Model save
Browse files- README.md +2 -2
- all_results.json +6 -2
- runs/Apr04_11-37-03_ab9c0969efb8/events.out.tfevents.1712233586.ab9c0969efb8.34.1 +2 -2
- test_results.json +6 -2
- trainer_state.json +396 -231
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [MCG-NJU/videomae-base-finetuned-ssv2](https://huggingface.co/MCG-NJU/videomae-base-finetuned-ssv2) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 3.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [MCG-NJU/videomae-base-finetuned-ssv2](https://huggingface.co/MCG-NJU/videomae-base-finetuned-ssv2) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 3.1259
|
| 21 |
+
- Accuracy: 0.41
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
CHANGED
|
@@ -1,4 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.12,
|
| 3 |
+
"eval_accuracy": 0.375,
|
| 4 |
+
"eval_loss": 3.811370849609375,
|
| 5 |
+
"eval_runtime": 14.4555,
|
| 6 |
+
"eval_samples_per_second": 1.107,
|
| 7 |
+
"eval_steps_per_second": 0.553
|
| 8 |
}
|
runs/Apr04_11-37-03_ab9c0969efb8/events.out.tfevents.1712233586.ab9c0969efb8.34.1
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39ab275d0def41e8a2ce80e3c3037ca98ef5be9fe1d94c3b26fc0dfe97a66bf6
|
| 3 |
+
size 1057
|
test_results.json
CHANGED
|
@@ -1,4 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.12,
|
| 3 |
+
"eval_accuracy": 0.375,
|
| 4 |
+
"eval_loss": 3.811370849609375,
|
| 5 |
+
"eval_runtime": 14.4555,
|
| 6 |
+
"eval_samples_per_second": 1.107,
|
| 7 |
+
"eval_steps_per_second": 0.553
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,374 +1,539 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "videomae-base-finetuned-ssv2-finetuned-traffic-dataset-mae/checkpoint-
|
| 4 |
-
"epoch": 7.
|
| 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": 0.02,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 4.
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"
|
| 49 |
-
"
|
| 50 |
-
"
|
| 51 |
-
"eval_samples_per_second": 0.576,
|
| 52 |
-
"eval_steps_per_second": 0.288,
|
| 53 |
-
"step": 56
|
| 54 |
-
},
|
| 55 |
-
{
|
| 56 |
-
"epoch": 1.01,
|
| 57 |
-
"grad_norm": 16.718944549560547,
|
| 58 |
-
"learning_rate": 4.8138957816377175e-05,
|
| 59 |
-
"loss": 0.4883,
|
| 60 |
"step": 60
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch":
|
| 64 |
-
"grad_norm": 0.
|
| 65 |
-
"learning_rate": 4.
|
| 66 |
-
"loss": 0.
|
| 67 |
"step": 70
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 1.
|
| 78 |
-
"grad_norm":
|
| 79 |
-
"learning_rate": 4.
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 90
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 1.
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate": 4.
|
| 87 |
-
"loss": 0.
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"epoch": 1.
|
| 92 |
-
"grad_norm": 0.
|
| 93 |
-
"learning_rate": 4.
|
| 94 |
-
"loss": 0.
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch": 1.
|
| 99 |
-
"
|
| 100 |
-
"
|
| 101 |
-
"
|
| 102 |
-
"eval_samples_per_second": 0.583,
|
| 103 |
-
"eval_steps_per_second": 0.291,
|
| 104 |
-
"step": 112
|
| 105 |
-
},
|
| 106 |
-
{
|
| 107 |
-
"epoch": 2.02,
|
| 108 |
-
"grad_norm": 0.009479613974690437,
|
| 109 |
-
"learning_rate": 4.069478908188586e-05,
|
| 110 |
-
"loss": 0.2538,
|
| 111 |
"step": 120
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"grad_norm": 0.
|
| 116 |
-
"learning_rate":
|
| 117 |
-
"loss": 0.
|
| 118 |
"step": 130
|
| 119 |
},
|
| 120 |
{
|
| 121 |
-
"epoch":
|
| 122 |
-
"grad_norm": 0.
|
| 123 |
-
"learning_rate":
|
| 124 |
-
"loss": 0.
|
| 125 |
"step": 140
|
| 126 |
},
|
| 127 |
{
|
| 128 |
-
"epoch":
|
| 129 |
-
"grad_norm":
|
| 130 |
-
"learning_rate":
|
| 131 |
-
"loss": 0.
|
| 132 |
"step": 150
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"
|
| 137 |
-
"
|
| 138 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 139 |
"step": 160
|
| 140 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 141 |
{
|
| 142 |
"epoch": 2.12,
|
| 143 |
-
"eval_accuracy":
|
| 144 |
-
"eval_loss":
|
| 145 |
-
"eval_runtime":
|
| 146 |
-
"eval_samples_per_second": 0.
|
| 147 |
-
"eval_steps_per_second": 0.
|
| 148 |
-
"step":
|
| 149 |
},
|
| 150 |
{
|
| 151 |
"epoch": 3.0,
|
| 152 |
-
"grad_norm": 0.
|
| 153 |
-
"learning_rate": 3.
|
| 154 |
-
"loss": 0.
|
| 155 |
-
"step":
|
| 156 |
},
|
| 157 |
{
|
| 158 |
-
"epoch": 3.
|
| 159 |
-
"grad_norm": 0.
|
| 160 |
-
"learning_rate": 3.
|
| 161 |
-
"loss": 0.
|
| 162 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 163 |
},
|
| 164 |
{
|
| 165 |
"epoch": 3.05,
|
| 166 |
-
"grad_norm":
|
| 167 |
-
"learning_rate": 3.
|
| 168 |
-
"loss": 0.
|
| 169 |
-
"step":
|
| 170 |
},
|
| 171 |
{
|
| 172 |
"epoch": 3.07,
|
| 173 |
-
"grad_norm": 0.
|
| 174 |
-
"learning_rate": 3.
|
| 175 |
-
"loss": 0.
|
| 176 |
-
"step":
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 3.09,
|
| 180 |
-
"grad_norm": 0.
|
| 181 |
-
"learning_rate": 2.
|
| 182 |
-
"loss": 0.
|
| 183 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 184 |
},
|
| 185 |
{
|
| 186 |
"epoch": 3.12,
|
| 187 |
-
"grad_norm": 0.
|
| 188 |
-
"learning_rate": 2.
|
| 189 |
-
"loss": 0.
|
| 190 |
-
"step":
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 3.12,
|
| 194 |
-
"eval_accuracy":
|
| 195 |
-
"eval_loss":
|
| 196 |
-
"eval_runtime":
|
| 197 |
-
"eval_samples_per_second": 0.
|
| 198 |
-
"eval_steps_per_second": 0.
|
| 199 |
-
"step":
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 4.01,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate": 2.
|
| 205 |
-
"loss": 0.
|
| 206 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 4.04,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate": 2.
|
| 212 |
-
"loss": 0.
|
| 213 |
-
"step":
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 4.06,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate": 2.
|
| 219 |
-
"loss": 0.
|
| 220 |
-
"step":
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 4.08,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate": 2.
|
| 226 |
-
"loss": 0.
|
| 227 |
-
"step":
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch": 4.
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate": 2.
|
| 233 |
-
"loss": 0.
|
| 234 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 4.12,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate": 2.
|
| 240 |
-
"loss": 0.
|
| 241 |
-
"step":
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 4.12,
|
| 245 |
-
"eval_accuracy": 0.
|
| 246 |
-
"eval_loss":
|
| 247 |
-
"eval_runtime":
|
| 248 |
-
"eval_samples_per_second": 0.
|
| 249 |
-
"eval_steps_per_second": 0.
|
| 250 |
-
"step":
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 5.02,
|
| 254 |
-
"grad_norm":
|
| 255 |
-
"learning_rate": 1.
|
| 256 |
-
"loss": 0.
|
| 257 |
-
"step":
|
| 258 |
},
|
| 259 |
{
|
| 260 |
-
"epoch": 5.
|
| 261 |
-
"grad_norm": 0.
|
| 262 |
-
"learning_rate": 1.
|
| 263 |
-
"loss": 0.
|
| 264 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 5.07,
|
| 268 |
-
"grad_norm": 0.
|
| 269 |
-
"learning_rate": 1.
|
| 270 |
-
"loss": 0.
|
| 271 |
-
"step":
|
| 272 |
},
|
| 273 |
{
|
| 274 |
-
"epoch": 5.
|
| 275 |
-
"grad_norm": 0.
|
| 276 |
-
"learning_rate": 1.
|
| 277 |
-
"loss": 0.
|
| 278 |
-
"step":
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch": 5.
|
| 282 |
-
"grad_norm": 0.
|
| 283 |
-
"learning_rate": 1.
|
| 284 |
-
"loss": 0.
|
| 285 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 5.12,
|
| 289 |
-
"eval_accuracy": 0.
|
| 290 |
-
"eval_loss":
|
| 291 |
-
"eval_runtime":
|
| 292 |
-
"eval_samples_per_second": 0.
|
| 293 |
-
"eval_steps_per_second": 0.
|
| 294 |
-
"step":
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 6.01,
|
| 298 |
-
"grad_norm": 0.
|
| 299 |
-
"learning_rate": 1.
|
| 300 |
-
"loss": 0.
|
| 301 |
-
"step":
|
| 302 |
},
|
| 303 |
{
|
| 304 |
-
"epoch": 6.
|
| 305 |
-
"grad_norm": 0.
|
| 306 |
-
"learning_rate": 1.
|
| 307 |
-
"loss": 0.
|
| 308 |
-
"step":
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch": 6.
|
| 312 |
-
"grad_norm": 0.
|
| 313 |
-
"learning_rate": 1.
|
| 314 |
-
"loss": 0.
|
| 315 |
-
"step":
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch": 6.
|
| 319 |
-
"grad_norm": 0.
|
| 320 |
-
"learning_rate":
|
| 321 |
-
"loss": 0.
|
| 322 |
-
"step":
|
| 323 |
},
|
| 324 |
{
|
| 325 |
-
"epoch": 6.
|
| 326 |
-
"grad_norm": 0.
|
| 327 |
-
"learning_rate":
|
| 328 |
-
"loss": 0.
|
| 329 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 330 |
},
|
| 331 |
{
|
| 332 |
"epoch": 6.12,
|
| 333 |
-
"grad_norm": 0.
|
| 334 |
-
"learning_rate": 7.
|
| 335 |
-
"loss": 0.
|
| 336 |
-
"step":
|
| 337 |
},
|
| 338 |
{
|
| 339 |
"epoch": 6.12,
|
| 340 |
-
"eval_accuracy": 0.
|
| 341 |
-
"eval_loss": 0.
|
| 342 |
-
"eval_runtime":
|
| 343 |
-
"eval_samples_per_second": 0.
|
| 344 |
-
"eval_steps_per_second": 0.
|
| 345 |
-
"step":
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 7.01,
|
| 349 |
-
"
|
| 350 |
-
"
|
| 351 |
-
"
|
| 352 |
-
"
|
| 353 |
-
"eval_steps_per_second": 0.343,
|
| 354 |
-
"step": 397
|
| 355 |
},
|
| 356 |
{
|
| 357 |
-
"epoch": 7.
|
| 358 |
-
"
|
| 359 |
-
"
|
| 360 |
-
"
|
| 361 |
-
"
|
| 362 |
-
|
| 363 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 364 |
}
|
| 365 |
],
|
| 366 |
"logging_steps": 10,
|
| 367 |
-
"max_steps":
|
| 368 |
"num_input_tokens_seen": 0,
|
| 369 |
"num_train_epochs": 9223372036854775807,
|
| 370 |
"save_steps": 500,
|
| 371 |
-
"total_flos":
|
| 372 |
"train_batch_size": 2,
|
| 373 |
"trial_name": null,
|
| 374 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.927536231884058,
|
| 3 |
+
"best_model_checkpoint": "videomae-base-finetuned-ssv2-finetuned-traffic-dataset-mae/checkpoint-456",
|
| 4 |
+
"epoch": 7.125,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 608,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 10.86192798614502,
|
| 14 |
+
"learning_rate": 8.196721311475409e-06,
|
| 15 |
+
"loss": 0.6065,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.03,
|
| 20 |
+
"grad_norm": 5.969607830047607,
|
| 21 |
+
"learning_rate": 1.6393442622950818e-05,
|
| 22 |
+
"loss": 0.5224,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 7.2913336753845215,
|
| 28 |
+
"learning_rate": 2.459016393442623e-05,
|
| 29 |
+
"loss": 0.5223,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.07,
|
| 34 |
+
"grad_norm": 1.6693388223648071,
|
| 35 |
+
"learning_rate": 3.2786885245901635e-05,
|
| 36 |
+
"loss": 0.1888,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.08,
|
| 41 |
+
"grad_norm": 1.9622797966003418,
|
| 42 |
+
"learning_rate": 4.098360655737705e-05,
|
| 43 |
+
"loss": 0.2521,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.1,
|
| 48 |
+
"grad_norm": 1.1212589740753174,
|
| 49 |
+
"learning_rate": 4.918032786885246e-05,
|
| 50 |
+
"loss": 0.0138,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"grad_norm": 0.07918030023574829,
|
| 56 |
+
"learning_rate": 4.917733089579525e-05,
|
| 57 |
+
"loss": 0.0309,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.12,
|
| 62 |
+
"eval_accuracy": 0.9130434782608695,
|
| 63 |
+
"eval_loss": 0.30077555775642395,
|
| 64 |
+
"eval_runtime": 121.2516,
|
| 65 |
+
"eval_samples_per_second": 0.569,
|
| 66 |
+
"eval_steps_per_second": 0.289,
|
| 67 |
+
"step": 76
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.01,
|
| 71 |
+
"grad_norm": 0.09372598677873611,
|
| 72 |
+
"learning_rate": 4.826325411334552e-05,
|
| 73 |
+
"loss": 0.0042,
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 1.02,
|
| 78 |
+
"grad_norm": 0.016796614974737167,
|
| 79 |
+
"learning_rate": 4.73491773308958e-05,
|
| 80 |
+
"loss": 0.0007,
|
| 81 |
"step": 90
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.04,
|
| 85 |
+
"grad_norm": 0.020755194127559662,
|
| 86 |
+
"learning_rate": 4.643510054844607e-05,
|
| 87 |
+
"loss": 0.3019,
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
{
|
| 91 |
+
"epoch": 1.06,
|
| 92 |
+
"grad_norm": 0.011669596657156944,
|
| 93 |
+
"learning_rate": 4.5521023765996346e-05,
|
| 94 |
+
"loss": 0.0003,
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
+
"epoch": 1.07,
|
| 99 |
+
"grad_norm": 0.003766011679545045,
|
| 100 |
+
"learning_rate": 4.460694698354662e-05,
|
| 101 |
+
"loss": 0.0002,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 1.09,
|
| 106 |
+
"grad_norm": 0.08908724784851074,
|
| 107 |
+
"learning_rate": 4.369287020109689e-05,
|
| 108 |
+
"loss": 0.0002,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 1.11,
|
| 113 |
+
"grad_norm": 0.005499520804733038,
|
| 114 |
+
"learning_rate": 4.2778793418647164e-05,
|
| 115 |
+
"loss": 0.0002,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 1.12,
|
| 120 |
+
"grad_norm": 0.005470686126500368,
|
| 121 |
+
"learning_rate": 4.1864716636197444e-05,
|
| 122 |
+
"loss": 0.0002,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 1.12,
|
| 127 |
+
"eval_accuracy": 0.6666666666666666,
|
| 128 |
+
"eval_loss": 2.1030023097991943,
|
| 129 |
+
"eval_runtime": 100.6515,
|
| 130 |
+
"eval_samples_per_second": 0.686,
|
| 131 |
+
"eval_steps_per_second": 0.348,
|
| 132 |
+
"step": 152
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.01,
|
| 136 |
+
"grad_norm": 0.0026209524367004633,
|
| 137 |
+
"learning_rate": 4.095063985374772e-05,
|
| 138 |
+
"loss": 0.0001,
|
| 139 |
"step": 160
|
| 140 |
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.03,
|
| 143 |
+
"grad_norm": 0.009242965839803219,
|
| 144 |
+
"learning_rate": 4.003656307129799e-05,
|
| 145 |
+
"loss": 0.0001,
|
| 146 |
+
"step": 170
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.05,
|
| 150 |
+
"grad_norm": 0.0026453358586877584,
|
| 151 |
+
"learning_rate": 3.912248628884826e-05,
|
| 152 |
+
"loss": 0.0001,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.06,
|
| 157 |
+
"grad_norm": 0.0033517158590257168,
|
| 158 |
+
"learning_rate": 3.820840950639854e-05,
|
| 159 |
+
"loss": 0.0001,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.08,
|
| 164 |
+
"grad_norm": 0.003627925645560026,
|
| 165 |
+
"learning_rate": 3.7294332723948815e-05,
|
| 166 |
+
"loss": 0.0001,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.1,
|
| 171 |
+
"grad_norm": 0.002879067324101925,
|
| 172 |
+
"learning_rate": 3.638025594149909e-05,
|
| 173 |
+
"loss": 0.0001,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.11,
|
| 178 |
+
"grad_norm": 0.0036734691821038723,
|
| 179 |
+
"learning_rate": 3.546617915904936e-05,
|
| 180 |
+
"loss": 0.0001,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
{
|
| 184 |
"epoch": 2.12,
|
| 185 |
+
"eval_accuracy": 0.7101449275362319,
|
| 186 |
+
"eval_loss": 1.84578275680542,
|
| 187 |
+
"eval_runtime": 105.6436,
|
| 188 |
+
"eval_samples_per_second": 0.653,
|
| 189 |
+
"eval_steps_per_second": 0.331,
|
| 190 |
+
"step": 228
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 3.0,
|
| 194 |
+
"grad_norm": 0.0011968453181907535,
|
| 195 |
+
"learning_rate": 3.455210237659964e-05,
|
| 196 |
+
"loss": 0.0001,
|
| 197 |
+
"step": 230
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 3.02,
|
| 201 |
+
"grad_norm": 0.0024677019100636244,
|
| 202 |
+
"learning_rate": 3.3638025594149906e-05,
|
| 203 |
+
"loss": 0.0001,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.04,
|
| 208 |
+
"grad_norm": 0.0011525729205459356,
|
| 209 |
+
"learning_rate": 3.2723948811700186e-05,
|
| 210 |
+
"loss": 0.0001,
|
| 211 |
+
"step": 250
|
| 212 |
},
|
| 213 |
{
|
| 214 |
"epoch": 3.05,
|
| 215 |
+
"grad_norm": 0.003028567647561431,
|
| 216 |
+
"learning_rate": 3.180987202925046e-05,
|
| 217 |
+
"loss": 0.0001,
|
| 218 |
+
"step": 260
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 3.07,
|
| 222 |
+
"grad_norm": 0.0027707908302545547,
|
| 223 |
+
"learning_rate": 3.089579524680074e-05,
|
| 224 |
+
"loss": 0.0001,
|
| 225 |
+
"step": 270
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 3.09,
|
| 229 |
+
"grad_norm": 0.0017000396037474275,
|
| 230 |
+
"learning_rate": 2.9981718464351004e-05,
|
| 231 |
+
"loss": 0.0001,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.1,
|
| 236 |
+
"grad_norm": 0.002608270151540637,
|
| 237 |
+
"learning_rate": 2.906764168190128e-05,
|
| 238 |
+
"loss": 0.0001,
|
| 239 |
+
"step": 290
|
| 240 |
},
|
| 241 |
{
|
| 242 |
"epoch": 3.12,
|
| 243 |
+
"grad_norm": 0.0012394236400723457,
|
| 244 |
+
"learning_rate": 2.8153564899451557e-05,
|
| 245 |
+
"loss": 0.0,
|
| 246 |
+
"step": 300
|
| 247 |
},
|
| 248 |
{
|
| 249 |
"epoch": 3.12,
|
| 250 |
+
"eval_accuracy": 0.7391304347826086,
|
| 251 |
+
"eval_loss": 1.5200165510177612,
|
| 252 |
+
"eval_runtime": 117.7366,
|
| 253 |
+
"eval_samples_per_second": 0.586,
|
| 254 |
+
"eval_steps_per_second": 0.297,
|
| 255 |
+
"step": 304
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 4.01,
|
| 259 |
+
"grad_norm": 0.00229161255992949,
|
| 260 |
+
"learning_rate": 2.7239488117001826e-05,
|
| 261 |
+
"loss": 0.0001,
|
| 262 |
+
"step": 310
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 4.03,
|
| 266 |
+
"grad_norm": 0.0013108043931424618,
|
| 267 |
+
"learning_rate": 2.6325411334552102e-05,
|
| 268 |
+
"loss": 0.0001,
|
| 269 |
+
"step": 320
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 4.04,
|
| 273 |
+
"grad_norm": 0.0012365940492600203,
|
| 274 |
+
"learning_rate": 2.541133455210238e-05,
|
| 275 |
+
"loss": 0.0,
|
| 276 |
+
"step": 330
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 4.06,
|
| 280 |
+
"grad_norm": 0.002082814695313573,
|
| 281 |
+
"learning_rate": 2.449725776965265e-05,
|
| 282 |
+
"loss": 0.0001,
|
| 283 |
+
"step": 340
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 4.08,
|
| 287 |
+
"grad_norm": 0.0035416895989328623,
|
| 288 |
+
"learning_rate": 2.3583180987202927e-05,
|
| 289 |
+
"loss": 0.0001,
|
| 290 |
+
"step": 350
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 4.09,
|
| 294 |
+
"grad_norm": 0.0009142422350123525,
|
| 295 |
+
"learning_rate": 2.26691042047532e-05,
|
| 296 |
+
"loss": 0.0001,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.11,
|
| 301 |
+
"grad_norm": 0.0014322904171422124,
|
| 302 |
+
"learning_rate": 2.1755027422303476e-05,
|
| 303 |
+
"loss": 0.0001,
|
| 304 |
+
"step": 370
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 4.12,
|
| 308 |
+
"grad_norm": 0.0017242878675460815,
|
| 309 |
+
"learning_rate": 2.084095063985375e-05,
|
| 310 |
+
"loss": 0.0001,
|
| 311 |
+
"step": 380
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 4.12,
|
| 315 |
+
"eval_accuracy": 0.7536231884057971,
|
| 316 |
+
"eval_loss": 1.4568772315979004,
|
| 317 |
+
"eval_runtime": 106.8623,
|
| 318 |
+
"eval_samples_per_second": 0.646,
|
| 319 |
+
"eval_steps_per_second": 0.328,
|
| 320 |
+
"step": 380
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 5.02,
|
| 324 |
+
"grad_norm": 0.0011396174086257815,
|
| 325 |
+
"learning_rate": 1.9926873857404025e-05,
|
| 326 |
+
"loss": 0.0001,
|
| 327 |
+
"step": 390
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 5.03,
|
| 331 |
+
"grad_norm": 0.0009178342879749835,
|
| 332 |
+
"learning_rate": 1.90127970749543e-05,
|
| 333 |
+
"loss": 0.0,
|
| 334 |
+
"step": 400
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 5.05,
|
| 338 |
+
"grad_norm": 0.0034428227227181196,
|
| 339 |
+
"learning_rate": 1.809872029250457e-05,
|
| 340 |
+
"loss": 0.0,
|
| 341 |
+
"step": 410
|
| 342 |
},
|
| 343 |
{
|
| 344 |
"epoch": 5.07,
|
| 345 |
+
"grad_norm": 0.0011987154139205813,
|
| 346 |
+
"learning_rate": 1.7184643510054847e-05,
|
| 347 |
+
"loss": 0.0,
|
| 348 |
+
"step": 420
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 5.08,
|
| 352 |
+
"grad_norm": 0.000779572525061667,
|
| 353 |
+
"learning_rate": 1.627056672760512e-05,
|
| 354 |
+
"loss": 0.0,
|
| 355 |
+
"step": 430
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 5.1,
|
| 359 |
+
"grad_norm": 0.0013000366743654013,
|
| 360 |
+
"learning_rate": 1.5356489945155393e-05,
|
| 361 |
+
"loss": 0.0003,
|
| 362 |
+
"step": 440
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.12,
|
| 366 |
+
"grad_norm": 0.0019034247379750013,
|
| 367 |
+
"learning_rate": 1.4442413162705667e-05,
|
| 368 |
+
"loss": 0.0,
|
| 369 |
+
"step": 450
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 5.12,
|
| 373 |
+
"eval_accuracy": 0.927536231884058,
|
| 374 |
+
"eval_loss": 0.39408108592033386,
|
| 375 |
+
"eval_runtime": 103.5083,
|
| 376 |
+
"eval_samples_per_second": 0.667,
|
| 377 |
+
"eval_steps_per_second": 0.338,
|
| 378 |
+
"step": 456
|
| 379 |
},
|
| 380 |
{
|
| 381 |
"epoch": 6.01,
|
| 382 |
+
"grad_norm": 0.0016774114919826388,
|
| 383 |
+
"learning_rate": 1.3528336380255944e-05,
|
| 384 |
+
"loss": 0.0005,
|
| 385 |
+
"step": 460
|
| 386 |
},
|
| 387 |
{
|
| 388 |
+
"epoch": 6.02,
|
| 389 |
+
"grad_norm": 0.0013074069283902645,
|
| 390 |
+
"learning_rate": 1.2614259597806216e-05,
|
| 391 |
+
"loss": 0.0001,
|
| 392 |
+
"step": 470
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 6.04,
|
| 396 |
+
"grad_norm": 0.0015878825215622783,
|
| 397 |
+
"learning_rate": 1.1700182815356491e-05,
|
| 398 |
+
"loss": 0.0001,
|
| 399 |
+
"step": 480
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 6.06,
|
| 403 |
+
"grad_norm": 0.0008045001304708421,
|
| 404 |
+
"learning_rate": 1.0786106032906766e-05,
|
| 405 |
+
"loss": 0.0001,
|
| 406 |
+
"step": 490
|
| 407 |
},
|
| 408 |
{
|
| 409 |
+
"epoch": 6.07,
|
| 410 |
+
"grad_norm": 0.0012260900111868978,
|
| 411 |
+
"learning_rate": 9.872029250457038e-06,
|
| 412 |
+
"loss": 0.0002,
|
| 413 |
+
"step": 500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 6.09,
|
| 417 |
+
"grad_norm": 0.001507502980530262,
|
| 418 |
+
"learning_rate": 8.957952468007313e-06,
|
| 419 |
+
"loss": 0.0001,
|
| 420 |
+
"step": 510
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 6.11,
|
| 424 |
+
"grad_norm": 0.001074848580174148,
|
| 425 |
+
"learning_rate": 8.043875685557587e-06,
|
| 426 |
+
"loss": 0.0001,
|
| 427 |
+
"step": 520
|
| 428 |
},
|
| 429 |
{
|
| 430 |
"epoch": 6.12,
|
| 431 |
+
"grad_norm": 0.0012807522434741259,
|
| 432 |
+
"learning_rate": 7.129798903107861e-06,
|
| 433 |
+
"loss": 0.0001,
|
| 434 |
+
"step": 530
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 6.12,
|
| 438 |
+
"eval_accuracy": 0.8695652173913043,
|
| 439 |
+
"eval_loss": 0.9657596945762634,
|
| 440 |
+
"eval_runtime": 105.1521,
|
| 441 |
+
"eval_samples_per_second": 0.656,
|
| 442 |
+
"eval_steps_per_second": 0.333,
|
| 443 |
+
"step": 532
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 7.01,
|
| 447 |
+
"grad_norm": 0.0008385140681639314,
|
| 448 |
+
"learning_rate": 6.2157221206581355e-06,
|
| 449 |
+
"loss": 0.0,
|
| 450 |
+
"step": 540
|
|
|
|
|
|
|
| 451 |
},
|
| 452 |
{
|
| 453 |
+
"epoch": 7.03,
|
| 454 |
+
"grad_norm": 0.001427665469236672,
|
| 455 |
+
"learning_rate": 5.301645338208409e-06,
|
| 456 |
+
"loss": 0.0,
|
| 457 |
+
"step": 550
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.05,
|
| 461 |
+
"grad_norm": 0.001054079388268292,
|
| 462 |
+
"learning_rate": 4.387568555758684e-06,
|
| 463 |
+
"loss": 0.0001,
|
| 464 |
+
"step": 560
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.06,
|
| 468 |
+
"grad_norm": 0.0016839707968756557,
|
| 469 |
+
"learning_rate": 3.4734917733089582e-06,
|
| 470 |
+
"loss": 0.0001,
|
| 471 |
+
"step": 570
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.08,
|
| 475 |
+
"grad_norm": 0.0011065697763115168,
|
| 476 |
+
"learning_rate": 2.5594149908592323e-06,
|
| 477 |
+
"loss": 0.0,
|
| 478 |
+
"step": 580
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.1,
|
| 482 |
+
"grad_norm": 0.0027244570665061474,
|
| 483 |
+
"learning_rate": 1.6453382084095064e-06,
|
| 484 |
+
"loss": 0.0,
|
| 485 |
+
"step": 590
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.11,
|
| 489 |
+
"grad_norm": 0.002320781582966447,
|
| 490 |
+
"learning_rate": 7.312614259597807e-07,
|
| 491 |
+
"loss": 0.0001,
|
| 492 |
+
"step": 600
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.12,
|
| 496 |
+
"eval_accuracy": 0.8405797101449275,
|
| 497 |
+
"eval_loss": 0.983595609664917,
|
| 498 |
+
"eval_runtime": 105.1841,
|
| 499 |
+
"eval_samples_per_second": 0.656,
|
| 500 |
+
"eval_steps_per_second": 0.333,
|
| 501 |
+
"step": 608
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 7.12,
|
| 505 |
+
"step": 608,
|
| 506 |
+
"total_flos": 1.5152145407976407e+18,
|
| 507 |
+
"train_loss": 0.040272084387700156,
|
| 508 |
+
"train_runtime": 2933.2373,
|
| 509 |
+
"train_samples_per_second": 0.415,
|
| 510 |
+
"train_steps_per_second": 0.207
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 7.12,
|
| 514 |
+
"eval_accuracy": 0.375,
|
| 515 |
+
"eval_loss": 3.811370849609375,
|
| 516 |
+
"eval_runtime": 15.0567,
|
| 517 |
+
"eval_samples_per_second": 1.063,
|
| 518 |
+
"eval_steps_per_second": 0.531,
|
| 519 |
+
"step": 608
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 7.12,
|
| 523 |
+
"eval_accuracy": 0.375,
|
| 524 |
+
"eval_loss": 3.811370849609375,
|
| 525 |
+
"eval_runtime": 14.4555,
|
| 526 |
+
"eval_samples_per_second": 1.107,
|
| 527 |
+
"eval_steps_per_second": 0.553,
|
| 528 |
+
"step": 608
|
| 529 |
}
|
| 530 |
],
|
| 531 |
"logging_steps": 10,
|
| 532 |
+
"max_steps": 608,
|
| 533 |
"num_input_tokens_seen": 0,
|
| 534 |
"num_train_epochs": 9223372036854775807,
|
| 535 |
"save_steps": 500,
|
| 536 |
+
"total_flos": 1.5152145407976407e+18,
|
| 537 |
"train_batch_size": 2,
|
| 538 |
"trial_name": null,
|
| 539 |
"trial_params": null
|