Commit
•
6f6fdf2
1
Parent(s):
15cc8fb
End of training
Browse files- README.md +1 -1
- all_results.json +12 -12
- eval_results.json +6 -6
- runs/Nov14_21-46-14_58f76309bdc2/events.out.tfevents.1731681181.58f76309bdc2.157.1 +3 -0
- train_results.json +7 -7
- trainer_state.json +822 -822
README.md
CHANGED
|
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# whisper-large-v3-ft-btb-ca-cy
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.3810
|
| 22 |
- Wer: 0.2750
|
|
|
|
| 16 |
|
| 17 |
# whisper-large-v3-ft-btb-ca-cy
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the DewiBrynJones/banc-trawsgrifiadau-bangor-clean train main, cymen-arfor/25awr train+dev main dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.3810
|
| 22 |
- Wer: 0.2750
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
"eval_samples": 3901,
|
| 6 |
-
"eval_samples_per_second": 2.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
-
"total_flos": 5.
|
| 10 |
-
"train_loss": 0.
|
| 11 |
-
"train_runtime":
|
| 12 |
-
"train_samples":
|
| 13 |
-
"train_samples_per_second": 2.
|
| 14 |
-
"train_steps_per_second": 0.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.7056277056277054,
|
| 3 |
+
"eval_loss": 0.38099274039268494,
|
| 4 |
+
"eval_runtime": 1765.5662,
|
| 5 |
"eval_samples": 3901,
|
| 6 |
+
"eval_samples_per_second": 2.209,
|
| 7 |
+
"eval_steps_per_second": 0.138,
|
| 8 |
+
"eval_wer": 0.27501690331304934,
|
| 9 |
+
"total_flos": 5.434978041004032e+20,
|
| 10 |
+
"train_loss": 0.39443905401229856,
|
| 11 |
+
"train_runtime": 57517.0908,
|
| 12 |
+
"train_samples": 59121,
|
| 13 |
+
"train_samples_per_second": 2.782,
|
| 14 |
+
"train_steps_per_second": 0.087
|
| 15 |
}
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
"eval_samples": 3901,
|
| 6 |
-
"eval_samples_per_second": 2.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.7056277056277054,
|
| 3 |
+
"eval_loss": 0.38099274039268494,
|
| 4 |
+
"eval_runtime": 1765.5662,
|
| 5 |
"eval_samples": 3901,
|
| 6 |
+
"eval_samples_per_second": 2.209,
|
| 7 |
+
"eval_steps_per_second": 0.138,
|
| 8 |
+
"eval_wer": 0.27501690331304934
|
| 9 |
}
|
runs/Nov14_21-46-14_58f76309bdc2/events.out.tfevents.1731681181.58f76309bdc2.157.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19d6d2eb00b08b2db0681e4f1fcd7dd77fe49df073c304dc4c8ec65cc4a443f1
|
| 3 |
+
size 406
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 5.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second": 2.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.7056277056277054,
|
| 3 |
+
"total_flos": 5.434978041004032e+20,
|
| 4 |
+
"train_loss": 0.39443905401229856,
|
| 5 |
+
"train_runtime": 57517.0908,
|
| 6 |
+
"train_samples": 59121,
|
| 7 |
+
"train_samples_per_second": 2.782,
|
| 8 |
+
"train_steps_per_second": 0.087
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -9,1464 +9,1464 @@
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 4.800000000000001e-07,
|
| 15 |
-
"loss":
|
| 16 |
"step": 25
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 11.
|
| 21 |
"learning_rate": 9.800000000000001e-07,
|
| 22 |
-
"loss":
|
| 23 |
"step": 50
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm": 10.
|
| 28 |
"learning_rate": 1.48e-06,
|
| 29 |
-
"loss":
|
| 30 |
"step": 75
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
"learning_rate": 1.98e-06,
|
| 36 |
-
"loss":
|
| 37 |
"step": 100
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 2.4800000000000004e-06,
|
| 43 |
-
"loss":
|
| 44 |
"step": 125
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 2.9800000000000003e-06,
|
| 50 |
-
"loss":
|
| 51 |
"step": 150
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 3.48e-06,
|
| 57 |
-
"loss":
|
| 58 |
"step": 175
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 3.980000000000001e-06,
|
| 64 |
-
"loss":
|
| 65 |
"step": 200
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm": 9.
|
| 70 |
"learning_rate": 4.48e-06,
|
| 71 |
-
"loss":
|
| 72 |
"step": 225
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 4.980000000000001e-06,
|
| 78 |
-
"loss":
|
| 79 |
"step": 250
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm": 9.
|
| 84 |
"learning_rate": 5.480000000000001e-06,
|
| 85 |
-
"loss":
|
| 86 |
"step": 275
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm": 9.
|
| 91 |
"learning_rate": 5.98e-06,
|
| 92 |
-
"loss":
|
| 93 |
"step": 300
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm":
|
| 98 |
"learning_rate": 6.480000000000001e-06,
|
| 99 |
-
"loss":
|
| 100 |
"step": 325
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm":
|
| 105 |
"learning_rate": 6.98e-06,
|
| 106 |
-
"loss":
|
| 107 |
"step": 350
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm":
|
| 112 |
"learning_rate": 7.48e-06,
|
| 113 |
-
"loss":
|
| 114 |
"step": 375
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 7.980000000000002e-06,
|
| 120 |
-
"loss":
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm": 9.
|
| 126 |
"learning_rate": 8.48e-06,
|
| 127 |
-
"loss":
|
| 128 |
"step": 425
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm": 8.
|
| 133 |
-
"learning_rate": 8.
|
| 134 |
-
"loss":
|
| 135 |
"step": 450
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm": 8.
|
| 140 |
-
"learning_rate": 9.
|
| 141 |
-
"loss":
|
| 142 |
"step": 475
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate": 9.
|
| 148 |
-
"loss":
|
| 149 |
"step": 500
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate": 9.
|
| 155 |
-
"loss":
|
| 156 |
"step": 525
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 9.
|
| 162 |
-
"loss":
|
| 163 |
"step": 550
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 9.
|
| 169 |
-
"loss":
|
| 170 |
"step": 575
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm": 8.
|
| 175 |
-
"learning_rate": 9.
|
| 176 |
-
"loss":
|
| 177 |
"step": 600
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm": 8.
|
| 182 |
-
"learning_rate": 9.
|
| 183 |
-
"loss":
|
| 184 |
"step": 625
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm":
|
| 189 |
-
"learning_rate": 9.
|
| 190 |
-
"loss":
|
| 191 |
"step": 650
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate": 9.
|
| 197 |
-
"loss":
|
| 198 |
"step": 675
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm": 8.
|
| 203 |
-
"learning_rate": 9.
|
| 204 |
-
"loss":
|
| 205 |
"step": 700
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm":
|
| 210 |
-
"learning_rate": 9.
|
| 211 |
-
"loss":
|
| 212 |
"step": 725
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm":
|
| 217 |
-
"learning_rate": 9.
|
| 218 |
-
"loss":
|
| 219 |
"step": 750
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate": 9.
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 775
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate": 9.
|
| 232 |
-
"loss":
|
| 233 |
"step": 800
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm": 7.
|
| 238 |
-
"learning_rate": 9.
|
| 239 |
-
"loss":
|
| 240 |
"step": 825
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate": 9.
|
| 246 |
-
"loss":
|
| 247 |
"step": 850
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch": 0.
|
| 251 |
-
"grad_norm":
|
| 252 |
-
"learning_rate": 9.
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 875
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate": 9.
|
| 260 |
-
"loss":
|
| 261 |
"step": 900
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate": 9.
|
| 267 |
-
"loss":
|
| 268 |
"step": 925
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch": 0.
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 9.
|
| 274 |
-
"loss":
|
| 275 |
"step": 950
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"grad_norm":
|
| 280 |
-
"learning_rate": 8.
|
| 281 |
-
"loss":
|
| 282 |
"step": 975
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 0.
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate": 8.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 1000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 0.
|
| 293 |
-
"eval_loss": 0.
|
| 294 |
-
"eval_runtime":
|
| 295 |
-
"eval_samples_per_second": 2.
|
| 296 |
-
"eval_steps_per_second": 0.
|
| 297 |
-
"eval_wer": 0.
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
-
"epoch": 0.
|
| 302 |
-
"grad_norm":
|
| 303 |
-
"learning_rate": 8.
|
| 304 |
-
"loss":
|
| 305 |
"step": 1025
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch": 0.
|
| 309 |
-
"grad_norm": 6.
|
| 310 |
-
"learning_rate": 8.
|
| 311 |
-
"loss": 0.
|
| 312 |
"step": 1050
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"epoch": 0.
|
| 316 |
-
"grad_norm": 7.
|
| 317 |
-
"learning_rate": 8.
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 1075
|
| 320 |
},
|
| 321 |
{
|
| 322 |
-
"epoch": 0.
|
| 323 |
-
"grad_norm":
|
| 324 |
-
"learning_rate": 8.
|
| 325 |
-
"loss":
|
| 326 |
"step": 1100
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch": 0.
|
| 330 |
-
"grad_norm":
|
| 331 |
-
"learning_rate": 8.
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 1125
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 0.
|
| 337 |
-
"grad_norm": 6.
|
| 338 |
-
"learning_rate": 8.
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 1150
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"epoch": 0.
|
| 344 |
-
"grad_norm":
|
| 345 |
-
"learning_rate": 8.
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 1175
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch": 0.
|
| 351 |
-
"grad_norm":
|
| 352 |
-
"learning_rate": 8.
|
| 353 |
-
"loss": 0.
|
| 354 |
"step": 1200
|
| 355 |
},
|
| 356 |
{
|
| 357 |
-
"epoch": 0.
|
| 358 |
-
"grad_norm": 6.
|
| 359 |
-
"learning_rate": 8.
|
| 360 |
-
"loss": 0.
|
| 361 |
"step": 1225
|
| 362 |
},
|
| 363 |
{
|
| 364 |
-
"epoch": 0.
|
| 365 |
-
"grad_norm": 7.
|
| 366 |
-
"learning_rate": 8.
|
| 367 |
-
"loss": 0.
|
| 368 |
"step": 1250
|
| 369 |
},
|
| 370 |
{
|
| 371 |
-
"epoch": 0.
|
| 372 |
-
"grad_norm":
|
| 373 |
-
"learning_rate": 8.
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 1275
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 0.
|
| 379 |
-
"grad_norm":
|
| 380 |
-
"learning_rate": 8.
|
| 381 |
-
"loss": 0.
|
| 382 |
"step": 1300
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"epoch": 0.
|
| 386 |
-
"grad_norm":
|
| 387 |
-
"learning_rate": 8.
|
| 388 |
-
"loss": 0.
|
| 389 |
"step": 1325
|
| 390 |
},
|
| 391 |
{
|
| 392 |
-
"epoch": 0.
|
| 393 |
-
"grad_norm":
|
| 394 |
-
"learning_rate": 8.
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 1350
|
| 397 |
},
|
| 398 |
{
|
| 399 |
-
"epoch": 0.
|
| 400 |
-
"grad_norm": 6.
|
| 401 |
-
"learning_rate": 8.
|
| 402 |
-
"loss": 0.
|
| 403 |
"step": 1375
|
| 404 |
},
|
| 405 |
{
|
| 406 |
-
"epoch": 0.
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 8.
|
| 409 |
-
"loss": 0.
|
| 410 |
"step": 1400
|
| 411 |
},
|
| 412 |
{
|
| 413 |
-
"epoch": 0.
|
| 414 |
-
"grad_norm":
|
| 415 |
-
"learning_rate": 7.
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 1425
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 0.
|
| 421 |
-
"grad_norm":
|
| 422 |
-
"learning_rate": 7.
|
| 423 |
-
"loss": 0.
|
| 424 |
"step": 1450
|
| 425 |
},
|
| 426 |
{
|
| 427 |
-
"epoch": 0.
|
| 428 |
-
"grad_norm": 6.
|
| 429 |
-
"learning_rate": 7.
|
| 430 |
-
"loss": 0.
|
| 431 |
"step": 1475
|
| 432 |
},
|
| 433 |
{
|
| 434 |
-
"epoch": 0.
|
| 435 |
-
"grad_norm":
|
| 436 |
-
"learning_rate": 7.
|
| 437 |
-
"loss": 0.
|
| 438 |
"step": 1500
|
| 439 |
},
|
| 440 |
{
|
| 441 |
-
"epoch": 0.
|
| 442 |
-
"grad_norm": 6.
|
| 443 |
-
"learning_rate": 7.
|
| 444 |
-
"loss": 0.
|
| 445 |
"step": 1525
|
| 446 |
},
|
| 447 |
{
|
| 448 |
-
"epoch":
|
| 449 |
-
"grad_norm":
|
| 450 |
-
"learning_rate": 7.
|
| 451 |
-
"loss": 0.
|
| 452 |
"step": 1550
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch":
|
| 456 |
-
"grad_norm":
|
| 457 |
-
"learning_rate": 7.
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 1575
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"grad_norm":
|
| 464 |
-
"learning_rate": 7.
|
| 465 |
-
"loss": 0.
|
| 466 |
"step": 1600
|
| 467 |
},
|
| 468 |
{
|
| 469 |
-
"epoch":
|
| 470 |
-
"grad_norm": 6.
|
| 471 |
-
"learning_rate": 7.
|
| 472 |
-
"loss": 0.
|
| 473 |
"step": 1625
|
| 474 |
},
|
| 475 |
{
|
| 476 |
-
"epoch":
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 7.
|
| 479 |
-
"loss": 0.
|
| 480 |
"step": 1650
|
| 481 |
},
|
| 482 |
{
|
| 483 |
-
"epoch":
|
| 484 |
-
"grad_norm":
|
| 485 |
-
"learning_rate": 7.
|
| 486 |
-
"loss": 0.
|
| 487 |
"step": 1675
|
| 488 |
},
|
| 489 |
{
|
| 490 |
-
"epoch":
|
| 491 |
-
"grad_norm":
|
| 492 |
-
"learning_rate": 7.
|
| 493 |
-
"loss": 0.
|
| 494 |
"step": 1700
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch":
|
| 498 |
-
"grad_norm":
|
| 499 |
-
"learning_rate": 7.
|
| 500 |
-
"loss": 0.
|
| 501 |
"step": 1725
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
-
"grad_norm": 6.
|
| 506 |
-
"learning_rate": 7.
|
| 507 |
-
"loss": 0.
|
| 508 |
"step": 1750
|
| 509 |
},
|
| 510 |
{
|
| 511 |
-
"epoch":
|
| 512 |
-
"grad_norm":
|
| 513 |
-
"learning_rate": 7.
|
| 514 |
-
"loss": 0.
|
| 515 |
"step": 1775
|
| 516 |
},
|
| 517 |
{
|
| 518 |
-
"epoch":
|
| 519 |
-
"grad_norm":
|
| 520 |
-
"learning_rate": 7.
|
| 521 |
-
"loss": 0.
|
| 522 |
"step": 1800
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"epoch":
|
| 526 |
-
"grad_norm": 5.
|
| 527 |
-
"learning_rate": 7.
|
| 528 |
-
"loss": 0.
|
| 529 |
"step": 1825
|
| 530 |
},
|
| 531 |
{
|
| 532 |
-
"epoch": 1.
|
| 533 |
-
"grad_norm":
|
| 534 |
-
"learning_rate": 7.
|
| 535 |
-
"loss": 0.
|
| 536 |
"step": 1850
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch": 1.
|
| 540 |
-
"grad_norm":
|
| 541 |
-
"learning_rate": 6.
|
| 542 |
-
"loss": 0.
|
| 543 |
"step": 1875
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch": 1.
|
| 547 |
-
"grad_norm": 6.
|
| 548 |
-
"learning_rate": 6.
|
| 549 |
-
"loss": 0.
|
| 550 |
"step": 1900
|
| 551 |
},
|
| 552 |
{
|
| 553 |
-
"epoch": 1.
|
| 554 |
-
"grad_norm":
|
| 555 |
-
"learning_rate": 6.
|
| 556 |
-
"loss": 0.
|
| 557 |
"step": 1925
|
| 558 |
},
|
| 559 |
{
|
| 560 |
-
"epoch": 1.
|
| 561 |
-
"grad_norm": 5.
|
| 562 |
-
"learning_rate": 6.
|
| 563 |
-
"loss": 0.
|
| 564 |
"step": 1950
|
| 565 |
},
|
| 566 |
{
|
| 567 |
-
"epoch": 1.
|
| 568 |
-
"grad_norm": 5.
|
| 569 |
-
"learning_rate": 6.
|
| 570 |
-
"loss": 0.
|
| 571 |
"step": 1975
|
| 572 |
},
|
| 573 |
{
|
| 574 |
-
"epoch": 1.
|
| 575 |
-
"grad_norm":
|
| 576 |
-
"learning_rate": 6.
|
| 577 |
-
"loss": 0.
|
| 578 |
"step": 2000
|
| 579 |
},
|
| 580 |
{
|
| 581 |
-
"epoch": 1.
|
| 582 |
-
"eval_loss": 0.
|
| 583 |
-
"eval_runtime":
|
| 584 |
-
"eval_samples_per_second": 2.
|
| 585 |
-
"eval_steps_per_second": 0.
|
| 586 |
-
"eval_wer": 0.
|
| 587 |
"step": 2000
|
| 588 |
},
|
| 589 |
{
|
| 590 |
-
"epoch": 1.
|
| 591 |
-
"grad_norm":
|
| 592 |
-
"learning_rate": 6.
|
| 593 |
-
"loss": 0.
|
| 594 |
"step": 2025
|
| 595 |
},
|
| 596 |
{
|
| 597 |
-
"epoch": 1.
|
| 598 |
-
"grad_norm": 5.
|
| 599 |
-
"learning_rate": 6.
|
| 600 |
-
"loss": 0.
|
| 601 |
"step": 2050
|
| 602 |
},
|
| 603 |
{
|
| 604 |
-
"epoch": 1.
|
| 605 |
-
"grad_norm": 5.
|
| 606 |
-
"learning_rate": 6.
|
| 607 |
-
"loss": 0.
|
| 608 |
"step": 2075
|
| 609 |
},
|
| 610 |
{
|
| 611 |
-
"epoch": 1.
|
| 612 |
-
"grad_norm":
|
| 613 |
-
"learning_rate": 6.
|
| 614 |
-
"loss": 0.
|
| 615 |
"step": 2100
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch": 1.
|
| 619 |
-
"grad_norm": 5.
|
| 620 |
-
"learning_rate": 6.
|
| 621 |
-
"loss": 0.
|
| 622 |
"step": 2125
|
| 623 |
},
|
| 624 |
{
|
| 625 |
-
"epoch": 1.
|
| 626 |
-
"grad_norm": 5.
|
| 627 |
-
"learning_rate": 6.
|
| 628 |
-
"loss": 0.
|
| 629 |
"step": 2150
|
| 630 |
},
|
| 631 |
{
|
| 632 |
-
"epoch": 1.
|
| 633 |
-
"grad_norm": 5.
|
| 634 |
-
"learning_rate": 6.
|
| 635 |
-
"loss": 0.
|
| 636 |
"step": 2175
|
| 637 |
},
|
| 638 |
{
|
| 639 |
-
"epoch": 1.
|
| 640 |
-
"grad_norm":
|
| 641 |
-
"learning_rate": 6.
|
| 642 |
-
"loss": 0.
|
| 643 |
"step": 2200
|
| 644 |
},
|
| 645 |
{
|
| 646 |
-
"epoch": 1.
|
| 647 |
-
"grad_norm":
|
| 648 |
-
"learning_rate": 6.
|
| 649 |
-
"loss": 0.
|
| 650 |
"step": 2225
|
| 651 |
},
|
| 652 |
{
|
| 653 |
-
"epoch": 1.
|
| 654 |
-
"grad_norm": 5.
|
| 655 |
-
"learning_rate": 6.
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 2250
|
| 658 |
},
|
| 659 |
{
|
| 660 |
-
"epoch": 1.
|
| 661 |
-
"grad_norm":
|
| 662 |
-
"learning_rate": 6.
|
| 663 |
-
"loss": 0.
|
| 664 |
"step": 2275
|
| 665 |
},
|
| 666 |
{
|
| 667 |
-
"epoch": 1.
|
| 668 |
-
"grad_norm": 5.
|
| 669 |
-
"learning_rate": 6.
|
| 670 |
-
"loss": 0.
|
| 671 |
"step": 2300
|
| 672 |
},
|
| 673 |
{
|
| 674 |
-
"epoch": 1.
|
| 675 |
-
"grad_norm":
|
| 676 |
-
"learning_rate": 5.
|
| 677 |
-
"loss": 0.
|
| 678 |
"step": 2325
|
| 679 |
},
|
| 680 |
{
|
| 681 |
-
"epoch": 1.
|
| 682 |
-
"grad_norm": 5.
|
| 683 |
-
"learning_rate": 5.
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 2350
|
| 686 |
},
|
| 687 |
{
|
| 688 |
-
"epoch": 1.
|
| 689 |
-
"grad_norm":
|
| 690 |
-
"learning_rate": 5.
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 2375
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"epoch": 1.
|
| 696 |
-
"grad_norm":
|
| 697 |
-
"learning_rate": 5.
|
| 698 |
-
"loss": 0.
|
| 699 |
"step": 2400
|
| 700 |
},
|
| 701 |
{
|
| 702 |
-
"epoch": 1.
|
| 703 |
-
"grad_norm": 5.
|
| 704 |
-
"learning_rate": 5.
|
| 705 |
-
"loss": 0.
|
| 706 |
"step": 2425
|
| 707 |
},
|
| 708 |
{
|
| 709 |
-
"epoch": 1.
|
| 710 |
-
"grad_norm":
|
| 711 |
-
"learning_rate": 5.
|
| 712 |
-
"loss": 0.
|
| 713 |
"step": 2450
|
| 714 |
},
|
| 715 |
{
|
| 716 |
-
"epoch": 1.
|
| 717 |
-
"grad_norm":
|
| 718 |
-
"learning_rate": 5.
|
| 719 |
-
"loss": 0.
|
| 720 |
"step": 2475
|
| 721 |
},
|
| 722 |
{
|
| 723 |
-
"epoch": 1.
|
| 724 |
-
"grad_norm":
|
| 725 |
-
"learning_rate": 5.
|
| 726 |
-
"loss": 0.
|
| 727 |
"step": 2500
|
| 728 |
},
|
| 729 |
{
|
| 730 |
-
"epoch": 1.
|
| 731 |
-
"grad_norm":
|
| 732 |
-
"learning_rate": 5.
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 2525
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"epoch": 1.
|
| 738 |
-
"grad_norm": 5.
|
| 739 |
-
"learning_rate": 5.
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 2550
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch": 1.
|
| 745 |
-
"grad_norm":
|
| 746 |
-
"learning_rate": 5.
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 2575
|
| 749 |
},
|
| 750 |
{
|
| 751 |
-
"epoch": 1.
|
| 752 |
-
"grad_norm":
|
| 753 |
-
"learning_rate": 5.
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 2600
|
| 756 |
},
|
| 757 |
{
|
| 758 |
-
"epoch": 1.
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate": 5.
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 2625
|
| 763 |
},
|
| 764 |
{
|
| 765 |
-
"epoch": 1.
|
| 766 |
-
"grad_norm":
|
| 767 |
-
"learning_rate": 5.
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 2650
|
| 770 |
},
|
| 771 |
{
|
| 772 |
-
"epoch": 1.
|
| 773 |
-
"grad_norm": 5.
|
| 774 |
-
"learning_rate": 5.
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 2675
|
| 777 |
},
|
| 778 |
{
|
| 779 |
-
"epoch": 1.
|
| 780 |
-
"grad_norm":
|
| 781 |
-
"learning_rate": 5.
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 2700
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch": 1.
|
| 787 |
-
"grad_norm":
|
| 788 |
-
"learning_rate": 5.
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 2725
|
| 791 |
},
|
| 792 |
{
|
| 793 |
-
"epoch": 1.
|
| 794 |
-
"grad_norm":
|
| 795 |
-
"learning_rate": 5.
|
| 796 |
-
"loss": 0.
|
| 797 |
"step": 2750
|
| 798 |
},
|
| 799 |
{
|
| 800 |
-
"epoch": 1.
|
| 801 |
-
"grad_norm":
|
| 802 |
-
"learning_rate": 4.
|
| 803 |
-
"loss": 0.
|
| 804 |
"step": 2775
|
| 805 |
},
|
| 806 |
{
|
| 807 |
-
"epoch": 1.
|
| 808 |
-
"grad_norm":
|
| 809 |
-
"learning_rate": 4.
|
| 810 |
-
"loss": 0.
|
| 811 |
"step": 2800
|
| 812 |
},
|
| 813 |
{
|
| 814 |
-
"epoch": 1.
|
| 815 |
-
"grad_norm": 5.
|
| 816 |
-
"learning_rate": 4.
|
| 817 |
-
"loss": 0.
|
| 818 |
"step": 2825
|
| 819 |
},
|
| 820 |
{
|
| 821 |
-
"epoch": 1.
|
| 822 |
-
"grad_norm": 5.
|
| 823 |
-
"learning_rate": 4.
|
| 824 |
-
"loss": 0.
|
| 825 |
"step": 2850
|
| 826 |
},
|
| 827 |
{
|
| 828 |
-
"epoch": 1.
|
| 829 |
-
"grad_norm":
|
| 830 |
-
"learning_rate": 4.
|
| 831 |
-
"loss": 0.
|
| 832 |
"step": 2875
|
| 833 |
},
|
| 834 |
{
|
| 835 |
-
"epoch": 1.
|
| 836 |
-
"grad_norm":
|
| 837 |
-
"learning_rate": 4.
|
| 838 |
-
"loss": 0.
|
| 839 |
"step": 2900
|
| 840 |
},
|
| 841 |
{
|
| 842 |
-
"epoch": 1.
|
| 843 |
-
"grad_norm":
|
| 844 |
-
"learning_rate": 4.
|
| 845 |
-
"loss": 0.
|
| 846 |
"step": 2925
|
| 847 |
},
|
| 848 |
{
|
| 849 |
-
"epoch": 1.
|
| 850 |
-
"grad_norm":
|
| 851 |
-
"learning_rate": 4.
|
| 852 |
-
"loss": 0.
|
| 853 |
"step": 2950
|
| 854 |
},
|
| 855 |
{
|
| 856 |
-
"epoch": 1.
|
| 857 |
-
"grad_norm": 5.
|
| 858 |
-
"learning_rate": 4.
|
| 859 |
-
"loss": 0.
|
| 860 |
"step": 2975
|
| 861 |
},
|
| 862 |
{
|
| 863 |
-
"epoch": 1.
|
| 864 |
-
"grad_norm": 5.
|
| 865 |
-
"learning_rate": 4.
|
| 866 |
-
"loss": 0.
|
| 867 |
"step": 3000
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch": 1.
|
| 871 |
-
"eval_loss": 0.
|
| 872 |
-
"eval_runtime":
|
| 873 |
-
"eval_samples_per_second": 2.
|
| 874 |
-
"eval_steps_per_second": 0.
|
| 875 |
-
"eval_wer": 0.
|
| 876 |
"step": 3000
|
| 877 |
},
|
| 878 |
{
|
| 879 |
-
"epoch": 1.
|
| 880 |
-
"grad_norm":
|
| 881 |
-
"learning_rate": 4.
|
| 882 |
-
"loss": 0.
|
| 883 |
"step": 3025
|
| 884 |
},
|
| 885 |
{
|
| 886 |
-
"epoch": 1.
|
| 887 |
-
"grad_norm": 5.
|
| 888 |
-
"learning_rate": 4.
|
| 889 |
-
"loss": 0.
|
| 890 |
"step": 3050
|
| 891 |
},
|
| 892 |
{
|
| 893 |
-
"epoch":
|
| 894 |
-
"grad_norm":
|
| 895 |
-
"learning_rate": 4.
|
| 896 |
-
"loss": 0.
|
| 897 |
"step": 3075
|
| 898 |
},
|
| 899 |
{
|
| 900 |
-
"epoch":
|
| 901 |
-
"grad_norm": 5.
|
| 902 |
-
"learning_rate": 4.
|
| 903 |
-
"loss": 0.
|
| 904 |
"step": 3100
|
| 905 |
},
|
| 906 |
{
|
| 907 |
-
"epoch":
|
| 908 |
-
"grad_norm":
|
| 909 |
-
"learning_rate": 4.
|
| 910 |
-
"loss": 0.
|
| 911 |
"step": 3125
|
| 912 |
},
|
| 913 |
{
|
| 914 |
-
"epoch":
|
| 915 |
-
"grad_norm": 5.
|
| 916 |
-
"learning_rate": 4.
|
| 917 |
-
"loss": 0.
|
| 918 |
"step": 3150
|
| 919 |
},
|
| 920 |
{
|
| 921 |
-
"epoch":
|
| 922 |
-
"grad_norm": 5.
|
| 923 |
-
"learning_rate": 4.
|
| 924 |
-
"loss": 0.
|
| 925 |
"step": 3175
|
| 926 |
},
|
| 927 |
{
|
| 928 |
-
"epoch":
|
| 929 |
-
"grad_norm":
|
| 930 |
-
"learning_rate": 4.
|
| 931 |
-
"loss": 0.
|
| 932 |
"step": 3200
|
| 933 |
},
|
| 934 |
{
|
| 935 |
-
"epoch":
|
| 936 |
-
"grad_norm": 5.
|
| 937 |
-
"learning_rate": 3.
|
| 938 |
-
"loss": 0.
|
| 939 |
"step": 3225
|
| 940 |
},
|
| 941 |
{
|
| 942 |
-
"epoch":
|
| 943 |
-
"grad_norm": 5.
|
| 944 |
-
"learning_rate": 3.
|
| 945 |
-
"loss": 0.
|
| 946 |
"step": 3250
|
| 947 |
},
|
| 948 |
{
|
| 949 |
-
"epoch":
|
| 950 |
-
"grad_norm":
|
| 951 |
-
"learning_rate": 3.
|
| 952 |
-
"loss": 0.
|
| 953 |
"step": 3275
|
| 954 |
},
|
| 955 |
{
|
| 956 |
-
"epoch":
|
| 957 |
-
"grad_norm":
|
| 958 |
-
"learning_rate": 3.
|
| 959 |
-
"loss": 0.
|
| 960 |
"step": 3300
|
| 961 |
},
|
| 962 |
{
|
| 963 |
-
"epoch":
|
| 964 |
-
"grad_norm": 5.
|
| 965 |
-
"learning_rate": 3.
|
| 966 |
-
"loss": 0.
|
| 967 |
"step": 3325
|
| 968 |
},
|
| 969 |
{
|
| 970 |
-
"epoch":
|
| 971 |
-
"grad_norm":
|
| 972 |
-
"learning_rate": 3.
|
| 973 |
-
"loss": 0.
|
| 974 |
"step": 3350
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"epoch":
|
| 978 |
-
"grad_norm": 5.
|
| 979 |
-
"learning_rate": 3.
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 3375
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch":
|
| 985 |
-
"grad_norm":
|
| 986 |
-
"learning_rate": 3.
|
| 987 |
-
"loss": 0.
|
| 988 |
"step": 3400
|
| 989 |
},
|
| 990 |
{
|
| 991 |
-
"epoch":
|
| 992 |
-
"grad_norm":
|
| 993 |
-
"learning_rate": 3.
|
| 994 |
-
"loss": 0.
|
| 995 |
"step": 3425
|
| 996 |
},
|
| 997 |
{
|
| 998 |
-
"epoch":
|
| 999 |
-
"grad_norm": 5.
|
| 1000 |
-
"learning_rate": 3.
|
| 1001 |
-
"loss": 0.
|
| 1002 |
"step": 3450
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
-
"epoch":
|
| 1006 |
-
"grad_norm":
|
| 1007 |
-
"learning_rate": 3.
|
| 1008 |
-
"loss": 0.
|
| 1009 |
"step": 3475
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
-
"epoch":
|
| 1013 |
-
"grad_norm": 4.
|
| 1014 |
-
"learning_rate": 3.
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 3500
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
-
"epoch":
|
| 1020 |
-
"grad_norm": 5.
|
| 1021 |
-
"learning_rate": 3.
|
| 1022 |
-
"loss": 0.
|
| 1023 |
"step": 3525
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
-
"epoch":
|
| 1027 |
-
"grad_norm": 5.
|
| 1028 |
-
"learning_rate": 3.
|
| 1029 |
-
"loss": 0.
|
| 1030 |
"step": 3550
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
-
"epoch":
|
| 1034 |
-
"grad_norm":
|
| 1035 |
-
"learning_rate": 3.
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 3575
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
-
"epoch":
|
| 1041 |
-
"grad_norm":
|
| 1042 |
-
"learning_rate": 3.
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 3600
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
-
"epoch":
|
| 1048 |
-
"grad_norm": 5.
|
| 1049 |
-
"learning_rate": 3.
|
| 1050 |
-
"loss": 0.
|
| 1051 |
"step": 3625
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
-
"epoch":
|
| 1055 |
-
"grad_norm":
|
| 1056 |
-
"learning_rate": 3.
|
| 1057 |
-
"loss": 0.
|
| 1058 |
"step": 3650
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
-
"epoch":
|
| 1062 |
-
"grad_norm":
|
| 1063 |
-
"learning_rate": 2.
|
| 1064 |
-
"loss": 0.
|
| 1065 |
"step": 3675
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
-
"epoch": 2.
|
| 1069 |
-
"grad_norm":
|
| 1070 |
-
"learning_rate": 2.
|
| 1071 |
-
"loss": 0.
|
| 1072 |
"step": 3700
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
-
"epoch": 2.
|
| 1076 |
-
"grad_norm":
|
| 1077 |
-
"learning_rate": 2.
|
| 1078 |
-
"loss": 0.
|
| 1079 |
"step": 3725
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
-
"epoch": 2.
|
| 1083 |
-
"grad_norm":
|
| 1084 |
-
"learning_rate": 2.
|
| 1085 |
-
"loss": 0.
|
| 1086 |
"step": 3750
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
-
"epoch": 2.
|
| 1090 |
-
"grad_norm":
|
| 1091 |
-
"learning_rate": 2.
|
| 1092 |
-
"loss": 0.
|
| 1093 |
"step": 3775
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
-
"epoch": 2.
|
| 1097 |
-
"grad_norm":
|
| 1098 |
-
"learning_rate": 2.
|
| 1099 |
-
"loss": 0.
|
| 1100 |
"step": 3800
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
-
"epoch": 2.
|
| 1104 |
-
"grad_norm":
|
| 1105 |
-
"learning_rate": 2.
|
| 1106 |
-
"loss": 0.
|
| 1107 |
"step": 3825
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
-
"epoch": 2.
|
| 1111 |
-
"grad_norm":
|
| 1112 |
-
"learning_rate": 2.
|
| 1113 |
-
"loss": 0.
|
| 1114 |
"step": 3850
|
| 1115 |
},
|
| 1116 |
{
|
| 1117 |
-
"epoch": 2.
|
| 1118 |
-
"grad_norm": 5.
|
| 1119 |
-
"learning_rate": 2.
|
| 1120 |
-
"loss": 0.
|
| 1121 |
"step": 3875
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
-
"epoch": 2.
|
| 1125 |
-
"grad_norm":
|
| 1126 |
-
"learning_rate": 2.
|
| 1127 |
-
"loss": 0.
|
| 1128 |
"step": 3900
|
| 1129 |
},
|
| 1130 |
{
|
| 1131 |
-
"epoch": 2.
|
| 1132 |
-
"grad_norm":
|
| 1133 |
-
"learning_rate": 2.
|
| 1134 |
-
"loss": 0.
|
| 1135 |
"step": 3925
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
-
"epoch": 2.
|
| 1139 |
-
"grad_norm":
|
| 1140 |
-
"learning_rate": 2.
|
| 1141 |
-
"loss": 0.
|
| 1142 |
"step": 3950
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
-
"epoch": 2.
|
| 1146 |
-
"grad_norm": 4.
|
| 1147 |
-
"learning_rate": 2.
|
| 1148 |
-
"loss": 0.
|
| 1149 |
"step": 3975
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
-
"epoch": 2.
|
| 1153 |
-
"grad_norm": 4.
|
| 1154 |
-
"learning_rate": 2.
|
| 1155 |
-
"loss": 0.
|
| 1156 |
"step": 4000
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
-
"epoch": 2.
|
| 1160 |
-
"eval_loss": 0.
|
| 1161 |
-
"eval_runtime":
|
| 1162 |
-
"eval_samples_per_second": 2.
|
| 1163 |
-
"eval_steps_per_second": 0.
|
| 1164 |
-
"eval_wer": 0.
|
| 1165 |
"step": 4000
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
-
"epoch": 2.
|
| 1169 |
-
"grad_norm": 5.
|
| 1170 |
-
"learning_rate": 2.
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 4025
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
-
"epoch": 2.
|
| 1176 |
-
"grad_norm":
|
| 1177 |
-
"learning_rate": 2.
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 4050
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch": 2.
|
| 1183 |
-
"grad_norm": 5.
|
| 1184 |
-
"learning_rate": 2.
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 4075
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
-
"epoch": 2.
|
| 1190 |
-
"grad_norm": 4.
|
| 1191 |
-
"learning_rate": 2.
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 4100
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch": 2.
|
| 1197 |
-
"grad_norm": 4.
|
| 1198 |
-
"learning_rate": 1.
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 4125
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
-
"epoch": 2.
|
| 1204 |
-
"grad_norm":
|
| 1205 |
-
"learning_rate": 1.
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 4150
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
-
"epoch": 2.
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate": 1.
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 4175
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
-
"epoch": 2.
|
| 1218 |
-
"grad_norm": 4.
|
| 1219 |
-
"learning_rate": 1.
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 4200
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
-
"epoch": 2.
|
| 1225 |
-
"grad_norm": 4.
|
| 1226 |
-
"learning_rate": 1.
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 4225
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
-
"epoch": 2.
|
| 1232 |
-
"grad_norm":
|
| 1233 |
-
"learning_rate": 1.
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 4250
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
-
"epoch": 2.
|
| 1239 |
-
"grad_norm": 4.
|
| 1240 |
-
"learning_rate": 1.
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 4275
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
-
"epoch": 2.
|
| 1246 |
-
"grad_norm":
|
| 1247 |
-
"learning_rate": 1.
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 4300
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
-
"epoch": 2.
|
| 1253 |
-
"grad_norm":
|
| 1254 |
-
"learning_rate": 1.
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 4325
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
-
"epoch": 2.
|
| 1260 |
-
"grad_norm":
|
| 1261 |
-
"learning_rate": 1.
|
| 1262 |
-
"loss": 0.
|
| 1263 |
"step": 4350
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
-
"epoch": 2.
|
| 1267 |
-
"grad_norm": 4.
|
| 1268 |
-
"learning_rate": 1.
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 4375
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
-
"epoch": 2.
|
| 1274 |
-
"grad_norm":
|
| 1275 |
-
"learning_rate": 1.
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 4400
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
-
"epoch": 2.
|
| 1281 |
-
"grad_norm":
|
| 1282 |
-
"learning_rate": 1.
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 4425
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
-
"epoch": 2.
|
| 1288 |
-
"grad_norm": 5.
|
| 1289 |
-
"learning_rate": 1.
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 4450
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
-
"epoch": 2.
|
| 1295 |
-
"grad_norm": 4.
|
| 1296 |
-
"learning_rate": 1.
|
| 1297 |
-
"loss": 0.
|
| 1298 |
"step": 4475
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
-
"epoch": 2.
|
| 1302 |
-
"grad_norm":
|
| 1303 |
-
"learning_rate": 1.
|
| 1304 |
-
"loss": 0.
|
| 1305 |
"step": 4500
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
-
"epoch": 2.
|
| 1309 |
-
"grad_norm": 5.
|
| 1310 |
-
"learning_rate": 1.
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 4525
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
-
"epoch": 2.
|
| 1316 |
-
"grad_norm":
|
| 1317 |
-
"learning_rate": 1.
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 4550
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
-
"epoch": 2.
|
| 1323 |
-
"grad_norm":
|
| 1324 |
-
"learning_rate": 9.
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 4575
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
-
"epoch":
|
| 1330 |
-
"grad_norm": 4.
|
| 1331 |
-
"learning_rate": 8.
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 4600
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
-
"epoch":
|
| 1337 |
-
"grad_norm":
|
| 1338 |
-
"learning_rate": 8.
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 4625
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
-
"epoch":
|
| 1344 |
-
"grad_norm": 4.
|
| 1345 |
-
"learning_rate": 7.
|
| 1346 |
-
"loss": 0.
|
| 1347 |
"step": 4650
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
-
"epoch":
|
| 1351 |
-
"grad_norm":
|
| 1352 |
-
"learning_rate": 7.
|
| 1353 |
-
"loss": 0.
|
| 1354 |
"step": 4675
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
-
"epoch":
|
| 1358 |
-
"grad_norm": 4.
|
| 1359 |
-
"learning_rate": 6.
|
| 1360 |
-
"loss": 0.
|
| 1361 |
"step": 4700
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
-
"epoch":
|
| 1365 |
-
"grad_norm":
|
| 1366 |
-
"learning_rate": 6.
|
| 1367 |
-
"loss": 0.
|
| 1368 |
"step": 4725
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
-
"epoch":
|
| 1372 |
-
"grad_norm":
|
| 1373 |
-
"learning_rate": 5.
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 4750
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
-
"epoch":
|
| 1379 |
-
"grad_norm": 4.
|
| 1380 |
-
"learning_rate": 5.
|
| 1381 |
-
"loss": 0.
|
| 1382 |
"step": 4775
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
-
"epoch":
|
| 1386 |
-
"grad_norm":
|
| 1387 |
-
"learning_rate": 4.
|
| 1388 |
-
"loss": 0.
|
| 1389 |
"step": 4800
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
-
"epoch":
|
| 1393 |
-
"grad_norm":
|
| 1394 |
-
"learning_rate": 3.
|
| 1395 |
-
"loss": 0.
|
| 1396 |
"step": 4825
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
-
"epoch":
|
| 1400 |
-
"grad_norm": 5.
|
| 1401 |
-
"learning_rate": 3.
|
| 1402 |
-
"loss": 0.
|
| 1403 |
"step": 4850
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
-
"epoch":
|
| 1407 |
-
"grad_norm":
|
| 1408 |
-
"learning_rate": 2.
|
| 1409 |
-
"loss": 0.
|
| 1410 |
"step": 4875
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
-
"epoch":
|
| 1414 |
-
"grad_norm":
|
| 1415 |
-
"learning_rate": 2.
|
| 1416 |
-
"loss": 0.
|
| 1417 |
"step": 4900
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
-
"epoch":
|
| 1421 |
-
"grad_norm":
|
| 1422 |
-
"learning_rate": 1.
|
| 1423 |
-
"loss": 0.
|
| 1424 |
"step": 4925
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
-
"epoch":
|
| 1428 |
-
"grad_norm": 4.
|
| 1429 |
-
"learning_rate": 1.
|
| 1430 |
-
"loss": 0.
|
| 1431 |
"step": 4950
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
-
"epoch":
|
| 1435 |
-
"grad_norm":
|
| 1436 |
-
"learning_rate":
|
| 1437 |
-
"loss": 0.
|
| 1438 |
"step": 4975
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
-
"epoch":
|
| 1442 |
-
"grad_norm":
|
| 1443 |
-
"learning_rate":
|
| 1444 |
-
"loss": 0.
|
| 1445 |
"step": 5000
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
-
"epoch":
|
| 1449 |
-
"eval_loss": 0.
|
| 1450 |
-
"eval_runtime":
|
| 1451 |
-
"eval_samples_per_second": 2.
|
| 1452 |
-
"eval_steps_per_second": 0.
|
| 1453 |
-
"eval_wer": 0.
|
| 1454 |
"step": 5000
|
| 1455 |
},
|
| 1456 |
{
|
| 1457 |
-
"epoch":
|
| 1458 |
"step": 5000,
|
| 1459 |
-
"total_flos": 5.
|
| 1460 |
-
"train_loss": 0.
|
| 1461 |
-
"train_runtime":
|
| 1462 |
-
"train_samples_per_second": 2.
|
| 1463 |
-
"train_steps_per_second": 0.
|
| 1464 |
}
|
| 1465 |
],
|
| 1466 |
"logging_steps": 25,
|
| 1467 |
"max_steps": 5000,
|
| 1468 |
"num_input_tokens_seen": 0,
|
| 1469 |
-
"num_train_epochs":
|
| 1470 |
"save_steps": 1000,
|
| 1471 |
"stateful_callbacks": {
|
| 1472 |
"TrainerControl": {
|
|
@@ -1480,7 +1480,7 @@
|
|
| 1480 |
"attributes": {}
|
| 1481 |
}
|
| 1482 |
},
|
| 1483 |
-
"total_flos": 5.
|
| 1484 |
"train_batch_size": 16,
|
| 1485 |
"trial_name": null,
|
| 1486 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.7056277056277054,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.013528138528138528,
|
| 13 |
+
"grad_norm": 18.23776626586914,
|
| 14 |
"learning_rate": 4.800000000000001e-07,
|
| 15 |
+
"loss": 1.7881,
|
| 16 |
"step": 25
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.027056277056277056,
|
| 20 |
+
"grad_norm": 11.490796089172363,
|
| 21 |
"learning_rate": 9.800000000000001e-07,
|
| 22 |
+
"loss": 1.4477,
|
| 23 |
"step": 50
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.040584415584415584,
|
| 27 |
+
"grad_norm": 10.601150512695312,
|
| 28 |
"learning_rate": 1.48e-06,
|
| 29 |
+
"loss": 1.0989,
|
| 30 |
"step": 75
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.05411255411255411,
|
| 34 |
+
"grad_norm": 12.294251441955566,
|
| 35 |
"learning_rate": 1.98e-06,
|
| 36 |
+
"loss": 0.9367,
|
| 37 |
"step": 100
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.06764069264069264,
|
| 41 |
+
"grad_norm": 11.920494079589844,
|
| 42 |
"learning_rate": 2.4800000000000004e-06,
|
| 43 |
+
"loss": 0.8698,
|
| 44 |
"step": 125
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.08116883116883117,
|
| 48 |
+
"grad_norm": 11.758705139160156,
|
| 49 |
"learning_rate": 2.9800000000000003e-06,
|
| 50 |
+
"loss": 0.8531,
|
| 51 |
"step": 150
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.0946969696969697,
|
| 55 |
+
"grad_norm": 11.037556648254395,
|
| 56 |
"learning_rate": 3.48e-06,
|
| 57 |
+
"loss": 0.812,
|
| 58 |
"step": 175
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.10822510822510822,
|
| 62 |
+
"grad_norm": 10.065262794494629,
|
| 63 |
"learning_rate": 3.980000000000001e-06,
|
| 64 |
+
"loss": 0.7987,
|
| 65 |
"step": 200
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.12175324675324675,
|
| 69 |
+
"grad_norm": 9.124336242675781,
|
| 70 |
"learning_rate": 4.48e-06,
|
| 71 |
+
"loss": 0.7455,
|
| 72 |
"step": 225
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.13528138528138528,
|
| 76 |
+
"grad_norm": 10.971399307250977,
|
| 77 |
"learning_rate": 4.980000000000001e-06,
|
| 78 |
+
"loss": 0.7564,
|
| 79 |
"step": 250
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.1488095238095238,
|
| 83 |
+
"grad_norm": 9.226423263549805,
|
| 84 |
"learning_rate": 5.480000000000001e-06,
|
| 85 |
+
"loss": 0.7163,
|
| 86 |
"step": 275
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.16233766233766234,
|
| 90 |
+
"grad_norm": 9.523015022277832,
|
| 91 |
"learning_rate": 5.98e-06,
|
| 92 |
+
"loss": 0.7341,
|
| 93 |
"step": 300
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.17586580086580086,
|
| 97 |
+
"grad_norm": 8.390256881713867,
|
| 98 |
"learning_rate": 6.480000000000001e-06,
|
| 99 |
+
"loss": 0.7301,
|
| 100 |
"step": 325
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.1893939393939394,
|
| 104 |
+
"grad_norm": 9.996743202209473,
|
| 105 |
"learning_rate": 6.98e-06,
|
| 106 |
+
"loss": 0.6897,
|
| 107 |
"step": 350
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.20292207792207792,
|
| 111 |
+
"grad_norm": 9.470787048339844,
|
| 112 |
"learning_rate": 7.48e-06,
|
| 113 |
+
"loss": 0.6728,
|
| 114 |
"step": 375
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.21645021645021645,
|
| 118 |
+
"grad_norm": 8.221435546875,
|
| 119 |
"learning_rate": 7.980000000000002e-06,
|
| 120 |
+
"loss": 0.6853,
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.22997835497835498,
|
| 125 |
+
"grad_norm": 9.243407249450684,
|
| 126 |
"learning_rate": 8.48e-06,
|
| 127 |
+
"loss": 0.6899,
|
| 128 |
"step": 425
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.2435064935064935,
|
| 132 |
+
"grad_norm": 8.308032989501953,
|
| 133 |
+
"learning_rate": 8.96e-06,
|
| 134 |
+
"loss": 0.634,
|
| 135 |
"step": 450
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.25703463203463206,
|
| 139 |
+
"grad_norm": 8.970362663269043,
|
| 140 |
+
"learning_rate": 9.460000000000001e-06,
|
| 141 |
+
"loss": 0.6191,
|
| 142 |
"step": 475
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.27056277056277056,
|
| 146 |
+
"grad_norm": 9.167222023010254,
|
| 147 |
+
"learning_rate": 9.960000000000001e-06,
|
| 148 |
+
"loss": 0.6348,
|
| 149 |
"step": 500
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.2840909090909091,
|
| 153 |
+
"grad_norm": 8.824662208557129,
|
| 154 |
+
"learning_rate": 9.94888888888889e-06,
|
| 155 |
+
"loss": 0.6228,
|
| 156 |
"step": 525
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.2976190476190476,
|
| 160 |
+
"grad_norm": 10.600458145141602,
|
| 161 |
+
"learning_rate": 9.893333333333334e-06,
|
| 162 |
+
"loss": 0.622,
|
| 163 |
"step": 550
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.31114718614718617,
|
| 167 |
+
"grad_norm": 10.680913925170898,
|
| 168 |
+
"learning_rate": 9.837777777777778e-06,
|
| 169 |
+
"loss": 0.6115,
|
| 170 |
"step": 575
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.3246753246753247,
|
| 174 |
+
"grad_norm": 8.21044635772705,
|
| 175 |
+
"learning_rate": 9.782222222222222e-06,
|
| 176 |
+
"loss": 0.6219,
|
| 177 |
"step": 600
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.33820346320346323,
|
| 181 |
+
"grad_norm": 8.558358192443848,
|
| 182 |
+
"learning_rate": 9.726666666666668e-06,
|
| 183 |
+
"loss": 0.5936,
|
| 184 |
"step": 625
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.35173160173160173,
|
| 188 |
+
"grad_norm": 6.6742095947265625,
|
| 189 |
+
"learning_rate": 9.671111111111112e-06,
|
| 190 |
+
"loss": 0.5871,
|
| 191 |
"step": 650
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.3652597402597403,
|
| 195 |
+
"grad_norm": 8.321733474731445,
|
| 196 |
+
"learning_rate": 9.615555555555558e-06,
|
| 197 |
+
"loss": 0.5767,
|
| 198 |
"step": 675
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.3787878787878788,
|
| 202 |
+
"grad_norm": 8.88350772857666,
|
| 203 |
+
"learning_rate": 9.56e-06,
|
| 204 |
+
"loss": 0.5387,
|
| 205 |
"step": 700
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.39231601731601734,
|
| 209 |
+
"grad_norm": 8.155245780944824,
|
| 210 |
+
"learning_rate": 9.504444444444446e-06,
|
| 211 |
+
"loss": 0.5809,
|
| 212 |
"step": 725
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.40584415584415584,
|
| 216 |
+
"grad_norm": 8.907155990600586,
|
| 217 |
+
"learning_rate": 9.44888888888889e-06,
|
| 218 |
+
"loss": 0.5778,
|
| 219 |
"step": 750
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.4193722943722944,
|
| 223 |
+
"grad_norm": 9.427032470703125,
|
| 224 |
+
"learning_rate": 9.393333333333334e-06,
|
| 225 |
+
"loss": 0.5776,
|
| 226 |
"step": 775
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.4329004329004329,
|
| 230 |
+
"grad_norm": 6.904598236083984,
|
| 231 |
+
"learning_rate": 9.33777777777778e-06,
|
| 232 |
+
"loss": 0.5287,
|
| 233 |
"step": 800
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.44642857142857145,
|
| 237 |
+
"grad_norm": 7.866734504699707,
|
| 238 |
+
"learning_rate": 9.282222222222222e-06,
|
| 239 |
+
"loss": 0.5555,
|
| 240 |
"step": 825
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.45995670995670995,
|
| 244 |
+
"grad_norm": 7.301151752471924,
|
| 245 |
+
"learning_rate": 9.226666666666668e-06,
|
| 246 |
+
"loss": 0.5296,
|
| 247 |
"step": 850
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 0.4734848484848485,
|
| 251 |
+
"grad_norm": 9.370705604553223,
|
| 252 |
+
"learning_rate": 9.171111111111112e-06,
|
| 253 |
+
"loss": 0.5516,
|
| 254 |
"step": 875
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.487012987012987,
|
| 258 |
+
"grad_norm": 7.46251916885376,
|
| 259 |
+
"learning_rate": 9.115555555555556e-06,
|
| 260 |
+
"loss": 0.534,
|
| 261 |
"step": 900
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.5005411255411255,
|
| 265 |
+
"grad_norm": 7.885534286499023,
|
| 266 |
+
"learning_rate": 9.060000000000001e-06,
|
| 267 |
+
"loss": 0.5113,
|
| 268 |
"step": 925
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 0.5140692640692641,
|
| 272 |
+
"grad_norm": 6.12823486328125,
|
| 273 |
+
"learning_rate": 9.004444444444445e-06,
|
| 274 |
+
"loss": 0.5274,
|
| 275 |
"step": 950
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.5275974025974026,
|
| 279 |
+
"grad_norm": 7.1373515129089355,
|
| 280 |
+
"learning_rate": 8.94888888888889e-06,
|
| 281 |
+
"loss": 0.5241,
|
| 282 |
"step": 975
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 0.5411255411255411,
|
| 286 |
+
"grad_norm": 7.099331378936768,
|
| 287 |
+
"learning_rate": 8.893333333333333e-06,
|
| 288 |
+
"loss": 0.5152,
|
| 289 |
"step": 1000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 0.5411255411255411,
|
| 293 |
+
"eval_loss": 0.4954243302345276,
|
| 294 |
+
"eval_runtime": 1779.4215,
|
| 295 |
+
"eval_samples_per_second": 2.192,
|
| 296 |
+
"eval_steps_per_second": 0.137,
|
| 297 |
+
"eval_wer": 0.3535207186322805,
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 0.5546536796536796,
|
| 302 |
+
"grad_norm": 8.30470085144043,
|
| 303 |
+
"learning_rate": 8.83777777777778e-06,
|
| 304 |
+
"loss": 0.531,
|
| 305 |
"step": 1025
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 0.5681818181818182,
|
| 309 |
+
"grad_norm": 6.959774971008301,
|
| 310 |
+
"learning_rate": 8.782222222222223e-06,
|
| 311 |
+
"loss": 0.5038,
|
| 312 |
"step": 1050
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 0.5817099567099567,
|
| 316 |
+
"grad_norm": 7.844577789306641,
|
| 317 |
+
"learning_rate": 8.726666666666667e-06,
|
| 318 |
+
"loss": 0.5196,
|
| 319 |
"step": 1075
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 0.5952380952380952,
|
| 323 |
+
"grad_norm": 6.599257946014404,
|
| 324 |
+
"learning_rate": 8.671111111111113e-06,
|
| 325 |
+
"loss": 0.4982,
|
| 326 |
"step": 1100
|
| 327 |
},
|
| 328 |
{
|
| 329 |
+
"epoch": 0.6087662337662337,
|
| 330 |
+
"grad_norm": 5.671600818634033,
|
| 331 |
+
"learning_rate": 8.615555555555555e-06,
|
| 332 |
+
"loss": 0.497,
|
| 333 |
"step": 1125
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 0.6222943722943723,
|
| 337 |
+
"grad_norm": 6.545307636260986,
|
| 338 |
+
"learning_rate": 8.560000000000001e-06,
|
| 339 |
+
"loss": 0.4875,
|
| 340 |
"step": 1150
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"epoch": 0.6358225108225108,
|
| 344 |
+
"grad_norm": 6.877360820770264,
|
| 345 |
+
"learning_rate": 8.504444444444445e-06,
|
| 346 |
+
"loss": 0.5162,
|
| 347 |
"step": 1175
|
| 348 |
},
|
| 349 |
{
|
| 350 |
+
"epoch": 0.6493506493506493,
|
| 351 |
+
"grad_norm": 7.325205326080322,
|
| 352 |
+
"learning_rate": 8.448888888888889e-06,
|
| 353 |
+
"loss": 0.5151,
|
| 354 |
"step": 1200
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 0.6628787878787878,
|
| 358 |
+
"grad_norm": 6.775233745574951,
|
| 359 |
+
"learning_rate": 8.393333333333335e-06,
|
| 360 |
+
"loss": 0.498,
|
| 361 |
"step": 1225
|
| 362 |
},
|
| 363 |
{
|
| 364 |
+
"epoch": 0.6764069264069265,
|
| 365 |
+
"grad_norm": 7.457151412963867,
|
| 366 |
+
"learning_rate": 8.337777777777777e-06,
|
| 367 |
+
"loss": 0.5012,
|
| 368 |
"step": 1250
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 0.689935064935065,
|
| 372 |
+
"grad_norm": 7.285881042480469,
|
| 373 |
+
"learning_rate": 8.282222222222223e-06,
|
| 374 |
+
"loss": 0.4684,
|
| 375 |
"step": 1275
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 0.7034632034632035,
|
| 379 |
+
"grad_norm": 9.163443565368652,
|
| 380 |
+
"learning_rate": 8.226666666666667e-06,
|
| 381 |
+
"loss": 0.5079,
|
| 382 |
"step": 1300
|
| 383 |
},
|
| 384 |
{
|
| 385 |
+
"epoch": 0.716991341991342,
|
| 386 |
+
"grad_norm": 7.168745994567871,
|
| 387 |
+
"learning_rate": 8.171111111111113e-06,
|
| 388 |
+
"loss": 0.475,
|
| 389 |
"step": 1325
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 0.7305194805194806,
|
| 393 |
+
"grad_norm": 7.457499027252197,
|
| 394 |
+
"learning_rate": 8.115555555555557e-06,
|
| 395 |
+
"loss": 0.488,
|
| 396 |
"step": 1350
|
| 397 |
},
|
| 398 |
{
|
| 399 |
+
"epoch": 0.7440476190476191,
|
| 400 |
+
"grad_norm": 6.2372846603393555,
|
| 401 |
+
"learning_rate": 8.06e-06,
|
| 402 |
+
"loss": 0.4822,
|
| 403 |
"step": 1375
|
| 404 |
},
|
| 405 |
{
|
| 406 |
+
"epoch": 0.7575757575757576,
|
| 407 |
+
"grad_norm": 5.880990505218506,
|
| 408 |
+
"learning_rate": 8.004444444444445e-06,
|
| 409 |
+
"loss": 0.531,
|
| 410 |
"step": 1400
|
| 411 |
},
|
| 412 |
{
|
| 413 |
+
"epoch": 0.7711038961038961,
|
| 414 |
+
"grad_norm": 7.057967185974121,
|
| 415 |
+
"learning_rate": 7.948888888888889e-06,
|
| 416 |
+
"loss": 0.4872,
|
| 417 |
"step": 1425
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 0.7846320346320347,
|
| 421 |
+
"grad_norm": 7.299345970153809,
|
| 422 |
+
"learning_rate": 7.893333333333335e-06,
|
| 423 |
+
"loss": 0.4749,
|
| 424 |
"step": 1450
|
| 425 |
},
|
| 426 |
{
|
| 427 |
+
"epoch": 0.7981601731601732,
|
| 428 |
+
"grad_norm": 6.807291030883789,
|
| 429 |
+
"learning_rate": 7.837777777777779e-06,
|
| 430 |
+
"loss": 0.4676,
|
| 431 |
"step": 1475
|
| 432 |
},
|
| 433 |
{
|
| 434 |
+
"epoch": 0.8116883116883117,
|
| 435 |
+
"grad_norm": 5.556617736816406,
|
| 436 |
+
"learning_rate": 7.782222222222223e-06,
|
| 437 |
+
"loss": 0.4614,
|
| 438 |
"step": 1500
|
| 439 |
},
|
| 440 |
{
|
| 441 |
+
"epoch": 0.8252164502164502,
|
| 442 |
+
"grad_norm": 6.165937900543213,
|
| 443 |
+
"learning_rate": 7.726666666666667e-06,
|
| 444 |
+
"loss": 0.4459,
|
| 445 |
"step": 1525
|
| 446 |
},
|
| 447 |
{
|
| 448 |
+
"epoch": 0.8387445887445888,
|
| 449 |
+
"grad_norm": 6.99851655960083,
|
| 450 |
+
"learning_rate": 7.67111111111111e-06,
|
| 451 |
+
"loss": 0.4734,
|
| 452 |
"step": 1550
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 0.8522727272727273,
|
| 456 |
+
"grad_norm": 7.385776519775391,
|
| 457 |
+
"learning_rate": 7.6155555555555564e-06,
|
| 458 |
+
"loss": 0.4547,
|
| 459 |
"step": 1575
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 0.8658008658008658,
|
| 463 |
+
"grad_norm": 6.626092910766602,
|
| 464 |
+
"learning_rate": 7.5600000000000005e-06,
|
| 465 |
+
"loss": 0.4462,
|
| 466 |
"step": 1600
|
| 467 |
},
|
| 468 |
{
|
| 469 |
+
"epoch": 0.8793290043290043,
|
| 470 |
+
"grad_norm": 6.563342094421387,
|
| 471 |
+
"learning_rate": 7.504444444444445e-06,
|
| 472 |
+
"loss": 0.4511,
|
| 473 |
"step": 1625
|
| 474 |
},
|
| 475 |
{
|
| 476 |
+
"epoch": 0.8928571428571429,
|
| 477 |
+
"grad_norm": 9.904861450195312,
|
| 478 |
+
"learning_rate": 7.44888888888889e-06,
|
| 479 |
+
"loss": 0.4728,
|
| 480 |
"step": 1650
|
| 481 |
},
|
| 482 |
{
|
| 483 |
+
"epoch": 0.9063852813852814,
|
| 484 |
+
"grad_norm": 7.5107622146606445,
|
| 485 |
+
"learning_rate": 7.393333333333333e-06,
|
| 486 |
+
"loss": 0.4867,
|
| 487 |
"step": 1675
|
| 488 |
},
|
| 489 |
{
|
| 490 |
+
"epoch": 0.9199134199134199,
|
| 491 |
+
"grad_norm": 6.618627548217773,
|
| 492 |
+
"learning_rate": 7.337777777777778e-06,
|
| 493 |
+
"loss": 0.4512,
|
| 494 |
"step": 1700
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 0.9334415584415584,
|
| 498 |
+
"grad_norm": 7.19182014465332,
|
| 499 |
+
"learning_rate": 7.282222222222222e-06,
|
| 500 |
+
"loss": 0.4657,
|
| 501 |
"step": 1725
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 0.946969696969697,
|
| 505 |
+
"grad_norm": 6.207240104675293,
|
| 506 |
+
"learning_rate": 7.226666666666667e-06,
|
| 507 |
+
"loss": 0.4455,
|
| 508 |
"step": 1750
|
| 509 |
},
|
| 510 |
{
|
| 511 |
+
"epoch": 0.9604978354978355,
|
| 512 |
+
"grad_norm": 8.109068870544434,
|
| 513 |
+
"learning_rate": 7.171111111111112e-06,
|
| 514 |
+
"loss": 0.4744,
|
| 515 |
"step": 1775
|
| 516 |
},
|
| 517 |
{
|
| 518 |
+
"epoch": 0.974025974025974,
|
| 519 |
+
"grad_norm": 7.550827503204346,
|
| 520 |
+
"learning_rate": 7.115555555555557e-06,
|
| 521 |
+
"loss": 0.4565,
|
| 522 |
"step": 1800
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"epoch": 0.9875541125541125,
|
| 526 |
+
"grad_norm": 5.667859077453613,
|
| 527 |
+
"learning_rate": 7.06e-06,
|
| 528 |
+
"loss": 0.4368,
|
| 529 |
"step": 1825
|
| 530 |
},
|
| 531 |
{
|
| 532 |
+
"epoch": 1.001082251082251,
|
| 533 |
+
"grad_norm": 5.609886646270752,
|
| 534 |
+
"learning_rate": 7.004444444444445e-06,
|
| 535 |
+
"loss": 0.4406,
|
| 536 |
"step": 1850
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 1.0146103896103895,
|
| 540 |
+
"grad_norm": 4.862238883972168,
|
| 541 |
+
"learning_rate": 6.948888888888889e-06,
|
| 542 |
+
"loss": 0.3379,
|
| 543 |
"step": 1875
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 1.0281385281385282,
|
| 547 |
+
"grad_norm": 6.2563066482543945,
|
| 548 |
+
"learning_rate": 6.893333333333334e-06,
|
| 549 |
+
"loss": 0.3386,
|
| 550 |
"step": 1900
|
| 551 |
},
|
| 552 |
{
|
| 553 |
+
"epoch": 1.0416666666666667,
|
| 554 |
+
"grad_norm": 6.764842987060547,
|
| 555 |
+
"learning_rate": 6.837777777777779e-06,
|
| 556 |
+
"loss": 0.3534,
|
| 557 |
"step": 1925
|
| 558 |
},
|
| 559 |
{
|
| 560 |
+
"epoch": 1.0551948051948052,
|
| 561 |
+
"grad_norm": 5.962332248687744,
|
| 562 |
+
"learning_rate": 6.782222222222222e-06,
|
| 563 |
+
"loss": 0.3212,
|
| 564 |
"step": 1950
|
| 565 |
},
|
| 566 |
{
|
| 567 |
+
"epoch": 1.0687229437229437,
|
| 568 |
+
"grad_norm": 5.471970081329346,
|
| 569 |
+
"learning_rate": 6.726666666666667e-06,
|
| 570 |
+
"loss": 0.3572,
|
| 571 |
"step": 1975
|
| 572 |
},
|
| 573 |
{
|
| 574 |
+
"epoch": 1.0822510822510822,
|
| 575 |
+
"grad_norm": 6.054861545562744,
|
| 576 |
+
"learning_rate": 6.671111111111112e-06,
|
| 577 |
+
"loss": 0.3339,
|
| 578 |
"step": 2000
|
| 579 |
},
|
| 580 |
{
|
| 581 |
+
"epoch": 1.0822510822510822,
|
| 582 |
+
"eval_loss": 0.42054763436317444,
|
| 583 |
+
"eval_runtime": 1782.8878,
|
| 584 |
+
"eval_samples_per_second": 2.188,
|
| 585 |
+
"eval_steps_per_second": 0.137,
|
| 586 |
+
"eval_wer": 0.3197865353037767,
|
| 587 |
"step": 2000
|
| 588 |
},
|
| 589 |
{
|
| 590 |
+
"epoch": 1.0957792207792207,
|
| 591 |
+
"grad_norm": 6.194203853607178,
|
| 592 |
+
"learning_rate": 6.615555555555556e-06,
|
| 593 |
+
"loss": 0.339,
|
| 594 |
"step": 2025
|
| 595 |
},
|
| 596 |
{
|
| 597 |
+
"epoch": 1.1093073593073592,
|
| 598 |
+
"grad_norm": 5.470515727996826,
|
| 599 |
+
"learning_rate": 6.560000000000001e-06,
|
| 600 |
+
"loss": 0.3375,
|
| 601 |
"step": 2050
|
| 602 |
},
|
| 603 |
{
|
| 604 |
+
"epoch": 1.1228354978354977,
|
| 605 |
+
"grad_norm": 5.1414618492126465,
|
| 606 |
+
"learning_rate": 6.504444444444446e-06,
|
| 607 |
+
"loss": 0.3348,
|
| 608 |
"step": 2075
|
| 609 |
},
|
| 610 |
{
|
| 611 |
+
"epoch": 1.1363636363636362,
|
| 612 |
+
"grad_norm": 5.000445365905762,
|
| 613 |
+
"learning_rate": 6.448888888888889e-06,
|
| 614 |
+
"loss": 0.3255,
|
| 615 |
"step": 2100
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 1.149891774891775,
|
| 619 |
+
"grad_norm": 5.545360088348389,
|
| 620 |
+
"learning_rate": 6.393333333333334e-06,
|
| 621 |
+
"loss": 0.3296,
|
| 622 |
"step": 2125
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"epoch": 1.1634199134199135,
|
| 626 |
+
"grad_norm": 5.920198440551758,
|
| 627 |
+
"learning_rate": 6.3377777777777786e-06,
|
| 628 |
+
"loss": 0.3436,
|
| 629 |
"step": 2150
|
| 630 |
},
|
| 631 |
{
|
| 632 |
+
"epoch": 1.176948051948052,
|
| 633 |
+
"grad_norm": 5.722521781921387,
|
| 634 |
+
"learning_rate": 6.282222222222223e-06,
|
| 635 |
+
"loss": 0.3366,
|
| 636 |
"step": 2175
|
| 637 |
},
|
| 638 |
{
|
| 639 |
+
"epoch": 1.1904761904761905,
|
| 640 |
+
"grad_norm": 6.066483020782471,
|
| 641 |
+
"learning_rate": 6.2266666666666675e-06,
|
| 642 |
+
"loss": 0.332,
|
| 643 |
"step": 2200
|
| 644 |
},
|
| 645 |
{
|
| 646 |
+
"epoch": 1.204004329004329,
|
| 647 |
+
"grad_norm": 6.301929473876953,
|
| 648 |
+
"learning_rate": 6.171111111111112e-06,
|
| 649 |
+
"loss": 0.3207,
|
| 650 |
"step": 2225
|
| 651 |
},
|
| 652 |
{
|
| 653 |
+
"epoch": 1.2175324675324675,
|
| 654 |
+
"grad_norm": 5.607754230499268,
|
| 655 |
+
"learning_rate": 6.1155555555555555e-06,
|
| 656 |
+
"loss": 0.3338,
|
| 657 |
"step": 2250
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 1.231060606060606,
|
| 661 |
+
"grad_norm": 5.145053863525391,
|
| 662 |
+
"learning_rate": 6.0600000000000004e-06,
|
| 663 |
+
"loss": 0.3268,
|
| 664 |
"step": 2275
|
| 665 |
},
|
| 666 |
{
|
| 667 |
+
"epoch": 1.2445887445887447,
|
| 668 |
+
"grad_norm": 5.448360443115234,
|
| 669 |
+
"learning_rate": 6.004444444444445e-06,
|
| 670 |
+
"loss": 0.3365,
|
| 671 |
"step": 2300
|
| 672 |
},
|
| 673 |
{
|
| 674 |
+
"epoch": 1.2581168831168832,
|
| 675 |
+
"grad_norm": 5.5156474113464355,
|
| 676 |
+
"learning_rate": 5.948888888888889e-06,
|
| 677 |
+
"loss": 0.3268,
|
| 678 |
"step": 2325
|
| 679 |
},
|
| 680 |
{
|
| 681 |
+
"epoch": 1.2716450216450217,
|
| 682 |
+
"grad_norm": 5.252381324768066,
|
| 683 |
+
"learning_rate": 5.893333333333334e-06,
|
| 684 |
+
"loss": 0.3228,
|
| 685 |
"step": 2350
|
| 686 |
},
|
| 687 |
{
|
| 688 |
+
"epoch": 1.2851731601731602,
|
| 689 |
+
"grad_norm": 5.7689313888549805,
|
| 690 |
+
"learning_rate": 5.837777777777777e-06,
|
| 691 |
+
"loss": 0.3304,
|
| 692 |
"step": 2375
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"epoch": 1.2987012987012987,
|
| 696 |
+
"grad_norm": 4.822956085205078,
|
| 697 |
+
"learning_rate": 5.782222222222222e-06,
|
| 698 |
+
"loss": 0.3066,
|
| 699 |
"step": 2400
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 1.3122294372294372,
|
| 703 |
+
"grad_norm": 5.012087345123291,
|
| 704 |
+
"learning_rate": 5.726666666666667e-06,
|
| 705 |
+
"loss": 0.3323,
|
| 706 |
"step": 2425
|
| 707 |
},
|
| 708 |
{
|
| 709 |
+
"epoch": 1.3257575757575757,
|
| 710 |
+
"grad_norm": 5.262439250946045,
|
| 711 |
+
"learning_rate": 5.671111111111112e-06,
|
| 712 |
+
"loss": 0.3256,
|
| 713 |
"step": 2450
|
| 714 |
},
|
| 715 |
{
|
| 716 |
+
"epoch": 1.3392857142857144,
|
| 717 |
+
"grad_norm": 5.300339221954346,
|
| 718 |
+
"learning_rate": 5.615555555555556e-06,
|
| 719 |
+
"loss": 0.3287,
|
| 720 |
"step": 2475
|
| 721 |
},
|
| 722 |
{
|
| 723 |
+
"epoch": 1.3528138528138527,
|
| 724 |
+
"grad_norm": 6.058621883392334,
|
| 725 |
+
"learning_rate": 5.560000000000001e-06,
|
| 726 |
+
"loss": 0.3283,
|
| 727 |
"step": 2500
|
| 728 |
},
|
| 729 |
{
|
| 730 |
+
"epoch": 1.3663419913419914,
|
| 731 |
+
"grad_norm": 6.223220348358154,
|
| 732 |
+
"learning_rate": 5.504444444444444e-06,
|
| 733 |
+
"loss": 0.3288,
|
| 734 |
"step": 2525
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"epoch": 1.37987012987013,
|
| 738 |
+
"grad_norm": 5.865265369415283,
|
| 739 |
+
"learning_rate": 5.448888888888889e-06,
|
| 740 |
+
"loss": 0.3303,
|
| 741 |
"step": 2550
|
| 742 |
},
|
| 743 |
{
|
| 744 |
+
"epoch": 1.3933982683982684,
|
| 745 |
+
"grad_norm": 4.715255260467529,
|
| 746 |
+
"learning_rate": 5.393333333333334e-06,
|
| 747 |
+
"loss": 0.3432,
|
| 748 |
"step": 2575
|
| 749 |
},
|
| 750 |
{
|
| 751 |
+
"epoch": 1.406926406926407,
|
| 752 |
+
"grad_norm": 5.57729434967041,
|
| 753 |
+
"learning_rate": 5.337777777777779e-06,
|
| 754 |
+
"loss": 0.3174,
|
| 755 |
"step": 2600
|
| 756 |
},
|
| 757 |
{
|
| 758 |
+
"epoch": 1.4204545454545454,
|
| 759 |
+
"grad_norm": 6.372653484344482,
|
| 760 |
+
"learning_rate": 5.282222222222223e-06,
|
| 761 |
+
"loss": 0.3251,
|
| 762 |
"step": 2625
|
| 763 |
},
|
| 764 |
{
|
| 765 |
+
"epoch": 1.433982683982684,
|
| 766 |
+
"grad_norm": 6.7026848793029785,
|
| 767 |
+
"learning_rate": 5.226666666666667e-06,
|
| 768 |
+
"loss": 0.3258,
|
| 769 |
"step": 2650
|
| 770 |
},
|
| 771 |
{
|
| 772 |
+
"epoch": 1.4475108225108224,
|
| 773 |
+
"grad_norm": 5.12203311920166,
|
| 774 |
+
"learning_rate": 5.171111111111111e-06,
|
| 775 |
+
"loss": 0.3217,
|
| 776 |
"step": 2675
|
| 777 |
},
|
| 778 |
{
|
| 779 |
+
"epoch": 1.4610389610389611,
|
| 780 |
+
"grad_norm": 7.778601169586182,
|
| 781 |
+
"learning_rate": 5.115555555555556e-06,
|
| 782 |
+
"loss": 0.3182,
|
| 783 |
"step": 2700
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 1.4745670995670996,
|
| 787 |
+
"grad_norm": 4.994805335998535,
|
| 788 |
+
"learning_rate": 5.060000000000001e-06,
|
| 789 |
+
"loss": 0.3142,
|
| 790 |
"step": 2725
|
| 791 |
},
|
| 792 |
{
|
| 793 |
+
"epoch": 1.4880952380952381,
|
| 794 |
+
"grad_norm": 6.392801761627197,
|
| 795 |
+
"learning_rate": 5.004444444444445e-06,
|
| 796 |
+
"loss": 0.3289,
|
| 797 |
"step": 2750
|
| 798 |
},
|
| 799 |
{
|
| 800 |
+
"epoch": 1.5016233766233766,
|
| 801 |
+
"grad_norm": 5.523842811584473,
|
| 802 |
+
"learning_rate": 4.94888888888889e-06,
|
| 803 |
+
"loss": 0.3248,
|
| 804 |
"step": 2775
|
| 805 |
},
|
| 806 |
{
|
| 807 |
+
"epoch": 1.5151515151515151,
|
| 808 |
+
"grad_norm": 5.348546981811523,
|
| 809 |
+
"learning_rate": 4.893333333333334e-06,
|
| 810 |
+
"loss": 0.303,
|
| 811 |
"step": 2800
|
| 812 |
},
|
| 813 |
{
|
| 814 |
+
"epoch": 1.5286796536796536,
|
| 815 |
+
"grad_norm": 5.714568614959717,
|
| 816 |
+
"learning_rate": 4.837777777777778e-06,
|
| 817 |
+
"loss": 0.3144,
|
| 818 |
"step": 2825
|
| 819 |
},
|
| 820 |
{
|
| 821 |
+
"epoch": 1.5422077922077921,
|
| 822 |
+
"grad_norm": 5.544715404510498,
|
| 823 |
+
"learning_rate": 4.7822222222222226e-06,
|
| 824 |
+
"loss": 0.3183,
|
| 825 |
"step": 2850
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 1.5557359307359309,
|
| 829 |
+
"grad_norm": 6.49782133102417,
|
| 830 |
+
"learning_rate": 4.7266666666666674e-06,
|
| 831 |
+
"loss": 0.2981,
|
| 832 |
"step": 2875
|
| 833 |
},
|
| 834 |
{
|
| 835 |
+
"epoch": 1.5692640692640691,
|
| 836 |
+
"grad_norm": 5.356492042541504,
|
| 837 |
+
"learning_rate": 4.6711111111111115e-06,
|
| 838 |
+
"loss": 0.3159,
|
| 839 |
"step": 2900
|
| 840 |
},
|
| 841 |
{
|
| 842 |
+
"epoch": 1.5827922077922079,
|
| 843 |
+
"grad_norm": 5.4491119384765625,
|
| 844 |
+
"learning_rate": 4.6155555555555555e-06,
|
| 845 |
+
"loss": 0.3333,
|
| 846 |
"step": 2925
|
| 847 |
},
|
| 848 |
{
|
| 849 |
+
"epoch": 1.5963203463203464,
|
| 850 |
+
"grad_norm": 5.832214832305908,
|
| 851 |
+
"learning_rate": 4.56e-06,
|
| 852 |
+
"loss": 0.3141,
|
| 853 |
"step": 2950
|
| 854 |
},
|
| 855 |
{
|
| 856 |
+
"epoch": 1.6098484848484849,
|
| 857 |
+
"grad_norm": 5.139626979827881,
|
| 858 |
+
"learning_rate": 4.504444444444444e-06,
|
| 859 |
+
"loss": 0.308,
|
| 860 |
"step": 2975
|
| 861 |
},
|
| 862 |
{
|
| 863 |
+
"epoch": 1.6233766233766234,
|
| 864 |
+
"grad_norm": 5.6519999504089355,
|
| 865 |
+
"learning_rate": 4.448888888888889e-06,
|
| 866 |
+
"loss": 0.3189,
|
| 867 |
"step": 3000
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 1.6233766233766234,
|
| 871 |
+
"eval_loss": 0.3910607397556305,
|
| 872 |
+
"eval_runtime": 1770.6955,
|
| 873 |
+
"eval_samples_per_second": 2.203,
|
| 874 |
+
"eval_steps_per_second": 0.138,
|
| 875 |
+
"eval_wer": 0.29134067420071474,
|
| 876 |
"step": 3000
|
| 877 |
},
|
| 878 |
{
|
| 879 |
+
"epoch": 1.6369047619047619,
|
| 880 |
+
"grad_norm": 5.791579246520996,
|
| 881 |
+
"learning_rate": 4.393333333333334e-06,
|
| 882 |
+
"loss": 0.3122,
|
| 883 |
"step": 3025
|
| 884 |
},
|
| 885 |
{
|
| 886 |
+
"epoch": 1.6504329004329006,
|
| 887 |
+
"grad_norm": 5.640200614929199,
|
| 888 |
+
"learning_rate": 4.337777777777778e-06,
|
| 889 |
+
"loss": 0.3062,
|
| 890 |
"step": 3050
|
| 891 |
},
|
| 892 |
{
|
| 893 |
+
"epoch": 1.6639610389610389,
|
| 894 |
+
"grad_norm": 5.585713863372803,
|
| 895 |
+
"learning_rate": 4.282222222222222e-06,
|
| 896 |
+
"loss": 0.3208,
|
| 897 |
"step": 3075
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 1.6774891774891776,
|
| 901 |
+
"grad_norm": 5.871087074279785,
|
| 902 |
+
"learning_rate": 4.226666666666667e-06,
|
| 903 |
+
"loss": 0.3071,
|
| 904 |
"step": 3100
|
| 905 |
},
|
| 906 |
{
|
| 907 |
+
"epoch": 1.6910173160173159,
|
| 908 |
+
"grad_norm": 5.412327766418457,
|
| 909 |
+
"learning_rate": 4.171111111111111e-06,
|
| 910 |
+
"loss": 0.3167,
|
| 911 |
"step": 3125
|
| 912 |
},
|
| 913 |
{
|
| 914 |
+
"epoch": 1.7045454545454546,
|
| 915 |
+
"grad_norm": 5.0698561668396,
|
| 916 |
+
"learning_rate": 4.115555555555556e-06,
|
| 917 |
+
"loss": 0.3231,
|
| 918 |
"step": 3150
|
| 919 |
},
|
| 920 |
{
|
| 921 |
+
"epoch": 1.718073593073593,
|
| 922 |
+
"grad_norm": 5.63693904876709,
|
| 923 |
+
"learning_rate": 4.060000000000001e-06,
|
| 924 |
+
"loss": 0.3128,
|
| 925 |
"step": 3175
|
| 926 |
},
|
| 927 |
{
|
| 928 |
+
"epoch": 1.7316017316017316,
|
| 929 |
+
"grad_norm": 5.766589164733887,
|
| 930 |
+
"learning_rate": 4.004444444444445e-06,
|
| 931 |
+
"loss": 0.3229,
|
| 932 |
"step": 3200
|
| 933 |
},
|
| 934 |
{
|
| 935 |
+
"epoch": 1.74512987012987,
|
| 936 |
+
"grad_norm": 5.414788246154785,
|
| 937 |
+
"learning_rate": 3.948888888888889e-06,
|
| 938 |
+
"loss": 0.2917,
|
| 939 |
"step": 3225
|
| 940 |
},
|
| 941 |
{
|
| 942 |
+
"epoch": 1.7586580086580086,
|
| 943 |
+
"grad_norm": 5.106072902679443,
|
| 944 |
+
"learning_rate": 3.893333333333333e-06,
|
| 945 |
+
"loss": 0.311,
|
| 946 |
"step": 3250
|
| 947 |
},
|
| 948 |
{
|
| 949 |
+
"epoch": 1.7721861471861473,
|
| 950 |
+
"grad_norm": 4.694611549377441,
|
| 951 |
+
"learning_rate": 3.837777777777778e-06,
|
| 952 |
+
"loss": 0.3228,
|
| 953 |
"step": 3275
|
| 954 |
},
|
| 955 |
{
|
| 956 |
+
"epoch": 1.7857142857142856,
|
| 957 |
+
"grad_norm": 6.422979354858398,
|
| 958 |
+
"learning_rate": 3.782222222222223e-06,
|
| 959 |
+
"loss": 0.314,
|
| 960 |
"step": 3300
|
| 961 |
},
|
| 962 |
{
|
| 963 |
+
"epoch": 1.7992424242424243,
|
| 964 |
+
"grad_norm": 5.5537567138671875,
|
| 965 |
+
"learning_rate": 3.726666666666667e-06,
|
| 966 |
+
"loss": 0.303,
|
| 967 |
"step": 3325
|
| 968 |
},
|
| 969 |
{
|
| 970 |
+
"epoch": 1.8127705627705628,
|
| 971 |
+
"grad_norm": 6.503033638000488,
|
| 972 |
+
"learning_rate": 3.6711111111111113e-06,
|
| 973 |
+
"loss": 0.336,
|
| 974 |
"step": 3350
|
| 975 |
},
|
| 976 |
{
|
| 977 |
+
"epoch": 1.8262987012987013,
|
| 978 |
+
"grad_norm": 5.406898021697998,
|
| 979 |
+
"learning_rate": 3.615555555555556e-06,
|
| 980 |
+
"loss": 0.3031,
|
| 981 |
"step": 3375
|
| 982 |
},
|
| 983 |
{
|
| 984 |
+
"epoch": 1.8398268398268398,
|
| 985 |
+
"grad_norm": 6.486941337585449,
|
| 986 |
+
"learning_rate": 3.5600000000000002e-06,
|
| 987 |
+
"loss": 0.3203,
|
| 988 |
"step": 3400
|
| 989 |
},
|
| 990 |
{
|
| 991 |
+
"epoch": 1.8533549783549783,
|
| 992 |
+
"grad_norm": 7.027703285217285,
|
| 993 |
+
"learning_rate": 3.5044444444444447e-06,
|
| 994 |
+
"loss": 0.3159,
|
| 995 |
"step": 3425
|
| 996 |
},
|
| 997 |
{
|
| 998 |
+
"epoch": 1.866883116883117,
|
| 999 |
+
"grad_norm": 5.475865364074707,
|
| 1000 |
+
"learning_rate": 3.4488888888888896e-06,
|
| 1001 |
+
"loss": 0.3239,
|
| 1002 |
"step": 3450
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
+
"epoch": 1.8804112554112553,
|
| 1006 |
+
"grad_norm": 6.124994277954102,
|
| 1007 |
+
"learning_rate": 3.3933333333333336e-06,
|
| 1008 |
+
"loss": 0.2928,
|
| 1009 |
"step": 3475
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
+
"epoch": 1.893939393939394,
|
| 1013 |
+
"grad_norm": 4.759301662445068,
|
| 1014 |
+
"learning_rate": 3.337777777777778e-06,
|
| 1015 |
+
"loss": 0.2862,
|
| 1016 |
"step": 3500
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
+
"epoch": 1.9074675324675323,
|
| 1020 |
+
"grad_norm": 5.548280239105225,
|
| 1021 |
+
"learning_rate": 3.282222222222223e-06,
|
| 1022 |
+
"loss": 0.312,
|
| 1023 |
"step": 3525
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
+
"epoch": 1.920995670995671,
|
| 1027 |
+
"grad_norm": 5.691162109375,
|
| 1028 |
+
"learning_rate": 3.226666666666667e-06,
|
| 1029 |
+
"loss": 0.3161,
|
| 1030 |
"step": 3550
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
+
"epoch": 1.9345238095238095,
|
| 1034 |
+
"grad_norm": 6.089394569396973,
|
| 1035 |
+
"learning_rate": 3.1711111111111114e-06,
|
| 1036 |
+
"loss": 0.3028,
|
| 1037 |
"step": 3575
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
+
"epoch": 1.948051948051948,
|
| 1041 |
+
"grad_norm": 5.725650310516357,
|
| 1042 |
+
"learning_rate": 3.1155555555555555e-06,
|
| 1043 |
+
"loss": 0.3058,
|
| 1044 |
"step": 3600
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
+
"epoch": 1.9615800865800865,
|
| 1048 |
+
"grad_norm": 5.124326705932617,
|
| 1049 |
+
"learning_rate": 3.0600000000000003e-06,
|
| 1050 |
+
"loss": 0.2947,
|
| 1051 |
"step": 3625
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
+
"epoch": 1.975108225108225,
|
| 1055 |
+
"grad_norm": 6.62967586517334,
|
| 1056 |
+
"learning_rate": 3.004444444444445e-06,
|
| 1057 |
+
"loss": 0.318,
|
| 1058 |
"step": 3650
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
+
"epoch": 1.9886363636363638,
|
| 1062 |
+
"grad_norm": 6.150094985961914,
|
| 1063 |
+
"learning_rate": 2.948888888888889e-06,
|
| 1064 |
+
"loss": 0.3257,
|
| 1065 |
"step": 3675
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 2.002164502164502,
|
| 1069 |
+
"grad_norm": 3.962730884552002,
|
| 1070 |
+
"learning_rate": 2.8933333333333337e-06,
|
| 1071 |
+
"loss": 0.29,
|
| 1072 |
"step": 3700
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
+
"epoch": 2.0156926406926408,
|
| 1076 |
+
"grad_norm": 3.999758005142212,
|
| 1077 |
+
"learning_rate": 2.837777777777778e-06,
|
| 1078 |
+
"loss": 0.2127,
|
| 1079 |
"step": 3725
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
+
"epoch": 2.029220779220779,
|
| 1083 |
+
"grad_norm": 4.3916015625,
|
| 1084 |
+
"learning_rate": 2.7822222222222222e-06,
|
| 1085 |
+
"loss": 0.2073,
|
| 1086 |
"step": 3750
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
+
"epoch": 2.0427489177489178,
|
| 1090 |
+
"grad_norm": 4.647676944732666,
|
| 1091 |
+
"learning_rate": 2.726666666666667e-06,
|
| 1092 |
+
"loss": 0.2065,
|
| 1093 |
"step": 3775
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
+
"epoch": 2.0562770562770565,
|
| 1097 |
+
"grad_norm": 5.331233501434326,
|
| 1098 |
+
"learning_rate": 2.6711111111111116e-06,
|
| 1099 |
+
"loss": 0.208,
|
| 1100 |
"step": 3800
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
+
"epoch": 2.0698051948051948,
|
| 1104 |
+
"grad_norm": 4.8974995613098145,
|
| 1105 |
+
"learning_rate": 2.6155555555555556e-06,
|
| 1106 |
+
"loss": 0.2007,
|
| 1107 |
"step": 3825
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 2.0833333333333335,
|
| 1111 |
+
"grad_norm": 4.972061634063721,
|
| 1112 |
+
"learning_rate": 2.56e-06,
|
| 1113 |
+
"loss": 0.2017,
|
| 1114 |
"step": 3850
|
| 1115 |
},
|
| 1116 |
{
|
| 1117 |
+
"epoch": 2.0968614718614718,
|
| 1118 |
+
"grad_norm": 5.035933494567871,
|
| 1119 |
+
"learning_rate": 2.504444444444445e-06,
|
| 1120 |
+
"loss": 0.2075,
|
| 1121 |
"step": 3875
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
+
"epoch": 2.1103896103896105,
|
| 1125 |
+
"grad_norm": 6.620712757110596,
|
| 1126 |
+
"learning_rate": 2.448888888888889e-06,
|
| 1127 |
+
"loss": 0.2074,
|
| 1128 |
"step": 3900
|
| 1129 |
},
|
| 1130 |
{
|
| 1131 |
+
"epoch": 2.1239177489177488,
|
| 1132 |
+
"grad_norm": 7.161535739898682,
|
| 1133 |
+
"learning_rate": 2.3933333333333334e-06,
|
| 1134 |
+
"loss": 0.2075,
|
| 1135 |
"step": 3925
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
+
"epoch": 2.1374458874458875,
|
| 1139 |
+
"grad_norm": 5.531479358673096,
|
| 1140 |
+
"learning_rate": 2.337777777777778e-06,
|
| 1141 |
+
"loss": 0.2088,
|
| 1142 |
"step": 3950
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
+
"epoch": 2.150974025974026,
|
| 1146 |
+
"grad_norm": 4.983880043029785,
|
| 1147 |
+
"learning_rate": 2.2822222222222223e-06,
|
| 1148 |
+
"loss": 0.2103,
|
| 1149 |
"step": 3975
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
+
"epoch": 2.1645021645021645,
|
| 1153 |
+
"grad_norm": 4.373054504394531,
|
| 1154 |
+
"learning_rate": 2.226666666666667e-06,
|
| 1155 |
+
"loss": 0.2051,
|
| 1156 |
"step": 4000
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
+
"epoch": 2.1645021645021645,
|
| 1160 |
+
"eval_loss": 0.38633546233177185,
|
| 1161 |
+
"eval_runtime": 1768.38,
|
| 1162 |
+
"eval_samples_per_second": 2.206,
|
| 1163 |
+
"eval_steps_per_second": 0.138,
|
| 1164 |
+
"eval_wer": 0.27895296049454266,
|
| 1165 |
"step": 4000
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
+
"epoch": 2.178030303030303,
|
| 1169 |
+
"grad_norm": 5.22249174118042,
|
| 1170 |
+
"learning_rate": 2.1711111111111113e-06,
|
| 1171 |
+
"loss": 0.1988,
|
| 1172 |
"step": 4025
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
+
"epoch": 2.1915584415584415,
|
| 1176 |
+
"grad_norm": 3.8173792362213135,
|
| 1177 |
+
"learning_rate": 2.1155555555555557e-06,
|
| 1178 |
+
"loss": 0.2241,
|
| 1179 |
"step": 4050
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
+
"epoch": 2.20508658008658,
|
| 1183 |
+
"grad_norm": 5.271940231323242,
|
| 1184 |
+
"learning_rate": 2.06e-06,
|
| 1185 |
+
"loss": 0.201,
|
| 1186 |
"step": 4075
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
+
"epoch": 2.2186147186147185,
|
| 1190 |
+
"grad_norm": 4.359199523925781,
|
| 1191 |
+
"learning_rate": 2.0044444444444446e-06,
|
| 1192 |
+
"loss": 0.1946,
|
| 1193 |
"step": 4100
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
+
"epoch": 2.232142857142857,
|
| 1197 |
+
"grad_norm": 4.675993919372559,
|
| 1198 |
+
"learning_rate": 1.948888888888889e-06,
|
| 1199 |
+
"loss": 0.2123,
|
| 1200 |
"step": 4125
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
+
"epoch": 2.2456709956709955,
|
| 1204 |
+
"grad_norm": 4.090628147125244,
|
| 1205 |
+
"learning_rate": 1.8933333333333333e-06,
|
| 1206 |
+
"loss": 0.2045,
|
| 1207 |
"step": 4150
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
+
"epoch": 2.259199134199134,
|
| 1211 |
+
"grad_norm": 3.5872035026550293,
|
| 1212 |
+
"learning_rate": 1.837777777777778e-06,
|
| 1213 |
+
"loss": 0.2042,
|
| 1214 |
"step": 4175
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
+
"epoch": 2.2727272727272725,
|
| 1218 |
+
"grad_norm": 4.375498294830322,
|
| 1219 |
+
"learning_rate": 1.7822222222222225e-06,
|
| 1220 |
+
"loss": 0.2056,
|
| 1221 |
"step": 4200
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
+
"epoch": 2.286255411255411,
|
| 1225 |
+
"grad_norm": 4.972550868988037,
|
| 1226 |
+
"learning_rate": 1.7266666666666667e-06,
|
| 1227 |
+
"loss": 0.2024,
|
| 1228 |
"step": 4225
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
+
"epoch": 2.29978354978355,
|
| 1232 |
+
"grad_norm": 4.7940168380737305,
|
| 1233 |
+
"learning_rate": 1.6711111111111112e-06,
|
| 1234 |
+
"loss": 0.1996,
|
| 1235 |
"step": 4250
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
+
"epoch": 2.313311688311688,
|
| 1239 |
+
"grad_norm": 4.399414539337158,
|
| 1240 |
+
"learning_rate": 1.6155555555555559e-06,
|
| 1241 |
+
"loss": 0.2101,
|
| 1242 |
"step": 4275
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
+
"epoch": 2.326839826839827,
|
| 1246 |
+
"grad_norm": 5.292896747589111,
|
| 1247 |
+
"learning_rate": 1.56e-06,
|
| 1248 |
+
"loss": 0.1929,
|
| 1249 |
"step": 4300
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
+
"epoch": 2.340367965367965,
|
| 1253 |
+
"grad_norm": 4.333370685577393,
|
| 1254 |
+
"learning_rate": 1.5044444444444446e-06,
|
| 1255 |
+
"loss": 0.208,
|
| 1256 |
"step": 4325
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
+
"epoch": 2.353896103896104,
|
| 1260 |
+
"grad_norm": 4.69057035446167,
|
| 1261 |
+
"learning_rate": 1.4488888888888892e-06,
|
| 1262 |
+
"loss": 0.1905,
|
| 1263 |
"step": 4350
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
+
"epoch": 2.367424242424242,
|
| 1267 |
+
"grad_norm": 4.56622838973999,
|
| 1268 |
+
"learning_rate": 1.3933333333333335e-06,
|
| 1269 |
+
"loss": 0.2051,
|
| 1270 |
"step": 4375
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
+
"epoch": 2.380952380952381,
|
| 1274 |
+
"grad_norm": 4.605253219604492,
|
| 1275 |
+
"learning_rate": 1.337777777777778e-06,
|
| 1276 |
+
"loss": 0.1978,
|
| 1277 |
"step": 4400
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
+
"epoch": 2.3944805194805197,
|
| 1281 |
+
"grad_norm": 5.727032661437988,
|
| 1282 |
+
"learning_rate": 1.2822222222222222e-06,
|
| 1283 |
+
"loss": 0.2002,
|
| 1284 |
"step": 4425
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
+
"epoch": 2.408008658008658,
|
| 1288 |
+
"grad_norm": 5.882457256317139,
|
| 1289 |
+
"learning_rate": 1.2266666666666666e-06,
|
| 1290 |
+
"loss": 0.202,
|
| 1291 |
"step": 4450
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
+
"epoch": 2.4215367965367967,
|
| 1295 |
+
"grad_norm": 4.464743614196777,
|
| 1296 |
+
"learning_rate": 1.171111111111111e-06,
|
| 1297 |
+
"loss": 0.2145,
|
| 1298 |
"step": 4475
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
+
"epoch": 2.435064935064935,
|
| 1302 |
+
"grad_norm": 4.503987789154053,
|
| 1303 |
+
"learning_rate": 1.1155555555555558e-06,
|
| 1304 |
+
"loss": 0.2101,
|
| 1305 |
"step": 4500
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
+
"epoch": 2.4485930735930737,
|
| 1309 |
+
"grad_norm": 5.735741138458252,
|
| 1310 |
+
"learning_rate": 1.06e-06,
|
| 1311 |
+
"loss": 0.1913,
|
| 1312 |
"step": 4525
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
+
"epoch": 2.462121212121212,
|
| 1316 |
+
"grad_norm": 4.6319098472595215,
|
| 1317 |
+
"learning_rate": 1.0044444444444445e-06,
|
| 1318 |
+
"loss": 0.2001,
|
| 1319 |
"step": 4550
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
+
"epoch": 2.4756493506493507,
|
| 1323 |
+
"grad_norm": 5.589540958404541,
|
| 1324 |
+
"learning_rate": 9.488888888888889e-07,
|
| 1325 |
+
"loss": 0.1981,
|
| 1326 |
"step": 4575
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
+
"epoch": 2.4891774891774894,
|
| 1330 |
+
"grad_norm": 4.481135845184326,
|
| 1331 |
+
"learning_rate": 8.933333333333334e-07,
|
| 1332 |
+
"loss": 0.198,
|
| 1333 |
"step": 4600
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
+
"epoch": 2.5027056277056277,
|
| 1337 |
+
"grad_norm": 6.087165355682373,
|
| 1338 |
+
"learning_rate": 8.37777777777778e-07,
|
| 1339 |
+
"loss": 0.2131,
|
| 1340 |
"step": 4625
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
+
"epoch": 2.5162337662337664,
|
| 1344 |
+
"grad_norm": 4.635289669036865,
|
| 1345 |
+
"learning_rate": 7.822222222222223e-07,
|
| 1346 |
+
"loss": 0.2088,
|
| 1347 |
"step": 4650
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
+
"epoch": 2.5297619047619047,
|
| 1351 |
+
"grad_norm": 4.698585510253906,
|
| 1352 |
+
"learning_rate": 7.266666666666668e-07,
|
| 1353 |
+
"loss": 0.2057,
|
| 1354 |
"step": 4675
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
+
"epoch": 2.5432900432900434,
|
| 1358 |
+
"grad_norm": 4.562716960906982,
|
| 1359 |
+
"learning_rate": 6.711111111111111e-07,
|
| 1360 |
+
"loss": 0.2117,
|
| 1361 |
"step": 4700
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
+
"epoch": 2.5568181818181817,
|
| 1365 |
+
"grad_norm": 5.381985187530518,
|
| 1366 |
+
"learning_rate": 6.155555555555556e-07,
|
| 1367 |
+
"loss": 0.1975,
|
| 1368 |
"step": 4725
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
+
"epoch": 2.5703463203463204,
|
| 1372 |
+
"grad_norm": 5.667773723602295,
|
| 1373 |
+
"learning_rate": 5.6e-07,
|
| 1374 |
+
"loss": 0.2409,
|
| 1375 |
"step": 4750
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
+
"epoch": 2.583874458874459,
|
| 1379 |
+
"grad_norm": 4.565330982208252,
|
| 1380 |
+
"learning_rate": 5.044444444444445e-07,
|
| 1381 |
+
"loss": 0.1915,
|
| 1382 |
"step": 4775
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
+
"epoch": 2.5974025974025974,
|
| 1386 |
+
"grad_norm": 5.17742395401001,
|
| 1387 |
+
"learning_rate": 4.488888888888889e-07,
|
| 1388 |
+
"loss": 0.1973,
|
| 1389 |
"step": 4800
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
+
"epoch": 2.6109307359307357,
|
| 1393 |
+
"grad_norm": 4.878474712371826,
|
| 1394 |
+
"learning_rate": 3.9333333333333336e-07,
|
| 1395 |
+
"loss": 0.2209,
|
| 1396 |
"step": 4825
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
+
"epoch": 2.6244588744588744,
|
| 1400 |
+
"grad_norm": 5.2556328773498535,
|
| 1401 |
+
"learning_rate": 3.3777777777777777e-07,
|
| 1402 |
+
"loss": 0.204,
|
| 1403 |
"step": 4850
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
+
"epoch": 2.637987012987013,
|
| 1407 |
+
"grad_norm": 3.8071792125701904,
|
| 1408 |
+
"learning_rate": 2.822222222222222e-07,
|
| 1409 |
+
"loss": 0.2052,
|
| 1410 |
"step": 4875
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
+
"epoch": 2.6515151515151514,
|
| 1414 |
+
"grad_norm": 4.218277454376221,
|
| 1415 |
+
"learning_rate": 2.266666666666667e-07,
|
| 1416 |
+
"loss": 0.1989,
|
| 1417 |
"step": 4900
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
+
"epoch": 2.66504329004329,
|
| 1421 |
+
"grad_norm": 5.260907173156738,
|
| 1422 |
+
"learning_rate": 1.7111111111111114e-07,
|
| 1423 |
+
"loss": 0.1915,
|
| 1424 |
"step": 4925
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
+
"epoch": 2.678571428571429,
|
| 1428 |
+
"grad_norm": 4.497314453125,
|
| 1429 |
+
"learning_rate": 1.1555555555555556e-07,
|
| 1430 |
+
"loss": 0.1997,
|
| 1431 |
"step": 4950
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
+
"epoch": 2.692099567099567,
|
| 1435 |
+
"grad_norm": 4.543353080749512,
|
| 1436 |
+
"learning_rate": 6.000000000000001e-08,
|
| 1437 |
+
"loss": 0.1866,
|
| 1438 |
"step": 4975
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
+
"epoch": 2.7056277056277054,
|
| 1442 |
+
"grad_norm": 6.265724182128906,
|
| 1443 |
+
"learning_rate": 4.444444444444445e-09,
|
| 1444 |
+
"loss": 0.202,
|
| 1445 |
"step": 5000
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
+
"epoch": 2.7056277056277054,
|
| 1449 |
+
"eval_loss": 0.38099274039268494,
|
| 1450 |
+
"eval_runtime": 1768.0451,
|
| 1451 |
+
"eval_samples_per_second": 2.206,
|
| 1452 |
+
"eval_steps_per_second": 0.138,
|
| 1453 |
+
"eval_wer": 0.27501690331304934,
|
| 1454 |
"step": 5000
|
| 1455 |
},
|
| 1456 |
{
|
| 1457 |
+
"epoch": 2.7056277056277054,
|
| 1458 |
"step": 5000,
|
| 1459 |
+
"total_flos": 5.434978041004032e+20,
|
| 1460 |
+
"train_loss": 0.39443905401229856,
|
| 1461 |
+
"train_runtime": 57517.0908,
|
| 1462 |
+
"train_samples_per_second": 2.782,
|
| 1463 |
+
"train_steps_per_second": 0.087
|
| 1464 |
}
|
| 1465 |
],
|
| 1466 |
"logging_steps": 25,
|
| 1467 |
"max_steps": 5000,
|
| 1468 |
"num_input_tokens_seen": 0,
|
| 1469 |
+
"num_train_epochs": 3,
|
| 1470 |
"save_steps": 1000,
|
| 1471 |
"stateful_callbacks": {
|
| 1472 |
"TrainerControl": {
|
|
|
|
| 1480 |
"attributes": {}
|
| 1481 |
}
|
| 1482 |
},
|
| 1483 |
+
"total_flos": 5.434978041004032e+20,
|
| 1484 |
"train_batch_size": 16,
|
| 1485 |
"trial_name": null,
|
| 1486 |
"trial_params": null
|