Commit
•
9b0bc2a
1
Parent(s):
f1a2c70
End of training
Browse files- README.md +1 -1
- all_results.json +10 -10
- eval_results.json +6 -6
- train_results.json +5 -5
- trainer_state.json +1983 -367
README.md
CHANGED
|
@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
- Accuracy: 0.7968
|
| 37 |
|
| 38 |
## Model description
|
|
|
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 0.6838
|
| 36 |
- Accuracy: 0.7968
|
| 37 |
|
| 38 |
## Model description
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 3.
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second": 53.
|
| 12 |
"train_steps_per_second": 0.414
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.948571428571428,
|
| 3 |
+
"eval_accuracy": 0.796756082345602,
|
| 4 |
+
"eval_loss": 0.6838445663452148,
|
| 5 |
+
"eval_runtime": 70.7289,
|
| 6 |
+
"eval_samples_per_second": 113.32,
|
| 7 |
+
"eval_steps_per_second": 3.549,
|
| 8 |
+
"total_flos": 1.0403963301155365e+19,
|
| 9 |
+
"train_loss": 0.9863547705729073,
|
| 10 |
+
"train_runtime": 7901.2254,
|
| 11 |
+
"train_samples_per_second": 53.135,
|
| 12 |
"train_steps_per_second": 0.414
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 3.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.948571428571428,
|
| 3 |
+
"eval_accuracy": 0.796756082345602,
|
| 4 |
+
"eval_loss": 0.6838445663452148,
|
| 5 |
+
"eval_runtime": 70.7289,
|
| 6 |
+
"eval_samples_per_second": 113.32,
|
| 7 |
+
"eval_steps_per_second": 3.549
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second": 53.
|
| 7 |
"train_steps_per_second": 0.414
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.948571428571428,
|
| 3 |
+
"total_flos": 1.0403963301155365e+19,
|
| 4 |
+
"train_loss": 0.9863547705729073,
|
| 5 |
+
"train_runtime": 7901.2254,
|
| 6 |
+
"train_samples_per_second": 53.135,
|
| 7 |
"train_steps_per_second": 0.414
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,835 +1,2451 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat\\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": 0.045714285714285714,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.09142857142857143,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 2.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.13714285714285715,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 2.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.18285714285714286,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 2.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.22857142857142856,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 2.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.2742857142857143,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.32,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss":
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3657142857142857,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss":
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.4114285714285714,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss":
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.45714285714285713,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.5028571428571429,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.5485714285714286,
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.5942857142857143,
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.64,
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.6857142857142857,
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.7314285714285714,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.7771428571428571,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.8228571428571428,
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.8685714285714285,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.9142857142857143,
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.96,
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.9965714285714286,
|
| 160 |
-
"eval_accuracy": 0.
|
| 161 |
-
"eval_loss": 1.
|
| 162 |
-
"eval_runtime": 74.
|
| 163 |
-
"eval_samples_per_second": 107.
|
| 164 |
-
"eval_steps_per_second": 3.
|
| 165 |
"step": 218
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.0057142857142858,
|
| 169 |
-
"grad_norm":
|
| 170 |
-
"learning_rate":
|
| 171 |
-
"loss": 1.
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.0514285714285714,
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate":
|
| 178 |
-
"loss": 1.
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 1.0971428571428572,
|
| 183 |
-
"grad_norm":
|
| 184 |
-
"learning_rate":
|
| 185 |
-
"loss": 1.
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 1.1428571428571428,
|
| 190 |
-
"grad_norm":
|
| 191 |
-
"learning_rate":
|
| 192 |
-
"loss": 1.
|
| 193 |
"step": 250
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 1.1885714285714286,
|
| 197 |
-
"grad_norm":
|
| 198 |
-
"learning_rate":
|
| 199 |
-
"loss": 1.
|
| 200 |
"step": 260
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.2342857142857142,
|
| 204 |
-
"grad_norm": 8.
|
| 205 |
-
"learning_rate": 4.
|
| 206 |
-
"loss": 1.
|
| 207 |
"step": 270
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.28,
|
| 211 |
-
"grad_norm":
|
| 212 |
-
"learning_rate": 4.
|
| 213 |
-
"loss": 1.
|
| 214 |
"step": 280
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 1.3257142857142856,
|
| 218 |
-
"grad_norm":
|
| 219 |
-
"learning_rate": 4.
|
| 220 |
-
"loss": 1.
|
| 221 |
"step": 290
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 1.3714285714285714,
|
| 225 |
-
"grad_norm": 9.
|
| 226 |
-
"learning_rate": 4.
|
| 227 |
-
"loss": 1.
|
| 228 |
"step": 300
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 1.4171428571428573,
|
| 232 |
-
"grad_norm":
|
| 233 |
-
"learning_rate":
|
| 234 |
-
"loss": 1.
|
| 235 |
"step": 310
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 1.4628571428571429,
|
| 239 |
-
"grad_norm":
|
| 240 |
-
"learning_rate":
|
| 241 |
-
"loss": 1.
|
| 242 |
"step": 320
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 1.5085714285714285,
|
| 246 |
-
"grad_norm":
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss": 1.
|
| 249 |
"step": 330
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.5542857142857143,
|
| 253 |
-
"grad_norm":
|
| 254 |
-
"learning_rate":
|
| 255 |
-
"loss": 1.
|
| 256 |
"step": 340
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 1.6,
|
| 260 |
-
"grad_norm": 7.
|
| 261 |
-
"learning_rate":
|
| 262 |
-
"loss": 1.
|
| 263 |
"step": 350
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 1.6457142857142857,
|
| 267 |
-
"grad_norm":
|
| 268 |
-
"learning_rate":
|
| 269 |
-
"loss": 1.
|
| 270 |
"step": 360
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 1.6914285714285713,
|
| 274 |
-
"grad_norm":
|
| 275 |
-
"learning_rate":
|
| 276 |
-
"loss": 1.
|
| 277 |
"step": 370
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 1.737142857142857,
|
| 281 |
-
"grad_norm":
|
| 282 |
-
"learning_rate":
|
| 283 |
-
"loss": 1.
|
| 284 |
"step": 380
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 1.782857142857143,
|
| 288 |
-
"grad_norm":
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss": 1.
|
| 291 |
"step": 390
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 1.8285714285714287,
|
| 295 |
-
"grad_norm": 9.
|
| 296 |
-
"learning_rate":
|
| 297 |
-
"loss": 1.
|
| 298 |
"step": 400
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 1.8742857142857143,
|
| 302 |
-
"grad_norm":
|
| 303 |
-
"learning_rate":
|
| 304 |
-
"loss": 1.
|
| 305 |
"step": 410
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 1.92,
|
| 309 |
-
"grad_norm":
|
| 310 |
-
"learning_rate":
|
| 311 |
-
"loss": 1.
|
| 312 |
"step": 420
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 1.9657142857142857,
|
| 316 |
-
"grad_norm":
|
| 317 |
-
"learning_rate":
|
| 318 |
-
"loss": 1.
|
| 319 |
"step": 430
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 1.9977142857142858,
|
| 323 |
-
"eval_accuracy": 0.
|
| 324 |
-
"eval_loss": 1.
|
| 325 |
-
"eval_runtime": 74.
|
| 326 |
-
"eval_samples_per_second":
|
| 327 |
-
"eval_steps_per_second": 3.
|
| 328 |
"step": 437
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 2.0114285714285716,
|
| 332 |
-
"grad_norm":
|
| 333 |
-
"learning_rate":
|
| 334 |
-
"loss": 1.
|
| 335 |
"step": 440
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 2.057142857142857,
|
| 339 |
-
"grad_norm":
|
| 340 |
-
"learning_rate":
|
| 341 |
-
"loss": 1.
|
| 342 |
"step": 450
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 2.1028571428571428,
|
| 346 |
-
"grad_norm":
|
| 347 |
-
"learning_rate":
|
| 348 |
-
"loss": 1.
|
| 349 |
"step": 460
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 2.1485714285714286,
|
| 353 |
-
"grad_norm":
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss": 1.
|
| 356 |
"step": 470
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 2.1942857142857144,
|
| 360 |
-
"grad_norm":
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss": 1.
|
| 363 |
"step": 480
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 2.24,
|
| 367 |
-
"grad_norm": 10.
|
| 368 |
-
"learning_rate":
|
| 369 |
-
"loss": 1.
|
| 370 |
"step": 490
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 2.2857142857142856,
|
| 374 |
-
"grad_norm":
|
| 375 |
-
"learning_rate":
|
| 376 |
-
"loss": 1.
|
| 377 |
"step": 500
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 2.3314285714285714,
|
| 381 |
-
"grad_norm": 7.
|
| 382 |
-
"learning_rate":
|
| 383 |
-
"loss": 1.
|
| 384 |
"step": 510
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 2.3771428571428572,
|
| 388 |
-
"grad_norm":
|
| 389 |
-
"learning_rate":
|
| 390 |
-
"loss": 1.
|
| 391 |
"step": 520
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 2.422857142857143,
|
| 395 |
-
"grad_norm":
|
| 396 |
-
"learning_rate":
|
| 397 |
-
"loss": 1.
|
| 398 |
"step": 530
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 2.4685714285714284,
|
| 402 |
-
"grad_norm":
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss": 1.
|
| 405 |
"step": 540
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 2.5142857142857142,
|
| 409 |
-
"grad_norm":
|
| 410 |
-
"learning_rate":
|
| 411 |
-
"loss": 1.
|
| 412 |
"step": 550
|
| 413 |
},
|
| 414 |
{
|
| 415 |
"epoch": 2.56,
|
| 416 |
-
"grad_norm": 8.
|
| 417 |
-
"learning_rate":
|
| 418 |
-
"loss": 1.
|
| 419 |
"step": 560
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 2.605714285714286,
|
| 423 |
-
"grad_norm": 7.
|
| 424 |
-
"learning_rate":
|
| 425 |
-
"loss": 1.
|
| 426 |
"step": 570
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 2.6514285714285712,
|
| 430 |
-
"grad_norm":
|
| 431 |
-
"learning_rate":
|
| 432 |
-
"loss": 1.
|
| 433 |
"step": 580
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 2.697142857142857,
|
| 437 |
-
"grad_norm":
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss": 1.
|
| 440 |
"step": 590
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 2.742857142857143,
|
| 444 |
-
"grad_norm":
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 1.
|
| 447 |
"step": 600
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 2.7885714285714287,
|
| 451 |
-
"grad_norm":
|
| 452 |
-
"learning_rate":
|
| 453 |
-
"loss": 1.
|
| 454 |
"step": 610
|
| 455 |
},
|
| 456 |
{
|
| 457 |
"epoch": 2.8342857142857145,
|
| 458 |
-
"grad_norm":
|
| 459 |
-
"learning_rate":
|
| 460 |
-
"loss": 1.
|
| 461 |
"step": 620
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 2.88,
|
| 465 |
-
"grad_norm":
|
| 466 |
-
"learning_rate":
|
| 467 |
-
"loss": 1.
|
| 468 |
"step": 630
|
| 469 |
},
|
| 470 |
{
|
| 471 |
"epoch": 2.9257142857142857,
|
| 472 |
-
"grad_norm": 8.
|
| 473 |
-
"learning_rate":
|
| 474 |
-
"loss": 1.
|
| 475 |
"step": 640
|
| 476 |
},
|
| 477 |
{
|
| 478 |
"epoch": 2.9714285714285715,
|
| 479 |
-
"grad_norm": 7.
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 1.
|
| 482 |
"step": 650
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 2.998857142857143,
|
| 486 |
-
"eval_accuracy": 0.
|
| 487 |
-
"eval_loss": 0.
|
| 488 |
-
"eval_runtime":
|
| 489 |
-
"eval_samples_per_second":
|
| 490 |
-
"eval_steps_per_second": 3.
|
| 491 |
"step": 656
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 3.0171428571428573,
|
| 495 |
-
"grad_norm":
|
| 496 |
-
"learning_rate":
|
| 497 |
-
"loss": 1.
|
| 498 |
"step": 660
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 3.0628571428571427,
|
| 502 |
-
"grad_norm":
|
| 503 |
-
"learning_rate":
|
| 504 |
-
"loss": 1.
|
| 505 |
"step": 670
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 3.1085714285714285,
|
| 509 |
-
"grad_norm": 7.
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss":
|
| 512 |
"step": 680
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 3.1542857142857144,
|
| 516 |
-
"grad_norm":
|
| 517 |
-
"learning_rate":
|
| 518 |
-
"loss": 1.
|
| 519 |
"step": 690
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 3.2,
|
| 523 |
-
"grad_norm":
|
| 524 |
-
"learning_rate":
|
| 525 |
-
"loss": 1.
|
| 526 |
"step": 700
|
| 527 |
},
|
| 528 |
{
|
| 529 |
"epoch": 3.2457142857142856,
|
| 530 |
-
"grad_norm":
|
| 531 |
-
"learning_rate":
|
| 532 |
-
"loss": 1.
|
| 533 |
"step": 710
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 3.2914285714285714,
|
| 537 |
-
"grad_norm":
|
| 538 |
-
"learning_rate":
|
| 539 |
-
"loss": 1.
|
| 540 |
"step": 720
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 3.337142857142857,
|
| 544 |
-
"grad_norm":
|
| 545 |
-
"learning_rate":
|
| 546 |
-
"loss": 1.
|
| 547 |
"step": 730
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 3.382857142857143,
|
| 551 |
-
"grad_norm":
|
| 552 |
-
"learning_rate":
|
| 553 |
-
"loss": 1.
|
| 554 |
"step": 740
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 3.4285714285714284,
|
| 558 |
-
"grad_norm":
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 1.
|
| 561 |
"step": 750
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 3.474285714285714,
|
| 565 |
-
"grad_norm":
|
| 566 |
-
"learning_rate":
|
| 567 |
-
"loss": 1.
|
| 568 |
"step": 760
|
| 569 |
},
|
| 570 |
{
|
| 571 |
"epoch": 3.52,
|
| 572 |
-
"grad_norm":
|
| 573 |
-
"learning_rate":
|
| 574 |
-
"loss":
|
| 575 |
"step": 770
|
| 576 |
},
|
| 577 |
{
|
| 578 |
"epoch": 3.565714285714286,
|
| 579 |
-
"grad_norm":
|
| 580 |
-
"learning_rate":
|
| 581 |
-
"loss": 1.
|
| 582 |
"step": 780
|
| 583 |
},
|
| 584 |
{
|
| 585 |
"epoch": 3.611428571428571,
|
| 586 |
-
"grad_norm": 8.
|
| 587 |
-
"learning_rate":
|
| 588 |
-
"loss": 1.
|
| 589 |
"step": 790
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 3.657142857142857,
|
| 593 |
-
"grad_norm": 7.
|
| 594 |
-
"learning_rate":
|
| 595 |
-
"loss": 1.
|
| 596 |
"step": 800
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 3.702857142857143,
|
| 600 |
-
"grad_norm":
|
| 601 |
-
"learning_rate":
|
| 602 |
-
"loss": 1.
|
| 603 |
"step": 810
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 3.7485714285714287,
|
| 607 |
-
"grad_norm":
|
| 608 |
-
"learning_rate":
|
| 609 |
-
"loss": 1.
|
| 610 |
"step": 820
|
| 611 |
},
|
| 612 |
{
|
| 613 |
"epoch": 3.7942857142857145,
|
| 614 |
-
"grad_norm":
|
| 615 |
-
"learning_rate":
|
| 616 |
-
"loss": 1.
|
| 617 |
"step": 830
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 3.84,
|
| 621 |
-
"grad_norm":
|
| 622 |
-
"learning_rate":
|
| 623 |
-
"loss": 1.
|
| 624 |
"step": 840
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 3.8857142857142857,
|
| 628 |
-
"grad_norm": 6.
|
| 629 |
-
"learning_rate":
|
| 630 |
-
"loss": 1.
|
| 631 |
"step": 850
|
| 632 |
},
|
| 633 |
{
|
| 634 |
"epoch": 3.9314285714285715,
|
| 635 |
-
"grad_norm":
|
| 636 |
-
"learning_rate":
|
| 637 |
-
"loss": 1.
|
| 638 |
"step": 860
|
| 639 |
},
|
| 640 |
{
|
| 641 |
"epoch": 3.977142857142857,
|
| 642 |
-
"grad_norm":
|
| 643 |
-
"learning_rate":
|
| 644 |
-
"loss":
|
| 645 |
"step": 870
|
| 646 |
},
|
| 647 |
{
|
| 648 |
"epoch": 4.0,
|
| 649 |
-
"eval_accuracy": 0.
|
| 650 |
-
"eval_loss": 0.
|
| 651 |
-
"eval_runtime": 74.
|
| 652 |
-
"eval_samples_per_second":
|
| 653 |
-
"eval_steps_per_second": 3.
|
| 654 |
"step": 875
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 4.022857142857143,
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"loss": 1.
|
| 661 |
"step": 880
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 4.0685714285714285,
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss": 1.
|
| 668 |
"step": 890
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 4.114285714285714,
|
| 672 |
-
"grad_norm": 8.
|
| 673 |
-
"learning_rate":
|
| 674 |
-
"loss": 1.
|
| 675 |
"step": 900
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 4.16,
|
| 679 |
-
"grad_norm": 7.
|
| 680 |
-
"learning_rate":
|
| 681 |
-
"loss": 1.
|
| 682 |
"step": 910
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 4.2057142857142855,
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate":
|
| 688 |
-
"loss": 1.
|
| 689 |
"step": 920
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 4.251428571428572,
|
| 693 |
-
"grad_norm": 8.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss": 1.
|
| 696 |
"step": 930
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 4.297142857142857,
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate":
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 940
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 4.3428571428571425,
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss":
|
| 710 |
"step": 950
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 4.388571428571429,
|
| 714 |
-
"grad_norm": 6.
|
| 715 |
-
"learning_rate":
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 960
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 4.434285714285714,
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate":
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 970
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 4.48,
|
| 728 |
-
"grad_norm":
|
| 729 |
-
"learning_rate":
|
| 730 |
-
"loss": 1.
|
| 731 |
"step": 980
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 4.525714285714286,
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss": 1.
|
| 738 |
"step": 990
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 4.571428571428571,
|
| 742 |
-
"grad_norm":
|
| 743 |
-
"learning_rate":
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 1000
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 4.617142857142857,
|
| 749 |
-
"grad_norm": 6.
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss": 0.
|
| 752 |
"step": 1010
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 4.662857142857143,
|
| 756 |
-
"grad_norm":
|
| 757 |
-
"learning_rate": 3.
|
| 758 |
-
"loss": 1.
|
| 759 |
"step": 1020
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 4.708571428571428,
|
| 763 |
-
"grad_norm":
|
| 764 |
-
"learning_rate": 3.
|
| 765 |
-
"loss": 1.
|
| 766 |
"step": 1030
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 4.7542857142857144,
|
| 770 |
-
"grad_norm":
|
| 771 |
-
"learning_rate":
|
| 772 |
-
"loss":
|
| 773 |
"step": 1040
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 4.8,
|
| 777 |
-
"grad_norm":
|
| 778 |
-
"learning_rate":
|
| 779 |
-
"loss":
|
| 780 |
"step": 1050
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 4.845714285714286,
|
| 784 |
-
"grad_norm":
|
| 785 |
-
"learning_rate":
|
| 786 |
-
"loss": 0.
|
| 787 |
"step": 1060
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 4.8914285714285715,
|
| 791 |
-
"grad_norm":
|
| 792 |
-
"learning_rate":
|
| 793 |
-
"loss":
|
| 794 |
"step": 1070
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 4.937142857142857,
|
| 798 |
-
"grad_norm":
|
| 799 |
-
"learning_rate":
|
| 800 |
-
"loss": 0.
|
| 801 |
"step": 1080
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 4.982857142857143,
|
| 805 |
-
"grad_norm": 7.
|
| 806 |
-
"learning_rate":
|
| 807 |
-
"loss": 0.
|
| 808 |
"step": 1090
|
| 809 |
},
|
| 810 |
{
|
| 811 |
-
"epoch": 4.
|
| 812 |
-
"eval_accuracy": 0.
|
| 813 |
-
"eval_loss": 0.
|
| 814 |
-
"eval_runtime":
|
| 815 |
-
"eval_samples_per_second":
|
| 816 |
-
"eval_steps_per_second": 3.
|
| 817 |
-
"step":
|
| 818 |
},
|
| 819 |
{
|
| 820 |
-
"epoch":
|
| 821 |
-
"
|
| 822 |
-
"
|
| 823 |
-
"
|
| 824 |
-
"
|
| 825 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 826 |
"train_steps_per_second": 0.414
|
| 827 |
}
|
| 828 |
],
|
| 829 |
"logging_steps": 10,
|
| 830 |
-
"max_steps":
|
| 831 |
"num_input_tokens_seen": 0,
|
| 832 |
-
"num_train_epochs":
|
| 833 |
"save_steps": 500,
|
| 834 |
"stateful_callbacks": {
|
| 835 |
"TrainerControl": {
|
|
@@ -843,7 +2459,7 @@
|
|
| 843 |
"attributes": {}
|
| 844 |
}
|
| 845 |
},
|
| 846 |
-
"total_flos":
|
| 847 |
"train_batch_size": 32,
|
| 848 |
"trial_name": null,
|
| 849 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.796756082345602,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat\\checkpoint-3062",
|
| 4 |
+
"epoch": 14.948571428571428,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3270,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.045714285714285714,
|
| 13 |
+
"grad_norm": 13.994832992553711,
|
| 14 |
+
"learning_rate": 1.5290519877675841e-06,
|
| 15 |
+
"loss": 2.7494,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.09142857142857143,
|
| 20 |
+
"grad_norm": 7.420119762420654,
|
| 21 |
+
"learning_rate": 3.0581039755351682e-06,
|
| 22 |
+
"loss": 2.7395,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.13714285714285715,
|
| 27 |
+
"grad_norm": 5.467243194580078,
|
| 28 |
+
"learning_rate": 4.587155963302753e-06,
|
| 29 |
+
"loss": 2.7049,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.18285714285714286,
|
| 34 |
+
"grad_norm": 7.835602760314941,
|
| 35 |
+
"learning_rate": 6.1162079510703365e-06,
|
| 36 |
+
"loss": 2.6367,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.22857142857142856,
|
| 41 |
+
"grad_norm": 13.763399124145508,
|
| 42 |
+
"learning_rate": 7.645259938837921e-06,
|
| 43 |
+
"loss": 2.5592,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.2742857142857143,
|
| 48 |
+
"grad_norm": 10.583270072937012,
|
| 49 |
+
"learning_rate": 9.174311926605506e-06,
|
| 50 |
+
"loss": 2.4444,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.32,
|
| 55 |
+
"grad_norm": 19.168621063232422,
|
| 56 |
+
"learning_rate": 1.070336391437309e-05,
|
| 57 |
+
"loss": 2.3336,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3657142857142857,
|
| 62 |
+
"grad_norm": 20.13259506225586,
|
| 63 |
+
"learning_rate": 1.2232415902140673e-05,
|
| 64 |
+
"loss": 2.1877,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.4114285714285714,
|
| 69 |
+
"grad_norm": 28.141359329223633,
|
| 70 |
+
"learning_rate": 1.3761467889908258e-05,
|
| 71 |
+
"loss": 2.0578,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.45714285714285713,
|
| 76 |
+
"grad_norm": 36.60035705566406,
|
| 77 |
+
"learning_rate": 1.5290519877675842e-05,
|
| 78 |
+
"loss": 1.9729,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.5028571428571429,
|
| 83 |
+
"grad_norm": 27.167247772216797,
|
| 84 |
+
"learning_rate": 1.6819571865443427e-05,
|
| 85 |
+
"loss": 1.884,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.5485714285714286,
|
| 90 |
+
"grad_norm": 21.56601905822754,
|
| 91 |
+
"learning_rate": 1.834862385321101e-05,
|
| 92 |
+
"loss": 1.7821,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.5942857142857143,
|
| 97 |
+
"grad_norm": 18.523468017578125,
|
| 98 |
+
"learning_rate": 1.9877675840978592e-05,
|
| 99 |
+
"loss": 1.8148,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.64,
|
| 104 |
+
"grad_norm": 14.976977348327637,
|
| 105 |
+
"learning_rate": 2.140672782874618e-05,
|
| 106 |
+
"loss": 1.6829,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.6857142857142857,
|
| 111 |
+
"grad_norm": 14.441283226013184,
|
| 112 |
+
"learning_rate": 2.2935779816513765e-05,
|
| 113 |
+
"loss": 1.6312,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.7314285714285714,
|
| 118 |
+
"grad_norm": 17.649080276489258,
|
| 119 |
+
"learning_rate": 2.4464831804281346e-05,
|
| 120 |
+
"loss": 1.6482,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.7771428571428571,
|
| 125 |
+
"grad_norm": 19.691490173339844,
|
| 126 |
+
"learning_rate": 2.5993883792048927e-05,
|
| 127 |
+
"loss": 1.5693,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.8228571428571428,
|
| 132 |
+
"grad_norm": 12.245804786682129,
|
| 133 |
+
"learning_rate": 2.7522935779816515e-05,
|
| 134 |
+
"loss": 1.6267,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.8685714285714285,
|
| 139 |
+
"grad_norm": 24.526466369628906,
|
| 140 |
+
"learning_rate": 2.90519877675841e-05,
|
| 141 |
+
"loss": 1.5964,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.9142857142857143,
|
| 146 |
+
"grad_norm": 14.706289291381836,
|
| 147 |
+
"learning_rate": 3.0581039755351684e-05,
|
| 148 |
+
"loss": 1.6017,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.96,
|
| 153 |
+
"grad_norm": 12.513092041015625,
|
| 154 |
+
"learning_rate": 3.211009174311927e-05,
|
| 155 |
+
"loss": 1.5891,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.9965714285714286,
|
| 160 |
+
"eval_accuracy": 0.5723019338739863,
|
| 161 |
+
"eval_loss": 1.383291244506836,
|
| 162 |
+
"eval_runtime": 74.6009,
|
| 163 |
+
"eval_samples_per_second": 107.438,
|
| 164 |
+
"eval_steps_per_second": 3.365,
|
| 165 |
"step": 218
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.0057142857142858,
|
| 169 |
+
"grad_norm": 9.570089340209961,
|
| 170 |
+
"learning_rate": 3.363914373088685e-05,
|
| 171 |
+
"loss": 1.4337,
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.0514285714285714,
|
| 176 |
+
"grad_norm": 29.601459503173828,
|
| 177 |
+
"learning_rate": 3.516819571865443e-05,
|
| 178 |
+
"loss": 1.4987,
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 1.0971428571428572,
|
| 183 |
+
"grad_norm": 14.568526268005371,
|
| 184 |
+
"learning_rate": 3.669724770642202e-05,
|
| 185 |
+
"loss": 1.4594,
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 1.1428571428571428,
|
| 190 |
+
"grad_norm": 9.901498794555664,
|
| 191 |
+
"learning_rate": 3.822629969418961e-05,
|
| 192 |
+
"loss": 1.4964,
|
| 193 |
"step": 250
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 1.1885714285714286,
|
| 197 |
+
"grad_norm": 11.662745475769043,
|
| 198 |
+
"learning_rate": 3.9755351681957185e-05,
|
| 199 |
+
"loss": 1.4872,
|
| 200 |
"step": 260
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.2342857142857142,
|
| 204 |
+
"grad_norm": 8.338129043579102,
|
| 205 |
+
"learning_rate": 4.1284403669724776e-05,
|
| 206 |
+
"loss": 1.3592,
|
| 207 |
"step": 270
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.28,
|
| 211 |
+
"grad_norm": 12.640501022338867,
|
| 212 |
+
"learning_rate": 4.281345565749236e-05,
|
| 213 |
+
"loss": 1.4165,
|
| 214 |
"step": 280
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 1.3257142857142856,
|
| 218 |
+
"grad_norm": 9.851750373840332,
|
| 219 |
+
"learning_rate": 4.434250764525994e-05,
|
| 220 |
+
"loss": 1.4671,
|
| 221 |
"step": 290
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 1.3714285714285714,
|
| 225 |
+
"grad_norm": 9.80663013458252,
|
| 226 |
+
"learning_rate": 4.587155963302753e-05,
|
| 227 |
+
"loss": 1.4069,
|
| 228 |
"step": 300
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 1.4171428571428573,
|
| 232 |
+
"grad_norm": 10.75167179107666,
|
| 233 |
+
"learning_rate": 4.740061162079511e-05,
|
| 234 |
+
"loss": 1.3966,
|
| 235 |
"step": 310
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 1.4628571428571429,
|
| 239 |
+
"grad_norm": 9.546366691589355,
|
| 240 |
+
"learning_rate": 4.892966360856269e-05,
|
| 241 |
+
"loss": 1.3365,
|
| 242 |
"step": 320
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 1.5085714285714285,
|
| 246 |
+
"grad_norm": 11.726640701293945,
|
| 247 |
+
"learning_rate": 4.994903160040775e-05,
|
| 248 |
+
"loss": 1.4017,
|
| 249 |
"step": 330
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.5542857142857143,
|
| 253 |
+
"grad_norm": 9.12558650970459,
|
| 254 |
+
"learning_rate": 4.977913693510024e-05,
|
| 255 |
+
"loss": 1.3402,
|
| 256 |
"step": 340
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 1.6,
|
| 260 |
+
"grad_norm": 7.8366804122924805,
|
| 261 |
+
"learning_rate": 4.9609242269792734e-05,
|
| 262 |
+
"loss": 1.3675,
|
| 263 |
"step": 350
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 1.6457142857142857,
|
| 267 |
+
"grad_norm": 7.4768805503845215,
|
| 268 |
+
"learning_rate": 4.943934760448522e-05,
|
| 269 |
+
"loss": 1.3559,
|
| 270 |
"step": 360
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 1.6914285714285713,
|
| 274 |
+
"grad_norm": 8.65304183959961,
|
| 275 |
+
"learning_rate": 4.926945293917771e-05,
|
| 276 |
+
"loss": 1.3973,
|
| 277 |
"step": 370
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 1.737142857142857,
|
| 281 |
+
"grad_norm": 8.800814628601074,
|
| 282 |
+
"learning_rate": 4.9099558273870206e-05,
|
| 283 |
+
"loss": 1.2501,
|
| 284 |
"step": 380
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 1.782857142857143,
|
| 288 |
+
"grad_norm": 14.658060073852539,
|
| 289 |
+
"learning_rate": 4.892966360856269e-05,
|
| 290 |
+
"loss": 1.287,
|
| 291 |
"step": 390
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 1.8285714285714287,
|
| 295 |
+
"grad_norm": 9.586373329162598,
|
| 296 |
+
"learning_rate": 4.8759768943255185e-05,
|
| 297 |
+
"loss": 1.3249,
|
| 298 |
"step": 400
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 1.8742857142857143,
|
| 302 |
+
"grad_norm": 12.57496166229248,
|
| 303 |
+
"learning_rate": 4.858987427794767e-05,
|
| 304 |
+
"loss": 1.3564,
|
| 305 |
"step": 410
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 1.92,
|
| 309 |
+
"grad_norm": 9.330473899841309,
|
| 310 |
+
"learning_rate": 4.8419979612640164e-05,
|
| 311 |
+
"loss": 1.2967,
|
| 312 |
"step": 420
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 1.9657142857142857,
|
| 316 |
+
"grad_norm": 8.545197486877441,
|
| 317 |
+
"learning_rate": 4.8250084947332657e-05,
|
| 318 |
+
"loss": 1.2997,
|
| 319 |
"step": 430
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 1.9977142857142858,
|
| 323 |
+
"eval_accuracy": 0.6699937616968185,
|
| 324 |
+
"eval_loss": 1.0830751657485962,
|
| 325 |
+
"eval_runtime": 74.7614,
|
| 326 |
+
"eval_samples_per_second": 107.208,
|
| 327 |
+
"eval_steps_per_second": 3.357,
|
| 328 |
"step": 437
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 2.0114285714285716,
|
| 332 |
+
"grad_norm": 7.490045547485352,
|
| 333 |
+
"learning_rate": 4.808019028202514e-05,
|
| 334 |
+
"loss": 1.258,
|
| 335 |
"step": 440
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 2.057142857142857,
|
| 339 |
+
"grad_norm": 8.006925582885742,
|
| 340 |
+
"learning_rate": 4.7910295616717635e-05,
|
| 341 |
+
"loss": 1.168,
|
| 342 |
"step": 450
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 2.1028571428571428,
|
| 346 |
+
"grad_norm": 12.528056144714355,
|
| 347 |
+
"learning_rate": 4.774040095141013e-05,
|
| 348 |
+
"loss": 1.2195,
|
| 349 |
"step": 460
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 2.1485714285714286,
|
| 353 |
+
"grad_norm": 9.448918342590332,
|
| 354 |
+
"learning_rate": 4.7570506286102614e-05,
|
| 355 |
+
"loss": 1.2242,
|
| 356 |
"step": 470
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 2.1942857142857144,
|
| 360 |
+
"grad_norm": 7.972462177276611,
|
| 361 |
+
"learning_rate": 4.740061162079511e-05,
|
| 362 |
+
"loss": 1.1925,
|
| 363 |
"step": 480
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 2.24,
|
| 367 |
+
"grad_norm": 10.068849563598633,
|
| 368 |
+
"learning_rate": 4.72307169554876e-05,
|
| 369 |
+
"loss": 1.2521,
|
| 370 |
"step": 490
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 2.2857142857142856,
|
| 374 |
+
"grad_norm": 9.536931037902832,
|
| 375 |
+
"learning_rate": 4.706082229018009e-05,
|
| 376 |
+
"loss": 1.2375,
|
| 377 |
"step": 500
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 2.3314285714285714,
|
| 381 |
+
"grad_norm": 7.058177471160889,
|
| 382 |
+
"learning_rate": 4.6890927624872586e-05,
|
| 383 |
+
"loss": 1.1762,
|
| 384 |
"step": 510
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 2.3771428571428572,
|
| 388 |
+
"grad_norm": 8.200982093811035,
|
| 389 |
+
"learning_rate": 4.672103295956507e-05,
|
| 390 |
+
"loss": 1.259,
|
| 391 |
"step": 520
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 2.422857142857143,
|
| 395 |
+
"grad_norm": 6.11261510848999,
|
| 396 |
+
"learning_rate": 4.6551138294257565e-05,
|
| 397 |
+
"loss": 1.2291,
|
| 398 |
"step": 530
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 2.4685714285714284,
|
| 402 |
+
"grad_norm": 8.771502494812012,
|
| 403 |
+
"learning_rate": 4.638124362895006e-05,
|
| 404 |
+
"loss": 1.2237,
|
| 405 |
"step": 540
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 2.5142857142857142,
|
| 409 |
+
"grad_norm": 11.995628356933594,
|
| 410 |
+
"learning_rate": 4.6211348963642544e-05,
|
| 411 |
+
"loss": 1.2349,
|
| 412 |
"step": 550
|
| 413 |
},
|
| 414 |
{
|
| 415 |
"epoch": 2.56,
|
| 416 |
+
"grad_norm": 8.15999698638916,
|
| 417 |
+
"learning_rate": 4.604145429833504e-05,
|
| 418 |
+
"loss": 1.1786,
|
| 419 |
"step": 560
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 2.605714285714286,
|
| 423 |
+
"grad_norm": 7.252849578857422,
|
| 424 |
+
"learning_rate": 4.587155963302753e-05,
|
| 425 |
+
"loss": 1.1285,
|
| 426 |
"step": 570
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 2.6514285714285712,
|
| 430 |
+
"grad_norm": 8.760148048400879,
|
| 431 |
+
"learning_rate": 4.5701664967720016e-05,
|
| 432 |
+
"loss": 1.074,
|
| 433 |
"step": 580
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 2.697142857142857,
|
| 437 |
+
"grad_norm": 5.591124057769775,
|
| 438 |
+
"learning_rate": 4.553177030241251e-05,
|
| 439 |
+
"loss": 1.104,
|
| 440 |
"step": 590
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 2.742857142857143,
|
| 444 |
+
"grad_norm": 7.468674659729004,
|
| 445 |
+
"learning_rate": 4.5361875637104995e-05,
|
| 446 |
+
"loss": 1.1936,
|
| 447 |
"step": 600
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 2.7885714285714287,
|
| 451 |
+
"grad_norm": 7.683737754821777,
|
| 452 |
+
"learning_rate": 4.519198097179749e-05,
|
| 453 |
+
"loss": 1.1574,
|
| 454 |
"step": 610
|
| 455 |
},
|
| 456 |
{
|
| 457 |
"epoch": 2.8342857142857145,
|
| 458 |
+
"grad_norm": 12.986687660217285,
|
| 459 |
+
"learning_rate": 4.502208630648998e-05,
|
| 460 |
+
"loss": 1.1513,
|
| 461 |
"step": 620
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 2.88,
|
| 465 |
+
"grad_norm": 8.037307739257812,
|
| 466 |
+
"learning_rate": 4.4852191641182466e-05,
|
| 467 |
+
"loss": 1.1537,
|
| 468 |
"step": 630
|
| 469 |
},
|
| 470 |
{
|
| 471 |
"epoch": 2.9257142857142857,
|
| 472 |
+
"grad_norm": 8.727031707763672,
|
| 473 |
+
"learning_rate": 4.468229697587496e-05,
|
| 474 |
+
"loss": 1.1823,
|
| 475 |
"step": 640
|
| 476 |
},
|
| 477 |
{
|
| 478 |
"epoch": 2.9714285714285715,
|
| 479 |
+
"grad_norm": 7.474994659423828,
|
| 480 |
+
"learning_rate": 4.451240231056745e-05,
|
| 481 |
+
"loss": 1.1166,
|
| 482 |
"step": 650
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 2.998857142857143,
|
| 486 |
+
"eval_accuracy": 0.6958203368683719,
|
| 487 |
+
"eval_loss": 0.9937148094177246,
|
| 488 |
+
"eval_runtime": 75.7204,
|
| 489 |
+
"eval_samples_per_second": 105.85,
|
| 490 |
+
"eval_steps_per_second": 3.315,
|
| 491 |
"step": 656
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 3.0171428571428573,
|
| 495 |
+
"grad_norm": 5.9896016120910645,
|
| 496 |
+
"learning_rate": 4.434250764525994e-05,
|
| 497 |
+
"loss": 1.1378,
|
| 498 |
"step": 660
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 3.0628571428571427,
|
| 502 |
+
"grad_norm": 5.852768898010254,
|
| 503 |
+
"learning_rate": 4.417261297995243e-05,
|
| 504 |
+
"loss": 1.0542,
|
| 505 |
"step": 670
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 3.1085714285714285,
|
| 509 |
+
"grad_norm": 7.916086673736572,
|
| 510 |
+
"learning_rate": 4.400271831464492e-05,
|
| 511 |
+
"loss": 1.0331,
|
| 512 |
"step": 680
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 3.1542857142857144,
|
| 516 |
+
"grad_norm": 8.959929466247559,
|
| 517 |
+
"learning_rate": 4.383282364933741e-05,
|
| 518 |
+
"loss": 1.0943,
|
| 519 |
"step": 690
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 3.2,
|
| 523 |
+
"grad_norm": 6.69697904586792,
|
| 524 |
+
"learning_rate": 4.36629289840299e-05,
|
| 525 |
+
"loss": 1.0978,
|
| 526 |
"step": 700
|
| 527 |
},
|
| 528 |
{
|
| 529 |
"epoch": 3.2457142857142856,
|
| 530 |
+
"grad_norm": 7.420085430145264,
|
| 531 |
+
"learning_rate": 4.349303431872239e-05,
|
| 532 |
+
"loss": 1.0588,
|
| 533 |
"step": 710
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 3.2914285714285714,
|
| 537 |
+
"grad_norm": 10.829694747924805,
|
| 538 |
+
"learning_rate": 4.332313965341488e-05,
|
| 539 |
+
"loss": 1.0838,
|
| 540 |
"step": 720
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 3.337142857142857,
|
| 544 |
+
"grad_norm": 6.954169273376465,
|
| 545 |
+
"learning_rate": 4.3153244988107375e-05,
|
| 546 |
+
"loss": 1.1126,
|
| 547 |
"step": 730
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 3.382857142857143,
|
| 551 |
+
"grad_norm": 6.538149833679199,
|
| 552 |
+
"learning_rate": 4.298335032279987e-05,
|
| 553 |
+
"loss": 1.1049,
|
| 554 |
"step": 740
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 3.4285714285714284,
|
| 558 |
+
"grad_norm": 11.640555381774902,
|
| 559 |
+
"learning_rate": 4.281345565749236e-05,
|
| 560 |
+
"loss": 1.0577,
|
| 561 |
"step": 750
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 3.474285714285714,
|
| 565 |
+
"grad_norm": 7.022554397583008,
|
| 566 |
+
"learning_rate": 4.264356099218485e-05,
|
| 567 |
+
"loss": 1.1454,
|
| 568 |
"step": 760
|
| 569 |
},
|
| 570 |
{
|
| 571 |
"epoch": 3.52,
|
| 572 |
+
"grad_norm": 9.619643211364746,
|
| 573 |
+
"learning_rate": 4.247366632687734e-05,
|
| 574 |
+
"loss": 1.0374,
|
| 575 |
"step": 770
|
| 576 |
},
|
| 577 |
{
|
| 578 |
"epoch": 3.565714285714286,
|
| 579 |
+
"grad_norm": 8.229997634887695,
|
| 580 |
+
"learning_rate": 4.230377166156983e-05,
|
| 581 |
+
"loss": 1.0354,
|
| 582 |
"step": 780
|
| 583 |
},
|
| 584 |
{
|
| 585 |
"epoch": 3.611428571428571,
|
| 586 |
+
"grad_norm": 8.94089126586914,
|
| 587 |
+
"learning_rate": 4.213387699626232e-05,
|
| 588 |
+
"loss": 1.1144,
|
| 589 |
"step": 790
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 3.657142857142857,
|
| 593 |
+
"grad_norm": 7.949928283691406,
|
| 594 |
+
"learning_rate": 4.196398233095481e-05,
|
| 595 |
+
"loss": 1.0678,
|
| 596 |
"step": 800
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 3.702857142857143,
|
| 600 |
+
"grad_norm": 6.979115009307861,
|
| 601 |
+
"learning_rate": 4.1794087665647304e-05,
|
| 602 |
+
"loss": 1.0945,
|
| 603 |
"step": 810
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 3.7485714285714287,
|
| 607 |
+
"grad_norm": 8.669636726379395,
|
| 608 |
+
"learning_rate": 4.162419300033979e-05,
|
| 609 |
+
"loss": 1.0521,
|
| 610 |
"step": 820
|
| 611 |
},
|
| 612 |
{
|
| 613 |
"epoch": 3.7942857142857145,
|
| 614 |
+
"grad_norm": 6.849328994750977,
|
| 615 |
+
"learning_rate": 4.145429833503228e-05,
|
| 616 |
+
"loss": 1.0513,
|
| 617 |
"step": 830
|
| 618 |
},
|
| 619 |
{
|
| 620 |
"epoch": 3.84,
|
| 621 |
+
"grad_norm": 7.107924461364746,
|
| 622 |
+
"learning_rate": 4.1284403669724776e-05,
|
| 623 |
+
"loss": 1.0477,
|
| 624 |
"step": 840
|
| 625 |
},
|
| 626 |
{
|
| 627 |
"epoch": 3.8857142857142857,
|
| 628 |
+
"grad_norm": 6.533900737762451,
|
| 629 |
+
"learning_rate": 4.111450900441726e-05,
|
| 630 |
+
"loss": 1.0513,
|
| 631 |
"step": 850
|
| 632 |
},
|
| 633 |
{
|
| 634 |
"epoch": 3.9314285714285715,
|
| 635 |
+
"grad_norm": 6.826801776885986,
|
| 636 |
+
"learning_rate": 4.0944614339109755e-05,
|
| 637 |
+
"loss": 1.0932,
|
| 638 |
"step": 860
|
| 639 |
},
|
| 640 |
{
|
| 641 |
"epoch": 3.977142857142857,
|
| 642 |
+
"grad_norm": 6.449113845825195,
|
| 643 |
+
"learning_rate": 4.077471967380224e-05,
|
| 644 |
+
"loss": 1.0464,
|
| 645 |
"step": 870
|
| 646 |
},
|
| 647 |
{
|
| 648 |
"epoch": 4.0,
|
| 649 |
+
"eval_accuracy": 0.7231441048034934,
|
| 650 |
+
"eval_loss": 0.9180014729499817,
|
| 651 |
+
"eval_runtime": 74.9581,
|
| 652 |
+
"eval_samples_per_second": 106.926,
|
| 653 |
+
"eval_steps_per_second": 3.349,
|
| 654 |
"step": 875
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 4.022857142857143,
|
| 658 |
+
"grad_norm": 6.743358135223389,
|
| 659 |
+
"learning_rate": 4.0604825008494734e-05,
|
| 660 |
+
"loss": 1.0845,
|
| 661 |
"step": 880
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 4.0685714285714285,
|
| 665 |
+
"grad_norm": 6.779079437255859,
|
| 666 |
+
"learning_rate": 4.043493034318723e-05,
|
| 667 |
+
"loss": 1.0415,
|
| 668 |
"step": 890
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 4.114285714285714,
|
| 672 |
+
"grad_norm": 8.849162101745605,
|
| 673 |
+
"learning_rate": 4.026503567787971e-05,
|
| 674 |
+
"loss": 1.0487,
|
| 675 |
"step": 900
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 4.16,
|
| 679 |
+
"grad_norm": 7.157181739807129,
|
| 680 |
+
"learning_rate": 4.0095141012572206e-05,
|
| 681 |
+
"loss": 1.0647,
|
| 682 |
"step": 910
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 4.2057142857142855,
|
| 686 |
+
"grad_norm": 8.016735076904297,
|
| 687 |
+
"learning_rate": 3.99252463472647e-05,
|
| 688 |
+
"loss": 1.0493,
|
| 689 |
"step": 920
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 4.251428571428572,
|
| 693 |
+
"grad_norm": 8.14920711517334,
|
| 694 |
+
"learning_rate": 3.9755351681957185e-05,
|
| 695 |
+
"loss": 1.0228,
|
| 696 |
"step": 930
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 4.297142857142857,
|
| 700 |
+
"grad_norm": 6.980301856994629,
|
| 701 |
+
"learning_rate": 3.958545701664968e-05,
|
| 702 |
+
"loss": 0.9977,
|
| 703 |
"step": 940
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 4.3428571428571425,
|
| 707 |
+
"grad_norm": 6.754297256469727,
|
| 708 |
+
"learning_rate": 3.941556235134217e-05,
|
| 709 |
+
"loss": 1.025,
|
| 710 |
"step": 950
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 4.388571428571429,
|
| 714 |
+
"grad_norm": 6.082598686218262,
|
| 715 |
+
"learning_rate": 3.9245667686034656e-05,
|
| 716 |
+
"loss": 0.9734,
|
| 717 |
"step": 960
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 4.434285714285714,
|
| 721 |
+
"grad_norm": 9.236204147338867,
|
| 722 |
+
"learning_rate": 3.907577302072715e-05,
|
| 723 |
+
"loss": 0.9897,
|
| 724 |
"step": 970
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 4.48,
|
| 728 |
+
"grad_norm": 6.07050085067749,
|
| 729 |
+
"learning_rate": 3.890587835541964e-05,
|
| 730 |
+
"loss": 1.0549,
|
| 731 |
"step": 980
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 4.525714285714286,
|
| 735 |
+
"grad_norm": 8.898210525512695,
|
| 736 |
+
"learning_rate": 3.8735983690112135e-05,
|
| 737 |
+
"loss": 1.0703,
|
| 738 |
"step": 990
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 4.571428571428571,
|
| 742 |
+
"grad_norm": 6.365227699279785,
|
| 743 |
+
"learning_rate": 3.856608902480463e-05,
|
| 744 |
+
"loss": 0.9674,
|
| 745 |
"step": 1000
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 4.617142857142857,
|
| 749 |
+
"grad_norm": 6.180821895599365,
|
| 750 |
+
"learning_rate": 3.8396194359497114e-05,
|
| 751 |
+
"loss": 0.9661,
|
| 752 |
"step": 1010
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 4.662857142857143,
|
| 756 |
+
"grad_norm": 7.927034854888916,
|
| 757 |
+
"learning_rate": 3.822629969418961e-05,
|
| 758 |
+
"loss": 1.025,
|
| 759 |
"step": 1020
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 4.708571428571428,
|
| 763 |
+
"grad_norm": 6.064193248748779,
|
| 764 |
+
"learning_rate": 3.80564050288821e-05,
|
| 765 |
+
"loss": 1.0494,
|
| 766 |
"step": 1030
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 4.7542857142857144,
|
| 770 |
+
"grad_norm": 7.2729668617248535,
|
| 771 |
+
"learning_rate": 3.7886510363574586e-05,
|
| 772 |
+
"loss": 1.0002,
|
| 773 |
"step": 1040
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 4.8,
|
| 777 |
+
"grad_norm": 8.703302383422852,
|
| 778 |
+
"learning_rate": 3.771661569826708e-05,
|
| 779 |
+
"loss": 1.001,
|
| 780 |
"step": 1050
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 4.845714285714286,
|
| 784 |
+
"grad_norm": 5.9366583824157715,
|
| 785 |
+
"learning_rate": 3.7546721032959565e-05,
|
| 786 |
+
"loss": 0.8903,
|
| 787 |
"step": 1060
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 4.8914285714285715,
|
| 791 |
+
"grad_norm": 10.52004337310791,
|
| 792 |
+
"learning_rate": 3.737682636765206e-05,
|
| 793 |
+
"loss": 1.0021,
|
| 794 |
"step": 1070
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 4.937142857142857,
|
| 798 |
+
"grad_norm": 6.183305740356445,
|
| 799 |
+
"learning_rate": 3.720693170234455e-05,
|
| 800 |
+
"loss": 0.9574,
|
| 801 |
"step": 1080
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 4.982857142857143,
|
| 805 |
+
"grad_norm": 7.366205215454102,
|
| 806 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 807 |
+
"loss": 0.982,
|
| 808 |
"step": 1090
|
| 809 |
},
|
| 810 |
{
|
| 811 |
+
"epoch": 4.996571428571428,
|
| 812 |
+
"eval_accuracy": 0.743231441048035,
|
| 813 |
+
"eval_loss": 0.8399370312690735,
|
| 814 |
+
"eval_runtime": 75.6387,
|
| 815 |
+
"eval_samples_per_second": 105.964,
|
| 816 |
+
"eval_steps_per_second": 3.318,
|
| 817 |
+
"step": 1093
|
| 818 |
},
|
| 819 |
{
|
| 820 |
+
"epoch": 5.0285714285714285,
|
| 821 |
+
"grad_norm": 6.932504177093506,
|
| 822 |
+
"learning_rate": 3.686714237172953e-05,
|
| 823 |
+
"loss": 1.0264,
|
| 824 |
+
"step": 1100
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 5.074285714285715,
|
| 828 |
+
"grad_norm": 5.84490966796875,
|
| 829 |
+
"learning_rate": 3.669724770642202e-05,
|
| 830 |
+
"loss": 0.9594,
|
| 831 |
+
"step": 1110
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 5.12,
|
| 835 |
+
"grad_norm": 6.8914475440979,
|
| 836 |
+
"learning_rate": 3.652735304111451e-05,
|
| 837 |
+
"loss": 0.918,
|
| 838 |
+
"step": 1120
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 5.1657142857142855,
|
| 842 |
+
"grad_norm": 6.1007161140441895,
|
| 843 |
+
"learning_rate": 3.6357458375807e-05,
|
| 844 |
+
"loss": 1.0036,
|
| 845 |
+
"step": 1130
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 5.211428571428572,
|
| 849 |
+
"grad_norm": 8.02859115600586,
|
| 850 |
+
"learning_rate": 3.6187563710499494e-05,
|
| 851 |
+
"loss": 0.9655,
|
| 852 |
+
"step": 1140
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 5.257142857142857,
|
| 856 |
+
"grad_norm": 7.857163906097412,
|
| 857 |
+
"learning_rate": 3.601766904519198e-05,
|
| 858 |
+
"loss": 0.9567,
|
| 859 |
+
"step": 1150
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 5.3028571428571425,
|
| 863 |
+
"grad_norm": 7.495018005371094,
|
| 864 |
+
"learning_rate": 3.584777437988447e-05,
|
| 865 |
+
"loss": 0.9226,
|
| 866 |
+
"step": 1160
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.348571428571429,
|
| 870 |
+
"grad_norm": 8.65245532989502,
|
| 871 |
+
"learning_rate": 3.567787971457696e-05,
|
| 872 |
+
"loss": 0.8992,
|
| 873 |
+
"step": 1170
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 5.394285714285714,
|
| 877 |
+
"grad_norm": 8.458373069763184,
|
| 878 |
+
"learning_rate": 3.550798504926945e-05,
|
| 879 |
+
"loss": 0.9539,
|
| 880 |
+
"step": 1180
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 5.44,
|
| 884 |
+
"grad_norm": 6.742293357849121,
|
| 885 |
+
"learning_rate": 3.5338090383961945e-05,
|
| 886 |
+
"loss": 0.9382,
|
| 887 |
+
"step": 1190
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 5.485714285714286,
|
| 891 |
+
"grad_norm": 6.219183921813965,
|
| 892 |
+
"learning_rate": 3.516819571865443e-05,
|
| 893 |
+
"loss": 0.9673,
|
| 894 |
+
"step": 1200
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 5.531428571428571,
|
| 898 |
+
"grad_norm": 6.372159004211426,
|
| 899 |
+
"learning_rate": 3.4998301053346924e-05,
|
| 900 |
+
"loss": 0.9558,
|
| 901 |
+
"step": 1210
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 5.577142857142857,
|
| 905 |
+
"grad_norm": 9.132979393005371,
|
| 906 |
+
"learning_rate": 3.482840638803942e-05,
|
| 907 |
+
"loss": 0.9738,
|
| 908 |
+
"step": 1220
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 5.622857142857143,
|
| 912 |
+
"grad_norm": 10.370519638061523,
|
| 913 |
+
"learning_rate": 3.465851172273191e-05,
|
| 914 |
+
"loss": 0.9576,
|
| 915 |
+
"step": 1230
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 5.668571428571429,
|
| 919 |
+
"grad_norm": 6.819110870361328,
|
| 920 |
+
"learning_rate": 3.44886170574244e-05,
|
| 921 |
+
"loss": 0.9086,
|
| 922 |
+
"step": 1240
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 5.714285714285714,
|
| 926 |
+
"grad_norm": 5.857959747314453,
|
| 927 |
+
"learning_rate": 3.431872239211689e-05,
|
| 928 |
+
"loss": 0.8758,
|
| 929 |
+
"step": 1250
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 5.76,
|
| 933 |
+
"grad_norm": 8.476204872131348,
|
| 934 |
+
"learning_rate": 3.414882772680938e-05,
|
| 935 |
+
"loss": 0.9147,
|
| 936 |
+
"step": 1260
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 5.805714285714286,
|
| 940 |
+
"grad_norm": 6.7867865562438965,
|
| 941 |
+
"learning_rate": 3.3978933061501874e-05,
|
| 942 |
+
"loss": 0.9843,
|
| 943 |
+
"step": 1270
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 5.851428571428571,
|
| 947 |
+
"grad_norm": 8.020210266113281,
|
| 948 |
+
"learning_rate": 3.380903839619436e-05,
|
| 949 |
+
"loss": 1.0334,
|
| 950 |
+
"step": 1280
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.897142857142857,
|
| 954 |
+
"grad_norm": 7.1987199783325195,
|
| 955 |
+
"learning_rate": 3.363914373088685e-05,
|
| 956 |
+
"loss": 1.0116,
|
| 957 |
+
"step": 1290
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 5.942857142857143,
|
| 961 |
+
"grad_norm": 6.633023738861084,
|
| 962 |
+
"learning_rate": 3.3469249065579346e-05,
|
| 963 |
+
"loss": 0.9476,
|
| 964 |
+
"step": 1300
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 5.988571428571428,
|
| 968 |
+
"grad_norm": 7.582513332366943,
|
| 969 |
+
"learning_rate": 3.329935440027183e-05,
|
| 970 |
+
"loss": 0.9472,
|
| 971 |
+
"step": 1310
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 5.997714285714285,
|
| 975 |
+
"eval_accuracy": 0.7535870243293824,
|
| 976 |
+
"eval_loss": 0.8126731514930725,
|
| 977 |
+
"eval_runtime": 71.4321,
|
| 978 |
+
"eval_samples_per_second": 112.204,
|
| 979 |
+
"eval_steps_per_second": 3.514,
|
| 980 |
+
"step": 1312
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 6.034285714285715,
|
| 984 |
+
"grad_norm": 8.917913436889648,
|
| 985 |
+
"learning_rate": 3.3129459734964325e-05,
|
| 986 |
+
"loss": 0.8892,
|
| 987 |
+
"step": 1320
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 6.08,
|
| 991 |
+
"grad_norm": 11.498839378356934,
|
| 992 |
+
"learning_rate": 3.295956506965682e-05,
|
| 993 |
+
"loss": 0.8989,
|
| 994 |
+
"step": 1330
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 6.1257142857142854,
|
| 998 |
+
"grad_norm": 6.905866622924805,
|
| 999 |
+
"learning_rate": 3.2789670404349304e-05,
|
| 1000 |
+
"loss": 0.8923,
|
| 1001 |
+
"step": 1340
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 6.171428571428572,
|
| 1005 |
+
"grad_norm": 5.6410417556762695,
|
| 1006 |
+
"learning_rate": 3.26197757390418e-05,
|
| 1007 |
+
"loss": 0.9276,
|
| 1008 |
+
"step": 1350
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 6.217142857142857,
|
| 1012 |
+
"grad_norm": 8.219051361083984,
|
| 1013 |
+
"learning_rate": 3.244988107373428e-05,
|
| 1014 |
+
"loss": 0.9223,
|
| 1015 |
+
"step": 1360
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 6.2628571428571425,
|
| 1019 |
+
"grad_norm": 6.403746128082275,
|
| 1020 |
+
"learning_rate": 3.2279986408426776e-05,
|
| 1021 |
+
"loss": 0.9082,
|
| 1022 |
+
"step": 1370
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 6.308571428571429,
|
| 1026 |
+
"grad_norm": 5.092130184173584,
|
| 1027 |
+
"learning_rate": 3.211009174311927e-05,
|
| 1028 |
+
"loss": 0.8609,
|
| 1029 |
+
"step": 1380
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 6.354285714285714,
|
| 1033 |
+
"grad_norm": 7.074675559997559,
|
| 1034 |
+
"learning_rate": 3.1940197077811755e-05,
|
| 1035 |
+
"loss": 0.908,
|
| 1036 |
+
"step": 1390
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 6.4,
|
| 1040 |
+
"grad_norm": 7.3398332595825195,
|
| 1041 |
+
"learning_rate": 3.177030241250425e-05,
|
| 1042 |
+
"loss": 0.9756,
|
| 1043 |
+
"step": 1400
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 6.445714285714286,
|
| 1047 |
+
"grad_norm": 8.28463363647461,
|
| 1048 |
+
"learning_rate": 3.160040774719674e-05,
|
| 1049 |
+
"loss": 0.9251,
|
| 1050 |
+
"step": 1410
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 6.491428571428571,
|
| 1054 |
+
"grad_norm": 11.502705574035645,
|
| 1055 |
+
"learning_rate": 3.1430513081889227e-05,
|
| 1056 |
+
"loss": 0.8992,
|
| 1057 |
+
"step": 1420
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 6.537142857142857,
|
| 1061 |
+
"grad_norm": 6.942375659942627,
|
| 1062 |
+
"learning_rate": 3.126061841658172e-05,
|
| 1063 |
+
"loss": 0.878,
|
| 1064 |
+
"step": 1430
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.582857142857143,
|
| 1068 |
+
"grad_norm": 6.975459098815918,
|
| 1069 |
+
"learning_rate": 3.1090723751274206e-05,
|
| 1070 |
+
"loss": 0.928,
|
| 1071 |
+
"step": 1440
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 6.628571428571428,
|
| 1075 |
+
"grad_norm": 6.6826372146606445,
|
| 1076 |
+
"learning_rate": 3.09208290859667e-05,
|
| 1077 |
+
"loss": 0.9126,
|
| 1078 |
+
"step": 1450
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 6.674285714285714,
|
| 1082 |
+
"grad_norm": 6.843193054199219,
|
| 1083 |
+
"learning_rate": 3.075093442065919e-05,
|
| 1084 |
+
"loss": 0.94,
|
| 1085 |
+
"step": 1460
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 6.72,
|
| 1089 |
+
"grad_norm": 9.00674819946289,
|
| 1090 |
+
"learning_rate": 3.0581039755351684e-05,
|
| 1091 |
+
"loss": 0.9406,
|
| 1092 |
+
"step": 1470
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 6.765714285714286,
|
| 1096 |
+
"grad_norm": 7.362587928771973,
|
| 1097 |
+
"learning_rate": 3.0411145090044174e-05,
|
| 1098 |
+
"loss": 0.9414,
|
| 1099 |
+
"step": 1480
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 6.811428571428571,
|
| 1103 |
+
"grad_norm": 10.040757179260254,
|
| 1104 |
+
"learning_rate": 3.0241250424736666e-05,
|
| 1105 |
+
"loss": 0.9372,
|
| 1106 |
+
"step": 1490
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 6.857142857142857,
|
| 1110 |
+
"grad_norm": 6.630922794342041,
|
| 1111 |
+
"learning_rate": 3.0071355759429153e-05,
|
| 1112 |
+
"loss": 0.918,
|
| 1113 |
+
"step": 1500
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 6.902857142857143,
|
| 1117 |
+
"grad_norm": 6.996939182281494,
|
| 1118 |
+
"learning_rate": 2.9901461094121645e-05,
|
| 1119 |
+
"loss": 0.8634,
|
| 1120 |
+
"step": 1510
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 6.948571428571428,
|
| 1124 |
+
"grad_norm": 6.364753723144531,
|
| 1125 |
+
"learning_rate": 2.9731566428814138e-05,
|
| 1126 |
+
"loss": 0.94,
|
| 1127 |
+
"step": 1520
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 6.994285714285715,
|
| 1131 |
+
"grad_norm": 7.604902267456055,
|
| 1132 |
+
"learning_rate": 2.9561671763506628e-05,
|
| 1133 |
+
"loss": 0.8751,
|
| 1134 |
+
"step": 1530
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 6.998857142857143,
|
| 1138 |
+
"eval_accuracy": 0.7639426076107299,
|
| 1139 |
+
"eval_loss": 0.7851645350456238,
|
| 1140 |
+
"eval_runtime": 71.3545,
|
| 1141 |
+
"eval_samples_per_second": 112.327,
|
| 1142 |
+
"eval_steps_per_second": 3.518,
|
| 1143 |
+
"step": 1531
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 7.04,
|
| 1147 |
+
"grad_norm": 6.300110816955566,
|
| 1148 |
+
"learning_rate": 2.939177709819912e-05,
|
| 1149 |
+
"loss": 0.9023,
|
| 1150 |
+
"step": 1540
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 7.085714285714285,
|
| 1154 |
+
"grad_norm": 6.873245716094971,
|
| 1155 |
+
"learning_rate": 2.9221882432891607e-05,
|
| 1156 |
+
"loss": 0.8445,
|
| 1157 |
+
"step": 1550
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 7.131428571428572,
|
| 1161 |
+
"grad_norm": 6.225979328155518,
|
| 1162 |
+
"learning_rate": 2.90519877675841e-05,
|
| 1163 |
+
"loss": 0.8429,
|
| 1164 |
+
"step": 1560
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 7.177142857142857,
|
| 1168 |
+
"grad_norm": 9.389466285705566,
|
| 1169 |
+
"learning_rate": 2.8882093102276592e-05,
|
| 1170 |
+
"loss": 0.8686,
|
| 1171 |
+
"step": 1570
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 7.222857142857142,
|
| 1175 |
+
"grad_norm": 6.56587028503418,
|
| 1176 |
+
"learning_rate": 2.871219843696908e-05,
|
| 1177 |
+
"loss": 0.9168,
|
| 1178 |
+
"step": 1580
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 7.268571428571429,
|
| 1182 |
+
"grad_norm": 7.2763447761535645,
|
| 1183 |
+
"learning_rate": 2.854230377166157e-05,
|
| 1184 |
+
"loss": 0.9107,
|
| 1185 |
+
"step": 1590
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 7.314285714285714,
|
| 1189 |
+
"grad_norm": 10.647456169128418,
|
| 1190 |
+
"learning_rate": 2.8372409106354064e-05,
|
| 1191 |
+
"loss": 0.8785,
|
| 1192 |
+
"step": 1600
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 7.36,
|
| 1196 |
+
"grad_norm": 9.847870826721191,
|
| 1197 |
+
"learning_rate": 2.820251444104655e-05,
|
| 1198 |
+
"loss": 0.8282,
|
| 1199 |
+
"step": 1610
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 7.405714285714286,
|
| 1203 |
+
"grad_norm": 6.836136341094971,
|
| 1204 |
+
"learning_rate": 2.8032619775739043e-05,
|
| 1205 |
+
"loss": 0.881,
|
| 1206 |
+
"step": 1620
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 7.451428571428571,
|
| 1210 |
+
"grad_norm": 7.8010687828063965,
|
| 1211 |
+
"learning_rate": 2.7862725110431533e-05,
|
| 1212 |
+
"loss": 0.8969,
|
| 1213 |
+
"step": 1630
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 7.497142857142857,
|
| 1217 |
+
"grad_norm": 8.551609992980957,
|
| 1218 |
+
"learning_rate": 2.7692830445124026e-05,
|
| 1219 |
+
"loss": 0.9146,
|
| 1220 |
+
"step": 1640
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 7.542857142857143,
|
| 1224 |
+
"grad_norm": 6.829668998718262,
|
| 1225 |
+
"learning_rate": 2.7522935779816515e-05,
|
| 1226 |
+
"loss": 0.8377,
|
| 1227 |
+
"step": 1650
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 7.588571428571429,
|
| 1231 |
+
"grad_norm": 5.782020568847656,
|
| 1232 |
+
"learning_rate": 2.7353041114509004e-05,
|
| 1233 |
+
"loss": 0.8956,
|
| 1234 |
+
"step": 1660
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 7.634285714285714,
|
| 1238 |
+
"grad_norm": 7.081970691680908,
|
| 1239 |
+
"learning_rate": 2.7183146449201497e-05,
|
| 1240 |
+
"loss": 0.8707,
|
| 1241 |
+
"step": 1670
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 7.68,
|
| 1245 |
+
"grad_norm": 7.2627739906311035,
|
| 1246 |
+
"learning_rate": 2.701325178389399e-05,
|
| 1247 |
+
"loss": 0.8865,
|
| 1248 |
+
"step": 1680
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 7.725714285714286,
|
| 1252 |
+
"grad_norm": 6.740649700164795,
|
| 1253 |
+
"learning_rate": 2.6843357118586476e-05,
|
| 1254 |
+
"loss": 0.8183,
|
| 1255 |
+
"step": 1690
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 7.771428571428571,
|
| 1259 |
+
"grad_norm": 6.93267822265625,
|
| 1260 |
+
"learning_rate": 2.667346245327897e-05,
|
| 1261 |
+
"loss": 0.8898,
|
| 1262 |
+
"step": 1700
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.817142857142857,
|
| 1266 |
+
"grad_norm": 8.19253921508789,
|
| 1267 |
+
"learning_rate": 2.6503567787971462e-05,
|
| 1268 |
+
"loss": 0.8982,
|
| 1269 |
+
"step": 1710
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 7.862857142857143,
|
| 1273 |
+
"grad_norm": 5.707729816436768,
|
| 1274 |
+
"learning_rate": 2.6333673122663948e-05,
|
| 1275 |
+
"loss": 0.888,
|
| 1276 |
+
"step": 1720
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 7.908571428571428,
|
| 1280 |
+
"grad_norm": 6.924159049987793,
|
| 1281 |
+
"learning_rate": 2.616377845735644e-05,
|
| 1282 |
+
"loss": 0.8658,
|
| 1283 |
+
"step": 1730
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 7.954285714285715,
|
| 1287 |
+
"grad_norm": 8.230816841125488,
|
| 1288 |
+
"learning_rate": 2.5993883792048927e-05,
|
| 1289 |
+
"loss": 0.8772,
|
| 1290 |
+
"step": 1740
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 8.0,
|
| 1294 |
+
"grad_norm": 7.3947906494140625,
|
| 1295 |
+
"learning_rate": 2.582398912674142e-05,
|
| 1296 |
+
"loss": 0.9107,
|
| 1297 |
+
"step": 1750
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 8.0,
|
| 1301 |
+
"eval_accuracy": 0.7713038053649407,
|
| 1302 |
+
"eval_loss": 0.7643583416938782,
|
| 1303 |
+
"eval_runtime": 72.4816,
|
| 1304 |
+
"eval_samples_per_second": 110.58,
|
| 1305 |
+
"eval_steps_per_second": 3.463,
|
| 1306 |
+
"step": 1750
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 8.045714285714286,
|
| 1310 |
+
"grad_norm": 6.424936294555664,
|
| 1311 |
+
"learning_rate": 2.5654094461433913e-05,
|
| 1312 |
+
"loss": 0.8515,
|
| 1313 |
+
"step": 1760
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 8.09142857142857,
|
| 1317 |
+
"grad_norm": 7.372068881988525,
|
| 1318 |
+
"learning_rate": 2.5484199796126402e-05,
|
| 1319 |
+
"loss": 0.7997,
|
| 1320 |
+
"step": 1770
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 8.137142857142857,
|
| 1324 |
+
"grad_norm": 6.683503150939941,
|
| 1325 |
+
"learning_rate": 2.5314305130818895e-05,
|
| 1326 |
+
"loss": 0.8021,
|
| 1327 |
+
"step": 1780
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 8.182857142857143,
|
| 1331 |
+
"grad_norm": 7.226657390594482,
|
| 1332 |
+
"learning_rate": 2.5144410465511388e-05,
|
| 1333 |
+
"loss": 0.8367,
|
| 1334 |
+
"step": 1790
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 8.228571428571428,
|
| 1338 |
+
"grad_norm": 6.432008743286133,
|
| 1339 |
+
"learning_rate": 2.4974515800203874e-05,
|
| 1340 |
+
"loss": 0.8582,
|
| 1341 |
+
"step": 1800
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 8.274285714285714,
|
| 1345 |
+
"grad_norm": 6.051323890686035,
|
| 1346 |
+
"learning_rate": 2.4804621134896367e-05,
|
| 1347 |
+
"loss": 0.8166,
|
| 1348 |
+
"step": 1810
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 8.32,
|
| 1352 |
+
"grad_norm": 10.92369270324707,
|
| 1353 |
+
"learning_rate": 2.4634726469588856e-05,
|
| 1354 |
+
"loss": 0.8797,
|
| 1355 |
+
"step": 1820
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 8.365714285714287,
|
| 1359 |
+
"grad_norm": 9.526762962341309,
|
| 1360 |
+
"learning_rate": 2.4464831804281346e-05,
|
| 1361 |
+
"loss": 0.8032,
|
| 1362 |
+
"step": 1830
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 8.411428571428571,
|
| 1366 |
+
"grad_norm": 7.965165138244629,
|
| 1367 |
+
"learning_rate": 2.4294937138973835e-05,
|
| 1368 |
+
"loss": 0.8453,
|
| 1369 |
+
"step": 1840
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 8.457142857142857,
|
| 1373 |
+
"grad_norm": 7.171777248382568,
|
| 1374 |
+
"learning_rate": 2.4125042473666328e-05,
|
| 1375 |
+
"loss": 0.8668,
|
| 1376 |
+
"step": 1850
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 8.502857142857144,
|
| 1380 |
+
"grad_norm": 7.443463325500488,
|
| 1381 |
+
"learning_rate": 2.3955147808358818e-05,
|
| 1382 |
+
"loss": 0.8412,
|
| 1383 |
+
"step": 1860
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 8.548571428571428,
|
| 1387 |
+
"grad_norm": 5.963488578796387,
|
| 1388 |
+
"learning_rate": 2.3785253143051307e-05,
|
| 1389 |
+
"loss": 0.8126,
|
| 1390 |
+
"step": 1870
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 8.594285714285714,
|
| 1394 |
+
"grad_norm": 7.679189682006836,
|
| 1395 |
+
"learning_rate": 2.36153584777438e-05,
|
| 1396 |
+
"loss": 0.8548,
|
| 1397 |
+
"step": 1880
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 8.64,
|
| 1401 |
+
"grad_norm": 7.266505718231201,
|
| 1402 |
+
"learning_rate": 2.3445463812436293e-05,
|
| 1403 |
+
"loss": 0.8217,
|
| 1404 |
+
"step": 1890
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 8.685714285714285,
|
| 1408 |
+
"grad_norm": 6.643305778503418,
|
| 1409 |
+
"learning_rate": 2.3275569147128782e-05,
|
| 1410 |
+
"loss": 0.8284,
|
| 1411 |
+
"step": 1900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 8.731428571428571,
|
| 1415 |
+
"grad_norm": 6.7400922775268555,
|
| 1416 |
+
"learning_rate": 2.3105674481821272e-05,
|
| 1417 |
+
"loss": 0.8595,
|
| 1418 |
+
"step": 1910
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 8.777142857142858,
|
| 1422 |
+
"grad_norm": 9.218864440917969,
|
| 1423 |
+
"learning_rate": 2.2935779816513765e-05,
|
| 1424 |
+
"loss": 0.8636,
|
| 1425 |
+
"step": 1920
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 8.822857142857142,
|
| 1429 |
+
"grad_norm": 6.2901434898376465,
|
| 1430 |
+
"learning_rate": 2.2765885151206254e-05,
|
| 1431 |
+
"loss": 0.8293,
|
| 1432 |
+
"step": 1930
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 8.868571428571428,
|
| 1436 |
+
"grad_norm": 7.57029390335083,
|
| 1437 |
+
"learning_rate": 2.2595990485898744e-05,
|
| 1438 |
+
"loss": 0.8275,
|
| 1439 |
+
"step": 1940
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 8.914285714285715,
|
| 1443 |
+
"grad_norm": 6.863813400268555,
|
| 1444 |
+
"learning_rate": 2.2426095820591233e-05,
|
| 1445 |
+
"loss": 0.7866,
|
| 1446 |
+
"step": 1950
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 8.96,
|
| 1450 |
+
"grad_norm": 7.338642597198486,
|
| 1451 |
+
"learning_rate": 2.2256201155283726e-05,
|
| 1452 |
+
"loss": 0.8464,
|
| 1453 |
+
"step": 1960
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 8.996571428571428,
|
| 1457 |
+
"eval_accuracy": 0.7830318153462258,
|
| 1458 |
+
"eval_loss": 0.7322039604187012,
|
| 1459 |
+
"eval_runtime": 69.8506,
|
| 1460 |
+
"eval_samples_per_second": 114.745,
|
| 1461 |
+
"eval_steps_per_second": 3.593,
|
| 1462 |
+
"step": 1968
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 9.005714285714285,
|
| 1466 |
+
"grad_norm": 5.92184591293335,
|
| 1467 |
+
"learning_rate": 2.2086306489976216e-05,
|
| 1468 |
+
"loss": 0.7986,
|
| 1469 |
+
"step": 1970
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 9.051428571428572,
|
| 1473 |
+
"grad_norm": 8.04157829284668,
|
| 1474 |
+
"learning_rate": 2.1916411824668705e-05,
|
| 1475 |
+
"loss": 0.8227,
|
| 1476 |
+
"step": 1980
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 9.097142857142858,
|
| 1480 |
+
"grad_norm": 6.019360065460205,
|
| 1481 |
+
"learning_rate": 2.1746517159361194e-05,
|
| 1482 |
+
"loss": 0.7563,
|
| 1483 |
+
"step": 1990
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 9.142857142857142,
|
| 1487 |
+
"grad_norm": 7.545748233795166,
|
| 1488 |
+
"learning_rate": 2.1576622494053687e-05,
|
| 1489 |
+
"loss": 0.8087,
|
| 1490 |
+
"step": 2000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 9.188571428571429,
|
| 1494 |
+
"grad_norm": 7.692215442657471,
|
| 1495 |
+
"learning_rate": 2.140672782874618e-05,
|
| 1496 |
+
"loss": 0.7899,
|
| 1497 |
+
"step": 2010
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 9.234285714285715,
|
| 1501 |
+
"grad_norm": 7.4185566902160645,
|
| 1502 |
+
"learning_rate": 2.123683316343867e-05,
|
| 1503 |
+
"loss": 0.7823,
|
| 1504 |
+
"step": 2020
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 9.28,
|
| 1508 |
+
"grad_norm": 8.05014705657959,
|
| 1509 |
+
"learning_rate": 2.106693849813116e-05,
|
| 1510 |
+
"loss": 0.767,
|
| 1511 |
+
"step": 2030
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 9.325714285714286,
|
| 1515 |
+
"grad_norm": 7.103221893310547,
|
| 1516 |
+
"learning_rate": 2.0897043832823652e-05,
|
| 1517 |
+
"loss": 0.7722,
|
| 1518 |
+
"step": 2040
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 9.371428571428572,
|
| 1522 |
+
"grad_norm": 6.289785861968994,
|
| 1523 |
+
"learning_rate": 2.072714916751614e-05,
|
| 1524 |
+
"loss": 0.7829,
|
| 1525 |
+
"step": 2050
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 9.417142857142856,
|
| 1529 |
+
"grad_norm": 5.672107696533203,
|
| 1530 |
+
"learning_rate": 2.055725450220863e-05,
|
| 1531 |
+
"loss": 0.7582,
|
| 1532 |
+
"step": 2060
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 9.462857142857143,
|
| 1536 |
+
"grad_norm": 7.584166049957275,
|
| 1537 |
+
"learning_rate": 2.038735983690112e-05,
|
| 1538 |
+
"loss": 0.8066,
|
| 1539 |
+
"step": 2070
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 9.508571428571429,
|
| 1543 |
+
"grad_norm": 6.826247215270996,
|
| 1544 |
+
"learning_rate": 2.0217465171593613e-05,
|
| 1545 |
+
"loss": 0.8539,
|
| 1546 |
+
"step": 2080
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 9.554285714285715,
|
| 1550 |
+
"grad_norm": 7.450297832489014,
|
| 1551 |
+
"learning_rate": 2.0047570506286103e-05,
|
| 1552 |
+
"loss": 0.8866,
|
| 1553 |
+
"step": 2090
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 9.6,
|
| 1557 |
+
"grad_norm": 7.801323890686035,
|
| 1558 |
+
"learning_rate": 1.9877675840978592e-05,
|
| 1559 |
+
"loss": 0.7857,
|
| 1560 |
+
"step": 2100
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 9.645714285714286,
|
| 1564 |
+
"grad_norm": 5.812144756317139,
|
| 1565 |
+
"learning_rate": 1.9707781175671085e-05,
|
| 1566 |
+
"loss": 0.7891,
|
| 1567 |
+
"step": 2110
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 9.691428571428572,
|
| 1571 |
+
"grad_norm": 6.948761940002441,
|
| 1572 |
+
"learning_rate": 1.9537886510363575e-05,
|
| 1573 |
+
"loss": 0.8282,
|
| 1574 |
+
"step": 2120
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 9.737142857142857,
|
| 1578 |
+
"grad_norm": 7.478163719177246,
|
| 1579 |
+
"learning_rate": 1.9367991845056068e-05,
|
| 1580 |
+
"loss": 0.7716,
|
| 1581 |
+
"step": 2130
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 9.782857142857143,
|
| 1585 |
+
"grad_norm": 7.083802700042725,
|
| 1586 |
+
"learning_rate": 1.9198097179748557e-05,
|
| 1587 |
+
"loss": 0.8072,
|
| 1588 |
+
"step": 2140
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 9.82857142857143,
|
| 1592 |
+
"grad_norm": 8.924737930297852,
|
| 1593 |
+
"learning_rate": 1.902820251444105e-05,
|
| 1594 |
+
"loss": 0.795,
|
| 1595 |
+
"step": 2150
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 9.874285714285714,
|
| 1599 |
+
"grad_norm": 6.02655553817749,
|
| 1600 |
+
"learning_rate": 1.885830784913354e-05,
|
| 1601 |
+
"loss": 0.8785,
|
| 1602 |
+
"step": 2160
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 9.92,
|
| 1606 |
+
"grad_norm": 6.204999923706055,
|
| 1607 |
+
"learning_rate": 1.868841318382603e-05,
|
| 1608 |
+
"loss": 0.8836,
|
| 1609 |
+
"step": 2170
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 9.965714285714286,
|
| 1613 |
+
"grad_norm": 9.425383567810059,
|
| 1614 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 1615 |
+
"loss": 0.8398,
|
| 1616 |
+
"step": 2180
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 9.997714285714286,
|
| 1620 |
+
"eval_accuracy": 0.7797878976918278,
|
| 1621 |
+
"eval_loss": 0.7243014574050903,
|
| 1622 |
+
"eval_runtime": 70.1724,
|
| 1623 |
+
"eval_samples_per_second": 114.219,
|
| 1624 |
+
"eval_steps_per_second": 3.577,
|
| 1625 |
+
"step": 2187
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 10.01142857142857,
|
| 1629 |
+
"grad_norm": 6.657647132873535,
|
| 1630 |
+
"learning_rate": 1.834862385321101e-05,
|
| 1631 |
+
"loss": 0.7659,
|
| 1632 |
+
"step": 2190
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 10.057142857142857,
|
| 1636 |
+
"grad_norm": 6.205779075622559,
|
| 1637 |
+
"learning_rate": 1.81787291879035e-05,
|
| 1638 |
+
"loss": 0.7612,
|
| 1639 |
+
"step": 2200
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 10.102857142857143,
|
| 1643 |
+
"grad_norm": 4.940152168273926,
|
| 1644 |
+
"learning_rate": 1.800883452259599e-05,
|
| 1645 |
+
"loss": 0.7277,
|
| 1646 |
+
"step": 2210
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 10.14857142857143,
|
| 1650 |
+
"grad_norm": 6.750416278839111,
|
| 1651 |
+
"learning_rate": 1.783893985728848e-05,
|
| 1652 |
+
"loss": 0.7334,
|
| 1653 |
+
"step": 2220
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 10.194285714285714,
|
| 1657 |
+
"grad_norm": 8.511019706726074,
|
| 1658 |
+
"learning_rate": 1.7669045191980972e-05,
|
| 1659 |
+
"loss": 0.846,
|
| 1660 |
+
"step": 2230
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 10.24,
|
| 1664 |
+
"grad_norm": 6.814949989318848,
|
| 1665 |
+
"learning_rate": 1.7499150526673462e-05,
|
| 1666 |
+
"loss": 0.7721,
|
| 1667 |
+
"step": 2240
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 10.285714285714286,
|
| 1671 |
+
"grad_norm": 8.27193546295166,
|
| 1672 |
+
"learning_rate": 1.7329255861365955e-05,
|
| 1673 |
+
"loss": 0.7794,
|
| 1674 |
+
"step": 2250
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 10.331428571428571,
|
| 1678 |
+
"grad_norm": 6.475657939910889,
|
| 1679 |
+
"learning_rate": 1.7159361196058444e-05,
|
| 1680 |
+
"loss": 0.7965,
|
| 1681 |
+
"step": 2260
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 10.377142857142857,
|
| 1685 |
+
"grad_norm": 7.63599967956543,
|
| 1686 |
+
"learning_rate": 1.6989466530750937e-05,
|
| 1687 |
+
"loss": 0.768,
|
| 1688 |
+
"step": 2270
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 10.422857142857143,
|
| 1692 |
+
"grad_norm": 6.7202043533325195,
|
| 1693 |
+
"learning_rate": 1.6819571865443427e-05,
|
| 1694 |
+
"loss": 0.7671,
|
| 1695 |
+
"step": 2280
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 10.468571428571428,
|
| 1699 |
+
"grad_norm": 11.612401962280273,
|
| 1700 |
+
"learning_rate": 1.6649677200135916e-05,
|
| 1701 |
+
"loss": 0.7752,
|
| 1702 |
+
"step": 2290
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 10.514285714285714,
|
| 1706 |
+
"grad_norm": 10.456415176391602,
|
| 1707 |
+
"learning_rate": 1.647978253482841e-05,
|
| 1708 |
+
"loss": 0.8036,
|
| 1709 |
+
"step": 2300
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 10.56,
|
| 1713 |
+
"grad_norm": 5.8782782554626465,
|
| 1714 |
+
"learning_rate": 1.63098878695209e-05,
|
| 1715 |
+
"loss": 0.7652,
|
| 1716 |
+
"step": 2310
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 10.605714285714285,
|
| 1720 |
+
"grad_norm": 8.532390594482422,
|
| 1721 |
+
"learning_rate": 1.6139993204213388e-05,
|
| 1722 |
+
"loss": 0.8296,
|
| 1723 |
+
"step": 2320
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 10.651428571428571,
|
| 1727 |
+
"grad_norm": 8.59847354888916,
|
| 1728 |
+
"learning_rate": 1.5970098538905877e-05,
|
| 1729 |
+
"loss": 0.8239,
|
| 1730 |
+
"step": 2330
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 10.697142857142858,
|
| 1734 |
+
"grad_norm": 7.728725910186768,
|
| 1735 |
+
"learning_rate": 1.580020387359837e-05,
|
| 1736 |
+
"loss": 0.7657,
|
| 1737 |
+
"step": 2340
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 10.742857142857144,
|
| 1741 |
+
"grad_norm": 7.599369049072266,
|
| 1742 |
+
"learning_rate": 1.563030920829086e-05,
|
| 1743 |
+
"loss": 0.734,
|
| 1744 |
+
"step": 2350
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 10.788571428571428,
|
| 1748 |
+
"grad_norm": 7.236623287200928,
|
| 1749 |
+
"learning_rate": 1.546041454298335e-05,
|
| 1750 |
+
"loss": 0.7992,
|
| 1751 |
+
"step": 2360
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 10.834285714285715,
|
| 1755 |
+
"grad_norm": 9.501078605651855,
|
| 1756 |
+
"learning_rate": 1.5290519877675842e-05,
|
| 1757 |
+
"loss": 0.7977,
|
| 1758 |
+
"step": 2370
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 10.88,
|
| 1762 |
+
"grad_norm": 7.179370403289795,
|
| 1763 |
+
"learning_rate": 1.5120625212368333e-05,
|
| 1764 |
+
"loss": 0.8284,
|
| 1765 |
+
"step": 2380
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 10.925714285714285,
|
| 1769 |
+
"grad_norm": 6.750663757324219,
|
| 1770 |
+
"learning_rate": 1.4950730547060823e-05,
|
| 1771 |
+
"loss": 0.8197,
|
| 1772 |
+
"step": 2390
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 10.971428571428572,
|
| 1776 |
+
"grad_norm": 8.519887924194336,
|
| 1777 |
+
"learning_rate": 1.4780835881753314e-05,
|
| 1778 |
+
"loss": 0.7534,
|
| 1779 |
+
"step": 2400
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 10.998857142857142,
|
| 1783 |
+
"eval_accuracy": 0.7845290081097941,
|
| 1784 |
+
"eval_loss": 0.708789050579071,
|
| 1785 |
+
"eval_runtime": 70.0816,
|
| 1786 |
+
"eval_samples_per_second": 114.367,
|
| 1787 |
+
"eval_steps_per_second": 3.582,
|
| 1788 |
+
"step": 2406
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 11.017142857142858,
|
| 1792 |
+
"grad_norm": 8.514420509338379,
|
| 1793 |
+
"learning_rate": 1.4610941216445803e-05,
|
| 1794 |
+
"loss": 0.804,
|
| 1795 |
+
"step": 2410
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 11.062857142857142,
|
| 1799 |
+
"grad_norm": 7.491757869720459,
|
| 1800 |
+
"learning_rate": 1.4441046551138296e-05,
|
| 1801 |
+
"loss": 0.8208,
|
| 1802 |
+
"step": 2420
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 11.108571428571429,
|
| 1806 |
+
"grad_norm": 8.859854698181152,
|
| 1807 |
+
"learning_rate": 1.4271151885830786e-05,
|
| 1808 |
+
"loss": 0.7703,
|
| 1809 |
+
"step": 2430
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 11.154285714285715,
|
| 1813 |
+
"grad_norm": 7.596364498138428,
|
| 1814 |
+
"learning_rate": 1.4101257220523275e-05,
|
| 1815 |
+
"loss": 0.7777,
|
| 1816 |
+
"step": 2440
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 11.2,
|
| 1820 |
+
"grad_norm": 4.977694988250732,
|
| 1821 |
+
"learning_rate": 1.3931362555215766e-05,
|
| 1822 |
+
"loss": 0.7842,
|
| 1823 |
+
"step": 2450
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 11.245714285714286,
|
| 1827 |
+
"grad_norm": 8.050721168518066,
|
| 1828 |
+
"learning_rate": 1.3761467889908258e-05,
|
| 1829 |
+
"loss": 0.7335,
|
| 1830 |
+
"step": 2460
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 11.291428571428572,
|
| 1834 |
+
"grad_norm": 6.315252304077148,
|
| 1835 |
+
"learning_rate": 1.3591573224600749e-05,
|
| 1836 |
+
"loss": 0.7898,
|
| 1837 |
+
"step": 2470
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 11.337142857142856,
|
| 1841 |
+
"grad_norm": 6.380728721618652,
|
| 1842 |
+
"learning_rate": 1.3421678559293238e-05,
|
| 1843 |
+
"loss": 0.7531,
|
| 1844 |
+
"step": 2480
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 11.382857142857143,
|
| 1848 |
+
"grad_norm": 6.871670246124268,
|
| 1849 |
+
"learning_rate": 1.3251783893985731e-05,
|
| 1850 |
+
"loss": 0.7604,
|
| 1851 |
+
"step": 2490
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 11.428571428571429,
|
| 1855 |
+
"grad_norm": 7.661756992340088,
|
| 1856 |
+
"learning_rate": 1.308188922867822e-05,
|
| 1857 |
+
"loss": 0.7772,
|
| 1858 |
+
"step": 2500
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 11.474285714285715,
|
| 1862 |
+
"grad_norm": 8.346705436706543,
|
| 1863 |
+
"learning_rate": 1.291199456337071e-05,
|
| 1864 |
+
"loss": 0.7372,
|
| 1865 |
+
"step": 2510
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 11.52,
|
| 1869 |
+
"grad_norm": 5.545016288757324,
|
| 1870 |
+
"learning_rate": 1.2742099898063201e-05,
|
| 1871 |
+
"loss": 0.7428,
|
| 1872 |
+
"step": 2520
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 11.565714285714286,
|
| 1876 |
+
"grad_norm": 8.73310661315918,
|
| 1877 |
+
"learning_rate": 1.2572205232755694e-05,
|
| 1878 |
+
"loss": 0.7566,
|
| 1879 |
+
"step": 2530
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 11.611428571428572,
|
| 1883 |
+
"grad_norm": 7.3945794105529785,
|
| 1884 |
+
"learning_rate": 1.2402310567448183e-05,
|
| 1885 |
+
"loss": 0.7988,
|
| 1886 |
+
"step": 2540
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 11.657142857142857,
|
| 1890 |
+
"grad_norm": 5.604186534881592,
|
| 1891 |
+
"learning_rate": 1.2232415902140673e-05,
|
| 1892 |
+
"loss": 0.6912,
|
| 1893 |
+
"step": 2550
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 11.702857142857143,
|
| 1897 |
+
"grad_norm": 6.901523590087891,
|
| 1898 |
+
"learning_rate": 1.2062521236833164e-05,
|
| 1899 |
+
"loss": 0.7438,
|
| 1900 |
+
"step": 2560
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 11.748571428571429,
|
| 1904 |
+
"grad_norm": 7.090272426605225,
|
| 1905 |
+
"learning_rate": 1.1892626571525654e-05,
|
| 1906 |
+
"loss": 0.7502,
|
| 1907 |
+
"step": 2570
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 11.794285714285714,
|
| 1911 |
+
"grad_norm": 9.939922332763672,
|
| 1912 |
+
"learning_rate": 1.1722731906218146e-05,
|
| 1913 |
+
"loss": 0.7574,
|
| 1914 |
+
"step": 2580
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 11.84,
|
| 1918 |
+
"grad_norm": 7.832808017730713,
|
| 1919 |
+
"learning_rate": 1.1552837240910636e-05,
|
| 1920 |
+
"loss": 0.7763,
|
| 1921 |
+
"step": 2590
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 11.885714285714286,
|
| 1925 |
+
"grad_norm": 7.921093940734863,
|
| 1926 |
+
"learning_rate": 1.1382942575603127e-05,
|
| 1927 |
+
"loss": 0.8136,
|
| 1928 |
+
"step": 2600
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 11.93142857142857,
|
| 1932 |
+
"grad_norm": 8.13971996307373,
|
| 1933 |
+
"learning_rate": 1.1213047910295617e-05,
|
| 1934 |
+
"loss": 0.7735,
|
| 1935 |
+
"step": 2610
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 11.977142857142857,
|
| 1939 |
+
"grad_norm": 5.870953559875488,
|
| 1940 |
+
"learning_rate": 1.1043153244988108e-05,
|
| 1941 |
+
"loss": 0.7051,
|
| 1942 |
+
"step": 2620
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 12.0,
|
| 1946 |
+
"eval_accuracy": 0.793512164691204,
|
| 1947 |
+
"eval_loss": 0.6982392072677612,
|
| 1948 |
+
"eval_runtime": 74.4508,
|
| 1949 |
+
"eval_samples_per_second": 107.655,
|
| 1950 |
+
"eval_steps_per_second": 3.371,
|
| 1951 |
+
"step": 2625
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 12.022857142857143,
|
| 1955 |
+
"grad_norm": 9.655831336975098,
|
| 1956 |
+
"learning_rate": 1.0873258579680597e-05,
|
| 1957 |
+
"loss": 0.7839,
|
| 1958 |
+
"step": 2630
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 12.06857142857143,
|
| 1962 |
+
"grad_norm": 6.195824146270752,
|
| 1963 |
+
"learning_rate": 1.070336391437309e-05,
|
| 1964 |
+
"loss": 0.7916,
|
| 1965 |
+
"step": 2640
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 12.114285714285714,
|
| 1969 |
+
"grad_norm": 7.92185115814209,
|
| 1970 |
+
"learning_rate": 1.053346924906558e-05,
|
| 1971 |
+
"loss": 0.7016,
|
| 1972 |
+
"step": 2650
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 12.16,
|
| 1976 |
+
"grad_norm": 5.990954875946045,
|
| 1977 |
+
"learning_rate": 1.036357458375807e-05,
|
| 1978 |
+
"loss": 0.6903,
|
| 1979 |
+
"step": 2660
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 12.205714285714286,
|
| 1983 |
+
"grad_norm": 5.883810520172119,
|
| 1984 |
+
"learning_rate": 1.019367991845056e-05,
|
| 1985 |
+
"loss": 0.7879,
|
| 1986 |
+
"step": 2670
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 12.251428571428571,
|
| 1990 |
+
"grad_norm": 6.014761447906494,
|
| 1991 |
+
"learning_rate": 1.0023785253143051e-05,
|
| 1992 |
+
"loss": 0.7569,
|
| 1993 |
+
"step": 2680
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 12.297142857142857,
|
| 1997 |
+
"grad_norm": 6.2539191246032715,
|
| 1998 |
+
"learning_rate": 9.853890587835543e-06,
|
| 1999 |
+
"loss": 0.7547,
|
| 2000 |
+
"step": 2690
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 12.342857142857143,
|
| 2004 |
+
"grad_norm": 8.04623031616211,
|
| 2005 |
+
"learning_rate": 9.683995922528034e-06,
|
| 2006 |
+
"loss": 0.7636,
|
| 2007 |
+
"step": 2700
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 12.388571428571428,
|
| 2011 |
+
"grad_norm": 7.5707106590271,
|
| 2012 |
+
"learning_rate": 9.514101257220525e-06,
|
| 2013 |
+
"loss": 0.7976,
|
| 2014 |
+
"step": 2710
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 12.434285714285714,
|
| 2018 |
+
"grad_norm": 7.271738529205322,
|
| 2019 |
+
"learning_rate": 9.344206591913014e-06,
|
| 2020 |
+
"loss": 0.6919,
|
| 2021 |
+
"step": 2720
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 12.48,
|
| 2025 |
+
"grad_norm": 6.238006591796875,
|
| 2026 |
+
"learning_rate": 9.174311926605506e-06,
|
| 2027 |
+
"loss": 0.7343,
|
| 2028 |
+
"step": 2730
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 12.525714285714285,
|
| 2032 |
+
"grad_norm": 6.735348701477051,
|
| 2033 |
+
"learning_rate": 9.004417261297995e-06,
|
| 2034 |
+
"loss": 0.6853,
|
| 2035 |
+
"step": 2740
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 12.571428571428571,
|
| 2039 |
+
"grad_norm": 7.480915069580078,
|
| 2040 |
+
"learning_rate": 8.834522595990486e-06,
|
| 2041 |
+
"loss": 0.7797,
|
| 2042 |
+
"step": 2750
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 12.617142857142857,
|
| 2046 |
+
"grad_norm": 7.131129741668701,
|
| 2047 |
+
"learning_rate": 8.664627930682977e-06,
|
| 2048 |
+
"loss": 0.7023,
|
| 2049 |
+
"step": 2760
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 12.662857142857142,
|
| 2053 |
+
"grad_norm": 6.145063400268555,
|
| 2054 |
+
"learning_rate": 8.494733265375469e-06,
|
| 2055 |
+
"loss": 0.801,
|
| 2056 |
+
"step": 2770
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 12.708571428571428,
|
| 2060 |
+
"grad_norm": 8.46693229675293,
|
| 2061 |
+
"learning_rate": 8.324838600067958e-06,
|
| 2062 |
+
"loss": 0.7693,
|
| 2063 |
+
"step": 2780
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 12.754285714285714,
|
| 2067 |
+
"grad_norm": 6.422353744506836,
|
| 2068 |
+
"learning_rate": 8.15494393476045e-06,
|
| 2069 |
+
"loss": 0.7248,
|
| 2070 |
+
"step": 2790
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 12.8,
|
| 2074 |
+
"grad_norm": 10.295066833496094,
|
| 2075 |
+
"learning_rate": 7.985049269452939e-06,
|
| 2076 |
+
"loss": 0.8283,
|
| 2077 |
+
"step": 2800
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 12.845714285714285,
|
| 2081 |
+
"grad_norm": 7.5840959548950195,
|
| 2082 |
+
"learning_rate": 7.81515460414543e-06,
|
| 2083 |
+
"loss": 0.7081,
|
| 2084 |
+
"step": 2810
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 12.891428571428571,
|
| 2088 |
+
"grad_norm": 6.741255760192871,
|
| 2089 |
+
"learning_rate": 7.645259938837921e-06,
|
| 2090 |
+
"loss": 0.7079,
|
| 2091 |
+
"step": 2820
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 12.937142857142858,
|
| 2095 |
+
"grad_norm": 7.525110721588135,
|
| 2096 |
+
"learning_rate": 7.475365273530411e-06,
|
| 2097 |
+
"loss": 0.7374,
|
| 2098 |
+
"step": 2830
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 12.982857142857142,
|
| 2102 |
+
"grad_norm": 7.784526348114014,
|
| 2103 |
+
"learning_rate": 7.305470608222902e-06,
|
| 2104 |
+
"loss": 0.7359,
|
| 2105 |
+
"step": 2840
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 12.996571428571428,
|
| 2109 |
+
"eval_accuracy": 0.7916406737367436,
|
| 2110 |
+
"eval_loss": 0.6984859704971313,
|
| 2111 |
+
"eval_runtime": 73.0055,
|
| 2112 |
+
"eval_samples_per_second": 109.786,
|
| 2113 |
+
"eval_steps_per_second": 3.438,
|
| 2114 |
+
"step": 2843
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 13.028571428571428,
|
| 2118 |
+
"grad_norm": 7.309814929962158,
|
| 2119 |
+
"learning_rate": 7.135575942915393e-06,
|
| 2120 |
+
"loss": 0.709,
|
| 2121 |
+
"step": 2850
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 13.074285714285715,
|
| 2125 |
+
"grad_norm": 10.355072021484375,
|
| 2126 |
+
"learning_rate": 6.965681277607883e-06,
|
| 2127 |
+
"loss": 0.7387,
|
| 2128 |
+
"step": 2860
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 13.12,
|
| 2132 |
+
"grad_norm": 7.303361415863037,
|
| 2133 |
+
"learning_rate": 6.795786612300374e-06,
|
| 2134 |
+
"loss": 0.6999,
|
| 2135 |
+
"step": 2870
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 13.165714285714285,
|
| 2139 |
+
"grad_norm": 8.104413032531738,
|
| 2140 |
+
"learning_rate": 6.6258919469928655e-06,
|
| 2141 |
+
"loss": 0.7602,
|
| 2142 |
+
"step": 2880
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 13.211428571428572,
|
| 2146 |
+
"grad_norm": 6.764256000518799,
|
| 2147 |
+
"learning_rate": 6.455997281685355e-06,
|
| 2148 |
+
"loss": 0.7195,
|
| 2149 |
+
"step": 2890
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 13.257142857142856,
|
| 2153 |
+
"grad_norm": 6.039324760437012,
|
| 2154 |
+
"learning_rate": 6.286102616377847e-06,
|
| 2155 |
+
"loss": 0.6927,
|
| 2156 |
+
"step": 2900
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 13.302857142857142,
|
| 2160 |
+
"grad_norm": 5.957076549530029,
|
| 2161 |
+
"learning_rate": 6.1162079510703365e-06,
|
| 2162 |
+
"loss": 0.7369,
|
| 2163 |
+
"step": 2910
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 13.348571428571429,
|
| 2167 |
+
"grad_norm": 6.061888694763184,
|
| 2168 |
+
"learning_rate": 5.946313285762827e-06,
|
| 2169 |
+
"loss": 0.734,
|
| 2170 |
+
"step": 2920
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 13.394285714285715,
|
| 2174 |
+
"grad_norm": 8.270124435424805,
|
| 2175 |
+
"learning_rate": 5.776418620455318e-06,
|
| 2176 |
+
"loss": 0.7389,
|
| 2177 |
+
"step": 2930
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 13.44,
|
| 2181 |
+
"grad_norm": 5.897438049316406,
|
| 2182 |
+
"learning_rate": 5.606523955147808e-06,
|
| 2183 |
+
"loss": 0.772,
|
| 2184 |
+
"step": 2940
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 13.485714285714286,
|
| 2188 |
+
"grad_norm": 10.048014640808105,
|
| 2189 |
+
"learning_rate": 5.436629289840299e-06,
|
| 2190 |
+
"loss": 0.7129,
|
| 2191 |
+
"step": 2950
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 13.531428571428572,
|
| 2195 |
+
"grad_norm": 6.532159805297852,
|
| 2196 |
+
"learning_rate": 5.26673462453279e-06,
|
| 2197 |
+
"loss": 0.7329,
|
| 2198 |
+
"step": 2960
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 13.577142857142857,
|
| 2202 |
+
"grad_norm": 7.861076831817627,
|
| 2203 |
+
"learning_rate": 5.09683995922528e-06,
|
| 2204 |
+
"loss": 0.659,
|
| 2205 |
+
"step": 2970
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 13.622857142857143,
|
| 2209 |
+
"grad_norm": 7.23004150390625,
|
| 2210 |
+
"learning_rate": 4.926945293917771e-06,
|
| 2211 |
+
"loss": 0.7383,
|
| 2212 |
+
"step": 2980
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 13.668571428571429,
|
| 2216 |
+
"grad_norm": 7.737166881561279,
|
| 2217 |
+
"learning_rate": 4.7570506286102625e-06,
|
| 2218 |
+
"loss": 0.7243,
|
| 2219 |
+
"step": 2990
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 13.714285714285714,
|
| 2223 |
+
"grad_norm": 6.211805820465088,
|
| 2224 |
+
"learning_rate": 4.587155963302753e-06,
|
| 2225 |
+
"loss": 0.748,
|
| 2226 |
+
"step": 3000
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 13.76,
|
| 2230 |
+
"grad_norm": 6.718061447143555,
|
| 2231 |
+
"learning_rate": 4.417261297995243e-06,
|
| 2232 |
+
"loss": 0.7005,
|
| 2233 |
+
"step": 3010
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 13.805714285714286,
|
| 2237 |
+
"grad_norm": 6.454185485839844,
|
| 2238 |
+
"learning_rate": 4.247366632687734e-06,
|
| 2239 |
+
"loss": 0.806,
|
| 2240 |
+
"step": 3020
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 13.85142857142857,
|
| 2244 |
+
"grad_norm": 8.608590126037598,
|
| 2245 |
+
"learning_rate": 4.077471967380225e-06,
|
| 2246 |
+
"loss": 0.7342,
|
| 2247 |
+
"step": 3030
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 13.897142857142857,
|
| 2251 |
+
"grad_norm": 6.505426406860352,
|
| 2252 |
+
"learning_rate": 3.907577302072715e-06,
|
| 2253 |
+
"loss": 0.7148,
|
| 2254 |
+
"step": 3040
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 13.942857142857143,
|
| 2258 |
+
"grad_norm": 7.760762691497803,
|
| 2259 |
+
"learning_rate": 3.7376826367652057e-06,
|
| 2260 |
+
"loss": 0.7208,
|
| 2261 |
+
"step": 3050
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 13.98857142857143,
|
| 2265 |
+
"grad_norm": 6.736624240875244,
|
| 2266 |
+
"learning_rate": 3.5677879714576964e-06,
|
| 2267 |
+
"loss": 0.7641,
|
| 2268 |
+
"step": 3060
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 13.997714285714286,
|
| 2272 |
+
"eval_accuracy": 0.796756082345602,
|
| 2273 |
+
"eval_loss": 0.6838445663452148,
|
| 2274 |
+
"eval_runtime": 74.6047,
|
| 2275 |
+
"eval_samples_per_second": 107.433,
|
| 2276 |
+
"eval_steps_per_second": 3.364,
|
| 2277 |
+
"step": 3062
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 14.034285714285714,
|
| 2281 |
+
"grad_norm": 6.665014266967773,
|
| 2282 |
+
"learning_rate": 3.397893306150187e-06,
|
| 2283 |
+
"loss": 0.6983,
|
| 2284 |
+
"step": 3070
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 14.08,
|
| 2288 |
+
"grad_norm": 6.425662994384766,
|
| 2289 |
+
"learning_rate": 3.2279986408426775e-06,
|
| 2290 |
+
"loss": 0.6563,
|
| 2291 |
+
"step": 3080
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 14.125714285714286,
|
| 2295 |
+
"grad_norm": 6.079023838043213,
|
| 2296 |
+
"learning_rate": 3.0581039755351682e-06,
|
| 2297 |
+
"loss": 0.6838,
|
| 2298 |
+
"step": 3090
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 14.17142857142857,
|
| 2302 |
+
"grad_norm": 6.835630416870117,
|
| 2303 |
+
"learning_rate": 2.888209310227659e-06,
|
| 2304 |
+
"loss": 0.6923,
|
| 2305 |
+
"step": 3100
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 14.217142857142857,
|
| 2309 |
+
"grad_norm": 7.853431224822998,
|
| 2310 |
+
"learning_rate": 2.7183146449201493e-06,
|
| 2311 |
+
"loss": 0.7237,
|
| 2312 |
+
"step": 3110
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 14.262857142857143,
|
| 2316 |
+
"grad_norm": 6.308788299560547,
|
| 2317 |
+
"learning_rate": 2.54841997961264e-06,
|
| 2318 |
+
"loss": 0.7469,
|
| 2319 |
+
"step": 3120
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 14.308571428571428,
|
| 2323 |
+
"grad_norm": 7.8088202476501465,
|
| 2324 |
+
"learning_rate": 2.3785253143051312e-06,
|
| 2325 |
+
"loss": 0.73,
|
| 2326 |
+
"step": 3130
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 14.354285714285714,
|
| 2330 |
+
"grad_norm": 6.700632095336914,
|
| 2331 |
+
"learning_rate": 2.2086306489976216e-06,
|
| 2332 |
+
"loss": 0.7349,
|
| 2333 |
+
"step": 3140
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 14.4,
|
| 2337 |
+
"grad_norm": 7.2199320793151855,
|
| 2338 |
+
"learning_rate": 2.0387359836901123e-06,
|
| 2339 |
+
"loss": 0.7093,
|
| 2340 |
+
"step": 3150
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 14.445714285714285,
|
| 2344 |
+
"grad_norm": 5.401243686676025,
|
| 2345 |
+
"learning_rate": 1.8688413183826028e-06,
|
| 2346 |
+
"loss": 0.6662,
|
| 2347 |
+
"step": 3160
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 14.491428571428571,
|
| 2351 |
+
"grad_norm": 6.991000652313232,
|
| 2352 |
+
"learning_rate": 1.6989466530750936e-06,
|
| 2353 |
+
"loss": 0.6859,
|
| 2354 |
+
"step": 3170
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 14.537142857142857,
|
| 2358 |
+
"grad_norm": 5.750977993011475,
|
| 2359 |
+
"learning_rate": 1.5290519877675841e-06,
|
| 2360 |
+
"loss": 0.6917,
|
| 2361 |
+
"step": 3180
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 14.582857142857144,
|
| 2365 |
+
"grad_norm": 8.291132926940918,
|
| 2366 |
+
"learning_rate": 1.3591573224600747e-06,
|
| 2367 |
+
"loss": 0.7237,
|
| 2368 |
+
"step": 3190
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 14.628571428571428,
|
| 2372 |
+
"grad_norm": 8.148652076721191,
|
| 2373 |
+
"learning_rate": 1.1892626571525656e-06,
|
| 2374 |
+
"loss": 0.795,
|
| 2375 |
+
"step": 3200
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 14.674285714285714,
|
| 2379 |
+
"grad_norm": 8.365762710571289,
|
| 2380 |
+
"learning_rate": 1.0193679918450562e-06,
|
| 2381 |
+
"loss": 0.7038,
|
| 2382 |
+
"step": 3210
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 14.72,
|
| 2386 |
+
"grad_norm": 8.118247032165527,
|
| 2387 |
+
"learning_rate": 8.494733265375468e-07,
|
| 2388 |
+
"loss": 0.7334,
|
| 2389 |
+
"step": 3220
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 14.765714285714285,
|
| 2393 |
+
"grad_norm": 7.585115432739258,
|
| 2394 |
+
"learning_rate": 6.795786612300373e-07,
|
| 2395 |
+
"loss": 0.7761,
|
| 2396 |
+
"step": 3230
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 14.811428571428571,
|
| 2400 |
+
"grad_norm": 7.77222204208374,
|
| 2401 |
+
"learning_rate": 5.096839959225281e-07,
|
| 2402 |
+
"loss": 0.6876,
|
| 2403 |
+
"step": 3240
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 14.857142857142858,
|
| 2407 |
+
"grad_norm": 7.689059734344482,
|
| 2408 |
+
"learning_rate": 3.3978933061501866e-07,
|
| 2409 |
+
"loss": 0.6827,
|
| 2410 |
+
"step": 3250
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 14.902857142857142,
|
| 2414 |
+
"grad_norm": 7.31766414642334,
|
| 2415 |
+
"learning_rate": 1.6989466530750933e-07,
|
| 2416 |
+
"loss": 0.7361,
|
| 2417 |
+
"step": 3260
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 14.948571428571428,
|
| 2421 |
+
"grad_norm": 6.6061015129089355,
|
| 2422 |
+
"learning_rate": 0.0,
|
| 2423 |
+
"loss": 0.7372,
|
| 2424 |
+
"step": 3270
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 14.948571428571428,
|
| 2428 |
+
"eval_accuracy": 0.796756082345602,
|
| 2429 |
+
"eval_loss": 0.6781123280525208,
|
| 2430 |
+
"eval_runtime": 74.4538,
|
| 2431 |
+
"eval_samples_per_second": 107.651,
|
| 2432 |
+
"eval_steps_per_second": 3.371,
|
| 2433 |
+
"step": 3270
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 14.948571428571428,
|
| 2437 |
+
"step": 3270,
|
| 2438 |
+
"total_flos": 1.0403963301155365e+19,
|
| 2439 |
+
"train_loss": 0.9863547705729073,
|
| 2440 |
+
"train_runtime": 7901.2254,
|
| 2441 |
+
"train_samples_per_second": 53.135,
|
| 2442 |
"train_steps_per_second": 0.414
|
| 2443 |
}
|
| 2444 |
],
|
| 2445 |
"logging_steps": 10,
|
| 2446 |
+
"max_steps": 3270,
|
| 2447 |
"num_input_tokens_seen": 0,
|
| 2448 |
+
"num_train_epochs": 15,
|
| 2449 |
"save_steps": 500,
|
| 2450 |
"stateful_callbacks": {
|
| 2451 |
"TrainerControl": {
|
|
|
|
| 2459 |
"attributes": {}
|
| 2460 |
}
|
| 2461 |
},
|
| 2462 |
+
"total_flos": 1.0403963301155365e+19,
|
| 2463 |
"train_batch_size": 32,
|
| 2464 |
"trial_name": null,
|
| 2465 |
"trial_params": null
|