End of training
Browse files- README.md +15 -2
- all_results.json +13 -13
- eval_results.json +7 -7
- runs/Aug06_03-27-27_4853604300ab/events.out.tfevents.1754477125.4853604300ab.1267.1 +3 -0
- train_results.json +7 -7
- trainer_state.json +1070 -492
README.md
CHANGED
|
@@ -4,11 +4,24 @@ license: apache-2.0
|
|
| 4 |
base_model: openai/whisper-small
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
model-index:
|
| 10 |
- name: whisper-medium-test
|
| 11 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -16,7 +29,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# whisper-medium-test
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.4044
|
| 22 |
- Wer: 0.1675
|
|
|
|
| 4 |
base_model: openai/whisper-small
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
datasets:
|
| 8 |
+
- rbcurzon/ph_dialect_asr
|
| 9 |
metrics:
|
| 10 |
- wer
|
| 11 |
model-index:
|
| 12 |
- name: whisper-medium-test
|
| 13 |
+
results:
|
| 14 |
+
- task:
|
| 15 |
+
name: Automatic Speech Recognition
|
| 16 |
+
type: automatic-speech-recognition
|
| 17 |
+
dataset:
|
| 18 |
+
name: rbcurzon/ph_dialect_asr all
|
| 19 |
+
type: rbcurzon/ph_dialect_asr
|
| 20 |
+
args: all
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Wer
|
| 23 |
+
type: wer
|
| 24 |
+
value: 0.1675184165997231
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 29 |
|
| 30 |
# whisper-medium-test
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the rbcurzon/ph_dialect_asr all dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
- Loss: 0.4044
|
| 35 |
- Wer: 0.1675
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
-
"total_flos": 4.
|
| 10 |
-
"train_loss": 0.
|
| 11 |
-
"train_runtime":
|
| 12 |
-
"train_samples":
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second": 0.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.79466271312083,
|
| 3 |
+
"eval_loss": 0.4044080376625061,
|
| 4 |
+
"eval_runtime": 942.1523,
|
| 5 |
+
"eval_samples": 2885,
|
| 6 |
+
"eval_samples_per_second": 3.062,
|
| 7 |
+
"eval_steps_per_second": 0.383,
|
| 8 |
+
"eval_wer": 0.1675184165997231,
|
| 9 |
+
"total_flos": 4.607669973179538e+19,
|
| 10 |
+
"train_loss": 0.11535389684215187,
|
| 11 |
+
"train_runtime": 25285.0125,
|
| 12 |
+
"train_samples": 10787,
|
| 13 |
+
"train_samples_per_second": 6.328,
|
| 14 |
+
"train_steps_per_second": 0.198
|
| 15 |
}
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.79466271312083,
|
| 3 |
+
"eval_loss": 0.4044080376625061,
|
| 4 |
+
"eval_runtime": 942.1523,
|
| 5 |
+
"eval_samples": 2885,
|
| 6 |
+
"eval_samples_per_second": 3.062,
|
| 7 |
+
"eval_steps_per_second": 0.383,
|
| 8 |
+
"eval_wer": 0.1675184165997231
|
| 9 |
}
|
runs/Aug06_03-27-27_4853604300ab/events.out.tfevents.1754477125.4853604300ab.1267.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ed7322a5f0a14fed47a4d413f46ab0f6368b93ab016914b91f347c32ec3bc1f
|
| 3 |
+
size 406
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 4.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 14.79466271312083,
|
| 3 |
+
"total_flos": 4.607669973179538e+19,
|
| 4 |
+
"train_loss": 0.11535389684215187,
|
| 5 |
+
"train_runtime": 25285.0125,
|
| 6 |
+
"train_samples": 10787,
|
| 7 |
+
"train_samples_per_second": 6.328,
|
| 8 |
+
"train_steps_per_second": 0.198
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -2,894 +2,1472 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 4.
|
| 16 |
-
"loss": 1.
|
| 17 |
"step": 25
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate": 9.
|
| 23 |
-
"loss": 1.
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 1.44e-06,
|
| 30 |
-
"loss":
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
"learning_rate": 1.94e-06,
|
| 37 |
-
"loss":
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
"learning_rate": 2.4400000000000004e-06,
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 2.9400000000000002e-06,
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
"learning_rate": 3.44e-06,
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm":
|
| 64 |
"learning_rate": 3.94e-06,
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm":
|
| 71 |
"learning_rate": 4.440000000000001e-06,
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 4.94e-06,
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm":
|
| 85 |
"learning_rate": 5.4400000000000004e-06,
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm": 5.
|
| 92 |
"learning_rate": 5.94e-06,
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm": 6.
|
| 99 |
"learning_rate": 6.440000000000001e-06,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch":
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 6.9400000000000005e-06,
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch":
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 7.440000000000001e-06,
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch":
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 7.94e-06,
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch":
|
| 126 |
-
"grad_norm":
|
| 127 |
"learning_rate": 8.44e-06,
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"grad_norm":
|
| 134 |
"learning_rate": 8.94e-06,
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch":
|
| 140 |
-
"grad_norm":
|
| 141 |
"learning_rate": 9.440000000000001e-06,
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch":
|
| 147 |
-
"grad_norm":
|
| 148 |
"learning_rate": 9.940000000000001e-06,
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch":
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate": 9.
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 525
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch":
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate": 9.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 550
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate": 9.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 575
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate": 9.
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 600
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 1.
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate": 9.
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 625
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 1.
|
| 189 |
-
"grad_norm": 4.
|
| 190 |
-
"learning_rate": 9.
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 650
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch": 1.
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate": 9.
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 675
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"grad_norm":
|
| 204 |
-
"learning_rate": 9.
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 700
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate": 9.
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 725
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"grad_norm": 3.
|
| 218 |
-
"learning_rate": 9.
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 750
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch":
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 775
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch":
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 800
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch":
|
| 238 |
-
"grad_norm": 3.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 825
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch":
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 850
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch":
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 875
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 900
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch":
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 925
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch":
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 950
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch":
|
| 280 |
-
"grad_norm":
|
| 281 |
-
"learning_rate": 8.
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 975
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch":
|
| 287 |
-
"grad_norm":
|
| 288 |
-
"learning_rate": 8.
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 1000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch":
|
| 294 |
-
"eval_loss": 0.
|
| 295 |
-
"eval_runtime":
|
| 296 |
-
"eval_samples_per_second":
|
| 297 |
-
"eval_steps_per_second": 0.
|
| 298 |
-
"eval_wer": 0.
|
| 299 |
"step": 1000
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch":
|
| 303 |
-
"grad_norm":
|
| 304 |
-
"learning_rate":
|
| 305 |
-
"loss": 0.
|
| 306 |
"step": 1025
|
| 307 |
},
|
| 308 |
{
|
| 309 |
-
"epoch":
|
| 310 |
-
"grad_norm":
|
| 311 |
-
"learning_rate":
|
| 312 |
-
"loss": 0.
|
| 313 |
"step": 1050
|
| 314 |
},
|
| 315 |
{
|
| 316 |
-
"epoch":
|
| 317 |
-
"grad_norm": 3.
|
| 318 |
-
"learning_rate":
|
| 319 |
-
"loss": 0.
|
| 320 |
"step": 1075
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch":
|
| 324 |
-
"grad_norm":
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss": 0.
|
| 327 |
"step": 1100
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch":
|
| 331 |
-
"grad_norm":
|
| 332 |
-
"learning_rate":
|
| 333 |
-
"loss": 0.
|
| 334 |
"step": 1125
|
| 335 |
},
|
| 336 |
{
|
| 337 |
-
"epoch":
|
| 338 |
-
"grad_norm":
|
| 339 |
-
"learning_rate":
|
| 340 |
-
"loss": 0.
|
| 341 |
"step": 1150
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch":
|
| 345 |
-
"grad_norm": 3.
|
| 346 |
-
"learning_rate":
|
| 347 |
-
"loss": 0.
|
| 348 |
"step": 1175
|
| 349 |
},
|
| 350 |
{
|
| 351 |
-
"epoch":
|
| 352 |
-
"grad_norm":
|
| 353 |
-
"learning_rate":
|
| 354 |
-
"loss": 0.
|
| 355 |
"step": 1200
|
| 356 |
},
|
| 357 |
{
|
| 358 |
-
"epoch":
|
| 359 |
-
"grad_norm": 2.
|
| 360 |
-
"learning_rate":
|
| 361 |
-
"loss": 0.
|
| 362 |
"step": 1225
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch":
|
| 366 |
-
"grad_norm":
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss": 0.
|
| 369 |
"step": 1250
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch":
|
| 373 |
-
"grad_norm":
|
| 374 |
-
"learning_rate":
|
| 375 |
-
"loss": 0.
|
| 376 |
"step": 1275
|
| 377 |
},
|
| 378 |
{
|
| 379 |
-
"epoch":
|
| 380 |
-
"grad_norm": 2.
|
| 381 |
-
"learning_rate":
|
| 382 |
-
"loss": 0.
|
| 383 |
"step": 1300
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch":
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate":
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 1325
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch":
|
| 394 |
-
"grad_norm": 3.
|
| 395 |
-
"learning_rate":
|
| 396 |
-
"loss": 0.
|
| 397 |
"step": 1350
|
| 398 |
},
|
| 399 |
{
|
| 400 |
-
"epoch":
|
| 401 |
-
"grad_norm": 1.
|
| 402 |
-
"learning_rate":
|
| 403 |
-
"loss": 0.
|
| 404 |
"step": 1375
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch":
|
| 408 |
-
"grad_norm":
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 1400
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
-
"grad_norm":
|
| 416 |
-
"learning_rate":
|
| 417 |
-
"loss": 0.
|
| 418 |
"step": 1425
|
| 419 |
},
|
| 420 |
{
|
| 421 |
-
"epoch":
|
| 422 |
-
"grad_norm":
|
| 423 |
-
"learning_rate":
|
| 424 |
-
"loss": 0.
|
| 425 |
"step": 1450
|
| 426 |
},
|
| 427 |
{
|
| 428 |
-
"epoch":
|
| 429 |
-
"grad_norm": 1.
|
| 430 |
-
"learning_rate":
|
| 431 |
-
"loss": 0.
|
| 432 |
"step": 1475
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"epoch":
|
| 436 |
-
"grad_norm":
|
| 437 |
-
"learning_rate":
|
| 438 |
-
"loss": 0.
|
| 439 |
"step": 1500
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch":
|
| 443 |
-
"grad_norm": 2.
|
| 444 |
-
"learning_rate":
|
| 445 |
-
"loss": 0.
|
| 446 |
"step": 1525
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch":
|
| 450 |
-
"grad_norm":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 1550
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"grad_norm": 1.
|
| 458 |
-
"learning_rate":
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 1575
|
| 461 |
},
|
| 462 |
{
|
| 463 |
-
"epoch":
|
| 464 |
-
"grad_norm":
|
| 465 |
-
"learning_rate":
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 1600
|
| 468 |
},
|
| 469 |
{
|
| 470 |
-
"epoch":
|
| 471 |
-
"grad_norm":
|
| 472 |
-
"learning_rate":
|
| 473 |
-
"loss": 0.
|
| 474 |
"step": 1625
|
| 475 |
},
|
| 476 |
{
|
| 477 |
-
"epoch":
|
| 478 |
-
"grad_norm":
|
| 479 |
-
"learning_rate":
|
| 480 |
-
"loss": 0.
|
| 481 |
"step": 1650
|
| 482 |
},
|
| 483 |
{
|
| 484 |
-
"epoch":
|
| 485 |
-
"grad_norm": 1.
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss": 0.
|
| 488 |
"step": 1675
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch":
|
| 492 |
-
"grad_norm":
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 1700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"grad_norm":
|
| 500 |
-
"learning_rate":
|
| 501 |
-
"loss": 0.
|
| 502 |
"step": 1725
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"epoch":
|
| 506 |
-
"grad_norm":
|
| 507 |
-
"learning_rate":
|
| 508 |
-
"loss": 0.
|
| 509 |
"step": 1750
|
| 510 |
},
|
| 511 |
{
|
| 512 |
-
"epoch":
|
| 513 |
-
"grad_norm":
|
| 514 |
-
"learning_rate":
|
| 515 |
-
"loss": 0.
|
| 516 |
"step": 1775
|
| 517 |
},
|
| 518 |
{
|
| 519 |
-
"epoch":
|
| 520 |
-
"grad_norm":
|
| 521 |
-
"learning_rate":
|
| 522 |
-
"loss": 0.
|
| 523 |
"step": 1800
|
| 524 |
},
|
| 525 |
{
|
| 526 |
-
"epoch":
|
| 527 |
-
"grad_norm": 1.
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 1825
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch":
|
| 534 |
-
"grad_norm": 1.
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 1850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch":
|
| 541 |
-
"grad_norm": 1.
|
| 542 |
-
"learning_rate":
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 1875
|
| 545 |
},
|
| 546 |
{
|
| 547 |
-
"epoch":
|
| 548 |
-
"grad_norm": 1.
|
| 549 |
-
"learning_rate":
|
| 550 |
-
"loss": 0.
|
| 551 |
"step": 1900
|
| 552 |
},
|
| 553 |
{
|
| 554 |
-
"epoch":
|
| 555 |
-
"grad_norm":
|
| 556 |
-
"learning_rate":
|
| 557 |
-
"loss": 0.
|
| 558 |
"step": 1925
|
| 559 |
},
|
| 560 |
{
|
| 561 |
-
"epoch":
|
| 562 |
-
"grad_norm": 1.
|
| 563 |
-
"learning_rate":
|
| 564 |
-
"loss": 0.
|
| 565 |
"step": 1950
|
| 566 |
},
|
| 567 |
{
|
| 568 |
-
"epoch":
|
| 569 |
-
"grad_norm": 1.
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss": 0.
|
| 572 |
"step": 1975
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch":
|
| 576 |
-
"grad_norm": 1.
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 2000
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch":
|
| 583 |
-
"eval_loss": 0.
|
| 584 |
-
"eval_runtime":
|
| 585 |
-
"eval_samples_per_second":
|
| 586 |
-
"eval_steps_per_second": 0.
|
| 587 |
-
"eval_wer": 0.
|
| 588 |
"step": 2000
|
| 589 |
},
|
| 590 |
{
|
| 591 |
-
"epoch":
|
| 592 |
-
"grad_norm":
|
| 593 |
-
"learning_rate":
|
| 594 |
-
"loss": 0.
|
| 595 |
"step": 2025
|
| 596 |
},
|
| 597 |
{
|
| 598 |
-
"epoch":
|
| 599 |
-
"grad_norm":
|
| 600 |
-
"learning_rate":
|
| 601 |
-
"loss": 0.
|
| 602 |
"step": 2050
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch":
|
| 606 |
-
"grad_norm": 0.
|
| 607 |
-
"learning_rate":
|
| 608 |
-
"loss": 0.
|
| 609 |
"step": 2075
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch":
|
| 613 |
-
"grad_norm": 1.
|
| 614 |
-
"learning_rate":
|
| 615 |
-
"loss": 0.
|
| 616 |
"step": 2100
|
| 617 |
},
|
| 618 |
{
|
| 619 |
-
"epoch":
|
| 620 |
-
"grad_norm": 1.
|
| 621 |
-
"learning_rate":
|
| 622 |
-
"loss": 0.
|
| 623 |
"step": 2125
|
| 624 |
},
|
| 625 |
{
|
| 626 |
-
"epoch":
|
| 627 |
-
"grad_norm":
|
| 628 |
-
"learning_rate":
|
| 629 |
-
"loss": 0.
|
| 630 |
"step": 2150
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch":
|
| 634 |
-
"grad_norm":
|
| 635 |
-
"learning_rate":
|
| 636 |
-
"loss": 0.
|
| 637 |
"step": 2175
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch":
|
| 641 |
-
"grad_norm":
|
| 642 |
-
"learning_rate":
|
| 643 |
-
"loss": 0.
|
| 644 |
"step": 2200
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch":
|
| 648 |
-
"grad_norm": 0.
|
| 649 |
-
"learning_rate":
|
| 650 |
-
"loss": 0.
|
| 651 |
"step": 2225
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch":
|
| 655 |
-
"grad_norm":
|
| 656 |
-
"learning_rate":
|
| 657 |
-
"loss": 0.
|
| 658 |
"step": 2250
|
| 659 |
},
|
| 660 |
{
|
| 661 |
-
"epoch":
|
| 662 |
-
"grad_norm": 1.
|
| 663 |
-
"learning_rate":
|
| 664 |
-
"loss": 0.
|
| 665 |
"step": 2275
|
| 666 |
},
|
| 667 |
{
|
| 668 |
-
"epoch":
|
| 669 |
-
"grad_norm": 0.
|
| 670 |
-
"learning_rate":
|
| 671 |
-
"loss": 0.
|
| 672 |
"step": 2300
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"epoch":
|
| 676 |
-
"grad_norm":
|
| 677 |
-
"learning_rate":
|
| 678 |
-
"loss": 0.
|
| 679 |
"step": 2325
|
| 680 |
},
|
| 681 |
{
|
| 682 |
-
"epoch":
|
| 683 |
-
"grad_norm":
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss": 0.
|
| 686 |
"step": 2350
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch":
|
| 690 |
-
"grad_norm": 0.
|
| 691 |
-
"learning_rate":
|
| 692 |
-
"loss": 0.
|
| 693 |
"step": 2375
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch":
|
| 697 |
-
"grad_norm": 1.
|
| 698 |
-
"learning_rate":
|
| 699 |
-
"loss": 0.
|
| 700 |
"step": 2400
|
| 701 |
},
|
| 702 |
{
|
| 703 |
-
"epoch":
|
| 704 |
-
"grad_norm": 0.
|
| 705 |
-
"learning_rate":
|
| 706 |
-
"loss": 0.
|
| 707 |
"step": 2425
|
| 708 |
},
|
| 709 |
{
|
| 710 |
-
"epoch":
|
| 711 |
-
"grad_norm": 0.
|
| 712 |
-
"learning_rate":
|
| 713 |
-
"loss": 0.
|
| 714 |
"step": 2450
|
| 715 |
},
|
| 716 |
{
|
| 717 |
-
"epoch":
|
| 718 |
-
"grad_norm": 0.
|
| 719 |
-
"learning_rate":
|
| 720 |
-
"loss": 0.
|
| 721 |
"step": 2475
|
| 722 |
},
|
| 723 |
{
|
| 724 |
-
"epoch":
|
| 725 |
-
"grad_norm": 0.
|
| 726 |
-
"learning_rate":
|
| 727 |
-
"loss": 0.
|
| 728 |
"step": 2500
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch":
|
| 732 |
-
"grad_norm":
|
| 733 |
-
"learning_rate":
|
| 734 |
-
"loss": 0.
|
| 735 |
"step": 2525
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
-
"grad_norm": 0.
|
| 740 |
-
"learning_rate":
|
| 741 |
-
"loss": 0.
|
| 742 |
"step": 2550
|
| 743 |
},
|
| 744 |
{
|
| 745 |
-
"epoch":
|
| 746 |
-
"grad_norm": 0.
|
| 747 |
-
"learning_rate":
|
| 748 |
-
"loss": 0.
|
| 749 |
"step": 2575
|
| 750 |
},
|
| 751 |
{
|
| 752 |
-
"epoch":
|
| 753 |
-
"grad_norm":
|
| 754 |
-
"learning_rate":
|
| 755 |
-
"loss": 0.
|
| 756 |
"step": 2600
|
| 757 |
},
|
| 758 |
{
|
| 759 |
-
"epoch":
|
| 760 |
-
"grad_norm": 0.
|
| 761 |
-
"learning_rate":
|
| 762 |
-
"loss": 0.
|
| 763 |
"step": 2625
|
| 764 |
},
|
| 765 |
{
|
| 766 |
-
"epoch":
|
| 767 |
-
"grad_norm": 0.
|
| 768 |
-
"learning_rate":
|
| 769 |
-
"loss": 0.
|
| 770 |
"step": 2650
|
| 771 |
},
|
| 772 |
{
|
| 773 |
-
"epoch":
|
| 774 |
-
"grad_norm": 0.
|
| 775 |
-
"learning_rate":
|
| 776 |
-
"loss": 0.
|
| 777 |
"step": 2675
|
| 778 |
},
|
| 779 |
{
|
| 780 |
-
"epoch":
|
| 781 |
-
"grad_norm": 1.
|
| 782 |
-
"learning_rate":
|
| 783 |
-
"loss": 0.
|
| 784 |
"step": 2700
|
| 785 |
},
|
| 786 |
{
|
| 787 |
-
"epoch":
|
| 788 |
-
"grad_norm": 0.
|
| 789 |
-
"learning_rate":
|
| 790 |
-
"loss": 0.
|
| 791 |
"step": 2725
|
| 792 |
},
|
| 793 |
{
|
| 794 |
-
"epoch":
|
| 795 |
-
"grad_norm": 0.
|
| 796 |
-
"learning_rate":
|
| 797 |
-
"loss": 0.
|
| 798 |
"step": 2750
|
| 799 |
},
|
| 800 |
{
|
| 801 |
-
"epoch":
|
| 802 |
-
"grad_norm": 0.
|
| 803 |
-
"learning_rate":
|
| 804 |
-
"loss": 0.
|
| 805 |
"step": 2775
|
| 806 |
},
|
| 807 |
{
|
| 808 |
-
"epoch":
|
| 809 |
-
"grad_norm":
|
| 810 |
-
"learning_rate":
|
| 811 |
-
"loss": 0.
|
| 812 |
"step": 2800
|
| 813 |
},
|
| 814 |
{
|
| 815 |
-
"epoch":
|
| 816 |
-
"grad_norm":
|
| 817 |
-
"learning_rate":
|
| 818 |
-
"loss": 0.
|
| 819 |
"step": 2825
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch":
|
| 823 |
-
"grad_norm": 0.
|
| 824 |
-
"learning_rate":
|
| 825 |
-
"loss": 0.
|
| 826 |
"step": 2850
|
| 827 |
},
|
| 828 |
{
|
| 829 |
-
"epoch":
|
| 830 |
-
"grad_norm": 0.
|
| 831 |
-
"learning_rate":
|
| 832 |
-
"loss": 0.
|
| 833 |
"step": 2875
|
| 834 |
},
|
| 835 |
{
|
| 836 |
-
"epoch":
|
| 837 |
-
"grad_norm": 0.
|
| 838 |
-
"learning_rate": 4.
|
| 839 |
-
"loss": 0.
|
| 840 |
"step": 2900
|
| 841 |
},
|
| 842 |
{
|
| 843 |
-
"epoch":
|
| 844 |
-
"grad_norm": 0.
|
| 845 |
-
"learning_rate":
|
| 846 |
-
"loss": 0.
|
| 847 |
"step": 2925
|
| 848 |
},
|
| 849 |
{
|
| 850 |
-
"epoch":
|
| 851 |
-
"grad_norm": 0.
|
| 852 |
-
"learning_rate":
|
| 853 |
-
"loss": 0.
|
| 854 |
"step": 2950
|
| 855 |
},
|
| 856 |
{
|
| 857 |
-
"epoch":
|
| 858 |
-
"grad_norm":
|
| 859 |
-
"learning_rate":
|
| 860 |
-
"loss": 0.
|
| 861 |
"step": 2975
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch":
|
| 865 |
-
"grad_norm": 0.
|
| 866 |
-
"learning_rate":
|
| 867 |
-
"loss": 0.
|
| 868 |
"step": 3000
|
| 869 |
},
|
| 870 |
{
|
| 871 |
-
"epoch":
|
| 872 |
-
"eval_loss": 0.
|
| 873 |
-
"eval_runtime":
|
| 874 |
-
"eval_samples_per_second":
|
| 875 |
-
"eval_steps_per_second": 0.
|
| 876 |
-
"eval_wer": 0.
|
| 877 |
"step": 3000
|
| 878 |
},
|
| 879 |
{
|
| 880 |
-
"epoch":
|
| 881 |
-
"
|
| 882 |
-
"
|
| 883 |
-
"
|
| 884 |
-
"
|
| 885 |
-
|
| 886 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 887 |
}
|
| 888 |
],
|
| 889 |
"logging_steps": 25,
|
| 890 |
-
"max_steps":
|
| 891 |
"num_input_tokens_seen": 0,
|
| 892 |
-
"num_train_epochs":
|
| 893 |
"save_steps": 1000,
|
| 894 |
"stateful_callbacks": {
|
| 895 |
"TrainerControl": {
|
|
@@ -903,7 +1481,7 @@
|
|
| 903 |
"attributes": {}
|
| 904 |
}
|
| 905 |
},
|
| 906 |
-
"total_flos": 4.
|
| 907 |
"train_batch_size": 4,
|
| 908 |
"trial_name": null,
|
| 909 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 14.79466271312083,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 5000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.07412898443291327,
|
| 14 |
+
"grad_norm": 11.929125785827637,
|
| 15 |
+
"learning_rate": 4.4e-07,
|
| 16 |
+
"loss": 1.6381,
|
| 17 |
"step": 25
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.14825796886582654,
|
| 21 |
+
"grad_norm": 8.018507957458496,
|
| 22 |
+
"learning_rate": 9.400000000000001e-07,
|
| 23 |
+
"loss": 1.3896,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.2223869532987398,
|
| 28 |
+
"grad_norm": 6.990742206573486,
|
| 29 |
"learning_rate": 1.44e-06,
|
| 30 |
+
"loss": 1.2335,
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.2965159377316531,
|
| 35 |
+
"grad_norm": 7.199395656585693,
|
| 36 |
"learning_rate": 1.94e-06,
|
| 37 |
+
"loss": 1.0625,
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.37064492216456635,
|
| 42 |
+
"grad_norm": 5.8580546379089355,
|
| 43 |
"learning_rate": 2.4400000000000004e-06,
|
| 44 |
+
"loss": 0.9354,
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.4447739065974796,
|
| 49 |
+
"grad_norm": 6.414823055267334,
|
| 50 |
"learning_rate": 2.9400000000000002e-06,
|
| 51 |
+
"loss": 0.8125,
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.5189028910303929,
|
| 56 |
+
"grad_norm": 6.512091159820557,
|
| 57 |
"learning_rate": 3.44e-06,
|
| 58 |
+
"loss": 0.7764,
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.5930318754633062,
|
| 63 |
+
"grad_norm": 6.021186351776123,
|
| 64 |
"learning_rate": 3.94e-06,
|
| 65 |
+
"loss": 0.7381,
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.6671608598962194,
|
| 70 |
+
"grad_norm": 5.814141273498535,
|
| 71 |
"learning_rate": 4.440000000000001e-06,
|
| 72 |
+
"loss": 0.6923,
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.7412898443291327,
|
| 77 |
+
"grad_norm": 5.2472968101501465,
|
| 78 |
"learning_rate": 4.94e-06,
|
| 79 |
+
"loss": 0.6586,
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.815418828762046,
|
| 84 |
+
"grad_norm": 5.4379401206970215,
|
| 85 |
"learning_rate": 5.4400000000000004e-06,
|
| 86 |
+
"loss": 0.6041,
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.8895478131949592,
|
| 91 |
+
"grad_norm": 5.554481029510498,
|
| 92 |
"learning_rate": 5.94e-06,
|
| 93 |
+
"loss": 0.5964,
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.9636767976278725,
|
| 98 |
+
"grad_norm": 6.5946502685546875,
|
| 99 |
"learning_rate": 6.440000000000001e-06,
|
| 100 |
+
"loss": 0.5586,
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 1.0355819125277983,
|
| 105 |
+
"grad_norm": 4.164698600769043,
|
| 106 |
"learning_rate": 6.9400000000000005e-06,
|
| 107 |
+
"loss": 0.4838,
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 1.1097108969607117,
|
| 112 |
+
"grad_norm": 6.609652519226074,
|
| 113 |
"learning_rate": 7.440000000000001e-06,
|
| 114 |
+
"loss": 0.4744,
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 1.1838398813936248,
|
| 119 |
+
"grad_norm": 4.401813507080078,
|
| 120 |
"learning_rate": 7.94e-06,
|
| 121 |
+
"loss": 0.4447,
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 1.2579688658265382,
|
| 126 |
+
"grad_norm": 4.387920379638672,
|
| 127 |
"learning_rate": 8.44e-06,
|
| 128 |
+
"loss": 0.4425,
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 1.3320978502594514,
|
| 133 |
+
"grad_norm": 4.439324855804443,
|
| 134 |
"learning_rate": 8.94e-06,
|
| 135 |
+
"loss": 0.3978,
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 1.4062268346923648,
|
| 140 |
+
"grad_norm": 4.0201826095581055,
|
| 141 |
"learning_rate": 9.440000000000001e-06,
|
| 142 |
+
"loss": 0.4059,
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 1.480355819125278,
|
| 147 |
+
"grad_norm": 4.707728385925293,
|
| 148 |
"learning_rate": 9.940000000000001e-06,
|
| 149 |
+
"loss": 0.3959,
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 1.5544848035581913,
|
| 154 |
+
"grad_norm": 4.224379062652588,
|
| 155 |
+
"learning_rate": 9.951111111111111e-06,
|
| 156 |
+
"loss": 0.3747,
|
| 157 |
"step": 525
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 1.6286137879911045,
|
| 161 |
+
"grad_norm": 4.27752161026001,
|
| 162 |
+
"learning_rate": 9.895555555555557e-06,
|
| 163 |
+
"loss": 0.3749,
|
| 164 |
"step": 550
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 1.7027427724240178,
|
| 168 |
+
"grad_norm": 4.656911373138428,
|
| 169 |
+
"learning_rate": 9.84e-06,
|
| 170 |
+
"loss": 0.3667,
|
| 171 |
"step": 575
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 1.7768717568569312,
|
| 175 |
+
"grad_norm": 4.724191188812256,
|
| 176 |
+
"learning_rate": 9.784444444444445e-06,
|
| 177 |
+
"loss": 0.3474,
|
| 178 |
"step": 600
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 1.8510007412898444,
|
| 182 |
+
"grad_norm": 4.367356300354004,
|
| 183 |
+
"learning_rate": 9.72888888888889e-06,
|
| 184 |
+
"loss": 0.3445,
|
| 185 |
"step": 625
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 1.9251297257227575,
|
| 189 |
+
"grad_norm": 4.204794406890869,
|
| 190 |
+
"learning_rate": 9.673333333333334e-06,
|
| 191 |
+
"loss": 0.3353,
|
| 192 |
"step": 650
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 1.999258710155671,
|
| 196 |
+
"grad_norm": 3.796111583709717,
|
| 197 |
+
"learning_rate": 9.617777777777778e-06,
|
| 198 |
+
"loss": 0.3136,
|
| 199 |
"step": 675
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 2.0711638250555966,
|
| 203 |
+
"grad_norm": 3.014369249343872,
|
| 204 |
+
"learning_rate": 9.562222222222223e-06,
|
| 205 |
+
"loss": 0.2109,
|
| 206 |
"step": 700
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 2.14529280948851,
|
| 210 |
+
"grad_norm": 3.4962964057922363,
|
| 211 |
+
"learning_rate": 9.506666666666667e-06,
|
| 212 |
+
"loss": 0.2112,
|
| 213 |
"step": 725
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 2.2194217939214234,
|
| 217 |
+
"grad_norm": 3.688762903213501,
|
| 218 |
+
"learning_rate": 9.451111111111112e-06,
|
| 219 |
+
"loss": 0.2122,
|
| 220 |
"step": 750
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 2.2935507783543367,
|
| 224 |
+
"grad_norm": 4.557786464691162,
|
| 225 |
+
"learning_rate": 9.395555555555556e-06,
|
| 226 |
+
"loss": 0.2107,
|
| 227 |
"step": 775
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 2.3676797627872497,
|
| 231 |
+
"grad_norm": 3.3230926990509033,
|
| 232 |
+
"learning_rate": 9.340000000000002e-06,
|
| 233 |
+
"loss": 0.2115,
|
| 234 |
"step": 800
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 2.441808747220163,
|
| 238 |
+
"grad_norm": 3.251685380935669,
|
| 239 |
+
"learning_rate": 9.284444444444444e-06,
|
| 240 |
+
"loss": 0.2021,
|
| 241 |
"step": 825
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 2.5159377316530764,
|
| 245 |
+
"grad_norm": 3.852593421936035,
|
| 246 |
+
"learning_rate": 9.22888888888889e-06,
|
| 247 |
+
"loss": 0.1999,
|
| 248 |
"step": 850
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 2.59006671608599,
|
| 252 |
+
"grad_norm": 3.0748980045318604,
|
| 253 |
+
"learning_rate": 9.173333333333334e-06,
|
| 254 |
+
"loss": 0.1936,
|
| 255 |
"step": 875
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 2.6641957005189028,
|
| 259 |
+
"grad_norm": 2.834657907485962,
|
| 260 |
+
"learning_rate": 9.117777777777778e-06,
|
| 261 |
+
"loss": 0.2031,
|
| 262 |
"step": 900
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 2.738324684951816,
|
| 266 |
+
"grad_norm": 2.9382002353668213,
|
| 267 |
+
"learning_rate": 9.062222222222224e-06,
|
| 268 |
+
"loss": 0.199,
|
| 269 |
"step": 925
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 2.8124536693847295,
|
| 273 |
+
"grad_norm": 2.9481730461120605,
|
| 274 |
+
"learning_rate": 9.006666666666666e-06,
|
| 275 |
+
"loss": 0.1867,
|
| 276 |
"step": 950
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 2.8865826538176425,
|
| 280 |
+
"grad_norm": 3.9300479888916016,
|
| 281 |
+
"learning_rate": 8.951111111111112e-06,
|
| 282 |
+
"loss": 0.203,
|
| 283 |
"step": 975
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 2.960711638250556,
|
| 287 |
+
"grad_norm": 3.3928661346435547,
|
| 288 |
+
"learning_rate": 8.895555555555556e-06,
|
| 289 |
+
"loss": 0.2023,
|
| 290 |
"step": 1000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 2.960711638250556,
|
| 294 |
+
"eval_loss": 0.3293725550174713,
|
| 295 |
+
"eval_runtime": 936.3299,
|
| 296 |
+
"eval_samples_per_second": 3.081,
|
| 297 |
+
"eval_steps_per_second": 0.386,
|
| 298 |
+
"eval_wer": 0.18741347189225221,
|
| 299 |
"step": 1000
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 3.032616753150482,
|
| 303 |
+
"grad_norm": 2.452481269836426,
|
| 304 |
+
"learning_rate": 8.84e-06,
|
| 305 |
+
"loss": 0.152,
|
| 306 |
"step": 1025
|
| 307 |
},
|
| 308 |
{
|
| 309 |
+
"epoch": 3.106745737583395,
|
| 310 |
+
"grad_norm": 2.802464485168457,
|
| 311 |
+
"learning_rate": 8.784444444444446e-06,
|
| 312 |
+
"loss": 0.1076,
|
| 313 |
"step": 1050
|
| 314 |
},
|
| 315 |
{
|
| 316 |
+
"epoch": 3.1808747220163083,
|
| 317 |
+
"grad_norm": 3.293215751647949,
|
| 318 |
+
"learning_rate": 8.72888888888889e-06,
|
| 319 |
+
"loss": 0.1114,
|
| 320 |
"step": 1075
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 3.2550037064492217,
|
| 324 |
+
"grad_norm": 2.4673008918762207,
|
| 325 |
+
"learning_rate": 8.673333333333334e-06,
|
| 326 |
+
"loss": 0.1075,
|
| 327 |
"step": 1100
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 3.329132690882135,
|
| 331 |
+
"grad_norm": 2.8392369747161865,
|
| 332 |
+
"learning_rate": 8.617777777777778e-06,
|
| 333 |
+
"loss": 0.1112,
|
| 334 |
"step": 1125
|
| 335 |
},
|
| 336 |
{
|
| 337 |
+
"epoch": 3.403261675315048,
|
| 338 |
+
"grad_norm": 2.4692773818969727,
|
| 339 |
+
"learning_rate": 8.562222222222224e-06,
|
| 340 |
+
"loss": 0.1109,
|
| 341 |
"step": 1150
|
| 342 |
},
|
| 343 |
{
|
| 344 |
+
"epoch": 3.4773906597479614,
|
| 345 |
+
"grad_norm": 3.809983968734741,
|
| 346 |
+
"learning_rate": 8.506666666666668e-06,
|
| 347 |
+
"loss": 0.1173,
|
| 348 |
"step": 1175
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 3.5515196441808747,
|
| 352 |
+
"grad_norm": 2.7140321731567383,
|
| 353 |
+
"learning_rate": 8.451111111111112e-06,
|
| 354 |
+
"loss": 0.1131,
|
| 355 |
"step": 1200
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 3.625648628613788,
|
| 359 |
+
"grad_norm": 2.5393877029418945,
|
| 360 |
+
"learning_rate": 8.395555555555557e-06,
|
| 361 |
+
"loss": 0.1077,
|
| 362 |
"step": 1225
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 3.6997776130467015,
|
| 366 |
+
"grad_norm": 2.905407667160034,
|
| 367 |
+
"learning_rate": 8.34e-06,
|
| 368 |
+
"loss": 0.1054,
|
| 369 |
"step": 1250
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 3.7739065974796144,
|
| 373 |
+
"grad_norm": 2.8358991146087646,
|
| 374 |
+
"learning_rate": 8.284444444444446e-06,
|
| 375 |
+
"loss": 0.1073,
|
| 376 |
"step": 1275
|
| 377 |
},
|
| 378 |
{
|
| 379 |
+
"epoch": 3.848035581912528,
|
| 380 |
+
"grad_norm": 2.4956672191619873,
|
| 381 |
+
"learning_rate": 8.22888888888889e-06,
|
| 382 |
+
"loss": 0.1052,
|
| 383 |
"step": 1300
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 3.922164566345441,
|
| 387 |
+
"grad_norm": 3.1253950595855713,
|
| 388 |
+
"learning_rate": 8.173333333333334e-06,
|
| 389 |
+
"loss": 0.1064,
|
| 390 |
"step": 1325
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 3.996293550778354,
|
| 394 |
+
"grad_norm": 3.244260549545288,
|
| 395 |
+
"learning_rate": 8.11777777777778e-06,
|
| 396 |
+
"loss": 0.0973,
|
| 397 |
"step": 1350
|
| 398 |
},
|
| 399 |
{
|
| 400 |
+
"epoch": 4.06819866567828,
|
| 401 |
+
"grad_norm": 1.2010555267333984,
|
| 402 |
+
"learning_rate": 8.062222222222222e-06,
|
| 403 |
+
"loss": 0.0602,
|
| 404 |
"step": 1375
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 4.142327650111193,
|
| 408 |
+
"grad_norm": 1.965174913406372,
|
| 409 |
+
"learning_rate": 8.006666666666667e-06,
|
| 410 |
+
"loss": 0.0607,
|
| 411 |
"step": 1400
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 4.216456634544107,
|
| 415 |
+
"grad_norm": 1.7773408889770508,
|
| 416 |
+
"learning_rate": 7.951111111111111e-06,
|
| 417 |
+
"loss": 0.0539,
|
| 418 |
"step": 1425
|
| 419 |
},
|
| 420 |
{
|
| 421 |
+
"epoch": 4.29058561897702,
|
| 422 |
+
"grad_norm": 2.6301095485687256,
|
| 423 |
+
"learning_rate": 7.895555555555557e-06,
|
| 424 |
+
"loss": 0.0606,
|
| 425 |
"step": 1450
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 4.364714603409933,
|
| 429 |
+
"grad_norm": 1.90291428565979,
|
| 430 |
+
"learning_rate": 7.840000000000001e-06,
|
| 431 |
+
"loss": 0.0584,
|
| 432 |
"step": 1475
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"epoch": 4.438843587842847,
|
| 436 |
+
"grad_norm": 2.0360007286071777,
|
| 437 |
+
"learning_rate": 7.784444444444445e-06,
|
| 438 |
+
"loss": 0.0553,
|
| 439 |
"step": 1500
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 4.51297257227576,
|
| 443 |
+
"grad_norm": 2.2208218574523926,
|
| 444 |
+
"learning_rate": 7.72888888888889e-06,
|
| 445 |
+
"loss": 0.0656,
|
| 446 |
"step": 1525
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 4.5871015567086735,
|
| 450 |
+
"grad_norm": 1.9523277282714844,
|
| 451 |
+
"learning_rate": 7.673333333333333e-06,
|
| 452 |
+
"loss": 0.0537,
|
| 453 |
"step": 1550
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 4.661230541141586,
|
| 457 |
+
"grad_norm": 1.7108356952667236,
|
| 458 |
+
"learning_rate": 7.617777777777778e-06,
|
| 459 |
+
"loss": 0.0575,
|
| 460 |
"step": 1575
|
| 461 |
},
|
| 462 |
{
|
| 463 |
+
"epoch": 4.735359525574499,
|
| 464 |
+
"grad_norm": 1.8699008226394653,
|
| 465 |
+
"learning_rate": 7.562222222222223e-06,
|
| 466 |
+
"loss": 0.0548,
|
| 467 |
"step": 1600
|
| 468 |
},
|
| 469 |
{
|
| 470 |
+
"epoch": 4.809488510007413,
|
| 471 |
+
"grad_norm": 2.0190742015838623,
|
| 472 |
+
"learning_rate": 7.506666666666668e-06,
|
| 473 |
+
"loss": 0.0548,
|
| 474 |
"step": 1625
|
| 475 |
},
|
| 476 |
{
|
| 477 |
+
"epoch": 4.883617494440326,
|
| 478 |
+
"grad_norm": 1.592809796333313,
|
| 479 |
+
"learning_rate": 7.451111111111111e-06,
|
| 480 |
+
"loss": 0.0536,
|
| 481 |
"step": 1650
|
| 482 |
},
|
| 483 |
{
|
| 484 |
+
"epoch": 4.957746478873239,
|
| 485 |
+
"grad_norm": 1.7714742422103882,
|
| 486 |
+
"learning_rate": 7.395555555555556e-06,
|
| 487 |
+
"loss": 0.0522,
|
| 488 |
"step": 1675
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 5.029651593773165,
|
| 492 |
+
"grad_norm": 0.9238072037696838,
|
| 493 |
+
"learning_rate": 7.340000000000001e-06,
|
| 494 |
+
"loss": 0.0427,
|
| 495 |
"step": 1700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 5.103780578206079,
|
| 499 |
+
"grad_norm": 1.1125034093856812,
|
| 500 |
+
"learning_rate": 7.284444444444445e-06,
|
| 501 |
+
"loss": 0.0269,
|
| 502 |
"step": 1725
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"epoch": 5.177909562638992,
|
| 506 |
+
"grad_norm": 1.247423768043518,
|
| 507 |
+
"learning_rate": 7.22888888888889e-06,
|
| 508 |
+
"loss": 0.0257,
|
| 509 |
"step": 1750
|
| 510 |
},
|
| 511 |
{
|
| 512 |
+
"epoch": 5.252038547071905,
|
| 513 |
+
"grad_norm": 1.1966558694839478,
|
| 514 |
+
"learning_rate": 7.173333333333335e-06,
|
| 515 |
+
"loss": 0.029,
|
| 516 |
"step": 1775
|
| 517 |
},
|
| 518 |
{
|
| 519 |
+
"epoch": 5.326167531504819,
|
| 520 |
+
"grad_norm": 0.9568462371826172,
|
| 521 |
+
"learning_rate": 7.117777777777778e-06,
|
| 522 |
+
"loss": 0.0294,
|
| 523 |
"step": 1800
|
| 524 |
},
|
| 525 |
{
|
| 526 |
+
"epoch": 5.400296515937732,
|
| 527 |
+
"grad_norm": 1.0882899761199951,
|
| 528 |
+
"learning_rate": 7.062222222222223e-06,
|
| 529 |
+
"loss": 0.028,
|
| 530 |
"step": 1825
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 5.474425500370645,
|
| 534 |
+
"grad_norm": 1.1803560256958008,
|
| 535 |
+
"learning_rate": 7.006666666666667e-06,
|
| 536 |
+
"loss": 0.0295,
|
| 537 |
"step": 1850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 5.548554484803558,
|
| 541 |
+
"grad_norm": 1.2876580953598022,
|
| 542 |
+
"learning_rate": 6.951111111111112e-06,
|
| 543 |
+
"loss": 0.0334,
|
| 544 |
"step": 1875
|
| 545 |
},
|
| 546 |
{
|
| 547 |
+
"epoch": 5.622683469236471,
|
| 548 |
+
"grad_norm": 1.7007198333740234,
|
| 549 |
+
"learning_rate": 6.8955555555555565e-06,
|
| 550 |
+
"loss": 0.0294,
|
| 551 |
"step": 1900
|
| 552 |
},
|
| 553 |
{
|
| 554 |
+
"epoch": 5.696812453669384,
|
| 555 |
+
"grad_norm": 0.9945253729820251,
|
| 556 |
+
"learning_rate": 6.8400000000000014e-06,
|
| 557 |
+
"loss": 0.0287,
|
| 558 |
"step": 1925
|
| 559 |
},
|
| 560 |
{
|
| 561 |
+
"epoch": 5.770941438102298,
|
| 562 |
+
"grad_norm": 1.3555738925933838,
|
| 563 |
+
"learning_rate": 6.784444444444445e-06,
|
| 564 |
+
"loss": 0.0278,
|
| 565 |
"step": 1950
|
| 566 |
},
|
| 567 |
{
|
| 568 |
+
"epoch": 5.845070422535211,
|
| 569 |
+
"grad_norm": 1.6404248476028442,
|
| 570 |
+
"learning_rate": 6.7288888888888895e-06,
|
| 571 |
+
"loss": 0.027,
|
| 572 |
"step": 1975
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 5.919199406968125,
|
| 576 |
+
"grad_norm": 1.1645821332931519,
|
| 577 |
+
"learning_rate": 6.6733333333333335e-06,
|
| 578 |
+
"loss": 0.0309,
|
| 579 |
"step": 2000
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 5.919199406968125,
|
| 583 |
+
"eval_loss": 0.3413762152194977,
|
| 584 |
+
"eval_runtime": 911.1637,
|
| 585 |
+
"eval_samples_per_second": 3.166,
|
| 586 |
+
"eval_steps_per_second": 0.396,
|
| 587 |
+
"eval_wer": 0.17174013365922025,
|
| 588 |
"step": 2000
|
| 589 |
},
|
| 590 |
{
|
| 591 |
+
"epoch": 5.993328391401038,
|
| 592 |
+
"grad_norm": 1.239682674407959,
|
| 593 |
+
"learning_rate": 6.617777777777778e-06,
|
| 594 |
+
"loss": 0.0246,
|
| 595 |
"step": 2025
|
| 596 |
},
|
| 597 |
{
|
| 598 |
+
"epoch": 6.065233506300964,
|
| 599 |
+
"grad_norm": 0.5780806541442871,
|
| 600 |
+
"learning_rate": 6.562222222222223e-06,
|
| 601 |
+
"loss": 0.0158,
|
| 602 |
"step": 2050
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 6.139362490733877,
|
| 606 |
+
"grad_norm": 0.7804180383682251,
|
| 607 |
+
"learning_rate": 6.5066666666666665e-06,
|
| 608 |
+
"loss": 0.0149,
|
| 609 |
"step": 2075
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 6.21349147516679,
|
| 613 |
+
"grad_norm": 1.2513881921768188,
|
| 614 |
+
"learning_rate": 6.451111111111111e-06,
|
| 615 |
+
"loss": 0.0186,
|
| 616 |
"step": 2100
|
| 617 |
},
|
| 618 |
{
|
| 619 |
+
"epoch": 6.287620459599704,
|
| 620 |
+
"grad_norm": 1.0758005380630493,
|
| 621 |
+
"learning_rate": 6.395555555555556e-06,
|
| 622 |
+
"loss": 0.0176,
|
| 623 |
"step": 2125
|
| 624 |
},
|
| 625 |
{
|
| 626 |
+
"epoch": 6.3617494440326166,
|
| 627 |
+
"grad_norm": 1.6794716119766235,
|
| 628 |
+
"learning_rate": 6.34e-06,
|
| 629 |
+
"loss": 0.0166,
|
| 630 |
"step": 2150
|
| 631 |
},
|
| 632 |
{
|
| 633 |
+
"epoch": 6.43587842846553,
|
| 634 |
+
"grad_norm": 0.4483010768890381,
|
| 635 |
+
"learning_rate": 6.284444444444445e-06,
|
| 636 |
+
"loss": 0.0116,
|
| 637 |
"step": 2175
|
| 638 |
},
|
| 639 |
{
|
| 640 |
+
"epoch": 6.510007412898443,
|
| 641 |
+
"grad_norm": 0.6224784255027771,
|
| 642 |
+
"learning_rate": 6.22888888888889e-06,
|
| 643 |
+
"loss": 0.0122,
|
| 644 |
"step": 2200
|
| 645 |
},
|
| 646 |
{
|
| 647 |
+
"epoch": 6.584136397331356,
|
| 648 |
+
"grad_norm": 0.5704225897789001,
|
| 649 |
+
"learning_rate": 6.173333333333333e-06,
|
| 650 |
+
"loss": 0.0138,
|
| 651 |
"step": 2225
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 6.65826538176427,
|
| 655 |
+
"grad_norm": 0.976425051689148,
|
| 656 |
+
"learning_rate": 6.117777777777778e-06,
|
| 657 |
+
"loss": 0.0142,
|
| 658 |
"step": 2250
|
| 659 |
},
|
| 660 |
{
|
| 661 |
+
"epoch": 6.732394366197183,
|
| 662 |
+
"grad_norm": 1.829641580581665,
|
| 663 |
+
"learning_rate": 6.062222222222223e-06,
|
| 664 |
+
"loss": 0.0139,
|
| 665 |
"step": 2275
|
| 666 |
},
|
| 667 |
{
|
| 668 |
+
"epoch": 6.806523350630096,
|
| 669 |
+
"grad_norm": 0.645282506942749,
|
| 670 |
+
"learning_rate": 6.006666666666667e-06,
|
| 671 |
+
"loss": 0.0154,
|
| 672 |
"step": 2300
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"epoch": 6.88065233506301,
|
| 676 |
+
"grad_norm": 0.9580066204071045,
|
| 677 |
+
"learning_rate": 5.951111111111112e-06,
|
| 678 |
+
"loss": 0.0119,
|
| 679 |
"step": 2325
|
| 680 |
},
|
| 681 |
{
|
| 682 |
+
"epoch": 6.954781319495923,
|
| 683 |
+
"grad_norm": 1.3833215236663818,
|
| 684 |
+
"learning_rate": 5.895555555555557e-06,
|
| 685 |
+
"loss": 0.0163,
|
| 686 |
"step": 2350
|
| 687 |
},
|
| 688 |
{
|
| 689 |
+
"epoch": 7.026686434395849,
|
| 690 |
+
"grad_norm": 0.3075716495513916,
|
| 691 |
+
"learning_rate": 5.84e-06,
|
| 692 |
+
"loss": 0.0115,
|
| 693 |
"step": 2375
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 7.100815418828762,
|
| 697 |
+
"grad_norm": 1.460079312324524,
|
| 698 |
+
"learning_rate": 5.784444444444445e-06,
|
| 699 |
+
"loss": 0.0098,
|
| 700 |
"step": 2400
|
| 701 |
},
|
| 702 |
{
|
| 703 |
+
"epoch": 7.174944403261676,
|
| 704 |
+
"grad_norm": 0.24449419975280762,
|
| 705 |
+
"learning_rate": 5.72888888888889e-06,
|
| 706 |
+
"loss": 0.0082,
|
| 707 |
"step": 2425
|
| 708 |
},
|
| 709 |
{
|
| 710 |
+
"epoch": 7.2490733876945885,
|
| 711 |
+
"grad_norm": 0.4475979208946228,
|
| 712 |
+
"learning_rate": 5.673333333333334e-06,
|
| 713 |
+
"loss": 0.008,
|
| 714 |
"step": 2450
|
| 715 |
},
|
| 716 |
{
|
| 717 |
+
"epoch": 7.3232023721275015,
|
| 718 |
+
"grad_norm": 0.9485690593719482,
|
| 719 |
+
"learning_rate": 5.617777777777779e-06,
|
| 720 |
+
"loss": 0.0086,
|
| 721 |
"step": 2475
|
| 722 |
},
|
| 723 |
{
|
| 724 |
+
"epoch": 7.397331356560415,
|
| 725 |
+
"grad_norm": 0.4818005859851837,
|
| 726 |
+
"learning_rate": 5.562222222222222e-06,
|
| 727 |
+
"loss": 0.0079,
|
| 728 |
"step": 2500
|
| 729 |
},
|
| 730 |
{
|
| 731 |
+
"epoch": 7.471460340993328,
|
| 732 |
+
"grad_norm": 0.2728360891342163,
|
| 733 |
+
"learning_rate": 5.506666666666667e-06,
|
| 734 |
+
"loss": 0.0071,
|
| 735 |
"step": 2525
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 7.545589325426242,
|
| 739 |
+
"grad_norm": 0.36738815903663635,
|
| 740 |
+
"learning_rate": 5.451111111111112e-06,
|
| 741 |
+
"loss": 0.0071,
|
| 742 |
"step": 2550
|
| 743 |
},
|
| 744 |
{
|
| 745 |
+
"epoch": 7.619718309859155,
|
| 746 |
+
"grad_norm": 0.5995059609413147,
|
| 747 |
+
"learning_rate": 5.3955555555555565e-06,
|
| 748 |
+
"loss": 0.0071,
|
| 749 |
"step": 2575
|
| 750 |
},
|
| 751 |
{
|
| 752 |
+
"epoch": 7.693847294292068,
|
| 753 |
+
"grad_norm": 0.6667615175247192,
|
| 754 |
+
"learning_rate": 5.3400000000000005e-06,
|
| 755 |
+
"loss": 0.0071,
|
| 756 |
"step": 2600
|
| 757 |
},
|
| 758 |
{
|
| 759 |
+
"epoch": 7.767976278724982,
|
| 760 |
+
"grad_norm": 0.9709383845329285,
|
| 761 |
+
"learning_rate": 5.2844444444444454e-06,
|
| 762 |
+
"loss": 0.0077,
|
| 763 |
"step": 2625
|
| 764 |
},
|
| 765 |
{
|
| 766 |
+
"epoch": 7.842105263157895,
|
| 767 |
+
"grad_norm": 0.9174025058746338,
|
| 768 |
+
"learning_rate": 5.228888888888889e-06,
|
| 769 |
+
"loss": 0.0064,
|
| 770 |
"step": 2650
|
| 771 |
},
|
| 772 |
{
|
| 773 |
+
"epoch": 7.916234247590808,
|
| 774 |
+
"grad_norm": 0.4343569576740265,
|
| 775 |
+
"learning_rate": 5.1733333333333335e-06,
|
| 776 |
+
"loss": 0.006,
|
| 777 |
"step": 2675
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 7.9903632320237215,
|
| 781 |
+
"grad_norm": 1.9002019166946411,
|
| 782 |
+
"learning_rate": 5.117777777777778e-06,
|
| 783 |
+
"loss": 0.0071,
|
| 784 |
"step": 2700
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"epoch": 8.062268346923647,
|
| 788 |
+
"grad_norm": 0.3372269570827484,
|
| 789 |
+
"learning_rate": 5.062222222222222e-06,
|
| 790 |
+
"loss": 0.005,
|
| 791 |
"step": 2725
|
| 792 |
},
|
| 793 |
{
|
| 794 |
+
"epoch": 8.13639733135656,
|
| 795 |
+
"grad_norm": 0.8460608720779419,
|
| 796 |
+
"learning_rate": 5.006666666666667e-06,
|
| 797 |
+
"loss": 0.005,
|
| 798 |
"step": 2750
|
| 799 |
},
|
| 800 |
{
|
| 801 |
+
"epoch": 8.210526315789474,
|
| 802 |
+
"grad_norm": 0.24673226475715637,
|
| 803 |
+
"learning_rate": 4.951111111111111e-06,
|
| 804 |
+
"loss": 0.0066,
|
| 805 |
"step": 2775
|
| 806 |
},
|
| 807 |
{
|
| 808 |
+
"epoch": 8.284655300222386,
|
| 809 |
+
"grad_norm": 0.20106859505176544,
|
| 810 |
+
"learning_rate": 4.895555555555556e-06,
|
| 811 |
+
"loss": 0.0037,
|
| 812 |
"step": 2800
|
| 813 |
},
|
| 814 |
{
|
| 815 |
+
"epoch": 8.3587842846553,
|
| 816 |
+
"grad_norm": 0.1510283499956131,
|
| 817 |
+
"learning_rate": 4.84e-06,
|
| 818 |
+
"loss": 0.0044,
|
| 819 |
"step": 2825
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 8.432913269088214,
|
| 823 |
+
"grad_norm": 0.7809015512466431,
|
| 824 |
+
"learning_rate": 4.784444444444445e-06,
|
| 825 |
+
"loss": 0.0041,
|
| 826 |
"step": 2850
|
| 827 |
},
|
| 828 |
{
|
| 829 |
+
"epoch": 8.507042253521126,
|
| 830 |
+
"grad_norm": 0.3642624318599701,
|
| 831 |
+
"learning_rate": 4.728888888888889e-06,
|
| 832 |
+
"loss": 0.0048,
|
| 833 |
"step": 2875
|
| 834 |
},
|
| 835 |
{
|
| 836 |
+
"epoch": 8.58117123795404,
|
| 837 |
+
"grad_norm": 0.20495444536209106,
|
| 838 |
+
"learning_rate": 4.673333333333333e-06,
|
| 839 |
+
"loss": 0.0036,
|
| 840 |
"step": 2900
|
| 841 |
},
|
| 842 |
{
|
| 843 |
+
"epoch": 8.655300222386954,
|
| 844 |
+
"grad_norm": 0.14625044167041779,
|
| 845 |
+
"learning_rate": 4.617777777777778e-06,
|
| 846 |
+
"loss": 0.0041,
|
| 847 |
"step": 2925
|
| 848 |
},
|
| 849 |
{
|
| 850 |
+
"epoch": 8.729429206819866,
|
| 851 |
+
"grad_norm": 0.1600050926208496,
|
| 852 |
+
"learning_rate": 4.562222222222222e-06,
|
| 853 |
+
"loss": 0.0044,
|
| 854 |
"step": 2950
|
| 855 |
},
|
| 856 |
{
|
| 857 |
+
"epoch": 8.80355819125278,
|
| 858 |
+
"grad_norm": 1.0638575553894043,
|
| 859 |
+
"learning_rate": 4.506666666666667e-06,
|
| 860 |
+
"loss": 0.0036,
|
| 861 |
"step": 2975
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 8.877687175685693,
|
| 865 |
+
"grad_norm": 0.16608619689941406,
|
| 866 |
+
"learning_rate": 4.451111111111112e-06,
|
| 867 |
+
"loss": 0.0037,
|
| 868 |
"step": 3000
|
| 869 |
},
|
| 870 |
{
|
| 871 |
+
"epoch": 8.877687175685693,
|
| 872 |
+
"eval_loss": 0.37816545367240906,
|
| 873 |
+
"eval_runtime": 954.3156,
|
| 874 |
+
"eval_samples_per_second": 3.023,
|
| 875 |
+
"eval_steps_per_second": 0.378,
|
| 876 |
+
"eval_wer": 0.17080007520467635,
|
| 877 |
"step": 3000
|
| 878 |
},
|
| 879 |
{
|
| 880 |
+
"epoch": 8.951816160118607,
|
| 881 |
+
"grad_norm": 0.13698986172676086,
|
| 882 |
+
"learning_rate": 4.395555555555556e-06,
|
| 883 |
+
"loss": 0.0043,
|
| 884 |
+
"step": 3025
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 9.023721275018532,
|
| 888 |
+
"grad_norm": 0.15848596394062042,
|
| 889 |
+
"learning_rate": 4.34e-06,
|
| 890 |
+
"loss": 0.0044,
|
| 891 |
+
"step": 3050
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 9.097850259451446,
|
| 895 |
+
"grad_norm": 0.12134752422571182,
|
| 896 |
+
"learning_rate": 4.284444444444445e-06,
|
| 897 |
+
"loss": 0.0031,
|
| 898 |
+
"step": 3075
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 9.171979243884358,
|
| 902 |
+
"grad_norm": 0.13019242882728577,
|
| 903 |
+
"learning_rate": 4.228888888888889e-06,
|
| 904 |
+
"loss": 0.0027,
|
| 905 |
+
"step": 3100
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 9.246108228317272,
|
| 909 |
+
"grad_norm": 0.19179771840572357,
|
| 910 |
+
"learning_rate": 4.173333333333334e-06,
|
| 911 |
+
"loss": 0.0028,
|
| 912 |
+
"step": 3125
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 9.320237212750186,
|
| 916 |
+
"grad_norm": 0.20711077749729156,
|
| 917 |
+
"learning_rate": 4.117777777777779e-06,
|
| 918 |
+
"loss": 0.0027,
|
| 919 |
+
"step": 3150
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 9.394366197183098,
|
| 923 |
+
"grad_norm": 0.09309308230876923,
|
| 924 |
+
"learning_rate": 4.062222222222223e-06,
|
| 925 |
+
"loss": 0.0026,
|
| 926 |
+
"step": 3175
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 9.468495181616012,
|
| 930 |
+
"grad_norm": 0.09770819544792175,
|
| 931 |
+
"learning_rate": 4.006666666666667e-06,
|
| 932 |
+
"loss": 0.0035,
|
| 933 |
+
"step": 3200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 9.542624166048926,
|
| 937 |
+
"grad_norm": 0.1151503324508667,
|
| 938 |
+
"learning_rate": 3.951111111111112e-06,
|
| 939 |
+
"loss": 0.0027,
|
| 940 |
+
"step": 3225
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 9.616753150481838,
|
| 944 |
+
"grad_norm": 0.1251019835472107,
|
| 945 |
+
"learning_rate": 3.895555555555556e-06,
|
| 946 |
+
"loss": 0.0029,
|
| 947 |
+
"step": 3250
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 9.690882134914752,
|
| 951 |
+
"grad_norm": 0.10979008674621582,
|
| 952 |
+
"learning_rate": 3.8400000000000005e-06,
|
| 953 |
+
"loss": 0.0027,
|
| 954 |
+
"step": 3275
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 9.765011119347665,
|
| 958 |
+
"grad_norm": 0.12528590857982635,
|
| 959 |
+
"learning_rate": 3.784444444444445e-06,
|
| 960 |
+
"loss": 0.0037,
|
| 961 |
+
"step": 3300
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 9.839140103780577,
|
| 965 |
+
"grad_norm": 0.12703803181648254,
|
| 966 |
+
"learning_rate": 3.728888888888889e-06,
|
| 967 |
+
"loss": 0.0026,
|
| 968 |
+
"step": 3325
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 9.913269088213491,
|
| 972 |
+
"grad_norm": 0.10044371336698532,
|
| 973 |
+
"learning_rate": 3.673333333333334e-06,
|
| 974 |
+
"loss": 0.003,
|
| 975 |
+
"step": 3350
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 9.987398072646405,
|
| 979 |
+
"grad_norm": 0.3393096625804901,
|
| 980 |
+
"learning_rate": 3.617777777777778e-06,
|
| 981 |
+
"loss": 0.0032,
|
| 982 |
+
"step": 3375
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 10.05930318754633,
|
| 986 |
+
"grad_norm": 0.08865969628095627,
|
| 987 |
+
"learning_rate": 3.5622222222222224e-06,
|
| 988 |
+
"loss": 0.0034,
|
| 989 |
+
"step": 3400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 10.133432171979244,
|
| 993 |
+
"grad_norm": 0.07572653144598007,
|
| 994 |
+
"learning_rate": 3.5066666666666673e-06,
|
| 995 |
+
"loss": 0.0023,
|
| 996 |
+
"step": 3425
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 10.207561156412158,
|
| 1000 |
+
"grad_norm": 0.08611512184143066,
|
| 1001 |
+
"learning_rate": 3.4511111111111113e-06,
|
| 1002 |
+
"loss": 0.0021,
|
| 1003 |
+
"step": 3450
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 10.28169014084507,
|
| 1007 |
+
"grad_norm": 0.09255647659301758,
|
| 1008 |
+
"learning_rate": 3.3955555555555558e-06,
|
| 1009 |
+
"loss": 0.0021,
|
| 1010 |
+
"step": 3475
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 10.355819125277984,
|
| 1014 |
+
"grad_norm": 0.09140720963478088,
|
| 1015 |
+
"learning_rate": 3.3400000000000006e-06,
|
| 1016 |
+
"loss": 0.0021,
|
| 1017 |
+
"step": 3500
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 10.429948109710898,
|
| 1021 |
+
"grad_norm": 0.08117499947547913,
|
| 1022 |
+
"learning_rate": 3.2844444444444447e-06,
|
| 1023 |
+
"loss": 0.002,
|
| 1024 |
+
"step": 3525
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 10.50407709414381,
|
| 1028 |
+
"grad_norm": 0.07924593240022659,
|
| 1029 |
+
"learning_rate": 3.228888888888889e-06,
|
| 1030 |
+
"loss": 0.0021,
|
| 1031 |
+
"step": 3550
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 10.578206078576724,
|
| 1035 |
+
"grad_norm": 0.09375474601984024,
|
| 1036 |
+
"learning_rate": 3.173333333333334e-06,
|
| 1037 |
+
"loss": 0.0023,
|
| 1038 |
+
"step": 3575
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 10.652335063009637,
|
| 1042 |
+
"grad_norm": 0.08366698771715164,
|
| 1043 |
+
"learning_rate": 3.117777777777778e-06,
|
| 1044 |
+
"loss": 0.0022,
|
| 1045 |
+
"step": 3600
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 10.72646404744255,
|
| 1049 |
+
"grad_norm": 0.09042704850435257,
|
| 1050 |
+
"learning_rate": 3.0622222222222225e-06,
|
| 1051 |
+
"loss": 0.0023,
|
| 1052 |
+
"step": 3625
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 10.800593031875463,
|
| 1056 |
+
"grad_norm": 0.07148691266775131,
|
| 1057 |
+
"learning_rate": 3.0066666666666674e-06,
|
| 1058 |
+
"loss": 0.002,
|
| 1059 |
+
"step": 3650
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 10.874722016308377,
|
| 1063 |
+
"grad_norm": 0.3031787574291229,
|
| 1064 |
+
"learning_rate": 2.9511111111111114e-06,
|
| 1065 |
+
"loss": 0.0022,
|
| 1066 |
+
"step": 3675
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 10.94885100074129,
|
| 1070 |
+
"grad_norm": 0.09258675575256348,
|
| 1071 |
+
"learning_rate": 2.895555555555556e-06,
|
| 1072 |
+
"loss": 0.002,
|
| 1073 |
+
"step": 3700
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 11.020756115641216,
|
| 1077 |
+
"grad_norm": 0.06253425031900406,
|
| 1078 |
+
"learning_rate": 2.84e-06,
|
| 1079 |
+
"loss": 0.002,
|
| 1080 |
+
"step": 3725
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 11.094885100074128,
|
| 1084 |
+
"grad_norm": 0.057923316955566406,
|
| 1085 |
+
"learning_rate": 2.784444444444445e-06,
|
| 1086 |
+
"loss": 0.0017,
|
| 1087 |
+
"step": 3750
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 11.169014084507042,
|
| 1091 |
+
"grad_norm": 0.06114868074655533,
|
| 1092 |
+
"learning_rate": 2.7288888888888893e-06,
|
| 1093 |
+
"loss": 0.0018,
|
| 1094 |
+
"step": 3775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 11.243143068939956,
|
| 1098 |
+
"grad_norm": 0.061181288212537766,
|
| 1099 |
+
"learning_rate": 2.6733333333333333e-06,
|
| 1100 |
+
"loss": 0.0017,
|
| 1101 |
+
"step": 3800
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 11.317272053372868,
|
| 1105 |
+
"grad_norm": 0.06324835866689682,
|
| 1106 |
+
"learning_rate": 2.617777777777778e-06,
|
| 1107 |
+
"loss": 0.0018,
|
| 1108 |
+
"step": 3825
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 11.391401037805782,
|
| 1112 |
+
"grad_norm": 0.07077573984861374,
|
| 1113 |
+
"learning_rate": 2.5622222222222226e-06,
|
| 1114 |
+
"loss": 0.0019,
|
| 1115 |
+
"step": 3850
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 11.465530022238696,
|
| 1119 |
+
"grad_norm": 0.065274178981781,
|
| 1120 |
+
"learning_rate": 2.5066666666666667e-06,
|
| 1121 |
+
"loss": 0.0019,
|
| 1122 |
+
"step": 3875
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 11.53965900667161,
|
| 1126 |
+
"grad_norm": 0.08122989535331726,
|
| 1127 |
+
"learning_rate": 2.451111111111111e-06,
|
| 1128 |
+
"loss": 0.0019,
|
| 1129 |
+
"step": 3900
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 11.613787991104521,
|
| 1133 |
+
"grad_norm": 0.06203175708651543,
|
| 1134 |
+
"learning_rate": 2.3955555555555556e-06,
|
| 1135 |
+
"loss": 0.0018,
|
| 1136 |
+
"step": 3925
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 11.687916975537435,
|
| 1140 |
+
"grad_norm": 0.08688808977603912,
|
| 1141 |
+
"learning_rate": 2.3400000000000005e-06,
|
| 1142 |
+
"loss": 0.0018,
|
| 1143 |
+
"step": 3950
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 11.762045959970349,
|
| 1147 |
+
"grad_norm": 0.06490565836429596,
|
| 1148 |
+
"learning_rate": 2.2844444444444445e-06,
|
| 1149 |
+
"loss": 0.0017,
|
| 1150 |
+
"step": 3975
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 11.836174944403261,
|
| 1154 |
+
"grad_norm": 0.06775514036417007,
|
| 1155 |
+
"learning_rate": 2.228888888888889e-06,
|
| 1156 |
+
"loss": 0.0018,
|
| 1157 |
+
"step": 4000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 11.836174944403261,
|
| 1161 |
+
"eval_loss": 0.39641496539115906,
|
| 1162 |
+
"eval_runtime": 944.1469,
|
| 1163 |
+
"eval_samples_per_second": 3.056,
|
| 1164 |
+
"eval_steps_per_second": 0.382,
|
| 1165 |
+
"eval_wer": 0.16803117575674706,
|
| 1166 |
+
"step": 4000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 11.910303928836175,
|
| 1170 |
+
"grad_norm": 0.06309942156076431,
|
| 1171 |
+
"learning_rate": 2.1733333333333334e-06,
|
| 1172 |
+
"loss": 0.0017,
|
| 1173 |
+
"step": 4025
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 11.984432913269089,
|
| 1177 |
+
"grad_norm": 0.061430562287569046,
|
| 1178 |
+
"learning_rate": 2.117777777777778e-06,
|
| 1179 |
+
"loss": 0.0017,
|
| 1180 |
+
"step": 4050
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 12.056338028169014,
|
| 1184 |
+
"grad_norm": 0.05334235355257988,
|
| 1185 |
+
"learning_rate": 2.0622222222222223e-06,
|
| 1186 |
+
"loss": 0.0016,
|
| 1187 |
+
"step": 4075
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 12.130467012601928,
|
| 1191 |
+
"grad_norm": 0.057459134608507156,
|
| 1192 |
+
"learning_rate": 2.006666666666667e-06,
|
| 1193 |
+
"loss": 0.0016,
|
| 1194 |
+
"step": 4100
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 12.20459599703484,
|
| 1198 |
+
"grad_norm": 0.047365590929985046,
|
| 1199 |
+
"learning_rate": 1.9511111111111113e-06,
|
| 1200 |
+
"loss": 0.0017,
|
| 1201 |
+
"step": 4125
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 12.278724981467754,
|
| 1205 |
+
"grad_norm": 0.056059811264276505,
|
| 1206 |
+
"learning_rate": 1.8955555555555557e-06,
|
| 1207 |
+
"loss": 0.0015,
|
| 1208 |
+
"step": 4150
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 12.352853965900668,
|
| 1212 |
+
"grad_norm": 0.062216296792030334,
|
| 1213 |
+
"learning_rate": 1.8400000000000002e-06,
|
| 1214 |
+
"loss": 0.0016,
|
| 1215 |
+
"step": 4175
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 12.42698295033358,
|
| 1219 |
+
"grad_norm": 0.055066876113414764,
|
| 1220 |
+
"learning_rate": 1.7844444444444444e-06,
|
| 1221 |
+
"loss": 0.0015,
|
| 1222 |
+
"step": 4200
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 12.501111934766493,
|
| 1226 |
+
"grad_norm": 0.06531412899494171,
|
| 1227 |
+
"learning_rate": 1.728888888888889e-06,
|
| 1228 |
+
"loss": 0.0015,
|
| 1229 |
+
"step": 4225
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 12.575240919199407,
|
| 1233 |
+
"grad_norm": 0.04836783558130264,
|
| 1234 |
+
"learning_rate": 1.6733333333333335e-06,
|
| 1235 |
+
"loss": 0.0016,
|
| 1236 |
+
"step": 4250
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 12.649369903632321,
|
| 1240 |
+
"grad_norm": 0.050022318959236145,
|
| 1241 |
+
"learning_rate": 1.6177777777777778e-06,
|
| 1242 |
+
"loss": 0.0016,
|
| 1243 |
+
"step": 4275
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 12.723498888065233,
|
| 1247 |
+
"grad_norm": 0.05791894719004631,
|
| 1248 |
+
"learning_rate": 1.5622222222222225e-06,
|
| 1249 |
+
"loss": 0.0015,
|
| 1250 |
+
"step": 4300
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 12.797627872498147,
|
| 1254 |
+
"grad_norm": 0.050857629626989365,
|
| 1255 |
+
"learning_rate": 1.506666666666667e-06,
|
| 1256 |
+
"loss": 0.0016,
|
| 1257 |
+
"step": 4325
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 12.87175685693106,
|
| 1261 |
+
"grad_norm": 0.05395572632551193,
|
| 1262 |
+
"learning_rate": 1.4511111111111112e-06,
|
| 1263 |
+
"loss": 0.0016,
|
| 1264 |
+
"step": 4350
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 12.945885841363973,
|
| 1268 |
+
"grad_norm": 0.06613746285438538,
|
| 1269 |
+
"learning_rate": 1.3955555555555556e-06,
|
| 1270 |
+
"loss": 0.0015,
|
| 1271 |
+
"step": 4375
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 13.0177909562639,
|
| 1275 |
+
"grad_norm": 0.048592809587717056,
|
| 1276 |
+
"learning_rate": 1.34e-06,
|
| 1277 |
+
"loss": 0.0014,
|
| 1278 |
+
"step": 4400
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 13.091919940696812,
|
| 1282 |
+
"grad_norm": 0.055479902774095535,
|
| 1283 |
+
"learning_rate": 1.2844444444444445e-06,
|
| 1284 |
+
"loss": 0.0014,
|
| 1285 |
+
"step": 4425
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 13.166048925129726,
|
| 1289 |
+
"grad_norm": 0.045032110065221786,
|
| 1290 |
+
"learning_rate": 1.228888888888889e-06,
|
| 1291 |
+
"loss": 0.0014,
|
| 1292 |
+
"step": 4450
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 13.24017790956264,
|
| 1296 |
+
"grad_norm": 0.06555484235286713,
|
| 1297 |
+
"learning_rate": 1.1733333333333335e-06,
|
| 1298 |
+
"loss": 0.0014,
|
| 1299 |
+
"step": 4475
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 13.314306893995552,
|
| 1303 |
+
"grad_norm": 0.048962488770484924,
|
| 1304 |
+
"learning_rate": 1.117777777777778e-06,
|
| 1305 |
+
"loss": 0.0014,
|
| 1306 |
+
"step": 4500
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 13.388435878428465,
|
| 1310 |
+
"grad_norm": 0.047078557312488556,
|
| 1311 |
+
"learning_rate": 1.0622222222222222e-06,
|
| 1312 |
+
"loss": 0.0014,
|
| 1313 |
+
"step": 4525
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 13.46256486286138,
|
| 1317 |
+
"grad_norm": 0.050719983875751495,
|
| 1318 |
+
"learning_rate": 1.0066666666666668e-06,
|
| 1319 |
+
"loss": 0.0014,
|
| 1320 |
+
"step": 4550
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 13.536693847294291,
|
| 1324 |
+
"grad_norm": 0.060897134244441986,
|
| 1325 |
+
"learning_rate": 9.511111111111111e-07,
|
| 1326 |
+
"loss": 0.0014,
|
| 1327 |
+
"step": 4575
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 13.610822831727205,
|
| 1331 |
+
"grad_norm": 0.0608171783387661,
|
| 1332 |
+
"learning_rate": 8.955555555555557e-07,
|
| 1333 |
+
"loss": 0.0016,
|
| 1334 |
+
"step": 4600
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 13.684951816160119,
|
| 1338 |
+
"grad_norm": 0.050377678126096725,
|
| 1339 |
+
"learning_rate": 8.400000000000001e-07,
|
| 1340 |
+
"loss": 0.0014,
|
| 1341 |
+
"step": 4625
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 13.759080800593033,
|
| 1345 |
+
"grad_norm": 0.04982956126332283,
|
| 1346 |
+
"learning_rate": 7.844444444444445e-07,
|
| 1347 |
+
"loss": 0.0013,
|
| 1348 |
+
"step": 4650
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 13.833209785025945,
|
| 1352 |
+
"grad_norm": 0.049559298902750015,
|
| 1353 |
+
"learning_rate": 7.28888888888889e-07,
|
| 1354 |
+
"loss": 0.0014,
|
| 1355 |
+
"step": 4675
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 13.907338769458859,
|
| 1359 |
+
"grad_norm": 0.045101020485162735,
|
| 1360 |
+
"learning_rate": 6.733333333333334e-07,
|
| 1361 |
+
"loss": 0.0014,
|
| 1362 |
+
"step": 4700
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 13.981467753891772,
|
| 1366 |
+
"grad_norm": 0.04930800572037697,
|
| 1367 |
+
"learning_rate": 6.177777777777778e-07,
|
| 1368 |
+
"loss": 0.0014,
|
| 1369 |
+
"step": 4725
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 14.053372868791698,
|
| 1373 |
+
"grad_norm": 0.046362608671188354,
|
| 1374 |
+
"learning_rate": 5.622222222222223e-07,
|
| 1375 |
+
"loss": 0.0013,
|
| 1376 |
+
"step": 4750
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 14.127501853224611,
|
| 1380 |
+
"grad_norm": 0.05173536762595177,
|
| 1381 |
+
"learning_rate": 5.066666666666667e-07,
|
| 1382 |
+
"loss": 0.0015,
|
| 1383 |
+
"step": 4775
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 14.201630837657524,
|
| 1387 |
+
"grad_norm": 0.045450981706380844,
|
| 1388 |
+
"learning_rate": 4.511111111111111e-07,
|
| 1389 |
+
"loss": 0.0013,
|
| 1390 |
+
"step": 4800
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 14.275759822090437,
|
| 1394 |
+
"grad_norm": 0.04437384009361267,
|
| 1395 |
+
"learning_rate": 3.9555555555555557e-07,
|
| 1396 |
+
"loss": 0.0013,
|
| 1397 |
+
"step": 4825
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 14.349888806523351,
|
| 1401 |
+
"grad_norm": 0.048076264560222626,
|
| 1402 |
+
"learning_rate": 3.4000000000000003e-07,
|
| 1403 |
+
"loss": 0.0013,
|
| 1404 |
+
"step": 4850
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 14.424017790956263,
|
| 1408 |
+
"grad_norm": 0.047441426664590836,
|
| 1409 |
+
"learning_rate": 2.844444444444445e-07,
|
| 1410 |
+
"loss": 0.0014,
|
| 1411 |
+
"step": 4875
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 14.498146775389177,
|
| 1415 |
+
"grad_norm": 0.04609740898013115,
|
| 1416 |
+
"learning_rate": 2.2888888888888892e-07,
|
| 1417 |
+
"loss": 0.0013,
|
| 1418 |
+
"step": 4900
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 14.572275759822091,
|
| 1422 |
+
"grad_norm": 0.05346500128507614,
|
| 1423 |
+
"learning_rate": 1.7333333333333335e-07,
|
| 1424 |
+
"loss": 0.0013,
|
| 1425 |
+
"step": 4925
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 14.646404744255003,
|
| 1429 |
+
"grad_norm": 0.04578370600938797,
|
| 1430 |
+
"learning_rate": 1.1777777777777778e-07,
|
| 1431 |
+
"loss": 0.0013,
|
| 1432 |
+
"step": 4950
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 14.720533728687917,
|
| 1436 |
+
"grad_norm": 0.045380473136901855,
|
| 1437 |
+
"learning_rate": 6.222222222222223e-08,
|
| 1438 |
+
"loss": 0.0014,
|
| 1439 |
+
"step": 4975
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 14.79466271312083,
|
| 1443 |
+
"grad_norm": 0.04897478222846985,
|
| 1444 |
+
"learning_rate": 6.666666666666667e-09,
|
| 1445 |
+
"loss": 0.0013,
|
| 1446 |
+
"step": 5000
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 14.79466271312083,
|
| 1450 |
+
"eval_loss": 0.4044080376625061,
|
| 1451 |
+
"eval_runtime": 939.8085,
|
| 1452 |
+
"eval_samples_per_second": 3.07,
|
| 1453 |
+
"eval_steps_per_second": 0.384,
|
| 1454 |
+
"eval_wer": 0.1675184165997231,
|
| 1455 |
+
"step": 5000
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 14.79466271312083,
|
| 1459 |
+
"step": 5000,
|
| 1460 |
+
"total_flos": 4.607669973179538e+19,
|
| 1461 |
+
"train_loss": 0.11535389684215187,
|
| 1462 |
+
"train_runtime": 25285.0125,
|
| 1463 |
+
"train_samples_per_second": 6.328,
|
| 1464 |
+
"train_steps_per_second": 0.198
|
| 1465 |
}
|
| 1466 |
],
|
| 1467 |
"logging_steps": 25,
|
| 1468 |
+
"max_steps": 5000,
|
| 1469 |
"num_input_tokens_seen": 0,
|
| 1470 |
+
"num_train_epochs": 15,
|
| 1471 |
"save_steps": 1000,
|
| 1472 |
"stateful_callbacks": {
|
| 1473 |
"TrainerControl": {
|
|
|
|
| 1481 |
"attributes": {}
|
| 1482 |
}
|
| 1483 |
},
|
| 1484 |
+
"total_flos": 4.607669973179538e+19,
|
| 1485 |
"train_batch_size": 4,
|
| 1486 |
"trial_name": null,
|
| 1487 |
"trial_params": null
|