End of training
Browse files- README.md +3 -3
- all_results.json +11 -6
- eval_results.json +6 -6
- runs/Dec09_18-28-48_c9ba084c85bf/events.out.tfevents.1733769728.c9ba084c85bf.5668.3 +3 -0
- train_results.json +6 -6
- trainer_state.json +609 -390
README.md
CHANGED
|
@@ -23,7 +23,7 @@ model-index:
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
-
value: 0.
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Accuracy: 0.
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
|
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
+
value: 0.8071570576540755
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 0.4697
|
| 37 |
+
- Accuracy: 0.8072
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
all_results.json
CHANGED
|
@@ -1,8 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 53.333333333333336,
|
| 3 |
+
"eval_accuracy": 0.8071570576540755,
|
| 4 |
+
"eval_loss": 0.46969088912010193,
|
| 5 |
+
"eval_runtime": 3.0372,
|
| 6 |
+
"eval_samples_per_second": 165.611,
|
| 7 |
+
"eval_steps_per_second": 5.268,
|
| 8 |
+
"total_flos": 1.140094502803243e+18,
|
| 9 |
+
"train_loss": 0.34168325927522447,
|
| 10 |
+
"train_runtime": 709.0525,
|
| 11 |
+
"train_samples_per_second": 72.773,
|
| 12 |
+
"train_steps_per_second": 0.508
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 5.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 53.333333333333336,
|
| 3 |
+
"eval_accuracy": 0.8071570576540755,
|
| 4 |
+
"eval_loss": 0.46969088912010193,
|
| 5 |
+
"eval_runtime": 3.0372,
|
| 6 |
+
"eval_samples_per_second": 165.611,
|
| 7 |
+
"eval_steps_per_second": 5.268
|
| 8 |
}
|
runs/Dec09_18-28-48_c9ba084c85bf/events.out.tfevents.1733769728.c9ba084c85bf.5668.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:566e64b29ae31d91cbd718c6da762c24c347a2cf668ad17054c84f497b0a6681
|
| 3 |
+
size 411
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 53.333333333333336,
|
| 3 |
+
"total_flos": 1.140094502803243e+18,
|
| 4 |
+
"train_loss": 0.34168325927522447,
|
| 5 |
+
"train_runtime": 709.0525,
|
| 6 |
+
"train_samples_per_second": 72.773,
|
| 7 |
+
"train_steps_per_second": 0.508
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,546 +1,765 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"eval_accuracy": 0.
|
| 14 |
-
"eval_loss": 0.
|
| 15 |
-
"eval_runtime":
|
| 16 |
-
"eval_samples_per_second":
|
| 17 |
-
"eval_steps_per_second":
|
| 18 |
"step": 6
|
| 19 |
},
|
| 20 |
{
|
| 21 |
-
"epoch": 1.
|
| 22 |
-
"grad_norm":
|
| 23 |
-
"learning_rate":
|
| 24 |
-
"loss": 0.
|
| 25 |
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
-
"epoch": 1.
|
| 29 |
-
"eval_accuracy": 0.
|
| 30 |
-
"eval_loss": 0.
|
| 31 |
-
"eval_runtime":
|
| 32 |
-
"eval_samples_per_second":
|
| 33 |
-
"eval_steps_per_second":
|
| 34 |
-
"step":
|
| 35 |
},
|
| 36 |
{
|
| 37 |
-
"epoch": 2.
|
| 38 |
-
"
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
"eval_steps_per_second": 6.564,
|
| 43 |
-
"step": 18
|
| 44 |
},
|
| 45 |
{
|
| 46 |
-
"epoch":
|
| 47 |
-
"
|
| 48 |
-
"
|
| 49 |
-
"
|
|
|
|
|
|
|
| 50 |
"step": 20
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 4.0,
|
| 54 |
-
"eval_accuracy": 0.
|
| 55 |
-
"eval_loss": 0.
|
| 56 |
-
"eval_runtime":
|
| 57 |
-
"eval_samples_per_second":
|
| 58 |
-
"eval_steps_per_second":
|
| 59 |
-
"step":
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 4.8,
|
| 63 |
-
"grad_norm": 52.168453216552734,
|
| 64 |
-
"learning_rate": 4.8611111111111115e-05,
|
| 65 |
-
"loss": 0.3901,
|
| 66 |
-
"step": 30
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 4.
|
| 70 |
-
"
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
| 74 |
-
"eval_steps_per_second": 6.551,
|
| 75 |
-
"step": 31
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"eval_accuracy": 0.
|
| 80 |
-
"eval_loss": 0.
|
| 81 |
-
"eval_runtime":
|
| 82 |
-
"eval_samples_per_second":
|
| 83 |
-
"eval_steps_per_second": 6.
|
| 84 |
-
"step":
|
| 85 |
},
|
| 86 |
{
|
| 87 |
-
"epoch":
|
| 88 |
-
"grad_norm":
|
| 89 |
-
"learning_rate": 4.
|
| 90 |
-
"loss": 0.
|
| 91 |
"step": 40
|
| 92 |
},
|
| 93 |
{
|
| 94 |
-
"epoch":
|
| 95 |
-
"eval_accuracy": 0.
|
| 96 |
-
"eval_loss": 0.
|
| 97 |
-
"eval_runtime":
|
| 98 |
-
"eval_samples_per_second":
|
| 99 |
-
"eval_steps_per_second":
|
| 100 |
-
"step":
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch":
|
| 104 |
-
"
|
| 105 |
-
"
|
| 106 |
-
"
|
| 107 |
-
"
|
|
|
|
|
|
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch":
|
| 111 |
-
"
|
| 112 |
-
"
|
| 113 |
-
"
|
| 114 |
-
"eval_samples_per_second": 181.688,
|
| 115 |
-
"eval_steps_per_second": 6.338,
|
| 116 |
"step": 50
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 8.
|
| 120 |
-
"eval_accuracy": 0.
|
| 121 |
-
"eval_loss": 0.
|
| 122 |
-
"eval_runtime":
|
| 123 |
-
"eval_samples_per_second":
|
| 124 |
-
"eval_steps_per_second":
|
| 125 |
-
"step":
|
| 126 |
},
|
| 127 |
{
|
| 128 |
-
"epoch":
|
| 129 |
-
"grad_norm":
|
| 130 |
-
"learning_rate": 4.
|
| 131 |
-
"loss": 0.
|
| 132 |
"step": 60
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"eval_accuracy": 0.
|
| 137 |
-
"eval_loss": 0.
|
| 138 |
-
"eval_runtime":
|
| 139 |
-
"eval_samples_per_second":
|
| 140 |
-
"eval_steps_per_second":
|
| 141 |
-
"step":
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"eval_accuracy": 0.
|
| 146 |
-
"eval_loss": 0.
|
| 147 |
-
"eval_runtime":
|
| 148 |
-
"eval_samples_per_second":
|
| 149 |
-
"eval_steps_per_second":
|
| 150 |
-
"step":
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch":
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 70
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch":
|
| 161 |
-
"eval_accuracy": 0.
|
| 162 |
-
"eval_loss": 0.
|
| 163 |
-
"eval_runtime":
|
| 164 |
-
"eval_samples_per_second":
|
| 165 |
-
"eval_steps_per_second":
|
| 166 |
-
"step":
|
| 167 |
},
|
| 168 |
{
|
| 169 |
-
"epoch":
|
| 170 |
-
"grad_norm":
|
| 171 |
-
"learning_rate":
|
| 172 |
-
"loss": 0.
|
| 173 |
"step": 80
|
| 174 |
},
|
| 175 |
{
|
| 176 |
-
"epoch": 12.
|
| 177 |
-
"eval_accuracy": 0.
|
| 178 |
-
"eval_loss": 0.
|
| 179 |
-
"eval_runtime":
|
| 180 |
-
"eval_samples_per_second":
|
| 181 |
-
"eval_steps_per_second":
|
| 182 |
"step": 81
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch":
|
| 186 |
-
"eval_accuracy": 0.
|
| 187 |
-
"eval_loss": 0.
|
| 188 |
-
"eval_runtime":
|
| 189 |
-
"eval_samples_per_second": 183.
|
| 190 |
-
"eval_steps_per_second":
|
| 191 |
"step": 87
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch":
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate":
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 90
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch":
|
| 202 |
-
"eval_accuracy": 0.
|
| 203 |
-
"eval_loss": 0.
|
| 204 |
-
"eval_runtime":
|
| 205 |
-
"eval_samples_per_second":
|
| 206 |
-
"eval_steps_per_second":
|
| 207 |
-
"step":
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"grad_norm":
|
| 212 |
-
"learning_rate":
|
| 213 |
-
"loss": 0.
|
| 214 |
"step": 100
|
| 215 |
},
|
| 216 |
{
|
| 217 |
-
"epoch":
|
| 218 |
-
"eval_accuracy": 0.
|
| 219 |
-
"eval_loss": 0.
|
| 220 |
-
"eval_runtime":
|
| 221 |
-
"eval_samples_per_second":
|
| 222 |
-
"eval_steps_per_second":
|
| 223 |
-
"step":
|
| 224 |
},
|
| 225 |
{
|
| 226 |
-
"epoch": 16.
|
| 227 |
-
"eval_accuracy": 0.
|
| 228 |
-
"eval_loss": 0.
|
| 229 |
-
"eval_runtime":
|
| 230 |
-
"eval_samples_per_second":
|
| 231 |
-
"eval_steps_per_second": 6.
|
| 232 |
-
"step":
|
| 233 |
-
},
|
| 234 |
-
{
|
| 235 |
-
"epoch":
|
| 236 |
-
"grad_norm":
|
| 237 |
-
"learning_rate": 3.
|
| 238 |
-
"loss": 0.
|
| 239 |
"step": 110
|
| 240 |
},
|
| 241 |
{
|
| 242 |
-
"epoch":
|
| 243 |
-
"eval_accuracy": 0.
|
| 244 |
-
"eval_loss": 0.
|
| 245 |
-
"eval_runtime":
|
| 246 |
-
"eval_samples_per_second":
|
| 247 |
-
"eval_steps_per_second":
|
| 248 |
-
"step":
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch":
|
| 252 |
-
"
|
| 253 |
-
"
|
| 254 |
-
"
|
| 255 |
-
"eval_samples_per_second": 176.154,
|
| 256 |
-
"eval_steps_per_second": 6.145,
|
| 257 |
-
"step": 118
|
| 258 |
-
},
|
| 259 |
-
{
|
| 260 |
-
"epoch": 19.2,
|
| 261 |
-
"grad_norm": 18.67655372619629,
|
| 262 |
-
"learning_rate": 2.777777777777778e-05,
|
| 263 |
-
"loss": 0.3382,
|
| 264 |
"step": 120
|
| 265 |
},
|
| 266 |
{
|
| 267 |
-
"epoch":
|
| 268 |
-
"eval_accuracy": 0.
|
| 269 |
-
"eval_loss": 0.
|
| 270 |
-
"eval_runtime":
|
| 271 |
-
"eval_samples_per_second":
|
| 272 |
-
"eval_steps_per_second":
|
| 273 |
-
"step":
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 20.8,
|
| 277 |
-
"grad_norm": 16.18861198425293,
|
| 278 |
-
"learning_rate": 2.5462962962962965e-05,
|
| 279 |
-
"loss": 0.3534,
|
| 280 |
-
"step": 130
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"epoch":
|
| 284 |
-
"eval_accuracy": 0.
|
| 285 |
-
"eval_loss": 0.
|
| 286 |
-
"eval_runtime":
|
| 287 |
-
"eval_samples_per_second":
|
| 288 |
-
"eval_steps_per_second":
|
| 289 |
-
"step":
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch":
|
| 293 |
-
"
|
| 294 |
-
"
|
| 295 |
-
"
|
| 296 |
-
"
|
| 297 |
-
"eval_steps_per_second": 6.111,
|
| 298 |
-
"step": 137
|
| 299 |
},
|
| 300 |
{
|
| 301 |
-
"epoch":
|
| 302 |
-
"
|
| 303 |
-
"
|
| 304 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 305 |
"step": 140
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch":
|
| 309 |
-
"eval_accuracy": 0.
|
| 310 |
-
"eval_loss": 0.
|
| 311 |
-
"eval_runtime":
|
| 312 |
-
"eval_samples_per_second":
|
| 313 |
-
"eval_steps_per_second":
|
| 314 |
-
"step":
|
| 315 |
},
|
| 316 |
{
|
| 317 |
-
"epoch":
|
| 318 |
-
"
|
| 319 |
-
"
|
| 320 |
-
"
|
| 321 |
-
"
|
|
|
|
|
|
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
-
"
|
| 326 |
-
"
|
| 327 |
-
"
|
| 328 |
-
"eval_samples_per_second": 179.607,
|
| 329 |
-
"eval_steps_per_second": 6.265,
|
| 330 |
"step": 150
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"epoch":
|
| 334 |
-
"eval_accuracy": 0.
|
| 335 |
-
"eval_loss": 0.
|
| 336 |
-
"eval_runtime":
|
| 337 |
-
"eval_samples_per_second":
|
| 338 |
-
"eval_steps_per_second":
|
| 339 |
-
"step":
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 160
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch":
|
| 350 |
-
"eval_accuracy": 0.
|
| 351 |
-
"eval_loss": 0.
|
| 352 |
-
"eval_runtime":
|
| 353 |
-
"eval_samples_per_second":
|
| 354 |
-
"eval_steps_per_second":
|
| 355 |
"step": 162
|
| 356 |
},
|
| 357 |
{
|
| 358 |
-
"epoch":
|
| 359 |
-
"eval_accuracy": 0.
|
| 360 |
-
"eval_loss": 0.
|
| 361 |
-
"eval_runtime":
|
| 362 |
-
"eval_samples_per_second":
|
| 363 |
-
"eval_steps_per_second":
|
| 364 |
"step": 168
|
| 365 |
},
|
| 366 |
{
|
| 367 |
-
"epoch":
|
| 368 |
-
"grad_norm":
|
| 369 |
-
"learning_rate":
|
| 370 |
-
"loss": 0.
|
| 371 |
"step": 170
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch":
|
| 375 |
-
"eval_accuracy": 0.
|
| 376 |
-
"eval_loss": 0.
|
| 377 |
-
"eval_runtime":
|
| 378 |
-
"eval_samples_per_second":
|
| 379 |
-
"eval_steps_per_second":
|
| 380 |
"step": 175
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch":
|
| 384 |
-
"grad_norm":
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 180
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch":
|
| 391 |
-
"eval_accuracy": 0.
|
| 392 |
-
"eval_loss": 0.
|
| 393 |
-
"eval_runtime":
|
| 394 |
-
"eval_samples_per_second":
|
| 395 |
-
"eval_steps_per_second":
|
| 396 |
-
"step":
|
| 397 |
},
|
| 398 |
{
|
| 399 |
-
"epoch":
|
| 400 |
-
"eval_accuracy": 0.
|
| 401 |
-
"eval_loss": 0.
|
| 402 |
-
"eval_runtime":
|
| 403 |
-
"eval_samples_per_second":
|
| 404 |
-
"eval_steps_per_second":
|
| 405 |
-
"step":
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"grad_norm":
|
| 410 |
-
"learning_rate":
|
| 411 |
-
"loss": 0.
|
| 412 |
"step": 190
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"epoch":
|
| 416 |
-
"eval_accuracy": 0.
|
| 417 |
-
"eval_loss": 0.
|
| 418 |
-
"eval_runtime":
|
| 419 |
-
"eval_samples_per_second":
|
| 420 |
-
"eval_steps_per_second":
|
| 421 |
-
"step":
|
| 422 |
},
|
| 423 |
{
|
| 424 |
-
"epoch":
|
| 425 |
-
"grad_norm":
|
| 426 |
-
"learning_rate":
|
| 427 |
-
"loss": 0.
|
| 428 |
"step": 200
|
| 429 |
},
|
| 430 |
{
|
| 431 |
-
"epoch":
|
| 432 |
-
"eval_accuracy": 0.
|
| 433 |
-
"eval_loss": 0.
|
| 434 |
-
"eval_runtime":
|
| 435 |
-
"eval_samples_per_second":
|
| 436 |
-
"eval_steps_per_second": 5.
|
| 437 |
-
"step":
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"eval_accuracy": 0.
|
| 442 |
-
"eval_loss": 0.
|
| 443 |
-
"eval_runtime":
|
| 444 |
-
"eval_samples_per_second":
|
| 445 |
-
"eval_steps_per_second":
|
| 446 |
-
"step":
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch":
|
| 450 |
-
"grad_norm":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 210
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"eval_accuracy": 0.
|
| 458 |
-
"eval_loss": 0.
|
| 459 |
-
"eval_runtime":
|
| 460 |
-
"eval_samples_per_second": 179.
|
| 461 |
-
"eval_steps_per_second":
|
| 462 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 463 |
},
|
| 464 |
{
|
| 465 |
-
"epoch":
|
| 466 |
-
"eval_accuracy": 0.
|
| 467 |
-
"eval_loss": 0.
|
| 468 |
-
"eval_runtime":
|
| 469 |
-
"eval_samples_per_second":
|
| 470 |
-
"eval_steps_per_second":
|
| 471 |
-
"step":
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"
|
| 476 |
-
"
|
| 477 |
-
"
|
| 478 |
-
"
|
|
|
|
|
|
|
| 479 |
},
|
| 480 |
{
|
| 481 |
-
"epoch":
|
| 482 |
-
"
|
| 483 |
-
"
|
| 484 |
-
"
|
| 485 |
-
"eval_samples_per_second": 173.392,
|
| 486 |
-
"eval_steps_per_second": 6.049,
|
| 487 |
-
"step": 225
|
| 488 |
-
},
|
| 489 |
-
{
|
| 490 |
-
"epoch": 36.8,
|
| 491 |
-
"grad_norm": 18.135772705078125,
|
| 492 |
-
"learning_rate": 2.3148148148148148e-06,
|
| 493 |
-
"loss": 0.2944,
|
| 494 |
"step": 230
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch":
|
| 498 |
-
"eval_accuracy": 0.
|
| 499 |
-
"eval_loss": 0.
|
| 500 |
-
"eval_runtime":
|
| 501 |
-
"eval_samples_per_second":
|
| 502 |
-
"eval_steps_per_second":
|
| 503 |
-
"step":
|
| 504 |
},
|
| 505 |
{
|
| 506 |
-
"epoch":
|
| 507 |
-
"
|
| 508 |
-
"
|
| 509 |
-
"
|
| 510 |
-
"
|
| 511 |
-
"eval_steps_per_second": 6.418,
|
| 512 |
-
"step": 237
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"
|
| 517 |
-
"
|
| 518 |
-
"
|
| 519 |
-
"
|
|
|
|
|
|
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"eval_accuracy": 0.
|
| 524 |
-
"eval_loss": 0.
|
| 525 |
-
"eval_runtime":
|
| 526 |
-
"eval_samples_per_second":
|
| 527 |
-
"eval_steps_per_second":
|
| 528 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch":
|
| 532 |
-
"
|
| 533 |
-
"
|
| 534 |
-
"
|
| 535 |
-
"
|
| 536 |
-
"
|
| 537 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 538 |
}
|
| 539 |
],
|
| 540 |
"logging_steps": 10,
|
| 541 |
-
"max_steps":
|
| 542 |
"num_input_tokens_seen": 0,
|
| 543 |
-
"num_train_epochs":
|
| 544 |
"save_steps": 500,
|
| 545 |
"stateful_callbacks": {
|
| 546 |
"TrainerControl": {
|
|
@@ -554,7 +773,7 @@
|
|
| 554 |
"attributes": {}
|
| 555 |
}
|
| 556 |
},
|
| 557 |
-
"total_flos":
|
| 558 |
"train_batch_size": 32,
|
| 559 |
"trial_name": null,
|
| 560 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8071570576540755,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-101",
|
| 4 |
+
"epoch": 53.333333333333336,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 360,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.8888888888888888,
|
| 13 |
+
"eval_accuracy": 0.6898608349900597,
|
| 14 |
+
"eval_loss": 0.6376240849494934,
|
| 15 |
+
"eval_runtime": 2.8661,
|
| 16 |
+
"eval_samples_per_second": 175.501,
|
| 17 |
+
"eval_steps_per_second": 5.583,
|
| 18 |
"step": 6
|
| 19 |
},
|
| 20 |
{
|
| 21 |
+
"epoch": 1.4814814814814814,
|
| 22 |
+
"grad_norm": 9.892236709594727,
|
| 23 |
+
"learning_rate": 1.388888888888889e-05,
|
| 24 |
+
"loss": 0.6757,
|
| 25 |
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
+
"epoch": 1.925925925925926,
|
| 29 |
+
"eval_accuracy": 0.6938369781312127,
|
| 30 |
+
"eval_loss": 0.6052560806274414,
|
| 31 |
+
"eval_runtime": 2.8417,
|
| 32 |
+
"eval_samples_per_second": 177.006,
|
| 33 |
+
"eval_steps_per_second": 5.63,
|
| 34 |
+
"step": 13
|
| 35 |
},
|
| 36 |
{
|
| 37 |
+
"epoch": 2.962962962962963,
|
| 38 |
+
"grad_norm": 8.984474182128906,
|
| 39 |
+
"learning_rate": 2.777777777777778e-05,
|
| 40 |
+
"loss": 0.5472,
|
| 41 |
+
"step": 20
|
|
|
|
|
|
|
| 42 |
},
|
| 43 |
{
|
| 44 |
+
"epoch": 2.962962962962963,
|
| 45 |
+
"eval_accuracy": 0.7256461232604374,
|
| 46 |
+
"eval_loss": 0.5903410315513611,
|
| 47 |
+
"eval_runtime": 2.7461,
|
| 48 |
+
"eval_samples_per_second": 183.169,
|
| 49 |
+
"eval_steps_per_second": 5.826,
|
| 50 |
"step": 20
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 4.0,
|
| 54 |
+
"eval_accuracy": 0.731610337972167,
|
| 55 |
+
"eval_loss": 0.5782421827316284,
|
| 56 |
+
"eval_runtime": 2.7061,
|
| 57 |
+
"eval_samples_per_second": 185.878,
|
| 58 |
+
"eval_steps_per_second": 5.913,
|
| 59 |
+
"step": 27
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 4.444444444444445,
|
| 63 |
+
"grad_norm": 13.92780590057373,
|
| 64 |
+
"learning_rate": 4.166666666666667e-05,
|
| 65 |
+
"loss": 0.4628,
|
| 66 |
+
"step": 30
|
|
|
|
|
|
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 4.888888888888889,
|
| 70 |
+
"eval_accuracy": 0.7455268389662028,
|
| 71 |
+
"eval_loss": 0.5979239344596863,
|
| 72 |
+
"eval_runtime": 2.6102,
|
| 73 |
+
"eval_samples_per_second": 192.706,
|
| 74 |
+
"eval_steps_per_second": 6.13,
|
| 75 |
+
"step": 33
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 5.925925925925926,
|
| 79 |
+
"grad_norm": 25.355466842651367,
|
| 80 |
+
"learning_rate": 4.938271604938271e-05,
|
| 81 |
+
"loss": 0.4181,
|
| 82 |
"step": 40
|
| 83 |
},
|
| 84 |
{
|
| 85 |
+
"epoch": 5.925925925925926,
|
| 86 |
+
"eval_accuracy": 0.7614314115308151,
|
| 87 |
+
"eval_loss": 0.5735302567481995,
|
| 88 |
+
"eval_runtime": 2.7235,
|
| 89 |
+
"eval_samples_per_second": 184.688,
|
| 90 |
+
"eval_steps_per_second": 5.875,
|
| 91 |
+
"step": 40
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 6.962962962962963,
|
| 95 |
+
"eval_accuracy": 0.7495029821073559,
|
| 96 |
+
"eval_loss": 0.5251761078834534,
|
| 97 |
+
"eval_runtime": 2.662,
|
| 98 |
+
"eval_samples_per_second": 188.956,
|
| 99 |
+
"eval_steps_per_second": 6.011,
|
| 100 |
+
"step": 47
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 7.407407407407407,
|
| 104 |
+
"grad_norm": 46.14825439453125,
|
| 105 |
+
"learning_rate": 4.783950617283951e-05,
|
| 106 |
+
"loss": 0.4079,
|
|
|
|
|
|
|
| 107 |
"step": 50
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 8.0,
|
| 111 |
+
"eval_accuracy": 0.7475149105367793,
|
| 112 |
+
"eval_loss": 0.5362666249275208,
|
| 113 |
+
"eval_runtime": 2.6877,
|
| 114 |
+
"eval_samples_per_second": 187.151,
|
| 115 |
+
"eval_steps_per_second": 5.953,
|
| 116 |
+
"step": 54
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 8.88888888888889,
|
| 120 |
+
"grad_norm": 22.70941925048828,
|
| 121 |
+
"learning_rate": 4.62962962962963e-05,
|
| 122 |
+
"loss": 0.4102,
|
| 123 |
"step": 60
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 8.88888888888889,
|
| 127 |
+
"eval_accuracy": 0.7495029821073559,
|
| 128 |
+
"eval_loss": 0.5288791060447693,
|
| 129 |
+
"eval_runtime": 2.7002,
|
| 130 |
+
"eval_samples_per_second": 186.285,
|
| 131 |
+
"eval_steps_per_second": 5.926,
|
| 132 |
+
"step": 60
|
| 133 |
},
|
| 134 |
{
|
| 135 |
+
"epoch": 9.925925925925926,
|
| 136 |
+
"eval_accuracy": 0.7534791252485089,
|
| 137 |
+
"eval_loss": 0.522729754447937,
|
| 138 |
+
"eval_runtime": 2.7539,
|
| 139 |
+
"eval_samples_per_second": 182.652,
|
| 140 |
+
"eval_steps_per_second": 5.81,
|
| 141 |
+
"step": 67
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 10.37037037037037,
|
| 145 |
+
"grad_norm": 92.61900329589844,
|
| 146 |
+
"learning_rate": 4.4753086419753084e-05,
|
| 147 |
+
"loss": 0.373,
|
| 148 |
"step": 70
|
| 149 |
},
|
| 150 |
{
|
| 151 |
+
"epoch": 10.962962962962964,
|
| 152 |
+
"eval_accuracy": 0.7773359840954275,
|
| 153 |
+
"eval_loss": 0.46767404675483704,
|
| 154 |
+
"eval_runtime": 2.7441,
|
| 155 |
+
"eval_samples_per_second": 183.301,
|
| 156 |
+
"eval_steps_per_second": 5.831,
|
| 157 |
+
"step": 74
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 11.851851851851851,
|
| 161 |
+
"grad_norm": 13.839974403381348,
|
| 162 |
+
"learning_rate": 4.3209876543209875e-05,
|
| 163 |
+
"loss": 0.3639,
|
| 164 |
"step": 80
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 12.0,
|
| 168 |
+
"eval_accuracy": 0.7813121272365805,
|
| 169 |
+
"eval_loss": 0.4978141188621521,
|
| 170 |
+
"eval_runtime": 2.7627,
|
| 171 |
+
"eval_samples_per_second": 182.069,
|
| 172 |
+
"eval_steps_per_second": 5.791,
|
| 173 |
"step": 81
|
| 174 |
},
|
| 175 |
{
|
| 176 |
+
"epoch": 12.88888888888889,
|
| 177 |
+
"eval_accuracy": 0.7992047713717694,
|
| 178 |
+
"eval_loss": 0.465084969997406,
|
| 179 |
+
"eval_runtime": 2.7393,
|
| 180 |
+
"eval_samples_per_second": 183.625,
|
| 181 |
+
"eval_steps_per_second": 5.841,
|
| 182 |
"step": 87
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 13.333333333333334,
|
| 186 |
+
"grad_norm": 12.678291320800781,
|
| 187 |
+
"learning_rate": 4.166666666666667e-05,
|
| 188 |
+
"loss": 0.3779,
|
| 189 |
"step": 90
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 13.925925925925926,
|
| 193 |
+
"eval_accuracy": 0.7912524850894632,
|
| 194 |
+
"eval_loss": 0.4737919569015503,
|
| 195 |
+
"eval_runtime": 2.7321,
|
| 196 |
+
"eval_samples_per_second": 184.107,
|
| 197 |
+
"eval_steps_per_second": 5.856,
|
| 198 |
+
"step": 94
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 14.814814814814815,
|
| 202 |
+
"grad_norm": 13.37376594543457,
|
| 203 |
+
"learning_rate": 4.012345679012346e-05,
|
| 204 |
+
"loss": 0.3476,
|
| 205 |
"step": 100
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 14.962962962962964,
|
| 209 |
+
"eval_accuracy": 0.8071570576540755,
|
| 210 |
+
"eval_loss": 0.46969088912010193,
|
| 211 |
+
"eval_runtime": 2.704,
|
| 212 |
+
"eval_samples_per_second": 186.017,
|
| 213 |
+
"eval_steps_per_second": 5.917,
|
| 214 |
+
"step": 101
|
| 215 |
},
|
| 216 |
{
|
| 217 |
+
"epoch": 16.0,
|
| 218 |
+
"eval_accuracy": 0.7952286282306164,
|
| 219 |
+
"eval_loss": 0.47190144658088684,
|
| 220 |
+
"eval_runtime": 2.651,
|
| 221 |
+
"eval_samples_per_second": 189.741,
|
| 222 |
+
"eval_steps_per_second": 6.036,
|
| 223 |
+
"step": 108
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 16.296296296296298,
|
| 227 |
+
"grad_norm": 36.415714263916016,
|
| 228 |
+
"learning_rate": 3.8580246913580246e-05,
|
| 229 |
+
"loss": 0.3467,
|
| 230 |
"step": 110
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 16.88888888888889,
|
| 234 |
+
"eval_accuracy": 0.7892644135188867,
|
| 235 |
+
"eval_loss": 0.4551805257797241,
|
| 236 |
+
"eval_runtime": 2.7016,
|
| 237 |
+
"eval_samples_per_second": 186.184,
|
| 238 |
+
"eval_steps_per_second": 5.922,
|
| 239 |
+
"step": 114
|
| 240 |
},
|
| 241 |
{
|
| 242 |
+
"epoch": 17.77777777777778,
|
| 243 |
+
"grad_norm": 21.425716400146484,
|
| 244 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 245 |
+
"loss": 0.3496,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 246 |
"step": 120
|
| 247 |
},
|
| 248 |
{
|
| 249 |
+
"epoch": 17.925925925925927,
|
| 250 |
+
"eval_accuracy": 0.7713717693836978,
|
| 251 |
+
"eval_loss": 0.5186495780944824,
|
| 252 |
+
"eval_runtime": 2.9189,
|
| 253 |
+
"eval_samples_per_second": 172.324,
|
| 254 |
+
"eval_steps_per_second": 5.481,
|
| 255 |
+
"step": 121
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 18.962962962962962,
|
| 259 |
+
"eval_accuracy": 0.7952286282306164,
|
| 260 |
+
"eval_loss": 0.4575484097003937,
|
| 261 |
+
"eval_runtime": 2.8214,
|
| 262 |
+
"eval_samples_per_second": 178.277,
|
| 263 |
+
"eval_steps_per_second": 5.671,
|
| 264 |
+
"step": 128
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"epoch": 19.25925925925926,
|
| 268 |
+
"grad_norm": 25.632354736328125,
|
| 269 |
+
"learning_rate": 3.5493827160493834e-05,
|
| 270 |
+
"loss": 0.3657,
|
| 271 |
+
"step": 130
|
|
|
|
|
|
|
| 272 |
},
|
| 273 |
{
|
| 274 |
+
"epoch": 20.0,
|
| 275 |
+
"eval_accuracy": 0.7793240556660039,
|
| 276 |
+
"eval_loss": 0.476377010345459,
|
| 277 |
+
"eval_runtime": 2.6973,
|
| 278 |
+
"eval_samples_per_second": 186.481,
|
| 279 |
+
"eval_steps_per_second": 5.932,
|
| 280 |
+
"step": 135
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 20.74074074074074,
|
| 284 |
+
"grad_norm": 37.57546615600586,
|
| 285 |
+
"learning_rate": 3.395061728395062e-05,
|
| 286 |
+
"loss": 0.3888,
|
| 287 |
"step": 140
|
| 288 |
},
|
| 289 |
{
|
| 290 |
+
"epoch": 20.88888888888889,
|
| 291 |
+
"eval_accuracy": 0.7713717693836978,
|
| 292 |
+
"eval_loss": 0.5008699297904968,
|
| 293 |
+
"eval_runtime": 2.7366,
|
| 294 |
+
"eval_samples_per_second": 183.807,
|
| 295 |
+
"eval_steps_per_second": 5.847,
|
| 296 |
+
"step": 141
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 21.925925925925927,
|
| 300 |
+
"eval_accuracy": 0.7813121272365805,
|
| 301 |
+
"eval_loss": 0.46731534600257874,
|
| 302 |
+
"eval_runtime": 2.8548,
|
| 303 |
+
"eval_samples_per_second": 176.192,
|
| 304 |
+
"eval_steps_per_second": 5.605,
|
| 305 |
+
"step": 148
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 22.22222222222222,
|
| 309 |
+
"grad_norm": 23.367645263671875,
|
| 310 |
+
"learning_rate": 3.240740740740741e-05,
|
| 311 |
+
"loss": 0.3236,
|
|
|
|
|
|
|
| 312 |
"step": 150
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 22.962962962962962,
|
| 316 |
+
"eval_accuracy": 0.7753479125248509,
|
| 317 |
+
"eval_loss": 0.493118554353714,
|
| 318 |
+
"eval_runtime": 2.8057,
|
| 319 |
+
"eval_samples_per_second": 179.277,
|
| 320 |
+
"eval_steps_per_second": 5.703,
|
| 321 |
+
"step": 155
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 23.703703703703702,
|
| 325 |
+
"grad_norm": 41.31706237792969,
|
| 326 |
+
"learning_rate": 3.08641975308642e-05,
|
| 327 |
+
"loss": 0.3179,
|
| 328 |
"step": 160
|
| 329 |
},
|
| 330 |
{
|
| 331 |
+
"epoch": 24.0,
|
| 332 |
+
"eval_accuracy": 0.7654075546719682,
|
| 333 |
+
"eval_loss": 0.4837174415588379,
|
| 334 |
+
"eval_runtime": 2.8061,
|
| 335 |
+
"eval_samples_per_second": 179.252,
|
| 336 |
+
"eval_steps_per_second": 5.702,
|
| 337 |
"step": 162
|
| 338 |
},
|
| 339 |
{
|
| 340 |
+
"epoch": 24.88888888888889,
|
| 341 |
+
"eval_accuracy": 0.7693836978131213,
|
| 342 |
+
"eval_loss": 0.4651556611061096,
|
| 343 |
+
"eval_runtime": 2.7801,
|
| 344 |
+
"eval_samples_per_second": 180.927,
|
| 345 |
+
"eval_steps_per_second": 5.755,
|
| 346 |
"step": 168
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 25.185185185185187,
|
| 350 |
+
"grad_norm": 48.94294357299805,
|
| 351 |
+
"learning_rate": 2.9320987654320992e-05,
|
| 352 |
+
"loss": 0.327,
|
| 353 |
"step": 170
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 25.925925925925927,
|
| 357 |
+
"eval_accuracy": 0.7495029821073559,
|
| 358 |
+
"eval_loss": 0.5107513070106506,
|
| 359 |
+
"eval_runtime": 2.746,
|
| 360 |
+
"eval_samples_per_second": 183.173,
|
| 361 |
+
"eval_steps_per_second": 5.827,
|
| 362 |
"step": 175
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 26.666666666666668,
|
| 366 |
+
"grad_norm": 17.72180938720703,
|
| 367 |
+
"learning_rate": 2.777777777777778e-05,
|
| 368 |
+
"loss": 0.3253,
|
| 369 |
"step": 180
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 26.962962962962962,
|
| 373 |
+
"eval_accuracy": 0.7833001988071571,
|
| 374 |
+
"eval_loss": 0.4423621594905853,
|
| 375 |
+
"eval_runtime": 2.851,
|
| 376 |
+
"eval_samples_per_second": 176.432,
|
| 377 |
+
"eval_steps_per_second": 5.612,
|
| 378 |
+
"step": 182
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 28.0,
|
| 382 |
+
"eval_accuracy": 0.7335984095427436,
|
| 383 |
+
"eval_loss": 0.5621975064277649,
|
| 384 |
+
"eval_runtime": 2.8441,
|
| 385 |
+
"eval_samples_per_second": 176.859,
|
| 386 |
+
"eval_steps_per_second": 5.626,
|
| 387 |
+
"step": 189
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 28.14814814814815,
|
| 391 |
+
"grad_norm": 26.241844177246094,
|
| 392 |
+
"learning_rate": 2.623456790123457e-05,
|
| 393 |
+
"loss": 0.3382,
|
| 394 |
"step": 190
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 28.88888888888889,
|
| 398 |
+
"eval_accuracy": 0.7693836978131213,
|
| 399 |
+
"eval_loss": 0.5067819952964783,
|
| 400 |
+
"eval_runtime": 2.8675,
|
| 401 |
+
"eval_samples_per_second": 175.415,
|
| 402 |
+
"eval_steps_per_second": 5.58,
|
| 403 |
+
"step": 195
|
| 404 |
},
|
| 405 |
{
|
| 406 |
+
"epoch": 29.62962962962963,
|
| 407 |
+
"grad_norm": 12.02319049835205,
|
| 408 |
+
"learning_rate": 2.4691358024691357e-05,
|
| 409 |
+
"loss": 0.331,
|
| 410 |
"step": 200
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 29.925925925925927,
|
| 414 |
+
"eval_accuracy": 0.7693836978131213,
|
| 415 |
+
"eval_loss": 0.45300325751304626,
|
| 416 |
+
"eval_runtime": 2.7755,
|
| 417 |
+
"eval_samples_per_second": 181.232,
|
| 418 |
+
"eval_steps_per_second": 5.765,
|
| 419 |
+
"step": 202
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 30.962962962962962,
|
| 423 |
+
"eval_accuracy": 0.731610337972167,
|
| 424 |
+
"eval_loss": 0.5205386877059937,
|
| 425 |
+
"eval_runtime": 2.8057,
|
| 426 |
+
"eval_samples_per_second": 179.276,
|
| 427 |
+
"eval_steps_per_second": 5.703,
|
| 428 |
+
"step": 209
|
| 429 |
},
|
| 430 |
{
|
| 431 |
+
"epoch": 31.11111111111111,
|
| 432 |
+
"grad_norm": 14.40517807006836,
|
| 433 |
+
"learning_rate": 2.314814814814815e-05,
|
| 434 |
+
"loss": 0.3302,
|
| 435 |
"step": 210
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 32.0,
|
| 439 |
+
"eval_accuracy": 0.7852882703777336,
|
| 440 |
+
"eval_loss": 0.4385511875152588,
|
| 441 |
+
"eval_runtime": 2.8045,
|
| 442 |
+
"eval_samples_per_second": 179.357,
|
| 443 |
+
"eval_steps_per_second": 5.705,
|
| 444 |
+
"step": 216
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 32.592592592592595,
|
| 448 |
+
"grad_norm": 18.574872970581055,
|
| 449 |
+
"learning_rate": 2.1604938271604937e-05,
|
| 450 |
+
"loss": 0.2972,
|
| 451 |
+
"step": 220
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 32.888888888888886,
|
| 455 |
+
"eval_accuracy": 0.7773359840954275,
|
| 456 |
+
"eval_loss": 0.5030562281608582,
|
| 457 |
+
"eval_runtime": 2.6994,
|
| 458 |
+
"eval_samples_per_second": 186.336,
|
| 459 |
+
"eval_steps_per_second": 5.927,
|
| 460 |
+
"step": 222
|
| 461 |
},
|
| 462 |
{
|
| 463 |
+
"epoch": 33.925925925925924,
|
| 464 |
+
"eval_accuracy": 0.757455268389662,
|
| 465 |
+
"eval_loss": 0.49088525772094727,
|
| 466 |
+
"eval_runtime": 2.709,
|
| 467 |
+
"eval_samples_per_second": 185.676,
|
| 468 |
+
"eval_steps_per_second": 5.906,
|
| 469 |
+
"step": 229
|
| 470 |
},
|
| 471 |
{
|
| 472 |
+
"epoch": 34.074074074074076,
|
| 473 |
+
"grad_norm": 21.681509017944336,
|
| 474 |
+
"learning_rate": 2.006172839506173e-05,
|
| 475 |
+
"loss": 0.3121,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 476 |
"step": 230
|
| 477 |
},
|
| 478 |
{
|
| 479 |
+
"epoch": 34.96296296296296,
|
| 480 |
+
"eval_accuracy": 0.7793240556660039,
|
| 481 |
+
"eval_loss": 0.47658684849739075,
|
| 482 |
+
"eval_runtime": 2.7374,
|
| 483 |
+
"eval_samples_per_second": 183.751,
|
| 484 |
+
"eval_steps_per_second": 5.845,
|
| 485 |
+
"step": 236
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 35.55555555555556,
|
| 489 |
+
"grad_norm": 29.247716903686523,
|
| 490 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 491 |
+
"loss": 0.2956,
|
| 492 |
+
"step": 240
|
|
|
|
|
|
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"epoch": 36.0,
|
| 496 |
+
"eval_accuracy": 0.7415506958250497,
|
| 497 |
+
"eval_loss": 0.5262213945388794,
|
| 498 |
+
"eval_runtime": 2.8701,
|
| 499 |
+
"eval_samples_per_second": 175.256,
|
| 500 |
+
"eval_steps_per_second": 5.575,
|
| 501 |
+
"step": 243
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 36.888888888888886,
|
| 505 |
+
"eval_accuracy": 0.731610337972167,
|
| 506 |
+
"eval_loss": 0.5373868942260742,
|
| 507 |
+
"eval_runtime": 2.878,
|
| 508 |
+
"eval_samples_per_second": 174.771,
|
| 509 |
+
"eval_steps_per_second": 5.559,
|
| 510 |
+
"step": 249
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 37.03703703703704,
|
| 514 |
+
"grad_norm": 13.55726146697998,
|
| 515 |
+
"learning_rate": 1.697530864197531e-05,
|
| 516 |
+
"loss": 0.2947,
|
| 517 |
+
"step": 250
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 37.925925925925924,
|
| 521 |
+
"eval_accuracy": 0.7673956262425448,
|
| 522 |
+
"eval_loss": 0.48880261182785034,
|
| 523 |
+
"eval_runtime": 2.7134,
|
| 524 |
+
"eval_samples_per_second": 185.373,
|
| 525 |
+
"eval_steps_per_second": 5.897,
|
| 526 |
+
"step": 256
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 38.51851851851852,
|
| 530 |
+
"grad_norm": 8.393943786621094,
|
| 531 |
+
"learning_rate": 1.54320987654321e-05,
|
| 532 |
+
"loss": 0.2662,
|
| 533 |
+
"step": 260
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 38.96296296296296,
|
| 537 |
+
"eval_accuracy": 0.7693836978131213,
|
| 538 |
+
"eval_loss": 0.4880698323249817,
|
| 539 |
+
"eval_runtime": 2.7354,
|
| 540 |
+
"eval_samples_per_second": 183.888,
|
| 541 |
+
"eval_steps_per_second": 5.849,
|
| 542 |
+
"step": 263
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 40.0,
|
| 546 |
+
"grad_norm": 24.057205200195312,
|
| 547 |
+
"learning_rate": 1.388888888888889e-05,
|
| 548 |
+
"loss": 0.2826,
|
| 549 |
+
"step": 270
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 40.0,
|
| 553 |
+
"eval_accuracy": 0.7892644135188867,
|
| 554 |
+
"eval_loss": 0.46687519550323486,
|
| 555 |
+
"eval_runtime": 2.7865,
|
| 556 |
+
"eval_samples_per_second": 180.514,
|
| 557 |
+
"eval_steps_per_second": 5.742,
|
| 558 |
+
"step": 270
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 40.888888888888886,
|
| 562 |
+
"eval_accuracy": 0.7972166998011928,
|
| 563 |
+
"eval_loss": 0.45914533734321594,
|
| 564 |
+
"eval_runtime": 2.8041,
|
| 565 |
+
"eval_samples_per_second": 179.381,
|
| 566 |
+
"eval_steps_per_second": 5.706,
|
| 567 |
+
"step": 276
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 41.48148148148148,
|
| 571 |
+
"grad_norm": 20.68549346923828,
|
| 572 |
+
"learning_rate": 1.2345679012345678e-05,
|
| 573 |
+
"loss": 0.2768,
|
| 574 |
+
"step": 280
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 41.925925925925924,
|
| 578 |
+
"eval_accuracy": 0.757455268389662,
|
| 579 |
+
"eval_loss": 0.5089908838272095,
|
| 580 |
+
"eval_runtime": 2.9736,
|
| 581 |
+
"eval_samples_per_second": 169.157,
|
| 582 |
+
"eval_steps_per_second": 5.381,
|
| 583 |
+
"step": 283
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 42.96296296296296,
|
| 587 |
+
"grad_norm": 9.9234619140625,
|
| 588 |
+
"learning_rate": 1.0802469135802469e-05,
|
| 589 |
+
"loss": 0.2836,
|
| 590 |
+
"step": 290
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 42.96296296296296,
|
| 594 |
+
"eval_accuracy": 0.7495029821073559,
|
| 595 |
+
"eval_loss": 0.5249876379966736,
|
| 596 |
+
"eval_runtime": 2.8985,
|
| 597 |
+
"eval_samples_per_second": 173.54,
|
| 598 |
+
"eval_steps_per_second": 5.52,
|
| 599 |
+
"step": 290
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 44.0,
|
| 603 |
+
"eval_accuracy": 0.7654075546719682,
|
| 604 |
+
"eval_loss": 0.4747855067253113,
|
| 605 |
+
"eval_runtime": 2.9057,
|
| 606 |
+
"eval_samples_per_second": 173.108,
|
| 607 |
+
"eval_steps_per_second": 5.506,
|
| 608 |
+
"step": 297
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 44.44444444444444,
|
| 612 |
+
"grad_norm": 19.76637840270996,
|
| 613 |
+
"learning_rate": 9.259259259259259e-06,
|
| 614 |
+
"loss": 0.2724,
|
| 615 |
+
"step": 300
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 44.888888888888886,
|
| 619 |
+
"eval_accuracy": 0.7833001988071571,
|
| 620 |
+
"eval_loss": 0.44288724660873413,
|
| 621 |
+
"eval_runtime": 2.7865,
|
| 622 |
+
"eval_samples_per_second": 180.51,
|
| 623 |
+
"eval_steps_per_second": 5.742,
|
| 624 |
+
"step": 303
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 45.925925925925924,
|
| 628 |
+
"grad_norm": 12.390064239501953,
|
| 629 |
+
"learning_rate": 7.71604938271605e-06,
|
| 630 |
+
"loss": 0.2498,
|
| 631 |
+
"step": 310
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 45.925925925925924,
|
| 635 |
+
"eval_accuracy": 0.7892644135188867,
|
| 636 |
+
"eval_loss": 0.4459961950778961,
|
| 637 |
+
"eval_runtime": 2.8017,
|
| 638 |
+
"eval_samples_per_second": 179.537,
|
| 639 |
+
"eval_steps_per_second": 5.711,
|
| 640 |
+
"step": 310
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 46.96296296296296,
|
| 644 |
+
"eval_accuracy": 0.7793240556660039,
|
| 645 |
+
"eval_loss": 0.4721997082233429,
|
| 646 |
+
"eval_runtime": 2.8302,
|
| 647 |
+
"eval_samples_per_second": 177.729,
|
| 648 |
+
"eval_steps_per_second": 5.653,
|
| 649 |
+
"step": 317
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 47.407407407407405,
|
| 653 |
+
"grad_norm": 11.559773445129395,
|
| 654 |
+
"learning_rate": 6.172839506172839e-06,
|
| 655 |
+
"loss": 0.2893,
|
| 656 |
+
"step": 320
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 48.0,
|
| 660 |
+
"eval_accuracy": 0.7713717693836978,
|
| 661 |
+
"eval_loss": 0.47993555665016174,
|
| 662 |
+
"eval_runtime": 2.9229,
|
| 663 |
+
"eval_samples_per_second": 172.087,
|
| 664 |
+
"eval_steps_per_second": 5.474,
|
| 665 |
+
"step": 324
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 48.888888888888886,
|
| 669 |
+
"grad_norm": 11.171250343322754,
|
| 670 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 671 |
+
"loss": 0.2618,
|
| 672 |
+
"step": 330
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 48.888888888888886,
|
| 676 |
+
"eval_accuracy": 0.7713717693836978,
|
| 677 |
+
"eval_loss": 0.4849596321582794,
|
| 678 |
+
"eval_runtime": 2.8546,
|
| 679 |
+
"eval_samples_per_second": 176.209,
|
| 680 |
+
"eval_steps_per_second": 5.605,
|
| 681 |
+
"step": 330
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 49.925925925925924,
|
| 685 |
+
"eval_accuracy": 0.7495029821073559,
|
| 686 |
+
"eval_loss": 0.5151545405387878,
|
| 687 |
+
"eval_runtime": 2.8154,
|
| 688 |
+
"eval_samples_per_second": 178.658,
|
| 689 |
+
"eval_steps_per_second": 5.683,
|
| 690 |
+
"step": 337
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 50.370370370370374,
|
| 694 |
+
"grad_norm": 16.287992477416992,
|
| 695 |
+
"learning_rate": 3.0864197530864196e-06,
|
| 696 |
+
"loss": 0.2664,
|
| 697 |
+
"step": 340
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 50.96296296296296,
|
| 701 |
+
"eval_accuracy": 0.7395626242544732,
|
| 702 |
+
"eval_loss": 0.5347036123275757,
|
| 703 |
+
"eval_runtime": 2.8326,
|
| 704 |
+
"eval_samples_per_second": 177.574,
|
| 705 |
+
"eval_steps_per_second": 5.648,
|
| 706 |
+
"step": 344
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 51.851851851851855,
|
| 710 |
+
"grad_norm": 15.289400100708008,
|
| 711 |
+
"learning_rate": 1.5432098765432098e-06,
|
| 712 |
+
"loss": 0.27,
|
| 713 |
+
"step": 350
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 52.0,
|
| 717 |
+
"eval_accuracy": 0.7415506958250497,
|
| 718 |
+
"eval_loss": 0.5342876315116882,
|
| 719 |
+
"eval_runtime": 2.8365,
|
| 720 |
+
"eval_samples_per_second": 177.329,
|
| 721 |
+
"eval_steps_per_second": 5.641,
|
| 722 |
+
"step": 351
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 52.888888888888886,
|
| 726 |
+
"eval_accuracy": 0.7415506958250497,
|
| 727 |
+
"eval_loss": 0.5330411195755005,
|
| 728 |
+
"eval_runtime": 2.823,
|
| 729 |
+
"eval_samples_per_second": 178.181,
|
| 730 |
+
"eval_steps_per_second": 5.668,
|
| 731 |
+
"step": 357
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 53.333333333333336,
|
| 735 |
+
"grad_norm": 15.15584945678711,
|
| 736 |
+
"learning_rate": 0.0,
|
| 737 |
+
"loss": 0.2539,
|
| 738 |
+
"step": 360
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 53.333333333333336,
|
| 742 |
+
"eval_accuracy": 0.7395626242544732,
|
| 743 |
+
"eval_loss": 0.5319550037384033,
|
| 744 |
+
"eval_runtime": 2.8191,
|
| 745 |
+
"eval_samples_per_second": 178.427,
|
| 746 |
+
"eval_steps_per_second": 5.676,
|
| 747 |
+
"step": 360
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 53.333333333333336,
|
| 751 |
+
"step": 360,
|
| 752 |
+
"total_flos": 1.140094502803243e+18,
|
| 753 |
+
"train_loss": 0.34168325927522447,
|
| 754 |
+
"train_runtime": 709.0525,
|
| 755 |
+
"train_samples_per_second": 72.773,
|
| 756 |
+
"train_steps_per_second": 0.508
|
| 757 |
}
|
| 758 |
],
|
| 759 |
"logging_steps": 10,
|
| 760 |
+
"max_steps": 360,
|
| 761 |
"num_input_tokens_seen": 0,
|
| 762 |
+
"num_train_epochs": 60,
|
| 763 |
"save_steps": 500,
|
| 764 |
"stateful_callbacks": {
|
| 765 |
"TrainerControl": {
|
|
|
|
| 773 |
"attributes": {}
|
| 774 |
}
|
| 775 |
},
|
| 776 |
+
"total_flos": 1.140094502803243e+18,
|
| 777 |
"train_batch_size": 32,
|
| 778 |
"trial_name": null,
|
| 779 |
"trial_params": null
|