End of training
Browse files- README.md +8 -6
- all_results.json +13 -11
- eval_results.json +9 -6
- train_results.json +5 -6
- trainer_state.json +908 -1348
README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: facebook/dinov2-small-imagenet1k-1-layer
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -18,13 +20,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# dinov2-small-imagenet1k-1-layer-finetuned-galaxy10-decals
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [facebook/dinov2-small-imagenet1k-1-layer](https://huggingface.co/facebook/dinov2-small-imagenet1k-1-layer) on
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
-
- Accuracy: 0.
|
| 25 |
-
- Precision: 0.
|
| 26 |
-
- Recall: 0.
|
| 27 |
-
- F1: 0.
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: facebook/dinov2-small-imagenet1k-1-layer
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
+
- vision
|
| 7 |
- generated_from_trainer
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
|
|
|
| 20 |
|
| 21 |
# dinov2-small-imagenet1k-1-layer-finetuned-galaxy10-decals
|
| 22 |
|
| 23 |
+
This model is a fine-tuned version of [facebook/dinov2-small-imagenet1k-1-layer](https://huggingface.co/facebook/dinov2-small-imagenet1k-1-layer) on the matthieulel/galaxy10_decals dataset.
|
| 24 |
It achieves the following results on the evaluation set:
|
| 25 |
+
- Loss: 0.5373
|
| 26 |
+
- Accuracy: 0.8563
|
| 27 |
+
- Precision: 0.8536
|
| 28 |
+
- Recall: 0.8563
|
| 29 |
+
- F1: 0.8543
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
| 8 |
-
"
|
| 9 |
-
"
|
| 10 |
-
"
|
| 11 |
-
"
|
| 12 |
-
"
|
|
|
|
|
|
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"eval_accuracy": 0.8562570462232244,
|
| 4 |
+
"eval_f1": 0.8543013145108364,
|
| 5 |
+
"eval_loss": 0.5373367667198181,
|
| 6 |
+
"eval_precision": 0.8535856101978109,
|
| 7 |
+
"eval_recall": 0.8562570462232244,
|
| 8 |
+
"eval_runtime": 9.4749,
|
| 9 |
+
"eval_samples_per_second": 187.232,
|
| 10 |
+
"eval_steps_per_second": 2.955,
|
| 11 |
+
"train_loss": 0.4955275515074371,
|
| 12 |
+
"train_runtime": 4173.3159,
|
| 13 |
+
"train_samples_per_second": 114.743,
|
| 14 |
+
"train_steps_per_second": 0.446
|
| 15 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,11 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"eval_accuracy": 0.8562570462232244,
|
| 4 |
+
"eval_f1": 0.8543013145108364,
|
| 5 |
+
"eval_loss": 0.5373367667198181,
|
| 6 |
+
"eval_precision": 0.8535856101978109,
|
| 7 |
+
"eval_recall": 0.8562570462232244,
|
| 8 |
+
"eval_runtime": 9.4749,
|
| 9 |
+
"eval_samples_per_second": 187.232,
|
| 10 |
+
"eval_steps_per_second": 2.955
|
| 11 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"train_steps_per_second": 0.347
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"train_loss": 0.4955275515074371,
|
| 4 |
+
"train_runtime": 4173.3159,
|
| 5 |
+
"train_samples_per_second": 114.743,
|
| 6 |
+
"train_steps_per_second": 0.446
|
|
|
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1946 +1,1506 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "dinov2-small-imagenet1k-1-layer-finetuned-galaxy10-decals/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 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"loss": 2.8233,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"
|
| 21 |
-
"
|
| 22 |
-
"loss": 2.1402,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"loss": 1.8321,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"loss": 1.5098,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"
|
| 42 |
-
"
|
| 43 |
-
"loss": 1.3558,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"
|
| 49 |
-
"
|
| 50 |
-
"loss": 1.2094,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"
|
| 56 |
-
"
|
| 57 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch":
|
| 62 |
-
"
|
| 63 |
-
"
|
| 64 |
-
"loss": 1.1189,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch":
|
| 69 |
-
"
|
| 70 |
-
"
|
| 71 |
-
"loss": 1.0405,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch":
|
| 76 |
-
"
|
| 77 |
-
"
|
| 78 |
-
"loss": 1.0129,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"loss": 0.9405,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch":
|
| 90 |
-
"
|
| 91 |
-
"
|
| 92 |
-
"loss": 0.924,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0
|
| 97 |
-
"eval_accuracy": 0.
|
| 98 |
-
"
|
| 99 |
-
"
|
| 100 |
-
"
|
| 101 |
-
"
|
| 102 |
-
"
|
|
|
|
|
|
|
|
|
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"
|
| 107 |
-
"
|
| 108 |
-
"loss": 0.8607,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
-
"epoch":
|
| 113 |
-
"
|
| 114 |
-
"
|
| 115 |
-
"loss": 0.9175,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
-
"
|
| 121 |
-
"
|
| 122 |
-
"loss": 0.8886,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"
|
| 128 |
-
"
|
| 129 |
-
"loss": 0.9134,
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"epoch":
|
| 134 |
-
"
|
| 135 |
-
"
|
| 136 |
-
"loss": 0.8889,
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
-
"epoch":
|
| 141 |
-
"
|
| 142 |
-
"
|
| 143 |
-
"loss": 0.9035,
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
-
"epoch":
|
| 148 |
-
"
|
| 149 |
-
"
|
| 150 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
-
"epoch":
|
| 155 |
-
"
|
| 156 |
-
"
|
| 157 |
-
"loss": 0.9545,
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch":
|
| 162 |
-
"
|
| 163 |
-
"
|
| 164 |
-
"loss": 0.9195,
|
| 165 |
"step": 210
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"
|
| 170 |
-
"
|
| 171 |
-
"loss": 0.9062,
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
-
"epoch":
|
| 176 |
-
"
|
| 177 |
-
"
|
| 178 |
-
"loss": 0.8817,
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
-
"epoch":
|
| 183 |
-
"grad_norm": 59.98125457763672,
|
| 184 |
"learning_rate": 4.8387096774193554e-05,
|
| 185 |
-
"loss": 0.
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
-
"epoch":
|
| 190 |
-
"
|
| 191 |
-
"
|
| 192 |
-
"
|
| 193 |
-
"eval_samples_per_second": 108.05,
|
| 194 |
-
"eval_steps_per_second": 3.411,
|
| 195 |
-
"step": 249
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
-
"
|
| 200 |
-
"
|
| 201 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 202 |
"step": 250
|
| 203 |
},
|
| 204 |
{
|
| 205 |
-
"epoch":
|
| 206 |
-
"
|
| 207 |
-
"
|
| 208 |
-
"loss": 0.8578,
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
-
"epoch":
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"loss": 0.9822,
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
-
"epoch":
|
| 220 |
-
"
|
| 221 |
-
"
|
| 222 |
-
"loss": 0.9396,
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
-
"epoch":
|
| 227 |
-
"
|
| 228 |
-
"
|
| 229 |
-
"loss": 0.9457,
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch":
|
| 234 |
-
"
|
| 235 |
-
"
|
| 236 |
-
"loss": 0.8546,
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
-
"
|
| 242 |
-
"
|
| 243 |
-
"loss": 0.9152,
|
| 244 |
"step": 310
|
| 245 |
},
|
| 246 |
{
|
| 247 |
-
"epoch":
|
| 248 |
-
"
|
| 249 |
-
"
|
| 250 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 251 |
"step": 320
|
| 252 |
},
|
| 253 |
{
|
| 254 |
-
"epoch":
|
| 255 |
-
"
|
| 256 |
-
"
|
| 257 |
-
"loss": 0.7998,
|
| 258 |
"step": 330
|
| 259 |
},
|
| 260 |
{
|
| 261 |
-
"epoch":
|
| 262 |
-
"
|
| 263 |
-
"
|
| 264 |
-
"loss": 0.7981,
|
| 265 |
"step": 340
|
| 266 |
},
|
| 267 |
{
|
| 268 |
-
"epoch":
|
| 269 |
-
"
|
| 270 |
-
"
|
| 271 |
-
"loss": 0.8871,
|
| 272 |
"step": 350
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch":
|
| 276 |
-
"
|
| 277 |
-
"
|
| 278 |
-
"loss": 0.7689,
|
| 279 |
"step": 360
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"
|
| 284 |
-
"
|
| 285 |
-
"loss": 0.805,
|
| 286 |
"step": 370
|
| 287 |
},
|
| 288 |
{
|
| 289 |
-
"epoch":
|
| 290 |
-
"eval_accuracy": 0.
|
| 291 |
-
"
|
| 292 |
-
"
|
| 293 |
-
"
|
| 294 |
-
"
|
| 295 |
-
"
|
|
|
|
|
|
|
|
|
|
| 296 |
},
|
| 297 |
{
|
| 298 |
-
"epoch":
|
| 299 |
-
"
|
| 300 |
-
"
|
| 301 |
-
"loss": 0.7415,
|
| 302 |
"step": 380
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch":
|
| 306 |
-
"
|
| 307 |
-
"
|
| 308 |
-
"loss": 0.7203,
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"
|
| 314 |
-
"
|
| 315 |
-
"loss": 0.6812,
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
-
"epoch":
|
| 320 |
-
"
|
| 321 |
-
"
|
| 322 |
-
"loss": 0.7374,
|
| 323 |
"step": 410
|
| 324 |
},
|
| 325 |
{
|
| 326 |
-
"epoch":
|
| 327 |
-
"
|
| 328 |
-
"
|
| 329 |
-
"loss": 0.7994,
|
| 330 |
"step": 420
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"epoch":
|
| 334 |
-
"
|
| 335 |
-
"
|
| 336 |
-
"loss": 0.7888,
|
| 337 |
"step": 430
|
| 338 |
},
|
| 339 |
{
|
| 340 |
-
"epoch":
|
| 341 |
-
"
|
| 342 |
-
"
|
| 343 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 344 |
"step": 440
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch":
|
| 348 |
-
"
|
| 349 |
-
"
|
| 350 |
-
"loss": 0.778,
|
| 351 |
"step": 450
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch":
|
| 355 |
-
"
|
| 356 |
-
"
|
| 357 |
-
"loss": 0.7487,
|
| 358 |
"step": 460
|
| 359 |
},
|
| 360 |
{
|
| 361 |
-
"epoch":
|
| 362 |
-
"
|
| 363 |
-
"
|
| 364 |
-
"loss": 0.7287,
|
| 365 |
"step": 470
|
| 366 |
},
|
| 367 |
{
|
| 368 |
-
"epoch":
|
| 369 |
-
"
|
| 370 |
-
"
|
| 371 |
-
"loss": 0.7762,
|
| 372 |
"step": 480
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"epoch":
|
| 376 |
-
"
|
| 377 |
-
"
|
| 378 |
-
"loss": 0.752,
|
| 379 |
"step": 490
|
| 380 |
},
|
| 381 |
{
|
| 382 |
-
"epoch":
|
| 383 |
-
"
|
| 384 |
-
"
|
| 385 |
-
"
|
| 386 |
-
|
| 387 |
-
|
| 388 |
-
"
|
| 389 |
-
|
| 390 |
-
|
| 391 |
-
"
|
| 392 |
-
"
|
| 393 |
-
"
|
| 394 |
-
"
|
|
|
|
|
|
|
| 395 |
"step": 500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch":
|
| 399 |
-
"
|
| 400 |
-
"
|
| 401 |
-
"loss": 0.7361,
|
| 402 |
"step": 510
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch":
|
| 406 |
-
"
|
| 407 |
-
"
|
| 408 |
-
"loss": 0.7795,
|
| 409 |
"step": 520
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch":
|
| 413 |
-
"
|
| 414 |
-
"
|
| 415 |
-
"loss": 0.7467,
|
| 416 |
"step": 530
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"
|
| 421 |
-
"
|
| 422 |
-
"loss": 0.6427,
|
| 423 |
"step": 540
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"
|
| 428 |
-
"
|
| 429 |
-
"loss": 0.6988,
|
| 430 |
"step": 550
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch":
|
| 434 |
-
"
|
| 435 |
-
"
|
| 436 |
-
"loss": 0.6696,
|
| 437 |
"step": 560
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"
|
| 442 |
-
"
|
| 443 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 444 |
"step": 570
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch":
|
| 448 |
-
"
|
| 449 |
-
"
|
| 450 |
-
"loss": 0.6775,
|
| 451 |
"step": 580
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch":
|
| 455 |
-
"
|
| 456 |
-
"
|
| 457 |
-
"loss": 0.6479,
|
| 458 |
"step": 590
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"
|
| 463 |
-
"
|
| 464 |
-
"loss": 0.7208,
|
| 465 |
"step": 600
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"
|
| 470 |
-
"
|
| 471 |
-
"loss": 0.6655,
|
| 472 |
"step": 610
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"
|
| 477 |
-
"
|
| 478 |
-
"loss": 0.6193,
|
| 479 |
"step": 620
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"eval_accuracy": 0.
|
| 484 |
-
"
|
| 485 |
-
"
|
| 486 |
-
"
|
| 487 |
-
"
|
| 488 |
-
"
|
|
|
|
|
|
|
|
|
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch":
|
| 492 |
-
"
|
| 493 |
-
"
|
| 494 |
-
"loss": 0.6843,
|
| 495 |
"step": 630
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"
|
| 500 |
-
"
|
| 501 |
-
"loss": 0.6394,
|
| 502 |
"step": 640
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"epoch":
|
| 506 |
-
"
|
| 507 |
-
"
|
| 508 |
-
"loss": 0.6629,
|
| 509 |
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
-
"epoch":
|
| 513 |
-
"
|
| 514 |
-
"
|
| 515 |
-
"loss": 0.6814,
|
| 516 |
"step": 660
|
| 517 |
},
|
| 518 |
{
|
| 519 |
-
"epoch":
|
| 520 |
-
"
|
| 521 |
-
"
|
| 522 |
-
"loss": 0.7281,
|
| 523 |
"step": 670
|
| 524 |
},
|
| 525 |
{
|
| 526 |
-
"epoch":
|
| 527 |
-
"
|
| 528 |
-
"
|
| 529 |
-
"loss": 0.6512,
|
| 530 |
"step": 680
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch":
|
| 534 |
-
"
|
| 535 |
-
"
|
| 536 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 537 |
"step": 690
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch":
|
| 541 |
-
"
|
| 542 |
-
"
|
| 543 |
-
"loss": 0.7244,
|
| 544 |
"step": 700
|
| 545 |
},
|
| 546 |
{
|
| 547 |
-
"epoch":
|
| 548 |
-
"
|
| 549 |
-
"
|
| 550 |
-
"loss": 0.7093,
|
| 551 |
"step": 710
|
| 552 |
},
|
| 553 |
{
|
| 554 |
-
"epoch":
|
| 555 |
-
"
|
| 556 |
-
"
|
| 557 |
-
"loss": 0.6039,
|
| 558 |
"step": 720
|
| 559 |
},
|
| 560 |
{
|
| 561 |
-
"epoch":
|
| 562 |
-
"
|
| 563 |
-
"
|
| 564 |
-
"loss": 0.6434,
|
| 565 |
"step": 730
|
| 566 |
},
|
| 567 |
{
|
| 568 |
-
"epoch":
|
| 569 |
-
"
|
| 570 |
-
"
|
| 571 |
-
"loss": 0.5822,
|
| 572 |
"step": 740
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch":
|
| 576 |
-
"
|
| 577 |
-
"
|
| 578 |
-
"
|
| 579 |
-
"eval_samples_per_second": 107.284,
|
| 580 |
-
"eval_steps_per_second": 3.387,
|
| 581 |
-
"step": 748
|
| 582 |
},
|
| 583 |
{
|
| 584 |
-
"epoch":
|
| 585 |
-
"
|
| 586 |
-
"
|
| 587 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 588 |
"step": 750
|
| 589 |
},
|
| 590 |
{
|
| 591 |
-
"epoch":
|
| 592 |
-
"
|
| 593 |
-
"
|
| 594 |
-
"loss": 0.6292,
|
| 595 |
"step": 760
|
| 596 |
},
|
| 597 |
{
|
| 598 |
-
"epoch":
|
| 599 |
-
"
|
| 600 |
-
"
|
| 601 |
-
"loss": 0.6334,
|
| 602 |
"step": 770
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch":
|
| 606 |
-
"
|
| 607 |
-
"
|
| 608 |
-
"loss": 0.6187,
|
| 609 |
"step": 780
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch":
|
| 613 |
-
"
|
| 614 |
-
"
|
| 615 |
-
"loss": 0.6301,
|
| 616 |
"step": 790
|
| 617 |
},
|
| 618 |
{
|
| 619 |
-
"epoch":
|
| 620 |
-
"
|
| 621 |
-
"
|
| 622 |
-
"loss": 0.5877,
|
| 623 |
"step": 800
|
| 624 |
},
|
| 625 |
{
|
| 626 |
-
"epoch":
|
| 627 |
-
"
|
| 628 |
-
"
|
| 629 |
-
"loss": 0.5988,
|
| 630 |
"step": 810
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch":
|
| 634 |
-
"
|
| 635 |
-
"
|
| 636 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 637 |
"step": 820
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch":
|
| 641 |
-
"
|
| 642 |
-
"
|
| 643 |
-
"loss": 0.6522,
|
| 644 |
"step": 830
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch":
|
| 648 |
-
"
|
| 649 |
-
"
|
| 650 |
-
"loss": 0.6015,
|
| 651 |
"step": 840
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch": 6
|
| 655 |
-
"
|
| 656 |
-
"
|
| 657 |
-
"loss": 0.6019,
|
| 658 |
"step": 850
|
| 659 |
},
|
| 660 |
{
|
| 661 |
-
"epoch":
|
| 662 |
-
"
|
| 663 |
-
"
|
| 664 |
-
"loss": 0.6806,
|
| 665 |
"step": 860
|
| 666 |
},
|
| 667 |
{
|
| 668 |
-
"epoch":
|
| 669 |
-
"
|
| 670 |
-
"
|
| 671 |
-
"loss": 0.6227,
|
| 672 |
"step": 870
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"epoch":
|
| 676 |
-
"eval_accuracy": 0.
|
| 677 |
-
"
|
| 678 |
-
"
|
| 679 |
-
"
|
| 680 |
-
"
|
| 681 |
-
"
|
|
|
|
|
|
|
|
|
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch":
|
| 685 |
-
"
|
| 686 |
-
"
|
| 687 |
-
"loss": 0.5791,
|
| 688 |
"step": 880
|
| 689 |
},
|
| 690 |
{
|
| 691 |
-
"epoch":
|
| 692 |
-
"
|
| 693 |
-
"
|
| 694 |
-
"loss": 0.6049,
|
| 695 |
"step": 890
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch":
|
| 699 |
-
"
|
| 700 |
-
"
|
| 701 |
-
"loss": 0.6177,
|
| 702 |
"step": 900
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch":
|
| 706 |
-
"
|
| 707 |
-
"
|
| 708 |
-
"loss": 0.5685,
|
| 709 |
"step": 910
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch":
|
| 713 |
-
"
|
| 714 |
-
"
|
| 715 |
-
"loss": 0.5914,
|
| 716 |
"step": 920
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch":
|
| 720 |
-
"
|
| 721 |
-
"
|
| 722 |
-
"loss": 0.6416,
|
| 723 |
"step": 930
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch":
|
| 727 |
-
"
|
| 728 |
-
"
|
| 729 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 730 |
"step": 940
|
| 731 |
},
|
| 732 |
{
|
| 733 |
-
"epoch":
|
| 734 |
-
"
|
| 735 |
-
"
|
| 736 |
-
"loss": 0.6198,
|
| 737 |
"step": 950
|
| 738 |
},
|
| 739 |
{
|
| 740 |
-
"epoch":
|
| 741 |
-
"
|
| 742 |
-
"
|
| 743 |
-
"loss": 0.5378,
|
| 744 |
"step": 960
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch":
|
| 748 |
-
"
|
| 749 |
-
"
|
| 750 |
-
"loss": 0.607,
|
| 751 |
"step": 970
|
| 752 |
},
|
| 753 |
{
|
| 754 |
-
"epoch":
|
| 755 |
-
"
|
| 756 |
-
"
|
| 757 |
-
"loss": 0.5962,
|
| 758 |
"step": 980
|
| 759 |
},
|
| 760 |
{
|
| 761 |
-
"epoch":
|
| 762 |
-
"
|
| 763 |
-
"
|
| 764 |
-
"loss": 0.6067,
|
| 765 |
"step": 990
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
-
"
|
| 770 |
-
"
|
| 771 |
-
"
|
| 772 |
-
"eval_samples_per_second": 107.07,
|
| 773 |
-
"eval_steps_per_second": 3.38,
|
| 774 |
-
"step": 998
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"epoch":
|
| 778 |
-
"
|
| 779 |
-
"
|
| 780 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 781 |
"step": 1000
|
| 782 |
},
|
| 783 |
{
|
| 784 |
-
"epoch":
|
| 785 |
-
"
|
| 786 |
-
"
|
| 787 |
-
"loss": 0.5297,
|
| 788 |
"step": 1010
|
| 789 |
},
|
| 790 |
{
|
| 791 |
-
"epoch":
|
| 792 |
-
"
|
| 793 |
-
"
|
| 794 |
-
"loss": 0.5678,
|
| 795 |
"step": 1020
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch":
|
| 799 |
-
"
|
| 800 |
-
"
|
| 801 |
-
"loss": 0.5404,
|
| 802 |
"step": 1030
|
| 803 |
},
|
| 804 |
{
|
| 805 |
-
"epoch":
|
| 806 |
-
"
|
| 807 |
-
"
|
| 808 |
-
"loss": 0.535,
|
| 809 |
"step": 1040
|
| 810 |
},
|
| 811 |
{
|
| 812 |
-
"epoch": 8
|
| 813 |
-
"
|
| 814 |
-
"
|
| 815 |
-
"loss": 0.5622,
|
| 816 |
"step": 1050
|
| 817 |
},
|
| 818 |
{
|
| 819 |
-
"epoch":
|
| 820 |
-
"
|
| 821 |
-
"
|
| 822 |
-
"loss": 0.5527,
|
| 823 |
"step": 1060
|
| 824 |
},
|
| 825 |
{
|
| 826 |
-
"epoch":
|
| 827 |
-
"
|
| 828 |
-
"
|
| 829 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 830 |
"step": 1070
|
| 831 |
},
|
| 832 |
{
|
| 833 |
-
"epoch":
|
| 834 |
-
"
|
| 835 |
-
"
|
| 836 |
-
"loss": 0.608,
|
| 837 |
"step": 1080
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch":
|
| 841 |
-
"
|
| 842 |
-
"
|
| 843 |
-
"loss": 0.5628,
|
| 844 |
"step": 1090
|
| 845 |
},
|
| 846 |
{
|
| 847 |
-
"epoch":
|
| 848 |
-
"
|
| 849 |
-
"
|
| 850 |
-
"loss": 0.5076,
|
| 851 |
"step": 1100
|
| 852 |
},
|
| 853 |
{
|
| 854 |
-
"epoch":
|
| 855 |
-
"
|
| 856 |
-
"
|
| 857 |
-
"loss": 0.5299,
|
| 858 |
"step": 1110
|
| 859 |
},
|
| 860 |
{
|
| 861 |
-
"epoch":
|
| 862 |
-
"
|
| 863 |
-
"
|
| 864 |
-
"loss": 0.533,
|
| 865 |
"step": 1120
|
| 866 |
},
|
| 867 |
{
|
| 868 |
-
"epoch":
|
| 869 |
-
"eval_accuracy": 0.
|
| 870 |
-
"
|
| 871 |
-
"
|
| 872 |
-
"
|
| 873 |
-
"
|
| 874 |
-
"
|
|
|
|
|
|
|
|
|
|
| 875 |
},
|
| 876 |
{
|
| 877 |
-
"epoch":
|
| 878 |
-
"
|
| 879 |
-
"
|
| 880 |
-
"loss": 0.6276,
|
| 881 |
"step": 1130
|
| 882 |
},
|
| 883 |
{
|
| 884 |
-
"epoch":
|
| 885 |
-
"
|
| 886 |
-
"
|
| 887 |
-
"loss": 0.5674,
|
| 888 |
"step": 1140
|
| 889 |
},
|
| 890 |
{
|
| 891 |
-
"epoch":
|
| 892 |
-
"
|
| 893 |
-
"
|
| 894 |
-
"loss": 0.5189,
|
| 895 |
"step": 1150
|
| 896 |
},
|
| 897 |
{
|
| 898 |
-
"epoch":
|
| 899 |
-
"
|
| 900 |
-
"
|
| 901 |
-
"loss": 0.5086,
|
| 902 |
"step": 1160
|
| 903 |
},
|
| 904 |
{
|
| 905 |
-
"epoch":
|
| 906 |
-
"
|
| 907 |
-
"
|
| 908 |
-
"loss": 0.5137,
|
| 909 |
"step": 1170
|
| 910 |
},
|
| 911 |
{
|
| 912 |
-
"epoch":
|
| 913 |
-
"
|
| 914 |
-
"
|
| 915 |
-
"loss": 0.4647,
|
| 916 |
"step": 1180
|
| 917 |
},
|
| 918 |
{
|
| 919 |
-
"epoch":
|
| 920 |
-
"
|
| 921 |
-
"
|
| 922 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 923 |
"step": 1190
|
| 924 |
},
|
| 925 |
{
|
| 926 |
-
"epoch":
|
| 927 |
-
"
|
| 928 |
-
"
|
| 929 |
-
"loss": 0.5256,
|
| 930 |
"step": 1200
|
| 931 |
},
|
| 932 |
{
|
| 933 |
-
"epoch":
|
| 934 |
-
"
|
| 935 |
-
"
|
| 936 |
-
"loss": 0.5121,
|
| 937 |
"step": 1210
|
| 938 |
},
|
| 939 |
{
|
| 940 |
-
"epoch":
|
| 941 |
-
"
|
| 942 |
-
"
|
| 943 |
-
"loss": 0.5393,
|
| 944 |
"step": 1220
|
| 945 |
},
|
| 946 |
{
|
| 947 |
-
"epoch":
|
| 948 |
-
"
|
| 949 |
-
"
|
| 950 |
-
"loss": 0.478,
|
| 951 |
"step": 1230
|
| 952 |
},
|
| 953 |
{
|
| 954 |
-
"epoch":
|
| 955 |
-
"
|
| 956 |
-
"
|
| 957 |
-
"loss": 0.5423,
|
| 958 |
"step": 1240
|
| 959 |
},
|
| 960 |
{
|
| 961 |
-
"epoch":
|
| 962 |
-
"
|
| 963 |
-
"
|
| 964 |
-
"
|
| 965 |
-
"eval_samples_per_second": 106.94,
|
| 966 |
-
"eval_steps_per_second": 3.376,
|
| 967 |
-
"step": 1247
|
| 968 |
},
|
| 969 |
{
|
| 970 |
-
"epoch":
|
| 971 |
-
"
|
| 972 |
-
"
|
| 973 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 974 |
"step": 1250
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"epoch":
|
| 978 |
-
"
|
| 979 |
-
"
|
| 980 |
-
"loss": 0.4834,
|
| 981 |
"step": 1260
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch":
|
| 985 |
-
"
|
| 986 |
-
"
|
| 987 |
-
"loss": 0.4799,
|
| 988 |
"step": 1270
|
| 989 |
},
|
| 990 |
{
|
| 991 |
-
"epoch":
|
| 992 |
-
"
|
| 993 |
-
"
|
| 994 |
-
"loss": 0.5015,
|
| 995 |
"step": 1280
|
| 996 |
},
|
| 997 |
{
|
| 998 |
-
"epoch":
|
| 999 |
-
"
|
| 1000 |
-
"
|
| 1001 |
-
"loss": 0.5561,
|
| 1002 |
"step": 1290
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
-
"epoch":
|
| 1006 |
-
"
|
| 1007 |
-
"
|
| 1008 |
-
"loss": 0.479,
|
| 1009 |
"step": 1300
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
-
"epoch":
|
| 1013 |
-
"
|
| 1014 |
-
"
|
| 1015 |
-
"loss": 0.5283,
|
| 1016 |
"step": 1310
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
-
"epoch":
|
| 1020 |
-
"
|
| 1021 |
-
"
|
| 1022 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1023 |
"step": 1320
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
-
"epoch":
|
| 1027 |
-
"
|
| 1028 |
-
"
|
| 1029 |
-
"loss": 0.4783,
|
| 1030 |
"step": 1330
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
-
"epoch":
|
| 1034 |
-
"
|
| 1035 |
-
"
|
| 1036 |
-
"loss": 0.4997,
|
| 1037 |
"step": 1340
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
-
"epoch":
|
| 1041 |
-
"
|
| 1042 |
-
"
|
| 1043 |
-
"loss": 0.4518,
|
| 1044 |
"step": 1350
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"epoch":
|
| 1048 |
-
"
|
| 1049 |
-
"
|
| 1050 |
-
"loss": 0.5313,
|
| 1051 |
"step": 1360
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
-
"epoch":
|
| 1055 |
-
"
|
| 1056 |
-
"
|
| 1057 |
-
"loss": 0.4747,
|
| 1058 |
"step": 1370
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
-
"epoch":
|
| 1062 |
-
"eval_accuracy": 0.
|
| 1063 |
-
"
|
| 1064 |
-
"
|
| 1065 |
-
"
|
| 1066 |
-
"
|
| 1067 |
-
"
|
|
|
|
|
|
|
|
|
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch":
|
| 1071 |
-
"
|
| 1072 |
-
"
|
| 1073 |
-
"loss": 0.4848,
|
| 1074 |
"step": 1380
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
-
"epoch":
|
| 1078 |
-
"
|
| 1079 |
-
"
|
| 1080 |
-
"loss": 0.4434,
|
| 1081 |
"step": 1390
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
-
"epoch":
|
| 1085 |
-
"
|
| 1086 |
-
"
|
| 1087 |
-
"loss": 0.4665,
|
| 1088 |
"step": 1400
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
-
"epoch":
|
| 1092 |
-
"
|
| 1093 |
-
"
|
| 1094 |
-
"loss": 0.45,
|
| 1095 |
"step": 1410
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
-
"
|
| 1100 |
-
"
|
| 1101 |
-
"loss": 0.5073,
|
| 1102 |
"step": 1420
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
-
"epoch":
|
| 1106 |
-
"
|
| 1107 |
-
"
|
| 1108 |
-
"loss": 0.5099,
|
| 1109 |
"step": 1430
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch":
|
| 1113 |
-
"
|
| 1114 |
-
"
|
| 1115 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1116 |
"step": 1440
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
-
"epoch":
|
| 1120 |
-
"
|
| 1121 |
-
"
|
| 1122 |
-
"loss": 0.4723,
|
| 1123 |
"step": 1450
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
-
"epoch":
|
| 1127 |
-
"
|
| 1128 |
-
"
|
| 1129 |
-
"loss": 0.4613,
|
| 1130 |
"step": 1460
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
-
"epoch":
|
| 1134 |
-
"
|
| 1135 |
-
"
|
| 1136 |
-
"loss": 0.4884,
|
| 1137 |
"step": 1470
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
-
"epoch":
|
| 1141 |
-
"
|
| 1142 |
-
"
|
| 1143 |
-
"loss": 0.4842,
|
| 1144 |
"step": 1480
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
-
"epoch":
|
| 1148 |
-
"
|
| 1149 |
-
"
|
| 1150 |
-
"loss": 0.4367,
|
| 1151 |
"step": 1490
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
-
"epoch":
|
| 1155 |
-
"
|
| 1156 |
-
"
|
| 1157 |
-
"eval_runtime": 16.5802,
|
| 1158 |
-
"eval_samples_per_second": 106.995,
|
| 1159 |
-
"eval_steps_per_second": 3.378,
|
| 1160 |
-
"step": 1497
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 12.024048096192384,
|
| 1164 |
-
"grad_norm": 21.091793060302734,
|
| 1165 |
-
"learning_rate": 2.1953405017921146e-05,
|
| 1166 |
-
"loss": 0.4565,
|
| 1167 |
"step": 1500
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
-
"epoch":
|
| 1171 |
-
"
|
| 1172 |
-
"
|
| 1173 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1174 |
"step": 1510
|
| 1175 |
},
|
| 1176 |
{
|
| 1177 |
-
"epoch":
|
| 1178 |
-
"
|
| 1179 |
-
"
|
| 1180 |
-
"loss": 0.4109,
|
| 1181 |
"step": 1520
|
| 1182 |
},
|
| 1183 |
{
|
| 1184 |
-
"epoch":
|
| 1185 |
-
"
|
| 1186 |
-
"
|
| 1187 |
-
"loss": 0.4572,
|
| 1188 |
"step": 1530
|
| 1189 |
},
|
| 1190 |
{
|
| 1191 |
-
"epoch":
|
| 1192 |
-
"
|
| 1193 |
-
"
|
| 1194 |
-
"loss": 0.4693,
|
| 1195 |
"step": 1540
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
-
"epoch":
|
| 1199 |
-
"
|
| 1200 |
-
"
|
| 1201 |
-
"loss": 0.4446,
|
| 1202 |
"step": 1550
|
| 1203 |
},
|
| 1204 |
{
|
| 1205 |
-
"epoch":
|
| 1206 |
-
"
|
| 1207 |
-
"
|
| 1208 |
-
"loss": 0.4272,
|
| 1209 |
"step": 1560
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
-
"epoch":
|
| 1213 |
-
"
|
| 1214 |
-
"
|
| 1215 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1216 |
"step": 1570
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
-
"epoch":
|
| 1220 |
-
"
|
| 1221 |
-
"
|
| 1222 |
-
"loss": 0.4416,
|
| 1223 |
"step": 1580
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
-
"epoch":
|
| 1227 |
-
"
|
| 1228 |
-
"
|
| 1229 |
-
"loss": 0.4517,
|
| 1230 |
"step": 1590
|
| 1231 |
},
|
| 1232 |
{
|
| 1233 |
-
"epoch":
|
| 1234 |
-
"
|
| 1235 |
-
"
|
| 1236 |
-
"loss": 0.4458,
|
| 1237 |
"step": 1600
|
| 1238 |
},
|
| 1239 |
{
|
| 1240 |
-
"epoch":
|
| 1241 |
-
"
|
| 1242 |
-
"
|
| 1243 |
-
"loss": 0.4362,
|
| 1244 |
"step": 1610
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
-
"epoch":
|
| 1248 |
-
"
|
| 1249 |
-
"
|
| 1250 |
-
"loss": 0.4907,
|
| 1251 |
"step": 1620
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
-
"epoch":
|
| 1255 |
-
"eval_accuracy": 0.
|
| 1256 |
-
"
|
| 1257 |
-
"
|
| 1258 |
-
"
|
| 1259 |
-
"
|
| 1260 |
-
"
|
| 1261 |
-
|
| 1262 |
-
|
| 1263 |
-
"
|
| 1264 |
-
|
| 1265 |
-
|
| 1266 |
-
"
|
|
|
|
|
|
|
| 1267 |
"step": 1630
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
-
"epoch":
|
| 1271 |
-
"
|
| 1272 |
-
"
|
| 1273 |
-
"loss": 0.4026,
|
| 1274 |
"step": 1640
|
| 1275 |
},
|
| 1276 |
{
|
| 1277 |
-
"epoch":
|
| 1278 |
-
"
|
| 1279 |
-
"
|
| 1280 |
-
"loss": 0.4015,
|
| 1281 |
"step": 1650
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
-
"epoch":
|
| 1285 |
-
"
|
| 1286 |
-
"
|
| 1287 |
-
"loss": 0.4238,
|
| 1288 |
"step": 1660
|
| 1289 |
},
|
| 1290 |
{
|
| 1291 |
-
"epoch":
|
| 1292 |
-
"
|
| 1293 |
-
"
|
| 1294 |
-
"loss": 0.4483,
|
| 1295 |
"step": 1670
|
| 1296 |
},
|
| 1297 |
{
|
| 1298 |
-
"epoch":
|
| 1299 |
-
"
|
| 1300 |
-
"
|
| 1301 |
-
"loss": 0.3972,
|
| 1302 |
"step": 1680
|
| 1303 |
},
|
| 1304 |
{
|
| 1305 |
-
"epoch":
|
| 1306 |
-
"
|
| 1307 |
-
"
|
| 1308 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1309 |
"step": 1690
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
-
"epoch":
|
| 1313 |
-
"
|
| 1314 |
-
"
|
| 1315 |
-
"loss": 0.3985,
|
| 1316 |
"step": 1700
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
-
"epoch":
|
| 1320 |
-
"
|
| 1321 |
-
"
|
| 1322 |
-
"loss": 0.4334,
|
| 1323 |
"step": 1710
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
-
"epoch":
|
| 1327 |
-
"
|
| 1328 |
-
"
|
| 1329 |
-
"loss": 0.4494,
|
| 1330 |
"step": 1720
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
-
"epoch":
|
| 1334 |
-
"
|
| 1335 |
-
"
|
| 1336 |
-
"loss": 0.4272,
|
| 1337 |
"step": 1730
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
-
"epoch":
|
| 1341 |
-
"
|
| 1342 |
-
"
|
| 1343 |
-
"loss": 0.4269,
|
| 1344 |
"step": 1740
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
-
"epoch":
|
| 1348 |
-
"
|
| 1349 |
-
"
|
| 1350 |
-
"
|
| 1351 |
-
"eval_samples_per_second": 106.972,
|
| 1352 |
-
"eval_steps_per_second": 3.377,
|
| 1353 |
-
"step": 1746
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
-
"epoch":
|
| 1357 |
-
"
|
| 1358 |
-
"
|
| 1359 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1360 |
"step": 1750
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
-
"epoch":
|
| 1364 |
-
"
|
| 1365 |
-
"
|
| 1366 |
-
"loss": 0.4569,
|
| 1367 |
"step": 1760
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
-
"epoch":
|
| 1371 |
-
"
|
| 1372 |
-
"
|
| 1373 |
-
"loss": 0.3897,
|
| 1374 |
"step": 1770
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
-
"epoch":
|
| 1378 |
-
"
|
| 1379 |
-
"
|
| 1380 |
-
"loss": 0.375,
|
| 1381 |
"step": 1780
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
-
"epoch":
|
| 1385 |
-
"
|
| 1386 |
-
"
|
| 1387 |
-
"loss": 0.3994,
|
| 1388 |
"step": 1790
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
-
"epoch":
|
| 1392 |
-
"
|
| 1393 |
-
"
|
| 1394 |
-
"loss": 0.4204,
|
| 1395 |
"step": 1800
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
-
"epoch":
|
| 1399 |
-
"
|
| 1400 |
-
"
|
| 1401 |
-
"loss": 0.4208,
|
| 1402 |
"step": 1810
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
-
"epoch":
|
| 1406 |
-
"
|
| 1407 |
-
"
|
| 1408 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1409 |
"step": 1820
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
-
"epoch":
|
| 1413 |
-
"
|
| 1414 |
-
"
|
| 1415 |
-
"loss": 0.3895,
|
| 1416 |
"step": 1830
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
-
"epoch":
|
| 1420 |
-
"
|
| 1421 |
-
"
|
| 1422 |
-
"loss": 0.3871,
|
| 1423 |
"step": 1840
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
-
"epoch":
|
| 1427 |
-
"
|
| 1428 |
-
"
|
| 1429 |
-
"loss": 0.4037,
|
| 1430 |
"step": 1850
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
-
"epoch":
|
| 1434 |
-
"
|
| 1435 |
-
"
|
| 1436 |
-
"loss": 0.409,
|
| 1437 |
"step": 1860
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
-
"epoch":
|
| 1441 |
-
"
|
| 1442 |
-
"
|
| 1443 |
-
"
|
| 1444 |
-
"
|
| 1445 |
-
|
| 1446 |
-
|
| 1447 |
-
"
|
| 1448 |
-
"
|
| 1449 |
-
"
|
| 1450 |
-
"eval_runtime": 16.5967,
|
| 1451 |
-
"eval_samples_per_second": 106.889,
|
| 1452 |
-
"eval_steps_per_second": 3.374,
|
| 1453 |
-
"step": 1871
|
| 1454 |
-
},
|
| 1455 |
-
{
|
| 1456 |
-
"epoch": 15.070140280561123,
|
| 1457 |
-
"grad_norm": 18.869020462036133,
|
| 1458 |
-
"learning_rate": 1.3440860215053763e-05,
|
| 1459 |
-
"loss": 0.3546,
|
| 1460 |
-
"step": 1880
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 15.150300601202405,
|
| 1464 |
-
"grad_norm": 18.53528594970703,
|
| 1465 |
-
"learning_rate": 1.3216845878136202e-05,
|
| 1466 |
-
"loss": 0.3785,
|
| 1467 |
-
"step": 1890
|
| 1468 |
-
},
|
| 1469 |
-
{
|
| 1470 |
-
"epoch": 15.230460921843687,
|
| 1471 |
-
"grad_norm": 25.095417022705078,
|
| 1472 |
-
"learning_rate": 1.2992831541218639e-05,
|
| 1473 |
-
"loss": 0.3921,
|
| 1474 |
-
"step": 1900
|
| 1475 |
-
},
|
| 1476 |
-
{
|
| 1477 |
-
"epoch": 15.31062124248497,
|
| 1478 |
-
"grad_norm": 19.071975708007812,
|
| 1479 |
-
"learning_rate": 1.2768817204301076e-05,
|
| 1480 |
-
"loss": 0.3924,
|
| 1481 |
-
"step": 1910
|
| 1482 |
-
},
|
| 1483 |
-
{
|
| 1484 |
-
"epoch": 15.390781563126252,
|
| 1485 |
-
"grad_norm": 20.396650314331055,
|
| 1486 |
-
"learning_rate": 1.2544802867383513e-05,
|
| 1487 |
-
"loss": 0.3877,
|
| 1488 |
-
"step": 1920
|
| 1489 |
-
},
|
| 1490 |
-
{
|
| 1491 |
-
"epoch": 15.470941883767535,
|
| 1492 |
-
"grad_norm": 26.20189094543457,
|
| 1493 |
-
"learning_rate": 1.232078853046595e-05,
|
| 1494 |
-
"loss": 0.3408,
|
| 1495 |
-
"step": 1930
|
| 1496 |
-
},
|
| 1497 |
-
{
|
| 1498 |
-
"epoch": 15.551102204408817,
|
| 1499 |
-
"grad_norm": 25.609718322753906,
|
| 1500 |
-
"learning_rate": 1.2096774193548388e-05,
|
| 1501 |
-
"loss": 0.3554,
|
| 1502 |
-
"step": 1940
|
| 1503 |
-
},
|
| 1504 |
-
{
|
| 1505 |
-
"epoch": 15.6312625250501,
|
| 1506 |
-
"grad_norm": 22.300622940063477,
|
| 1507 |
-
"learning_rate": 1.1872759856630825e-05,
|
| 1508 |
-
"loss": 0.3408,
|
| 1509 |
-
"step": 1950
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 15.711422845691382,
|
| 1513 |
-
"grad_norm": 21.68485450744629,
|
| 1514 |
-
"learning_rate": 1.1648745519713262e-05,
|
| 1515 |
-
"loss": 0.3534,
|
| 1516 |
-
"step": 1960
|
| 1517 |
-
},
|
| 1518 |
-
{
|
| 1519 |
-
"epoch": 15.791583166332666,
|
| 1520 |
-
"grad_norm": 21.4897518157959,
|
| 1521 |
-
"learning_rate": 1.14247311827957e-05,
|
| 1522 |
-
"loss": 0.3685,
|
| 1523 |
-
"step": 1970
|
| 1524 |
-
},
|
| 1525 |
-
{
|
| 1526 |
-
"epoch": 15.871743486973948,
|
| 1527 |
-
"grad_norm": 23.654247283935547,
|
| 1528 |
-
"learning_rate": 1.1200716845878136e-05,
|
| 1529 |
-
"loss": 0.4038,
|
| 1530 |
-
"step": 1980
|
| 1531 |
-
},
|
| 1532 |
-
{
|
| 1533 |
-
"epoch": 15.951903807615231,
|
| 1534 |
-
"grad_norm": 24.013349533081055,
|
| 1535 |
-
"learning_rate": 1.0976702508960573e-05,
|
| 1536 |
-
"loss": 0.3724,
|
| 1537 |
-
"step": 1990
|
| 1538 |
-
},
|
| 1539 |
-
{
|
| 1540 |
-
"epoch": 16.0,
|
| 1541 |
-
"eval_accuracy": 0.8714768883878241,
|
| 1542 |
-
"eval_loss": 0.42415910959243774,
|
| 1543 |
-
"eval_runtime": 16.5547,
|
| 1544 |
-
"eval_samples_per_second": 107.16,
|
| 1545 |
-
"eval_steps_per_second": 3.383,
|
| 1546 |
-
"step": 1996
|
| 1547 |
-
},
|
| 1548 |
-
{
|
| 1549 |
-
"epoch": 16.03206412825651,
|
| 1550 |
-
"grad_norm": 19.16984748840332,
|
| 1551 |
-
"learning_rate": 1.0752688172043012e-05,
|
| 1552 |
-
"loss": 0.3318,
|
| 1553 |
-
"step": 2000
|
| 1554 |
-
},
|
| 1555 |
-
{
|
| 1556 |
-
"epoch": 16.112224448897795,
|
| 1557 |
-
"grad_norm": 25.895992279052734,
|
| 1558 |
-
"learning_rate": 1.0528673835125449e-05,
|
| 1559 |
-
"loss": 0.3325,
|
| 1560 |
-
"step": 2010
|
| 1561 |
-
},
|
| 1562 |
-
{
|
| 1563 |
-
"epoch": 16.19238476953908,
|
| 1564 |
-
"grad_norm": 22.940765380859375,
|
| 1565 |
-
"learning_rate": 1.0304659498207886e-05,
|
| 1566 |
-
"loss": 0.3384,
|
| 1567 |
-
"step": 2020
|
| 1568 |
-
},
|
| 1569 |
-
{
|
| 1570 |
-
"epoch": 16.272545090180362,
|
| 1571 |
-
"grad_norm": 16.38847541809082,
|
| 1572 |
-
"learning_rate": 1.0080645161290323e-05,
|
| 1573 |
-
"loss": 0.3289,
|
| 1574 |
-
"step": 2030
|
| 1575 |
-
},
|
| 1576 |
-
{
|
| 1577 |
-
"epoch": 16.352705410821642,
|
| 1578 |
-
"grad_norm": 15.670344352722168,
|
| 1579 |
-
"learning_rate": 9.856630824372761e-06,
|
| 1580 |
-
"loss": 0.3289,
|
| 1581 |
-
"step": 2040
|
| 1582 |
-
},
|
| 1583 |
-
{
|
| 1584 |
-
"epoch": 16.432865731462925,
|
| 1585 |
-
"grad_norm": 16.80422019958496,
|
| 1586 |
-
"learning_rate": 9.632616487455196e-06,
|
| 1587 |
-
"loss": 0.3387,
|
| 1588 |
-
"step": 2050
|
| 1589 |
-
},
|
| 1590 |
-
{
|
| 1591 |
-
"epoch": 16.51302605210421,
|
| 1592 |
-
"grad_norm": 17.764860153198242,
|
| 1593 |
-
"learning_rate": 9.408602150537635e-06,
|
| 1594 |
-
"loss": 0.3979,
|
| 1595 |
-
"step": 2060
|
| 1596 |
-
},
|
| 1597 |
-
{
|
| 1598 |
-
"epoch": 16.593186372745492,
|
| 1599 |
-
"grad_norm": 18.336956024169922,
|
| 1600 |
-
"learning_rate": 9.184587813620072e-06,
|
| 1601 |
-
"loss": 0.4002,
|
| 1602 |
-
"step": 2070
|
| 1603 |
-
},
|
| 1604 |
-
{
|
| 1605 |
-
"epoch": 16.673346693386772,
|
| 1606 |
-
"grad_norm": 16.667768478393555,
|
| 1607 |
-
"learning_rate": 8.960573476702509e-06,
|
| 1608 |
-
"loss": 0.3401,
|
| 1609 |
-
"step": 2080
|
| 1610 |
-
},
|
| 1611 |
-
{
|
| 1612 |
-
"epoch": 16.753507014028056,
|
| 1613 |
-
"grad_norm": 15.367273330688477,
|
| 1614 |
-
"learning_rate": 8.736559139784948e-06,
|
| 1615 |
-
"loss": 0.4036,
|
| 1616 |
-
"step": 2090
|
| 1617 |
-
},
|
| 1618 |
-
{
|
| 1619 |
-
"epoch": 16.83366733466934,
|
| 1620 |
-
"grad_norm": 22.505924224853516,
|
| 1621 |
-
"learning_rate": 8.512544802867385e-06,
|
| 1622 |
-
"loss": 0.3287,
|
| 1623 |
-
"step": 2100
|
| 1624 |
-
},
|
| 1625 |
-
{
|
| 1626 |
-
"epoch": 16.91382765531062,
|
| 1627 |
-
"grad_norm": 17.039518356323242,
|
| 1628 |
-
"learning_rate": 8.288530465949821e-06,
|
| 1629 |
-
"loss": 0.3105,
|
| 1630 |
-
"step": 2110
|
| 1631 |
-
},
|
| 1632 |
-
{
|
| 1633 |
-
"epoch": 16.993987975951903,
|
| 1634 |
-
"grad_norm": 24.149869918823242,
|
| 1635 |
-
"learning_rate": 8.064516129032258e-06,
|
| 1636 |
-
"loss": 0.354,
|
| 1637 |
-
"step": 2120
|
| 1638 |
-
},
|
| 1639 |
-
{
|
| 1640 |
-
"epoch": 16.993987975951903,
|
| 1641 |
-
"eval_accuracy": 0.8528748590755355,
|
| 1642 |
-
"eval_loss": 0.45945772528648376,
|
| 1643 |
-
"eval_runtime": 16.5552,
|
| 1644 |
-
"eval_samples_per_second": 107.156,
|
| 1645 |
-
"eval_steps_per_second": 3.383,
|
| 1646 |
-
"step": 2120
|
| 1647 |
-
},
|
| 1648 |
-
{
|
| 1649 |
-
"epoch": 17.074148296593187,
|
| 1650 |
-
"grad_norm": 23.050357818603516,
|
| 1651 |
-
"learning_rate": 7.840501792114695e-06,
|
| 1652 |
-
"loss": 0.3518,
|
| 1653 |
-
"step": 2130
|
| 1654 |
-
},
|
| 1655 |
-
{
|
| 1656 |
-
"epoch": 17.15430861723447,
|
| 1657 |
-
"grad_norm": 17.574710845947266,
|
| 1658 |
-
"learning_rate": 7.616487455197132e-06,
|
| 1659 |
-
"loss": 0.3024,
|
| 1660 |
-
"step": 2140
|
| 1661 |
-
},
|
| 1662 |
-
{
|
| 1663 |
-
"epoch": 17.23446893787575,
|
| 1664 |
-
"grad_norm": 27.150959014892578,
|
| 1665 |
-
"learning_rate": 7.392473118279571e-06,
|
| 1666 |
-
"loss": 0.3482,
|
| 1667 |
-
"step": 2150
|
| 1668 |
-
},
|
| 1669 |
-
{
|
| 1670 |
-
"epoch": 17.314629258517034,
|
| 1671 |
-
"grad_norm": 25.671140670776367,
|
| 1672 |
-
"learning_rate": 7.168458781362007e-06,
|
| 1673 |
-
"loss": 0.3461,
|
| 1674 |
-
"step": 2160
|
| 1675 |
-
},
|
| 1676 |
-
{
|
| 1677 |
-
"epoch": 17.394789579158317,
|
| 1678 |
-
"grad_norm": 23.23431968688965,
|
| 1679 |
-
"learning_rate": 6.944444444444445e-06,
|
| 1680 |
-
"loss": 0.3502,
|
| 1681 |
-
"step": 2170
|
| 1682 |
-
},
|
| 1683 |
-
{
|
| 1684 |
-
"epoch": 17.4749498997996,
|
| 1685 |
-
"grad_norm": 25.251195907592773,
|
| 1686 |
-
"learning_rate": 6.720430107526882e-06,
|
| 1687 |
-
"loss": 0.3505,
|
| 1688 |
-
"step": 2180
|
| 1689 |
-
},
|
| 1690 |
-
{
|
| 1691 |
-
"epoch": 17.55511022044088,
|
| 1692 |
-
"grad_norm": 18.16839027404785,
|
| 1693 |
-
"learning_rate": 6.4964157706093195e-06,
|
| 1694 |
-
"loss": 0.3401,
|
| 1695 |
-
"step": 2190
|
| 1696 |
-
},
|
| 1697 |
-
{
|
| 1698 |
-
"epoch": 17.635270541082164,
|
| 1699 |
-
"grad_norm": 17.334238052368164,
|
| 1700 |
-
"learning_rate": 6.2724014336917564e-06,
|
| 1701 |
-
"loss": 0.3182,
|
| 1702 |
-
"step": 2200
|
| 1703 |
-
},
|
| 1704 |
-
{
|
| 1705 |
-
"epoch": 17.715430861723448,
|
| 1706 |
-
"grad_norm": 23.427167892456055,
|
| 1707 |
-
"learning_rate": 6.048387096774194e-06,
|
| 1708 |
-
"loss": 0.2838,
|
| 1709 |
-
"step": 2210
|
| 1710 |
-
},
|
| 1711 |
-
{
|
| 1712 |
-
"epoch": 17.79559118236473,
|
| 1713 |
-
"grad_norm": 25.35896110534668,
|
| 1714 |
-
"learning_rate": 5.824372759856631e-06,
|
| 1715 |
-
"loss": 0.31,
|
| 1716 |
-
"step": 2220
|
| 1717 |
-
},
|
| 1718 |
-
{
|
| 1719 |
-
"epoch": 17.87575150300601,
|
| 1720 |
-
"grad_norm": 22.105119705200195,
|
| 1721 |
-
"learning_rate": 5.600358422939068e-06,
|
| 1722 |
-
"loss": 0.3231,
|
| 1723 |
-
"step": 2230
|
| 1724 |
-
},
|
| 1725 |
-
{
|
| 1726 |
-
"epoch": 17.955911823647295,
|
| 1727 |
-
"grad_norm": 33.91689682006836,
|
| 1728 |
-
"learning_rate": 5.376344086021506e-06,
|
| 1729 |
-
"loss": 0.3304,
|
| 1730 |
-
"step": 2240
|
| 1731 |
-
},
|
| 1732 |
-
{
|
| 1733 |
-
"epoch": 17.995991983967937,
|
| 1734 |
-
"eval_accuracy": 0.8562570462232244,
|
| 1735 |
-
"eval_loss": 0.45429307222366333,
|
| 1736 |
-
"eval_runtime": 17.1301,
|
| 1737 |
-
"eval_samples_per_second": 103.561,
|
| 1738 |
-
"eval_steps_per_second": 3.269,
|
| 1739 |
-
"step": 2245
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 18.03607214428858,
|
| 1743 |
-
"grad_norm": 17.72748374938965,
|
| 1744 |
-
"learning_rate": 5.152329749103943e-06,
|
| 1745 |
-
"loss": 0.3273,
|
| 1746 |
-
"step": 2250
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 18.11623246492986,
|
| 1750 |
-
"grad_norm": 15.051362991333008,
|
| 1751 |
-
"learning_rate": 4.928315412186381e-06,
|
| 1752 |
-
"loss": 0.2583,
|
| 1753 |
-
"step": 2260
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 18.196392785571142,
|
| 1757 |
-
"grad_norm": 15.139473915100098,
|
| 1758 |
-
"learning_rate": 4.7043010752688175e-06,
|
| 1759 |
-
"loss": 0.2711,
|
| 1760 |
-
"step": 2270
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 18.276553106212425,
|
| 1764 |
-
"grad_norm": 23.201719284057617,
|
| 1765 |
-
"learning_rate": 4.4802867383512545e-06,
|
| 1766 |
-
"loss": 0.3015,
|
| 1767 |
-
"step": 2280
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 18.35671342685371,
|
| 1771 |
-
"grad_norm": 30.31717872619629,
|
| 1772 |
-
"learning_rate": 4.256272401433692e-06,
|
| 1773 |
-
"loss": 0.321,
|
| 1774 |
-
"step": 2290
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 18.43687374749499,
|
| 1778 |
-
"grad_norm": 21.005664825439453,
|
| 1779 |
-
"learning_rate": 4.032258064516129e-06,
|
| 1780 |
-
"loss": 0.2887,
|
| 1781 |
-
"step": 2300
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 18.517034068136272,
|
| 1785 |
-
"grad_norm": 14.847023963928223,
|
| 1786 |
-
"learning_rate": 3.808243727598566e-06,
|
| 1787 |
-
"loss": 0.2821,
|
| 1788 |
-
"step": 2310
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 18.597194388777556,
|
| 1792 |
-
"grad_norm": 20.269981384277344,
|
| 1793 |
-
"learning_rate": 3.5842293906810035e-06,
|
| 1794 |
-
"loss": 0.2938,
|
| 1795 |
-
"step": 2320
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 18.677354709418836,
|
| 1799 |
-
"grad_norm": 18.943777084350586,
|
| 1800 |
-
"learning_rate": 3.360215053763441e-06,
|
| 1801 |
-
"loss": 0.2936,
|
| 1802 |
-
"step": 2330
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 18.75751503006012,
|
| 1806 |
-
"grad_norm": 17.58114242553711,
|
| 1807 |
-
"learning_rate": 3.1362007168458782e-06,
|
| 1808 |
-
"loss": 0.2801,
|
| 1809 |
-
"step": 2340
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 18.837675350701403,
|
| 1813 |
-
"grad_norm": 18.223283767700195,
|
| 1814 |
-
"learning_rate": 2.9121863799283156e-06,
|
| 1815 |
-
"loss": 0.2792,
|
| 1816 |
-
"step": 2350
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 18.917835671342687,
|
| 1820 |
-
"grad_norm": 19.910411834716797,
|
| 1821 |
-
"learning_rate": 2.688172043010753e-06,
|
| 1822 |
-
"loss": 0.3133,
|
| 1823 |
-
"step": 2360
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 18.997995991983966,
|
| 1827 |
-
"grad_norm": 18.749902725219727,
|
| 1828 |
-
"learning_rate": 2.4641577060931903e-06,
|
| 1829 |
-
"loss": 0.2849,
|
| 1830 |
-
"step": 2370
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 18.997995991983966,
|
| 1834 |
-
"eval_accuracy": 0.8607666290868095,
|
| 1835 |
-
"eval_loss": 0.4698711931705475,
|
| 1836 |
-
"eval_runtime": 16.5801,
|
| 1837 |
-
"eval_samples_per_second": 106.996,
|
| 1838 |
-
"eval_steps_per_second": 3.378,
|
| 1839 |
-
"step": 2370
|
| 1840 |
-
},
|
| 1841 |
-
{
|
| 1842 |
-
"epoch": 19.07815631262525,
|
| 1843 |
-
"grad_norm": 16.722375869750977,
|
| 1844 |
-
"learning_rate": 2.2401433691756272e-06,
|
| 1845 |
-
"loss": 0.3041,
|
| 1846 |
-
"step": 2380
|
| 1847 |
-
},
|
| 1848 |
-
{
|
| 1849 |
-
"epoch": 19.158316633266534,
|
| 1850 |
-
"grad_norm": 23.04833221435547,
|
| 1851 |
-
"learning_rate": 2.0161290322580646e-06,
|
| 1852 |
-
"loss": 0.301,
|
| 1853 |
-
"step": 2390
|
| 1854 |
-
},
|
| 1855 |
-
{
|
| 1856 |
-
"epoch": 19.238476953907817,
|
| 1857 |
-
"grad_norm": 20.021759033203125,
|
| 1858 |
-
"learning_rate": 1.7921146953405017e-06,
|
| 1859 |
-
"loss": 0.2723,
|
| 1860 |
-
"step": 2400
|
| 1861 |
-
},
|
| 1862 |
-
{
|
| 1863 |
-
"epoch": 19.318637274549097,
|
| 1864 |
-
"grad_norm": 18.34398078918457,
|
| 1865 |
-
"learning_rate": 1.5681003584229391e-06,
|
| 1866 |
-
"loss": 0.2484,
|
| 1867 |
-
"step": 2410
|
| 1868 |
-
},
|
| 1869 |
-
{
|
| 1870 |
-
"epoch": 19.39879759519038,
|
| 1871 |
-
"grad_norm": 18.83694839477539,
|
| 1872 |
-
"learning_rate": 1.3440860215053765e-06,
|
| 1873 |
-
"loss": 0.3204,
|
| 1874 |
-
"step": 2420
|
| 1875 |
-
},
|
| 1876 |
-
{
|
| 1877 |
-
"epoch": 19.478957915831664,
|
| 1878 |
-
"grad_norm": 16.674007415771484,
|
| 1879 |
-
"learning_rate": 1.1200716845878136e-06,
|
| 1880 |
-
"loss": 0.3034,
|
| 1881 |
-
"step": 2430
|
| 1882 |
-
},
|
| 1883 |
-
{
|
| 1884 |
-
"epoch": 19.559118236472948,
|
| 1885 |
-
"grad_norm": 28.710783004760742,
|
| 1886 |
-
"learning_rate": 8.960573476702509e-07,
|
| 1887 |
-
"loss": 0.3078,
|
| 1888 |
-
"step": 2440
|
| 1889 |
-
},
|
| 1890 |
-
{
|
| 1891 |
-
"epoch": 19.639278557114228,
|
| 1892 |
-
"grad_norm": 18.92753791809082,
|
| 1893 |
-
"learning_rate": 6.720430107526882e-07,
|
| 1894 |
-
"loss": 0.2644,
|
| 1895 |
-
"step": 2450
|
| 1896 |
-
},
|
| 1897 |
-
{
|
| 1898 |
-
"epoch": 19.71943887775551,
|
| 1899 |
-
"grad_norm": 17.13638687133789,
|
| 1900 |
-
"learning_rate": 4.4802867383512544e-07,
|
| 1901 |
-
"loss": 0.2673,
|
| 1902 |
-
"step": 2460
|
| 1903 |
-
},
|
| 1904 |
-
{
|
| 1905 |
-
"epoch": 19.799599198396795,
|
| 1906 |
-
"grad_norm": 18.35292625427246,
|
| 1907 |
-
"learning_rate": 2.2401433691756272e-07,
|
| 1908 |
-
"loss": 0.2569,
|
| 1909 |
-
"step": 2470
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
-
"epoch":
|
| 1913 |
-
"
|
| 1914 |
-
"
|
| 1915 |
-
"
|
| 1916 |
-
"
|
| 1917 |
-
|
| 1918 |
-
|
| 1919 |
-
"epoch": 19.879759519038075,
|
| 1920 |
-
"eval_accuracy": 0.8664036076662909,
|
| 1921 |
-
"eval_loss": 0.4505213499069214,
|
| 1922 |
-
"eval_runtime": 16.6944,
|
| 1923 |
-
"eval_samples_per_second": 106.263,
|
| 1924 |
-
"eval_steps_per_second": 3.354,
|
| 1925 |
-
"step": 2480
|
| 1926 |
-
},
|
| 1927 |
-
{
|
| 1928 |
-
"epoch": 19.879759519038075,
|
| 1929 |
-
"step": 2480,
|
| 1930 |
-
"total_flos": 8.259382470828884e+18,
|
| 1931 |
-
"train_loss": 0.5719680822664691,
|
| 1932 |
-
"train_runtime": 7149.119,
|
| 1933 |
-
"train_samples_per_second": 44.654,
|
| 1934 |
-
"train_steps_per_second": 0.347
|
| 1935 |
}
|
| 1936 |
],
|
| 1937 |
"logging_steps": 10,
|
| 1938 |
-
"max_steps":
|
| 1939 |
"num_input_tokens_seen": 0,
|
| 1940 |
-
"num_train_epochs":
|
| 1941 |
"save_steps": 500,
|
| 1942 |
-
"total_flos":
|
| 1943 |
-
"train_batch_size":
|
| 1944 |
"trial_name": null,
|
| 1945 |
"trial_params": null
|
| 1946 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8562570462232244,
|
| 3 |
+
"best_model_checkpoint": "dinov2-small-imagenet1k-1-layer-finetuned-galaxy10-decals/checkpoint-1750",
|
| 4 |
+
"epoch": 29.76,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1860,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.16,
|
| 13 |
+
"learning_rate": 2.688172043010753e-06,
|
| 14 |
+
"loss": 2.558,
|
|
|
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.32,
|
| 19 |
+
"learning_rate": 5.376344086021506e-06,
|
| 20 |
+
"loss": 2.0637,
|
|
|
|
| 21 |
"step": 20
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.48,
|
| 25 |
+
"learning_rate": 8.064516129032258e-06,
|
| 26 |
+
"loss": 1.6881,
|
|
|
|
| 27 |
"step": 30
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.64,
|
| 31 |
+
"learning_rate": 1.0752688172043012e-05,
|
| 32 |
+
"loss": 1.4019,
|
|
|
|
| 33 |
"step": 40
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.8,
|
| 37 |
+
"learning_rate": 1.3440860215053763e-05,
|
| 38 |
+
"loss": 1.2171,
|
|
|
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.96,
|
| 43 |
+
"learning_rate": 1.6129032258064517e-05,
|
| 44 |
+
"loss": 1.0027,
|
|
|
|
| 45 |
"step": 60
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.99,
|
| 49 |
+
"eval_accuracy": 0.7080045095828635,
|
| 50 |
+
"eval_f1": 0.7016603472901116,
|
| 51 |
+
"eval_loss": 0.8262113928794861,
|
| 52 |
+
"eval_precision": 0.7230853006708634,
|
| 53 |
+
"eval_recall": 0.7080045095828635,
|
| 54 |
+
"eval_runtime": 10.7016,
|
| 55 |
+
"eval_samples_per_second": 165.77,
|
| 56 |
+
"eval_steps_per_second": 2.616,
|
| 57 |
+
"step": 62
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.12,
|
| 61 |
+
"learning_rate": 1.881720430107527e-05,
|
| 62 |
+
"loss": 0.9389,
|
| 63 |
"step": 70
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 1.28,
|
| 67 |
+
"learning_rate": 2.1505376344086024e-05,
|
| 68 |
+
"loss": 0.9166,
|
|
|
|
| 69 |
"step": 80
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 1.44,
|
| 73 |
+
"learning_rate": 2.4193548387096777e-05,
|
| 74 |
+
"loss": 0.8792,
|
|
|
|
| 75 |
"step": 90
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 1.6,
|
| 79 |
+
"learning_rate": 2.6881720430107527e-05,
|
| 80 |
+
"loss": 0.8329,
|
|
|
|
| 81 |
"step": 100
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.76,
|
| 85 |
+
"learning_rate": 2.9569892473118284e-05,
|
| 86 |
+
"loss": 0.7826,
|
|
|
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 1.92,
|
| 91 |
+
"learning_rate": 3.2258064516129034e-05,
|
| 92 |
+
"loss": 0.8374,
|
|
|
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 2.0,
|
| 97 |
+
"eval_accuracy": 0.7948139797068771,
|
| 98 |
+
"eval_f1": 0.789920738225672,
|
| 99 |
+
"eval_loss": 0.612916111946106,
|
| 100 |
+
"eval_precision": 0.7960304264137663,
|
| 101 |
+
"eval_recall": 0.7948139797068771,
|
| 102 |
+
"eval_runtime": 9.6957,
|
| 103 |
+
"eval_samples_per_second": 182.968,
|
| 104 |
+
"eval_steps_per_second": 2.888,
|
| 105 |
+
"step": 125
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 2.08,
|
| 109 |
+
"learning_rate": 3.494623655913979e-05,
|
| 110 |
+
"loss": 0.804,
|
|
|
|
| 111 |
"step": 130
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 2.24,
|
| 115 |
+
"learning_rate": 3.763440860215054e-05,
|
| 116 |
+
"loss": 0.8107,
|
|
|
|
| 117 |
"step": 140
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 2.4,
|
| 121 |
+
"learning_rate": 4.032258064516129e-05,
|
| 122 |
+
"loss": 0.791,
|
|
|
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 2.56,
|
| 127 |
+
"learning_rate": 4.301075268817205e-05,
|
| 128 |
+
"loss": 0.7538,
|
|
|
|
| 129 |
"step": 160
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 2.72,
|
| 133 |
+
"learning_rate": 4.56989247311828e-05,
|
| 134 |
+
"loss": 0.7667,
|
|
|
|
| 135 |
"step": 170
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 2.88,
|
| 139 |
+
"learning_rate": 4.8387096774193554e-05,
|
| 140 |
+
"loss": 0.7753,
|
|
|
|
| 141 |
"step": 180
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 2.99,
|
| 145 |
+
"eval_accuracy": 0.7891770011273957,
|
| 146 |
+
"eval_f1": 0.7786862536985929,
|
| 147 |
+
"eval_loss": 0.6555072665214539,
|
| 148 |
+
"eval_precision": 0.792054417969131,
|
| 149 |
+
"eval_recall": 0.7891770011273957,
|
| 150 |
+
"eval_runtime": 9.4242,
|
| 151 |
+
"eval_samples_per_second": 188.238,
|
| 152 |
+
"eval_steps_per_second": 2.971,
|
| 153 |
+
"step": 187
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 3.04,
|
| 157 |
+
"learning_rate": 4.98805256869773e-05,
|
| 158 |
+
"loss": 0.8543,
|
| 159 |
"step": 190
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 3.2,
|
| 163 |
+
"learning_rate": 4.9581839904420555e-05,
|
| 164 |
+
"loss": 0.7868,
|
|
|
|
| 165 |
"step": 200
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 3.36,
|
| 169 |
+
"learning_rate": 4.92831541218638e-05,
|
| 170 |
+
"loss": 0.7524,
|
|
|
|
| 171 |
"step": 210
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 3.52,
|
| 175 |
+
"learning_rate": 4.898446833930705e-05,
|
| 176 |
+
"loss": 0.803,
|
|
|
|
| 177 |
"step": 220
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 3.68,
|
| 181 |
+
"learning_rate": 4.86857825567503e-05,
|
| 182 |
+
"loss": 0.7192,
|
|
|
|
| 183 |
"step": 230
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 3.84,
|
|
|
|
| 187 |
"learning_rate": 4.8387096774193554e-05,
|
| 188 |
+
"loss": 0.7813,
|
| 189 |
"step": 240
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 4.0,
|
| 193 |
+
"learning_rate": 4.80884109916368e-05,
|
| 194 |
+
"loss": 0.7165,
|
| 195 |
+
"step": 250
|
|
|
|
|
|
|
|
|
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 4.0,
|
| 199 |
+
"eval_accuracy": 0.8004509582863585,
|
| 200 |
+
"eval_f1": 0.7970428033398866,
|
| 201 |
+
"eval_loss": 0.5862058401107788,
|
| 202 |
+
"eval_precision": 0.8052638185635553,
|
| 203 |
+
"eval_recall": 0.8004509582863585,
|
| 204 |
+
"eval_runtime": 9.4111,
|
| 205 |
+
"eval_samples_per_second": 188.5,
|
| 206 |
+
"eval_steps_per_second": 2.975,
|
| 207 |
"step": 250
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 4.16,
|
| 211 |
+
"learning_rate": 4.778972520908005e-05,
|
| 212 |
+
"loss": 0.728,
|
|
|
|
| 213 |
"step": 260
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 4.32,
|
| 217 |
+
"learning_rate": 4.74910394265233e-05,
|
| 218 |
+
"loss": 0.7121,
|
|
|
|
| 219 |
"step": 270
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 4.48,
|
| 223 |
+
"learning_rate": 4.7192353643966546e-05,
|
| 224 |
+
"loss": 0.6889,
|
|
|
|
| 225 |
"step": 280
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 4.64,
|
| 229 |
+
"learning_rate": 4.6893667861409805e-05,
|
| 230 |
+
"loss": 0.6748,
|
|
|
|
| 231 |
"step": 290
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 4.8,
|
| 235 |
+
"learning_rate": 4.659498207885305e-05,
|
| 236 |
+
"loss": 0.7071,
|
|
|
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 4.96,
|
| 241 |
+
"learning_rate": 4.62962962962963e-05,
|
| 242 |
+
"loss": 0.6477,
|
|
|
|
| 243 |
"step": 310
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 4.99,
|
| 247 |
+
"eval_accuracy": 0.7965050732807215,
|
| 248 |
+
"eval_f1": 0.7985277928131062,
|
| 249 |
+
"eval_loss": 0.6183481216430664,
|
| 250 |
+
"eval_precision": 0.8118777372534622,
|
| 251 |
+
"eval_recall": 0.7965050732807215,
|
| 252 |
+
"eval_runtime": 9.6246,
|
| 253 |
+
"eval_samples_per_second": 184.32,
|
| 254 |
+
"eval_steps_per_second": 2.909,
|
| 255 |
+
"step": 312
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 5.12,
|
| 259 |
+
"learning_rate": 4.5997610513739546e-05,
|
| 260 |
+
"loss": 0.6682,
|
| 261 |
"step": 320
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 5.28,
|
| 265 |
+
"learning_rate": 4.56989247311828e-05,
|
| 266 |
+
"loss": 0.6595,
|
|
|
|
| 267 |
"step": 330
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 5.44,
|
| 271 |
+
"learning_rate": 4.540023894862604e-05,
|
| 272 |
+
"loss": 0.6427,
|
|
|
|
| 273 |
"step": 340
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 5.6,
|
| 277 |
+
"learning_rate": 4.51015531660693e-05,
|
| 278 |
+
"loss": 0.6044,
|
|
|
|
| 279 |
"step": 350
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 5.76,
|
| 283 |
+
"learning_rate": 4.4802867383512545e-05,
|
| 284 |
+
"loss": 0.6392,
|
|
|
|
| 285 |
"step": 360
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 5.92,
|
| 289 |
+
"learning_rate": 4.4504181600955796e-05,
|
| 290 |
+
"loss": 0.6892,
|
|
|
|
| 291 |
"step": 370
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 6.0,
|
| 295 |
+
"eval_accuracy": 0.8246899661781285,
|
| 296 |
+
"eval_f1": 0.8194968508495551,
|
| 297 |
+
"eval_loss": 0.530994176864624,
|
| 298 |
+
"eval_precision": 0.8274763562734486,
|
| 299 |
+
"eval_recall": 0.8246899661781285,
|
| 300 |
+
"eval_runtime": 9.4059,
|
| 301 |
+
"eval_samples_per_second": 188.605,
|
| 302 |
+
"eval_steps_per_second": 2.977,
|
| 303 |
+
"step": 375
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 6.08,
|
| 307 |
+
"learning_rate": 4.420549581839905e-05,
|
| 308 |
+
"loss": 0.6048,
|
|
|
|
| 309 |
"step": 380
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 6.24,
|
| 313 |
+
"learning_rate": 4.390681003584229e-05,
|
| 314 |
+
"loss": 0.6048,
|
|
|
|
| 315 |
"step": 390
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 6.4,
|
| 319 |
+
"learning_rate": 4.360812425328555e-05,
|
| 320 |
+
"loss": 0.5979,
|
|
|
|
| 321 |
"step": 400
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 6.56,
|
| 325 |
+
"learning_rate": 4.3309438470728796e-05,
|
| 326 |
+
"loss": 0.599,
|
|
|
|
| 327 |
"step": 410
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 6.72,
|
| 331 |
+
"learning_rate": 4.301075268817205e-05,
|
| 332 |
+
"loss": 0.5938,
|
|
|
|
| 333 |
"step": 420
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 6.88,
|
| 337 |
+
"learning_rate": 4.271206690561529e-05,
|
| 338 |
+
"loss": 0.6171,
|
|
|
|
| 339 |
"step": 430
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 6.99,
|
| 343 |
+
"eval_accuracy": 0.8083427282976324,
|
| 344 |
+
"eval_f1": 0.8021700130993371,
|
| 345 |
+
"eval_loss": 0.5677684545516968,
|
| 346 |
+
"eval_precision": 0.8157285448675808,
|
| 347 |
+
"eval_recall": 0.8083427282976324,
|
| 348 |
+
"eval_runtime": 9.4301,
|
| 349 |
+
"eval_samples_per_second": 188.121,
|
| 350 |
+
"eval_steps_per_second": 2.969,
|
| 351 |
+
"step": 437
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 7.04,
|
| 355 |
+
"learning_rate": 4.241338112305854e-05,
|
| 356 |
+
"loss": 0.6715,
|
| 357 |
"step": 440
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 7.2,
|
| 361 |
+
"learning_rate": 4.2114695340501795e-05,
|
| 362 |
+
"loss": 0.6297,
|
|
|
|
| 363 |
"step": 450
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 7.36,
|
| 367 |
+
"learning_rate": 4.1816009557945046e-05,
|
| 368 |
+
"loss": 0.5852,
|
|
|
|
| 369 |
"step": 460
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 7.52,
|
| 373 |
+
"learning_rate": 4.15173237753883e-05,
|
| 374 |
+
"loss": 0.6125,
|
|
|
|
| 375 |
"step": 470
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 7.68,
|
| 379 |
+
"learning_rate": 4.121863799283154e-05,
|
| 380 |
+
"loss": 0.5598,
|
|
|
|
| 381 |
"step": 480
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 7.84,
|
| 385 |
+
"learning_rate": 4.0919952210274794e-05,
|
| 386 |
+
"loss": 0.5845,
|
|
|
|
| 387 |
"step": 490
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 8.0,
|
| 391 |
+
"learning_rate": 4.062126642771804e-05,
|
| 392 |
+
"loss": 0.55,
|
| 393 |
+
"step": 500
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 8.0,
|
| 397 |
+
"eval_accuracy": 0.8325817361894025,
|
| 398 |
+
"eval_f1": 0.8316011722287234,
|
| 399 |
+
"eval_loss": 0.49609655141830444,
|
| 400 |
+
"eval_precision": 0.8352811642513641,
|
| 401 |
+
"eval_recall": 0.8325817361894025,
|
| 402 |
+
"eval_runtime": 9.446,
|
| 403 |
+
"eval_samples_per_second": 187.804,
|
| 404 |
+
"eval_steps_per_second": 2.964,
|
| 405 |
"step": 500
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 8.16,
|
| 409 |
+
"learning_rate": 4.032258064516129e-05,
|
| 410 |
+
"loss": 0.5259,
|
|
|
|
| 411 |
"step": 510
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 8.32,
|
| 415 |
+
"learning_rate": 4.002389486260454e-05,
|
| 416 |
+
"loss": 0.5585,
|
|
|
|
| 417 |
"step": 520
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 8.48,
|
| 421 |
+
"learning_rate": 3.972520908004779e-05,
|
| 422 |
+
"loss": 0.5927,
|
|
|
|
| 423 |
"step": 530
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 8.64,
|
| 427 |
+
"learning_rate": 3.9426523297491045e-05,
|
| 428 |
+
"loss": 0.5585,
|
|
|
|
| 429 |
"step": 540
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 8.8,
|
| 433 |
+
"learning_rate": 3.912783751493429e-05,
|
| 434 |
+
"loss": 0.5732,
|
|
|
|
| 435 |
"step": 550
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 8.96,
|
| 439 |
+
"learning_rate": 3.882915173237754e-05,
|
| 440 |
+
"loss": 0.5615,
|
|
|
|
| 441 |
"step": 560
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 8.99,
|
| 445 |
+
"eval_accuracy": 0.830890642615558,
|
| 446 |
+
"eval_f1": 0.8274241002805047,
|
| 447 |
+
"eval_loss": 0.5032677054405212,
|
| 448 |
+
"eval_precision": 0.831185175749917,
|
| 449 |
+
"eval_recall": 0.830890642615558,
|
| 450 |
+
"eval_runtime": 9.45,
|
| 451 |
+
"eval_samples_per_second": 187.725,
|
| 452 |
+
"eval_steps_per_second": 2.963,
|
| 453 |
+
"step": 562
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 9.12,
|
| 457 |
+
"learning_rate": 3.8530465949820786e-05,
|
| 458 |
+
"loss": 0.5505,
|
| 459 |
"step": 570
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 9.28,
|
| 463 |
+
"learning_rate": 3.8231780167264044e-05,
|
| 464 |
+
"loss": 0.5101,
|
|
|
|
| 465 |
"step": 580
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 9.44,
|
| 469 |
+
"learning_rate": 3.793309438470729e-05,
|
| 470 |
+
"loss": 0.5215,
|
|
|
|
| 471 |
"step": 590
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 9.6,
|
| 475 |
+
"learning_rate": 3.763440860215054e-05,
|
| 476 |
+
"loss": 0.5503,
|
|
|
|
| 477 |
"step": 600
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 9.76,
|
| 481 |
+
"learning_rate": 3.7335722819593785e-05,
|
| 482 |
+
"loss": 0.5045,
|
|
|
|
| 483 |
"step": 610
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 9.92,
|
| 487 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 488 |
+
"loss": 0.5107,
|
|
|
|
| 489 |
"step": 620
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 10.0,
|
| 493 |
+
"eval_accuracy": 0.8190529875986471,
|
| 494 |
+
"eval_f1": 0.8152076243657725,
|
| 495 |
+
"eval_loss": 0.5162058472633362,
|
| 496 |
+
"eval_precision": 0.8163659955266428,
|
| 497 |
+
"eval_recall": 0.8190529875986471,
|
| 498 |
+
"eval_runtime": 9.4412,
|
| 499 |
+
"eval_samples_per_second": 187.9,
|
| 500 |
+
"eval_steps_per_second": 2.966,
|
| 501 |
+
"step": 625
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 10.08,
|
| 505 |
+
"learning_rate": 3.673835125448029e-05,
|
| 506 |
+
"loss": 0.5196,
|
|
|
|
| 507 |
"step": 630
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 10.24,
|
| 511 |
+
"learning_rate": 3.643966547192354e-05,
|
| 512 |
+
"loss": 0.5125,
|
|
|
|
| 513 |
"step": 640
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 10.4,
|
| 517 |
+
"learning_rate": 3.614097968936679e-05,
|
| 518 |
+
"loss": 0.5879,
|
|
|
|
| 519 |
"step": 650
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 10.56,
|
| 523 |
+
"learning_rate": 3.5842293906810036e-05,
|
| 524 |
+
"loss": 0.5195,
|
|
|
|
| 525 |
"step": 660
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 10.72,
|
| 529 |
+
"learning_rate": 3.554360812425329e-05,
|
| 530 |
+
"loss": 0.4909,
|
|
|
|
| 531 |
"step": 670
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 10.88,
|
| 535 |
+
"learning_rate": 3.524492234169653e-05,
|
| 536 |
+
"loss": 0.5237,
|
|
|
|
| 537 |
"step": 680
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 10.99,
|
| 541 |
+
"eval_accuracy": 0.8421645997745209,
|
| 542 |
+
"eval_f1": 0.8381375195281238,
|
| 543 |
+
"eval_loss": 0.4789753556251526,
|
| 544 |
+
"eval_precision": 0.8451751547830025,
|
| 545 |
+
"eval_recall": 0.8421645997745209,
|
| 546 |
+
"eval_runtime": 9.4165,
|
| 547 |
+
"eval_samples_per_second": 188.392,
|
| 548 |
+
"eval_steps_per_second": 2.973,
|
| 549 |
+
"step": 687
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 11.04,
|
| 553 |
+
"learning_rate": 3.494623655913979e-05,
|
| 554 |
+
"loss": 0.4974,
|
| 555 |
"step": 690
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 11.2,
|
| 559 |
+
"learning_rate": 3.4647550776583035e-05,
|
| 560 |
+
"loss": 0.4889,
|
|
|
|
| 561 |
"step": 700
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 11.36,
|
| 565 |
+
"learning_rate": 3.4348864994026287e-05,
|
| 566 |
+
"loss": 0.4849,
|
|
|
|
| 567 |
"step": 710
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 11.52,
|
| 571 |
+
"learning_rate": 3.405017921146954e-05,
|
| 572 |
+
"loss": 0.4964,
|
|
|
|
| 573 |
"step": 720
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 11.68,
|
| 577 |
+
"learning_rate": 3.375149342891278e-05,
|
| 578 |
+
"loss": 0.5049,
|
|
|
|
| 579 |
"step": 730
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 11.84,
|
| 583 |
+
"learning_rate": 3.3452807646356034e-05,
|
| 584 |
+
"loss": 0.4907,
|
|
|
|
| 585 |
"step": 740
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 12.0,
|
| 589 |
+
"learning_rate": 3.3154121863799286e-05,
|
| 590 |
+
"loss": 0.4954,
|
| 591 |
+
"step": 750
|
|
|
|
|
|
|
|
|
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 12.0,
|
| 595 |
+
"eval_accuracy": 0.8421645997745209,
|
| 596 |
+
"eval_f1": 0.8372920175463381,
|
| 597 |
+
"eval_loss": 0.4781884253025055,
|
| 598 |
+
"eval_precision": 0.8430155601755086,
|
| 599 |
+
"eval_recall": 0.8421645997745209,
|
| 600 |
+
"eval_runtime": 9.4751,
|
| 601 |
+
"eval_samples_per_second": 187.227,
|
| 602 |
+
"eval_steps_per_second": 2.955,
|
| 603 |
"step": 750
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 12.16,
|
| 607 |
+
"learning_rate": 3.285543608124254e-05,
|
| 608 |
+
"loss": 0.4693,
|
|
|
|
| 609 |
"step": 760
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 12.32,
|
| 613 |
+
"learning_rate": 3.255675029868578e-05,
|
| 614 |
+
"loss": 0.4387,
|
|
|
|
| 615 |
"step": 770
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 12.48,
|
| 619 |
+
"learning_rate": 3.2258064516129034e-05,
|
| 620 |
+
"loss": 0.4934,
|
|
|
|
| 621 |
"step": 780
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 12.64,
|
| 625 |
+
"learning_rate": 3.195937873357228e-05,
|
| 626 |
+
"loss": 0.471,
|
|
|
|
| 627 |
"step": 790
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 12.8,
|
| 631 |
+
"learning_rate": 3.1660692951015537e-05,
|
| 632 |
+
"loss": 0.4691,
|
|
|
|
| 633 |
"step": 800
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 12.96,
|
| 637 |
+
"learning_rate": 3.136200716845878e-05,
|
| 638 |
+
"loss": 0.4887,
|
|
|
|
| 639 |
"step": 810
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 12.99,
|
| 643 |
+
"eval_accuracy": 0.8370913190529876,
|
| 644 |
+
"eval_f1": 0.8358084111631787,
|
| 645 |
+
"eval_loss": 0.46888935565948486,
|
| 646 |
+
"eval_precision": 0.8394938653593303,
|
| 647 |
+
"eval_recall": 0.8370913190529876,
|
| 648 |
+
"eval_runtime": 9.4084,
|
| 649 |
+
"eval_samples_per_second": 188.555,
|
| 650 |
+
"eval_steps_per_second": 2.976,
|
| 651 |
+
"step": 812
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 13.12,
|
| 655 |
+
"learning_rate": 3.106332138590203e-05,
|
| 656 |
+
"loss": 0.4613,
|
| 657 |
"step": 820
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 13.28,
|
| 661 |
+
"learning_rate": 3.0764635603345284e-05,
|
| 662 |
+
"loss": 0.4439,
|
|
|
|
| 663 |
"step": 830
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 13.44,
|
| 667 |
+
"learning_rate": 3.046594982078853e-05,
|
| 668 |
+
"loss": 0.5005,
|
|
|
|
| 669 |
"step": 840
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 13.6,
|
| 673 |
+
"learning_rate": 3.016726403823178e-05,
|
| 674 |
+
"loss": 0.4262,
|
|
|
|
| 675 |
"step": 850
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 13.76,
|
| 679 |
+
"learning_rate": 2.9868578255675032e-05,
|
| 680 |
+
"loss": 0.4866,
|
|
|
|
| 681 |
"step": 860
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 13.92,
|
| 685 |
+
"learning_rate": 2.9569892473118284e-05,
|
| 686 |
+
"loss": 0.4629,
|
|
|
|
| 687 |
"step": 870
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 14.0,
|
| 691 |
+
"eval_accuracy": 0.8523111612175873,
|
| 692 |
+
"eval_f1": 0.8502117923643765,
|
| 693 |
+
"eval_loss": 0.45412173867225647,
|
| 694 |
+
"eval_precision": 0.8500083028083808,
|
| 695 |
+
"eval_recall": 0.8523111612175873,
|
| 696 |
+
"eval_runtime": 9.4205,
|
| 697 |
+
"eval_samples_per_second": 188.313,
|
| 698 |
+
"eval_steps_per_second": 2.972,
|
| 699 |
+
"step": 875
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 14.08,
|
| 703 |
+
"learning_rate": 2.9271206690561532e-05,
|
| 704 |
+
"loss": 0.4421,
|
|
|
|
| 705 |
"step": 880
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 14.24,
|
| 709 |
+
"learning_rate": 2.897252090800478e-05,
|
| 710 |
+
"loss": 0.3844,
|
|
|
|
| 711 |
"step": 890
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 14.4,
|
| 715 |
+
"learning_rate": 2.8673835125448028e-05,
|
| 716 |
+
"loss": 0.4462,
|
|
|
|
| 717 |
"step": 900
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 14.56,
|
| 721 |
+
"learning_rate": 2.8375149342891276e-05,
|
| 722 |
+
"loss": 0.4703,
|
|
|
|
| 723 |
"step": 910
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 14.72,
|
| 727 |
+
"learning_rate": 2.807646356033453e-05,
|
| 728 |
+
"loss": 0.4755,
|
|
|
|
| 729 |
"step": 920
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 14.88,
|
| 733 |
+
"learning_rate": 2.777777777777778e-05,
|
| 734 |
+
"loss": 0.4486,
|
|
|
|
| 735 |
"step": 930
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 14.99,
|
| 739 |
+
"eval_accuracy": 0.8404735062006764,
|
| 740 |
+
"eval_f1": 0.8394262955931711,
|
| 741 |
+
"eval_loss": 0.4754864275455475,
|
| 742 |
+
"eval_precision": 0.8400472182964517,
|
| 743 |
+
"eval_recall": 0.8404735062006764,
|
| 744 |
+
"eval_runtime": 9.6429,
|
| 745 |
+
"eval_samples_per_second": 183.97,
|
| 746 |
+
"eval_steps_per_second": 2.904,
|
| 747 |
+
"step": 937
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 15.04,
|
| 751 |
+
"learning_rate": 2.747909199522103e-05,
|
| 752 |
+
"loss": 0.4446,
|
| 753 |
"step": 940
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 15.2,
|
| 757 |
+
"learning_rate": 2.718040621266428e-05,
|
| 758 |
+
"loss": 0.4243,
|
|
|
|
| 759 |
"step": 950
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 15.36,
|
| 763 |
+
"learning_rate": 2.6881720430107527e-05,
|
| 764 |
+
"loss": 0.4136,
|
|
|
|
| 765 |
"step": 960
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 15.52,
|
| 769 |
+
"learning_rate": 2.6583034647550775e-05,
|
| 770 |
+
"loss": 0.4471,
|
|
|
|
| 771 |
"step": 970
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 15.68,
|
| 775 |
+
"learning_rate": 2.628434886499403e-05,
|
| 776 |
+
"loss": 0.3961,
|
|
|
|
| 777 |
"step": 980
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 15.84,
|
| 781 |
+
"learning_rate": 2.5985663082437278e-05,
|
| 782 |
+
"loss": 0.4334,
|
|
|
|
| 783 |
"step": 990
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 16.0,
|
| 787 |
+
"learning_rate": 2.5686977299880526e-05,
|
| 788 |
+
"loss": 0.4361,
|
| 789 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 16.0,
|
| 793 |
+
"eval_accuracy": 0.8370913190529876,
|
| 794 |
+
"eval_f1": 0.8369915773538048,
|
| 795 |
+
"eval_loss": 0.47626104950904846,
|
| 796 |
+
"eval_precision": 0.8391572358322159,
|
| 797 |
+
"eval_recall": 0.8370913190529876,
|
| 798 |
+
"eval_runtime": 9.6357,
|
| 799 |
+
"eval_samples_per_second": 184.107,
|
| 800 |
+
"eval_steps_per_second": 2.906,
|
| 801 |
"step": 1000
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 16.16,
|
| 805 |
+
"learning_rate": 2.5388291517323774e-05,
|
| 806 |
+
"loss": 0.4227,
|
|
|
|
| 807 |
"step": 1010
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 16.32,
|
| 811 |
+
"learning_rate": 2.5089605734767026e-05,
|
| 812 |
+
"loss": 0.3709,
|
|
|
|
| 813 |
"step": 1020
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 16.48,
|
| 817 |
+
"learning_rate": 2.4790919952210277e-05,
|
| 818 |
+
"loss": 0.3767,
|
|
|
|
| 819 |
"step": 1030
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 16.64,
|
| 823 |
+
"learning_rate": 2.4492234169653525e-05,
|
| 824 |
+
"loss": 0.4753,
|
|
|
|
| 825 |
"step": 1040
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 16.8,
|
| 829 |
+
"learning_rate": 2.4193548387096777e-05,
|
| 830 |
+
"loss": 0.41,
|
|
|
|
| 831 |
"step": 1050
|
| 832 |
},
|
| 833 |
{
|
| 834 |
+
"epoch": 16.96,
|
| 835 |
+
"learning_rate": 2.3894862604540025e-05,
|
| 836 |
+
"loss": 0.3833,
|
|
|
|
| 837 |
"step": 1060
|
| 838 |
},
|
| 839 |
{
|
| 840 |
+
"epoch": 16.99,
|
| 841 |
+
"eval_accuracy": 0.8416009019165727,
|
| 842 |
+
"eval_f1": 0.8396254917954661,
|
| 843 |
+
"eval_loss": 0.4982186257839203,
|
| 844 |
+
"eval_precision": 0.8428664297953069,
|
| 845 |
+
"eval_recall": 0.8416009019165727,
|
| 846 |
+
"eval_runtime": 9.4953,
|
| 847 |
+
"eval_samples_per_second": 186.829,
|
| 848 |
+
"eval_steps_per_second": 2.949,
|
| 849 |
+
"step": 1062
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 17.12,
|
| 853 |
+
"learning_rate": 2.3596176821983273e-05,
|
| 854 |
+
"loss": 0.4147,
|
| 855 |
"step": 1070
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 17.28,
|
| 859 |
+
"learning_rate": 2.3297491039426525e-05,
|
| 860 |
+
"loss": 0.411,
|
|
|
|
| 861 |
"step": 1080
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 17.44,
|
| 865 |
+
"learning_rate": 2.2998805256869773e-05,
|
| 866 |
+
"loss": 0.4258,
|
|
|
|
| 867 |
"step": 1090
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 17.6,
|
| 871 |
+
"learning_rate": 2.270011947431302e-05,
|
| 872 |
+
"loss": 0.4165,
|
|
|
|
| 873 |
"step": 1100
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 17.76,
|
| 877 |
+
"learning_rate": 2.2401433691756272e-05,
|
| 878 |
+
"loss": 0.3852,
|
|
|
|
| 879 |
"step": 1110
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 17.92,
|
| 883 |
+
"learning_rate": 2.2102747909199524e-05,
|
| 884 |
+
"loss": 0.3788,
|
|
|
|
| 885 |
"step": 1120
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 18.0,
|
| 889 |
+
"eval_accuracy": 0.8291995490417137,
|
| 890 |
+
"eval_f1": 0.826730633772571,
|
| 891 |
+
"eval_loss": 0.5631851553916931,
|
| 892 |
+
"eval_precision": 0.836503982215401,
|
| 893 |
+
"eval_recall": 0.8291995490417137,
|
| 894 |
+
"eval_runtime": 9.4,
|
| 895 |
+
"eval_samples_per_second": 188.724,
|
| 896 |
+
"eval_steps_per_second": 2.979,
|
| 897 |
+
"step": 1125
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 18.08,
|
| 901 |
+
"learning_rate": 2.1804062126642775e-05,
|
| 902 |
+
"loss": 0.4057,
|
|
|
|
| 903 |
"step": 1130
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 18.24,
|
| 907 |
+
"learning_rate": 2.1505376344086024e-05,
|
| 908 |
+
"loss": 0.356,
|
|
|
|
| 909 |
"step": 1140
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 18.4,
|
| 913 |
+
"learning_rate": 2.120669056152927e-05,
|
| 914 |
+
"loss": 0.3455,
|
|
|
|
| 915 |
"step": 1150
|
| 916 |
},
|
| 917 |
{
|
| 918 |
+
"epoch": 18.56,
|
| 919 |
+
"learning_rate": 2.0908004778972523e-05,
|
| 920 |
+
"loss": 0.3733,
|
|
|
|
| 921 |
"step": 1160
|
| 922 |
},
|
| 923 |
{
|
| 924 |
+
"epoch": 18.72,
|
| 925 |
+
"learning_rate": 2.060931899641577e-05,
|
| 926 |
+
"loss": 0.3674,
|
|
|
|
| 927 |
"step": 1170
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 18.88,
|
| 931 |
+
"learning_rate": 2.031063321385902e-05,
|
| 932 |
+
"loss": 0.3722,
|
|
|
|
| 933 |
"step": 1180
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 18.99,
|
| 937 |
+
"eval_accuracy": 0.8387824126268321,
|
| 938 |
+
"eval_f1": 0.8356991631458416,
|
| 939 |
+
"eval_loss": 0.5161823034286499,
|
| 940 |
+
"eval_precision": 0.8364077463920019,
|
| 941 |
+
"eval_recall": 0.8387824126268321,
|
| 942 |
+
"eval_runtime": 9.4455,
|
| 943 |
+
"eval_samples_per_second": 187.814,
|
| 944 |
+
"eval_steps_per_second": 2.964,
|
| 945 |
+
"step": 1187
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 19.04,
|
| 949 |
+
"learning_rate": 2.001194743130227e-05,
|
| 950 |
+
"loss": 0.377,
|
| 951 |
"step": 1190
|
| 952 |
},
|
| 953 |
{
|
| 954 |
+
"epoch": 19.2,
|
| 955 |
+
"learning_rate": 1.9713261648745522e-05,
|
| 956 |
+
"loss": 0.3445,
|
|
|
|
| 957 |
"step": 1200
|
| 958 |
},
|
| 959 |
{
|
| 960 |
+
"epoch": 19.36,
|
| 961 |
+
"learning_rate": 1.941457586618877e-05,
|
| 962 |
+
"loss": 0.3634,
|
|
|
|
| 963 |
"step": 1210
|
| 964 |
},
|
| 965 |
{
|
| 966 |
+
"epoch": 19.52,
|
| 967 |
+
"learning_rate": 1.9115890083632022e-05,
|
| 968 |
+
"loss": 0.37,
|
|
|
|
| 969 |
"step": 1220
|
| 970 |
},
|
| 971 |
{
|
| 972 |
+
"epoch": 19.68,
|
| 973 |
+
"learning_rate": 1.881720430107527e-05,
|
| 974 |
+
"loss": 0.3587,
|
|
|
|
| 975 |
"step": 1230
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 19.84,
|
| 979 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 980 |
+
"loss": 0.3676,
|
|
|
|
| 981 |
"step": 1240
|
| 982 |
},
|
| 983 |
{
|
| 984 |
+
"epoch": 20.0,
|
| 985 |
+
"learning_rate": 1.821983273596177e-05,
|
| 986 |
+
"loss": 0.3467,
|
| 987 |
+
"step": 1250
|
|
|
|
|
|
|
|
|
|
| 988 |
},
|
| 989 |
{
|
| 990 |
+
"epoch": 20.0,
|
| 991 |
+
"eval_accuracy": 0.8399098083427283,
|
| 992 |
+
"eval_f1": 0.8341879072603243,
|
| 993 |
+
"eval_loss": 0.512520968914032,
|
| 994 |
+
"eval_precision": 0.8356956407334528,
|
| 995 |
+
"eval_recall": 0.8399098083427283,
|
| 996 |
+
"eval_runtime": 9.3979,
|
| 997 |
+
"eval_samples_per_second": 188.765,
|
| 998 |
+
"eval_steps_per_second": 2.979,
|
| 999 |
"step": 1250
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
+
"epoch": 20.16,
|
| 1003 |
+
"learning_rate": 1.7921146953405018e-05,
|
| 1004 |
+
"loss": 0.3306,
|
|
|
|
| 1005 |
"step": 1260
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
+
"epoch": 20.32,
|
| 1009 |
+
"learning_rate": 1.7622461170848266e-05,
|
| 1010 |
+
"loss": 0.3441,
|
|
|
|
| 1011 |
"step": 1270
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
+
"epoch": 20.48,
|
| 1015 |
+
"learning_rate": 1.7323775388291518e-05,
|
| 1016 |
+
"loss": 0.3275,
|
|
|
|
| 1017 |
"step": 1280
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 20.64,
|
| 1021 |
+
"learning_rate": 1.702508960573477e-05,
|
| 1022 |
+
"loss": 0.3437,
|
|
|
|
| 1023 |
"step": 1290
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
+
"epoch": 20.8,
|
| 1027 |
+
"learning_rate": 1.6726403823178017e-05,
|
| 1028 |
+
"loss": 0.3654,
|
|
|
|
| 1029 |
"step": 1300
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
+
"epoch": 20.96,
|
| 1033 |
+
"learning_rate": 1.642771804062127e-05,
|
| 1034 |
+
"loss": 0.3518,
|
|
|
|
| 1035 |
"step": 1310
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
+
"epoch": 20.99,
|
| 1039 |
+
"eval_accuracy": 0.830890642615558,
|
| 1040 |
+
"eval_f1": 0.8276459982195031,
|
| 1041 |
+
"eval_loss": 0.5569064617156982,
|
| 1042 |
+
"eval_precision": 0.8326824734829679,
|
| 1043 |
+
"eval_recall": 0.830890642615558,
|
| 1044 |
+
"eval_runtime": 9.4092,
|
| 1045 |
+
"eval_samples_per_second": 188.538,
|
| 1046 |
+
"eval_steps_per_second": 2.976,
|
| 1047 |
+
"step": 1312
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 21.12,
|
| 1051 |
+
"learning_rate": 1.6129032258064517e-05,
|
| 1052 |
+
"loss": 0.3096,
|
| 1053 |
"step": 1320
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 21.28,
|
| 1057 |
+
"learning_rate": 1.5830346475507768e-05,
|
| 1058 |
+
"loss": 0.3105,
|
|
|
|
| 1059 |
"step": 1330
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 21.44,
|
| 1063 |
+
"learning_rate": 1.5531660692951016e-05,
|
| 1064 |
+
"loss": 0.3231,
|
|
|
|
| 1065 |
"step": 1340
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 21.6,
|
| 1069 |
+
"learning_rate": 1.5232974910394265e-05,
|
| 1070 |
+
"loss": 0.3348,
|
|
|
|
| 1071 |
"step": 1350
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 21.76,
|
| 1075 |
+
"learning_rate": 1.4934289127837516e-05,
|
| 1076 |
+
"loss": 0.3267,
|
|
|
|
| 1077 |
"step": 1360
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
+
"epoch": 21.92,
|
| 1081 |
+
"learning_rate": 1.4635603345280766e-05,
|
| 1082 |
+
"loss": 0.3432,
|
|
|
|
| 1083 |
"step": 1370
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
+
"epoch": 22.0,
|
| 1087 |
+
"eval_accuracy": 0.8483652762119503,
|
| 1088 |
+
"eval_f1": 0.8454482921463644,
|
| 1089 |
+
"eval_loss": 0.503221333026886,
|
| 1090 |
+
"eval_precision": 0.8451494566314203,
|
| 1091 |
+
"eval_recall": 0.8483652762119503,
|
| 1092 |
+
"eval_runtime": 9.6561,
|
| 1093 |
+
"eval_samples_per_second": 183.718,
|
| 1094 |
+
"eval_steps_per_second": 2.9,
|
| 1095 |
+
"step": 1375
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 22.08,
|
| 1099 |
+
"learning_rate": 1.4336917562724014e-05,
|
| 1100 |
+
"loss": 0.2949,
|
|
|
|
| 1101 |
"step": 1380
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 22.24,
|
| 1105 |
+
"learning_rate": 1.4038231780167265e-05,
|
| 1106 |
+
"loss": 0.2799,
|
|
|
|
| 1107 |
"step": 1390
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 22.4,
|
| 1111 |
+
"learning_rate": 1.3739545997610515e-05,
|
| 1112 |
+
"loss": 0.3187,
|
|
|
|
| 1113 |
"step": 1400
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 22.56,
|
| 1117 |
+
"learning_rate": 1.3440860215053763e-05,
|
| 1118 |
+
"loss": 0.3097,
|
|
|
|
| 1119 |
"step": 1410
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
+
"epoch": 22.72,
|
| 1123 |
+
"learning_rate": 1.3142174432497015e-05,
|
| 1124 |
+
"loss": 0.3246,
|
|
|
|
| 1125 |
"step": 1420
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
+
"epoch": 22.88,
|
| 1129 |
+
"learning_rate": 1.2843488649940263e-05,
|
| 1130 |
+
"loss": 0.3067,
|
|
|
|
| 1131 |
"step": 1430
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
+
"epoch": 22.99,
|
| 1135 |
+
"eval_accuracy": 0.8432919954904171,
|
| 1136 |
+
"eval_f1": 0.8432965588108505,
|
| 1137 |
+
"eval_loss": 0.5245735049247742,
|
| 1138 |
+
"eval_precision": 0.8461645036328044,
|
| 1139 |
+
"eval_recall": 0.8432919954904171,
|
| 1140 |
+
"eval_runtime": 9.4406,
|
| 1141 |
+
"eval_samples_per_second": 187.912,
|
| 1142 |
+
"eval_steps_per_second": 2.966,
|
| 1143 |
+
"step": 1437
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 23.04,
|
| 1147 |
+
"learning_rate": 1.2544802867383513e-05,
|
| 1148 |
+
"loss": 0.3102,
|
| 1149 |
"step": 1440
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
+
"epoch": 23.2,
|
| 1153 |
+
"learning_rate": 1.2246117084826763e-05,
|
| 1154 |
+
"loss": 0.2936,
|
|
|
|
| 1155 |
"step": 1450
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
+
"epoch": 23.36,
|
| 1159 |
+
"learning_rate": 1.1947431302270013e-05,
|
| 1160 |
+
"loss": 0.2917,
|
|
|
|
| 1161 |
"step": 1460
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
+
"epoch": 23.52,
|
| 1165 |
+
"learning_rate": 1.1648745519713262e-05,
|
| 1166 |
+
"loss": 0.278,
|
|
|
|
| 1167 |
"step": 1470
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
+
"epoch": 23.68,
|
| 1171 |
+
"learning_rate": 1.135005973715651e-05,
|
| 1172 |
+
"loss": 0.3334,
|
|
|
|
| 1173 |
"step": 1480
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
+
"epoch": 23.84,
|
| 1177 |
+
"learning_rate": 1.1051373954599762e-05,
|
| 1178 |
+
"loss": 0.2626,
|
|
|
|
| 1179 |
"step": 1490
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
+
"epoch": 24.0,
|
| 1183 |
+
"learning_rate": 1.0752688172043012e-05,
|
| 1184 |
+
"loss": 0.2923,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1185 |
"step": 1500
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 24.0,
|
| 1189 |
+
"eval_accuracy": 0.846674182638106,
|
| 1190 |
+
"eval_f1": 0.846389580543113,
|
| 1191 |
+
"eval_loss": 0.5363038778305054,
|
| 1192 |
+
"eval_precision": 0.84823995361272,
|
| 1193 |
+
"eval_recall": 0.846674182638106,
|
| 1194 |
+
"eval_runtime": 9.4508,
|
| 1195 |
+
"eval_samples_per_second": 187.708,
|
| 1196 |
+
"eval_steps_per_second": 2.963,
|
| 1197 |
+
"step": 1500
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 24.16,
|
| 1201 |
+
"learning_rate": 1.0454002389486262e-05,
|
| 1202 |
+
"loss": 0.2874,
|
| 1203 |
"step": 1510
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
+
"epoch": 24.32,
|
| 1207 |
+
"learning_rate": 1.015531660692951e-05,
|
| 1208 |
+
"loss": 0.2991,
|
|
|
|
| 1209 |
"step": 1520
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
+
"epoch": 24.48,
|
| 1213 |
+
"learning_rate": 9.856630824372761e-06,
|
| 1214 |
+
"loss": 0.2951,
|
|
|
|
| 1215 |
"step": 1530
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
+
"epoch": 24.64,
|
| 1219 |
+
"learning_rate": 9.557945041816011e-06,
|
| 1220 |
+
"loss": 0.2664,
|
|
|
|
| 1221 |
"step": 1540
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
+
"epoch": 24.8,
|
| 1225 |
+
"learning_rate": 9.259259259259259e-06,
|
| 1226 |
+
"loss": 0.2643,
|
|
|
|
| 1227 |
"step": 1550
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
+
"epoch": 24.96,
|
| 1231 |
+
"learning_rate": 8.960573476702509e-06,
|
| 1232 |
+
"loss": 0.303,
|
|
|
|
| 1233 |
"step": 1560
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
+
"epoch": 24.99,
|
| 1237 |
+
"eval_accuracy": 0.8483652762119503,
|
| 1238 |
+
"eval_f1": 0.8456910500695487,
|
| 1239 |
+
"eval_loss": 0.543497622013092,
|
| 1240 |
+
"eval_precision": 0.8453137649338565,
|
| 1241 |
+
"eval_recall": 0.8483652762119503,
|
| 1242 |
+
"eval_runtime": 9.4055,
|
| 1243 |
+
"eval_samples_per_second": 188.613,
|
| 1244 |
+
"eval_steps_per_second": 2.977,
|
| 1245 |
+
"step": 1562
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 25.12,
|
| 1249 |
+
"learning_rate": 8.661887694145759e-06,
|
| 1250 |
+
"loss": 0.2598,
|
| 1251 |
"step": 1570
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
+
"epoch": 25.28,
|
| 1255 |
+
"learning_rate": 8.363201911589009e-06,
|
| 1256 |
+
"loss": 0.2491,
|
|
|
|
| 1257 |
"step": 1580
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
+
"epoch": 25.44,
|
| 1261 |
+
"learning_rate": 8.064516129032258e-06,
|
| 1262 |
+
"loss": 0.2478,
|
|
|
|
| 1263 |
"step": 1590
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
+
"epoch": 25.6,
|
| 1267 |
+
"learning_rate": 7.765830346475508e-06,
|
| 1268 |
+
"loss": 0.2817,
|
|
|
|
| 1269 |
"step": 1600
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
+
"epoch": 25.76,
|
| 1273 |
+
"learning_rate": 7.467144563918758e-06,
|
| 1274 |
+
"loss": 0.2645,
|
|
|
|
| 1275 |
"step": 1610
|
| 1276 |
},
|
| 1277 |
{
|
| 1278 |
+
"epoch": 25.92,
|
| 1279 |
+
"learning_rate": 7.168458781362007e-06,
|
| 1280 |
+
"loss": 0.2523,
|
|
|
|
| 1281 |
"step": 1620
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
+
"epoch": 26.0,
|
| 1285 |
+
"eval_accuracy": 0.8444193912063134,
|
| 1286 |
+
"eval_f1": 0.8419142587663883,
|
| 1287 |
+
"eval_loss": 0.5499712824821472,
|
| 1288 |
+
"eval_precision": 0.8421658227924053,
|
| 1289 |
+
"eval_recall": 0.8444193912063134,
|
| 1290 |
+
"eval_runtime": 9.4664,
|
| 1291 |
+
"eval_samples_per_second": 187.399,
|
| 1292 |
+
"eval_steps_per_second": 2.958,
|
| 1293 |
+
"step": 1625
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 26.08,
|
| 1297 |
+
"learning_rate": 6.869772998805258e-06,
|
| 1298 |
+
"loss": 0.2821,
|
| 1299 |
"step": 1630
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
+
"epoch": 26.24,
|
| 1303 |
+
"learning_rate": 6.5710872162485075e-06,
|
| 1304 |
+
"loss": 0.2845,
|
|
|
|
| 1305 |
"step": 1640
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
+
"epoch": 26.4,
|
| 1309 |
+
"learning_rate": 6.2724014336917564e-06,
|
| 1310 |
+
"loss": 0.2578,
|
|
|
|
| 1311 |
"step": 1650
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
+
"epoch": 26.56,
|
| 1315 |
+
"learning_rate": 5.973715651135006e-06,
|
| 1316 |
+
"loss": 0.2487,
|
|
|
|
| 1317 |
"step": 1660
|
| 1318 |
},
|
| 1319 |
{
|
| 1320 |
+
"epoch": 26.72,
|
| 1321 |
+
"learning_rate": 5.675029868578255e-06,
|
| 1322 |
+
"loss": 0.2626,
|
|
|
|
| 1323 |
"step": 1670
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
+
"epoch": 26.88,
|
| 1327 |
+
"learning_rate": 5.376344086021506e-06,
|
| 1328 |
+
"loss": 0.2523,
|
|
|
|
| 1329 |
"step": 1680
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
+
"epoch": 26.99,
|
| 1333 |
+
"eval_accuracy": 0.8528748590755355,
|
| 1334 |
+
"eval_f1": 0.8529064904562935,
|
| 1335 |
+
"eval_loss": 0.5368608832359314,
|
| 1336 |
+
"eval_precision": 0.85325189512013,
|
| 1337 |
+
"eval_recall": 0.8528748590755355,
|
| 1338 |
+
"eval_runtime": 9.4466,
|
| 1339 |
+
"eval_samples_per_second": 187.792,
|
| 1340 |
+
"eval_steps_per_second": 2.964,
|
| 1341 |
+
"step": 1687
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 27.04,
|
| 1345 |
+
"learning_rate": 5.077658303464755e-06,
|
| 1346 |
+
"loss": 0.2445,
|
| 1347 |
"step": 1690
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
+
"epoch": 27.2,
|
| 1351 |
+
"learning_rate": 4.7789725209080055e-06,
|
| 1352 |
+
"loss": 0.254,
|
|
|
|
| 1353 |
"step": 1700
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
+
"epoch": 27.36,
|
| 1357 |
+
"learning_rate": 4.4802867383512545e-06,
|
| 1358 |
+
"loss": 0.2609,
|
|
|
|
| 1359 |
"step": 1710
|
| 1360 |
},
|
| 1361 |
{
|
| 1362 |
+
"epoch": 27.52,
|
| 1363 |
+
"learning_rate": 4.181600955794504e-06,
|
| 1364 |
+
"loss": 0.2406,
|
|
|
|
| 1365 |
"step": 1720
|
| 1366 |
},
|
| 1367 |
{
|
| 1368 |
+
"epoch": 27.68,
|
| 1369 |
+
"learning_rate": 3.882915173237754e-06,
|
| 1370 |
+
"loss": 0.217,
|
|
|
|
| 1371 |
"step": 1730
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
+
"epoch": 27.84,
|
| 1375 |
+
"learning_rate": 3.5842293906810035e-06,
|
| 1376 |
+
"loss": 0.2385,
|
|
|
|
| 1377 |
"step": 1740
|
| 1378 |
},
|
| 1379 |
{
|
| 1380 |
+
"epoch": 28.0,
|
| 1381 |
+
"learning_rate": 3.2855436081242537e-06,
|
| 1382 |
+
"loss": 0.262,
|
| 1383 |
+
"step": 1750
|
|
|
|
|
|
|
|
|
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
+
"epoch": 28.0,
|
| 1387 |
+
"eval_accuracy": 0.8562570462232244,
|
| 1388 |
+
"eval_f1": 0.8543013145108364,
|
| 1389 |
+
"eval_loss": 0.5373367667198181,
|
| 1390 |
+
"eval_precision": 0.8535856101978109,
|
| 1391 |
+
"eval_recall": 0.8562570462232244,
|
| 1392 |
+
"eval_runtime": 9.4681,
|
| 1393 |
+
"eval_samples_per_second": 187.366,
|
| 1394 |
+
"eval_steps_per_second": 2.957,
|
| 1395 |
"step": 1750
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
+
"epoch": 28.16,
|
| 1399 |
+
"learning_rate": 2.986857825567503e-06,
|
| 1400 |
+
"loss": 0.2366,
|
|
|
|
| 1401 |
"step": 1760
|
| 1402 |
},
|
| 1403 |
{
|
| 1404 |
+
"epoch": 28.32,
|
| 1405 |
+
"learning_rate": 2.688172043010753e-06,
|
| 1406 |
+
"loss": 0.2395,
|
|
|
|
| 1407 |
"step": 1770
|
| 1408 |
},
|
| 1409 |
{
|
| 1410 |
+
"epoch": 28.48,
|
| 1411 |
+
"learning_rate": 2.3894862604540028e-06,
|
| 1412 |
+
"loss": 0.221,
|
|
|
|
| 1413 |
"step": 1780
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
+
"epoch": 28.64,
|
| 1417 |
+
"learning_rate": 2.090800477897252e-06,
|
| 1418 |
+
"loss": 0.2437,
|
|
|
|
| 1419 |
"step": 1790
|
| 1420 |
},
|
| 1421 |
{
|
| 1422 |
+
"epoch": 28.8,
|
| 1423 |
+
"learning_rate": 1.7921146953405017e-06,
|
| 1424 |
+
"loss": 0.2405,
|
|
|
|
| 1425 |
"step": 1800
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
+
"epoch": 28.96,
|
| 1429 |
+
"learning_rate": 1.4934289127837516e-06,
|
| 1430 |
+
"loss": 0.232,
|
|
|
|
| 1431 |
"step": 1810
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
+
"epoch": 28.99,
|
| 1435 |
+
"eval_accuracy": 0.8528748590755355,
|
| 1436 |
+
"eval_f1": 0.8515517670522557,
|
| 1437 |
+
"eval_loss": 0.5384246706962585,
|
| 1438 |
+
"eval_precision": 0.8508958605112941,
|
| 1439 |
+
"eval_recall": 0.8528748590755355,
|
| 1440 |
+
"eval_runtime": 9.4835,
|
| 1441 |
+
"eval_samples_per_second": 187.062,
|
| 1442 |
+
"eval_steps_per_second": 2.953,
|
| 1443 |
+
"step": 1812
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 29.12,
|
| 1447 |
+
"learning_rate": 1.1947431302270014e-06,
|
| 1448 |
+
"loss": 0.2338,
|
| 1449 |
"step": 1820
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
+
"epoch": 29.28,
|
| 1453 |
+
"learning_rate": 8.960573476702509e-07,
|
| 1454 |
+
"loss": 0.2173,
|
|
|
|
| 1455 |
"step": 1830
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
+
"epoch": 29.44,
|
| 1459 |
+
"learning_rate": 5.973715651135007e-07,
|
| 1460 |
+
"loss": 0.2244,
|
|
|
|
| 1461 |
"step": 1840
|
| 1462 |
},
|
| 1463 |
{
|
| 1464 |
+
"epoch": 29.6,
|
| 1465 |
+
"learning_rate": 2.9868578255675034e-07,
|
| 1466 |
+
"loss": 0.2391,
|
|
|
|
| 1467 |
"step": 1850
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
+
"epoch": 29.76,
|
| 1471 |
+
"learning_rate": 0.0,
|
| 1472 |
+
"loss": 0.2278,
|
|
|
|
| 1473 |
"step": 1860
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
+
"epoch": 29.76,
|
| 1477 |
+
"eval_accuracy": 0.8511837655016911,
|
| 1478 |
+
"eval_f1": 0.849520928911175,
|
| 1479 |
+
"eval_loss": 0.5428664088249207,
|
| 1480 |
+
"eval_precision": 0.848929825944967,
|
| 1481 |
+
"eval_recall": 0.8511837655016911,
|
| 1482 |
+
"eval_runtime": 9.3955,
|
| 1483 |
+
"eval_samples_per_second": 188.814,
|
| 1484 |
+
"eval_steps_per_second": 2.98,
|
| 1485 |
+
"step": 1860
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1486 |
},
|
| 1487 |
{
|
| 1488 |
+
"epoch": 29.76,
|
| 1489 |
+
"step": 1860,
|
| 1490 |
+
"total_flos": 1.2364841575625785e+19,
|
| 1491 |
+
"train_loss": 0.4955275515074371,
|
| 1492 |
+
"train_runtime": 4173.3159,
|
| 1493 |
+
"train_samples_per_second": 114.743,
|
| 1494 |
+
"train_steps_per_second": 0.446
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1495 |
}
|
| 1496 |
],
|
| 1497 |
"logging_steps": 10,
|
| 1498 |
+
"max_steps": 1860,
|
| 1499 |
"num_input_tokens_seen": 0,
|
| 1500 |
+
"num_train_epochs": 30,
|
| 1501 |
"save_steps": 500,
|
| 1502 |
+
"total_flos": 1.2364841575625785e+19,
|
| 1503 |
+
"train_batch_size": 64,
|
| 1504 |
"trial_name": null,
|
| 1505 |
"trial_params": null
|
| 1506 |
}
|