End of training
Browse files- README.md +8 -6
- all_results.json +13 -13
- eval_results.json +9 -9
- train_results.json +5 -5
- trainer_state.json +645 -2319
README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-large-patch16-224-in21k
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -18,13 +20,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# vit-large-patch16-224-in21k-finetuned-galaxy10-decals
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [google/vit-large-patch16-224-in21k](https://huggingface.co/google/vit-large-patch16-224-in21k) on
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
-
- Accuracy: 0.
|
| 25 |
-
- Precision: 0.
|
| 26 |
-
- Recall: 0.
|
| 27 |
-
- F1: 0.
|
| 28 |
|
| 29 |
## Model description
|
| 30 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-large-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
+
- vision
|
| 7 |
- generated_from_trainer
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
|
|
|
| 20 |
|
| 21 |
# vit-large-patch16-224-in21k-finetuned-galaxy10-decals
|
| 22 |
|
| 23 |
+
This model is a fine-tuned version of [google/vit-large-patch16-224-in21k](https://huggingface.co/google/vit-large-patch16-224-in21k) on the matthieulel/galaxy10_decals dataset.
|
| 24 |
It achieves the following results on the evaluation set:
|
| 25 |
+
- Loss: 0.6167
|
| 26 |
+
- Accuracy: 0.8455
|
| 27 |
+
- Precision: 0.8422
|
| 28 |
+
- Recall: 0.8455
|
| 29 |
+
- F1: 0.8432
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 29.
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_f1": 0.
|
| 5 |
-
"eval_loss": 0.
|
| 6 |
-
"eval_precision": 0.
|
| 7 |
-
"eval_recall": 0.
|
| 8 |
-
"eval_runtime":
|
| 9 |
-
"eval_samples_per_second":
|
| 10 |
-
"eval_steps_per_second":
|
| 11 |
-
"train_loss": 0.
|
| 12 |
-
"train_runtime":
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second": 0.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"eval_accuracy": 0.8455467869222097,
|
| 4 |
+
"eval_f1": 0.8432046349153324,
|
| 5 |
+
"eval_loss": 0.6166961193084717,
|
| 6 |
+
"eval_precision": 0.8421630294947093,
|
| 7 |
+
"eval_recall": 0.8455467869222097,
|
| 8 |
+
"eval_runtime": 17.7734,
|
| 9 |
+
"eval_samples_per_second": 99.812,
|
| 10 |
+
"eval_steps_per_second": 0.788,
|
| 11 |
+
"train_loss": 0.37854229006716,
|
| 12 |
+
"train_runtime": 10019.4589,
|
| 13 |
+
"train_samples_per_second": 47.793,
|
| 14 |
+
"train_steps_per_second": 0.093
|
| 15 |
}
|
eval_results.json
CHANGED
|
@@ -1,11 +1,11 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 29.
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_f1": 0.
|
| 5 |
-
"eval_loss": 0.
|
| 6 |
-
"eval_precision": 0.
|
| 7 |
-
"eval_recall": 0.
|
| 8 |
-
"eval_runtime":
|
| 9 |
-
"eval_samples_per_second":
|
| 10 |
-
"eval_steps_per_second":
|
| 11 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"eval_accuracy": 0.8455467869222097,
|
| 4 |
+
"eval_f1": 0.8432046349153324,
|
| 5 |
+
"eval_loss": 0.6166961193084717,
|
| 6 |
+
"eval_precision": 0.8421630294947093,
|
| 7 |
+
"eval_recall": 0.8455467869222097,
|
| 8 |
+
"eval_runtime": 17.7734,
|
| 9 |
+
"eval_samples_per_second": 99.812,
|
| 10 |
+
"eval_steps_per_second": 0.788
|
| 11 |
}
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 29.
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 29.76,
|
| 3 |
+
"train_loss": 0.37854229006716,
|
| 4 |
+
"train_runtime": 10019.4589,
|
| 5 |
+
"train_samples_per_second": 47.793,
|
| 6 |
+
"train_steps_per_second": 0.093
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,2622 +1,948 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "vit-large-patch16-224-in21k-finetuned-galaxy10-decals/checkpoint-
|
| 4 |
-
"epoch": 29.
|
| 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.
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 2.
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate":
|
| 20 |
-
"loss":
|
| 21 |
"step": 20
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate":
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 30
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"
|
| 32 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
"step": 40
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch":
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss":
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss":
|
| 45 |
"step": 60
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch":
|
| 49 |
-
"
|
| 50 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
"step": 70
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch":
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 80
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch":
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 90
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch":
|
| 67 |
-
"
|
| 68 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
"step": 100
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 110
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 120
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0
|
| 85 |
-
"eval_accuracy": 0.
|
| 86 |
-
"eval_f1": 0.
|
| 87 |
-
"eval_loss": 0.
|
| 88 |
-
"eval_precision": 0.
|
| 89 |
-
"eval_recall": 0.
|
| 90 |
-
"eval_runtime":
|
| 91 |
-
"eval_samples_per_second":
|
| 92 |
-
"eval_steps_per_second":
|
| 93 |
-
"step":
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 130
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 140
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 150
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"
|
| 116 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
"step": 160
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 170
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 180
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"
|
| 134 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 135 |
"step": 190
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 200
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"learning_rate":
|
| 146 |
-
"loss": 0.
|
| 147 |
"step": 210
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
-
"
|
| 152 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 153 |
"step": 220
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 230
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"learning_rate":
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 240
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"
|
| 170 |
-
"
|
| 171 |
-
"
|
| 172 |
-
"eval_precision": 0.7920273624057349,
|
| 173 |
-
"eval_recall": 0.7852311161217588,
|
| 174 |
-
"eval_runtime": 22.421,
|
| 175 |
-
"eval_samples_per_second": 79.122,
|
| 176 |
-
"eval_steps_per_second": 2.498,
|
| 177 |
-
"step": 249
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch":
|
| 181 |
-
"
|
| 182 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 183 |
"step": 250
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch":
|
| 187 |
-
"learning_rate":
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 260
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
-
"learning_rate": 7.
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 270
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
-
"learning_rate": 7.
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 280
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"
|
| 206 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 207 |
"step": 290
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 300
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 310
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch":
|
| 223 |
-
"
|
| 224 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 225 |
"step": 320
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch":
|
| 229 |
-
"learning_rate":
|
| 230 |
-
"loss": 0.
|
| 231 |
"step": 330
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch":
|
| 235 |
-
"learning_rate":
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 340
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
-
"
|
| 242 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
"step": 350
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
-
"learning_rate":
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 360
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 370
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"eval_accuracy": 0.
|
| 260 |
-
"eval_f1": 0.
|
| 261 |
-
"eval_loss": 0.
|
| 262 |
-
"eval_precision": 0.
|
| 263 |
-
"eval_recall": 0.
|
| 264 |
-
"eval_runtime":
|
| 265 |
-
"eval_samples_per_second":
|
| 266 |
-
"eval_steps_per_second":
|
| 267 |
-
"step":
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
-
"learning_rate":
|
| 272 |
-
"loss": 0.
|
| 273 |
"step": 380
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
-
"learning_rate":
|
| 278 |
-
"loss": 0.
|
| 279 |
"step": 390
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"learning_rate":
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 400
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch":
|
| 289 |
-
"
|
| 290 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 291 |
"step": 410
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 420
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 430
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"
|
| 308 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 309 |
"step": 440
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"learning_rate":
|
| 314 |
-
"loss": 0.
|
| 315 |
"step": 450
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch":
|
| 319 |
-
"learning_rate":
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 460
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
-
"
|
| 326 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 327 |
"step": 470
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch":
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 480
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch":
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 490
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
-
"
|
| 344 |
-
"
|
| 345 |
-
"
|
| 346 |
-
|
| 347 |
-
|
| 348 |
-
"
|
| 349 |
-
"
|
| 350 |
-
"
|
| 351 |
-
"
|
| 352 |
-
|
| 353 |
-
|
| 354 |
-
"
|
| 355 |
-
"
|
| 356 |
-
"
|
| 357 |
"step": 500
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch":
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 510
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch":
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss": 0.
|
| 369 |
"step": 520
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch":
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 530
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch":
|
| 379 |
-
"
|
| 380 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 381 |
"step": 540
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch":
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 550
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch":
|
| 391 |
-
"learning_rate":
|
| 392 |
-
"loss": 0.
|
| 393 |
"step": 560
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch":
|
| 397 |
-
"
|
| 398 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 399 |
"step": 570
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch":
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 580
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 590
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
-
"
|
| 416 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 417 |
"step": 600
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch":
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 610
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 620
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"eval_accuracy": 0.
|
| 434 |
-
"eval_f1": 0.
|
| 435 |
-
"eval_loss": 0.
|
| 436 |
-
"eval_precision": 0.
|
| 437 |
-
"eval_recall": 0.
|
| 438 |
-
"eval_runtime":
|
| 439 |
-
"eval_samples_per_second":
|
| 440 |
-
"eval_steps_per_second":
|
| 441 |
-
"step":
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch":
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 630
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 640
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 650
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"
|
| 464 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 465 |
"step": 660
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 670
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"learning_rate":
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 680
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"
|
| 482 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 483 |
"step": 690
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss": 0.
|
| 489 |
"step": 700
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch":
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 710
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"
|
| 500 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 501 |
"step": 720
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 730
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 740
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
-
"
|
| 518 |
-
"
|
| 519 |
-
"
|
| 520 |
-
"eval_precision": 0.8342187707845001,
|
| 521 |
-
"eval_recall": 0.8376550169109357,
|
| 522 |
-
"eval_runtime": 22.3308,
|
| 523 |
-
"eval_samples_per_second": 79.442,
|
| 524 |
-
"eval_steps_per_second": 2.508,
|
| 525 |
-
"step": 748
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch":
|
| 529 |
-
"
|
| 530 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 531 |
"step": 750
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch":
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 760
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch":
|
| 541 |
-
"learning_rate":
|
| 542 |
-
"loss": 0.
|
| 543 |
"step": 770
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch":
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 780
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
-
"
|
| 554 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 555 |
"step": 790
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 6
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 0.
|
| 561 |
"step": 800
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
-
"learning_rate":
|
| 566 |
-
"loss": 0.
|
| 567 |
"step": 810
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch":
|
| 571 |
-
"
|
| 572 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 573 |
"step": 820
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch":
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 830
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch":
|
| 583 |
-
"learning_rate":
|
| 584 |
-
"loss": 0.
|
| 585 |
"step": 840
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch":
|
| 589 |
-
"
|
| 590 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 591 |
"step": 850
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"learning_rate": 8.
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 860
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"learning_rate":
|
| 602 |
-
"loss": 0.
|
| 603 |
"step": 870
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch":
|
| 607 |
-
"eval_accuracy": 0.
|
| 608 |
-
"eval_f1": 0.
|
| 609 |
-
"eval_loss": 0.
|
| 610 |
-
"eval_precision": 0.
|
| 611 |
-
"eval_recall": 0.
|
| 612 |
-
"eval_runtime":
|
| 613 |
-
"eval_samples_per_second":
|
| 614 |
-
"eval_steps_per_second":
|
| 615 |
-
"step":
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch":
|
| 619 |
-
"learning_rate":
|
| 620 |
-
"loss": 0.
|
| 621 |
"step": 880
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
-
"learning_rate":
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 890
|
| 628 |
},
|
| 629 |
{
|
| 630 |
-
"epoch":
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 900
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"
|
| 638 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 639 |
"step": 910
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch":
|
| 643 |
-
"learning_rate":
|
| 644 |
-
"loss": 0.
|
| 645 |
"step": 920
|
| 646 |
},
|
| 647 |
{
|
| 648 |
-
"epoch":
|
| 649 |
-
"learning_rate":
|
| 650 |
-
"loss": 0.
|
| 651 |
"step": 930
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch":
|
| 655 |
-
"
|
| 656 |
-
"
|
| 657 |
-
"
|
| 658 |
-
|
| 659 |
-
|
| 660 |
-
"
|
| 661 |
-
"
|
| 662 |
-
"
|
| 663 |
-
"step":
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 7.7,
|
| 667 |
-
"learning_rate": 8.243727598566309e-05,
|
| 668 |
-
"loss": 0.4404,
|
| 669 |
-
"step": 960
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 7.78,
|
| 673 |
-
"learning_rate": 8.213859020310634e-05,
|
| 674 |
-
"loss": 0.4221,
|
| 675 |
-
"step": 970
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 7.86,
|
| 679 |
-
"learning_rate": 8.183990442054959e-05,
|
| 680 |
-
"loss": 0.4426,
|
| 681 |
-
"step": 980
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 7.94,
|
| 685 |
-
"learning_rate": 8.154121863799284e-05,
|
| 686 |
-
"loss": 0.4833,
|
| 687 |
-
"step": 990
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 8.0,
|
| 691 |
-
"eval_accuracy": 0.8167981961668546,
|
| 692 |
-
"eval_f1": 0.8170327325832576,
|
| 693 |
-
"eval_loss": 0.5622426271438599,
|
| 694 |
-
"eval_precision": 0.8256514745396676,
|
| 695 |
-
"eval_recall": 0.8167981961668546,
|
| 696 |
-
"eval_runtime": 22.3523,
|
| 697 |
-
"eval_samples_per_second": 79.365,
|
| 698 |
-
"eval_steps_per_second": 2.505,
|
| 699 |
-
"step": 998
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 8.02,
|
| 703 |
-
"learning_rate": 8.124253285543608e-05,
|
| 704 |
-
"loss": 0.4741,
|
| 705 |
-
"step": 1000
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 8.1,
|
| 709 |
-
"learning_rate": 8.094384707287933e-05,
|
| 710 |
-
"loss": 0.4037,
|
| 711 |
-
"step": 1010
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 8.18,
|
| 715 |
-
"learning_rate": 8.064516129032258e-05,
|
| 716 |
-
"loss": 0.4119,
|
| 717 |
-
"step": 1020
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 8.26,
|
| 721 |
-
"learning_rate": 8.034647550776585e-05,
|
| 722 |
-
"loss": 0.4307,
|
| 723 |
-
"step": 1030
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 8.34,
|
| 727 |
-
"learning_rate": 8.004778972520908e-05,
|
| 728 |
-
"loss": 0.3923,
|
| 729 |
-
"step": 1040
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
-
"
|
| 734 |
-
"
|
| 735 |
-
"
|
| 736 |
-
|
| 737 |
-
|
| 738 |
-
"
|
| 739 |
-
"learning_rate": 7.945041816009559e-05,
|
| 740 |
-
"loss": 0.4772,
|
| 741 |
-
"step": 1060
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 8.58,
|
| 745 |
-
"learning_rate": 7.915173237753884e-05,
|
| 746 |
-
"loss": 0.4463,
|
| 747 |
-
"step": 1070
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 8.66,
|
| 751 |
-
"learning_rate": 7.885304659498209e-05,
|
| 752 |
-
"loss": 0.4204,
|
| 753 |
-
"step": 1080
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 8.74,
|
| 757 |
-
"learning_rate": 7.855436081242533e-05,
|
| 758 |
-
"loss": 0.4722,
|
| 759 |
-
"step": 1090
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 8.82,
|
| 763 |
-
"learning_rate": 7.825567502986858e-05,
|
| 764 |
-
"loss": 0.406,
|
| 765 |
-
"step": 1100
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 8.9,
|
| 769 |
-
"learning_rate": 7.795698924731183e-05,
|
| 770 |
-
"loss": 0.4359,
|
| 771 |
-
"step": 1110
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 8.98,
|
| 775 |
-
"learning_rate": 7.765830346475508e-05,
|
| 776 |
-
"loss": 0.4371,
|
| 777 |
-
"step": 1120
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 8.99,
|
| 781 |
-
"eval_accuracy": 0.8348365276211951,
|
| 782 |
-
"eval_f1": 0.8339605863139328,
|
| 783 |
-
"eval_loss": 0.5137929320335388,
|
| 784 |
-
"eval_precision": 0.8387012336946776,
|
| 785 |
-
"eval_recall": 0.8348365276211951,
|
| 786 |
-
"eval_runtime": 22.3049,
|
| 787 |
-
"eval_samples_per_second": 79.534,
|
| 788 |
-
"eval_steps_per_second": 2.511,
|
| 789 |
-
"step": 1122
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 9.06,
|
| 793 |
-
"learning_rate": 7.735961768219832e-05,
|
| 794 |
-
"loss": 0.4301,
|
| 795 |
-
"step": 1130
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 9.14,
|
| 799 |
-
"learning_rate": 7.706093189964157e-05,
|
| 800 |
-
"loss": 0.3969,
|
| 801 |
-
"step": 1140
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 9.22,
|
| 805 |
-
"learning_rate": 7.676224611708484e-05,
|
| 806 |
-
"loss": 0.4139,
|
| 807 |
-
"step": 1150
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 9.3,
|
| 811 |
-
"learning_rate": 7.646356033452809e-05,
|
| 812 |
-
"loss": 0.3784,
|
| 813 |
-
"step": 1160
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 9.38,
|
| 817 |
-
"learning_rate": 7.616487455197133e-05,
|
| 818 |
-
"loss": 0.4135,
|
| 819 |
-
"step": 1170
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 9.46,
|
| 823 |
-
"learning_rate": 7.586618876941458e-05,
|
| 824 |
-
"loss": 0.3923,
|
| 825 |
-
"step": 1180
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 9.54,
|
| 829 |
-
"learning_rate": 7.556750298685783e-05,
|
| 830 |
-
"loss": 0.4297,
|
| 831 |
-
"step": 1190
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 9.62,
|
| 835 |
-
"learning_rate": 7.526881720430108e-05,
|
| 836 |
-
"loss": 0.399,
|
| 837 |
-
"step": 1200
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 9.7,
|
| 841 |
-
"learning_rate": 7.497013142174433e-05,
|
| 842 |
-
"loss": 0.3844,
|
| 843 |
-
"step": 1210
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 9.78,
|
| 847 |
-
"learning_rate": 7.467144563918757e-05,
|
| 848 |
-
"loss": 0.3905,
|
| 849 |
-
"step": 1220
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 9.86,
|
| 853 |
-
"learning_rate": 7.437275985663082e-05,
|
| 854 |
-
"loss": 0.3798,
|
| 855 |
-
"step": 1230
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 9.94,
|
| 859 |
-
"learning_rate": 7.407407407407407e-05,
|
| 860 |
-
"loss": 0.3489,
|
| 861 |
-
"step": 1240
|
| 862 |
-
},
|
| 863 |
-
{
|
| 864 |
-
"epoch": 10.0,
|
| 865 |
-
"eval_accuracy": 0.8325817361894025,
|
| 866 |
-
"eval_f1": 0.8306589317057965,
|
| 867 |
-
"eval_loss": 0.524004340171814,
|
| 868 |
-
"eval_precision": 0.832321948078395,
|
| 869 |
-
"eval_recall": 0.8325817361894025,
|
| 870 |
-
"eval_runtime": 22.3492,
|
| 871 |
-
"eval_samples_per_second": 79.377,
|
| 872 |
-
"eval_steps_per_second": 2.506,
|
| 873 |
-
"step": 1247
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 10.02,
|
| 877 |
-
"learning_rate": 7.377538829151732e-05,
|
| 878 |
-
"loss": 0.4086,
|
| 879 |
-
"step": 1250
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 10.1,
|
| 883 |
-
"learning_rate": 7.347670250896058e-05,
|
| 884 |
-
"loss": 0.3827,
|
| 885 |
-
"step": 1260
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 10.18,
|
| 889 |
-
"learning_rate": 7.317801672640383e-05,
|
| 890 |
-
"loss": 0.3463,
|
| 891 |
-
"step": 1270
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 10.26,
|
| 895 |
-
"learning_rate": 7.287933094384708e-05,
|
| 896 |
-
"loss": 0.3523,
|
| 897 |
-
"step": 1280
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 10.34,
|
| 901 |
-
"learning_rate": 7.258064516129033e-05,
|
| 902 |
-
"loss": 0.4188,
|
| 903 |
-
"step": 1290
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 10.42,
|
| 907 |
-
"learning_rate": 7.228195937873358e-05,
|
| 908 |
-
"loss": 0.3877,
|
| 909 |
-
"step": 1300
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 10.5,
|
| 913 |
-
"learning_rate": 7.198327359617682e-05,
|
| 914 |
-
"loss": 0.3776,
|
| 915 |
-
"step": 1310
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 10.58,
|
| 919 |
-
"learning_rate": 7.168458781362007e-05,
|
| 920 |
-
"loss": 0.3994,
|
| 921 |
-
"step": 1320
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 10.66,
|
| 925 |
-
"learning_rate": 7.138590203106332e-05,
|
| 926 |
-
"loss": 0.3845,
|
| 927 |
-
"step": 1330
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 10.74,
|
| 931 |
-
"learning_rate": 7.108721624850657e-05,
|
| 932 |
-
"loss": 0.3806,
|
| 933 |
-
"step": 1340
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 10.82,
|
| 937 |
-
"learning_rate": 7.078853046594983e-05,
|
| 938 |
-
"loss": 0.3752,
|
| 939 |
-
"step": 1350
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 10.9,
|
| 943 |
-
"learning_rate": 7.048984468339306e-05,
|
| 944 |
-
"loss": 0.3604,
|
| 945 |
-
"step": 1360
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 10.98,
|
| 949 |
-
"learning_rate": 7.019115890083633e-05,
|
| 950 |
-
"loss": 0.3732,
|
| 951 |
-
"step": 1370
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 11.0,
|
| 955 |
-
"eval_accuracy": 0.8291995490417137,
|
| 956 |
-
"eval_f1": 0.8271920748636273,
|
| 957 |
-
"eval_loss": 0.5577002167701721,
|
| 958 |
-
"eval_precision": 0.8300846529332921,
|
| 959 |
-
"eval_recall": 0.8291995490417137,
|
| 960 |
-
"eval_runtime": 22.4478,
|
| 961 |
-
"eval_samples_per_second": 79.028,
|
| 962 |
-
"eval_steps_per_second": 2.495,
|
| 963 |
-
"step": 1372
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 11.06,
|
| 967 |
-
"learning_rate": 6.989247311827958e-05,
|
| 968 |
-
"loss": 0.3541,
|
| 969 |
-
"step": 1380
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 11.14,
|
| 973 |
-
"learning_rate": 6.959378733572283e-05,
|
| 974 |
-
"loss": 0.3494,
|
| 975 |
-
"step": 1390
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 11.22,
|
| 979 |
-
"learning_rate": 6.929510155316607e-05,
|
| 980 |
-
"loss": 0.3505,
|
| 981 |
-
"step": 1400
|
| 982 |
-
},
|
| 983 |
-
{
|
| 984 |
-
"epoch": 11.3,
|
| 985 |
-
"learning_rate": 6.899641577060932e-05,
|
| 986 |
-
"loss": 0.3407,
|
| 987 |
-
"step": 1410
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 11.38,
|
| 991 |
-
"learning_rate": 6.869772998805257e-05,
|
| 992 |
-
"loss": 0.3604,
|
| 993 |
-
"step": 1420
|
| 994 |
-
},
|
| 995 |
-
{
|
| 996 |
-
"epoch": 11.46,
|
| 997 |
-
"learning_rate": 6.839904420549582e-05,
|
| 998 |
-
"loss": 0.3551,
|
| 999 |
-
"step": 1430
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 11.54,
|
| 1003 |
-
"learning_rate": 6.810035842293908e-05,
|
| 1004 |
-
"loss": 0.3581,
|
| 1005 |
-
"step": 1440
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 11.62,
|
| 1009 |
-
"learning_rate": 6.780167264038231e-05,
|
| 1010 |
-
"loss": 0.3377,
|
| 1011 |
-
"step": 1450
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 11.7,
|
| 1015 |
-
"learning_rate": 6.750298685782557e-05,
|
| 1016 |
-
"loss": 0.3816,
|
| 1017 |
-
"step": 1460
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 11.78,
|
| 1021 |
-
"learning_rate": 6.720430107526882e-05,
|
| 1022 |
-
"loss": 0.3666,
|
| 1023 |
-
"step": 1470
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 11.86,
|
| 1027 |
-
"learning_rate": 6.690561529271207e-05,
|
| 1028 |
-
"loss": 0.3493,
|
| 1029 |
-
"step": 1480
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 11.94,
|
| 1033 |
-
"learning_rate": 6.660692951015532e-05,
|
| 1034 |
-
"loss": 0.3226,
|
| 1035 |
-
"step": 1490
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 12.0,
|
| 1039 |
-
"eval_accuracy": 0.826381059751973,
|
| 1040 |
-
"eval_f1": 0.8228837160691321,
|
| 1041 |
-
"eval_loss": 0.5933008193969727,
|
| 1042 |
-
"eval_precision": 0.8214741549096344,
|
| 1043 |
-
"eval_recall": 0.826381059751973,
|
| 1044 |
-
"eval_runtime": 22.3433,
|
| 1045 |
-
"eval_samples_per_second": 79.397,
|
| 1046 |
-
"eval_steps_per_second": 2.506,
|
| 1047 |
-
"step": 1497
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 12.02,
|
| 1051 |
-
"learning_rate": 6.630824372759857e-05,
|
| 1052 |
-
"loss": 0.3357,
|
| 1053 |
-
"step": 1500
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 12.1,
|
| 1057 |
-
"learning_rate": 6.600955794504182e-05,
|
| 1058 |
-
"loss": 0.3632,
|
| 1059 |
-
"step": 1510
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 12.18,
|
| 1063 |
-
"learning_rate": 6.571087216248507e-05,
|
| 1064 |
-
"loss": 0.3133,
|
| 1065 |
-
"step": 1520
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 12.26,
|
| 1069 |
-
"learning_rate": 6.541218637992833e-05,
|
| 1070 |
-
"loss": 0.316,
|
| 1071 |
-
"step": 1530
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 12.34,
|
| 1075 |
-
"learning_rate": 6.511350059737156e-05,
|
| 1076 |
-
"loss": 0.3276,
|
| 1077 |
-
"step": 1540
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 12.42,
|
| 1081 |
-
"learning_rate": 6.481481481481482e-05,
|
| 1082 |
-
"loss": 0.3606,
|
| 1083 |
-
"step": 1550
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 12.51,
|
| 1087 |
-
"learning_rate": 6.451612903225807e-05,
|
| 1088 |
-
"loss": 0.2919,
|
| 1089 |
-
"step": 1560
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 12.59,
|
| 1093 |
-
"learning_rate": 6.421744324970132e-05,
|
| 1094 |
-
"loss": 0.3633,
|
| 1095 |
-
"step": 1570
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 12.67,
|
| 1099 |
-
"learning_rate": 6.391875746714456e-05,
|
| 1100 |
-
"loss": 0.3687,
|
| 1101 |
-
"step": 1580
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 12.75,
|
| 1105 |
-
"learning_rate": 6.362007168458781e-05,
|
| 1106 |
-
"loss": 0.3241,
|
| 1107 |
-
"step": 1590
|
| 1108 |
-
},
|
| 1109 |
-
{
|
| 1110 |
-
"epoch": 12.83,
|
| 1111 |
-
"learning_rate": 6.332138590203107e-05,
|
| 1112 |
-
"loss": 0.3542,
|
| 1113 |
-
"step": 1600
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 12.91,
|
| 1117 |
-
"learning_rate": 6.302270011947432e-05,
|
| 1118 |
-
"loss": 0.3242,
|
| 1119 |
-
"step": 1610
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 12.99,
|
| 1123 |
-
"learning_rate": 6.272401433691756e-05,
|
| 1124 |
-
"loss": 0.3324,
|
| 1125 |
-
"step": 1620
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 12.99,
|
| 1129 |
-
"eval_accuracy": 0.8342728297632469,
|
| 1130 |
-
"eval_f1": 0.8346366092452495,
|
| 1131 |
-
"eval_loss": 0.5755845308303833,
|
| 1132 |
-
"eval_precision": 0.8404500158146665,
|
| 1133 |
-
"eval_recall": 0.8342728297632469,
|
| 1134 |
-
"eval_runtime": 22.4039,
|
| 1135 |
-
"eval_samples_per_second": 79.183,
|
| 1136 |
-
"eval_steps_per_second": 2.5,
|
| 1137 |
-
"step": 1621
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 13.07,
|
| 1141 |
-
"learning_rate": 6.242532855436081e-05,
|
| 1142 |
-
"loss": 0.3174,
|
| 1143 |
-
"step": 1630
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 13.15,
|
| 1147 |
-
"learning_rate": 6.212664277180407e-05,
|
| 1148 |
-
"loss": 0.2962,
|
| 1149 |
-
"step": 1640
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 13.23,
|
| 1153 |
-
"learning_rate": 6.182795698924732e-05,
|
| 1154 |
-
"loss": 0.2916,
|
| 1155 |
-
"step": 1650
|
| 1156 |
-
},
|
| 1157 |
-
{
|
| 1158 |
-
"epoch": 13.31,
|
| 1159 |
-
"learning_rate": 6.152927120669057e-05,
|
| 1160 |
-
"loss": 0.3415,
|
| 1161 |
-
"step": 1660
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 13.39,
|
| 1165 |
-
"learning_rate": 6.12305854241338e-05,
|
| 1166 |
-
"loss": 0.3441,
|
| 1167 |
-
"step": 1670
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 13.47,
|
| 1171 |
-
"learning_rate": 6.093189964157706e-05,
|
| 1172 |
-
"loss": 0.3185,
|
| 1173 |
-
"step": 1680
|
| 1174 |
-
},
|
| 1175 |
-
{
|
| 1176 |
-
"epoch": 13.55,
|
| 1177 |
-
"learning_rate": 6.063321385902031e-05,
|
| 1178 |
-
"loss": 0.3167,
|
| 1179 |
-
"step": 1690
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 13.63,
|
| 1183 |
-
"learning_rate": 6.033452807646356e-05,
|
| 1184 |
-
"loss": 0.3053,
|
| 1185 |
-
"step": 1700
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 13.71,
|
| 1189 |
-
"learning_rate": 6.0035842293906806e-05,
|
| 1190 |
-
"loss": 0.3156,
|
| 1191 |
-
"step": 1710
|
| 1192 |
-
},
|
| 1193 |
-
{
|
| 1194 |
-
"epoch": 13.79,
|
| 1195 |
-
"learning_rate": 5.9737156511350064e-05,
|
| 1196 |
-
"loss": 0.3179,
|
| 1197 |
-
"step": 1720
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 13.87,
|
| 1201 |
-
"learning_rate": 5.9438470728793316e-05,
|
| 1202 |
-
"loss": 0.3221,
|
| 1203 |
-
"step": 1730
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 13.95,
|
| 1207 |
-
"learning_rate": 5.913978494623657e-05,
|
| 1208 |
-
"loss": 0.3255,
|
| 1209 |
-
"step": 1740
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 14.0,
|
| 1213 |
-
"eval_accuracy": 0.8314543404735062,
|
| 1214 |
-
"eval_f1": 0.8294315197830872,
|
| 1215 |
-
"eval_loss": 0.5854434370994568,
|
| 1216 |
-
"eval_precision": 0.832619836621521,
|
| 1217 |
-
"eval_recall": 0.8314543404735062,
|
| 1218 |
-
"eval_runtime": 22.2776,
|
| 1219 |
-
"eval_samples_per_second": 79.631,
|
| 1220 |
-
"eval_steps_per_second": 2.514,
|
| 1221 |
-
"step": 1746
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 14.03,
|
| 1225 |
-
"learning_rate": 5.884109916367981e-05,
|
| 1226 |
-
"loss": 0.2823,
|
| 1227 |
-
"step": 1750
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 14.11,
|
| 1231 |
-
"learning_rate": 5.8542413381123063e-05,
|
| 1232 |
-
"loss": 0.2985,
|
| 1233 |
-
"step": 1760
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 14.19,
|
| 1237 |
-
"learning_rate": 5.824372759856631e-05,
|
| 1238 |
-
"loss": 0.2966,
|
| 1239 |
-
"step": 1770
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 14.27,
|
| 1243 |
-
"learning_rate": 5.794504181600956e-05,
|
| 1244 |
-
"loss": 0.3052,
|
| 1245 |
-
"step": 1780
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 14.35,
|
| 1249 |
-
"learning_rate": 5.764635603345281e-05,
|
| 1250 |
-
"loss": 0.2914,
|
| 1251 |
-
"step": 1790
|
| 1252 |
-
},
|
| 1253 |
-
{
|
| 1254 |
-
"epoch": 14.43,
|
| 1255 |
-
"learning_rate": 5.7347670250896056e-05,
|
| 1256 |
-
"loss": 0.3019,
|
| 1257 |
-
"step": 1800
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 14.51,
|
| 1261 |
-
"learning_rate": 5.704898446833931e-05,
|
| 1262 |
-
"loss": 0.338,
|
| 1263 |
-
"step": 1810
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 14.59,
|
| 1267 |
-
"learning_rate": 5.675029868578255e-05,
|
| 1268 |
-
"loss": 0.3303,
|
| 1269 |
-
"step": 1820
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 14.67,
|
| 1273 |
-
"learning_rate": 5.645161290322582e-05,
|
| 1274 |
-
"loss": 0.2695,
|
| 1275 |
-
"step": 1830
|
| 1276 |
-
},
|
| 1277 |
-
{
|
| 1278 |
-
"epoch": 14.75,
|
| 1279 |
-
"learning_rate": 5.615292712066906e-05,
|
| 1280 |
-
"loss": 0.2996,
|
| 1281 |
-
"step": 1840
|
| 1282 |
-
},
|
| 1283 |
-
{
|
| 1284 |
-
"epoch": 14.83,
|
| 1285 |
-
"learning_rate": 5.5854241338112313e-05,
|
| 1286 |
-
"loss": 0.3115,
|
| 1287 |
-
"step": 1850
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 14.91,
|
| 1291 |
-
"learning_rate": 5.555555555555556e-05,
|
| 1292 |
-
"loss": 0.2974,
|
| 1293 |
-
"step": 1860
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 14.99,
|
| 1297 |
-
"learning_rate": 5.525686977299881e-05,
|
| 1298 |
-
"loss": 0.3083,
|
| 1299 |
-
"step": 1870
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 15.0,
|
| 1303 |
-
"eval_accuracy": 0.8410372040586246,
|
| 1304 |
-
"eval_f1": 0.838198074050418,
|
| 1305 |
-
"eval_loss": 0.5341351628303528,
|
| 1306 |
-
"eval_precision": 0.8383772092535509,
|
| 1307 |
-
"eval_recall": 0.8410372040586246,
|
| 1308 |
-
"eval_runtime": 22.2997,
|
| 1309 |
-
"eval_samples_per_second": 79.553,
|
| 1310 |
-
"eval_steps_per_second": 2.511,
|
| 1311 |
-
"step": 1871
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 15.07,
|
| 1315 |
-
"learning_rate": 5.495818399044206e-05,
|
| 1316 |
-
"loss": 0.2895,
|
| 1317 |
-
"step": 1880
|
| 1318 |
-
},
|
| 1319 |
-
{
|
| 1320 |
-
"epoch": 15.15,
|
| 1321 |
-
"learning_rate": 5.4659498207885306e-05,
|
| 1322 |
-
"loss": 0.2776,
|
| 1323 |
-
"step": 1890
|
| 1324 |
-
},
|
| 1325 |
-
{
|
| 1326 |
-
"epoch": 15.23,
|
| 1327 |
-
"learning_rate": 5.436081242532856e-05,
|
| 1328 |
-
"loss": 0.292,
|
| 1329 |
-
"step": 1900
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 15.31,
|
| 1333 |
-
"learning_rate": 5.40621266427718e-05,
|
| 1334 |
-
"loss": 0.2776,
|
| 1335 |
-
"step": 1910
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 15.39,
|
| 1339 |
-
"learning_rate": 5.3763440860215054e-05,
|
| 1340 |
-
"loss": 0.2689,
|
| 1341 |
-
"step": 1920
|
| 1342 |
-
},
|
| 1343 |
-
{
|
| 1344 |
-
"epoch": 15.47,
|
| 1345 |
-
"learning_rate": 5.34647550776583e-05,
|
| 1346 |
-
"loss": 0.3141,
|
| 1347 |
-
"step": 1930
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 15.55,
|
| 1351 |
-
"learning_rate": 5.316606929510155e-05,
|
| 1352 |
-
"loss": 0.2717,
|
| 1353 |
-
"step": 1940
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 15.63,
|
| 1357 |
-
"learning_rate": 5.286738351254481e-05,
|
| 1358 |
-
"loss": 0.267,
|
| 1359 |
-
"step": 1950
|
| 1360 |
-
},
|
| 1361 |
-
{
|
| 1362 |
-
"epoch": 15.71,
|
| 1363 |
-
"learning_rate": 5.256869772998806e-05,
|
| 1364 |
-
"loss": 0.2692,
|
| 1365 |
-
"step": 1960
|
| 1366 |
-
},
|
| 1367 |
-
{
|
| 1368 |
-
"epoch": 15.79,
|
| 1369 |
-
"learning_rate": 5.2270011947431304e-05,
|
| 1370 |
-
"loss": 0.2697,
|
| 1371 |
-
"step": 1970
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 15.87,
|
| 1375 |
-
"learning_rate": 5.1971326164874556e-05,
|
| 1376 |
-
"loss": 0.2845,
|
| 1377 |
-
"step": 1980
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 15.95,
|
| 1381 |
-
"learning_rate": 5.167264038231781e-05,
|
| 1382 |
-
"loss": 0.2967,
|
| 1383 |
-
"step": 1990
|
| 1384 |
-
},
|
| 1385 |
-
{
|
| 1386 |
-
"epoch": 16.0,
|
| 1387 |
-
"eval_accuracy": 0.826381059751973,
|
| 1388 |
-
"eval_f1": 0.825177966989534,
|
| 1389 |
-
"eval_loss": 0.5956478118896484,
|
| 1390 |
-
"eval_precision": 0.8267605731966666,
|
| 1391 |
-
"eval_recall": 0.826381059751973,
|
| 1392 |
-
"eval_runtime": 22.3228,
|
| 1393 |
-
"eval_samples_per_second": 79.47,
|
| 1394 |
-
"eval_steps_per_second": 2.509,
|
| 1395 |
-
"step": 1996
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 16.03,
|
| 1399 |
-
"learning_rate": 5.137395459976105e-05,
|
| 1400 |
-
"loss": 0.2905,
|
| 1401 |
-
"step": 2000
|
| 1402 |
-
},
|
| 1403 |
-
{
|
| 1404 |
-
"epoch": 16.11,
|
| 1405 |
-
"learning_rate": 5.1075268817204304e-05,
|
| 1406 |
-
"loss": 0.281,
|
| 1407 |
-
"step": 2010
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 16.19,
|
| 1411 |
-
"learning_rate": 5.077658303464755e-05,
|
| 1412 |
-
"loss": 0.2657,
|
| 1413 |
-
"step": 2020
|
| 1414 |
-
},
|
| 1415 |
-
{
|
| 1416 |
-
"epoch": 16.27,
|
| 1417 |
-
"learning_rate": 5.04778972520908e-05,
|
| 1418 |
-
"loss": 0.2548,
|
| 1419 |
-
"step": 2030
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 16.35,
|
| 1423 |
-
"learning_rate": 5.017921146953405e-05,
|
| 1424 |
-
"loss": 0.2586,
|
| 1425 |
-
"step": 2040
|
| 1426 |
-
},
|
| 1427 |
-
{
|
| 1428 |
-
"epoch": 16.43,
|
| 1429 |
-
"learning_rate": 4.98805256869773e-05,
|
| 1430 |
-
"loss": 0.2401,
|
| 1431 |
-
"step": 2050
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 16.51,
|
| 1435 |
-
"learning_rate": 4.9581839904420555e-05,
|
| 1436 |
-
"loss": 0.2839,
|
| 1437 |
-
"step": 2060
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 16.59,
|
| 1441 |
-
"learning_rate": 4.92831541218638e-05,
|
| 1442 |
-
"loss": 0.3153,
|
| 1443 |
-
"step": 2070
|
| 1444 |
-
},
|
| 1445 |
-
{
|
| 1446 |
-
"epoch": 16.67,
|
| 1447 |
-
"learning_rate": 4.898446833930705e-05,
|
| 1448 |
-
"loss": 0.2806,
|
| 1449 |
-
"step": 2080
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 16.75,
|
| 1453 |
-
"learning_rate": 4.86857825567503e-05,
|
| 1454 |
-
"loss": 0.2824,
|
| 1455 |
-
"step": 2090
|
| 1456 |
-
},
|
| 1457 |
-
{
|
| 1458 |
-
"epoch": 16.83,
|
| 1459 |
-
"learning_rate": 4.8387096774193554e-05,
|
| 1460 |
-
"loss": 0.2502,
|
| 1461 |
-
"step": 2100
|
| 1462 |
-
},
|
| 1463 |
-
{
|
| 1464 |
-
"epoch": 16.91,
|
| 1465 |
-
"learning_rate": 4.80884109916368e-05,
|
| 1466 |
-
"loss": 0.2224,
|
| 1467 |
-
"step": 2110
|
| 1468 |
-
},
|
| 1469 |
-
{
|
| 1470 |
-
"epoch": 16.99,
|
| 1471 |
-
"learning_rate": 4.778972520908005e-05,
|
| 1472 |
-
"loss": 0.2613,
|
| 1473 |
-
"step": 2120
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 16.99,
|
| 1477 |
-
"eval_accuracy": 0.8314543404735062,
|
| 1478 |
-
"eval_f1": 0.8282517076467648,
|
| 1479 |
-
"eval_loss": 0.635787308216095,
|
| 1480 |
-
"eval_precision": 0.8336110282521515,
|
| 1481 |
-
"eval_recall": 0.8314543404735062,
|
| 1482 |
-
"eval_runtime": 22.3672,
|
| 1483 |
-
"eval_samples_per_second": 79.313,
|
| 1484 |
-
"eval_steps_per_second": 2.504,
|
| 1485 |
-
"step": 2120
|
| 1486 |
-
},
|
| 1487 |
-
{
|
| 1488 |
-
"epoch": 17.07,
|
| 1489 |
-
"learning_rate": 4.74910394265233e-05,
|
| 1490 |
-
"loss": 0.2752,
|
| 1491 |
-
"step": 2130
|
| 1492 |
-
},
|
| 1493 |
-
{
|
| 1494 |
-
"epoch": 17.15,
|
| 1495 |
-
"learning_rate": 4.7192353643966546e-05,
|
| 1496 |
-
"loss": 0.2435,
|
| 1497 |
-
"step": 2140
|
| 1498 |
-
},
|
| 1499 |
-
{
|
| 1500 |
-
"epoch": 17.23,
|
| 1501 |
-
"learning_rate": 4.6893667861409805e-05,
|
| 1502 |
-
"loss": 0.2905,
|
| 1503 |
-
"step": 2150
|
| 1504 |
-
},
|
| 1505 |
-
{
|
| 1506 |
-
"epoch": 17.31,
|
| 1507 |
-
"learning_rate": 4.659498207885305e-05,
|
| 1508 |
-
"loss": 0.272,
|
| 1509 |
-
"step": 2160
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 17.39,
|
| 1513 |
-
"learning_rate": 4.62962962962963e-05,
|
| 1514 |
-
"loss": 0.2832,
|
| 1515 |
-
"step": 2170
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 17.47,
|
| 1519 |
-
"learning_rate": 4.5997610513739546e-05,
|
| 1520 |
-
"loss": 0.2636,
|
| 1521 |
-
"step": 2180
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 17.56,
|
| 1525 |
-
"learning_rate": 4.56989247311828e-05,
|
| 1526 |
-
"loss": 0.2889,
|
| 1527 |
-
"step": 2190
|
| 1528 |
-
},
|
| 1529 |
-
{
|
| 1530 |
-
"epoch": 17.64,
|
| 1531 |
-
"learning_rate": 4.540023894862604e-05,
|
| 1532 |
-
"loss": 0.2393,
|
| 1533 |
-
"step": 2200
|
| 1534 |
-
},
|
| 1535 |
-
{
|
| 1536 |
-
"epoch": 17.72,
|
| 1537 |
-
"learning_rate": 4.51015531660693e-05,
|
| 1538 |
-
"loss": 0.249,
|
| 1539 |
-
"step": 2210
|
| 1540 |
-
},
|
| 1541 |
-
{
|
| 1542 |
-
"epoch": 17.8,
|
| 1543 |
-
"learning_rate": 4.4802867383512545e-05,
|
| 1544 |
-
"loss": 0.2431,
|
| 1545 |
-
"step": 2220
|
| 1546 |
-
},
|
| 1547 |
-
{
|
| 1548 |
-
"epoch": 17.88,
|
| 1549 |
-
"learning_rate": 4.4504181600955796e-05,
|
| 1550 |
-
"loss": 0.2559,
|
| 1551 |
-
"step": 2230
|
| 1552 |
-
},
|
| 1553 |
-
{
|
| 1554 |
-
"epoch": 17.96,
|
| 1555 |
-
"learning_rate": 4.420549581839905e-05,
|
| 1556 |
-
"loss": 0.2544,
|
| 1557 |
-
"step": 2240
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 18.0,
|
| 1561 |
-
"eval_accuracy": 0.8291995490417137,
|
| 1562 |
-
"eval_f1": 0.827361466855668,
|
| 1563 |
-
"eval_loss": 0.6696366667747498,
|
| 1564 |
-
"eval_precision": 0.8286890088274868,
|
| 1565 |
-
"eval_recall": 0.8291995490417137,
|
| 1566 |
-
"eval_runtime": 22.2886,
|
| 1567 |
-
"eval_samples_per_second": 79.592,
|
| 1568 |
-
"eval_steps_per_second": 2.512,
|
| 1569 |
-
"step": 2245
|
| 1570 |
-
},
|
| 1571 |
-
{
|
| 1572 |
-
"epoch": 18.04,
|
| 1573 |
-
"learning_rate": 4.390681003584229e-05,
|
| 1574 |
-
"loss": 0.2831,
|
| 1575 |
-
"step": 2250
|
| 1576 |
-
},
|
| 1577 |
-
{
|
| 1578 |
-
"epoch": 18.12,
|
| 1579 |
-
"learning_rate": 4.360812425328555e-05,
|
| 1580 |
-
"loss": 0.2324,
|
| 1581 |
-
"step": 2260
|
| 1582 |
-
},
|
| 1583 |
-
{
|
| 1584 |
-
"epoch": 18.2,
|
| 1585 |
-
"learning_rate": 4.3309438470728796e-05,
|
| 1586 |
-
"loss": 0.2137,
|
| 1587 |
-
"step": 2270
|
| 1588 |
-
},
|
| 1589 |
-
{
|
| 1590 |
-
"epoch": 18.28,
|
| 1591 |
-
"learning_rate": 4.301075268817205e-05,
|
| 1592 |
-
"loss": 0.247,
|
| 1593 |
-
"step": 2280
|
| 1594 |
-
},
|
| 1595 |
-
{
|
| 1596 |
-
"epoch": 18.36,
|
| 1597 |
-
"learning_rate": 4.271206690561529e-05,
|
| 1598 |
-
"loss": 0.2379,
|
| 1599 |
-
"step": 2290
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 18.44,
|
| 1603 |
-
"learning_rate": 4.241338112305854e-05,
|
| 1604 |
-
"loss": 0.2134,
|
| 1605 |
-
"step": 2300
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 18.52,
|
| 1609 |
-
"learning_rate": 4.2114695340501795e-05,
|
| 1610 |
-
"loss": 0.2371,
|
| 1611 |
-
"step": 2310
|
| 1612 |
-
},
|
| 1613 |
-
{
|
| 1614 |
-
"epoch": 18.6,
|
| 1615 |
-
"learning_rate": 4.1816009557945046e-05,
|
| 1616 |
-
"loss": 0.2537,
|
| 1617 |
-
"step": 2320
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 18.68,
|
| 1621 |
-
"learning_rate": 4.15173237753883e-05,
|
| 1622 |
-
"loss": 0.2364,
|
| 1623 |
-
"step": 2330
|
| 1624 |
-
},
|
| 1625 |
-
{
|
| 1626 |
-
"epoch": 18.76,
|
| 1627 |
-
"learning_rate": 4.121863799283154e-05,
|
| 1628 |
-
"loss": 0.2097,
|
| 1629 |
-
"step": 2340
|
| 1630 |
-
},
|
| 1631 |
-
{
|
| 1632 |
-
"epoch": 18.84,
|
| 1633 |
-
"learning_rate": 4.0919952210274794e-05,
|
| 1634 |
-
"loss": 0.2263,
|
| 1635 |
-
"step": 2350
|
| 1636 |
-
},
|
| 1637 |
-
{
|
| 1638 |
-
"epoch": 18.92,
|
| 1639 |
-
"learning_rate": 4.062126642771804e-05,
|
| 1640 |
-
"loss": 0.227,
|
| 1641 |
-
"step": 2360
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 19.0,
|
| 1645 |
-
"learning_rate": 4.032258064516129e-05,
|
| 1646 |
-
"loss": 0.2829,
|
| 1647 |
-
"step": 2370
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 19.0,
|
| 1651 |
-
"eval_accuracy": 0.830890642615558,
|
| 1652 |
-
"eval_f1": 0.8316074924684721,
|
| 1653 |
-
"eval_loss": 0.6282095909118652,
|
| 1654 |
-
"eval_precision": 0.8356856994643999,
|
| 1655 |
-
"eval_recall": 0.830890642615558,
|
| 1656 |
-
"eval_runtime": 22.3661,
|
| 1657 |
-
"eval_samples_per_second": 79.316,
|
| 1658 |
-
"eval_steps_per_second": 2.504,
|
| 1659 |
-
"step": 2370
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 19.08,
|
| 1663 |
-
"learning_rate": 4.002389486260454e-05,
|
| 1664 |
-
"loss": 0.2362,
|
| 1665 |
-
"step": 2380
|
| 1666 |
-
},
|
| 1667 |
-
{
|
| 1668 |
-
"epoch": 19.16,
|
| 1669 |
-
"learning_rate": 3.972520908004779e-05,
|
| 1670 |
-
"loss": 0.2488,
|
| 1671 |
-
"step": 2390
|
| 1672 |
-
},
|
| 1673 |
-
{
|
| 1674 |
-
"epoch": 19.24,
|
| 1675 |
-
"learning_rate": 3.9426523297491045e-05,
|
| 1676 |
-
"loss": 0.2132,
|
| 1677 |
-
"step": 2400
|
| 1678 |
-
},
|
| 1679 |
-
{
|
| 1680 |
-
"epoch": 19.32,
|
| 1681 |
-
"learning_rate": 3.912783751493429e-05,
|
| 1682 |
-
"loss": 0.2237,
|
| 1683 |
-
"step": 2410
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 19.4,
|
| 1687 |
-
"learning_rate": 3.882915173237754e-05,
|
| 1688 |
-
"loss": 0.2559,
|
| 1689 |
-
"step": 2420
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 19.48,
|
| 1693 |
-
"learning_rate": 3.8530465949820786e-05,
|
| 1694 |
-
"loss": 0.264,
|
| 1695 |
-
"step": 2430
|
| 1696 |
-
},
|
| 1697 |
-
{
|
| 1698 |
-
"epoch": 19.56,
|
| 1699 |
-
"learning_rate": 3.8231780167264044e-05,
|
| 1700 |
-
"loss": 0.1993,
|
| 1701 |
-
"step": 2440
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 19.64,
|
| 1705 |
-
"learning_rate": 3.793309438470729e-05,
|
| 1706 |
-
"loss": 0.2292,
|
| 1707 |
-
"step": 2450
|
| 1708 |
-
},
|
| 1709 |
-
{
|
| 1710 |
-
"epoch": 19.72,
|
| 1711 |
-
"learning_rate": 3.763440860215054e-05,
|
| 1712 |
-
"loss": 0.2444,
|
| 1713 |
-
"step": 2460
|
| 1714 |
-
},
|
| 1715 |
-
{
|
| 1716 |
-
"epoch": 19.8,
|
| 1717 |
-
"learning_rate": 3.7335722819593785e-05,
|
| 1718 |
-
"loss": 0.2228,
|
| 1719 |
-
"step": 2470
|
| 1720 |
-
},
|
| 1721 |
-
{
|
| 1722 |
-
"epoch": 19.88,
|
| 1723 |
-
"learning_rate": 3.7037037037037037e-05,
|
| 1724 |
-
"loss": 0.2329,
|
| 1725 |
-
"step": 2480
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 19.96,
|
| 1729 |
-
"learning_rate": 3.673835125448029e-05,
|
| 1730 |
-
"loss": 0.2241,
|
| 1731 |
-
"step": 2490
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 20.0,
|
| 1735 |
-
"eval_accuracy": 0.8291995490417137,
|
| 1736 |
-
"eval_f1": 0.8258985946222193,
|
| 1737 |
-
"eval_loss": 0.6849871277809143,
|
| 1738 |
-
"eval_precision": 0.8253536027438152,
|
| 1739 |
-
"eval_recall": 0.8291995490417137,
|
| 1740 |
-
"eval_runtime": 22.299,
|
| 1741 |
-
"eval_samples_per_second": 79.555,
|
| 1742 |
-
"eval_steps_per_second": 2.511,
|
| 1743 |
-
"step": 2495
|
| 1744 |
-
},
|
| 1745 |
-
{
|
| 1746 |
-
"epoch": 20.04,
|
| 1747 |
-
"learning_rate": 3.643966547192354e-05,
|
| 1748 |
-
"loss": 0.2276,
|
| 1749 |
-
"step": 2500
|
| 1750 |
-
},
|
| 1751 |
-
{
|
| 1752 |
-
"epoch": 20.12,
|
| 1753 |
-
"learning_rate": 3.614097968936679e-05,
|
| 1754 |
-
"loss": 0.1957,
|
| 1755 |
-
"step": 2510
|
| 1756 |
-
},
|
| 1757 |
-
{
|
| 1758 |
-
"epoch": 20.2,
|
| 1759 |
-
"learning_rate": 3.5842293906810036e-05,
|
| 1760 |
-
"loss": 0.229,
|
| 1761 |
-
"step": 2520
|
| 1762 |
-
},
|
| 1763 |
-
{
|
| 1764 |
-
"epoch": 20.28,
|
| 1765 |
-
"learning_rate": 3.554360812425329e-05,
|
| 1766 |
-
"loss": 0.2424,
|
| 1767 |
-
"step": 2530
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 20.36,
|
| 1771 |
-
"learning_rate": 3.524492234169653e-05,
|
| 1772 |
-
"loss": 0.2265,
|
| 1773 |
-
"step": 2540
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 20.44,
|
| 1777 |
-
"learning_rate": 3.494623655913979e-05,
|
| 1778 |
-
"loss": 0.1893,
|
| 1779 |
-
"step": 2550
|
| 1780 |
-
},
|
| 1781 |
-
{
|
| 1782 |
-
"epoch": 20.52,
|
| 1783 |
-
"learning_rate": 3.4647550776583035e-05,
|
| 1784 |
-
"loss": 0.2186,
|
| 1785 |
-
"step": 2560
|
| 1786 |
-
},
|
| 1787 |
-
{
|
| 1788 |
-
"epoch": 20.6,
|
| 1789 |
-
"learning_rate": 3.4348864994026287e-05,
|
| 1790 |
-
"loss": 0.2116,
|
| 1791 |
-
"step": 2570
|
| 1792 |
-
},
|
| 1793 |
-
{
|
| 1794 |
-
"epoch": 20.68,
|
| 1795 |
-
"learning_rate": 3.405017921146954e-05,
|
| 1796 |
-
"loss": 0.2059,
|
| 1797 |
-
"step": 2580
|
| 1798 |
-
},
|
| 1799 |
-
{
|
| 1800 |
-
"epoch": 20.76,
|
| 1801 |
-
"learning_rate": 3.375149342891278e-05,
|
| 1802 |
-
"loss": 0.2522,
|
| 1803 |
-
"step": 2590
|
| 1804 |
-
},
|
| 1805 |
-
{
|
| 1806 |
-
"epoch": 20.84,
|
| 1807 |
-
"learning_rate": 3.3452807646356034e-05,
|
| 1808 |
-
"loss": 0.2517,
|
| 1809 |
-
"step": 2600
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 20.92,
|
| 1813 |
-
"learning_rate": 3.3154121863799286e-05,
|
| 1814 |
-
"loss": 0.214,
|
| 1815 |
-
"step": 2610
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 20.99,
|
| 1819 |
-
"eval_accuracy": 0.8387824126268321,
|
| 1820 |
-
"eval_f1": 0.833896519330686,
|
| 1821 |
-
"eval_loss": 0.6962023377418518,
|
| 1822 |
-
"eval_precision": 0.8376423558431443,
|
| 1823 |
-
"eval_recall": 0.8387824126268321,
|
| 1824 |
-
"eval_runtime": 22.3087,
|
| 1825 |
-
"eval_samples_per_second": 79.521,
|
| 1826 |
-
"eval_steps_per_second": 2.51,
|
| 1827 |
-
"step": 2619
|
| 1828 |
-
},
|
| 1829 |
-
{
|
| 1830 |
-
"epoch": 21.0,
|
| 1831 |
-
"learning_rate": 3.285543608124254e-05,
|
| 1832 |
-
"loss": 0.2014,
|
| 1833 |
-
"step": 2620
|
| 1834 |
-
},
|
| 1835 |
-
{
|
| 1836 |
-
"epoch": 21.08,
|
| 1837 |
-
"learning_rate": 3.255675029868578e-05,
|
| 1838 |
-
"loss": 0.2092,
|
| 1839 |
-
"step": 2630
|
| 1840 |
-
},
|
| 1841 |
-
{
|
| 1842 |
-
"epoch": 21.16,
|
| 1843 |
-
"learning_rate": 3.2258064516129034e-05,
|
| 1844 |
-
"loss": 0.2134,
|
| 1845 |
-
"step": 2640
|
| 1846 |
-
},
|
| 1847 |
-
{
|
| 1848 |
-
"epoch": 21.24,
|
| 1849 |
-
"learning_rate": 3.195937873357228e-05,
|
| 1850 |
-
"loss": 0.2116,
|
| 1851 |
-
"step": 2650
|
| 1852 |
-
},
|
| 1853 |
-
{
|
| 1854 |
-
"epoch": 21.32,
|
| 1855 |
-
"learning_rate": 3.1660692951015537e-05,
|
| 1856 |
-
"loss": 0.1958,
|
| 1857 |
-
"step": 2660
|
| 1858 |
-
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 21.4,
|
| 1861 |
-
"learning_rate": 3.136200716845878e-05,
|
| 1862 |
-
"loss": 0.2259,
|
| 1863 |
-
"step": 2670
|
| 1864 |
-
},
|
| 1865 |
-
{
|
| 1866 |
-
"epoch": 21.48,
|
| 1867 |
-
"learning_rate": 3.106332138590203e-05,
|
| 1868 |
-
"loss": 0.2121,
|
| 1869 |
-
"step": 2680
|
| 1870 |
-
},
|
| 1871 |
-
{
|
| 1872 |
-
"epoch": 21.56,
|
| 1873 |
-
"learning_rate": 3.0764635603345284e-05,
|
| 1874 |
-
"loss": 0.2077,
|
| 1875 |
-
"step": 2690
|
| 1876 |
-
},
|
| 1877 |
-
{
|
| 1878 |
-
"epoch": 21.64,
|
| 1879 |
-
"learning_rate": 3.046594982078853e-05,
|
| 1880 |
-
"loss": 0.2141,
|
| 1881 |
-
"step": 2700
|
| 1882 |
-
},
|
| 1883 |
-
{
|
| 1884 |
-
"epoch": 21.72,
|
| 1885 |
-
"learning_rate": 3.016726403823178e-05,
|
| 1886 |
-
"loss": 0.219,
|
| 1887 |
-
"step": 2710
|
| 1888 |
-
},
|
| 1889 |
-
{
|
| 1890 |
-
"epoch": 21.8,
|
| 1891 |
-
"learning_rate": 2.9868578255675032e-05,
|
| 1892 |
-
"loss": 0.2477,
|
| 1893 |
-
"step": 2720
|
| 1894 |
-
},
|
| 1895 |
-
{
|
| 1896 |
-
"epoch": 21.88,
|
| 1897 |
-
"learning_rate": 2.9569892473118284e-05,
|
| 1898 |
-
"loss": 0.2114,
|
| 1899 |
-
"step": 2730
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 21.96,
|
| 1903 |
-
"learning_rate": 2.9271206690561532e-05,
|
| 1904 |
-
"loss": 0.2019,
|
| 1905 |
-
"step": 2740
|
| 1906 |
-
},
|
| 1907 |
-
{
|
| 1908 |
-
"epoch": 22.0,
|
| 1909 |
-
"eval_accuracy": 0.830890642615558,
|
| 1910 |
-
"eval_f1": 0.8294472852659055,
|
| 1911 |
-
"eval_loss": 0.6705580949783325,
|
| 1912 |
-
"eval_precision": 0.831426218316801,
|
| 1913 |
-
"eval_recall": 0.830890642615558,
|
| 1914 |
-
"eval_runtime": 22.3451,
|
| 1915 |
-
"eval_samples_per_second": 79.391,
|
| 1916 |
-
"eval_steps_per_second": 2.506,
|
| 1917 |
-
"step": 2744
|
| 1918 |
-
},
|
| 1919 |
-
{
|
| 1920 |
-
"epoch": 22.04,
|
| 1921 |
-
"learning_rate": 2.897252090800478e-05,
|
| 1922 |
-
"loss": 0.2186,
|
| 1923 |
-
"step": 2750
|
| 1924 |
-
},
|
| 1925 |
-
{
|
| 1926 |
-
"epoch": 22.12,
|
| 1927 |
-
"learning_rate": 2.8673835125448028e-05,
|
| 1928 |
-
"loss": 0.2436,
|
| 1929 |
-
"step": 2760
|
| 1930 |
-
},
|
| 1931 |
-
{
|
| 1932 |
-
"epoch": 22.2,
|
| 1933 |
-
"learning_rate": 2.8375149342891276e-05,
|
| 1934 |
-
"loss": 0.1869,
|
| 1935 |
-
"step": 2770
|
| 1936 |
-
},
|
| 1937 |
-
{
|
| 1938 |
-
"epoch": 22.28,
|
| 1939 |
-
"learning_rate": 2.807646356033453e-05,
|
| 1940 |
-
"loss": 0.1906,
|
| 1941 |
-
"step": 2780
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 22.36,
|
| 1945 |
-
"learning_rate": 2.777777777777778e-05,
|
| 1946 |
-
"loss": 0.1993,
|
| 1947 |
-
"step": 2790
|
| 1948 |
-
},
|
| 1949 |
-
{
|
| 1950 |
-
"epoch": 22.44,
|
| 1951 |
-
"learning_rate": 2.747909199522103e-05,
|
| 1952 |
-
"loss": 0.2213,
|
| 1953 |
-
"step": 2800
|
| 1954 |
-
},
|
| 1955 |
-
{
|
| 1956 |
-
"epoch": 22.53,
|
| 1957 |
-
"learning_rate": 2.718040621266428e-05,
|
| 1958 |
-
"loss": 0.1938,
|
| 1959 |
-
"step": 2810
|
| 1960 |
-
},
|
| 1961 |
-
{
|
| 1962 |
-
"epoch": 22.61,
|
| 1963 |
-
"learning_rate": 2.6881720430107527e-05,
|
| 1964 |
-
"loss": 0.194,
|
| 1965 |
-
"step": 2820
|
| 1966 |
-
},
|
| 1967 |
-
{
|
| 1968 |
-
"epoch": 22.69,
|
| 1969 |
-
"learning_rate": 2.6583034647550775e-05,
|
| 1970 |
-
"loss": 0.2092,
|
| 1971 |
-
"step": 2830
|
| 1972 |
-
},
|
| 1973 |
-
{
|
| 1974 |
-
"epoch": 22.77,
|
| 1975 |
-
"learning_rate": 2.628434886499403e-05,
|
| 1976 |
-
"loss": 0.1998,
|
| 1977 |
-
"step": 2840
|
| 1978 |
-
},
|
| 1979 |
-
{
|
| 1980 |
-
"epoch": 22.85,
|
| 1981 |
-
"learning_rate": 2.5985663082437278e-05,
|
| 1982 |
-
"loss": 0.1966,
|
| 1983 |
-
"step": 2850
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 22.93,
|
| 1987 |
-
"learning_rate": 2.5686977299880526e-05,
|
| 1988 |
-
"loss": 0.227,
|
| 1989 |
-
"step": 2860
|
| 1990 |
-
},
|
| 1991 |
-
{
|
| 1992 |
-
"epoch": 23.0,
|
| 1993 |
-
"eval_accuracy": 0.8399098083427283,
|
| 1994 |
-
"eval_f1": 0.837257957402222,
|
| 1995 |
-
"eval_loss": 0.6810197234153748,
|
| 1996 |
-
"eval_precision": 0.8362119147881629,
|
| 1997 |
-
"eval_recall": 0.8399098083427283,
|
| 1998 |
-
"eval_runtime": 22.3602,
|
| 1999 |
-
"eval_samples_per_second": 79.337,
|
| 2000 |
-
"eval_steps_per_second": 2.504,
|
| 2001 |
-
"step": 2869
|
| 2002 |
-
},
|
| 2003 |
-
{
|
| 2004 |
-
"epoch": 23.01,
|
| 2005 |
-
"learning_rate": 2.5388291517323774e-05,
|
| 2006 |
-
"loss": 0.2024,
|
| 2007 |
-
"step": 2870
|
| 2008 |
-
},
|
| 2009 |
-
{
|
| 2010 |
-
"epoch": 23.09,
|
| 2011 |
-
"learning_rate": 2.5089605734767026e-05,
|
| 2012 |
-
"loss": 0.2197,
|
| 2013 |
-
"step": 2880
|
| 2014 |
-
},
|
| 2015 |
-
{
|
| 2016 |
-
"epoch": 23.17,
|
| 2017 |
-
"learning_rate": 2.4790919952210277e-05,
|
| 2018 |
-
"loss": 0.2037,
|
| 2019 |
-
"step": 2890
|
| 2020 |
-
},
|
| 2021 |
-
{
|
| 2022 |
-
"epoch": 23.25,
|
| 2023 |
-
"learning_rate": 2.4492234169653525e-05,
|
| 2024 |
-
"loss": 0.1667,
|
| 2025 |
-
"step": 2900
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 23.33,
|
| 2029 |
-
"learning_rate": 2.4193548387096777e-05,
|
| 2030 |
-
"loss": 0.212,
|
| 2031 |
-
"step": 2910
|
| 2032 |
-
},
|
| 2033 |
-
{
|
| 2034 |
-
"epoch": 23.41,
|
| 2035 |
-
"learning_rate": 2.3894862604540025e-05,
|
| 2036 |
-
"loss": 0.1947,
|
| 2037 |
-
"step": 2920
|
| 2038 |
-
},
|
| 2039 |
-
{
|
| 2040 |
-
"epoch": 23.49,
|
| 2041 |
-
"learning_rate": 2.3596176821983273e-05,
|
| 2042 |
-
"loss": 0.188,
|
| 2043 |
-
"step": 2930
|
| 2044 |
-
},
|
| 2045 |
-
{
|
| 2046 |
-
"epoch": 23.57,
|
| 2047 |
-
"learning_rate": 2.3297491039426525e-05,
|
| 2048 |
-
"loss": 0.2253,
|
| 2049 |
-
"step": 2940
|
| 2050 |
-
},
|
| 2051 |
-
{
|
| 2052 |
-
"epoch": 23.65,
|
| 2053 |
-
"learning_rate": 2.2998805256869773e-05,
|
| 2054 |
-
"loss": 0.2012,
|
| 2055 |
-
"step": 2950
|
| 2056 |
-
},
|
| 2057 |
-
{
|
| 2058 |
-
"epoch": 23.73,
|
| 2059 |
-
"learning_rate": 2.270011947431302e-05,
|
| 2060 |
-
"loss": 0.1907,
|
| 2061 |
-
"step": 2960
|
| 2062 |
-
},
|
| 2063 |
-
{
|
| 2064 |
-
"epoch": 23.81,
|
| 2065 |
-
"learning_rate": 2.2401433691756272e-05,
|
| 2066 |
-
"loss": 0.1828,
|
| 2067 |
-
"step": 2970
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 23.89,
|
| 2071 |
-
"learning_rate": 2.2102747909199524e-05,
|
| 2072 |
-
"loss": 0.1983,
|
| 2073 |
-
"step": 2980
|
| 2074 |
-
},
|
| 2075 |
-
{
|
| 2076 |
-
"epoch": 23.97,
|
| 2077 |
-
"learning_rate": 2.1804062126642775e-05,
|
| 2078 |
-
"loss": 0.181,
|
| 2079 |
-
"step": 2990
|
| 2080 |
-
},
|
| 2081 |
-
{
|
| 2082 |
-
"epoch": 24.0,
|
| 2083 |
-
"eval_accuracy": 0.8483652762119503,
|
| 2084 |
-
"eval_f1": 0.845187740150188,
|
| 2085 |
-
"eval_loss": 0.6782332062721252,
|
| 2086 |
-
"eval_precision": 0.8449457293571202,
|
| 2087 |
-
"eval_recall": 0.8483652762119503,
|
| 2088 |
-
"eval_runtime": 22.3679,
|
| 2089 |
-
"eval_samples_per_second": 79.31,
|
| 2090 |
-
"eval_steps_per_second": 2.504,
|
| 2091 |
-
"step": 2994
|
| 2092 |
-
},
|
| 2093 |
-
{
|
| 2094 |
-
"epoch": 24.05,
|
| 2095 |
-
"learning_rate": 2.1505376344086024e-05,
|
| 2096 |
-
"loss": 0.197,
|
| 2097 |
-
"step": 3000
|
| 2098 |
-
},
|
| 2099 |
-
{
|
| 2100 |
-
"epoch": 24.13,
|
| 2101 |
-
"learning_rate": 2.120669056152927e-05,
|
| 2102 |
-
"loss": 0.2024,
|
| 2103 |
-
"step": 3010
|
| 2104 |
-
},
|
| 2105 |
-
{
|
| 2106 |
-
"epoch": 24.21,
|
| 2107 |
-
"learning_rate": 2.0908004778972523e-05,
|
| 2108 |
-
"loss": 0.1791,
|
| 2109 |
-
"step": 3020
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 24.29,
|
| 2113 |
-
"learning_rate": 2.060931899641577e-05,
|
| 2114 |
-
"loss": 0.2235,
|
| 2115 |
-
"step": 3030
|
| 2116 |
-
},
|
| 2117 |
-
{
|
| 2118 |
-
"epoch": 24.37,
|
| 2119 |
-
"learning_rate": 2.031063321385902e-05,
|
| 2120 |
-
"loss": 0.1916,
|
| 2121 |
-
"step": 3040
|
| 2122 |
-
},
|
| 2123 |
-
{
|
| 2124 |
-
"epoch": 24.45,
|
| 2125 |
-
"learning_rate": 2.001194743130227e-05,
|
| 2126 |
-
"loss": 0.2072,
|
| 2127 |
-
"step": 3050
|
| 2128 |
-
},
|
| 2129 |
-
{
|
| 2130 |
-
"epoch": 24.53,
|
| 2131 |
-
"learning_rate": 1.9713261648745522e-05,
|
| 2132 |
-
"loss": 0.1842,
|
| 2133 |
-
"step": 3060
|
| 2134 |
-
},
|
| 2135 |
-
{
|
| 2136 |
-
"epoch": 24.61,
|
| 2137 |
-
"learning_rate": 1.941457586618877e-05,
|
| 2138 |
-
"loss": 0.1834,
|
| 2139 |
-
"step": 3070
|
| 2140 |
-
},
|
| 2141 |
-
{
|
| 2142 |
-
"epoch": 24.69,
|
| 2143 |
-
"learning_rate": 1.9115890083632022e-05,
|
| 2144 |
-
"loss": 0.1873,
|
| 2145 |
-
"step": 3080
|
| 2146 |
-
},
|
| 2147 |
-
{
|
| 2148 |
-
"epoch": 24.77,
|
| 2149 |
-
"learning_rate": 1.881720430107527e-05,
|
| 2150 |
-
"loss": 0.1894,
|
| 2151 |
-
"step": 3090
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 24.85,
|
| 2155 |
-
"learning_rate": 1.8518518518518518e-05,
|
| 2156 |
-
"loss": 0.1985,
|
| 2157 |
-
"step": 3100
|
| 2158 |
-
},
|
| 2159 |
-
{
|
| 2160 |
-
"epoch": 24.93,
|
| 2161 |
-
"learning_rate": 1.821983273596177e-05,
|
| 2162 |
-
"loss": 0.1905,
|
| 2163 |
-
"step": 3110
|
| 2164 |
-
},
|
| 2165 |
-
{
|
| 2166 |
-
"epoch": 24.99,
|
| 2167 |
-
"eval_accuracy": 0.8399098083427283,
|
| 2168 |
-
"eval_f1": 0.8384726457864934,
|
| 2169 |
-
"eval_loss": 0.726107120513916,
|
| 2170 |
-
"eval_precision": 0.8389807600396787,
|
| 2171 |
-
"eval_recall": 0.8399098083427283,
|
| 2172 |
-
"eval_runtime": 22.3158,
|
| 2173 |
-
"eval_samples_per_second": 79.495,
|
| 2174 |
-
"eval_steps_per_second": 2.509,
|
| 2175 |
-
"step": 3118
|
| 2176 |
-
},
|
| 2177 |
-
{
|
| 2178 |
-
"epoch": 25.01,
|
| 2179 |
-
"learning_rate": 1.7921146953405018e-05,
|
| 2180 |
-
"loss": 0.1772,
|
| 2181 |
-
"step": 3120
|
| 2182 |
-
},
|
| 2183 |
-
{
|
| 2184 |
-
"epoch": 25.09,
|
| 2185 |
-
"learning_rate": 1.7622461170848266e-05,
|
| 2186 |
-
"loss": 0.1892,
|
| 2187 |
-
"step": 3130
|
| 2188 |
-
},
|
| 2189 |
-
{
|
| 2190 |
-
"epoch": 25.17,
|
| 2191 |
-
"learning_rate": 1.7323775388291518e-05,
|
| 2192 |
-
"loss": 0.1698,
|
| 2193 |
-
"step": 3140
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 25.25,
|
| 2197 |
-
"learning_rate": 1.702508960573477e-05,
|
| 2198 |
-
"loss": 0.1546,
|
| 2199 |
-
"step": 3150
|
| 2200 |
-
},
|
| 2201 |
-
{
|
| 2202 |
-
"epoch": 25.33,
|
| 2203 |
-
"learning_rate": 1.6726403823178017e-05,
|
| 2204 |
-
"loss": 0.1623,
|
| 2205 |
-
"step": 3160
|
| 2206 |
-
},
|
| 2207 |
-
{
|
| 2208 |
-
"epoch": 25.41,
|
| 2209 |
-
"learning_rate": 1.642771804062127e-05,
|
| 2210 |
-
"loss": 0.1833,
|
| 2211 |
-
"step": 3170
|
| 2212 |
-
},
|
| 2213 |
-
{
|
| 2214 |
-
"epoch": 25.49,
|
| 2215 |
-
"learning_rate": 1.6129032258064517e-05,
|
| 2216 |
-
"loss": 0.1987,
|
| 2217 |
-
"step": 3180
|
| 2218 |
-
},
|
| 2219 |
-
{
|
| 2220 |
-
"epoch": 25.57,
|
| 2221 |
-
"learning_rate": 1.5830346475507768e-05,
|
| 2222 |
-
"loss": 0.197,
|
| 2223 |
-
"step": 3190
|
| 2224 |
-
},
|
| 2225 |
-
{
|
| 2226 |
-
"epoch": 25.65,
|
| 2227 |
-
"learning_rate": 1.5531660692951016e-05,
|
| 2228 |
-
"loss": 0.175,
|
| 2229 |
-
"step": 3200
|
| 2230 |
-
},
|
| 2231 |
-
{
|
| 2232 |
-
"epoch": 25.73,
|
| 2233 |
-
"learning_rate": 1.5232974910394265e-05,
|
| 2234 |
-
"loss": 0.193,
|
| 2235 |
-
"step": 3210
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 25.81,
|
| 2239 |
-
"learning_rate": 1.4934289127837516e-05,
|
| 2240 |
-
"loss": 0.1519,
|
| 2241 |
-
"step": 3220
|
| 2242 |
-
},
|
| 2243 |
-
{
|
| 2244 |
-
"epoch": 25.89,
|
| 2245 |
-
"learning_rate": 1.4635603345280766e-05,
|
| 2246 |
-
"loss": 0.1663,
|
| 2247 |
-
"step": 3230
|
| 2248 |
-
},
|
| 2249 |
-
{
|
| 2250 |
-
"epoch": 25.97,
|
| 2251 |
-
"learning_rate": 1.4336917562724014e-05,
|
| 2252 |
-
"loss": 0.1953,
|
| 2253 |
-
"step": 3240
|
| 2254 |
-
},
|
| 2255 |
-
{
|
| 2256 |
-
"epoch": 26.0,
|
| 2257 |
-
"eval_accuracy": 0.8494926719278467,
|
| 2258 |
-
"eval_f1": 0.8474091705502425,
|
| 2259 |
-
"eval_loss": 0.6963042616844177,
|
| 2260 |
-
"eval_precision": 0.8477969804807908,
|
| 2261 |
-
"eval_recall": 0.8494926719278467,
|
| 2262 |
-
"eval_runtime": 22.3057,
|
| 2263 |
-
"eval_samples_per_second": 79.531,
|
| 2264 |
-
"eval_steps_per_second": 2.511,
|
| 2265 |
-
"step": 3243
|
| 2266 |
-
},
|
| 2267 |
-
{
|
| 2268 |
-
"epoch": 26.05,
|
| 2269 |
-
"learning_rate": 1.4038231780167265e-05,
|
| 2270 |
-
"loss": 0.1705,
|
| 2271 |
-
"step": 3250
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 26.13,
|
| 2275 |
-
"learning_rate": 1.3739545997610515e-05,
|
| 2276 |
-
"loss": 0.2156,
|
| 2277 |
-
"step": 3260
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 26.21,
|
| 2281 |
-
"learning_rate": 1.3440860215053763e-05,
|
| 2282 |
-
"loss": 0.1703,
|
| 2283 |
-
"step": 3270
|
| 2284 |
-
},
|
| 2285 |
-
{
|
| 2286 |
-
"epoch": 26.29,
|
| 2287 |
-
"learning_rate": 1.3142174432497015e-05,
|
| 2288 |
-
"loss": 0.165,
|
| 2289 |
-
"step": 3280
|
| 2290 |
-
},
|
| 2291 |
-
{
|
| 2292 |
-
"epoch": 26.37,
|
| 2293 |
-
"learning_rate": 1.2843488649940263e-05,
|
| 2294 |
-
"loss": 0.2087,
|
| 2295 |
-
"step": 3290
|
| 2296 |
-
},
|
| 2297 |
-
{
|
| 2298 |
-
"epoch": 26.45,
|
| 2299 |
-
"learning_rate": 1.2544802867383513e-05,
|
| 2300 |
-
"loss": 0.1819,
|
| 2301 |
-
"step": 3300
|
| 2302 |
-
},
|
| 2303 |
-
{
|
| 2304 |
-
"epoch": 26.53,
|
| 2305 |
-
"learning_rate": 1.2246117084826763e-05,
|
| 2306 |
-
"loss": 0.1938,
|
| 2307 |
-
"step": 3310
|
| 2308 |
-
},
|
| 2309 |
-
{
|
| 2310 |
-
"epoch": 26.61,
|
| 2311 |
-
"learning_rate": 1.1947431302270013e-05,
|
| 2312 |
-
"loss": 0.1777,
|
| 2313 |
-
"step": 3320
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 26.69,
|
| 2317 |
-
"learning_rate": 1.1648745519713262e-05,
|
| 2318 |
-
"loss": 0.172,
|
| 2319 |
-
"step": 3330
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 26.77,
|
| 2323 |
-
"learning_rate": 1.135005973715651e-05,
|
| 2324 |
-
"loss": 0.1974,
|
| 2325 |
-
"step": 3340
|
| 2326 |
-
},
|
| 2327 |
-
{
|
| 2328 |
-
"epoch": 26.85,
|
| 2329 |
-
"learning_rate": 1.1051373954599762e-05,
|
| 2330 |
-
"loss": 0.1784,
|
| 2331 |
-
"step": 3350
|
| 2332 |
-
},
|
| 2333 |
-
{
|
| 2334 |
-
"epoch": 26.93,
|
| 2335 |
-
"learning_rate": 1.0752688172043012e-05,
|
| 2336 |
-
"loss": 0.1853,
|
| 2337 |
-
"step": 3360
|
| 2338 |
-
},
|
| 2339 |
-
{
|
| 2340 |
-
"epoch": 27.0,
|
| 2341 |
-
"eval_accuracy": 0.8432919954904171,
|
| 2342 |
-
"eval_f1": 0.8441059528157674,
|
| 2343 |
-
"eval_loss": 0.6968245506286621,
|
| 2344 |
-
"eval_precision": 0.8460210470553858,
|
| 2345 |
-
"eval_recall": 0.8432919954904171,
|
| 2346 |
-
"eval_runtime": 22.2612,
|
| 2347 |
-
"eval_samples_per_second": 79.69,
|
| 2348 |
-
"eval_steps_per_second": 2.516,
|
| 2349 |
-
"step": 3368
|
| 2350 |
-
},
|
| 2351 |
-
{
|
| 2352 |
-
"epoch": 27.01,
|
| 2353 |
-
"learning_rate": 1.0454002389486262e-05,
|
| 2354 |
-
"loss": 0.2023,
|
| 2355 |
-
"step": 3370
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 27.09,
|
| 2359 |
-
"learning_rate": 1.015531660692951e-05,
|
| 2360 |
-
"loss": 0.1866,
|
| 2361 |
-
"step": 3380
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 27.17,
|
| 2365 |
-
"learning_rate": 9.856630824372761e-06,
|
| 2366 |
-
"loss": 0.1627,
|
| 2367 |
-
"step": 3390
|
| 2368 |
-
},
|
| 2369 |
-
{
|
| 2370 |
-
"epoch": 27.25,
|
| 2371 |
-
"learning_rate": 9.557945041816011e-06,
|
| 2372 |
-
"loss": 0.2159,
|
| 2373 |
-
"step": 3400
|
| 2374 |
-
},
|
| 2375 |
-
{
|
| 2376 |
-
"epoch": 27.33,
|
| 2377 |
-
"learning_rate": 9.259259259259259e-06,
|
| 2378 |
-
"loss": 0.1907,
|
| 2379 |
-
"step": 3410
|
| 2380 |
-
},
|
| 2381 |
-
{
|
| 2382 |
-
"epoch": 27.41,
|
| 2383 |
-
"learning_rate": 8.960573476702509e-06,
|
| 2384 |
-
"loss": 0.1811,
|
| 2385 |
-
"step": 3420
|
| 2386 |
-
},
|
| 2387 |
-
{
|
| 2388 |
-
"epoch": 27.49,
|
| 2389 |
-
"learning_rate": 8.661887694145759e-06,
|
| 2390 |
-
"loss": 0.1746,
|
| 2391 |
-
"step": 3430
|
| 2392 |
-
},
|
| 2393 |
-
{
|
| 2394 |
-
"epoch": 27.58,
|
| 2395 |
-
"learning_rate": 8.363201911589009e-06,
|
| 2396 |
-
"loss": 0.1535,
|
| 2397 |
-
"step": 3440
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 27.66,
|
| 2401 |
-
"learning_rate": 8.064516129032258e-06,
|
| 2402 |
-
"loss": 0.1547,
|
| 2403 |
-
"step": 3450
|
| 2404 |
-
},
|
| 2405 |
-
{
|
| 2406 |
-
"epoch": 27.74,
|
| 2407 |
-
"learning_rate": 7.765830346475508e-06,
|
| 2408 |
-
"loss": 0.1775,
|
| 2409 |
-
"step": 3460
|
| 2410 |
-
},
|
| 2411 |
-
{
|
| 2412 |
-
"epoch": 27.82,
|
| 2413 |
-
"learning_rate": 7.467144563918758e-06,
|
| 2414 |
-
"loss": 0.186,
|
| 2415 |
-
"step": 3470
|
| 2416 |
-
},
|
| 2417 |
-
{
|
| 2418 |
-
"epoch": 27.9,
|
| 2419 |
-
"learning_rate": 7.168458781362007e-06,
|
| 2420 |
-
"loss": 0.1797,
|
| 2421 |
-
"step": 3480
|
| 2422 |
-
},
|
| 2423 |
-
{
|
| 2424 |
-
"epoch": 27.98,
|
| 2425 |
-
"learning_rate": 6.869772998805258e-06,
|
| 2426 |
-
"loss": 0.201,
|
| 2427 |
-
"step": 3490
|
| 2428 |
-
},
|
| 2429 |
-
{
|
| 2430 |
-
"epoch": 28.0,
|
| 2431 |
-
"eval_accuracy": 0.8416009019165727,
|
| 2432 |
-
"eval_f1": 0.8401630929550372,
|
| 2433 |
-
"eval_loss": 0.7103535532951355,
|
| 2434 |
-
"eval_precision": 0.8408996757640382,
|
| 2435 |
-
"eval_recall": 0.8416009019165727,
|
| 2436 |
-
"eval_runtime": 22.3276,
|
| 2437 |
-
"eval_samples_per_second": 79.453,
|
| 2438 |
-
"eval_steps_per_second": 2.508,
|
| 2439 |
-
"step": 3493
|
| 2440 |
-
},
|
| 2441 |
-
{
|
| 2442 |
-
"epoch": 28.06,
|
| 2443 |
-
"learning_rate": 6.5710872162485075e-06,
|
| 2444 |
-
"loss": 0.1946,
|
| 2445 |
-
"step": 3500
|
| 2446 |
-
},
|
| 2447 |
-
{
|
| 2448 |
-
"epoch": 28.14,
|
| 2449 |
-
"learning_rate": 6.2724014336917564e-06,
|
| 2450 |
-
"loss": 0.1747,
|
| 2451 |
-
"step": 3510
|
| 2452 |
-
},
|
| 2453 |
-
{
|
| 2454 |
-
"epoch": 28.22,
|
| 2455 |
-
"learning_rate": 5.973715651135006e-06,
|
| 2456 |
-
"loss": 0.167,
|
| 2457 |
-
"step": 3520
|
| 2458 |
-
},
|
| 2459 |
-
{
|
| 2460 |
-
"epoch": 28.3,
|
| 2461 |
-
"learning_rate": 5.675029868578255e-06,
|
| 2462 |
-
"loss": 0.1577,
|
| 2463 |
-
"step": 3530
|
| 2464 |
-
},
|
| 2465 |
-
{
|
| 2466 |
-
"epoch": 28.38,
|
| 2467 |
-
"learning_rate": 5.376344086021506e-06,
|
| 2468 |
-
"loss": 0.171,
|
| 2469 |
-
"step": 3540
|
| 2470 |
-
},
|
| 2471 |
-
{
|
| 2472 |
-
"epoch": 28.46,
|
| 2473 |
-
"learning_rate": 5.077658303464755e-06,
|
| 2474 |
-
"loss": 0.1671,
|
| 2475 |
-
"step": 3550
|
| 2476 |
-
},
|
| 2477 |
-
{
|
| 2478 |
-
"epoch": 28.54,
|
| 2479 |
-
"learning_rate": 4.7789725209080055e-06,
|
| 2480 |
-
"loss": 0.1497,
|
| 2481 |
-
"step": 3560
|
| 2482 |
-
},
|
| 2483 |
-
{
|
| 2484 |
-
"epoch": 28.62,
|
| 2485 |
-
"learning_rate": 4.4802867383512545e-06,
|
| 2486 |
-
"loss": 0.1839,
|
| 2487 |
-
"step": 3570
|
| 2488 |
-
},
|
| 2489 |
-
{
|
| 2490 |
-
"epoch": 28.7,
|
| 2491 |
-
"learning_rate": 4.181600955794504e-06,
|
| 2492 |
-
"loss": 0.173,
|
| 2493 |
-
"step": 3580
|
| 2494 |
-
},
|
| 2495 |
-
{
|
| 2496 |
-
"epoch": 28.78,
|
| 2497 |
-
"learning_rate": 3.882915173237754e-06,
|
| 2498 |
-
"loss": 0.181,
|
| 2499 |
-
"step": 3590
|
| 2500 |
-
},
|
| 2501 |
-
{
|
| 2502 |
-
"epoch": 28.86,
|
| 2503 |
-
"learning_rate": 3.5842293906810035e-06,
|
| 2504 |
-
"loss": 0.1862,
|
| 2505 |
-
"step": 3600
|
| 2506 |
-
},
|
| 2507 |
-
{
|
| 2508 |
-
"epoch": 28.94,
|
| 2509 |
-
"learning_rate": 3.2855436081242537e-06,
|
| 2510 |
-
"loss": 0.1613,
|
| 2511 |
-
"step": 3610
|
| 2512 |
-
},
|
| 2513 |
-
{
|
| 2514 |
-
"epoch": 28.99,
|
| 2515 |
-
"eval_accuracy": 0.8432919954904171,
|
| 2516 |
-
"eval_f1": 0.8414849178968039,
|
| 2517 |
-
"eval_loss": 0.7158181667327881,
|
| 2518 |
-
"eval_precision": 0.8414280361847442,
|
| 2519 |
-
"eval_recall": 0.8432919954904171,
|
| 2520 |
-
"eval_runtime": 22.2521,
|
| 2521 |
-
"eval_samples_per_second": 79.723,
|
| 2522 |
-
"eval_steps_per_second": 2.517,
|
| 2523 |
-
"step": 3617
|
| 2524 |
-
},
|
| 2525 |
-
{
|
| 2526 |
-
"epoch": 29.02,
|
| 2527 |
-
"learning_rate": 2.986857825567503e-06,
|
| 2528 |
-
"loss": 0.1771,
|
| 2529 |
-
"step": 3620
|
| 2530 |
-
},
|
| 2531 |
-
{
|
| 2532 |
-
"epoch": 29.1,
|
| 2533 |
-
"learning_rate": 2.688172043010753e-06,
|
| 2534 |
-
"loss": 0.1558,
|
| 2535 |
-
"step": 3630
|
| 2536 |
-
},
|
| 2537 |
-
{
|
| 2538 |
-
"epoch": 29.18,
|
| 2539 |
-
"learning_rate": 2.3894862604540028e-06,
|
| 2540 |
-
"loss": 0.1654,
|
| 2541 |
-
"step": 3640
|
| 2542 |
-
},
|
| 2543 |
-
{
|
| 2544 |
-
"epoch": 29.26,
|
| 2545 |
-
"learning_rate": 2.090800477897252e-06,
|
| 2546 |
-
"loss": 0.165,
|
| 2547 |
-
"step": 3650
|
| 2548 |
-
},
|
| 2549 |
-
{
|
| 2550 |
-
"epoch": 29.34,
|
| 2551 |
-
"learning_rate": 1.7921146953405017e-06,
|
| 2552 |
-
"loss": 0.1649,
|
| 2553 |
-
"step": 3660
|
| 2554 |
-
},
|
| 2555 |
-
{
|
| 2556 |
-
"epoch": 29.42,
|
| 2557 |
-
"learning_rate": 1.4934289127837516e-06,
|
| 2558 |
-
"loss": 0.1795,
|
| 2559 |
-
"step": 3670
|
| 2560 |
-
},
|
| 2561 |
-
{
|
| 2562 |
-
"epoch": 29.5,
|
| 2563 |
-
"learning_rate": 1.1947431302270014e-06,
|
| 2564 |
-
"loss": 0.1468,
|
| 2565 |
-
"step": 3680
|
| 2566 |
-
},
|
| 2567 |
-
{
|
| 2568 |
-
"epoch": 29.58,
|
| 2569 |
-
"learning_rate": 8.960573476702509e-07,
|
| 2570 |
-
"loss": 0.1846,
|
| 2571 |
-
"step": 3690
|
| 2572 |
-
},
|
| 2573 |
-
{
|
| 2574 |
-
"epoch": 29.66,
|
| 2575 |
-
"learning_rate": 5.973715651135007e-07,
|
| 2576 |
-
"loss": 0.1509,
|
| 2577 |
-
"step": 3700
|
| 2578 |
-
},
|
| 2579 |
-
{
|
| 2580 |
-
"epoch": 29.74,
|
| 2581 |
-
"learning_rate": 2.9868578255675034e-07,
|
| 2582 |
-
"loss": 0.1678,
|
| 2583 |
-
"step": 3710
|
| 2584 |
-
},
|
| 2585 |
-
{
|
| 2586 |
-
"epoch": 29.82,
|
| 2587 |
-
"learning_rate": 0.0,
|
| 2588 |
-
"loss": 0.1509,
|
| 2589 |
-
"step": 3720
|
| 2590 |
-
},
|
| 2591 |
-
{
|
| 2592 |
-
"epoch": 29.82,
|
| 2593 |
-
"eval_accuracy": 0.8494926719278467,
|
| 2594 |
-
"eval_f1": 0.8470597592760953,
|
| 2595 |
-
"eval_loss": 0.7118731737136841,
|
| 2596 |
-
"eval_precision": 0.8471993952398931,
|
| 2597 |
-
"eval_recall": 0.8494926719278467,
|
| 2598 |
-
"eval_runtime": 22.3288,
|
| 2599 |
-
"eval_samples_per_second": 79.449,
|
| 2600 |
-
"eval_steps_per_second": 2.508,
|
| 2601 |
-
"step": 3720
|
| 2602 |
-
},
|
| 2603 |
-
{
|
| 2604 |
-
"epoch": 29.82,
|
| 2605 |
-
"step": 3720,
|
| 2606 |
-
"total_flos": 1.3039235967867388e+20,
|
| 2607 |
-
"train_loss": 0.3687121340984939,
|
| 2608 |
-
"train_runtime": 15093.0186,
|
| 2609 |
-
"train_samples_per_second": 31.727,
|
| 2610 |
-
"train_steps_per_second": 0.246
|
| 2611 |
}
|
| 2612 |
],
|
| 2613 |
"logging_steps": 10,
|
| 2614 |
-
"max_steps":
|
| 2615 |
"num_input_tokens_seen": 0,
|
| 2616 |
"num_train_epochs": 30,
|
| 2617 |
"save_steps": 500,
|
| 2618 |
-
"total_flos": 1.
|
| 2619 |
-
"train_batch_size":
|
| 2620 |
"trial_name": null,
|
| 2621 |
"trial_params": null
|
| 2622 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8455467869222097,
|
| 3 |
+
"best_model_checkpoint": "vit-large-patch16-224-in21k-finetuned-galaxy10-decals/checkpoint-750",
|
| 4 |
+
"epoch": 29.76,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 930,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.32,
|
| 13 |
+
"learning_rate": 1.0752688172043012e-05,
|
| 14 |
+
"loss": 2.2203,
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.64,
|
| 19 |
+
"learning_rate": 2.1505376344086024e-05,
|
| 20 |
+
"loss": 1.7376,
|
| 21 |
"step": 20
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.96,
|
| 25 |
+
"learning_rate": 3.2258064516129034e-05,
|
| 26 |
+
"loss": 1.1548,
|
| 27 |
"step": 30
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.99,
|
| 31 |
+
"eval_accuracy": 0.7237880496054115,
|
| 32 |
+
"eval_f1": 0.6998518427941565,
|
| 33 |
+
"eval_loss": 0.8377441763877869,
|
| 34 |
+
"eval_precision": 0.7319527073905499,
|
| 35 |
+
"eval_recall": 0.7237880496054115,
|
| 36 |
+
"eval_runtime": 17.5371,
|
| 37 |
+
"eval_samples_per_second": 101.157,
|
| 38 |
+
"eval_steps_per_second": 0.798,
|
| 39 |
+
"step": 31
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.28,
|
| 43 |
+
"learning_rate": 4.301075268817205e-05,
|
| 44 |
+
"loss": 0.8731,
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 1.6,
|
| 49 |
+
"learning_rate": 5.3763440860215054e-05,
|
| 50 |
+
"loss": 0.7472,
|
| 51 |
"step": 50
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 1.92,
|
| 55 |
+
"learning_rate": 6.451612903225807e-05,
|
| 56 |
+
"loss": 0.7125,
|
| 57 |
"step": 60
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 1.98,
|
| 61 |
+
"eval_accuracy": 0.7846674182638106,
|
| 62 |
+
"eval_f1": 0.7807446374625957,
|
| 63 |
+
"eval_loss": 0.6551325917243958,
|
| 64 |
+
"eval_precision": 0.7913522140271126,
|
| 65 |
+
"eval_recall": 0.7846674182638106,
|
| 66 |
+
"eval_runtime": 17.5599,
|
| 67 |
+
"eval_samples_per_second": 101.026,
|
| 68 |
+
"eval_steps_per_second": 0.797,
|
| 69 |
+
"step": 62
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.24,
|
| 73 |
+
"learning_rate": 7.526881720430108e-05,
|
| 74 |
+
"loss": 0.6856,
|
| 75 |
"step": 70
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 2.56,
|
| 79 |
+
"learning_rate": 8.60215053763441e-05,
|
| 80 |
+
"loss": 0.6692,
|
| 81 |
"step": 80
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 2.88,
|
| 85 |
+
"learning_rate": 9.677419354838711e-05,
|
| 86 |
+
"loss": 0.6646,
|
| 87 |
"step": 90
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 2.98,
|
| 91 |
+
"eval_accuracy": 0.8105975197294251,
|
| 92 |
+
"eval_f1": 0.8024237083317286,
|
| 93 |
+
"eval_loss": 0.5388627648353577,
|
| 94 |
+
"eval_precision": 0.807880727787592,
|
| 95 |
+
"eval_recall": 0.8105975197294251,
|
| 96 |
+
"eval_runtime": 17.3546,
|
| 97 |
+
"eval_samples_per_second": 102.22,
|
| 98 |
+
"eval_steps_per_second": 0.807,
|
| 99 |
+
"step": 93
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 3.2,
|
| 103 |
+
"learning_rate": 9.916367980884111e-05,
|
| 104 |
+
"loss": 0.6433,
|
| 105 |
"step": 100
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 3.52,
|
| 109 |
+
"learning_rate": 9.79689366786141e-05,
|
| 110 |
+
"loss": 0.6301,
|
| 111 |
"step": 110
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 3.84,
|
| 115 |
+
"learning_rate": 9.677419354838711e-05,
|
| 116 |
+
"loss": 0.5941,
|
| 117 |
"step": 120
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 4.0,
|
| 121 |
+
"eval_accuracy": 0.8246899661781285,
|
| 122 |
+
"eval_f1": 0.8133527459199387,
|
| 123 |
+
"eval_loss": 0.5208512544631958,
|
| 124 |
+
"eval_precision": 0.8318616556747246,
|
| 125 |
+
"eval_recall": 0.8246899661781285,
|
| 126 |
+
"eval_runtime": 17.6749,
|
| 127 |
+
"eval_samples_per_second": 100.368,
|
| 128 |
+
"eval_steps_per_second": 0.792,
|
| 129 |
+
"step": 125
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 4.16,
|
| 133 |
+
"learning_rate": 9.55794504181601e-05,
|
| 134 |
+
"loss": 0.5646,
|
| 135 |
"step": 130
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 4.48,
|
| 139 |
+
"learning_rate": 9.438470728793309e-05,
|
| 140 |
+
"loss": 0.5586,
|
| 141 |
"step": 140
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 4.8,
|
| 145 |
+
"learning_rate": 9.31899641577061e-05,
|
| 146 |
+
"loss": 0.5485,
|
| 147 |
"step": 150
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 4.99,
|
| 151 |
+
"eval_accuracy": 0.8252536640360767,
|
| 152 |
+
"eval_f1": 0.8249336069302634,
|
| 153 |
+
"eval_loss": 0.522233784198761,
|
| 154 |
+
"eval_precision": 0.8317369051500769,
|
| 155 |
+
"eval_recall": 0.8252536640360767,
|
| 156 |
+
"eval_runtime": 17.7182,
|
| 157 |
+
"eval_samples_per_second": 100.123,
|
| 158 |
+
"eval_steps_per_second": 0.79,
|
| 159 |
+
"step": 156
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 5.12,
|
| 163 |
+
"learning_rate": 9.199522102747909e-05,
|
| 164 |
+
"loss": 0.5084,
|
| 165 |
"step": 160
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 5.44,
|
| 169 |
+
"learning_rate": 9.080047789725208e-05,
|
| 170 |
+
"loss": 0.5295,
|
| 171 |
"step": 170
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 5.76,
|
| 175 |
+
"learning_rate": 8.960573476702509e-05,
|
| 176 |
+
"loss": 0.4856,
|
| 177 |
"step": 180
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 5.98,
|
| 181 |
+
"eval_accuracy": 0.8297632468996617,
|
| 182 |
+
"eval_f1": 0.8240850037114443,
|
| 183 |
+
"eval_loss": 0.506437361240387,
|
| 184 |
+
"eval_precision": 0.8324418933740947,
|
| 185 |
+
"eval_recall": 0.8297632468996617,
|
| 186 |
+
"eval_runtime": 17.4786,
|
| 187 |
+
"eval_samples_per_second": 101.496,
|
| 188 |
+
"eval_steps_per_second": 0.801,
|
| 189 |
+
"step": 187
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 6.08,
|
| 193 |
+
"learning_rate": 8.84109916367981e-05,
|
| 194 |
+
"loss": 0.4901,
|
| 195 |
"step": 190
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 6.4,
|
| 199 |
+
"learning_rate": 8.72162485065711e-05,
|
| 200 |
+
"loss": 0.4703,
|
| 201 |
"step": 200
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 6.72,
|
| 205 |
+
"learning_rate": 8.60215053763441e-05,
|
| 206 |
+
"loss": 0.4613,
|
| 207 |
"step": 210
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 6.98,
|
| 211 |
+
"eval_accuracy": 0.8438556933483653,
|
| 212 |
+
"eval_f1": 0.8417548677243999,
|
| 213 |
+
"eval_loss": 0.4738723635673523,
|
| 214 |
+
"eval_precision": 0.8463534578606328,
|
| 215 |
+
"eval_recall": 0.8438556933483653,
|
| 216 |
+
"eval_runtime": 17.8615,
|
| 217 |
+
"eval_samples_per_second": 99.32,
|
| 218 |
+
"eval_steps_per_second": 0.784,
|
| 219 |
+
"step": 218
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 7.04,
|
| 223 |
+
"learning_rate": 8.482676224611709e-05,
|
| 224 |
+
"loss": 0.4871,
|
| 225 |
"step": 220
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 7.36,
|
| 229 |
+
"learning_rate": 8.363201911589009e-05,
|
| 230 |
+
"loss": 0.4725,
|
| 231 |
"step": 230
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 7.68,
|
| 235 |
+
"learning_rate": 8.243727598566309e-05,
|
| 236 |
+
"loss": 0.4634,
|
| 237 |
"step": 240
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 8.0,
|
| 241 |
+
"learning_rate": 8.124253285543608e-05,
|
| 242 |
+
"loss": 0.4395,
|
| 243 |
+
"step": 250
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 8.0,
|
| 247 |
+
"eval_accuracy": 0.8348365276211951,
|
| 248 |
+
"eval_f1": 0.8359412222619289,
|
| 249 |
+
"eval_loss": 0.49929434061050415,
|
| 250 |
+
"eval_precision": 0.8449314341267409,
|
| 251 |
+
"eval_recall": 0.8348365276211951,
|
| 252 |
+
"eval_runtime": 17.4538,
|
| 253 |
+
"eval_samples_per_second": 101.64,
|
| 254 |
+
"eval_steps_per_second": 0.802,
|
| 255 |
"step": 250
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 8.32,
|
| 259 |
+
"learning_rate": 8.004778972520908e-05,
|
| 260 |
+
"loss": 0.4077,
|
| 261 |
"step": 260
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 8.64,
|
| 265 |
+
"learning_rate": 7.885304659498209e-05,
|
| 266 |
+
"loss": 0.4297,
|
| 267 |
"step": 270
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 8.96,
|
| 271 |
+
"learning_rate": 7.765830346475508e-05,
|
| 272 |
+
"loss": 0.4184,
|
| 273 |
"step": 280
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 8.99,
|
| 277 |
+
"eval_accuracy": 0.8387824126268321,
|
| 278 |
+
"eval_f1": 0.8338694468987556,
|
| 279 |
+
"eval_loss": 0.5139430165290833,
|
| 280 |
+
"eval_precision": 0.838610429701314,
|
| 281 |
+
"eval_recall": 0.8387824126268321,
|
| 282 |
+
"eval_runtime": 17.7086,
|
| 283 |
+
"eval_samples_per_second": 100.177,
|
| 284 |
+
"eval_steps_per_second": 0.791,
|
| 285 |
+
"step": 281
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 9.28,
|
| 289 |
+
"learning_rate": 7.646356033452809e-05,
|
| 290 |
+
"loss": 0.394,
|
| 291 |
"step": 290
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 9.6,
|
| 295 |
+
"learning_rate": 7.526881720430108e-05,
|
| 296 |
+
"loss": 0.3982,
|
| 297 |
"step": 300
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 9.92,
|
| 301 |
+
"learning_rate": 7.407407407407407e-05,
|
| 302 |
+
"loss": 0.3635,
|
| 303 |
"step": 310
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 9.98,
|
| 307 |
+
"eval_accuracy": 0.8348365276211951,
|
| 308 |
+
"eval_f1": 0.8348602782924321,
|
| 309 |
+
"eval_loss": 0.5030932426452637,
|
| 310 |
+
"eval_precision": 0.837210745863444,
|
| 311 |
+
"eval_recall": 0.8348365276211951,
|
| 312 |
+
"eval_runtime": 17.8339,
|
| 313 |
+
"eval_samples_per_second": 99.473,
|
| 314 |
+
"eval_steps_per_second": 0.785,
|
| 315 |
+
"step": 312
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 10.24,
|
| 319 |
+
"learning_rate": 7.287933094384708e-05,
|
| 320 |
+
"loss": 0.3705,
|
| 321 |
"step": 320
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 10.56,
|
| 325 |
+
"learning_rate": 7.168458781362007e-05,
|
| 326 |
+
"loss": 0.3796,
|
| 327 |
"step": 330
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 10.88,
|
| 331 |
+
"learning_rate": 7.048984468339306e-05,
|
| 332 |
+
"loss": 0.3665,
|
| 333 |
"step": 340
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 10.98,
|
| 337 |
+
"eval_accuracy": 0.8337091319052987,
|
| 338 |
+
"eval_f1": 0.8318888968480479,
|
| 339 |
+
"eval_loss": 0.5234501361846924,
|
| 340 |
+
"eval_precision": 0.8330118494746035,
|
| 341 |
+
"eval_recall": 0.8337091319052987,
|
| 342 |
+
"eval_runtime": 17.5178,
|
| 343 |
+
"eval_samples_per_second": 101.269,
|
| 344 |
+
"eval_steps_per_second": 0.799,
|
| 345 |
+
"step": 343
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 11.2,
|
| 349 |
+
"learning_rate": 6.929510155316607e-05,
|
| 350 |
+
"loss": 0.3336,
|
| 351 |
"step": 350
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 11.52,
|
| 355 |
+
"learning_rate": 6.810035842293908e-05,
|
| 356 |
+
"loss": 0.3372,
|
| 357 |
"step": 360
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 11.84,
|
| 361 |
+
"learning_rate": 6.690561529271207e-05,
|
| 362 |
+
"loss": 0.3639,
|
| 363 |
"step": 370
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 12.0,
|
| 367 |
+
"eval_accuracy": 0.8432919954904171,
|
| 368 |
+
"eval_f1": 0.8428614745351065,
|
| 369 |
+
"eval_loss": 0.5032076835632324,
|
| 370 |
+
"eval_precision": 0.8470585262796811,
|
| 371 |
+
"eval_recall": 0.8432919954904171,
|
| 372 |
+
"eval_runtime": 17.6285,
|
| 373 |
+
"eval_samples_per_second": 100.633,
|
| 374 |
+
"eval_steps_per_second": 0.794,
|
| 375 |
+
"step": 375
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 12.16,
|
| 379 |
+
"learning_rate": 6.571087216248507e-05,
|
| 380 |
+
"loss": 0.3233,
|
| 381 |
"step": 380
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 12.48,
|
| 385 |
+
"learning_rate": 6.451612903225807e-05,
|
| 386 |
+
"loss": 0.3063,
|
| 387 |
"step": 390
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 12.8,
|
| 391 |
+
"learning_rate": 6.332138590203107e-05,
|
| 392 |
+
"loss": 0.344,
|
| 393 |
"step": 400
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 12.99,
|
| 397 |
+
"eval_accuracy": 0.8325817361894025,
|
| 398 |
+
"eval_f1": 0.8294643664020839,
|
| 399 |
+
"eval_loss": 0.5632408857345581,
|
| 400 |
+
"eval_precision": 0.8365617728975507,
|
| 401 |
+
"eval_recall": 0.8325817361894025,
|
| 402 |
+
"eval_runtime": 17.7706,
|
| 403 |
+
"eval_samples_per_second": 99.828,
|
| 404 |
+
"eval_steps_per_second": 0.788,
|
| 405 |
+
"step": 406
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 13.12,
|
| 409 |
+
"learning_rate": 6.212664277180407e-05,
|
| 410 |
+
"loss": 0.316,
|
| 411 |
"step": 410
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 13.44,
|
| 415 |
+
"learning_rate": 6.093189964157706e-05,
|
| 416 |
+
"loss": 0.3063,
|
| 417 |
"step": 420
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 13.76,
|
| 421 |
+
"learning_rate": 5.9737156511350064e-05,
|
| 422 |
+
"loss": 0.3089,
|
| 423 |
"step": 430
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 13.98,
|
| 427 |
+
"eval_accuracy": 0.8432919954904171,
|
| 428 |
+
"eval_f1": 0.8415075700592501,
|
| 429 |
+
"eval_loss": 0.535535991191864,
|
| 430 |
+
"eval_precision": 0.8459262463992824,
|
| 431 |
+
"eval_recall": 0.8432919954904171,
|
| 432 |
+
"eval_runtime": 17.7154,
|
| 433 |
+
"eval_samples_per_second": 100.139,
|
| 434 |
+
"eval_steps_per_second": 0.79,
|
| 435 |
+
"step": 437
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 14.08,
|
| 439 |
+
"learning_rate": 5.8542413381123063e-05,
|
| 440 |
+
"loss": 0.2896,
|
| 441 |
"step": 440
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 14.4,
|
| 445 |
+
"learning_rate": 5.7347670250896056e-05,
|
| 446 |
+
"loss": 0.2721,
|
| 447 |
"step": 450
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 14.72,
|
| 451 |
+
"learning_rate": 5.615292712066906e-05,
|
| 452 |
+
"loss": 0.2947,
|
| 453 |
"step": 460
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 14.98,
|
| 457 |
+
"eval_accuracy": 0.8370913190529876,
|
| 458 |
+
"eval_f1": 0.8363308006944028,
|
| 459 |
+
"eval_loss": 0.5595812797546387,
|
| 460 |
+
"eval_precision": 0.8379180596620824,
|
| 461 |
+
"eval_recall": 0.8370913190529876,
|
| 462 |
+
"eval_runtime": 17.5313,
|
| 463 |
+
"eval_samples_per_second": 101.191,
|
| 464 |
+
"eval_steps_per_second": 0.799,
|
| 465 |
+
"step": 468
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 15.04,
|
| 469 |
+
"learning_rate": 5.495818399044206e-05,
|
| 470 |
+
"loss": 0.2814,
|
| 471 |
"step": 470
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 15.36,
|
| 475 |
+
"learning_rate": 5.3763440860215054e-05,
|
| 476 |
+
"loss": 0.2692,
|
| 477 |
"step": 480
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 15.68,
|
| 481 |
+
"learning_rate": 5.256869772998806e-05,
|
| 482 |
+
"loss": 0.2825,
|
| 483 |
"step": 490
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 16.0,
|
| 487 |
+
"learning_rate": 5.137395459976105e-05,
|
| 488 |
+
"loss": 0.279,
|
| 489 |
+
"step": 500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 16.0,
|
| 493 |
+
"eval_accuracy": 0.8432919954904171,
|
| 494 |
+
"eval_f1": 0.8395658040260672,
|
| 495 |
+
"eval_loss": 0.5363758206367493,
|
| 496 |
+
"eval_precision": 0.840323192777568,
|
| 497 |
+
"eval_recall": 0.8432919954904171,
|
| 498 |
+
"eval_runtime": 17.807,
|
| 499 |
+
"eval_samples_per_second": 99.624,
|
| 500 |
+
"eval_steps_per_second": 0.786,
|
| 501 |
"step": 500
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 16.32,
|
| 505 |
+
"learning_rate": 5.017921146953405e-05,
|
| 506 |
+
"loss": 0.2523,
|
| 507 |
"step": 510
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 16.64,
|
| 511 |
+
"learning_rate": 4.898446833930705e-05,
|
| 512 |
+
"loss": 0.2701,
|
| 513 |
"step": 520
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 16.96,
|
| 517 |
+
"learning_rate": 4.778972520908005e-05,
|
| 518 |
+
"loss": 0.2465,
|
| 519 |
"step": 530
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 16.99,
|
| 523 |
+
"eval_accuracy": 0.8376550169109357,
|
| 524 |
+
"eval_f1": 0.8359085558912757,
|
| 525 |
+
"eval_loss": 0.5635837912559509,
|
| 526 |
+
"eval_precision": 0.8360044836481326,
|
| 527 |
+
"eval_recall": 0.8376550169109357,
|
| 528 |
+
"eval_runtime": 17.4977,
|
| 529 |
+
"eval_samples_per_second": 101.385,
|
| 530 |
+
"eval_steps_per_second": 0.8,
|
| 531 |
+
"step": 531
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 17.28,
|
| 535 |
+
"learning_rate": 4.659498207885305e-05,
|
| 536 |
+
"loss": 0.2564,
|
| 537 |
"step": 540
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 17.6,
|
| 541 |
+
"learning_rate": 4.540023894862604e-05,
|
| 542 |
+
"loss": 0.2599,
|
| 543 |
"step": 550
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 17.92,
|
| 547 |
+
"learning_rate": 4.420549581839905e-05,
|
| 548 |
+
"loss": 0.2404,
|
| 549 |
"step": 560
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 17.98,
|
| 553 |
+
"eval_accuracy": 0.8342728297632469,
|
| 554 |
+
"eval_f1": 0.8333504634521319,
|
| 555 |
+
"eval_loss": 0.6249589920043945,
|
| 556 |
+
"eval_precision": 0.8391413533867828,
|
| 557 |
+
"eval_recall": 0.8342728297632469,
|
| 558 |
+
"eval_runtime": 17.5339,
|
| 559 |
+
"eval_samples_per_second": 101.175,
|
| 560 |
+
"eval_steps_per_second": 0.798,
|
| 561 |
+
"step": 562
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 18.24,
|
| 565 |
+
"learning_rate": 4.301075268817205e-05,
|
| 566 |
+
"loss": 0.2481,
|
| 567 |
"step": 570
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 18.56,
|
| 571 |
+
"learning_rate": 4.1816009557945046e-05,
|
| 572 |
+
"loss": 0.2362,
|
| 573 |
"step": 580
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 18.88,
|
| 577 |
+
"learning_rate": 4.062126642771804e-05,
|
| 578 |
+
"loss": 0.2365,
|
| 579 |
"step": 590
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 18.98,
|
| 583 |
+
"eval_accuracy": 0.8410372040586246,
|
| 584 |
+
"eval_f1": 0.8404754925508872,
|
| 585 |
+
"eval_loss": 0.5921120047569275,
|
| 586 |
+
"eval_precision": 0.8415669743057491,
|
| 587 |
+
"eval_recall": 0.8410372040586246,
|
| 588 |
+
"eval_runtime": 17.5929,
|
| 589 |
+
"eval_samples_per_second": 100.836,
|
| 590 |
+
"eval_steps_per_second": 0.796,
|
| 591 |
+
"step": 593
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 19.2,
|
| 595 |
+
"learning_rate": 3.9426523297491045e-05,
|
| 596 |
+
"loss": 0.2357,
|
| 597 |
"step": 600
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 19.52,
|
| 601 |
+
"learning_rate": 3.8231780167264044e-05,
|
| 602 |
+
"loss": 0.2353,
|
| 603 |
"step": 610
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 19.84,
|
| 607 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 608 |
+
"loss": 0.2307,
|
| 609 |
"step": 620
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 20.0,
|
| 613 |
+
"eval_accuracy": 0.8393461104847801,
|
| 614 |
+
"eval_f1": 0.8377580101631056,
|
| 615 |
+
"eval_loss": 0.6118177771568298,
|
| 616 |
+
"eval_precision": 0.8383117076517037,
|
| 617 |
+
"eval_recall": 0.8393461104847801,
|
| 618 |
+
"eval_runtime": 17.357,
|
| 619 |
+
"eval_samples_per_second": 102.207,
|
| 620 |
+
"eval_steps_per_second": 0.807,
|
| 621 |
+
"step": 625
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 20.16,
|
| 625 |
+
"learning_rate": 3.5842293906810036e-05,
|
| 626 |
+
"loss": 0.2227,
|
| 627 |
"step": 630
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 20.48,
|
| 631 |
+
"learning_rate": 3.4647550776583035e-05,
|
| 632 |
+
"loss": 0.2123,
|
| 633 |
"step": 640
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 20.8,
|
| 637 |
+
"learning_rate": 3.3452807646356034e-05,
|
| 638 |
+
"loss": 0.2217,
|
| 639 |
"step": 650
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 20.99,
|
| 643 |
+
"eval_accuracy": 0.8410372040586246,
|
| 644 |
+
"eval_f1": 0.8389205543716994,
|
| 645 |
+
"eval_loss": 0.6195398569107056,
|
| 646 |
+
"eval_precision": 0.8398862557581964,
|
| 647 |
+
"eval_recall": 0.8410372040586246,
|
| 648 |
+
"eval_runtime": 17.7844,
|
| 649 |
+
"eval_samples_per_second": 99.75,
|
| 650 |
+
"eval_steps_per_second": 0.787,
|
| 651 |
+
"step": 656
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 21.12,
|
| 655 |
+
"learning_rate": 3.2258064516129034e-05,
|
| 656 |
+
"loss": 0.2091,
|
| 657 |
"step": 660
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 21.44,
|
| 661 |
+
"learning_rate": 3.106332138590203e-05,
|
| 662 |
+
"loss": 0.2191,
|
| 663 |
"step": 670
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 21.76,
|
| 667 |
+
"learning_rate": 2.9868578255675032e-05,
|
| 668 |
+
"loss": 0.2104,
|
| 669 |
"step": 680
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 21.98,
|
| 673 |
+
"eval_accuracy": 0.8325817361894025,
|
| 674 |
+
"eval_f1": 0.8315304985796867,
|
| 675 |
+
"eval_loss": 0.6107810735702515,
|
| 676 |
+
"eval_precision": 0.8330093624186226,
|
| 677 |
+
"eval_recall": 0.8325817361894025,
|
| 678 |
+
"eval_runtime": 17.9136,
|
| 679 |
+
"eval_samples_per_second": 99.031,
|
| 680 |
+
"eval_steps_per_second": 0.782,
|
| 681 |
+
"step": 687
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 22.08,
|
| 685 |
+
"learning_rate": 2.8673835125448028e-05,
|
| 686 |
+
"loss": 0.2175,
|
| 687 |
"step": 690
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 22.4,
|
| 691 |
+
"learning_rate": 2.747909199522103e-05,
|
| 692 |
+
"loss": 0.202,
|
| 693 |
"step": 700
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 22.72,
|
| 697 |
+
"learning_rate": 2.628434886499403e-05,
|
| 698 |
+
"loss": 0.2079,
|
| 699 |
"step": 710
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 22.98,
|
| 703 |
+
"eval_accuracy": 0.8404735062006764,
|
| 704 |
+
"eval_f1": 0.8372749251321038,
|
| 705 |
+
"eval_loss": 0.6308367252349854,
|
| 706 |
+
"eval_precision": 0.8374860539003082,
|
| 707 |
+
"eval_recall": 0.8404735062006764,
|
| 708 |
+
"eval_runtime": 17.5068,
|
| 709 |
+
"eval_samples_per_second": 101.332,
|
| 710 |
+
"eval_steps_per_second": 0.8,
|
| 711 |
+
"step": 718
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 23.04,
|
| 715 |
+
"learning_rate": 2.5089605734767026e-05,
|
| 716 |
+
"loss": 0.2058,
|
| 717 |
"step": 720
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 23.36,
|
| 721 |
+
"learning_rate": 2.3894862604540025e-05,
|
| 722 |
+
"loss": 0.1959,
|
| 723 |
"step": 730
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 23.68,
|
| 727 |
+
"learning_rate": 2.270011947431302e-05,
|
| 728 |
+
"loss": 0.2099,
|
| 729 |
"step": 740
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 24.0,
|
| 733 |
+
"learning_rate": 2.1505376344086024e-05,
|
| 734 |
+
"loss": 0.1784,
|
| 735 |
+
"step": 750
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 24.0,
|
| 739 |
+
"eval_accuracy": 0.8455467869222097,
|
| 740 |
+
"eval_f1": 0.8432046349153324,
|
| 741 |
+
"eval_loss": 0.6166961193084717,
|
| 742 |
+
"eval_precision": 0.8421630294947093,
|
| 743 |
+
"eval_recall": 0.8455467869222097,
|
| 744 |
+
"eval_runtime": 17.9099,
|
| 745 |
+
"eval_samples_per_second": 99.051,
|
| 746 |
+
"eval_steps_per_second": 0.782,
|
| 747 |
"step": 750
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 24.32,
|
| 751 |
+
"learning_rate": 2.031063321385902e-05,
|
| 752 |
+
"loss": 0.2061,
|
| 753 |
"step": 760
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 24.64,
|
| 757 |
+
"learning_rate": 1.9115890083632022e-05,
|
| 758 |
+
"loss": 0.1931,
|
| 759 |
"step": 770
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 24.96,
|
| 763 |
+
"learning_rate": 1.7921146953405018e-05,
|
| 764 |
+
"loss": 0.1901,
|
| 765 |
"step": 780
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 24.99,
|
| 769 |
+
"eval_accuracy": 0.8438556933483653,
|
| 770 |
+
"eval_f1": 0.841616389445069,
|
| 771 |
+
"eval_loss": 0.6393969058990479,
|
| 772 |
+
"eval_precision": 0.841081716577,
|
| 773 |
+
"eval_recall": 0.8438556933483653,
|
| 774 |
+
"eval_runtime": 17.3226,
|
| 775 |
+
"eval_samples_per_second": 102.41,
|
| 776 |
+
"eval_steps_per_second": 0.808,
|
| 777 |
+
"step": 781
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 25.28,
|
| 781 |
+
"learning_rate": 1.6726403823178017e-05,
|
| 782 |
+
"loss": 0.1776,
|
| 783 |
"step": 790
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 25.6,
|
| 787 |
+
"learning_rate": 1.5531660692951016e-05,
|
| 788 |
+
"loss": 0.1737,
|
| 789 |
"step": 800
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 25.92,
|
| 793 |
+
"learning_rate": 1.4336917562724014e-05,
|
| 794 |
+
"loss": 0.1801,
|
| 795 |
"step": 810
|
| 796 |
},
|
| 797 |
{
|
| 798 |
+
"epoch": 25.98,
|
| 799 |
+
"eval_accuracy": 0.8348365276211951,
|
| 800 |
+
"eval_f1": 0.8341671282721944,
|
| 801 |
+
"eval_loss": 0.6667414903640747,
|
| 802 |
+
"eval_precision": 0.834899664272928,
|
| 803 |
+
"eval_recall": 0.8348365276211951,
|
| 804 |
+
"eval_runtime": 17.7502,
|
| 805 |
+
"eval_samples_per_second": 99.942,
|
| 806 |
+
"eval_steps_per_second": 0.789,
|
| 807 |
+
"step": 812
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 26.24,
|
| 811 |
+
"learning_rate": 1.3142174432497015e-05,
|
| 812 |
+
"loss": 0.1871,
|
| 813 |
"step": 820
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 26.56,
|
| 817 |
+
"learning_rate": 1.1947431302270013e-05,
|
| 818 |
+
"loss": 0.1822,
|
| 819 |
"step": 830
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 26.88,
|
| 823 |
+
"learning_rate": 1.0752688172043012e-05,
|
| 824 |
+
"loss": 0.1892,
|
| 825 |
"step": 840
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 26.98,
|
| 829 |
+
"eval_accuracy": 0.8393461104847801,
|
| 830 |
+
"eval_f1": 0.8393631075518366,
|
| 831 |
+
"eval_loss": 0.6335396766662598,
|
| 832 |
+
"eval_precision": 0.8409729646771817,
|
| 833 |
+
"eval_recall": 0.8393461104847801,
|
| 834 |
+
"eval_runtime": 17.5202,
|
| 835 |
+
"eval_samples_per_second": 101.255,
|
| 836 |
+
"eval_steps_per_second": 0.799,
|
| 837 |
+
"step": 843
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 27.2,
|
| 841 |
+
"learning_rate": 9.557945041816011e-06,
|
| 842 |
+
"loss": 0.1867,
|
| 843 |
"step": 850
|
| 844 |
},
|
| 845 |
{
|
| 846 |
+
"epoch": 27.52,
|
| 847 |
+
"learning_rate": 8.363201911589009e-06,
|
| 848 |
+
"loss": 0.1903,
|
| 849 |
"step": 860
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 27.84,
|
| 853 |
+
"learning_rate": 7.168458781362007e-06,
|
| 854 |
+
"loss": 0.1756,
|
| 855 |
"step": 870
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 28.0,
|
| 859 |
+
"eval_accuracy": 0.8455467869222097,
|
| 860 |
+
"eval_f1": 0.8440049278026865,
|
| 861 |
+
"eval_loss": 0.6430765986442566,
|
| 862 |
+
"eval_precision": 0.8439254080309072,
|
| 863 |
+
"eval_recall": 0.8455467869222097,
|
| 864 |
+
"eval_runtime": 17.9226,
|
| 865 |
+
"eval_samples_per_second": 98.981,
|
| 866 |
+
"eval_steps_per_second": 0.781,
|
| 867 |
+
"step": 875
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 28.16,
|
| 871 |
+
"learning_rate": 5.973715651135006e-06,
|
| 872 |
+
"loss": 0.1913,
|
| 873 |
"step": 880
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 28.48,
|
| 877 |
+
"learning_rate": 4.7789725209080055e-06,
|
| 878 |
+
"loss": 0.1764,
|
| 879 |
"step": 890
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 28.8,
|
| 883 |
+
"learning_rate": 3.5842293906810035e-06,
|
| 884 |
+
"loss": 0.1791,
|
| 885 |
"step": 900
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 28.99,
|
| 889 |
+
"eval_accuracy": 0.8444193912063134,
|
| 890 |
+
"eval_f1": 0.8427748522925546,
|
| 891 |
+
"eval_loss": 0.6444693207740784,
|
| 892 |
+
"eval_precision": 0.8423422781414887,
|
| 893 |
+
"eval_recall": 0.8444193912063134,
|
| 894 |
+
"eval_runtime": 17.6456,
|
| 895 |
+
"eval_samples_per_second": 100.535,
|
| 896 |
+
"eval_steps_per_second": 0.793,
|
| 897 |
+
"step": 906
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 29.12,
|
| 901 |
+
"learning_rate": 2.3894862604540028e-06,
|
| 902 |
+
"loss": 0.1792,
|
| 903 |
"step": 910
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 29.44,
|
| 907 |
+
"learning_rate": 1.1947431302270014e-06,
|
| 908 |
+
"loss": 0.1649,
|
| 909 |
"step": 920
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 29.76,
|
| 913 |
+
"learning_rate": 0.0,
|
| 914 |
+
"loss": 0.1728,
|
| 915 |
"step": 930
|
| 916 |
},
|
| 917 |
{
|
| 918 |
+
"epoch": 29.76,
|
| 919 |
+
"eval_accuracy": 0.8438556933483653,
|
| 920 |
+
"eval_f1": 0.8418033278629379,
|
| 921 |
+
"eval_loss": 0.6450782418251038,
|
| 922 |
+
"eval_precision": 0.8413926709950079,
|
| 923 |
+
"eval_recall": 0.8438556933483653,
|
| 924 |
+
"eval_runtime": 17.2969,
|
| 925 |
+
"eval_samples_per_second": 102.562,
|
| 926 |
+
"eval_steps_per_second": 0.809,
|
| 927 |
+
"step": 930
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 29.76,
|
| 931 |
+
"step": 930,
|
| 932 |
+
"total_flos": 1.3013814188701234e+20,
|
| 933 |
+
"train_loss": 0.37854229006716,
|
| 934 |
+
"train_runtime": 10019.4589,
|
| 935 |
+
"train_samples_per_second": 47.793,
|
| 936 |
+
"train_steps_per_second": 0.093
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 937 |
}
|
| 938 |
],
|
| 939 |
"logging_steps": 10,
|
| 940 |
+
"max_steps": 930,
|
| 941 |
"num_input_tokens_seen": 0,
|
| 942 |
"num_train_epochs": 30,
|
| 943 |
"save_steps": 500,
|
| 944 |
+
"total_flos": 1.3013814188701234e+20,
|
| 945 |
+
"train_batch_size": 128,
|
| 946 |
"trial_name": null,
|
| 947 |
"trial_params": null
|
| 948 |
}
|