Commit
·
79978f0
1
Parent(s):
a091c3b
huggingartists
Browse files- README.md +3 -3
- config.json +1 -1
- 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 +241 -1371
- 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/1qrb0dkt/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/3htsm9rh) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/3htsm9rh/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
config.json
CHANGED
|
@@ -35,7 +35,7 @@
|
|
| 35 |
}
|
| 36 |
},
|
| 37 |
"torch_dtype": "float32",
|
| 38 |
-
"transformers_version": "4.
|
| 39 |
"use_cache": true,
|
| 40 |
"vocab_size": 50257
|
| 41 |
}
|
|
|
|
| 35 |
}
|
| 36 |
},
|
| 37 |
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.10.0",
|
| 39 |
"use_cache": true,
|
| 40 |
"vocab_size": 50257
|
| 41 |
}
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss": 3.
|
|
|
|
| 1 |
+
{"eval_loss": 3.2645678520202637, "eval_runtime": 27.753, "eval_samples_per_second": 22.16, "eval_steps_per_second": 2.774, "epoch": 4.0}
|
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 sha256:1378031fdf9482b5db27b2bc00af093181f9681f90ee767c0cf9b58c15f34e2d
|
| 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 sha256:71f6347398b9b70ec58ef5870588fabc343892795992a9cc55b88838578f9cb0
|
| 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 510403817
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2634e3dc54e0fcf5b59c873d6c31c539bd50f6a8ee764ed20d3596bee9a7a6b
|
| 3 |
size 510403817
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83f1ad5e80a1e13954ef55b306eeb22095173b0c8f2dc6ea02fadde78742155f
|
| 3 |
size 14503
|
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:d351be0cf1fb195e4b3d5eaeb5b8bdf81c2295b6997969fad1395e04446b38a1
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,1700 +1,570 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 3.
|
| 3 |
-
"best_model_checkpoint": "output/eminem/checkpoint-
|
| 4 |
-
"epoch":
|
| 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": 0.
|
| 13 |
-
"loss":
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.02,
|
| 18 |
-
"learning_rate": 0.
|
| 19 |
-
"loss":
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.03,
|
| 24 |
-
"learning_rate": 0.
|
| 25 |
-
"loss":
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.04,
|
| 30 |
-
"learning_rate": 0.
|
| 31 |
-
"loss": 3.
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss":
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.07,
|
| 42 |
-
"learning_rate": 0.
|
| 43 |
-
"loss":
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.08,
|
| 48 |
-
"learning_rate": 0.
|
| 49 |
-
"loss": 3.
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.09,
|
| 54 |
-
"learning_rate": 0.
|
| 55 |
-
"loss":
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.1,
|
| 60 |
-
"learning_rate": 0.
|
| 61 |
-
"loss":
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.11,
|
| 66 |
-
"learning_rate": 0.
|
| 67 |
-
"loss": 3.
|
| 68 |
"step": 50
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.12,
|
| 72 |
-
"learning_rate": 0.
|
| 73 |
-
"loss":
|
| 74 |
"step": 55
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.13,
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 3.
|
| 80 |
"step": 60
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.14,
|
| 84 |
-
"learning_rate": 0.
|
| 85 |
-
"loss": 3.
|
| 86 |
"step": 65
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.15,
|
| 90 |
-
"learning_rate": 0.
|
| 91 |
-
"loss": 3.
|
| 92 |
"step": 70
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
-
"learning_rate": 0.
|
| 97 |
-
"loss": 3.
|
| 98 |
"step": 75
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch": 0.
|
| 102 |
-
"learning_rate": 0.
|
| 103 |
-
"loss": 3.
|
| 104 |
"step": 80
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.19,
|
| 108 |
-
"learning_rate": 0.
|
| 109 |
-
"loss": 3.
|
| 110 |
"step": 85
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.2,
|
| 114 |
-
"learning_rate": 0.
|
| 115 |
-
"loss": 3.
|
| 116 |
"step": 90
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.21,
|
| 120 |
-
"learning_rate": 0.
|
| 121 |
-
"loss":
|
| 122 |
"step": 95
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.22,
|
| 126 |
-
"learning_rate": 0.
|
| 127 |
-
"loss": 3.
|
| 128 |
"step": 100
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.23,
|
| 132 |
-
"learning_rate": 0.
|
| 133 |
-
"loss":
|
| 134 |
"step": 105
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.24,
|
| 138 |
-
"learning_rate": 0.
|
| 139 |
-
"loss":
|
| 140 |
"step": 110
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.25,
|
| 144 |
-
"learning_rate": 0.
|
| 145 |
-
"loss": 3.
|
| 146 |
"step": 115
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.26,
|
| 150 |
-
"learning_rate": 0.
|
| 151 |
-
"loss": 3.
|
| 152 |
"step": 120
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
-
"learning_rate": 0.
|
| 157 |
-
"loss":
|
| 158 |
"step": 125
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 0.
|
| 162 |
-
"learning_rate": 0.
|
| 163 |
-
"loss": 3.
|
| 164 |
"step": 130
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"learning_rate": 0.
|
| 169 |
-
"loss": 3.
|
| 170 |
"step": 135
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.31,
|
| 174 |
-
"learning_rate": 0.
|
| 175 |
-
"loss": 3.
|
| 176 |
"step": 140
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.32,
|
| 180 |
-
"learning_rate": 0.
|
| 181 |
-
"loss": 3.
|
| 182 |
"step": 145
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.33,
|
| 186 |
-
"learning_rate": 0.
|
| 187 |
-
"loss":
|
| 188 |
"step": 150
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.34,
|
| 192 |
-
"learning_rate": 0.
|
| 193 |
-
"loss": 3.
|
| 194 |
"step": 155
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.35,
|
| 198 |
-
"learning_rate": 9.
|
| 199 |
-
"loss": 3.
|
| 200 |
"step": 160
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.36,
|
| 204 |
-
"learning_rate": 9.
|
| 205 |
-
"loss": 3.
|
| 206 |
"step": 165
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"learning_rate": 9.
|
| 211 |
-
"loss":
|
| 212 |
"step": 170
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"learning_rate": 9.
|
| 217 |
-
"loss": 3.
|
| 218 |
"step": 175
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
-
"learning_rate": 9.
|
| 223 |
-
"loss": 3.
|
| 224 |
"step": 180
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch": 0.
|
| 228 |
-
"learning_rate": 8.
|
| 229 |
-
"loss": 3.
|
| 230 |
"step": 185
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 0.
|
| 234 |
-
"learning_rate": 8.
|
| 235 |
-
"loss": 3.
|
| 236 |
"step": 190
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.43,
|
| 240 |
-
"learning_rate": 8.
|
| 241 |
-
"loss": 3.
|
| 242 |
"step": 195
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.44,
|
| 246 |
-
"learning_rate": 8.
|
| 247 |
-
"loss": 3.
|
| 248 |
"step": 200
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.45,
|
| 252 |
-
"learning_rate": 7.
|
| 253 |
-
"loss": 3.
|
| 254 |
"step": 205
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.46,
|
| 258 |
-
"learning_rate": 7.
|
| 259 |
-
"loss": 3.
|
| 260 |
"step": 210
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.47,
|
| 264 |
-
"learning_rate": 7.
|
| 265 |
-
"loss": 3.
|
| 266 |
"step": 215
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch": 0.
|
| 270 |
-
"learning_rate": 7.
|
| 271 |
-
"loss": 3.
|
| 272 |
"step": 220
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch": 0.
|
| 276 |
-
"learning_rate":
|
| 277 |
-
"loss": 3.
|
| 278 |
"step": 225
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch": 0.
|
| 282 |
-
"learning_rate": 6.
|
| 283 |
-
"loss":
|
| 284 |
"step": 230
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch": 0.
|
| 288 |
-
"learning_rate": 6.
|
| 289 |
-
"loss": 3.
|
| 290 |
"step": 235
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
-
"learning_rate": 6.
|
| 295 |
-
"loss": 3.
|
| 296 |
"step": 240
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"learning_rate":
|
| 301 |
-
"loss":
|
| 302 |
"step": 245
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.55,
|
| 306 |
-
"learning_rate": 5.
|
| 307 |
-
"loss": 3.
|
| 308 |
"step": 250
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.56,
|
| 312 |
-
"learning_rate": 5.
|
| 313 |
-
"loss": 3.
|
| 314 |
"step": 255
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.57,
|
| 318 |
-
"learning_rate": 5.
|
| 319 |
-
"loss": 3.
|
| 320 |
"step": 260
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.58,
|
| 324 |
-
"learning_rate": 5.
|
| 325 |
-
"loss":
|
| 326 |
"step": 265
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch": 0.
|
| 330 |
-
"learning_rate": 4.
|
| 331 |
-
"loss": 3.
|
| 332 |
"step": 270
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"learning_rate": 4.
|
| 337 |
-
"loss": 3.
|
| 338 |
"step": 275
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"learning_rate": 4.
|
| 343 |
-
"loss": 3.
|
| 344 |
"step": 280
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch": 0.
|
| 348 |
-
"learning_rate": 4.
|
| 349 |
-
"loss":
|
| 350 |
"step": 285
|
| 351 |
},
|
| 352 |
{
|
| 353 |
-
"epoch": 0.
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss": 3.
|
| 356 |
"step": 290
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch": 0.
|
| 360 |
-
"learning_rate": 3.
|
| 361 |
-
"loss": 3.
|
| 362 |
"step": 295
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.66,
|
| 366 |
-
"learning_rate": 3.
|
| 367 |
-
"loss": 3.
|
| 368 |
"step": 300
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.67,
|
| 372 |
-
"learning_rate": 3.
|
| 373 |
-
"loss": 3.
|
| 374 |
"step": 305
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.68,
|
| 378 |
-
"learning_rate": 3.
|
| 379 |
-
"loss":
|
| 380 |
"step": 310
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
-
"learning_rate":
|
| 385 |
-
"loss": 3.
|
| 386 |
"step": 315
|
| 387 |
},
|
| 388 |
{
|
| 389 |
-
"epoch": 0.
|
| 390 |
-
"learning_rate": 2.
|
| 391 |
-
"loss":
|
| 392 |
"step": 320
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch": 0.
|
| 396 |
-
"learning_rate": 2.
|
| 397 |
-
"loss":
|
| 398 |
"step": 325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
-
"epoch": 0.
|
| 402 |
-
"learning_rate": 2.
|
| 403 |
-
"loss":
|
| 404 |
"step": 330
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch": 0.
|
| 408 |
-
"learning_rate": 2.
|
| 409 |
-
"loss": 3.
|
| 410 |
"step": 335
|
| 411 |
},
|
| 412 |
{
|
| 413 |
-
"epoch": 0.
|
| 414 |
-
"learning_rate": 2.
|
| 415 |
-
"loss": 3.
|
| 416 |
"step": 340
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch": 0.
|
| 420 |
-
"learning_rate": 1.
|
| 421 |
-
"loss": 3.
|
| 422 |
"step": 345
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch": 0.
|
| 426 |
-
"learning_rate": 1.
|
| 427 |
-
"loss": 3.
|
| 428 |
"step": 350
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.78,
|
| 432 |
-
"learning_rate": 1.
|
| 433 |
-
"loss":
|
| 434 |
"step": 355
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.79,
|
| 438 |
-
"learning_rate": 1.
|
| 439 |
-
"loss":
|
| 440 |
"step": 360
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch": 0.
|
| 444 |
-
"learning_rate": 1.
|
| 445 |
-
"loss": 3.
|
| 446 |
"step": 365
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch": 0.
|
| 450 |
-
"learning_rate": 1.
|
| 451 |
-
"loss": 3.
|
| 452 |
"step": 370
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 3.
|
| 458 |
"step": 375
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss":
|
| 464 |
"step": 380
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"learning_rate":
|
| 469 |
-
"loss": 3.
|
| 470 |
"step": 385
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch": 0.
|
| 474 |
-
"learning_rate":
|
| 475 |
-
"loss": 3.
|
| 476 |
"step": 390
|
| 477 |
},
|
| 478 |
{
|
| 479 |
-
"epoch": 0.
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 3.
|
| 482 |
"step": 395
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"epoch": 0.
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss":
|
| 488 |
"step": 400
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch": 0.
|
| 492 |
-
"learning_rate":
|
| 493 |
-
"loss":
|
| 494 |
"step": 405
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch": 0.
|
| 498 |
-
"learning_rate": 3.
|
| 499 |
-
"loss": 3.
|
| 500 |
"step": 410
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch": 0.
|
| 504 |
-
"learning_rate": 2.
|
| 505 |
-
"loss": 3.
|
| 506 |
"step": 415
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss": 3.
|
| 512 |
"step": 420
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch": 0.
|
| 516 |
-
"learning_rate": 1.
|
| 517 |
-
"loss": 3.
|
| 518 |
"step": 425
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch": 0.
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 3.
|
| 524 |
"step": 430
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch": 0.
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss": 3.
|
| 530 |
"step": 435
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch": 0.
|
| 534 |
-
"learning_rate":
|
| 535 |
-
"loss": 3.
|
| 536 |
"step": 440
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch": 0.
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss":
|
| 542 |
"step": 445
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch": 0.
|
| 546 |
-
"learning_rate": 1.
|
| 547 |
-
"loss":
|
| 548 |
"step": 450
|
| 549 |
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 1.0,
|
| 552 |
-
"eval_loss": 3.7048158645629883,
|
| 553 |
-
"eval_runtime": 29.3019,
|
| 554 |
-
"eval_samples_per_second": 22.388,
|
| 555 |
-
"eval_steps_per_second": 2.798,
|
| 556 |
-
"step": 453
|
| 557 |
-
},
|
| 558 |
{
|
| 559 |
"epoch": 0.99,
|
| 560 |
-
"learning_rate":
|
| 561 |
-
"loss": 3.
|
| 562 |
"step": 455
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 1.0,
|
| 566 |
-
"
|
| 567 |
-
"
|
| 568 |
-
"
|
| 569 |
-
|
| 570 |
-
|
| 571 |
-
"epoch": 1.0,
|
| 572 |
-
"eval_loss": 3.5555710792541504,
|
| 573 |
-
"eval_runtime": 13.5198,
|
| 574 |
-
"eval_samples_per_second": 44.675,
|
| 575 |
-
"eval_steps_per_second": 5.621,
|
| 576 |
-
"step": 460
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 1.01,
|
| 580 |
-
"learning_rate": 3.9992265680461966e-08,
|
| 581 |
-
"loss": 3.5614,
|
| 582 |
-
"step": 465
|
| 583 |
-
},
|
| 584 |
-
{
|
| 585 |
-
"epoch": 1.02,
|
| 586 |
-
"learning_rate": 1.5992243352901425e-07,
|
| 587 |
-
"loss": 3.7249,
|
| 588 |
-
"step": 470
|
| 589 |
-
},
|
| 590 |
-
{
|
| 591 |
-
"epoch": 1.03,
|
| 592 |
-
"learning_rate": 3.5965067033469397e-07,
|
| 593 |
-
"loss": 3.4991,
|
| 594 |
-
"step": 475
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 1.04,
|
| 598 |
-
"learning_rate": 6.389441019077102e-07,
|
| 599 |
-
"loss": 3.439,
|
| 600 |
-
"step": 480
|
| 601 |
-
},
|
| 602 |
-
{
|
| 603 |
-
"epoch": 1.05,
|
| 604 |
-
"learning_rate": 9.97477084600295e-07,
|
| 605 |
-
"loss": 3.6568,
|
| 606 |
-
"step": 485
|
| 607 |
-
},
|
| 608 |
-
{
|
| 609 |
-
"epoch": 1.07,
|
| 610 |
-
"learning_rate": 1.4348315849926483e-06,
|
| 611 |
-
"loss": 3.5489,
|
| 612 |
-
"step": 490
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 1.08,
|
| 616 |
-
"learning_rate": 1.9504976673012086e-06,
|
| 617 |
-
"loss": 3.6268,
|
| 618 |
-
"step": 495
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 1.09,
|
| 622 |
-
"learning_rate": 2.5438740879409643e-06,
|
| 623 |
-
"loss": 3.8246,
|
| 624 |
-
"step": 500
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 1.1,
|
| 628 |
-
"learning_rate": 3.2142689965485674e-06,
|
| 629 |
-
"loss": 3.5789,
|
| 630 |
-
"step": 505
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 1.11,
|
| 634 |
-
"learning_rate": 3.960900742648913e-06,
|
| 635 |
-
"loss": 3.618,
|
| 636 |
-
"step": 510
|
| 637 |
-
},
|
| 638 |
-
{
|
| 639 |
-
"epoch": 1.12,
|
| 640 |
-
"learning_rate": 4.782898787024646e-06,
|
| 641 |
-
"loss": 3.6802,
|
| 642 |
-
"step": 515
|
| 643 |
-
},
|
| 644 |
-
{
|
| 645 |
-
"epoch": 1.13,
|
| 646 |
-
"learning_rate": 5.679304716725914e-06,
|
| 647 |
-
"loss": 3.6302,
|
| 648 |
-
"step": 520
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 1.14,
|
| 652 |
-
"learning_rate": 6.649073362537048e-06,
|
| 653 |
-
"loss": 3.5926,
|
| 654 |
-
"step": 525
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 1.15,
|
| 658 |
-
"learning_rate": 7.691074017597052e-06,
|
| 659 |
-
"loss": 3.5005,
|
| 660 |
-
"step": 530
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 1.16,
|
| 664 |
-
"learning_rate": 8.804091755753263e-06,
|
| 665 |
-
"loss": 3.6247,
|
| 666 |
-
"step": 535
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"epoch": 1.17,
|
| 670 |
-
"learning_rate": 9.986828848110884e-06,
|
| 671 |
-
"loss": 3.5699,
|
| 672 |
-
"step": 540
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 1.18,
|
| 676 |
-
"learning_rate": 1.1237906276126821e-05,
|
| 677 |
-
"loss": 3.5435,
|
| 678 |
-
"step": 545
|
| 679 |
-
},
|
| 680 |
-
{
|
| 681 |
-
"epoch": 1.2,
|
| 682 |
-
"learning_rate": 1.2555865339483672e-05,
|
| 683 |
-
"loss": 3.5681,
|
| 684 |
-
"step": 550
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 1.21,
|
| 688 |
-
"learning_rate": 1.3939169356868945e-05,
|
| 689 |
-
"loss": 3.5438,
|
| 690 |
-
"step": 555
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 1.22,
|
| 694 |
-
"learning_rate": 1.5386205457676803e-05,
|
| 695 |
-
"loss": 3.6615,
|
| 696 |
-
"step": 560
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 1.23,
|
| 700 |
-
"learning_rate": 1.6895286462543014e-05,
|
| 701 |
-
"loss": 3.5334,
|
| 702 |
-
"step": 565
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 1.24,
|
| 706 |
-
"learning_rate": 1.8464652850520678e-05,
|
| 707 |
-
"loss": 3.5256,
|
| 708 |
-
"step": 570
|
| 709 |
-
},
|
| 710 |
-
{
|
| 711 |
-
"epoch": 1.25,
|
| 712 |
-
"learning_rate": 2.009247481060283e-05,
|
| 713 |
-
"loss": 3.5908,
|
| 714 |
-
"step": 575
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 1.26,
|
| 718 |
-
"learning_rate": 2.1776854375200328e-05,
|
| 719 |
-
"loss": 3.6817,
|
| 720 |
-
"step": 580
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 1.27,
|
| 724 |
-
"learning_rate": 2.351582763308709e-05,
|
| 725 |
-
"loss": 3.6254,
|
| 726 |
-
"step": 585
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 1.28,
|
| 730 |
-
"learning_rate": 2.5307367019232758e-05,
|
| 731 |
-
"loss": 3.5371,
|
| 732 |
-
"step": 590
|
| 733 |
-
},
|
| 734 |
-
{
|
| 735 |
-
"epoch": 1.29,
|
| 736 |
-
"learning_rate": 2.714938367885288e-05,
|
| 737 |
-
"loss": 3.6406,
|
| 738 |
-
"step": 595
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 1.3,
|
| 742 |
-
"learning_rate": 2.9039729902920112e-05,
|
| 743 |
-
"loss": 3.5787,
|
| 744 |
-
"step": 600
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 1.32,
|
| 748 |
-
"learning_rate": 3.097620163229676e-05,
|
| 749 |
-
"loss": 3.5732,
|
| 750 |
-
"step": 605
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 1.33,
|
| 754 |
-
"learning_rate": 3.295654102756921e-05,
|
| 755 |
-
"loss": 3.5091,
|
| 756 |
-
"step": 610
|
| 757 |
-
},
|
| 758 |
-
{
|
| 759 |
-
"epoch": 1.34,
|
| 760 |
-
"learning_rate": 3.4978439101588006e-05,
|
| 761 |
-
"loss": 3.617,
|
| 762 |
-
"step": 615
|
| 763 |
-
},
|
| 764 |
-
{
|
| 765 |
-
"epoch": 1.35,
|
| 766 |
-
"learning_rate": 3.703953841164292e-05,
|
| 767 |
-
"loss": 3.582,
|
| 768 |
-
"step": 620
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 1.36,
|
| 772 |
-
"learning_rate": 3.913743580813637e-05,
|
| 773 |
-
"loss": 3.6594,
|
| 774 |
-
"step": 625
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 1.37,
|
| 778 |
-
"learning_rate": 4.126968523654786e-05,
|
| 779 |
-
"loss": 3.6063,
|
| 780 |
-
"step": 630
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 1.38,
|
| 784 |
-
"learning_rate": 4.343380058942427e-05,
|
| 785 |
-
"loss": 3.6137,
|
| 786 |
-
"step": 635
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 1.39,
|
| 790 |
-
"learning_rate": 4.562725860507034e-05,
|
| 791 |
-
"loss": 3.5565,
|
| 792 |
-
"step": 640
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 1.4,
|
| 796 |
-
"learning_rate": 4.784750180955822e-05,
|
| 797 |
-
"loss": 3.6099,
|
| 798 |
-
"step": 645
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 1.41,
|
| 802 |
-
"learning_rate": 5.009194149862813e-05,
|
| 803 |
-
"loss": 3.6417,
|
| 804 |
-
"step": 650
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 1.42,
|
| 808 |
-
"learning_rate": 5.235796075600178e-05,
|
| 809 |
-
"loss": 3.7049,
|
| 810 |
-
"step": 655
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 1.43,
|
| 814 |
-
"learning_rate": 5.4642917504589275e-05,
|
| 815 |
-
"loss": 3.7319,
|
| 816 |
-
"step": 660
|
| 817 |
-
},
|
| 818 |
-
{
|
| 819 |
-
"epoch": 1.45,
|
| 820 |
-
"learning_rate": 5.694414758703346e-05,
|
| 821 |
-
"loss": 3.554,
|
| 822 |
-
"step": 665
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 1.46,
|
| 826 |
-
"learning_rate": 5.92589678719975e-05,
|
| 827 |
-
"loss": 3.5522,
|
| 828 |
-
"step": 670
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 1.47,
|
| 832 |
-
"learning_rate": 6.158467938257645e-05,
|
| 833 |
-
"loss": 3.6115,
|
| 834 |
-
"step": 675
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 1.48,
|
| 838 |
-
"learning_rate": 6.391857044318355e-05,
|
| 839 |
-
"loss": 3.6475,
|
| 840 |
-
"step": 680
|
| 841 |
-
},
|
| 842 |
-
{
|
| 843 |
-
"epoch": 1.49,
|
| 844 |
-
"learning_rate": 6.625791984124255e-05,
|
| 845 |
-
"loss": 3.5034,
|
| 846 |
-
"step": 685
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"epoch": 1.5,
|
| 850 |
-
"learning_rate": 6.859999999999999e-05,
|
| 851 |
-
"loss": 3.5767,
|
| 852 |
-
"step": 690
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 1.51,
|
| 856 |
-
"learning_rate": 7.094208015875743e-05,
|
| 857 |
-
"loss": 3.6471,
|
| 858 |
-
"step": 695
|
| 859 |
-
},
|
| 860 |
-
{
|
| 861 |
-
"epoch": 1.52,
|
| 862 |
-
"learning_rate": 7.328142955681643e-05,
|
| 863 |
-
"loss": 3.6515,
|
| 864 |
-
"step": 700
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 1.53,
|
| 868 |
-
"learning_rate": 7.561532061742353e-05,
|
| 869 |
-
"loss": 3.6649,
|
| 870 |
-
"step": 705
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 1.54,
|
| 874 |
-
"learning_rate": 7.794103212800247e-05,
|
| 875 |
-
"loss": 3.4103,
|
| 876 |
-
"step": 710
|
| 877 |
-
},
|
| 878 |
-
{
|
| 879 |
-
"epoch": 1.55,
|
| 880 |
-
"learning_rate": 8.025585241296653e-05,
|
| 881 |
-
"loss": 3.5488,
|
| 882 |
-
"step": 715
|
| 883 |
-
},
|
| 884 |
-
{
|
| 885 |
-
"epoch": 1.57,
|
| 886 |
-
"learning_rate": 8.25570824954107e-05,
|
| 887 |
-
"loss": 3.5076,
|
| 888 |
-
"step": 720
|
| 889 |
-
},
|
| 890 |
-
{
|
| 891 |
-
"epoch": 1.58,
|
| 892 |
-
"learning_rate": 8.484203924399819e-05,
|
| 893 |
-
"loss": 3.5708,
|
| 894 |
-
"step": 725
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 1.59,
|
| 898 |
-
"learning_rate": 8.710805850137184e-05,
|
| 899 |
-
"loss": 3.7272,
|
| 900 |
-
"step": 730
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"epoch": 1.6,
|
| 904 |
-
"learning_rate": 8.935249819044176e-05,
|
| 905 |
-
"loss": 3.6548,
|
| 906 |
-
"step": 735
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 1.61,
|
| 910 |
-
"learning_rate": 9.157274139492964e-05,
|
| 911 |
-
"loss": 3.6463,
|
| 912 |
-
"step": 740
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 1.62,
|
| 916 |
-
"learning_rate": 9.376619941057571e-05,
|
| 917 |
-
"loss": 3.484,
|
| 918 |
-
"step": 745
|
| 919 |
-
},
|
| 920 |
-
{
|
| 921 |
-
"epoch": 1.63,
|
| 922 |
-
"learning_rate": 9.593031476345212e-05,
|
| 923 |
-
"loss": 3.5106,
|
| 924 |
-
"step": 750
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 1.64,
|
| 928 |
-
"learning_rate": 9.80625641918636e-05,
|
| 929 |
-
"loss": 3.6845,
|
| 930 |
-
"step": 755
|
| 931 |
-
},
|
| 932 |
-
{
|
| 933 |
-
"epoch": 1.65,
|
| 934 |
-
"learning_rate": 0.00010016046158835706,
|
| 935 |
-
"loss": 3.5078,
|
| 936 |
-
"step": 760
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"epoch": 1.66,
|
| 940 |
-
"learning_rate": 0.00010222156089841198,
|
| 941 |
-
"loss": 3.62,
|
| 942 |
-
"step": 765
|
| 943 |
-
},
|
| 944 |
-
{
|
| 945 |
-
"epoch": 1.67,
|
| 946 |
-
"learning_rate": 0.00010424345897243078,
|
| 947 |
-
"loss": 3.644,
|
| 948 |
-
"step": 770
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 1.68,
|
| 952 |
-
"learning_rate": 0.00010622379836770322,
|
| 953 |
-
"loss": 3.7684,
|
| 954 |
-
"step": 775
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 1.7,
|
| 958 |
-
"learning_rate": 0.00010816027009707987,
|
| 959 |
-
"loss": 3.5978,
|
| 960 |
-
"step": 780
|
| 961 |
-
},
|
| 962 |
-
{
|
| 963 |
-
"epoch": 1.71,
|
| 964 |
-
"learning_rate": 0.0001100506163211471,
|
| 965 |
-
"loss": 3.5937,
|
| 966 |
-
"step": 785
|
| 967 |
-
},
|
| 968 |
-
{
|
| 969 |
-
"epoch": 1.72,
|
| 970 |
-
"learning_rate": 0.00011189263298076723,
|
| 971 |
-
"loss": 3.6434,
|
| 972 |
-
"step": 790
|
| 973 |
-
},
|
| 974 |
-
{
|
| 975 |
-
"epoch": 1.73,
|
| 976 |
-
"learning_rate": 0.00011368417236691289,
|
| 977 |
-
"loss": 3.6893,
|
| 978 |
-
"step": 795
|
| 979 |
-
},
|
| 980 |
-
{
|
| 981 |
-
"epoch": 1.74,
|
| 982 |
-
"learning_rate": 0.00011542314562479964,
|
| 983 |
-
"loss": 3.6739,
|
| 984 |
-
"step": 800
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 1.75,
|
| 988 |
-
"learning_rate": 0.00011710752518939715,
|
| 989 |
-
"loss": 3.5122,
|
| 990 |
-
"step": 805
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 1.76,
|
| 994 |
-
"learning_rate": 0.00011873534714947934,
|
| 995 |
-
"loss": 3.6633,
|
| 996 |
-
"step": 810
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 1.77,
|
| 1000 |
-
"learning_rate": 0.00012030471353745696,
|
| 1001 |
-
"loss": 3.5101,
|
| 1002 |
-
"step": 815
|
| 1003 |
-
},
|
| 1004 |
-
{
|
| 1005 |
-
"epoch": 1.78,
|
| 1006 |
-
"learning_rate": 0.00012181379454232318,
|
| 1007 |
-
"loss": 3.6736,
|
| 1008 |
-
"step": 820
|
| 1009 |
-
},
|
| 1010 |
-
{
|
| 1011 |
-
"epoch": 1.79,
|
| 1012 |
-
"learning_rate": 0.00012326083064313103,
|
| 1013 |
-
"loss": 3.4271,
|
| 1014 |
-
"step": 825
|
| 1015 |
-
},
|
| 1016 |
-
{
|
| 1017 |
-
"epoch": 1.8,
|
| 1018 |
-
"learning_rate": 0.0001246441346605163,
|
| 1019 |
-
"loss": 3.5088,
|
| 1020 |
-
"step": 830
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 1.82,
|
| 1024 |
-
"learning_rate": 0.00012596209372387317,
|
| 1025 |
-
"loss": 3.5845,
|
| 1026 |
-
"step": 835
|
| 1027 |
-
},
|
| 1028 |
-
{
|
| 1029 |
-
"epoch": 1.83,
|
| 1030 |
-
"learning_rate": 0.0001272131711518891,
|
| 1031 |
-
"loss": 3.6172,
|
| 1032 |
-
"step": 840
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 1.84,
|
| 1036 |
-
"learning_rate": 0.00012839590824424672,
|
| 1037 |
-
"loss": 3.5649,
|
| 1038 |
-
"step": 845
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 1.85,
|
| 1042 |
-
"learning_rate": 0.00012950892598240292,
|
| 1043 |
-
"loss": 3.5506,
|
| 1044 |
-
"step": 850
|
| 1045 |
-
},
|
| 1046 |
-
{
|
| 1047 |
-
"epoch": 1.86,
|
| 1048 |
-
"learning_rate": 0.00013055092663746294,
|
| 1049 |
-
"loss": 3.4779,
|
| 1050 |
-
"step": 855
|
| 1051 |
-
},
|
| 1052 |
-
{
|
| 1053 |
-
"epoch": 1.87,
|
| 1054 |
-
"learning_rate": 0.00013152069528327408,
|
| 1055 |
-
"loss": 3.5707,
|
| 1056 |
-
"step": 860
|
| 1057 |
-
},
|
| 1058 |
-
{
|
| 1059 |
-
"epoch": 1.88,
|
| 1060 |
-
"learning_rate": 0.00013241710121297533,
|
| 1061 |
-
"loss": 3.497,
|
| 1062 |
-
"step": 865
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 1.89,
|
| 1066 |
-
"learning_rate": 0.0001332390992573511,
|
| 1067 |
-
"loss": 3.6343,
|
| 1068 |
-
"step": 870
|
| 1069 |
-
},
|
| 1070 |
-
{
|
| 1071 |
-
"epoch": 1.9,
|
| 1072 |
-
"learning_rate": 0.00013398573100345144,
|
| 1073 |
-
"loss": 3.5251,
|
| 1074 |
-
"step": 875
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 1.91,
|
| 1078 |
-
"learning_rate": 0.00013465612591205902,
|
| 1079 |
-
"loss": 3.7421,
|
| 1080 |
-
"step": 880
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 1.92,
|
| 1084 |
-
"learning_rate": 0.00013524950233269879,
|
| 1085 |
-
"loss": 3.5967,
|
| 1086 |
-
"step": 885
|
| 1087 |
-
},
|
| 1088 |
-
{
|
| 1089 |
-
"epoch": 1.93,
|
| 1090 |
-
"learning_rate": 0.00013576516841500732,
|
| 1091 |
-
"loss": 3.5853,
|
| 1092 |
-
"step": 890
|
| 1093 |
-
},
|
| 1094 |
-
{
|
| 1095 |
-
"epoch": 1.95,
|
| 1096 |
-
"learning_rate": 0.0001362025229153997,
|
| 1097 |
-
"loss": 3.7016,
|
| 1098 |
-
"step": 895
|
| 1099 |
-
},
|
| 1100 |
-
{
|
| 1101 |
-
"epoch": 1.96,
|
| 1102 |
-
"learning_rate": 0.00013656105589809228,
|
| 1103 |
-
"loss": 3.7649,
|
| 1104 |
-
"step": 900
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 1.97,
|
| 1108 |
-
"learning_rate": 0.0001368403493296653,
|
| 1109 |
-
"loss": 3.6488,
|
| 1110 |
-
"step": 905
|
| 1111 |
-
},
|
| 1112 |
-
{
|
| 1113 |
-
"epoch": 1.98,
|
| 1114 |
-
"learning_rate": 0.000137040077566471,
|
| 1115 |
-
"loss": 3.5113,
|
| 1116 |
-
"step": 910
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 1.99,
|
| 1120 |
-
"learning_rate": 0.00013716000773431953,
|
| 1121 |
-
"loss": 3.4688,
|
| 1122 |
-
"step": 915
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 2.0,
|
| 1126 |
-
"learning_rate": 0.0001372,
|
| 1127 |
-
"loss": 3.5363,
|
| 1128 |
-
"step": 920
|
| 1129 |
-
},
|
| 1130 |
-
{
|
| 1131 |
-
"epoch": 2.0,
|
| 1132 |
-
"eval_loss": 3.550342321395874,
|
| 1133 |
-
"eval_runtime": 15.5302,
|
| 1134 |
-
"eval_samples_per_second": 38.892,
|
| 1135 |
-
"eval_steps_per_second": 4.894,
|
| 1136 |
-
"step": 920
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 2.04,
|
| 1140 |
-
"learning_rate": 0.0001366053297599336,
|
| 1141 |
-
"loss": 3.3872,
|
| 1142 |
-
"step": 925
|
| 1143 |
-
},
|
| 1144 |
-
{
|
| 1145 |
-
"epoch": 2.05,
|
| 1146 |
-
"learning_rate": 0.00013625198084925744,
|
| 1147 |
-
"loss": 3.3661,
|
| 1148 |
-
"step": 930
|
| 1149 |
-
},
|
| 1150 |
-
{
|
| 1151 |
-
"epoch": 2.06,
|
| 1152 |
-
"learning_rate": 0.00013581729639919083,
|
| 1153 |
-
"loss": 3.3705,
|
| 1154 |
-
"step": 935
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 2.08,
|
| 1158 |
-
"learning_rate": 0.00013530179901518854,
|
| 1159 |
-
"loss": 3.3747,
|
| 1160 |
-
"step": 940
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 2.09,
|
| 1164 |
-
"learning_rate": 0.00013470610846118432,
|
| 1165 |
-
"loss": 3.3343,
|
| 1166 |
-
"step": 945
|
| 1167 |
-
},
|
| 1168 |
-
{
|
| 1169 |
-
"epoch": 2.1,
|
| 1170 |
-
"learning_rate": 0.00013403094091447133,
|
| 1171 |
-
"loss": 3.4331,
|
| 1172 |
-
"step": 950
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 2.11,
|
| 1176 |
-
"learning_rate": 0.00013327710810466788,
|
| 1177 |
-
"loss": 3.3831,
|
| 1178 |
-
"step": 955
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 2.12,
|
| 1182 |
-
"learning_rate": 0.00013244551633780405,
|
| 1183 |
-
"loss": 3.4936,
|
| 1184 |
-
"step": 960
|
| 1185 |
-
},
|
| 1186 |
-
{
|
| 1187 |
-
"epoch": 2.13,
|
| 1188 |
-
"learning_rate": 0.00013153716540670286,
|
| 1189 |
-
"loss": 3.5632,
|
| 1190 |
-
"step": 965
|
| 1191 |
-
},
|
| 1192 |
-
{
|
| 1193 |
-
"epoch": 2.14,
|
| 1194 |
-
"learning_rate": 0.0001305531473889651,
|
| 1195 |
-
"loss": 3.5966,
|
| 1196 |
-
"step": 970
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 2.15,
|
| 1200 |
-
"learning_rate": 0.00012949464533400386,
|
| 1201 |
-
"loss": 3.5279,
|
| 1202 |
-
"step": 975
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 2.16,
|
| 1206 |
-
"learning_rate": 0.00012836293184070708,
|
| 1207 |
-
"loss": 3.3265,
|
| 1208 |
-
"step": 980
|
| 1209 |
-
},
|
| 1210 |
-
{
|
| 1211 |
-
"epoch": 2.17,
|
| 1212 |
-
"learning_rate": 0.00012715936752743734,
|
| 1213 |
-
"loss": 3.3291,
|
| 1214 |
-
"step": 985
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 2.19,
|
| 1218 |
-
"learning_rate": 0.00012588539939620955,
|
| 1219 |
-
"loss": 3.4152,
|
| 1220 |
-
"step": 990
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 2.2,
|
| 1224 |
-
"learning_rate": 0.000124542559093013,
|
| 1225 |
-
"loss": 3.3681,
|
| 1226 |
-
"step": 995
|
| 1227 |
-
},
|
| 1228 |
-
{
|
| 1229 |
-
"epoch": 2.21,
|
| 1230 |
-
"learning_rate": 0.00012313246106636776,
|
| 1231 |
-
"loss": 3.4529,
|
| 1232 |
-
"step": 1000
|
| 1233 |
-
},
|
| 1234 |
-
{
|
| 1235 |
-
"epoch": 2.22,
|
| 1236 |
-
"learning_rate": 0.00012165680062633217,
|
| 1237 |
-
"loss": 3.3251,
|
| 1238 |
-
"step": 1005
|
| 1239 |
-
},
|
| 1240 |
-
{
|
| 1241 |
-
"epoch": 2.23,
|
| 1242 |
-
"learning_rate": 0.00012011735190629213,
|
| 1243 |
-
"loss": 3.5049,
|
| 1244 |
-
"step": 1010
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 2.24,
|
| 1248 |
-
"learning_rate": 0.00011851596572998445,
|
| 1249 |
-
"loss": 3.3393,
|
| 1250 |
-
"step": 1015
|
| 1251 |
-
},
|
| 1252 |
-
{
|
| 1253 |
-
"epoch": 2.25,
|
| 1254 |
-
"learning_rate": 0.00011685456738631861,
|
| 1255 |
-
"loss": 3.2768,
|
| 1256 |
-
"step": 1020
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 2.26,
|
| 1260 |
-
"learning_rate": 0.00011513515431467074,
|
| 1261 |
-
"loss": 3.4727,
|
| 1262 |
-
"step": 1025
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 2.27,
|
| 1266 |
-
"learning_rate": 0.0001133597937034344,
|
| 1267 |
-
"loss": 3.5473,
|
| 1268 |
-
"step": 1030
|
| 1269 |
-
},
|
| 1270 |
-
{
|
| 1271 |
-
"epoch": 2.28,
|
| 1272 |
-
"learning_rate": 0.00011153062000471465,
|
| 1273 |
-
"loss": 3.363,
|
| 1274 |
-
"step": 1035
|
| 1275 |
-
},
|
| 1276 |
-
{
|
| 1277 |
-
"epoch": 2.3,
|
| 1278 |
-
"learning_rate": 0.00010964983236815272,
|
| 1279 |
-
"loss": 3.5201,
|
| 1280 |
-
"step": 1040
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 2.31,
|
| 1284 |
-
"learning_rate": 0.00010771969199696821,
|
| 1285 |
-
"loss": 3.3989,
|
| 1286 |
-
"step": 1045
|
| 1287 |
-
},
|
| 1288 |
-
{
|
| 1289 |
-
"epoch": 2.32,
|
| 1290 |
-
"learning_rate": 0.00010574251942939544,
|
| 1291 |
-
"loss": 3.4949,
|
| 1292 |
-
"step": 1050
|
| 1293 |
-
},
|
| 1294 |
-
{
|
| 1295 |
-
"epoch": 2.33,
|
| 1296 |
-
"learning_rate": 0.00010372069174878399,
|
| 1297 |
-
"loss": 3.5509,
|
| 1298 |
-
"step": 1055
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 2.34,
|
| 1302 |
-
"learning_rate": 0.00010165663972571744,
|
| 1303 |
-
"loss": 3.3905,
|
| 1304 |
-
"step": 1060
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 2.35,
|
| 1308 |
-
"learning_rate": 9.955284489558486e-05,
|
| 1309 |
-
"loss": 3.2992,
|
| 1310 |
-
"step": 1065
|
| 1311 |
-
},
|
| 1312 |
-
{
|
| 1313 |
-
"epoch": 2.36,
|
| 1314 |
-
"learning_rate": 9.741183657512063e-05,
|
| 1315 |
-
"loss": 3.4254,
|
| 1316 |
-
"step": 1070
|
| 1317 |
-
},
|
| 1318 |
-
{
|
| 1319 |
-
"epoch": 2.37,
|
| 1320 |
-
"learning_rate": 9.523618882149793e-05,
|
| 1321 |
-
"loss": 3.5512,
|
| 1322 |
-
"step": 1075
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 2.38,
|
| 1326 |
-
"learning_rate": 9.30285173376325e-05,
|
| 1327 |
-
"loss": 3.5264,
|
| 1328 |
-
"step": 1080
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 2.4,
|
| 1332 |
-
"learning_rate": 9.079147632741743e-05,
|
| 1333 |
-
"loss": 3.5711,
|
| 1334 |
-
"step": 1085
|
| 1335 |
-
},
|
| 1336 |
-
{
|
| 1337 |
-
"epoch": 2.41,
|
| 1338 |
-
"learning_rate": 8.852775530466907e-05,
|
| 1339 |
-
"loss": 3.3514,
|
| 1340 |
-
"step": 1090
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 2.42,
|
| 1344 |
-
"learning_rate": 8.624007585962169e-05,
|
| 1345 |
-
"loss": 3.4614,
|
| 1346 |
-
"step": 1095
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 2.43,
|
| 1350 |
-
"learning_rate": 8.393118838685784e-05,
|
| 1351 |
-
"loss": 3.4247,
|
| 1352 |
-
"step": 1100
|
| 1353 |
-
},
|
| 1354 |
-
{
|
| 1355 |
-
"epoch": 2.44,
|
| 1356 |
-
"learning_rate": 8.160386877860755e-05,
|
| 1357 |
-
"loss": 3.6425,
|
| 1358 |
-
"step": 1105
|
| 1359 |
-
},
|
| 1360 |
-
{
|
| 1361 |
-
"epoch": 2.45,
|
| 1362 |
-
"learning_rate": 7.926091508739327e-05,
|
| 1363 |
-
"loss": 3.4344,
|
| 1364 |
-
"step": 1110
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 2.46,
|
| 1368 |
-
"learning_rate": 7.690514416203246e-05,
|
| 1369 |
-
"loss": 3.2589,
|
| 1370 |
-
"step": 1115
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 2.47,
|
| 1374 |
-
"learning_rate": 7.453938826104105e-05,
|
| 1375 |
-
"loss": 3.4692,
|
| 1376 |
-
"step": 1120
|
| 1377 |
-
},
|
| 1378 |
-
{
|
| 1379 |
-
"epoch": 2.48,
|
| 1380 |
-
"learning_rate": 7.216649164751185e-05,
|
| 1381 |
-
"loss": 3.1535,
|
| 1382 |
-
"step": 1125
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 2.49,
|
| 1386 |
-
"learning_rate": 6.978930716955893e-05,
|
| 1387 |
-
"loss": 3.2569,
|
| 1388 |
-
"step": 1130
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 2.51,
|
| 1392 |
-
"learning_rate": 6.741069283044111e-05,
|
| 1393 |
-
"loss": 3.2807,
|
| 1394 |
-
"step": 1135
|
| 1395 |
-
},
|
| 1396 |
-
{
|
| 1397 |
-
"epoch": 2.52,
|
| 1398 |
-
"learning_rate": 6.50335083524882e-05,
|
| 1399 |
-
"loss": 3.618,
|
| 1400 |
-
"step": 1140
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 2.53,
|
| 1404 |
-
"learning_rate": 6.2660611738959e-05,
|
| 1405 |
-
"loss": 3.2624,
|
| 1406 |
-
"step": 1145
|
| 1407 |
-
},
|
| 1408 |
-
{
|
| 1409 |
-
"epoch": 2.54,
|
| 1410 |
-
"learning_rate": 6.029485583796758e-05,
|
| 1411 |
-
"loss": 3.5394,
|
| 1412 |
-
"step": 1150
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 2.55,
|
| 1416 |
-
"learning_rate": 5.793908491260678e-05,
|
| 1417 |
-
"loss": 3.3858,
|
| 1418 |
-
"step": 1155
|
| 1419 |
-
},
|
| 1420 |
-
{
|
| 1421 |
-
"epoch": 2.56,
|
| 1422 |
-
"learning_rate": 5.5596131221392484e-05,
|
| 1423 |
-
"loss": 3.553,
|
| 1424 |
-
"step": 1160
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 2.57,
|
| 1428 |
-
"learning_rate": 5.326881161314222e-05,
|
| 1429 |
-
"loss": 3.533,
|
| 1430 |
-
"step": 1165
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 2.58,
|
| 1434 |
-
"learning_rate": 5.09599241403784e-05,
|
| 1435 |
-
"loss": 3.4779,
|
| 1436 |
-
"step": 1170
|
| 1437 |
-
},
|
| 1438 |
-
{
|
| 1439 |
-
"epoch": 2.59,
|
| 1440 |
-
"learning_rate": 4.8672244695331025e-05,
|
| 1441 |
-
"loss": 3.3822,
|
| 1442 |
-
"step": 1175
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 2.6,
|
| 1446 |
-
"learning_rate": 4.640852367258268e-05,
|
| 1447 |
-
"loss": 3.5514,
|
| 1448 |
-
"step": 1180
|
| 1449 |
-
},
|
| 1450 |
-
{
|
| 1451 |
-
"epoch": 2.62,
|
| 1452 |
-
"learning_rate": 4.417148266236749e-05,
|
| 1453 |
-
"loss": 3.3215,
|
| 1454 |
-
"step": 1185
|
| 1455 |
-
},
|
| 1456 |
-
{
|
| 1457 |
-
"epoch": 2.63,
|
| 1458 |
-
"learning_rate": 4.1963811178502046e-05,
|
| 1459 |
-
"loss": 3.384,
|
| 1460 |
-
"step": 1190
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 2.64,
|
| 1464 |
-
"learning_rate": 3.978816342487941e-05,
|
| 1465 |
-
"loss": 3.2548,
|
| 1466 |
-
"step": 1195
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 2.65,
|
| 1470 |
-
"learning_rate": 3.7647155104415125e-05,
|
| 1471 |
-
"loss": 3.5365,
|
| 1472 |
-
"step": 1200
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 2.66,
|
| 1476 |
-
"learning_rate": 3.5543360274282583e-05,
|
| 1477 |
-
"loss": 3.2376,
|
| 1478 |
-
"step": 1205
|
| 1479 |
-
},
|
| 1480 |
-
{
|
| 1481 |
-
"epoch": 2.67,
|
| 1482 |
-
"learning_rate": 3.3479308251216045e-05,
|
| 1483 |
-
"loss": 3.3611,
|
| 1484 |
-
"step": 1210
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 2.68,
|
| 1488 |
-
"learning_rate": 3.14574805706046e-05,
|
| 1489 |
-
"loss": 3.3802,
|
| 1490 |
-
"step": 1215
|
| 1491 |
-
},
|
| 1492 |
-
{
|
| 1493 |
-
"epoch": 2.69,
|
| 1494 |
-
"learning_rate": 2.9480308003031837e-05,
|
| 1495 |
-
"loss": 3.4986,
|
| 1496 |
-
"step": 1220
|
| 1497 |
-
},
|
| 1498 |
-
{
|
| 1499 |
-
"epoch": 2.7,
|
| 1500 |
-
"learning_rate": 2.7550167631847363e-05,
|
| 1501 |
-
"loss": 3.2628,
|
| 1502 |
-
"step": 1225
|
| 1503 |
-
},
|
| 1504 |
-
{
|
| 1505 |
-
"epoch": 2.72,
|
| 1506 |
-
"learning_rate": 2.5669379995285432e-05,
|
| 1507 |
-
"loss": 3.3669,
|
| 1508 |
-
"step": 1230
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 2.73,
|
| 1512 |
-
"learning_rate": 2.3840206296565597e-05,
|
| 1513 |
-
"loss": 3.2283,
|
| 1514 |
-
"step": 1235
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 2.74,
|
| 1518 |
-
"learning_rate": 2.2064845685329227e-05,
|
| 1519 |
-
"loss": 3.3849,
|
| 1520 |
-
"step": 1240
|
| 1521 |
-
},
|
| 1522 |
-
{
|
| 1523 |
-
"epoch": 2.75,
|
| 1524 |
-
"learning_rate": 2.0345432613681382e-05,
|
| 1525 |
-
"loss": 3.3535,
|
| 1526 |
-
"step": 1245
|
| 1527 |
-
},
|
| 1528 |
-
{
|
| 1529 |
-
"epoch": 2.76,
|
| 1530 |
-
"learning_rate": 1.8684034270015573e-05,
|
| 1531 |
-
"loss": 3.4427,
|
| 1532 |
-
"step": 1250
|
| 1533 |
-
},
|
| 1534 |
-
{
|
| 1535 |
-
"epoch": 2.77,
|
| 1536 |
-
"learning_rate": 1.70826480937079e-05,
|
| 1537 |
-
"loss": 3.4215,
|
| 1538 |
-
"step": 1255
|
| 1539 |
-
},
|
| 1540 |
-
{
|
| 1541 |
-
"epoch": 2.78,
|
| 1542 |
-
"learning_rate": 1.554319937366785e-05,
|
| 1543 |
-
"loss": 3.218,
|
| 1544 |
-
"step": 1260
|
| 1545 |
-
},
|
| 1546 |
-
{
|
| 1547 |
-
"epoch": 2.79,
|
| 1548 |
-
"learning_rate": 1.4067538933632289e-05,
|
| 1549 |
-
"loss": 3.3675,
|
| 1550 |
-
"step": 1265
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 2.8,
|
| 1554 |
-
"learning_rate": 1.2657440906987054e-05,
|
| 1555 |
-
"loss": 3.3894,
|
| 1556 |
-
"step": 1270
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 2.81,
|
| 1560 |
-
"learning_rate": 1.1314600603790494e-05,
|
| 1561 |
-
"loss": 3.3839,
|
| 1562 |
-
"step": 1275
|
| 1563 |
-
},
|
| 1564 |
-
{
|
| 1565 |
-
"epoch": 2.83,
|
| 1566 |
-
"learning_rate": 1.0040632472562733e-05,
|
| 1567 |
-
"loss": 3.6158,
|
| 1568 |
-
"step": 1280
|
| 1569 |
-
},
|
| 1570 |
-
{
|
| 1571 |
-
"epoch": 2.84,
|
| 1572 |
-
"learning_rate": 8.837068159292907e-06,
|
| 1573 |
-
"loss": 3.3993,
|
| 1574 |
-
"step": 1285
|
| 1575 |
-
},
|
| 1576 |
-
{
|
| 1577 |
-
"epoch": 2.85,
|
| 1578 |
-
"learning_rate": 7.705354665996124e-06,
|
| 1579 |
-
"loss": 3.5461,
|
| 1580 |
-
"step": 1290
|
| 1581 |
-
},
|
| 1582 |
-
{
|
| 1583 |
-
"epoch": 2.86,
|
| 1584 |
-
"learning_rate": 6.646852611034896e-06,
|
| 1585 |
-
"loss": 3.429,
|
| 1586 |
-
"step": 1295
|
| 1587 |
-
},
|
| 1588 |
-
{
|
| 1589 |
-
"epoch": 2.87,
|
| 1590 |
-
"learning_rate": 5.662834593297144e-06,
|
| 1591 |
-
"loss": 3.2661,
|
| 1592 |
-
"step": 1300
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 2.88,
|
| 1596 |
-
"learning_rate": 4.754483662195961e-06,
|
| 1597 |
-
"loss": 3.4591,
|
| 1598 |
-
"step": 1305
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 2.89,
|
| 1602 |
-
"learning_rate": 3.922891895332133e-06,
|
| 1603 |
-
"loss": 3.3605,
|
| 1604 |
-
"step": 1310
|
| 1605 |
-
},
|
| 1606 |
-
{
|
| 1607 |
-
"epoch": 2.9,
|
| 1608 |
-
"learning_rate": 3.1690590855286824e-06,
|
| 1609 |
-
"loss": 3.377,
|
| 1610 |
-
"step": 1315
|
| 1611 |
-
},
|
| 1612 |
-
{
|
| 1613 |
-
"epoch": 2.91,
|
| 1614 |
-
"learning_rate": 2.493891538815714e-06,
|
| 1615 |
-
"loss": 3.3229,
|
| 1616 |
-
"step": 1320
|
| 1617 |
-
},
|
| 1618 |
-
{
|
| 1619 |
-
"epoch": 2.92,
|
| 1620 |
-
"learning_rate": 1.898200984811481e-06,
|
| 1621 |
-
"loss": 3.3711,
|
| 1622 |
-
"step": 1325
|
| 1623 |
-
},
|
| 1624 |
-
{
|
| 1625 |
-
"epoch": 2.94,
|
| 1626 |
-
"learning_rate": 1.3827036008091848e-06,
|
| 1627 |
-
"loss": 3.3335,
|
| 1628 |
-
"step": 1330
|
| 1629 |
-
},
|
| 1630 |
-
{
|
| 1631 |
-
"epoch": 2.95,
|
| 1632 |
-
"learning_rate": 9.48019150742587e-07,
|
| 1633 |
-
"loss": 3.2538,
|
| 1634 |
-
"step": 1335
|
| 1635 |
-
},
|
| 1636 |
-
{
|
| 1637 |
-
"epoch": 2.96,
|
| 1638 |
-
"learning_rate": 5.946702400663892e-07,
|
| 1639 |
-
"loss": 3.2806,
|
| 1640 |
-
"step": 1340
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 2.97,
|
| 1644 |
-
"learning_rate": 3.2308168744640525e-07,
|
| 1645 |
-
"loss": 3.5042,
|
| 1646 |
-
"step": 1345
|
| 1647 |
-
},
|
| 1648 |
-
{
|
| 1649 |
-
"epoch": 2.98,
|
| 1650 |
-
"learning_rate": 1.335800140153434e-07,
|
| 1651 |
-
"loss": 3.4461,
|
| 1652 |
-
"step": 1350
|
| 1653 |
-
},
|
| 1654 |
-
{
|
| 1655 |
-
"epoch": 2.99,
|
| 1656 |
-
"learning_rate": 2.6393050808329943e-08,
|
| 1657 |
-
"loss": 3.3588,
|
| 1658 |
-
"step": 1355
|
| 1659 |
-
},
|
| 1660 |
-
{
|
| 1661 |
-
"epoch": 3.0,
|
| 1662 |
-
"eval_loss": 3.3743550777435303,
|
| 1663 |
-
"eval_runtime": 31.7563,
|
| 1664 |
-
"eval_samples_per_second": 20.815,
|
| 1665 |
-
"eval_steps_per_second": 2.614,
|
| 1666 |
-
"step": 1359
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 2.97,
|
| 1670 |
-
"learning_rate": 3.160713999402438e-07,
|
| 1671 |
-
"loss": 2.9957,
|
| 1672 |
-
"step": 1360
|
| 1673 |
-
},
|
| 1674 |
-
{
|
| 1675 |
-
"epoch": 2.98,
|
| 1676 |
-
"learning_rate": 1.3068026139647436e-07,
|
| 1677 |
-
"loss": 3.3681,
|
| 1678 |
-
"step": 1365
|
| 1679 |
-
},
|
| 1680 |
-
{
|
| 1681 |
-
"epoch": 2.99,
|
| 1682 |
-
"learning_rate": 2.5819964843398836e-08,
|
| 1683 |
-
"loss": 3.333,
|
| 1684 |
-
"step": 1370
|
| 1685 |
-
},
|
| 1686 |
-
{
|
| 1687 |
-
"epoch": 3.0,
|
| 1688 |
-
"eval_loss": 3.107706069946289,
|
| 1689 |
-
"eval_runtime": 7.9683,
|
| 1690 |
-
"eval_samples_per_second": 77.055,
|
| 1691 |
-
"eval_steps_per_second": 9.663,
|
| 1692 |
-
"step": 1374
|
| 1693 |
}
|
| 1694 |
],
|
| 1695 |
"max_steps": 1832,
|
| 1696 |
"num_train_epochs": 4,
|
| 1697 |
-
"total_flos":
|
| 1698 |
"trial_name": null,
|
| 1699 |
"trial_params": null
|
| 1700 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 3.2645678520202637,
|
| 3 |
+
"best_model_checkpoint": "output/eminem/checkpoint-458",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"global_step": 458,
|
| 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": 0.00013715965772858815,
|
| 13 |
+
"loss": 3.1417,
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 0.000137038678363299,
|
| 19 |
+
"loss": 2.9863,
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.03,
|
| 24 |
+
"learning_rate": 0.00013683720419516428,
|
| 25 |
+
"loss": 3.1237,
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.04,
|
| 30 |
+
"learning_rate": 0.00013655547218994425,
|
| 31 |
+
"loss": 3.0636,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.05,
|
| 36 |
+
"learning_rate": 0.00013619381370941783,
|
| 37 |
+
"loss": 3.1631,
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.0001357526541216486,
|
| 43 |
+
"loss": 3.017,
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 0.000135232512300684,
|
| 49 |
+
"loss": 3.1919,
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.09,
|
| 54 |
+
"learning_rate": 0.00013463400001627657,
|
| 55 |
+
"loss": 3.404,
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.1,
|
| 60 |
+
"learning_rate": 0.0001339578212143452,
|
| 61 |
+
"loss": 3.2587,
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.11,
|
| 66 |
+
"learning_rate": 0.00013320477118902276,
|
| 67 |
+
"loss": 3.2812,
|
| 68 |
"step": 50
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.12,
|
| 72 |
+
"learning_rate": 0.00013237573564726276,
|
| 73 |
+
"loss": 2.9708,
|
| 74 |
"step": 55
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.13,
|
| 78 |
+
"learning_rate": 0.000131471689667107,
|
| 79 |
+
"loss": 3.2539,
|
| 80 |
"step": 60
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.14,
|
| 84 |
+
"learning_rate": 0.00013049369655083827,
|
| 85 |
+
"loss": 3.3207,
|
| 86 |
"step": 65
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.15,
|
| 90 |
+
"learning_rate": 0.00012944290657436735,
|
| 91 |
+
"loss": 3.1984,
|
| 92 |
"step": 70
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.16,
|
| 96 |
+
"learning_rate": 0.00012832055563432473,
|
| 97 |
+
"loss": 3.2094,
|
| 98 |
"step": 75
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 0.17,
|
| 102 |
+
"learning_rate": 0.0001271279637944492,
|
| 103 |
+
"loss": 3.2116,
|
| 104 |
"step": 80
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.19,
|
| 108 |
+
"learning_rate": 0.00012586653373298248,
|
| 109 |
+
"loss": 3.2721,
|
| 110 |
"step": 85
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.2,
|
| 114 |
+
"learning_rate": 0.00012453774909289557,
|
| 115 |
+
"loss": 3.2313,
|
| 116 |
"step": 90
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.21,
|
| 120 |
+
"learning_rate": 0.00012314317273688812,
|
| 121 |
+
"loss": 3.3427,
|
| 122 |
"step": 95
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.22,
|
| 126 |
+
"learning_rate": 0.00012168444490921243,
|
| 127 |
+
"loss": 3.3434,
|
| 128 |
"step": 100
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.23,
|
| 132 |
+
"learning_rate": 0.00012016328130648513,
|
| 133 |
+
"loss": 3.0684,
|
| 134 |
"step": 105
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.24,
|
| 138 |
+
"learning_rate": 0.00011858147105975337,
|
| 139 |
+
"loss": 2.9787,
|
| 140 |
"step": 110
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.25,
|
| 144 |
+
"learning_rate": 0.0001169408746301919,
|
| 145 |
+
"loss": 3.0232,
|
| 146 |
"step": 115
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.26,
|
| 150 |
+
"learning_rate": 0.00011524342162090394,
|
| 151 |
+
"loss": 3.097,
|
| 152 |
"step": 120
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.27,
|
| 156 |
+
"learning_rate": 0.00011349110850740006,
|
| 157 |
+
"loss": 2.9875,
|
| 158 |
"step": 125
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.28,
|
| 162 |
+
"learning_rate": 0.00011168599628942434,
|
| 163 |
+
"loss": 3.069,
|
| 164 |
"step": 130
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.29,
|
| 168 |
+
"learning_rate": 0.00010983020806688995,
|
| 169 |
+
"loss": 3.2752,
|
| 170 |
"step": 135
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.31,
|
| 174 |
+
"learning_rate": 0.00010792592654277545,
|
| 175 |
+
"loss": 3.0414,
|
| 176 |
"step": 140
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.32,
|
| 180 |
+
"learning_rate": 0.00010597539145591715,
|
| 181 |
+
"loss": 3.2296,
|
| 182 |
"step": 145
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.33,
|
| 186 |
+
"learning_rate": 0.0001039808969467193,
|
| 187 |
+
"loss": 3.2247,
|
| 188 |
"step": 150
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.34,
|
| 192 |
+
"learning_rate": 0.00010194478885887969,
|
| 193 |
+
"loss": 3.0732,
|
| 194 |
"step": 155
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.35,
|
| 198 |
+
"learning_rate": 9.986946198030201e-05,
|
| 199 |
+
"loss": 3.0828,
|
| 200 |
"step": 160
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.36,
|
| 204 |
+
"learning_rate": 9.775735722644407e-05,
|
| 205 |
+
"loss": 3.0514,
|
| 206 |
"step": 165
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.37,
|
| 210 |
+
"learning_rate": 9.561095876941162e-05,
|
| 211 |
+
"loss": 2.9164,
|
| 212 |
"step": 170
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.38,
|
| 216 |
+
"learning_rate": 9.343279111617608e-05,
|
| 217 |
+
"loss": 3.0357,
|
| 218 |
"step": 175
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.39,
|
| 222 |
+
"learning_rate": 9.122541613935095e-05,
|
| 223 |
+
"loss": 3.2488,
|
| 224 |
"step": 180
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.4,
|
| 228 |
+
"learning_rate": 8.89914300640212e-05,
|
| 229 |
+
"loss": 3.017,
|
| 230 |
"step": 185
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 0.41,
|
| 234 |
+
"learning_rate": 8.673346041416928e-05,
|
| 235 |
+
"loss": 3.1126,
|
| 236 |
"step": 190
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.43,
|
| 240 |
+
"learning_rate": 8.445416292228631e-05,
|
| 241 |
+
"loss": 3.2453,
|
| 242 |
"step": 195
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.44,
|
| 246 |
+
"learning_rate": 8.215621840580805e-05,
|
| 247 |
+
"loss": 3.0444,
|
| 248 |
"step": 200
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.45,
|
| 252 |
+
"learning_rate": 7.984232961404542e-05,
|
| 253 |
+
"loss": 3.1687,
|
| 254 |
"step": 205
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.46,
|
| 258 |
+
"learning_rate": 7.75152180493212e-05,
|
| 259 |
+
"loss": 3.225,
|
| 260 |
"step": 210
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.47,
|
| 264 |
+
"learning_rate": 7.517762076604763e-05,
|
| 265 |
+
"loss": 3.0953,
|
| 266 |
"step": 215
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 0.48,
|
| 270 |
+
"learning_rate": 7.283228715151494e-05,
|
| 271 |
+
"loss": 3.25,
|
| 272 |
"step": 220
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 0.49,
|
| 276 |
+
"learning_rate": 7.048197569217281e-05,
|
| 277 |
+
"loss": 3.1282,
|
| 278 |
"step": 225
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 0.5,
|
| 282 |
+
"learning_rate": 6.812945072921086e-05,
|
| 283 |
+
"loss": 2.9809,
|
| 284 |
"step": 230
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 0.51,
|
| 288 |
+
"learning_rate": 6.577747920725168e-05,
|
| 289 |
+
"loss": 3.0873,
|
| 290 |
"step": 235
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.52,
|
| 294 |
+
"learning_rate": 6.342882741998283e-05,
|
| 295 |
+
"loss": 3.0845,
|
| 296 |
"step": 240
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.53,
|
| 300 |
+
"learning_rate": 6.108625775655505e-05,
|
| 301 |
+
"loss": 2.9657,
|
| 302 |
"step": 245
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.55,
|
| 306 |
+
"learning_rate": 5.875252545257054e-05,
|
| 307 |
+
"loss": 3.0202,
|
| 308 |
"step": 250
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.56,
|
| 312 |
+
"learning_rate": 5.6430375349487304e-05,
|
| 313 |
+
"loss": 3.1417,
|
| 314 |
"step": 255
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.57,
|
| 318 |
+
"learning_rate": 5.412253866624731e-05,
|
| 319 |
+
"loss": 3.0013,
|
| 320 |
"step": 260
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.58,
|
| 324 |
+
"learning_rate": 5.183172978692829e-05,
|
| 325 |
+
"loss": 2.8449,
|
| 326 |
"step": 265
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 0.59,
|
| 330 |
+
"learning_rate": 4.956064306819361e-05,
|
| 331 |
+
"loss": 3.2741,
|
| 332 |
"step": 270
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.6,
|
| 336 |
+
"learning_rate": 4.731194967029994e-05,
|
| 337 |
+
"loss": 3.024,
|
| 338 |
"step": 275
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.61,
|
| 342 |
+
"learning_rate": 4.508829441538685e-05,
|
| 343 |
+
"loss": 3.0632,
|
| 344 |
"step": 280
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"epoch": 0.62,
|
| 348 |
+
"learning_rate": 4.289229267674345e-05,
|
| 349 |
+
"loss": 2.8501,
|
| 350 |
"step": 285
|
| 351 |
},
|
| 352 |
{
|
| 353 |
+
"epoch": 0.63,
|
| 354 |
+
"learning_rate": 4.072652730271264e-05,
|
| 355 |
+
"loss": 3.2079,
|
| 356 |
"step": 290
|
| 357 |
},
|
| 358 |
{
|
| 359 |
+
"epoch": 0.64,
|
| 360 |
+
"learning_rate": 3.859354557884909e-05,
|
| 361 |
+
"loss": 3.1717,
|
| 362 |
"step": 295
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.66,
|
| 366 |
+
"learning_rate": 3.649585623190606e-05,
|
| 367 |
+
"loss": 3.1905,
|
| 368 |
"step": 300
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.67,
|
| 372 |
+
"learning_rate": 3.443592647917112e-05,
|
| 373 |
+
"loss": 3.0304,
|
| 374 |
"step": 305
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.68,
|
| 378 |
+
"learning_rate": 3.2416179126625834e-05,
|
| 379 |
+
"loss": 2.9144,
|
| 380 |
"step": 310
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.69,
|
| 384 |
+
"learning_rate": 3.0438989719340004e-05,
|
| 385 |
+
"loss": 3.2034,
|
| 386 |
"step": 315
|
| 387 |
},
|
| 388 |
{
|
| 389 |
+
"epoch": 0.7,
|
| 390 |
+
"learning_rate": 2.850668374745048e-05,
|
| 391 |
+
"loss": 2.9562,
|
| 392 |
"step": 320
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.71,
|
| 396 |
+
"learning_rate": 2.662153391101465e-05,
|
| 397 |
+
"loss": 2.9282,
|
| 398 |
"step": 325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 0.72,
|
| 402 |
+
"learning_rate": 2.4785757446952362e-05,
|
| 403 |
+
"loss": 2.9689,
|
| 404 |
"step": 330
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 0.73,
|
| 408 |
+
"learning_rate": 2.3001513521221957e-05,
|
| 409 |
+
"loss": 3.1242,
|
| 410 |
"step": 335
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 0.74,
|
| 414 |
+
"learning_rate": 2.1270900689296127e-05,
|
| 415 |
+
"loss": 3.174,
|
| 416 |
"step": 340
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.75,
|
| 420 |
+
"learning_rate": 1.959595442792614e-05,
|
| 421 |
+
"loss": 3.0034,
|
| 422 |
"step": 345
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.76,
|
| 426 |
+
"learning_rate": 1.797864474109709e-05,
|
| 427 |
+
"loss": 3.1004,
|
| 428 |
"step": 350
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.78,
|
| 432 |
+
"learning_rate": 1.642087384298789e-05,
|
| 433 |
+
"loss": 2.9905,
|
| 434 |
"step": 355
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.79,
|
| 438 |
+
"learning_rate": 1.4924473920664698e-05,
|
| 439 |
+
"loss": 2.984,
|
| 440 |
"step": 360
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 0.8,
|
| 444 |
+
"learning_rate": 1.3491204979136316e-05,
|
| 445 |
+
"loss": 3.1477,
|
| 446 |
"step": 365
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 0.81,
|
| 450 |
+
"learning_rate": 1.2122752771308157e-05,
|
| 451 |
+
"loss": 3.3952,
|
| 452 |
"step": 370
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.82,
|
| 456 |
+
"learning_rate": 1.0820726815267017e-05,
|
| 457 |
+
"loss": 3.2005,
|
| 458 |
"step": 375
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.83,
|
| 462 |
+
"learning_rate": 9.586658501231604e-06,
|
| 463 |
+
"loss": 3.1182,
|
| 464 |
"step": 380
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.84,
|
| 468 |
+
"learning_rate": 8.421999290393463e-06,
|
| 469 |
+
"loss": 3.1384,
|
| 470 |
"step": 385
|
| 471 |
},
|
| 472 |
{
|
| 473 |
+
"epoch": 0.85,
|
| 474 |
+
"learning_rate": 7.328119007766703e-06,
|
| 475 |
+
"loss": 3.2064,
|
| 476 |
"step": 390
|
| 477 |
},
|
| 478 |
{
|
| 479 |
+
"epoch": 0.86,
|
| 480 |
+
"learning_rate": 6.306304231055513e-06,
|
| 481 |
+
"loss": 3.1691,
|
| 482 |
"step": 395
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 0.87,
|
| 486 |
+
"learning_rate": 5.357756777433396e-06,
|
| 487 |
+
"loss": 2.9847,
|
| 488 |
"step": 400
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 0.88,
|
| 492 |
+
"learning_rate": 4.4835922900148836e-06,
|
| 493 |
+
"loss": 2.919,
|
| 494 |
"step": 405
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 0.9,
|
| 498 |
+
"learning_rate": 3.6848389256806193e-06,
|
| 499 |
+
"loss": 3.0809,
|
| 500 |
"step": 410
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.91,
|
| 504 |
+
"learning_rate": 2.9624361458012783e-06,
|
| 505 |
+
"loss": 3.1271,
|
| 506 |
"step": 415
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.92,
|
| 510 |
+
"learning_rate": 2.317233611281141e-06,
|
| 511 |
+
"loss": 3.2683,
|
| 512 |
"step": 420
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.93,
|
| 516 |
+
"learning_rate": 1.7499901832211207e-06,
|
| 517 |
+
"loss": 3.0606,
|
| 518 |
"step": 425
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.94,
|
| 522 |
+
"learning_rate": 1.2613730303772807e-06,
|
| 523 |
+
"loss": 3.1985,
|
| 524 |
"step": 430
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.95,
|
| 528 |
+
"learning_rate": 8.519568444637701e-07,
|
| 529 |
+
"loss": 3.0785,
|
| 530 |
"step": 435
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.96,
|
| 534 |
+
"learning_rate": 5.222231642238413e-07,
|
| 535 |
+
"loss": 3.2775,
|
| 536 |
"step": 440
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 0.97,
|
| 540 |
+
"learning_rate": 2.7255980906336113e-07,
|
| 541 |
+
"loss": 2.8361,
|
| 542 |
"step": 445
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.98,
|
| 546 |
+
"learning_rate": 1.0326042291338553e-07,
|
| 547 |
+
"loss": 2.9579,
|
| 548 |
"step": 450
|
| 549 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 550 |
{
|
| 551 |
"epoch": 0.99,
|
| 552 |
+
"learning_rate": 1.4524128858071773e-08,
|
| 553 |
+
"loss": 3.1667,
|
| 554 |
"step": 455
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 1.0,
|
| 558 |
+
"eval_loss": 3.2645678520202637,
|
| 559 |
+
"eval_runtime": 27.8254,
|
| 560 |
+
"eval_samples_per_second": 22.102,
|
| 561 |
+
"eval_steps_per_second": 2.767,
|
| 562 |
+
"step": 458
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 563 |
}
|
| 564 |
],
|
| 565 |
"max_steps": 1832,
|
| 566 |
"num_train_epochs": 4,
|
| 567 |
+
"total_flos": 478425710592000.0,
|
| 568 |
"trial_name": null,
|
| 569 |
"trial_params": null
|
| 570 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2671
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:932ce4ce440f60f3536cbafb5b7a0426c27973f99d940115f605b96d124d42e2
|
| 3 |
size 2671
|