End of training
Browse files- README.md +15 -2
- all_results.json +13 -13
- eval_results.json +7 -7
- runs/Aug09_05-15-33_30e2a19b1988/events.out.tfevents.1754729893.30e2a19b1988.3712.1 +3 -0
- train_results.json +7 -7
- trainer_state.json +1212 -345
README.md
CHANGED
|
@@ -4,11 +4,24 @@ license: apache-2.0
|
|
| 4 |
base_model: openai/whisper-medium
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
model-index:
|
| 10 |
- name: whisper-medium-ph
|
| 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-ph
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.2901
|
| 22 |
- Wer: 0.1147
|
|
|
|
| 4 |
base_model: openai/whisper-medium
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
datasets:
|
| 8 |
+
- rbcurzon/ph_dialect_asr
|
| 9 |
metrics:
|
| 10 |
- wer
|
| 11 |
model-index:
|
| 12 |
- name: whisper-medium-ph
|
| 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: 0.1146545827633379
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 29 |
|
| 30 |
# whisper-medium-ph
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the rbcurzon/ph_dialect_asr all dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
- Loss: 0.2901
|
| 35 |
- Wer: 0.1147
|
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": 3.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
-
"total_flos":
|
| 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": 7.407709414381023,
|
| 3 |
+
"eval_loss": 0.29011788964271545,
|
| 4 |
+
"eval_runtime": 736.3871,
|
| 5 |
+
"eval_samples": 2885,
|
| 6 |
+
"eval_samples_per_second": 3.918,
|
| 7 |
+
"eval_steps_per_second": 0.49,
|
| 8 |
+
"eval_wer": 0.1146545827633379,
|
| 9 |
+
"total_flos": 8.155551755501568e+19,
|
| 10 |
+
"train_loss": 0.10907779041565954,
|
| 11 |
+
"train_runtime": 12394.4337,
|
| 12 |
+
"train_samples": 10787,
|
| 13 |
+
"train_samples_per_second": 6.455,
|
| 14 |
+
"train_steps_per_second": 0.403
|
| 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": 3.
|
| 7 |
-
"eval_steps_per_second": 0.
|
| 8 |
-
"eval_wer": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 7.407709414381023,
|
| 3 |
+
"eval_loss": 0.29011788964271545,
|
| 4 |
+
"eval_runtime": 736.3871,
|
| 5 |
+
"eval_samples": 2885,
|
| 6 |
+
"eval_samples_per_second": 3.918,
|
| 7 |
+
"eval_steps_per_second": 0.49,
|
| 8 |
+
"eval_wer": 0.1146545827633379
|
| 9 |
}
|
runs/Aug09_05-15-33_30e2a19b1988/events.out.tfevents.1754729893.30e2a19b1988.3712.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4fc589040fcd11263525855db0ac5e3ae941083ed659ce852614b4c40555680d
|
| 3 |
+
size 406
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 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": 7.407709414381023,
|
| 3 |
+
"total_flos": 8.155551755501568e+19,
|
| 4 |
+
"train_loss": 0.10907779041565954,
|
| 5 |
+
"train_runtime": 12394.4337,
|
| 6 |
+
"train_samples": 10787,
|
| 7 |
+
"train_samples_per_second": 6.455,
|
| 8 |
+
"train_steps_per_second": 0.403
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -2,605 +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": 7.
|
| 29 |
-
"learning_rate": 1.
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate": 1.
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 2.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate": 2.
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 3.
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate": 3.
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 4.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 4.
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm": 6.
|
| 85 |
-
"learning_rate": 5.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm":
|
| 92 |
-
"learning_rate": 5.
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm":
|
| 99 |
-
"learning_rate": 6.
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 0.
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 6.
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 0.
|
| 112 |
-
"grad_norm":
|
| 113 |
-
"learning_rate": 7.
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 0.
|
| 119 |
-
"grad_norm": 5.
|
| 120 |
-
"learning_rate": 7.
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate": 8.
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 8.
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 0.
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate": 9.
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 0.
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate": 9.
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch": 0.
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate": 9.
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 525
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch": 0.
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate": 9.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 550
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"grad_norm": 5.
|
| 169 |
-
"learning_rate": 9.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 575
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate": 9.
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 600
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 0.
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate": 9.
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 625
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 0.
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate": 9.
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 650
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch":
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 675
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"grad_norm": 4.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 700
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 725
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 750
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch":
|
| 224 |
-
"grad_norm": 3.
|
| 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": 1.
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 825
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch": 1.
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 850
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 1.
|
| 252 |
-
"grad_norm": 3.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 875
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 1.
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 900
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch": 1.
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 925
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 1.
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 950
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch": 1.
|
| 280 |
-
"grad_norm": 4.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 975
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch": 1.
|
| 287 |
-
"grad_norm": 3.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 1000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 1.
|
| 294 |
-
"eval_loss": 0.
|
| 295 |
-
"eval_runtime":
|
| 296 |
-
"eval_samples_per_second": 3.
|
| 297 |
-
"eval_steps_per_second": 0.
|
| 298 |
-
"eval_wer": 0.
|
| 299 |
"step": 1000
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch": 1.
|
| 303 |
-
"grad_norm":
|
| 304 |
-
"learning_rate":
|
| 305 |
-
"loss": 0.
|
| 306 |
"step": 1025
|
| 307 |
},
|
| 308 |
{
|
| 309 |
-
"epoch": 1.
|
| 310 |
-
"grad_norm": 3.
|
| 311 |
-
"learning_rate":
|
| 312 |
-
"loss": 0.
|
| 313 |
"step": 1050
|
| 314 |
},
|
| 315 |
{
|
| 316 |
-
"epoch": 1.
|
| 317 |
-
"grad_norm":
|
| 318 |
-
"learning_rate":
|
| 319 |
-
"loss": 0.
|
| 320 |
"step": 1075
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch": 1.
|
| 324 |
-
"grad_norm":
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss": 0.
|
| 327 |
"step": 1100
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 1.
|
| 331 |
-
"grad_norm":
|
| 332 |
-
"learning_rate":
|
| 333 |
-
"loss": 0.
|
| 334 |
"step": 1125
|
| 335 |
},
|
| 336 |
{
|
| 337 |
-
"epoch": 1.
|
| 338 |
-
"grad_norm":
|
| 339 |
-
"learning_rate":
|
| 340 |
-
"loss": 0.
|
| 341 |
"step": 1150
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch": 1.
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate":
|
| 347 |
-
"loss": 0.
|
| 348 |
"step": 1175
|
| 349 |
},
|
| 350 |
{
|
| 351 |
-
"epoch": 1.
|
| 352 |
-
"grad_norm":
|
| 353 |
-
"learning_rate":
|
| 354 |
-
"loss": 0.
|
| 355 |
"step": 1200
|
| 356 |
},
|
| 357 |
{
|
| 358 |
-
"epoch": 1.
|
| 359 |
-
"grad_norm": 3.
|
| 360 |
-
"learning_rate":
|
| 361 |
-
"loss": 0.
|
| 362 |
"step": 1225
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch": 1.
|
| 366 |
-
"grad_norm":
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss": 0.
|
| 369 |
"step": 1250
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 1.
|
| 373 |
-
"grad_norm":
|
| 374 |
-
"learning_rate":
|
| 375 |
-
"loss": 0.
|
| 376 |
"step": 1275
|
| 377 |
},
|
| 378 |
{
|
| 379 |
-
"epoch": 1.
|
| 380 |
-
"grad_norm":
|
| 381 |
-
"learning_rate":
|
| 382 |
-
"loss": 0.
|
| 383 |
"step": 1300
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch": 1.
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate":
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 1325
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch":
|
| 394 |
-
"grad_norm":
|
| 395 |
-
"learning_rate":
|
| 396 |
-
"loss": 0.
|
| 397 |
"step": 1350
|
| 398 |
},
|
| 399 |
{
|
| 400 |
-
"epoch":
|
| 401 |
-
"grad_norm": 2.
|
| 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": 2.
|
| 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":
|
| 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": 2.
|
| 458 |
-
"learning_rate":
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 1575
|
| 461 |
},
|
| 462 |
{
|
| 463 |
-
"epoch":
|
| 464 |
-
"grad_norm": 2.
|
| 465 |
-
"learning_rate":
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 1600
|
| 468 |
},
|
| 469 |
{
|
| 470 |
-
"epoch": 2.
|
| 471 |
-
"grad_norm":
|
| 472 |
-
"learning_rate":
|
| 473 |
-
"loss": 0.
|
| 474 |
"step": 1625
|
| 475 |
},
|
| 476 |
{
|
| 477 |
-
"epoch": 2.
|
| 478 |
-
"grad_norm":
|
| 479 |
-
"learning_rate":
|
| 480 |
-
"loss": 0.
|
| 481 |
"step": 1650
|
| 482 |
},
|
| 483 |
{
|
| 484 |
-
"epoch": 2.
|
| 485 |
-
"grad_norm":
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss": 0.
|
| 488 |
"step": 1675
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch": 2.
|
| 492 |
-
"grad_norm":
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 1700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 2.
|
| 499 |
-
"grad_norm":
|
| 500 |
-
"learning_rate":
|
| 501 |
-
"loss": 0.
|
| 502 |
"step": 1725
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"epoch": 2.
|
| 506 |
-
"grad_norm":
|
| 507 |
-
"learning_rate":
|
| 508 |
-
"loss": 0.
|
| 509 |
"step": 1750
|
| 510 |
},
|
| 511 |
{
|
| 512 |
-
"epoch": 2.
|
| 513 |
-
"grad_norm":
|
| 514 |
-
"learning_rate":
|
| 515 |
-
"loss": 0.
|
| 516 |
"step": 1775
|
| 517 |
},
|
| 518 |
{
|
| 519 |
-
"epoch": 2.
|
| 520 |
-
"grad_norm":
|
| 521 |
-
"learning_rate":
|
| 522 |
-
"loss": 0.
|
| 523 |
"step": 1800
|
| 524 |
},
|
| 525 |
{
|
| 526 |
-
"epoch": 2.
|
| 527 |
-
"grad_norm":
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 1825
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch": 2.
|
| 534 |
-
"grad_norm":
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 1850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch": 2.
|
| 541 |
-
"grad_norm":
|
| 542 |
-
"learning_rate":
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 1875
|
| 545 |
},
|
| 546 |
{
|
| 547 |
-
"epoch": 2.
|
| 548 |
-
"grad_norm": 1.
|
| 549 |
-
"learning_rate": 6.
|
| 550 |
-
"loss": 0.
|
| 551 |
"step": 1900
|
| 552 |
},
|
| 553 |
{
|
| 554 |
-
"epoch": 2.
|
| 555 |
-
"grad_norm":
|
| 556 |
-
"learning_rate":
|
| 557 |
-
"loss": 0.
|
| 558 |
"step": 1925
|
| 559 |
},
|
| 560 |
{
|
| 561 |
-
"epoch": 2.
|
| 562 |
-
"grad_norm":
|
| 563 |
-
"learning_rate":
|
| 564 |
-
"loss": 0.
|
| 565 |
"step": 1950
|
| 566 |
},
|
| 567 |
{
|
| 568 |
-
"epoch": 2.
|
| 569 |
-
"grad_norm":
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss": 0.
|
| 572 |
"step": 1975
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch": 2.
|
| 576 |
-
"grad_norm":
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 2000
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch": 2.
|
| 583 |
-
"eval_loss": 0.
|
| 584 |
-
"eval_runtime":
|
| 585 |
-
"eval_samples_per_second": 3.
|
| 586 |
-
"eval_steps_per_second": 0.
|
| 587 |
-
"eval_wer": 0.
|
| 588 |
"step": 2000
|
| 589 |
},
|
| 590 |
{
|
| 591 |
-
"epoch":
|
| 592 |
-
"
|
| 593 |
-
"
|
| 594 |
-
"
|
| 595 |
-
"
|
| 596 |
-
|
| 597 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 598 |
}
|
| 599 |
],
|
| 600 |
"logging_steps": 25,
|
| 601 |
-
"max_steps":
|
| 602 |
"num_input_tokens_seen": 0,
|
| 603 |
-
"num_train_epochs":
|
| 604 |
"save_steps": 1000,
|
| 605 |
"stateful_callbacks": {
|
| 606 |
"TrainerControl": {
|
|
@@ -614,8 +1481,8 @@
|
|
| 614 |
"attributes": {}
|
| 615 |
}
|
| 616 |
},
|
| 617 |
-
"total_flos":
|
| 618 |
-
"train_batch_size":
|
| 619 |
"trial_name": null,
|
| 620 |
"trial_params": null
|
| 621 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 7.407709414381023,
|
| 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": 0.037064492216456635,
|
| 14 |
+
"grad_norm": 14.615763664245605,
|
| 15 |
+
"learning_rate": 4.800000000000001e-07,
|
| 16 |
+
"loss": 1.4607,
|
| 17 |
"step": 25
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.07412898443291327,
|
| 21 |
+
"grad_norm": 10.59756851196289,
|
| 22 |
+
"learning_rate": 9.800000000000001e-07,
|
| 23 |
+
"loss": 1.176,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.1111934766493699,
|
| 28 |
+
"grad_norm": 7.142136573791504,
|
| 29 |
+
"learning_rate": 1.48e-06,
|
| 30 |
+
"loss": 0.9267,
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.14825796886582654,
|
| 35 |
+
"grad_norm": 9.191902160644531,
|
| 36 |
+
"learning_rate": 1.98e-06,
|
| 37 |
+
"loss": 0.7253,
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.18532246108228317,
|
| 42 |
+
"grad_norm": 10.320201873779297,
|
| 43 |
+
"learning_rate": 2.4800000000000004e-06,
|
| 44 |
+
"loss": 0.7047,
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.2223869532987398,
|
| 49 |
+
"grad_norm": 8.486912727355957,
|
| 50 |
+
"learning_rate": 2.9800000000000003e-06,
|
| 51 |
+
"loss": 0.6634,
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.25945144551519644,
|
| 56 |
+
"grad_norm": 9.802300453186035,
|
| 57 |
+
"learning_rate": 3.48e-06,
|
| 58 |
+
"loss": 0.5786,
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.2965159377316531,
|
| 63 |
+
"grad_norm": 9.568249702453613,
|
| 64 |
+
"learning_rate": 3.980000000000001e-06,
|
| 65 |
+
"loss": 0.5857,
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.3335804299481097,
|
| 70 |
+
"grad_norm": 7.968526840209961,
|
| 71 |
+
"learning_rate": 4.48e-06,
|
| 72 |
+
"loss": 0.5385,
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.37064492216456635,
|
| 77 |
+
"grad_norm": 7.507795810699463,
|
| 78 |
+
"learning_rate": 4.980000000000001e-06,
|
| 79 |
+
"loss": 0.5151,
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.407709414381023,
|
| 84 |
+
"grad_norm": 6.258375644683838,
|
| 85 |
+
"learning_rate": 5.480000000000001e-06,
|
| 86 |
+
"loss": 0.4649,
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.4447739065974796,
|
| 91 |
+
"grad_norm": 9.89697551727295,
|
| 92 |
+
"learning_rate": 5.98e-06,
|
| 93 |
+
"loss": 0.4209,
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.48183839881393625,
|
| 98 |
+
"grad_norm": 7.8507490158081055,
|
| 99 |
+
"learning_rate": 6.480000000000001e-06,
|
| 100 |
+
"loss": 0.4459,
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 0.5189028910303929,
|
| 105 |
+
"grad_norm": 5.835811138153076,
|
| 106 |
+
"learning_rate": 6.98e-06,
|
| 107 |
+
"loss": 0.4141,
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 0.5559673832468495,
|
| 112 |
+
"grad_norm": 6.767547607421875,
|
| 113 |
+
"learning_rate": 7.48e-06,
|
| 114 |
+
"loss": 0.4108,
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.5930318754633062,
|
| 119 |
+
"grad_norm": 5.9475884437561035,
|
| 120 |
+
"learning_rate": 7.980000000000002e-06,
|
| 121 |
+
"loss": 0.41,
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.6300963676797627,
|
| 126 |
+
"grad_norm": 7.767906188964844,
|
| 127 |
+
"learning_rate": 8.48e-06,
|
| 128 |
+
"loss": 0.3781,
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.6671608598962194,
|
| 133 |
+
"grad_norm": 6.990137100219727,
|
| 134 |
+
"learning_rate": 8.98e-06,
|
| 135 |
+
"loss": 0.39,
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.704225352112676,
|
| 140 |
+
"grad_norm": 5.607441425323486,
|
| 141 |
+
"learning_rate": 9.48e-06,
|
| 142 |
+
"loss": 0.3783,
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 0.7412898443291327,
|
| 147 |
+
"grad_norm": 6.288857936859131,
|
| 148 |
+
"learning_rate": 9.980000000000001e-06,
|
| 149 |
+
"loss": 0.3559,
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 0.7783543365455893,
|
| 154 |
+
"grad_norm": 6.985698699951172,
|
| 155 |
+
"learning_rate": 9.946666666666667e-06,
|
| 156 |
+
"loss": 0.3595,
|
| 157 |
"step": 525
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 0.815418828762046,
|
| 161 |
+
"grad_norm": 6.037854194641113,
|
| 162 |
+
"learning_rate": 9.891111111111113e-06,
|
| 163 |
+
"loss": 0.3163,
|
| 164 |
"step": 550
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.8524833209785025,
|
| 168 |
+
"grad_norm": 5.8710784912109375,
|
| 169 |
+
"learning_rate": 9.835555555555556e-06,
|
| 170 |
+
"loss": 0.3502,
|
| 171 |
"step": 575
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.8895478131949592,
|
| 175 |
+
"grad_norm": 6.342834949493408,
|
| 176 |
+
"learning_rate": 9.780000000000001e-06,
|
| 177 |
+
"loss": 0.317,
|
| 178 |
"step": 600
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 0.9266123054114158,
|
| 182 |
+
"grad_norm": 5.589534759521484,
|
| 183 |
+
"learning_rate": 9.724444444444445e-06,
|
| 184 |
+
"loss": 0.3228,
|
| 185 |
"step": 625
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 0.9636767976278725,
|
| 189 |
+
"grad_norm": 7.743918418884277,
|
| 190 |
+
"learning_rate": 9.66888888888889e-06,
|
| 191 |
+
"loss": 0.3144,
|
| 192 |
"step": 650
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 1.0,
|
| 196 |
+
"grad_norm": 10.073568344116211,
|
| 197 |
+
"learning_rate": 9.613333333333335e-06,
|
| 198 |
+
"loss": 0.2939,
|
| 199 |
"step": 675
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 1.0370644922164567,
|
| 203 |
+
"grad_norm": 4.640520095825195,
|
| 204 |
+
"learning_rate": 9.557777777777777e-06,
|
| 205 |
+
"loss": 0.1939,
|
| 206 |
"step": 700
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 1.0741289844329134,
|
| 210 |
+
"grad_norm": 3.2049508094787598,
|
| 211 |
+
"learning_rate": 9.502222222222223e-06,
|
| 212 |
+
"loss": 0.1929,
|
| 213 |
"step": 725
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 1.1111934766493698,
|
| 217 |
+
"grad_norm": 3.9065611362457275,
|
| 218 |
+
"learning_rate": 9.446666666666667e-06,
|
| 219 |
+
"loss": 0.1998,
|
| 220 |
"step": 750
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 1.1482579688658265,
|
| 224 |
+
"grad_norm": 3.7471649646759033,
|
| 225 |
+
"learning_rate": 9.391111111111111e-06,
|
| 226 |
+
"loss": 0.2007,
|
| 227 |
"step": 775
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 1.1853224610822832,
|
| 231 |
+
"grad_norm": 3.952751874923706,
|
| 232 |
+
"learning_rate": 9.335555555555557e-06,
|
| 233 |
+
"loss": 0.1863,
|
| 234 |
"step": 800
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 1.2223869532987397,
|
| 238 |
+
"grad_norm": 5.39549446105957,
|
| 239 |
+
"learning_rate": 9.280000000000001e-06,
|
| 240 |
+
"loss": 0.1953,
|
| 241 |
"step": 825
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 1.2594514455151964,
|
| 245 |
+
"grad_norm": 4.03216552734375,
|
| 246 |
+
"learning_rate": 9.224444444444445e-06,
|
| 247 |
+
"loss": 0.2065,
|
| 248 |
"step": 850
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 1.296515937731653,
|
| 252 |
+
"grad_norm": 3.854651689529419,
|
| 253 |
+
"learning_rate": 9.168888888888889e-06,
|
| 254 |
+
"loss": 0.1703,
|
| 255 |
"step": 875
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 1.3335804299481098,
|
| 259 |
+
"grad_norm": 4.835360050201416,
|
| 260 |
+
"learning_rate": 9.113333333333335e-06,
|
| 261 |
+
"loss": 0.1692,
|
| 262 |
"step": 900
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 1.3706449221645665,
|
| 266 |
+
"grad_norm": 5.247130393981934,
|
| 267 |
+
"learning_rate": 9.057777777777779e-06,
|
| 268 |
+
"loss": 0.1982,
|
| 269 |
"step": 925
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 1.407709414381023,
|
| 273 |
+
"grad_norm": 3.9537737369537354,
|
| 274 |
+
"learning_rate": 9.002222222222223e-06,
|
| 275 |
+
"loss": 0.1661,
|
| 276 |
"step": 950
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 1.4447739065974796,
|
| 280 |
+
"grad_norm": 4.887810230255127,
|
| 281 |
+
"learning_rate": 8.946666666666669e-06,
|
| 282 |
+
"loss": 0.1836,
|
| 283 |
"step": 975
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 1.4818383988139363,
|
| 287 |
+
"grad_norm": 3.6338751316070557,
|
| 288 |
+
"learning_rate": 8.891111111111111e-06,
|
| 289 |
+
"loss": 0.1822,
|
| 290 |
"step": 1000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 1.4818383988139363,
|
| 294 |
+
"eval_loss": 0.2655850648880005,
|
| 295 |
+
"eval_runtime": 730.9503,
|
| 296 |
+
"eval_samples_per_second": 3.947,
|
| 297 |
+
"eval_steps_per_second": 0.494,
|
| 298 |
+
"eval_wer": 0.14449384404924762,
|
| 299 |
"step": 1000
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 1.5189028910303928,
|
| 303 |
+
"grad_norm": 4.078255653381348,
|
| 304 |
+
"learning_rate": 8.835555555555557e-06,
|
| 305 |
+
"loss": 0.1661,
|
| 306 |
"step": 1025
|
| 307 |
},
|
| 308 |
{
|
| 309 |
+
"epoch": 1.5559673832468495,
|
| 310 |
+
"grad_norm": 3.9311952590942383,
|
| 311 |
+
"learning_rate": 8.78e-06,
|
| 312 |
+
"loss": 0.1725,
|
| 313 |
"step": 1050
|
| 314 |
},
|
| 315 |
{
|
| 316 |
+
"epoch": 1.5930318754633062,
|
| 317 |
+
"grad_norm": 4.800196170806885,
|
| 318 |
+
"learning_rate": 8.724444444444445e-06,
|
| 319 |
+
"loss": 0.1704,
|
| 320 |
"step": 1075
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 1.6300963676797626,
|
| 324 |
+
"grad_norm": 4.550530910491943,
|
| 325 |
+
"learning_rate": 8.66888888888889e-06,
|
| 326 |
+
"loss": 0.1793,
|
| 327 |
"step": 1100
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 1.6671608598962195,
|
| 331 |
+
"grad_norm": 6.508624076843262,
|
| 332 |
+
"learning_rate": 8.613333333333333e-06,
|
| 333 |
+
"loss": 0.1619,
|
| 334 |
"step": 1125
|
| 335 |
},
|
| 336 |
{
|
| 337 |
+
"epoch": 1.704225352112676,
|
| 338 |
+
"grad_norm": 4.16792106628418,
|
| 339 |
+
"learning_rate": 8.557777777777778e-06,
|
| 340 |
+
"loss": 0.1652,
|
| 341 |
"step": 1150
|
| 342 |
},
|
| 343 |
{
|
| 344 |
+
"epoch": 1.7412898443291327,
|
| 345 |
+
"grad_norm": 4.420657157897949,
|
| 346 |
+
"learning_rate": 8.502222222222223e-06,
|
| 347 |
+
"loss": 0.16,
|
| 348 |
"step": 1175
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 1.7783543365455894,
|
| 352 |
+
"grad_norm": 4.781569004058838,
|
| 353 |
+
"learning_rate": 8.446666666666668e-06,
|
| 354 |
+
"loss": 0.1695,
|
| 355 |
"step": 1200
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 1.8154188287620459,
|
| 359 |
+
"grad_norm": 3.877307176589966,
|
| 360 |
+
"learning_rate": 8.391111111111112e-06,
|
| 361 |
+
"loss": 0.1529,
|
| 362 |
"step": 1225
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 1.8524833209785025,
|
| 366 |
+
"grad_norm": 4.159163475036621,
|
| 367 |
+
"learning_rate": 8.335555555555556e-06,
|
| 368 |
+
"loss": 0.1619,
|
| 369 |
"step": 1250
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 1.8895478131949592,
|
| 373 |
+
"grad_norm": 3.6631579399108887,
|
| 374 |
+
"learning_rate": 8.28e-06,
|
| 375 |
+
"loss": 0.1654,
|
| 376 |
"step": 1275
|
| 377 |
},
|
| 378 |
{
|
| 379 |
+
"epoch": 1.9266123054114157,
|
| 380 |
+
"grad_norm": 4.1784210205078125,
|
| 381 |
+
"learning_rate": 8.224444444444444e-06,
|
| 382 |
+
"loss": 0.1494,
|
| 383 |
"step": 1300
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 1.9636767976278726,
|
| 387 |
+
"grad_norm": 5.867852210998535,
|
| 388 |
+
"learning_rate": 8.16888888888889e-06,
|
| 389 |
+
"loss": 0.1443,
|
| 390 |
"step": 1325
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 2.0,
|
| 394 |
+
"grad_norm": 5.817214012145996,
|
| 395 |
+
"learning_rate": 8.113333333333334e-06,
|
| 396 |
+
"loss": 0.139,
|
| 397 |
"step": 1350
|
| 398 |
},
|
| 399 |
{
|
| 400 |
+
"epoch": 2.0370644922164565,
|
| 401 |
+
"grad_norm": 2.3572022914886475,
|
| 402 |
+
"learning_rate": 8.057777777777778e-06,
|
| 403 |
+
"loss": 0.0614,
|
| 404 |
"step": 1375
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 2.0741289844329134,
|
| 408 |
+
"grad_norm": 2.2769412994384766,
|
| 409 |
+
"learning_rate": 8.002222222222222e-06,
|
| 410 |
+
"loss": 0.0606,
|
| 411 |
"step": 1400
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 2.11119347664937,
|
| 415 |
+
"grad_norm": 2.474583864212036,
|
| 416 |
+
"learning_rate": 7.946666666666666e-06,
|
| 417 |
+
"loss": 0.0716,
|
| 418 |
"step": 1425
|
| 419 |
},
|
| 420 |
{
|
| 421 |
+
"epoch": 2.1482579688658268,
|
| 422 |
+
"grad_norm": 2.5783841609954834,
|
| 423 |
+
"learning_rate": 7.891111111111112e-06,
|
| 424 |
+
"loss": 0.065,
|
| 425 |
"step": 1450
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 2.1853224610822832,
|
| 429 |
+
"grad_norm": 1.6132420301437378,
|
| 430 |
+
"learning_rate": 7.835555555555556e-06,
|
| 431 |
+
"loss": 0.067,
|
| 432 |
"step": 1475
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"epoch": 2.2223869532987397,
|
| 436 |
+
"grad_norm": 3.8042001724243164,
|
| 437 |
+
"learning_rate": 7.78e-06,
|
| 438 |
+
"loss": 0.0724,
|
| 439 |
"step": 1500
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 2.2594514455151966,
|
| 443 |
+
"grad_norm": 2.2419843673706055,
|
| 444 |
+
"learning_rate": 7.724444444444446e-06,
|
| 445 |
+
"loss": 0.0761,
|
| 446 |
"step": 1525
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 2.296515937731653,
|
| 450 |
+
"grad_norm": 2.706354856491089,
|
| 451 |
+
"learning_rate": 7.66888888888889e-06,
|
| 452 |
+
"loss": 0.0659,
|
| 453 |
"step": 1550
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 2.3335804299481095,
|
| 457 |
+
"grad_norm": 2.8394265174865723,
|
| 458 |
+
"learning_rate": 7.613333333333334e-06,
|
| 459 |
+
"loss": 0.0688,
|
| 460 |
"step": 1575
|
| 461 |
},
|
| 462 |
{
|
| 463 |
+
"epoch": 2.3706449221645665,
|
| 464 |
+
"grad_norm": 2.383784770965576,
|
| 465 |
+
"learning_rate": 7.557777777777779e-06,
|
| 466 |
+
"loss": 0.0729,
|
| 467 |
"step": 1600
|
| 468 |
},
|
| 469 |
{
|
| 470 |
+
"epoch": 2.407709414381023,
|
| 471 |
+
"grad_norm": 3.0959832668304443,
|
| 472 |
+
"learning_rate": 7.502222222222223e-06,
|
| 473 |
+
"loss": 0.0626,
|
| 474 |
"step": 1625
|
| 475 |
},
|
| 476 |
{
|
| 477 |
+
"epoch": 2.4447739065974794,
|
| 478 |
+
"grad_norm": 2.927393913269043,
|
| 479 |
+
"learning_rate": 7.446666666666668e-06,
|
| 480 |
+
"loss": 0.0677,
|
| 481 |
"step": 1650
|
| 482 |
},
|
| 483 |
{
|
| 484 |
+
"epoch": 2.4818383988139363,
|
| 485 |
+
"grad_norm": 2.644434928894043,
|
| 486 |
+
"learning_rate": 7.3911111111111125e-06,
|
| 487 |
+
"loss": 0.0644,
|
| 488 |
"step": 1675
|
| 489 |
},
|
| 490 |
{
|
| 491 |
+
"epoch": 2.5189028910303928,
|
| 492 |
+
"grad_norm": 2.9071755409240723,
|
| 493 |
+
"learning_rate": 7.335555555555556e-06,
|
| 494 |
+
"loss": 0.061,
|
| 495 |
"step": 1700
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 2.5559673832468492,
|
| 499 |
+
"grad_norm": 2.6862034797668457,
|
| 500 |
+
"learning_rate": 7.280000000000001e-06,
|
| 501 |
+
"loss": 0.0615,
|
| 502 |
"step": 1725
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"epoch": 2.593031875463306,
|
| 506 |
+
"grad_norm": 3.1184046268463135,
|
| 507 |
+
"learning_rate": 7.224444444444445e-06,
|
| 508 |
+
"loss": 0.0714,
|
| 509 |
"step": 1750
|
| 510 |
},
|
| 511 |
{
|
| 512 |
+
"epoch": 2.6300963676797626,
|
| 513 |
+
"grad_norm": 1.7592053413391113,
|
| 514 |
+
"learning_rate": 7.1688888888888895e-06,
|
| 515 |
+
"loss": 0.0704,
|
| 516 |
"step": 1775
|
| 517 |
},
|
| 518 |
{
|
| 519 |
+
"epoch": 2.6671608598962195,
|
| 520 |
+
"grad_norm": 2.9316508769989014,
|
| 521 |
+
"learning_rate": 7.113333333333334e-06,
|
| 522 |
+
"loss": 0.0689,
|
| 523 |
"step": 1800
|
| 524 |
},
|
| 525 |
{
|
| 526 |
+
"epoch": 2.704225352112676,
|
| 527 |
+
"grad_norm": 2.1934666633605957,
|
| 528 |
+
"learning_rate": 7.057777777777778e-06,
|
| 529 |
+
"loss": 0.0721,
|
| 530 |
"step": 1825
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 2.741289844329133,
|
| 534 |
+
"grad_norm": 3.4919371604919434,
|
| 535 |
+
"learning_rate": 7.0022222222222225e-06,
|
| 536 |
+
"loss": 0.0638,
|
| 537 |
"step": 1850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 2.7783543365455894,
|
| 541 |
+
"grad_norm": 2.723252058029175,
|
| 542 |
+
"learning_rate": 6.946666666666667e-06,
|
| 543 |
+
"loss": 0.0598,
|
| 544 |
"step": 1875
|
| 545 |
},
|
| 546 |
{
|
| 547 |
+
"epoch": 2.815418828762046,
|
| 548 |
+
"grad_norm": 1.8668267726898193,
|
| 549 |
+
"learning_rate": 6.891111111111111e-06,
|
| 550 |
+
"loss": 0.0607,
|
| 551 |
"step": 1900
|
| 552 |
},
|
| 553 |
{
|
| 554 |
+
"epoch": 2.8524833209785028,
|
| 555 |
+
"grad_norm": 2.0989866256713867,
|
| 556 |
+
"learning_rate": 6.835555555555556e-06,
|
| 557 |
+
"loss": 0.0821,
|
| 558 |
"step": 1925
|
| 559 |
},
|
| 560 |
{
|
| 561 |
+
"epoch": 2.8895478131949592,
|
| 562 |
+
"grad_norm": 2.9375364780426025,
|
| 563 |
+
"learning_rate": 6.780000000000001e-06,
|
| 564 |
+
"loss": 0.0636,
|
| 565 |
"step": 1950
|
| 566 |
},
|
| 567 |
{
|
| 568 |
+
"epoch": 2.9266123054114157,
|
| 569 |
+
"grad_norm": 2.1375315189361572,
|
| 570 |
+
"learning_rate": 6.724444444444444e-06,
|
| 571 |
+
"loss": 0.0723,
|
| 572 |
"step": 1975
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 2.9636767976278726,
|
| 576 |
+
"grad_norm": 2.5874264240264893,
|
| 577 |
+
"learning_rate": 6.668888888888889e-06,
|
| 578 |
+
"loss": 0.0706,
|
| 579 |
"step": 2000
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 2.9636767976278726,
|
| 583 |
+
"eval_loss": 0.2490690052509308,
|
| 584 |
+
"eval_runtime": 730.2087,
|
| 585 |
+
"eval_samples_per_second": 3.951,
|
| 586 |
+
"eval_steps_per_second": 0.494,
|
| 587 |
+
"eval_wer": 0.12696648426812585,
|
| 588 |
"step": 2000
|
| 589 |
},
|
| 590 |
{
|
| 591 |
+
"epoch": 3.0,
|
| 592 |
+
"grad_norm": 6.509148597717285,
|
| 593 |
+
"learning_rate": 6.613333333333334e-06,
|
| 594 |
+
"loss": 0.0587,
|
| 595 |
+
"step": 2025
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 3.0370644922164565,
|
| 599 |
+
"grad_norm": 1.9590086936950684,
|
| 600 |
+
"learning_rate": 6.557777777777778e-06,
|
| 601 |
+
"loss": 0.0241,
|
| 602 |
+
"step": 2050
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.0741289844329134,
|
| 606 |
+
"grad_norm": 1.4612740278244019,
|
| 607 |
+
"learning_rate": 6.502222222222223e-06,
|
| 608 |
+
"loss": 0.0267,
|
| 609 |
+
"step": 2075
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.11119347664937,
|
| 613 |
+
"grad_norm": 0.9522780179977417,
|
| 614 |
+
"learning_rate": 6.446666666666668e-06,
|
| 615 |
+
"loss": 0.023,
|
| 616 |
+
"step": 2100
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 3.1482579688658268,
|
| 620 |
+
"grad_norm": 1.891400694847107,
|
| 621 |
+
"learning_rate": 6.391111111111111e-06,
|
| 622 |
+
"loss": 0.0281,
|
| 623 |
+
"step": 2125
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 3.1853224610822832,
|
| 627 |
+
"grad_norm": 1.0783302783966064,
|
| 628 |
+
"learning_rate": 6.335555555555556e-06,
|
| 629 |
+
"loss": 0.0246,
|
| 630 |
+
"step": 2150
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 3.2223869532987397,
|
| 634 |
+
"grad_norm": 1.3504562377929688,
|
| 635 |
+
"learning_rate": 6.280000000000001e-06,
|
| 636 |
+
"loss": 0.0244,
|
| 637 |
+
"step": 2175
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 3.2594514455151966,
|
| 641 |
+
"grad_norm": 1.8768439292907715,
|
| 642 |
+
"learning_rate": 6.224444444444445e-06,
|
| 643 |
+
"loss": 0.0264,
|
| 644 |
+
"step": 2200
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 3.296515937731653,
|
| 648 |
+
"grad_norm": 1.5083887577056885,
|
| 649 |
+
"learning_rate": 6.16888888888889e-06,
|
| 650 |
+
"loss": 0.0248,
|
| 651 |
+
"step": 2225
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 3.3335804299481095,
|
| 655 |
+
"grad_norm": 3.5768120288848877,
|
| 656 |
+
"learning_rate": 6.113333333333333e-06,
|
| 657 |
+
"loss": 0.0316,
|
| 658 |
+
"step": 2250
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 3.3706449221645665,
|
| 662 |
+
"grad_norm": 1.1493444442749023,
|
| 663 |
+
"learning_rate": 6.057777777777778e-06,
|
| 664 |
+
"loss": 0.0294,
|
| 665 |
+
"step": 2275
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 3.407709414381023,
|
| 669 |
+
"grad_norm": 2.3746306896209717,
|
| 670 |
+
"learning_rate": 6.002222222222223e-06,
|
| 671 |
+
"loss": 0.0263,
|
| 672 |
+
"step": 2300
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 3.4447739065974794,
|
| 676 |
+
"grad_norm": 2.144634485244751,
|
| 677 |
+
"learning_rate": 5.946666666666668e-06,
|
| 678 |
+
"loss": 0.0348,
|
| 679 |
+
"step": 2325
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 3.4818383988139363,
|
| 683 |
+
"grad_norm": 1.5002686977386475,
|
| 684 |
+
"learning_rate": 5.891111111111112e-06,
|
| 685 |
+
"loss": 0.0228,
|
| 686 |
+
"step": 2350
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 3.5189028910303928,
|
| 690 |
+
"grad_norm": 1.6059187650680542,
|
| 691 |
+
"learning_rate": 5.8355555555555565e-06,
|
| 692 |
+
"loss": 0.0239,
|
| 693 |
+
"step": 2375
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 3.5559673832468492,
|
| 697 |
+
"grad_norm": 2.757420778274536,
|
| 698 |
+
"learning_rate": 5.78e-06,
|
| 699 |
+
"loss": 0.0277,
|
| 700 |
+
"step": 2400
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 3.593031875463306,
|
| 704 |
+
"grad_norm": 1.3977222442626953,
|
| 705 |
+
"learning_rate": 5.724444444444445e-06,
|
| 706 |
+
"loss": 0.0224,
|
| 707 |
+
"step": 2425
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 3.6300963676797626,
|
| 711 |
+
"grad_norm": 1.9618048667907715,
|
| 712 |
+
"learning_rate": 5.6688888888888895e-06,
|
| 713 |
+
"loss": 0.026,
|
| 714 |
+
"step": 2450
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 3.6671608598962195,
|
| 718 |
+
"grad_norm": 0.898245632648468,
|
| 719 |
+
"learning_rate": 5.613333333333334e-06,
|
| 720 |
+
"loss": 0.0326,
|
| 721 |
+
"step": 2475
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 3.704225352112676,
|
| 725 |
+
"grad_norm": 1.8148616552352905,
|
| 726 |
+
"learning_rate": 5.557777777777778e-06,
|
| 727 |
+
"loss": 0.0213,
|
| 728 |
+
"step": 2500
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 3.741289844329133,
|
| 732 |
+
"grad_norm": 1.308030366897583,
|
| 733 |
+
"learning_rate": 5.5022222222222224e-06,
|
| 734 |
+
"loss": 0.0192,
|
| 735 |
+
"step": 2525
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 3.7783543365455894,
|
| 739 |
+
"grad_norm": 1.6680744886398315,
|
| 740 |
+
"learning_rate": 5.4466666666666665e-06,
|
| 741 |
+
"loss": 0.027,
|
| 742 |
+
"step": 2550
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 3.815418828762046,
|
| 746 |
+
"grad_norm": 3.235917568206787,
|
| 747 |
+
"learning_rate": 5.391111111111111e-06,
|
| 748 |
+
"loss": 0.0242,
|
| 749 |
+
"step": 2575
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 3.8524833209785028,
|
| 753 |
+
"grad_norm": 2.096780300140381,
|
| 754 |
+
"learning_rate": 5.335555555555556e-06,
|
| 755 |
+
"loss": 0.0243,
|
| 756 |
+
"step": 2600
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 3.8895478131949592,
|
| 760 |
+
"grad_norm": 1.8445031642913818,
|
| 761 |
+
"learning_rate": 5.28e-06,
|
| 762 |
+
"loss": 0.024,
|
| 763 |
+
"step": 2625
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 3.9266123054114157,
|
| 767 |
+
"grad_norm": 1.357937216758728,
|
| 768 |
+
"learning_rate": 5.224444444444445e-06,
|
| 769 |
+
"loss": 0.0244,
|
| 770 |
+
"step": 2650
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 3.9636767976278726,
|
| 774 |
+
"grad_norm": 1.0413466691970825,
|
| 775 |
+
"learning_rate": 5.168888888888889e-06,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 2675
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 4.0,
|
| 781 |
+
"grad_norm": 3.0572996139526367,
|
| 782 |
+
"learning_rate": 5.113333333333333e-06,
|
| 783 |
+
"loss": 0.0206,
|
| 784 |
+
"step": 2700
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 4.037064492216457,
|
| 788 |
+
"grad_norm": 0.9961848258972168,
|
| 789 |
+
"learning_rate": 5.057777777777778e-06,
|
| 790 |
+
"loss": 0.0136,
|
| 791 |
+
"step": 2725
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 4.074128984432913,
|
| 795 |
+
"grad_norm": 1.0248702764511108,
|
| 796 |
+
"learning_rate": 5.002222222222223e-06,
|
| 797 |
+
"loss": 0.009,
|
| 798 |
+
"step": 2750
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 4.11119347664937,
|
| 802 |
+
"grad_norm": 0.6142157912254333,
|
| 803 |
+
"learning_rate": 4.946666666666667e-06,
|
| 804 |
+
"loss": 0.0113,
|
| 805 |
+
"step": 2775
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 4.148257968865827,
|
| 809 |
+
"grad_norm": 0.27292531728744507,
|
| 810 |
+
"learning_rate": 4.891111111111111e-06,
|
| 811 |
+
"loss": 0.009,
|
| 812 |
+
"step": 2800
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.185322461082283,
|
| 816 |
+
"grad_norm": 2.2906312942504883,
|
| 817 |
+
"learning_rate": 4.835555555555556e-06,
|
| 818 |
+
"loss": 0.0073,
|
| 819 |
+
"step": 2825
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.22238695329874,
|
| 823 |
+
"grad_norm": 1.0498850345611572,
|
| 824 |
+
"learning_rate": 4.78e-06,
|
| 825 |
+
"loss": 0.0093,
|
| 826 |
+
"step": 2850
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 4.259451445515197,
|
| 830 |
+
"grad_norm": 1.1574844121932983,
|
| 831 |
+
"learning_rate": 4.724444444444445e-06,
|
| 832 |
+
"loss": 0.0159,
|
| 833 |
+
"step": 2875
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 4.2965159377316535,
|
| 837 |
+
"grad_norm": 0.7209671139717102,
|
| 838 |
+
"learning_rate": 4.66888888888889e-06,
|
| 839 |
+
"loss": 0.0088,
|
| 840 |
+
"step": 2900
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 4.3335804299481095,
|
| 844 |
+
"grad_norm": 1.168841004371643,
|
| 845 |
+
"learning_rate": 4.613333333333334e-06,
|
| 846 |
+
"loss": 0.0094,
|
| 847 |
+
"step": 2925
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 4.3706449221645665,
|
| 851 |
+
"grad_norm": 0.6153778433799744,
|
| 852 |
+
"learning_rate": 4.557777777777778e-06,
|
| 853 |
+
"loss": 0.009,
|
| 854 |
+
"step": 2950
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 4.407709414381023,
|
| 858 |
+
"grad_norm": 1.5705232620239258,
|
| 859 |
+
"learning_rate": 4.502222222222223e-06,
|
| 860 |
+
"loss": 0.0085,
|
| 861 |
+
"step": 2975
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 4.444773906597479,
|
| 865 |
+
"grad_norm": 0.24448032677173615,
|
| 866 |
+
"learning_rate": 4.446666666666667e-06,
|
| 867 |
+
"loss": 0.0072,
|
| 868 |
+
"step": 3000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 4.444773906597479,
|
| 872 |
+
"eval_loss": 0.27286583185195923,
|
| 873 |
+
"eval_runtime": 739.8615,
|
| 874 |
+
"eval_samples_per_second": 3.899,
|
| 875 |
+
"eval_steps_per_second": 0.488,
|
| 876 |
+
"eval_wer": 0.11913474692202462,
|
| 877 |
+
"step": 3000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.481838398813936,
|
| 881 |
+
"grad_norm": 1.2278587818145752,
|
| 882 |
+
"learning_rate": 4.391111111111112e-06,
|
| 883 |
+
"loss": 0.0146,
|
| 884 |
+
"step": 3025
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.518902891030393,
|
| 888 |
+
"grad_norm": 0.6478213667869568,
|
| 889 |
+
"learning_rate": 4.3355555555555565e-06,
|
| 890 |
+
"loss": 0.014,
|
| 891 |
+
"step": 3050
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.555967383246849,
|
| 895 |
+
"grad_norm": 0.7865190505981445,
|
| 896 |
+
"learning_rate": 4.2800000000000005e-06,
|
| 897 |
+
"loss": 0.0079,
|
| 898 |
+
"step": 3075
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.593031875463306,
|
| 902 |
+
"grad_norm": 2.3078877925872803,
|
| 903 |
+
"learning_rate": 4.2244444444444446e-06,
|
| 904 |
+
"loss": 0.009,
|
| 905 |
+
"step": 3100
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.630096367679763,
|
| 909 |
+
"grad_norm": 0.9625842571258545,
|
| 910 |
+
"learning_rate": 4.168888888888889e-06,
|
| 911 |
+
"loss": 0.0096,
|
| 912 |
+
"step": 3125
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.667160859896219,
|
| 916 |
+
"grad_norm": 0.7619579434394836,
|
| 917 |
+
"learning_rate": 4.1133333333333335e-06,
|
| 918 |
+
"loss": 0.0096,
|
| 919 |
+
"step": 3150
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.704225352112676,
|
| 923 |
+
"grad_norm": 1.5049270391464233,
|
| 924 |
+
"learning_rate": 4.057777777777778e-06,
|
| 925 |
+
"loss": 0.0099,
|
| 926 |
+
"step": 3175
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.741289844329133,
|
| 930 |
+
"grad_norm": 1.1056573390960693,
|
| 931 |
+
"learning_rate": 4.002222222222222e-06,
|
| 932 |
+
"loss": 0.0065,
|
| 933 |
+
"step": 3200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.778354336545589,
|
| 937 |
+
"grad_norm": 0.7983392477035522,
|
| 938 |
+
"learning_rate": 3.946666666666667e-06,
|
| 939 |
+
"loss": 0.0105,
|
| 940 |
+
"step": 3225
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.815418828762046,
|
| 944 |
+
"grad_norm": 1.1153795719146729,
|
| 945 |
+
"learning_rate": 3.891111111111111e-06,
|
| 946 |
+
"loss": 0.0075,
|
| 947 |
+
"step": 3250
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.852483320978503,
|
| 951 |
+
"grad_norm": 0.9730608463287354,
|
| 952 |
+
"learning_rate": 3.835555555555555e-06,
|
| 953 |
+
"loss": 0.0087,
|
| 954 |
+
"step": 3275
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.889547813194959,
|
| 958 |
+
"grad_norm": 0.5694206953048706,
|
| 959 |
+
"learning_rate": 3.7800000000000002e-06,
|
| 960 |
+
"loss": 0.0071,
|
| 961 |
+
"step": 3300
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.926612305411416,
|
| 965 |
+
"grad_norm": 0.2520028352737427,
|
| 966 |
+
"learning_rate": 3.724444444444445e-06,
|
| 967 |
+
"loss": 0.0081,
|
| 968 |
+
"step": 3325
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.963676797627873,
|
| 972 |
+
"grad_norm": 0.436355322599411,
|
| 973 |
+
"learning_rate": 3.668888888888889e-06,
|
| 974 |
+
"loss": 0.0078,
|
| 975 |
+
"step": 3350
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 5.0,
|
| 979 |
+
"grad_norm": 0.798361599445343,
|
| 980 |
+
"learning_rate": 3.6133333333333336e-06,
|
| 981 |
+
"loss": 0.0075,
|
| 982 |
+
"step": 3375
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 5.037064492216457,
|
| 986 |
+
"grad_norm": 1.3702267408370972,
|
| 987 |
+
"learning_rate": 3.5577777777777785e-06,
|
| 988 |
+
"loss": 0.005,
|
| 989 |
+
"step": 3400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 5.074128984432913,
|
| 993 |
+
"grad_norm": 0.2790464162826538,
|
| 994 |
+
"learning_rate": 3.5022222222222225e-06,
|
| 995 |
+
"loss": 0.0032,
|
| 996 |
+
"step": 3425
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.11119347664937,
|
| 1000 |
+
"grad_norm": 0.15111476182937622,
|
| 1001 |
+
"learning_rate": 3.446666666666667e-06,
|
| 1002 |
+
"loss": 0.0046,
|
| 1003 |
+
"step": 3450
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 5.148257968865827,
|
| 1007 |
+
"grad_norm": 0.09985285252332687,
|
| 1008 |
+
"learning_rate": 3.391111111111111e-06,
|
| 1009 |
+
"loss": 0.0035,
|
| 1010 |
+
"step": 3475
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.185322461082283,
|
| 1014 |
+
"grad_norm": 0.5352105498313904,
|
| 1015 |
+
"learning_rate": 3.335555555555556e-06,
|
| 1016 |
+
"loss": 0.0031,
|
| 1017 |
+
"step": 3500
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.22238695329874,
|
| 1021 |
+
"grad_norm": 0.9406213760375977,
|
| 1022 |
+
"learning_rate": 3.2800000000000004e-06,
|
| 1023 |
+
"loss": 0.0035,
|
| 1024 |
+
"step": 3525
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 5.259451445515197,
|
| 1028 |
+
"grad_norm": 0.7073507905006409,
|
| 1029 |
+
"learning_rate": 3.2244444444444444e-06,
|
| 1030 |
+
"loss": 0.0035,
|
| 1031 |
+
"step": 3550
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 5.2965159377316535,
|
| 1035 |
+
"grad_norm": 0.07916448265314102,
|
| 1036 |
+
"learning_rate": 3.1688888888888893e-06,
|
| 1037 |
+
"loss": 0.0035,
|
| 1038 |
+
"step": 3575
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.3335804299481095,
|
| 1042 |
+
"grad_norm": 0.5285120606422424,
|
| 1043 |
+
"learning_rate": 3.1133333333333337e-06,
|
| 1044 |
+
"loss": 0.0027,
|
| 1045 |
+
"step": 3600
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.3706449221645665,
|
| 1049 |
+
"grad_norm": 0.09832775592803955,
|
| 1050 |
+
"learning_rate": 3.0577777777777778e-06,
|
| 1051 |
+
"loss": 0.0036,
|
| 1052 |
+
"step": 3625
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.407709414381023,
|
| 1056 |
+
"grad_norm": 0.21083103120326996,
|
| 1057 |
+
"learning_rate": 3.0022222222222227e-06,
|
| 1058 |
+
"loss": 0.0041,
|
| 1059 |
+
"step": 3650
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.444773906597479,
|
| 1063 |
+
"grad_norm": 0.6747980713844299,
|
| 1064 |
+
"learning_rate": 2.946666666666667e-06,
|
| 1065 |
+
"loss": 0.003,
|
| 1066 |
+
"step": 3675
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.481838398813936,
|
| 1070 |
+
"grad_norm": 0.5111549496650696,
|
| 1071 |
+
"learning_rate": 2.891111111111111e-06,
|
| 1072 |
+
"loss": 0.0028,
|
| 1073 |
+
"step": 3700
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.518902891030393,
|
| 1077 |
+
"grad_norm": 0.6502516269683838,
|
| 1078 |
+
"learning_rate": 2.835555555555556e-06,
|
| 1079 |
+
"loss": 0.0045,
|
| 1080 |
+
"step": 3725
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.555967383246849,
|
| 1084 |
+
"grad_norm": 0.4688964784145355,
|
| 1085 |
+
"learning_rate": 2.7800000000000005e-06,
|
| 1086 |
+
"loss": 0.0036,
|
| 1087 |
+
"step": 3750
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.593031875463306,
|
| 1091 |
+
"grad_norm": 0.281994104385376,
|
| 1092 |
+
"learning_rate": 2.7244444444444445e-06,
|
| 1093 |
+
"loss": 0.0021,
|
| 1094 |
+
"step": 3775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.630096367679763,
|
| 1098 |
+
"grad_norm": 0.11583279073238373,
|
| 1099 |
+
"learning_rate": 2.6688888888888894e-06,
|
| 1100 |
+
"loss": 0.0041,
|
| 1101 |
+
"step": 3800
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.667160859896219,
|
| 1105 |
+
"grad_norm": 0.22941534221172333,
|
| 1106 |
+
"learning_rate": 2.6133333333333334e-06,
|
| 1107 |
+
"loss": 0.0022,
|
| 1108 |
+
"step": 3825
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.704225352112676,
|
| 1112 |
+
"grad_norm": 0.13950073719024658,
|
| 1113 |
+
"learning_rate": 2.557777777777778e-06,
|
| 1114 |
+
"loss": 0.003,
|
| 1115 |
+
"step": 3850
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.741289844329133,
|
| 1119 |
+
"grad_norm": 0.6869206428527832,
|
| 1120 |
+
"learning_rate": 2.5022222222222224e-06,
|
| 1121 |
+
"loss": 0.0024,
|
| 1122 |
+
"step": 3875
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.778354336545589,
|
| 1126 |
+
"grad_norm": 0.09893081337213516,
|
| 1127 |
+
"learning_rate": 2.446666666666667e-06,
|
| 1128 |
+
"loss": 0.0029,
|
| 1129 |
+
"step": 3900
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.815418828762046,
|
| 1133 |
+
"grad_norm": 0.1264762133359909,
|
| 1134 |
+
"learning_rate": 2.3911111111111113e-06,
|
| 1135 |
+
"loss": 0.0033,
|
| 1136 |
+
"step": 3925
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.852483320978503,
|
| 1140 |
+
"grad_norm": 0.15489889681339264,
|
| 1141 |
+
"learning_rate": 2.3355555555555557e-06,
|
| 1142 |
+
"loss": 0.003,
|
| 1143 |
+
"step": 3950
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.889547813194959,
|
| 1147 |
+
"grad_norm": 0.5875250697135925,
|
| 1148 |
+
"learning_rate": 2.28e-06,
|
| 1149 |
+
"loss": 0.0022,
|
| 1150 |
+
"step": 3975
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.926612305411416,
|
| 1154 |
+
"grad_norm": 0.06691984087228775,
|
| 1155 |
+
"learning_rate": 2.2244444444444447e-06,
|
| 1156 |
+
"loss": 0.005,
|
| 1157 |
+
"step": 4000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.926612305411416,
|
| 1161 |
+
"eval_loss": 0.28099098801612854,
|
| 1162 |
+
"eval_runtime": 734.9707,
|
| 1163 |
+
"eval_samples_per_second": 3.925,
|
| 1164 |
+
"eval_steps_per_second": 0.491,
|
| 1165 |
+
"eval_wer": 0.11566347469220246,
|
| 1166 |
+
"step": 4000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 5.963676797627873,
|
| 1170 |
+
"grad_norm": 0.2645249664783478,
|
| 1171 |
+
"learning_rate": 2.168888888888889e-06,
|
| 1172 |
+
"loss": 0.0026,
|
| 1173 |
+
"step": 4025
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 6.0,
|
| 1177 |
+
"grad_norm": 0.3361597955226898,
|
| 1178 |
+
"learning_rate": 2.1133333333333336e-06,
|
| 1179 |
+
"loss": 0.0023,
|
| 1180 |
+
"step": 4050
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 6.037064492216457,
|
| 1184 |
+
"grad_norm": 0.059147898107767105,
|
| 1185 |
+
"learning_rate": 2.057777777777778e-06,
|
| 1186 |
+
"loss": 0.0015,
|
| 1187 |
+
"step": 4075
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 6.074128984432913,
|
| 1191 |
+
"grad_norm": 0.1158735603094101,
|
| 1192 |
+
"learning_rate": 2.0022222222222225e-06,
|
| 1193 |
+
"loss": 0.0016,
|
| 1194 |
+
"step": 4100
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 6.11119347664937,
|
| 1198 |
+
"grad_norm": 1.3564985990524292,
|
| 1199 |
+
"learning_rate": 1.9466666666666665e-06,
|
| 1200 |
+
"loss": 0.0014,
|
| 1201 |
+
"step": 4125
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 6.148257968865827,
|
| 1205 |
+
"grad_norm": 0.5956087112426758,
|
| 1206 |
+
"learning_rate": 1.8911111111111114e-06,
|
| 1207 |
+
"loss": 0.0018,
|
| 1208 |
+
"step": 4150
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 6.185322461082283,
|
| 1212 |
+
"grad_norm": 0.09224885702133179,
|
| 1213 |
+
"learning_rate": 1.8355555555555557e-06,
|
| 1214 |
+
"loss": 0.0017,
|
| 1215 |
+
"step": 4175
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 6.22238695329874,
|
| 1219 |
+
"grad_norm": 0.06868930906057358,
|
| 1220 |
+
"learning_rate": 1.7800000000000001e-06,
|
| 1221 |
+
"loss": 0.0017,
|
| 1222 |
+
"step": 4200
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 6.259451445515197,
|
| 1226 |
+
"grad_norm": 0.06657718122005463,
|
| 1227 |
+
"learning_rate": 1.7244444444444448e-06,
|
| 1228 |
+
"loss": 0.0014,
|
| 1229 |
+
"step": 4225
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 6.2965159377316535,
|
| 1233 |
+
"grad_norm": 0.05459928885102272,
|
| 1234 |
+
"learning_rate": 1.668888888888889e-06,
|
| 1235 |
+
"loss": 0.0017,
|
| 1236 |
+
"step": 4250
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 6.3335804299481095,
|
| 1240 |
+
"grad_norm": 0.05795517563819885,
|
| 1241 |
+
"learning_rate": 1.6133333333333335e-06,
|
| 1242 |
+
"loss": 0.0027,
|
| 1243 |
+
"step": 4275
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 6.3706449221645665,
|
| 1247 |
+
"grad_norm": 0.06204914301633835,
|
| 1248 |
+
"learning_rate": 1.5577777777777777e-06,
|
| 1249 |
+
"loss": 0.0012,
|
| 1250 |
+
"step": 4300
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 6.407709414381023,
|
| 1254 |
+
"grad_norm": 0.0820712074637413,
|
| 1255 |
+
"learning_rate": 1.5022222222222224e-06,
|
| 1256 |
+
"loss": 0.0012,
|
| 1257 |
+
"step": 4325
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 6.444773906597479,
|
| 1261 |
+
"grad_norm": 0.056523606181144714,
|
| 1262 |
+
"learning_rate": 1.4466666666666669e-06,
|
| 1263 |
+
"loss": 0.0013,
|
| 1264 |
+
"step": 4350
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 6.481838398813936,
|
| 1268 |
+
"grad_norm": 0.07985592633485794,
|
| 1269 |
+
"learning_rate": 1.3911111111111111e-06,
|
| 1270 |
+
"loss": 0.0014,
|
| 1271 |
+
"step": 4375
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 6.518902891030393,
|
| 1275 |
+
"grad_norm": 0.044111426919698715,
|
| 1276 |
+
"learning_rate": 1.3355555555555558e-06,
|
| 1277 |
+
"loss": 0.0012,
|
| 1278 |
+
"step": 4400
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 6.555967383246849,
|
| 1282 |
+
"grad_norm": 0.05683915689587593,
|
| 1283 |
+
"learning_rate": 1.28e-06,
|
| 1284 |
+
"loss": 0.0014,
|
| 1285 |
+
"step": 4425
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 6.593031875463306,
|
| 1289 |
+
"grad_norm": 0.08568093180656433,
|
| 1290 |
+
"learning_rate": 1.2244444444444445e-06,
|
| 1291 |
+
"loss": 0.0012,
|
| 1292 |
+
"step": 4450
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 6.630096367679763,
|
| 1296 |
+
"grad_norm": 0.054062824696302414,
|
| 1297 |
+
"learning_rate": 1.168888888888889e-06,
|
| 1298 |
+
"loss": 0.0011,
|
| 1299 |
+
"step": 4475
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 6.667160859896219,
|
| 1303 |
+
"grad_norm": 0.0509476363658905,
|
| 1304 |
+
"learning_rate": 1.1133333333333334e-06,
|
| 1305 |
+
"loss": 0.0013,
|
| 1306 |
+
"step": 4500
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 6.704225352112676,
|
| 1310 |
+
"grad_norm": 0.04927874356508255,
|
| 1311 |
+
"learning_rate": 1.0577777777777779e-06,
|
| 1312 |
+
"loss": 0.0012,
|
| 1313 |
+
"step": 4525
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 6.741289844329133,
|
| 1317 |
+
"grad_norm": 0.08598697185516357,
|
| 1318 |
+
"learning_rate": 1.0022222222222223e-06,
|
| 1319 |
+
"loss": 0.0011,
|
| 1320 |
+
"step": 4550
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 6.778354336545589,
|
| 1324 |
+
"grad_norm": 0.3571934700012207,
|
| 1325 |
+
"learning_rate": 9.466666666666667e-07,
|
| 1326 |
+
"loss": 0.0016,
|
| 1327 |
+
"step": 4575
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 6.815418828762046,
|
| 1331 |
+
"grad_norm": 0.05977300554513931,
|
| 1332 |
+
"learning_rate": 8.911111111111112e-07,
|
| 1333 |
+
"loss": 0.001,
|
| 1334 |
+
"step": 4600
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 6.852483320978503,
|
| 1338 |
+
"grad_norm": 0.05966237559914589,
|
| 1339 |
+
"learning_rate": 8.355555555555556e-07,
|
| 1340 |
+
"loss": 0.001,
|
| 1341 |
+
"step": 4625
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 6.889547813194959,
|
| 1345 |
+
"grad_norm": 0.05432112514972687,
|
| 1346 |
+
"learning_rate": 7.8e-07,
|
| 1347 |
+
"loss": 0.001,
|
| 1348 |
+
"step": 4650
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 6.926612305411416,
|
| 1352 |
+
"grad_norm": 0.06741122156381607,
|
| 1353 |
+
"learning_rate": 7.244444444444446e-07,
|
| 1354 |
+
"loss": 0.0019,
|
| 1355 |
+
"step": 4675
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 6.963676797627873,
|
| 1359 |
+
"grad_norm": 0.04723643884062767,
|
| 1360 |
+
"learning_rate": 6.68888888888889e-07,
|
| 1361 |
+
"loss": 0.0012,
|
| 1362 |
+
"step": 4700
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 7.0,
|
| 1366 |
+
"grad_norm": 0.07329325377941132,
|
| 1367 |
+
"learning_rate": 6.133333333333333e-07,
|
| 1368 |
+
"loss": 0.001,
|
| 1369 |
+
"step": 4725
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 7.037064492216457,
|
| 1373 |
+
"grad_norm": 0.06389188766479492,
|
| 1374 |
+
"learning_rate": 5.577777777777779e-07,
|
| 1375 |
+
"loss": 0.001,
|
| 1376 |
+
"step": 4750
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 7.074128984432913,
|
| 1380 |
+
"grad_norm": 0.03797365352511406,
|
| 1381 |
+
"learning_rate": 5.022222222222222e-07,
|
| 1382 |
+
"loss": 0.001,
|
| 1383 |
+
"step": 4775
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 7.11119347664937,
|
| 1387 |
+
"grad_norm": 0.04686768725514412,
|
| 1388 |
+
"learning_rate": 4.466666666666667e-07,
|
| 1389 |
+
"loss": 0.0009,
|
| 1390 |
+
"step": 4800
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 7.148257968865827,
|
| 1394 |
+
"grad_norm": 0.06883518397808075,
|
| 1395 |
+
"learning_rate": 3.9111111111111115e-07,
|
| 1396 |
+
"loss": 0.001,
|
| 1397 |
+
"step": 4825
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 7.185322461082283,
|
| 1401 |
+
"grad_norm": 0.02842629700899124,
|
| 1402 |
+
"learning_rate": 3.3555555555555556e-07,
|
| 1403 |
+
"loss": 0.0009,
|
| 1404 |
+
"step": 4850
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 7.22238695329874,
|
| 1408 |
+
"grad_norm": 0.04749394953250885,
|
| 1409 |
+
"learning_rate": 2.8e-07,
|
| 1410 |
+
"loss": 0.001,
|
| 1411 |
+
"step": 4875
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 7.259451445515197,
|
| 1415 |
+
"grad_norm": 0.04491546377539635,
|
| 1416 |
+
"learning_rate": 2.2444444444444445e-07,
|
| 1417 |
+
"loss": 0.001,
|
| 1418 |
+
"step": 4900
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 7.2965159377316535,
|
| 1422 |
+
"grad_norm": 0.056013334542512894,
|
| 1423 |
+
"learning_rate": 1.6888888888888888e-07,
|
| 1424 |
+
"loss": 0.001,
|
| 1425 |
+
"step": 4925
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 7.3335804299481095,
|
| 1429 |
+
"grad_norm": 0.057778194546699524,
|
| 1430 |
+
"learning_rate": 1.1333333333333336e-07,
|
| 1431 |
+
"loss": 0.0011,
|
| 1432 |
+
"step": 4950
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 7.3706449221645665,
|
| 1436 |
+
"grad_norm": 0.051241885870695114,
|
| 1437 |
+
"learning_rate": 5.777777777777778e-08,
|
| 1438 |
+
"loss": 0.0011,
|
| 1439 |
+
"step": 4975
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 7.407709414381023,
|
| 1443 |
+
"grad_norm": 0.06301814317703247,
|
| 1444 |
+
"learning_rate": 2.2222222222222225e-09,
|
| 1445 |
+
"loss": 0.0009,
|
| 1446 |
+
"step": 5000
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 7.407709414381023,
|
| 1450 |
+
"eval_loss": 0.29011788964271545,
|
| 1451 |
+
"eval_runtime": 732.4342,
|
| 1452 |
+
"eval_samples_per_second": 3.939,
|
| 1453 |
+
"eval_steps_per_second": 0.493,
|
| 1454 |
+
"eval_wer": 0.1146545827633379,
|
| 1455 |
+
"step": 5000
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 7.407709414381023,
|
| 1459 |
+
"step": 5000,
|
| 1460 |
+
"total_flos": 8.155551755501568e+19,
|
| 1461 |
+
"train_loss": 0.10907779041565954,
|
| 1462 |
+
"train_runtime": 12394.4337,
|
| 1463 |
+
"train_samples_per_second": 6.455,
|
| 1464 |
+
"train_steps_per_second": 0.403
|
| 1465 |
}
|
| 1466 |
],
|
| 1467 |
"logging_steps": 25,
|
| 1468 |
+
"max_steps": 5000,
|
| 1469 |
"num_input_tokens_seen": 0,
|
| 1470 |
+
"num_train_epochs": 8,
|
| 1471 |
"save_steps": 1000,
|
| 1472 |
"stateful_callbacks": {
|
| 1473 |
"TrainerControl": {
|
|
|
|
| 1481 |
"attributes": {}
|
| 1482 |
}
|
| 1483 |
},
|
| 1484 |
+
"total_flos": 8.155551755501568e+19,
|
| 1485 |
+
"train_batch_size": 8,
|
| 1486 |
"trial_name": null,
|
| 1487 |
"trial_params": null
|
| 1488 |
}
|