End of training
Browse files- README.md +12 -10
- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- trainer_state.json +235 -111
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 1.
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -45,20 +45,22 @@ The following hyperparameters were used during training:
|
|
| 45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
- lr_scheduler_type: cosine
|
| 47 |
- lr_scheduler_warmup_ratio: 0.03
|
| 48 |
-
- num_epochs:
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
|
|
|
|
|
|
| 62 |
|
| 63 |
|
| 64 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [liuhaotian/llava-v1.5-13b](https://huggingface.co/liuhaotian/llava-v1.5-13b) on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 1.6014
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
- lr_scheduler_type: cosine
|
| 47 |
- lr_scheduler_warmup_ratio: 0.03
|
| 48 |
+
- num_epochs: 10.0
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 54 |
+
| 0.9019 | 1.0 | 9 | 1.0177 |
|
| 55 |
+
| 0.2356 | 2.0 | 18 | 1.1177 |
|
| 56 |
+
| 0.0761 | 3.0 | 27 | 1.1808 |
|
| 57 |
+
| 0.0304 | 4.0 | 36 | 1.3987 |
|
| 58 |
+
| 0.0092 | 5.0 | 45 | 1.4471 |
|
| 59 |
+
| 0.009 | 6.0 | 54 | 1.4910 |
|
| 60 |
+
| 0.0055 | 7.0 | 63 | 1.5329 |
|
| 61 |
+
| 0.0022 | 8.0 | 72 | 1.5343 |
|
| 62 |
+
| 0.002 | 9.0 | 81 | 1.5295 |
|
| 63 |
+
| 0.0011 | 10.0 | 90 | 1.6014 |
|
| 64 |
|
| 65 |
|
| 66 |
### Framework versions
|
adapter_config.json
CHANGED
|
@@ -20,13 +20,13 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
| 23 |
"v_proj",
|
| 24 |
"k_proj",
|
| 25 |
-
"q_proj",
|
| 26 |
"o_proj",
|
| 27 |
-
"down_proj",
|
| 28 |
"up_proj",
|
| 29 |
-
"
|
|
|
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"gate_proj",
|
| 24 |
"v_proj",
|
| 25 |
"k_proj",
|
|
|
|
| 26 |
"o_proj",
|
|
|
|
| 27 |
"up_proj",
|
| 28 |
+
"q_proj",
|
| 29 |
+
"down_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1001466944
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5cde74d154b0ceab1269ba61d5a4b6fb8824ca53401ce2d7883ea4260038dbb7
|
| 3 |
size 1001466944
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 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,
|
|
@@ -17,199 +17,199 @@
|
|
| 17 |
{
|
| 18 |
"epoch": 0.22,
|
| 19 |
"learning_rate": 0.00012618595071429146,
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 2
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.33,
|
| 25 |
"learning_rate": 0.0002,
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 3
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.44,
|
| 31 |
"learning_rate": 0.0002,
|
| 32 |
-
"loss": 1.
|
| 33 |
"step": 4
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.56,
|
| 37 |
"learning_rate": 0.0002,
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 5
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.67,
|
| 43 |
"learning_rate": 0.0002,
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 6
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.78,
|
| 49 |
"learning_rate": 0.0002,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 7
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.89,
|
| 55 |
"learning_rate": 0.0002,
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 8
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 1.0,
|
| 61 |
"learning_rate": 0.0002,
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 9
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 1.0,
|
| 67 |
-
"eval_loss": 1.
|
| 68 |
-
"eval_runtime":
|
| 69 |
-
"eval_samples_per_second":
|
| 70 |
-
"eval_steps_per_second": 0.
|
| 71 |
"step": 9
|
| 72 |
},
|
| 73 |
{
|
| 74 |
"epoch": 1.11,
|
| 75 |
"learning_rate": 0.0002,
|
| 76 |
-
"loss": 0.
|
| 77 |
"step": 10
|
| 78 |
},
|
| 79 |
{
|
| 80 |
"epoch": 1.22,
|
| 81 |
"learning_rate": 0.0002,
|
| 82 |
-
"loss": 0.
|
| 83 |
"step": 11
|
| 84 |
},
|
| 85 |
{
|
| 86 |
"epoch": 1.33,
|
| 87 |
"learning_rate": 0.0002,
|
| 88 |
-
"loss": 0.
|
| 89 |
"step": 12
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 1.44,
|
| 93 |
"learning_rate": 0.0002,
|
| 94 |
-
"loss": 0.
|
| 95 |
"step": 13
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.56,
|
| 99 |
"learning_rate": 0.0002,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 14
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 1.67,
|
| 105 |
"learning_rate": 0.0002,
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 15
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 1.78,
|
| 111 |
"learning_rate": 0.0002,
|
| 112 |
-
"loss": 0.
|
| 113 |
"step": 16
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"epoch": 1.89,
|
| 117 |
"learning_rate": 0.0002,
|
| 118 |
-
"loss": 0.
|
| 119 |
"step": 17
|
| 120 |
},
|
| 121 |
{
|
| 122 |
"epoch": 2.0,
|
| 123 |
"learning_rate": 0.0002,
|
| 124 |
-
"loss": 0.
|
| 125 |
"step": 18
|
| 126 |
},
|
| 127 |
{
|
| 128 |
"epoch": 2.0,
|
| 129 |
-
"eval_loss": 1.
|
| 130 |
-
"eval_runtime":
|
| 131 |
-
"eval_samples_per_second":
|
| 132 |
-
"eval_steps_per_second": 0.
|
| 133 |
"step": 18
|
| 134 |
},
|
| 135 |
{
|
| 136 |
"epoch": 2.11,
|
| 137 |
"learning_rate": 0.0002,
|
| 138 |
-
"loss": 0.
|
| 139 |
"step": 19
|
| 140 |
},
|
| 141 |
{
|
| 142 |
"epoch": 2.22,
|
| 143 |
"learning_rate": 0.0002,
|
| 144 |
-
"loss": 0.
|
| 145 |
"step": 20
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 2.33,
|
| 149 |
"learning_rate": 0.0002,
|
| 150 |
-
"loss": 0.
|
| 151 |
"step": 21
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 2.44,
|
| 155 |
"learning_rate": 0.0002,
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 22
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 2.56,
|
| 161 |
"learning_rate": 0.0002,
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 23
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 2.67,
|
| 167 |
"learning_rate": 0.0002,
|
| 168 |
-
"loss": 0.
|
| 169 |
"step": 24
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 2.78,
|
| 173 |
"learning_rate": 0.0002,
|
| 174 |
-
"loss": 0.
|
| 175 |
"step": 25
|
| 176 |
},
|
| 177 |
{
|
| 178 |
"epoch": 2.89,
|
| 179 |
"learning_rate": 0.0002,
|
| 180 |
-
"loss": 0.
|
| 181 |
"step": 26
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 3.0,
|
| 185 |
"learning_rate": 0.0002,
|
| 186 |
-
"loss": 0.
|
| 187 |
"step": 27
|
| 188 |
},
|
| 189 |
{
|
| 190 |
"epoch": 3.0,
|
| 191 |
-
"eval_loss": 1.
|
| 192 |
-
"eval_runtime":
|
| 193 |
-
"eval_samples_per_second":
|
| 194 |
-
"eval_steps_per_second": 0.
|
| 195 |
"step": 27
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 3.11,
|
| 199 |
"learning_rate": 0.0002,
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 28
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 3.22,
|
| 205 |
"learning_rate": 0.0002,
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 29
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 3.33,
|
| 211 |
"learning_rate": 0.0002,
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 30
|
| 214 |
},
|
| 215 |
{
|
|
@@ -221,81 +221,81 @@
|
|
| 221 |
{
|
| 222 |
"epoch": 3.56,
|
| 223 |
"learning_rate": 0.0002,
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 32
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 3.67,
|
| 229 |
"learning_rate": 0.0002,
|
| 230 |
-
"loss": 0.
|
| 231 |
"step": 33
|
| 232 |
},
|
| 233 |
{
|
| 234 |
"epoch": 3.78,
|
| 235 |
"learning_rate": 0.0002,
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 34
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 3.89,
|
| 241 |
"learning_rate": 0.0002,
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 35
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 4.0,
|
| 247 |
"learning_rate": 0.0002,
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 36
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 4.0,
|
| 253 |
-
"eval_loss": 1.
|
| 254 |
-
"eval_runtime":
|
| 255 |
-
"eval_samples_per_second":
|
| 256 |
-
"eval_steps_per_second": 0.
|
| 257 |
"step": 36
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 4.11,
|
| 261 |
"learning_rate": 0.0002,
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 37
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 4.22,
|
| 267 |
"learning_rate": 0.0002,
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 38
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 4.33,
|
| 273 |
"learning_rate": 0.0002,
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 39
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 4.44,
|
| 279 |
"learning_rate": 0.0002,
|
| 280 |
-
"loss": 0.
|
| 281 |
"step": 40
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 4.56,
|
| 285 |
"learning_rate": 0.0002,
|
| 286 |
-
"loss": 0.
|
| 287 |
"step": 41
|
| 288 |
},
|
| 289 |
{
|
| 290 |
"epoch": 4.67,
|
| 291 |
"learning_rate": 0.0002,
|
| 292 |
-
"loss": 0.
|
| 293 |
"step": 42
|
| 294 |
},
|
| 295 |
{
|
| 296 |
"epoch": 4.78,
|
| 297 |
"learning_rate": 0.0002,
|
| 298 |
-
"loss": 0.
|
| 299 |
"step": 43
|
| 300 |
},
|
| 301 |
{
|
|
@@ -307,45 +307,45 @@
|
|
| 307 |
{
|
| 308 |
"epoch": 5.0,
|
| 309 |
"learning_rate": 0.0002,
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 45
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 5.0,
|
| 315 |
-
"eval_loss": 1.
|
| 316 |
-
"eval_runtime":
|
| 317 |
-
"eval_samples_per_second":
|
| 318 |
-
"eval_steps_per_second": 0.
|
| 319 |
"step": 45
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 5.11,
|
| 323 |
"learning_rate": 0.0002,
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 46
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 5.22,
|
| 329 |
"learning_rate": 0.0002,
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 47
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 5.33,
|
| 335 |
"learning_rate": 0.0002,
|
| 336 |
-
"loss": 0.
|
| 337 |
"step": 48
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 5.44,
|
| 341 |
"learning_rate": 0.0002,
|
| 342 |
-
"loss": 0.
|
| 343 |
"step": 49
|
| 344 |
},
|
| 345 |
{
|
| 346 |
"epoch": 5.56,
|
| 347 |
"learning_rate": 0.0002,
|
| 348 |
-
"loss": 0.
|
| 349 |
"step": 50
|
| 350 |
},
|
| 351 |
{
|
|
@@ -357,95 +357,95 @@
|
|
| 357 |
{
|
| 358 |
"epoch": 5.78,
|
| 359 |
"learning_rate": 0.0002,
|
| 360 |
-
"loss": 0.
|
| 361 |
"step": 52
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 5.89,
|
| 365 |
"learning_rate": 0.0002,
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 53
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 6.0,
|
| 371 |
"learning_rate": 0.0002,
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 54
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 6.0,
|
| 377 |
-
"eval_loss": 1.
|
| 378 |
-
"eval_runtime":
|
| 379 |
-
"eval_samples_per_second":
|
| 380 |
-
"eval_steps_per_second": 0.
|
| 381 |
"step": 54
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 6.11,
|
| 385 |
"learning_rate": 0.0002,
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 55
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 6.22,
|
| 391 |
"learning_rate": 0.0002,
|
| 392 |
-
"loss": 0.
|
| 393 |
"step": 56
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 6.33,
|
| 397 |
"learning_rate": 0.0002,
|
| 398 |
-
"loss": 0.
|
| 399 |
"step": 57
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 6.44,
|
| 403 |
"learning_rate": 0.0002,
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 58
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 6.56,
|
| 409 |
"learning_rate": 0.0002,
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 59
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 6.67,
|
| 415 |
"learning_rate": 0.0002,
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 60
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 6.78,
|
| 421 |
"learning_rate": 0.0002,
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 61
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 6.89,
|
| 427 |
"learning_rate": 0.0002,
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 62
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 7.0,
|
| 433 |
"learning_rate": 0.0002,
|
| 434 |
-
"loss": 0.
|
| 435 |
"step": 63
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 7.0,
|
| 439 |
-
"eval_loss": 1.
|
| 440 |
-
"eval_runtime":
|
| 441 |
-
"eval_samples_per_second":
|
| 442 |
-
"eval_steps_per_second": 0.
|
| 443 |
"step": 63
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 7.11,
|
| 447 |
"learning_rate": 0.0002,
|
| 448 |
-
"loss": 0.
|
| 449 |
"step": 64
|
| 450 |
},
|
| 451 |
{
|
|
@@ -457,69 +457,193 @@
|
|
| 457 |
{
|
| 458 |
"epoch": 7.33,
|
| 459 |
"learning_rate": 0.0002,
|
| 460 |
-
"loss": 0.
|
| 461 |
"step": 66
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 7.44,
|
| 465 |
"learning_rate": 0.0002,
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 67
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 7.56,
|
| 471 |
"learning_rate": 0.0002,
|
| 472 |
-
"loss": 0.
|
| 473 |
"step": 68
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 7.67,
|
| 477 |
"learning_rate": 0.0002,
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 69
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 7.78,
|
| 483 |
"learning_rate": 0.0002,
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 70
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 7.89,
|
| 489 |
"learning_rate": 0.0002,
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 71
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 8.0,
|
| 495 |
"learning_rate": 0.0002,
|
| 496 |
-
"loss": 0.
|
| 497 |
"step": 72
|
| 498 |
},
|
| 499 |
{
|
| 500 |
"epoch": 8.0,
|
| 501 |
-
"eval_loss": 1.
|
| 502 |
-
"eval_runtime":
|
| 503 |
-
"eval_samples_per_second":
|
| 504 |
-
"eval_steps_per_second": 0.
|
| 505 |
"step": 72
|
| 506 |
},
|
| 507 |
{
|
| 508 |
-
"epoch": 8.
|
| 509 |
-
"
|
| 510 |
-
"
|
| 511 |
-
"
|
| 512 |
-
|
| 513 |
-
|
| 514 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 515 |
}
|
| 516 |
],
|
| 517 |
"logging_steps": 1.0,
|
| 518 |
-
"max_steps":
|
| 519 |
"num_input_tokens_seen": 0,
|
| 520 |
-
"num_train_epochs":
|
| 521 |
"save_steps": 50000,
|
| 522 |
-
"total_flos": 1.
|
| 523 |
"train_batch_size": 4,
|
| 524 |
"trial_name": null,
|
| 525 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 90,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 17 |
{
|
| 18 |
"epoch": 0.22,
|
| 19 |
"learning_rate": 0.00012618595071429146,
|
| 20 |
+
"loss": 1.3352,
|
| 21 |
"step": 2
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.33,
|
| 25 |
"learning_rate": 0.0002,
|
| 26 |
+
"loss": 1.4468,
|
| 27 |
"step": 3
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.44,
|
| 31 |
"learning_rate": 0.0002,
|
| 32 |
+
"loss": 1.2444,
|
| 33 |
"step": 4
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.56,
|
| 37 |
"learning_rate": 0.0002,
|
| 38 |
+
"loss": 1.1357,
|
| 39 |
"step": 5
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.67,
|
| 43 |
"learning_rate": 0.0002,
|
| 44 |
+
"loss": 1.0412,
|
| 45 |
"step": 6
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.78,
|
| 49 |
"learning_rate": 0.0002,
|
| 50 |
+
"loss": 0.9966,
|
| 51 |
"step": 7
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.89,
|
| 55 |
"learning_rate": 0.0002,
|
| 56 |
+
"loss": 0.8682,
|
| 57 |
"step": 8
|
| 58 |
},
|
| 59 |
{
|
| 60 |
"epoch": 1.0,
|
| 61 |
"learning_rate": 0.0002,
|
| 62 |
+
"loss": 0.9019,
|
| 63 |
"step": 9
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 1.0,
|
| 67 |
+
"eval_loss": 1.0177268981933594,
|
| 68 |
+
"eval_runtime": 11.4956,
|
| 69 |
+
"eval_samples_per_second": 4.349,
|
| 70 |
+
"eval_steps_per_second": 0.174,
|
| 71 |
"step": 9
|
| 72 |
},
|
| 73 |
{
|
| 74 |
"epoch": 1.11,
|
| 75 |
"learning_rate": 0.0002,
|
| 76 |
+
"loss": 0.6961,
|
| 77 |
"step": 10
|
| 78 |
},
|
| 79 |
{
|
| 80 |
"epoch": 1.22,
|
| 81 |
"learning_rate": 0.0002,
|
| 82 |
+
"loss": 0.5139,
|
| 83 |
"step": 11
|
| 84 |
},
|
| 85 |
{
|
| 86 |
"epoch": 1.33,
|
| 87 |
"learning_rate": 0.0002,
|
| 88 |
+
"loss": 0.4883,
|
| 89 |
"step": 12
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 1.44,
|
| 93 |
"learning_rate": 0.0002,
|
| 94 |
+
"loss": 0.5516,
|
| 95 |
"step": 13
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 1.56,
|
| 99 |
"learning_rate": 0.0002,
|
| 100 |
+
"loss": 0.4948,
|
| 101 |
"step": 14
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 1.67,
|
| 105 |
"learning_rate": 0.0002,
|
| 106 |
+
"loss": 0.4729,
|
| 107 |
"step": 15
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 1.78,
|
| 111 |
"learning_rate": 0.0002,
|
| 112 |
+
"loss": 0.3157,
|
| 113 |
"step": 16
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"epoch": 1.89,
|
| 117 |
"learning_rate": 0.0002,
|
| 118 |
+
"loss": 0.2823,
|
| 119 |
"step": 17
|
| 120 |
},
|
| 121 |
{
|
| 122 |
"epoch": 2.0,
|
| 123 |
"learning_rate": 0.0002,
|
| 124 |
+
"loss": 0.2356,
|
| 125 |
"step": 18
|
| 126 |
},
|
| 127 |
{
|
| 128 |
"epoch": 2.0,
|
| 129 |
+
"eval_loss": 1.1176843643188477,
|
| 130 |
+
"eval_runtime": 6.3692,
|
| 131 |
+
"eval_samples_per_second": 7.85,
|
| 132 |
+
"eval_steps_per_second": 0.314,
|
| 133 |
"step": 18
|
| 134 |
},
|
| 135 |
{
|
| 136 |
"epoch": 2.11,
|
| 137 |
"learning_rate": 0.0002,
|
| 138 |
+
"loss": 0.1731,
|
| 139 |
"step": 19
|
| 140 |
},
|
| 141 |
{
|
| 142 |
"epoch": 2.22,
|
| 143 |
"learning_rate": 0.0002,
|
| 144 |
+
"loss": 0.1148,
|
| 145 |
"step": 20
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 2.33,
|
| 149 |
"learning_rate": 0.0002,
|
| 150 |
+
"loss": 0.1148,
|
| 151 |
"step": 21
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 2.44,
|
| 155 |
"learning_rate": 0.0002,
|
| 156 |
+
"loss": 0.133,
|
| 157 |
"step": 22
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 2.56,
|
| 161 |
"learning_rate": 0.0002,
|
| 162 |
+
"loss": 0.0924,
|
| 163 |
"step": 23
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 2.67,
|
| 167 |
"learning_rate": 0.0002,
|
| 168 |
+
"loss": 0.0582,
|
| 169 |
"step": 24
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 2.78,
|
| 173 |
"learning_rate": 0.0002,
|
| 174 |
+
"loss": 0.0998,
|
| 175 |
"step": 25
|
| 176 |
},
|
| 177 |
{
|
| 178 |
"epoch": 2.89,
|
| 179 |
"learning_rate": 0.0002,
|
| 180 |
+
"loss": 0.099,
|
| 181 |
"step": 26
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 3.0,
|
| 185 |
"learning_rate": 0.0002,
|
| 186 |
+
"loss": 0.0761,
|
| 187 |
"step": 27
|
| 188 |
},
|
| 189 |
{
|
| 190 |
"epoch": 3.0,
|
| 191 |
+
"eval_loss": 1.1807804107666016,
|
| 192 |
+
"eval_runtime": 6.295,
|
| 193 |
+
"eval_samples_per_second": 7.943,
|
| 194 |
+
"eval_steps_per_second": 0.318,
|
| 195 |
"step": 27
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 3.11,
|
| 199 |
"learning_rate": 0.0002,
|
| 200 |
+
"loss": 0.0424,
|
| 201 |
"step": 28
|
| 202 |
},
|
| 203 |
{
|
| 204 |
"epoch": 3.22,
|
| 205 |
"learning_rate": 0.0002,
|
| 206 |
+
"loss": 0.0367,
|
| 207 |
"step": 29
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 3.33,
|
| 211 |
"learning_rate": 0.0002,
|
| 212 |
+
"loss": 0.0423,
|
| 213 |
"step": 30
|
| 214 |
},
|
| 215 |
{
|
|
|
|
| 221 |
{
|
| 222 |
"epoch": 3.56,
|
| 223 |
"learning_rate": 0.0002,
|
| 224 |
+
"loss": 0.0347,
|
| 225 |
"step": 32
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 3.67,
|
| 229 |
"learning_rate": 0.0002,
|
| 230 |
+
"loss": 0.0352,
|
| 231 |
"step": 33
|
| 232 |
},
|
| 233 |
{
|
| 234 |
"epoch": 3.78,
|
| 235 |
"learning_rate": 0.0002,
|
| 236 |
+
"loss": 0.0333,
|
| 237 |
"step": 34
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 3.89,
|
| 241 |
"learning_rate": 0.0002,
|
| 242 |
+
"loss": 0.0491,
|
| 243 |
"step": 35
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 4.0,
|
| 247 |
"learning_rate": 0.0002,
|
| 248 |
+
"loss": 0.0304,
|
| 249 |
"step": 36
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 4.0,
|
| 253 |
+
"eval_loss": 1.3986564874649048,
|
| 254 |
+
"eval_runtime": 6.3879,
|
| 255 |
+
"eval_samples_per_second": 7.827,
|
| 256 |
+
"eval_steps_per_second": 0.313,
|
| 257 |
"step": 36
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 4.11,
|
| 261 |
"learning_rate": 0.0002,
|
| 262 |
+
"loss": 0.0099,
|
| 263 |
"step": 37
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 4.22,
|
| 267 |
"learning_rate": 0.0002,
|
| 268 |
+
"loss": 0.016,
|
| 269 |
"step": 38
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 4.33,
|
| 273 |
"learning_rate": 0.0002,
|
| 274 |
+
"loss": 0.0133,
|
| 275 |
"step": 39
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 4.44,
|
| 279 |
"learning_rate": 0.0002,
|
| 280 |
+
"loss": 0.0128,
|
| 281 |
"step": 40
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 4.56,
|
| 285 |
"learning_rate": 0.0002,
|
| 286 |
+
"loss": 0.0101,
|
| 287 |
"step": 41
|
| 288 |
},
|
| 289 |
{
|
| 290 |
"epoch": 4.67,
|
| 291 |
"learning_rate": 0.0002,
|
| 292 |
+
"loss": 0.0204,
|
| 293 |
"step": 42
|
| 294 |
},
|
| 295 |
{
|
| 296 |
"epoch": 4.78,
|
| 297 |
"learning_rate": 0.0002,
|
| 298 |
+
"loss": 0.0263,
|
| 299 |
"step": 43
|
| 300 |
},
|
| 301 |
{
|
|
|
|
| 307 |
{
|
| 308 |
"epoch": 5.0,
|
| 309 |
"learning_rate": 0.0002,
|
| 310 |
+
"loss": 0.0092,
|
| 311 |
"step": 45
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 5.0,
|
| 315 |
+
"eval_loss": 1.4470524787902832,
|
| 316 |
+
"eval_runtime": 6.2792,
|
| 317 |
+
"eval_samples_per_second": 7.963,
|
| 318 |
+
"eval_steps_per_second": 0.319,
|
| 319 |
"step": 45
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 5.11,
|
| 323 |
"learning_rate": 0.0002,
|
| 324 |
+
"loss": 0.0043,
|
| 325 |
"step": 46
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 5.22,
|
| 329 |
"learning_rate": 0.0002,
|
| 330 |
+
"loss": 0.0111,
|
| 331 |
"step": 47
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 5.33,
|
| 335 |
"learning_rate": 0.0002,
|
| 336 |
+
"loss": 0.0084,
|
| 337 |
"step": 48
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 5.44,
|
| 341 |
"learning_rate": 0.0002,
|
| 342 |
+
"loss": 0.0075,
|
| 343 |
"step": 49
|
| 344 |
},
|
| 345 |
{
|
| 346 |
"epoch": 5.56,
|
| 347 |
"learning_rate": 0.0002,
|
| 348 |
+
"loss": 0.0048,
|
| 349 |
"step": 50
|
| 350 |
},
|
| 351 |
{
|
|
|
|
| 357 |
{
|
| 358 |
"epoch": 5.78,
|
| 359 |
"learning_rate": 0.0002,
|
| 360 |
+
"loss": 0.0069,
|
| 361 |
"step": 52
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 5.89,
|
| 365 |
"learning_rate": 0.0002,
|
| 366 |
+
"loss": 0.0058,
|
| 367 |
"step": 53
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 6.0,
|
| 371 |
"learning_rate": 0.0002,
|
| 372 |
+
"loss": 0.009,
|
| 373 |
"step": 54
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 6.0,
|
| 377 |
+
"eval_loss": 1.4909963607788086,
|
| 378 |
+
"eval_runtime": 6.398,
|
| 379 |
+
"eval_samples_per_second": 7.815,
|
| 380 |
+
"eval_steps_per_second": 0.313,
|
| 381 |
"step": 54
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 6.11,
|
| 385 |
"learning_rate": 0.0002,
|
| 386 |
+
"loss": 0.0041,
|
| 387 |
"step": 55
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 6.22,
|
| 391 |
"learning_rate": 0.0002,
|
| 392 |
+
"loss": 0.0048,
|
| 393 |
"step": 56
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 6.33,
|
| 397 |
"learning_rate": 0.0002,
|
| 398 |
+
"loss": 0.0053,
|
| 399 |
"step": 57
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 6.44,
|
| 403 |
"learning_rate": 0.0002,
|
| 404 |
+
"loss": 0.0045,
|
| 405 |
"step": 58
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 6.56,
|
| 409 |
"learning_rate": 0.0002,
|
| 410 |
+
"loss": 0.0109,
|
| 411 |
"step": 59
|
| 412 |
},
|
| 413 |
{
|
| 414 |
"epoch": 6.67,
|
| 415 |
"learning_rate": 0.0002,
|
| 416 |
+
"loss": 0.0064,
|
| 417 |
"step": 60
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 6.78,
|
| 421 |
"learning_rate": 0.0002,
|
| 422 |
+
"loss": 0.0056,
|
| 423 |
"step": 61
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 6.89,
|
| 427 |
"learning_rate": 0.0002,
|
| 428 |
+
"loss": 0.0038,
|
| 429 |
"step": 62
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 7.0,
|
| 433 |
"learning_rate": 0.0002,
|
| 434 |
+
"loss": 0.0055,
|
| 435 |
"step": 63
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 7.0,
|
| 439 |
+
"eval_loss": 1.5328795909881592,
|
| 440 |
+
"eval_runtime": 6.3671,
|
| 441 |
+
"eval_samples_per_second": 7.853,
|
| 442 |
+
"eval_steps_per_second": 0.314,
|
| 443 |
"step": 63
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 7.11,
|
| 447 |
"learning_rate": 0.0002,
|
| 448 |
+
"loss": 0.0022,
|
| 449 |
"step": 64
|
| 450 |
},
|
| 451 |
{
|
|
|
|
| 457 |
{
|
| 458 |
"epoch": 7.33,
|
| 459 |
"learning_rate": 0.0002,
|
| 460 |
+
"loss": 0.0089,
|
| 461 |
"step": 66
|
| 462 |
},
|
| 463 |
{
|
| 464 |
"epoch": 7.44,
|
| 465 |
"learning_rate": 0.0002,
|
| 466 |
+
"loss": 0.0057,
|
| 467 |
"step": 67
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 7.56,
|
| 471 |
"learning_rate": 0.0002,
|
| 472 |
+
"loss": 0.0073,
|
| 473 |
"step": 68
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 7.67,
|
| 477 |
"learning_rate": 0.0002,
|
| 478 |
+
"loss": 0.0038,
|
| 479 |
"step": 69
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 7.78,
|
| 483 |
"learning_rate": 0.0002,
|
| 484 |
+
"loss": 0.0137,
|
| 485 |
"step": 70
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 7.89,
|
| 489 |
"learning_rate": 0.0002,
|
| 490 |
+
"loss": 0.0055,
|
| 491 |
"step": 71
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 8.0,
|
| 495 |
"learning_rate": 0.0002,
|
| 496 |
+
"loss": 0.0022,
|
| 497 |
"step": 72
|
| 498 |
},
|
| 499 |
{
|
| 500 |
"epoch": 8.0,
|
| 501 |
+
"eval_loss": 1.534345269203186,
|
| 502 |
+
"eval_runtime": 6.3998,
|
| 503 |
+
"eval_samples_per_second": 7.813,
|
| 504 |
+
"eval_steps_per_second": 0.313,
|
| 505 |
"step": 72
|
| 506 |
},
|
| 507 |
{
|
| 508 |
+
"epoch": 8.11,
|
| 509 |
+
"learning_rate": 0.0002,
|
| 510 |
+
"loss": 0.0016,
|
| 511 |
+
"step": 73
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.22,
|
| 515 |
+
"learning_rate": 0.0002,
|
| 516 |
+
"loss": 0.0041,
|
| 517 |
+
"step": 74
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 8.33,
|
| 521 |
+
"learning_rate": 0.0002,
|
| 522 |
+
"loss": 0.0069,
|
| 523 |
+
"step": 75
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 8.44,
|
| 527 |
+
"learning_rate": 0.0002,
|
| 528 |
+
"loss": 0.0036,
|
| 529 |
+
"step": 76
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 8.56,
|
| 533 |
+
"learning_rate": 0.0002,
|
| 534 |
+
"loss": 0.0043,
|
| 535 |
+
"step": 77
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 8.67,
|
| 539 |
+
"learning_rate": 0.0002,
|
| 540 |
+
"loss": 0.0019,
|
| 541 |
+
"step": 78
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.78,
|
| 545 |
+
"learning_rate": 0.0002,
|
| 546 |
+
"loss": 0.0027,
|
| 547 |
+
"step": 79
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 8.89,
|
| 551 |
+
"learning_rate": 0.0002,
|
| 552 |
+
"loss": 0.0081,
|
| 553 |
+
"step": 80
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"learning_rate": 0.0002,
|
| 558 |
+
"loss": 0.002,
|
| 559 |
+
"step": 81
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 9.0,
|
| 563 |
+
"eval_loss": 1.5295281410217285,
|
| 564 |
+
"eval_runtime": 6.3871,
|
| 565 |
+
"eval_samples_per_second": 7.828,
|
| 566 |
+
"eval_steps_per_second": 0.313,
|
| 567 |
+
"step": 81
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.11,
|
| 571 |
+
"learning_rate": 0.0002,
|
| 572 |
+
"loss": 0.003,
|
| 573 |
+
"step": 82
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 9.22,
|
| 577 |
+
"learning_rate": 0.0002,
|
| 578 |
+
"loss": 0.0029,
|
| 579 |
+
"step": 83
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 9.33,
|
| 583 |
+
"learning_rate": 0.0002,
|
| 584 |
+
"loss": 0.0024,
|
| 585 |
+
"step": 84
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.44,
|
| 589 |
+
"learning_rate": 0.0002,
|
| 590 |
+
"loss": 0.002,
|
| 591 |
+
"step": 85
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 9.56,
|
| 595 |
+
"learning_rate": 0.0002,
|
| 596 |
+
"loss": 0.0028,
|
| 597 |
+
"step": 86
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 9.67,
|
| 601 |
+
"learning_rate": 0.0002,
|
| 602 |
+
"loss": 0.0036,
|
| 603 |
+
"step": 87
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 9.78,
|
| 607 |
+
"learning_rate": 0.0002,
|
| 608 |
+
"loss": 0.0021,
|
| 609 |
+
"step": 88
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 9.89,
|
| 613 |
+
"learning_rate": 0.0002,
|
| 614 |
+
"loss": 0.0033,
|
| 615 |
+
"step": 89
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 10.0,
|
| 619 |
+
"learning_rate": 0.0002,
|
| 620 |
+
"loss": 0.0011,
|
| 621 |
+
"step": 90
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 10.0,
|
| 625 |
+
"eval_loss": 1.6014364957809448,
|
| 626 |
+
"eval_runtime": 6.3819,
|
| 627 |
+
"eval_samples_per_second": 7.835,
|
| 628 |
+
"eval_steps_per_second": 0.313,
|
| 629 |
+
"step": 90
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 10.0,
|
| 633 |
+
"step": 90,
|
| 634 |
+
"total_flos": 1.6878235281679974e+17,
|
| 635 |
+
"train_loss": 0.18350568334701367,
|
| 636 |
+
"train_runtime": 641.931,
|
| 637 |
+
"train_samples_per_second": 4.409,
|
| 638 |
+
"train_steps_per_second": 0.14
|
| 639 |
}
|
| 640 |
],
|
| 641 |
"logging_steps": 1.0,
|
| 642 |
+
"max_steps": 90,
|
| 643 |
"num_input_tokens_seen": 0,
|
| 644 |
+
"num_train_epochs": 10,
|
| 645 |
"save_steps": 50000,
|
| 646 |
+
"total_flos": 1.6878235281679974e+17,
|
| 647 |
"train_batch_size": 4,
|
| 648 |
"trial_name": null,
|
| 649 |
"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 6776
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8ba27cdb68d41d4f248028ec3da3e3cb2d03762a55467bbf3f3c07ea03c0280
|
| 3 |
size 6776
|