Commit
•
d740899
1
Parent(s):
8d525cd
End of training
Browse files- README.md +3 -3
- all_results.json +11 -11
- eval_results.json +6 -6
- train_eval_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +672 -242
- val_eval_results.json +6 -6
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 [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 1.
|
| 37 |
-
- Accuracy: 0.
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
|
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
+
value: 0.9483333333333334
|
| 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 [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 1.0056
|
| 37 |
+
- Accuracy: 0.9483
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime": 4.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 2.
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 80.0,
|
| 3 |
+
"eval_accuracy": 0.9483333333333334,
|
| 4 |
+
"eval_loss": 1.0056288242340088,
|
| 5 |
+
"eval_runtime": 4.803,
|
| 6 |
+
"eval_samples_per_second": 249.845,
|
| 7 |
+
"eval_steps_per_second": 2.082,
|
| 8 |
+
"total_flos": 7.440697863438336e+18,
|
| 9 |
+
"train_loss": 1.5909362745285034,
|
| 10 |
+
"train_runtime": 1036.0381,
|
| 11 |
+
"train_samples_per_second": 115.826,
|
| 12 |
+
"train_steps_per_second": 0.193
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime": 3.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 80.0,
|
| 3 |
+
"eval_accuracy": 0.8638888888888889,
|
| 4 |
+
"eval_loss": 1.2210274934768677,
|
| 5 |
+
"eval_runtime": 3.2287,
|
| 6 |
+
"eval_samples_per_second": 223.003,
|
| 7 |
+
"eval_steps_per_second": 1.858
|
| 8 |
}
|
train_eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime": 4.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 2.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 80.0,
|
| 3 |
+
"eval_accuracy": 0.9483333333333334,
|
| 4 |
+
"eval_loss": 1.0056288242340088,
|
| 5 |
+
"eval_runtime": 4.803,
|
| 6 |
+
"eval_samples_per_second": 249.845,
|
| 7 |
+
"eval_steps_per_second": 2.082
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 80.0,
|
| 3 |
+
"total_flos": 7.440697863438336e+18,
|
| 4 |
+
"train_loss": 1.5909362745285034,
|
| 5 |
+
"train_runtime": 1036.0381,
|
| 6 |
+
"train_samples_per_second": 115.826,
|
| 7 |
+
"train_steps_per_second": 0.193
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,457 +1,887 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-inaturalist/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": 0.8,
|
| 13 |
-
"eval_accuracy": 0.
|
| 14 |
-
"eval_loss": 3.
|
| 15 |
-
"eval_runtime": 3.
|
| 16 |
-
"eval_samples_per_second":
|
| 17 |
-
"eval_steps_per_second": 1.
|
| 18 |
"step": 2
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 2.0,
|
| 22 |
-
"eval_accuracy": 0.
|
| 23 |
-
"eval_loss": 3.
|
| 24 |
-
"eval_runtime": 3.
|
| 25 |
-
"eval_samples_per_second":
|
| 26 |
-
"eval_steps_per_second": 1.
|
| 27 |
"step": 5
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 2.8,
|
| 31 |
-
"eval_accuracy": 0.
|
| 32 |
-
"eval_loss": 3.
|
| 33 |
-
"eval_runtime": 3.
|
| 34 |
-
"eval_samples_per_second":
|
| 35 |
-
"eval_steps_per_second": 1.
|
| 36 |
"step": 7
|
| 37 |
},
|
| 38 |
{
|
| 39 |
"epoch": 4.0,
|
| 40 |
-
"grad_norm": 0.
|
| 41 |
-
"learning_rate": 5e-05,
|
| 42 |
-
"loss": 3.
|
| 43 |
"step": 10
|
| 44 |
},
|
| 45 |
{
|
| 46 |
"epoch": 4.0,
|
| 47 |
-
"eval_accuracy": 0.
|
| 48 |
-
"eval_loss": 3.
|
| 49 |
-
"eval_runtime": 3.
|
| 50 |
-
"eval_samples_per_second":
|
| 51 |
-
"eval_steps_per_second": 1.
|
| 52 |
"step": 10
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 4.8,
|
| 56 |
-
"eval_accuracy": 0.
|
| 57 |
-
"eval_loss": 3.
|
| 58 |
-
"eval_runtime": 3.
|
| 59 |
-
"eval_samples_per_second":
|
| 60 |
-
"eval_steps_per_second": 1.
|
| 61 |
"step": 12
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 6.0,
|
| 65 |
-
"eval_accuracy": 0.
|
| 66 |
-
"eval_loss":
|
| 67 |
-
"eval_runtime": 3.
|
| 68 |
-
"eval_samples_per_second":
|
| 69 |
-
"eval_steps_per_second": 1.
|
| 70 |
"step": 15
|
| 71 |
},
|
| 72 |
{
|
| 73 |
"epoch": 6.8,
|
| 74 |
-
"eval_accuracy": 0.
|
| 75 |
-
"eval_loss":
|
| 76 |
-
"eval_runtime": 3.
|
| 77 |
-
"eval_samples_per_second":
|
| 78 |
-
"eval_steps_per_second": 1.
|
| 79 |
"step": 17
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 8.0,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss":
|
| 86 |
"step": 20
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 8.0,
|
| 90 |
-
"eval_accuracy": 0.
|
| 91 |
-
"eval_loss": 2.
|
| 92 |
-
"eval_runtime": 3.
|
| 93 |
-
"eval_samples_per_second":
|
| 94 |
-
"eval_steps_per_second": 1.
|
| 95 |
"step": 20
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 8.8,
|
| 99 |
-
"eval_accuracy": 0.
|
| 100 |
-
"eval_loss": 2.
|
| 101 |
-
"eval_runtime": 3.
|
| 102 |
-
"eval_samples_per_second":
|
| 103 |
-
"eval_steps_per_second": 1.
|
| 104 |
"step": 22
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 10.0,
|
| 108 |
-
"eval_accuracy": 0.
|
| 109 |
-
"eval_loss": 2.
|
| 110 |
-
"eval_runtime": 3.
|
| 111 |
-
"eval_samples_per_second":
|
| 112 |
-
"eval_steps_per_second": 1.
|
| 113 |
"step": 25
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"epoch": 10.8,
|
| 117 |
-
"eval_accuracy": 0.
|
| 118 |
-
"eval_loss": 2.
|
| 119 |
-
"eval_runtime": 3.
|
| 120 |
-
"eval_samples_per_second":
|
| 121 |
-
"eval_steps_per_second": 1.
|
| 122 |
"step": 27
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 12.0,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 2.
|
| 129 |
"step": 30
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 12.0,
|
| 133 |
-
"eval_accuracy": 0.
|
| 134 |
-
"eval_loss": 2.
|
| 135 |
-
"eval_runtime": 3.
|
| 136 |
-
"eval_samples_per_second":
|
| 137 |
-
"eval_steps_per_second": 1.
|
| 138 |
"step": 30
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"epoch": 12.8,
|
| 142 |
-
"eval_accuracy": 0.
|
| 143 |
-
"eval_loss": 2.
|
| 144 |
-
"eval_runtime": 3.
|
| 145 |
-
"eval_samples_per_second":
|
| 146 |
-
"eval_steps_per_second": 1.
|
| 147 |
"step": 32
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"epoch": 14.0,
|
| 151 |
-
"eval_accuracy": 0.
|
| 152 |
-
"eval_loss": 2.
|
| 153 |
-
"eval_runtime": 3.
|
| 154 |
-
"eval_samples_per_second":
|
| 155 |
-
"eval_steps_per_second": 1.
|
| 156 |
"step": 35
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 14.8,
|
| 160 |
-
"eval_accuracy": 0.
|
| 161 |
-
"eval_loss": 2.
|
| 162 |
-
"eval_runtime":
|
| 163 |
-
"eval_samples_per_second":
|
| 164 |
-
"eval_steps_per_second":
|
| 165 |
"step": 37
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 16.0,
|
| 169 |
-
"grad_norm": 0.
|
| 170 |
-
"learning_rate":
|
| 171 |
-
"loss": 2.
|
| 172 |
"step": 40
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 16.0,
|
| 176 |
-
"eval_accuracy": 0.
|
| 177 |
-
"eval_loss": 2.
|
| 178 |
-
"eval_runtime": 3.
|
| 179 |
-
"eval_samples_per_second":
|
| 180 |
-
"eval_steps_per_second": 1.
|
| 181 |
"step": 40
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 16.8,
|
| 185 |
-
"eval_accuracy": 0.
|
| 186 |
-
"eval_loss": 2.
|
| 187 |
-
"eval_runtime": 3.
|
| 188 |
-
"eval_samples_per_second":
|
| 189 |
-
"eval_steps_per_second": 1.
|
| 190 |
"step": 42
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 18.0,
|
| 194 |
-
"eval_accuracy": 0.
|
| 195 |
-
"eval_loss": 2.
|
| 196 |
-
"eval_runtime": 3.
|
| 197 |
-
"eval_samples_per_second":
|
| 198 |
-
"eval_steps_per_second": 1.
|
| 199 |
"step": 45
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 18.8,
|
| 203 |
-
"eval_accuracy": 0.
|
| 204 |
-
"eval_loss": 2.
|
| 205 |
-
"eval_runtime": 3.
|
| 206 |
-
"eval_samples_per_second":
|
| 207 |
-
"eval_steps_per_second": 1.
|
| 208 |
"step": 47
|
| 209 |
},
|
| 210 |
{
|
| 211 |
"epoch": 20.0,
|
| 212 |
-
"grad_norm": 0.
|
| 213 |
-
"learning_rate":
|
| 214 |
-
"loss": 2.
|
| 215 |
"step": 50
|
| 216 |
},
|
| 217 |
{
|
| 218 |
"epoch": 20.0,
|
| 219 |
-
"eval_accuracy": 0.
|
| 220 |
-
"eval_loss": 2.
|
| 221 |
-
"eval_runtime": 3.
|
| 222 |
-
"eval_samples_per_second":
|
| 223 |
-
"eval_steps_per_second": 1.
|
| 224 |
"step": 50
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 20.8,
|
| 228 |
-
"eval_accuracy": 0.
|
| 229 |
-
"eval_loss": 2.
|
| 230 |
-
"eval_runtime": 3.
|
| 231 |
-
"eval_samples_per_second": 235.
|
| 232 |
-
"eval_steps_per_second": 1.
|
| 233 |
"step": 52
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 22.0,
|
| 237 |
"eval_accuracy": 0.7930555555555555,
|
| 238 |
-
"eval_loss": 2.
|
| 239 |
-
"eval_runtime": 3.
|
| 240 |
-
"eval_samples_per_second":
|
| 241 |
-
"eval_steps_per_second": 1.
|
| 242 |
"step": 55
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 22.8,
|
| 246 |
-
"eval_accuracy": 0.
|
| 247 |
-
"eval_loss": 2.
|
| 248 |
-
"eval_runtime": 3.
|
| 249 |
-
"eval_samples_per_second": 235.
|
| 250 |
-
"eval_steps_per_second": 1.
|
| 251 |
"step": 57
|
| 252 |
},
|
| 253 |
{
|
| 254 |
"epoch": 24.0,
|
| 255 |
-
"grad_norm": 0.
|
| 256 |
-
"learning_rate":
|
| 257 |
-
"loss": 1.
|
| 258 |
"step": 60
|
| 259 |
},
|
| 260 |
{
|
| 261 |
"epoch": 24.0,
|
| 262 |
-
"eval_accuracy": 0.
|
| 263 |
-
"eval_loss":
|
| 264 |
-
"eval_runtime": 3.
|
| 265 |
-
"eval_samples_per_second":
|
| 266 |
-
"eval_steps_per_second": 1.
|
| 267 |
"step": 60
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 24.8,
|
| 271 |
-
"eval_accuracy": 0.
|
| 272 |
-
"eval_loss": 1.
|
| 273 |
-
"eval_runtime": 3.
|
| 274 |
-
"eval_samples_per_second":
|
| 275 |
-
"eval_steps_per_second": 1.
|
| 276 |
"step": 62
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 26.0,
|
| 280 |
-
"eval_accuracy": 0.
|
| 281 |
-
"eval_loss": 1.
|
| 282 |
-
"eval_runtime": 3.
|
| 283 |
-
"eval_samples_per_second":
|
| 284 |
-
"eval_steps_per_second": 1.
|
| 285 |
"step": 65
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 26.8,
|
| 289 |
-
"eval_accuracy": 0.
|
| 290 |
-
"eval_loss": 1.
|
| 291 |
-
"eval_runtime": 3.
|
| 292 |
-
"eval_samples_per_second":
|
| 293 |
-
"eval_steps_per_second": 1.
|
| 294 |
"step": 67
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 28.0,
|
| 298 |
-
"grad_norm": 0.
|
| 299 |
-
"learning_rate":
|
| 300 |
-
"loss": 1.
|
| 301 |
"step": 70
|
| 302 |
},
|
| 303 |
{
|
| 304 |
"epoch": 28.0,
|
| 305 |
-
"eval_accuracy": 0.
|
| 306 |
-
"eval_loss": 1.
|
| 307 |
-
"eval_runtime": 3.
|
| 308 |
-
"eval_samples_per_second":
|
| 309 |
-
"eval_steps_per_second": 1.
|
| 310 |
"step": 70
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 28.8,
|
| 314 |
-
"eval_accuracy": 0.
|
| 315 |
-
"eval_loss": 1.
|
| 316 |
-
"eval_runtime": 3.
|
| 317 |
-
"eval_samples_per_second": 233.
|
| 318 |
-
"eval_steps_per_second": 1.
|
| 319 |
"step": 72
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 30.0,
|
| 323 |
-
"eval_accuracy": 0.
|
| 324 |
-
"eval_loss": 1.
|
| 325 |
-
"eval_runtime": 3.
|
| 326 |
-
"eval_samples_per_second":
|
| 327 |
-
"eval_steps_per_second": 1.
|
| 328 |
"step": 75
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 30.8,
|
| 332 |
-
"eval_accuracy": 0.
|
| 333 |
-
"eval_loss": 1.
|
| 334 |
-
"eval_runtime": 3.
|
| 335 |
-
"eval_samples_per_second":
|
| 336 |
-
"eval_steps_per_second": 1.
|
| 337 |
"step": 77
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 32.0,
|
| 341 |
-
"grad_norm": 0.
|
| 342 |
-
"learning_rate":
|
| 343 |
-
"loss": 1.
|
| 344 |
"step": 80
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 32.0,
|
| 348 |
-
"eval_accuracy": 0.
|
| 349 |
-
"eval_loss": 1.
|
| 350 |
-
"eval_runtime": 3.
|
| 351 |
-
"eval_samples_per_second":
|
| 352 |
-
"eval_steps_per_second": 1.
|
| 353 |
"step": 80
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 32.8,
|
| 357 |
-
"eval_accuracy": 0.
|
| 358 |
-
"eval_loss": 1.
|
| 359 |
-
"eval_runtime": 3.
|
| 360 |
-
"eval_samples_per_second":
|
| 361 |
-
"eval_steps_per_second": 1.
|
| 362 |
"step": 82
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 34.0,
|
| 366 |
-
"eval_accuracy": 0.
|
| 367 |
-
"eval_loss": 1.
|
| 368 |
-
"eval_runtime": 3.
|
| 369 |
-
"eval_samples_per_second":
|
| 370 |
-
"eval_steps_per_second": 1.
|
| 371 |
"step": 85
|
| 372 |
},
|
| 373 |
{
|
| 374 |
"epoch": 34.8,
|
| 375 |
"eval_accuracy": 0.8347222222222223,
|
| 376 |
-
"eval_loss": 1.
|
| 377 |
-
"eval_runtime": 3.
|
| 378 |
-
"eval_samples_per_second":
|
| 379 |
-
"eval_steps_per_second": 1.
|
| 380 |
"step": 87
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 36.0,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 90
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 36.0,
|
| 391 |
-
"eval_accuracy": 0.
|
| 392 |
-
"eval_loss": 1.
|
| 393 |
-
"eval_runtime": 3.
|
| 394 |
-
"eval_samples_per_second":
|
| 395 |
-
"eval_steps_per_second": 1.
|
| 396 |
"step": 90
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 36.8,
|
| 400 |
-
"eval_accuracy": 0.
|
| 401 |
-
"eval_loss": 1.
|
| 402 |
-
"eval_runtime": 3.
|
| 403 |
-
"eval_samples_per_second":
|
| 404 |
-
"eval_steps_per_second": 1.
|
| 405 |
"step": 92
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 38.0,
|
| 409 |
-
"eval_accuracy": 0.
|
| 410 |
-
"eval_loss": 1.
|
| 411 |
-
"eval_runtime": 3.
|
| 412 |
-
"eval_samples_per_second":
|
| 413 |
-
"eval_steps_per_second": 1.
|
| 414 |
"step": 95
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 38.8,
|
| 418 |
-
"eval_accuracy": 0.
|
| 419 |
-
"eval_loss": 1.
|
| 420 |
-
"eval_runtime": 3.
|
| 421 |
-
"eval_samples_per_second":
|
| 422 |
-
"eval_steps_per_second": 1.
|
| 423 |
"step": 97
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 40.0,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 100
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 40.0,
|
| 434 |
-
"eval_accuracy": 0.
|
| 435 |
-
"eval_loss": 1.
|
| 436 |
-
"eval_runtime": 3.
|
| 437 |
-
"eval_samples_per_second":
|
| 438 |
-
"eval_steps_per_second": 1.
|
| 439 |
"step": 100
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch": 40.
|
| 443 |
-
"
|
| 444 |
-
"
|
| 445 |
-
"
|
| 446 |
-
"
|
| 447 |
-
"
|
| 448 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 449 |
}
|
| 450 |
],
|
| 451 |
"logging_steps": 10,
|
| 452 |
-
"max_steps":
|
| 453 |
"num_input_tokens_seen": 0,
|
| 454 |
-
"num_train_epochs":
|
| 455 |
"save_steps": 500,
|
| 456 |
"stateful_callbacks": {
|
| 457 |
"TrainerControl": {
|
|
@@ -465,7 +895,7 @@
|
|
| 465 |
"attributes": {}
|
| 466 |
}
|
| 467 |
},
|
| 468 |
-
"total_flos":
|
| 469 |
"train_batch_size": 128,
|
| 470 |
"trial_name": null,
|
| 471 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8638888888888889,
|
| 3 |
+
"best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-inaturalist/checkpoint-140",
|
| 4 |
+
"epoch": 80.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.8,
|
| 13 |
+
"eval_accuracy": 0.03333333333333333,
|
| 14 |
+
"eval_loss": 3.200693368911743,
|
| 15 |
+
"eval_runtime": 3.9107,
|
| 16 |
+
"eval_samples_per_second": 184.111,
|
| 17 |
+
"eval_steps_per_second": 1.534,
|
| 18 |
"step": 2
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 2.0,
|
| 22 |
+
"eval_accuracy": 0.044444444444444446,
|
| 23 |
+
"eval_loss": 3.188920021057129,
|
| 24 |
+
"eval_runtime": 3.5624,
|
| 25 |
+
"eval_samples_per_second": 202.11,
|
| 26 |
+
"eval_steps_per_second": 1.684,
|
| 27 |
"step": 5
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 2.8,
|
| 31 |
+
"eval_accuracy": 0.06388888888888888,
|
| 32 |
+
"eval_loss": 3.1746575832366943,
|
| 33 |
+
"eval_runtime": 3.358,
|
| 34 |
+
"eval_samples_per_second": 214.412,
|
| 35 |
+
"eval_steps_per_second": 1.787,
|
| 36 |
"step": 7
|
| 37 |
},
|
| 38 |
{
|
| 39 |
"epoch": 4.0,
|
| 40 |
+
"grad_norm": 0.6863054633140564,
|
| 41 |
+
"learning_rate": 2.5e-05,
|
| 42 |
+
"loss": 3.1888,
|
| 43 |
"step": 10
|
| 44 |
},
|
| 45 |
{
|
| 46 |
"epoch": 4.0,
|
| 47 |
+
"eval_accuracy": 0.10972222222222222,
|
| 48 |
+
"eval_loss": 3.144199848175049,
|
| 49 |
+
"eval_runtime": 3.3224,
|
| 50 |
+
"eval_samples_per_second": 216.708,
|
| 51 |
+
"eval_steps_per_second": 1.806,
|
| 52 |
"step": 10
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 4.8,
|
| 56 |
+
"eval_accuracy": 0.14583333333333334,
|
| 57 |
+
"eval_loss": 3.1183247566223145,
|
| 58 |
+
"eval_runtime": 3.3659,
|
| 59 |
+
"eval_samples_per_second": 213.913,
|
| 60 |
+
"eval_steps_per_second": 1.783,
|
| 61 |
"step": 12
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 6.0,
|
| 65 |
+
"eval_accuracy": 0.21944444444444444,
|
| 66 |
+
"eval_loss": 3.071033477783203,
|
| 67 |
+
"eval_runtime": 3.1013,
|
| 68 |
+
"eval_samples_per_second": 232.158,
|
| 69 |
+
"eval_steps_per_second": 1.935,
|
| 70 |
"step": 15
|
| 71 |
},
|
| 72 |
{
|
| 73 |
"epoch": 6.8,
|
| 74 |
+
"eval_accuracy": 0.30416666666666664,
|
| 75 |
+
"eval_loss": 3.0330796241760254,
|
| 76 |
+
"eval_runtime": 3.0337,
|
| 77 |
+
"eval_samples_per_second": 237.333,
|
| 78 |
+
"eval_steps_per_second": 1.978,
|
| 79 |
"step": 17
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 8.0,
|
| 83 |
+
"grad_norm": 0.7356524467468262,
|
| 84 |
+
"learning_rate": 5e-05,
|
| 85 |
+
"loss": 3.0673,
|
| 86 |
"step": 20
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 8.0,
|
| 90 |
+
"eval_accuracy": 0.4388888888888889,
|
| 91 |
+
"eval_loss": 2.96268367767334,
|
| 92 |
+
"eval_runtime": 3.2316,
|
| 93 |
+
"eval_samples_per_second": 222.8,
|
| 94 |
+
"eval_steps_per_second": 1.857,
|
| 95 |
"step": 20
|
| 96 |
},
|
| 97 |
{
|
| 98 |
"epoch": 8.8,
|
| 99 |
+
"eval_accuracy": 0.49444444444444446,
|
| 100 |
+
"eval_loss": 2.9109385013580322,
|
| 101 |
+
"eval_runtime": 3.1222,
|
| 102 |
+
"eval_samples_per_second": 230.605,
|
| 103 |
+
"eval_steps_per_second": 1.922,
|
| 104 |
"step": 22
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 10.0,
|
| 108 |
+
"eval_accuracy": 0.5763888888888888,
|
| 109 |
+
"eval_loss": 2.8359875679016113,
|
| 110 |
+
"eval_runtime": 3.0808,
|
| 111 |
+
"eval_samples_per_second": 233.708,
|
| 112 |
+
"eval_steps_per_second": 1.948,
|
| 113 |
"step": 25
|
| 114 |
},
|
| 115 |
{
|
| 116 |
"epoch": 10.8,
|
| 117 |
+
"eval_accuracy": 0.6055555555555555,
|
| 118 |
+
"eval_loss": 2.780921459197998,
|
| 119 |
+
"eval_runtime": 3.0798,
|
| 120 |
+
"eval_samples_per_second": 233.781,
|
| 121 |
+
"eval_steps_per_second": 1.948,
|
| 122 |
"step": 27
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 12.0,
|
| 126 |
+
"grad_norm": 0.76748126745224,
|
| 127 |
+
"learning_rate": 4.722222222222222e-05,
|
| 128 |
+
"loss": 2.8151,
|
| 129 |
"step": 30
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 12.0,
|
| 133 |
+
"eval_accuracy": 0.6541666666666667,
|
| 134 |
+
"eval_loss": 2.695770025253296,
|
| 135 |
+
"eval_runtime": 3.3541,
|
| 136 |
+
"eval_samples_per_second": 214.662,
|
| 137 |
+
"eval_steps_per_second": 1.789,
|
| 138 |
"step": 30
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"epoch": 12.8,
|
| 142 |
+
"eval_accuracy": 0.6763888888888889,
|
| 143 |
+
"eval_loss": 2.640125036239624,
|
| 144 |
+
"eval_runtime": 3.1085,
|
| 145 |
+
"eval_samples_per_second": 231.622,
|
| 146 |
+
"eval_steps_per_second": 1.93,
|
| 147 |
"step": 32
|
| 148 |
},
|
| 149 |
{
|
| 150 |
"epoch": 14.0,
|
| 151 |
+
"eval_accuracy": 0.6944444444444444,
|
| 152 |
+
"eval_loss": 2.558335304260254,
|
| 153 |
+
"eval_runtime": 3.0528,
|
| 154 |
+
"eval_samples_per_second": 235.852,
|
| 155 |
+
"eval_steps_per_second": 1.965,
|
| 156 |
"step": 35
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 14.8,
|
| 160 |
+
"eval_accuracy": 0.7083333333333334,
|
| 161 |
+
"eval_loss": 2.503415822982788,
|
| 162 |
+
"eval_runtime": 3.1925,
|
| 163 |
+
"eval_samples_per_second": 225.525,
|
| 164 |
+
"eval_steps_per_second": 1.879,
|
| 165 |
"step": 37
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 16.0,
|
| 169 |
+
"grad_norm": 0.8363860249519348,
|
| 170 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 171 |
+
"loss": 2.5143,
|
| 172 |
"step": 40
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 16.0,
|
| 176 |
+
"eval_accuracy": 0.7347222222222223,
|
| 177 |
+
"eval_loss": 2.4201643466949463,
|
| 178 |
+
"eval_runtime": 3.0856,
|
| 179 |
+
"eval_samples_per_second": 233.339,
|
| 180 |
+
"eval_steps_per_second": 1.944,
|
| 181 |
"step": 40
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 16.8,
|
| 185 |
+
"eval_accuracy": 0.7375,
|
| 186 |
+
"eval_loss": 2.3662188053131104,
|
| 187 |
+
"eval_runtime": 3.0872,
|
| 188 |
+
"eval_samples_per_second": 233.218,
|
| 189 |
+
"eval_steps_per_second": 1.943,
|
| 190 |
"step": 42
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 18.0,
|
| 194 |
+
"eval_accuracy": 0.7444444444444445,
|
| 195 |
+
"eval_loss": 2.2883973121643066,
|
| 196 |
+
"eval_runtime": 3.1262,
|
| 197 |
+
"eval_samples_per_second": 230.314,
|
| 198 |
+
"eval_steps_per_second": 1.919,
|
| 199 |
"step": 45
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 18.8,
|
| 203 |
+
"eval_accuracy": 0.7569444444444444,
|
| 204 |
+
"eval_loss": 2.237414598464966,
|
| 205 |
+
"eval_runtime": 3.0834,
|
| 206 |
+
"eval_samples_per_second": 233.51,
|
| 207 |
+
"eval_steps_per_second": 1.946,
|
| 208 |
"step": 47
|
| 209 |
},
|
| 210 |
{
|
| 211 |
"epoch": 20.0,
|
| 212 |
+
"grad_norm": 0.8517465591430664,
|
| 213 |
+
"learning_rate": 4.166666666666667e-05,
|
| 214 |
+
"loss": 2.2236,
|
| 215 |
"step": 50
|
| 216 |
},
|
| 217 |
{
|
| 218 |
"epoch": 20.0,
|
| 219 |
+
"eval_accuracy": 0.7777777777777778,
|
| 220 |
+
"eval_loss": 2.1632001399993896,
|
| 221 |
+
"eval_runtime": 3.0769,
|
| 222 |
+
"eval_samples_per_second": 234.001,
|
| 223 |
+
"eval_steps_per_second": 1.95,
|
| 224 |
"step": 50
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 20.8,
|
| 228 |
+
"eval_accuracy": 0.7833333333333333,
|
| 229 |
+
"eval_loss": 2.1174519062042236,
|
| 230 |
+
"eval_runtime": 3.0592,
|
| 231 |
+
"eval_samples_per_second": 235.356,
|
| 232 |
+
"eval_steps_per_second": 1.961,
|
| 233 |
"step": 52
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 22.0,
|
| 237 |
"eval_accuracy": 0.7930555555555555,
|
| 238 |
+
"eval_loss": 2.052760124206543,
|
| 239 |
+
"eval_runtime": 3.21,
|
| 240 |
+
"eval_samples_per_second": 224.301,
|
| 241 |
+
"eval_steps_per_second": 1.869,
|
| 242 |
"step": 55
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 22.8,
|
| 246 |
+
"eval_accuracy": 0.7958333333333333,
|
| 247 |
+
"eval_loss": 2.009880304336548,
|
| 248 |
+
"eval_runtime": 3.0534,
|
| 249 |
+
"eval_samples_per_second": 235.801,
|
| 250 |
+
"eval_steps_per_second": 1.965,
|
| 251 |
"step": 57
|
| 252 |
},
|
| 253 |
{
|
| 254 |
"epoch": 24.0,
|
| 255 |
+
"grad_norm": 0.8587987422943115,
|
| 256 |
+
"learning_rate": 3.888888888888889e-05,
|
| 257 |
+
"loss": 1.9677,
|
| 258 |
"step": 60
|
| 259 |
},
|
| 260 |
{
|
| 261 |
"epoch": 24.0,
|
| 262 |
+
"eval_accuracy": 0.8013888888888889,
|
| 263 |
+
"eval_loss": 1.9488461017608643,
|
| 264 |
+
"eval_runtime": 3.1017,
|
| 265 |
+
"eval_samples_per_second": 232.129,
|
| 266 |
+
"eval_steps_per_second": 1.934,
|
| 267 |
"step": 60
|
| 268 |
},
|
| 269 |
{
|
| 270 |
"epoch": 24.8,
|
| 271 |
+
"eval_accuracy": 0.8097222222222222,
|
| 272 |
+
"eval_loss": 1.9112929105758667,
|
| 273 |
+
"eval_runtime": 3.0366,
|
| 274 |
+
"eval_samples_per_second": 237.104,
|
| 275 |
+
"eval_steps_per_second": 1.976,
|
| 276 |
"step": 62
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 26.0,
|
| 280 |
+
"eval_accuracy": 0.8138888888888889,
|
| 281 |
+
"eval_loss": 1.8581663370132446,
|
| 282 |
+
"eval_runtime": 3.0592,
|
| 283 |
+
"eval_samples_per_second": 235.359,
|
| 284 |
+
"eval_steps_per_second": 1.961,
|
| 285 |
"step": 65
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 26.8,
|
| 289 |
+
"eval_accuracy": 0.8138888888888889,
|
| 290 |
+
"eval_loss": 1.8241873979568481,
|
| 291 |
+
"eval_runtime": 3.2249,
|
| 292 |
+
"eval_samples_per_second": 223.26,
|
| 293 |
+
"eval_steps_per_second": 1.861,
|
| 294 |
"step": 67
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 28.0,
|
| 298 |
+
"grad_norm": 0.8794375061988831,
|
| 299 |
+
"learning_rate": 3.611111111111111e-05,
|
| 300 |
+
"loss": 1.7467,
|
| 301 |
"step": 70
|
| 302 |
},
|
| 303 |
{
|
| 304 |
"epoch": 28.0,
|
| 305 |
+
"eval_accuracy": 0.8111111111111111,
|
| 306 |
+
"eval_loss": 1.7740373611450195,
|
| 307 |
+
"eval_runtime": 3.1535,
|
| 308 |
+
"eval_samples_per_second": 228.319,
|
| 309 |
+
"eval_steps_per_second": 1.903,
|
| 310 |
"step": 70
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 28.8,
|
| 314 |
+
"eval_accuracy": 0.8055555555555556,
|
| 315 |
+
"eval_loss": 1.7457906007766724,
|
| 316 |
+
"eval_runtime": 3.0886,
|
| 317 |
+
"eval_samples_per_second": 233.114,
|
| 318 |
+
"eval_steps_per_second": 1.943,
|
| 319 |
"step": 72
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 30.0,
|
| 323 |
+
"eval_accuracy": 0.8180555555555555,
|
| 324 |
+
"eval_loss": 1.7013169527053833,
|
| 325 |
+
"eval_runtime": 3.0478,
|
| 326 |
+
"eval_samples_per_second": 236.234,
|
| 327 |
+
"eval_steps_per_second": 1.969,
|
| 328 |
"step": 75
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 30.8,
|
| 332 |
+
"eval_accuracy": 0.8194444444444444,
|
| 333 |
+
"eval_loss": 1.6714116334915161,
|
| 334 |
+
"eval_runtime": 3.1668,
|
| 335 |
+
"eval_samples_per_second": 227.361,
|
| 336 |
+
"eval_steps_per_second": 1.895,
|
| 337 |
"step": 77
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 32.0,
|
| 341 |
+
"grad_norm": 0.9110932946205139,
|
| 342 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 343 |
+
"loss": 1.5765,
|
| 344 |
"step": 80
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 32.0,
|
| 348 |
+
"eval_accuracy": 0.8263888888888888,
|
| 349 |
+
"eval_loss": 1.631589412689209,
|
| 350 |
+
"eval_runtime": 3.3174,
|
| 351 |
+
"eval_samples_per_second": 217.038,
|
| 352 |
+
"eval_steps_per_second": 1.809,
|
| 353 |
"step": 80
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 32.8,
|
| 357 |
+
"eval_accuracy": 0.8236111111111111,
|
| 358 |
+
"eval_loss": 1.6082607507705688,
|
| 359 |
+
"eval_runtime": 3.107,
|
| 360 |
+
"eval_samples_per_second": 231.734,
|
| 361 |
+
"eval_steps_per_second": 1.931,
|
| 362 |
"step": 82
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 34.0,
|
| 366 |
+
"eval_accuracy": 0.8291666666666667,
|
| 367 |
+
"eval_loss": 1.5738186836242676,
|
| 368 |
+
"eval_runtime": 3.1316,
|
| 369 |
+
"eval_samples_per_second": 229.912,
|
| 370 |
+
"eval_steps_per_second": 1.916,
|
| 371 |
"step": 85
|
| 372 |
},
|
| 373 |
{
|
| 374 |
"epoch": 34.8,
|
| 375 |
"eval_accuracy": 0.8347222222222223,
|
| 376 |
+
"eval_loss": 1.553105354309082,
|
| 377 |
+
"eval_runtime": 3.2321,
|
| 378 |
+
"eval_samples_per_second": 222.767,
|
| 379 |
+
"eval_steps_per_second": 1.856,
|
| 380 |
"step": 87
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 36.0,
|
| 384 |
+
"grad_norm": 0.9191176891326904,
|
| 385 |
+
"learning_rate": 3.055555555555556e-05,
|
| 386 |
+
"loss": 1.4431,
|
| 387 |
"step": 90
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 36.0,
|
| 391 |
+
"eval_accuracy": 0.8430555555555556,
|
| 392 |
+
"eval_loss": 1.5228244066238403,
|
| 393 |
+
"eval_runtime": 3.0649,
|
| 394 |
+
"eval_samples_per_second": 234.921,
|
| 395 |
+
"eval_steps_per_second": 1.958,
|
| 396 |
"step": 90
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 36.8,
|
| 400 |
+
"eval_accuracy": 0.8444444444444444,
|
| 401 |
+
"eval_loss": 1.5046004056930542,
|
| 402 |
+
"eval_runtime": 3.2047,
|
| 403 |
+
"eval_samples_per_second": 224.669,
|
| 404 |
+
"eval_steps_per_second": 1.872,
|
| 405 |
"step": 92
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 38.0,
|
| 409 |
+
"eval_accuracy": 0.8472222222222222,
|
| 410 |
+
"eval_loss": 1.4780092239379883,
|
| 411 |
+
"eval_runtime": 3.0681,
|
| 412 |
+
"eval_samples_per_second": 234.673,
|
| 413 |
+
"eval_steps_per_second": 1.956,
|
| 414 |
"step": 95
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 38.8,
|
| 418 |
+
"eval_accuracy": 0.8458333333333333,
|
| 419 |
+
"eval_loss": 1.4607552289962769,
|
| 420 |
+
"eval_runtime": 3.3287,
|
| 421 |
+
"eval_samples_per_second": 216.303,
|
| 422 |
+
"eval_steps_per_second": 1.803,
|
| 423 |
"step": 97
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 40.0,
|
| 427 |
+
"grad_norm": 0.896637499332428,
|
| 428 |
+
"learning_rate": 2.777777777777778e-05,
|
| 429 |
+
"loss": 1.3049,
|
| 430 |
"step": 100
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 40.0,
|
| 434 |
+
"eval_accuracy": 0.8458333333333333,
|
| 435 |
+
"eval_loss": 1.435728907585144,
|
| 436 |
+
"eval_runtime": 3.1142,
|
| 437 |
+
"eval_samples_per_second": 231.201,
|
| 438 |
+
"eval_steps_per_second": 1.927,
|
| 439 |
"step": 100
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 40.8,
|
| 443 |
+
"eval_accuracy": 0.85,
|
| 444 |
+
"eval_loss": 1.4187902212142944,
|
| 445 |
+
"eval_runtime": 3.144,
|
| 446 |
+
"eval_samples_per_second": 229.01,
|
| 447 |
+
"eval_steps_per_second": 1.908,
|
| 448 |
+
"step": 102
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 42.0,
|
| 452 |
+
"eval_accuracy": 0.8527777777777777,
|
| 453 |
+
"eval_loss": 1.3949499130249023,
|
| 454 |
+
"eval_runtime": 3.0808,
|
| 455 |
+
"eval_samples_per_second": 233.709,
|
| 456 |
+
"eval_steps_per_second": 1.948,
|
| 457 |
+
"step": 105
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 42.8,
|
| 461 |
+
"eval_accuracy": 0.8527777777777777,
|
| 462 |
+
"eval_loss": 1.3807573318481445,
|
| 463 |
+
"eval_runtime": 3.1783,
|
| 464 |
+
"eval_samples_per_second": 226.537,
|
| 465 |
+
"eval_steps_per_second": 1.888,
|
| 466 |
+
"step": 107
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 44.0,
|
| 470 |
+
"grad_norm": 0.8933643102645874,
|
| 471 |
+
"learning_rate": 2.5e-05,
|
| 472 |
+
"loss": 1.2312,
|
| 473 |
+
"step": 110
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 44.0,
|
| 477 |
+
"eval_accuracy": 0.8458333333333333,
|
| 478 |
+
"eval_loss": 1.3636168241500854,
|
| 479 |
+
"eval_runtime": 3.356,
|
| 480 |
+
"eval_samples_per_second": 214.543,
|
| 481 |
+
"eval_steps_per_second": 1.788,
|
| 482 |
+
"step": 110
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 44.8,
|
| 486 |
+
"eval_accuracy": 0.8486111111111111,
|
| 487 |
+
"eval_loss": 1.3513400554656982,
|
| 488 |
+
"eval_runtime": 3.2597,
|
| 489 |
+
"eval_samples_per_second": 220.878,
|
| 490 |
+
"eval_steps_per_second": 1.841,
|
| 491 |
+
"step": 112
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 46.0,
|
| 495 |
+
"eval_accuracy": 0.8527777777777777,
|
| 496 |
+
"eval_loss": 1.3329037427902222,
|
| 497 |
+
"eval_runtime": 3.0929,
|
| 498 |
+
"eval_samples_per_second": 232.788,
|
| 499 |
+
"eval_steps_per_second": 1.94,
|
| 500 |
+
"step": 115
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 46.8,
|
| 504 |
+
"eval_accuracy": 0.8527777777777777,
|
| 505 |
+
"eval_loss": 1.319313883781433,
|
| 506 |
+
"eval_runtime": 3.1007,
|
| 507 |
+
"eval_samples_per_second": 232.203,
|
| 508 |
+
"eval_steps_per_second": 1.935,
|
| 509 |
+
"step": 117
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 48.0,
|
| 513 |
+
"grad_norm": 0.9017526507377625,
|
| 514 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 515 |
+
"loss": 1.1368,
|
| 516 |
+
"step": 120
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 48.0,
|
| 520 |
+
"eval_accuracy": 0.8527777777777777,
|
| 521 |
+
"eval_loss": 1.3025320768356323,
|
| 522 |
+
"eval_runtime": 3.1234,
|
| 523 |
+
"eval_samples_per_second": 230.52,
|
| 524 |
+
"eval_steps_per_second": 1.921,
|
| 525 |
+
"step": 120
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 48.8,
|
| 529 |
+
"eval_accuracy": 0.8541666666666666,
|
| 530 |
+
"eval_loss": 1.2945308685302734,
|
| 531 |
+
"eval_runtime": 3.0692,
|
| 532 |
+
"eval_samples_per_second": 234.59,
|
| 533 |
+
"eval_steps_per_second": 1.955,
|
| 534 |
+
"step": 122
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 50.0,
|
| 538 |
+
"eval_accuracy": 0.8527777777777777,
|
| 539 |
+
"eval_loss": 1.2819503545761108,
|
| 540 |
+
"eval_runtime": 3.1162,
|
| 541 |
+
"eval_samples_per_second": 231.052,
|
| 542 |
+
"eval_steps_per_second": 1.925,
|
| 543 |
+
"step": 125
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 50.8,
|
| 547 |
+
"eval_accuracy": 0.8569444444444444,
|
| 548 |
+
"eval_loss": 1.2704639434814453,
|
| 549 |
+
"eval_runtime": 3.2771,
|
| 550 |
+
"eval_samples_per_second": 219.706,
|
| 551 |
+
"eval_steps_per_second": 1.831,
|
| 552 |
+
"step": 127
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 52.0,
|
| 556 |
+
"grad_norm": 0.8930371999740601,
|
| 557 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 558 |
+
"loss": 1.0821,
|
| 559 |
+
"step": 130
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 52.0,
|
| 563 |
+
"eval_accuracy": 0.8583333333333333,
|
| 564 |
+
"eval_loss": 1.2615665197372437,
|
| 565 |
+
"eval_runtime": 3.1151,
|
| 566 |
+
"eval_samples_per_second": 231.134,
|
| 567 |
+
"eval_steps_per_second": 1.926,
|
| 568 |
+
"step": 130
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 52.8,
|
| 572 |
+
"eval_accuracy": 0.8555555555555555,
|
| 573 |
+
"eval_loss": 1.2545220851898193,
|
| 574 |
+
"eval_runtime": 3.0635,
|
| 575 |
+
"eval_samples_per_second": 235.024,
|
| 576 |
+
"eval_steps_per_second": 1.959,
|
| 577 |
+
"step": 132
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 54.0,
|
| 581 |
+
"eval_accuracy": 0.8541666666666666,
|
| 582 |
+
"eval_loss": 1.2422840595245361,
|
| 583 |
+
"eval_runtime": 3.0988,
|
| 584 |
+
"eval_samples_per_second": 232.348,
|
| 585 |
+
"eval_steps_per_second": 1.936,
|
| 586 |
+
"step": 135
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 54.8,
|
| 590 |
+
"eval_accuracy": 0.8597222222222223,
|
| 591 |
+
"eval_loss": 1.233168125152588,
|
| 592 |
+
"eval_runtime": 3.1982,
|
| 593 |
+
"eval_samples_per_second": 225.127,
|
| 594 |
+
"eval_steps_per_second": 1.876,
|
| 595 |
+
"step": 137
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 56.0,
|
| 599 |
+
"grad_norm": 0.8568278551101685,
|
| 600 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 601 |
+
"loss": 1.0232,
|
| 602 |
+
"step": 140
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 56.0,
|
| 606 |
+
"eval_accuracy": 0.8638888888888889,
|
| 607 |
+
"eval_loss": 1.2210274934768677,
|
| 608 |
+
"eval_runtime": 3.355,
|
| 609 |
+
"eval_samples_per_second": 214.604,
|
| 610 |
+
"eval_steps_per_second": 1.788,
|
| 611 |
+
"step": 140
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 56.8,
|
| 615 |
+
"eval_accuracy": 0.8625,
|
| 616 |
+
"eval_loss": 1.2160966396331787,
|
| 617 |
+
"eval_runtime": 3.127,
|
| 618 |
+
"eval_samples_per_second": 230.252,
|
| 619 |
+
"eval_steps_per_second": 1.919,
|
| 620 |
+
"step": 142
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 58.0,
|
| 624 |
+
"eval_accuracy": 0.8569444444444444,
|
| 625 |
+
"eval_loss": 1.209418535232544,
|
| 626 |
+
"eval_runtime": 3.1685,
|
| 627 |
+
"eval_samples_per_second": 227.239,
|
| 628 |
+
"eval_steps_per_second": 1.894,
|
| 629 |
+
"step": 145
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 58.8,
|
| 633 |
+
"eval_accuracy": 0.8541666666666666,
|
| 634 |
+
"eval_loss": 1.205717921257019,
|
| 635 |
+
"eval_runtime": 3.1918,
|
| 636 |
+
"eval_samples_per_second": 225.576,
|
| 637 |
+
"eval_steps_per_second": 1.88,
|
| 638 |
+
"step": 147
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 60.0,
|
| 642 |
+
"grad_norm": 0.8556333780288696,
|
| 643 |
+
"learning_rate": 1.388888888888889e-05,
|
| 644 |
+
"loss": 0.9814,
|
| 645 |
+
"step": 150
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 60.0,
|
| 649 |
+
"eval_accuracy": 0.85,
|
| 650 |
+
"eval_loss": 1.1972941160202026,
|
| 651 |
+
"eval_runtime": 3.1515,
|
| 652 |
+
"eval_samples_per_second": 228.465,
|
| 653 |
+
"eval_steps_per_second": 1.904,
|
| 654 |
+
"step": 150
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 60.8,
|
| 658 |
+
"eval_accuracy": 0.8486111111111111,
|
| 659 |
+
"eval_loss": 1.1918764114379883,
|
| 660 |
+
"eval_runtime": 3.349,
|
| 661 |
+
"eval_samples_per_second": 214.987,
|
| 662 |
+
"eval_steps_per_second": 1.792,
|
| 663 |
+
"step": 152
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 62.0,
|
| 667 |
+
"eval_accuracy": 0.8625,
|
| 668 |
+
"eval_loss": 1.1825212240219116,
|
| 669 |
+
"eval_runtime": 3.1401,
|
| 670 |
+
"eval_samples_per_second": 229.295,
|
| 671 |
+
"eval_steps_per_second": 1.911,
|
| 672 |
+
"step": 155
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 62.8,
|
| 676 |
+
"eval_accuracy": 0.8597222222222223,
|
| 677 |
+
"eval_loss": 1.179900050163269,
|
| 678 |
+
"eval_runtime": 3.239,
|
| 679 |
+
"eval_samples_per_second": 222.293,
|
| 680 |
+
"eval_steps_per_second": 1.852,
|
| 681 |
+
"step": 157
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 64.0,
|
| 685 |
+
"grad_norm": 0.827712893486023,
|
| 686 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 687 |
+
"loss": 0.9415,
|
| 688 |
+
"step": 160
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 64.0,
|
| 692 |
+
"eval_accuracy": 0.8597222222222223,
|
| 693 |
+
"eval_loss": 1.1716293096542358,
|
| 694 |
+
"eval_runtime": 3.1517,
|
| 695 |
+
"eval_samples_per_second": 228.447,
|
| 696 |
+
"eval_steps_per_second": 1.904,
|
| 697 |
+
"step": 160
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 64.8,
|
| 701 |
+
"eval_accuracy": 0.8625,
|
| 702 |
+
"eval_loss": 1.166538119316101,
|
| 703 |
+
"eval_runtime": 3.22,
|
| 704 |
+
"eval_samples_per_second": 223.601,
|
| 705 |
+
"eval_steps_per_second": 1.863,
|
| 706 |
+
"step": 162
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 66.0,
|
| 710 |
+
"eval_accuracy": 0.8638888888888889,
|
| 711 |
+
"eval_loss": 1.1611206531524658,
|
| 712 |
+
"eval_runtime": 3.323,
|
| 713 |
+
"eval_samples_per_second": 216.67,
|
| 714 |
+
"eval_steps_per_second": 1.806,
|
| 715 |
+
"step": 165
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 66.8,
|
| 719 |
+
"eval_accuracy": 0.8625,
|
| 720 |
+
"eval_loss": 1.1600357294082642,
|
| 721 |
+
"eval_runtime": 3.2253,
|
| 722 |
+
"eval_samples_per_second": 223.233,
|
| 723 |
+
"eval_steps_per_second": 1.86,
|
| 724 |
+
"step": 167
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 68.0,
|
| 728 |
+
"grad_norm": 0.9616327285766602,
|
| 729 |
+
"learning_rate": 8.333333333333334e-06,
|
| 730 |
+
"loss": 0.9135,
|
| 731 |
+
"step": 170
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 68.0,
|
| 735 |
+
"eval_accuracy": 0.8638888888888889,
|
| 736 |
+
"eval_loss": 1.1577341556549072,
|
| 737 |
+
"eval_runtime": 3.1095,
|
| 738 |
+
"eval_samples_per_second": 231.548,
|
| 739 |
+
"eval_steps_per_second": 1.93,
|
| 740 |
+
"step": 170
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 68.8,
|
| 744 |
+
"eval_accuracy": 0.8638888888888889,
|
| 745 |
+
"eval_loss": 1.1546900272369385,
|
| 746 |
+
"eval_runtime": 3.2289,
|
| 747 |
+
"eval_samples_per_second": 222.988,
|
| 748 |
+
"eval_steps_per_second": 1.858,
|
| 749 |
+
"step": 172
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 70.0,
|
| 753 |
+
"eval_accuracy": 0.8638888888888889,
|
| 754 |
+
"eval_loss": 1.1493217945098877,
|
| 755 |
+
"eval_runtime": 3.103,
|
| 756 |
+
"eval_samples_per_second": 232.033,
|
| 757 |
+
"eval_steps_per_second": 1.934,
|
| 758 |
+
"step": 175
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 70.8,
|
| 762 |
+
"eval_accuracy": 0.8611111111111112,
|
| 763 |
+
"eval_loss": 1.1463639736175537,
|
| 764 |
+
"eval_runtime": 3.3191,
|
| 765 |
+
"eval_samples_per_second": 216.926,
|
| 766 |
+
"eval_steps_per_second": 1.808,
|
| 767 |
+
"step": 177
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 72.0,
|
| 771 |
+
"grad_norm": 0.8919360637664795,
|
| 772 |
+
"learning_rate": 5.555555555555556e-06,
|
| 773 |
+
"loss": 0.8946,
|
| 774 |
+
"step": 180
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 72.0,
|
| 778 |
+
"eval_accuracy": 0.8555555555555555,
|
| 779 |
+
"eval_loss": 1.1423206329345703,
|
| 780 |
+
"eval_runtime": 3.0106,
|
| 781 |
+
"eval_samples_per_second": 239.152,
|
| 782 |
+
"eval_steps_per_second": 1.993,
|
| 783 |
+
"step": 180
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 72.8,
|
| 787 |
+
"eval_accuracy": 0.8611111111111112,
|
| 788 |
+
"eval_loss": 1.1402053833007812,
|
| 789 |
+
"eval_runtime": 3.4453,
|
| 790 |
+
"eval_samples_per_second": 208.981,
|
| 791 |
+
"eval_steps_per_second": 1.742,
|
| 792 |
+
"step": 182
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 74.0,
|
| 796 |
+
"eval_accuracy": 0.8583333333333333,
|
| 797 |
+
"eval_loss": 1.1375410556793213,
|
| 798 |
+
"eval_runtime": 3.1143,
|
| 799 |
+
"eval_samples_per_second": 231.188,
|
| 800 |
+
"eval_steps_per_second": 1.927,
|
| 801 |
+
"step": 185
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 74.8,
|
| 805 |
+
"eval_accuracy": 0.8597222222222223,
|
| 806 |
+
"eval_loss": 1.1360384225845337,
|
| 807 |
+
"eval_runtime": 3.2332,
|
| 808 |
+
"eval_samples_per_second": 222.688,
|
| 809 |
+
"eval_steps_per_second": 1.856,
|
| 810 |
+
"step": 187
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 76.0,
|
| 814 |
+
"grad_norm": 0.9335024356842041,
|
| 815 |
+
"learning_rate": 2.777777777777778e-06,
|
| 816 |
+
"loss": 0.8866,
|
| 817 |
+
"step": 190
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 76.0,
|
| 821 |
+
"eval_accuracy": 0.8625,
|
| 822 |
+
"eval_loss": 1.134353756904602,
|
| 823 |
+
"eval_runtime": 3.1331,
|
| 824 |
+
"eval_samples_per_second": 229.808,
|
| 825 |
+
"eval_steps_per_second": 1.915,
|
| 826 |
+
"step": 190
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 76.8,
|
| 830 |
+
"eval_accuracy": 0.8638888888888889,
|
| 831 |
+
"eval_loss": 1.1333543062210083,
|
| 832 |
+
"eval_runtime": 3.2199,
|
| 833 |
+
"eval_samples_per_second": 223.611,
|
| 834 |
+
"eval_steps_per_second": 1.863,
|
| 835 |
+
"step": 192
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 78.0,
|
| 839 |
+
"eval_accuracy": 0.8638888888888889,
|
| 840 |
+
"eval_loss": 1.1323890686035156,
|
| 841 |
+
"eval_runtime": 3.0815,
|
| 842 |
+
"eval_samples_per_second": 233.653,
|
| 843 |
+
"eval_steps_per_second": 1.947,
|
| 844 |
+
"step": 195
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 78.8,
|
| 848 |
+
"eval_accuracy": 0.8638888888888889,
|
| 849 |
+
"eval_loss": 1.1320444345474243,
|
| 850 |
+
"eval_runtime": 3.2638,
|
| 851 |
+
"eval_samples_per_second": 220.604,
|
| 852 |
+
"eval_steps_per_second": 1.838,
|
| 853 |
+
"step": 197
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 80.0,
|
| 857 |
+
"grad_norm": 0.8473469614982605,
|
| 858 |
+
"learning_rate": 0.0,
|
| 859 |
+
"loss": 0.8798,
|
| 860 |
+
"step": 200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 80.0,
|
| 864 |
+
"eval_accuracy": 0.8638888888888889,
|
| 865 |
+
"eval_loss": 1.1318904161453247,
|
| 866 |
+
"eval_runtime": 3.2096,
|
| 867 |
+
"eval_samples_per_second": 224.325,
|
| 868 |
+
"eval_steps_per_second": 1.869,
|
| 869 |
+
"step": 200
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 80.0,
|
| 873 |
+
"step": 200,
|
| 874 |
+
"total_flos": 7.440697863438336e+18,
|
| 875 |
+
"train_loss": 1.5909362745285034,
|
| 876 |
+
"train_runtime": 1036.0381,
|
| 877 |
+
"train_samples_per_second": 115.826,
|
| 878 |
+
"train_steps_per_second": 0.193
|
| 879 |
}
|
| 880 |
],
|
| 881 |
"logging_steps": 10,
|
| 882 |
+
"max_steps": 200,
|
| 883 |
"num_input_tokens_seen": 0,
|
| 884 |
+
"num_train_epochs": 100,
|
| 885 |
"save_steps": 500,
|
| 886 |
"stateful_callbacks": {
|
| 887 |
"TrainerControl": {
|
|
|
|
| 895 |
"attributes": {}
|
| 896 |
}
|
| 897 |
},
|
| 898 |
+
"total_flos": 7.440697863438336e+18,
|
| 899 |
"train_batch_size": 128,
|
| 900 |
"trial_name": null,
|
| 901 |
"trial_params": null
|
val_eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 1.
|
| 5 |
-
"eval_runtime": 3.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 80.0,
|
| 3 |
+
"eval_accuracy": 0.8638888888888889,
|
| 4 |
+
"eval_loss": 1.2210274934768677,
|
| 5 |
+
"eval_runtime": 3.2732,
|
| 6 |
+
"eval_samples_per_second": 219.966,
|
| 7 |
+
"eval_steps_per_second": 1.833
|
| 8 |
}
|