Zheng Li commited on
End of training
Browse files- README.md +1 -0
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/May15_01-09-48_cs-Precision-7960-Tower/events.out.tfevents.1747286441.cs-Precision-7960-Tower.142382.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +495 -890
README.md
CHANGED
|
@@ -3,6 +3,7 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- superb
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
| 6 |
+
- audio-classification
|
| 7 |
- generated_from_trainer
|
| 8 |
datasets:
|
| 9 |
- superb
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 5.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 38.
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.9811709326272433,
|
| 4 |
+
"eval_loss": 0.12430207431316376,
|
| 5 |
+
"eval_runtime": 5.5067,
|
| 6 |
+
"eval_samples_per_second": 1234.495,
|
| 7 |
+
"eval_steps_per_second": 38.68,
|
| 8 |
+
"total_flos": 4.72566865822464e+18,
|
| 9 |
+
"train_loss": 0.7669839228391647,
|
| 10 |
+
"train_runtime": 621.093,
|
| 11 |
+
"train_samples_per_second": 822.646,
|
| 12 |
+
"train_steps_per_second": 1.61
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 5.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 38.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.9811709326272433,
|
| 4 |
+
"eval_loss": 0.12430207431316376,
|
| 5 |
+
"eval_runtime": 5.5067,
|
| 6 |
+
"eval_samples_per_second": 1234.495,
|
| 7 |
+
"eval_steps_per_second": 38.68
|
| 8 |
}
|
runs/May15_01-09-48_cs-Precision-7960-Tower/events.out.tfevents.1747286441.cs-Precision-7960-Tower.142382.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b4a01e10ace34301715b9656dc9fd42cad89a638c9363f5b4a92b534d963ea4
|
| 3 |
+
size 411
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 4.72566865822464e+18,
|
| 4 |
+
"train_loss": 0.7669839228391647,
|
| 5 |
+
"train_runtime": 621.093,
|
| 6 |
+
"train_samples_per_second": 822.646,
|
| 7 |
+
"train_steps_per_second": 1.61
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1212 +1,817 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 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": 2.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 4.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 3.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 4.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 3.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 2.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 2.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss":
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch":
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss":
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch":
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch":
|
| 104 |
-
"grad_norm": 2.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch":
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch":
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch":
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch":
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"grad_norm": 2.
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch":
|
| 146 |
-
"
|
| 147 |
-
"
|
| 148 |
-
"
|
| 149 |
-
"
|
| 150 |
-
"eval_steps_per_second": 42.786,
|
| 151 |
-
"step": 199
|
| 152 |
},
|
| 153 |
{
|
| 154 |
-
"epoch":
|
| 155 |
-
"
|
| 156 |
-
"
|
| 157 |
-
"
|
|
|
|
|
|
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch":
|
| 162 |
-
"grad_norm":
|
| 163 |
-
"learning_rate":
|
| 164 |
-
"loss": 1.
|
| 165 |
"step": 210
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"grad_norm": 2.
|
| 170 |
-
"learning_rate":
|
| 171 |
-
"loss": 1.
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
-
"epoch":
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate":
|
| 178 |
-
"loss": 1.
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
-
"epoch":
|
| 183 |
-
"grad_norm":
|
| 184 |
-
"learning_rate":
|
| 185 |
-
"loss": 1.
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
-
"epoch":
|
| 190 |
-
"grad_norm":
|
| 191 |
-
"learning_rate":
|
| 192 |
-
"loss": 1.
|
| 193 |
"step": 250
|
| 194 |
},
|
| 195 |
{
|
| 196 |
-
"epoch":
|
| 197 |
-
"grad_norm":
|
| 198 |
-
"learning_rate":
|
| 199 |
-
"loss": 1.
|
| 200 |
"step": 260
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch":
|
| 204 |
-
"grad_norm":
|
| 205 |
-
"learning_rate":
|
| 206 |
-
"loss": 1.
|
| 207 |
"step": 270
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"grad_norm":
|
| 212 |
-
"learning_rate":
|
| 213 |
-
"loss": 1.
|
| 214 |
"step": 280
|
| 215 |
},
|
| 216 |
{
|
| 217 |
-
"epoch":
|
| 218 |
-
"grad_norm":
|
| 219 |
-
"learning_rate":
|
| 220 |
-
"loss":
|
| 221 |
"step": 290
|
| 222 |
},
|
| 223 |
{
|
| 224 |
-
"epoch":
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate":
|
| 227 |
-
"loss":
|
| 228 |
"step": 300
|
| 229 |
},
|
| 230 |
{
|
| 231 |
-
"epoch":
|
| 232 |
-
"
|
| 233 |
-
"
|
| 234 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 235 |
"step": 310
|
| 236 |
},
|
| 237 |
{
|
| 238 |
-
"epoch":
|
| 239 |
-
"grad_norm":
|
| 240 |
-
"learning_rate":
|
| 241 |
-
"loss":
|
| 242 |
"step": 320
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch":
|
| 246 |
-
"grad_norm":
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss":
|
| 249 |
"step": 330
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
-
"grad_norm":
|
| 254 |
-
"learning_rate":
|
| 255 |
-
"loss":
|
| 256 |
"step": 340
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"epoch":
|
| 260 |
-
"grad_norm":
|
| 261 |
-
"learning_rate":
|
| 262 |
-
"loss":
|
| 263 |
"step": 350
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch":
|
| 267 |
-
"grad_norm":
|
| 268 |
-
"learning_rate":
|
| 269 |
-
"loss":
|
| 270 |
"step": 360
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch":
|
| 274 |
-
"grad_norm":
|
| 275 |
-
"learning_rate":
|
| 276 |
-
"loss":
|
| 277 |
"step": 370
|
| 278 |
},
|
| 279 |
{
|
| 280 |
-
"epoch":
|
| 281 |
-
"grad_norm":
|
| 282 |
-
"learning_rate":
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 380
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch":
|
| 288 |
-
"grad_norm":
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss":
|
| 291 |
"step": 390
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"
|
| 296 |
-
"
|
| 297 |
-
"
|
| 298 |
-
"
|
| 299 |
-
"eval_steps_per_second": 41.924,
|
| 300 |
-
"step": 398
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"epoch":
|
| 304 |
-
"
|
| 305 |
-
"
|
| 306 |
-
"
|
|
|
|
|
|
|
| 307 |
"step": 400
|
| 308 |
},
|
| 309 |
{
|
| 310 |
-
"epoch":
|
| 311 |
-
"grad_norm":
|
| 312 |
-
"learning_rate": 1.
|
| 313 |
-
"loss": 0.
|
| 314 |
"step": 410
|
| 315 |
},
|
| 316 |
{
|
| 317 |
-
"epoch":
|
| 318 |
-
"grad_norm":
|
| 319 |
-
"learning_rate": 1.
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 420
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
-
"grad_norm":
|
| 326 |
-
"learning_rate": 1.
|
| 327 |
-
"loss": 0.
|
| 328 |
"step": 430
|
| 329 |
},
|
| 330 |
{
|
| 331 |
-
"epoch":
|
| 332 |
-
"grad_norm":
|
| 333 |
-
"learning_rate": 1.
|
| 334 |
-
"loss": 0.
|
| 335 |
"step": 440
|
| 336 |
},
|
| 337 |
{
|
| 338 |
-
"epoch":
|
| 339 |
-
"grad_norm":
|
| 340 |
-
"learning_rate": 1.
|
| 341 |
-
"loss": 0.
|
| 342 |
"step": 450
|
| 343 |
},
|
| 344 |
{
|
| 345 |
-
"epoch":
|
| 346 |
-
"grad_norm":
|
| 347 |
-
"learning_rate": 1.
|
| 348 |
-
"loss": 0.
|
| 349 |
"step": 460
|
| 350 |
},
|
| 351 |
{
|
| 352 |
-
"epoch":
|
| 353 |
-
"grad_norm":
|
| 354 |
-
"learning_rate": 1.
|
| 355 |
-
"loss": 0.
|
| 356 |
"step": 470
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch":
|
| 360 |
-
"grad_norm":
|
| 361 |
-
"learning_rate": 1.
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 480
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch":
|
| 367 |
-
"grad_norm":
|
| 368 |
-
"learning_rate": 1.
|
| 369 |
-
"loss": 0.
|
| 370 |
"step": 490
|
| 371 |
},
|
| 372 |
{
|
| 373 |
-
"epoch":
|
| 374 |
-
"grad_norm":
|
| 375 |
-
"learning_rate": 1.
|
| 376 |
-
"loss": 0.
|
| 377 |
"step": 500
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch":
|
| 381 |
-
"
|
| 382 |
-
"
|
| 383 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 384 |
"step": 510
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch":
|
| 388 |
-
"grad_norm":
|
| 389 |
-
"learning_rate": 1.
|
| 390 |
-
"loss": 0.
|
| 391 |
"step": 520
|
| 392 |
},
|
| 393 |
{
|
| 394 |
-
"epoch":
|
| 395 |
-
"grad_norm":
|
| 396 |
-
"learning_rate": 1.
|
| 397 |
-
"loss": 0.
|
| 398 |
"step": 530
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch":
|
| 402 |
-
"grad_norm": 2.
|
| 403 |
-
"learning_rate": 1.
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 540
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"grad_norm":
|
| 410 |
-
"learning_rate": 1.
|
| 411 |
-
"loss": 0.
|
| 412 |
"step": 550
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"epoch":
|
| 416 |
-
"grad_norm":
|
| 417 |
-
"learning_rate": 1.
|
| 418 |
-
"loss": 0.
|
| 419 |
"step": 560
|
| 420 |
},
|
| 421 |
{
|
| 422 |
-
"epoch":
|
| 423 |
-
"grad_norm":
|
| 424 |
-
"learning_rate": 1.
|
| 425 |
-
"loss": 0.
|
| 426 |
"step": 570
|
| 427 |
},
|
| 428 |
{
|
| 429 |
-
"epoch":
|
| 430 |
-
"grad_norm":
|
| 431 |
-
"learning_rate": 1.
|
| 432 |
-
"loss": 0.
|
| 433 |
"step": 580
|
| 434 |
},
|
| 435 |
{
|
| 436 |
-
"epoch":
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss": 0.
|
| 440 |
"step": 590
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch":
|
| 444 |
-
"
|
| 445 |
-
"
|
| 446 |
-
"
|
| 447 |
-
"
|
| 448 |
-
"eval_steps_per_second": 43.082,
|
| 449 |
-
"step": 597
|
| 450 |
},
|
| 451 |
{
|
| 452 |
-
"epoch":
|
| 453 |
-
"
|
| 454 |
-
"
|
| 455 |
-
"
|
|
|
|
|
|
|
| 456 |
"step": 600
|
| 457 |
},
|
| 458 |
{
|
| 459 |
-
"epoch":
|
| 460 |
-
"grad_norm":
|
| 461 |
-
"learning_rate": 1.
|
| 462 |
-
"loss": 0.
|
| 463 |
"step": 610
|
| 464 |
},
|
| 465 |
{
|
| 466 |
-
"epoch":
|
| 467 |
-
"grad_norm": 2.
|
| 468 |
-
"learning_rate": 1.
|
| 469 |
-
"loss": 0.
|
| 470 |
"step": 620
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch":
|
| 474 |
-
"grad_norm":
|
| 475 |
-
"learning_rate": 1.
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 630
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"grad_norm":
|
| 482 |
-
"learning_rate": 1.
|
| 483 |
-
"loss": 0.
|
| 484 |
"step": 640
|
| 485 |
},
|
| 486 |
{
|
| 487 |
-
"epoch":
|
| 488 |
-
"grad_norm":
|
| 489 |
-
"learning_rate": 1.
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 650
|
| 492 |
},
|
| 493 |
{
|
| 494 |
-
"epoch":
|
| 495 |
-
"grad_norm": 2.
|
| 496 |
-
"learning_rate": 1.
|
| 497 |
-
"loss": 0.
|
| 498 |
"step": 660
|
| 499 |
},
|
| 500 |
{
|
| 501 |
-
"epoch":
|
| 502 |
-
"grad_norm":
|
| 503 |
-
"learning_rate": 1.
|
| 504 |
-
"loss": 0.
|
| 505 |
"step": 670
|
| 506 |
},
|
| 507 |
{
|
| 508 |
-
"epoch":
|
| 509 |
-
"grad_norm": 2.
|
| 510 |
-
"learning_rate": 1.
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 680
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"grad_norm": 2.
|
| 517 |
-
"learning_rate": 1.
|
| 518 |
-
"loss": 0.
|
| 519 |
"step": 690
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"grad_norm":
|
| 524 |
-
"learning_rate":
|
| 525 |
-
"loss": 0.
|
| 526 |
"step": 700
|
| 527 |
},
|
| 528 |
{
|
| 529 |
-
"epoch":
|
| 530 |
-
"
|
| 531 |
-
"
|
| 532 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 533 |
"step": 710
|
| 534 |
},
|
| 535 |
{
|
| 536 |
-
"epoch":
|
| 537 |
-
"grad_norm":
|
| 538 |
-
"learning_rate":
|
| 539 |
-
"loss": 0.
|
| 540 |
"step": 720
|
| 541 |
},
|
| 542 |
{
|
| 543 |
-
"epoch":
|
| 544 |
-
"grad_norm":
|
| 545 |
-
"learning_rate":
|
| 546 |
-
"loss": 0.
|
| 547 |
"step": 730
|
| 548 |
},
|
| 549 |
{
|
| 550 |
-
"epoch":
|
| 551 |
-
"grad_norm":
|
| 552 |
-
"learning_rate":
|
| 553 |
-
"loss": 0.
|
| 554 |
"step": 740
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch":
|
| 558 |
-
"grad_norm":
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 0.
|
| 561 |
"step": 750
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
-
"grad_norm":
|
| 566 |
-
"learning_rate":
|
| 567 |
-
"loss": 0.
|
| 568 |
"step": 760
|
| 569 |
},
|
| 570 |
{
|
| 571 |
-
"epoch":
|
| 572 |
-
"grad_norm":
|
| 573 |
-
"learning_rate":
|
| 574 |
-
"loss": 0.
|
| 575 |
"step": 770
|
| 576 |
},
|
| 577 |
{
|
| 578 |
-
"epoch":
|
| 579 |
-
"grad_norm":
|
| 580 |
-
"learning_rate":
|
| 581 |
-
"loss": 0.
|
| 582 |
"step": 780
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"epoch":
|
| 586 |
-
"grad_norm":
|
| 587 |
-
"learning_rate":
|
| 588 |
-
"loss": 0.
|
| 589 |
"step": 790
|
| 590 |
},
|
| 591 |
{
|
| 592 |
-
"epoch":
|
| 593 |
-
"
|
| 594 |
-
"
|
| 595 |
-
"
|
| 596 |
-
"
|
| 597 |
-
"eval_steps_per_second": 43.262,
|
| 598 |
-
"step": 796
|
| 599 |
},
|
| 600 |
{
|
| 601 |
-
"epoch":
|
| 602 |
-
"
|
| 603 |
-
"
|
| 604 |
-
"
|
|
|
|
|
|
|
| 605 |
"step": 800
|
| 606 |
},
|
| 607 |
{
|
| 608 |
-
"epoch":
|
| 609 |
-
"grad_norm": 2.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 810
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch":
|
| 616 |
-
"grad_norm":
|
| 617 |
-
"learning_rate":
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 820
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch":
|
| 623 |
-
"grad_norm":
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 830
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"grad_norm":
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 840
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"grad_norm":
|
| 638 |
-
"learning_rate":
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 850
|
| 641 |
},
|
| 642 |
{
|
| 643 |
-
"epoch":
|
| 644 |
-
"grad_norm":
|
| 645 |
-
"learning_rate":
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 860
|
| 648 |
},
|
| 649 |
{
|
| 650 |
-
"epoch":
|
| 651 |
-
"grad_norm":
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 870
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch":
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 880
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 890
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"grad_norm":
|
| 673 |
-
"learning_rate":
|
| 674 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 675 |
"step": 900
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch":
|
| 679 |
-
"grad_norm":
|
| 680 |
-
"learning_rate":
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 910
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"epoch":
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate":
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 920
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"grad_norm": 2.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 930
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch":
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate":
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 940
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch":
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 950
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"grad_norm":
|
| 715 |
-
"learning_rate":
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 960
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate":
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 970
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch":
|
| 728 |
-
"grad_norm":
|
| 729 |
-
"learning_rate":
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 980
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch":
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 990
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch":
|
| 742 |
-
"
|
| 743 |
-
"
|
| 744 |
-
"
|
| 745 |
-
"eval_samples_per_second": 1334.339,
|
| 746 |
-
"eval_steps_per_second": 41.809,
|
| 747 |
-
"step": 995
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 5.025031289111389,
|
| 751 |
-
"grad_norm": 3.607999563217163,
|
| 752 |
-
"learning_rate": 8.268156424581007e-06,
|
| 753 |
-
"loss": 0.261,
|
| 754 |
"step": 1000
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"epoch":
|
| 758 |
-
"
|
| 759 |
-
"
|
| 760 |
-
"
|
| 761 |
-
"
|
| 762 |
-
|
| 763 |
-
|
| 764 |
-
"epoch": 5.1251564455569465,
|
| 765 |
-
"grad_norm": 3.3627846240997314,
|
| 766 |
-
"learning_rate": 7.988826815642458e-06,
|
| 767 |
-
"loss": 0.2363,
|
| 768 |
-
"step": 1020
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 5.175219023779725,
|
| 772 |
-
"grad_norm": 3.3233959674835205,
|
| 773 |
-
"learning_rate": 7.849162011173185e-06,
|
| 774 |
-
"loss": 0.2275,
|
| 775 |
-
"step": 1030
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 5.225281602002503,
|
| 779 |
-
"grad_norm": 3.21708083152771,
|
| 780 |
-
"learning_rate": 7.709497206703911e-06,
|
| 781 |
-
"loss": 0.2247,
|
| 782 |
-
"step": 1040
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 5.275344180225281,
|
| 786 |
-
"grad_norm": 2.4232187271118164,
|
| 787 |
-
"learning_rate": 7.569832402234637e-06,
|
| 788 |
-
"loss": 0.2044,
|
| 789 |
-
"step": 1050
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 5.32540675844806,
|
| 793 |
-
"grad_norm": 2.9417362213134766,
|
| 794 |
-
"learning_rate": 7.430167597765364e-06,
|
| 795 |
-
"loss": 0.2156,
|
| 796 |
-
"step": 1060
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 5.375469336670839,
|
| 800 |
-
"grad_norm": 5.781579971313477,
|
| 801 |
-
"learning_rate": 7.290502793296091e-06,
|
| 802 |
-
"loss": 0.2479,
|
| 803 |
-
"step": 1070
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 5.425531914893617,
|
| 807 |
-
"grad_norm": 2.417659044265747,
|
| 808 |
-
"learning_rate": 7.150837988826816e-06,
|
| 809 |
-
"loss": 0.2349,
|
| 810 |
-
"step": 1080
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 5.475594493116396,
|
| 814 |
-
"grad_norm": 3.229393482208252,
|
| 815 |
-
"learning_rate": 7.011173184357543e-06,
|
| 816 |
-
"loss": 0.218,
|
| 817 |
-
"step": 1090
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 5.5256570713391735,
|
| 821 |
-
"grad_norm": 3.5707740783691406,
|
| 822 |
-
"learning_rate": 6.871508379888268e-06,
|
| 823 |
-
"loss": 0.2287,
|
| 824 |
-
"step": 1100
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 5.575719649561952,
|
| 828 |
-
"grad_norm": 2.99178147315979,
|
| 829 |
-
"learning_rate": 6.731843575418995e-06,
|
| 830 |
-
"loss": 0.2474,
|
| 831 |
-
"step": 1110
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 5.625782227784731,
|
| 835 |
-
"grad_norm": 4.041738986968994,
|
| 836 |
-
"learning_rate": 6.592178770949721e-06,
|
| 837 |
-
"loss": 0.254,
|
| 838 |
-
"step": 1120
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 5.675844806007509,
|
| 842 |
-
"grad_norm": 3.7862696647644043,
|
| 843 |
-
"learning_rate": 6.452513966480447e-06,
|
| 844 |
-
"loss": 0.2336,
|
| 845 |
-
"step": 1130
|
| 846 |
-
},
|
| 847 |
-
{
|
| 848 |
-
"epoch": 5.725907384230288,
|
| 849 |
-
"grad_norm": 3.772305488586426,
|
| 850 |
-
"learning_rate": 6.312849162011173e-06,
|
| 851 |
-
"loss": 0.2366,
|
| 852 |
-
"step": 1140
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 5.7759699624530665,
|
| 856 |
-
"grad_norm": 3.721914291381836,
|
| 857 |
-
"learning_rate": 6.173184357541901e-06,
|
| 858 |
-
"loss": 0.2056,
|
| 859 |
-
"step": 1150
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 5.826032540675845,
|
| 863 |
-
"grad_norm": 3.321104049682617,
|
| 864 |
-
"learning_rate": 6.033519553072626e-06,
|
| 865 |
-
"loss": 0.2083,
|
| 866 |
-
"step": 1160
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 5.876095118898624,
|
| 870 |
-
"grad_norm": 2.396998405456543,
|
| 871 |
-
"learning_rate": 5.893854748603353e-06,
|
| 872 |
-
"loss": 0.2413,
|
| 873 |
-
"step": 1170
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 5.926157697121401,
|
| 877 |
-
"grad_norm": 2.6875176429748535,
|
| 878 |
-
"learning_rate": 5.754189944134078e-06,
|
| 879 |
-
"loss": 0.2049,
|
| 880 |
-
"step": 1180
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 5.97622027534418,
|
| 884 |
-
"grad_norm": 3.377181053161621,
|
| 885 |
-
"learning_rate": 5.614525139664805e-06,
|
| 886 |
-
"loss": 0.214,
|
| 887 |
-
"step": 1190
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 5.996245306633291,
|
| 891 |
-
"eval_accuracy": 0.9807296263606943,
|
| 892 |
-
"eval_loss": 0.13626398146152496,
|
| 893 |
-
"eval_runtime": 4.9535,
|
| 894 |
-
"eval_samples_per_second": 1372.375,
|
| 895 |
-
"eval_steps_per_second": 43.0,
|
| 896 |
-
"step": 1194
|
| 897 |
-
},
|
| 898 |
-
{
|
| 899 |
-
"epoch": 6.030037546933667,
|
| 900 |
-
"grad_norm": 2.5127053260803223,
|
| 901 |
-
"learning_rate": 5.474860335195531e-06,
|
| 902 |
-
"loss": 0.2374,
|
| 903 |
-
"step": 1200
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 6.080100125156446,
|
| 907 |
-
"grad_norm": 2.9415111541748047,
|
| 908 |
-
"learning_rate": 5.335195530726257e-06,
|
| 909 |
-
"loss": 0.203,
|
| 910 |
-
"step": 1210
|
| 911 |
-
},
|
| 912 |
-
{
|
| 913 |
-
"epoch": 6.130162703379224,
|
| 914 |
-
"grad_norm": 1.5115219354629517,
|
| 915 |
-
"learning_rate": 5.195530726256983e-06,
|
| 916 |
-
"loss": 0.2151,
|
| 917 |
-
"step": 1220
|
| 918 |
-
},
|
| 919 |
-
{
|
| 920 |
-
"epoch": 6.180225281602002,
|
| 921 |
-
"grad_norm": 1.8936338424682617,
|
| 922 |
-
"learning_rate": 5.055865921787711e-06,
|
| 923 |
-
"loss": 0.2159,
|
| 924 |
-
"step": 1230
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 6.230287859824781,
|
| 928 |
-
"grad_norm": 2.6654956340789795,
|
| 929 |
-
"learning_rate": 4.916201117318436e-06,
|
| 930 |
-
"loss": 0.2013,
|
| 931 |
-
"step": 1240
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 6.280350438047559,
|
| 935 |
-
"grad_norm": 3.0140798091888428,
|
| 936 |
-
"learning_rate": 4.776536312849163e-06,
|
| 937 |
-
"loss": 0.1985,
|
| 938 |
-
"step": 1250
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 6.330413016270338,
|
| 942 |
-
"grad_norm": 3.579763412475586,
|
| 943 |
-
"learning_rate": 4.636871508379888e-06,
|
| 944 |
-
"loss": 0.207,
|
| 945 |
-
"step": 1260
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 6.380475594493117,
|
| 949 |
-
"grad_norm": 2.5355074405670166,
|
| 950 |
-
"learning_rate": 4.497206703910615e-06,
|
| 951 |
-
"loss": 0.193,
|
| 952 |
-
"step": 1270
|
| 953 |
-
},
|
| 954 |
-
{
|
| 955 |
-
"epoch": 6.430538172715895,
|
| 956 |
-
"grad_norm": 2.3967621326446533,
|
| 957 |
-
"learning_rate": 4.357541899441341e-06,
|
| 958 |
-
"loss": 0.2139,
|
| 959 |
-
"step": 1280
|
| 960 |
-
},
|
| 961 |
-
{
|
| 962 |
-
"epoch": 6.480600750938673,
|
| 963 |
-
"grad_norm": 2.9278903007507324,
|
| 964 |
-
"learning_rate": 4.217877094972068e-06,
|
| 965 |
-
"loss": 0.2048,
|
| 966 |
-
"step": 1290
|
| 967 |
-
},
|
| 968 |
-
{
|
| 969 |
-
"epoch": 6.5306633291614515,
|
| 970 |
-
"grad_norm": 5.488080024719238,
|
| 971 |
-
"learning_rate": 4.078212290502794e-06,
|
| 972 |
-
"loss": 0.2098,
|
| 973 |
-
"step": 1300
|
| 974 |
-
},
|
| 975 |
-
{
|
| 976 |
-
"epoch": 6.58072590738423,
|
| 977 |
-
"grad_norm": 4.781853199005127,
|
| 978 |
-
"learning_rate": 3.93854748603352e-06,
|
| 979 |
-
"loss": 0.208,
|
| 980 |
-
"step": 1310
|
| 981 |
-
},
|
| 982 |
-
{
|
| 983 |
-
"epoch": 6.630788485607009,
|
| 984 |
-
"grad_norm": 2.9733641147613525,
|
| 985 |
-
"learning_rate": 3.798882681564246e-06,
|
| 986 |
-
"loss": 0.1923,
|
| 987 |
-
"step": 1320
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 6.680851063829787,
|
| 991 |
-
"grad_norm": 2.823633909225464,
|
| 992 |
-
"learning_rate": 3.6592178770949723e-06,
|
| 993 |
-
"loss": 0.185,
|
| 994 |
-
"step": 1330
|
| 995 |
-
},
|
| 996 |
-
{
|
| 997 |
-
"epoch": 6.730913642052566,
|
| 998 |
-
"grad_norm": 2.7353272438049316,
|
| 999 |
-
"learning_rate": 3.5195530726256988e-06,
|
| 1000 |
-
"loss": 0.1977,
|
| 1001 |
-
"step": 1340
|
| 1002 |
-
},
|
| 1003 |
-
{
|
| 1004 |
-
"epoch": 6.7809762202753445,
|
| 1005 |
-
"grad_norm": 2.489300489425659,
|
| 1006 |
-
"learning_rate": 3.3798882681564248e-06,
|
| 1007 |
-
"loss": 0.1999,
|
| 1008 |
-
"step": 1350
|
| 1009 |
-
},
|
| 1010 |
-
{
|
| 1011 |
-
"epoch": 6.831038798498122,
|
| 1012 |
-
"grad_norm": 2.8995063304901123,
|
| 1013 |
-
"learning_rate": 3.240223463687151e-06,
|
| 1014 |
-
"loss": 0.2111,
|
| 1015 |
-
"step": 1360
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 6.881101376720901,
|
| 1019 |
-
"grad_norm": 2.4604063034057617,
|
| 1020 |
-
"learning_rate": 3.1005586592178773e-06,
|
| 1021 |
-
"loss": 0.2043,
|
| 1022 |
-
"step": 1370
|
| 1023 |
-
},
|
| 1024 |
-
{
|
| 1025 |
-
"epoch": 6.931163954943679,
|
| 1026 |
-
"grad_norm": 2.1975669860839844,
|
| 1027 |
-
"learning_rate": 2.9608938547486037e-06,
|
| 1028 |
-
"loss": 0.203,
|
| 1029 |
-
"step": 1380
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 6.981226533166458,
|
| 1033 |
-
"grad_norm": 2.8632118701934814,
|
| 1034 |
-
"learning_rate": 2.8212290502793298e-06,
|
| 1035 |
-
"loss": 0.1826,
|
| 1036 |
-
"step": 1390
|
| 1037 |
-
},
|
| 1038 |
-
{
|
| 1039 |
-
"epoch": 6.996245306633291,
|
| 1040 |
-
"eval_accuracy": 0.9814651368049426,
|
| 1041 |
-
"eval_loss": 0.12438357621431351,
|
| 1042 |
-
"eval_runtime": 4.9437,
|
| 1043 |
-
"eval_samples_per_second": 1375.072,
|
| 1044 |
-
"eval_steps_per_second": 43.085,
|
| 1045 |
-
"step": 1393
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 7.035043804755945,
|
| 1049 |
-
"grad_norm": 2.0548951625823975,
|
| 1050 |
-
"learning_rate": 2.6815642458100562e-06,
|
| 1051 |
-
"loss": 0.2089,
|
| 1052 |
-
"step": 1400
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 7.085106382978723,
|
| 1056 |
-
"grad_norm": 2.192319631576538,
|
| 1057 |
-
"learning_rate": 2.5418994413407823e-06,
|
| 1058 |
-
"loss": 0.1803,
|
| 1059 |
-
"step": 1410
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 7.135168961201502,
|
| 1063 |
-
"grad_norm": 3.190675973892212,
|
| 1064 |
-
"learning_rate": 2.4022346368715087e-06,
|
| 1065 |
-
"loss": 0.1985,
|
| 1066 |
-
"step": 1420
|
| 1067 |
-
},
|
| 1068 |
-
{
|
| 1069 |
-
"epoch": 7.18523153942428,
|
| 1070 |
-
"grad_norm": 3.319995880126953,
|
| 1071 |
-
"learning_rate": 2.2625698324022348e-06,
|
| 1072 |
-
"loss": 0.1815,
|
| 1073 |
-
"step": 1430
|
| 1074 |
-
},
|
| 1075 |
-
{
|
| 1076 |
-
"epoch": 7.235294117647059,
|
| 1077 |
-
"grad_norm": 3.1613974571228027,
|
| 1078 |
-
"learning_rate": 2.1229050279329612e-06,
|
| 1079 |
-
"loss": 0.1874,
|
| 1080 |
-
"step": 1440
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 7.2853566958698375,
|
| 1084 |
-
"grad_norm": 5.030276298522949,
|
| 1085 |
-
"learning_rate": 1.9832402234636873e-06,
|
| 1086 |
-
"loss": 0.2316,
|
| 1087 |
-
"step": 1450
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 7.335419274092616,
|
| 1091 |
-
"grad_norm": 4.310720443725586,
|
| 1092 |
-
"learning_rate": 1.8435754189944135e-06,
|
| 1093 |
-
"loss": 0.187,
|
| 1094 |
-
"step": 1460
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 7.385481852315394,
|
| 1098 |
-
"grad_norm": 3.277670383453369,
|
| 1099 |
-
"learning_rate": 1.7039106145251397e-06,
|
| 1100 |
-
"loss": 0.189,
|
| 1101 |
-
"step": 1470
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 7.435544430538172,
|
| 1105 |
-
"grad_norm": 2.657017707824707,
|
| 1106 |
-
"learning_rate": 1.564245810055866e-06,
|
| 1107 |
-
"loss": 0.1963,
|
| 1108 |
-
"step": 1480
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 7.485607008760951,
|
| 1112 |
-
"grad_norm": 1.9806122779846191,
|
| 1113 |
-
"learning_rate": 1.4245810055865922e-06,
|
| 1114 |
-
"loss": 0.1736,
|
| 1115 |
-
"step": 1490
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 7.53566958698373,
|
| 1119 |
-
"grad_norm": 2.318047285079956,
|
| 1120 |
-
"learning_rate": 1.2849162011173185e-06,
|
| 1121 |
-
"loss": 0.192,
|
| 1122 |
-
"step": 1500
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 7.585732165206508,
|
| 1126 |
-
"grad_norm": 1.8546510934829712,
|
| 1127 |
-
"learning_rate": 1.1452513966480447e-06,
|
| 1128 |
-
"loss": 0.1915,
|
| 1129 |
-
"step": 1510
|
| 1130 |
-
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 7.635794743429287,
|
| 1133 |
-
"grad_norm": 2.3725483417510986,
|
| 1134 |
-
"learning_rate": 1.005586592178771e-06,
|
| 1135 |
-
"loss": 0.1924,
|
| 1136 |
-
"step": 1520
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 7.685857321652065,
|
| 1140 |
-
"grad_norm": 2.3467774391174316,
|
| 1141 |
-
"learning_rate": 8.659217877094973e-07,
|
| 1142 |
-
"loss": 0.1997,
|
| 1143 |
-
"step": 1530
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 7.735919899874844,
|
| 1147 |
-
"grad_norm": 2.653728485107422,
|
| 1148 |
-
"learning_rate": 7.262569832402236e-07,
|
| 1149 |
-
"loss": 0.2081,
|
| 1150 |
-
"step": 1540
|
| 1151 |
-
},
|
| 1152 |
-
{
|
| 1153 |
-
"epoch": 7.785982478097622,
|
| 1154 |
-
"grad_norm": 3.015143394470215,
|
| 1155 |
-
"learning_rate": 5.865921787709498e-07,
|
| 1156 |
-
"loss": 0.2194,
|
| 1157 |
-
"step": 1550
|
| 1158 |
-
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 7.8360450563204,
|
| 1161 |
-
"grad_norm": 1.836916446685791,
|
| 1162 |
-
"learning_rate": 4.46927374301676e-07,
|
| 1163 |
-
"loss": 0.2127,
|
| 1164 |
-
"step": 1560
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 7.886107634543179,
|
| 1168 |
-
"grad_norm": 1.9544241428375244,
|
| 1169 |
-
"learning_rate": 3.0726256983240227e-07,
|
| 1170 |
-
"loss": 0.1995,
|
| 1171 |
-
"step": 1570
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 7.9361702127659575,
|
| 1175 |
-
"grad_norm": 1.683449625968933,
|
| 1176 |
-
"learning_rate": 1.6759776536312851e-07,
|
| 1177 |
-
"loss": 0.1974,
|
| 1178 |
-
"step": 1580
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 7.986232790988736,
|
| 1182 |
-
"grad_norm": 2.348433494567871,
|
| 1183 |
-
"learning_rate": 2.793296089385475e-08,
|
| 1184 |
-
"loss": 0.1982,
|
| 1185 |
-
"step": 1590
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 7.996245306633291,
|
| 1189 |
-
"eval_accuracy": 0.9814651368049426,
|
| 1190 |
-
"eval_loss": 0.12064122408628464,
|
| 1191 |
-
"eval_runtime": 5.5563,
|
| 1192 |
-
"eval_samples_per_second": 1223.467,
|
| 1193 |
-
"eval_steps_per_second": 38.335,
|
| 1194 |
-
"step": 1592
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
-
"epoch":
|
| 1198 |
-
"step":
|
| 1199 |
-
"total_flos":
|
| 1200 |
-
"train_loss": 0.
|
| 1201 |
-
"train_runtime":
|
| 1202 |
-
"train_samples_per_second":
|
| 1203 |
-
"train_steps_per_second":
|
| 1204 |
}
|
| 1205 |
],
|
| 1206 |
"logging_steps": 10,
|
| 1207 |
-
"max_steps":
|
| 1208 |
"num_input_tokens_seen": 0,
|
| 1209 |
-
"num_train_epochs":
|
| 1210 |
"save_steps": 500,
|
| 1211 |
"stateful_callbacks": {
|
| 1212 |
"TrainerControl": {
|
|
@@ -1220,8 +825,8 @@
|
|
| 1220 |
"attributes": {}
|
| 1221 |
}
|
| 1222 |
},
|
| 1223 |
-
"total_flos":
|
| 1224 |
-
"train_batch_size":
|
| 1225 |
"trial_name": null,
|
| 1226 |
"trial_params": null
|
| 1227 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9811709326272433,
|
| 3 |
+
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1000",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"grad_norm": 2.6891753673553467,
|
| 14 |
+
"learning_rate": 3e-06,
|
| 15 |
+
"loss": 4.1933,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.2,
|
| 20 |
+
"grad_norm": 3.1144065856933594,
|
| 21 |
+
"learning_rate": 6e-06,
|
| 22 |
+
"loss": 4.1225,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.3,
|
| 27 |
+
"grad_norm": 4.312869071960449,
|
| 28 |
+
"learning_rate": 9e-06,
|
| 29 |
+
"loss": 3.8887,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.4,
|
| 34 |
+
"grad_norm": 5.83105993270874,
|
| 35 |
+
"learning_rate": 1.2e-05,
|
| 36 |
+
"loss": 3.4137,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.5,
|
| 41 |
+
"grad_norm": 5.6602678298950195,
|
| 42 |
+
"learning_rate": 1.5e-05,
|
| 43 |
+
"loss": 2.7921,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.6,
|
| 48 |
+
"grad_norm": 4.747939586639404,
|
| 49 |
+
"learning_rate": 1.8e-05,
|
| 50 |
+
"loss": 2.3948,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.7,
|
| 55 |
+
"grad_norm": 4.421939849853516,
|
| 56 |
+
"learning_rate": 2.1e-05,
|
| 57 |
+
"loss": 2.1514,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.8,
|
| 62 |
+
"grad_norm": 3.393974542617798,
|
| 63 |
+
"learning_rate": 2.4e-05,
|
| 64 |
+
"loss": 2.0222,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.9,
|
| 69 |
+
"grad_norm": 2.5288150310516357,
|
| 70 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 71 |
+
"loss": 1.8712,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 1.0,
|
| 76 |
+
"grad_norm": 1.0866141319274902,
|
| 77 |
+
"learning_rate": 3e-05,
|
| 78 |
+
"loss": 1.8178,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 1.0,
|
| 83 |
+
"eval_accuracy": 0.6209179170344219,
|
| 84 |
+
"eval_loss": 1.7386223077774048,
|
| 85 |
+
"eval_runtime": 5.0695,
|
| 86 |
+
"eval_samples_per_second": 1340.971,
|
| 87 |
+
"eval_steps_per_second": 42.016,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.1,
|
| 92 |
+
"grad_norm": 0.4499356150627136,
|
| 93 |
+
"learning_rate": 2.966666666666667e-05,
|
| 94 |
+
"loss": 1.7568,
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
+
"epoch": 1.2,
|
| 99 |
+
"grad_norm": 5.13858699798584,
|
| 100 |
+
"learning_rate": 2.9333333333333333e-05,
|
| 101 |
+
"loss": 1.7009,
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 1.3,
|
| 106 |
+
"grad_norm": 1.0781441926956177,
|
| 107 |
+
"learning_rate": 2.9e-05,
|
| 108 |
+
"loss": 1.7067,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 1.4,
|
| 113 |
+
"grad_norm": 2.4858882427215576,
|
| 114 |
+
"learning_rate": 2.8666666666666668e-05,
|
| 115 |
+
"loss": 1.6577,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 1.5,
|
| 120 |
+
"grad_norm": 3.1494271755218506,
|
| 121 |
+
"learning_rate": 2.8333333333333332e-05,
|
| 122 |
+
"loss": 1.5608,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 1.6,
|
| 127 |
+
"grad_norm": 2.8912174701690674,
|
| 128 |
+
"learning_rate": 2.8e-05,
|
| 129 |
+
"loss": 1.5741,
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"epoch": 1.7,
|
| 134 |
+
"grad_norm": 1.4826748371124268,
|
| 135 |
+
"learning_rate": 2.766666666666667e-05,
|
| 136 |
+
"loss": 1.5054,
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
+
"epoch": 1.8,
|
| 141 |
+
"grad_norm": 5.051352024078369,
|
| 142 |
+
"learning_rate": 2.7333333333333335e-05,
|
| 143 |
+
"loss": 1.4572,
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
+
"epoch": 1.9,
|
| 148 |
+
"grad_norm": 2.021836042404175,
|
| 149 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 150 |
+
"loss": 1.3724,
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
+
"epoch": 2.0,
|
| 155 |
+
"grad_norm": 3.871109962463379,
|
| 156 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 157 |
+
"loss": 1.3138,
|
| 158 |
+
"step": 200
|
|
|
|
|
|
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 2.0,
|
| 162 |
+
"eval_accuracy": 0.6510738452486026,
|
| 163 |
+
"eval_loss": 1.1778711080551147,
|
| 164 |
+
"eval_runtime": 5.0951,
|
| 165 |
+
"eval_samples_per_second": 1334.21,
|
| 166 |
+
"eval_steps_per_second": 41.804,
|
| 167 |
"step": 200
|
| 168 |
},
|
| 169 |
{
|
| 170 |
+
"epoch": 2.1,
|
| 171 |
+
"grad_norm": 5.006422519683838,
|
| 172 |
+
"learning_rate": 2.6333333333333334e-05,
|
| 173 |
+
"loss": 1.2749,
|
| 174 |
"step": 210
|
| 175 |
},
|
| 176 |
{
|
| 177 |
+
"epoch": 2.2,
|
| 178 |
+
"grad_norm": 2.877138614654541,
|
| 179 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 180 |
+
"loss": 1.2645,
|
| 181 |
"step": 220
|
| 182 |
},
|
| 183 |
{
|
| 184 |
+
"epoch": 2.3,
|
| 185 |
+
"grad_norm": 4.362603187561035,
|
| 186 |
+
"learning_rate": 2.5666666666666666e-05,
|
| 187 |
+
"loss": 1.1934,
|
| 188 |
"step": 230
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 2.4,
|
| 192 |
+
"grad_norm": 3.1135756969451904,
|
| 193 |
+
"learning_rate": 2.5333333333333334e-05,
|
| 194 |
+
"loss": 1.153,
|
| 195 |
"step": 240
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 2.5,
|
| 199 |
+
"grad_norm": 4.02156400680542,
|
| 200 |
+
"learning_rate": 2.5e-05,
|
| 201 |
+
"loss": 1.0804,
|
| 202 |
"step": 250
|
| 203 |
},
|
| 204 |
{
|
| 205 |
+
"epoch": 2.6,
|
| 206 |
+
"grad_norm": 3.90006160736084,
|
| 207 |
+
"learning_rate": 2.4666666666666665e-05,
|
| 208 |
+
"loss": 1.0651,
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
+
"epoch": 2.7,
|
| 213 |
+
"grad_norm": 2.3373022079467773,
|
| 214 |
+
"learning_rate": 2.4333333333333333e-05,
|
| 215 |
+
"loss": 1.0293,
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
+
"epoch": 2.8,
|
| 220 |
+
"grad_norm": 2.3652572631835938,
|
| 221 |
+
"learning_rate": 2.4e-05,
|
| 222 |
+
"loss": 1.0356,
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
+
"epoch": 2.9,
|
| 227 |
+
"grad_norm": 2.4576399326324463,
|
| 228 |
+
"learning_rate": 2.3666666666666665e-05,
|
| 229 |
+
"loss": 0.9779,
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 3.0,
|
| 234 |
+
"grad_norm": 4.176488876342773,
|
| 235 |
+
"learning_rate": 2.3333333333333336e-05,
|
| 236 |
+
"loss": 0.9632,
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 3.0,
|
| 241 |
+
"eval_accuracy": 0.8679023242130038,
|
| 242 |
+
"eval_loss": 0.8326017260551453,
|
| 243 |
+
"eval_runtime": 5.1431,
|
| 244 |
+
"eval_samples_per_second": 1321.771,
|
| 245 |
+
"eval_steps_per_second": 41.415,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.1,
|
| 250 |
+
"grad_norm": 2.0200586318969727,
|
| 251 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 252 |
+
"loss": 0.8637,
|
| 253 |
"step": 310
|
| 254 |
},
|
| 255 |
{
|
| 256 |
+
"epoch": 3.2,
|
| 257 |
+
"grad_norm": 2.459822416305542,
|
| 258 |
+
"learning_rate": 2.2666666666666668e-05,
|
| 259 |
+
"loss": 0.8304,
|
| 260 |
"step": 320
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"epoch": 3.3,
|
| 264 |
+
"grad_norm": 2.8096423149108887,
|
| 265 |
+
"learning_rate": 2.2333333333333335e-05,
|
| 266 |
+
"loss": 0.7811,
|
| 267 |
"step": 330
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 3.4,
|
| 271 |
+
"grad_norm": 2.188521385192871,
|
| 272 |
+
"learning_rate": 2.2e-05,
|
| 273 |
+
"loss": 0.7311,
|
| 274 |
"step": 340
|
| 275 |
},
|
| 276 |
{
|
| 277 |
+
"epoch": 3.5,
|
| 278 |
+
"grad_norm": 2.1101114749908447,
|
| 279 |
+
"learning_rate": 2.1666666666666667e-05,
|
| 280 |
+
"loss": 0.6899,
|
| 281 |
"step": 350
|
| 282 |
},
|
| 283 |
{
|
| 284 |
+
"epoch": 3.6,
|
| 285 |
+
"grad_norm": 3.0662026405334473,
|
| 286 |
+
"learning_rate": 2.1333333333333335e-05,
|
| 287 |
+
"loss": 0.6275,
|
| 288 |
"step": 360
|
| 289 |
},
|
| 290 |
{
|
| 291 |
+
"epoch": 3.7,
|
| 292 |
+
"grad_norm": 2.7955899238586426,
|
| 293 |
+
"learning_rate": 2.1e-05,
|
| 294 |
+
"loss": 0.5982,
|
| 295 |
"step": 370
|
| 296 |
},
|
| 297 |
{
|
| 298 |
+
"epoch": 3.8,
|
| 299 |
+
"grad_norm": 2.069448947906494,
|
| 300 |
+
"learning_rate": 2.0666666666666666e-05,
|
| 301 |
+
"loss": 0.5618,
|
| 302 |
"step": 380
|
| 303 |
},
|
| 304 |
{
|
| 305 |
+
"epoch": 3.9,
|
| 306 |
+
"grad_norm": 2.3345484733581543,
|
| 307 |
+
"learning_rate": 2.0333333333333334e-05,
|
| 308 |
+
"loss": 0.5223,
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 4.0,
|
| 313 |
+
"grad_norm": 2.2423112392425537,
|
| 314 |
+
"learning_rate": 1.9999999999999998e-05,
|
| 315 |
+
"loss": 0.499,
|
| 316 |
+
"step": 400
|
|
|
|
|
|
|
| 317 |
},
|
| 318 |
{
|
| 319 |
+
"epoch": 4.0,
|
| 320 |
+
"eval_accuracy": 0.9724919093851133,
|
| 321 |
+
"eval_loss": 0.3696895241737366,
|
| 322 |
+
"eval_runtime": 4.9717,
|
| 323 |
+
"eval_samples_per_second": 1367.332,
|
| 324 |
+
"eval_steps_per_second": 42.842,
|
| 325 |
"step": 400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 4.1,
|
| 329 |
+
"grad_norm": 2.581549644470215,
|
| 330 |
+
"learning_rate": 1.9666666666666666e-05,
|
| 331 |
+
"loss": 0.4643,
|
| 332 |
"step": 410
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 4.2,
|
| 336 |
+
"grad_norm": 2.2802562713623047,
|
| 337 |
+
"learning_rate": 1.9333333333333333e-05,
|
| 338 |
+
"loss": 0.4679,
|
| 339 |
"step": 420
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 4.3,
|
| 343 |
+
"grad_norm": 2.000401496887207,
|
| 344 |
+
"learning_rate": 1.9e-05,
|
| 345 |
+
"loss": 0.4143,
|
| 346 |
"step": 430
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 4.4,
|
| 350 |
+
"grad_norm": 1.913470983505249,
|
| 351 |
+
"learning_rate": 1.866666666666667e-05,
|
| 352 |
+
"loss": 0.3969,
|
| 353 |
"step": 440
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 4.5,
|
| 357 |
+
"grad_norm": 1.812354564666748,
|
| 358 |
+
"learning_rate": 1.8333333333333336e-05,
|
| 359 |
+
"loss": 0.3903,
|
| 360 |
"step": 450
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 4.6,
|
| 364 |
+
"grad_norm": 1.731414556503296,
|
| 365 |
+
"learning_rate": 1.8e-05,
|
| 366 |
+
"loss": 0.3536,
|
| 367 |
"step": 460
|
| 368 |
},
|
| 369 |
{
|
| 370 |
+
"epoch": 4.7,
|
| 371 |
+
"grad_norm": 2.4097819328308105,
|
| 372 |
+
"learning_rate": 1.7666666666666668e-05,
|
| 373 |
+
"loss": 0.3537,
|
| 374 |
"step": 470
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 4.8,
|
| 378 |
+
"grad_norm": 2.2438652515411377,
|
| 379 |
+
"learning_rate": 1.7333333333333332e-05,
|
| 380 |
+
"loss": 0.3417,
|
| 381 |
"step": 480
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 4.9,
|
| 385 |
+
"grad_norm": 1.8313252925872803,
|
| 386 |
+
"learning_rate": 1.7e-05,
|
| 387 |
+
"loss": 0.3269,
|
| 388 |
"step": 490
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 5.0,
|
| 392 |
+
"grad_norm": 3.7112584114074707,
|
| 393 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 394 |
+
"loss": 0.3191,
|
| 395 |
"step": 500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 5.0,
|
| 399 |
+
"eval_accuracy": 0.9760223595175052,
|
| 400 |
+
"eval_loss": 0.22396039962768555,
|
| 401 |
+
"eval_runtime": 5.0015,
|
| 402 |
+
"eval_samples_per_second": 1359.188,
|
| 403 |
+
"eval_steps_per_second": 42.587,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.1,
|
| 408 |
+
"grad_norm": 2.2409660816192627,
|
| 409 |
+
"learning_rate": 1.633333333333333e-05,
|
| 410 |
+
"loss": 0.3044,
|
| 411 |
"step": 510
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 5.2,
|
| 415 |
+
"grad_norm": 1.6726042032241821,
|
| 416 |
+
"learning_rate": 1.6e-05,
|
| 417 |
+
"loss": 0.2886,
|
| 418 |
"step": 520
|
| 419 |
},
|
| 420 |
{
|
| 421 |
+
"epoch": 5.3,
|
| 422 |
+
"grad_norm": 2.055961847305298,
|
| 423 |
+
"learning_rate": 1.5666666666666667e-05,
|
| 424 |
+
"loss": 0.2745,
|
| 425 |
"step": 530
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 5.4,
|
| 429 |
+
"grad_norm": 2.2671661376953125,
|
| 430 |
+
"learning_rate": 1.533333333333333e-05,
|
| 431 |
+
"loss": 0.2943,
|
| 432 |
"step": 540
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"epoch": 5.5,
|
| 436 |
+
"grad_norm": 2.010511636734009,
|
| 437 |
+
"learning_rate": 1.5e-05,
|
| 438 |
+
"loss": 0.246,
|
| 439 |
"step": 550
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 5.6,
|
| 443 |
+
"grad_norm": 2.027716636657715,
|
| 444 |
+
"learning_rate": 1.4666666666666666e-05,
|
| 445 |
+
"loss": 0.2962,
|
| 446 |
"step": 560
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 5.7,
|
| 450 |
+
"grad_norm": 1.839068055152893,
|
| 451 |
+
"learning_rate": 1.4333333333333334e-05,
|
| 452 |
+
"loss": 0.2817,
|
| 453 |
"step": 570
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 5.8,
|
| 457 |
+
"grad_norm": 2.8274471759796143,
|
| 458 |
+
"learning_rate": 1.4e-05,
|
| 459 |
+
"loss": 0.2512,
|
| 460 |
"step": 580
|
| 461 |
},
|
| 462 |
{
|
| 463 |
+
"epoch": 5.9,
|
| 464 |
+
"grad_norm": 2.0226399898529053,
|
| 465 |
+
"learning_rate": 1.3666666666666667e-05,
|
| 466 |
+
"loss": 0.2756,
|
| 467 |
"step": 590
|
| 468 |
},
|
| 469 |
{
|
| 470 |
+
"epoch": 6.0,
|
| 471 |
+
"grad_norm": 2.212151050567627,
|
| 472 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 473 |
+
"loss": 0.242,
|
| 474 |
+
"step": 600
|
|
|
|
|
|
|
| 475 |
},
|
| 476 |
{
|
| 477 |
+
"epoch": 6.0,
|
| 478 |
+
"eval_accuracy": 0.9792586054721977,
|
| 479 |
+
"eval_loss": 0.1708967536687851,
|
| 480 |
+
"eval_runtime": 4.9642,
|
| 481 |
+
"eval_samples_per_second": 1369.415,
|
| 482 |
+
"eval_steps_per_second": 42.908,
|
| 483 |
"step": 600
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 6.1,
|
| 487 |
+
"grad_norm": 2.019993305206299,
|
| 488 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 489 |
+
"loss": 0.2557,
|
| 490 |
"step": 610
|
| 491 |
},
|
| 492 |
{
|
| 493 |
+
"epoch": 6.2,
|
| 494 |
+
"grad_norm": 2.0448718070983887,
|
| 495 |
+
"learning_rate": 1.2666666666666667e-05,
|
| 496 |
+
"loss": 0.2455,
|
| 497 |
"step": 620
|
| 498 |
},
|
| 499 |
{
|
| 500 |
+
"epoch": 6.3,
|
| 501 |
+
"grad_norm": 2.4769129753112793,
|
| 502 |
+
"learning_rate": 1.2333333333333333e-05,
|
| 503 |
+
"loss": 0.2435,
|
| 504 |
"step": 630
|
| 505 |
},
|
| 506 |
{
|
| 507 |
+
"epoch": 6.4,
|
| 508 |
+
"grad_norm": 2.217947483062744,
|
| 509 |
+
"learning_rate": 1.2e-05,
|
| 510 |
+
"loss": 0.2392,
|
| 511 |
"step": 640
|
| 512 |
},
|
| 513 |
{
|
| 514 |
+
"epoch": 6.5,
|
| 515 |
+
"grad_norm": 2.198836326599121,
|
| 516 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 517 |
+
"loss": 0.2194,
|
| 518 |
"step": 650
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 6.6,
|
| 522 |
+
"grad_norm": 2.2629637718200684,
|
| 523 |
+
"learning_rate": 1.1333333333333334e-05,
|
| 524 |
+
"loss": 0.2165,
|
| 525 |
"step": 660
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 6.7,
|
| 529 |
+
"grad_norm": 1.9289292097091675,
|
| 530 |
+
"learning_rate": 1.1e-05,
|
| 531 |
+
"loss": 0.2264,
|
| 532 |
"step": 670
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"epoch": 6.8,
|
| 536 |
+
"grad_norm": 2.585972547531128,
|
| 537 |
+
"learning_rate": 1.0666666666666667e-05,
|
| 538 |
+
"loss": 0.2254,
|
| 539 |
"step": 680
|
| 540 |
},
|
| 541 |
{
|
| 542 |
+
"epoch": 6.9,
|
| 543 |
+
"grad_norm": 2.3876583576202393,
|
| 544 |
+
"learning_rate": 1.0333333333333333e-05,
|
| 545 |
+
"loss": 0.2264,
|
| 546 |
"step": 690
|
| 547 |
},
|
| 548 |
{
|
| 549 |
+
"epoch": 7.0,
|
| 550 |
+
"grad_norm": 2.297879695892334,
|
| 551 |
+
"learning_rate": 9.999999999999999e-06,
|
| 552 |
+
"loss": 0.2144,
|
| 553 |
"step": 700
|
| 554 |
},
|
| 555 |
{
|
| 556 |
+
"epoch": 7.0,
|
| 557 |
+
"eval_accuracy": 0.9805825242718447,
|
| 558 |
+
"eval_loss": 0.14598147571086884,
|
| 559 |
+
"eval_runtime": 4.9644,
|
| 560 |
+
"eval_samples_per_second": 1369.359,
|
| 561 |
+
"eval_steps_per_second": 42.906,
|
| 562 |
+
"step": 700
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.1,
|
| 566 |
+
"grad_norm": 2.2874860763549805,
|
| 567 |
+
"learning_rate": 9.666666666666667e-06,
|
| 568 |
+
"loss": 0.2217,
|
| 569 |
"step": 710
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 7.2,
|
| 573 |
+
"grad_norm": 1.6444002389907837,
|
| 574 |
+
"learning_rate": 9.333333333333334e-06,
|
| 575 |
+
"loss": 0.2048,
|
| 576 |
"step": 720
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 7.3,
|
| 580 |
+
"grad_norm": 2.211688280105591,
|
| 581 |
+
"learning_rate": 9e-06,
|
| 582 |
+
"loss": 0.2122,
|
| 583 |
"step": 730
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 7.4,
|
| 587 |
+
"grad_norm": 2.334533929824829,
|
| 588 |
+
"learning_rate": 8.666666666666666e-06,
|
| 589 |
+
"loss": 0.2142,
|
| 590 |
"step": 740
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 7.5,
|
| 594 |
+
"grad_norm": 1.83983314037323,
|
| 595 |
+
"learning_rate": 8.333333333333334e-06,
|
| 596 |
+
"loss": 0.1873,
|
| 597 |
"step": 750
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 7.6,
|
| 601 |
+
"grad_norm": 1.8291276693344116,
|
| 602 |
+
"learning_rate": 8e-06,
|
| 603 |
+
"loss": 0.2003,
|
| 604 |
"step": 760
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 7.7,
|
| 608 |
+
"grad_norm": 2.4695987701416016,
|
| 609 |
+
"learning_rate": 7.666666666666666e-06,
|
| 610 |
+
"loss": 0.2075,
|
| 611 |
"step": 770
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 7.8,
|
| 615 |
+
"grad_norm": 1.7656598091125488,
|
| 616 |
+
"learning_rate": 7.333333333333333e-06,
|
| 617 |
+
"loss": 0.2089,
|
| 618 |
"step": 780
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 7.9,
|
| 622 |
+
"grad_norm": 1.8952357769012451,
|
| 623 |
+
"learning_rate": 7e-06,
|
| 624 |
+
"loss": 0.2013,
|
| 625 |
"step": 790
|
| 626 |
},
|
| 627 |
{
|
| 628 |
+
"epoch": 8.0,
|
| 629 |
+
"grad_norm": 1.6189404726028442,
|
| 630 |
+
"learning_rate": 6.666666666666667e-06,
|
| 631 |
+
"loss": 0.1977,
|
| 632 |
+
"step": 800
|
|
|
|
|
|
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 8.0,
|
| 636 |
+
"eval_accuracy": 0.979994115916446,
|
| 637 |
+
"eval_loss": 0.13441701233386993,
|
| 638 |
+
"eval_runtime": 5.0435,
|
| 639 |
+
"eval_samples_per_second": 1347.881,
|
| 640 |
+
"eval_steps_per_second": 42.233,
|
| 641 |
"step": 800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
+
"epoch": 8.1,
|
| 645 |
+
"grad_norm": 2.2733678817749023,
|
| 646 |
+
"learning_rate": 6.333333333333333e-06,
|
| 647 |
+
"loss": 0.1908,
|
| 648 |
"step": 810
|
| 649 |
},
|
| 650 |
{
|
| 651 |
+
"epoch": 8.2,
|
| 652 |
+
"grad_norm": 1.6874712705612183,
|
| 653 |
+
"learning_rate": 6e-06,
|
| 654 |
+
"loss": 0.1819,
|
| 655 |
"step": 820
|
| 656 |
},
|
| 657 |
{
|
| 658 |
+
"epoch": 8.3,
|
| 659 |
+
"grad_norm": 1.4649100303649902,
|
| 660 |
+
"learning_rate": 5.666666666666667e-06,
|
| 661 |
+
"loss": 0.2037,
|
| 662 |
"step": 830
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"epoch": 8.4,
|
| 666 |
+
"grad_norm": 1.7231727838516235,
|
| 667 |
+
"learning_rate": 5.333333333333334e-06,
|
| 668 |
+
"loss": 0.1935,
|
| 669 |
"step": 840
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 8.5,
|
| 673 |
+
"grad_norm": 1.6908072233200073,
|
| 674 |
+
"learning_rate": 4.9999999999999996e-06,
|
| 675 |
+
"loss": 0.1957,
|
| 676 |
"step": 850
|
| 677 |
},
|
| 678 |
{
|
| 679 |
+
"epoch": 8.6,
|
| 680 |
+
"grad_norm": 2.1523678302764893,
|
| 681 |
+
"learning_rate": 4.666666666666667e-06,
|
| 682 |
+
"loss": 0.1955,
|
| 683 |
"step": 860
|
| 684 |
},
|
| 685 |
{
|
| 686 |
+
"epoch": 8.7,
|
| 687 |
+
"grad_norm": 2.5662710666656494,
|
| 688 |
+
"learning_rate": 4.333333333333333e-06,
|
| 689 |
+
"loss": 0.2008,
|
| 690 |
"step": 870
|
| 691 |
},
|
| 692 |
{
|
| 693 |
+
"epoch": 8.8,
|
| 694 |
+
"grad_norm": 1.984096884727478,
|
| 695 |
+
"learning_rate": 4e-06,
|
| 696 |
+
"loss": 0.186,
|
| 697 |
"step": 880
|
| 698 |
},
|
| 699 |
{
|
| 700 |
+
"epoch": 8.9,
|
| 701 |
+
"grad_norm": 1.4925553798675537,
|
| 702 |
+
"learning_rate": 3.6666666666666666e-06,
|
| 703 |
+
"loss": 0.1905,
|
| 704 |
"step": 890
|
| 705 |
},
|
| 706 |
{
|
| 707 |
+
"epoch": 9.0,
|
| 708 |
+
"grad_norm": 4.499406814575195,
|
| 709 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 710 |
+
"loss": 0.1742,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 9.0,
|
| 715 |
+
"eval_accuracy": 0.9796999117387467,
|
| 716 |
+
"eval_loss": 0.12820769846439362,
|
| 717 |
+
"eval_runtime": 5.0266,
|
| 718 |
+
"eval_samples_per_second": 1352.417,
|
| 719 |
+
"eval_steps_per_second": 42.375,
|
| 720 |
"step": 900
|
| 721 |
},
|
| 722 |
{
|
| 723 |
+
"epoch": 9.1,
|
| 724 |
+
"grad_norm": 1.9934319257736206,
|
| 725 |
+
"learning_rate": 3e-06,
|
| 726 |
+
"loss": 0.1915,
|
| 727 |
"step": 910
|
| 728 |
},
|
| 729 |
{
|
| 730 |
+
"epoch": 9.2,
|
| 731 |
+
"grad_norm": 1.9417580366134644,
|
| 732 |
+
"learning_rate": 2.666666666666667e-06,
|
| 733 |
+
"loss": 0.1759,
|
| 734 |
"step": 920
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"epoch": 9.3,
|
| 738 |
+
"grad_norm": 2.1848134994506836,
|
| 739 |
+
"learning_rate": 2.3333333333333336e-06,
|
| 740 |
+
"loss": 0.1811,
|
| 741 |
"step": 930
|
| 742 |
},
|
| 743 |
{
|
| 744 |
+
"epoch": 9.4,
|
| 745 |
+
"grad_norm": 1.6830646991729736,
|
| 746 |
+
"learning_rate": 2e-06,
|
| 747 |
+
"loss": 0.1905,
|
| 748 |
"step": 940
|
| 749 |
},
|
| 750 |
{
|
| 751 |
+
"epoch": 9.5,
|
| 752 |
+
"grad_norm": 2.51246976852417,
|
| 753 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 754 |
+
"loss": 0.1889,
|
| 755 |
"step": 950
|
| 756 |
},
|
| 757 |
{
|
| 758 |
+
"epoch": 9.6,
|
| 759 |
+
"grad_norm": 2.315727949142456,
|
| 760 |
+
"learning_rate": 1.3333333333333334e-06,
|
| 761 |
+
"loss": 0.1755,
|
| 762 |
"step": 960
|
| 763 |
},
|
| 764 |
{
|
| 765 |
+
"epoch": 9.7,
|
| 766 |
+
"grad_norm": 1.6819000244140625,
|
| 767 |
+
"learning_rate": 1e-06,
|
| 768 |
+
"loss": 0.1717,
|
| 769 |
"step": 970
|
| 770 |
},
|
| 771 |
{
|
| 772 |
+
"epoch": 9.8,
|
| 773 |
+
"grad_norm": 1.7445892095565796,
|
| 774 |
+
"learning_rate": 6.666666666666667e-07,
|
| 775 |
+
"loss": 0.1756,
|
| 776 |
"step": 980
|
| 777 |
},
|
| 778 |
{
|
| 779 |
+
"epoch": 9.9,
|
| 780 |
+
"grad_norm": 1.542716145515442,
|
| 781 |
+
"learning_rate": 3.3333333333333335e-07,
|
| 782 |
+
"loss": 0.181,
|
| 783 |
"step": 990
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 10.0,
|
| 787 |
+
"grad_norm": 2.6007721424102783,
|
| 788 |
+
"learning_rate": 0.0,
|
| 789 |
+
"loss": 0.1867,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 790 |
"step": 1000
|
| 791 |
},
|
| 792 |
{
|
| 793 |
+
"epoch": 10.0,
|
| 794 |
+
"eval_accuracy": 0.9811709326272433,
|
| 795 |
+
"eval_loss": 0.12430207431316376,
|
| 796 |
+
"eval_runtime": 5.3794,
|
| 797 |
+
"eval_samples_per_second": 1263.701,
|
| 798 |
+
"eval_steps_per_second": 39.595,
|
| 799 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 800 |
},
|
| 801 |
{
|
| 802 |
+
"epoch": 10.0,
|
| 803 |
+
"step": 1000,
|
| 804 |
+
"total_flos": 4.72566865822464e+18,
|
| 805 |
+
"train_loss": 0.7669839228391647,
|
| 806 |
+
"train_runtime": 621.093,
|
| 807 |
+
"train_samples_per_second": 822.646,
|
| 808 |
+
"train_steps_per_second": 1.61
|
| 809 |
}
|
| 810 |
],
|
| 811 |
"logging_steps": 10,
|
| 812 |
+
"max_steps": 1000,
|
| 813 |
"num_input_tokens_seen": 0,
|
| 814 |
+
"num_train_epochs": 10,
|
| 815 |
"save_steps": 500,
|
| 816 |
"stateful_callbacks": {
|
| 817 |
"TrainerControl": {
|
|
|
|
| 825 |
"attributes": {}
|
| 826 |
}
|
| 827 |
},
|
| 828 |
+
"total_flos": 4.72566865822464e+18,
|
| 829 |
+
"train_batch_size": 128,
|
| 830 |
"trial_name": null,
|
| 831 |
"trial_params": null
|
| 832 |
}
|