Model save
Browse files- README.md +1 -1
- all_results.json +13 -13
- generated_predictions (copy).txt +0 -0
- generated_predictions.txt +0 -0
- mbart_generated_predictions (copy).txt +0 -0
- model.safetensors +1 -1
- predict_results.json +7 -7
- train_results.json +6 -6
- trainer_state.json +236 -908
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -37,7 +37,7 @@ The following hyperparameters were used during training:
|
|
| 37 |
- seed: 42
|
| 38 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 39 |
- lr_scheduler_type: linear
|
| 40 |
-
- num_epochs:
|
| 41 |
|
| 42 |
### Training results
|
| 43 |
|
|
|
|
| 37 |
- seed: 42
|
| 38 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 39 |
- lr_scheduler_type: linear
|
| 40 |
+
- num_epochs: 7.0
|
| 41 |
|
| 42 |
### Training results
|
| 43 |
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"predict_bleu":
|
| 4 |
-
"predict_gen_len":
|
| 5 |
-
"predict_loss": 1.
|
| 6 |
-
"predict_runtime":
|
| 7 |
-
"predict_samples":
|
| 8 |
-
"predict_samples_per_second":
|
| 9 |
-
"predict_steps_per_second": 9.
|
| 10 |
-
"train_loss":
|
| 11 |
-
"train_runtime":
|
| 12 |
-
"train_samples":
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second":
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"predict_bleu": 40.9816,
|
| 4 |
+
"predict_gen_len": 13.035,
|
| 5 |
+
"predict_loss": 1.048066258430481,
|
| 6 |
+
"predict_runtime": 51.9598,
|
| 7 |
+
"predict_samples": 2000,
|
| 8 |
+
"predict_samples_per_second": 38.491,
|
| 9 |
+
"predict_steps_per_second": 9.623,
|
| 10 |
+
"train_loss": 1.5174339204652274,
|
| 11 |
+
"train_runtime": 9956.4506,
|
| 12 |
+
"train_samples": 50418,
|
| 13 |
+
"train_samples_per_second": 15.192,
|
| 14 |
+
"train_steps_per_second": 3.798
|
| 15 |
}
|
generated_predictions (copy).txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
generated_predictions.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
mbart_generated_predictions (copy).txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2444578688
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc97ea71b777a76b3859b80fd38c68693292c37553bdcb67d37368d88334b6d0
|
| 3 |
size 2444578688
|
predict_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"predict_bleu":
|
| 3 |
-
"predict_gen_len":
|
| 4 |
-
"predict_loss": 1.
|
| 5 |
-
"predict_runtime":
|
| 6 |
-
"predict_samples":
|
| 7 |
-
"predict_samples_per_second":
|
| 8 |
-
"predict_steps_per_second": 9.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"predict_bleu": 40.9816,
|
| 3 |
+
"predict_gen_len": 13.035,
|
| 4 |
+
"predict_loss": 1.048066258430481,
|
| 5 |
+
"predict_runtime": 51.9598,
|
| 6 |
+
"predict_samples": 2000,
|
| 7 |
+
"predict_samples_per_second": 38.491,
|
| 8 |
+
"predict_steps_per_second": 9.623
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 1.5174339204652274,
|
| 4 |
+
"train_runtime": 9956.4506,
|
| 5 |
+
"train_samples": 50418,
|
| 6 |
+
"train_samples_per_second": 15.192,
|
| 7 |
+
"train_steps_per_second": 3.798
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1151 +1,479 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 3.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss":
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 2.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 2.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 2.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 2.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 2.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 2.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 2.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 2.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 2.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 2.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 2.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 2.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 2.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss":
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss":
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss":
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss":
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate":
|
| 146 |
-
"loss":
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate":
|
| 152 |
-
"loss":
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 3.
|
| 158 |
-
"loss":
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"learning_rate":
|
| 164 |
-
"loss":
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 1.
|
| 169 |
-
"learning_rate": 3.
|
| 170 |
-
"loss":
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 1.
|
| 175 |
-
"learning_rate": 3.
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 1.
|
| 181 |
-
"learning_rate": 3.
|
| 182 |
-
"loss": 1.
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 1.
|
| 187 |
-
"learning_rate": 3.
|
| 188 |
-
"loss": 1.
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 1.
|
| 193 |
-
"learning_rate":
|
| 194 |
-
"loss": 1.
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 1.
|
| 199 |
-
"learning_rate":
|
| 200 |
-
"loss": 1.
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 1.
|
| 205 |
-
"learning_rate":
|
| 206 |
-
"loss": 1.
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 1.
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 1.
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 1.
|
| 223 |
-
"learning_rate":
|
| 224 |
-
"loss": 1.
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 1.
|
| 229 |
-
"learning_rate":
|
| 230 |
-
"loss": 1.
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch": 1.
|
| 235 |
-
"learning_rate":
|
| 236 |
-
"loss": 1.
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch": 1.
|
| 241 |
-
"learning_rate":
|
| 242 |
-
"loss": 1.
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch": 1.
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss": 1.
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 1.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 1.
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 1.
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 1.
|
| 265 |
-
"learning_rate":
|
| 266 |
-
"loss": 1.
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch": 1.
|
| 271 |
-
"learning_rate":
|
| 272 |
-
"loss": 1.
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch": 1.
|
| 277 |
-
"learning_rate":
|
| 278 |
-
"loss": 1.
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch": 1.
|
| 283 |
-
"learning_rate":
|
| 284 |
-
"loss": 1.
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch": 1.
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss": 1.
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 1.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 1.
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 1.
|
| 307 |
-
"learning_rate":
|
| 308 |
-
"loss": 1.
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"learning_rate":
|
| 314 |
-
"loss": 1.
|
| 315 |
"step": 25500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch":
|
| 319 |
-
"learning_rate":
|
| 320 |
-
"loss":
|
| 321 |
"step": 26000
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch": 1
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss":
|
| 327 |
"step": 26500
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 2.
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss":
|
| 333 |
"step": 27000
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 2.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss":
|
| 339 |
"step": 27500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 2.
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss":
|
| 345 |
"step": 28000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 2.
|
| 349 |
-
"learning_rate":
|
| 350 |
-
"loss":
|
| 351 |
"step": 28500
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch": 2.
|
| 355 |
-
"learning_rate":
|
| 356 |
-
"loss":
|
| 357 |
"step": 29000
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch": 2.
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss":
|
| 363 |
"step": 29500
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch": 2.
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss":
|
| 369 |
"step": 30000
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 2.
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss":
|
| 375 |
"step": 30500
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 2.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss":
|
| 381 |
"step": 31000
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 2.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss":
|
| 387 |
"step": 31500
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 2.
|
| 391 |
-
"learning_rate":
|
| 392 |
-
"loss":
|
| 393 |
"step": 32000
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 2.
|
| 397 |
-
"learning_rate":
|
| 398 |
-
"loss":
|
| 399 |
"step": 32500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 2.
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss":
|
| 405 |
"step": 33000
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch": 2.
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss":
|
| 411 |
"step": 33500
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 2.
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss":
|
| 417 |
"step": 34000
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 2.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss":
|
| 423 |
"step": 34500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 2.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss":
|
| 429 |
"step": 35000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 2.
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss":
|
| 435 |
"step": 35500
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch": 2.
|
| 439 |
-
"learning_rate": 2.
|
| 440 |
-
"loss":
|
| 441 |
"step": 36000
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch": 2.
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss":
|
| 447 |
"step": 36500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch": 2.
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss":
|
| 453 |
"step": 37000
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch": 2.
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss":
|
| 459 |
"step": 37500
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"
|
| 464 |
-
"
|
| 465 |
-
"
|
| 466 |
-
|
| 467 |
-
|
| 468 |
-
"
|
| 469 |
-
"learning_rate": 2.1247199402539207e-05,
|
| 470 |
-
"loss": 1.3877,
|
| 471 |
-
"step": 38500
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 2.91,
|
| 475 |
-
"learning_rate": 2.0873786407766992e-05,
|
| 476 |
-
"loss": 1.4218,
|
| 477 |
-
"step": 39000
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 2.95,
|
| 481 |
-
"learning_rate": 2.0500373412994773e-05,
|
| 482 |
-
"loss": 1.4167,
|
| 483 |
-
"step": 39500
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 2.99,
|
| 487 |
-
"learning_rate": 2.0126960418222554e-05,
|
| 488 |
-
"loss": 1.3621,
|
| 489 |
-
"step": 40000
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 3.02,
|
| 493 |
-
"learning_rate": 1.9753547423450336e-05,
|
| 494 |
-
"loss": 1.0586,
|
| 495 |
-
"step": 40500
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 3.06,
|
| 499 |
-
"learning_rate": 1.9380134428678117e-05,
|
| 500 |
-
"loss": 0.9882,
|
| 501 |
-
"step": 41000
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 3.1,
|
| 505 |
-
"learning_rate": 1.9006721433905898e-05,
|
| 506 |
-
"loss": 0.9795,
|
| 507 |
-
"step": 41500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 3.14,
|
| 511 |
-
"learning_rate": 1.8633308439133683e-05,
|
| 512 |
-
"loss": 0.9992,
|
| 513 |
-
"step": 42000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 3.17,
|
| 517 |
-
"learning_rate": 1.8259895444361464e-05,
|
| 518 |
-
"loss": 0.9474,
|
| 519 |
-
"step": 42500
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 3.21,
|
| 523 |
-
"learning_rate": 1.7886482449589245e-05,
|
| 524 |
-
"loss": 0.9975,
|
| 525 |
-
"step": 43000
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 3.25,
|
| 529 |
-
"learning_rate": 1.7513069454817026e-05,
|
| 530 |
-
"loss": 0.9606,
|
| 531 |
-
"step": 43500
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 3.29,
|
| 535 |
-
"learning_rate": 1.713965646004481e-05,
|
| 536 |
-
"loss": 0.9781,
|
| 537 |
-
"step": 44000
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 3.32,
|
| 541 |
-
"learning_rate": 1.6766243465272592e-05,
|
| 542 |
-
"loss": 1.0437,
|
| 543 |
-
"step": 44500
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 3.36,
|
| 547 |
-
"learning_rate": 1.6392830470500377e-05,
|
| 548 |
-
"loss": 0.963,
|
| 549 |
-
"step": 45000
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 3.4,
|
| 553 |
-
"learning_rate": 1.6019417475728158e-05,
|
| 554 |
-
"loss": 1.0407,
|
| 555 |
-
"step": 45500
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 3.44,
|
| 559 |
-
"learning_rate": 1.564600448095594e-05,
|
| 560 |
-
"loss": 1.0054,
|
| 561 |
-
"step": 46000
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 3.47,
|
| 565 |
-
"learning_rate": 1.527259148618372e-05,
|
| 566 |
-
"loss": 1.0396,
|
| 567 |
-
"step": 46500
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 3.51,
|
| 571 |
-
"learning_rate": 1.4899178491411503e-05,
|
| 572 |
-
"loss": 1.0051,
|
| 573 |
-
"step": 47000
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 3.55,
|
| 577 |
-
"learning_rate": 1.4525765496639285e-05,
|
| 578 |
-
"loss": 0.9828,
|
| 579 |
-
"step": 47500
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 3.58,
|
| 583 |
-
"learning_rate": 1.4152352501867066e-05,
|
| 584 |
-
"loss": 0.9976,
|
| 585 |
-
"step": 48000
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 3.62,
|
| 589 |
-
"learning_rate": 1.3778939507094849e-05,
|
| 590 |
-
"loss": 1.0251,
|
| 591 |
-
"step": 48500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 3.66,
|
| 595 |
-
"learning_rate": 1.340552651232263e-05,
|
| 596 |
-
"loss": 1.0432,
|
| 597 |
-
"step": 49000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 3.7,
|
| 601 |
-
"learning_rate": 1.3032113517550413e-05,
|
| 602 |
-
"loss": 0.9727,
|
| 603 |
-
"step": 49500
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 3.73,
|
| 607 |
-
"learning_rate": 1.2658700522778194e-05,
|
| 608 |
-
"loss": 0.9822,
|
| 609 |
-
"step": 50000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 3.77,
|
| 613 |
-
"learning_rate": 1.2285287528005974e-05,
|
| 614 |
-
"loss": 1.0392,
|
| 615 |
-
"step": 50500
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch": 3.81,
|
| 619 |
-
"learning_rate": 1.1911874533233758e-05,
|
| 620 |
-
"loss": 0.9943,
|
| 621 |
-
"step": 51000
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 3.85,
|
| 625 |
-
"learning_rate": 1.153846153846154e-05,
|
| 626 |
-
"loss": 1.0256,
|
| 627 |
-
"step": 51500
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 3.88,
|
| 631 |
-
"learning_rate": 1.116504854368932e-05,
|
| 632 |
-
"loss": 1.0166,
|
| 633 |
-
"step": 52000
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 3.92,
|
| 637 |
-
"learning_rate": 1.0791635548917104e-05,
|
| 638 |
-
"loss": 0.9493,
|
| 639 |
-
"step": 52500
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 3.96,
|
| 643 |
-
"learning_rate": 1.0418222554144885e-05,
|
| 644 |
-
"loss": 1.0561,
|
| 645 |
-
"step": 53000
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 4.0,
|
| 649 |
-
"learning_rate": 1.0044809559372668e-05,
|
| 650 |
-
"loss": 0.976,
|
| 651 |
-
"step": 53500
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 4.03,
|
| 655 |
-
"learning_rate": 9.671396564600449e-06,
|
| 656 |
-
"loss": 0.7222,
|
| 657 |
-
"step": 54000
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 4.07,
|
| 661 |
-
"learning_rate": 9.29798356982823e-06,
|
| 662 |
-
"loss": 0.6606,
|
| 663 |
-
"step": 54500
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 4.11,
|
| 667 |
-
"learning_rate": 8.924570575056013e-06,
|
| 668 |
-
"loss": 0.6666,
|
| 669 |
-
"step": 55000
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 4.14,
|
| 673 |
-
"learning_rate": 8.551157580283794e-06,
|
| 674 |
-
"loss": 0.6524,
|
| 675 |
-
"step": 55500
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 4.18,
|
| 679 |
-
"learning_rate": 8.177744585511576e-06,
|
| 680 |
-
"loss": 0.6646,
|
| 681 |
-
"step": 56000
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 4.22,
|
| 685 |
-
"learning_rate": 7.804331590739359e-06,
|
| 686 |
-
"loss": 0.7429,
|
| 687 |
-
"step": 56500
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 4.26,
|
| 691 |
-
"learning_rate": 7.43091859596714e-06,
|
| 692 |
-
"loss": 0.711,
|
| 693 |
-
"step": 57000
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 4.29,
|
| 697 |
-
"learning_rate": 7.057505601194922e-06,
|
| 698 |
-
"loss": 0.6983,
|
| 699 |
-
"step": 57500
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 4.33,
|
| 703 |
-
"learning_rate": 6.684092606422704e-06,
|
| 704 |
-
"loss": 0.6639,
|
| 705 |
-
"step": 58000
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 4.37,
|
| 709 |
-
"learning_rate": 6.310679611650486e-06,
|
| 710 |
-
"loss": 0.7254,
|
| 711 |
-
"step": 58500
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 4.41,
|
| 715 |
-
"learning_rate": 5.937266616878267e-06,
|
| 716 |
-
"loss": 0.701,
|
| 717 |
-
"step": 59000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 4.44,
|
| 721 |
-
"learning_rate": 5.563853622106049e-06,
|
| 722 |
-
"loss": 0.7424,
|
| 723 |
-
"step": 59500
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 4.48,
|
| 727 |
-
"learning_rate": 5.190440627333831e-06,
|
| 728 |
-
"loss": 0.6679,
|
| 729 |
-
"step": 60000
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 4.52,
|
| 733 |
-
"learning_rate": 4.817027632561613e-06,
|
| 734 |
-
"loss": 0.6413,
|
| 735 |
-
"step": 60500
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 4.56,
|
| 739 |
-
"learning_rate": 4.443614637789395e-06,
|
| 740 |
-
"loss": 0.7014,
|
| 741 |
-
"step": 61000
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 4.59,
|
| 745 |
-
"learning_rate": 4.070201643017177e-06,
|
| 746 |
-
"loss": 0.6786,
|
| 747 |
-
"step": 61500
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 4.63,
|
| 751 |
-
"learning_rate": 3.6967886482449596e-06,
|
| 752 |
-
"loss": 0.6792,
|
| 753 |
-
"step": 62000
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 4.67,
|
| 757 |
-
"learning_rate": 3.3233756534727413e-06,
|
| 758 |
-
"loss": 0.7148,
|
| 759 |
-
"step": 62500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 4.71,
|
| 763 |
-
"learning_rate": 2.949962658700523e-06,
|
| 764 |
-
"loss": 0.6512,
|
| 765 |
-
"step": 63000
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 4.74,
|
| 769 |
-
"learning_rate": 2.5765496639283046e-06,
|
| 770 |
-
"loss": 0.681,
|
| 771 |
-
"step": 63500
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 4.78,
|
| 775 |
-
"learning_rate": 2.203136669156087e-06,
|
| 776 |
-
"loss": 0.7019,
|
| 777 |
-
"step": 64000
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 4.82,
|
| 781 |
-
"learning_rate": 1.8297236743838687e-06,
|
| 782 |
-
"loss": 0.5755,
|
| 783 |
-
"step": 64500
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 4.85,
|
| 787 |
-
"learning_rate": 1.4563106796116506e-06,
|
| 788 |
-
"loss": 0.5506,
|
| 789 |
-
"step": 65000
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 4.89,
|
| 793 |
-
"learning_rate": 1.0828976848394324e-06,
|
| 794 |
-
"loss": 0.5695,
|
| 795 |
-
"step": 65500
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 4.93,
|
| 799 |
-
"learning_rate": 7.094846900672144e-07,
|
| 800 |
-
"loss": 0.5534,
|
| 801 |
-
"step": 66000
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 4.97,
|
| 805 |
-
"learning_rate": 3.360716952949963e-07,
|
| 806 |
-
"loss": 0.5473,
|
| 807 |
-
"step": 66500
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 5.0,
|
| 811 |
-
"learning_rate": 1.425904192894484e-05,
|
| 812 |
-
"loss": 0.5644,
|
| 813 |
-
"step": 67000
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 5.04,
|
| 817 |
-
"learning_rate": 1.39923183612504e-05,
|
| 818 |
-
"loss": 0.574,
|
| 819 |
-
"step": 67500
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 5.08,
|
| 823 |
-
"learning_rate": 1.3725594793555959e-05,
|
| 824 |
-
"loss": 0.5951,
|
| 825 |
-
"step": 68000
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 5.12,
|
| 829 |
-
"learning_rate": 1.345887122586152e-05,
|
| 830 |
-
"loss": 0.6167,
|
| 831 |
-
"step": 68500
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 5.15,
|
| 835 |
-
"learning_rate": 1.3192147658167076e-05,
|
| 836 |
-
"loss": 0.6116,
|
| 837 |
-
"step": 69000
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 5.19,
|
| 841 |
-
"learning_rate": 1.2925424090472635e-05,
|
| 842 |
-
"loss": 0.6739,
|
| 843 |
-
"step": 69500
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 5.23,
|
| 847 |
-
"learning_rate": 1.2658700522778194e-05,
|
| 848 |
-
"loss": 0.6638,
|
| 849 |
-
"step": 70000
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 5.27,
|
| 853 |
-
"learning_rate": 1.2391976955083752e-05,
|
| 854 |
-
"loss": 0.6339,
|
| 855 |
-
"step": 70500
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 5.3,
|
| 859 |
-
"learning_rate": 1.212525338738931e-05,
|
| 860 |
-
"loss": 0.6601,
|
| 861 |
-
"step": 71000
|
| 862 |
-
},
|
| 863 |
-
{
|
| 864 |
-
"epoch": 5.34,
|
| 865 |
-
"learning_rate": 1.1858529819694868e-05,
|
| 866 |
-
"loss": 0.6563,
|
| 867 |
-
"step": 71500
|
| 868 |
-
},
|
| 869 |
-
{
|
| 870 |
-
"epoch": 5.38,
|
| 871 |
-
"learning_rate": 1.1591806252000428e-05,
|
| 872 |
-
"loss": 0.6491,
|
| 873 |
-
"step": 72000
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 5.41,
|
| 877 |
-
"learning_rate": 1.1325082684305987e-05,
|
| 878 |
-
"loss": 0.6026,
|
| 879 |
-
"step": 72500
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 5.45,
|
| 883 |
-
"learning_rate": 1.1058359116611544e-05,
|
| 884 |
-
"loss": 0.6851,
|
| 885 |
-
"step": 73000
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 5.49,
|
| 889 |
-
"learning_rate": 1.0791635548917104e-05,
|
| 890 |
-
"loss": 0.6337,
|
| 891 |
-
"step": 73500
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 5.53,
|
| 895 |
-
"learning_rate": 1.0524911981222661e-05,
|
| 896 |
-
"loss": 0.6659,
|
| 897 |
-
"step": 74000
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 5.56,
|
| 901 |
-
"learning_rate": 1.025818841352822e-05,
|
| 902 |
-
"loss": 0.6691,
|
| 903 |
-
"step": 74500
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 5.6,
|
| 907 |
-
"learning_rate": 9.991464845833778e-06,
|
| 908 |
-
"loss": 0.6611,
|
| 909 |
-
"step": 75000
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 5.64,
|
| 913 |
-
"learning_rate": 9.724741278139336e-06,
|
| 914 |
-
"loss": 0.6542,
|
| 915 |
-
"step": 75500
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 5.68,
|
| 919 |
-
"learning_rate": 9.458017710444895e-06,
|
| 920 |
-
"loss": 0.6849,
|
| 921 |
-
"step": 76000
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 5.71,
|
| 925 |
-
"learning_rate": 9.191294142750454e-06,
|
| 926 |
-
"loss": 0.6552,
|
| 927 |
-
"step": 76500
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 5.75,
|
| 931 |
-
"learning_rate": 8.924570575056013e-06,
|
| 932 |
-
"loss": 0.6636,
|
| 933 |
-
"step": 77000
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 5.79,
|
| 937 |
-
"learning_rate": 8.65784700736157e-06,
|
| 938 |
-
"loss": 0.6323,
|
| 939 |
-
"step": 77500
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 5.83,
|
| 943 |
-
"learning_rate": 8.39112343966713e-06,
|
| 944 |
-
"loss": 0.6581,
|
| 945 |
-
"step": 78000
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 5.86,
|
| 949 |
-
"learning_rate": 8.124399871972687e-06,
|
| 950 |
-
"loss": 0.6613,
|
| 951 |
-
"step": 78500
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 5.9,
|
| 955 |
-
"learning_rate": 7.857676304278247e-06,
|
| 956 |
-
"loss": 0.6816,
|
| 957 |
-
"step": 79000
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 5.94,
|
| 961 |
-
"learning_rate": 7.590952736583805e-06,
|
| 962 |
-
"loss": 0.6908,
|
| 963 |
-
"step": 79500
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 5.97,
|
| 967 |
-
"learning_rate": 7.324229168889363e-06,
|
| 968 |
-
"loss": 0.6854,
|
| 969 |
-
"step": 80000
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 6.01,
|
| 973 |
-
"learning_rate": 7.057505601194922e-06,
|
| 974 |
-
"loss": 0.5705,
|
| 975 |
-
"step": 80500
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 6.05,
|
| 979 |
-
"learning_rate": 6.79078203350048e-06,
|
| 980 |
-
"loss": 0.4532,
|
| 981 |
-
"step": 81000
|
| 982 |
-
},
|
| 983 |
-
{
|
| 984 |
-
"epoch": 6.09,
|
| 985 |
-
"learning_rate": 6.5240584658060394e-06,
|
| 986 |
-
"loss": 0.4785,
|
| 987 |
-
"step": 81500
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 6.12,
|
| 991 |
-
"learning_rate": 6.257334898111597e-06,
|
| 992 |
-
"loss": 0.4568,
|
| 993 |
-
"step": 82000
|
| 994 |
-
},
|
| 995 |
-
{
|
| 996 |
-
"epoch": 6.16,
|
| 997 |
-
"learning_rate": 5.990611330417156e-06,
|
| 998 |
-
"loss": 0.4415,
|
| 999 |
-
"step": 82500
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 6.2,
|
| 1003 |
-
"learning_rate": 5.723887762722714e-06,
|
| 1004 |
-
"loss": 0.4598,
|
| 1005 |
-
"step": 83000
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 6.24,
|
| 1009 |
-
"learning_rate": 5.457164195028273e-06,
|
| 1010 |
-
"loss": 0.4661,
|
| 1011 |
-
"step": 83500
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 6.27,
|
| 1015 |
-
"learning_rate": 5.190440627333831e-06,
|
| 1016 |
-
"loss": 0.437,
|
| 1017 |
-
"step": 84000
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 6.31,
|
| 1021 |
-
"learning_rate": 4.92371705963939e-06,
|
| 1022 |
-
"loss": 0.4782,
|
| 1023 |
-
"step": 84500
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 6.35,
|
| 1027 |
-
"learning_rate": 4.656993491944948e-06,
|
| 1028 |
-
"loss": 0.465,
|
| 1029 |
-
"step": 85000
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 6.39,
|
| 1033 |
-
"learning_rate": 4.390269924250507e-06,
|
| 1034 |
-
"loss": 0.4491,
|
| 1035 |
-
"step": 85500
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 6.42,
|
| 1039 |
-
"learning_rate": 4.123546356556066e-06,
|
| 1040 |
-
"loss": 0.4727,
|
| 1041 |
-
"step": 86000
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 6.46,
|
| 1045 |
-
"learning_rate": 3.856822788861624e-06,
|
| 1046 |
-
"loss": 0.4576,
|
| 1047 |
-
"step": 86500
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 6.5,
|
| 1051 |
-
"learning_rate": 3.590099221167183e-06,
|
| 1052 |
-
"loss": 0.4565,
|
| 1053 |
-
"step": 87000
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 6.53,
|
| 1057 |
-
"learning_rate": 3.3233756534727413e-06,
|
| 1058 |
-
"loss": 0.4646,
|
| 1059 |
-
"step": 87500
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 6.57,
|
| 1063 |
-
"learning_rate": 3.0566520857782996e-06,
|
| 1064 |
-
"loss": 0.4749,
|
| 1065 |
-
"step": 88000
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 6.61,
|
| 1069 |
-
"learning_rate": 2.789928518083858e-06,
|
| 1070 |
-
"loss": 0.3821,
|
| 1071 |
-
"step": 88500
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 6.65,
|
| 1075 |
-
"learning_rate": 2.523204950389417e-06,
|
| 1076 |
-
"loss": 0.4083,
|
| 1077 |
-
"step": 89000
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 6.68,
|
| 1081 |
-
"learning_rate": 2.2564813826949748e-06,
|
| 1082 |
-
"loss": 0.4041,
|
| 1083 |
-
"step": 89500
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 6.72,
|
| 1087 |
-
"learning_rate": 1.9897578150005336e-06,
|
| 1088 |
-
"loss": 0.4081,
|
| 1089 |
-
"step": 90000
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 6.76,
|
| 1093 |
-
"learning_rate": 1.723034247306092e-06,
|
| 1094 |
-
"loss": 0.4027,
|
| 1095 |
-
"step": 90500
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 6.8,
|
| 1099 |
-
"learning_rate": 1.4563106796116506e-06,
|
| 1100 |
-
"loss": 0.4354,
|
| 1101 |
-
"step": 91000
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 6.83,
|
| 1105 |
-
"learning_rate": 1.189587111917209e-06,
|
| 1106 |
-
"loss": 0.4193,
|
| 1107 |
-
"step": 91500
|
| 1108 |
-
},
|
| 1109 |
-
{
|
| 1110 |
-
"epoch": 6.87,
|
| 1111 |
-
"learning_rate": 9.228635442227675e-07,
|
| 1112 |
-
"loss": 0.3877,
|
| 1113 |
-
"step": 92000
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 6.91,
|
| 1117 |
-
"learning_rate": 6.56139976528326e-07,
|
| 1118 |
-
"loss": 0.4322,
|
| 1119 |
-
"step": 92500
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 6.95,
|
| 1123 |
-
"learning_rate": 3.894164088338846e-07,
|
| 1124 |
-
"loss": 0.4068,
|
| 1125 |
-
"step": 93000
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 6.98,
|
| 1129 |
-
"learning_rate": 1.226928411394431e-07,
|
| 1130 |
-
"loss": 0.3968,
|
| 1131 |
-
"step": 93500
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 7.0,
|
| 1135 |
-
"step": 93730,
|
| 1136 |
-
"total_flos": 2.490638010831667e+16,
|
| 1137 |
-
"train_loss": 0.024883948061263487,
|
| 1138 |
-
"train_runtime": 1768.5377,
|
| 1139 |
-
"train_samples_per_second": 211.99,
|
| 1140 |
-
"train_steps_per_second": 52.999
|
| 1141 |
}
|
| 1142 |
],
|
| 1143 |
"logging_steps": 500,
|
| 1144 |
-
"max_steps":
|
| 1145 |
"num_input_tokens_seen": 0,
|
| 1146 |
-
"num_train_epochs":
|
| 1147 |
"save_steps": 500,
|
| 1148 |
-
"total_flos":
|
| 1149 |
"train_batch_size": 4,
|
| 1150 |
"trial_name": null,
|
| 1151 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 37815,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.04,
|
| 13 |
+
"learning_rate": 4.9338886685177844e-05,
|
| 14 |
+
"loss": 3.2991,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.08,
|
| 19 |
+
"learning_rate": 4.8677773370355686e-05,
|
| 20 |
+
"loss": 2.7662,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.12,
|
| 25 |
+
"learning_rate": 4.801666005553352e-05,
|
| 26 |
+
"loss": 2.662,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.16,
|
| 31 |
+
"learning_rate": 4.735554674071136e-05,
|
| 32 |
+
"loss": 2.5305,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.2,
|
| 37 |
+
"learning_rate": 4.66944334258892e-05,
|
| 38 |
+
"loss": 2.4368,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.24,
|
| 43 |
+
"learning_rate": 4.603332011106704e-05,
|
| 44 |
+
"loss": 2.4286,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.28,
|
| 49 |
+
"learning_rate": 4.5372206796244874e-05,
|
| 50 |
+
"loss": 2.3404,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.32,
|
| 55 |
+
"learning_rate": 4.4711093481422716e-05,
|
| 56 |
+
"loss": 2.2398,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.36,
|
| 61 |
+
"learning_rate": 4.404998016660056e-05,
|
| 62 |
+
"loss": 2.2966,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.4,
|
| 67 |
+
"learning_rate": 4.33888668517784e-05,
|
| 68 |
+
"loss": 2.2255,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.44,
|
| 73 |
+
"learning_rate": 4.272775353695624e-05,
|
| 74 |
+
"loss": 2.2713,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"learning_rate": 4.2066640222134076e-05,
|
| 80 |
+
"loss": 2.2018,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.52,
|
| 85 |
+
"learning_rate": 4.140552690731192e-05,
|
| 86 |
+
"loss": 2.1203,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.56,
|
| 91 |
+
"learning_rate": 4.074441359248975e-05,
|
| 92 |
+
"loss": 2.1148,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.6,
|
| 97 |
+
"learning_rate": 4.0083300277667595e-05,
|
| 98 |
+
"loss": 2.1149,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.63,
|
| 103 |
+
"learning_rate": 3.942218696284543e-05,
|
| 104 |
+
"loss": 2.0219,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.67,
|
| 109 |
+
"learning_rate": 3.876107364802327e-05,
|
| 110 |
+
"loss": 2.0354,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.71,
|
| 115 |
+
"learning_rate": 3.809996033320111e-05,
|
| 116 |
+
"loss": 2.0341,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.75,
|
| 121 |
+
"learning_rate": 3.7438847018378955e-05,
|
| 122 |
+
"loss": 1.9972,
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.79,
|
| 127 |
+
"learning_rate": 3.6777733703556796e-05,
|
| 128 |
+
"loss": 1.9654,
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.83,
|
| 133 |
+
"learning_rate": 3.611662038873463e-05,
|
| 134 |
+
"loss": 1.9853,
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.87,
|
| 139 |
+
"learning_rate": 3.5455507073912466e-05,
|
| 140 |
+
"loss": 1.9487,
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.91,
|
| 145 |
+
"learning_rate": 3.479439375909031e-05,
|
| 146 |
+
"loss": 1.9498,
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.95,
|
| 151 |
+
"learning_rate": 3.413328044426815e-05,
|
| 152 |
+
"loss": 1.8963,
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.99,
|
| 157 |
+
"learning_rate": 3.3472167129445985e-05,
|
| 158 |
+
"loss": 1.9259,
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 1.03,
|
| 163 |
+
"learning_rate": 3.2811053814623827e-05,
|
| 164 |
+
"loss": 1.5878,
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 1.07,
|
| 169 |
+
"learning_rate": 3.214994049980167e-05,
|
| 170 |
+
"loss": 1.4017,
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 1.11,
|
| 175 |
+
"learning_rate": 3.148882718497951e-05,
|
| 176 |
+
"loss": 1.4809,
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 1.15,
|
| 181 |
+
"learning_rate": 3.082771387015735e-05,
|
| 182 |
+
"loss": 1.4646,
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 1.19,
|
| 187 |
+
"learning_rate": 3.0166600555335183e-05,
|
| 188 |
+
"loss": 1.5017,
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 1.23,
|
| 193 |
+
"learning_rate": 2.9505487240513025e-05,
|
| 194 |
+
"loss": 1.4745,
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 1.27,
|
| 199 |
+
"learning_rate": 2.8844373925690867e-05,
|
| 200 |
+
"loss": 1.4496,
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 1.31,
|
| 205 |
+
"learning_rate": 2.8183260610868705e-05,
|
| 206 |
+
"loss": 1.4599,
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 1.35,
|
| 211 |
+
"learning_rate": 2.752214729604654e-05,
|
| 212 |
+
"loss": 1.3974,
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 1.39,
|
| 217 |
+
"learning_rate": 2.6861033981224382e-05,
|
| 218 |
+
"loss": 1.397,
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 1.43,
|
| 223 |
+
"learning_rate": 2.6199920666402224e-05,
|
| 224 |
+
"loss": 1.436,
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 1.47,
|
| 229 |
+
"learning_rate": 2.5538807351580062e-05,
|
| 230 |
+
"loss": 1.4359,
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 1.51,
|
| 235 |
+
"learning_rate": 2.48776940367579e-05,
|
| 236 |
+
"loss": 1.4215,
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 1.55,
|
| 241 |
+
"learning_rate": 2.4216580721935742e-05,
|
| 242 |
+
"loss": 1.3611,
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 1.59,
|
| 247 |
+
"learning_rate": 2.355546740711358e-05,
|
| 248 |
+
"loss": 1.4515,
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 1.63,
|
| 253 |
+
"learning_rate": 2.289435409229142e-05,
|
| 254 |
+
"loss": 1.3923,
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 1.67,
|
| 259 |
+
"learning_rate": 2.2233240777469257e-05,
|
| 260 |
+
"loss": 1.3968,
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 1.71,
|
| 265 |
+
"learning_rate": 2.15721274626471e-05,
|
| 266 |
+
"loss": 1.4511,
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 1.75,
|
| 271 |
+
"learning_rate": 2.091101414782494e-05,
|
| 272 |
+
"loss": 1.3736,
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 1.79,
|
| 277 |
+
"learning_rate": 2.024990083300278e-05,
|
| 278 |
+
"loss": 1.4196,
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 1.82,
|
| 283 |
+
"learning_rate": 1.9588787518180617e-05,
|
| 284 |
+
"loss": 1.4012,
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 1.86,
|
| 289 |
+
"learning_rate": 1.8927674203358456e-05,
|
| 290 |
+
"loss": 1.3995,
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 1.9,
|
| 295 |
+
"learning_rate": 1.8266560888536297e-05,
|
| 296 |
+
"loss": 1.3706,
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 1.94,
|
| 301 |
+
"learning_rate": 1.7605447573714136e-05,
|
| 302 |
+
"loss": 1.357,
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 1.98,
|
| 307 |
+
"learning_rate": 1.6944334258891974e-05,
|
| 308 |
+
"loss": 1.3894,
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 2.02,
|
| 313 |
+
"learning_rate": 1.6283220944069812e-05,
|
| 314 |
+
"loss": 1.1266,
|
| 315 |
"step": 25500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 2.06,
|
| 319 |
+
"learning_rate": 1.5622107629247654e-05,
|
| 320 |
+
"loss": 0.9034,
|
| 321 |
"step": 26000
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 2.1,
|
| 325 |
+
"learning_rate": 1.4960994314425494e-05,
|
| 326 |
+
"loss": 0.88,
|
| 327 |
"step": 26500
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 2.14,
|
| 331 |
+
"learning_rate": 1.4299880999603333e-05,
|
| 332 |
+
"loss": 0.9379,
|
| 333 |
"step": 27000
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 2.18,
|
| 337 |
+
"learning_rate": 1.3638767684781173e-05,
|
| 338 |
+
"loss": 0.9224,
|
| 339 |
"step": 27500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 2.22,
|
| 343 |
+
"learning_rate": 1.2977654369959011e-05,
|
| 344 |
+
"loss": 0.9108,
|
| 345 |
"step": 28000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 2.26,
|
| 349 |
+
"learning_rate": 1.2316541055136851e-05,
|
| 350 |
+
"loss": 0.9068,
|
| 351 |
"step": 28500
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 2.3,
|
| 355 |
+
"learning_rate": 1.1655427740314691e-05,
|
| 356 |
+
"loss": 0.9158,
|
| 357 |
"step": 29000
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 2.34,
|
| 361 |
+
"learning_rate": 1.099431442549253e-05,
|
| 362 |
+
"loss": 0.901,
|
| 363 |
"step": 29500
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 2.38,
|
| 367 |
+
"learning_rate": 1.033320111067037e-05,
|
| 368 |
+
"loss": 0.8898,
|
| 369 |
"step": 30000
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 2.42,
|
| 373 |
+
"learning_rate": 9.67208779584821e-06,
|
| 374 |
+
"loss": 0.9295,
|
| 375 |
"step": 30500
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 2.46,
|
| 379 |
+
"learning_rate": 9.010974481026048e-06,
|
| 380 |
+
"loss": 0.9325,
|
| 381 |
"step": 31000
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 2.5,
|
| 385 |
+
"learning_rate": 8.349861166203888e-06,
|
| 386 |
+
"loss": 0.9357,
|
| 387 |
"step": 31500
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 2.54,
|
| 391 |
+
"learning_rate": 7.688747851381726e-06,
|
| 392 |
+
"loss": 0.8832,
|
| 393 |
"step": 32000
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 2.58,
|
| 397 |
+
"learning_rate": 7.027634536559567e-06,
|
| 398 |
+
"loss": 0.9101,
|
| 399 |
"step": 32500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 2.62,
|
| 403 |
+
"learning_rate": 6.366521221737406e-06,
|
| 404 |
+
"loss": 0.9018,
|
| 405 |
"step": 33000
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 2.66,
|
| 409 |
+
"learning_rate": 5.7054079069152455e-06,
|
| 410 |
+
"loss": 0.8886,
|
| 411 |
"step": 33500
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 2.7,
|
| 415 |
+
"learning_rate": 5.044294592093085e-06,
|
| 416 |
+
"loss": 0.8771,
|
| 417 |
"step": 34000
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 2.74,
|
| 421 |
+
"learning_rate": 4.383181277270925e-06,
|
| 422 |
+
"loss": 0.8956,
|
| 423 |
"step": 34500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 2.78,
|
| 427 |
+
"learning_rate": 3.7220679624487635e-06,
|
| 428 |
+
"loss": 0.8586,
|
| 429 |
"step": 35000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 2.82,
|
| 433 |
+
"learning_rate": 3.060954647626603e-06,
|
| 434 |
+
"loss": 0.9039,
|
| 435 |
"step": 35500
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 2.86,
|
| 439 |
+
"learning_rate": 2.3998413328044427e-06,
|
| 440 |
+
"loss": 0.8817,
|
| 441 |
"step": 36000
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 2.9,
|
| 445 |
+
"learning_rate": 1.7387280179822822e-06,
|
| 446 |
+
"loss": 0.8601,
|
| 447 |
"step": 36500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 2.94,
|
| 451 |
+
"learning_rate": 1.0776147031601218e-06,
|
| 452 |
+
"loss": 0.8837,
|
| 453 |
"step": 37000
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 2.98,
|
| 457 |
+
"learning_rate": 4.165013883379611e-07,
|
| 458 |
+
"loss": 0.8894,
|
| 459 |
"step": 37500
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 3.0,
|
| 463 |
+
"step": 37815,
|
| 464 |
+
"total_flos": 9288563680542720.0,
|
| 465 |
+
"train_loss": 1.5174339204652274,
|
| 466 |
+
"train_runtime": 9956.4506,
|
| 467 |
+
"train_samples_per_second": 15.192,
|
| 468 |
+
"train_steps_per_second": 3.798
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 469 |
}
|
| 470 |
],
|
| 471 |
"logging_steps": 500,
|
| 472 |
+
"max_steps": 37815,
|
| 473 |
"num_input_tokens_seen": 0,
|
| 474 |
+
"num_train_epochs": 3,
|
| 475 |
"save_steps": 500,
|
| 476 |
+
"total_flos": 9288563680542720.0,
|
| 477 |
"train_batch_size": 4,
|
| 478 |
"trial_name": null,
|
| 479 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a56bee8fa4e14938cb95192439c644b615e9dd68f07ea8d99031a4d251338efc
|
| 3 |
size 5048
|