1758 |
-
{
|
| 1759 |
-
"epoch": 2.0865139949109412,
|
| 1760 |
-
"grad_norm": 0.26441746950149536,
|
| 1761 |
-
"learning_rate": 1.2843448143678175e-05,
|
| 1762 |
-
"loss": 1.5314,
|
| 1763 |
-
"step": 2460
|
| 1764 |
-
},
|
| 1765 |
-
{
|
| 1766 |
-
"epoch": 2.0949957591178965,
|
| 1767 |
-
"grad_norm": 0.2639044523239136,
|
| 1768 |
-
"learning_rate": 1.2628432338195423e-05,
|
| 1769 |
-
"loss": 1.5769,
|
| 1770 |
-
"step": 2470
|
| 1771 |
-
},
|
| 1772 |
-
{
|
| 1773 |
-
"epoch": 2.1034775233248517,
|
| 1774 |
-
"grad_norm": 0.26254355907440186,
|
| 1775 |
-
"learning_rate": 1.2414621700950307e-05,
|
| 1776 |
-
"loss": 1.5392,
|
| 1777 |
-
"step": 2480
|
| 1778 |
-
},
|
| 1779 |
-
{
|
| 1780 |
-
"epoch": 2.1119592875318065,
|
| 1781 |
-
"grad_norm": 0.2559792697429657,
|
| 1782 |
-
"learning_rate": 1.2202037060167338e-05,
|
| 1783 |
-
"loss": 1.5482,
|
| 1784 |
-
"step": 2490
|
| 1785 |
-
},
|
| 1786 |
-
{
|
| 1787 |
-
"epoch": 2.1204410517387617,
|
| 1788 |
-
"grad_norm": 0.2633630037307739,
|
| 1789 |
-
"learning_rate": 1.1990699124641385e-05,
|
| 1790 |
-
"loss": 1.5305,
|
| 1791 |
-
"step": 2500
|
| 1792 |
-
},
|
| 1793 |
-
{
|
| 1794 |
-
"epoch": 2.1289228159457165,
|
| 1795 |
-
"grad_norm": 0.26149672269821167,
|
| 1796 |
-
"learning_rate": 1.1780628481720333e-05,
|
| 1797 |
-
"loss": 1.5205,
|
| 1798 |
-
"step": 2510
|
| 1799 |
-
},
|
| 1800 |
-
{
|
| 1801 |
-
"epoch": 2.1374045801526718,
|
| 1802 |
-
"grad_norm": 0.2706843912601471,
|
| 1803 |
-
"learning_rate": 1.1571845595299571e-05,
|
| 1804 |
-
"loss": 1.5377,
|
| 1805 |
-
"step": 2520
|
| 1806 |
-
},
|
| 1807 |
-
{
|
| 1808 |
-
"epoch": 2.145886344359627,
|
| 1809 |
-
"grad_norm": 0.2620759606361389,
|
| 1810 |
-
"learning_rate": 1.1364370803828536e-05,
|
| 1811 |
-
"loss": 1.5274,
|
| 1812 |
-
"step": 2530
|
| 1813 |
-
},
|
| 1814 |
-
{
|
| 1815 |
-
"epoch": 2.154368108566582,
|
| 1816 |
-
"grad_norm": 0.2704004645347595,
|
| 1817 |
-
"learning_rate": 1.1158224318329439e-05,
|
| 1818 |
-
"loss": 1.5422,
|
| 1819 |
-
"step": 2540
|
| 1820 |
-
},
|
| 1821 |
-
{
|
| 1822 |
-
"epoch": 2.162849872773537,
|
| 1823 |
-
"grad_norm": 0.2624742090702057,
|
| 1824 |
-
"learning_rate": 1.0953426220428417e-05,
|
| 1825 |
-
"loss": 1.5189,
|
| 1826 |
-
"step": 2550
|
| 1827 |
-
},
|
| 1828 |
-
{
|
| 1829 |
-
"epoch": 2.171331636980492,
|
| 1830 |
-
"grad_norm": 0.2630760669708252,
|
| 1831 |
-
"learning_rate": 1.0749996460399334e-05,
|
| 1832 |
-
"loss": 1.545,
|
| 1833 |
-
"step": 2560
|
| 1834 |
-
},
|
| 1835 |
-
{
|
| 1836 |
-
"epoch": 2.179813401187447,
|
| 1837 |
-
"grad_norm": 0.2676001489162445,
|
| 1838 |
-
"learning_rate": 1.0547954855220258e-05,
|
| 1839 |
-
"loss": 1.5395,
|
| 1840 |
-
"step": 2570
|
| 1841 |
-
},
|
| 1842 |
-
{
|
| 1843 |
-
"epoch": 2.188295165394402,
|
| 1844 |
-
"grad_norm": 0.2664811611175537,
|
| 1845 |
-
"learning_rate": 1.0347321086643098e-05,
|
| 1846 |
-
"loss": 1.5427,
|
| 1847 |
-
"step": 2580
|
| 1848 |
-
},
|
| 1849 |
-
{
|
| 1850 |
-
"epoch": 2.196776929601357,
|
| 1851 |
-
"grad_norm": 0.28437891602516174,
|
| 1852 |
-
"learning_rate": 1.0148114699276254e-05,
|
| 1853 |
-
"loss": 1.5424,
|
| 1854 |
-
"step": 2590
|
| 1855 |
-
},
|
| 1856 |
-
{
|
| 1857 |
-
"epoch": 2.2052586938083123,
|
| 1858 |
-
"grad_norm": 0.27211999893188477,
|
| 1859 |
-
"learning_rate": 9.950355098680716e-06,
|
| 1860 |
-
"loss": 1.5274,
|
| 1861 |
-
"step": 2600
|
| 1862 |
-
},
|
| 1863 |
-
{
|
| 1864 |
-
"epoch": 2.213740458015267,
|
| 1865 |
-
"grad_norm": 0.2823927104473114,
|
| 1866 |
-
"learning_rate": 9.754061549479684e-06,
|
| 1867 |
-
"loss": 1.5466,
|
| 1868 |
-
"step": 2610
|
| 1869 |
-
},
|
| 1870 |
-
{
|
| 1871 |
-
"epoch": 2.2222222222222223,
|
| 1872 |
-
"grad_norm": 0.26873740553855896,
|
| 1873 |
-
"learning_rate": 9.559253173481894e-06,
|
| 1874 |
-
"loss": 1.5228,
|
| 1875 |
-
"step": 2620
|
| 1876 |
-
},
|
| 1877 |
-
{
|
| 1878 |
-
"epoch": 2.230703986429177,
|
| 1879 |
-
"grad_norm": 0.27045539021492004,
|
| 1880 |
-
"learning_rate": 9.365948947818917e-06,
|
| 1881 |
-
"loss": 1.5392,
|
| 1882 |
-
"step": 2630
|
| 1883 |
-
},
|
| 1884 |
-
{
|
| 1885 |
-
"epoch": 2.2391857506361323,
|
| 1886 |
-
"grad_norm": 0.2740824222564697,
|
| 1887 |
-
"learning_rate": 9.174167703096487e-06,
|
| 1888 |
-
"loss": 1.5312,
|
| 1889 |
-
"step": 2640
|
| 1890 |
-
},
|
| 1891 |
-
{
|
| 1892 |
-
"epoch": 2.247667514843087,
|
| 1893 |
-
"grad_norm": 0.2652507424354553,
|
| 1894 |
-
"learning_rate": 8.983928121560142e-06,
|
| 1895 |
-
"loss": 1.5506,
|
| 1896 |
-
"step": 2650
|
| 1897 |
-
},
|
| 1898 |
-
{
|
| 1899 |
-
"epoch": 2.2561492790500424,
|
| 1900 |
-
"grad_norm": 0.27051135897636414,
|
| 1901 |
-
"learning_rate": 8.795248735275295e-06,
|
| 1902 |
-
"loss": 1.5392,
|
| 1903 |
-
"step": 2660
|
| 1904 |
-
},
|
| 1905 |
-
{
|
| 1906 |
-
"epoch": 2.2646310432569976,
|
| 1907 |
-
"grad_norm": 0.26868537068367004,
|
| 1908 |
-
"learning_rate": 8.608147924321954e-06,
|
| 1909 |
-
"loss": 1.5313,
|
| 1910 |
-
"step": 2670
|
| 1911 |
-
},
|
| 1912 |
-
{
|
| 1913 |
-
"epoch": 2.2731128074639524,
|
| 1914 |
-
"grad_norm": 0.27669626474380493,
|
| 1915 |
-
"learning_rate": 8.422643915004242e-06,
|
| 1916 |
-
"loss": 1.5379,
|
| 1917 |
-
"step": 2680
|
| 1918 |
-
},
|
| 1919 |
-
{
|
| 1920 |
-
"epoch": 2.2815945716709076,
|
| 1921 |
-
"grad_norm": 0.2646745443344116,
|
| 1922 |
-
"learning_rate": 8.238754778074869e-06,
|
| 1923 |
-
"loss": 1.5187,
|
| 1924 |
-
"step": 2690
|
| 1925 |
-
},
|
| 1926 |
-
{
|
| 1927 |
-
"epoch": 2.2900763358778624,
|
| 1928 |
-
"grad_norm": 0.2804393470287323,
|
| 1929 |
-
"learning_rate": 8.056498426974817e-06,
|
| 1930 |
-
"loss": 1.5436,
|
| 1931 |
-
"step": 2700
|
| 1932 |
-
},
|
| 1933 |
-
{
|
| 1934 |
-
"epoch": 2.2985581000848176,
|
| 1935 |
-
"grad_norm": 0.2765372097492218,
|
| 1936 |
-
"learning_rate": 7.875892616088293e-06,
|
| 1937 |
-
"loss": 1.5349,
|
| 1938 |
-
"step": 2710
|
| 1939 |
-
},
|
| 1940 |
-
{
|
| 1941 |
-
"epoch": 2.307039864291773,
|
| 1942 |
-
"grad_norm": 0.26837730407714844,
|
| 1943 |
-
"learning_rate": 7.696954939013182e-06,
|
| 1944 |
-
"loss": 1.5157,
|
| 1945 |
-
"step": 2720
|
| 1946 |
-
},
|
| 1947 |
-
{
|
| 1948 |
-
"epoch": 2.3155216284987277,
|
| 1949 |
-
"grad_norm": 0.27085524797439575,
|
| 1950 |
-
"learning_rate": 7.5197028268472134e-06,
|
| 1951 |
-
"loss": 1.5115,
|
| 1952 |
-
"step": 2730
|
| 1953 |
-
},
|
| 1954 |
-
{
|
| 1955 |
-
"epoch": 2.324003392705683,
|
| 1956 |
-
"grad_norm": 0.2781482934951782,
|
| 1957 |
-
"learning_rate": 7.344153546489904e-06,
|
| 1958 |
-
"loss": 1.5391,
|
| 1959 |
-
"step": 2740
|
| 1960 |
-
},
|
| 1961 |
-
{
|
| 1962 |
-
"epoch": 2.3324851569126377,
|
| 1963 |
-
"grad_norm": 0.2616721987724304,
|
| 1964 |
-
"learning_rate": 7.17032419896051e-06,
|
| 1965 |
-
"loss": 1.5305,
|
| 1966 |
-
"step": 2750
|
| 1967 |
-
},
|
| 1968 |
-
{
|
| 1969 |
-
"epoch": 2.340966921119593,
|
| 1970 |
-
"grad_norm": 0.2856769859790802,
|
| 1971 |
-
"learning_rate": 6.998231717732148e-06,
|
| 1972 |
-
"loss": 1.5011,
|
| 1973 |
-
"step": 2760
|
| 1974 |
-
},
|
| 1975 |
-
{
|
| 1976 |
-
"epoch": 2.349448685326548,
|
| 1977 |
-
"grad_norm": 0.27664417028427124,
|
| 1978 |
-
"learning_rate": 6.827892867082236e-06,
|
| 1979 |
-
"loss": 1.503,
|
| 1980 |
-
"step": 2770
|
| 1981 |
-
},
|
| 1982 |
-
{
|
| 1983 |
-
"epoch": 2.357930449533503,
|
| 1984 |
-
"grad_norm": 0.2847400903701782,
|
| 1985 |
-
"learning_rate": 6.659324240459389e-06,
|
| 1986 |
-
"loss": 1.5295,
|
| 1987 |
-
"step": 2780
|
| 1988 |
-
},
|
| 1989 |
-
{
|
| 1990 |
-
"epoch": 2.366412213740458,
|
| 1991 |
-
"grad_norm": 0.2821629047393799,
|
| 1992 |
-
"learning_rate": 6.492542258867013e-06,
|
| 1993 |
-
"loss": 1.539,
|
| 1994 |
-
"step": 2790
|
| 1995 |
-
},
|
| 1996 |
-
{
|
| 1997 |
-
"epoch": 2.374893977947413,
|
| 1998 |
-
"grad_norm": 0.28155988454818726,
|
| 1999 |
-
"learning_rate": 6.327563169263651e-06,
|
| 2000 |
-
"loss": 1.5195,
|
| 2001 |
-
"step": 2800
|
| 2002 |
-
},
|
| 2003 |
-
{
|
| 2004 |
-
"epoch": 2.383375742154368,
|
| 2005 |
-
"grad_norm": 0.275939017534256,
|
| 2006 |
-
"learning_rate": 6.164403042980277e-06,
|
| 2007 |
-
"loss": 1.5154,
|
| 2008 |
-
"step": 2810
|
| 2009 |
-
},
|
| 2010 |
-
{
|
| 2011 |
-
"epoch": 2.391857506361323,
|
| 2012 |
-
"grad_norm": 0.28075718879699707,
|
| 2013 |
-
"learning_rate": 6.00307777415475e-06,
|
| 2014 |
-
"loss": 1.5294,
|
| 2015 |
-
"step": 2820
|
| 2016 |
-
},
|
| 2017 |
-
{
|
| 2018 |
-
"epoch": 2.4003392705682782,
|
| 2019 |
-
"grad_norm": 0.2765721082687378,
|
| 2020 |
-
"learning_rate": 5.843603078183465e-06,
|
| 2021 |
-
"loss": 1.5282,
|
| 2022 |
-
"step": 2830
|
| 2023 |
-
},
|
| 2024 |
-
{
|
| 2025 |
-
"epoch": 2.408821034775233,
|
| 2026 |
-
"grad_norm": 0.27522534132003784,
|
| 2027 |
-
"learning_rate": 5.685994490190474e-06,
|
| 2028 |
-
"loss": 1.5159,
|
| 2029 |
-
"step": 2840
|
| 2030 |
-
},
|
| 2031 |
-
{
|
| 2032 |
-
"epoch": 2.4173027989821882,
|
| 2033 |
-
"grad_norm": 0.276583731174469,
|
| 2034 |
-
"learning_rate": 5.530267363514124e-06,
|
| 2035 |
-
"loss": 1.5402,
|
| 2036 |
-
"step": 2850
|
| 2037 |
-
},
|
| 2038 |
-
{
|
| 2039 |
-
"epoch": 2.4257845631891435,
|
| 2040 |
-
"grad_norm": 0.27267441153526306,
|
| 2041 |
-
"learning_rate": 5.3764368682114315e-06,
|
| 2042 |
-
"loss": 1.5255,
|
| 2043 |
-
"step": 2860
|
| 2044 |
-
},
|
| 2045 |
-
{
|
| 2046 |
-
"epoch": 2.4342663273960983,
|
| 2047 |
-
"grad_norm": 0.27260467410087585,
|
| 2048 |
-
"learning_rate": 5.2245179895803044e-06,
|
| 2049 |
-
"loss": 1.5335,
|
| 2050 |
-
"step": 2870
|
| 2051 |
-
},
|
| 2052 |
-
{
|
| 2053 |
-
"epoch": 2.4427480916030535,
|
| 2054 |
-
"grad_norm": 0.2702862620353699,
|
| 2055 |
-
"learning_rate": 5.074525526699736e-06,
|
| 2056 |
-
"loss": 1.5305,
|
| 2057 |
-
"step": 2880
|
| 2058 |
-
},
|
| 2059 |
-
{
|
| 2060 |
-
"epoch": 2.4512298558100083,
|
| 2061 |
-
"grad_norm": 0.2800707519054413,
|
| 2062 |
-
"learning_rate": 4.926474090988193e-06,
|
| 2063 |
-
"loss": 1.5533,
|
| 2064 |
-
"step": 2890
|
| 2065 |
-
},
|
| 2066 |
-
{
|
| 2067 |
-
"epoch": 2.4597116200169635,
|
| 2068 |
-
"grad_norm": 0.2815302908420563,
|
| 2069 |
-
"learning_rate": 4.780378104780234e-06,
|
| 2070 |
-
"loss": 1.5158,
|
| 2071 |
-
"step": 2900
|
| 2072 |
-
},
|
| 2073 |
-
{
|
| 2074 |
-
"epoch": 2.4681933842239188,
|
| 2075 |
-
"grad_norm": 0.2695189416408539,
|
| 2076 |
-
"learning_rate": 4.636251799921579e-06,
|
| 2077 |
-
"loss": 1.525,
|
| 2078 |
-
"step": 2910
|
| 2079 |
-
},
|
| 2080 |
-
{
|
| 2081 |
-
"epoch": 2.4766751484308736,
|
| 2082 |
-
"grad_norm": 0.27255234122276306,
|
| 2083 |
-
"learning_rate": 4.4941092163827135e-06,
|
| 2084 |
-
"loss": 1.514,
|
| 2085 |
-
"step": 2920
|
| 2086 |
-
},
|
| 2087 |
-
{
|
| 2088 |
-
"epoch": 2.485156912637829,
|
| 2089 |
-
"grad_norm": 0.27873003482818604,
|
| 2090 |
-
"learning_rate": 4.353964200891192e-06,
|
| 2091 |
-
"loss": 1.5393,
|
| 2092 |
-
"step": 2930
|
| 2093 |
-
},
|
| 2094 |
-
{
|
| 2095 |
-
"epoch": 2.4936386768447836,
|
| 2096 |
-
"grad_norm": 0.2896173298358917,
|
| 2097 |
-
"learning_rate": 4.215830405582777e-06,
|
| 2098 |
-
"loss": 1.5401,
|
| 2099 |
-
"step": 2940
|
| 2100 |
-
},
|
| 2101 |
-
{
|
| 2102 |
-
"epoch": 2.502120441051739,
|
| 2103 |
-
"grad_norm": 0.2764180302619934,
|
| 2104 |
-
"learning_rate": 4.079721286671528e-06,
|
| 2105 |
-
"loss": 1.5417,
|
| 2106 |
-
"step": 2950
|
| 2107 |
-
},
|
| 2108 |
-
{
|
| 2109 |
-
"epoch": 2.510602205258694,
|
| 2110 |
-
"grad_norm": 0.27540919184684753,
|
| 2111 |
-
"learning_rate": 3.945650103138962e-06,
|
| 2112 |
-
"loss": 1.5291,
|
| 2113 |
-
"step": 2960
|
| 2114 |
-
},
|
| 2115 |
-
{
|
| 2116 |
-
"epoch": 2.519083969465649,
|
| 2117 |
-
"grad_norm": 0.2762376070022583,
|
| 2118 |
-
"learning_rate": 3.813629915442435e-06,
|
| 2119 |
-
"loss": 1.5305,
|
| 2120 |
-
"step": 2970
|
| 2121 |
-
},
|
| 2122 |
-
{
|
| 2123 |
-
"epoch": 2.527565733672604,
|
| 2124 |
-
"grad_norm": 0.2932542562484741,
|
| 2125 |
-
"learning_rate": 3.6836735842429027e-06,
|
| 2126 |
-
"loss": 1.5355,
|
| 2127 |
-
"step": 2980
|
| 2128 |
-
},
|
| 2129 |
-
{
|
| 2130 |
-
"epoch": 2.536047497879559,
|
| 2131 |
-
"grad_norm": 0.28812217712402344,
|
| 2132 |
-
"learning_rate": 3.5557937691520382e-06,
|
| 2133 |
-
"loss": 1.5618,
|
| 2134 |
-
"step": 2990
|
| 2135 |
-
},
|
| 2136 |
-
{
|
| 2137 |
-
"epoch": 2.544529262086514,
|
| 2138 |
-
"grad_norm": 0.28096404671669006,
|
| 2139 |
-
"learning_rate": 3.4300029274990734e-06,
|
| 2140 |
-
"loss": 1.5356,
|
| 2141 |
-
"step": 3000
|
| 2142 |
-
},
|
| 2143 |
-
{
|
| 2144 |
-
"epoch": 2.544529262086514,
|
| 2145 |
-
"eval_loss": 1.6116546392440796,
|
| 2146 |
-
"eval_runtime": 8.3692,
|
| 2147 |
-
"eval_samples_per_second": 45.524,
|
| 2148 |
-
"eval_steps_per_second": 11.471,
|
| 2149 |
-
"step": 3000
|
| 2150 |
-
},
|
| 2151 |
-
{
|
| 2152 |
-
"epoch": 2.553011026293469,
|
| 2153 |
-
"grad_norm": 0.2974391579627991,
|
| 2154 |
-
"learning_rate": 3.3063133131172403e-06,
|
| 2155 |
-
"loss": 1.5331,
|
| 2156 |
-
"step": 3010
|
| 2157 |
-
},
|
| 2158 |
-
{
|
| 2159 |
-
"epoch": 2.561492790500424,
|
| 2160 |
-
"grad_norm": 0.2741416394710541,
|
| 2161 |
-
"learning_rate": 3.1847369751500766e-06,
|
| 2162 |
-
"loss": 1.5251,
|
| 2163 |
-
"step": 3020
|
| 2164 |
-
},
|
| 2165 |
-
{
|
| 2166 |
-
"epoch": 2.569974554707379,
|
| 2167 |
-
"grad_norm": 0.27761682868003845,
|
| 2168 |
-
"learning_rate": 3.065285756877684e-06,
|
| 2169 |
-
"loss": 1.5296,
|
| 2170 |
-
"step": 3030
|
| 2171 |
-
},
|
| 2172 |
-
{
|
| 2173 |
-
"epoch": 2.578456318914334,
|
| 2174 |
-
"grad_norm": 0.2878273129463196,
|
| 2175 |
-
"learning_rate": 2.947971294562976e-06,
|
| 2176 |
-
"loss": 1.5432,
|
| 2177 |
-
"step": 3040
|
| 2178 |
-
},
|
| 2179 |
-
{
|
| 2180 |
-
"epoch": 2.5869380831212894,
|
| 2181 |
-
"grad_norm": 0.2833128869533539,
|
| 2182 |
-
"learning_rate": 2.8328050163182017e-06,
|
| 2183 |
-
"loss": 1.5209,
|
| 2184 |
-
"step": 3050
|
| 2185 |
-
},
|
| 2186 |
-
{
|
| 2187 |
-
"epoch": 2.595419847328244,
|
| 2188 |
-
"grad_norm": 0.3064183294773102,
|
| 2189 |
-
"learning_rate": 2.7197981409916416e-06,
|
| 2190 |
-
"loss": 1.5096,
|
| 2191 |
-
"step": 3060
|
| 2192 |
-
},
|
| 2193 |
-
{
|
| 2194 |
-
"epoch": 2.6039016115351994,
|
| 2195 |
-
"grad_norm": 0.2813494801521301,
|
| 2196 |
-
"learning_rate": 2.6089616770747394e-06,
|
| 2197 |
-
"loss": 1.5133,
|
| 2198 |
-
"step": 3070
|
| 2199 |
-
},
|
| 2200 |
-
{
|
| 2201 |
-
"epoch": 2.612383375742154,
|
| 2202 |
-
"grad_norm": 0.28296294808387756,
|
| 2203 |
-
"learning_rate": 2.500306421629728e-06,
|
| 2204 |
-
"loss": 1.5477,
|
| 2205 |
-
"step": 3080
|
| 2206 |
-
},
|
| 2207 |
-
{
|
| 2208 |
-
"epoch": 2.6208651399491094,
|
| 2209 |
-
"grad_norm": 0.2628287672996521,
|
| 2210 |
-
"learning_rate": 2.393842959237813e-06,
|
| 2211 |
-
"loss": 1.5082,
|
| 2212 |
-
"step": 3090
|
| 2213 |
-
},
|
| 2214 |
-
{
|
| 2215 |
-
"epoch": 2.6293469041560646,
|
| 2216 |
-
"grad_norm": 0.2780253291130066,
|
| 2217 |
-
"learning_rate": 2.289581660968121e-06,
|
| 2218 |
-
"loss": 1.5568,
|
| 2219 |
-
"step": 3100
|
| 2220 |
-
},
|
| 2221 |
-
{
|
| 2222 |
-
"epoch": 2.6378286683630194,
|
| 2223 |
-
"grad_norm": 0.27985766530036926,
|
| 2224 |
-
"learning_rate": 2.187532683367385e-06,
|
| 2225 |
-
"loss": 1.5065,
|
| 2226 |
-
"step": 3110
|
| 2227 |
-
},
|
| 2228 |
-
{
|
| 2229 |
-
"epoch": 2.6463104325699747,
|
| 2230 |
-
"grad_norm": 0.28796765208244324,
|
| 2231 |
-
"learning_rate": 2.0877059674705464e-06,
|
| 2232 |
-
"loss": 1.5217,
|
| 2233 |
-
"step": 3120
|
| 2234 |
-
},
|
| 2235 |
-
{
|
| 2236 |
-
"epoch": 2.6547921967769295,
|
| 2237 |
-
"grad_norm": 0.28680670261383057,
|
| 2238 |
-
"learning_rate": 1.9901112378323846e-06,
|
| 2239 |
-
"loss": 1.5044,
|
| 2240 |
-
"step": 3130
|
| 2241 |
-
},
|
| 2242 |
-
{
|
| 2243 |
-
"epoch": 2.6632739609838847,
|
| 2244 |
-
"grad_norm": 0.2811967730522156,
|
| 2245 |
-
"learning_rate": 1.8947580015801575e-06,
|
| 2246 |
-
"loss": 1.536,
|
| 2247 |
-
"step": 3140
|
| 2248 |
-
},
|
| 2249 |
-
{
|
| 2250 |
-
"epoch": 2.67175572519084,
|
| 2251 |
-
"grad_norm": 0.2895421087741852,
|
| 2252 |
-
"learning_rate": 1.8016555474875213e-06,
|
| 2253 |
-
"loss": 1.5325,
|
| 2254 |
-
"step": 3150
|
| 2255 |
-
},
|
| 2256 |
-
{
|
| 2257 |
-
"epoch": 2.6802374893977947,
|
| 2258 |
-
"grad_norm": 0.284627228975296,
|
| 2259 |
-
"learning_rate": 1.710812945069637e-06,
|
| 2260 |
-
"loss": 1.5302,
|
| 2261 |
-
"step": 3160
|
| 2262 |
-
},
|
| 2263 |
-
{
|
| 2264 |
-
"epoch": 2.68871925360475,
|
| 2265 |
-
"grad_norm": 0.27630600333213806,
|
| 2266 |
-
"learning_rate": 1.622239043699686e-06,
|
| 2267 |
-
"loss": 1.5353,
|
| 2268 |
-
"step": 3170
|
| 2269 |
-
},
|
| 2270 |
-
{
|
| 2271 |
-
"epoch": 2.6972010178117047,
|
| 2272 |
-
"grad_norm": 0.28358718752861023,
|
| 2273 |
-
"learning_rate": 1.535942471746807e-06,
|
| 2274 |
-
"loss": 1.5171,
|
| 2275 |
-
"step": 3180
|
| 2276 |
-
},
|
| 2277 |
-
{
|
| 2278 |
-
"epoch": 2.70568278201866,
|
| 2279 |
-
"grad_norm": 0.2867433428764343,
|
| 2280 |
-
"learning_rate": 1.4519316357355733e-06,
|
| 2281 |
-
"loss": 1.5308,
|
| 2282 |
-
"step": 3190
|
| 2283 |
-
},
|
| 2284 |
-
{
|
| 2285 |
-
"epoch": 2.714164546225615,
|
| 2286 |
-
"grad_norm": 0.29057636857032776,
|
| 2287 |
-
"learning_rate": 1.3702147195270687e-06,
|
| 2288 |
-
"loss": 1.5366,
|
| 2289 |
-
"step": 3200
|
| 2290 |
-
},
|
| 2291 |
-
{
|
| 2292 |
-
"epoch": 2.72264631043257,
|
| 2293 |
-
"grad_norm": 0.28410962224006653,
|
| 2294 |
-
"learning_rate": 1.2907996835216734e-06,
|
| 2295 |
-
"loss": 1.5148,
|
| 2296 |
-
"step": 3210
|
| 2297 |
-
},
|
| 2298 |
-
{
|
| 2299 |
-
"epoch": 2.731128074639525,
|
| 2300 |
-
"grad_norm": 0.2737291157245636,
|
| 2301 |
-
"learning_rate": 1.2136942638835995e-06,
|
| 2302 |
-
"loss": 1.5228,
|
| 2303 |
-
"step": 3220
|
| 2304 |
-
},
|
| 2305 |
-
{
|
| 2306 |
-
"epoch": 2.73960983884648,
|
| 2307 |
-
"grad_norm": 0.28176769614219666,
|
| 2308 |
-
"learning_rate": 1.1389059717872741e-06,
|
| 2309 |
-
"loss": 1.5197,
|
| 2310 |
-
"step": 3230
|
| 2311 |
-
},
|
| 2312 |
-
{
|
| 2313 |
-
"epoch": 2.7480916030534353,
|
| 2314 |
-
"grad_norm": 0.27323269844055176,
|
| 2315 |
-
"learning_rate": 1.06644209268566e-06,
|
| 2316 |
-
"loss": 1.5125,
|
| 2317 |
-
"step": 3240
|
| 2318 |
-
},
|
| 2319 |
-
{
|
| 2320 |
-
"epoch": 2.75657336726039,
|
| 2321 |
-
"grad_norm": 0.2849061191082001,
|
| 2322 |
-
"learning_rate": 9.963096856005216e-07,
|
| 2323 |
-
"loss": 1.5549,
|
| 2324 |
-
"step": 3250
|
| 2325 |
-
},
|
| 2326 |
-
{
|
| 2327 |
-
"epoch": 2.7650551314673453,
|
| 2328 |
-
"grad_norm": 0.27499040961265564,
|
| 2329 |
-
"learning_rate": 9.285155824347969e-07,
|
| 2330 |
-
"loss": 1.5183,
|
| 2331 |
-
"step": 3260
|
| 2332 |
-
},
|
| 2333 |
-
{
|
| 2334 |
-
"epoch": 2.7735368956743,
|
| 2335 |
-
"grad_norm": 0.2779834270477295,
|
| 2336 |
-
"learning_rate": 8.630663873070627e-07,
|
| 2337 |
-
"loss": 1.5323,
|
| 2338 |
-
"step": 3270
|
| 2339 |
-
},
|
| 2340 |
-
{
|
| 2341 |
-
"epoch": 2.7820186598812553,
|
| 2342 |
-
"grad_norm": 0.2841133177280426,
|
| 2343 |
-
"learning_rate": 7.999684759082083e-07,
|
| 2344 |
-
"loss": 1.5202,
|
| 2345 |
-
"step": 3280
|
| 2346 |
-
},
|
| 2347 |
-
{
|
| 2348 |
-
"epoch": 2.7905004240882105,
|
| 2349 |
-
"grad_norm": 0.2832767367362976,
|
| 2350 |
-
"learning_rate": 7.392279948803371e-07,
|
| 2351 |
-
"loss": 1.541,
|
| 2352 |
-
"step": 3290
|
| 2353 |
-
},
|
| 2354 |
-
{
|
| 2355 |
-
"epoch": 2.7989821882951653,
|
| 2356 |
-
"grad_norm": 0.29252272844314575,
|
| 2357 |
-
"learning_rate": 6.808508612179937e-07,
|
| 2358 |
-
"loss": 1.5437,
|
| 2359 |
-
"step": 3300
|
| 2360 |
-
},
|
| 2361 |
-
{
|
| 2362 |
-
"epoch": 2.8074639525021206,
|
| 2363 |
-
"grad_norm": 0.28466904163360596,
|
| 2364 |
-
"learning_rate": 6.248427616917907e-07,
|
| 2365 |
-
"loss": 1.5489,
|
| 2366 |
-
"step": 3310
|
| 2367 |
-
},
|
| 2368 |
-
{
|
| 2369 |
-
"epoch": 2.8159457167090753,
|
| 2370 |
-
"grad_norm": 0.2813858091831207,
|
| 2371 |
-
"learning_rate": 5.712091522943996e-07,
|
| 2372 |
-
"loss": 1.5159,
|
| 2373 |
-
"step": 3320
|
| 2374 |
-
},
|
| 2375 |
-
{
|
| 2376 |
-
"epoch": 2.8244274809160306,
|
| 2377 |
-
"grad_norm": 0.2804953157901764,
|
| 2378 |
-
"learning_rate": 5.199552577090927e-07,
|
| 2379 |
-
"loss": 1.5359,
|
| 2380 |
-
"step": 3330
|
| 2381 |
-
},
|
| 2382 |
-
{
|
| 2383 |
-
"epoch": 2.832909245122986,
|
| 2384 |
-
"grad_norm": 0.2917366027832031,
|
| 2385 |
-
"learning_rate": 4.710860708007553e-07,
|
| 2386 |
-
"loss": 1.5281,
|
| 2387 |
-
"step": 3340
|
| 2388 |
-
},
|
| 2389 |
-
{
|
| 2390 |
-
"epoch": 2.8413910093299406,
|
| 2391 |
-
"grad_norm": 0.2769799530506134,
|
| 2392 |
-
"learning_rate": 4.2460635212952515e-07,
|
| 2393 |
-
"loss": 1.511,
|
| 2394 |
-
"step": 3350
|
| 2395 |
-
},
|
| 2396 |
-
{
|
| 2397 |
-
"epoch": 2.849872773536896,
|
| 2398 |
-
"grad_norm": 0.29529932141304016,
|
| 2399 |
-
"learning_rate": 3.805206294870439e-07,
|
| 2400 |
-
"loss": 1.5074,
|
| 2401 |
-
"step": 3360
|
| 2402 |
-
},
|
| 2403 |
-
{
|
| 2404 |
-
"epoch": 2.8583545377438506,
|
| 2405 |
-
"grad_norm": 0.2867056131362915,
|
| 2406 |
-
"learning_rate": 3.388331974553821e-07,
|
| 2407 |
-
"loss": 1.5001,
|
| 2408 |
-
"step": 3370
|
| 2409 |
-
},
|
| 2410 |
-
{
|
| 2411 |
-
"epoch": 2.866836301950806,
|
| 2412 |
-
"grad_norm": 0.2916123569011688,
|
| 2413 |
-
"learning_rate": 2.995481169886849e-07,
|
| 2414 |
-
"loss": 1.5362,
|
| 2415 |
-
"step": 3380
|
| 2416 |
-
},
|
| 2417 |
-
{
|
| 2418 |
-
"epoch": 2.875318066157761,
|
| 2419 |
-
"grad_norm": 0.2851467430591583,
|
| 2420 |
-
"learning_rate": 2.6266921501758024e-07,
|
| 2421 |
-
"loss": 1.5205,
|
| 2422 |
-
"step": 3390
|
| 2423 |
-
},
|
| 2424 |
-
{
|
| 2425 |
-
"epoch": 2.883799830364716,
|
| 2426 |
-
"grad_norm": 0.29022783041000366,
|
| 2427 |
-
"learning_rate": 2.2820008407637983e-07,
|
| 2428 |
-
"loss": 1.5117,
|
| 2429 |
-
"step": 3400
|
| 2430 |
-
},
|
| 2431 |
-
{
|
| 2432 |
-
"epoch": 2.8922815945716707,
|
| 2433 |
-
"grad_norm": 0.2804555296897888,
|
| 2434 |
-
"learning_rate": 1.9614408195310907e-07,
|
| 2435 |
-
"loss": 1.5184,
|
| 2436 |
-
"step": 3410
|
| 2437 |
-
},
|
| 2438 |
-
{
|
| 2439 |
-
"epoch": 2.900763358778626,
|
| 2440 |
-
"grad_norm": 0.2887100279331207,
|
| 2441 |
-
"learning_rate": 1.6650433136241872e-07,
|
| 2442 |
-
"loss": 1.5141,
|
| 2443 |
-
"step": 3420
|
| 2444 |
-
},
|
| 2445 |
-
{
|
| 2446 |
-
"epoch": 2.909245122985581,
|
| 2447 |
-
"grad_norm": 0.27130427956581116,
|
| 2448 |
-
"learning_rate": 1.3928371964138653e-07,
|
| 2449 |
-
"loss": 1.5328,
|
| 2450 |
-
"step": 3430
|
| 2451 |
-
},
|
| 2452 |
-
{
|
| 2453 |
-
"epoch": 2.917726887192536,
|
| 2454 |
-
"grad_norm": 0.2768704891204834,
|
| 2455 |
-
"learning_rate": 1.1448489846824228e-07,
|
| 2456 |
-
"loss": 1.5251,
|
| 2457 |
-
"step": 3440
|
| 2458 |
-
},
|
| 2459 |
-
{
|
| 2460 |
-
"epoch": 2.926208651399491,
|
| 2461 |
-
"grad_norm": 0.28594911098480225,
|
| 2462 |
-
"learning_rate": 9.211028360406326e-08,
|
| 2463 |
-
"loss": 1.5303,
|
| 2464 |
-
"step": 3450
|
| 2465 |
-
},
|
| 2466 |
-
{
|
| 2467 |
-
"epoch": 2.934690415606446,
|
| 2468 |
-
"grad_norm": 0.2850119471549988,
|
| 2469 |
-
"learning_rate": 7.216205465743753e-08,
|
| 2470 |
-
"loss": 1.5128,
|
| 2471 |
-
"step": 3460
|
| 2472 |
-
},
|
| 2473 |
-
{
|
| 2474 |
-
"epoch": 2.943172179813401,
|
| 2475 |
-
"grad_norm": 0.28492864966392517,
|
| 2476 |
-
"learning_rate": 5.464215487214763e-08,
|
| 2477 |
-
"loss": 1.5382,
|
| 2478 |
-
"step": 3470
|
| 2479 |
-
},
|
| 2480 |
-
{
|
| 2481 |
-
"epoch": 2.9516539440203564,
|
| 2482 |
-
"grad_norm": 0.2814091444015503,
|
| 2483 |
-
"learning_rate": 3.95522909378665e-08,
|
| 2484 |
-
"loss": 1.5374,
|
| 2485 |
-
"step": 3480
|
| 2486 |
-
},
|
| 2487 |
-
{
|
| 2488 |
-
"epoch": 2.960135708227311,
|
| 2489 |
-
"grad_norm": 0.287667453289032,
|
| 2490 |
-
"learning_rate": 2.6893932823898738e-08,
|
| 2491 |
-
"loss": 1.5138,
|
| 2492 |
-
"step": 3490
|
| 2493 |
-
},
|
| 2494 |
-
{
|
| 2495 |
-
"epoch": 2.9686174724342664,
|
| 2496 |
-
"grad_norm": 0.2752053737640381,
|
| 2497 |
-
"learning_rate": 1.666831363598964e-08,
|
| 2498 |
-
"loss": 1.5293,
|
| 2499 |
-
"step": 3500
|
| 2500 |
-
},
|
| 2501 |
-
{
|
| 2502 |
-
"epoch": 2.9770992366412212,
|
| 2503 |
-
"grad_norm": 0.2834467887878418,
|
| 2504 |
-
"learning_rate": 8.876429496193494e-09,
|
| 2505 |
-
"loss": 1.545,
|
| 2506 |
-
"step": 3510
|
| 2507 |
-
},
|
| 2508 |
-
{
|
| 2509 |
-
"epoch": 2.9855810008481765,
|
| 2510 |
-
"grad_norm": 0.28982028365135193,
|
| 2511 |
-
"learning_rate": 3.519039445851191e-09,
|
| 2512 |
-
"loss": 1.5387,
|
| 2513 |
-
"step": 3520
|
| 2514 |
-
},
|
| 2515 |
-
{
|
| 2516 |
-
"epoch": 2.9940627650551317,
|
| 2517 |
-
"grad_norm": 0.2901974320411682,
|
| 2518 |
-
"learning_rate": 5.966653716410476e-10,
|
| 2519 |
-
"loss": 1.533,
|
| 2520 |
-
"step": 3530
|
| 2521 |
-
},
|
| 2522 |
{
|
| 2523 |
"epoch": 3.0,
|
| 2524 |
-
"step":
|
| 2525 |
-
"total_flos": 1.
|
| 2526 |
-
"train_loss": 1.
|
| 2527 |
-
"train_runtime":
|
| 2528 |
-
"train_samples_per_second":
|
| 2529 |
-
"train_steps_per_second": 0.
|
| 2530 |
}
|
| 2531 |
],
|
| 2532 |
"logging_steps": 10,
|
| 2533 |
-
"max_steps":
|
| 2534 |
"num_input_tokens_seen": 0,
|
| 2535 |
"num_train_epochs": 3,
|
| 2536 |
"save_steps": 500,
|
|
@@ -2546,7 +1290,7 @@
|
|
| 2546 |
"attributes": {}
|
| 2547 |
}
|
| 2548 |
},
|
| 2549 |
-
"total_flos": 1.
|
| 2550 |
"train_batch_size": 8,
|
| 2551 |
"trial_name": null,
|
| 2552 |
"trial_params": null
|
|
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 3.0,
|
| 5 |
"eval_steps": 600,
|
| 6 |
+
"global_step": 1770,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.01694915254237288,
|
| 13 |
+
"grad_norm": 0.9151965379714966,
|
| 14 |
+
"learning_rate": 2.824858757062147e-06,
|
| 15 |
+
"loss": 2.0144,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.03389830508474576,
|
| 20 |
+
"grad_norm": 0.3579045534133911,
|
| 21 |
+
"learning_rate": 5.649717514124294e-06,
|
| 22 |
+
"loss": 1.9709,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.05084745762711865,
|
| 27 |
+
"grad_norm": 0.26751402020454407,
|
| 28 |
+
"learning_rate": 8.47457627118644e-06,
|
| 29 |
+
"loss": 1.8976,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.06779661016949153,
|
| 34 |
+
"grad_norm": 0.17199155688285828,
|
| 35 |
+
"learning_rate": 1.1299435028248587e-05,
|
| 36 |
+
"loss": 1.832,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.0847457627118644,
|
| 41 |
+
"grad_norm": 0.15522244572639465,
|
| 42 |
+
"learning_rate": 1.4124293785310736e-05,
|
| 43 |
+
"loss": 1.7838,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.1016949152542373,
|
| 48 |
+
"grad_norm": 0.1401771456003189,
|
| 49 |
+
"learning_rate": 1.694915254237288e-05,
|
| 50 |
+
"loss": 1.7742,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.11864406779661017,
|
| 55 |
+
"grad_norm": 0.13913491368293762,
|
| 56 |
+
"learning_rate": 1.977401129943503e-05,
|
| 57 |
+
"loss": 1.7601,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.13559322033898305,
|
| 62 |
+
"grad_norm": 0.13437530398368835,
|
| 63 |
+
"learning_rate": 2.2598870056497175e-05,
|
| 64 |
+
"loss": 1.7553,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.15254237288135594,
|
| 69 |
+
"grad_norm": 0.14121650159358978,
|
| 70 |
+
"learning_rate": 2.5423728813559322e-05,
|
| 71 |
+
"loss": 1.7318,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.1694915254237288,
|
| 76 |
+
"grad_norm": 0.15103434026241302,
|
| 77 |
+
"learning_rate": 2.8248587570621472e-05,
|
| 78 |
+
"loss": 1.7234,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.1864406779661017,
|
| 83 |
+
"grad_norm": 0.1516140103340149,
|
| 84 |
+
"learning_rate": 3.107344632768362e-05,
|
| 85 |
+
"loss": 1.731,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.2033898305084746,
|
| 90 |
+
"grad_norm": 0.14474578201770782,
|
| 91 |
+
"learning_rate": 3.389830508474576e-05,
|
| 92 |
+
"loss": 1.7048,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.22033898305084745,
|
| 97 |
+
"grad_norm": 0.1441410928964615,
|
| 98 |
+
"learning_rate": 3.672316384180791e-05,
|
| 99 |
+
"loss": 1.7091,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.23728813559322035,
|
| 104 |
+
"grad_norm": 0.14374051988124847,
|
| 105 |
+
"learning_rate": 3.954802259887006e-05,
|
| 106 |
+
"loss": 1.7163,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.2542372881355932,
|
| 111 |
+
"grad_norm": 0.15048211812973022,
|
| 112 |
+
"learning_rate": 4.2372881355932206e-05,
|
| 113 |
+
"loss": 1.7085,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.2711864406779661,
|
| 118 |
+
"grad_norm": 0.13899974524974823,
|
| 119 |
+
"learning_rate": 4.519774011299435e-05,
|
| 120 |
+
"loss": 1.6966,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.288135593220339,
|
| 125 |
+
"grad_norm": 0.14798595011234283,
|
| 126 |
+
"learning_rate": 4.80225988700565e-05,
|
| 127 |
+
"loss": 1.7086,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.3050847457627119,
|
| 132 |
+
"grad_norm": 0.1401553452014923,
|
| 133 |
+
"learning_rate": 4.999956245830044e-05,
|
| 134 |
+
"loss": 1.7086,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.3220338983050847,
|
| 139 |
+
"grad_norm": 0.14466150104999542,
|
| 140 |
+
"learning_rate": 4.9991784365248394e-05,
|
| 141 |
+
"loss": 1.6885,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.3389830508474576,
|
| 146 |
+
"grad_norm": 0.1339222490787506,
|
| 147 |
+
"learning_rate": 4.997428660526443e-05,
|
| 148 |
+
"loss": 1.705,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.3559322033898305,
|
| 153 |
+
"grad_norm": 0.13571041822433472,
|
| 154 |
+
"learning_rate": 4.994707598348085e-05,
|
| 155 |
+
"loss": 1.6889,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.3728813559322034,
|
| 160 |
+
"grad_norm": 0.12768986821174622,
|
| 161 |
+
"learning_rate": 4.991016308250258e-05,
|
| 162 |
+
"loss": 1.6939,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.3898305084745763,
|
| 167 |
+
"grad_norm": 0.1278401017189026,
|
| 168 |
+
"learning_rate": 4.9863562258291494e-05,
|
| 169 |
+
"loss": 1.699,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.4067796610169492,
|
| 174 |
+
"grad_norm": 0.1307385265827179,
|
| 175 |
+
"learning_rate": 4.980729163458312e-05,
|
| 176 |
+
"loss": 1.6969,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.423728813559322,
|
| 181 |
+
"grad_norm": 0.12608782947063446,
|
| 182 |
+
"learning_rate": 4.974137309583804e-05,
|
| 183 |
+
"loss": 1.6931,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.4406779661016949,
|
| 188 |
+
"grad_norm": 0.9130762815475464,
|
| 189 |
+
"learning_rate": 4.966583227873079e-05,
|
| 190 |
+
"loss": 1.7061,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.4576271186440678,
|
| 195 |
+
"grad_norm": 0.12975232303142548,
|
| 196 |
+
"learning_rate": 4.9580698562179297e-05,
|
| 197 |
+
"loss": 1.6956,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.4745762711864407,
|
| 202 |
+
"grad_norm": 0.1291048377752304,
|
| 203 |
+
"learning_rate": 4.948600505591905e-05,
|
| 204 |
+
"loss": 1.7004,
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.4915254237288136,
|
| 209 |
+
"grad_norm": 0.12629887461662292,
|
| 210 |
+
"learning_rate": 4.938178858762622e-05,
|
| 211 |
+
"loss": 1.693,
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.5084745762711864,
|
| 216 |
+
"grad_norm": 0.12622225284576416,
|
| 217 |
+
"learning_rate": 4.926808968859483e-05,
|
| 218 |
+
"loss": 1.6819,
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.5254237288135594,
|
| 223 |
+
"grad_norm": 0.12781433761119843,
|
| 224 |
+
"learning_rate": 4.9144952577973596e-05,
|
| 225 |
+
"loss": 1.6984,
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.5423728813559322,
|
| 230 |
+
"grad_norm": 0.1239006370306015,
|
| 231 |
+
"learning_rate": 4.90124251455684e-05,
|
| 232 |
+
"loss": 1.6875,
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.559322033898305,
|
| 237 |
+
"grad_norm": 0.1296393871307373,
|
| 238 |
+
"learning_rate": 4.88705589332173e-05,
|
| 239 |
+
"loss": 1.6758,
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.576271186440678,
|
| 244 |
+
"grad_norm": 0.12318539619445801,
|
| 245 |
+
"learning_rate": 4.871940911474513e-05,
|
| 246 |
+
"loss": 1.695,
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 0.5932203389830508,
|
| 251 |
+
"grad_norm": 0.12802407145500183,
|
| 252 |
+
"learning_rate": 4.8559034474505614e-05,
|
| 253 |
+
"loss": 1.6862,
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.6101694915254238,
|
| 258 |
+
"grad_norm": 0.12415086477994919,
|
| 259 |
+
"learning_rate": 4.838949738451929e-05,
|
| 260 |
+
"loss": 1.6652,
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.6271186440677966,
|
| 265 |
+
"grad_norm": 0.12488778680562973,
|
| 266 |
+
"learning_rate": 4.821086378021608e-05,
|
| 267 |
+
"loss": 1.6898,
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 0.6440677966101694,
|
| 272 |
+
"grad_norm": 0.12341786175966263,
|
| 273 |
+
"learning_rate": 4.802320313479214e-05,
|
| 274 |
+
"loss": 1.6766,
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.6610169491525424,
|
| 279 |
+
"grad_norm": 0.12222524732351303,
|
| 280 |
+
"learning_rate": 4.7826588432190614e-05,
|
| 281 |
+
"loss": 1.6838,
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 0.6779661016949152,
|
| 286 |
+
"grad_norm": 0.19455482065677643,
|
| 287 |
+
"learning_rate": 4.76210961387172e-05,
|
| 288 |
+
"loss": 1.6702,
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 0.6949152542372882,
|
| 293 |
+
"grad_norm": 0.12282032519578934,
|
| 294 |
+
"learning_rate": 4.7406806173301285e-05,
|
| 295 |
+
"loss": 1.6793,
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.711864406779661,
|
| 300 |
+
"grad_norm": 0.12586097419261932,
|
| 301 |
+
"learning_rate": 4.7183801876414294e-05,
|
| 302 |
+
"loss": 1.6611,
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.7288135593220338,
|
| 307 |
+
"grad_norm": 0.12690460681915283,
|
| 308 |
+
"learning_rate": 4.695216997765737e-05,
|
| 309 |
+
"loss": 1.6896,
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 0.7457627118644068,
|
| 314 |
+
"grad_norm": 0.1262531876564026,
|
| 315 |
+
"learning_rate": 4.6712000562031e-05,
|
| 316 |
+
"loss": 1.6756,
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 0.7627118644067796,
|
| 321 |
+
"grad_norm": 0.12429999560117722,
|
| 322 |
+
"learning_rate": 4.6463387034899645e-05,
|
| 323 |
+
"loss": 1.6658,
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 0.7796610169491526,
|
| 328 |
+
"grad_norm": 0.12122652679681778,
|
| 329 |
+
"learning_rate": 4.6206426085665046e-05,
|
| 330 |
+
"loss": 1.6841,
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 0.7966101694915254,
|
| 335 |
+
"grad_norm": 0.1276836097240448,
|
| 336 |
+
"learning_rate": 4.594121765016229e-05,
|
| 337 |
+
"loss": 1.6947,
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.8135593220338984,
|
| 342 |
+
"grad_norm": 0.12912869453430176,
|
| 343 |
+
"learning_rate": 4.5667864871793345e-05,
|
| 344 |
+
"loss": 1.6888,
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.8305084745762712,
|
| 349 |
+
"grad_norm": 0.12462390959262848,
|
| 350 |
+
"learning_rate": 4.538647406141308e-05,
|
| 351 |
+
"loss": 1.6675,
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 0.847457627118644,
|
| 356 |
+
"grad_norm": 0.12297125160694122,
|
| 357 |
+
"learning_rate": 4.509715465598344e-05,
|
| 358 |
+
"loss": 1.6828,
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 0.864406779661017,
|
| 363 |
+
"grad_norm": 0.12475843727588654,
|
| 364 |
+
"learning_rate": 4.480001917601185e-05,
|
| 365 |
+
"loss": 1.6713,
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 0.8813559322033898,
|
| 370 |
+
"grad_norm": 0.1244443878531456,
|
| 371 |
+
"learning_rate": 4.449518318179029e-05,
|
| 372 |
+
"loss": 1.6544,
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 0.8983050847457628,
|
| 377 |
+
"grad_norm": 0.12166355550289154,
|
| 378 |
+
"learning_rate": 4.418276522845231e-05,
|
| 379 |
+
"loss": 1.6679,
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.9152542372881356,
|
| 384 |
+
"grad_norm": 0.1233435645699501,
|
| 385 |
+
"learning_rate": 4.386288681986516e-05,
|
| 386 |
+
"loss": 1.6704,
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.9322033898305084,
|
| 391 |
+
"grad_norm": 0.12299688160419464,
|
| 392 |
+
"learning_rate": 4.353567236137521e-05,
|
| 393 |
+
"loss": 1.6759,
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 0.9491525423728814,
|
| 398 |
+
"grad_norm": 0.12501336634159088,
|
| 399 |
+
"learning_rate": 4.3201249111424877e-05,
|
| 400 |
+
"loss": 1.6617,
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.9661016949152542,
|
| 405 |
+
"grad_norm": 0.1281379759311676,
|
| 406 |
+
"learning_rate": 4.2859747132060006e-05,
|
| 407 |
+
"loss": 1.6632,
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 0.9830508474576272,
|
| 412 |
+
"grad_norm": 0.12356545031070709,
|
| 413 |
+
"learning_rate": 4.251129923834685e-05,
|
| 414 |
+
"loss": 1.6841,
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
+
"epoch": 1.0,
|
| 419 |
+
"grad_norm": 0.12595489621162415,
|
| 420 |
+
"learning_rate": 4.215604094671835e-05,
|
| 421 |
+
"loss": 1.6448,
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 1.0169491525423728,
|
| 426 |
+
"grad_norm": 0.1292751580476761,
|
| 427 |
+
"learning_rate": 4.1794110422269825e-05,
|
| 428 |
+
"loss": 1.6261,
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 1.0169491525423728,
|
| 433 |
+
"eval_loss": 1.6485204696655273,
|
| 434 |
+
"eval_runtime": 4.4104,
|
| 435 |
+
"eval_samples_per_second": 86.387,
|
| 436 |
+
"eval_steps_per_second": 10.883,
|
| 437 |
"step": 600
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 1.0338983050847457,
|
| 441 |
+
"grad_norm": 0.13464362919330597,
|
| 442 |
+
"learning_rate": 4.142564842502461e-05,
|
| 443 |
+
"loss": 1.6501,
|
| 444 |
"step": 610
|
| 445 |
},
|
| 446 |
{
|
| 447 |
+
"epoch": 1.0508474576271187,
|
| 448 |
+
"grad_norm": 0.13430601358413696,
|
| 449 |
+
"learning_rate": 4.1050798255190405e-05,
|
| 450 |
+
"loss": 1.6435,
|
| 451 |
"step": 620
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 1.0677966101694916,
|
| 455 |
+
"grad_norm": 0.1346493363380432,
|
| 456 |
+
"learning_rate": 4.0669705697427754e-05,
|
| 457 |
+
"loss": 1.6567,
|
| 458 |
"step": 630
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 1.0847457627118644,
|
| 462 |
+
"grad_norm": 0.13532420992851257,
|
| 463 |
+
"learning_rate": 4.0282518964152296e-05,
|
| 464 |
+
"loss": 1.6527,
|
| 465 |
"step": 640
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 1.1016949152542372,
|
| 469 |
+
"grad_norm": 0.1375868022441864,
|
| 470 |
+
"learning_rate": 3.988938863789278e-05,
|
| 471 |
+
"loss": 1.6408,
|
| 472 |
"step": 650
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 1.11864406779661,
|
| 476 |
+
"grad_norm": 0.13993752002716064,
|
| 477 |
+
"learning_rate": 3.949046761272736e-05,
|
| 478 |
+
"loss": 1.6441,
|
| 479 |
"step": 660
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 1.1355932203389831,
|
| 483 |
+
"grad_norm": 0.1394631415605545,
|
| 484 |
+
"learning_rate": 3.908591103482088e-05,
|
| 485 |
+
"loss": 1.6454,
|
| 486 |
"step": 670
|
| 487 |
},
|
| 488 |
{
|
| 489 |
+
"epoch": 1.152542372881356,
|
| 490 |
+
"grad_norm": 0.1412103921175003,
|
| 491 |
+
"learning_rate": 3.867587624208627e-05,
|
| 492 |
+
"loss": 1.6689,
|
| 493 |
"step": 680
|
| 494 |
},
|
| 495 |
{
|
| 496 |
+
"epoch": 1.1694915254237288,
|
| 497 |
+
"grad_norm": 0.1373807042837143,
|
| 498 |
+
"learning_rate": 3.826052270299356e-05,
|
| 499 |
+
"loss": 1.6475,
|
| 500 |
"step": 690
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 1.1864406779661016,
|
| 504 |
+
"grad_norm": 0.14468206465244293,
|
| 505 |
+
"learning_rate": 3.784001195455027e-05,
|
| 506 |
+
"loss": 1.6469,
|
| 507 |
"step": 700
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 1.2033898305084745,
|
| 511 |
+
"grad_norm": 0.14381623268127441,
|
| 512 |
+
"learning_rate": 3.7414507539477315e-05,
|
| 513 |
+
"loss": 1.6335,
|
| 514 |
"step": 710
|
| 515 |
},
|
| 516 |
{
|
| 517 |
+
"epoch": 1.2203389830508475,
|
| 518 |
+
"grad_norm": 0.1429220587015152,
|
| 519 |
+
"learning_rate": 3.698417494260494e-05,
|
| 520 |
+
"loss": 1.6289,
|
| 521 |
"step": 720
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 1.2372881355932204,
|
| 525 |
+
"grad_norm": 0.14810022711753845,
|
| 526 |
+
"learning_rate": 3.654918152651319e-05,
|
| 527 |
+
"loss": 1.6428,
|
| 528 |
"step": 730
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 1.2542372881355932,
|
| 532 |
+
"grad_norm": 0.14096328616142273,
|
| 533 |
+
"learning_rate": 3.610969646644222e-05,
|
| 534 |
+
"loss": 1.6592,
|
| 535 |
"step": 740
|
| 536 |
},
|
| 537 |
{
|
| 538 |
+
"epoch": 1.271186440677966,
|
| 539 |
+
"grad_norm": 0.14148060977458954,
|
| 540 |
+
"learning_rate": 3.566589068449761e-05,
|
| 541 |
+
"loss": 1.6251,
|
| 542 |
"step": 750
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 1.288135593220339,
|
| 546 |
+
"grad_norm": 0.14239969849586487,
|
| 547 |
+
"learning_rate": 3.5217936783176216e-05,
|
| 548 |
+
"loss": 1.6708,
|
| 549 |
"step": 760
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 1.305084745762712,
|
| 553 |
+
"grad_norm": 0.14554665982723236,
|
| 554 |
+
"learning_rate": 3.476600897823864e-05,
|
| 555 |
+
"loss": 1.6334,
|
| 556 |
"step": 770
|
| 557 |
},
|
| 558 |
{
|
| 559 |
+
"epoch": 1.3220338983050848,
|
| 560 |
+
"grad_norm": 0.14560729265213013,
|
| 561 |
+
"learning_rate": 3.431028303095415e-05,
|
| 562 |
+
"loss": 1.6475,
|
| 563 |
"step": 780
|
| 564 |
},
|
| 565 |
{
|
| 566 |
+
"epoch": 1.3389830508474576,
|
| 567 |
+
"grad_norm": 0.14253520965576172,
|
| 568 |
+
"learning_rate": 3.385093617974458e-05,
|
| 569 |
+
"loss": 1.6418,
|
| 570 |
"step": 790
|
| 571 |
},
|
| 572 |
{
|
| 573 |
+
"epoch": 1.3559322033898304,
|
| 574 |
+
"grad_norm": 0.1438617706298828,
|
| 575 |
+
"learning_rate": 3.338814707125377e-05,
|
| 576 |
+
"loss": 1.6296,
|
| 577 |
"step": 800
|
| 578 |
},
|
| 579 |
{
|
| 580 |
+
"epoch": 1.3728813559322033,
|
| 581 |
+
"grad_norm": 0.14633050560951233,
|
| 582 |
+
"learning_rate": 3.2922095690869224e-05,
|
| 583 |
+
"loss": 1.636,
|
| 584 |
"step": 810
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 1.3898305084745763,
|
| 588 |
+
"grad_norm": 0.1484278440475464,
|
| 589 |
+
"learning_rate": 3.2452963292723305e-05,
|
| 590 |
+
"loss": 1.6156,
|
| 591 |
"step": 820
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 1.4067796610169492,
|
| 595 |
+
"grad_norm": 0.14433981478214264,
|
| 596 |
+
"learning_rate": 3.1980932329200756e-05,
|
| 597 |
+
"loss": 1.6474,
|
| 598 |
"step": 830
|
| 599 |
},
|
| 600 |
{
|
| 601 |
+
"epoch": 1.423728813559322,
|
| 602 |
+
"grad_norm": 0.1485067456960678,
|
| 603 |
+
"learning_rate": 3.150618637998041e-05,
|
| 604 |
+
"loss": 1.6356,
|
| 605 |
"step": 840
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 1.4406779661016949,
|
| 609 |
+
"grad_norm": 0.14338724315166473,
|
| 610 |
+
"learning_rate": 3.1028910080638354e-05,
|
| 611 |
+
"loss": 1.6396,
|
| 612 |
"step": 850
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"epoch": 1.457627118644068,
|
| 616 |
+
"grad_norm": 0.14848238229751587,
|
| 617 |
+
"learning_rate": 3.054928905084049e-05,
|
| 618 |
+
"loss": 1.6402,
|
| 619 |
"step": 860
|
| 620 |
},
|
| 621 |
{
|
| 622 |
+
"epoch": 1.4745762711864407,
|
| 623 |
+
"grad_norm": 0.14883729815483093,
|
| 624 |
+
"learning_rate": 3.006750982215234e-05,
|
| 625 |
+
"loss": 1.6282,
|
| 626 |
"step": 870
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 1.4915254237288136,
|
| 630 |
+
"grad_norm": 0.15096916258335114,
|
| 631 |
+
"learning_rate": 2.9583759765494267e-05,
|
| 632 |
+
"loss": 1.6402,
|
| 633 |
"step": 880
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 1.5084745762711864,
|
| 637 |
+
"grad_norm": 0.1506650447845459,
|
| 638 |
+
"learning_rate": 2.9098227018270134e-05,
|
| 639 |
+
"loss": 1.6392,
|
| 640 |
"step": 890
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 1.5254237288135593,
|
| 644 |
+
"grad_norm": 0.14937371015548706,
|
| 645 |
+
"learning_rate": 2.8611100411198037e-05,
|
| 646 |
+
"loss": 1.6472,
|
| 647 |
"step": 900
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 1.542372881355932,
|
| 651 |
+
"grad_norm": 0.14744701981544495,
|
| 652 |
+
"learning_rate": 2.8122569394871225e-05,
|
| 653 |
+
"loss": 1.6215,
|
| 654 |
"step": 910
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 1.559322033898305,
|
| 658 |
+
"grad_norm": 0.14676077663898468,
|
| 659 |
+
"learning_rate": 2.763282396607803e-05,
|
| 660 |
+
"loss": 1.6313,
|
| 661 |
"step": 920
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 1.576271186440678,
|
| 665 |
+
"grad_norm": 0.14905640482902527,
|
| 666 |
+
"learning_rate": 2.7142054593909422e-05,
|
| 667 |
+
"loss": 1.6452,
|
| 668 |
"step": 930
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 1.5932203389830508,
|
| 672 |
+
"grad_norm": 0.14924922585487366,
|
| 673 |
+
"learning_rate": 2.6650452145682762e-05,
|
| 674 |
+
"loss": 1.6386,
|
| 675 |
"step": 940
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 1.6101694915254239,
|
| 679 |
+
"grad_norm": 0.15311497449874878,
|
| 680 |
+
"learning_rate": 2.615820781271075e-05,
|
| 681 |
+
"loss": 1.6301,
|
| 682 |
"step": 950
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 1.6271186440677967,
|
| 686 |
+
"grad_norm": 0.15127715468406677,
|
| 687 |
+
"learning_rate": 2.566551303594437e-05,
|
| 688 |
+
"loss": 1.6247,
|
| 689 |
"step": 960
|
| 690 |
},
|
| 691 |
{
|
| 692 |
+
"epoch": 1.6440677966101696,
|
| 693 |
+
"grad_norm": 0.15096339583396912,
|
| 694 |
+
"learning_rate": 2.5172559431518727e-05,
|
| 695 |
+
"loss": 1.6387,
|
| 696 |
"step": 970
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 1.6610169491525424,
|
| 700 |
+
"grad_norm": 0.15415184199810028,
|
| 701 |
+
"learning_rate": 2.467953871623073e-05,
|
| 702 |
+
"loss": 1.6289,
|
| 703 |
"step": 980
|
| 704 |
},
|
| 705 |
{
|
| 706 |
+
"epoch": 1.6779661016949152,
|
| 707 |
+
"grad_norm": 0.14868630468845367,
|
| 708 |
+
"learning_rate": 2.4186642632977697e-05,
|
| 709 |
+
"loss": 1.6339,
|
| 710 |
"step": 990
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 1.694915254237288,
|
| 714 |
+
"grad_norm": 0.1516588181257248,
|
| 715 |
+
"learning_rate": 2.3694062876185738e-05,
|
| 716 |
+
"loss": 1.6172,
|
| 717 |
"step": 1000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 1.711864406779661,
|
| 721 |
+
"grad_norm": 0.15194956958293915,
|
| 722 |
+
"learning_rate": 2.3201991017257007e-05,
|
| 723 |
+
"loss": 1.6169,
|
| 724 |
"step": 1010
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"epoch": 1.7288135593220337,
|
| 728 |
+
"grad_norm": 0.15378808975219727,
|
| 729 |
+
"learning_rate": 2.2710618430064843e-05,
|
| 730 |
+
"loss": 1.6156,
|
| 731 |
"step": 1020
|
| 732 |
},
|
| 733 |
{
|
| 734 |
+
"epoch": 1.7457627118644068,
|
| 735 |
+
"grad_norm": 0.15048706531524658,
|
| 736 |
+
"learning_rate": 2.222013621652565e-05,
|
| 737 |
+
"loss": 1.6348,
|
| 738 |
"step": 1030
|
| 739 |
},
|
| 740 |
{
|
| 741 |
+
"epoch": 1.7627118644067796,
|
| 742 |
+
"grad_norm": 0.1514016091823578,
|
| 743 |
+
"learning_rate": 2.173073513227667e-05,
|
| 744 |
+
"loss": 1.6162,
|
| 745 |
"step": 1040
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 1.7796610169491527,
|
| 749 |
+
"grad_norm": 0.15484623610973358,
|
| 750 |
+
"learning_rate": 2.1242605512488248e-05,
|
| 751 |
+
"loss": 1.6162,
|
| 752 |
"step": 1050
|
| 753 |
},
|
| 754 |
{
|
| 755 |
+
"epoch": 1.7966101694915255,
|
| 756 |
+
"grad_norm": 0.1544799506664276,
|
| 757 |
+
"learning_rate": 2.0755937197839802e-05,
|
| 758 |
+
"loss": 1.6146,
|
| 759 |
"step": 1060
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 1.8135593220338984,
|
| 763 |
+
"grad_norm": 0.15184859931468964,
|
| 764 |
+
"learning_rate": 2.0270919460688055e-05,
|
| 765 |
+
"loss": 1.6178,
|
| 766 |
"step": 1070
|
| 767 |
},
|
| 768 |
{
|
| 769 |
+
"epoch": 1.8305084745762712,
|
| 770 |
+
"grad_norm": 0.154934361577034,
|
| 771 |
+
"learning_rate": 1.9787740931456165e-05,
|
| 772 |
+
"loss": 1.6287,
|
| 773 |
"step": 1080
|
| 774 |
},
|
| 775 |
{
|
| 776 |
+
"epoch": 1.847457627118644,
|
| 777 |
+
"grad_norm": 0.15421293675899506,
|
| 778 |
+
"learning_rate": 1.9306589525272756e-05,
|
| 779 |
+
"loss": 1.6348,
|
| 780 |
"step": 1090
|
| 781 |
},
|
| 782 |
{
|
| 783 |
+
"epoch": 1.8644067796610169,
|
| 784 |
+
"grad_norm": 0.15633705258369446,
|
| 785 |
+
"learning_rate": 1.8827652368888888e-05,
|
| 786 |
+
"loss": 1.6295,
|
| 787 |
"step": 1100
|
| 788 |
},
|
| 789 |
{
|
| 790 |
+
"epoch": 1.8813559322033897,
|
| 791 |
+
"grad_norm": 0.15599501132965088,
|
| 792 |
+
"learning_rate": 1.835111572790183e-05,
|
| 793 |
+
"loss": 1.629,
|
| 794 |
"step": 1110
|
| 795 |
},
|
| 796 |
{
|
| 797 |
+
"epoch": 1.8983050847457628,
|
| 798 |
+
"grad_norm": 0.1540609449148178,
|
| 799 |
+
"learning_rate": 1.7877164934313628e-05,
|
| 800 |
+
"loss": 1.6127,
|
| 801 |
"step": 1120
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 1.9152542372881356,
|
| 805 |
+
"grad_norm": 0.1557774394750595,
|
| 806 |
+
"learning_rate": 1.740598431445282e-05,
|
| 807 |
+
"loss": 1.6345,
|
| 808 |
"step": 1130
|
| 809 |
},
|
| 810 |
{
|
| 811 |
+
"epoch": 1.9322033898305084,
|
| 812 |
+
"grad_norm": 0.15810750424861908,
|
| 813 |
+
"learning_rate": 1.6937757117287278e-05,
|
| 814 |
+
"loss": 1.6291,
|
| 815 |
"step": 1140
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 1.9491525423728815,
|
| 819 |
+
"grad_norm": 0.15605640411376953,
|
| 820 |
+
"learning_rate": 1.6472665443155983e-05,
|
| 821 |
+
"loss": 1.6311,
|
| 822 |
"step": 1150
|
| 823 |
},
|
| 824 |
{
|
| 825 |
+
"epoch": 1.9661016949152543,
|
| 826 |
+
"grad_norm": 0.15914802253246307,
|
| 827 |
+
"learning_rate": 1.6010890172947606e-05,
|
| 828 |
+
"loss": 1.6082,
|
| 829 |
"step": 1160
|
| 830 |
},
|
| 831 |
{
|
| 832 |
+
"epoch": 1.9830508474576272,
|
| 833 |
+
"grad_norm": 0.15677359700202942,
|
| 834 |
+
"learning_rate": 1.5552610897753292e-05,
|
| 835 |
+
"loss": 1.6266,
|
| 836 |
"step": 1170
|
| 837 |
},
|
| 838 |
{
|
| 839 |
+
"epoch": 2.0,
|
| 840 |
+
"grad_norm": 0.15751083195209503,
|
| 841 |
+
"learning_rate": 1.509800584902108e-05,
|
| 842 |
+
"loss": 1.6265,
|
| 843 |
"step": 1180
|
| 844 |
},
|
| 845 |
{
|
| 846 |
+
"epoch": 2.016949152542373,
|
| 847 |
+
"grad_norm": 0.17009219527244568,
|
| 848 |
+
"learning_rate": 1.4647251829239139e-05,
|
| 849 |
+
"loss": 1.5776,
|
| 850 |
"step": 1190
|
| 851 |
},
|
| 852 |
{
|
| 853 |
+
"epoch": 2.0338983050847457,
|
| 854 |
+
"grad_norm": 0.16682426631450653,
|
| 855 |
+
"learning_rate": 1.4200524143174677e-05,
|
| 856 |
+
"loss": 1.5922,
|
| 857 |
"step": 1200
|
| 858 |
},
|
| 859 |
{
|
| 860 |
+
"epoch": 2.0338983050847457,
|
| 861 |
+
"eval_loss": 1.6305924654006958,
|
| 862 |
+
"eval_runtime": 4.3076,
|
| 863 |
+
"eval_samples_per_second": 88.449,
|
| 864 |
+
"eval_steps_per_second": 11.143,
|
| 865 |
"step": 1200
|
| 866 |
},
|
| 867 |
{
|
| 868 |
+
"epoch": 2.0508474576271185,
|
| 869 |
+
"grad_norm": 0.16748355329036713,
|
| 870 |
+
"learning_rate": 1.3757996529695411e-05,
|
| 871 |
+
"loss": 1.5948,
|
| 872 |
"step": 1210
|
| 873 |
},
|
| 874 |
{
|
| 875 |
+
"epoch": 2.0677966101694913,
|
| 876 |
+
"grad_norm": 0.17169494926929474,
|
| 877 |
+
"learning_rate": 1.331984109420001e-05,
|
| 878 |
+
"loss": 1.5884,
|
| 879 |
"step": 1220
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 2.084745762711864,
|
| 883 |
+
"grad_norm": 0.17144790291786194,
|
| 884 |
+
"learning_rate": 1.2886228241683749e-05,
|
| 885 |
+
"loss": 1.6067,
|
| 886 |
"step": 1230
|
| 887 |
},
|
| 888 |
{
|
| 889 |
+
"epoch": 2.1016949152542375,
|
| 890 |
+
"grad_norm": 0.16989319026470184,
|
| 891 |
+
"learning_rate": 1.2457326610465642e-05,
|
| 892 |
+
"loss": 1.6103,
|
| 893 |
"step": 1240
|
| 894 |
},
|
| 895 |
{
|
| 896 |
+
"epoch": 2.1186440677966103,
|
| 897 |
+
"grad_norm": 0.1745283454656601,
|
| 898 |
+
"learning_rate": 1.2033303006602444e-05,
|
| 899 |
+
"loss": 1.5979,
|
| 900 |
"step": 1250
|
| 901 |
},
|
| 902 |
{
|
| 903 |
+
"epoch": 2.135593220338983,
|
| 904 |
+
"grad_norm": 0.17248323559761047,
|
| 905 |
+
"learning_rate": 1.1614322339015484e-05,
|
| 906 |
+
"loss": 1.5786,
|
| 907 |
"step": 1260
|
| 908 |
},
|
| 909 |
{
|
| 910 |
+
"epoch": 2.152542372881356,
|
| 911 |
+
"grad_norm": 0.16860243678092957,
|
| 912 |
+
"learning_rate": 1.1200547555355067e-05,
|
| 913 |
+
"loss": 1.5903,
|
| 914 |
"step": 1270
|
| 915 |
},
|
| 916 |
{
|
| 917 |
+
"epoch": 2.169491525423729,
|
| 918 |
+
"grad_norm": 0.17552419006824493,
|
| 919 |
+
"learning_rate": 1.0792139578627865e-05,
|
| 920 |
+
"loss": 1.5913,
|
| 921 |
"step": 1280
|
| 922 |
},
|
| 923 |
{
|
| 924 |
+
"epoch": 2.1864406779661016,
|
| 925 |
+
"grad_norm": 0.17908188700675964,
|
| 926 |
+
"learning_rate": 1.0389257244611602e-05,
|
| 927 |
+
"loss": 1.592,
|
| 928 |
"step": 1290
|
| 929 |
},
|
| 930 |
{
|
| 931 |
+
"epoch": 2.2033898305084745,
|
| 932 |
+
"grad_norm": 0.17985224723815918,
|
| 933 |
+
"learning_rate": 9.992057240081487e-06,
|
| 934 |
+
"loss": 1.5968,
|
| 935 |
"step": 1300
|
| 936 |
},
|
| 937 |
{
|
| 938 |
+
"epoch": 2.2203389830508473,
|
| 939 |
+
"grad_norm": 0.174930602312088,
|
| 940 |
+
"learning_rate": 9.600694041872518e-06,
|
| 941 |
+
"loss": 1.5893,
|
| 942 |
"step": 1310
|
| 943 |
},
|
| 944 |
{
|
| 945 |
+
"epoch": 2.23728813559322,
|
| 946 |
+
"grad_norm": 0.17611059546470642,
|
| 947 |
+
"learning_rate": 9.215319856801158e-06,
|
| 948 |
+
"loss": 1.5861,
|
| 949 |
"step": 1320
|
| 950 |
},
|
| 951 |
{
|
| 952 |
+
"epoch": 2.2542372881355934,
|
| 953 |
+
"grad_norm": 0.1744425892829895,
|
| 954 |
+
"learning_rate": 8.836084562469921e-06,
|
| 955 |
+
"loss": 1.606,
|
| 956 |
"step": 1330
|
| 957 |
},
|
| 958 |
{
|
| 959 |
+
"epoch": 2.2711864406779663,
|
| 960 |
+
"grad_norm": 0.17969445884227753,
|
| 961 |
+
"learning_rate": 8.463135648977773e-06,
|
| 962 |
+
"loss": 1.5911,
|
| 963 |
"step": 1340
|
| 964 |
},
|
| 965 |
{
|
| 966 |
+
"epoch": 2.288135593220339,
|
| 967 |
+
"grad_norm": 0.1790507286787033,
|
| 968 |
+
"learning_rate": 8.096618161559142e-06,
|
| 969 |
+
"loss": 1.5832,
|
| 970 |
"step": 1350
|
| 971 |
},
|
| 972 |
{
|
| 973 |
+
"epoch": 2.305084745762712,
|
| 974 |
+
"grad_norm": 0.17973437905311584,
|
| 975 |
+
"learning_rate": 7.736674644173695e-06,
|
| 976 |
+
"loss": 1.5868,
|
| 977 |
"step": 1360
|
| 978 |
},
|
| 979 |
{
|
| 980 |
+
"epoch": 2.3220338983050848,
|
| 981 |
+
"grad_norm": 0.17950280010700226,
|
| 982 |
+
"learning_rate": 7.383445084068907e-06,
|
| 983 |
+
"loss": 1.5788,
|
| 984 |
"step": 1370
|
| 985 |
},
|
| 986 |
{
|
| 987 |
+
"epoch": 2.3389830508474576,
|
| 988 |
+
"grad_norm": 0.17697365581989288,
|
| 989 |
+
"learning_rate": 7.037066857337058e-06,
|
| 990 |
+
"loss": 1.5769,
|
| 991 |
"step": 1380
|
| 992 |
},
|
| 993 |
{
|
| 994 |
+
"epoch": 2.3559322033898304,
|
| 995 |
+
"grad_norm": 0.17717047035694122,
|
| 996 |
+
"learning_rate": 6.69767467548762e-06,
|
| 997 |
+
"loss": 1.5676,
|
| 998 |
"step": 1390
|
| 999 |
},
|
| 1000 |
{
|
| 1001 |
+
"epoch": 2.3728813559322033,
|
| 1002 |
+
"grad_norm": 0.17924557626247406,
|
| 1003 |
+
"learning_rate": 6.365400533056012e-06,
|
| 1004 |
+
"loss": 1.5896,
|
| 1005 |
"step": 1400
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
+
"epoch": 2.389830508474576,
|
| 1009 |
+
"grad_norm": 0.17875994741916656,
|
| 1010 |
+
"learning_rate": 6.040373656269041e-06,
|
| 1011 |
+
"loss": 1.5698,
|
| 1012 |
"step": 1410
|
| 1013 |
},
|
| 1014 |
{
|
| 1015 |
+
"epoch": 2.406779661016949,
|
| 1016 |
+
"grad_norm": 0.18066538870334625,
|
| 1017 |
+
"learning_rate": 5.722720452786928e-06,
|
| 1018 |
+
"loss": 1.5881,
|
| 1019 |
"step": 1420
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
+
"epoch": 2.423728813559322,
|
| 1023 |
+
"grad_norm": 0.18073728680610657,
|
| 1024 |
+
"learning_rate": 5.412564462541544e-06,
|
| 1025 |
+
"loss": 1.5864,
|
| 1026 |
"step": 1430
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
+
"epoch": 2.440677966101695,
|
| 1030 |
+
"grad_norm": 0.18556879460811615,
|
| 1031 |
+
"learning_rate": 5.110026309689922e-06,
|
| 1032 |
+
"loss": 1.592,
|
| 1033 |
"step": 1440
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
+
"epoch": 2.457627118644068,
|
| 1037 |
+
"grad_norm": 0.17982180416584015,
|
| 1038 |
+
"learning_rate": 4.815223655701812e-06,
|
| 1039 |
+
"loss": 1.5912,
|
| 1040 |
"step": 1450
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
+
"epoch": 2.4745762711864407,
|
| 1044 |
+
"grad_norm": 0.1827942430973053,
|
| 1045 |
+
"learning_rate": 4.5282711535994115e-06,
|
| 1046 |
+
"loss": 1.5732,
|
| 1047 |
"step": 1460
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
+
"epoch": 2.4915254237288136,
|
| 1051 |
+
"grad_norm": 0.18152937293052673,
|
| 1052 |
+
"learning_rate": 4.2492804033671145e-06,
|
| 1053 |
+
"loss": 1.5994,
|
| 1054 |
"step": 1470
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
+
"epoch": 2.5084745762711864,
|
| 1058 |
+
"grad_norm": 0.18349309265613556,
|
| 1059 |
+
"learning_rate": 3.978359908548746e-06,
|
| 1060 |
+
"loss": 1.5906,
|
| 1061 |
"step": 1480
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
+
"epoch": 2.5254237288135593,
|
| 1065 |
+
"grad_norm": 0.18495327234268188,
|
| 1066 |
+
"learning_rate": 3.71561503404885e-06,
|
| 1067 |
+
"loss": 1.593,
|
| 1068 |
"step": 1490
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
+
"epoch": 2.542372881355932,
|
| 1072 |
+
"grad_norm": 0.18758471310138702,
|
| 1073 |
+
"learning_rate": 3.4611479651548457e-06,
|
| 1074 |
+
"loss": 1.6067,
|
| 1075 |
"step": 1500
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
+
"epoch": 2.559322033898305,
|
| 1079 |
+
"grad_norm": 0.18525028228759766,
|
| 1080 |
+
"learning_rate": 3.2150576677956458e-06,
|
| 1081 |
+
"loss": 1.5926,
|
| 1082 |
"step": 1510
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
+
"epoch": 2.576271186440678,
|
| 1086 |
+
"grad_norm": 0.17536324262619019,
|
| 1087 |
+
"learning_rate": 2.977439850052366e-06,
|
| 1088 |
+
"loss": 1.5849,
|
| 1089 |
"step": 1520
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
+
"epoch": 2.593220338983051,
|
| 1093 |
+
"grad_norm": 0.18296369910240173,
|
| 1094 |
+
"learning_rate": 2.7483869249360915e-06,
|
| 1095 |
+
"loss": 1.5828,
|
| 1096 |
"step": 1530
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
+
"epoch": 2.610169491525424,
|
| 1100 |
+
"grad_norm": 0.1857452690601349,
|
| 1101 |
+
"learning_rate": 2.527987974447113e-06,
|
| 1102 |
+
"loss": 1.5802,
|
| 1103 |
"step": 1540
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
+
"epoch": 2.6271186440677967,
|
| 1107 |
+
"grad_norm": 0.18499261140823364,
|
| 1108 |
+
"learning_rate": 2.316328714929633e-06,
|
| 1109 |
+
"loss": 1.5869,
|
| 1110 |
"step": 1550
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
+
"epoch": 2.6440677966101696,
|
| 1114 |
+
"grad_norm": 0.18661075830459595,
|
| 1115 |
+
"learning_rate": 2.113491463735437e-06,
|
| 1116 |
+
"loss": 1.5813,
|
| 1117 |
"step": 1560
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
+
"epoch": 2.6610169491525424,
|
| 1121 |
+
"grad_norm": 0.18289625644683838,
|
| 1122 |
+
"learning_rate": 1.9195551072095086e-06,
|
| 1123 |
+
"loss": 1.5751,
|
| 1124 |
"step": 1570
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
+
"epoch": 2.6779661016949152,
|
| 1128 |
+
"grad_norm": 0.1815192848443985,
|
| 1129 |
+
"learning_rate": 1.7345950700099617e-06,
|
| 1130 |
+
"loss": 1.5887,
|
| 1131 |
"step": 1580
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
+
"epoch": 2.694915254237288,
|
| 1135 |
+
"grad_norm": 0.17822951078414917,
|
| 1136 |
+
"learning_rate": 1.558683285774304e-06,
|
| 1137 |
+
"loss": 1.5868,
|
| 1138 |
"step": 1590
|
| 1139 |
},
|
| 1140 |
{
|
| 1141 |
+
"epoch": 2.711864406779661,
|
| 1142 |
+
"grad_norm": 0.17891579866409302,
|
| 1143 |
+
"learning_rate": 1.3918881691434128e-06,
|
| 1144 |
+
"loss": 1.5841,
|
| 1145 |
"step": 1600
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
+
"epoch": 2.7288135593220337,
|
| 1149 |
+
"grad_norm": 0.18717658519744873,
|
| 1150 |
+
"learning_rate": 1.2342745891540857e-06,
|
| 1151 |
+
"loss": 1.5783,
|
| 1152 |
"step": 1610
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
+
"epoch": 2.7457627118644066,
|
| 1156 |
+
"grad_norm": 0.1843055635690689,
|
| 1157 |
+
"learning_rate": 1.0859038440105164e-06,
|
| 1158 |
+
"loss": 1.5761,
|
| 1159 |
"step": 1620
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
+
"epoch": 2.7627118644067794,
|
| 1163 |
+
"grad_norm": 0.18406164646148682,
|
| 1164 |
+
"learning_rate": 9.46833637244568e-07,
|
| 1165 |
+
"loss": 1.5948,
|
| 1166 |
"step": 1630
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
+
"epoch": 2.7796610169491527,
|
| 1170 |
+
"grad_norm": 0.1829424947500229,
|
| 1171 |
+
"learning_rate": 8.171180552740048e-07,
|
| 1172 |
+
"loss": 1.5791,
|
| 1173 |
"step": 1640
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
+
"epoch": 2.7966101694915255,
|
| 1177 |
+
"grad_norm": 0.18232209980487823,
|
| 1178 |
+
"learning_rate": 6.9680754636752e-07,
|
| 1179 |
+
"loss": 1.6034,
|
| 1180 |
"step": 1650
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
+
"epoch": 2.8135593220338984,
|
| 1184 |
+
"grad_norm": 0.18290942907333374,
|
| 1185 |
+
"learning_rate": 5.859489010246494e-07,
|
| 1186 |
+
"loss": 1.5902,
|
| 1187 |
"step": 1660
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
+
"epoch": 2.830508474576271,
|
| 1191 |
+
"grad_norm": 0.17814739048480988,
|
| 1192 |
+
"learning_rate": 4.845852337782953e-07,
|
| 1193 |
+
"loss": 1.5887,
|
| 1194 |
"step": 1670
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
+
"epoch": 2.847457627118644,
|
| 1198 |
+
"grad_norm": 0.1855529248714447,
|
| 1199 |
+
"learning_rate": 3.9275596642685543e-07,
|
| 1200 |
+
"loss": 1.5685,
|
| 1201 |
"step": 1680
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
+
"epoch": 2.864406779661017,
|
| 1205 |
+
"grad_norm": 0.1889604926109314,
|
| 1206 |
+
"learning_rate": 3.1049681270249044e-07,
|
| 1207 |
+
"loss": 1.5738,
|
| 1208 |
"step": 1690
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
+
"epoch": 2.8813559322033897,
|
| 1212 |
+
"grad_norm": 0.18516628444194794,
|
| 1213 |
+
"learning_rate": 2.3783976438156442e-07,
|
| 1214 |
+
"loss": 1.5753,
|
| 1215 |
"step": 1700
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
+
"epoch": 2.898305084745763,
|
| 1219 |
+
"grad_norm": 0.18402761220932007,
|
| 1220 |
+
"learning_rate": 1.7481307884256727e-07,
|
| 1221 |
+
"loss": 1.5773,
|
| 1222 |
"step": 1710
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
+
"epoch": 2.915254237288136,
|
| 1226 |
+
"grad_norm": 0.18513433635234833,
|
| 1227 |
+
"learning_rate": 1.2144126807641376e-07,
|
| 1228 |
+
"loss": 1.5865,
|
| 1229 |
"step": 1720
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
+
"epoch": 2.9322033898305087,
|
| 1233 |
+
"grad_norm": 0.1811605542898178,
|
| 1234 |
+
"learning_rate": 7.774508915337209e-08,
|
| 1235 |
+
"loss": 1.5831,
|
| 1236 |
"step": 1730
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
+
"epoch": 2.9491525423728815,
|
| 1240 |
+
"grad_norm": 0.17931722104549408,
|
| 1241 |
+
"learning_rate": 4.374153615033794e-08,
|
| 1242 |
+
"loss": 1.5919,
|
| 1243 |
"step": 1740
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
+
"epoch": 2.9661016949152543,
|
| 1247 |
+
"grad_norm": 0.18402303755283356,
|
| 1248 |
+
"learning_rate": 1.9443833541599178e-08,
|
| 1249 |
+
"loss": 1.5807,
|
| 1250 |
"step": 1750
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
+
"epoch": 2.983050847457627,
|
| 1254 |
+
"grad_norm": 0.18316687643527985,
|
| 1255 |
+
"learning_rate": 4.8614310556388544e-09,
|
| 1256 |
+
"loss": 1.5941,
|
| 1257 |
"step": 1760
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
+
"epoch": 3.0,
|
| 1261 |
+
"grad_norm": 0.1795782446861267,
|
| 1262 |
+
"learning_rate": 0.0,
|
| 1263 |
+
"loss": 1.5941,
|
| 1264 |
"step": 1770
|
| 1265 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1266 |
{
|
| 1267 |
"epoch": 3.0,
|
| 1268 |
+
"step": 1770,
|
| 1269 |
+
"total_flos": 1.0291750314285466e+19,
|
| 1270 |
+
"train_loss": 1.6438454094579664,
|
| 1271 |
+
"train_runtime": 3711.4901,
|
| 1272 |
+
"train_samples_per_second": 30.48,
|
| 1273 |
+
"train_steps_per_second": 0.477
|
| 1274 |
}
|
| 1275 |
],
|
| 1276 |
"logging_steps": 10,
|
| 1277 |
+
"max_steps": 1770,
|
| 1278 |
"num_input_tokens_seen": 0,
|
| 1279 |
"num_train_epochs": 3,
|
| 1280 |
"save_steps": 500,
|
|
|
|
| 1290 |
"attributes": {}
|
| 1291 |
}
|
| 1292 |
},
|
| 1293 |
+
"total_flos": 1.0291750314285466e+19,
|
| 1294 |
"train_batch_size": 8,
|
| 1295 |
"trial_name": null,
|
| 1296 |
"trial_params": null
|
training_eval_loss.png
CHANGED
|
|
training_loss.png
CHANGED
|
|