Commit
·
22862bd
1
Parent(s):
a589ae8
huggingartists
Browse files- README.md +3 -3
- evaluation.txt +1 -1
- flax_model.msgpack +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +489 -509
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/eminem")
|
| 46 |
```
|
| 47 |
|
| 48 |
-
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Eminem's lyrics.
|
| 53 |
|
| 54 |
-
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 55 |
|
| 56 |
-
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
|
|
|
| 45 |
dataset = load_dataset("huggingartists/eminem")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/s2mjtgo8/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Eminem's lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/zikdyrc6) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/zikdyrc6/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss": 0.
|
|
|
|
| 1 |
+
{"eval_loss": 0.20774193108081818, "eval_runtime": 15.0542, "eval_samples_per_second": 42.978, "eval_steps_per_second": 5.381, "epoch": 3.0}
|
flax_model.msgpack
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 497764120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:758c6b87d8212253a5334ae7f98913538370445564928cae53eff8b1f96ba0d1
|
| 3 |
size 497764120
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 995604017
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a24e99d4dc7013f43cbadb32badf86893275bcd52f36e3a3e6188836728497e1
|
| 3 |
size 995604017
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 510396521
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:470bb080f7f86a03fc1b9714d4c1fcfd018e4a30715683ad366a847fc90b25a7
|
| 3 |
size 510396521
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14567
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1dd4f3adbf918816f8bc24c02487c37270992d6dc1de5e29ef3b17931ced7d9e
|
| 3 |
size 14567
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2eae8c7bd95e47c6b3eb60b363caee41d6c38cf405dd6ff23467a076f503128e
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,1138 +1,1118 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "output/eminem/checkpoint-
|
| 4 |
"epoch": 2.0,
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.01,
|
| 12 |
-
"learning_rate":
|
| 13 |
-
"loss": 0.
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.02,
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 0.
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.03,
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss": 0.
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.04,
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss": 0.
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.07,
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.08,
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss": 0.
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.09,
|
| 54 |
-
"learning_rate":
|
| 55 |
-
"loss": 0.
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.1,
|
| 60 |
-
"learning_rate":
|
| 61 |
-
"loss": 0.
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.11,
|
| 66 |
-
"learning_rate":
|
| 67 |
-
"loss": 0.
|
| 68 |
"step": 50
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.12,
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss": 0.
|
| 74 |
"step": 55
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.13,
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 60
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.14,
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 65
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.15,
|
| 90 |
-
"learning_rate":
|
| 91 |
-
"loss": 0.
|
| 92 |
"step": 70
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
-
"learning_rate":
|
| 97 |
-
"loss": 0.
|
| 98 |
"step": 75
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.18,
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 0.
|
| 104 |
"step": 80
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.19,
|
| 108 |
-
"learning_rate":
|
| 109 |
-
"loss": 0.
|
| 110 |
"step": 85
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.2,
|
| 114 |
-
"learning_rate":
|
| 115 |
-
"loss": 0.
|
| 116 |
"step": 90
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.21,
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 95
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.22,
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 100
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.23,
|
| 132 |
-
"learning_rate":
|
| 133 |
-
"loss": 0.
|
| 134 |
"step": 105
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.24,
|
| 138 |
-
"learning_rate":
|
| 139 |
-
"loss": 0.
|
| 140 |
"step": 110
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.25,
|
| 144 |
-
"learning_rate":
|
| 145 |
-
"loss": 0.
|
| 146 |
"step": 115
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.26,
|
| 150 |
-
"learning_rate":
|
| 151 |
-
"loss": 0.
|
| 152 |
"step": 120
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
-
"learning_rate":
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 125
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.29,
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 130
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.3,
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 135
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.31,
|
| 174 |
-
"learning_rate":
|
| 175 |
-
"loss": 0.
|
| 176 |
"step": 140
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.32,
|
| 180 |
-
"learning_rate":
|
| 181 |
-
"loss": 0.
|
| 182 |
"step": 145
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.33,
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss": 0.
|
| 188 |
"step": 150
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.34,
|
| 192 |
-
"learning_rate":
|
| 193 |
-
"loss": 0.
|
| 194 |
"step": 155
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.35,
|
| 198 |
-
"learning_rate":
|
| 199 |
-
"loss": 0.
|
| 200 |
"step": 160
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.36,
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 165
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.37,
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 170
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"learning_rate":
|
| 217 |
-
"loss": 0.
|
| 218 |
"step": 175
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
-
"learning_rate":
|
| 223 |
-
"loss": 0.
|
| 224 |
"step": 180
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.41,
|
| 228 |
-
"learning_rate":
|
| 229 |
-
"loss": 0.
|
| 230 |
"step": 185
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.42,
|
| 234 |
-
"learning_rate":
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 190
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.43,
|
| 240 |
-
"learning_rate":
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 195
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.44,
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 200
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.45,
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 205
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.46,
|
| 258 |
-
"learning_rate":
|
| 259 |
-
"loss": 0.
|
| 260 |
"step": 210
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.47,
|
| 264 |
-
"learning_rate":
|
| 265 |
-
"loss": 0.
|
| 266 |
"step": 215
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.48,
|
| 270 |
-
"learning_rate":
|
| 271 |
-
"loss": 0.
|
| 272 |
"step": 220
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch": 0.
|
| 276 |
-
"learning_rate":
|
| 277 |
-
"loss": 0.
|
| 278 |
"step": 225
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch": 0.
|
| 282 |
-
"learning_rate":
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 230
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.52,
|
| 288 |
-
"learning_rate": 9.
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 235
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.53,
|
| 294 |
-
"learning_rate": 9.
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 240
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.54,
|
| 300 |
-
"learning_rate":
|
| 301 |
-
"loss": 0.
|
| 302 |
"step": 245
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.55,
|
| 306 |
-
"learning_rate":
|
| 307 |
-
"loss": 0.
|
| 308 |
"step": 250
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.56,
|
| 312 |
-
"learning_rate":
|
| 313 |
-
"loss": 0.
|
| 314 |
"step": 255
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.57,
|
| 318 |
-
"learning_rate":
|
| 319 |
-
"loss": 0.
|
| 320 |
"step": 260
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.58,
|
| 324 |
-
"learning_rate":
|
| 325 |
-
"loss": 0.
|
| 326 |
"step": 265
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.59,
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 270
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"learning_rate":
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 275
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"learning_rate":
|
| 343 |
-
"loss": 0.
|
| 344 |
"step": 280
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch": 0.
|
| 348 |
-
"learning_rate":
|
| 349 |
-
"loss": 0.
|
| 350 |
"step": 285
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.64,
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss": 0.
|
| 356 |
"step": 290
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.65,
|
| 360 |
-
"learning_rate":
|
| 361 |
-
"loss": 0.
|
| 362 |
"step": 295
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.66,
|
| 366 |
-
"learning_rate":
|
| 367 |
-
"loss": 0.
|
| 368 |
"step": 300
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.67,
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 305
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.68,
|
| 378 |
-
"learning_rate":
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 310
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.69,
|
| 384 |
-
"learning_rate":
|
| 385 |
-
"loss": 0.
|
| 386 |
"step": 315
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.7,
|
| 390 |
-
"learning_rate":
|
| 391 |
-
"loss": 0.
|
| 392 |
"step": 320
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch": 0.
|
| 396 |
-
"learning_rate":
|
| 397 |
-
"loss": 0.
|
| 398 |
"step": 325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch": 0.
|
| 402 |
-
"learning_rate":
|
| 403 |
-
"loss": 0.
|
| 404 |
"step": 330
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch": 0.
|
| 408 |
-
"learning_rate":
|
| 409 |
-
"loss": 0.
|
| 410 |
"step": 335
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.75,
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 340
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.76,
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 345
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.77,
|
| 426 |
-
"learning_rate":
|
| 427 |
-
"loss": 0.
|
| 428 |
"step": 350
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.78,
|
| 432 |
-
"learning_rate":
|
| 433 |
-
"loss": 0.
|
| 434 |
"step": 355
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.79,
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss": 0.
|
| 440 |
"step": 360
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.8,
|
| 444 |
-
"learning_rate":
|
| 445 |
-
"loss": 0.
|
| 446 |
"step": 365
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.81,
|
| 450 |
-
"learning_rate":
|
| 451 |
-
"loss": 0.
|
| 452 |
"step": 370
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 375
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 380
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"learning_rate":
|
| 469 |
-
"loss": 0.
|
| 470 |
"step": 385
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.86,
|
| 474 |
-
"learning_rate":
|
| 475 |
-
"loss": 0.
|
| 476 |
"step": 390
|
| 477 |
},
|
| 478 |
{
|
| 479 |
"epoch": 0.87,
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 0.
|
| 482 |
"step": 395
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.88,
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss": 0.
|
| 488 |
"step": 400
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.89,
|
| 492 |
-
"learning_rate":
|
| 493 |
-
"loss": 0.
|
| 494 |
"step": 405
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.9,
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 410
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.91,
|
| 504 |
-
"learning_rate":
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 415
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 420
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch": 0.
|
| 516 |
-
"learning_rate":
|
| 517 |
-
"loss": 0.
|
| 518 |
"step": 425
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch": 0.
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 0.
|
| 524 |
"step": 430
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch": 0.
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 435
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch": 0.
|
| 534 |
-
"learning_rate":
|
| 535 |
-
"loss": 0.
|
| 536 |
"step": 440
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.98,
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 445
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.99,
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 450
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 1.0,
|
| 552 |
-
"
|
| 553 |
-
"
|
| 554 |
-
"
|
| 555 |
-
|
| 556 |
-
|
| 557 |
-
"epoch": 1.0,
|
| 558 |
-
"eval_loss": 0.200975701212883,
|
| 559 |
-
"eval_runtime": 14.7931,
|
| 560 |
-
"eval_samples_per_second": 42.993,
|
| 561 |
-
"eval_steps_per_second": 5.408,
|
| 562 |
-
"step": 456
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 1.0,
|
| 566 |
-
"
|
| 567 |
-
"
|
| 568 |
-
"
|
| 569 |
-
"eval_steps_per_second": 5.98,
|
| 570 |
-
"step": 459
|
| 571 |
},
|
| 572 |
{
|
| 573 |
-
"epoch": 1.
|
| 574 |
-
"learning_rate":
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 460
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 1.02,
|
| 580 |
-
"learning_rate":
|
| 581 |
-
"loss": 0.
|
| 582 |
"step": 465
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"epoch": 1.
|
| 586 |
-
"learning_rate":
|
| 587 |
-
"loss": 0.
|
| 588 |
"step": 470
|
| 589 |
},
|
| 590 |
{
|
| 591 |
-
"epoch": 1.
|
| 592 |
-
"learning_rate":
|
| 593 |
-
"loss": 0.
|
| 594 |
"step": 475
|
| 595 |
},
|
| 596 |
{
|
| 597 |
-
"epoch": 1.
|
| 598 |
-
"learning_rate":
|
| 599 |
-
"loss": 0.
|
| 600 |
"step": 480
|
| 601 |
},
|
| 602 |
{
|
| 603 |
-
"epoch": 1.
|
| 604 |
-
"learning_rate":
|
| 605 |
-
"loss": 0.
|
| 606 |
"step": 485
|
| 607 |
},
|
| 608 |
{
|
| 609 |
-
"epoch": 1.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 490
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch": 1.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 495
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch": 1.
|
| 622 |
-
"learning_rate":
|
| 623 |
-
"loss": 0.
|
| 624 |
"step": 500
|
| 625 |
},
|
| 626 |
{
|
| 627 |
-
"epoch": 1.
|
| 628 |
-
"learning_rate":
|
| 629 |
-
"loss": 0.
|
| 630 |
"step": 505
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch": 1.
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss": 0.
|
| 636 |
"step": 510
|
| 637 |
},
|
| 638 |
{
|
| 639 |
-
"epoch": 1.
|
| 640 |
-
"learning_rate":
|
| 641 |
-
"loss": 0.
|
| 642 |
"step": 515
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"epoch": 1.
|
| 646 |
-
"learning_rate":
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 520
|
| 649 |
},
|
| 650 |
{
|
| 651 |
-
"epoch": 1.
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 525
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch": 1.
|
| 658 |
-
"learning_rate":
|
| 659 |
-
"loss": 0.
|
| 660 |
"step": 530
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch": 1.
|
| 664 |
-
"learning_rate":
|
| 665 |
-
"loss": 0.
|
| 666 |
"step": 535
|
| 667 |
},
|
| 668 |
{
|
| 669 |
-
"epoch": 1.
|
| 670 |
-
"learning_rate":
|
| 671 |
-
"loss": 0.
|
| 672 |
"step": 540
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"epoch": 1.
|
| 676 |
-
"learning_rate":
|
| 677 |
-
"loss": 0.
|
| 678 |
"step": 545
|
| 679 |
},
|
| 680 |
{
|
| 681 |
-
"epoch": 1.
|
| 682 |
-
"learning_rate":
|
| 683 |
-
"loss": 0.
|
| 684 |
"step": 550
|
| 685 |
},
|
| 686 |
{
|
| 687 |
-
"epoch": 1.
|
| 688 |
-
"learning_rate":
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 555
|
| 691 |
},
|
| 692 |
{
|
| 693 |
-
"epoch": 1.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 560
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch": 1.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 565
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch": 1.
|
| 706 |
-
"learning_rate":
|
| 707 |
-
"loss": 0.
|
| 708 |
"step": 570
|
| 709 |
},
|
| 710 |
{
|
| 711 |
-
"epoch": 1.
|
| 712 |
-
"learning_rate":
|
| 713 |
-
"loss": 0.
|
| 714 |
"step": 575
|
| 715 |
},
|
| 716 |
{
|
| 717 |
-
"epoch": 1.
|
| 718 |
-
"learning_rate":
|
| 719 |
-
"loss": 0.
|
| 720 |
"step": 580
|
| 721 |
},
|
| 722 |
{
|
| 723 |
-
"epoch": 1.
|
| 724 |
-
"learning_rate": 8.
|
| 725 |
-
"loss": 0.
|
| 726 |
"step": 585
|
| 727 |
},
|
| 728 |
{
|
| 729 |
-
"epoch": 1.
|
| 730 |
-
"learning_rate":
|
| 731 |
-
"loss": 0.
|
| 732 |
"step": 590
|
| 733 |
},
|
| 734 |
{
|
| 735 |
-
"epoch": 1.
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 595
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch": 1.
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 600
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch": 1.
|
| 748 |
-
"learning_rate":
|
| 749 |
-
"loss": 0.
|
| 750 |
"step": 605
|
| 751 |
},
|
| 752 |
{
|
| 753 |
-
"epoch": 1.
|
| 754 |
-
"learning_rate":
|
| 755 |
-
"loss": 0.
|
| 756 |
"step": 610
|
| 757 |
},
|
| 758 |
{
|
| 759 |
-
"epoch": 1.
|
| 760 |
-
"learning_rate":
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 615
|
| 763 |
},
|
| 764 |
{
|
| 765 |
-
"epoch": 1.
|
| 766 |
-
"learning_rate":
|
| 767 |
-
"loss": 0.
|
| 768 |
"step": 620
|
| 769 |
},
|
| 770 |
{
|
| 771 |
-
"epoch": 1.
|
| 772 |
-
"learning_rate":
|
| 773 |
-
"loss": 0.
|
| 774 |
"step": 625
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"epoch": 1.
|
| 778 |
-
"learning_rate":
|
| 779 |
-
"loss": 0.
|
| 780 |
"step": 630
|
| 781 |
},
|
| 782 |
{
|
| 783 |
-
"epoch": 1.
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 635
|
| 787 |
},
|
| 788 |
{
|
| 789 |
-
"epoch": 1.
|
| 790 |
-
"learning_rate": 6.
|
| 791 |
-
"loss": 0.
|
| 792 |
"step": 640
|
| 793 |
},
|
| 794 |
{
|
| 795 |
-
"epoch": 1.
|
| 796 |
-
"learning_rate":
|
| 797 |
-
"loss": 0.
|
| 798 |
"step": 645
|
| 799 |
},
|
| 800 |
{
|
| 801 |
-
"epoch": 1.
|
| 802 |
-
"learning_rate":
|
| 803 |
-
"loss": 0.
|
| 804 |
"step": 650
|
| 805 |
},
|
| 806 |
{
|
| 807 |
-
"epoch": 1.
|
| 808 |
-
"learning_rate":
|
| 809 |
-
"loss": 0.
|
| 810 |
"step": 655
|
| 811 |
},
|
| 812 |
{
|
| 813 |
-
"epoch": 1.
|
| 814 |
-
"learning_rate":
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 660
|
| 817 |
},
|
| 818 |
{
|
| 819 |
-
"epoch": 1.
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 0.
|
| 822 |
"step": 665
|
| 823 |
},
|
| 824 |
{
|
| 825 |
-
"epoch": 1.
|
| 826 |
-
"learning_rate":
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 670
|
| 829 |
},
|
| 830 |
{
|
| 831 |
-
"epoch": 1.
|
| 832 |
-
"learning_rate":
|
| 833 |
-
"loss": 0.
|
| 834 |
"step": 675
|
| 835 |
},
|
| 836 |
{
|
| 837 |
-
"epoch": 1.
|
| 838 |
-
"learning_rate":
|
| 839 |
-
"loss": 0.
|
| 840 |
"step": 680
|
| 841 |
},
|
| 842 |
{
|
| 843 |
-
"epoch": 1.
|
| 844 |
-
"learning_rate":
|
| 845 |
-
"loss": 0.
|
| 846 |
"step": 685
|
| 847 |
},
|
| 848 |
{
|
| 849 |
-
"epoch": 1.
|
| 850 |
-
"learning_rate":
|
| 851 |
-
"loss": 0.
|
| 852 |
"step": 690
|
| 853 |
},
|
| 854 |
{
|
| 855 |
-
"epoch": 1.
|
| 856 |
-
"learning_rate":
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 695
|
| 859 |
},
|
| 860 |
{
|
| 861 |
-
"epoch": 1.
|
| 862 |
-
"learning_rate":
|
| 863 |
-
"loss": 0.
|
| 864 |
"step": 700
|
| 865 |
},
|
| 866 |
{
|
| 867 |
-
"epoch": 1.
|
| 868 |
-
"learning_rate": 3.
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 705
|
| 871 |
},
|
| 872 |
{
|
| 873 |
-
"epoch": 1.
|
| 874 |
-
"learning_rate": 3.
|
| 875 |
-
"loss": 0.
|
| 876 |
"step": 710
|
| 877 |
},
|
| 878 |
{
|
| 879 |
-
"epoch": 1.
|
| 880 |
-
"learning_rate":
|
| 881 |
-
"loss": 0.
|
| 882 |
"step": 715
|
| 883 |
},
|
| 884 |
{
|
| 885 |
-
"epoch": 1.
|
| 886 |
-
"learning_rate":
|
| 887 |
-
"loss": 0.
|
| 888 |
"step": 720
|
| 889 |
},
|
| 890 |
{
|
| 891 |
-
"epoch": 1.
|
| 892 |
-
"learning_rate":
|
| 893 |
-
"loss": 0.
|
| 894 |
"step": 725
|
| 895 |
},
|
| 896 |
{
|
| 897 |
-
"epoch": 1.
|
| 898 |
-
"learning_rate":
|
| 899 |
-
"loss": 0.
|
| 900 |
"step": 730
|
| 901 |
},
|
| 902 |
{
|
| 903 |
-
"epoch": 1.
|
| 904 |
-
"learning_rate":
|
| 905 |
-
"loss": 0.
|
| 906 |
"step": 735
|
| 907 |
},
|
| 908 |
{
|
| 909 |
-
"epoch": 1.
|
| 910 |
-
"learning_rate":
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 740
|
| 913 |
},
|
| 914 |
{
|
| 915 |
-
"epoch": 1.
|
| 916 |
-
"learning_rate": 1.
|
| 917 |
-
"loss": 0.
|
| 918 |
"step": 745
|
| 919 |
},
|
| 920 |
{
|
| 921 |
-
"epoch": 1.
|
| 922 |
-
"learning_rate": 1.
|
| 923 |
-
"loss": 0.
|
| 924 |
"step": 750
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"epoch": 1.
|
| 928 |
-
"learning_rate": 1.
|
| 929 |
-
"loss": 0.
|
| 930 |
"step": 755
|
| 931 |
},
|
| 932 |
{
|
| 933 |
-
"epoch": 1.
|
| 934 |
-
"learning_rate": 1.
|
| 935 |
-
"loss": 0.
|
| 936 |
"step": 760
|
| 937 |
},
|
| 938 |
{
|
| 939 |
-
"epoch": 1.
|
| 940 |
-
"learning_rate": 1.
|
| 941 |
-
"loss": 0.
|
| 942 |
"step": 765
|
| 943 |
},
|
| 944 |
{
|
| 945 |
-
"epoch": 1.
|
| 946 |
-
"learning_rate": 1.
|
| 947 |
-
"loss": 0.
|
| 948 |
"step": 770
|
| 949 |
},
|
| 950 |
{
|
| 951 |
-
"epoch": 1.
|
| 952 |
-
"learning_rate":
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 775
|
| 955 |
},
|
| 956 |
{
|
| 957 |
-
"epoch": 1.
|
| 958 |
-
"learning_rate":
|
| 959 |
-
"loss": 0.
|
| 960 |
"step": 780
|
| 961 |
},
|
| 962 |
{
|
| 963 |
-
"epoch": 1.
|
| 964 |
-
"learning_rate":
|
| 965 |
-
"loss": 0.
|
| 966 |
"step": 785
|
| 967 |
},
|
| 968 |
{
|
| 969 |
-
"epoch": 1.
|
| 970 |
-
"learning_rate":
|
| 971 |
-
"loss": 0.
|
| 972 |
"step": 790
|
| 973 |
},
|
| 974 |
{
|
| 975 |
-
"epoch": 1.
|
| 976 |
-
"learning_rate":
|
| 977 |
-
"loss": 0.
|
| 978 |
"step": 795
|
| 979 |
},
|
| 980 |
{
|
| 981 |
-
"epoch": 1.
|
| 982 |
-
"learning_rate":
|
| 983 |
-
"loss": 0.
|
| 984 |
"step": 800
|
| 985 |
},
|
| 986 |
{
|
| 987 |
-
"epoch": 1.
|
| 988 |
-
"learning_rate": 3.
|
| 989 |
-
"loss": 0.
|
| 990 |
"step": 805
|
| 991 |
},
|
| 992 |
{
|
| 993 |
-
"epoch": 1.
|
| 994 |
-
"learning_rate": 3.
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 810
|
| 997 |
},
|
| 998 |
{
|
| 999 |
-
"epoch": 1.
|
| 1000 |
-
"learning_rate":
|
| 1001 |
-
"loss": 0.
|
| 1002 |
"step": 815
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
-
"epoch": 1.
|
| 1006 |
-
"learning_rate":
|
| 1007 |
-
"loss": 0.
|
| 1008 |
"step": 820
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
-
"epoch": 1.
|
| 1012 |
-
"learning_rate":
|
| 1013 |
-
"loss": 0.
|
| 1014 |
"step": 825
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
-
"epoch": 1.
|
| 1018 |
-
"learning_rate":
|
| 1019 |
-
"loss": 0.
|
| 1020 |
"step": 830
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
-
"epoch": 1.
|
| 1024 |
-
"learning_rate":
|
| 1025 |
-
"loss": 0.
|
| 1026 |
"step": 835
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
-
"epoch": 1.
|
| 1030 |
-
"learning_rate":
|
| 1031 |
-
"loss": 0.
|
| 1032 |
"step": 840
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
-
"epoch": 1.
|
| 1036 |
-
"learning_rate":
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 845
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
-
"epoch": 1.
|
| 1042 |
-
"learning_rate":
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 850
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"epoch": 1.
|
| 1048 |
-
"learning_rate":
|
| 1049 |
-
"loss": 0.
|
| 1050 |
"step": 855
|
| 1051 |
},
|
| 1052 |
{
|
| 1053 |
-
"epoch": 1.
|
| 1054 |
-
"learning_rate": 5.
|
| 1055 |
-
"loss": 0.
|
| 1056 |
"step": 860
|
| 1057 |
},
|
| 1058 |
{
|
| 1059 |
-
"epoch": 1.
|
| 1060 |
-
"learning_rate":
|
| 1061 |
-
"loss": 0.
|
| 1062 |
"step": 865
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
-
"epoch": 1.
|
| 1066 |
-
"learning_rate":
|
| 1067 |
-
"loss": 0.
|
| 1068 |
"step": 870
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
-
"epoch": 1.
|
| 1072 |
-
"learning_rate":
|
| 1073 |
-
"loss": 0.
|
| 1074 |
"step": 875
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
-
"epoch": 1.
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 880
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch": 1.
|
| 1084 |
-
"learning_rate":
|
| 1085 |
-
"loss": 0.
|
| 1086 |
"step": 885
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
-
"epoch": 1.
|
| 1090 |
-
"learning_rate":
|
| 1091 |
-
"loss": 0.
|
| 1092 |
"step": 890
|
| 1093 |
},
|
| 1094 |
-
{
|
| 1095 |
-
"epoch": 1.95,
|
| 1096 |
-
"learning_rate": 7.142252079274891e-05,
|
| 1097 |
-
"loss": 0.4096,
|
| 1098 |
-
"step": 895
|
| 1099 |
-
},
|
| 1100 |
{
|
| 1101 |
"epoch": 1.97,
|
| 1102 |
-
"learning_rate":
|
| 1103 |
-
"loss": 0.
|
| 1104 |
-
"step":
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 1.98,
|
| 1108 |
-
"learning_rate":
|
| 1109 |
-
"loss": 0.
|
| 1110 |
-
"step":
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
"epoch": 1.99,
|
| 1114 |
-
"learning_rate":
|
| 1115 |
-
"loss": 0.
|
| 1116 |
-
"step":
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 2.0,
|
| 1120 |
-
"learning_rate": 8.076962465051111e-05,
|
| 1121 |
-
"loss": 0.3749,
|
| 1122 |
-
"step": 915
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 2.0,
|
| 1126 |
-
"eval_loss": 0.
|
| 1127 |
-
"eval_runtime": 15.
|
| 1128 |
-
"eval_samples_per_second":
|
| 1129 |
-
"eval_steps_per_second": 5.
|
| 1130 |
-
"step":
|
| 1131 |
}
|
| 1132 |
],
|
| 1133 |
-
"max_steps":
|
| 1134 |
-
"num_train_epochs":
|
| 1135 |
-
"total_flos":
|
| 1136 |
"trial_name": null,
|
| 1137 |
"trial_params": null
|
| 1138 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.20774193108081818,
|
| 3 |
+
"best_model_checkpoint": "output/eminem/checkpoint-908",
|
| 4 |
"epoch": 2.0,
|
| 5 |
+
"global_step": 908,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 5.638262768466246e-06,
|
| 13 |
+
"loss": 0.3685,
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 6.618083216571314e-06,
|
| 19 |
+
"loss": 0.32,
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.03,
|
| 24 |
+
"learning_rate": 7.672094408674214e-06,
|
| 25 |
+
"loss": 0.3424,
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.04,
|
| 30 |
+
"learning_rate": 8.79903472075949e-06,
|
| 31 |
+
"loss": 0.3602,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 9.997555234556058e-06,
|
| 37 |
+
"loss": 0.3777,
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 1.126622135214757e-05,
|
| 43 |
+
"loss": 0.374,
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 1.2603514513152213e-05,
|
| 49 |
+
"loss": 0.3274,
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.09,
|
| 54 |
+
"learning_rate": 1.4007834012404736e-05,
|
| 55 |
+
"loss": 0.3659,
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.1,
|
| 60 |
+
"learning_rate": 1.5477498915945268e-05,
|
| 61 |
+
"loss": 0.3498,
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.11,
|
| 66 |
+
"learning_rate": 1.7010750073075836e-05,
|
| 67 |
+
"loss": 0.3689,
|
| 68 |
"step": 50
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.12,
|
| 72 |
+
"learning_rate": 1.8605752221991583e-05,
|
| 73 |
+
"loss": 0.3778,
|
| 74 |
"step": 55
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.13,
|
| 78 |
+
"learning_rate": 2.026059618656492e-05,
|
| 79 |
+
"loss": 0.3858,
|
| 80 |
"step": 60
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.14,
|
| 84 |
+
"learning_rate": 2.1973301161563144e-05,
|
| 85 |
+
"loss": 0.3659,
|
| 86 |
"step": 65
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.15,
|
| 90 |
+
"learning_rate": 2.3741817083632926e-05,
|
| 91 |
+
"loss": 0.3721,
|
| 92 |
"step": 70
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.17,
|
| 96 |
+
"learning_rate": 2.5564027085161653e-05,
|
| 97 |
+
"loss": 0.3581,
|
| 98 |
"step": 75
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.18,
|
| 102 |
+
"learning_rate": 2.7437750028124533e-05,
|
| 103 |
+
"loss": 0.3966,
|
| 104 |
"step": 80
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.19,
|
| 108 |
+
"learning_rate": 2.9360743114838212e-05,
|
| 109 |
+
"loss": 0.4075,
|
| 110 |
"step": 85
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.2,
|
| 114 |
+
"learning_rate": 3.1330704572545265e-05,
|
| 115 |
+
"loss": 0.3213,
|
| 116 |
"step": 90
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.21,
|
| 120 |
+
"learning_rate": 3.334527640855557e-05,
|
| 121 |
+
"loss": 0.3886,
|
| 122 |
"step": 95
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.22,
|
| 126 |
+
"learning_rate": 3.540204723273326e-05,
|
| 127 |
+
"loss": 0.3776,
|
| 128 |
"step": 100
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.23,
|
| 132 |
+
"learning_rate": 3.7498555143833497e-05,
|
| 133 |
+
"loss": 0.369,
|
| 134 |
"step": 105
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.24,
|
| 138 |
+
"learning_rate": 3.963229067635737e-05,
|
| 139 |
+
"loss": 0.3304,
|
| 140 |
"step": 110
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.25,
|
| 144 |
+
"learning_rate": 4.180069980431396e-05,
|
| 145 |
+
"loss": 0.3778,
|
| 146 |
"step": 115
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.26,
|
| 150 |
+
"learning_rate": 4.400118699831243e-05,
|
| 151 |
+
"loss": 0.3336,
|
| 152 |
"step": 120
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.28,
|
| 156 |
+
"learning_rate": 4.623111833235935e-05,
|
| 157 |
+
"loss": 0.3594,
|
| 158 |
"step": 125
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.29,
|
| 162 |
+
"learning_rate": 4.848782463657604e-05,
|
| 163 |
+
"loss": 0.3779,
|
| 164 |
"step": 130
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.3,
|
| 168 |
+
"learning_rate": 5.076860469216043e-05,
|
| 169 |
+
"loss": 0.3485,
|
| 170 |
"step": 135
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.31,
|
| 174 |
+
"learning_rate": 5.3070728464635136e-05,
|
| 175 |
+
"loss": 0.3823,
|
| 176 |
"step": 140
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.32,
|
| 180 |
+
"learning_rate": 5.539144037166318e-05,
|
| 181 |
+
"loss": 0.3733,
|
| 182 |
"step": 145
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.33,
|
| 186 |
+
"learning_rate": 5.7727962581384004e-05,
|
| 187 |
+
"loss": 0.3473,
|
| 188 |
"step": 150
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.34,
|
| 192 |
+
"learning_rate": 6.007749833742314e-05,
|
| 193 |
+
"loss": 0.4018,
|
| 194 |
"step": 155
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.35,
|
| 198 |
+
"learning_rate": 6.243723530652164e-05,
|
| 199 |
+
"loss": 0.396,
|
| 200 |
"step": 160
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.36,
|
| 204 |
+
"learning_rate": 6.480434894484319e-05,
|
| 205 |
+
"loss": 0.3803,
|
| 206 |
"step": 165
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.37,
|
| 210 |
+
"learning_rate": 6.71760058788546e-05,
|
| 211 |
+
"loss": 0.3732,
|
| 212 |
"step": 170
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.39,
|
| 216 |
+
"learning_rate": 6.954936729683709e-05,
|
| 217 |
+
"loss": 0.3755,
|
| 218 |
"step": 175
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.4,
|
| 222 |
+
"learning_rate": 7.192159234682948e-05,
|
| 223 |
+
"loss": 0.3447,
|
| 224 |
"step": 180
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.41,
|
| 228 |
+
"learning_rate": 7.428984153708128e-05,
|
| 229 |
+
"loss": 0.3732,
|
| 230 |
"step": 185
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.42,
|
| 234 |
+
"learning_rate": 7.665128013484672e-05,
|
| 235 |
+
"loss": 0.3869,
|
| 236 |
"step": 190
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.43,
|
| 240 |
+
"learning_rate": 7.900308155947452e-05,
|
| 241 |
+
"loss": 0.387,
|
| 242 |
"step": 195
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.44,
|
| 246 |
+
"learning_rate": 8.134243076576889e-05,
|
| 247 |
+
"loss": 0.426,
|
| 248 |
"step": 200
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.45,
|
| 252 |
+
"learning_rate": 8.366652761350052e-05,
|
| 253 |
+
"loss": 0.3936,
|
| 254 |
"step": 205
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.46,
|
| 258 |
+
"learning_rate": 8.597259021913802e-05,
|
| 259 |
+
"loss": 0.4021,
|
| 260 |
"step": 210
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.47,
|
| 264 |
+
"learning_rate": 8.825785828564833e-05,
|
| 265 |
+
"loss": 0.3664,
|
| 266 |
"step": 215
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.48,
|
| 270 |
+
"learning_rate": 9.051959640653656e-05,
|
| 271 |
+
"loss": 0.3974,
|
| 272 |
"step": 220
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 0.5,
|
| 276 |
+
"learning_rate": 9.275509734003202e-05,
|
| 277 |
+
"loss": 0.3667,
|
| 278 |
"step": 225
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 0.51,
|
| 282 |
+
"learning_rate": 9.496168524960306e-05,
|
| 283 |
+
"loss": 0.4155,
|
| 284 |
"step": 230
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.52,
|
| 288 |
+
"learning_rate": 9.713671890684909e-05,
|
| 289 |
+
"loss": 0.371,
|
| 290 |
"step": 235
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.53,
|
| 294 |
+
"learning_rate": 9.927759485299701e-05,
|
| 295 |
+
"loss": 0.3468,
|
| 296 |
"step": 240
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.54,
|
| 300 |
+
"learning_rate": 0.00010138175051515928,
|
| 301 |
+
"loss": 0.4078,
|
| 302 |
"step": 245
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.55,
|
| 306 |
+
"learning_rate": 0.0001034466672736829,
|
| 307 |
+
"loss": 0.387,
|
| 308 |
"step": 250
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.56,
|
| 312 |
+
"learning_rate": 0.00010546987347685277,
|
| 313 |
+
"loss": 0.4011,
|
| 314 |
"step": 255
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.57,
|
| 318 |
+
"learning_rate": 0.00010744894739941007,
|
| 319 |
+
"loss": 0.4018,
|
| 320 |
"step": 260
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.58,
|
| 324 |
+
"learning_rate": 0.00010938152014129237,
|
| 325 |
+
"loss": 0.4338,
|
| 326 |
"step": 265
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.59,
|
| 330 |
+
"learning_rate": 0.00011126527846313597,
|
| 331 |
+
"loss": 0.4123,
|
| 332 |
"step": 270
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.61,
|
| 336 |
+
"learning_rate": 0.00011309796755517588,
|
| 337 |
+
"loss": 0.3855,
|
| 338 |
"step": 275
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.62,
|
| 342 |
+
"learning_rate": 0.00011487739373618224,
|
| 343 |
+
"loss": 0.4182,
|
| 344 |
"step": 280
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"epoch": 0.63,
|
| 348 |
+
"learning_rate": 0.00011660142707925318,
|
| 349 |
+
"loss": 0.3913,
|
| 350 |
"step": 285
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.64,
|
| 354 |
+
"learning_rate": 0.00011826800396126461,
|
| 355 |
+
"loss": 0.477,
|
| 356 |
"step": 290
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.65,
|
| 360 |
+
"learning_rate": 0.00011987512953299345,
|
| 361 |
+
"loss": 0.3962,
|
| 362 |
"step": 295
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.66,
|
| 366 |
+
"learning_rate": 0.00012142088010688345,
|
| 367 |
+
"loss": 0.4238,
|
| 368 |
"step": 300
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.67,
|
| 372 |
+
"learning_rate": 0.0001229034054596614,
|
| 373 |
+
"loss": 0.4407,
|
| 374 |
"step": 305
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.68,
|
| 378 |
+
"learning_rate": 0.00012432093104699607,
|
| 379 |
+
"loss": 0.4689,
|
| 380 |
"step": 310
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.69,
|
| 384 |
+
"learning_rate": 0.00012567176012759143,
|
| 385 |
+
"loss": 0.4264,
|
| 386 |
"step": 315
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.7,
|
| 390 |
+
"learning_rate": 0.0001269542757941326,
|
| 391 |
+
"loss": 0.3877,
|
| 392 |
"step": 320
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.72,
|
| 396 |
+
"learning_rate": 0.0001281669429086917,
|
| 397 |
+
"loss": 0.4585,
|
| 398 |
"step": 325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 0.73,
|
| 402 |
+
"learning_rate": 0.0001293083099402366,
|
| 403 |
+
"loss": 0.4342,
|
| 404 |
"step": 330
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 0.74,
|
| 408 |
+
"learning_rate": 0.0001303770107020854,
|
| 409 |
+
"loss": 0.4388,
|
| 410 |
"step": 335
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.75,
|
| 414 |
+
"learning_rate": 0.00013137176598719452,
|
| 415 |
+
"loss": 0.4933,
|
| 416 |
"step": 340
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.76,
|
| 420 |
+
"learning_rate": 0.00013229138509933145,
|
| 421 |
+
"loss": 0.4994,
|
| 422 |
"step": 345
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.77,
|
| 426 |
+
"learning_rate": 0.00013313476727831372,
|
| 427 |
+
"loss": 0.4215,
|
| 428 |
"step": 350
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.78,
|
| 432 |
+
"learning_rate": 0.00013390090301758416,
|
| 433 |
+
"loss": 0.4488,
|
| 434 |
"step": 355
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.79,
|
| 438 |
+
"learning_rate": 0.00013458887527257018,
|
| 439 |
+
"loss": 0.4785,
|
| 440 |
"step": 360
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.8,
|
| 444 |
+
"learning_rate": 0.0001351978605583545,
|
| 445 |
+
"loss": 0.444,
|
| 446 |
"step": 365
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.81,
|
| 450 |
+
"learning_rate": 0.00013572712993537543,
|
| 451 |
+
"loss": 0.409,
|
| 452 |
"step": 370
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.83,
|
| 456 |
+
"learning_rate": 0.00013617604988193778,
|
| 457 |
+
"loss": 0.4202,
|
| 458 |
"step": 375
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.84,
|
| 462 |
+
"learning_rate": 0.00013654408305253035,
|
| 463 |
+
"loss": 0.4465,
|
| 464 |
"step": 380
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.85,
|
| 468 |
+
"learning_rate": 0.0001368307889210095,
|
| 469 |
+
"loss": 0.487,
|
| 470 |
"step": 385
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.86,
|
| 474 |
+
"learning_rate": 0.0001370358243079002,
|
| 475 |
+
"loss": 0.4389,
|
| 476 |
"step": 390
|
| 477 |
},
|
| 478 |
{
|
| 479 |
"epoch": 0.87,
|
| 480 |
+
"learning_rate": 0.00013715894379117118,
|
| 481 |
+
"loss": 0.4761,
|
| 482 |
"step": 395
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.88,
|
| 486 |
+
"learning_rate": 0.0001372,
|
| 487 |
+
"loss": 0.4531,
|
| 488 |
"step": 400
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.89,
|
| 492 |
+
"learning_rate": 0.0001371589437911712,
|
| 493 |
+
"loss": 0.4367,
|
| 494 |
"step": 405
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.9,
|
| 498 |
+
"learning_rate": 0.0001370358243079003,
|
| 499 |
+
"loss": 0.4265,
|
| 500 |
"step": 410
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.91,
|
| 504 |
+
"learning_rate": 0.0001368307889210096,
|
| 505 |
+
"loss": 0.4721,
|
| 506 |
"step": 415
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.93,
|
| 510 |
+
"learning_rate": 0.00013654408305253054,
|
| 511 |
+
"loss": 0.3988,
|
| 512 |
"step": 420
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.94,
|
| 516 |
+
"learning_rate": 0.00013617604988193797,
|
| 517 |
+
"loss": 0.4344,
|
| 518 |
"step": 425
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.95,
|
| 522 |
+
"learning_rate": 0.00013572712993537567,
|
| 523 |
+
"loss": 0.4899,
|
| 524 |
"step": 430
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.96,
|
| 528 |
+
"learning_rate": 0.00013519786055835476,
|
| 529 |
+
"loss": 0.5019,
|
| 530 |
"step": 435
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.97,
|
| 534 |
+
"learning_rate": 0.0001345888752725705,
|
| 535 |
+
"loss": 0.4645,
|
| 536 |
"step": 440
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.98,
|
| 540 |
+
"learning_rate": 0.00013390090301758454,
|
| 541 |
+
"loss": 0.3885,
|
| 542 |
"step": 445
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.99,
|
| 546 |
+
"learning_rate": 0.0001331347672783141,
|
| 547 |
+
"loss": 0.426,
|
| 548 |
"step": 450
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 1.0,
|
| 552 |
+
"eval_loss": 0.21941067278385162,
|
| 553 |
+
"eval_runtime": 15.1157,
|
| 554 |
+
"eval_samples_per_second": 42.803,
|
| 555 |
+
"eval_steps_per_second": 5.359,
|
| 556 |
+
"step": 454
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.0,
|
| 560 |
+
"learning_rate": 0.00013229138509933188,
|
| 561 |
+
"loss": 0.4675,
|
| 562 |
+
"step": 455
|
|
|
|
|
|
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"epoch": 1.01,
|
| 566 |
+
"learning_rate": 0.00013137176598719498,
|
| 567 |
+
"loss": 0.3498,
|
| 568 |
"step": 460
|
| 569 |
},
|
| 570 |
{
|
| 571 |
"epoch": 1.02,
|
| 572 |
+
"learning_rate": 0.00013037701070208588,
|
| 573 |
+
"loss": 0.4028,
|
| 574 |
"step": 465
|
| 575 |
},
|
| 576 |
{
|
| 577 |
+
"epoch": 1.04,
|
| 578 |
+
"learning_rate": 0.00012930830994023715,
|
| 579 |
+
"loss": 0.3913,
|
| 580 |
"step": 470
|
| 581 |
},
|
| 582 |
{
|
| 583 |
+
"epoch": 1.05,
|
| 584 |
+
"learning_rate": 0.00012816694290869227,
|
| 585 |
+
"loss": 0.427,
|
| 586 |
"step": 475
|
| 587 |
},
|
| 588 |
{
|
| 589 |
+
"epoch": 1.06,
|
| 590 |
+
"learning_rate": 0.00012695427579413322,
|
| 591 |
+
"loss": 0.3914,
|
| 592 |
"step": 480
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"epoch": 1.07,
|
| 596 |
+
"learning_rate": 0.00012567176012759208,
|
| 597 |
+
"loss": 0.4128,
|
| 598 |
"step": 485
|
| 599 |
},
|
| 600 |
{
|
| 601 |
+
"epoch": 1.08,
|
| 602 |
+
"learning_rate": 0.00012432093104699674,
|
| 603 |
+
"loss": 0.3783,
|
| 604 |
"step": 490
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 1.09,
|
| 608 |
+
"learning_rate": 0.00012290340545966207,
|
| 609 |
+
"loss": 0.4026,
|
| 610 |
"step": 495
|
| 611 |
},
|
| 612 |
{
|
| 613 |
+
"epoch": 1.1,
|
| 614 |
+
"learning_rate": 0.00012142088010688418,
|
| 615 |
+
"loss": 0.3649,
|
| 616 |
"step": 500
|
| 617 |
},
|
| 618 |
{
|
| 619 |
+
"epoch": 1.11,
|
| 620 |
+
"learning_rate": 0.00011987512953299421,
|
| 621 |
+
"loss": 0.4037,
|
| 622 |
"step": 505
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"epoch": 1.12,
|
| 626 |
+
"learning_rate": 0.00011826800396126541,
|
| 627 |
+
"loss": 0.3905,
|
| 628 |
"step": 510
|
| 629 |
},
|
| 630 |
{
|
| 631 |
+
"epoch": 1.13,
|
| 632 |
+
"learning_rate": 0.000116601427079254,
|
| 633 |
+
"loss": 0.3741,
|
| 634 |
"step": 515
|
| 635 |
},
|
| 636 |
{
|
| 637 |
+
"epoch": 1.15,
|
| 638 |
+
"learning_rate": 0.00011487739373618308,
|
| 639 |
+
"loss": 0.3776,
|
| 640 |
"step": 520
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 1.16,
|
| 644 |
+
"learning_rate": 0.00011309796755517674,
|
| 645 |
+
"loss": 0.3915,
|
| 646 |
"step": 525
|
| 647 |
},
|
| 648 |
{
|
| 649 |
+
"epoch": 1.17,
|
| 650 |
+
"learning_rate": 0.00011126527846313685,
|
| 651 |
+
"loss": 0.3885,
|
| 652 |
"step": 530
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"epoch": 1.18,
|
| 656 |
+
"learning_rate": 0.00010938152014129329,
|
| 657 |
+
"loss": 0.3599,
|
| 658 |
"step": 535
|
| 659 |
},
|
| 660 |
{
|
| 661 |
+
"epoch": 1.19,
|
| 662 |
+
"learning_rate": 0.00010744894739941099,
|
| 663 |
+
"loss": 0.3955,
|
| 664 |
"step": 540
|
| 665 |
},
|
| 666 |
{
|
| 667 |
+
"epoch": 1.2,
|
| 668 |
+
"learning_rate": 0.00010546987347685372,
|
| 669 |
+
"loss": 0.3638,
|
| 670 |
"step": 545
|
| 671 |
},
|
| 672 |
{
|
| 673 |
+
"epoch": 1.21,
|
| 674 |
+
"learning_rate": 0.00010344666727368387,
|
| 675 |
+
"loss": 0.4055,
|
| 676 |
"step": 550
|
| 677 |
},
|
| 678 |
{
|
| 679 |
+
"epoch": 1.22,
|
| 680 |
+
"learning_rate": 0.00010138175051516028,
|
| 681 |
+
"loss": 0.3965,
|
| 682 |
"step": 555
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 1.23,
|
| 686 |
+
"learning_rate": 9.927759485299804e-05,
|
| 687 |
+
"loss": 0.3767,
|
| 688 |
"step": 560
|
| 689 |
},
|
| 690 |
{
|
| 691 |
+
"epoch": 1.24,
|
| 692 |
+
"learning_rate": 9.713671890685015e-05,
|
| 693 |
+
"loss": 0.4088,
|
| 694 |
"step": 565
|
| 695 |
},
|
| 696 |
{
|
| 697 |
+
"epoch": 1.26,
|
| 698 |
+
"learning_rate": 9.49616852496041e-05,
|
| 699 |
+
"loss": 0.4329,
|
| 700 |
"step": 570
|
| 701 |
},
|
| 702 |
{
|
| 703 |
+
"epoch": 1.27,
|
| 704 |
+
"learning_rate": 9.275509734003309e-05,
|
| 705 |
+
"loss": 0.3581,
|
| 706 |
"step": 575
|
| 707 |
},
|
| 708 |
{
|
| 709 |
+
"epoch": 1.28,
|
| 710 |
+
"learning_rate": 9.051959640653763e-05,
|
| 711 |
+
"loss": 0.4084,
|
| 712 |
"step": 580
|
| 713 |
},
|
| 714 |
{
|
| 715 |
+
"epoch": 1.29,
|
| 716 |
+
"learning_rate": 8.825785828564943e-05,
|
| 717 |
+
"loss": 0.3755,
|
| 718 |
"step": 585
|
| 719 |
},
|
| 720 |
{
|
| 721 |
+
"epoch": 1.3,
|
| 722 |
+
"learning_rate": 8.597259021913913e-05,
|
| 723 |
+
"loss": 0.4321,
|
| 724 |
"step": 590
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"epoch": 1.31,
|
| 728 |
+
"learning_rate": 8.366652761350163e-05,
|
| 729 |
+
"loss": 0.4021,
|
| 730 |
"step": 595
|
| 731 |
},
|
| 732 |
{
|
| 733 |
+
"epoch": 1.32,
|
| 734 |
+
"learning_rate": 8.134243076577001e-05,
|
| 735 |
+
"loss": 0.4115,
|
| 736 |
"step": 600
|
| 737 |
},
|
| 738 |
{
|
| 739 |
+
"epoch": 1.33,
|
| 740 |
+
"learning_rate": 7.900308155947565e-05,
|
| 741 |
+
"loss": 0.3358,
|
| 742 |
"step": 605
|
| 743 |
},
|
| 744 |
{
|
| 745 |
+
"epoch": 1.34,
|
| 746 |
+
"learning_rate": 7.665128013484786e-05,
|
| 747 |
+
"loss": 0.3576,
|
| 748 |
"step": 610
|
| 749 |
},
|
| 750 |
{
|
| 751 |
+
"epoch": 1.35,
|
| 752 |
+
"learning_rate": 7.428984153708243e-05,
|
| 753 |
+
"loss": 0.4326,
|
| 754 |
"step": 615
|
| 755 |
},
|
| 756 |
{
|
| 757 |
+
"epoch": 1.37,
|
| 758 |
+
"learning_rate": 7.192159234683063e-05,
|
| 759 |
+
"loss": 0.3627,
|
| 760 |
"step": 620
|
| 761 |
},
|
| 762 |
{
|
| 763 |
+
"epoch": 1.38,
|
| 764 |
+
"learning_rate": 6.954936729683823e-05,
|
| 765 |
+
"loss": 0.3891,
|
| 766 |
"step": 625
|
| 767 |
},
|
| 768 |
{
|
| 769 |
+
"epoch": 1.39,
|
| 770 |
+
"learning_rate": 6.717600587885573e-05,
|
| 771 |
+
"loss": 0.3983,
|
| 772 |
"step": 630
|
| 773 |
},
|
| 774 |
{
|
| 775 |
+
"epoch": 1.4,
|
| 776 |
+
"learning_rate": 6.480434894484433e-05,
|
| 777 |
+
"loss": 0.3504,
|
| 778 |
"step": 635
|
| 779 |
},
|
| 780 |
{
|
| 781 |
+
"epoch": 1.41,
|
| 782 |
+
"learning_rate": 6.243723530652277e-05,
|
| 783 |
+
"loss": 0.3806,
|
| 784 |
"step": 640
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"epoch": 1.42,
|
| 788 |
+
"learning_rate": 6.007749833742428e-05,
|
| 789 |
+
"loss": 0.3616,
|
| 790 |
"step": 645
|
| 791 |
},
|
| 792 |
{
|
| 793 |
+
"epoch": 1.43,
|
| 794 |
+
"learning_rate": 5.772796258138513e-05,
|
| 795 |
+
"loss": 0.4417,
|
| 796 |
"step": 650
|
| 797 |
},
|
| 798 |
{
|
| 799 |
+
"epoch": 1.44,
|
| 800 |
+
"learning_rate": 5.53914403716643e-05,
|
| 801 |
+
"loss": 0.4272,
|
| 802 |
"step": 655
|
| 803 |
},
|
| 804 |
{
|
| 805 |
+
"epoch": 1.45,
|
| 806 |
+
"learning_rate": 5.307072846463625e-05,
|
| 807 |
+
"loss": 0.3909,
|
| 808 |
"step": 660
|
| 809 |
},
|
| 810 |
{
|
| 811 |
+
"epoch": 1.46,
|
| 812 |
+
"learning_rate": 5.076860469216153e-05,
|
| 813 |
+
"loss": 0.3833,
|
| 814 |
"step": 665
|
| 815 |
},
|
| 816 |
{
|
| 817 |
+
"epoch": 1.48,
|
| 818 |
+
"learning_rate": 4.848782463657713e-05,
|
| 819 |
+
"loss": 0.3611,
|
| 820 |
"step": 670
|
| 821 |
},
|
| 822 |
{
|
| 823 |
+
"epoch": 1.49,
|
| 824 |
+
"learning_rate": 4.6231118332360436e-05,
|
| 825 |
+
"loss": 0.415,
|
| 826 |
"step": 675
|
| 827 |
},
|
| 828 |
{
|
| 829 |
+
"epoch": 1.5,
|
| 830 |
+
"learning_rate": 4.40011869983135e-05,
|
| 831 |
+
"loss": 0.4037,
|
| 832 |
"step": 680
|
| 833 |
},
|
| 834 |
{
|
| 835 |
+
"epoch": 1.51,
|
| 836 |
+
"learning_rate": 4.1800699804315014e-05,
|
| 837 |
+
"loss": 0.3836,
|
| 838 |
"step": 685
|
| 839 |
},
|
| 840 |
{
|
| 841 |
+
"epoch": 1.52,
|
| 842 |
+
"learning_rate": 3.9632290676358406e-05,
|
| 843 |
+
"loss": 0.3564,
|
| 844 |
"step": 690
|
| 845 |
},
|
| 846 |
{
|
| 847 |
+
"epoch": 1.53,
|
| 848 |
+
"learning_rate": 3.749855514383451e-05,
|
| 849 |
+
"loss": 0.4255,
|
| 850 |
"step": 695
|
| 851 |
},
|
| 852 |
{
|
| 853 |
+
"epoch": 1.54,
|
| 854 |
+
"learning_rate": 3.5402047232734255e-05,
|
| 855 |
+
"loss": 0.3377,
|
| 856 |
"step": 700
|
| 857 |
},
|
| 858 |
{
|
| 859 |
+
"epoch": 1.55,
|
| 860 |
+
"learning_rate": 3.334527640855654e-05,
|
| 861 |
+
"loss": 0.3671,
|
| 862 |
"step": 705
|
| 863 |
},
|
| 864 |
{
|
| 865 |
+
"epoch": 1.56,
|
| 866 |
+
"learning_rate": 3.133070457254623e-05,
|
| 867 |
+
"loss": 0.3735,
|
| 868 |
"step": 710
|
| 869 |
},
|
| 870 |
{
|
| 871 |
+
"epoch": 1.57,
|
| 872 |
+
"learning_rate": 2.9360743114839147e-05,
|
| 873 |
+
"loss": 0.3614,
|
| 874 |
"step": 715
|
| 875 |
},
|
| 876 |
{
|
| 877 |
+
"epoch": 1.59,
|
| 878 |
+
"learning_rate": 2.7437750028125448e-05,
|
| 879 |
+
"loss": 0.373,
|
| 880 |
"step": 720
|
| 881 |
},
|
| 882 |
{
|
| 883 |
+
"epoch": 1.6,
|
| 884 |
+
"learning_rate": 2.5564027085162544e-05,
|
| 885 |
+
"loss": 0.3795,
|
| 886 |
"step": 725
|
| 887 |
},
|
| 888 |
{
|
| 889 |
+
"epoch": 1.61,
|
| 890 |
+
"learning_rate": 2.3741817083633794e-05,
|
| 891 |
+
"loss": 0.3621,
|
| 892 |
"step": 730
|
| 893 |
},
|
| 894 |
{
|
| 895 |
+
"epoch": 1.62,
|
| 896 |
+
"learning_rate": 2.197330116156398e-05,
|
| 897 |
+
"loss": 0.3882,
|
| 898 |
"step": 735
|
| 899 |
},
|
| 900 |
{
|
| 901 |
+
"epoch": 1.63,
|
| 902 |
+
"learning_rate": 2.026059618656573e-05,
|
| 903 |
+
"loss": 0.3836,
|
| 904 |
"step": 740
|
| 905 |
},
|
| 906 |
{
|
| 907 |
+
"epoch": 1.64,
|
| 908 |
+
"learning_rate": 1.860575222199237e-05,
|
| 909 |
+
"loss": 0.3413,
|
| 910 |
"step": 745
|
| 911 |
},
|
| 912 |
{
|
| 913 |
+
"epoch": 1.65,
|
| 914 |
+
"learning_rate": 1.701075007307659e-05,
|
| 915 |
+
"loss": 0.3664,
|
| 916 |
"step": 750
|
| 917 |
},
|
| 918 |
{
|
| 919 |
+
"epoch": 1.66,
|
| 920 |
+
"learning_rate": 1.5477498915945983e-05,
|
| 921 |
+
"loss": 0.3463,
|
| 922 |
"step": 755
|
| 923 |
},
|
| 924 |
{
|
| 925 |
+
"epoch": 1.67,
|
| 926 |
+
"learning_rate": 1.4007834012405429e-05,
|
| 927 |
+
"loss": 0.3331,
|
| 928 |
"step": 760
|
| 929 |
},
|
| 930 |
{
|
| 931 |
+
"epoch": 1.69,
|
| 932 |
+
"learning_rate": 1.2603514513152868e-05,
|
| 933 |
+
"loss": 0.3815,
|
| 934 |
"step": 765
|
| 935 |
},
|
| 936 |
{
|
| 937 |
+
"epoch": 1.7,
|
| 938 |
+
"learning_rate": 1.1266221352148201e-05,
|
| 939 |
+
"loss": 0.3769,
|
| 940 |
"step": 770
|
| 941 |
},
|
| 942 |
{
|
| 943 |
+
"epoch": 1.71,
|
| 944 |
+
"learning_rate": 9.997555234556651e-06,
|
| 945 |
+
"loss": 0.4108,
|
| 946 |
"step": 775
|
| 947 |
},
|
| 948 |
{
|
| 949 |
+
"epoch": 1.72,
|
| 950 |
+
"learning_rate": 8.799034720760045e-06,
|
| 951 |
+
"loss": 0.3722,
|
| 952 |
"step": 780
|
| 953 |
},
|
| 954 |
{
|
| 955 |
+
"epoch": 1.73,
|
| 956 |
+
"learning_rate": 7.672094408674733e-06,
|
| 957 |
+
"loss": 0.4017,
|
| 958 |
"step": 785
|
| 959 |
},
|
| 960 |
{
|
| 961 |
+
"epoch": 1.74,
|
| 962 |
+
"learning_rate": 6.6180832165718085e-06,
|
| 963 |
+
"loss": 0.4178,
|
| 964 |
"step": 790
|
| 965 |
},
|
| 966 |
{
|
| 967 |
+
"epoch": 1.75,
|
| 968 |
+
"learning_rate": 5.638262768466695e-06,
|
| 969 |
+
"loss": 0.3899,
|
| 970 |
"step": 795
|
| 971 |
},
|
| 972 |
{
|
| 973 |
+
"epoch": 1.76,
|
| 974 |
+
"learning_rate": 4.733805883975914e-06,
|
| 975 |
+
"loss": 0.3741,
|
| 976 |
"step": 800
|
| 977 |
},
|
| 978 |
{
|
| 979 |
+
"epoch": 1.77,
|
| 980 |
+
"learning_rate": 3.905795174492382e-06,
|
| 981 |
+
"loss": 0.3831,
|
| 982 |
"step": 805
|
| 983 |
},
|
| 984 |
{
|
| 985 |
+
"epoch": 1.78,
|
| 986 |
+
"learning_rate": 3.15522174732089e-06,
|
| 987 |
+
"loss": 0.3682,
|
| 988 |
"step": 810
|
| 989 |
},
|
| 990 |
{
|
| 991 |
+
"epoch": 1.8,
|
| 992 |
+
"learning_rate": 2.4829840193538357e-06,
|
| 993 |
+
"loss": 0.391,
|
| 994 |
"step": 815
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"epoch": 1.81,
|
| 998 |
+
"learning_rate": 1.88988664168465e-06,
|
| 999 |
+
"loss": 0.3597,
|
| 1000 |
"step": 820
|
| 1001 |
},
|
| 1002 |
{
|
| 1003 |
+
"epoch": 1.82,
|
| 1004 |
+
"learning_rate": 1.3766395364657029e-06,
|
| 1005 |
+
"loss": 0.379,
|
| 1006 |
"step": 825
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
+
"epoch": 1.83,
|
| 1010 |
+
"learning_rate": 9.43857047145762e-07,
|
| 1011 |
+
"loss": 0.4247,
|
| 1012 |
"step": 830
|
| 1013 |
},
|
| 1014 |
{
|
| 1015 |
+
"epoch": 1.84,
|
| 1016 |
+
"learning_rate": 5.920572031218176e-07,
|
| 1017 |
+
"loss": 0.3705,
|
| 1018 |
"step": 835
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
+
"epoch": 1.85,
|
| 1022 |
+
"learning_rate": 3.2166109966507624e-07,
|
| 1023 |
+
"loss": 0.3148,
|
| 1024 |
"step": 840
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
+
"epoch": 1.86,
|
| 1028 |
+
"learning_rate": 1.3299239388725518e-07,
|
| 1029 |
+
"loss": 0.3331,
|
| 1030 |
"step": 845
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
+
"epoch": 1.87,
|
| 1034 |
+
"learning_rate": 2.6276917326658224e-08,
|
| 1035 |
+
"loss": 0.3802,
|
| 1036 |
"step": 850
|
| 1037 |
},
|
| 1038 |
{
|
| 1039 |
+
"epoch": 1.88,
|
| 1040 |
+
"learning_rate": 1.6424056362936757e-09,
|
| 1041 |
+
"loss": 0.324,
|
| 1042 |
"step": 855
|
| 1043 |
},
|
| 1044 |
{
|
| 1045 |
+
"epoch": 1.89,
|
| 1046 |
+
"learning_rate": 5.911834568650871e-08,
|
| 1047 |
+
"loss": 0.3751,
|
| 1048 |
"step": 860
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
+
"epoch": 1.91,
|
| 1052 |
+
"learning_rate": 1.9863594027063992e-07,
|
| 1053 |
+
"loss": 0.3792,
|
| 1054 |
"step": 865
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
+
"epoch": 1.92,
|
| 1058 |
+
"learning_rate": 4.200281904521944e-07,
|
| 1059 |
+
"loss": 0.3776,
|
| 1060 |
"step": 870
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
+
"epoch": 1.93,
|
| 1064 |
+
"learning_rate": 7.230300954609269e-07,
|
| 1065 |
+
"loss": 0.3469,
|
| 1066 |
"step": 875
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
+
"epoch": 1.94,
|
| 1070 |
+
"learning_rate": 1.1072789698879036e-06,
|
| 1071 |
+
"loss": 0.3464,
|
| 1072 |
"step": 880
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
+
"epoch": 1.95,
|
| 1076 |
+
"learning_rate": 1.572314877814697e-06,
|
| 1077 |
+
"loss": 0.3726,
|
| 1078 |
"step": 885
|
| 1079 |
},
|
| 1080 |
{
|
| 1081 |
+
"epoch": 1.96,
|
| 1082 |
+
"learning_rate": 2.11758118334118e-06,
|
| 1083 |
+
"loss": 0.3881,
|
| 1084 |
"step": 890
|
| 1085 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1086 |
{
|
| 1087 |
"epoch": 1.97,
|
| 1088 |
+
"learning_rate": 2.742425216867453e-06,
|
| 1089 |
+
"loss": 0.3518,
|
| 1090 |
+
"step": 895
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 1.98,
|
| 1094 |
+
"learning_rate": 3.4460990563188536e-06,
|
| 1095 |
+
"loss": 0.3839,
|
| 1096 |
+
"step": 900
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
"epoch": 1.99,
|
| 1100 |
+
"learning_rate": 4.227760422393333e-06,
|
| 1101 |
+
"loss": 0.347,
|
| 1102 |
+
"step": 905
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 2.0,
|
| 1106 |
+
"eval_loss": 0.20774193108081818,
|
| 1107 |
+
"eval_runtime": 15.111,
|
| 1108 |
+
"eval_samples_per_second": 42.816,
|
| 1109 |
+
"eval_steps_per_second": 5.36,
|
| 1110 |
+
"step": 908
|
| 1111 |
}
|
| 1112 |
],
|
| 1113 |
+
"max_steps": 1362,
|
| 1114 |
+
"num_train_epochs": 3,
|
| 1115 |
+
"total_flos": 948490076160000.0,
|
| 1116 |
"trial_name": null,
|
| 1117 |
"trial_params": null
|
| 1118 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3311
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59697cb8fe6b03cd91b8ebe02c7f7dde6782a826ae481ef82af6bcc802bdb131
|
| 3 |
size 3311
|