End of training
Browse files- README.md +25 -15
- pytorch_model.bin +1 -1
- trainer_state.json +983 -883
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [alexue4/text-normalization-ru-new](https://huggingface.co/alexue4/text-normalization-ru-new) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.
|
| 19 |
- Mean Distance: 0
|
| 20 |
-
- Max Distance:
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
@@ -43,22 +43,32 @@ The following hyperparameters were used during training:
|
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
-
- num_epochs:
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
-
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
-
|:-------------:|:-----:|:------:|:---------------:|:------------
|
| 52 |
-
| 0.
|
| 53 |
-
| 0.
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
|
| 63 |
|
| 64 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [alexue4/text-normalization-ru-new](https://huggingface.co/alexue4/text-normalization-ru-new) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.0405
|
| 19 |
- Mean Distance: 0
|
| 20 |
+
- Max Distance: 6
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
+
- num_epochs: 20
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss | Max Distance | Mean Distance |
|
| 51 |
+
|:-------------:|:-----:|:------:|:---------------:|:------------:|:-------------:|
|
| 52 |
+
| 0.0052 | 1.0 | 14041 | 0.0272 | 9 | 0 |
|
| 53 |
+
| 0.0045 | 2.0 | 28088 | 0.0327 | 0 | 9 |
|
| 54 |
+
| 0.0043 | 3.0 | 42132 | 0.0317 | 0 | 6 |
|
| 55 |
+
| 0.0042 | 4.0 | 56176 | 0.0316 | 0 | 6 |
|
| 56 |
+
| 0.0035 | 5.0 | 70220 | 0.0357 | 0 | 6 |
|
| 57 |
+
| 0.0032 | 6.0 | 84264 | 0.0365 | 0 | 6 |
|
| 58 |
+
| 0.0027 | 7.0 | 98308 | 0.0403 | 0 | 6 |
|
| 59 |
+
| 0.0027 | 8.0 | 112352 | 0.0398 | 0 | 6 |
|
| 60 |
+
| 0.0023 | 9.0 | 126396 | 0.0404 | 0 | 6 |
|
| 61 |
+
| 0.0023 | 10.0 | 140440 | 0.0385 | 0 | 6 |
|
| 62 |
+
| 0.002 | 11.0 | 154484 | 0.0407 | 0 | 6 |
|
| 63 |
+
| 0.0018 | 12.0 | 168528 | 0.0426 | 0 | 9 |
|
| 64 |
+
| 0.0018 | 13.0 | 182572 | 0.0422 | 0 | 6 |
|
| 65 |
+
| 0.0016 | 14.0 | 196616 | 0.0421 | 0 | 6 |
|
| 66 |
+
| 0.0016 | 15.0 | 210660 | 0.0402 | 0 | 6 |
|
| 67 |
+
| 0.0014 | 16.0 | 224704 | 0.0407 | 0 | 6 |
|
| 68 |
+
| 0.0014 | 17.0 | 238748 | 0.0427 | 0 | 6 |
|
| 69 |
+
| 0.0014 | 18.0 | 252792 | 0.0411 | 0 | 6 |
|
| 70 |
+
| 0.0013 | 19.0 | 266836 | 0.0406 | 0 | 6 |
|
| 71 |
+
| 0.0013 | 20.0 | 280880 | 0.0405 | 0 | 6 |
|
| 72 |
|
| 73 |
|
| 74 |
### Framework versions
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 258643461
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5969f5d1763ec4c33ef6b8b477385b4628340069265990d31d83adc01ff3d90
|
| 3 |
size 258643461
|
trainer_state.json
CHANGED
|
@@ -1,1328 +1,1428 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 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": 0.0,
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 0.
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 5.
|
| 20 |
-
"loss": 0.
|
| 21 |
-
"step":
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 1.
|
| 26 |
-
"loss": 0.
|
| 27 |
-
"step":
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss": 0.
|
| 33 |
-
"step":
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 2.
|
| 38 |
-
"loss": 0.
|
| 39 |
-
"step":
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 2.
|
| 44 |
-
"loss": 0.
|
| 45 |
-
"step":
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 3.
|
| 50 |
-
"loss": 0.
|
| 51 |
-
"step":
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 3.
|
| 56 |
-
"loss": 0.
|
| 57 |
-
"step":
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 0.
|
| 63 |
-
"step":
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 4.
|
| 68 |
-
"loss": 0.
|
| 69 |
-
"step":
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
-
"
|
| 74 |
-
"
|
| 75 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"learning_rate": 5.
|
| 80 |
-
"loss": 0.
|
| 81 |
-
"step":
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 0.
|
| 87 |
-
"step":
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch":
|
| 91 |
-
"learning_rate": 6.
|
| 92 |
-
"loss": 0.
|
| 93 |
-
"step":
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 0.
|
| 99 |
-
"step":
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
-
"learning_rate": 7.
|
| 104 |
-
"loss": 0.
|
| 105 |
-
"step":
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
-
"step":
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"learning_rate": 8.
|
| 116 |
-
"loss": 0.
|
| 117 |
-
"step":
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 0.
|
| 123 |
-
"step":
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"learning_rate": 9.
|
| 128 |
-
"loss": 0.
|
| 129 |
-
"step":
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 1.
|
| 133 |
-
"
|
| 134 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 135 |
"eval_mean_distance": 0,
|
| 136 |
-
"eval_runtime":
|
| 137 |
-
"eval_samples_per_second":
|
| 138 |
-
"eval_steps_per_second": 1.
|
| 139 |
-
"step":
|
| 140 |
},
|
| 141 |
{
|
| 142 |
-
"epoch":
|
| 143 |
-
"learning_rate": 9.
|
| 144 |
-
"loss": 0.
|
| 145 |
-
"step":
|
| 146 |
},
|
| 147 |
{
|
| 148 |
-
"epoch":
|
| 149 |
-
"learning_rate": 9.
|
| 150 |
-
"loss": 0.
|
| 151 |
-
"step":
|
| 152 |
},
|
| 153 |
{
|
| 154 |
-
"epoch":
|
| 155 |
-
"learning_rate": 9.
|
| 156 |
-
"loss": 0.
|
| 157 |
-
"step":
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch":
|
| 161 |
-
"learning_rate": 9.
|
| 162 |
-
"loss": 0.
|
| 163 |
-
"step":
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch":
|
| 167 |
-
"learning_rate": 9.
|
| 168 |
-
"loss": 0.
|
| 169 |
-
"step":
|
| 170 |
},
|
| 171 |
{
|
| 172 |
-
"epoch":
|
| 173 |
-
"learning_rate": 9.
|
| 174 |
-
"loss": 0.
|
| 175 |
-
"step":
|
| 176 |
},
|
| 177 |
{
|
| 178 |
-
"epoch":
|
| 179 |
-
"learning_rate": 9.
|
| 180 |
-
"loss": 0.
|
| 181 |
-
"step":
|
| 182 |
},
|
| 183 |
{
|
| 184 |
-
"epoch":
|
| 185 |
-
"learning_rate": 9.
|
| 186 |
-
"loss": 0.
|
| 187 |
-
"step":
|
| 188 |
},
|
| 189 |
{
|
| 190 |
-
"epoch":
|
| 191 |
-
"learning_rate": 9.
|
| 192 |
-
"loss": 0.
|
| 193 |
-
"step":
|
| 194 |
},
|
| 195 |
{
|
| 196 |
-
"epoch":
|
| 197 |
-
"learning_rate": 9.
|
| 198 |
-
"loss": 0.
|
| 199 |
-
"step":
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"
|
| 204 |
-
"
|
| 205 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch":
|
| 209 |
-
"learning_rate": 9.
|
| 210 |
-
"loss": 0.
|
| 211 |
-
"step":
|
| 212 |
},
|
| 213 |
{
|
| 214 |
-
"epoch":
|
| 215 |
-
"learning_rate": 9.
|
| 216 |
-
"loss": 0.
|
| 217 |
-
"step":
|
| 218 |
},
|
| 219 |
{
|
| 220 |
-
"epoch":
|
| 221 |
-
"learning_rate": 9.
|
| 222 |
-
"loss": 0.
|
| 223 |
-
"step":
|
| 224 |
},
|
| 225 |
{
|
| 226 |
-
"epoch":
|
| 227 |
-
"learning_rate": 9.
|
| 228 |
-
"loss": 0.
|
| 229 |
-
"step":
|
| 230 |
},
|
| 231 |
{
|
| 232 |
-
"epoch":
|
| 233 |
-
"learning_rate": 9.
|
| 234 |
-
"loss": 0.
|
| 235 |
-
"step":
|
| 236 |
},
|
| 237 |
{
|
| 238 |
-
"epoch":
|
| 239 |
-
"learning_rate": 9.
|
| 240 |
-
"loss": 0.
|
| 241 |
-
"step":
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch":
|
| 245 |
-
"learning_rate": 9.
|
| 246 |
-
"loss": 0.
|
| 247 |
-
"step":
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch":
|
| 251 |
-
"learning_rate":
|
| 252 |
-
"loss": 0.
|
| 253 |
-
"step":
|
| 254 |
},
|
| 255 |
{
|
| 256 |
-
"epoch":
|
| 257 |
-
"learning_rate": 8.
|
| 258 |
-
"loss": 0.
|
| 259 |
-
"step":
|
| 260 |
},
|
| 261 |
{
|
| 262 |
-
"epoch":
|
| 263 |
-
"
|
| 264 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 265 |
"eval_mean_distance": 0,
|
| 266 |
-
"eval_runtime":
|
| 267 |
-
"eval_samples_per_second":
|
| 268 |
-
"eval_steps_per_second": 1.
|
| 269 |
-
"step":
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch":
|
| 273 |
-
"learning_rate": 8.
|
| 274 |
-
"loss": 0.
|
| 275 |
-
"step":
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch":
|
| 279 |
-
"learning_rate": 8.
|
| 280 |
-
"loss": 0.
|
| 281 |
-
"step":
|
| 282 |
},
|
| 283 |
{
|
| 284 |
-
"epoch":
|
| 285 |
-
"learning_rate": 8.
|
| 286 |
-
"loss": 0.
|
| 287 |
-
"step":
|
| 288 |
},
|
| 289 |
{
|
| 290 |
-
"epoch":
|
| 291 |
-
"learning_rate": 8.
|
| 292 |
-
"loss": 0.
|
| 293 |
-
"step":
|
| 294 |
},
|
| 295 |
{
|
| 296 |
-
"epoch":
|
| 297 |
-
"learning_rate": 8.
|
| 298 |
-
"loss": 0.
|
| 299 |
-
"step":
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch":
|
| 303 |
-
"learning_rate": 8.
|
| 304 |
-
"loss": 0.
|
| 305 |
-
"step":
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch":
|
| 309 |
-
"learning_rate": 8.
|
| 310 |
-
"loss": 0.
|
| 311 |
-
"step":
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch":
|
| 315 |
-
"learning_rate": 8.
|
| 316 |
-
"loss": 0.
|
| 317 |
-
"step":
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch":
|
| 321 |
-
"learning_rate": 8.
|
| 322 |
-
"loss": 0.
|
| 323 |
-
"step":
|
| 324 |
},
|
| 325 |
{
|
| 326 |
-
"epoch":
|
| 327 |
-
"learning_rate": 8.
|
| 328 |
-
"loss": 0.
|
| 329 |
-
"step":
|
| 330 |
},
|
| 331 |
{
|
| 332 |
-
"epoch":
|
| 333 |
-
"
|
| 334 |
-
"
|
| 335 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 336 |
},
|
| 337 |
{
|
| 338 |
-
"epoch":
|
| 339 |
-
"learning_rate": 8.
|
| 340 |
-
"loss": 0.
|
| 341 |
-
"step":
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch":
|
| 345 |
-
"learning_rate": 8.
|
| 346 |
-
"loss": 0.
|
| 347 |
-
"step":
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch":
|
| 351 |
-
"learning_rate": 8.
|
| 352 |
-
"loss": 0.
|
| 353 |
-
"step":
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch":
|
| 357 |
-
"learning_rate": 8.
|
| 358 |
-
"loss": 0.
|
| 359 |
-
"step":
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch":
|
| 363 |
-
"learning_rate": 8.
|
| 364 |
-
"loss": 0.
|
| 365 |
-
"step":
|
| 366 |
},
|
| 367 |
{
|
| 368 |
-
"epoch":
|
| 369 |
-
"learning_rate":
|
| 370 |
-
"loss": 0.
|
| 371 |
-
"step":
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch":
|
| 375 |
-
"learning_rate": 7.
|
| 376 |
-
"loss": 0.
|
| 377 |
-
"step":
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch":
|
| 381 |
-
"learning_rate": 7.
|
| 382 |
-
"loss": 0.
|
| 383 |
-
"step":
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch":
|
| 387 |
-
"learning_rate": 7.
|
| 388 |
-
"loss": 0.
|
| 389 |
-
"step":
|
| 390 |
},
|
| 391 |
{
|
| 392 |
-
"epoch":
|
| 393 |
-
"
|
| 394 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 395 |
"eval_mean_distance": 0,
|
| 396 |
-
"eval_runtime":
|
| 397 |
-
"eval_samples_per_second":
|
| 398 |
-
"eval_steps_per_second": 1.
|
| 399 |
-
"step":
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch":
|
| 403 |
-
"learning_rate": 7.
|
| 404 |
-
"loss": 0.
|
| 405 |
-
"step":
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"learning_rate": 7.
|
| 410 |
-
"loss": 0.
|
| 411 |
-
"step":
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
-
"learning_rate": 7.
|
| 416 |
-
"loss": 0.
|
| 417 |
-
"step":
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch":
|
| 421 |
-
"learning_rate": 7.
|
| 422 |
-
"loss": 0.
|
| 423 |
-
"step":
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"learning_rate": 7.
|
| 428 |
-
"loss": 0.
|
| 429 |
-
"step":
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"learning_rate": 7.
|
| 434 |
-
"loss": 0.
|
| 435 |
-
"step":
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch":
|
| 439 |
-
"learning_rate": 7.
|
| 440 |
-
"loss": 0.
|
| 441 |
-
"step":
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch":
|
| 445 |
-
"learning_rate": 7.
|
| 446 |
-
"loss": 0.
|
| 447 |
-
"step":
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch":
|
| 451 |
-
"learning_rate": 7.
|
| 452 |
-
"loss": 0.
|
| 453 |
-
"step":
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"learning_rate": 7.
|
| 458 |
-
"loss": 0.
|
| 459 |
-
"step":
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"
|
| 464 |
-
"
|
| 465 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"learning_rate": 7.
|
| 470 |
-
"loss": 0.
|
| 471 |
-
"step":
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"learning_rate": 7.
|
| 476 |
-
"loss": 0.
|
| 477 |
-
"step":
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"learning_rate": 7.
|
| 482 |
-
"loss": 0.
|
| 483 |
-
"step":
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss": 0.
|
| 489 |
-
"step":
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch":
|
| 493 |
-
"learning_rate": 6.
|
| 494 |
-
"loss": 0.
|
| 495 |
-
"step":
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"learning_rate": 6.
|
| 500 |
-
"loss": 0.
|
| 501 |
-
"step":
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
-
"learning_rate": 6.
|
| 506 |
-
"loss": 0.
|
| 507 |
-
"step":
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"learning_rate": 6.
|
| 512 |
-
"loss": 0.
|
| 513 |
-
"step":
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
-
"learning_rate": 6.
|
| 518 |
-
"loss": 0.
|
| 519 |
-
"step":
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"
|
| 524 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 525 |
"eval_mean_distance": 0,
|
| 526 |
-
"eval_runtime":
|
| 527 |
-
"eval_samples_per_second":
|
| 528 |
-
"eval_steps_per_second": 1.
|
| 529 |
-
"step":
|
| 530 |
},
|
| 531 |
{
|
| 532 |
-
"epoch":
|
| 533 |
-
"learning_rate": 6.
|
| 534 |
-
"loss": 0.
|
| 535 |
-
"step":
|
| 536 |
},
|
| 537 |
{
|
| 538 |
-
"epoch":
|
| 539 |
-
"learning_rate": 6.
|
| 540 |
-
"loss": 0.
|
| 541 |
-
"step":
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch":
|
| 545 |
-
"learning_rate": 6.
|
| 546 |
-
"loss": 0.
|
| 547 |
-
"step":
|
| 548 |
},
|
| 549 |
{
|
| 550 |
-
"epoch":
|
| 551 |
-
"learning_rate": 6.
|
| 552 |
-
"loss": 0.
|
| 553 |
-
"step":
|
| 554 |
},
|
| 555 |
{
|
| 556 |
-
"epoch":
|
| 557 |
-
"learning_rate": 6.
|
| 558 |
-
"loss": 0.
|
| 559 |
-
"step":
|
| 560 |
},
|
| 561 |
{
|
| 562 |
-
"epoch":
|
| 563 |
-
"learning_rate": 6.
|
| 564 |
-
"loss": 0.
|
| 565 |
-
"step":
|
| 566 |
},
|
| 567 |
{
|
| 568 |
-
"epoch":
|
| 569 |
-
"learning_rate": 6.
|
| 570 |
-
"loss": 0.
|
| 571 |
-
"step":
|
| 572 |
},
|
| 573 |
{
|
| 574 |
-
"epoch":
|
| 575 |
-
"learning_rate": 6.
|
| 576 |
-
"loss": 0.
|
| 577 |
-
"step":
|
| 578 |
},
|
| 579 |
{
|
| 580 |
-
"epoch":
|
| 581 |
-
"learning_rate": 6.
|
| 582 |
-
"loss": 0.
|
| 583 |
-
"step":
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch":
|
| 587 |
-
"learning_rate": 6.
|
| 588 |
-
"loss": 0.
|
| 589 |
-
"step":
|
| 590 |
},
|
| 591 |
{
|
| 592 |
-
"epoch":
|
| 593 |
-
"
|
| 594 |
-
"
|
| 595 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 596 |
},
|
| 597 |
{
|
| 598 |
-
"epoch":
|
| 599 |
-
"learning_rate": 6.
|
| 600 |
-
"loss": 0.
|
| 601 |
-
"step":
|
| 602 |
},
|
| 603 |
{
|
| 604 |
-
"epoch":
|
| 605 |
-
"learning_rate":
|
| 606 |
-
"loss": 0.
|
| 607 |
-
"step":
|
| 608 |
},
|
| 609 |
{
|
| 610 |
-
"epoch":
|
| 611 |
-
"learning_rate": 5.
|
| 612 |
-
"loss": 0.
|
| 613 |
-
"step":
|
| 614 |
},
|
| 615 |
{
|
| 616 |
-
"epoch":
|
| 617 |
-
"learning_rate": 5.
|
| 618 |
-
"loss": 0.
|
| 619 |
-
"step":
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch":
|
| 623 |
-
"learning_rate": 5.
|
| 624 |
-
"loss": 0.
|
| 625 |
-
"step":
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch":
|
| 629 |
-
"learning_rate": 5.
|
| 630 |
-
"loss": 0.
|
| 631 |
-
"step":
|
| 632 |
},
|
| 633 |
{
|
| 634 |
-
"epoch":
|
| 635 |
-
"learning_rate": 5.
|
| 636 |
-
"loss": 0.
|
| 637 |
-
"step":
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch":
|
| 641 |
-
"learning_rate": 5.
|
| 642 |
-
"loss": 0.
|
| 643 |
-
"step":
|
| 644 |
},
|
| 645 |
{
|
| 646 |
-
"epoch":
|
| 647 |
-
"learning_rate": 5.
|
| 648 |
-
"loss": 0.
|
| 649 |
-
"step":
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch":
|
| 653 |
-
"
|
| 654 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 655 |
"eval_mean_distance": 0,
|
| 656 |
-
"eval_runtime":
|
| 657 |
-
"eval_samples_per_second":
|
| 658 |
-
"eval_steps_per_second": 1.
|
| 659 |
-
"step":
|
| 660 |
},
|
| 661 |
{
|
| 662 |
-
"epoch":
|
| 663 |
-
"learning_rate": 5.
|
| 664 |
-
"loss": 0.
|
| 665 |
-
"step":
|
| 666 |
},
|
| 667 |
{
|
| 668 |
-
"epoch":
|
| 669 |
-
"learning_rate": 5.
|
| 670 |
-
"loss": 0.
|
| 671 |
-
"step":
|
| 672 |
},
|
| 673 |
{
|
| 674 |
-
"epoch":
|
| 675 |
-
"learning_rate": 5.
|
| 676 |
-
"loss": 0.
|
| 677 |
-
"step":
|
| 678 |
},
|
| 679 |
{
|
| 680 |
-
"epoch":
|
| 681 |
-
"learning_rate": 5.
|
| 682 |
-
"loss": 0.
|
| 683 |
-
"step":
|
| 684 |
},
|
| 685 |
{
|
| 686 |
-
"epoch":
|
| 687 |
-
"learning_rate": 5.
|
| 688 |
-
"loss": 0.
|
| 689 |
-
"step":
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"learning_rate": 5.
|
| 694 |
-
"loss": 0.
|
| 695 |
-
"step":
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch":
|
| 699 |
-
"learning_rate": 5.
|
| 700 |
-
"loss": 0.
|
| 701 |
-
"step":
|
| 702 |
},
|
| 703 |
{
|
| 704 |
-
"epoch":
|
| 705 |
-
"learning_rate": 5.
|
| 706 |
-
"loss": 0.
|
| 707 |
-
"step":
|
| 708 |
},
|
| 709 |
{
|
| 710 |
-
"epoch":
|
| 711 |
-
"learning_rate": 5.
|
| 712 |
-
"loss": 0.
|
| 713 |
-
"step":
|
| 714 |
},
|
| 715 |
{
|
| 716 |
-
"epoch":
|
| 717 |
-
"learning_rate": 5.
|
| 718 |
-
"loss": 0.
|
| 719 |
-
"step":
|
| 720 |
},
|
| 721 |
{
|
| 722 |
-
"epoch":
|
| 723 |
-
"
|
| 724 |
-
"
|
| 725 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 726 |
},
|
| 727 |
{
|
| 728 |
-
"epoch":
|
| 729 |
-
"learning_rate": 4.
|
| 730 |
-
"loss": 0.
|
| 731 |
-
"step":
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch":
|
| 735 |
-
"learning_rate": 4.
|
| 736 |
-
"loss": 0.
|
| 737 |
-
"step":
|
| 738 |
},
|
| 739 |
{
|
| 740 |
-
"epoch":
|
| 741 |
-
"learning_rate": 4.
|
| 742 |
-
"loss": 0.
|
| 743 |
-
"step":
|
| 744 |
},
|
| 745 |
{
|
| 746 |
-
"epoch":
|
| 747 |
-
"learning_rate": 4.
|
| 748 |
-
"loss": 0.
|
| 749 |
-
"step":
|
| 750 |
},
|
| 751 |
{
|
| 752 |
-
"epoch":
|
| 753 |
-
"learning_rate": 4.
|
| 754 |
-
"loss": 0.
|
| 755 |
-
"step":
|
| 756 |
},
|
| 757 |
{
|
| 758 |
-
"epoch":
|
| 759 |
-
"learning_rate": 4.
|
| 760 |
-
"loss": 0.
|
| 761 |
-
"step":
|
| 762 |
},
|
| 763 |
{
|
| 764 |
-
"epoch":
|
| 765 |
-
"learning_rate": 4.
|
| 766 |
-
"loss": 0.
|
| 767 |
-
"step":
|
| 768 |
},
|
| 769 |
{
|
| 770 |
-
"epoch":
|
| 771 |
-
"learning_rate": 4.
|
| 772 |
-
"loss": 0.
|
| 773 |
-
"step":
|
| 774 |
},
|
| 775 |
{
|
| 776 |
-
"epoch":
|
| 777 |
-
"learning_rate": 4.
|
| 778 |
-
"loss": 0.
|
| 779 |
-
"step":
|
| 780 |
},
|
| 781 |
{
|
| 782 |
-
"epoch":
|
| 783 |
-
"
|
| 784 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 785 |
"eval_mean_distance": 0,
|
| 786 |
-
"eval_runtime":
|
| 787 |
-
"eval_samples_per_second":
|
| 788 |
-
"eval_steps_per_second": 1.
|
| 789 |
-
"step":
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch":
|
| 793 |
-
"learning_rate": 4.
|
| 794 |
-
"loss": 0.
|
| 795 |
-
"step":
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch":
|
| 799 |
-
"learning_rate": 4.
|
| 800 |
-
"loss": 0.
|
| 801 |
-
"step":
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch":
|
| 805 |
-
"learning_rate": 4.
|
| 806 |
-
"loss": 0.
|
| 807 |
-
"step":
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch":
|
| 811 |
-
"learning_rate": 4.
|
| 812 |
-
"loss": 0.
|
| 813 |
-
"step":
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch":
|
| 817 |
-
"learning_rate": 4.
|
| 818 |
-
"loss": 0.
|
| 819 |
-
"step":
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch":
|
| 823 |
-
"learning_rate": 4.
|
| 824 |
-
"loss": 0.
|
| 825 |
-
"step":
|
| 826 |
},
|
| 827 |
{
|
| 828 |
-
"epoch":
|
| 829 |
-
"learning_rate": 4.
|
| 830 |
-
"loss": 0.
|
| 831 |
-
"step":
|
| 832 |
},
|
| 833 |
{
|
| 834 |
-
"epoch":
|
| 835 |
-
"learning_rate": 4.
|
| 836 |
-
"loss": 0.
|
| 837 |
-
"step":
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch":
|
| 841 |
-
"learning_rate": 3.
|
| 842 |
-
"loss": 0.
|
| 843 |
-
"step":
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch":
|
| 847 |
-
"learning_rate": 3.
|
| 848 |
-
"loss": 0.
|
| 849 |
-
"step":
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch":
|
| 853 |
-
"
|
| 854 |
-
"
|
| 855 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch":
|
| 859 |
-
"learning_rate": 3.
|
| 860 |
-
"loss": 0.
|
| 861 |
-
"step":
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch":
|
| 865 |
-
"learning_rate": 3.
|
| 866 |
-
"loss": 0.
|
| 867 |
-
"step":
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch":
|
| 871 |
-
"learning_rate": 3.
|
| 872 |
-
"loss": 0.
|
| 873 |
-
"step":
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch":
|
| 877 |
-
"learning_rate": 3.
|
| 878 |
-
"loss": 0.
|
| 879 |
-
"step":
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch":
|
| 883 |
-
"learning_rate": 3.
|
| 884 |
-
"loss": 0.
|
| 885 |
-
"step":
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch":
|
| 889 |
-
"learning_rate": 3.
|
| 890 |
-
"loss": 0.
|
| 891 |
-
"step":
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch":
|
| 895 |
-
"learning_rate": 3.
|
| 896 |
-
"loss": 0.
|
| 897 |
-
"step":
|
| 898 |
},
|
| 899 |
{
|
| 900 |
-
"epoch":
|
| 901 |
-
"learning_rate": 3.
|
| 902 |
-
"loss": 0.
|
| 903 |
-
"step":
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch":
|
| 907 |
-
"learning_rate": 3.
|
| 908 |
-
"loss": 0.
|
| 909 |
-
"step":
|
| 910 |
},
|
| 911 |
{
|
| 912 |
-
"epoch":
|
| 913 |
-
"
|
| 914 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 915 |
"eval_mean_distance": 0,
|
| 916 |
-
"eval_runtime":
|
| 917 |
-
"eval_samples_per_second":
|
| 918 |
-
"eval_steps_per_second": 1.
|
| 919 |
-
"step":
|
| 920 |
},
|
| 921 |
{
|
| 922 |
-
"epoch":
|
| 923 |
-
"learning_rate": 3.
|
| 924 |
-
"loss": 0.
|
| 925 |
-
"step":
|
| 926 |
},
|
| 927 |
{
|
| 928 |
-
"epoch":
|
| 929 |
-
"learning_rate": 3.
|
| 930 |
-
"loss": 0.
|
| 931 |
-
"step":
|
| 932 |
},
|
| 933 |
{
|
| 934 |
-
"epoch":
|
| 935 |
-
"learning_rate": 3.
|
| 936 |
-
"loss": 0.
|
| 937 |
-
"step":
|
| 938 |
},
|
| 939 |
{
|
| 940 |
-
"epoch":
|
| 941 |
-
"learning_rate": 3.
|
| 942 |
-
"loss": 0.
|
| 943 |
-
"step":
|
| 944 |
},
|
| 945 |
{
|
| 946 |
-
"epoch":
|
| 947 |
-
"learning_rate": 3.
|
| 948 |
-
"loss": 0.
|
| 949 |
-
"step":
|
| 950 |
},
|
| 951 |
{
|
| 952 |
-
"epoch":
|
| 953 |
-
"learning_rate": 3.
|
| 954 |
-
"loss": 0.
|
| 955 |
-
"step":
|
| 956 |
},
|
| 957 |
{
|
| 958 |
-
"epoch":
|
| 959 |
-
"learning_rate": 2.
|
| 960 |
-
"loss": 0.
|
| 961 |
-
"step":
|
| 962 |
},
|
| 963 |
{
|
| 964 |
-
"epoch":
|
| 965 |
-
"learning_rate": 2.
|
| 966 |
-
"loss": 0.
|
| 967 |
-
"step":
|
| 968 |
},
|
| 969 |
{
|
| 970 |
-
"epoch":
|
| 971 |
-
"learning_rate": 2.
|
| 972 |
-
"loss": 0.
|
| 973 |
-
"step":
|
| 974 |
},
|
| 975 |
{
|
| 976 |
-
"epoch":
|
| 977 |
-
"learning_rate": 2.
|
| 978 |
-
"loss": 0.
|
| 979 |
-
"step":
|
| 980 |
},
|
| 981 |
{
|
| 982 |
-
"epoch":
|
| 983 |
-
"
|
| 984 |
-
"
|
| 985 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 986 |
},
|
| 987 |
{
|
| 988 |
-
"epoch":
|
| 989 |
-
"learning_rate": 2.
|
| 990 |
-
"loss": 0.
|
| 991 |
-
"step":
|
| 992 |
},
|
| 993 |
{
|
| 994 |
-
"epoch":
|
| 995 |
-
"learning_rate": 2.
|
| 996 |
-
"loss": 0.
|
| 997 |
-
"step":
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
-
"epoch":
|
| 1001 |
-
"learning_rate": 2.
|
| 1002 |
-
"loss": 0.
|
| 1003 |
-
"step":
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
-
"epoch":
|
| 1007 |
-
"learning_rate": 2.
|
| 1008 |
-
"loss": 0.
|
| 1009 |
-
"step":
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
-
"epoch":
|
| 1013 |
-
"learning_rate": 2.
|
| 1014 |
-
"loss": 0.
|
| 1015 |
-
"step":
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
-
"epoch":
|
| 1019 |
-
"learning_rate": 2.
|
| 1020 |
-
"loss": 0.
|
| 1021 |
-
"step":
|
| 1022 |
},
|
| 1023 |
{
|
| 1024 |
-
"epoch":
|
| 1025 |
-
"learning_rate": 2.
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"step":
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
-
"epoch":
|
| 1031 |
-
"learning_rate": 2.
|
| 1032 |
-
"loss": 0.
|
| 1033 |
-
"step":
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
-
"epoch":
|
| 1037 |
-
"learning_rate": 2.
|
| 1038 |
-
"loss": 0.
|
| 1039 |
-
"step":
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
-
"epoch":
|
| 1043 |
-
"
|
| 1044 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1045 |
"eval_mean_distance": 0,
|
| 1046 |
-
"eval_runtime":
|
| 1047 |
-
"eval_samples_per_second":
|
| 1048 |
-
"eval_steps_per_second": 1.
|
| 1049 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
-
"epoch":
|
| 1053 |
-
"learning_rate": 2.
|
| 1054 |
-
"loss": 0.
|
| 1055 |
-
"step":
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
-
"epoch":
|
| 1059 |
-
"learning_rate": 2.
|
| 1060 |
-
"loss": 0.
|
| 1061 |
-
"step":
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
-
"epoch":
|
| 1065 |
-
"learning_rate": 2.
|
| 1066 |
-
"loss": 0.
|
| 1067 |
-
"step":
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch":
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
-
"step":
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch":
|
| 1077 |
-
"learning_rate": 1.
|
| 1078 |
-
"loss": 0.
|
| 1079 |
-
"step":
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
-
"epoch":
|
| 1083 |
-
"learning_rate": 1.
|
| 1084 |
-
"loss": 0.
|
| 1085 |
-
"step":
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
-
"epoch":
|
| 1089 |
-
"learning_rate": 1.
|
| 1090 |
-
"loss": 0.
|
| 1091 |
-
"step":
|
| 1092 |
},
|
| 1093 |
{
|
| 1094 |
-
"epoch":
|
| 1095 |
-
"learning_rate": 1.
|
| 1096 |
-
"loss": 0.
|
| 1097 |
-
"step":
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
-
"epoch":
|
| 1101 |
-
"learning_rate": 1.
|
| 1102 |
-
"loss": 0.
|
| 1103 |
-
"step":
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
-
"epoch":
|
| 1107 |
-
"
|
| 1108 |
-
"
|
| 1109 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
-
"epoch":
|
| 1113 |
-
"learning_rate": 1.
|
| 1114 |
-
"loss": 0.
|
| 1115 |
-
"step":
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch":
|
| 1119 |
-
"learning_rate": 1.
|
| 1120 |
-
"loss": 0.
|
| 1121 |
-
"step":
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
-
"epoch":
|
| 1125 |
-
"learning_rate": 1.
|
| 1126 |
-
"loss": 0.
|
| 1127 |
-
"step":
|
| 1128 |
},
|
| 1129 |
{
|
| 1130 |
-
"epoch":
|
| 1131 |
-
"learning_rate": 1.
|
| 1132 |
-
"loss": 0.
|
| 1133 |
-
"step":
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
-
"epoch":
|
| 1137 |
-
"learning_rate": 1.
|
| 1138 |
-
"loss": 0.
|
| 1139 |
-
"step":
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
-
"epoch":
|
| 1143 |
-
"learning_rate": 1.
|
| 1144 |
-
"loss": 0.
|
| 1145 |
-
"step":
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
-
"epoch":
|
| 1149 |
-
"learning_rate": 1.
|
| 1150 |
-
"loss": 0.
|
| 1151 |
-
"step":
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
-
"epoch":
|
| 1155 |
-
"learning_rate": 1.
|
| 1156 |
-
"loss": 0.
|
| 1157 |
-
"step":
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch":
|
| 1161 |
-
"learning_rate": 1.
|
| 1162 |
-
"loss": 0.
|
| 1163 |
-
"step":
|
| 1164 |
},
|
| 1165 |
{
|
| 1166 |
-
"epoch":
|
| 1167 |
-
"learning_rate": 1.
|
| 1168 |
-
"loss": 0.
|
| 1169 |
-
"step":
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
-
"epoch":
|
| 1173 |
-
"eval_loss": 0.
|
| 1174 |
-
"eval_max_distance":
|
| 1175 |
"eval_mean_distance": 0,
|
| 1176 |
-
"eval_runtime":
|
| 1177 |
-
"eval_samples_per_second":
|
| 1178 |
-
"eval_steps_per_second": 1.
|
| 1179 |
-
"step":
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch":
|
| 1183 |
-
"learning_rate": 1.
|
| 1184 |
-
"loss": 0.
|
| 1185 |
-
"step":
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch":
|
| 1189 |
-
"learning_rate": 1.
|
| 1190 |
-
"loss": 0.
|
| 1191 |
-
"step":
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
-
"epoch":
|
| 1195 |
-
"learning_rate": 9.
|
| 1196 |
-
"loss": 0.
|
| 1197 |
-
"step":
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
-
"epoch":
|
| 1201 |
-
"learning_rate": 9.
|
| 1202 |
-
"loss": 0.
|
| 1203 |
-
"step":
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
-
"epoch":
|
| 1207 |
-
"learning_rate": 8.
|
| 1208 |
-
"loss": 0.
|
| 1209 |
-
"step":
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
-
"epoch":
|
| 1213 |
-
"learning_rate": 8.
|
| 1214 |
-
"loss": 0.
|
| 1215 |
-
"step":
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
-
"epoch":
|
| 1219 |
-
"learning_rate": 7.
|
| 1220 |
-
"loss": 0.
|
| 1221 |
-
"step":
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
-
"epoch":
|
| 1225 |
-
"learning_rate": 7.
|
| 1226 |
-
"loss": 0.
|
| 1227 |
-
"step":
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
-
"epoch":
|
| 1231 |
-
"learning_rate": 6.
|
| 1232 |
-
"loss": 0.
|
| 1233 |
-
"step":
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
-
"epoch":
|
| 1237 |
-
"learning_rate": 6.
|
| 1238 |
-
"loss": 0.
|
| 1239 |
-
"step":
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
-
"epoch":
|
| 1243 |
-
"
|
| 1244 |
-
"
|
| 1245 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
-
"epoch":
|
| 1249 |
-
"learning_rate":
|
| 1250 |
-
"loss": 0.
|
| 1251 |
-
"step":
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
-
"epoch":
|
| 1255 |
-
"learning_rate": 4.
|
| 1256 |
-
"loss": 0.
|
| 1257 |
-
"step":
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
-
"epoch":
|
| 1261 |
-
"learning_rate":
|
| 1262 |
-
"loss": 0.
|
| 1263 |
-
"step":
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
-
"epoch":
|
| 1267 |
-
"learning_rate": 3.
|
| 1268 |
-
"loss": 0.
|
| 1269 |
-
"step":
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
-
"epoch":
|
| 1273 |
-
"learning_rate":
|
| 1274 |
-
"loss": 0.
|
| 1275 |
-
"step":
|
| 1276 |
},
|
| 1277 |
{
|
| 1278 |
-
"epoch":
|
| 1279 |
-
"learning_rate": 2.
|
| 1280 |
-
"loss": 0.
|
| 1281 |
-
"step":
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
-
"epoch":
|
| 1285 |
-
"learning_rate":
|
| 1286 |
-
"loss": 0.
|
| 1287 |
-
"step":
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
-
"epoch":
|
| 1291 |
-
"learning_rate": 1.
|
| 1292 |
-
"loss": 0.
|
| 1293 |
-
"step":
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
-
"epoch":
|
| 1297 |
-
"learning_rate":
|
| 1298 |
-
"loss": 0.
|
| 1299 |
-
"step":
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
-
"epoch":
|
| 1303 |
-
"
|
| 1304 |
-
"
|
| 1305 |
-
"
|
| 1306 |
-
"eval_runtime": 15.7597,
|
| 1307 |
-
"eval_samples_per_second": 15.863,
|
| 1308 |
-
"eval_steps_per_second": 1.079,
|
| 1309 |
-
"step": 237010
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
-
"epoch":
|
| 1313 |
-
"
|
| 1314 |
-
"
|
| 1315 |
-
"
|
| 1316 |
-
"
|
| 1317 |
-
"
|
| 1318 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1319 |
}
|
| 1320 |
],
|
| 1321 |
-
"logging_steps":
|
| 1322 |
-
"max_steps":
|
| 1323 |
-
"num_train_epochs":
|
| 1324 |
-
"save_steps":
|
| 1325 |
-
"total_flos":
|
| 1326 |
"trial_name": null,
|
| 1327 |
"trial_params": null
|
| 1328 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 280880,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 3.5609999287800016e-09,
|
| 14 |
+
"loss": 0.0011,
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 5.003204899935903e-06,
|
| 20 |
+
"loss": 0.0046,
|
| 21 |
+
"step": 1405
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.2,
|
| 25 |
+
"learning_rate": 1.0006409799871805e-05,
|
| 26 |
+
"loss": 0.005,
|
| 27 |
+
"step": 2810
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.3,
|
| 31 |
+
"learning_rate": 1.5009614699807706e-05,
|
| 32 |
+
"loss": 0.0052,
|
| 33 |
+
"step": 4215
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.4,
|
| 37 |
+
"learning_rate": 2.001281959974361e-05,
|
| 38 |
+
"loss": 0.005,
|
| 39 |
+
"step": 5620
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.5,
|
| 43 |
+
"learning_rate": 2.5016024499679513e-05,
|
| 44 |
+
"loss": 0.0048,
|
| 45 |
+
"step": 7025
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.6,
|
| 49 |
+
"learning_rate": 3.001922939961541e-05,
|
| 50 |
+
"loss": 0.0045,
|
| 51 |
+
"step": 8430
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.7,
|
| 55 |
+
"learning_rate": 3.5022434299551316e-05,
|
| 56 |
+
"loss": 0.0049,
|
| 57 |
+
"step": 9835
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.8,
|
| 61 |
+
"learning_rate": 4.002563919948722e-05,
|
| 62 |
+
"loss": 0.0051,
|
| 63 |
+
"step": 11240
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.9,
|
| 67 |
+
"learning_rate": 4.502884409942312e-05,
|
| 68 |
+
"loss": 0.0052,
|
| 69 |
+
"step": 12645
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 1.0,
|
| 73 |
+
"eval_loss": 0.027173755690455437,
|
| 74 |
+
"eval_max_distance": 9,
|
| 75 |
+
"eval_mean_distance": 0,
|
| 76 |
+
"eval_runtime": 13.945,
|
| 77 |
+
"eval_samples_per_second": 17.928,
|
| 78 |
+
"eval_steps_per_second": 1.219,
|
| 79 |
+
"step": 14041
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 1.0,
|
| 83 |
+
"learning_rate": 5.0032048999359025e-05,
|
| 84 |
+
"loss": 0.0045,
|
| 85 |
+
"step": 14050
|
| 86 |
},
|
| 87 |
{
|
| 88 |
+
"epoch": 1.1,
|
| 89 |
+
"learning_rate": 5.5035253899294924e-05,
|
| 90 |
+
"loss": 0.0047,
|
| 91 |
+
"step": 15455
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 1.2,
|
| 95 |
+
"learning_rate": 6.003845879923082e-05,
|
| 96 |
+
"loss": 0.0044,
|
| 97 |
+
"step": 16860
|
| 98 |
},
|
| 99 |
{
|
| 100 |
+
"epoch": 1.3,
|
| 101 |
+
"learning_rate": 6.504166369916673e-05,
|
| 102 |
+
"loss": 0.0045,
|
| 103 |
+
"step": 18265
|
| 104 |
},
|
| 105 |
{
|
| 106 |
+
"epoch": 1.4,
|
| 107 |
+
"learning_rate": 7.002990600968385e-05,
|
| 108 |
+
"loss": 0.0046,
|
| 109 |
+
"step": 19670
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 1.5,
|
| 113 |
+
"learning_rate": 7.50320421532327e-05,
|
| 114 |
+
"loss": 0.0043,
|
| 115 |
+
"step": 21075
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 1.6,
|
| 119 |
+
"learning_rate": 8.003417829678155e-05,
|
| 120 |
+
"loss": 0.0042,
|
| 121 |
+
"step": 22480
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 1.7,
|
| 125 |
+
"learning_rate": 8.503631444033039e-05,
|
| 126 |
+
"loss": 0.0046,
|
| 127 |
+
"step": 23885
|
| 128 |
},
|
| 129 |
{
|
| 130 |
+
"epoch": 1.8,
|
| 131 |
+
"learning_rate": 9.003845058387924e-05,
|
| 132 |
+
"loss": 0.0043,
|
| 133 |
+
"step": 25290
|
| 134 |
},
|
| 135 |
{
|
| 136 |
+
"epoch": 1.9,
|
| 137 |
+
"learning_rate": 9.504058672742809e-05,
|
| 138 |
+
"loss": 0.0045,
|
| 139 |
+
"step": 26695
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.0,
|
| 143 |
+
"eval_loss": 0.03272660821676254,
|
| 144 |
+
"eval_max_distance": 9,
|
| 145 |
"eval_mean_distance": 0,
|
| 146 |
+
"eval_runtime": 12.7866,
|
| 147 |
+
"eval_samples_per_second": 19.552,
|
| 148 |
+
"eval_steps_per_second": 1.33,
|
| 149 |
+
"step": 28088
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 2.0,
|
| 153 |
+
"learning_rate": 9.99952530143359e-05,
|
| 154 |
+
"loss": 0.0046,
|
| 155 |
+
"step": 28100
|
| 156 |
},
|
| 157 |
{
|
| 158 |
+
"epoch": 2.1,
|
| 159 |
+
"learning_rate": 9.943946010949714e-05,
|
| 160 |
+
"loss": 0.0042,
|
| 161 |
+
"step": 29505
|
| 162 |
},
|
| 163 |
{
|
| 164 |
+
"epoch": 2.2,
|
| 165 |
+
"learning_rate": 9.888366720465838e-05,
|
| 166 |
+
"loss": 0.0043,
|
| 167 |
+
"step": 30910
|
| 168 |
},
|
| 169 |
{
|
| 170 |
+
"epoch": 2.3,
|
| 171 |
+
"learning_rate": 9.832787429981962e-05,
|
| 172 |
+
"loss": 0.0044,
|
| 173 |
+
"step": 32315
|
| 174 |
},
|
| 175 |
{
|
| 176 |
+
"epoch": 2.4,
|
| 177 |
+
"learning_rate": 9.777208139498087e-05,
|
| 178 |
+
"loss": 0.0045,
|
| 179 |
+
"step": 33720
|
| 180 |
},
|
| 181 |
{
|
| 182 |
+
"epoch": 2.5,
|
| 183 |
+
"learning_rate": 9.721628849014209e-05,
|
| 184 |
+
"loss": 0.0041,
|
| 185 |
+
"step": 35125
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 2.6,
|
| 189 |
+
"learning_rate": 9.666049558530333e-05,
|
| 190 |
+
"loss": 0.0043,
|
| 191 |
+
"step": 36530
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 2.7,
|
| 195 |
+
"learning_rate": 9.610470268046457e-05,
|
| 196 |
+
"loss": 0.0046,
|
| 197 |
+
"step": 37935
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 2.8,
|
| 201 |
+
"learning_rate": 9.554890977562582e-05,
|
| 202 |
+
"loss": 0.0043,
|
| 203 |
+
"step": 39340
|
| 204 |
},
|
| 205 |
{
|
| 206 |
+
"epoch": 2.9,
|
| 207 |
+
"learning_rate": 9.499311687078706e-05,
|
| 208 |
+
"loss": 0.0043,
|
| 209 |
+
"step": 40745
|
| 210 |
},
|
| 211 |
{
|
| 212 |
+
"epoch": 3.0,
|
| 213 |
+
"eval_loss": 0.03169206902384758,
|
| 214 |
+
"eval_max_distance": 6,
|
| 215 |
+
"eval_mean_distance": 0,
|
| 216 |
+
"eval_runtime": 12.3553,
|
| 217 |
+
"eval_samples_per_second": 20.234,
|
| 218 |
+
"eval_steps_per_second": 1.376,
|
| 219 |
+
"step": 42132
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 3.0,
|
| 223 |
+
"learning_rate": 9.44373239659483e-05,
|
| 224 |
+
"loss": 0.0048,
|
| 225 |
+
"step": 42150
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 3.1,
|
| 229 |
+
"learning_rate": 9.388153106110953e-05,
|
| 230 |
+
"loss": 0.0034,
|
| 231 |
+
"step": 43555
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 3.2,
|
| 235 |
+
"learning_rate": 9.332573815627077e-05,
|
| 236 |
+
"loss": 0.0036,
|
| 237 |
+
"step": 44960
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 3.3,
|
| 241 |
+
"learning_rate": 9.276994525143201e-05,
|
| 242 |
+
"loss": 0.0037,
|
| 243 |
+
"step": 46365
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 3.4,
|
| 247 |
+
"learning_rate": 9.221415234659325e-05,
|
| 248 |
+
"loss": 0.004,
|
| 249 |
+
"step": 47770
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 3.5,
|
| 253 |
+
"learning_rate": 9.16583594417545e-05,
|
| 254 |
+
"loss": 0.0041,
|
| 255 |
+
"step": 49175
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 3.6,
|
| 259 |
+
"learning_rate": 9.110256653691574e-05,
|
| 260 |
+
"loss": 0.0039,
|
| 261 |
+
"step": 50580
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 3.7,
|
| 265 |
+
"learning_rate": 9.054677363207696e-05,
|
| 266 |
+
"loss": 0.0039,
|
| 267 |
+
"step": 51985
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 3.8,
|
| 271 |
+
"learning_rate": 8.99909807272382e-05,
|
| 272 |
+
"loss": 0.0039,
|
| 273 |
+
"step": 53390
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 3.9,
|
| 277 |
+
"learning_rate": 8.943518782239945e-05,
|
| 278 |
+
"loss": 0.0042,
|
| 279 |
+
"step": 54795
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.0,
|
| 283 |
+
"eval_loss": 0.031554438173770905,
|
| 284 |
+
"eval_max_distance": 6,
|
| 285 |
"eval_mean_distance": 0,
|
| 286 |
+
"eval_runtime": 12.1943,
|
| 287 |
+
"eval_samples_per_second": 20.501,
|
| 288 |
+
"eval_steps_per_second": 1.394,
|
| 289 |
+
"step": 56176
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 4.0,
|
| 293 |
+
"learning_rate": 8.887939491756069e-05,
|
| 294 |
+
"loss": 0.004,
|
| 295 |
+
"step": 56200
|
| 296 |
},
|
| 297 |
{
|
| 298 |
+
"epoch": 4.1,
|
| 299 |
+
"learning_rate": 8.832360201272193e-05,
|
| 300 |
+
"loss": 0.0034,
|
| 301 |
+
"step": 57605
|
| 302 |
},
|
| 303 |
{
|
| 304 |
+
"epoch": 4.2,
|
| 305 |
+
"learning_rate": 8.776780910788317e-05,
|
| 306 |
+
"loss": 0.0033,
|
| 307 |
+
"step": 59010
|
| 308 |
},
|
| 309 |
{
|
| 310 |
+
"epoch": 4.3,
|
| 311 |
+
"learning_rate": 8.72120162030444e-05,
|
| 312 |
+
"loss": 0.0034,
|
| 313 |
+
"step": 60415
|
| 314 |
},
|
| 315 |
{
|
| 316 |
+
"epoch": 4.4,
|
| 317 |
+
"learning_rate": 8.665622329820564e-05,
|
| 318 |
+
"loss": 0.0034,
|
| 319 |
+
"step": 61820
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 4.5,
|
| 323 |
+
"learning_rate": 8.610043039336688e-05,
|
| 324 |
+
"loss": 0.0032,
|
| 325 |
+
"step": 63225
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 4.6,
|
| 329 |
+
"learning_rate": 8.554463748852812e-05,
|
| 330 |
+
"loss": 0.0035,
|
| 331 |
+
"step": 64630
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 4.7,
|
| 335 |
+
"learning_rate": 8.498884458368937e-05,
|
| 336 |
+
"loss": 0.0035,
|
| 337 |
+
"step": 66035
|
| 338 |
},
|
| 339 |
{
|
| 340 |
+
"epoch": 4.8,
|
| 341 |
+
"learning_rate": 8.443305167885061e-05,
|
| 342 |
+
"loss": 0.0035,
|
| 343 |
+
"step": 67440
|
| 344 |
},
|
| 345 |
{
|
| 346 |
+
"epoch": 4.9,
|
| 347 |
+
"learning_rate": 8.387725877401183e-05,
|
| 348 |
+
"loss": 0.0035,
|
| 349 |
+
"step": 68845
|
| 350 |
},
|
| 351 |
{
|
| 352 |
+
"epoch": 5.0,
|
| 353 |
+
"eval_loss": 0.03568544238805771,
|
| 354 |
+
"eval_max_distance": 6,
|
| 355 |
+
"eval_mean_distance": 0,
|
| 356 |
+
"eval_runtime": 12.3273,
|
| 357 |
+
"eval_samples_per_second": 20.28,
|
| 358 |
+
"eval_steps_per_second": 1.379,
|
| 359 |
+
"step": 70220
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 5.0,
|
| 363 |
+
"learning_rate": 8.332146586917307e-05,
|
| 364 |
+
"loss": 0.0037,
|
| 365 |
+
"step": 70250
|
| 366 |
},
|
| 367 |
{
|
| 368 |
+
"epoch": 5.1,
|
| 369 |
+
"learning_rate": 8.276567296433432e-05,
|
| 370 |
+
"loss": 0.0028,
|
| 371 |
+
"step": 71655
|
| 372 |
},
|
| 373 |
{
|
| 374 |
+
"epoch": 5.2,
|
| 375 |
+
"learning_rate": 8.220988005949556e-05,
|
| 376 |
+
"loss": 0.0029,
|
| 377 |
+
"step": 73060
|
| 378 |
},
|
| 379 |
{
|
| 380 |
+
"epoch": 5.3,
|
| 381 |
+
"learning_rate": 8.16540871546568e-05,
|
| 382 |
+
"loss": 0.003,
|
| 383 |
+
"step": 74465
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 5.4,
|
| 387 |
+
"learning_rate": 8.109829424981804e-05,
|
| 388 |
+
"loss": 0.0031,
|
| 389 |
+
"step": 75870
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 5.5,
|
| 393 |
+
"learning_rate": 8.054250134497927e-05,
|
| 394 |
+
"loss": 0.0032,
|
| 395 |
+
"step": 77275
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 5.6,
|
| 399 |
+
"learning_rate": 7.998670844014051e-05,
|
| 400 |
+
"loss": 0.003,
|
| 401 |
+
"step": 78680
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 5.7,
|
| 405 |
+
"learning_rate": 7.943091553530175e-05,
|
| 406 |
+
"loss": 0.0032,
|
| 407 |
+
"step": 80085
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 5.8,
|
| 411 |
+
"learning_rate": 7.8875122630463e-05,
|
| 412 |
+
"loss": 0.0031,
|
| 413 |
+
"step": 81490
|
| 414 |
},
|
| 415 |
{
|
| 416 |
+
"epoch": 5.9,
|
| 417 |
+
"learning_rate": 7.831932972562424e-05,
|
| 418 |
+
"loss": 0.0032,
|
| 419 |
+
"step": 82895
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.0,
|
| 423 |
+
"eval_loss": 0.03648597374558449,
|
| 424 |
+
"eval_max_distance": 6,
|
| 425 |
"eval_mean_distance": 0,
|
| 426 |
+
"eval_runtime": 12.1121,
|
| 427 |
+
"eval_samples_per_second": 20.641,
|
| 428 |
+
"eval_steps_per_second": 1.404,
|
| 429 |
+
"step": 84264
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 6.0,
|
| 433 |
+
"learning_rate": 7.776353682078548e-05,
|
| 434 |
+
"loss": 0.0032,
|
| 435 |
+
"step": 84300
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 6.1,
|
| 439 |
+
"learning_rate": 7.72077439159467e-05,
|
| 440 |
+
"loss": 0.0027,
|
| 441 |
+
"step": 85705
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 6.2,
|
| 445 |
+
"learning_rate": 7.665195101110795e-05,
|
| 446 |
+
"loss": 0.0026,
|
| 447 |
+
"step": 87110
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 6.3,
|
| 451 |
+
"learning_rate": 7.609615810626919e-05,
|
| 452 |
+
"loss": 0.0028,
|
| 453 |
+
"step": 88515
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 6.4,
|
| 457 |
+
"learning_rate": 7.554036520143043e-05,
|
| 458 |
+
"loss": 0.0029,
|
| 459 |
+
"step": 89920
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 6.5,
|
| 463 |
+
"learning_rate": 7.498457229659167e-05,
|
| 464 |
+
"loss": 0.0026,
|
| 465 |
+
"step": 91325
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 6.6,
|
| 469 |
+
"learning_rate": 7.442877939175292e-05,
|
| 470 |
+
"loss": 0.0028,
|
| 471 |
+
"step": 92730
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 6.7,
|
| 475 |
+
"learning_rate": 7.387298648691414e-05,
|
| 476 |
+
"loss": 0.0031,
|
| 477 |
+
"step": 94135
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 6.8,
|
| 481 |
+
"learning_rate": 7.331719358207538e-05,
|
| 482 |
+
"loss": 0.0027,
|
| 483 |
+
"step": 95540
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 6.9,
|
| 487 |
+
"learning_rate": 7.276140067723662e-05,
|
| 488 |
+
"loss": 0.0027,
|
| 489 |
+
"step": 96945
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 7.0,
|
| 493 |
+
"eval_loss": 0.04028007388114929,
|
| 494 |
+
"eval_max_distance": 6,
|
| 495 |
+
"eval_mean_distance": 0,
|
| 496 |
+
"eval_runtime": 12.1827,
|
| 497 |
+
"eval_samples_per_second": 20.521,
|
| 498 |
+
"eval_steps_per_second": 1.395,
|
| 499 |
+
"step": 98308
|
| 500 |
},
|
| 501 |
{
|
| 502 |
+
"epoch": 7.0,
|
| 503 |
+
"learning_rate": 7.220560777239787e-05,
|
| 504 |
+
"loss": 0.0028,
|
| 505 |
+
"step": 98350
|
| 506 |
},
|
| 507 |
{
|
| 508 |
+
"epoch": 7.1,
|
| 509 |
+
"learning_rate": 7.164981486755911e-05,
|
| 510 |
+
"loss": 0.0027,
|
| 511 |
+
"step": 99755
|
| 512 |
},
|
| 513 |
{
|
| 514 |
+
"epoch": 7.2,
|
| 515 |
+
"learning_rate": 7.109402196272035e-05,
|
| 516 |
+
"loss": 0.0025,
|
| 517 |
+
"step": 101160
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 7.3,
|
| 521 |
+
"learning_rate": 7.053822905788158e-05,
|
| 522 |
+
"loss": 0.0025,
|
| 523 |
+
"step": 102565
|
| 524 |
},
|
| 525 |
{
|
| 526 |
+
"epoch": 7.4,
|
| 527 |
+
"learning_rate": 6.998243615304282e-05,
|
| 528 |
+
"loss": 0.0024,
|
| 529 |
+
"step": 103970
|
| 530 |
},
|
| 531 |
{
|
| 532 |
+
"epoch": 7.5,
|
| 533 |
+
"learning_rate": 6.942664324820406e-05,
|
| 534 |
+
"loss": 0.0026,
|
| 535 |
+
"step": 105375
|
| 536 |
},
|
| 537 |
{
|
| 538 |
+
"epoch": 7.6,
|
| 539 |
+
"learning_rate": 6.88708503433653e-05,
|
| 540 |
+
"loss": 0.0026,
|
| 541 |
+
"step": 106780
|
| 542 |
},
|
| 543 |
{
|
| 544 |
+
"epoch": 7.7,
|
| 545 |
+
"learning_rate": 6.831505743852654e-05,
|
| 546 |
+
"loss": 0.0026,
|
| 547 |
+
"step": 108185
|
| 548 |
},
|
| 549 |
{
|
| 550 |
+
"epoch": 7.8,
|
| 551 |
+
"learning_rate": 6.775926453368779e-05,
|
| 552 |
+
"loss": 0.0025,
|
| 553 |
+
"step": 109590
|
| 554 |
},
|
| 555 |
{
|
| 556 |
+
"epoch": 7.9,
|
| 557 |
+
"learning_rate": 6.720347162884901e-05,
|
| 558 |
+
"loss": 0.0027,
|
| 559 |
+
"step": 110995
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 8.0,
|
| 563 |
+
"eval_loss": 0.03977961093187332,
|
| 564 |
+
"eval_max_distance": 6,
|
| 565 |
"eval_mean_distance": 0,
|
| 566 |
+
"eval_runtime": 12.1864,
|
| 567 |
+
"eval_samples_per_second": 20.515,
|
| 568 |
+
"eval_steps_per_second": 1.395,
|
| 569 |
+
"step": 112352
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 8.0,
|
| 573 |
+
"learning_rate": 6.664767872401025e-05,
|
| 574 |
+
"loss": 0.0028,
|
| 575 |
+
"step": 112400
|
| 576 |
},
|
| 577 |
{
|
| 578 |
+
"epoch": 8.1,
|
| 579 |
+
"learning_rate": 6.60918858191715e-05,
|
| 580 |
+
"loss": 0.0023,
|
| 581 |
+
"step": 113805
|
| 582 |
},
|
| 583 |
{
|
| 584 |
+
"epoch": 8.2,
|
| 585 |
+
"learning_rate": 6.553609291433274e-05,
|
| 586 |
+
"loss": 0.0023,
|
| 587 |
+
"step": 115210
|
| 588 |
},
|
| 589 |
{
|
| 590 |
+
"epoch": 8.3,
|
| 591 |
+
"learning_rate": 6.498030000949398e-05,
|
| 592 |
+
"loss": 0.0022,
|
| 593 |
+
"step": 116615
|
| 594 |
},
|
| 595 |
{
|
| 596 |
+
"epoch": 8.4,
|
| 597 |
+
"learning_rate": 6.442450710465522e-05,
|
| 598 |
+
"loss": 0.0023,
|
| 599 |
+
"step": 118020
|
| 600 |
},
|
| 601 |
{
|
| 602 |
+
"epoch": 8.5,
|
| 603 |
+
"learning_rate": 6.386871419981645e-05,
|
| 604 |
+
"loss": 0.0024,
|
| 605 |
+
"step": 119425
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 8.6,
|
| 609 |
+
"learning_rate": 6.331292129497769e-05,
|
| 610 |
+
"loss": 0.0025,
|
| 611 |
+
"step": 120830
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 8.7,
|
| 615 |
+
"learning_rate": 6.275712839013893e-05,
|
| 616 |
+
"loss": 0.0024,
|
| 617 |
+
"step": 122235
|
| 618 |
},
|
| 619 |
{
|
| 620 |
+
"epoch": 8.8,
|
| 621 |
+
"learning_rate": 6.220133548530017e-05,
|
| 622 |
+
"loss": 0.0022,
|
| 623 |
+
"step": 123640
|
| 624 |
},
|
| 625 |
{
|
| 626 |
+
"epoch": 8.9,
|
| 627 |
+
"learning_rate": 6.164554258046142e-05,
|
| 628 |
+
"loss": 0.0023,
|
| 629 |
+
"step": 125045
|
| 630 |
},
|
| 631 |
{
|
| 632 |
+
"epoch": 9.0,
|
| 633 |
+
"eval_loss": 0.040445487946271896,
|
| 634 |
+
"eval_max_distance": 6,
|
| 635 |
+
"eval_mean_distance": 0,
|
| 636 |
+
"eval_runtime": 12.3201,
|
| 637 |
+
"eval_samples_per_second": 20.292,
|
| 638 |
+
"eval_steps_per_second": 1.38,
|
| 639 |
+
"step": 126396
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 9.0,
|
| 643 |
+
"learning_rate": 6.108974967562266e-05,
|
| 644 |
+
"loss": 0.0024,
|
| 645 |
+
"step": 126450
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 9.1,
|
| 649 |
+
"learning_rate": 6.053395677078388e-05,
|
| 650 |
+
"loss": 0.002,
|
| 651 |
+
"step": 127855
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 9.2,
|
| 655 |
+
"learning_rate": 5.9978163865945126e-05,
|
| 656 |
+
"loss": 0.0021,
|
| 657 |
+
"step": 129260
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 9.3,
|
| 661 |
+
"learning_rate": 5.9422370961106364e-05,
|
| 662 |
+
"loss": 0.0022,
|
| 663 |
+
"step": 130665
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 9.4,
|
| 667 |
+
"learning_rate": 5.886657805626761e-05,
|
| 668 |
+
"loss": 0.0023,
|
| 669 |
+
"step": 132070
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 9.5,
|
| 673 |
+
"learning_rate": 5.831078515142885e-05,
|
| 674 |
+
"loss": 0.0022,
|
| 675 |
+
"step": 133475
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 9.6,
|
| 679 |
+
"learning_rate": 5.775499224659009e-05,
|
| 680 |
+
"loss": 0.0021,
|
| 681 |
+
"step": 134880
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 9.7,
|
| 685 |
+
"learning_rate": 5.719919934175132e-05,
|
| 686 |
+
"loss": 0.0021,
|
| 687 |
+
"step": 136285
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 9.8,
|
| 691 |
+
"learning_rate": 5.664340643691256e-05,
|
| 692 |
+
"loss": 0.0021,
|
| 693 |
+
"step": 137690
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 9.9,
|
| 697 |
+
"learning_rate": 5.60876135320738e-05,
|
| 698 |
+
"loss": 0.0023,
|
| 699 |
+
"step": 139095
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 10.0,
|
| 703 |
+
"eval_loss": 0.03852245956659317,
|
| 704 |
+
"eval_max_distance": 6,
|
| 705 |
"eval_mean_distance": 0,
|
| 706 |
+
"eval_runtime": 12.131,
|
| 707 |
+
"eval_samples_per_second": 20.608,
|
| 708 |
+
"eval_steps_per_second": 1.401,
|
| 709 |
+
"step": 140440
|
| 710 |
},
|
| 711 |
{
|
| 712 |
+
"epoch": 10.0,
|
| 713 |
+
"learning_rate": 5.5531820627235044e-05,
|
| 714 |
+
"loss": 0.0023,
|
| 715 |
+
"step": 140500
|
| 716 |
},
|
| 717 |
{
|
| 718 |
+
"epoch": 10.1,
|
| 719 |
+
"learning_rate": 5.497602772239628e-05,
|
| 720 |
+
"loss": 0.0018,
|
| 721 |
+
"step": 141905
|
| 722 |
},
|
| 723 |
{
|
| 724 |
+
"epoch": 10.2,
|
| 725 |
+
"learning_rate": 5.4420234817557526e-05,
|
| 726 |
+
"loss": 0.0021,
|
| 727 |
+
"step": 143310
|
| 728 |
},
|
| 729 |
{
|
| 730 |
+
"epoch": 10.3,
|
| 731 |
+
"learning_rate": 5.386444191271876e-05,
|
| 732 |
+
"loss": 0.0021,
|
| 733 |
+
"step": 144715
|
| 734 |
},
|
| 735 |
{
|
| 736 |
+
"epoch": 10.4,
|
| 737 |
+
"learning_rate": 5.3308649007879995e-05,
|
| 738 |
+
"loss": 0.0021,
|
| 739 |
+
"step": 146120
|
| 740 |
},
|
| 741 |
{
|
| 742 |
+
"epoch": 10.5,
|
| 743 |
+
"learning_rate": 5.275285610304124e-05,
|
| 744 |
+
"loss": 0.0021,
|
| 745 |
+
"step": 147525
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 10.6,
|
| 749 |
+
"learning_rate": 5.219706319820248e-05,
|
| 750 |
+
"loss": 0.002,
|
| 751 |
+
"step": 148930
|
| 752 |
},
|
| 753 |
{
|
| 754 |
+
"epoch": 10.7,
|
| 755 |
+
"learning_rate": 5.1641270293363716e-05,
|
| 756 |
+
"loss": 0.002,
|
| 757 |
+
"step": 150335
|
| 758 |
},
|
| 759 |
{
|
| 760 |
+
"epoch": 10.8,
|
| 761 |
+
"learning_rate": 5.108547738852496e-05,
|
| 762 |
+
"loss": 0.0019,
|
| 763 |
+
"step": 151740
|
| 764 |
},
|
| 765 |
{
|
| 766 |
+
"epoch": 10.9,
|
| 767 |
+
"learning_rate": 5.052968448368619e-05,
|
| 768 |
+
"loss": 0.002,
|
| 769 |
+
"step": 153145
|
| 770 |
},
|
| 771 |
{
|
| 772 |
+
"epoch": 11.0,
|
| 773 |
+
"eval_loss": 0.040671207010746,
|
| 774 |
+
"eval_max_distance": 6,
|
| 775 |
+
"eval_mean_distance": 0,
|
| 776 |
+
"eval_runtime": 12.2718,
|
| 777 |
+
"eval_samples_per_second": 20.372,
|
| 778 |
+
"eval_steps_per_second": 1.385,
|
| 779 |
+
"step": 154484
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 11.0,
|
| 783 |
+
"learning_rate": 4.997389157884743e-05,
|
| 784 |
+
"loss": 0.0018,
|
| 785 |
+
"step": 154550
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 11.1,
|
| 789 |
+
"learning_rate": 4.9418098674008675e-05,
|
| 790 |
+
"loss": 0.002,
|
| 791 |
+
"step": 155955
|
| 792 |
},
|
| 793 |
{
|
| 794 |
+
"epoch": 11.2,
|
| 795 |
+
"learning_rate": 4.886230576916991e-05,
|
| 796 |
+
"loss": 0.0019,
|
| 797 |
+
"step": 157360
|
| 798 |
},
|
| 799 |
{
|
| 800 |
+
"epoch": 11.3,
|
| 801 |
+
"learning_rate": 4.830651286433115e-05,
|
| 802 |
+
"loss": 0.0018,
|
| 803 |
+
"step": 158765
|
| 804 |
},
|
| 805 |
{
|
| 806 |
+
"epoch": 11.4,
|
| 807 |
+
"learning_rate": 4.775071995949239e-05,
|
| 808 |
+
"loss": 0.0018,
|
| 809 |
+
"step": 160170
|
| 810 |
},
|
| 811 |
{
|
| 812 |
+
"epoch": 11.5,
|
| 813 |
+
"learning_rate": 4.7194927054653634e-05,
|
| 814 |
+
"loss": 0.002,
|
| 815 |
+
"step": 161575
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 11.6,
|
| 819 |
+
"learning_rate": 4.6639134149814865e-05,
|
| 820 |
+
"loss": 0.0019,
|
| 821 |
+
"step": 162980
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 11.7,
|
| 825 |
+
"learning_rate": 4.608334124497611e-05,
|
| 826 |
+
"loss": 0.0017,
|
| 827 |
+
"step": 164385
|
| 828 |
},
|
| 829 |
{
|
| 830 |
+
"epoch": 11.81,
|
| 831 |
+
"learning_rate": 4.552754834013735e-05,
|
| 832 |
+
"loss": 0.002,
|
| 833 |
+
"step": 165790
|
| 834 |
},
|
| 835 |
{
|
| 836 |
+
"epoch": 11.91,
|
| 837 |
+
"learning_rate": 4.4971755435298586e-05,
|
| 838 |
+
"loss": 0.0018,
|
| 839 |
+
"step": 167195
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 12.0,
|
| 843 |
+
"eval_loss": 0.04258317872881889,
|
| 844 |
+
"eval_max_distance": 9,
|
| 845 |
"eval_mean_distance": 0,
|
| 846 |
+
"eval_runtime": 12.8224,
|
| 847 |
+
"eval_samples_per_second": 19.497,
|
| 848 |
+
"eval_steps_per_second": 1.326,
|
| 849 |
+
"step": 168528
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 12.01,
|
| 853 |
+
"learning_rate": 4.4415962530459824e-05,
|
| 854 |
+
"loss": 0.0017,
|
| 855 |
+
"step": 168600
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 12.11,
|
| 859 |
+
"learning_rate": 4.386016962562107e-05,
|
| 860 |
+
"loss": 0.0016,
|
| 861 |
+
"step": 170005
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 12.21,
|
| 865 |
+
"learning_rate": 4.33043767207823e-05,
|
| 866 |
+
"loss": 0.0018,
|
| 867 |
+
"step": 171410
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 12.31,
|
| 871 |
+
"learning_rate": 4.2748583815943544e-05,
|
| 872 |
+
"loss": 0.0016,
|
| 873 |
+
"step": 172815
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 12.41,
|
| 877 |
+
"learning_rate": 4.219279091110478e-05,
|
| 878 |
+
"loss": 0.0018,
|
| 879 |
+
"step": 174220
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 12.51,
|
| 883 |
+
"learning_rate": 4.163699800626602e-05,
|
| 884 |
+
"loss": 0.0018,
|
| 885 |
+
"step": 175625
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 12.61,
|
| 889 |
+
"learning_rate": 4.108120510142726e-05,
|
| 890 |
+
"loss": 0.0018,
|
| 891 |
+
"step": 177030
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 12.71,
|
| 895 |
+
"learning_rate": 4.05254121965885e-05,
|
| 896 |
+
"loss": 0.0016,
|
| 897 |
+
"step": 178435
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 12.81,
|
| 901 |
+
"learning_rate": 3.996961929174974e-05,
|
| 902 |
+
"loss": 0.002,
|
| 903 |
+
"step": 179840
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 12.91,
|
| 907 |
+
"learning_rate": 3.941382638691098e-05,
|
| 908 |
+
"loss": 0.0018,
|
| 909 |
+
"step": 181245
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 13.0,
|
| 913 |
+
"eval_loss": 0.042234089225530624,
|
| 914 |
+
"eval_max_distance": 6,
|
| 915 |
+
"eval_mean_distance": 0,
|
| 916 |
+
"eval_runtime": 13.9619,
|
| 917 |
+
"eval_samples_per_second": 17.906,
|
| 918 |
+
"eval_steps_per_second": 1.218,
|
| 919 |
+
"step": 182572
|
| 920 |
},
|
| 921 |
{
|
| 922 |
+
"epoch": 13.01,
|
| 923 |
+
"learning_rate": 3.885803348207222e-05,
|
| 924 |
+
"loss": 0.0018,
|
| 925 |
+
"step": 182650
|
| 926 |
},
|
| 927 |
{
|
| 928 |
+
"epoch": 13.11,
|
| 929 |
+
"learning_rate": 3.8302240577233455e-05,
|
| 930 |
+
"loss": 0.0017,
|
| 931 |
+
"step": 184055
|
| 932 |
},
|
| 933 |
{
|
| 934 |
+
"epoch": 13.21,
|
| 935 |
+
"learning_rate": 3.77464476723947e-05,
|
| 936 |
+
"loss": 0.0017,
|
| 937 |
+
"step": 185460
|
| 938 |
},
|
| 939 |
{
|
| 940 |
+
"epoch": 13.31,
|
| 941 |
+
"learning_rate": 3.719065476755594e-05,
|
| 942 |
+
"loss": 0.0016,
|
| 943 |
+
"step": 186865
|
| 944 |
},
|
| 945 |
{
|
| 946 |
+
"epoch": 13.41,
|
| 947 |
+
"learning_rate": 3.6634861862717176e-05,
|
| 948 |
+
"loss": 0.0016,
|
| 949 |
+
"step": 188270
|
| 950 |
},
|
| 951 |
{
|
| 952 |
+
"epoch": 13.51,
|
| 953 |
+
"learning_rate": 3.6079068957878414e-05,
|
| 954 |
+
"loss": 0.0016,
|
| 955 |
+
"step": 189675
|
| 956 |
},
|
| 957 |
{
|
| 958 |
+
"epoch": 13.61,
|
| 959 |
+
"learning_rate": 3.552327605303966e-05,
|
| 960 |
+
"loss": 0.0017,
|
| 961 |
+
"step": 191080
|
| 962 |
},
|
| 963 |
{
|
| 964 |
+
"epoch": 13.71,
|
| 965 |
+
"learning_rate": 3.4967483148200896e-05,
|
| 966 |
+
"loss": 0.0017,
|
| 967 |
+
"step": 192485
|
| 968 |
},
|
| 969 |
{
|
| 970 |
+
"epoch": 13.81,
|
| 971 |
+
"learning_rate": 3.4411690243362134e-05,
|
| 972 |
+
"loss": 0.0015,
|
| 973 |
+
"step": 193890
|
| 974 |
},
|
| 975 |
{
|
| 976 |
+
"epoch": 13.91,
|
| 977 |
+
"learning_rate": 3.385589733852337e-05,
|
| 978 |
+
"loss": 0.0016,
|
| 979 |
+
"step": 195295
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 14.0,
|
| 983 |
+
"eval_loss": 0.04207869619131088,
|
| 984 |
+
"eval_max_distance": 6,
|
| 985 |
"eval_mean_distance": 0,
|
| 986 |
+
"eval_runtime": 12.3358,
|
| 987 |
+
"eval_samples_per_second": 20.266,
|
| 988 |
+
"eval_steps_per_second": 1.378,
|
| 989 |
+
"step": 196616
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 14.01,
|
| 993 |
+
"learning_rate": 3.330010443368462e-05,
|
| 994 |
+
"loss": 0.0016,
|
| 995 |
+
"step": 196700
|
| 996 |
},
|
| 997 |
{
|
| 998 |
+
"epoch": 14.11,
|
| 999 |
+
"learning_rate": 3.274431152884585e-05,
|
| 1000 |
+
"loss": 0.0016,
|
| 1001 |
+
"step": 198105
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
+
"epoch": 14.21,
|
| 1005 |
+
"learning_rate": 3.218851862400709e-05,
|
| 1006 |
+
"loss": 0.0015,
|
| 1007 |
+
"step": 199510
|
| 1008 |
},
|
| 1009 |
{
|
| 1010 |
+
"epoch": 14.31,
|
| 1011 |
+
"learning_rate": 3.163272571916833e-05,
|
| 1012 |
+
"loss": 0.0016,
|
| 1013 |
+
"step": 200915
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
+
"epoch": 14.41,
|
| 1017 |
+
"learning_rate": 3.107693281432957e-05,
|
| 1018 |
+
"loss": 0.0016,
|
| 1019 |
+
"step": 202320
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
+
"epoch": 14.51,
|
| 1023 |
+
"learning_rate": 3.052113990949081e-05,
|
| 1024 |
+
"loss": 0.0016,
|
| 1025 |
+
"step": 203725
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
+
"epoch": 14.61,
|
| 1029 |
+
"learning_rate": 2.9965347004652052e-05,
|
| 1030 |
+
"loss": 0.0016,
|
| 1031 |
+
"step": 205130
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 14.71,
|
| 1035 |
+
"learning_rate": 2.9409554099813286e-05,
|
| 1036 |
+
"loss": 0.0015,
|
| 1037 |
+
"step": 206535
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
+
"epoch": 14.81,
|
| 1041 |
+
"learning_rate": 2.8853761194974528e-05,
|
| 1042 |
+
"loss": 0.0016,
|
| 1043 |
+
"step": 207940
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
+
"epoch": 14.91,
|
| 1047 |
+
"learning_rate": 2.829796829013577e-05,
|
| 1048 |
+
"loss": 0.0016,
|
| 1049 |
+
"step": 209345
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
+
"epoch": 15.0,
|
| 1053 |
+
"eval_loss": 0.04015611857175827,
|
| 1054 |
+
"eval_max_distance": 6,
|
| 1055 |
+
"eval_mean_distance": 0,
|
| 1056 |
+
"eval_runtime": 12.1736,
|
| 1057 |
+
"eval_samples_per_second": 20.536,
|
| 1058 |
+
"eval_steps_per_second": 1.396,
|
| 1059 |
+
"step": 210660
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 15.01,
|
| 1063 |
+
"learning_rate": 2.7742175385297004e-05,
|
| 1064 |
+
"loss": 0.0015,
|
| 1065 |
+
"step": 210750
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 15.11,
|
| 1069 |
+
"learning_rate": 2.7186382480458245e-05,
|
| 1070 |
+
"loss": 0.0015,
|
| 1071 |
+
"step": 212155
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 15.21,
|
| 1075 |
+
"learning_rate": 2.6630589575619486e-05,
|
| 1076 |
+
"loss": 0.0014,
|
| 1077 |
+
"step": 213560
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
+
"epoch": 15.31,
|
| 1081 |
+
"learning_rate": 2.607479667078072e-05,
|
| 1082 |
+
"loss": 0.0016,
|
| 1083 |
+
"step": 214965
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
+
"epoch": 15.41,
|
| 1087 |
+
"learning_rate": 2.5519003765941962e-05,
|
| 1088 |
+
"loss": 0.0014,
|
| 1089 |
+
"step": 216370
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
+
"epoch": 15.51,
|
| 1093 |
+
"learning_rate": 2.49632108611032e-05,
|
| 1094 |
+
"loss": 0.0016,
|
| 1095 |
+
"step": 217775
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 15.61,
|
| 1099 |
+
"learning_rate": 2.440741795626444e-05,
|
| 1100 |
+
"loss": 0.0014,
|
| 1101 |
+
"step": 219180
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 15.71,
|
| 1105 |
+
"learning_rate": 2.385162505142568e-05,
|
| 1106 |
+
"loss": 0.0015,
|
| 1107 |
+
"step": 220585
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 15.81,
|
| 1111 |
+
"learning_rate": 2.3295832146586918e-05,
|
| 1112 |
+
"loss": 0.0015,
|
| 1113 |
+
"step": 221990
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 15.91,
|
| 1117 |
+
"learning_rate": 2.2740039241748156e-05,
|
| 1118 |
+
"loss": 0.0014,
|
| 1119 |
+
"step": 223395
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 16.0,
|
| 1123 |
+
"eval_loss": 0.04069029539823532,
|
| 1124 |
+
"eval_max_distance": 6,
|
| 1125 |
"eval_mean_distance": 0,
|
| 1126 |
+
"eval_runtime": 12.2719,
|
| 1127 |
+
"eval_samples_per_second": 20.372,
|
| 1128 |
+
"eval_steps_per_second": 1.385,
|
| 1129 |
+
"step": 224704
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 16.01,
|
| 1133 |
+
"learning_rate": 2.2184246336909397e-05,
|
| 1134 |
+
"loss": 0.0015,
|
| 1135 |
+
"step": 224800
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
+
"epoch": 16.11,
|
| 1139 |
+
"learning_rate": 2.1628453432070635e-05,
|
| 1140 |
+
"loss": 0.0013,
|
| 1141 |
+
"step": 226205
|
| 1142 |
},
|
| 1143 |
{
|
| 1144 |
+
"epoch": 16.21,
|
| 1145 |
+
"learning_rate": 2.1072660527231873e-05,
|
| 1146 |
+
"loss": 0.0014,
|
| 1147 |
+
"step": 227610
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
+
"epoch": 16.31,
|
| 1151 |
+
"learning_rate": 2.0516867622393114e-05,
|
| 1152 |
+
"loss": 0.0014,
|
| 1153 |
+
"step": 229015
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
+
"epoch": 16.41,
|
| 1157 |
+
"learning_rate": 1.9961074717554352e-05,
|
| 1158 |
+
"loss": 0.0015,
|
| 1159 |
+
"step": 230420
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
+
"epoch": 16.51,
|
| 1163 |
+
"learning_rate": 1.9405281812715594e-05,
|
| 1164 |
+
"loss": 0.0016,
|
| 1165 |
+
"step": 231825
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
+
"epoch": 16.61,
|
| 1169 |
+
"learning_rate": 1.8849488907876832e-05,
|
| 1170 |
+
"loss": 0.0015,
|
| 1171 |
+
"step": 233230
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 16.71,
|
| 1175 |
+
"learning_rate": 1.8293696003038073e-05,
|
| 1176 |
+
"loss": 0.0014,
|
| 1177 |
+
"step": 234635
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
+
"epoch": 16.81,
|
| 1181 |
+
"learning_rate": 1.773790309819931e-05,
|
| 1182 |
+
"loss": 0.0013,
|
| 1183 |
+
"step": 236040
|
| 1184 |
},
|
| 1185 |
{
|
| 1186 |
+
"epoch": 16.91,
|
| 1187 |
+
"learning_rate": 1.7182110193360553e-05,
|
| 1188 |
+
"loss": 0.0014,
|
| 1189 |
+
"step": 237445
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
+
"epoch": 17.0,
|
| 1193 |
+
"eval_loss": 0.042702946811914444,
|
| 1194 |
+
"eval_max_distance": 6,
|
| 1195 |
+
"eval_mean_distance": 0,
|
| 1196 |
+
"eval_runtime": 12.1336,
|
| 1197 |
+
"eval_samples_per_second": 20.604,
|
| 1198 |
+
"eval_steps_per_second": 1.401,
|
| 1199 |
+
"step": 238748
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
+
"epoch": 17.01,
|
| 1203 |
+
"learning_rate": 1.662631728852179e-05,
|
| 1204 |
+
"loss": 0.0013,
|
| 1205 |
+
"step": 238850
|
| 1206 |
},
|
| 1207 |
{
|
| 1208 |
+
"epoch": 17.11,
|
| 1209 |
+
"learning_rate": 1.607052438368303e-05,
|
| 1210 |
+
"loss": 0.0014,
|
| 1211 |
+
"step": 240255
|
| 1212 |
},
|
| 1213 |
{
|
| 1214 |
+
"epoch": 17.21,
|
| 1215 |
+
"learning_rate": 1.551473147884427e-05,
|
| 1216 |
+
"loss": 0.0014,
|
| 1217 |
+
"step": 241660
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
+
"epoch": 17.31,
|
| 1221 |
+
"learning_rate": 1.4958938574005508e-05,
|
| 1222 |
+
"loss": 0.0013,
|
| 1223 |
+
"step": 243065
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
+
"epoch": 17.41,
|
| 1227 |
+
"learning_rate": 1.4403145669166746e-05,
|
| 1228 |
+
"loss": 0.0014,
|
| 1229 |
+
"step": 244470
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
+
"epoch": 17.51,
|
| 1233 |
+
"learning_rate": 1.3847352764327987e-05,
|
| 1234 |
+
"loss": 0.0013,
|
| 1235 |
+
"step": 245875
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
+
"epoch": 17.61,
|
| 1239 |
+
"learning_rate": 1.3291559859489225e-05,
|
| 1240 |
+
"loss": 0.0013,
|
| 1241 |
+
"step": 247280
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
+
"epoch": 17.71,
|
| 1245 |
+
"learning_rate": 1.2735766954650463e-05,
|
| 1246 |
+
"loss": 0.0014,
|
| 1247 |
+
"step": 248685
|
| 1248 |
},
|
| 1249 |
{
|
| 1250 |
+
"epoch": 17.81,
|
| 1251 |
+
"learning_rate": 1.2179974049811703e-05,
|
| 1252 |
+
"loss": 0.0014,
|
| 1253 |
+
"step": 250090
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
+
"epoch": 17.91,
|
| 1257 |
+
"learning_rate": 1.1624181144972943e-05,
|
| 1258 |
+
"loss": 0.0014,
|
| 1259 |
+
"step": 251495
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
+
"epoch": 18.0,
|
| 1263 |
+
"eval_loss": 0.04110053926706314,
|
| 1264 |
+
"eval_max_distance": 6,
|
| 1265 |
"eval_mean_distance": 0,
|
| 1266 |
+
"eval_runtime": 12.0446,
|
| 1267 |
+
"eval_samples_per_second": 20.756,
|
| 1268 |
+
"eval_steps_per_second": 1.411,
|
| 1269 |
+
"step": 252792
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
+
"epoch": 18.01,
|
| 1273 |
+
"learning_rate": 1.1068388240134182e-05,
|
| 1274 |
+
"loss": 0.0014,
|
| 1275 |
+
"step": 252900
|
| 1276 |
},
|
| 1277 |
{
|
| 1278 |
+
"epoch": 18.11,
|
| 1279 |
+
"learning_rate": 1.051259533529542e-05,
|
| 1280 |
+
"loss": 0.0014,
|
| 1281 |
+
"step": 254305
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
+
"epoch": 18.21,
|
| 1285 |
+
"learning_rate": 9.95680243045666e-06,
|
| 1286 |
+
"loss": 0.0012,
|
| 1287 |
+
"step": 255710
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
+
"epoch": 18.31,
|
| 1291 |
+
"learning_rate": 9.4010095256179e-06,
|
| 1292 |
+
"loss": 0.0012,
|
| 1293 |
+
"step": 257115
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
+
"epoch": 18.41,
|
| 1297 |
+
"learning_rate": 8.84521662077914e-06,
|
| 1298 |
+
"loss": 0.0014,
|
| 1299 |
+
"step": 258520
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
+
"epoch": 18.51,
|
| 1303 |
+
"learning_rate": 8.289423715940379e-06,
|
| 1304 |
+
"loss": 0.0012,
|
| 1305 |
+
"step": 259925
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
+
"epoch": 18.61,
|
| 1309 |
+
"learning_rate": 7.733630811101619e-06,
|
| 1310 |
+
"loss": 0.0013,
|
| 1311 |
+
"step": 261330
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
+
"epoch": 18.71,
|
| 1315 |
+
"learning_rate": 7.1778379062628565e-06,
|
| 1316 |
+
"loss": 0.0014,
|
| 1317 |
+
"step": 262735
|
| 1318 |
},
|
| 1319 |
{
|
| 1320 |
+
"epoch": 18.81,
|
| 1321 |
+
"learning_rate": 6.622045001424096e-06,
|
| 1322 |
+
"loss": 0.0012,
|
| 1323 |
+
"step": 264140
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
+
"epoch": 18.91,
|
| 1327 |
+
"learning_rate": 6.066252096585336e-06,
|
| 1328 |
+
"loss": 0.0013,
|
| 1329 |
+
"step": 265545
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
+
"epoch": 19.0,
|
| 1333 |
+
"eval_loss": 0.040625352412462234,
|
| 1334 |
+
"eval_max_distance": 6,
|
| 1335 |
+
"eval_mean_distance": 0,
|
| 1336 |
+
"eval_runtime": 12.0757,
|
| 1337 |
+
"eval_samples_per_second": 20.703,
|
| 1338 |
+
"eval_steps_per_second": 1.408,
|
| 1339 |
+
"step": 266836
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
+
"epoch": 19.01,
|
| 1343 |
+
"learning_rate": 5.510459191746575e-06,
|
| 1344 |
+
"loss": 0.0013,
|
| 1345 |
+
"step": 266950
|
| 1346 |
},
|
| 1347 |
{
|
| 1348 |
+
"epoch": 19.11,
|
| 1349 |
+
"learning_rate": 4.9546662869078136e-06,
|
| 1350 |
+
"loss": 0.0014,
|
| 1351 |
+
"step": 268355
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
+
"epoch": 19.21,
|
| 1355 |
+
"learning_rate": 4.398873382069053e-06,
|
| 1356 |
+
"loss": 0.0013,
|
| 1357 |
+
"step": 269760
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
+
"epoch": 19.31,
|
| 1361 |
+
"learning_rate": 3.843080477230292e-06,
|
| 1362 |
+
"loss": 0.0013,
|
| 1363 |
+
"step": 271165
|
| 1364 |
},
|
| 1365 |
{
|
| 1366 |
+
"epoch": 19.41,
|
| 1367 |
+
"learning_rate": 3.2872875723915313e-06,
|
| 1368 |
+
"loss": 0.0013,
|
| 1369 |
+
"step": 272570
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
+
"epoch": 19.51,
|
| 1373 |
+
"learning_rate": 2.7314946675527706e-06,
|
| 1374 |
+
"loss": 0.0012,
|
| 1375 |
+
"step": 273975
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
+
"epoch": 19.61,
|
| 1379 |
+
"learning_rate": 2.1757017627140103e-06,
|
| 1380 |
+
"loss": 0.0012,
|
| 1381 |
+
"step": 275380
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
+
"epoch": 19.71,
|
| 1385 |
+
"learning_rate": 1.619908857875249e-06,
|
| 1386 |
+
"loss": 0.0013,
|
| 1387 |
+
"step": 276785
|
| 1388 |
},
|
| 1389 |
{
|
| 1390 |
+
"epoch": 19.81,
|
| 1391 |
+
"learning_rate": 1.0641159530364886e-06,
|
| 1392 |
+
"loss": 0.0014,
|
| 1393 |
+
"step": 278190
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
+
"epoch": 19.91,
|
| 1397 |
+
"learning_rate": 5.083230481977278e-07,
|
| 1398 |
+
"loss": 0.0013,
|
| 1399 |
+
"step": 279595
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
+
"epoch": 20.0,
|
| 1403 |
+
"eval_loss": 0.040508754551410675,
|
| 1404 |
+
"eval_max_distance": 6,
|
| 1405 |
+
"eval_mean_distance": 0,
|
| 1406 |
+
"eval_runtime": 12.1336,
|
| 1407 |
+
"eval_samples_per_second": 20.604,
|
| 1408 |
+
"eval_steps_per_second": 1.401,
|
| 1409 |
+
"step": 280880
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 20.0,
|
| 1413 |
+
"step": 280880,
|
| 1414 |
+
"total_flos": 7.269950341627085e+16,
|
| 1415 |
+
"train_loss": 0.0021448437322240947,
|
| 1416 |
+
"train_runtime": 20054.4611,
|
| 1417 |
+
"train_samples_per_second": 210.084,
|
| 1418 |
+
"train_steps_per_second": 14.006
|
| 1419 |
}
|
| 1420 |
],
|
| 1421 |
+
"logging_steps": 1405,
|
| 1422 |
+
"max_steps": 280880,
|
| 1423 |
+
"num_train_epochs": 20,
|
| 1424 |
+
"save_steps": 2809,
|
| 1425 |
+
"total_flos": 7.269950341627085e+16,
|
| 1426 |
"trial_name": null,
|
| 1427 |
"trial_params": null
|
| 1428 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:768220270743ed53ef0cb8a00d97f4eba8645c310d633cf26655e6848b4b3523
|
| 3 |
size 4091
|