Zheng Li
commited on
End of training
Browse files- README.md +4 -3
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/May15_02-50-35_cs-Precision-7960-Tower/events.out.tfevents.1747292652.cs-Precision-7960-Tower.145708.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +1268 -502
README.md
CHANGED
|
@@ -3,6 +3,7 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- superb
|
|
@@ -23,7 +24,7 @@ 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 +34,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Accuracy: 0.
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
| 6 |
+
- audio-classification
|
| 7 |
- generated_from_trainer
|
| 8 |
datasets:
|
| 9 |
- superb
|
|
|
|
| 24 |
metrics:
|
| 25 |
- name: Accuracy
|
| 26 |
type: accuracy
|
| 27 |
+
value: 0.9826419535157399
|
| 28 |
---
|
| 29 |
|
| 30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 34 |
|
| 35 |
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
|
| 36 |
It achieves the following results on the evaluation set:
|
| 37 |
+
- Loss: 0.0954
|
| 38 |
+
- Accuracy: 0.9826
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 5.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.972769953051643,
|
| 3 |
+
"eval_accuracy": 0.9826419535157399,
|
| 4 |
+
"eval_loss": 0.09542840719223022,
|
| 5 |
+
"eval_runtime": 5.5034,
|
| 6 |
+
"eval_samples_per_second": 1235.228,
|
| 7 |
+
"eval_steps_per_second": 38.703,
|
| 8 |
+
"total_flos": 3.767900833756416e+18,
|
| 9 |
+
"train_loss": 0.5178930132572812,
|
| 10 |
+
"train_runtime": 760.6734,
|
| 11 |
+
"train_samples_per_second": 537.355,
|
| 12 |
+
"train_steps_per_second": 2.798
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 5.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.972769953051643,
|
| 3 |
+
"eval_accuracy": 0.9826419535157399,
|
| 4 |
+
"eval_loss": 0.09542840719223022,
|
| 5 |
+
"eval_runtime": 5.5034,
|
| 6 |
+
"eval_samples_per_second": 1235.228,
|
| 7 |
+
"eval_steps_per_second": 38.703
|
| 8 |
}
|
runs/May15_02-50-35_cs-Precision-7960-Tower/events.out.tfevents.1747292652.cs-Precision-7960-Tower.145708.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:285677df20f84a09e3da60b417862a127a2e430c9b3f05ed1cbf7e75d8393024
|
| 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":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.972769953051643,
|
| 3 |
+
"total_flos": 3.767900833756416e+18,
|
| 4 |
+
"train_loss": 0.5178930132572812,
|
| 5 |
+
"train_runtime": 760.6734,
|
| 6 |
+
"train_samples_per_second": 537.355,
|
| 7 |
+
"train_steps_per_second": 2.798
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,817 +1,1583 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/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 |
-
"grad_norm": 2.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 3.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 4.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 3.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 5.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 2.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 2.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss":
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch":
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss":
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"
|
| 86 |
-
"eval_samples_per_second": 1335.617,
|
| 87 |
-
"eval_steps_per_second": 41.849,
|
| 88 |
-
"step": 100
|
| 89 |
-
},
|
| 90 |
-
{
|
| 91 |
-
"epoch": 1.1,
|
| 92 |
-
"grad_norm": 0.4499356150627136,
|
| 93 |
-
"learning_rate": 2.966666666666667e-05,
|
| 94 |
-
"loss": 1.7568,
|
| 95 |
"step": 110
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch":
|
| 99 |
-
"grad_norm":
|
| 100 |
-
"learning_rate":
|
| 101 |
-
"loss": 1.
|
| 102 |
"step": 120
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"grad_norm":
|
| 107 |
-
"learning_rate":
|
| 108 |
-
"loss": 1.
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
-
"epoch":
|
| 113 |
-
"grad_norm": 2.
|
| 114 |
-
"learning_rate":
|
| 115 |
-
"loss": 1.
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
-
"grad_norm":
|
| 121 |
-
"learning_rate": 2.
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"grad_norm":
|
| 128 |
-
"learning_rate": 2.
|
| 129 |
-
"loss": 1.
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"epoch":
|
| 134 |
-
"grad_norm": 1.
|
| 135 |
-
"learning_rate": 2.
|
| 136 |
-
"loss": 1.
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
-
"epoch":
|
| 141 |
-
"grad_norm":
|
| 142 |
-
"learning_rate": 2.
|
| 143 |
-
"loss": 1.
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
-
"epoch":
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"learning_rate": 2.
|
| 150 |
-
"loss": 1.
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
-
"epoch":
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 2.
|
| 157 |
-
"loss": 1.
|
| 158 |
-
"step": 200
|
| 159 |
-
},
|
| 160 |
-
{
|
| 161 |
-
"epoch": 2.0,
|
| 162 |
-
"eval_accuracy": 0.6510738452486026,
|
| 163 |
-
"eval_loss": 1.1778711080551147,
|
| 164 |
-
"eval_runtime": 4.9948,
|
| 165 |
-
"eval_samples_per_second": 1361.014,
|
| 166 |
-
"eval_steps_per_second": 42.644,
|
| 167 |
"step": 200
|
| 168 |
},
|
| 169 |
{
|
| 170 |
-
"epoch":
|
| 171 |
-
"grad_norm":
|
| 172 |
-
"learning_rate": 2.
|
| 173 |
-
"loss": 1.
|
| 174 |
"step": 210
|
| 175 |
},
|
| 176 |
{
|
| 177 |
-
"epoch":
|
| 178 |
-
"grad_norm":
|
| 179 |
-
"learning_rate": 2.
|
| 180 |
-
"loss": 1.
|
| 181 |
"step": 220
|
| 182 |
},
|
| 183 |
{
|
| 184 |
-
"epoch":
|
| 185 |
-
"grad_norm":
|
| 186 |
-
"learning_rate": 2.
|
| 187 |
-
"loss": 1.
|
| 188 |
"step": 230
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
| 192 |
-
"grad_norm":
|
| 193 |
-
"learning_rate": 2.
|
| 194 |
-
"loss": 1.
|
| 195 |
"step": 240
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
-
"grad_norm":
|
| 200 |
-
"learning_rate": 2.
|
| 201 |
-
"loss": 1.
|
| 202 |
"step": 250
|
| 203 |
},
|
| 204 |
{
|
| 205 |
-
"epoch":
|
| 206 |
-
"grad_norm":
|
| 207 |
-
"learning_rate": 2.
|
| 208 |
-
"loss": 1.
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
-
"epoch":
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
-
"epoch":
|
| 220 |
-
"grad_norm":
|
| 221 |
-
"learning_rate": 2.
|
| 222 |
-
"loss": 1.
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
-
"epoch":
|
| 227 |
-
"grad_norm": 2.
|
| 228 |
-
"learning_rate": 2.
|
| 229 |
-
"loss":
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch":
|
| 234 |
-
"grad_norm":
|
| 235 |
-
"learning_rate": 2.
|
| 236 |
-
"loss":
|
| 237 |
-
"step": 300
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 3.0,
|
| 241 |
-
"eval_accuracy": 0.8679023242130038,
|
| 242 |
-
"eval_loss": 0.8326017260551453,
|
| 243 |
-
"eval_runtime": 4.997,
|
| 244 |
-
"eval_samples_per_second": 1360.422,
|
| 245 |
-
"eval_steps_per_second": 42.626,
|
| 246 |
"step": 300
|
| 247 |
},
|
| 248 |
{
|
| 249 |
-
"epoch":
|
| 250 |
-
"grad_norm":
|
| 251 |
-
"learning_rate": 2.
|
| 252 |
-
"loss":
|
| 253 |
"step": 310
|
| 254 |
},
|
| 255 |
{
|
| 256 |
-
"epoch":
|
| 257 |
-
"grad_norm":
|
| 258 |
-
"learning_rate": 2.
|
| 259 |
-
"loss":
|
| 260 |
"step": 320
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"epoch":
|
| 264 |
-
"grad_norm": 2.
|
| 265 |
-
"learning_rate": 2.
|
| 266 |
-
"loss": 0.
|
| 267 |
"step": 330
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
-
"grad_norm":
|
| 272 |
-
"learning_rate": 2.
|
| 273 |
-
"loss":
|
| 274 |
"step": 340
|
| 275 |
},
|
| 276 |
{
|
| 277 |
-
"epoch":
|
| 278 |
-
"grad_norm":
|
| 279 |
-
"learning_rate": 2.
|
| 280 |
-
"loss": 0.
|
| 281 |
"step": 350
|
| 282 |
},
|
| 283 |
{
|
| 284 |
-
"epoch":
|
| 285 |
-
"grad_norm":
|
| 286 |
-
"learning_rate": 2.
|
| 287 |
-
"loss": 0.
|
| 288 |
"step": 360
|
| 289 |
},
|
| 290 |
{
|
| 291 |
-
"epoch":
|
| 292 |
-
"grad_norm":
|
| 293 |
-
"learning_rate": 2.
|
| 294 |
-
"loss": 0.
|
| 295 |
"step": 370
|
| 296 |
},
|
| 297 |
{
|
| 298 |
-
"epoch":
|
| 299 |
-
"grad_norm":
|
| 300 |
-
"learning_rate": 2.
|
| 301 |
-
"loss": 0.
|
| 302 |
"step": 380
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch":
|
| 306 |
-
"grad_norm": 2.
|
| 307 |
-
"learning_rate": 2.
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"grad_norm":
|
| 314 |
-
"learning_rate":
|
| 315 |
-
"loss": 0.
|
| 316 |
-
"step": 400
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"epoch": 4.0,
|
| 320 |
-
"eval_accuracy": 0.9724919093851133,
|
| 321 |
-
"eval_loss": 0.3696895241737366,
|
| 322 |
-
"eval_runtime": 5.064,
|
| 323 |
-
"eval_samples_per_second": 1342.416,
|
| 324 |
-
"eval_steps_per_second": 42.062,
|
| 325 |
"step": 400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
-
"epoch":
|
| 329 |
-
"grad_norm":
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 410
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch":
|
| 336 |
-
"grad_norm":
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 420
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 430
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch":
|
| 350 |
-
"grad_norm":
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 440
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch":
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 450
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch":
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 460
|
| 368 |
},
|
| 369 |
{
|
| 370 |
-
"epoch":
|
| 371 |
-
"grad_norm":
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 470
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch":
|
| 378 |
-
"grad_norm":
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 480
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch":
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 490
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch":
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss": 0.
|
| 395 |
-
"step": 500
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 5.0,
|
| 399 |
-
"eval_accuracy": 0.9760223595175052,
|
| 400 |
-
"eval_loss": 0.22396039962768555,
|
| 401 |
-
"eval_runtime": 4.8682,
|
| 402 |
-
"eval_samples_per_second": 1396.423,
|
| 403 |
-
"eval_steps_per_second": 43.754,
|
| 404 |
"step": 500
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch":
|
| 408 |
-
"grad_norm": 2.
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 510
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
-
"grad_norm":
|
| 416 |
-
"learning_rate":
|
| 417 |
-
"loss": 0.
|
| 418 |
"step": 520
|
| 419 |
},
|
| 420 |
{
|
| 421 |
-
"epoch":
|
| 422 |
-
"grad_norm":
|
| 423 |
-
"learning_rate":
|
| 424 |
-
"loss": 0.
|
| 425 |
"step": 530
|
| 426 |
},
|
| 427 |
{
|
| 428 |
-
"epoch":
|
| 429 |
-
"
|
| 430 |
-
"
|
| 431 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 432 |
"step": 540
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"epoch":
|
| 436 |
-
"grad_norm":
|
| 437 |
-
"learning_rate":
|
| 438 |
-
"loss": 0.
|
| 439 |
"step": 550
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch":
|
| 443 |
-
"grad_norm":
|
| 444 |
-
"learning_rate":
|
| 445 |
-
"loss": 0.
|
| 446 |
"step": 560
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch":
|
| 450 |
-
"grad_norm":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 570
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"grad_norm":
|
| 458 |
-
"learning_rate":
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 580
|
| 461 |
},
|
| 462 |
{
|
| 463 |
-
"epoch":
|
| 464 |
-
"grad_norm": 2.
|
| 465 |
-
"learning_rate":
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 590
|
| 468 |
},
|
| 469 |
{
|
| 470 |
-
"epoch":
|
| 471 |
-
"grad_norm": 2.
|
| 472 |
-
"learning_rate":
|
| 473 |
-
"loss": 0.
|
| 474 |
-
"step": 600
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 6.0,
|
| 478 |
-
"eval_accuracy": 0.9792586054721977,
|
| 479 |
-
"eval_loss": 0.1708967536687851,
|
| 480 |
-
"eval_runtime": 5.0298,
|
| 481 |
-
"eval_samples_per_second": 1351.551,
|
| 482 |
-
"eval_steps_per_second": 42.348,
|
| 483 |
"step": 600
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
-
"grad_norm":
|
| 488 |
-
"learning_rate":
|
| 489 |
-
"loss": 0.
|
| 490 |
"step": 610
|
| 491 |
},
|
| 492 |
{
|
| 493 |
-
"epoch":
|
| 494 |
-
"grad_norm":
|
| 495 |
-
"learning_rate":
|
| 496 |
-
"loss": 0.
|
| 497 |
"step": 620
|
| 498 |
},
|
| 499 |
{
|
| 500 |
-
"epoch":
|
| 501 |
-
"grad_norm": 2.
|
| 502 |
-
"learning_rate":
|
| 503 |
-
"loss": 0.
|
| 504 |
"step": 630
|
| 505 |
},
|
| 506 |
{
|
| 507 |
-
"epoch":
|
| 508 |
-
"grad_norm": 2.
|
| 509 |
-
"learning_rate":
|
| 510 |
-
"loss": 0.
|
| 511 |
"step": 640
|
| 512 |
},
|
| 513 |
{
|
| 514 |
-
"epoch":
|
| 515 |
-
"grad_norm":
|
| 516 |
-
"learning_rate":
|
| 517 |
-
"loss": 0.
|
| 518 |
"step": 650
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch":
|
| 522 |
-
"grad_norm":
|
| 523 |
-
"learning_rate":
|
| 524 |
-
"loss": 0.
|
| 525 |
"step": 660
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch":
|
| 529 |
-
"grad_norm":
|
| 530 |
-
"learning_rate":
|
| 531 |
-
"loss": 0.
|
| 532 |
"step": 670
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"epoch":
|
| 536 |
-
"grad_norm":
|
| 537 |
-
"learning_rate":
|
| 538 |
-
"loss": 0.
|
| 539 |
"step": 680
|
| 540 |
},
|
| 541 |
{
|
| 542 |
-
"epoch":
|
| 543 |
-
"grad_norm": 2.
|
| 544 |
-
"learning_rate":
|
| 545 |
-
"loss": 0.
|
| 546 |
"step": 690
|
| 547 |
},
|
| 548 |
{
|
| 549 |
-
"epoch":
|
| 550 |
-
"grad_norm": 2.
|
| 551 |
-
"learning_rate":
|
| 552 |
-
"loss": 0.
|
| 553 |
-
"step": 700
|
| 554 |
-
},
|
| 555 |
-
{
|
| 556 |
-
"epoch": 7.0,
|
| 557 |
-
"eval_accuracy": 0.9805825242718447,
|
| 558 |
-
"eval_loss": 0.14598147571086884,
|
| 559 |
-
"eval_runtime": 5.1023,
|
| 560 |
-
"eval_samples_per_second": 1332.35,
|
| 561 |
-
"eval_steps_per_second": 41.746,
|
| 562 |
"step": 700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch":
|
| 566 |
-
"grad_norm":
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 710
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch":
|
| 573 |
-
"grad_norm":
|
| 574 |
-
"learning_rate":
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 720
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch":
|
| 580 |
-
"grad_norm":
|
| 581 |
-
"learning_rate":
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 730
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch":
|
| 587 |
-
"grad_norm":
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 740
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"grad_norm":
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 750
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"grad_norm": 1.
|
| 602 |
-
"learning_rate":
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 760
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch":
|
| 608 |
-
"grad_norm": 2.
|
| 609 |
-
"learning_rate":
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 770
|
| 612 |
},
|
| 613 |
{
|
| 614 |
-
"epoch":
|
| 615 |
-
"grad_norm":
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 780
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch":
|
| 622 |
-
"grad_norm":
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 790
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch":
|
| 629 |
-
"grad_norm":
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 800
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch":
|
| 636 |
-
"eval_accuracy": 0.
|
| 637 |
-
"eval_loss": 0.
|
| 638 |
-
"eval_runtime":
|
| 639 |
-
"eval_samples_per_second":
|
| 640 |
-
"eval_steps_per_second":
|
| 641 |
-
"step":
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch":
|
| 645 |
-
"grad_norm":
|
| 646 |
-
"learning_rate":
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 810
|
| 649 |
},
|
| 650 |
{
|
| 651 |
-
"epoch":
|
| 652 |
-
"grad_norm":
|
| 653 |
-
"learning_rate":
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 820
|
| 656 |
},
|
| 657 |
{
|
| 658 |
-
"epoch":
|
| 659 |
-
"grad_norm":
|
| 660 |
-
"learning_rate":
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 830
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"epoch":
|
| 666 |
-
"grad_norm":
|
| 667 |
-
"learning_rate":
|
| 668 |
-
"loss": 0.
|
| 669 |
"step": 840
|
| 670 |
},
|
| 671 |
{
|
| 672 |
-
"epoch":
|
| 673 |
-
"grad_norm":
|
| 674 |
-
"learning_rate":
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 850
|
| 677 |
},
|
| 678 |
{
|
| 679 |
-
"epoch":
|
| 680 |
-
"grad_norm": 2.
|
| 681 |
-
"learning_rate":
|
| 682 |
-
"loss": 0.
|
| 683 |
"step": 860
|
| 684 |
},
|
| 685 |
{
|
| 686 |
-
"epoch":
|
| 687 |
-
"grad_norm":
|
| 688 |
-
"learning_rate":
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 870
|
| 691 |
},
|
| 692 |
{
|
| 693 |
-
"epoch":
|
| 694 |
-
"grad_norm":
|
| 695 |
-
"learning_rate":
|
| 696 |
-
"loss": 0.
|
| 697 |
"step": 880
|
| 698 |
},
|
| 699 |
{
|
| 700 |
-
"epoch":
|
| 701 |
-
"grad_norm":
|
| 702 |
-
"learning_rate":
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 890
|
| 705 |
},
|
| 706 |
{
|
| 707 |
-
"epoch":
|
| 708 |
-
"grad_norm":
|
| 709 |
-
"learning_rate":
|
| 710 |
-
"loss": 0.
|
| 711 |
"step": 900
|
| 712 |
},
|
| 713 |
{
|
| 714 |
-
"epoch":
|
| 715 |
-
"
|
| 716 |
-
"
|
| 717 |
-
"
|
| 718 |
-
"eval_samples_per_second": 1323.045,
|
| 719 |
-
"eval_steps_per_second": 41.455,
|
| 720 |
-
"step": 900
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 9.1,
|
| 724 |
-
"grad_norm": 1.9934319257736206,
|
| 725 |
-
"learning_rate": 3e-06,
|
| 726 |
-
"loss": 0.1915,
|
| 727 |
"step": 910
|
| 728 |
},
|
| 729 |
{
|
| 730 |
-
"epoch":
|
| 731 |
-
"grad_norm":
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 920
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"epoch":
|
| 738 |
-
"grad_norm":
|
| 739 |
-
"learning_rate":
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 930
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch":
|
| 745 |
-
"grad_norm":
|
| 746 |
-
"learning_rate":
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 940
|
| 749 |
},
|
| 750 |
{
|
| 751 |
-
"epoch":
|
| 752 |
-
"grad_norm":
|
| 753 |
-
"learning_rate": 1.
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 950
|
| 756 |
},
|
| 757 |
{
|
| 758 |
-
"epoch":
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate": 1.
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 960
|
| 763 |
},
|
| 764 |
{
|
| 765 |
-
"epoch":
|
| 766 |
-
"grad_norm":
|
| 767 |
-
"learning_rate":
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 970
|
| 770 |
},
|
| 771 |
{
|
| 772 |
-
"epoch":
|
| 773 |
-
"grad_norm":
|
| 774 |
-
"learning_rate":
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 980
|
| 777 |
},
|
| 778 |
{
|
| 779 |
-
"epoch":
|
| 780 |
-
"grad_norm":
|
| 781 |
-
"learning_rate":
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 990
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch":
|
| 787 |
-
"grad_norm":
|
| 788 |
-
"learning_rate":
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 1000
|
| 791 |
},
|
| 792 |
{
|
| 793 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 794 |
"eval_accuracy": 0.9811709326272433,
|
| 795 |
-
"eval_loss": 0.
|
| 796 |
-
"eval_runtime": 5.
|
| 797 |
-
"eval_samples_per_second":
|
| 798 |
-
"eval_steps_per_second": 41.
|
| 799 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 800 |
},
|
| 801 |
{
|
| 802 |
-
"epoch":
|
| 803 |
-
"
|
| 804 |
-
"
|
| 805 |
-
"
|
| 806 |
-
"
|
| 807 |
-
|
| 808 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 809 |
}
|
| 810 |
],
|
| 811 |
"logging_steps": 10,
|
| 812 |
-
"max_steps":
|
| 813 |
"num_input_tokens_seen": 0,
|
| 814 |
-
"num_train_epochs":
|
| 815 |
"save_steps": 500,
|
| 816 |
"stateful_callbacks": {
|
| 817 |
"TrainerControl": {
|
|
@@ -825,8 +1591,8 @@
|
|
| 825 |
"attributes": {}
|
| 826 |
}
|
| 827 |
},
|
| 828 |
-
"total_flos":
|
| 829 |
-
"train_batch_size":
|
| 830 |
"trial_name": null,
|
| 831 |
"trial_params": null
|
| 832 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9826419535157399,
|
| 3 |
+
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1335",
|
| 4 |
+
"epoch": 7.972769953051643,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 2128,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.03755868544600939,
|
| 13 |
+
"grad_norm": 2.0377416610717773,
|
| 14 |
+
"learning_rate": 1.4084507042253521e-06,
|
| 15 |
+
"loss": 3.8687,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.07511737089201878,
|
| 20 |
+
"grad_norm": 3.055781602859497,
|
| 21 |
+
"learning_rate": 2.8169014084507042e-06,
|
| 22 |
+
"loss": 4.1156,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.11267605633802817,
|
| 27 |
+
"grad_norm": 3.383268356323242,
|
| 28 |
+
"learning_rate": 4.225352112676057e-06,
|
| 29 |
+
"loss": 4.0885,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.15023474178403756,
|
| 34 |
+
"grad_norm": 3.8566606044769287,
|
| 35 |
+
"learning_rate": 5.6338028169014084e-06,
|
| 36 |
+
"loss": 3.9316,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.18779342723004694,
|
| 41 |
+
"grad_norm": 5.065456867218018,
|
| 42 |
+
"learning_rate": 7.042253521126761e-06,
|
| 43 |
+
"loss": 3.6474,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.22535211267605634,
|
| 48 |
+
"grad_norm": 5.89341926574707,
|
| 49 |
+
"learning_rate": 8.450704225352114e-06,
|
| 50 |
+
"loss": 3.2124,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.26291079812206575,
|
| 55 |
+
"grad_norm": 5.9929399490356445,
|
| 56 |
+
"learning_rate": 9.859154929577466e-06,
|
| 57 |
+
"loss": 2.756,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.3004694835680751,
|
| 62 |
+
"grad_norm": 5.689433574676514,
|
| 63 |
+
"learning_rate": 1.1267605633802817e-05,
|
| 64 |
+
"loss": 2.4596,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.3380281690140845,
|
| 69 |
+
"grad_norm": 4.89589262008667,
|
| 70 |
+
"learning_rate": 1.267605633802817e-05,
|
| 71 |
+
"loss": 2.2638,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.3755868544600939,
|
| 76 |
+
"grad_norm": 4.8666839599609375,
|
| 77 |
+
"learning_rate": 1.4084507042253522e-05,
|
| 78 |
+
"loss": 2.1166,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.4131455399061033,
|
| 83 |
+
"grad_norm": 4.466708660125732,
|
| 84 |
+
"learning_rate": 1.5492957746478876e-05,
|
| 85 |
+
"loss": 2.0048,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.4507042253521127,
|
| 90 |
+
"grad_norm": 3.676050901412964,
|
| 91 |
+
"learning_rate": 1.6901408450704228e-05,
|
| 92 |
+
"loss": 1.9138,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.48826291079812206,
|
| 97 |
+
"grad_norm": 2.183825731277466,
|
| 98 |
+
"learning_rate": 1.830985915492958e-05,
|
| 99 |
+
"loss": 1.863,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.5258215962441315,
|
| 104 |
+
"grad_norm": 2.075413465499878,
|
| 105 |
+
"learning_rate": 1.9718309859154933e-05,
|
| 106 |
+
"loss": 1.7616,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.5633802816901409,
|
| 111 |
+
"grad_norm": 0.8534318208694458,
|
| 112 |
+
"learning_rate": 2.112676056338028e-05,
|
| 113 |
+
"loss": 1.7185,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.6009389671361502,
|
| 118 |
+
"grad_norm": 0.9039830565452576,
|
| 119 |
+
"learning_rate": 2.2535211267605634e-05,
|
| 120 |
+
"loss": 1.8054,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.6384976525821596,
|
| 125 |
+
"grad_norm": 1.32124662399292,
|
| 126 |
+
"learning_rate": 2.3943661971830986e-05,
|
| 127 |
+
"loss": 1.7367,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.676056338028169,
|
| 132 |
+
"grad_norm": 1.232069969177246,
|
| 133 |
+
"learning_rate": 2.535211267605634e-05,
|
| 134 |
+
"loss": 1.7423,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.7136150234741784,
|
| 139 |
+
"grad_norm": 1.9570960998535156,
|
| 140 |
+
"learning_rate": 2.676056338028169e-05,
|
| 141 |
+
"loss": 1.6132,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.7511737089201878,
|
| 146 |
+
"grad_norm": 2.4463119506835938,
|
| 147 |
+
"learning_rate": 2.8169014084507043e-05,
|
| 148 |
+
"loss": 1.6099,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.7887323943661971,
|
| 153 |
+
"grad_norm": 6.601908206939697,
|
| 154 |
+
"learning_rate": 2.9577464788732395e-05,
|
| 155 |
+
"loss": 1.6043,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.8262910798122066,
|
| 160 |
+
"grad_norm": 3.225101947784424,
|
| 161 |
+
"learning_rate": 2.989033942558747e-05,
|
| 162 |
+
"loss": 1.5621,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.863849765258216,
|
| 167 |
+
"grad_norm": 3.698263645172119,
|
| 168 |
+
"learning_rate": 2.9733681462140994e-05,
|
| 169 |
+
"loss": 1.514,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.9014084507042254,
|
| 174 |
+
"grad_norm": 5.209756374359131,
|
| 175 |
+
"learning_rate": 2.9577023498694518e-05,
|
| 176 |
+
"loss": 1.4532,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.9389671361502347,
|
| 181 |
+
"grad_norm": 2.1304848194122314,
|
| 182 |
+
"learning_rate": 2.9420365535248042e-05,
|
| 183 |
+
"loss": 1.4312,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.9765258215962441,
|
| 188 |
+
"grad_norm": 4.837350368499756,
|
| 189 |
+
"learning_rate": 2.926370757180157e-05,
|
| 190 |
+
"loss": 1.3624,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 1.0,
|
| 195 |
+
"eval_accuracy": 0.6546042953809944,
|
| 196 |
+
"eval_loss": 1.19585382938385,
|
| 197 |
+
"eval_runtime": 4.8787,
|
| 198 |
+
"eval_samples_per_second": 1393.409,
|
| 199 |
+
"eval_steps_per_second": 43.659,
|
| 200 |
+
"step": 267
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.0112676056338028,
|
| 204 |
+
"grad_norm": 4.779292106628418,
|
| 205 |
+
"learning_rate": 2.9107049608355094e-05,
|
| 206 |
+
"loss": 1.2541,
|
| 207 |
"step": 270
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 1.0488262910798123,
|
| 211 |
+
"grad_norm": 3.60760498046875,
|
| 212 |
+
"learning_rate": 2.8950391644908618e-05,
|
| 213 |
+
"loss": 1.2271,
|
| 214 |
"step": 280
|
| 215 |
},
|
| 216 |
{
|
| 217 |
+
"epoch": 1.0863849765258216,
|
| 218 |
+
"grad_norm": 2.3788599967956543,
|
| 219 |
+
"learning_rate": 2.8793733681462142e-05,
|
| 220 |
+
"loss": 1.2335,
|
| 221 |
"step": 290
|
| 222 |
},
|
| 223 |
{
|
| 224 |
+
"epoch": 1.123943661971831,
|
| 225 |
+
"grad_norm": 3.353325843811035,
|
| 226 |
+
"learning_rate": 2.8637075718015666e-05,
|
| 227 |
+
"loss": 1.1613,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 228 |
"step": 300
|
| 229 |
},
|
| 230 |
{
|
| 231 |
+
"epoch": 1.1615023474178403,
|
| 232 |
+
"grad_norm": 4.326411247253418,
|
| 233 |
+
"learning_rate": 2.8480417754569193e-05,
|
| 234 |
+
"loss": 1.0754,
|
| 235 |
"step": 310
|
| 236 |
},
|
| 237 |
{
|
| 238 |
+
"epoch": 1.1990610328638498,
|
| 239 |
+
"grad_norm": 3.1939706802368164,
|
| 240 |
+
"learning_rate": 2.8323759791122717e-05,
|
| 241 |
+
"loss": 1.0353,
|
| 242 |
"step": 320
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 1.236619718309859,
|
| 246 |
+
"grad_norm": 2.8827011585235596,
|
| 247 |
+
"learning_rate": 2.816710182767624e-05,
|
| 248 |
+
"loss": 0.9806,
|
| 249 |
"step": 330
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 1.2741784037558685,
|
| 253 |
+
"grad_norm": 3.910698652267456,
|
| 254 |
+
"learning_rate": 2.8010443864229766e-05,
|
| 255 |
+
"loss": 1.0813,
|
| 256 |
"step": 340
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"epoch": 1.3117370892018778,
|
| 260 |
+
"grad_norm": 3.5916378498077393,
|
| 261 |
+
"learning_rate": 2.7853785900783293e-05,
|
| 262 |
+
"loss": 0.9792,
|
| 263 |
"step": 350
|
| 264 |
},
|
| 265 |
{
|
| 266 |
+
"epoch": 1.3492957746478873,
|
| 267 |
+
"grad_norm": 2.6981167793273926,
|
| 268 |
+
"learning_rate": 2.7697127937336817e-05,
|
| 269 |
+
"loss": 0.9231,
|
| 270 |
"step": 360
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 1.3868544600938968,
|
| 274 |
+
"grad_norm": 5.702897071838379,
|
| 275 |
+
"learning_rate": 2.754046997389034e-05,
|
| 276 |
+
"loss": 0.9435,
|
| 277 |
"step": 370
|
| 278 |
},
|
| 279 |
{
|
| 280 |
+
"epoch": 1.424413145539906,
|
| 281 |
+
"grad_norm": 4.622363090515137,
|
| 282 |
+
"learning_rate": 2.7383812010443865e-05,
|
| 283 |
+
"loss": 0.8449,
|
| 284 |
"step": 380
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 1.4619718309859155,
|
| 288 |
+
"grad_norm": 2.2103636264801025,
|
| 289 |
+
"learning_rate": 2.7227154046997393e-05,
|
| 290 |
+
"loss": 0.7713,
|
| 291 |
"step": 390
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 1.4995305164319248,
|
| 295 |
+
"grad_norm": 4.545182228088379,
|
| 296 |
+
"learning_rate": 2.7070496083550917e-05,
|
| 297 |
+
"loss": 0.7719,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 298 |
"step": 400
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 1.5370892018779343,
|
| 302 |
+
"grad_norm": 6.883026599884033,
|
| 303 |
+
"learning_rate": 2.691383812010444e-05,
|
| 304 |
+
"loss": 0.7564,
|
| 305 |
"step": 410
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 1.5746478873239438,
|
| 309 |
+
"grad_norm": 4.770920276641846,
|
| 310 |
+
"learning_rate": 2.6757180156657965e-05,
|
| 311 |
+
"loss": 0.6994,
|
| 312 |
"step": 420
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 1.612206572769953,
|
| 316 |
+
"grad_norm": 4.413459300994873,
|
| 317 |
+
"learning_rate": 2.660052219321149e-05,
|
| 318 |
+
"loss": 0.6313,
|
| 319 |
"step": 430
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 1.6497652582159623,
|
| 323 |
+
"grad_norm": 2.0261390209198,
|
| 324 |
+
"learning_rate": 2.6443864229765013e-05,
|
| 325 |
+
"loss": 0.6017,
|
| 326 |
"step": 440
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 1.6873239436619718,
|
| 330 |
+
"grad_norm": 5.67121696472168,
|
| 331 |
+
"learning_rate": 2.6287206266318537e-05,
|
| 332 |
+
"loss": 0.5792,
|
| 333 |
"step": 450
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 1.7248826291079813,
|
| 337 |
+
"grad_norm": 2.573594808578491,
|
| 338 |
+
"learning_rate": 2.6146214099216712e-05,
|
| 339 |
+
"loss": 0.545,
|
| 340 |
"step": 460
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"epoch": 1.7624413145539906,
|
| 344 |
+
"grad_norm": 4.145854949951172,
|
| 345 |
+
"learning_rate": 2.5989556135770236e-05,
|
| 346 |
+
"loss": 0.4907,
|
| 347 |
"step": 470
|
| 348 |
},
|
| 349 |
{
|
| 350 |
+
"epoch": 1.8,
|
| 351 |
+
"grad_norm": 1.7418975830078125,
|
| 352 |
+
"learning_rate": 2.583289817232376e-05,
|
| 353 |
+
"loss": 0.485,
|
| 354 |
"step": 480
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 1.8375586854460093,
|
| 358 |
+
"grad_norm": 4.651867866516113,
|
| 359 |
+
"learning_rate": 2.5676240208877287e-05,
|
| 360 |
+
"loss": 0.4572,
|
| 361 |
"step": 490
|
| 362 |
},
|
| 363 |
{
|
| 364 |
+
"epoch": 1.8751173708920188,
|
| 365 |
+
"grad_norm": 4.849829196929932,
|
| 366 |
+
"learning_rate": 2.551958224543081e-05,
|
| 367 |
+
"loss": 0.4864,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 368 |
"step": 500
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 1.9126760563380283,
|
| 372 |
+
"grad_norm": 2.631229877471924,
|
| 373 |
+
"learning_rate": 2.5362924281984335e-05,
|
| 374 |
+
"loss": 0.4035,
|
| 375 |
"step": 510
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 1.9502347417840376,
|
| 379 |
+
"grad_norm": 5.099828243255615,
|
| 380 |
+
"learning_rate": 2.520626631853786e-05,
|
| 381 |
+
"loss": 0.3818,
|
| 382 |
"step": 520
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"epoch": 1.9877934272300468,
|
| 386 |
+
"grad_norm": 3.25174617767334,
|
| 387 |
+
"learning_rate": 2.5049608355091387e-05,
|
| 388 |
+
"loss": 0.3854,
|
| 389 |
"step": 530
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 2.0,
|
| 393 |
+
"eval_accuracy": 0.9733745219182113,
|
| 394 |
+
"eval_loss": 0.2675245702266693,
|
| 395 |
+
"eval_runtime": 4.9697,
|
| 396 |
+
"eval_samples_per_second": 1367.882,
|
| 397 |
+
"eval_steps_per_second": 42.859,
|
| 398 |
+
"step": 534
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.0225352112676056,
|
| 402 |
+
"grad_norm": 4.533545017242432,
|
| 403 |
+
"learning_rate": 2.489295039164491e-05,
|
| 404 |
+
"loss": 0.3589,
|
| 405 |
"step": 540
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 2.060093896713615,
|
| 409 |
+
"grad_norm": 4.4245991706848145,
|
| 410 |
+
"learning_rate": 2.4736292428198435e-05,
|
| 411 |
+
"loss": 0.378,
|
| 412 |
"step": 550
|
| 413 |
},
|
| 414 |
{
|
| 415 |
+
"epoch": 2.0976525821596246,
|
| 416 |
+
"grad_norm": 5.778880596160889,
|
| 417 |
+
"learning_rate": 2.457963446475196e-05,
|
| 418 |
+
"loss": 0.3653,
|
| 419 |
"step": 560
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 2.1352112676056336,
|
| 423 |
+
"grad_norm": 3.5573890209198,
|
| 424 |
+
"learning_rate": 2.4422976501305487e-05,
|
| 425 |
+
"loss": 0.3107,
|
| 426 |
"step": 570
|
| 427 |
},
|
| 428 |
{
|
| 429 |
+
"epoch": 2.172769953051643,
|
| 430 |
+
"grad_norm": 3.655824899673462,
|
| 431 |
+
"learning_rate": 2.426631853785901e-05,
|
| 432 |
+
"loss": 0.3405,
|
| 433 |
"step": 580
|
| 434 |
},
|
| 435 |
{
|
| 436 |
+
"epoch": 2.2103286384976526,
|
| 437 |
+
"grad_norm": 2.430022954940796,
|
| 438 |
+
"learning_rate": 2.4109660574412535e-05,
|
| 439 |
+
"loss": 0.3298,
|
| 440 |
"step": 590
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 2.247887323943662,
|
| 444 |
+
"grad_norm": 2.9207568168640137,
|
| 445 |
+
"learning_rate": 2.3953002610966055e-05,
|
| 446 |
+
"loss": 0.3022,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 447 |
"step": 600
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 2.2854460093896716,
|
| 451 |
+
"grad_norm": 4.8787007331848145,
|
| 452 |
+
"learning_rate": 2.3796344647519583e-05,
|
| 453 |
+
"loss": 0.3991,
|
| 454 |
"step": 610
|
| 455 |
},
|
| 456 |
{
|
| 457 |
+
"epoch": 2.3230046948356806,
|
| 458 |
+
"grad_norm": 3.0268468856811523,
|
| 459 |
+
"learning_rate": 2.3639686684073107e-05,
|
| 460 |
+
"loss": 0.3159,
|
| 461 |
"step": 620
|
| 462 |
},
|
| 463 |
{
|
| 464 |
+
"epoch": 2.36056338028169,
|
| 465 |
+
"grad_norm": 2.6611557006835938,
|
| 466 |
+
"learning_rate": 2.348302872062663e-05,
|
| 467 |
+
"loss": 0.2868,
|
| 468 |
"step": 630
|
| 469 |
},
|
| 470 |
{
|
| 471 |
+
"epoch": 2.3981220657276996,
|
| 472 |
+
"grad_norm": 2.485551595687866,
|
| 473 |
+
"learning_rate": 2.3326370757180155e-05,
|
| 474 |
+
"loss": 0.3032,
|
| 475 |
"step": 640
|
| 476 |
},
|
| 477 |
{
|
| 478 |
+
"epoch": 2.435680751173709,
|
| 479 |
+
"grad_norm": 4.556153297424316,
|
| 480 |
+
"learning_rate": 2.316971279373368e-05,
|
| 481 |
+
"loss": 0.2985,
|
| 482 |
"step": 650
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 2.473239436619718,
|
| 486 |
+
"grad_norm": 5.270796298980713,
|
| 487 |
+
"learning_rate": 2.3013054830287207e-05,
|
| 488 |
+
"loss": 0.2839,
|
| 489 |
"step": 660
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 2.5107981220657276,
|
| 493 |
+
"grad_norm": 3.347005844116211,
|
| 494 |
+
"learning_rate": 2.285639686684073e-05,
|
| 495 |
+
"loss": 0.2871,
|
| 496 |
"step": 670
|
| 497 |
},
|
| 498 |
{
|
| 499 |
+
"epoch": 2.548356807511737,
|
| 500 |
+
"grad_norm": 5.236591815948486,
|
| 501 |
+
"learning_rate": 2.2699738903394255e-05,
|
| 502 |
+
"loss": 0.3028,
|
| 503 |
"step": 680
|
| 504 |
},
|
| 505 |
{
|
| 506 |
+
"epoch": 2.5859154929577466,
|
| 507 |
+
"grad_norm": 2.995059013366699,
|
| 508 |
+
"learning_rate": 2.254308093994778e-05,
|
| 509 |
+
"loss": 0.2537,
|
| 510 |
"step": 690
|
| 511 |
},
|
| 512 |
{
|
| 513 |
+
"epoch": 2.6234741784037556,
|
| 514 |
+
"grad_norm": 2.805640459060669,
|
| 515 |
+
"learning_rate": 2.2386422976501306e-05,
|
| 516 |
+
"loss": 0.297,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 517 |
"step": 700
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 2.661032863849765,
|
| 521 |
+
"grad_norm": 3.0646071434020996,
|
| 522 |
+
"learning_rate": 2.222976501305483e-05,
|
| 523 |
+
"loss": 0.2453,
|
| 524 |
"step": 710
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 2.6985915492957746,
|
| 528 |
+
"grad_norm": 3.6719613075256348,
|
| 529 |
+
"learning_rate": 2.2073107049608354e-05,
|
| 530 |
+
"loss": 0.2655,
|
| 531 |
"step": 720
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 2.736150234741784,
|
| 535 |
+
"grad_norm": 3.2248122692108154,
|
| 536 |
+
"learning_rate": 2.191644908616188e-05,
|
| 537 |
+
"loss": 0.2297,
|
| 538 |
"step": 730
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 2.7737089201877936,
|
| 542 |
+
"grad_norm": 3.769843578338623,
|
| 543 |
+
"learning_rate": 2.1759791122715406e-05,
|
| 544 |
+
"loss": 0.2548,
|
| 545 |
"step": 740
|
| 546 |
},
|
| 547 |
{
|
| 548 |
+
"epoch": 2.8112676056338026,
|
| 549 |
+
"grad_norm": 3.6679906845092773,
|
| 550 |
+
"learning_rate": 2.160313315926893e-05,
|
| 551 |
+
"loss": 0.2836,
|
| 552 |
"step": 750
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"epoch": 2.848826291079812,
|
| 556 |
+
"grad_norm": 1.6924936771392822,
|
| 557 |
+
"learning_rate": 2.1446475195822454e-05,
|
| 558 |
+
"loss": 0.2555,
|
| 559 |
"step": 760
|
| 560 |
},
|
| 561 |
{
|
| 562 |
+
"epoch": 2.8863849765258216,
|
| 563 |
+
"grad_norm": 2.1275901794433594,
|
| 564 |
+
"learning_rate": 2.1289817232375978e-05,
|
| 565 |
+
"loss": 0.2334,
|
| 566 |
"step": 770
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 2.923943661971831,
|
| 570 |
+
"grad_norm": 6.528135299682617,
|
| 571 |
+
"learning_rate": 2.1133159268929506e-05,
|
| 572 |
+
"loss": 0.2544,
|
| 573 |
"step": 780
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 2.9615023474178406,
|
| 577 |
+
"grad_norm": 2.4497199058532715,
|
| 578 |
+
"learning_rate": 2.097650130548303e-05,
|
| 579 |
+
"loss": 0.2628,
|
| 580 |
"step": 790
|
| 581 |
},
|
| 582 |
{
|
| 583 |
+
"epoch": 2.9990610328638496,
|
| 584 |
+
"grad_norm": 2.278947591781616,
|
| 585 |
+
"learning_rate": 2.0819843342036554e-05,
|
| 586 |
+
"loss": 0.2473,
|
| 587 |
"step": 800
|
| 588 |
},
|
| 589 |
{
|
| 590 |
+
"epoch": 3.0,
|
| 591 |
+
"eval_accuracy": 0.9767578699617535,
|
| 592 |
+
"eval_loss": 0.1461225152015686,
|
| 593 |
+
"eval_runtime": 5.0063,
|
| 594 |
+
"eval_samples_per_second": 1357.894,
|
| 595 |
+
"eval_steps_per_second": 42.547,
|
| 596 |
+
"step": 801
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.0338028169014084,
|
| 600 |
+
"grad_norm": 3.1185402870178223,
|
| 601 |
+
"learning_rate": 2.0663185378590078e-05,
|
| 602 |
+
"loss": 0.2245,
|
| 603 |
"step": 810
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 3.071361502347418,
|
| 607 |
+
"grad_norm": 2.456102132797241,
|
| 608 |
+
"learning_rate": 2.0506527415143602e-05,
|
| 609 |
+
"loss": 0.2423,
|
| 610 |
"step": 820
|
| 611 |
},
|
| 612 |
{
|
| 613 |
+
"epoch": 3.1089201877934274,
|
| 614 |
+
"grad_norm": 2.9463231563568115,
|
| 615 |
+
"learning_rate": 2.034986945169713e-05,
|
| 616 |
+
"loss": 0.2274,
|
| 617 |
"step": 830
|
| 618 |
},
|
| 619 |
{
|
| 620 |
+
"epoch": 3.1464788732394364,
|
| 621 |
+
"grad_norm": 3.5940473079681396,
|
| 622 |
+
"learning_rate": 2.0193211488250653e-05,
|
| 623 |
+
"loss": 0.2368,
|
| 624 |
"step": 840
|
| 625 |
},
|
| 626 |
{
|
| 627 |
+
"epoch": 3.184037558685446,
|
| 628 |
+
"grad_norm": 4.721577167510986,
|
| 629 |
+
"learning_rate": 2.0036553524804177e-05,
|
| 630 |
+
"loss": 0.2554,
|
| 631 |
"step": 850
|
| 632 |
},
|
| 633 |
{
|
| 634 |
+
"epoch": 3.2215962441314554,
|
| 635 |
+
"grad_norm": 2.496495485305786,
|
| 636 |
+
"learning_rate": 1.98798955613577e-05,
|
| 637 |
+
"loss": 0.2363,
|
| 638 |
"step": 860
|
| 639 |
},
|
| 640 |
{
|
| 641 |
+
"epoch": 3.259154929577465,
|
| 642 |
+
"grad_norm": 3.0665740966796875,
|
| 643 |
+
"learning_rate": 1.972323759791123e-05,
|
| 644 |
+
"loss": 0.2248,
|
| 645 |
"step": 870
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 3.2967136150234744,
|
| 649 |
+
"grad_norm": 4.336172580718994,
|
| 650 |
+
"learning_rate": 1.9566579634464753e-05,
|
| 651 |
+
"loss": 0.1922,
|
| 652 |
"step": 880
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"epoch": 3.3342723004694834,
|
| 656 |
+
"grad_norm": 4.110763072967529,
|
| 657 |
+
"learning_rate": 1.9409921671018277e-05,
|
| 658 |
+
"loss": 0.1965,
|
| 659 |
"step": 890
|
| 660 |
},
|
| 661 |
{
|
| 662 |
+
"epoch": 3.371830985915493,
|
| 663 |
+
"grad_norm": 1.9457247257232666,
|
| 664 |
+
"learning_rate": 1.92532637075718e-05,
|
| 665 |
+
"loss": 0.2258,
|
| 666 |
"step": 900
|
| 667 |
},
|
| 668 |
{
|
| 669 |
+
"epoch": 3.4093896713615024,
|
| 670 |
+
"grad_norm": 2.719369411468506,
|
| 671 |
+
"learning_rate": 1.909660574412533e-05,
|
| 672 |
+
"loss": 0.2184,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 673 |
"step": 910
|
| 674 |
},
|
| 675 |
{
|
| 676 |
+
"epoch": 3.446948356807512,
|
| 677 |
+
"grad_norm": 3.438279151916504,
|
| 678 |
+
"learning_rate": 1.8939947780678853e-05,
|
| 679 |
+
"loss": 0.1964,
|
| 680 |
"step": 920
|
| 681 |
},
|
| 682 |
{
|
| 683 |
+
"epoch": 3.4845070422535214,
|
| 684 |
+
"grad_norm": 3.2813045978546143,
|
| 685 |
+
"learning_rate": 1.8783289817232377e-05,
|
| 686 |
+
"loss": 0.2348,
|
| 687 |
"step": 930
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 3.5220657276995304,
|
| 691 |
+
"grad_norm": 4.151478290557861,
|
| 692 |
+
"learning_rate": 1.86266318537859e-05,
|
| 693 |
+
"loss": 0.2004,
|
| 694 |
"step": 940
|
| 695 |
},
|
| 696 |
{
|
| 697 |
+
"epoch": 3.55962441314554,
|
| 698 |
+
"grad_norm": 3.4271771907806396,
|
| 699 |
+
"learning_rate": 1.8469973890339425e-05,
|
| 700 |
+
"loss": 0.2039,
|
| 701 |
"step": 950
|
| 702 |
},
|
| 703 |
{
|
| 704 |
+
"epoch": 3.5971830985915494,
|
| 705 |
+
"grad_norm": 4.0341901779174805,
|
| 706 |
+
"learning_rate": 1.8313315926892952e-05,
|
| 707 |
+
"loss": 0.1997,
|
| 708 |
"step": 960
|
| 709 |
},
|
| 710 |
{
|
| 711 |
+
"epoch": 3.6347417840375584,
|
| 712 |
+
"grad_norm": 4.762091636657715,
|
| 713 |
+
"learning_rate": 1.8156657963446476e-05,
|
| 714 |
+
"loss": 0.2153,
|
| 715 |
"step": 970
|
| 716 |
},
|
| 717 |
{
|
| 718 |
+
"epoch": 3.672300469483568,
|
| 719 |
+
"grad_norm": 3.3214402198791504,
|
| 720 |
+
"learning_rate": 1.8e-05,
|
| 721 |
+
"loss": 0.1801,
|
| 722 |
"step": 980
|
| 723 |
},
|
| 724 |
{
|
| 725 |
+
"epoch": 3.7098591549295774,
|
| 726 |
+
"grad_norm": 3.84503173828125,
|
| 727 |
+
"learning_rate": 1.7843342036553525e-05,
|
| 728 |
+
"loss": 0.2106,
|
| 729 |
"step": 990
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 3.747417840375587,
|
| 733 |
+
"grad_norm": 3.303781747817993,
|
| 734 |
+
"learning_rate": 1.7686684073107052e-05,
|
| 735 |
+
"loss": 0.1965,
|
| 736 |
"step": 1000
|
| 737 |
},
|
| 738 |
{
|
| 739 |
+
"epoch": 3.7849765258215964,
|
| 740 |
+
"grad_norm": 2.691159248352051,
|
| 741 |
+
"learning_rate": 1.7530026109660576e-05,
|
| 742 |
+
"loss": 0.193,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.8225352112676054,
|
| 747 |
+
"grad_norm": 4.134768009185791,
|
| 748 |
+
"learning_rate": 1.73733681462141e-05,
|
| 749 |
+
"loss": 0.1908,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.860093896713615,
|
| 754 |
+
"grad_norm": 2.9195241928100586,
|
| 755 |
+
"learning_rate": 1.7216710182767624e-05,
|
| 756 |
+
"loss": 0.1886,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.8976525821596244,
|
| 761 |
+
"grad_norm": 3.795133352279663,
|
| 762 |
+
"learning_rate": 1.706005221932115e-05,
|
| 763 |
+
"loss": 0.2007,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.935211267605634,
|
| 768 |
+
"grad_norm": 3.9436607360839844,
|
| 769 |
+
"learning_rate": 1.6903394255874676e-05,
|
| 770 |
+
"loss": 0.1834,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.9727699530516434,
|
| 775 |
+
"grad_norm": 3.4115564823150635,
|
| 776 |
+
"learning_rate": 1.67467362924282e-05,
|
| 777 |
+
"loss": 0.1997,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 4.0,
|
| 782 |
+
"eval_accuracy": 0.980435422182995,
|
| 783 |
+
"eval_loss": 0.10877315700054169,
|
| 784 |
+
"eval_runtime": 4.881,
|
| 785 |
+
"eval_samples_per_second": 1392.737,
|
| 786 |
+
"eval_steps_per_second": 43.638,
|
| 787 |
+
"step": 1068
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 4.007511737089202,
|
| 791 |
+
"grad_norm": 5.121041774749756,
|
| 792 |
+
"learning_rate": 1.6590078328981724e-05,
|
| 793 |
+
"loss": 0.1785,
|
| 794 |
+
"step": 1070
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.045070422535211,
|
| 798 |
+
"grad_norm": 2.908527374267578,
|
| 799 |
+
"learning_rate": 1.643342036553525e-05,
|
| 800 |
+
"loss": 0.1678,
|
| 801 |
+
"step": 1080
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.08262910798122,
|
| 805 |
+
"grad_norm": 1.9687402248382568,
|
| 806 |
+
"learning_rate": 1.6276762402088775e-05,
|
| 807 |
+
"loss": 0.192,
|
| 808 |
+
"step": 1090
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 4.12018779342723,
|
| 812 |
+
"grad_norm": 2.722937822341919,
|
| 813 |
+
"learning_rate": 1.61201044386423e-05,
|
| 814 |
+
"loss": 0.1983,
|
| 815 |
+
"step": 1100
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 4.157746478873239,
|
| 819 |
+
"grad_norm": 2.3741490840911865,
|
| 820 |
+
"learning_rate": 1.5963446475195823e-05,
|
| 821 |
+
"loss": 0.2145,
|
| 822 |
+
"step": 1110
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 4.195305164319249,
|
| 826 |
+
"grad_norm": 2.653414011001587,
|
| 827 |
+
"learning_rate": 1.5806788511749348e-05,
|
| 828 |
+
"loss": 0.1701,
|
| 829 |
+
"step": 1120
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.232863849765258,
|
| 833 |
+
"grad_norm": 3.444087266921997,
|
| 834 |
+
"learning_rate": 1.5650130548302875e-05,
|
| 835 |
+
"loss": 0.2047,
|
| 836 |
+
"step": 1130
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.270422535211267,
|
| 840 |
+
"grad_norm": 2.024235486984253,
|
| 841 |
+
"learning_rate": 1.54934725848564e-05,
|
| 842 |
+
"loss": 0.1817,
|
| 843 |
+
"step": 1140
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.307981220657277,
|
| 847 |
+
"grad_norm": 2.742171049118042,
|
| 848 |
+
"learning_rate": 1.533681462140992e-05,
|
| 849 |
+
"loss": 0.1723,
|
| 850 |
+
"step": 1150
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.345539906103286,
|
| 854 |
+
"grad_norm": 3.3700480461120605,
|
| 855 |
+
"learning_rate": 1.5180156657963446e-05,
|
| 856 |
+
"loss": 0.17,
|
| 857 |
+
"step": 1160
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.383098591549296,
|
| 861 |
+
"grad_norm": 2.552915573120117,
|
| 862 |
+
"learning_rate": 1.5023498694516973e-05,
|
| 863 |
+
"loss": 0.1802,
|
| 864 |
+
"step": 1170
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.420657276995305,
|
| 868 |
+
"grad_norm": 3.3317511081695557,
|
| 869 |
+
"learning_rate": 1.4866840731070497e-05,
|
| 870 |
+
"loss": 0.1933,
|
| 871 |
+
"step": 1180
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.458215962441314,
|
| 875 |
+
"grad_norm": 1.9266548156738281,
|
| 876 |
+
"learning_rate": 1.4710182767624021e-05,
|
| 877 |
+
"loss": 0.1739,
|
| 878 |
+
"step": 1190
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.495774647887324,
|
| 882 |
+
"grad_norm": 2.1459243297576904,
|
| 883 |
+
"learning_rate": 1.4553524804177547e-05,
|
| 884 |
+
"loss": 0.1599,
|
| 885 |
+
"step": 1200
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.533333333333333,
|
| 889 |
+
"grad_norm": 3.9314770698547363,
|
| 890 |
+
"learning_rate": 1.4396866840731071e-05,
|
| 891 |
+
"loss": 0.1958,
|
| 892 |
+
"step": 1210
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.570892018779343,
|
| 896 |
+
"grad_norm": 2.6377363204956055,
|
| 897 |
+
"learning_rate": 1.4240208877284597e-05,
|
| 898 |
+
"loss": 0.1604,
|
| 899 |
+
"step": 1220
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.608450704225352,
|
| 903 |
+
"grad_norm": 2.810866594314575,
|
| 904 |
+
"learning_rate": 1.408355091383812e-05,
|
| 905 |
+
"loss": 0.1495,
|
| 906 |
+
"step": 1230
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.646009389671361,
|
| 910 |
+
"grad_norm": 2.2084455490112305,
|
| 911 |
+
"learning_rate": 1.3926892950391646e-05,
|
| 912 |
+
"loss": 0.185,
|
| 913 |
+
"step": 1240
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.683568075117371,
|
| 917 |
+
"grad_norm": 2.7217283248901367,
|
| 918 |
+
"learning_rate": 1.377023498694517e-05,
|
| 919 |
+
"loss": 0.1757,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.72112676056338,
|
| 924 |
+
"grad_norm": 3.075267791748047,
|
| 925 |
+
"learning_rate": 1.3613577023498696e-05,
|
| 926 |
+
"loss": 0.1814,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.758685446009389,
|
| 931 |
+
"grad_norm": 3.2452406883239746,
|
| 932 |
+
"learning_rate": 1.345691906005222e-05,
|
| 933 |
+
"loss": 0.1622,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.796244131455399,
|
| 938 |
+
"grad_norm": 2.712754487991333,
|
| 939 |
+
"learning_rate": 1.3300261096605744e-05,
|
| 940 |
+
"loss": 0.1714,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.833802816901408,
|
| 945 |
+
"grad_norm": 1.6795600652694702,
|
| 946 |
+
"learning_rate": 1.3143603133159269e-05,
|
| 947 |
+
"loss": 0.1519,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.871361502347418,
|
| 952 |
+
"grad_norm": 3.9085493087768555,
|
| 953 |
+
"learning_rate": 1.2986945169712793e-05,
|
| 954 |
+
"loss": 0.1758,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.908920187793427,
|
| 959 |
+
"grad_norm": 3.529478073120117,
|
| 960 |
+
"learning_rate": 1.2830287206266318e-05,
|
| 961 |
+
"loss": 0.1549,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.946478873239436,
|
| 966 |
+
"grad_norm": 2.559157609939575,
|
| 967 |
+
"learning_rate": 1.2673629242819842e-05,
|
| 968 |
+
"loss": 0.1824,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.984037558685446,
|
| 973 |
+
"grad_norm": 2.2350497245788574,
|
| 974 |
+
"learning_rate": 1.2516971279373368e-05,
|
| 975 |
+
"loss": 0.1723,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 5.0,
|
| 980 |
+
"eval_accuracy": 0.9826419535157399,
|
| 981 |
+
"eval_loss": 0.09542840719223022,
|
| 982 |
+
"eval_runtime": 4.9771,
|
| 983 |
+
"eval_samples_per_second": 1365.856,
|
| 984 |
+
"eval_steps_per_second": 42.796,
|
| 985 |
+
"step": 1335
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 5.018779342723005,
|
| 989 |
+
"grad_norm": 2.5073907375335693,
|
| 990 |
+
"learning_rate": 1.2360313315926892e-05,
|
| 991 |
+
"loss": 0.1401,
|
| 992 |
+
"step": 1340
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 5.056338028169014,
|
| 996 |
+
"grad_norm": 4.696757793426514,
|
| 997 |
+
"learning_rate": 1.2203655352480418e-05,
|
| 998 |
+
"loss": 0.1801,
|
| 999 |
+
"step": 1350
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 5.093896713615023,
|
| 1003 |
+
"grad_norm": 1.2180489301681519,
|
| 1004 |
+
"learning_rate": 1.2046997389033942e-05,
|
| 1005 |
+
"loss": 0.1335,
|
| 1006 |
+
"step": 1360
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 5.131455399061033,
|
| 1010 |
+
"grad_norm": 0.887860119342804,
|
| 1011 |
+
"learning_rate": 1.1890339425587468e-05,
|
| 1012 |
+
"loss": 0.1479,
|
| 1013 |
+
"step": 1370
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 5.169014084507042,
|
| 1017 |
+
"grad_norm": 3.6347432136535645,
|
| 1018 |
+
"learning_rate": 1.1733681462140992e-05,
|
| 1019 |
+
"loss": 0.1575,
|
| 1020 |
+
"step": 1380
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 5.206572769953052,
|
| 1024 |
+
"grad_norm": 2.901700496673584,
|
| 1025 |
+
"learning_rate": 1.1577023498694518e-05,
|
| 1026 |
+
"loss": 0.1367,
|
| 1027 |
+
"step": 1390
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 5.244131455399061,
|
| 1031 |
+
"grad_norm": 2.6395390033721924,
|
| 1032 |
+
"learning_rate": 1.1420365535248042e-05,
|
| 1033 |
+
"loss": 0.144,
|
| 1034 |
+
"step": 1400
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 5.28169014084507,
|
| 1038 |
+
"grad_norm": 3.923652172088623,
|
| 1039 |
+
"learning_rate": 1.1263707571801567e-05,
|
| 1040 |
+
"loss": 0.1576,
|
| 1041 |
+
"step": 1410
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 5.31924882629108,
|
| 1045 |
+
"grad_norm": 2.290224313735962,
|
| 1046 |
+
"learning_rate": 1.1107049608355092e-05,
|
| 1047 |
+
"loss": 0.16,
|
| 1048 |
+
"step": 1420
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 5.356807511737089,
|
| 1052 |
+
"grad_norm": 2.332317590713501,
|
| 1053 |
+
"learning_rate": 1.0950391644908617e-05,
|
| 1054 |
+
"loss": 0.1505,
|
| 1055 |
+
"step": 1430
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 5.394366197183099,
|
| 1059 |
+
"grad_norm": 3.474155902862549,
|
| 1060 |
+
"learning_rate": 1.0793733681462141e-05,
|
| 1061 |
+
"loss": 0.1828,
|
| 1062 |
+
"step": 1440
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 5.431924882629108,
|
| 1066 |
+
"grad_norm": 2.5219180583953857,
|
| 1067 |
+
"learning_rate": 1.0637075718015665e-05,
|
| 1068 |
+
"loss": 0.1563,
|
| 1069 |
+
"step": 1450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 5.469483568075117,
|
| 1073 |
+
"grad_norm": 4.863851547241211,
|
| 1074 |
+
"learning_rate": 1.0480417754569191e-05,
|
| 1075 |
+
"loss": 0.1308,
|
| 1076 |
+
"step": 1460
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 5.507042253521127,
|
| 1080 |
+
"grad_norm": 4.817688941955566,
|
| 1081 |
+
"learning_rate": 1.0323759791122715e-05,
|
| 1082 |
+
"loss": 0.1757,
|
| 1083 |
+
"step": 1470
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 5.544600938967136,
|
| 1087 |
+
"grad_norm": 3.194732189178467,
|
| 1088 |
+
"learning_rate": 1.0167101827676241e-05,
|
| 1089 |
+
"loss": 0.1577,
|
| 1090 |
+
"step": 1480
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 5.582159624413146,
|
| 1094 |
+
"grad_norm": 3.6605474948883057,
|
| 1095 |
+
"learning_rate": 1.0010443864229765e-05,
|
| 1096 |
+
"loss": 0.2044,
|
| 1097 |
+
"step": 1490
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 5.619718309859155,
|
| 1101 |
+
"grad_norm": 2.427701473236084,
|
| 1102 |
+
"learning_rate": 9.853785900783291e-06,
|
| 1103 |
+
"loss": 0.1574,
|
| 1104 |
+
"step": 1500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.657276995305164,
|
| 1108 |
+
"grad_norm": 2.8025519847869873,
|
| 1109 |
+
"learning_rate": 9.697127937336815e-06,
|
| 1110 |
+
"loss": 0.188,
|
| 1111 |
+
"step": 1510
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.694835680751174,
|
| 1115 |
+
"grad_norm": 2.042407989501953,
|
| 1116 |
+
"learning_rate": 9.54046997389034e-06,
|
| 1117 |
+
"loss": 0.1639,
|
| 1118 |
+
"step": 1520
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 5.732394366197183,
|
| 1122 |
+
"grad_norm": 4.5383477210998535,
|
| 1123 |
+
"learning_rate": 9.383812010443865e-06,
|
| 1124 |
+
"loss": 0.1641,
|
| 1125 |
+
"step": 1530
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 5.769953051643192,
|
| 1129 |
+
"grad_norm": 2.919588804244995,
|
| 1130 |
+
"learning_rate": 9.22715404699739e-06,
|
| 1131 |
+
"loss": 0.1374,
|
| 1132 |
+
"step": 1540
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 5.807511737089202,
|
| 1136 |
+
"grad_norm": 2.4344029426574707,
|
| 1137 |
+
"learning_rate": 9.070496083550915e-06,
|
| 1138 |
+
"loss": 0.1711,
|
| 1139 |
+
"step": 1550
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 5.845070422535211,
|
| 1143 |
+
"grad_norm": 1.5614906549453735,
|
| 1144 |
+
"learning_rate": 8.913838120104439e-06,
|
| 1145 |
+
"loss": 0.1624,
|
| 1146 |
+
"step": 1560
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 5.882629107981221,
|
| 1150 |
+
"grad_norm": 3.0189967155456543,
|
| 1151 |
+
"learning_rate": 8.757180156657963e-06,
|
| 1152 |
+
"loss": 0.1691,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.92018779342723,
|
| 1157 |
+
"grad_norm": 2.44000506401062,
|
| 1158 |
+
"learning_rate": 8.600522193211488e-06,
|
| 1159 |
+
"loss": 0.1513,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.957746478873239,
|
| 1164 |
+
"grad_norm": 2.4327423572540283,
|
| 1165 |
+
"learning_rate": 8.443864229765013e-06,
|
| 1166 |
+
"loss": 0.1538,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.995305164319249,
|
| 1171 |
+
"grad_norm": 2.1192240715026855,
|
| 1172 |
+
"learning_rate": 8.287206266318538e-06,
|
| 1173 |
+
"loss": 0.1442,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 6.0,
|
| 1178 |
+
"eval_accuracy": 0.981318034716093,
|
| 1179 |
+
"eval_loss": 0.09270217269659042,
|
| 1180 |
+
"eval_runtime": 4.992,
|
| 1181 |
+
"eval_samples_per_second": 1361.771,
|
| 1182 |
+
"eval_steps_per_second": 42.668,
|
| 1183 |
+
"step": 1602
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 6.030046948356808,
|
| 1187 |
+
"grad_norm": 1.8678548336029053,
|
| 1188 |
+
"learning_rate": 8.130548302872062e-06,
|
| 1189 |
+
"loss": 0.1328,
|
| 1190 |
+
"step": 1610
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 6.067605633802817,
|
| 1194 |
+
"grad_norm": 3.0712783336639404,
|
| 1195 |
+
"learning_rate": 7.973890339425586e-06,
|
| 1196 |
+
"loss": 0.1543,
|
| 1197 |
+
"step": 1620
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 6.105164319248826,
|
| 1201 |
+
"grad_norm": 4.49588680267334,
|
| 1202 |
+
"learning_rate": 7.817232375979112e-06,
|
| 1203 |
+
"loss": 0.1452,
|
| 1204 |
+
"step": 1630
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 6.142723004694836,
|
| 1208 |
+
"grad_norm": 3.9594759941101074,
|
| 1209 |
+
"learning_rate": 7.660574412532636e-06,
|
| 1210 |
+
"loss": 0.1513,
|
| 1211 |
+
"step": 1640
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 6.180281690140845,
|
| 1215 |
+
"grad_norm": 2.528153657913208,
|
| 1216 |
+
"learning_rate": 7.503916449086162e-06,
|
| 1217 |
+
"loss": 0.1589,
|
| 1218 |
+
"step": 1650
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 6.217840375586855,
|
| 1222 |
+
"grad_norm": 2.159458875656128,
|
| 1223 |
+
"learning_rate": 7.347258485639687e-06,
|
| 1224 |
+
"loss": 0.1443,
|
| 1225 |
+
"step": 1660
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 6.255399061032864,
|
| 1229 |
+
"grad_norm": 2.098022222518921,
|
| 1230 |
+
"learning_rate": 7.190600522193212e-06,
|
| 1231 |
+
"loss": 0.1564,
|
| 1232 |
+
"step": 1670
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 6.292957746478873,
|
| 1236 |
+
"grad_norm": 1.993698239326477,
|
| 1237 |
+
"learning_rate": 7.033942558746737e-06,
|
| 1238 |
+
"loss": 0.1401,
|
| 1239 |
+
"step": 1680
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 6.330516431924883,
|
| 1243 |
+
"grad_norm": 2.2639145851135254,
|
| 1244 |
+
"learning_rate": 6.877284595300262e-06,
|
| 1245 |
+
"loss": 0.1452,
|
| 1246 |
+
"step": 1690
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 6.368075117370892,
|
| 1250 |
+
"grad_norm": 2.5003936290740967,
|
| 1251 |
+
"learning_rate": 6.720626631853786e-06,
|
| 1252 |
+
"loss": 0.1439,
|
| 1253 |
+
"step": 1700
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 6.405633802816902,
|
| 1257 |
+
"grad_norm": 2.0841052532196045,
|
| 1258 |
+
"learning_rate": 6.563968668407311e-06,
|
| 1259 |
+
"loss": 0.1438,
|
| 1260 |
+
"step": 1710
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 6.443192488262911,
|
| 1264 |
+
"grad_norm": 3.550182819366455,
|
| 1265 |
+
"learning_rate": 6.4073107049608355e-06,
|
| 1266 |
+
"loss": 0.1433,
|
| 1267 |
+
"step": 1720
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 6.48075117370892,
|
| 1271 |
+
"grad_norm": 1.4857251644134521,
|
| 1272 |
+
"learning_rate": 6.2506527415143605e-06,
|
| 1273 |
+
"loss": 0.1404,
|
| 1274 |
+
"step": 1730
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 6.51830985915493,
|
| 1278 |
+
"grad_norm": 3.503309726715088,
|
| 1279 |
+
"learning_rate": 6.093994778067885e-06,
|
| 1280 |
+
"loss": 0.1493,
|
| 1281 |
+
"step": 1740
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 6.555868544600939,
|
| 1285 |
+
"grad_norm": 3.59545636177063,
|
| 1286 |
+
"learning_rate": 5.93733681462141e-06,
|
| 1287 |
+
"loss": 0.1563,
|
| 1288 |
+
"step": 1750
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 6.593427230046949,
|
| 1292 |
+
"grad_norm": 2.879582405090332,
|
| 1293 |
+
"learning_rate": 5.780678851174934e-06,
|
| 1294 |
+
"loss": 0.122,
|
| 1295 |
+
"step": 1760
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.630985915492958,
|
| 1299 |
+
"grad_norm": 1.7240543365478516,
|
| 1300 |
+
"learning_rate": 5.624020887728459e-06,
|
| 1301 |
+
"loss": 0.1404,
|
| 1302 |
+
"step": 1770
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.668544600938967,
|
| 1306 |
+
"grad_norm": 3.0438528060913086,
|
| 1307 |
+
"learning_rate": 5.467362924281984e-06,
|
| 1308 |
+
"loss": 0.1432,
|
| 1309 |
+
"step": 1780
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.706103286384977,
|
| 1313 |
+
"grad_norm": 2.496366024017334,
|
| 1314 |
+
"learning_rate": 5.310704960835509e-06,
|
| 1315 |
+
"loss": 0.1277,
|
| 1316 |
+
"step": 1790
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.743661971830986,
|
| 1320 |
+
"grad_norm": 1.7166277170181274,
|
| 1321 |
+
"learning_rate": 5.154046997389034e-06,
|
| 1322 |
+
"loss": 0.143,
|
| 1323 |
+
"step": 1800
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.781220657276995,
|
| 1327 |
+
"grad_norm": 2.4547784328460693,
|
| 1328 |
+
"learning_rate": 4.997389033942559e-06,
|
| 1329 |
+
"loss": 0.1198,
|
| 1330 |
+
"step": 1810
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.818779342723005,
|
| 1334 |
+
"grad_norm": 2.604220390319824,
|
| 1335 |
+
"learning_rate": 4.840731070496084e-06,
|
| 1336 |
+
"loss": 0.1705,
|
| 1337 |
+
"step": 1820
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.856338028169014,
|
| 1341 |
+
"grad_norm": 2.7237601280212402,
|
| 1342 |
+
"learning_rate": 4.684073107049609e-06,
|
| 1343 |
+
"loss": 0.1506,
|
| 1344 |
+
"step": 1830
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.893896713615024,
|
| 1348 |
+
"grad_norm": 2.638058662414551,
|
| 1349 |
+
"learning_rate": 4.527415143603134e-06,
|
| 1350 |
+
"loss": 0.154,
|
| 1351 |
+
"step": 1840
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.931455399061033,
|
| 1355 |
+
"grad_norm": 3.8382205963134766,
|
| 1356 |
+
"learning_rate": 4.3707571801566586e-06,
|
| 1357 |
+
"loss": 0.1553,
|
| 1358 |
+
"step": 1850
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.969014084507043,
|
| 1362 |
+
"grad_norm": 2.071164131164551,
|
| 1363 |
+
"learning_rate": 4.2140992167101835e-06,
|
| 1364 |
+
"loss": 0.1397,
|
| 1365 |
+
"step": 1860
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 7.0,
|
| 1369 |
"eval_accuracy": 0.9811709326272433,
|
| 1370 |
+
"eval_loss": 0.08920056372880936,
|
| 1371 |
+
"eval_runtime": 5.1325,
|
| 1372 |
+
"eval_samples_per_second": 1324.506,
|
| 1373 |
+
"eval_steps_per_second": 41.5,
|
| 1374 |
+
"step": 1869
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 7.003755868544601,
|
| 1378 |
+
"grad_norm": 2.5346381664276123,
|
| 1379 |
+
"learning_rate": 4.0574412532637075e-06,
|
| 1380 |
+
"loss": 0.1296,
|
| 1381 |
+
"step": 1870
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 7.041314553990611,
|
| 1385 |
+
"grad_norm": 2.575307846069336,
|
| 1386 |
+
"learning_rate": 3.9007832898172325e-06,
|
| 1387 |
+
"loss": 0.1389,
|
| 1388 |
+
"step": 1880
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 7.07887323943662,
|
| 1392 |
+
"grad_norm": 2.0408527851104736,
|
| 1393 |
+
"learning_rate": 3.7441253263707574e-06,
|
| 1394 |
+
"loss": 0.1521,
|
| 1395 |
+
"step": 1890
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 7.1164319248826295,
|
| 1399 |
+
"grad_norm": 3.2742061614990234,
|
| 1400 |
+
"learning_rate": 3.5874673629242823e-06,
|
| 1401 |
+
"loss": 0.1342,
|
| 1402 |
+
"step": 1900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 7.153990610328639,
|
| 1406 |
+
"grad_norm": 1.4502960443496704,
|
| 1407 |
+
"learning_rate": 3.4308093994778068e-06,
|
| 1408 |
+
"loss": 0.1204,
|
| 1409 |
+
"step": 1910
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 7.191549295774648,
|
| 1413 |
+
"grad_norm": 3.7600743770599365,
|
| 1414 |
+
"learning_rate": 3.2741514360313317e-06,
|
| 1415 |
+
"loss": 0.1431,
|
| 1416 |
+
"step": 1920
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.229107981220658,
|
| 1420 |
+
"grad_norm": 2.7332417964935303,
|
| 1421 |
+
"learning_rate": 3.1174934725848566e-06,
|
| 1422 |
+
"loss": 0.1281,
|
| 1423 |
+
"step": 1930
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 7.266666666666667,
|
| 1427 |
+
"grad_norm": 2.6618921756744385,
|
| 1428 |
+
"learning_rate": 2.960835509138381e-06,
|
| 1429 |
+
"loss": 0.141,
|
| 1430 |
+
"step": 1940
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 7.304225352112676,
|
| 1434 |
+
"grad_norm": 3.625688314437866,
|
| 1435 |
+
"learning_rate": 2.804177545691906e-06,
|
| 1436 |
+
"loss": 0.1455,
|
| 1437 |
+
"step": 1950
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
+
"epoch": 7.341784037558686,
|
| 1441 |
+
"grad_norm": 2.0667765140533447,
|
| 1442 |
+
"learning_rate": 2.647519582245431e-06,
|
| 1443 |
+
"loss": 0.1359,
|
| 1444 |
+
"step": 1960
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.379342723004695,
|
| 1448 |
+
"grad_norm": 2.369652509689331,
|
| 1449 |
+
"learning_rate": 2.490861618798956e-06,
|
| 1450 |
+
"loss": 0.1295,
|
| 1451 |
+
"step": 1970
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.416901408450705,
|
| 1455 |
+
"grad_norm": 3.836838722229004,
|
| 1456 |
+
"learning_rate": 2.3342036553524807e-06,
|
| 1457 |
+
"loss": 0.1489,
|
| 1458 |
+
"step": 1980
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.454460093896714,
|
| 1462 |
+
"grad_norm": 3.3261311054229736,
|
| 1463 |
+
"learning_rate": 2.1775456919060052e-06,
|
| 1464 |
+
"loss": 0.1289,
|
| 1465 |
+
"step": 1990
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.492018779342723,
|
| 1469 |
+
"grad_norm": 2.6514954566955566,
|
| 1470 |
+
"learning_rate": 2.0208877284595297e-06,
|
| 1471 |
+
"loss": 0.1185,
|
| 1472 |
+
"step": 2000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.529577464788733,
|
| 1476 |
+
"grad_norm": 2.1017005443573,
|
| 1477 |
+
"learning_rate": 1.8642297650130548e-06,
|
| 1478 |
+
"loss": 0.1472,
|
| 1479 |
+
"step": 2010
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.567136150234742,
|
| 1483 |
+
"grad_norm": 2.5104258060455322,
|
| 1484 |
+
"learning_rate": 1.7075718015665795e-06,
|
| 1485 |
+
"loss": 0.1467,
|
| 1486 |
+
"step": 2020
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.6046948356807516,
|
| 1490 |
+
"grad_norm": 1.7915935516357422,
|
| 1491 |
+
"learning_rate": 1.5509138381201045e-06,
|
| 1492 |
+
"loss": 0.1212,
|
| 1493 |
+
"step": 2030
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.642253521126761,
|
| 1497 |
+
"grad_norm": 2.4937989711761475,
|
| 1498 |
+
"learning_rate": 1.3942558746736294e-06,
|
| 1499 |
+
"loss": 0.1395,
|
| 1500 |
+
"step": 2040
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.67981220657277,
|
| 1504 |
+
"grad_norm": 2.758594274520874,
|
| 1505 |
+
"learning_rate": 1.237597911227154e-06,
|
| 1506 |
+
"loss": 0.1361,
|
| 1507 |
+
"step": 2050
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.71737089201878,
|
| 1511 |
+
"grad_norm": 2.291672468185425,
|
| 1512 |
+
"learning_rate": 1.0809399477806788e-06,
|
| 1513 |
+
"loss": 0.1182,
|
| 1514 |
+
"step": 2060
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.754929577464789,
|
| 1518 |
+
"grad_norm": 1.944736361503601,
|
| 1519 |
+
"learning_rate": 9.242819843342037e-07,
|
| 1520 |
+
"loss": 0.1307,
|
| 1521 |
+
"step": 2070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.792488262910798,
|
| 1525 |
+
"grad_norm": 1.448411226272583,
|
| 1526 |
+
"learning_rate": 7.676240208877285e-07,
|
| 1527 |
+
"loss": 0.1407,
|
| 1528 |
+
"step": 2080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.830046948356808,
|
| 1532 |
+
"grad_norm": 3.276000499725342,
|
| 1533 |
+
"learning_rate": 6.109660574412533e-07,
|
| 1534 |
+
"loss": 0.1361,
|
| 1535 |
+
"step": 2090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.867605633802817,
|
| 1539 |
+
"grad_norm": 3.627788543701172,
|
| 1540 |
+
"learning_rate": 4.5430809399477806e-07,
|
| 1541 |
+
"loss": 0.131,
|
| 1542 |
+
"step": 2100
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.905164319248827,
|
| 1546 |
+
"grad_norm": 1.2533661127090454,
|
| 1547 |
+
"learning_rate": 2.9765013054830287e-07,
|
| 1548 |
+
"loss": 0.1245,
|
| 1549 |
+
"step": 2110
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.942723004694836,
|
| 1553 |
+
"grad_norm": 1.472484827041626,
|
| 1554 |
+
"learning_rate": 1.409921671018277e-07,
|
| 1555 |
+
"loss": 0.1368,
|
| 1556 |
+
"step": 2120
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.972769953051643,
|
| 1560 |
+
"eval_accuracy": 0.9811709326272433,
|
| 1561 |
+
"eval_loss": 0.08957477658987045,
|
| 1562 |
+
"eval_runtime": 5.6752,
|
| 1563 |
+
"eval_samples_per_second": 1197.851,
|
| 1564 |
+
"eval_steps_per_second": 37.532,
|
| 1565 |
+
"step": 2128
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 7.972769953051643,
|
| 1569 |
+
"step": 2128,
|
| 1570 |
+
"total_flos": 3.767900833756416e+18,
|
| 1571 |
+
"train_loss": 0.5178930132572812,
|
| 1572 |
+
"train_runtime": 760.6734,
|
| 1573 |
+
"train_samples_per_second": 537.355,
|
| 1574 |
+
"train_steps_per_second": 2.798
|
| 1575 |
}
|
| 1576 |
],
|
| 1577 |
"logging_steps": 10,
|
| 1578 |
+
"max_steps": 2128,
|
| 1579 |
"num_input_tokens_seen": 0,
|
| 1580 |
+
"num_train_epochs": 8,
|
| 1581 |
"save_steps": 500,
|
| 1582 |
"stateful_callbacks": {
|
| 1583 |
"TrainerControl": {
|
|
|
|
| 1591 |
"attributes": {}
|
| 1592 |
}
|
| 1593 |
},
|
| 1594 |
+
"total_flos": 3.767900833756416e+18,
|
| 1595 |
+
"train_batch_size": 48,
|
| 1596 |
"trial_name": null,
|
| 1597 |
"trial_params": null
|
| 1598 |
}
|