End of training
Browse files- README.md +19 -6
- all_results.json +11 -11
- eval_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +1132 -1132
README.md
CHANGED
|
@@ -3,23 +3,36 @@ library_name: transformers
|
|
| 3 |
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:
|
| 11 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
|
| 17 |
-
#
|
| 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.
|
| 22 |
-
- Wer:
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: openai/whisper-medium
|
| 5 |
tags:
|
| 6 |
+
- whisper-event
|
| 7 |
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- asierhv/composite_corpus_eu_v2.1
|
| 10 |
metrics:
|
| 11 |
- wer
|
| 12 |
model-index:
|
| 13 |
+
- name: Whisper Medium Basque
|
| 14 |
+
results:
|
| 15 |
+
- task:
|
| 16 |
+
name: Automatic Speech Recognition
|
| 17 |
+
type: automatic-speech-recognition
|
| 18 |
+
dataset:
|
| 19 |
+
name: asierhv/composite_corpus_eu_v2.1
|
| 20 |
+
type: asierhv/composite_corpus_eu_v2.1
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Wer
|
| 23 |
+
type: wer
|
| 24 |
+
value: 9.98410769374591
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 28 |
should probably proofread and complete it, then remove this comment. -->
|
| 29 |
|
| 30 |
+
# Whisper Medium Basque
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the asierhv/composite_corpus_eu_v2.1 dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
+
- Loss: 0.2191
|
| 35 |
+
- Wer: 9.9841
|
| 36 |
|
| 37 |
## Model description
|
| 38 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 5.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer":
|
| 8 |
-
"total_flos": 1.
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 0.21914222836494446,
|
| 4 |
+
"eval_runtime": 416.2982,
|
| 5 |
+
"eval_samples_per_second": 5.054,
|
| 6 |
+
"eval_steps_per_second": 0.632,
|
| 7 |
+
"eval_wer": 9.98410769374591,
|
| 8 |
+
"total_flos": 1.3063742816256e+20,
|
| 9 |
+
"train_loss": 0.1566779458373785,
|
| 10 |
+
"train_runtime": 20473.0101,
|
| 11 |
+
"train_samples_per_second": 6.252,
|
| 12 |
+
"train_steps_per_second": 0.391
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 5.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 0.21914222836494446,
|
| 4 |
+
"eval_runtime": 416.2982,
|
| 5 |
+
"eval_samples_per_second": 5.054,
|
| 6 |
+
"eval_steps_per_second": 0.632,
|
| 7 |
+
"eval_wer": 9.98410769374591
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 1.3063742816256e+20,
|
| 4 |
+
"train_loss": 0.1566779458373785,
|
| 5 |
+
"train_runtime": 20473.0101,
|
| 6 |
+
"train_samples_per_second": 6.252,
|
| 7 |
+
"train_steps_per_second": 0.391
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 8000,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -10,2396 +10,2396 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.003125,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 2.875e-07,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 25
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.00625,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 6.000000000000001e-07,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 50
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009375,
|
| 27 |
-
"grad_norm":
|
| 28 |
"learning_rate": 9.124999999999999e-07,
|
| 29 |
-
"loss":
|
| 30 |
"step": 75
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.0125,
|
| 34 |
-
"grad_norm": 9.
|
| 35 |
"learning_rate": 1.2250000000000001e-06,
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 100
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.015625,
|
| 41 |
-
"grad_norm": 9.
|
| 42 |
"learning_rate": 1.5375e-06,
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 125
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.01875,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 1.85e-06,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 150
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.021875,
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 2.1625e-06,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 175
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.025,
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 2.4750000000000004e-06,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 200
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.028125,
|
| 69 |
-
"grad_norm":
|
| 70 |
"learning_rate": 2.7875e-06,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 225
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03125,
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 3.1e-06,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 250
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.034375,
|
| 83 |
-
"grad_norm":
|
| 84 |
"learning_rate": 3.4125000000000004e-06,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 275
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0375,
|
| 90 |
-
"grad_norm": 6.
|
| 91 |
"learning_rate": 3.725e-06,
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 300
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.040625,
|
| 97 |
-
"grad_norm":
|
| 98 |
"learning_rate": 4.037500000000001e-06,
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 325
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.04375,
|
| 104 |
-
"grad_norm":
|
| 105 |
"learning_rate": 4.35e-06,
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 350
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.046875,
|
| 111 |
-
"grad_norm":
|
| 112 |
"learning_rate": 4.6625e-06,
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 375
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.05,
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 4.975000000000001e-06,
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.053125,
|
| 125 |
-
"grad_norm": 6.
|
| 126 |
"learning_rate": 5.2875e-06,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 425
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.05625,
|
| 132 |
-
"grad_norm":
|
| 133 |
"learning_rate": 5.600000000000001e-06,
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 450
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.059375,
|
| 139 |
-
"grad_norm":
|
| 140 |
"learning_rate": 5.9125e-06,
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 475
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.0625,
|
| 146 |
-
"grad_norm":
|
| 147 |
"learning_rate": 6.2250000000000005e-06,
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 500
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.0625,
|
| 153 |
-
"eval_loss": 0.
|
| 154 |
-
"eval_runtime":
|
| 155 |
-
"eval_samples_per_second":
|
| 156 |
-
"eval_steps_per_second": 0.
|
| 157 |
-
"eval_wer":
|
| 158 |
"step": 500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.065625,
|
| 162 |
-
"grad_norm":
|
| 163 |
"learning_rate": 6.230833333333334e-06,
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 525
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.06875,
|
| 169 |
-
"grad_norm": 6.
|
| 170 |
"learning_rate": 6.210000000000001e-06,
|
| 171 |
-
"loss": 0.
|
| 172 |
"step": 550
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.071875,
|
| 176 |
-
"grad_norm":
|
| 177 |
"learning_rate": 6.1891666666666666e-06,
|
| 178 |
-
"loss": 0.
|
| 179 |
"step": 575
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.075,
|
| 183 |
-
"grad_norm":
|
| 184 |
"learning_rate": 6.168333333333333e-06,
|
| 185 |
-
"loss": 0.
|
| 186 |
"step": 600
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.078125,
|
| 190 |
-
"grad_norm": 6.
|
| 191 |
"learning_rate": 6.147500000000001e-06,
|
| 192 |
-
"loss": 0.
|
| 193 |
"step": 625
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.08125,
|
| 197 |
-
"grad_norm":
|
| 198 |
"learning_rate": 6.126666666666667e-06,
|
| 199 |
-
"loss": 0.
|
| 200 |
"step": 650
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.084375,
|
| 204 |
-
"grad_norm": 5.
|
| 205 |
"learning_rate": 6.1058333333333335e-06,
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 675
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.0875,
|
| 211 |
-
"grad_norm":
|
| 212 |
"learning_rate": 6.085e-06,
|
| 213 |
-
"loss": 0.
|
| 214 |
"step": 700
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 0.090625,
|
| 218 |
-
"grad_norm":
|
| 219 |
"learning_rate": 6.064166666666667e-06,
|
| 220 |
-
"loss": 0.
|
| 221 |
"step": 725
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.09375,
|
| 225 |
-
"grad_norm":
|
| 226 |
"learning_rate": 6.043333333333334e-06,
|
| 227 |
-
"loss": 0.
|
| 228 |
"step": 750
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 0.096875,
|
| 232 |
-
"grad_norm":
|
| 233 |
"learning_rate": 6.0225e-06,
|
| 234 |
-
"loss": 0.
|
| 235 |
"step": 775
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.1,
|
| 239 |
-
"grad_norm":
|
| 240 |
"learning_rate": 6.001666666666667e-06,
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 800
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.103125,
|
| 246 |
-
"grad_norm":
|
| 247 |
"learning_rate": 5.980833333333334e-06,
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 825
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.10625,
|
| 253 |
-
"grad_norm":
|
| 254 |
"learning_rate": 5.9600000000000005e-06,
|
| 255 |
-
"loss": 0.
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 0.109375,
|
| 260 |
-
"grad_norm":
|
| 261 |
"learning_rate": 5.939166666666667e-06,
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 875
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.1125,
|
| 267 |
-
"grad_norm":
|
| 268 |
"learning_rate": 5.918333333333333e-06,
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.115625,
|
| 274 |
-
"grad_norm":
|
| 275 |
"learning_rate": 5.8975e-06,
|
| 276 |
-
"loss": 0.
|
| 277 |
"step": 925
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.11875,
|
| 281 |
-
"grad_norm":
|
| 282 |
"learning_rate": 5.8766666666666674e-06,
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 950
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.121875,
|
| 288 |
-
"grad_norm": 4.
|
| 289 |
"learning_rate": 5.855833333333333e-06,
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 975
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.125,
|
| 295 |
-
"grad_norm":
|
| 296 |
"learning_rate": 5.835e-06,
|
| 297 |
-
"loss": 0.
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 0.125,
|
| 302 |
-
"eval_loss": 0.
|
| 303 |
-
"eval_runtime":
|
| 304 |
-
"eval_samples_per_second":
|
| 305 |
-
"eval_steps_per_second": 0.
|
| 306 |
-
"eval_wer":
|
| 307 |
"step": 1000
|
| 308 |
},
|
| 309 |
{
|
| 310 |
"epoch": 0.128125,
|
| 311 |
-
"grad_norm":
|
| 312 |
"learning_rate": 5.814166666666667e-06,
|
| 313 |
-
"loss": 0.
|
| 314 |
"step": 1025
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.13125,
|
| 318 |
-
"grad_norm":
|
| 319 |
"learning_rate": 5.793333333333334e-06,
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 1050
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 0.134375,
|
| 325 |
-
"grad_norm":
|
| 326 |
"learning_rate": 5.7725e-06,
|
| 327 |
-
"loss": 0.
|
| 328 |
"step": 1075
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 0.1375,
|
| 332 |
-
"grad_norm":
|
| 333 |
"learning_rate": 5.751666666666667e-06,
|
| 334 |
-
"loss": 0.
|
| 335 |
"step": 1100
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.140625,
|
| 339 |
-
"grad_norm":
|
| 340 |
"learning_rate": 5.730833333333334e-06,
|
| 341 |
-
"loss": 0.
|
| 342 |
"step": 1125
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 0.14375,
|
| 346 |
-
"grad_norm": 4.
|
| 347 |
"learning_rate": 5.71e-06,
|
| 348 |
-
"loss": 0.
|
| 349 |
"step": 1150
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 0.146875,
|
| 353 |
-
"grad_norm":
|
| 354 |
"learning_rate": 5.689166666666667e-06,
|
| 355 |
-
"loss": 0.
|
| 356 |
"step": 1175
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.15,
|
| 360 |
-
"grad_norm":
|
| 361 |
"learning_rate": 5.668333333333334e-06,
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 1200
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 0.153125,
|
| 367 |
-
"grad_norm":
|
| 368 |
"learning_rate": 5.6475e-06,
|
| 369 |
-
"loss": 0.
|
| 370 |
"step": 1225
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 0.15625,
|
| 374 |
-
"grad_norm":
|
| 375 |
"learning_rate": 5.626666666666667e-06,
|
| 376 |
-
"loss": 0.
|
| 377 |
"step": 1250
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.159375,
|
| 381 |
-
"grad_norm":
|
| 382 |
"learning_rate": 5.605833333333334e-06,
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 1275
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.1625,
|
| 388 |
-
"grad_norm":
|
| 389 |
"learning_rate": 5.585e-06,
|
| 390 |
-
"loss": 0.
|
| 391 |
"step": 1300
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 0.165625,
|
| 395 |
-
"grad_norm":
|
| 396 |
"learning_rate": 5.564166666666667e-06,
|
| 397 |
-
"loss": 0.
|
| 398 |
"step": 1325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.16875,
|
| 402 |
-
"grad_norm": 4.
|
| 403 |
"learning_rate": 5.543333333333333e-06,
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 1350
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.171875,
|
| 409 |
-
"grad_norm":
|
| 410 |
"learning_rate": 5.522500000000001e-06,
|
| 411 |
-
"loss": 0.
|
| 412 |
"step": 1375
|
| 413 |
},
|
| 414 |
{
|
| 415 |
"epoch": 0.175,
|
| 416 |
-
"grad_norm":
|
| 417 |
"learning_rate": 5.501666666666667e-06,
|
| 418 |
-
"loss": 0.
|
| 419 |
"step": 1400
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 0.178125,
|
| 423 |
-
"grad_norm": 3.
|
| 424 |
"learning_rate": 5.4808333333333335e-06,
|
| 425 |
-
"loss": 0.
|
| 426 |
"step": 1425
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 0.18125,
|
| 430 |
-
"grad_norm":
|
| 431 |
"learning_rate": 5.46e-06,
|
| 432 |
-
"loss": 0.
|
| 433 |
"step": 1450
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 0.184375,
|
| 437 |
-
"grad_norm":
|
| 438 |
"learning_rate": 5.439166666666667e-06,
|
| 439 |
-
"loss": 0.
|
| 440 |
"step": 1475
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.1875,
|
| 444 |
-
"grad_norm": 5.
|
| 445 |
"learning_rate": 5.418333333333334e-06,
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 1500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 0.1875,
|
| 451 |
-
"eval_loss": 0.
|
| 452 |
-
"eval_runtime":
|
| 453 |
-
"eval_samples_per_second":
|
| 454 |
-
"eval_steps_per_second": 0.
|
| 455 |
-
"eval_wer":
|
| 456 |
"step": 1500
|
| 457 |
},
|
| 458 |
{
|
| 459 |
"epoch": 0.190625,
|
| 460 |
-
"grad_norm":
|
| 461 |
"learning_rate": 5.3975e-06,
|
| 462 |
-
"loss": 0.
|
| 463 |
"step": 1525
|
| 464 |
},
|
| 465 |
{
|
| 466 |
"epoch": 0.19375,
|
| 467 |
-
"grad_norm": 5.
|
| 468 |
"learning_rate": 5.376666666666666e-06,
|
| 469 |
-
"loss": 0.
|
| 470 |
"step": 1550
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.196875,
|
| 474 |
-
"grad_norm":
|
| 475 |
"learning_rate": 5.355833333333334e-06,
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 1575
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 0.2,
|
| 481 |
-
"grad_norm":
|
| 482 |
"learning_rate": 5.335000000000001e-06,
|
| 483 |
-
"loss": 0.
|
| 484 |
"step": 1600
|
| 485 |
},
|
| 486 |
{
|
| 487 |
"epoch": 0.203125,
|
| 488 |
-
"grad_norm":
|
| 489 |
"learning_rate": 5.3141666666666665e-06,
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 1625
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.20625,
|
| 495 |
-
"grad_norm":
|
| 496 |
"learning_rate": 5.293333333333333e-06,
|
| 497 |
-
"loss": 0.
|
| 498 |
"step": 1650
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 0.209375,
|
| 502 |
-
"grad_norm":
|
| 503 |
"learning_rate": 5.272500000000001e-06,
|
| 504 |
-
"loss": 0.
|
| 505 |
"step": 1675
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.2125,
|
| 509 |
-
"grad_norm":
|
| 510 |
"learning_rate": 5.2516666666666675e-06,
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 1700
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"grad_norm":
|
| 517 |
"learning_rate": 5.230833333333333e-06,
|
| 518 |
-
"loss": 0.
|
| 519 |
"step": 1725
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"grad_norm":
|
| 524 |
"learning_rate": 5.21e-06,
|
| 525 |
-
"loss": 0.
|
| 526 |
"step": 1750
|
| 527 |
},
|
| 528 |
{
|
| 529 |
-
"epoch":
|
| 530 |
-
"grad_norm":
|
| 531 |
"learning_rate": 5.189166666666667e-06,
|
| 532 |
-
"loss": 0.
|
| 533 |
"step": 1775
|
| 534 |
},
|
| 535 |
{
|
| 536 |
-
"epoch":
|
| 537 |
-
"grad_norm": 4.
|
| 538 |
"learning_rate": 5.1683333333333335e-06,
|
| 539 |
-
"loss": 0.
|
| 540 |
"step": 1800
|
| 541 |
},
|
| 542 |
{
|
| 543 |
-
"epoch":
|
| 544 |
-
"grad_norm":
|
| 545 |
"learning_rate": 5.1475e-06,
|
| 546 |
-
"loss": 0.
|
| 547 |
"step": 1825
|
| 548 |
},
|
| 549 |
{
|
| 550 |
-
"epoch":
|
| 551 |
-
"grad_norm":
|
| 552 |
"learning_rate": 5.126666666666667e-06,
|
| 553 |
-
"loss": 0.
|
| 554 |
"step": 1850
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch":
|
| 558 |
-
"grad_norm":
|
| 559 |
"learning_rate": 5.105833333333334e-06,
|
| 560 |
-
"loss": 0.
|
| 561 |
"step": 1875
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
-
"grad_norm": 5.
|
| 566 |
"learning_rate": 5.0850000000000004e-06,
|
| 567 |
-
"loss": 0.
|
| 568 |
"step": 1900
|
| 569 |
},
|
| 570 |
{
|
| 571 |
-
"epoch":
|
| 572 |
-
"grad_norm":
|
| 573 |
"learning_rate": 5.064166666666667e-06,
|
| 574 |
-
"loss": 0.
|
| 575 |
"step": 1925
|
| 576 |
},
|
| 577 |
{
|
| 578 |
-
"epoch":
|
| 579 |
-
"grad_norm":
|
| 580 |
"learning_rate": 5.043333333333333e-06,
|
| 581 |
-
"loss": 0.
|
| 582 |
"step": 1950
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"epoch":
|
| 586 |
-
"grad_norm":
|
| 587 |
"learning_rate": 5.0225e-06,
|
| 588 |
-
"loss": 0.
|
| 589 |
"step": 1975
|
| 590 |
},
|
| 591 |
{
|
| 592 |
-
"epoch":
|
| 593 |
-
"grad_norm": 6.
|
| 594 |
"learning_rate": 5.001666666666667e-06,
|
| 595 |
-
"loss": 0.
|
| 596 |
"step": 2000
|
| 597 |
},
|
| 598 |
{
|
| 599 |
-
"epoch":
|
| 600 |
-
"eval_loss": 0.
|
| 601 |
-
"eval_runtime":
|
| 602 |
-
"eval_samples_per_second":
|
| 603 |
-
"eval_steps_per_second": 0.
|
| 604 |
-
"eval_wer":
|
| 605 |
"step": 2000
|
| 606 |
},
|
| 607 |
{
|
| 608 |
-
"epoch":
|
| 609 |
-
"grad_norm":
|
| 610 |
"learning_rate": 4.980833333333334e-06,
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 2025
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch":
|
| 616 |
-
"grad_norm":
|
| 617 |
"learning_rate": 4.96e-06,
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 2050
|
| 620 |
},
|
| 621 |
{
|
| 622 |
-
"epoch":
|
| 623 |
-
"grad_norm": 4.
|
| 624 |
"learning_rate": 4.939166666666667e-06,
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 2075
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"grad_norm":
|
| 631 |
"learning_rate": 4.918333333333334e-06,
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 2100
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"grad_norm":
|
| 638 |
"learning_rate": 4.8975e-06,
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 2125
|
| 641 |
},
|
| 642 |
{
|
| 643 |
-
"epoch":
|
| 644 |
-
"grad_norm":
|
| 645 |
"learning_rate": 4.876666666666667e-06,
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 2150
|
| 648 |
},
|
| 649 |
{
|
| 650 |
-
"epoch":
|
| 651 |
-
"grad_norm":
|
| 652 |
"learning_rate": 4.8558333333333336e-06,
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 2175
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch":
|
| 658 |
-
"grad_norm":
|
| 659 |
"learning_rate": 4.835e-06,
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 2200
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"grad_norm":
|
| 666 |
"learning_rate": 4.814166666666667e-06,
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 2225
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"grad_norm":
|
| 673 |
"learning_rate": 4.793333333333334e-06,
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 2250
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch":
|
| 679 |
-
"grad_norm": 4.
|
| 680 |
"learning_rate": 4.7725e-06,
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 2275
|
| 683 |
},
|
| 684 |
{
|
| 685 |
-
"epoch":
|
| 686 |
-
"grad_norm":
|
| 687 |
"learning_rate": 4.751666666666666e-06,
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 2300
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"grad_norm":
|
| 694 |
"learning_rate": 4.730833333333334e-06,
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 2325
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch":
|
| 700 |
-
"grad_norm":
|
| 701 |
"learning_rate": 4.710000000000001e-06,
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 2350
|
| 704 |
},
|
| 705 |
{
|
| 706 |
-
"epoch":
|
| 707 |
-
"grad_norm":
|
| 708 |
"learning_rate": 4.6891666666666665e-06,
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 2375
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"grad_norm":
|
| 715 |
"learning_rate": 4.668333333333333e-06,
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 2400
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
-
"grad_norm": 3.
|
| 722 |
"learning_rate": 4.647500000000001e-06,
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 2425
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch":
|
| 728 |
-
"grad_norm":
|
| 729 |
"learning_rate": 4.626666666666667e-06,
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 2450
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch":
|
| 735 |
-
"grad_norm":
|
| 736 |
"learning_rate": 4.605833333333333e-06,
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 2475
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch":
|
| 742 |
-
"grad_norm":
|
| 743 |
"learning_rate": 4.585e-06,
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 2500
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch":
|
| 749 |
-
"eval_loss": 0.
|
| 750 |
-
"eval_runtime":
|
| 751 |
-
"eval_samples_per_second":
|
| 752 |
-
"eval_steps_per_second": 0.
|
| 753 |
-
"eval_wer":
|
| 754 |
"step": 2500
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"epoch":
|
| 758 |
-
"grad_norm":
|
| 759 |
"learning_rate": 4.564166666666667e-06,
|
| 760 |
-
"loss": 0.
|
| 761 |
"step": 2525
|
| 762 |
},
|
| 763 |
{
|
| 764 |
-
"epoch":
|
| 765 |
-
"grad_norm":
|
| 766 |
"learning_rate": 4.543333333333334e-06,
|
| 767 |
-
"loss": 0.
|
| 768 |
"step": 2550
|
| 769 |
},
|
| 770 |
{
|
| 771 |
-
"epoch":
|
| 772 |
-
"grad_norm":
|
| 773 |
"learning_rate": 4.5225e-06,
|
| 774 |
-
"loss": 0.
|
| 775 |
"step": 2575
|
| 776 |
},
|
| 777 |
{
|
| 778 |
-
"epoch":
|
| 779 |
-
"grad_norm": 5.
|
| 780 |
"learning_rate": 4.501666666666667e-06,
|
| 781 |
-
"loss": 0.
|
| 782 |
"step": 2600
|
| 783 |
},
|
| 784 |
{
|
| 785 |
-
"epoch":
|
| 786 |
-
"grad_norm":
|
| 787 |
-
"learning_rate": 4.
|
| 788 |
-
"loss": 0.
|
| 789 |
"step": 2625
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch":
|
| 793 |
-
"grad_norm":
|
| 794 |
-
"learning_rate": 4.
|
| 795 |
-
"loss": 0.
|
| 796 |
"step": 2650
|
| 797 |
},
|
| 798 |
{
|
| 799 |
-
"epoch":
|
| 800 |
-
"grad_norm":
|
| 801 |
-
"learning_rate": 4.
|
| 802 |
-
"loss": 0.
|
| 803 |
"step": 2675
|
| 804 |
},
|
| 805 |
{
|
| 806 |
-
"epoch":
|
| 807 |
-
"grad_norm":
|
| 808 |
-
"learning_rate": 4.
|
| 809 |
-
"loss": 0.
|
| 810 |
"step": 2700
|
| 811 |
},
|
| 812 |
{
|
| 813 |
-
"epoch":
|
| 814 |
-
"grad_norm":
|
| 815 |
-
"learning_rate": 4.
|
| 816 |
-
"loss": 0.
|
| 817 |
"step": 2725
|
| 818 |
},
|
| 819 |
{
|
| 820 |
-
"epoch":
|
| 821 |
-
"grad_norm": 5.
|
| 822 |
-
"learning_rate": 4.
|
| 823 |
-
"loss": 0.
|
| 824 |
"step": 2750
|
| 825 |
},
|
| 826 |
{
|
| 827 |
-
"epoch":
|
| 828 |
-
"grad_norm":
|
| 829 |
-
"learning_rate": 4.
|
| 830 |
-
"loss": 0.
|
| 831 |
"step": 2775
|
| 832 |
},
|
| 833 |
{
|
| 834 |
-
"epoch":
|
| 835 |
-
"grad_norm":
|
| 836 |
-
"learning_rate": 4.
|
| 837 |
-
"loss": 0.
|
| 838 |
"step": 2800
|
| 839 |
},
|
| 840 |
{
|
| 841 |
-
"epoch":
|
| 842 |
-
"grad_norm":
|
| 843 |
-
"learning_rate": 4.
|
| 844 |
-
"loss": 0.
|
| 845 |
"step": 2825
|
| 846 |
},
|
| 847 |
{
|
| 848 |
-
"epoch":
|
| 849 |
-
"grad_norm":
|
| 850 |
-
"learning_rate": 4.
|
| 851 |
-
"loss": 0.
|
| 852 |
"step": 2850
|
| 853 |
},
|
| 854 |
{
|
| 855 |
-
"epoch":
|
| 856 |
-
"grad_norm":
|
| 857 |
-
"learning_rate": 4.
|
| 858 |
-
"loss": 0.
|
| 859 |
"step": 2875
|
| 860 |
},
|
| 861 |
{
|
| 862 |
-
"epoch":
|
| 863 |
-
"grad_norm": 4.
|
| 864 |
-
"learning_rate": 4.
|
| 865 |
-
"loss": 0.
|
| 866 |
"step": 2900
|
| 867 |
},
|
| 868 |
{
|
| 869 |
-
"epoch":
|
| 870 |
-
"grad_norm":
|
| 871 |
-
"learning_rate": 4.
|
| 872 |
-
"loss": 0.
|
| 873 |
"step": 2925
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch":
|
| 877 |
-
"grad_norm":
|
| 878 |
-
"learning_rate": 4.
|
| 879 |
-
"loss": 0.
|
| 880 |
"step": 2950
|
| 881 |
},
|
| 882 |
{
|
| 883 |
-
"epoch":
|
| 884 |
-
"grad_norm":
|
| 885 |
-
"learning_rate": 4.
|
| 886 |
-
"loss": 0.
|
| 887 |
"step": 2975
|
| 888 |
},
|
| 889 |
{
|
| 890 |
-
"epoch":
|
| 891 |
-
"grad_norm": 3.
|
| 892 |
-
"learning_rate": 4.
|
| 893 |
-
"loss": 0.
|
| 894 |
"step": 3000
|
| 895 |
},
|
| 896 |
{
|
| 897 |
-
"epoch":
|
| 898 |
-
"eval_loss": 0.
|
| 899 |
-
"eval_runtime":
|
| 900 |
-
"eval_samples_per_second":
|
| 901 |
-
"eval_steps_per_second": 0.
|
| 902 |
-
"eval_wer":
|
| 903 |
"step": 3000
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch":
|
| 907 |
-
"grad_norm":
|
| 908 |
-
"learning_rate": 4.
|
| 909 |
-
"loss": 0.
|
| 910 |
"step": 3025
|
| 911 |
},
|
| 912 |
{
|
| 913 |
-
"epoch":
|
| 914 |
-
"grad_norm": 4.
|
| 915 |
-
"learning_rate": 4.
|
| 916 |
-
"loss": 0.
|
| 917 |
"step": 3050
|
| 918 |
},
|
| 919 |
{
|
| 920 |
-
"epoch":
|
| 921 |
-
"grad_norm":
|
| 922 |
-
"learning_rate": 4.
|
| 923 |
-
"loss": 0.
|
| 924 |
"step": 3075
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"epoch":
|
| 928 |
-
"grad_norm":
|
| 929 |
-
"learning_rate": 4.
|
| 930 |
-
"loss": 0.
|
| 931 |
"step": 3100
|
| 932 |
},
|
| 933 |
{
|
| 934 |
-
"epoch":
|
| 935 |
-
"grad_norm":
|
| 936 |
-
"learning_rate": 4.
|
| 937 |
-
"loss": 0.
|
| 938 |
"step": 3125
|
| 939 |
},
|
| 940 |
{
|
| 941 |
-
"epoch":
|
| 942 |
-
"grad_norm":
|
| 943 |
-
"learning_rate": 4.
|
| 944 |
-
"loss": 0.
|
| 945 |
"step": 3150
|
| 946 |
},
|
| 947 |
{
|
| 948 |
-
"epoch":
|
| 949 |
-
"grad_norm":
|
| 950 |
-
"learning_rate": 4.
|
| 951 |
-
"loss": 0.
|
| 952 |
"step": 3175
|
| 953 |
},
|
| 954 |
{
|
| 955 |
-
"epoch":
|
| 956 |
-
"grad_norm":
|
| 957 |
-
"learning_rate": 4.
|
| 958 |
-
"loss": 0.
|
| 959 |
"step": 3200
|
| 960 |
},
|
| 961 |
{
|
| 962 |
-
"epoch":
|
| 963 |
-
"grad_norm":
|
| 964 |
-
"learning_rate": 3.
|
| 965 |
-
"loss": 0.
|
| 966 |
"step": 3225
|
| 967 |
},
|
| 968 |
{
|
| 969 |
-
"epoch":
|
| 970 |
-
"grad_norm":
|
| 971 |
-
"learning_rate": 3.
|
| 972 |
-
"loss": 0.
|
| 973 |
"step": 3250
|
| 974 |
},
|
| 975 |
{
|
| 976 |
-
"epoch":
|
| 977 |
-
"grad_norm":
|
| 978 |
-
"learning_rate": 3.
|
| 979 |
-
"loss": 0.
|
| 980 |
"step": 3275
|
| 981 |
},
|
| 982 |
{
|
| 983 |
-
"epoch":
|
| 984 |
-
"grad_norm":
|
| 985 |
-
"learning_rate": 3.
|
| 986 |
-
"loss": 0.
|
| 987 |
"step": 3300
|
| 988 |
},
|
| 989 |
{
|
| 990 |
-
"epoch":
|
| 991 |
-
"grad_norm":
|
| 992 |
-
"learning_rate": 3.
|
| 993 |
-
"loss": 0.
|
| 994 |
"step": 3325
|
| 995 |
},
|
| 996 |
{
|
| 997 |
-
"epoch":
|
| 998 |
-
"grad_norm":
|
| 999 |
-
"learning_rate": 3.
|
| 1000 |
-
"loss": 0.
|
| 1001 |
"step": 3350
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
-
"epoch":
|
| 1005 |
-
"grad_norm":
|
| 1006 |
-
"learning_rate": 3.
|
| 1007 |
-
"loss": 0.
|
| 1008 |
"step": 3375
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
-
"epoch":
|
| 1012 |
-
"grad_norm": 3.
|
| 1013 |
-
"learning_rate": 3.
|
| 1014 |
-
"loss": 0.
|
| 1015 |
"step": 3400
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
-
"epoch":
|
| 1019 |
-
"grad_norm":
|
| 1020 |
-
"learning_rate": 3.
|
| 1021 |
-
"loss": 0.
|
| 1022 |
"step": 3425
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
-
"epoch":
|
| 1026 |
-
"grad_norm":
|
| 1027 |
-
"learning_rate": 3.
|
| 1028 |
-
"loss": 0.
|
| 1029 |
"step": 3450
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
-
"epoch":
|
| 1033 |
-
"grad_norm":
|
| 1034 |
-
"learning_rate": 3.
|
| 1035 |
-
"loss": 0.
|
| 1036 |
"step": 3475
|
| 1037 |
},
|
| 1038 |
{
|
| 1039 |
-
"epoch":
|
| 1040 |
-
"grad_norm": 3.
|
| 1041 |
-
"learning_rate": 3.
|
| 1042 |
-
"loss": 0.
|
| 1043 |
"step": 3500
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
-
"epoch":
|
| 1047 |
-
"eval_loss": 0.
|
| 1048 |
-
"eval_runtime":
|
| 1049 |
-
"eval_samples_per_second":
|
| 1050 |
-
"eval_steps_per_second": 0.
|
| 1051 |
-
"eval_wer":
|
| 1052 |
"step": 3500
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
-
"epoch":
|
| 1056 |
-
"grad_norm": 3.
|
| 1057 |
-
"learning_rate": 3.
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 3525
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch":
|
| 1063 |
-
"grad_norm":
|
| 1064 |
-
"learning_rate": 3.
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 3550
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
-
"epoch":
|
| 1070 |
-
"grad_norm": 3.
|
| 1071 |
-
"learning_rate": 3.
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 3575
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch":
|
| 1077 |
-
"grad_norm":
|
| 1078 |
-
"learning_rate": 3.
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 3600
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch":
|
| 1084 |
-
"grad_norm":
|
| 1085 |
-
"learning_rate": 3.
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 3625
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
-
"epoch":
|
| 1091 |
-
"grad_norm":
|
| 1092 |
-
"learning_rate": 3.
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 3650
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch":
|
| 1098 |
-
"grad_norm": 3.
|
| 1099 |
-
"learning_rate": 3.
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 3675
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch":
|
| 1105 |
-
"grad_norm": 3.
|
| 1106 |
-
"learning_rate": 3.
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 3700
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
-
"epoch":
|
| 1112 |
-
"grad_norm": 2.
|
| 1113 |
-
"learning_rate": 3.
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 3725
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
-
"epoch":
|
| 1119 |
-
"grad_norm":
|
| 1120 |
-
"learning_rate": 3.
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 3750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
-
"epoch":
|
| 1126 |
-
"grad_norm":
|
| 1127 |
-
"learning_rate": 3.
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 3775
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch":
|
| 1133 |
-
"grad_norm": 2.
|
| 1134 |
-
"learning_rate": 3.
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 3800
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch":
|
| 1140 |
-
"grad_norm":
|
| 1141 |
-
"learning_rate": 3.
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 3825
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch":
|
| 1147 |
-
"grad_norm":
|
| 1148 |
-
"learning_rate": 3.
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 3850
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
-
"epoch":
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate": 3.
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 3875
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
-
"epoch":
|
| 1161 |
-
"grad_norm":
|
| 1162 |
-
"learning_rate": 3.
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 3900
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
-
"epoch":
|
| 1168 |
-
"grad_norm":
|
| 1169 |
-
"learning_rate": 3.
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 3925
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
-
"epoch":
|
| 1175 |
-
"grad_norm":
|
| 1176 |
-
"learning_rate": 3.
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 3950
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch":
|
| 1182 |
-
"grad_norm":
|
| 1183 |
-
"learning_rate": 3.
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 3975
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch":
|
| 1189 |
-
"grad_norm":
|
| 1190 |
-
"learning_rate": 3.
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 4000
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
-
"epoch":
|
| 1196 |
-
"eval_loss": 0.
|
| 1197 |
-
"eval_runtime":
|
| 1198 |
-
"eval_samples_per_second": 5.
|
| 1199 |
-
"eval_steps_per_second": 0.
|
| 1200 |
-
"eval_wer":
|
| 1201 |
"step": 4000
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
-
"epoch":
|
| 1205 |
-
"grad_norm":
|
| 1206 |
-
"learning_rate": 3.
|
| 1207 |
-
"loss": 0.
|
| 1208 |
"step": 4025
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
-
"epoch":
|
| 1212 |
-
"grad_norm":
|
| 1213 |
-
"learning_rate": 3.
|
| 1214 |
-
"loss": 0.
|
| 1215 |
"step": 4050
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
-
"epoch":
|
| 1219 |
-
"grad_norm": 3.
|
| 1220 |
-
"learning_rate": 3.
|
| 1221 |
-
"loss": 0.
|
| 1222 |
"step": 4075
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
-
"epoch":
|
| 1226 |
-
"grad_norm":
|
| 1227 |
-
"learning_rate": 3.
|
| 1228 |
-
"loss": 0.
|
| 1229 |
"step": 4100
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
-
"epoch":
|
| 1233 |
-
"grad_norm":
|
| 1234 |
-
"learning_rate": 3.
|
| 1235 |
-
"loss": 0.
|
| 1236 |
"step": 4125
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
-
"epoch":
|
| 1240 |
-
"grad_norm":
|
| 1241 |
-
"learning_rate": 3.
|
| 1242 |
-
"loss": 0.
|
| 1243 |
"step": 4150
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
-
"epoch":
|
| 1247 |
-
"grad_norm":
|
| 1248 |
-
"learning_rate": 3.
|
| 1249 |
-
"loss": 0.
|
| 1250 |
"step": 4175
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
-
"epoch":
|
| 1254 |
-
"grad_norm":
|
| 1255 |
-
"learning_rate": 3.
|
| 1256 |
-
"loss": 0.
|
| 1257 |
"step": 4200
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
-
"epoch":
|
| 1261 |
-
"grad_norm":
|
| 1262 |
-
"learning_rate": 3.
|
| 1263 |
-
"loss": 0.
|
| 1264 |
"step": 4225
|
| 1265 |
},
|
| 1266 |
{
|
| 1267 |
-
"epoch":
|
| 1268 |
-
"grad_norm":
|
| 1269 |
-
"learning_rate": 3.
|
| 1270 |
-
"loss": 0.
|
| 1271 |
"step": 4250
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
-
"epoch":
|
| 1275 |
-
"grad_norm":
|
| 1276 |
-
"learning_rate": 3.
|
| 1277 |
-
"loss": 0.
|
| 1278 |
"step": 4275
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
-
"epoch":
|
| 1282 |
-
"grad_norm": 3.
|
| 1283 |
-
"learning_rate": 3.
|
| 1284 |
-
"loss": 0.
|
| 1285 |
"step": 4300
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
-
"epoch":
|
| 1289 |
-
"grad_norm":
|
| 1290 |
-
"learning_rate": 3.
|
| 1291 |
-
"loss": 0.
|
| 1292 |
"step": 4325
|
| 1293 |
},
|
| 1294 |
{
|
| 1295 |
-
"epoch":
|
| 1296 |
-
"grad_norm":
|
| 1297 |
-
"learning_rate": 3.
|
| 1298 |
-
"loss": 0.
|
| 1299 |
"step": 4350
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
-
"epoch":
|
| 1303 |
-
"grad_norm": 4.
|
| 1304 |
-
"learning_rate": 3.
|
| 1305 |
-
"loss": 0.
|
| 1306 |
"step": 4375
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
-
"epoch":
|
| 1310 |
-
"grad_norm":
|
| 1311 |
-
"learning_rate": 3.
|
| 1312 |
-
"loss": 0.
|
| 1313 |
"step": 4400
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
-
"epoch":
|
| 1317 |
-
"grad_norm":
|
| 1318 |
-
"learning_rate": 2.
|
| 1319 |
-
"loss": 0.
|
| 1320 |
"step": 4425
|
| 1321 |
},
|
| 1322 |
{
|
| 1323 |
-
"epoch":
|
| 1324 |
-
"grad_norm":
|
| 1325 |
-
"learning_rate": 2.
|
| 1326 |
-
"loss": 0.
|
| 1327 |
"step": 4450
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
-
"epoch":
|
| 1331 |
-
"grad_norm":
|
| 1332 |
-
"learning_rate": 2.
|
| 1333 |
-
"loss": 0.
|
| 1334 |
"step": 4475
|
| 1335 |
},
|
| 1336 |
{
|
| 1337 |
-
"epoch":
|
| 1338 |
-
"grad_norm":
|
| 1339 |
-
"learning_rate": 2.
|
| 1340 |
-
"loss": 0.
|
| 1341 |
"step": 4500
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
-
"epoch":
|
| 1345 |
-
"eval_loss": 0.
|
| 1346 |
-
"eval_runtime":
|
| 1347 |
-
"eval_samples_per_second": 5.
|
| 1348 |
-
"eval_steps_per_second": 0.
|
| 1349 |
-
"eval_wer":
|
| 1350 |
"step": 4500
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
-
"epoch":
|
| 1354 |
-
"grad_norm":
|
| 1355 |
-
"learning_rate": 2.
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 4525
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
-
"epoch":
|
| 1361 |
-
"grad_norm":
|
| 1362 |
-
"learning_rate": 2.
|
| 1363 |
-
"loss": 0.
|
| 1364 |
"step": 4550
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
-
"epoch":
|
| 1368 |
-
"grad_norm": 2.
|
| 1369 |
-
"learning_rate": 2.
|
| 1370 |
-
"loss": 0.
|
| 1371 |
"step": 4575
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
-
"epoch":
|
| 1375 |
-
"grad_norm":
|
| 1376 |
-
"learning_rate": 2.
|
| 1377 |
-
"loss": 0.
|
| 1378 |
"step": 4600
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
-
"epoch":
|
| 1382 |
-
"grad_norm": 3.
|
| 1383 |
-
"learning_rate": 2.
|
| 1384 |
-
"loss": 0.
|
| 1385 |
"step": 4625
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
-
"epoch":
|
| 1389 |
-
"grad_norm":
|
| 1390 |
-
"learning_rate": 2.
|
| 1391 |
-
"loss": 0.
|
| 1392 |
"step": 4650
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
-
"epoch":
|
| 1396 |
-
"grad_norm": 2.
|
| 1397 |
-
"learning_rate": 2.
|
| 1398 |
-
"loss": 0.
|
| 1399 |
"step": 4675
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
-
"epoch":
|
| 1403 |
-
"grad_norm": 2.
|
| 1404 |
-
"learning_rate": 2.
|
| 1405 |
-
"loss": 0.
|
| 1406 |
"step": 4700
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
-
"epoch":
|
| 1410 |
-
"grad_norm": 3.
|
| 1411 |
-
"learning_rate": 2.
|
| 1412 |
-
"loss": 0.
|
| 1413 |
"step": 4725
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
-
"epoch":
|
| 1417 |
-
"grad_norm": 2.
|
| 1418 |
-
"learning_rate": 2.
|
| 1419 |
-
"loss": 0.
|
| 1420 |
"step": 4750
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
-
"epoch":
|
| 1424 |
-
"grad_norm": 2.
|
| 1425 |
-
"learning_rate": 2.
|
| 1426 |
-
"loss": 0.
|
| 1427 |
"step": 4775
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
-
"epoch":
|
| 1431 |
-
"grad_norm": 2.
|
| 1432 |
-
"learning_rate": 2.
|
| 1433 |
-
"loss": 0.
|
| 1434 |
"step": 4800
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
-
"epoch":
|
| 1438 |
-
"grad_norm":
|
| 1439 |
-
"learning_rate": 2.
|
| 1440 |
-
"loss": 0.
|
| 1441 |
"step": 4825
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
-
"epoch":
|
| 1445 |
-
"grad_norm": 2.
|
| 1446 |
-
"learning_rate": 2.
|
| 1447 |
-
"loss": 0.
|
| 1448 |
"step": 4850
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
-
"epoch":
|
| 1452 |
-
"grad_norm":
|
| 1453 |
-
"learning_rate": 2.
|
| 1454 |
-
"loss": 0.
|
| 1455 |
"step": 4875
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
-
"epoch":
|
| 1459 |
-
"grad_norm":
|
| 1460 |
-
"learning_rate": 2.
|
| 1461 |
-
"loss": 0.
|
| 1462 |
"step": 4900
|
| 1463 |
},
|
| 1464 |
{
|
| 1465 |
-
"epoch":
|
| 1466 |
-
"grad_norm":
|
| 1467 |
-
"learning_rate": 2.
|
| 1468 |
-
"loss": 0.
|
| 1469 |
"step": 4925
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
-
"epoch":
|
| 1473 |
-
"grad_norm": 2.
|
| 1474 |
-
"learning_rate": 2.
|
| 1475 |
-
"loss": 0.
|
| 1476 |
"step": 4950
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
-
"epoch":
|
| 1480 |
-
"grad_norm": 2.
|
| 1481 |
-
"learning_rate": 2.
|
| 1482 |
-
"loss": 0.
|
| 1483 |
"step": 4975
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
-
"epoch":
|
| 1487 |
-
"grad_norm":
|
| 1488 |
-
"learning_rate": 2.
|
| 1489 |
-
"loss": 0.
|
| 1490 |
"step": 5000
|
| 1491 |
},
|
| 1492 |
{
|
| 1493 |
-
"epoch":
|
| 1494 |
-
"eval_loss": 0.
|
| 1495 |
-
"eval_runtime":
|
| 1496 |
-
"eval_samples_per_second":
|
| 1497 |
-
"eval_steps_per_second": 0.
|
| 1498 |
-
"eval_wer":
|
| 1499 |
"step": 5000
|
| 1500 |
},
|
| 1501 |
{
|
| 1502 |
-
"epoch":
|
| 1503 |
-
"grad_norm": 3.
|
| 1504 |
-
"learning_rate": 2.
|
| 1505 |
-
"loss": 0.
|
| 1506 |
"step": 5025
|
| 1507 |
},
|
| 1508 |
{
|
| 1509 |
-
"epoch":
|
| 1510 |
-
"grad_norm":
|
| 1511 |
-
"learning_rate": 2.
|
| 1512 |
-
"loss": 0.
|
| 1513 |
"step": 5050
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
-
"epoch":
|
| 1517 |
-
"grad_norm":
|
| 1518 |
-
"learning_rate": 2.
|
| 1519 |
-
"loss": 0.
|
| 1520 |
"step": 5075
|
| 1521 |
},
|
| 1522 |
{
|
| 1523 |
-
"epoch":
|
| 1524 |
-
"grad_norm":
|
| 1525 |
-
"learning_rate": 2.
|
| 1526 |
-
"loss": 0.
|
| 1527 |
"step": 5100
|
| 1528 |
},
|
| 1529 |
{
|
| 1530 |
-
"epoch":
|
| 1531 |
-
"grad_norm":
|
| 1532 |
-
"learning_rate": 2.
|
| 1533 |
-
"loss": 0.
|
| 1534 |
"step": 5125
|
| 1535 |
},
|
| 1536 |
{
|
| 1537 |
-
"epoch":
|
| 1538 |
-
"grad_norm":
|
| 1539 |
-
"learning_rate": 2.
|
| 1540 |
-
"loss": 0.
|
| 1541 |
"step": 5150
|
| 1542 |
},
|
| 1543 |
{
|
| 1544 |
-
"epoch":
|
| 1545 |
-
"grad_norm":
|
| 1546 |
-
"learning_rate": 2.
|
| 1547 |
-
"loss": 0.
|
| 1548 |
"step": 5175
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
-
"epoch":
|
| 1552 |
-
"grad_norm":
|
| 1553 |
-
"learning_rate": 2.
|
| 1554 |
-
"loss": 0.
|
| 1555 |
"step": 5200
|
| 1556 |
},
|
| 1557 |
{
|
| 1558 |
-
"epoch":
|
| 1559 |
-
"grad_norm":
|
| 1560 |
-
"learning_rate": 2.
|
| 1561 |
-
"loss": 0.
|
| 1562 |
"step": 5225
|
| 1563 |
},
|
| 1564 |
{
|
| 1565 |
-
"epoch":
|
| 1566 |
-
"grad_norm":
|
| 1567 |
-
"learning_rate": 2.
|
| 1568 |
-
"loss": 0.
|
| 1569 |
"step": 5250
|
| 1570 |
},
|
| 1571 |
{
|
| 1572 |
-
"epoch":
|
| 1573 |
-
"grad_norm":
|
| 1574 |
-
"learning_rate": 2.
|
| 1575 |
-
"loss": 0.
|
| 1576 |
"step": 5275
|
| 1577 |
},
|
| 1578 |
{
|
| 1579 |
-
"epoch":
|
| 1580 |
-
"grad_norm": 2.
|
| 1581 |
-
"learning_rate": 2.
|
| 1582 |
-
"loss": 0.
|
| 1583 |
"step": 5300
|
| 1584 |
},
|
| 1585 |
{
|
| 1586 |
-
"epoch":
|
| 1587 |
-
"grad_norm":
|
| 1588 |
-
"learning_rate": 2.
|
| 1589 |
-
"loss": 0.
|
| 1590 |
"step": 5325
|
| 1591 |
},
|
| 1592 |
{
|
| 1593 |
-
"epoch":
|
| 1594 |
-
"grad_norm":
|
| 1595 |
-
"learning_rate": 2.
|
| 1596 |
-
"loss": 0.
|
| 1597 |
"step": 5350
|
| 1598 |
},
|
| 1599 |
{
|
| 1600 |
-
"epoch":
|
| 1601 |
-
"grad_norm": 2.
|
| 1602 |
-
"learning_rate": 2.
|
| 1603 |
-
"loss": 0.
|
| 1604 |
"step": 5375
|
| 1605 |
},
|
| 1606 |
{
|
| 1607 |
-
"epoch":
|
| 1608 |
-
"grad_norm": 1.
|
| 1609 |
-
"learning_rate": 2.
|
| 1610 |
-
"loss": 0.
|
| 1611 |
"step": 5400
|
| 1612 |
},
|
| 1613 |
{
|
| 1614 |
-
"epoch":
|
| 1615 |
-
"grad_norm":
|
| 1616 |
-
"learning_rate": 2.
|
| 1617 |
-
"loss": 0.
|
| 1618 |
"step": 5425
|
| 1619 |
},
|
| 1620 |
{
|
| 1621 |
-
"epoch":
|
| 1622 |
-
"grad_norm": 2.
|
| 1623 |
-
"learning_rate": 2.
|
| 1624 |
-
"loss": 0.
|
| 1625 |
"step": 5450
|
| 1626 |
},
|
| 1627 |
{
|
| 1628 |
-
"epoch":
|
| 1629 |
-
"grad_norm":
|
| 1630 |
-
"learning_rate": 2.
|
| 1631 |
-
"loss": 0.
|
| 1632 |
"step": 5475
|
| 1633 |
},
|
| 1634 |
{
|
| 1635 |
-
"epoch":
|
| 1636 |
-
"grad_norm":
|
| 1637 |
-
"learning_rate": 2.
|
| 1638 |
-
"loss": 0.
|
| 1639 |
"step": 5500
|
| 1640 |
},
|
| 1641 |
{
|
| 1642 |
-
"epoch":
|
| 1643 |
-
"eval_loss": 0.
|
| 1644 |
-
"eval_runtime":
|
| 1645 |
-
"eval_samples_per_second": 5.
|
| 1646 |
-
"eval_steps_per_second": 0.
|
| 1647 |
-
"eval_wer":
|
| 1648 |
"step": 5500
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
-
"epoch":
|
| 1652 |
-
"grad_norm":
|
| 1653 |
-
"learning_rate": 2.
|
| 1654 |
-
"loss": 0.
|
| 1655 |
"step": 5525
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
-
"epoch":
|
| 1659 |
-
"grad_norm":
|
| 1660 |
-
"learning_rate": 2.
|
| 1661 |
-
"loss": 0.
|
| 1662 |
"step": 5550
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
-
"epoch":
|
| 1666 |
-
"grad_norm":
|
| 1667 |
-
"learning_rate": 2.
|
| 1668 |
-
"loss": 0.
|
| 1669 |
"step": 5575
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
-
"epoch":
|
| 1673 |
-
"grad_norm":
|
| 1674 |
-
"learning_rate": 2.
|
| 1675 |
-
"loss": 0.
|
| 1676 |
"step": 5600
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
-
"epoch":
|
| 1680 |
-
"grad_norm":
|
| 1681 |
-
"learning_rate": 1.
|
| 1682 |
-
"loss": 0.
|
| 1683 |
"step": 5625
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
-
"epoch":
|
| 1687 |
-
"grad_norm":
|
| 1688 |
-
"learning_rate": 1.
|
| 1689 |
-
"loss": 0.
|
| 1690 |
"step": 5650
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
-
"epoch":
|
| 1694 |
-
"grad_norm":
|
| 1695 |
-
"learning_rate": 1.
|
| 1696 |
-
"loss": 0.
|
| 1697 |
"step": 5675
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
-
"epoch":
|
| 1701 |
-
"grad_norm": 3.
|
| 1702 |
-
"learning_rate": 1.
|
| 1703 |
-
"loss": 0.
|
| 1704 |
"step": 5700
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
-
"epoch":
|
| 1708 |
-
"grad_norm":
|
| 1709 |
-
"learning_rate": 1.
|
| 1710 |
-
"loss": 0.
|
| 1711 |
"step": 5725
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
-
"epoch":
|
| 1715 |
-
"grad_norm":
|
| 1716 |
-
"learning_rate": 1.
|
| 1717 |
-
"loss": 0.
|
| 1718 |
"step": 5750
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
-
"epoch":
|
| 1722 |
-
"grad_norm":
|
| 1723 |
-
"learning_rate": 1.
|
| 1724 |
-
"loss": 0.
|
| 1725 |
"step": 5775
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
-
"epoch":
|
| 1729 |
-
"grad_norm":
|
| 1730 |
-
"learning_rate": 1.
|
| 1731 |
-
"loss": 0.
|
| 1732 |
"step": 5800
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
-
"epoch":
|
| 1736 |
-
"grad_norm":
|
| 1737 |
-
"learning_rate": 1.
|
| 1738 |
-
"loss": 0.
|
| 1739 |
"step": 5825
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
-
"epoch":
|
| 1743 |
-
"grad_norm":
|
| 1744 |
-
"learning_rate": 1.
|
| 1745 |
-
"loss": 0.
|
| 1746 |
"step": 5850
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
-
"epoch":
|
| 1750 |
-
"grad_norm":
|
| 1751 |
-
"learning_rate": 1.
|
| 1752 |
-
"loss": 0.
|
| 1753 |
"step": 5875
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
-
"epoch":
|
| 1757 |
-
"grad_norm":
|
| 1758 |
-
"learning_rate": 1.
|
| 1759 |
-
"loss": 0.
|
| 1760 |
"step": 5900
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
-
"epoch":
|
| 1764 |
-
"grad_norm":
|
| 1765 |
-
"learning_rate": 1.
|
| 1766 |
-
"loss": 0.
|
| 1767 |
"step": 5925
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
-
"epoch":
|
| 1771 |
-
"grad_norm":
|
| 1772 |
-
"learning_rate": 1.
|
| 1773 |
-
"loss": 0.
|
| 1774 |
"step": 5950
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
-
"epoch":
|
| 1778 |
-
"grad_norm":
|
| 1779 |
-
"learning_rate": 1.
|
| 1780 |
-
"loss": 0.
|
| 1781 |
"step": 5975
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
-
"epoch":
|
| 1785 |
-
"grad_norm":
|
| 1786 |
-
"learning_rate": 1.
|
| 1787 |
-
"loss": 0.
|
| 1788 |
"step": 6000
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
-
"epoch":
|
| 1792 |
-
"eval_loss": 0.
|
| 1793 |
-
"eval_runtime":
|
| 1794 |
-
"eval_samples_per_second":
|
| 1795 |
-
"eval_steps_per_second": 0.
|
| 1796 |
-
"eval_wer": 9.
|
| 1797 |
"step": 6000
|
| 1798 |
},
|
| 1799 |
{
|
| 1800 |
-
"epoch":
|
| 1801 |
-
"grad_norm":
|
| 1802 |
"learning_rate": 1.6483333333333334e-06,
|
| 1803 |
-
"loss": 0.
|
| 1804 |
"step": 6025
|
| 1805 |
},
|
| 1806 |
{
|
| 1807 |
-
"epoch":
|
| 1808 |
-
"grad_norm":
|
| 1809 |
"learning_rate": 1.6275000000000003e-06,
|
| 1810 |
-
"loss": 0.
|
| 1811 |
"step": 6050
|
| 1812 |
},
|
| 1813 |
{
|
| 1814 |
-
"epoch":
|
| 1815 |
-
"grad_norm":
|
| 1816 |
"learning_rate": 1.6066666666666668e-06,
|
| 1817 |
-
"loss": 0.
|
| 1818 |
"step": 6075
|
| 1819 |
},
|
| 1820 |
{
|
| 1821 |
-
"epoch":
|
| 1822 |
-
"grad_norm":
|
| 1823 |
"learning_rate": 1.5858333333333333e-06,
|
| 1824 |
-
"loss": 0.
|
| 1825 |
"step": 6100
|
| 1826 |
},
|
| 1827 |
{
|
| 1828 |
-
"epoch":
|
| 1829 |
-
"grad_norm":
|
| 1830 |
"learning_rate": 1.565e-06,
|
| 1831 |
-
"loss": 0.
|
| 1832 |
"step": 6125
|
| 1833 |
},
|
| 1834 |
{
|
| 1835 |
-
"epoch":
|
| 1836 |
-
"grad_norm": 2.
|
| 1837 |
"learning_rate": 1.5441666666666668e-06,
|
| 1838 |
-
"loss": 0.
|
| 1839 |
"step": 6150
|
| 1840 |
},
|
| 1841 |
{
|
| 1842 |
-
"epoch":
|
| 1843 |
-
"grad_norm":
|
| 1844 |
"learning_rate": 1.5233333333333333e-06,
|
| 1845 |
-
"loss": 0.
|
| 1846 |
"step": 6175
|
| 1847 |
},
|
| 1848 |
{
|
| 1849 |
-
"epoch":
|
| 1850 |
-
"grad_norm":
|
| 1851 |
"learning_rate": 1.5025e-06,
|
| 1852 |
-
"loss": 0.
|
| 1853 |
"step": 6200
|
| 1854 |
},
|
| 1855 |
{
|
| 1856 |
-
"epoch":
|
| 1857 |
-
"grad_norm": 2.
|
| 1858 |
"learning_rate": 1.4816666666666667e-06,
|
| 1859 |
-
"loss": 0.
|
| 1860 |
"step": 6225
|
| 1861 |
},
|
| 1862 |
{
|
| 1863 |
-
"epoch":
|
| 1864 |
-
"grad_norm":
|
| 1865 |
"learning_rate": 1.4608333333333333e-06,
|
| 1866 |
-
"loss": 0.
|
| 1867 |
"step": 6250
|
| 1868 |
},
|
| 1869 |
{
|
| 1870 |
-
"epoch":
|
| 1871 |
-
"grad_norm":
|
| 1872 |
"learning_rate": 1.44e-06,
|
| 1873 |
-
"loss": 0.
|
| 1874 |
"step": 6275
|
| 1875 |
},
|
| 1876 |
{
|
| 1877 |
-
"epoch":
|
| 1878 |
-
"grad_norm":
|
| 1879 |
"learning_rate": 1.4191666666666667e-06,
|
| 1880 |
-
"loss": 0.
|
| 1881 |
"step": 6300
|
| 1882 |
},
|
| 1883 |
{
|
| 1884 |
-
"epoch":
|
| 1885 |
-
"grad_norm":
|
| 1886 |
"learning_rate": 1.3983333333333334e-06,
|
| 1887 |
-
"loss": 0.
|
| 1888 |
"step": 6325
|
| 1889 |
},
|
| 1890 |
{
|
| 1891 |
-
"epoch":
|
| 1892 |
-
"grad_norm":
|
| 1893 |
"learning_rate": 1.3775000000000002e-06,
|
| 1894 |
-
"loss": 0.
|
| 1895 |
"step": 6350
|
| 1896 |
},
|
| 1897 |
{
|
| 1898 |
-
"epoch":
|
| 1899 |
-
"grad_norm":
|
| 1900 |
"learning_rate": 1.3566666666666667e-06,
|
| 1901 |
-
"loss": 0.
|
| 1902 |
"step": 6375
|
| 1903 |
},
|
| 1904 |
{
|
| 1905 |
-
"epoch":
|
| 1906 |
-
"grad_norm":
|
| 1907 |
"learning_rate": 1.3358333333333334e-06,
|
| 1908 |
-
"loss": 0.
|
| 1909 |
"step": 6400
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
-
"epoch":
|
| 1913 |
-
"grad_norm":
|
| 1914 |
"learning_rate": 1.3150000000000001e-06,
|
| 1915 |
-
"loss": 0.
|
| 1916 |
"step": 6425
|
| 1917 |
},
|
| 1918 |
{
|
| 1919 |
-
"epoch":
|
| 1920 |
-
"grad_norm":
|
| 1921 |
"learning_rate": 1.2941666666666668e-06,
|
| 1922 |
-
"loss": 0.
|
| 1923 |
"step": 6450
|
| 1924 |
},
|
| 1925 |
{
|
| 1926 |
-
"epoch":
|
| 1927 |
-
"grad_norm":
|
| 1928 |
"learning_rate": 1.2733333333333334e-06,
|
| 1929 |
-
"loss": 0.
|
| 1930 |
"step": 6475
|
| 1931 |
},
|
| 1932 |
{
|
| 1933 |
-
"epoch":
|
| 1934 |
-
"grad_norm":
|
| 1935 |
"learning_rate": 1.2525e-06,
|
| 1936 |
-
"loss": 0.
|
| 1937 |
"step": 6500
|
| 1938 |
},
|
| 1939 |
{
|
| 1940 |
-
"epoch":
|
| 1941 |
-
"eval_loss": 0.
|
| 1942 |
-
"eval_runtime":
|
| 1943 |
-
"eval_samples_per_second":
|
| 1944 |
-
"eval_steps_per_second": 0.
|
| 1945 |
-
"eval_wer":
|
| 1946 |
"step": 6500
|
| 1947 |
},
|
| 1948 |
{
|
| 1949 |
-
"epoch":
|
| 1950 |
-
"grad_norm":
|
| 1951 |
"learning_rate": 1.2316666666666668e-06,
|
| 1952 |
-
"loss": 0.
|
| 1953 |
"step": 6525
|
| 1954 |
},
|
| 1955 |
{
|
| 1956 |
-
"epoch":
|
| 1957 |
-
"grad_norm":
|
| 1958 |
"learning_rate": 1.2108333333333335e-06,
|
| 1959 |
-
"loss": 0.
|
| 1960 |
"step": 6550
|
| 1961 |
},
|
| 1962 |
{
|
| 1963 |
-
"epoch":
|
| 1964 |
-
"grad_norm":
|
| 1965 |
"learning_rate": 1.19e-06,
|
| 1966 |
-
"loss": 0.
|
| 1967 |
"step": 6575
|
| 1968 |
},
|
| 1969 |
{
|
| 1970 |
-
"epoch":
|
| 1971 |
-
"grad_norm": 2.
|
| 1972 |
"learning_rate": 1.1691666666666668e-06,
|
| 1973 |
-
"loss": 0.
|
| 1974 |
"step": 6600
|
| 1975 |
},
|
| 1976 |
{
|
| 1977 |
-
"epoch":
|
| 1978 |
-
"grad_norm":
|
| 1979 |
-
"learning_rate": 1.
|
| 1980 |
-
"loss": 0.
|
| 1981 |
"step": 6625
|
| 1982 |
},
|
| 1983 |
{
|
| 1984 |
-
"epoch":
|
| 1985 |
-
"grad_norm":
|
| 1986 |
"learning_rate": 1.1283333333333333e-06,
|
| 1987 |
-
"loss": 0.
|
| 1988 |
"step": 6650
|
| 1989 |
},
|
| 1990 |
{
|
| 1991 |
-
"epoch":
|
| 1992 |
-
"grad_norm":
|
| 1993 |
"learning_rate": 1.1075e-06,
|
| 1994 |
-
"loss": 0.
|
| 1995 |
"step": 6675
|
| 1996 |
},
|
| 1997 |
{
|
| 1998 |
-
"epoch":
|
| 1999 |
-
"grad_norm":
|
| 2000 |
"learning_rate": 1.0866666666666667e-06,
|
| 2001 |
-
"loss": 0.
|
| 2002 |
"step": 6700
|
| 2003 |
},
|
| 2004 |
{
|
| 2005 |
-
"epoch":
|
| 2006 |
-
"grad_norm":
|
| 2007 |
"learning_rate": 1.0658333333333334e-06,
|
| 2008 |
-
"loss": 0.
|
| 2009 |
"step": 6725
|
| 2010 |
},
|
| 2011 |
{
|
| 2012 |
-
"epoch":
|
| 2013 |
-
"grad_norm":
|
| 2014 |
"learning_rate": 1.045e-06,
|
| 2015 |
-
"loss": 0.
|
| 2016 |
"step": 6750
|
| 2017 |
},
|
| 2018 |
{
|
| 2019 |
-
"epoch":
|
| 2020 |
-
"grad_norm":
|
| 2021 |
"learning_rate": 1.0241666666666667e-06,
|
| 2022 |
-
"loss": 0.
|
| 2023 |
"step": 6775
|
| 2024 |
},
|
| 2025 |
{
|
| 2026 |
-
"epoch":
|
| 2027 |
-
"grad_norm":
|
| 2028 |
"learning_rate": 1.0033333333333334e-06,
|
| 2029 |
-
"loss": 0.
|
| 2030 |
"step": 6800
|
| 2031 |
},
|
| 2032 |
{
|
| 2033 |
-
"epoch":
|
| 2034 |
-
"grad_norm":
|
| 2035 |
"learning_rate": 9.825000000000001e-07,
|
| 2036 |
-
"loss": 0.
|
| 2037 |
"step": 6825
|
| 2038 |
},
|
| 2039 |
{
|
| 2040 |
-
"epoch":
|
| 2041 |
-
"grad_norm":
|
| 2042 |
"learning_rate": 9.616666666666668e-07,
|
| 2043 |
-
"loss": 0.
|
| 2044 |
"step": 6850
|
| 2045 |
},
|
| 2046 |
{
|
| 2047 |
-
"epoch":
|
| 2048 |
-
"grad_norm":
|
| 2049 |
"learning_rate": 9.408333333333334e-07,
|
| 2050 |
-
"loss": 0.
|
| 2051 |
"step": 6875
|
| 2052 |
},
|
| 2053 |
{
|
| 2054 |
-
"epoch":
|
| 2055 |
-
"grad_norm":
|
| 2056 |
"learning_rate": 9.2e-07,
|
| 2057 |
-
"loss": 0.
|
| 2058 |
"step": 6900
|
| 2059 |
},
|
| 2060 |
{
|
| 2061 |
-
"epoch":
|
| 2062 |
-
"grad_norm":
|
| 2063 |
"learning_rate": 8.991666666666667e-07,
|
| 2064 |
-
"loss": 0.
|
| 2065 |
"step": 6925
|
| 2066 |
},
|
| 2067 |
{
|
| 2068 |
-
"epoch":
|
| 2069 |
-
"grad_norm":
|
| 2070 |
"learning_rate": 8.783333333333334e-07,
|
| 2071 |
-
"loss": 0.
|
| 2072 |
"step": 6950
|
| 2073 |
},
|
| 2074 |
{
|
| 2075 |
-
"epoch":
|
| 2076 |
-
"grad_norm":
|
| 2077 |
"learning_rate": 8.574999999999999e-07,
|
| 2078 |
-
"loss": 0.
|
| 2079 |
"step": 6975
|
| 2080 |
},
|
| 2081 |
{
|
| 2082 |
-
"epoch":
|
| 2083 |
-
"grad_norm":
|
| 2084 |
"learning_rate": 8.366666666666667e-07,
|
| 2085 |
-
"loss": 0.
|
| 2086 |
"step": 7000
|
| 2087 |
},
|
| 2088 |
{
|
| 2089 |
-
"epoch":
|
| 2090 |
-
"eval_loss": 0.
|
| 2091 |
-
"eval_runtime":
|
| 2092 |
-
"eval_samples_per_second":
|
| 2093 |
-
"eval_steps_per_second": 0.
|
| 2094 |
-
"eval_wer":
|
| 2095 |
"step": 7000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
-
"epoch":
|
| 2099 |
-
"grad_norm":
|
| 2100 |
"learning_rate": 8.158333333333334e-07,
|
| 2101 |
-
"loss": 0.
|
| 2102 |
"step": 7025
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
-
"epoch":
|
| 2106 |
-
"grad_norm":
|
| 2107 |
"learning_rate": 7.950000000000001e-07,
|
| 2108 |
-
"loss": 0.
|
| 2109 |
"step": 7050
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
-
"epoch":
|
| 2113 |
-
"grad_norm":
|
| 2114 |
"learning_rate": 7.741666666666667e-07,
|
| 2115 |
-
"loss": 0.
|
| 2116 |
"step": 7075
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
-
"epoch":
|
| 2120 |
-
"grad_norm":
|
| 2121 |
"learning_rate": 7.533333333333334e-07,
|
| 2122 |
-
"loss": 0.
|
| 2123 |
"step": 7100
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
-
"epoch":
|
| 2127 |
-
"grad_norm":
|
| 2128 |
"learning_rate": 7.325e-07,
|
| 2129 |
-
"loss": 0.
|
| 2130 |
"step": 7125
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
-
"epoch":
|
| 2134 |
-
"grad_norm":
|
| 2135 |
"learning_rate": 7.116666666666667e-07,
|
| 2136 |
-
"loss": 0.
|
| 2137 |
"step": 7150
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
-
"epoch":
|
| 2141 |
-
"grad_norm":
|
| 2142 |
"learning_rate": 6.908333333333333e-07,
|
| 2143 |
-
"loss": 0.
|
| 2144 |
"step": 7175
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
-
"epoch":
|
| 2148 |
-
"grad_norm":
|
| 2149 |
"learning_rate": 6.7e-07,
|
| 2150 |
-
"loss": 0.
|
| 2151 |
"step": 7200
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
-
"epoch":
|
| 2155 |
-
"grad_norm":
|
| 2156 |
"learning_rate": 6.491666666666667e-07,
|
| 2157 |
-
"loss": 0.
|
| 2158 |
"step": 7225
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
-
"epoch":
|
| 2162 |
-
"grad_norm":
|
| 2163 |
"learning_rate": 6.283333333333334e-07,
|
| 2164 |
-
"loss": 0.
|
| 2165 |
"step": 7250
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
-
"epoch":
|
| 2169 |
-
"grad_norm":
|
| 2170 |
"learning_rate": 6.075e-07,
|
| 2171 |
-
"loss": 0.
|
| 2172 |
"step": 7275
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
-
"epoch":
|
| 2176 |
-
"grad_norm":
|
| 2177 |
"learning_rate": 5.866666666666667e-07,
|
| 2178 |
-
"loss": 0.
|
| 2179 |
"step": 7300
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
-
"epoch":
|
| 2183 |
-
"grad_norm":
|
| 2184 |
"learning_rate": 5.658333333333334e-07,
|
| 2185 |
-
"loss": 0.
|
| 2186 |
"step": 7325
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
-
"epoch":
|
| 2190 |
-
"grad_norm":
|
| 2191 |
"learning_rate": 5.450000000000001e-07,
|
| 2192 |
-
"loss": 0.
|
| 2193 |
"step": 7350
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
-
"epoch":
|
| 2197 |
-
"grad_norm":
|
| 2198 |
"learning_rate": 5.241666666666667e-07,
|
| 2199 |
-
"loss": 0.
|
| 2200 |
"step": 7375
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
-
"epoch":
|
| 2204 |
-
"grad_norm":
|
| 2205 |
"learning_rate": 5.033333333333333e-07,
|
| 2206 |
-
"loss": 0.
|
| 2207 |
"step": 7400
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
-
"epoch":
|
| 2211 |
-
"grad_norm":
|
| 2212 |
"learning_rate": 4.825e-07,
|
| 2213 |
-
"loss": 0.
|
| 2214 |
"step": 7425
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
-
"epoch":
|
| 2218 |
-
"grad_norm":
|
| 2219 |
"learning_rate": 4.6166666666666666e-07,
|
| 2220 |
-
"loss": 0.
|
| 2221 |
"step": 7450
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
-
"epoch":
|
| 2225 |
-
"grad_norm":
|
| 2226 |
"learning_rate": 4.408333333333334e-07,
|
| 2227 |
-
"loss": 0.
|
| 2228 |
"step": 7475
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
-
"epoch":
|
| 2232 |
-
"grad_norm":
|
| 2233 |
"learning_rate": 4.2e-07,
|
| 2234 |
-
"loss": 0.
|
| 2235 |
"step": 7500
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
-
"epoch":
|
| 2239 |
-
"eval_loss": 0.
|
| 2240 |
-
"eval_runtime":
|
| 2241 |
-
"eval_samples_per_second":
|
| 2242 |
-
"eval_steps_per_second": 0.
|
| 2243 |
-
"eval_wer":
|
| 2244 |
"step": 7500
|
| 2245 |
},
|
| 2246 |
{
|
| 2247 |
-
"epoch":
|
| 2248 |
-
"grad_norm":
|
| 2249 |
"learning_rate": 3.991666666666667e-07,
|
| 2250 |
-
"loss": 0.
|
| 2251 |
"step": 7525
|
| 2252 |
},
|
| 2253 |
{
|
| 2254 |
-
"epoch":
|
| 2255 |
-
"grad_norm":
|
| 2256 |
"learning_rate": 3.7833333333333335e-07,
|
| 2257 |
-
"loss": 0.
|
| 2258 |
"step": 7550
|
| 2259 |
},
|
| 2260 |
{
|
| 2261 |
-
"epoch":
|
| 2262 |
-
"grad_norm":
|
| 2263 |
"learning_rate": 3.575e-07,
|
| 2264 |
-
"loss": 0.
|
| 2265 |
"step": 7575
|
| 2266 |
},
|
| 2267 |
{
|
| 2268 |
-
"epoch":
|
| 2269 |
-
"grad_norm":
|
| 2270 |
"learning_rate": 3.366666666666667e-07,
|
| 2271 |
-
"loss": 0.
|
| 2272 |
"step": 7600
|
| 2273 |
},
|
| 2274 |
{
|
| 2275 |
-
"epoch":
|
| 2276 |
-
"grad_norm":
|
| 2277 |
"learning_rate": 3.1583333333333336e-07,
|
| 2278 |
-
"loss": 0.
|
| 2279 |
"step": 7625
|
| 2280 |
},
|
| 2281 |
{
|
| 2282 |
-
"epoch":
|
| 2283 |
-
"grad_norm":
|
| 2284 |
"learning_rate": 2.95e-07,
|
| 2285 |
-
"loss": 0.
|
| 2286 |
"step": 7650
|
| 2287 |
},
|
| 2288 |
{
|
| 2289 |
-
"epoch":
|
| 2290 |
-
"grad_norm":
|
| 2291 |
"learning_rate": 2.7416666666666665e-07,
|
| 2292 |
-
"loss": 0.
|
| 2293 |
"step": 7675
|
| 2294 |
},
|
| 2295 |
{
|
| 2296 |
-
"epoch":
|
| 2297 |
-
"grad_norm":
|
| 2298 |
"learning_rate": 2.533333333333333e-07,
|
| 2299 |
-
"loss": 0.
|
| 2300 |
"step": 7700
|
| 2301 |
},
|
| 2302 |
{
|
| 2303 |
-
"epoch":
|
| 2304 |
-
"grad_norm":
|
| 2305 |
"learning_rate": 2.325e-07,
|
| 2306 |
-
"loss": 0.
|
| 2307 |
"step": 7725
|
| 2308 |
},
|
| 2309 |
{
|
| 2310 |
-
"epoch":
|
| 2311 |
-
"grad_norm":
|
| 2312 |
"learning_rate": 2.116666666666667e-07,
|
| 2313 |
-
"loss": 0.
|
| 2314 |
"step": 7750
|
| 2315 |
},
|
| 2316 |
{
|
| 2317 |
-
"epoch":
|
| 2318 |
-
"grad_norm":
|
| 2319 |
"learning_rate": 1.9083333333333334e-07,
|
| 2320 |
-
"loss": 0.
|
| 2321 |
"step": 7775
|
| 2322 |
},
|
| 2323 |
{
|
| 2324 |
-
"epoch":
|
| 2325 |
-
"grad_norm":
|
| 2326 |
"learning_rate": 1.7e-07,
|
| 2327 |
-
"loss": 0.
|
| 2328 |
"step": 7800
|
| 2329 |
},
|
| 2330 |
{
|
| 2331 |
-
"epoch":
|
| 2332 |
-
"grad_norm":
|
| 2333 |
"learning_rate": 1.4916666666666669e-07,
|
| 2334 |
-
"loss": 0.
|
| 2335 |
"step": 7825
|
| 2336 |
},
|
| 2337 |
{
|
| 2338 |
-
"epoch":
|
| 2339 |
-
"grad_norm":
|
| 2340 |
"learning_rate": 1.2833333333333333e-07,
|
| 2341 |
-
"loss": 0.
|
| 2342 |
"step": 7850
|
| 2343 |
},
|
| 2344 |
{
|
| 2345 |
-
"epoch":
|
| 2346 |
-
"grad_norm":
|
| 2347 |
"learning_rate": 1.075e-07,
|
| 2348 |
-
"loss": 0.
|
| 2349 |
"step": 7875
|
| 2350 |
},
|
| 2351 |
{
|
| 2352 |
-
"epoch":
|
| 2353 |
-
"grad_norm": 2.
|
| 2354 |
"learning_rate": 8.666666666666666e-08,
|
| 2355 |
-
"loss": 0.
|
| 2356 |
"step": 7900
|
| 2357 |
},
|
| 2358 |
{
|
| 2359 |
-
"epoch":
|
| 2360 |
-
"grad_norm":
|
| 2361 |
"learning_rate": 6.583333333333333e-08,
|
| 2362 |
-
"loss": 0.
|
| 2363 |
"step": 7925
|
| 2364 |
},
|
| 2365 |
{
|
| 2366 |
-
"epoch":
|
| 2367 |
-
"grad_norm":
|
| 2368 |
"learning_rate": 4.5e-08,
|
| 2369 |
-
"loss": 0.
|
| 2370 |
"step": 7950
|
| 2371 |
},
|
| 2372 |
{
|
| 2373 |
-
"epoch":
|
| 2374 |
-
"grad_norm":
|
| 2375 |
"learning_rate": 2.4166666666666668e-08,
|
| 2376 |
-
"loss": 0.
|
| 2377 |
"step": 7975
|
| 2378 |
},
|
| 2379 |
{
|
| 2380 |
-
"epoch":
|
| 2381 |
-
"grad_norm":
|
| 2382 |
"learning_rate": 3.333333333333334e-09,
|
| 2383 |
-
"loss": 0.
|
| 2384 |
"step": 8000
|
| 2385 |
},
|
| 2386 |
{
|
| 2387 |
-
"epoch":
|
| 2388 |
-
"eval_loss": 0.
|
| 2389 |
-
"eval_runtime":
|
| 2390 |
-
"eval_samples_per_second":
|
| 2391 |
-
"eval_steps_per_second": 0.
|
| 2392 |
-
"eval_wer":
|
| 2393 |
"step": 8000
|
| 2394 |
},
|
| 2395 |
{
|
| 2396 |
-
"epoch":
|
| 2397 |
"step": 8000,
|
| 2398 |
-
"total_flos": 1.
|
| 2399 |
-
"train_loss": 0.
|
| 2400 |
-
"train_runtime":
|
| 2401 |
-
"train_samples_per_second":
|
| 2402 |
-
"train_steps_per_second": 0.
|
| 2403 |
}
|
| 2404 |
],
|
| 2405 |
"logging_steps": 25,
|
|
@@ -2419,7 +2419,7 @@
|
|
| 2419 |
"attributes": {}
|
| 2420 |
}
|
| 2421 |
},
|
| 2422 |
-
"total_flos": 1.
|
| 2423 |
"train_batch_size": 16,
|
| 2424 |
"trial_name": null,
|
| 2425 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 9.98410769374591,
|
| 3 |
+
"best_model_checkpoint": "./checkpoint-6000",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 8000,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.003125,
|
| 13 |
+
"grad_norm": 18.986873626708984,
|
| 14 |
"learning_rate": 2.875e-07,
|
| 15 |
+
"loss": 1.979,
|
| 16 |
"step": 25
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.00625,
|
| 20 |
+
"grad_norm": 16.025917053222656,
|
| 21 |
"learning_rate": 6.000000000000001e-07,
|
| 22 |
+
"loss": 1.8343,
|
| 23 |
"step": 50
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009375,
|
| 27 |
+
"grad_norm": 11.7916841506958,
|
| 28 |
"learning_rate": 9.124999999999999e-07,
|
| 29 |
+
"loss": 1.4286,
|
| 30 |
"step": 75
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.0125,
|
| 34 |
+
"grad_norm": 9.848176002502441,
|
| 35 |
"learning_rate": 1.2250000000000001e-06,
|
| 36 |
+
"loss": 0.9655,
|
| 37 |
"step": 100
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.015625,
|
| 41 |
+
"grad_norm": 9.252968788146973,
|
| 42 |
"learning_rate": 1.5375e-06,
|
| 43 |
+
"loss": 0.7411,
|
| 44 |
"step": 125
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.01875,
|
| 48 |
+
"grad_norm": 9.25166130065918,
|
| 49 |
"learning_rate": 1.85e-06,
|
| 50 |
+
"loss": 0.7377,
|
| 51 |
"step": 150
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.021875,
|
| 55 |
+
"grad_norm": 9.127242088317871,
|
| 56 |
"learning_rate": 2.1625e-06,
|
| 57 |
+
"loss": 0.5628,
|
| 58 |
"step": 175
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.025,
|
| 62 |
+
"grad_norm": 7.140044212341309,
|
| 63 |
"learning_rate": 2.4750000000000004e-06,
|
| 64 |
+
"loss": 0.5043,
|
| 65 |
"step": 200
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.028125,
|
| 69 |
+
"grad_norm": 7.495832920074463,
|
| 70 |
"learning_rate": 2.7875e-06,
|
| 71 |
+
"loss": 0.4543,
|
| 72 |
"step": 225
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03125,
|
| 76 |
+
"grad_norm": 7.260509014129639,
|
| 77 |
"learning_rate": 3.1e-06,
|
| 78 |
+
"loss": 0.441,
|
| 79 |
"step": 250
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.034375,
|
| 83 |
+
"grad_norm": 7.78987455368042,
|
| 84 |
"learning_rate": 3.4125000000000004e-06,
|
| 85 |
+
"loss": 0.3709,
|
| 86 |
"step": 275
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0375,
|
| 90 |
+
"grad_norm": 6.570291042327881,
|
| 91 |
"learning_rate": 3.725e-06,
|
| 92 |
+
"loss": 0.3778,
|
| 93 |
"step": 300
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.040625,
|
| 97 |
+
"grad_norm": 6.600038051605225,
|
| 98 |
"learning_rate": 4.037500000000001e-06,
|
| 99 |
+
"loss": 0.342,
|
| 100 |
"step": 325
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.04375,
|
| 104 |
+
"grad_norm": 6.174134731292725,
|
| 105 |
"learning_rate": 4.35e-06,
|
| 106 |
+
"loss": 0.3479,
|
| 107 |
"step": 350
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.046875,
|
| 111 |
+
"grad_norm": 5.382558345794678,
|
| 112 |
"learning_rate": 4.6625e-06,
|
| 113 |
+
"loss": 0.3389,
|
| 114 |
"step": 375
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.05,
|
| 118 |
+
"grad_norm": 6.139970302581787,
|
| 119 |
"learning_rate": 4.975000000000001e-06,
|
| 120 |
+
"loss": 0.3001,
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.053125,
|
| 125 |
+
"grad_norm": 6.267641544342041,
|
| 126 |
"learning_rate": 5.2875e-06,
|
| 127 |
+
"loss": 0.2827,
|
| 128 |
"step": 425
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.05625,
|
| 132 |
+
"grad_norm": 4.701253414154053,
|
| 133 |
"learning_rate": 5.600000000000001e-06,
|
| 134 |
+
"loss": 0.252,
|
| 135 |
"step": 450
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.059375,
|
| 139 |
+
"grad_norm": 7.49235725402832,
|
| 140 |
"learning_rate": 5.9125e-06,
|
| 141 |
+
"loss": 0.2724,
|
| 142 |
"step": 475
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.0625,
|
| 146 |
+
"grad_norm": 8.219965934753418,
|
| 147 |
"learning_rate": 6.2250000000000005e-06,
|
| 148 |
+
"loss": 0.3412,
|
| 149 |
"step": 500
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.0625,
|
| 153 |
+
"eval_loss": 0.4569947421550751,
|
| 154 |
+
"eval_runtime": 421.2437,
|
| 155 |
+
"eval_samples_per_second": 4.995,
|
| 156 |
+
"eval_steps_per_second": 0.624,
|
| 157 |
+
"eval_wer": 28.21819201645321,
|
| 158 |
"step": 500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.065625,
|
| 162 |
+
"grad_norm": 7.646754741668701,
|
| 163 |
"learning_rate": 6.230833333333334e-06,
|
| 164 |
+
"loss": 0.3663,
|
| 165 |
"step": 525
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.06875,
|
| 169 |
+
"grad_norm": 6.3216094970703125,
|
| 170 |
"learning_rate": 6.210000000000001e-06,
|
| 171 |
+
"loss": 0.3401,
|
| 172 |
"step": 550
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.071875,
|
| 176 |
+
"grad_norm": 6.551443099975586,
|
| 177 |
"learning_rate": 6.1891666666666666e-06,
|
| 178 |
+
"loss": 0.3617,
|
| 179 |
"step": 575
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.075,
|
| 183 |
+
"grad_norm": 7.667718887329102,
|
| 184 |
"learning_rate": 6.168333333333333e-06,
|
| 185 |
+
"loss": 0.3339,
|
| 186 |
"step": 600
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.078125,
|
| 190 |
+
"grad_norm": 6.759644031524658,
|
| 191 |
"learning_rate": 6.147500000000001e-06,
|
| 192 |
+
"loss": 0.3236,
|
| 193 |
"step": 625
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.08125,
|
| 197 |
+
"grad_norm": 5.439680576324463,
|
| 198 |
"learning_rate": 6.126666666666667e-06,
|
| 199 |
+
"loss": 0.2588,
|
| 200 |
"step": 650
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.084375,
|
| 204 |
+
"grad_norm": 5.2066450119018555,
|
| 205 |
"learning_rate": 6.1058333333333335e-06,
|
| 206 |
+
"loss": 0.2389,
|
| 207 |
"step": 675
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.0875,
|
| 211 |
+
"grad_norm": 4.867877960205078,
|
| 212 |
"learning_rate": 6.085e-06,
|
| 213 |
+
"loss": 0.2178,
|
| 214 |
"step": 700
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 0.090625,
|
| 218 |
+
"grad_norm": 4.836174011230469,
|
| 219 |
"learning_rate": 6.064166666666667e-06,
|
| 220 |
+
"loss": 0.185,
|
| 221 |
"step": 725
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.09375,
|
| 225 |
+
"grad_norm": 4.676128387451172,
|
| 226 |
"learning_rate": 6.043333333333334e-06,
|
| 227 |
+
"loss": 0.1968,
|
| 228 |
"step": 750
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 0.096875,
|
| 232 |
+
"grad_norm": 3.805962324142456,
|
| 233 |
"learning_rate": 6.0225e-06,
|
| 234 |
+
"loss": 0.1987,
|
| 235 |
"step": 775
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.1,
|
| 239 |
+
"grad_norm": 4.182210922241211,
|
| 240 |
"learning_rate": 6.001666666666667e-06,
|
| 241 |
+
"loss": 0.1945,
|
| 242 |
"step": 800
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.103125,
|
| 246 |
+
"grad_norm": 4.078372001647949,
|
| 247 |
"learning_rate": 5.980833333333334e-06,
|
| 248 |
+
"loss": 0.1806,
|
| 249 |
"step": 825
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.10625,
|
| 253 |
+
"grad_norm": 4.926514148712158,
|
| 254 |
"learning_rate": 5.9600000000000005e-06,
|
| 255 |
+
"loss": 0.1655,
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 0.109375,
|
| 260 |
+
"grad_norm": 4.110943794250488,
|
| 261 |
"learning_rate": 5.939166666666667e-06,
|
| 262 |
+
"loss": 0.1711,
|
| 263 |
"step": 875
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.1125,
|
| 267 |
+
"grad_norm": 4.059125900268555,
|
| 268 |
"learning_rate": 5.918333333333333e-06,
|
| 269 |
+
"loss": 0.1721,
|
| 270 |
"step": 900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.115625,
|
| 274 |
+
"grad_norm": 4.911381244659424,
|
| 275 |
"learning_rate": 5.8975e-06,
|
| 276 |
+
"loss": 0.1617,
|
| 277 |
"step": 925
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.11875,
|
| 281 |
+
"grad_norm": 4.673095226287842,
|
| 282 |
"learning_rate": 5.8766666666666674e-06,
|
| 283 |
+
"loss": 0.1543,
|
| 284 |
"step": 950
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.121875,
|
| 288 |
+
"grad_norm": 4.966635704040527,
|
| 289 |
"learning_rate": 5.855833333333333e-06,
|
| 290 |
+
"loss": 0.1585,
|
| 291 |
"step": 975
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.125,
|
| 295 |
+
"grad_norm": 4.074397087097168,
|
| 296 |
"learning_rate": 5.835e-06,
|
| 297 |
+
"loss": 0.1462,
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 0.125,
|
| 302 |
+
"eval_loss": 0.3524394631385803,
|
| 303 |
+
"eval_runtime": 423.5166,
|
| 304 |
+
"eval_samples_per_second": 4.968,
|
| 305 |
+
"eval_steps_per_second": 0.621,
|
| 306 |
+
"eval_wer": 19.940170141161072,
|
| 307 |
"step": 1000
|
| 308 |
},
|
| 309 |
{
|
| 310 |
"epoch": 0.128125,
|
| 311 |
+
"grad_norm": 2.7621817588806152,
|
| 312 |
"learning_rate": 5.814166666666667e-06,
|
| 313 |
+
"loss": 0.1543,
|
| 314 |
"step": 1025
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.13125,
|
| 318 |
+
"grad_norm": 5.36495304107666,
|
| 319 |
"learning_rate": 5.793333333333334e-06,
|
| 320 |
+
"loss": 0.1356,
|
| 321 |
"step": 1050
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 0.134375,
|
| 325 |
+
"grad_norm": 4.922938346862793,
|
| 326 |
"learning_rate": 5.7725e-06,
|
| 327 |
+
"loss": 0.1459,
|
| 328 |
"step": 1075
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 0.1375,
|
| 332 |
+
"grad_norm": 3.6823904514312744,
|
| 333 |
"learning_rate": 5.751666666666667e-06,
|
| 334 |
+
"loss": 0.148,
|
| 335 |
"step": 1100
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.140625,
|
| 339 |
+
"grad_norm": 3.583972215652466,
|
| 340 |
"learning_rate": 5.730833333333334e-06,
|
| 341 |
+
"loss": 0.1408,
|
| 342 |
"step": 1125
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 0.14375,
|
| 346 |
+
"grad_norm": 4.794659614562988,
|
| 347 |
"learning_rate": 5.71e-06,
|
| 348 |
+
"loss": 0.1363,
|
| 349 |
"step": 1150
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 0.146875,
|
| 353 |
+
"grad_norm": 2.823441743850708,
|
| 354 |
"learning_rate": 5.689166666666667e-06,
|
| 355 |
+
"loss": 0.1378,
|
| 356 |
"step": 1175
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.15,
|
| 360 |
+
"grad_norm": 5.426193714141846,
|
| 361 |
"learning_rate": 5.668333333333334e-06,
|
| 362 |
+
"loss": 0.1454,
|
| 363 |
"step": 1200
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 0.153125,
|
| 367 |
+
"grad_norm": 3.506147861480713,
|
| 368 |
"learning_rate": 5.6475e-06,
|
| 369 |
+
"loss": 0.124,
|
| 370 |
"step": 1225
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 0.15625,
|
| 374 |
+
"grad_norm": 3.6159119606018066,
|
| 375 |
"learning_rate": 5.626666666666667e-06,
|
| 376 |
+
"loss": 0.1214,
|
| 377 |
"step": 1250
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.159375,
|
| 381 |
+
"grad_norm": 4.304079055786133,
|
| 382 |
"learning_rate": 5.605833333333334e-06,
|
| 383 |
+
"loss": 0.1266,
|
| 384 |
"step": 1275
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 0.1625,
|
| 388 |
+
"grad_norm": 3.570537805557251,
|
| 389 |
"learning_rate": 5.585e-06,
|
| 390 |
+
"loss": 0.1474,
|
| 391 |
"step": 1300
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 0.165625,
|
| 395 |
+
"grad_norm": 3.944892644882202,
|
| 396 |
"learning_rate": 5.564166666666667e-06,
|
| 397 |
+
"loss": 0.1327,
|
| 398 |
"step": 1325
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.16875,
|
| 402 |
+
"grad_norm": 4.3111572265625,
|
| 403 |
"learning_rate": 5.543333333333333e-06,
|
| 404 |
+
"loss": 0.1361,
|
| 405 |
"step": 1350
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.171875,
|
| 409 |
+
"grad_norm": 4.4654083251953125,
|
| 410 |
"learning_rate": 5.522500000000001e-06,
|
| 411 |
+
"loss": 0.1348,
|
| 412 |
"step": 1375
|
| 413 |
},
|
| 414 |
{
|
| 415 |
"epoch": 0.175,
|
| 416 |
+
"grad_norm": 4.158225059509277,
|
| 417 |
"learning_rate": 5.501666666666667e-06,
|
| 418 |
+
"loss": 0.1321,
|
| 419 |
"step": 1400
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 0.178125,
|
| 423 |
+
"grad_norm": 3.5653274059295654,
|
| 424 |
"learning_rate": 5.4808333333333335e-06,
|
| 425 |
+
"loss": 0.1392,
|
| 426 |
"step": 1425
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 0.18125,
|
| 430 |
+
"grad_norm": 8.204817771911621,
|
| 431 |
"learning_rate": 5.46e-06,
|
| 432 |
+
"loss": 0.1697,
|
| 433 |
"step": 1450
|
| 434 |
},
|
| 435 |
{
|
| 436 |
"epoch": 0.184375,
|
| 437 |
+
"grad_norm": 6.251273155212402,
|
| 438 |
"learning_rate": 5.439166666666667e-06,
|
| 439 |
+
"loss": 0.2126,
|
| 440 |
"step": 1475
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.1875,
|
| 444 |
+
"grad_norm": 5.315432071685791,
|
| 445 |
"learning_rate": 5.418333333333334e-06,
|
| 446 |
+
"loss": 0.2495,
|
| 447 |
"step": 1500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
"epoch": 0.1875,
|
| 451 |
+
"eval_loss": 0.30772972106933594,
|
| 452 |
+
"eval_runtime": 441.4634,
|
| 453 |
+
"eval_samples_per_second": 4.766,
|
| 454 |
+
"eval_steps_per_second": 0.596,
|
| 455 |
+
"eval_wer": 17.523604748995044,
|
| 456 |
"step": 1500
|
| 457 |
},
|
| 458 |
{
|
| 459 |
"epoch": 0.190625,
|
| 460 |
+
"grad_norm": 5.567235469818115,
|
| 461 |
"learning_rate": 5.3975e-06,
|
| 462 |
+
"loss": 0.2418,
|
| 463 |
"step": 1525
|
| 464 |
},
|
| 465 |
{
|
| 466 |
"epoch": 0.19375,
|
| 467 |
+
"grad_norm": 5.031063079833984,
|
| 468 |
"learning_rate": 5.376666666666666e-06,
|
| 469 |
+
"loss": 0.2061,
|
| 470 |
"step": 1550
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.196875,
|
| 474 |
+
"grad_norm": 6.948727130889893,
|
| 475 |
"learning_rate": 5.355833333333334e-06,
|
| 476 |
+
"loss": 0.2166,
|
| 477 |
"step": 1575
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 0.2,
|
| 481 |
+
"grad_norm": 7.243198394775391,
|
| 482 |
"learning_rate": 5.335000000000001e-06,
|
| 483 |
+
"loss": 0.278,
|
| 484 |
"step": 1600
|
| 485 |
},
|
| 486 |
{
|
| 487 |
"epoch": 0.203125,
|
| 488 |
+
"grad_norm": 4.161729335784912,
|
| 489 |
"learning_rate": 5.3141666666666665e-06,
|
| 490 |
+
"loss": 0.1822,
|
| 491 |
"step": 1625
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.20625,
|
| 495 |
+
"grad_norm": 4.0524516105651855,
|
| 496 |
"learning_rate": 5.293333333333333e-06,
|
| 497 |
+
"loss": 0.1439,
|
| 498 |
"step": 1650
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 0.209375,
|
| 502 |
+
"grad_norm": 4.263937473297119,
|
| 503 |
"learning_rate": 5.272500000000001e-06,
|
| 504 |
+
"loss": 0.1272,
|
| 505 |
"step": 1675
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.2125,
|
| 509 |
+
"grad_norm": 3.639070987701416,
|
| 510 |
"learning_rate": 5.2516666666666675e-06,
|
| 511 |
+
"loss": 0.1252,
|
| 512 |
"step": 1700
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.215625,
|
| 516 |
+
"grad_norm": 3.305999517440796,
|
| 517 |
"learning_rate": 5.230833333333333e-06,
|
| 518 |
+
"loss": 0.1183,
|
| 519 |
"step": 1725
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 0.21875,
|
| 523 |
+
"grad_norm": 2.888228178024292,
|
| 524 |
"learning_rate": 5.21e-06,
|
| 525 |
+
"loss": 0.1101,
|
| 526 |
"step": 1750
|
| 527 |
},
|
| 528 |
{
|
| 529 |
+
"epoch": 0.221875,
|
| 530 |
+
"grad_norm": 5.215433597564697,
|
| 531 |
"learning_rate": 5.189166666666667e-06,
|
| 532 |
+
"loss": 0.1562,
|
| 533 |
"step": 1775
|
| 534 |
},
|
| 535 |
{
|
| 536 |
+
"epoch": 0.225,
|
| 537 |
+
"grad_norm": 4.641208171844482,
|
| 538 |
"learning_rate": 5.1683333333333335e-06,
|
| 539 |
+
"loss": 0.192,
|
| 540 |
"step": 1800
|
| 541 |
},
|
| 542 |
{
|
| 543 |
+
"epoch": 0.228125,
|
| 544 |
+
"grad_norm": 6.326239109039307,
|
| 545 |
"learning_rate": 5.1475e-06,
|
| 546 |
+
"loss": 0.2123,
|
| 547 |
"step": 1825
|
| 548 |
},
|
| 549 |
{
|
| 550 |
+
"epoch": 0.23125,
|
| 551 |
+
"grad_norm": 7.012181282043457,
|
| 552 |
"learning_rate": 5.126666666666667e-06,
|
| 553 |
+
"loss": 0.2123,
|
| 554 |
"step": 1850
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 0.234375,
|
| 558 |
+
"grad_norm": 5.915773391723633,
|
| 559 |
"learning_rate": 5.105833333333334e-06,
|
| 560 |
+
"loss": 0.1966,
|
| 561 |
"step": 1875
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 0.2375,
|
| 565 |
+
"grad_norm": 5.602591037750244,
|
| 566 |
"learning_rate": 5.0850000000000004e-06,
|
| 567 |
+
"loss": 0.1913,
|
| 568 |
"step": 1900
|
| 569 |
},
|
| 570 |
{
|
| 571 |
+
"epoch": 0.240625,
|
| 572 |
+
"grad_norm": 5.096761703491211,
|
| 573 |
"learning_rate": 5.064166666666667e-06,
|
| 574 |
+
"loss": 0.2013,
|
| 575 |
"step": 1925
|
| 576 |
},
|
| 577 |
{
|
| 578 |
+
"epoch": 0.24375,
|
| 579 |
+
"grad_norm": 11.189478874206543,
|
| 580 |
"learning_rate": 5.043333333333333e-06,
|
| 581 |
+
"loss": 0.587,
|
| 582 |
"step": 1950
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"epoch": 0.246875,
|
| 586 |
+
"grad_norm": 6.038069248199463,
|
| 587 |
"learning_rate": 5.0225e-06,
|
| 588 |
+
"loss": 0.3751,
|
| 589 |
"step": 1975
|
| 590 |
},
|
| 591 |
{
|
| 592 |
+
"epoch": 0.25,
|
| 593 |
+
"grad_norm": 6.957058429718018,
|
| 594 |
"learning_rate": 5.001666666666667e-06,
|
| 595 |
+
"loss": 0.2617,
|
| 596 |
"step": 2000
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 0.25,
|
| 600 |
+
"eval_loss": 0.2810758948326111,
|
| 601 |
+
"eval_runtime": 423.474,
|
| 602 |
+
"eval_samples_per_second": 4.968,
|
| 603 |
+
"eval_steps_per_second": 0.621,
|
| 604 |
+
"eval_wer": 16.58408899691502,
|
| 605 |
"step": 2000
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 0.253125,
|
| 609 |
+
"grad_norm": 5.579774856567383,
|
| 610 |
"learning_rate": 4.980833333333334e-06,
|
| 611 |
+
"loss": 0.2469,
|
| 612 |
"step": 2025
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"epoch": 0.25625,
|
| 616 |
+
"grad_norm": 5.4091620445251465,
|
| 617 |
"learning_rate": 4.96e-06,
|
| 618 |
+
"loss": 0.2291,
|
| 619 |
"step": 2050
|
| 620 |
},
|
| 621 |
{
|
| 622 |
+
"epoch": 0.259375,
|
| 623 |
+
"grad_norm": 4.21276330947876,
|
| 624 |
"learning_rate": 4.939166666666667e-06,
|
| 625 |
+
"loss": 0.1921,
|
| 626 |
"step": 2075
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 0.2625,
|
| 630 |
+
"grad_norm": 3.7961809635162354,
|
| 631 |
"learning_rate": 4.918333333333334e-06,
|
| 632 |
+
"loss": 0.1341,
|
| 633 |
"step": 2100
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 0.265625,
|
| 637 |
+
"grad_norm": 3.204401969909668,
|
| 638 |
"learning_rate": 4.8975e-06,
|
| 639 |
+
"loss": 0.107,
|
| 640 |
"step": 2125
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 0.26875,
|
| 644 |
+
"grad_norm": 3.3962206840515137,
|
| 645 |
"learning_rate": 4.876666666666667e-06,
|
| 646 |
+
"loss": 0.1206,
|
| 647 |
"step": 2150
|
| 648 |
},
|
| 649 |
{
|
| 650 |
+
"epoch": 0.271875,
|
| 651 |
+
"grad_norm": 3.666205644607544,
|
| 652 |
"learning_rate": 4.8558333333333336e-06,
|
| 653 |
+
"loss": 0.1059,
|
| 654 |
"step": 2175
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 0.275,
|
| 658 |
+
"grad_norm": 3.0882160663604736,
|
| 659 |
"learning_rate": 4.835e-06,
|
| 660 |
+
"loss": 0.097,
|
| 661 |
"step": 2200
|
| 662 |
},
|
| 663 |
{
|
| 664 |
+
"epoch": 0.278125,
|
| 665 |
+
"grad_norm": 3.348127603530884,
|
| 666 |
"learning_rate": 4.814166666666667e-06,
|
| 667 |
+
"loss": 0.0965,
|
| 668 |
"step": 2225
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 0.28125,
|
| 672 |
+
"grad_norm": 2.4447720050811768,
|
| 673 |
"learning_rate": 4.793333333333334e-06,
|
| 674 |
+
"loss": 0.1003,
|
| 675 |
"step": 2250
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 0.284375,
|
| 679 |
+
"grad_norm": 4.615771293640137,
|
| 680 |
"learning_rate": 4.7725e-06,
|
| 681 |
+
"loss": 0.1012,
|
| 682 |
"step": 2275
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 0.2875,
|
| 686 |
+
"grad_norm": 3.513615131378174,
|
| 687 |
"learning_rate": 4.751666666666666e-06,
|
| 688 |
+
"loss": 0.107,
|
| 689 |
"step": 2300
|
| 690 |
},
|
| 691 |
{
|
| 692 |
+
"epoch": 0.290625,
|
| 693 |
+
"grad_norm": 3.266713857650757,
|
| 694 |
"learning_rate": 4.730833333333334e-06,
|
| 695 |
+
"loss": 0.1073,
|
| 696 |
"step": 2325
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 0.29375,
|
| 700 |
+
"grad_norm": 3.3199737071990967,
|
| 701 |
"learning_rate": 4.710000000000001e-06,
|
| 702 |
+
"loss": 0.1077,
|
| 703 |
"step": 2350
|
| 704 |
},
|
| 705 |
{
|
| 706 |
+
"epoch": 0.296875,
|
| 707 |
+
"grad_norm": 2.671046018600464,
|
| 708 |
"learning_rate": 4.6891666666666665e-06,
|
| 709 |
+
"loss": 0.1031,
|
| 710 |
"step": 2375
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 0.3,
|
| 714 |
+
"grad_norm": 3.041368007659912,
|
| 715 |
"learning_rate": 4.668333333333333e-06,
|
| 716 |
+
"loss": 0.1053,
|
| 717 |
"step": 2400
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 0.303125,
|
| 721 |
+
"grad_norm": 3.7318310737609863,
|
| 722 |
"learning_rate": 4.647500000000001e-06,
|
| 723 |
+
"loss": 0.1439,
|
| 724 |
"step": 2425
|
| 725 |
},
|
| 726 |
{
|
| 727 |
+
"epoch": 0.30625,
|
| 728 |
+
"grad_norm": 7.745230674743652,
|
| 729 |
"learning_rate": 4.626666666666667e-06,
|
| 730 |
+
"loss": 0.1686,
|
| 731 |
"step": 2450
|
| 732 |
},
|
| 733 |
{
|
| 734 |
+
"epoch": 0.309375,
|
| 735 |
+
"grad_norm": 5.621090412139893,
|
| 736 |
"learning_rate": 4.605833333333333e-06,
|
| 737 |
+
"loss": 0.1794,
|
| 738 |
"step": 2475
|
| 739 |
},
|
| 740 |
{
|
| 741 |
+
"epoch": 0.3125,
|
| 742 |
+
"grad_norm": 4.522252559661865,
|
| 743 |
"learning_rate": 4.585e-06,
|
| 744 |
+
"loss": 0.1646,
|
| 745 |
"step": 2500
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 0.3125,
|
| 749 |
+
"eval_loss": 0.2725737690925598,
|
| 750 |
+
"eval_runtime": 422.3202,
|
| 751 |
+
"eval_samples_per_second": 4.982,
|
| 752 |
+
"eval_steps_per_second": 0.623,
|
| 753 |
+
"eval_wer": 13.891745349163317,
|
| 754 |
"step": 2500
|
| 755 |
},
|
| 756 |
{
|
| 757 |
+
"epoch": 0.315625,
|
| 758 |
+
"grad_norm": 5.429344177246094,
|
| 759 |
"learning_rate": 4.564166666666667e-06,
|
| 760 |
+
"loss": 0.188,
|
| 761 |
"step": 2525
|
| 762 |
},
|
| 763 |
{
|
| 764 |
+
"epoch": 0.31875,
|
| 765 |
+
"grad_norm": 6.522274971008301,
|
| 766 |
"learning_rate": 4.543333333333334e-06,
|
| 767 |
+
"loss": 0.1988,
|
| 768 |
"step": 2550
|
| 769 |
},
|
| 770 |
{
|
| 771 |
+
"epoch": 0.321875,
|
| 772 |
+
"grad_norm": 6.914414405822754,
|
| 773 |
"learning_rate": 4.5225e-06,
|
| 774 |
+
"loss": 0.1915,
|
| 775 |
"step": 2575
|
| 776 |
},
|
| 777 |
{
|
| 778 |
+
"epoch": 0.325,
|
| 779 |
+
"grad_norm": 5.277989387512207,
|
| 780 |
"learning_rate": 4.501666666666667e-06,
|
| 781 |
+
"loss": 0.2011,
|
| 782 |
"step": 2600
|
| 783 |
},
|
| 784 |
{
|
| 785 |
+
"epoch": 0.328125,
|
| 786 |
+
"grad_norm": 5.887080192565918,
|
| 787 |
+
"learning_rate": 4.481666666666666e-06,
|
| 788 |
+
"loss": 0.1717,
|
| 789 |
"step": 2625
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 0.33125,
|
| 793 |
+
"grad_norm": 5.534679412841797,
|
| 794 |
+
"learning_rate": 4.460833333333333e-06,
|
| 795 |
+
"loss": 0.1696,
|
| 796 |
"step": 2650
|
| 797 |
},
|
| 798 |
{
|
| 799 |
+
"epoch": 0.334375,
|
| 800 |
+
"grad_norm": 5.044061183929443,
|
| 801 |
+
"learning_rate": 4.440000000000001e-06,
|
| 802 |
+
"loss": 0.1809,
|
| 803 |
"step": 2675
|
| 804 |
},
|
| 805 |
{
|
| 806 |
+
"epoch": 0.3375,
|
| 807 |
+
"grad_norm": 6.7157769203186035,
|
| 808 |
+
"learning_rate": 4.4191666666666666e-06,
|
| 809 |
+
"loss": 0.1763,
|
| 810 |
"step": 2700
|
| 811 |
},
|
| 812 |
{
|
| 813 |
+
"epoch": 0.340625,
|
| 814 |
+
"grad_norm": 5.815196990966797,
|
| 815 |
+
"learning_rate": 4.398333333333333e-06,
|
| 816 |
+
"loss": 0.1626,
|
| 817 |
"step": 2725
|
| 818 |
},
|
| 819 |
{
|
| 820 |
+
"epoch": 0.34375,
|
| 821 |
+
"grad_norm": 5.573261737823486,
|
| 822 |
+
"learning_rate": 4.3775e-06,
|
| 823 |
+
"loss": 0.1599,
|
| 824 |
"step": 2750
|
| 825 |
},
|
| 826 |
{
|
| 827 |
+
"epoch": 0.346875,
|
| 828 |
+
"grad_norm": 4.859520435333252,
|
| 829 |
+
"learning_rate": 4.356666666666667e-06,
|
| 830 |
+
"loss": 0.1523,
|
| 831 |
"step": 2775
|
| 832 |
},
|
| 833 |
{
|
| 834 |
+
"epoch": 0.35,
|
| 835 |
+
"grad_norm": 5.533599853515625,
|
| 836 |
+
"learning_rate": 4.3358333333333335e-06,
|
| 837 |
+
"loss": 0.1672,
|
| 838 |
"step": 2800
|
| 839 |
},
|
| 840 |
{
|
| 841 |
+
"epoch": 0.353125,
|
| 842 |
+
"grad_norm": 6.5196123123168945,
|
| 843 |
+
"learning_rate": 4.315e-06,
|
| 844 |
+
"loss": 0.1652,
|
| 845 |
"step": 2825
|
| 846 |
},
|
| 847 |
{
|
| 848 |
+
"epoch": 0.35625,
|
| 849 |
+
"grad_norm": 5.192087173461914,
|
| 850 |
+
"learning_rate": 4.294166666666667e-06,
|
| 851 |
+
"loss": 0.1587,
|
| 852 |
"step": 2850
|
| 853 |
},
|
| 854 |
{
|
| 855 |
+
"epoch": 0.359375,
|
| 856 |
+
"grad_norm": 4.697035789489746,
|
| 857 |
+
"learning_rate": 4.273333333333334e-06,
|
| 858 |
+
"loss": 0.1697,
|
| 859 |
"step": 2875
|
| 860 |
},
|
| 861 |
{
|
| 862 |
+
"epoch": 0.3625,
|
| 863 |
+
"grad_norm": 4.622238636016846,
|
| 864 |
+
"learning_rate": 4.2525e-06,
|
| 865 |
+
"loss": 0.1431,
|
| 866 |
"step": 2900
|
| 867 |
},
|
| 868 |
{
|
| 869 |
+
"epoch": 0.365625,
|
| 870 |
+
"grad_norm": 2.833672046661377,
|
| 871 |
+
"learning_rate": 4.231666666666667e-06,
|
| 872 |
+
"loss": 0.1187,
|
| 873 |
"step": 2925
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 0.36875,
|
| 877 |
+
"grad_norm": 2.756924867630005,
|
| 878 |
+
"learning_rate": 4.210833333333333e-06,
|
| 879 |
+
"loss": 0.098,
|
| 880 |
"step": 2950
|
| 881 |
},
|
| 882 |
{
|
| 883 |
+
"epoch": 0.371875,
|
| 884 |
+
"grad_norm": 4.0311598777771,
|
| 885 |
+
"learning_rate": 4.1900000000000005e-06,
|
| 886 |
+
"loss": 0.0956,
|
| 887 |
"step": 2975
|
| 888 |
},
|
| 889 |
{
|
| 890 |
+
"epoch": 0.375,
|
| 891 |
+
"grad_norm": 3.7465600967407227,
|
| 892 |
+
"learning_rate": 4.169166666666667e-06,
|
| 893 |
+
"loss": 0.0934,
|
| 894 |
"step": 3000
|
| 895 |
},
|
| 896 |
{
|
| 897 |
+
"epoch": 0.375,
|
| 898 |
+
"eval_loss": 0.2532876133918762,
|
| 899 |
+
"eval_runtime": 423.4639,
|
| 900 |
+
"eval_samples_per_second": 4.969,
|
| 901 |
+
"eval_steps_per_second": 0.621,
|
| 902 |
+
"eval_wer": 14.02729737309526,
|
| 903 |
"step": 3000
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 0.378125,
|
| 907 |
+
"grad_norm": 3.0190958976745605,
|
| 908 |
+
"learning_rate": 4.148333333333333e-06,
|
| 909 |
+
"loss": 0.1074,
|
| 910 |
"step": 3025
|
| 911 |
},
|
| 912 |
{
|
| 913 |
+
"epoch": 0.38125,
|
| 914 |
+
"grad_norm": 4.197691917419434,
|
| 915 |
+
"learning_rate": 4.1275e-06,
|
| 916 |
+
"loss": 0.0952,
|
| 917 |
"step": 3050
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 0.384375,
|
| 921 |
+
"grad_norm": 4.304806709289551,
|
| 922 |
+
"learning_rate": 4.106666666666667e-06,
|
| 923 |
+
"loss": 0.1005,
|
| 924 |
"step": 3075
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"epoch": 0.3875,
|
| 928 |
+
"grad_norm": 3.895831823348999,
|
| 929 |
+
"learning_rate": 4.085833333333334e-06,
|
| 930 |
+
"loss": 0.1027,
|
| 931 |
"step": 3100
|
| 932 |
},
|
| 933 |
{
|
| 934 |
+
"epoch": 0.390625,
|
| 935 |
+
"grad_norm": 2.959507703781128,
|
| 936 |
+
"learning_rate": 4.065e-06,
|
| 937 |
+
"loss": 0.0863,
|
| 938 |
"step": 3125
|
| 939 |
},
|
| 940 |
{
|
| 941 |
+
"epoch": 0.39375,
|
| 942 |
+
"grad_norm": 2.9817397594451904,
|
| 943 |
+
"learning_rate": 4.044166666666667e-06,
|
| 944 |
+
"loss": 0.0773,
|
| 945 |
"step": 3150
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 0.396875,
|
| 949 |
+
"grad_norm": 3.26869535446167,
|
| 950 |
+
"learning_rate": 4.0233333333333335e-06,
|
| 951 |
+
"loss": 0.0885,
|
| 952 |
"step": 3175
|
| 953 |
},
|
| 954 |
{
|
| 955 |
+
"epoch": 0.4,
|
| 956 |
+
"grad_norm": 4.37345552444458,
|
| 957 |
+
"learning_rate": 4.0025e-06,
|
| 958 |
+
"loss": 0.0899,
|
| 959 |
"step": 3200
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 0.403125,
|
| 963 |
+
"grad_norm": 4.3736371994018555,
|
| 964 |
+
"learning_rate": 3.981666666666667e-06,
|
| 965 |
+
"loss": 0.11,
|
| 966 |
"step": 3225
|
| 967 |
},
|
| 968 |
{
|
| 969 |
+
"epoch": 0.40625,
|
| 970 |
+
"grad_norm": 3.9477250576019287,
|
| 971 |
+
"learning_rate": 3.960833333333334e-06,
|
| 972 |
+
"loss": 0.1579,
|
| 973 |
"step": 3250
|
| 974 |
},
|
| 975 |
{
|
| 976 |
+
"epoch": 0.409375,
|
| 977 |
+
"grad_norm": 4.6521315574646,
|
| 978 |
+
"learning_rate": 3.9399999999999995e-06,
|
| 979 |
+
"loss": 0.1409,
|
| 980 |
"step": 3275
|
| 981 |
},
|
| 982 |
{
|
| 983 |
+
"epoch": 0.4125,
|
| 984 |
+
"grad_norm": 3.4862189292907715,
|
| 985 |
+
"learning_rate": 3.919166666666667e-06,
|
| 986 |
+
"loss": 0.1368,
|
| 987 |
"step": 3300
|
| 988 |
},
|
| 989 |
{
|
| 990 |
+
"epoch": 0.415625,
|
| 991 |
+
"grad_norm": 4.412058353424072,
|
| 992 |
+
"learning_rate": 3.898333333333334e-06,
|
| 993 |
+
"loss": 0.136,
|
| 994 |
"step": 3325
|
| 995 |
},
|
| 996 |
{
|
| 997 |
+
"epoch": 0.41875,
|
| 998 |
+
"grad_norm": 5.891563892364502,
|
| 999 |
+
"learning_rate": 3.8775e-06,
|
| 1000 |
+
"loss": 0.1514,
|
| 1001 |
"step": 3350
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
+
"epoch": 0.421875,
|
| 1005 |
+
"grad_norm": 4.186588287353516,
|
| 1006 |
+
"learning_rate": 3.8566666666666664e-06,
|
| 1007 |
+
"loss": 0.1574,
|
| 1008 |
"step": 3375
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
+
"epoch": 0.425,
|
| 1012 |
+
"grad_norm": 3.0448837280273438,
|
| 1013 |
+
"learning_rate": 3.835833333333334e-06,
|
| 1014 |
+
"loss": 0.1198,
|
| 1015 |
"step": 3400
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
+
"epoch": 0.428125,
|
| 1019 |
+
"grad_norm": 3.597623825073242,
|
| 1020 |
+
"learning_rate": 3.815000000000001e-06,
|
| 1021 |
+
"loss": 0.1079,
|
| 1022 |
"step": 3425
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
+
"epoch": 0.43125,
|
| 1026 |
+
"grad_norm": 4.119424819946289,
|
| 1027 |
+
"learning_rate": 3.7941666666666666e-06,
|
| 1028 |
+
"loss": 0.1015,
|
| 1029 |
"step": 3450
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
+
"epoch": 0.434375,
|
| 1033 |
+
"grad_norm": 4.269451141357422,
|
| 1034 |
+
"learning_rate": 3.7733333333333338e-06,
|
| 1035 |
+
"loss": 0.0931,
|
| 1036 |
"step": 3475
|
| 1037 |
},
|
| 1038 |
{
|
| 1039 |
+
"epoch": 0.4375,
|
| 1040 |
+
"grad_norm": 3.5041770935058594,
|
| 1041 |
+
"learning_rate": 3.7525000000000005e-06,
|
| 1042 |
+
"loss": 0.1016,
|
| 1043 |
"step": 3500
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
+
"epoch": 0.4375,
|
| 1047 |
+
"eval_loss": 0.23310637474060059,
|
| 1048 |
+
"eval_runtime": 422.6444,
|
| 1049 |
+
"eval_samples_per_second": 4.978,
|
| 1050 |
+
"eval_steps_per_second": 0.622,
|
| 1051 |
+
"eval_wer": 12.162288492100588,
|
| 1052 |
"step": 3500
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
+
"epoch": 0.440625,
|
| 1056 |
+
"grad_norm": 3.1294009685516357,
|
| 1057 |
+
"learning_rate": 3.731666666666667e-06,
|
| 1058 |
+
"loss": 0.0875,
|
| 1059 |
"step": 3525
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 0.44375,
|
| 1063 |
+
"grad_norm": 5.66436767578125,
|
| 1064 |
+
"learning_rate": 3.7108333333333335e-06,
|
| 1065 |
+
"loss": 0.0852,
|
| 1066 |
"step": 3550
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
+
"epoch": 0.446875,
|
| 1070 |
+
"grad_norm": 3.854020833969116,
|
| 1071 |
+
"learning_rate": 3.6900000000000002e-06,
|
| 1072 |
+
"loss": 0.1009,
|
| 1073 |
"step": 3575
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
+
"epoch": 0.45,
|
| 1077 |
+
"grad_norm": 2.3957531452178955,
|
| 1078 |
+
"learning_rate": 3.6691666666666665e-06,
|
| 1079 |
+
"loss": 0.0884,
|
| 1080 |
"step": 3600
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
+
"epoch": 0.453125,
|
| 1084 |
+
"grad_norm": 2.6974878311157227,
|
| 1085 |
+
"learning_rate": 3.6483333333333333e-06,
|
| 1086 |
+
"loss": 0.0878,
|
| 1087 |
"step": 3625
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
+
"epoch": 0.45625,
|
| 1091 |
+
"grad_norm": 2.4071967601776123,
|
| 1092 |
+
"learning_rate": 3.6275000000000004e-06,
|
| 1093 |
+
"loss": 0.0822,
|
| 1094 |
"step": 3650
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 0.459375,
|
| 1098 |
+
"grad_norm": 3.604412078857422,
|
| 1099 |
+
"learning_rate": 3.6066666666666667e-06,
|
| 1100 |
+
"loss": 0.0885,
|
| 1101 |
"step": 3675
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 0.4625,
|
| 1105 |
+
"grad_norm": 3.1856794357299805,
|
| 1106 |
+
"learning_rate": 3.5858333333333334e-06,
|
| 1107 |
+
"loss": 0.0795,
|
| 1108 |
"step": 3700
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
+
"epoch": 0.465625,
|
| 1112 |
+
"grad_norm": 2.82853102684021,
|
| 1113 |
+
"learning_rate": 3.565e-06,
|
| 1114 |
+
"loss": 0.0839,
|
| 1115 |
"step": 3725
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
+
"epoch": 0.46875,
|
| 1119 |
+
"grad_norm": 3.0090298652648926,
|
| 1120 |
+
"learning_rate": 3.5441666666666673e-06,
|
| 1121 |
+
"loss": 0.0861,
|
| 1122 |
"step": 3750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
+
"epoch": 0.471875,
|
| 1126 |
+
"grad_norm": 4.2935028076171875,
|
| 1127 |
+
"learning_rate": 3.523333333333333e-06,
|
| 1128 |
+
"loss": 0.0902,
|
| 1129 |
"step": 3775
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
+
"epoch": 0.475,
|
| 1133 |
+
"grad_norm": 2.894404172897339,
|
| 1134 |
+
"learning_rate": 3.5025000000000003e-06,
|
| 1135 |
+
"loss": 0.0897,
|
| 1136 |
"step": 3800
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"epoch": 0.478125,
|
| 1140 |
+
"grad_norm": 2.974064826965332,
|
| 1141 |
+
"learning_rate": 3.481666666666667e-06,
|
| 1142 |
+
"loss": 0.0846,
|
| 1143 |
"step": 3825
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 0.48125,
|
| 1147 |
+
"grad_norm": 1.8990185260772705,
|
| 1148 |
+
"learning_rate": 3.4608333333333334e-06,
|
| 1149 |
+
"loss": 0.0782,
|
| 1150 |
"step": 3850
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
+
"epoch": 0.484375,
|
| 1154 |
+
"grad_norm": 3.043663740158081,
|
| 1155 |
+
"learning_rate": 3.44e-06,
|
| 1156 |
+
"loss": 0.0923,
|
| 1157 |
"step": 3875
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
+
"epoch": 0.4875,
|
| 1161 |
+
"grad_norm": 4.826819896697998,
|
| 1162 |
+
"learning_rate": 3.4191666666666672e-06,
|
| 1163 |
+
"loss": 0.1242,
|
| 1164 |
"step": 3900
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
+
"epoch": 0.490625,
|
| 1168 |
+
"grad_norm": 4.978621006011963,
|
| 1169 |
+
"learning_rate": 3.398333333333333e-06,
|
| 1170 |
+
"loss": 0.1445,
|
| 1171 |
"step": 3925
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 0.49375,
|
| 1175 |
+
"grad_norm": 4.0232977867126465,
|
| 1176 |
+
"learning_rate": 3.3775000000000003e-06,
|
| 1177 |
+
"loss": 0.1514,
|
| 1178 |
"step": 3950
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
+
"epoch": 0.496875,
|
| 1182 |
+
"grad_norm": 5.176286220550537,
|
| 1183 |
+
"learning_rate": 3.356666666666667e-06,
|
| 1184 |
+
"loss": 0.1532,
|
| 1185 |
"step": 3975
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 0.5,
|
| 1189 |
+
"grad_norm": 4.363390922546387,
|
| 1190 |
+
"learning_rate": 3.3358333333333333e-06,
|
| 1191 |
+
"loss": 0.1454,
|
| 1192 |
"step": 4000
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
+
"epoch": 0.5,
|
| 1196 |
+
"eval_loss": 0.22985798120498657,
|
| 1197 |
+
"eval_runtime": 420.4785,
|
| 1198 |
+
"eval_samples_per_second": 5.004,
|
| 1199 |
+
"eval_steps_per_second": 0.625,
|
| 1200 |
+
"eval_wer": 11.55464148826774,
|
| 1201 |
"step": 4000
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
+
"epoch": 0.503125,
|
| 1205 |
+
"grad_norm": 4.654495716094971,
|
| 1206 |
+
"learning_rate": 3.315e-06,
|
| 1207 |
+
"loss": 0.1331,
|
| 1208 |
"step": 4025
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
+
"epoch": 0.50625,
|
| 1212 |
+
"grad_norm": 3.879624605178833,
|
| 1213 |
+
"learning_rate": 3.2941666666666667e-06,
|
| 1214 |
+
"loss": 0.1029,
|
| 1215 |
"step": 4050
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
+
"epoch": 0.509375,
|
| 1219 |
+
"grad_norm": 3.2961080074310303,
|
| 1220 |
+
"learning_rate": 3.273333333333334e-06,
|
| 1221 |
+
"loss": 0.0842,
|
| 1222 |
"step": 4075
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
+
"epoch": 0.5125,
|
| 1226 |
+
"grad_norm": 2.5998172760009766,
|
| 1227 |
+
"learning_rate": 3.2525e-06,
|
| 1228 |
+
"loss": 0.0846,
|
| 1229 |
"step": 4100
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
+
"epoch": 0.515625,
|
| 1233 |
+
"grad_norm": 2.965538501739502,
|
| 1234 |
+
"learning_rate": 3.231666666666667e-06,
|
| 1235 |
+
"loss": 0.0858,
|
| 1236 |
"step": 4125
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
+
"epoch": 0.51875,
|
| 1240 |
+
"grad_norm": 2.8639001846313477,
|
| 1241 |
+
"learning_rate": 3.2108333333333336e-06,
|
| 1242 |
+
"loss": 0.0836,
|
| 1243 |
"step": 4150
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
+
"epoch": 0.521875,
|
| 1247 |
+
"grad_norm": 3.651369094848633,
|
| 1248 |
+
"learning_rate": 3.19e-06,
|
| 1249 |
+
"loss": 0.0852,
|
| 1250 |
"step": 4175
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
+
"epoch": 0.525,
|
| 1254 |
+
"grad_norm": 4.119676113128662,
|
| 1255 |
+
"learning_rate": 3.1691666666666667e-06,
|
| 1256 |
+
"loss": 0.0886,
|
| 1257 |
"step": 4200
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
+
"epoch": 0.528125,
|
| 1261 |
+
"grad_norm": 2.3661563396453857,
|
| 1262 |
+
"learning_rate": 3.148333333333334e-06,
|
| 1263 |
+
"loss": 0.0732,
|
| 1264 |
"step": 4225
|
| 1265 |
},
|
| 1266 |
{
|
| 1267 |
+
"epoch": 0.53125,
|
| 1268 |
+
"grad_norm": 4.026716232299805,
|
| 1269 |
+
"learning_rate": 3.1274999999999997e-06,
|
| 1270 |
+
"loss": 0.0739,
|
| 1271 |
"step": 4250
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
+
"epoch": 0.534375,
|
| 1275 |
+
"grad_norm": 2.2528867721557617,
|
| 1276 |
+
"learning_rate": 3.106666666666667e-06,
|
| 1277 |
+
"loss": 0.0682,
|
| 1278 |
"step": 4275
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
+
"epoch": 0.5375,
|
| 1282 |
+
"grad_norm": 3.5776236057281494,
|
| 1283 |
+
"learning_rate": 3.0858333333333336e-06,
|
| 1284 |
+
"loss": 0.0729,
|
| 1285 |
"step": 4300
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
+
"epoch": 0.540625,
|
| 1289 |
+
"grad_norm": 2.0452890396118164,
|
| 1290 |
+
"learning_rate": 3.0650000000000003e-06,
|
| 1291 |
+
"loss": 0.0676,
|
| 1292 |
"step": 4325
|
| 1293 |
},
|
| 1294 |
{
|
| 1295 |
+
"epoch": 0.54375,
|
| 1296 |
+
"grad_norm": 3.9987590312957764,
|
| 1297 |
+
"learning_rate": 3.0441666666666666e-06,
|
| 1298 |
+
"loss": 0.081,
|
| 1299 |
"step": 4350
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
+
"epoch": 0.546875,
|
| 1303 |
+
"grad_norm": 4.795535087585449,
|
| 1304 |
+
"learning_rate": 3.0233333333333338e-06,
|
| 1305 |
+
"loss": 0.126,
|
| 1306 |
"step": 4375
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
+
"epoch": 0.55,
|
| 1310 |
+
"grad_norm": 4.841895580291748,
|
| 1311 |
+
"learning_rate": 3.0025e-06,
|
| 1312 |
+
"loss": 0.1363,
|
| 1313 |
"step": 4400
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
+
"epoch": 0.553125,
|
| 1317 |
+
"grad_norm": 6.109241962432861,
|
| 1318 |
+
"learning_rate": 2.9816666666666668e-06,
|
| 1319 |
+
"loss": 0.1325,
|
| 1320 |
"step": 4425
|
| 1321 |
},
|
| 1322 |
{
|
| 1323 |
+
"epoch": 0.55625,
|
| 1324 |
+
"grad_norm": 4.623290538787842,
|
| 1325 |
+
"learning_rate": 2.9608333333333335e-06,
|
| 1326 |
+
"loss": 0.123,
|
| 1327 |
"step": 4450
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
+
"epoch": 0.559375,
|
| 1331 |
+
"grad_norm": 3.5602827072143555,
|
| 1332 |
+
"learning_rate": 2.9400000000000002e-06,
|
| 1333 |
+
"loss": 0.1361,
|
| 1334 |
"step": 4475
|
| 1335 |
},
|
| 1336 |
{
|
| 1337 |
+
"epoch": 0.5625,
|
| 1338 |
+
"grad_norm": 4.564285755157471,
|
| 1339 |
+
"learning_rate": 2.919166666666667e-06,
|
| 1340 |
+
"loss": 0.1502,
|
| 1341 |
"step": 4500
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
+
"epoch": 0.5625,
|
| 1345 |
+
"eval_loss": 0.23331405222415924,
|
| 1346 |
+
"eval_runtime": 420.0095,
|
| 1347 |
+
"eval_samples_per_second": 5.009,
|
| 1348 |
+
"eval_steps_per_second": 0.626,
|
| 1349 |
+
"eval_wer": 12.400673085911938,
|
| 1350 |
"step": 4500
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
+
"epoch": 0.565625,
|
| 1354 |
+
"grad_norm": 3.0876269340515137,
|
| 1355 |
+
"learning_rate": 2.8983333333333333e-06,
|
| 1356 |
+
"loss": 0.1061,
|
| 1357 |
"step": 4525
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
+
"epoch": 0.56875,
|
| 1361 |
+
"grad_norm": 2.9698376655578613,
|
| 1362 |
+
"learning_rate": 2.8775e-06,
|
| 1363 |
+
"loss": 0.0809,
|
| 1364 |
"step": 4550
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
+
"epoch": 0.571875,
|
| 1368 |
+
"grad_norm": 2.491389274597168,
|
| 1369 |
+
"learning_rate": 2.8566666666666667e-06,
|
| 1370 |
+
"loss": 0.0769,
|
| 1371 |
"step": 4575
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
+
"epoch": 0.575,
|
| 1375 |
+
"grad_norm": 3.7412362098693848,
|
| 1376 |
+
"learning_rate": 2.8358333333333334e-06,
|
| 1377 |
+
"loss": 0.0632,
|
| 1378 |
"step": 4600
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
+
"epoch": 0.578125,
|
| 1382 |
+
"grad_norm": 3.1096959114074707,
|
| 1383 |
+
"learning_rate": 2.815e-06,
|
| 1384 |
+
"loss": 0.0731,
|
| 1385 |
"step": 4625
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
+
"epoch": 0.58125,
|
| 1389 |
+
"grad_norm": 3.5074427127838135,
|
| 1390 |
+
"learning_rate": 2.794166666666667e-06,
|
| 1391 |
+
"loss": 0.0724,
|
| 1392 |
"step": 4650
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
+
"epoch": 0.584375,
|
| 1396 |
+
"grad_norm": 2.3038923740386963,
|
| 1397 |
+
"learning_rate": 2.773333333333333e-06,
|
| 1398 |
+
"loss": 0.0663,
|
| 1399 |
"step": 4675
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
+
"epoch": 0.5875,
|
| 1403 |
+
"grad_norm": 2.202913522720337,
|
| 1404 |
+
"learning_rate": 2.7525000000000003e-06,
|
| 1405 |
+
"loss": 0.0624,
|
| 1406 |
"step": 4700
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
+
"epoch": 0.590625,
|
| 1410 |
+
"grad_norm": 3.385122060775757,
|
| 1411 |
+
"learning_rate": 2.7316666666666666e-06,
|
| 1412 |
+
"loss": 0.0764,
|
| 1413 |
"step": 4725
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
+
"epoch": 0.59375,
|
| 1417 |
+
"grad_norm": 2.382167100906372,
|
| 1418 |
+
"learning_rate": 2.7108333333333338e-06,
|
| 1419 |
+
"loss": 0.0735,
|
| 1420 |
"step": 4750
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
+
"epoch": 0.596875,
|
| 1424 |
+
"grad_norm": 2.670067071914673,
|
| 1425 |
+
"learning_rate": 2.69e-06,
|
| 1426 |
+
"loss": 0.0769,
|
| 1427 |
"step": 4775
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
+
"epoch": 0.6,
|
| 1431 |
+
"grad_norm": 2.72542667388916,
|
| 1432 |
+
"learning_rate": 2.669166666666667e-06,
|
| 1433 |
+
"loss": 0.0699,
|
| 1434 |
"step": 4800
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
+
"epoch": 0.603125,
|
| 1438 |
+
"grad_norm": 3.4631733894348145,
|
| 1439 |
+
"learning_rate": 2.6483333333333335e-06,
|
| 1440 |
+
"loss": 0.0748,
|
| 1441 |
"step": 4825
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
+
"epoch": 0.60625,
|
| 1445 |
+
"grad_norm": 2.1158626079559326,
|
| 1446 |
+
"learning_rate": 2.6275000000000003e-06,
|
| 1447 |
+
"loss": 0.0775,
|
| 1448 |
"step": 4850
|
| 1449 |
},
|
| 1450 |
{
|
| 1451 |
+
"epoch": 0.609375,
|
| 1452 |
+
"grad_norm": 3.326186418533325,
|
| 1453 |
+
"learning_rate": 2.6066666666666666e-06,
|
| 1454 |
+
"loss": 0.0781,
|
| 1455 |
"step": 4875
|
| 1456 |
},
|
| 1457 |
{
|
| 1458 |
+
"epoch": 0.6125,
|
| 1459 |
+
"grad_norm": 3.0408577919006348,
|
| 1460 |
+
"learning_rate": 2.5858333333333337e-06,
|
| 1461 |
+
"loss": 0.0712,
|
| 1462 |
"step": 4900
|
| 1463 |
},
|
| 1464 |
{
|
| 1465 |
+
"epoch": 0.615625,
|
| 1466 |
+
"grad_norm": 2.389796018600464,
|
| 1467 |
+
"learning_rate": 2.565e-06,
|
| 1468 |
+
"loss": 0.0728,
|
| 1469 |
"step": 4925
|
| 1470 |
},
|
| 1471 |
{
|
| 1472 |
+
"epoch": 0.61875,
|
| 1473 |
+
"grad_norm": 2.9459266662597656,
|
| 1474 |
+
"learning_rate": 2.5441666666666667e-06,
|
| 1475 |
+
"loss": 0.0818,
|
| 1476 |
"step": 4950
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
+
"epoch": 0.621875,
|
| 1480 |
+
"grad_norm": 2.216707706451416,
|
| 1481 |
+
"learning_rate": 2.5233333333333335e-06,
|
| 1482 |
+
"loss": 0.0887,
|
| 1483 |
"step": 4975
|
| 1484 |
},
|
| 1485 |
{
|
| 1486 |
+
"epoch": 0.625,
|
| 1487 |
+
"grad_norm": 4.172521114349365,
|
| 1488 |
+
"learning_rate": 2.5025e-06,
|
| 1489 |
+
"loss": 0.0916,
|
| 1490 |
"step": 5000
|
| 1491 |
},
|
| 1492 |
{
|
| 1493 |
+
"epoch": 0.625,
|
| 1494 |
+
"eval_loss": 0.22707216441631317,
|
| 1495 |
+
"eval_runtime": 421.6318,
|
| 1496 |
+
"eval_samples_per_second": 4.99,
|
| 1497 |
+
"eval_steps_per_second": 0.624,
|
| 1498 |
+
"eval_wer": 10.965691315322053,
|
| 1499 |
"step": 5000
|
| 1500 |
},
|
| 1501 |
{
|
| 1502 |
+
"epoch": 0.628125,
|
| 1503 |
+
"grad_norm": 3.083894968032837,
|
| 1504 |
+
"learning_rate": 2.481666666666667e-06,
|
| 1505 |
+
"loss": 0.1084,
|
| 1506 |
"step": 5025
|
| 1507 |
},
|
| 1508 |
{
|
| 1509 |
+
"epoch": 0.63125,
|
| 1510 |
+
"grad_norm": 3.9233384132385254,
|
| 1511 |
+
"learning_rate": 2.460833333333333e-06,
|
| 1512 |
+
"loss": 0.1161,
|
| 1513 |
"step": 5050
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
+
"epoch": 0.634375,
|
| 1517 |
+
"grad_norm": 4.261771202087402,
|
| 1518 |
+
"learning_rate": 2.4400000000000004e-06,
|
| 1519 |
+
"loss": 0.1032,
|
| 1520 |
"step": 5075
|
| 1521 |
},
|
| 1522 |
{
|
| 1523 |
+
"epoch": 0.6375,
|
| 1524 |
+
"grad_norm": 4.277398109436035,
|
| 1525 |
+
"learning_rate": 2.4191666666666667e-06,
|
| 1526 |
+
"loss": 0.1063,
|
| 1527 |
"step": 5100
|
| 1528 |
},
|
| 1529 |
{
|
| 1530 |
+
"epoch": 0.640625,
|
| 1531 |
+
"grad_norm": 4.196617126464844,
|
| 1532 |
+
"learning_rate": 2.3983333333333334e-06,
|
| 1533 |
+
"loss": 0.138,
|
| 1534 |
"step": 5125
|
| 1535 |
},
|
| 1536 |
{
|
| 1537 |
+
"epoch": 0.64375,
|
| 1538 |
+
"grad_norm": 4.253470420837402,
|
| 1539 |
+
"learning_rate": 2.3775e-06,
|
| 1540 |
+
"loss": 0.1308,
|
| 1541 |
"step": 5150
|
| 1542 |
},
|
| 1543 |
{
|
| 1544 |
+
"epoch": 0.646875,
|
| 1545 |
+
"grad_norm": 3.221344470977783,
|
| 1546 |
+
"learning_rate": 2.356666666666667e-06,
|
| 1547 |
+
"loss": 0.1118,
|
| 1548 |
"step": 5175
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
+
"epoch": 0.65,
|
| 1552 |
+
"grad_norm": 3.630119800567627,
|
| 1553 |
+
"learning_rate": 2.335833333333333e-06,
|
| 1554 |
+
"loss": 0.0882,
|
| 1555 |
"step": 5200
|
| 1556 |
},
|
| 1557 |
{
|
| 1558 |
+
"epoch": 0.653125,
|
| 1559 |
+
"grad_norm": 3.146988868713379,
|
| 1560 |
+
"learning_rate": 2.3150000000000003e-06,
|
| 1561 |
+
"loss": 0.08,
|
| 1562 |
"step": 5225
|
| 1563 |
},
|
| 1564 |
{
|
| 1565 |
+
"epoch": 0.65625,
|
| 1566 |
+
"grad_norm": 3.6203882694244385,
|
| 1567 |
+
"learning_rate": 2.2941666666666666e-06,
|
| 1568 |
+
"loss": 0.0769,
|
| 1569 |
"step": 5250
|
| 1570 |
},
|
| 1571 |
{
|
| 1572 |
+
"epoch": 0.659375,
|
| 1573 |
+
"grad_norm": 2.9987850189208984,
|
| 1574 |
+
"learning_rate": 2.2733333333333337e-06,
|
| 1575 |
+
"loss": 0.0775,
|
| 1576 |
"step": 5275
|
| 1577 |
},
|
| 1578 |
{
|
| 1579 |
+
"epoch": 0.6625,
|
| 1580 |
+
"grad_norm": 2.3863637447357178,
|
| 1581 |
+
"learning_rate": 2.2525e-06,
|
| 1582 |
+
"loss": 0.0694,
|
| 1583 |
"step": 5300
|
| 1584 |
},
|
| 1585 |
{
|
| 1586 |
+
"epoch": 0.665625,
|
| 1587 |
+
"grad_norm": 2.7977538108825684,
|
| 1588 |
+
"learning_rate": 2.2316666666666668e-06,
|
| 1589 |
+
"loss": 0.0677,
|
| 1590 |
"step": 5325
|
| 1591 |
},
|
| 1592 |
{
|
| 1593 |
+
"epoch": 0.66875,
|
| 1594 |
+
"grad_norm": 3.0015907287597656,
|
| 1595 |
+
"learning_rate": 2.2108333333333335e-06,
|
| 1596 |
+
"loss": 0.0737,
|
| 1597 |
"step": 5350
|
| 1598 |
},
|
| 1599 |
{
|
| 1600 |
+
"epoch": 0.671875,
|
| 1601 |
+
"grad_norm": 2.2097315788269043,
|
| 1602 |
+
"learning_rate": 2.19e-06,
|
| 1603 |
+
"loss": 0.0671,
|
| 1604 |
"step": 5375
|
| 1605 |
},
|
| 1606 |
{
|
| 1607 |
+
"epoch": 0.675,
|
| 1608 |
+
"grad_norm": 1.3963428735733032,
|
| 1609 |
+
"learning_rate": 2.169166666666667e-06,
|
| 1610 |
+
"loss": 0.0665,
|
| 1611 |
"step": 5400
|
| 1612 |
},
|
| 1613 |
{
|
| 1614 |
+
"epoch": 0.678125,
|
| 1615 |
+
"grad_norm": 4.170374393463135,
|
| 1616 |
+
"learning_rate": 2.1483333333333337e-06,
|
| 1617 |
+
"loss": 0.0807,
|
| 1618 |
"step": 5425
|
| 1619 |
},
|
| 1620 |
{
|
| 1621 |
+
"epoch": 0.68125,
|
| 1622 |
+
"grad_norm": 2.4668662548065186,
|
| 1623 |
+
"learning_rate": 2.1275e-06,
|
| 1624 |
+
"loss": 0.0887,
|
| 1625 |
"step": 5450
|
| 1626 |
},
|
| 1627 |
{
|
| 1628 |
+
"epoch": 0.684375,
|
| 1629 |
+
"grad_norm": 3.388085126876831,
|
| 1630 |
+
"learning_rate": 2.1066666666666667e-06,
|
| 1631 |
+
"loss": 0.0773,
|
| 1632 |
"step": 5475
|
| 1633 |
},
|
| 1634 |
{
|
| 1635 |
+
"epoch": 0.6875,
|
| 1636 |
+
"grad_norm": 3.437046527862549,
|
| 1637 |
+
"learning_rate": 2.0858333333333334e-06,
|
| 1638 |
+
"loss": 0.0914,
|
| 1639 |
"step": 5500
|
| 1640 |
},
|
| 1641 |
{
|
| 1642 |
+
"epoch": 0.6875,
|
| 1643 |
+
"eval_loss": 0.23427686095237732,
|
| 1644 |
+
"eval_runtime": 419.3101,
|
| 1645 |
+
"eval_samples_per_second": 5.018,
|
| 1646 |
+
"eval_steps_per_second": 0.627,
|
| 1647 |
+
"eval_wer": 10.502944750864728,
|
| 1648 |
"step": 5500
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
+
"epoch": 0.690625,
|
| 1652 |
+
"grad_norm": 4.649775981903076,
|
| 1653 |
+
"learning_rate": 2.065e-06,
|
| 1654 |
+
"loss": 0.1537,
|
| 1655 |
"step": 5525
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
+
"epoch": 0.69375,
|
| 1659 |
+
"grad_norm": 3.8799123764038086,
|
| 1660 |
+
"learning_rate": 2.044166666666667e-06,
|
| 1661 |
+
"loss": 0.1565,
|
| 1662 |
"step": 5550
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
+
"epoch": 0.696875,
|
| 1666 |
+
"grad_norm": 4.4588518142700195,
|
| 1667 |
+
"learning_rate": 2.023333333333333e-06,
|
| 1668 |
+
"loss": 0.1427,
|
| 1669 |
"step": 5575
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
+
"epoch": 0.7,
|
| 1673 |
+
"grad_norm": 3.449648141860962,
|
| 1674 |
+
"learning_rate": 2.0025000000000003e-06,
|
| 1675 |
+
"loss": 0.1312,
|
| 1676 |
"step": 5600
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
+
"epoch": 0.703125,
|
| 1680 |
+
"grad_norm": 4.717497825622559,
|
| 1681 |
+
"learning_rate": 1.9816666666666666e-06,
|
| 1682 |
+
"loss": 0.1093,
|
| 1683 |
"step": 5625
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
+
"epoch": 0.70625,
|
| 1687 |
+
"grad_norm": 4.465198040008545,
|
| 1688 |
+
"learning_rate": 1.9608333333333333e-06,
|
| 1689 |
+
"loss": 0.1237,
|
| 1690 |
"step": 5650
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
+
"epoch": 0.709375,
|
| 1694 |
+
"grad_norm": 4.121485233306885,
|
| 1695 |
+
"learning_rate": 1.94e-06,
|
| 1696 |
+
"loss": 0.1196,
|
| 1697 |
"step": 5675
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
+
"epoch": 0.7125,
|
| 1701 |
+
"grad_norm": 3.461294651031494,
|
| 1702 |
+
"learning_rate": 1.9191666666666668e-06,
|
| 1703 |
+
"loss": 0.1204,
|
| 1704 |
"step": 5700
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
+
"epoch": 0.715625,
|
| 1708 |
+
"grad_norm": 5.659304618835449,
|
| 1709 |
+
"learning_rate": 1.8983333333333335e-06,
|
| 1710 |
+
"loss": 0.1229,
|
| 1711 |
"step": 5725
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
+
"epoch": 0.71875,
|
| 1715 |
+
"grad_norm": 4.469748497009277,
|
| 1716 |
+
"learning_rate": 1.8775e-06,
|
| 1717 |
+
"loss": 0.1253,
|
| 1718 |
"step": 5750
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
+
"epoch": 0.721875,
|
| 1722 |
+
"grad_norm": 4.66398811340332,
|
| 1723 |
+
"learning_rate": 1.8566666666666665e-06,
|
| 1724 |
+
"loss": 0.1146,
|
| 1725 |
"step": 5775
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
+
"epoch": 0.725,
|
| 1729 |
+
"grad_norm": 4.326991558074951,
|
| 1730 |
+
"learning_rate": 1.8358333333333335e-06,
|
| 1731 |
+
"loss": 0.1213,
|
| 1732 |
"step": 5800
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
+
"epoch": 0.728125,
|
| 1736 |
+
"grad_norm": 3.054807662963867,
|
| 1737 |
+
"learning_rate": 1.815e-06,
|
| 1738 |
+
"loss": 0.0985,
|
| 1739 |
"step": 5825
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
+
"epoch": 0.73125,
|
| 1743 |
+
"grad_norm": 3.3157272338867188,
|
| 1744 |
+
"learning_rate": 1.794166666666667e-06,
|
| 1745 |
+
"loss": 0.0958,
|
| 1746 |
"step": 5850
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
+
"epoch": 0.734375,
|
| 1750 |
+
"grad_norm": 2.41603422164917,
|
| 1751 |
+
"learning_rate": 1.7733333333333334e-06,
|
| 1752 |
+
"loss": 0.069,
|
| 1753 |
"step": 5875
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
+
"epoch": 0.7375,
|
| 1757 |
+
"grad_norm": 3.2789347171783447,
|
| 1758 |
+
"learning_rate": 1.7525e-06,
|
| 1759 |
+
"loss": 0.0713,
|
| 1760 |
"step": 5900
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
+
"epoch": 0.740625,
|
| 1764 |
+
"grad_norm": 3.526294231414795,
|
| 1765 |
+
"learning_rate": 1.7316666666666669e-06,
|
| 1766 |
+
"loss": 0.0713,
|
| 1767 |
"step": 5925
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
+
"epoch": 0.74375,
|
| 1771 |
+
"grad_norm": 2.1724507808685303,
|
| 1772 |
+
"learning_rate": 1.7108333333333334e-06,
|
| 1773 |
+
"loss": 0.0793,
|
| 1774 |
"step": 5950
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
+
"epoch": 0.746875,
|
| 1778 |
+
"grad_norm": 3.1178219318389893,
|
| 1779 |
+
"learning_rate": 1.69e-06,
|
| 1780 |
+
"loss": 0.1,
|
| 1781 |
"step": 5975
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
+
"epoch": 0.75,
|
| 1785 |
+
"grad_norm": 4.3492889404296875,
|
| 1786 |
+
"learning_rate": 1.6691666666666668e-06,
|
| 1787 |
+
"loss": 0.1093,
|
| 1788 |
"step": 6000
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
+
"epoch": 0.75,
|
| 1792 |
+
"eval_loss": 0.21914222836494446,
|
| 1793 |
+
"eval_runtime": 421.5164,
|
| 1794 |
+
"eval_samples_per_second": 4.992,
|
| 1795 |
+
"eval_steps_per_second": 0.624,
|
| 1796 |
+
"eval_wer": 9.98410769374591,
|
| 1797 |
"step": 6000
|
| 1798 |
},
|
| 1799 |
{
|
| 1800 |
+
"epoch": 0.753125,
|
| 1801 |
+
"grad_norm": 4.253927707672119,
|
| 1802 |
"learning_rate": 1.6483333333333334e-06,
|
| 1803 |
+
"loss": 0.1182,
|
| 1804 |
"step": 6025
|
| 1805 |
},
|
| 1806 |
{
|
| 1807 |
+
"epoch": 0.75625,
|
| 1808 |
+
"grad_norm": 5.901562213897705,
|
| 1809 |
"learning_rate": 1.6275000000000003e-06,
|
| 1810 |
+
"loss": 0.1278,
|
| 1811 |
"step": 6050
|
| 1812 |
},
|
| 1813 |
{
|
| 1814 |
+
"epoch": 0.759375,
|
| 1815 |
+
"grad_norm": 5.934751510620117,
|
| 1816 |
"learning_rate": 1.6066666666666668e-06,
|
| 1817 |
+
"loss": 0.1411,
|
| 1818 |
"step": 6075
|
| 1819 |
},
|
| 1820 |
{
|
| 1821 |
+
"epoch": 0.7625,
|
| 1822 |
+
"grad_norm": 5.152722358703613,
|
| 1823 |
"learning_rate": 1.5858333333333333e-06,
|
| 1824 |
+
"loss": 0.1496,
|
| 1825 |
"step": 6100
|
| 1826 |
},
|
| 1827 |
{
|
| 1828 |
+
"epoch": 0.765625,
|
| 1829 |
+
"grad_norm": 4.460409641265869,
|
| 1830 |
"learning_rate": 1.565e-06,
|
| 1831 |
+
"loss": 0.1132,
|
| 1832 |
"step": 6125
|
| 1833 |
},
|
| 1834 |
{
|
| 1835 |
+
"epoch": 0.76875,
|
| 1836 |
+
"grad_norm": 2.9203169345855713,
|
| 1837 |
"learning_rate": 1.5441666666666668e-06,
|
| 1838 |
+
"loss": 0.0827,
|
| 1839 |
"step": 6150
|
| 1840 |
},
|
| 1841 |
{
|
| 1842 |
+
"epoch": 0.771875,
|
| 1843 |
+
"grad_norm": 2.982921600341797,
|
| 1844 |
"learning_rate": 1.5233333333333333e-06,
|
| 1845 |
+
"loss": 0.0708,
|
| 1846 |
"step": 6175
|
| 1847 |
},
|
| 1848 |
{
|
| 1849 |
+
"epoch": 0.775,
|
| 1850 |
+
"grad_norm": 3.2219862937927246,
|
| 1851 |
"learning_rate": 1.5025e-06,
|
| 1852 |
+
"loss": 0.0705,
|
| 1853 |
"step": 6200
|
| 1854 |
},
|
| 1855 |
{
|
| 1856 |
+
"epoch": 0.778125,
|
| 1857 |
+
"grad_norm": 2.3243980407714844,
|
| 1858 |
"learning_rate": 1.4816666666666667e-06,
|
| 1859 |
+
"loss": 0.0666,
|
| 1860 |
"step": 6225
|
| 1861 |
},
|
| 1862 |
{
|
| 1863 |
+
"epoch": 0.78125,
|
| 1864 |
+
"grad_norm": 2.5418896675109863,
|
| 1865 |
"learning_rate": 1.4608333333333333e-06,
|
| 1866 |
+
"loss": 0.0692,
|
| 1867 |
"step": 6250
|
| 1868 |
},
|
| 1869 |
{
|
| 1870 |
+
"epoch": 0.784375,
|
| 1871 |
+
"grad_norm": 3.7306313514709473,
|
| 1872 |
"learning_rate": 1.44e-06,
|
| 1873 |
+
"loss": 0.0772,
|
| 1874 |
"step": 6275
|
| 1875 |
},
|
| 1876 |
{
|
| 1877 |
+
"epoch": 0.7875,
|
| 1878 |
+
"grad_norm": 3.970508098602295,
|
| 1879 |
"learning_rate": 1.4191666666666667e-06,
|
| 1880 |
+
"loss": 0.0843,
|
| 1881 |
"step": 6300
|
| 1882 |
},
|
| 1883 |
{
|
| 1884 |
+
"epoch": 0.790625,
|
| 1885 |
+
"grad_norm": 3.8446128368377686,
|
| 1886 |
"learning_rate": 1.3983333333333334e-06,
|
| 1887 |
+
"loss": 0.1233,
|
| 1888 |
"step": 6325
|
| 1889 |
},
|
| 1890 |
{
|
| 1891 |
+
"epoch": 0.79375,
|
| 1892 |
+
"grad_norm": 4.295629501342773,
|
| 1893 |
"learning_rate": 1.3775000000000002e-06,
|
| 1894 |
+
"loss": 0.139,
|
| 1895 |
"step": 6350
|
| 1896 |
},
|
| 1897 |
{
|
| 1898 |
+
"epoch": 0.796875,
|
| 1899 |
+
"grad_norm": 4.724053382873535,
|
| 1900 |
"learning_rate": 1.3566666666666667e-06,
|
| 1901 |
+
"loss": 0.1107,
|
| 1902 |
"step": 6375
|
| 1903 |
},
|
| 1904 |
{
|
| 1905 |
+
"epoch": 0.8,
|
| 1906 |
+
"grad_norm": 5.828359127044678,
|
| 1907 |
"learning_rate": 1.3358333333333334e-06,
|
| 1908 |
+
"loss": 0.1304,
|
| 1909 |
"step": 6400
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
+
"epoch": 0.803125,
|
| 1913 |
+
"grad_norm": 4.042730331420898,
|
| 1914 |
"learning_rate": 1.3150000000000001e-06,
|
| 1915 |
+
"loss": 0.1277,
|
| 1916 |
"step": 6425
|
| 1917 |
},
|
| 1918 |
{
|
| 1919 |
+
"epoch": 0.80625,
|
| 1920 |
+
"grad_norm": 3.5273425579071045,
|
| 1921 |
"learning_rate": 1.2941666666666668e-06,
|
| 1922 |
+
"loss": 0.126,
|
| 1923 |
"step": 6450
|
| 1924 |
},
|
| 1925 |
{
|
| 1926 |
+
"epoch": 0.809375,
|
| 1927 |
+
"grad_norm": 2.901884078979492,
|
| 1928 |
"learning_rate": 1.2733333333333334e-06,
|
| 1929 |
+
"loss": 0.0929,
|
| 1930 |
"step": 6475
|
| 1931 |
},
|
| 1932 |
{
|
| 1933 |
+
"epoch": 0.8125,
|
| 1934 |
+
"grad_norm": 4.719346523284912,
|
| 1935 |
"learning_rate": 1.2525e-06,
|
| 1936 |
+
"loss": 0.0948,
|
| 1937 |
"step": 6500
|
| 1938 |
},
|
| 1939 |
{
|
| 1940 |
+
"epoch": 0.8125,
|
| 1941 |
+
"eval_loss": 0.2214665412902832,
|
| 1942 |
+
"eval_runtime": 423.0169,
|
| 1943 |
+
"eval_samples_per_second": 4.974,
|
| 1944 |
+
"eval_steps_per_second": 0.622,
|
| 1945 |
+
"eval_wer": 10.535664204917266,
|
| 1946 |
"step": 6500
|
| 1947 |
},
|
| 1948 |
{
|
| 1949 |
+
"epoch": 0.815625,
|
| 1950 |
+
"grad_norm": 4.184325218200684,
|
| 1951 |
"learning_rate": 1.2316666666666668e-06,
|
| 1952 |
+
"loss": 0.1045,
|
| 1953 |
"step": 6525
|
| 1954 |
},
|
| 1955 |
{
|
| 1956 |
+
"epoch": 0.81875,
|
| 1957 |
+
"grad_norm": 4.737115383148193,
|
| 1958 |
"learning_rate": 1.2108333333333335e-06,
|
| 1959 |
+
"loss": 0.112,
|
| 1960 |
"step": 6550
|
| 1961 |
},
|
| 1962 |
{
|
| 1963 |
+
"epoch": 0.821875,
|
| 1964 |
+
"grad_norm": 7.072691440582275,
|
| 1965 |
"learning_rate": 1.19e-06,
|
| 1966 |
+
"loss": 0.1326,
|
| 1967 |
"step": 6575
|
| 1968 |
},
|
| 1969 |
{
|
| 1970 |
+
"epoch": 0.825,
|
| 1971 |
+
"grad_norm": 2.486250162124634,
|
| 1972 |
"learning_rate": 1.1691666666666668e-06,
|
| 1973 |
+
"loss": 0.1154,
|
| 1974 |
"step": 6600
|
| 1975 |
},
|
| 1976 |
{
|
| 1977 |
+
"epoch": 0.828125,
|
| 1978 |
+
"grad_norm": 4.1657328605651855,
|
| 1979 |
+
"learning_rate": 1.1491666666666667e-06,
|
| 1980 |
+
"loss": 0.1102,
|
| 1981 |
"step": 6625
|
| 1982 |
},
|
| 1983 |
{
|
| 1984 |
+
"epoch": 0.83125,
|
| 1985 |
+
"grad_norm": 3.565835475921631,
|
| 1986 |
"learning_rate": 1.1283333333333333e-06,
|
| 1987 |
+
"loss": 0.0793,
|
| 1988 |
"step": 6650
|
| 1989 |
},
|
| 1990 |
{
|
| 1991 |
+
"epoch": 0.834375,
|
| 1992 |
+
"grad_norm": 2.6341757774353027,
|
| 1993 |
"learning_rate": 1.1075e-06,
|
| 1994 |
+
"loss": 0.0698,
|
| 1995 |
"step": 6675
|
| 1996 |
},
|
| 1997 |
{
|
| 1998 |
+
"epoch": 0.8375,
|
| 1999 |
+
"grad_norm": 3.431013584136963,
|
| 2000 |
"learning_rate": 1.0866666666666667e-06,
|
| 2001 |
+
"loss": 0.0757,
|
| 2002 |
"step": 6700
|
| 2003 |
},
|
| 2004 |
{
|
| 2005 |
+
"epoch": 0.840625,
|
| 2006 |
+
"grad_norm": 2.743353843688965,
|
| 2007 |
"learning_rate": 1.0658333333333334e-06,
|
| 2008 |
+
"loss": 0.0606,
|
| 2009 |
"step": 6725
|
| 2010 |
},
|
| 2011 |
{
|
| 2012 |
+
"epoch": 0.84375,
|
| 2013 |
+
"grad_norm": 2.702622890472412,
|
| 2014 |
"learning_rate": 1.045e-06,
|
| 2015 |
+
"loss": 0.0598,
|
| 2016 |
"step": 6750
|
| 2017 |
},
|
| 2018 |
{
|
| 2019 |
+
"epoch": 0.846875,
|
| 2020 |
+
"grad_norm": 2.669710636138916,
|
| 2021 |
"learning_rate": 1.0241666666666667e-06,
|
| 2022 |
+
"loss": 0.0641,
|
| 2023 |
"step": 6775
|
| 2024 |
},
|
| 2025 |
{
|
| 2026 |
+
"epoch": 0.85,
|
| 2027 |
+
"grad_norm": 3.356062173843384,
|
| 2028 |
"learning_rate": 1.0033333333333334e-06,
|
| 2029 |
+
"loss": 0.0717,
|
| 2030 |
"step": 6800
|
| 2031 |
},
|
| 2032 |
{
|
| 2033 |
+
"epoch": 0.853125,
|
| 2034 |
+
"grad_norm": 3.792300224304199,
|
| 2035 |
"learning_rate": 9.825000000000001e-07,
|
| 2036 |
+
"loss": 0.0755,
|
| 2037 |
"step": 6825
|
| 2038 |
},
|
| 2039 |
{
|
| 2040 |
+
"epoch": 0.85625,
|
| 2041 |
+
"grad_norm": 2.8524975776672363,
|
| 2042 |
"learning_rate": 9.616666666666668e-07,
|
| 2043 |
+
"loss": 0.074,
|
| 2044 |
"step": 6850
|
| 2045 |
},
|
| 2046 |
{
|
| 2047 |
+
"epoch": 0.859375,
|
| 2048 |
+
"grad_norm": 2.653484344482422,
|
| 2049 |
"learning_rate": 9.408333333333334e-07,
|
| 2050 |
+
"loss": 0.0732,
|
| 2051 |
"step": 6875
|
| 2052 |
},
|
| 2053 |
{
|
| 2054 |
+
"epoch": 0.8625,
|
| 2055 |
+
"grad_norm": 2.2992992401123047,
|
| 2056 |
"learning_rate": 9.2e-07,
|
| 2057 |
+
"loss": 0.0653,
|
| 2058 |
"step": 6900
|
| 2059 |
},
|
| 2060 |
{
|
| 2061 |
+
"epoch": 0.865625,
|
| 2062 |
+
"grad_norm": 2.3256728649139404,
|
| 2063 |
"learning_rate": 8.991666666666667e-07,
|
| 2064 |
+
"loss": 0.0612,
|
| 2065 |
"step": 6925
|
| 2066 |
},
|
| 2067 |
{
|
| 2068 |
+
"epoch": 0.86875,
|
| 2069 |
+
"grad_norm": 3.0332698822021484,
|
| 2070 |
"learning_rate": 8.783333333333334e-07,
|
| 2071 |
+
"loss": 0.0533,
|
| 2072 |
"step": 6950
|
| 2073 |
},
|
| 2074 |
{
|
| 2075 |
+
"epoch": 0.871875,
|
| 2076 |
+
"grad_norm": 4.096222877502441,
|
| 2077 |
"learning_rate": 8.574999999999999e-07,
|
| 2078 |
+
"loss": 0.0704,
|
| 2079 |
"step": 6975
|
| 2080 |
},
|
| 2081 |
{
|
| 2082 |
+
"epoch": 0.875,
|
| 2083 |
+
"grad_norm": 3.0144271850585938,
|
| 2084 |
"learning_rate": 8.366666666666667e-07,
|
| 2085 |
+
"loss": 0.0744,
|
| 2086 |
"step": 7000
|
| 2087 |
},
|
| 2088 |
{
|
| 2089 |
+
"epoch": 0.875,
|
| 2090 |
+
"eval_loss": 0.21083347499370575,
|
| 2091 |
+
"eval_runtime": 429.259,
|
| 2092 |
+
"eval_samples_per_second": 4.901,
|
| 2093 |
+
"eval_steps_per_second": 0.613,
|
| 2094 |
+
"eval_wer": 11.23679536318594,
|
| 2095 |
"step": 7000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
+
"epoch": 0.878125,
|
| 2099 |
+
"grad_norm": 2.6104869842529297,
|
| 2100 |
"learning_rate": 8.158333333333334e-07,
|
| 2101 |
+
"loss": 0.0669,
|
| 2102 |
"step": 7025
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
+
"epoch": 0.88125,
|
| 2106 |
+
"grad_norm": 2.6073827743530273,
|
| 2107 |
"learning_rate": 7.950000000000001e-07,
|
| 2108 |
+
"loss": 0.0651,
|
| 2109 |
"step": 7050
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
+
"epoch": 0.884375,
|
| 2113 |
+
"grad_norm": 2.584033966064453,
|
| 2114 |
"learning_rate": 7.741666666666667e-07,
|
| 2115 |
+
"loss": 0.0622,
|
| 2116 |
"step": 7075
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
+
"epoch": 0.8875,
|
| 2120 |
+
"grad_norm": 2.425426959991455,
|
| 2121 |
"learning_rate": 7.533333333333334e-07,
|
| 2122 |
+
"loss": 0.0735,
|
| 2123 |
"step": 7100
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
+
"epoch": 0.890625,
|
| 2127 |
+
"grad_norm": 3.056304931640625,
|
| 2128 |
"learning_rate": 7.325e-07,
|
| 2129 |
+
"loss": 0.0982,
|
| 2130 |
"step": 7125
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
+
"epoch": 0.89375,
|
| 2134 |
+
"grad_norm": 3.8735291957855225,
|
| 2135 |
"learning_rate": 7.116666666666667e-07,
|
| 2136 |
+
"loss": 0.1166,
|
| 2137 |
"step": 7150
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
+
"epoch": 0.896875,
|
| 2141 |
+
"grad_norm": 4.358260631561279,
|
| 2142 |
"learning_rate": 6.908333333333333e-07,
|
| 2143 |
+
"loss": 0.1212,
|
| 2144 |
"step": 7175
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
+
"epoch": 0.9,
|
| 2148 |
+
"grad_norm": 3.1888036727905273,
|
| 2149 |
"learning_rate": 6.7e-07,
|
| 2150 |
+
"loss": 0.1209,
|
| 2151 |
"step": 7200
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
+
"epoch": 0.903125,
|
| 2155 |
+
"grad_norm": 4.592689514160156,
|
| 2156 |
"learning_rate": 6.491666666666667e-07,
|
| 2157 |
+
"loss": 0.1265,
|
| 2158 |
"step": 7225
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
+
"epoch": 0.90625,
|
| 2162 |
+
"grad_norm": 4.064499378204346,
|
| 2163 |
"learning_rate": 6.283333333333334e-07,
|
| 2164 |
+
"loss": 0.1031,
|
| 2165 |
"step": 7250
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
+
"epoch": 0.909375,
|
| 2169 |
+
"grad_norm": 4.060811996459961,
|
| 2170 |
"learning_rate": 6.075e-07,
|
| 2171 |
+
"loss": 0.0694,
|
| 2172 |
"step": 7275
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
+
"epoch": 0.9125,
|
| 2176 |
+
"grad_norm": 2.3940441608428955,
|
| 2177 |
"learning_rate": 5.866666666666667e-07,
|
| 2178 |
+
"loss": 0.0717,
|
| 2179 |
"step": 7300
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
+
"epoch": 0.915625,
|
| 2183 |
+
"grad_norm": 3.34171199798584,
|
| 2184 |
"learning_rate": 5.658333333333334e-07,
|
| 2185 |
+
"loss": 0.0693,
|
| 2186 |
"step": 7325
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
+
"epoch": 0.91875,
|
| 2190 |
+
"grad_norm": 3.1847307682037354,
|
| 2191 |
"learning_rate": 5.450000000000001e-07,
|
| 2192 |
+
"loss": 0.0549,
|
| 2193 |
"step": 7350
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
+
"epoch": 0.921875,
|
| 2197 |
+
"grad_norm": 2.4034082889556885,
|
| 2198 |
"learning_rate": 5.241666666666667e-07,
|
| 2199 |
+
"loss": 0.0679,
|
| 2200 |
"step": 7375
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
+
"epoch": 0.925,
|
| 2204 |
+
"grad_norm": 2.2434728145599365,
|
| 2205 |
"learning_rate": 5.033333333333333e-07,
|
| 2206 |
+
"loss": 0.0677,
|
| 2207 |
"step": 7400
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
+
"epoch": 0.928125,
|
| 2211 |
+
"grad_norm": 2.655233144760132,
|
| 2212 |
"learning_rate": 4.825e-07,
|
| 2213 |
+
"loss": 0.0844,
|
| 2214 |
"step": 7425
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
+
"epoch": 0.93125,
|
| 2218 |
+
"grad_norm": 3.4723622798919678,
|
| 2219 |
"learning_rate": 4.6166666666666666e-07,
|
| 2220 |
+
"loss": 0.0766,
|
| 2221 |
"step": 7450
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
+
"epoch": 0.934375,
|
| 2225 |
+
"grad_norm": 3.544377326965332,
|
| 2226 |
"learning_rate": 4.408333333333334e-07,
|
| 2227 |
+
"loss": 0.0761,
|
| 2228 |
"step": 7475
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
+
"epoch": 0.9375,
|
| 2232 |
+
"grad_norm": 4.966926097869873,
|
| 2233 |
"learning_rate": 4.2e-07,
|
| 2234 |
+
"loss": 0.1269,
|
| 2235 |
"step": 7500
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
+
"epoch": 0.9375,
|
| 2239 |
+
"eval_loss": 0.21579307317733765,
|
| 2240 |
+
"eval_runtime": 453.0321,
|
| 2241 |
+
"eval_samples_per_second": 4.644,
|
| 2242 |
+
"eval_steps_per_second": 0.581,
|
| 2243 |
+
"eval_wer": 10.002804524633074,
|
| 2244 |
"step": 7500
|
| 2245 |
},
|
| 2246 |
{
|
| 2247 |
+
"epoch": 0.940625,
|
| 2248 |
+
"grad_norm": 4.923036098480225,
|
| 2249 |
"learning_rate": 3.991666666666667e-07,
|
| 2250 |
+
"loss": 0.1428,
|
| 2251 |
"step": 7525
|
| 2252 |
},
|
| 2253 |
{
|
| 2254 |
+
"epoch": 0.94375,
|
| 2255 |
+
"grad_norm": 3.906621217727661,
|
| 2256 |
"learning_rate": 3.7833333333333335e-07,
|
| 2257 |
+
"loss": 0.1148,
|
| 2258 |
"step": 7550
|
| 2259 |
},
|
| 2260 |
{
|
| 2261 |
+
"epoch": 0.946875,
|
| 2262 |
+
"grad_norm": 3.8426077365875244,
|
| 2263 |
"learning_rate": 3.575e-07,
|
| 2264 |
+
"loss": 0.1139,
|
| 2265 |
"step": 7575
|
| 2266 |
},
|
| 2267 |
{
|
| 2268 |
+
"epoch": 0.95,
|
| 2269 |
+
"grad_norm": 4.980268955230713,
|
| 2270 |
"learning_rate": 3.366666666666667e-07,
|
| 2271 |
+
"loss": 0.1389,
|
| 2272 |
"step": 7600
|
| 2273 |
},
|
| 2274 |
{
|
| 2275 |
+
"epoch": 0.953125,
|
| 2276 |
+
"grad_norm": 5.212338447570801,
|
| 2277 |
"learning_rate": 3.1583333333333336e-07,
|
| 2278 |
+
"loss": 0.1296,
|
| 2279 |
"step": 7625
|
| 2280 |
},
|
| 2281 |
{
|
| 2282 |
+
"epoch": 0.95625,
|
| 2283 |
+
"grad_norm": 4.394927024841309,
|
| 2284 |
"learning_rate": 2.95e-07,
|
| 2285 |
+
"loss": 0.1125,
|
| 2286 |
"step": 7650
|
| 2287 |
},
|
| 2288 |
{
|
| 2289 |
+
"epoch": 0.959375,
|
| 2290 |
+
"grad_norm": 5.008238792419434,
|
| 2291 |
"learning_rate": 2.7416666666666665e-07,
|
| 2292 |
+
"loss": 0.1323,
|
| 2293 |
"step": 7675
|
| 2294 |
},
|
| 2295 |
{
|
| 2296 |
+
"epoch": 0.9625,
|
| 2297 |
+
"grad_norm": 6.502245903015137,
|
| 2298 |
"learning_rate": 2.533333333333333e-07,
|
| 2299 |
+
"loss": 0.1401,
|
| 2300 |
"step": 7700
|
| 2301 |
},
|
| 2302 |
{
|
| 2303 |
+
"epoch": 0.965625,
|
| 2304 |
+
"grad_norm": 5.630526542663574,
|
| 2305 |
"learning_rate": 2.325e-07,
|
| 2306 |
+
"loss": 0.1358,
|
| 2307 |
"step": 7725
|
| 2308 |
},
|
| 2309 |
{
|
| 2310 |
+
"epoch": 0.96875,
|
| 2311 |
+
"grad_norm": 2.802499532699585,
|
| 2312 |
"learning_rate": 2.116666666666667e-07,
|
| 2313 |
+
"loss": 0.1068,
|
| 2314 |
"step": 7750
|
| 2315 |
},
|
| 2316 |
{
|
| 2317 |
+
"epoch": 0.971875,
|
| 2318 |
+
"grad_norm": 2.5446484088897705,
|
| 2319 |
"learning_rate": 1.9083333333333334e-07,
|
| 2320 |
+
"loss": 0.0862,
|
| 2321 |
"step": 7775
|
| 2322 |
},
|
| 2323 |
{
|
| 2324 |
+
"epoch": 0.975,
|
| 2325 |
+
"grad_norm": 5.739973068237305,
|
| 2326 |
"learning_rate": 1.7e-07,
|
| 2327 |
+
"loss": 0.0749,
|
| 2328 |
"step": 7800
|
| 2329 |
},
|
| 2330 |
{
|
| 2331 |
+
"epoch": 0.978125,
|
| 2332 |
+
"grad_norm": 2.2183213233947754,
|
| 2333 |
"learning_rate": 1.4916666666666669e-07,
|
| 2334 |
+
"loss": 0.0694,
|
| 2335 |
"step": 7825
|
| 2336 |
},
|
| 2337 |
{
|
| 2338 |
+
"epoch": 0.98125,
|
| 2339 |
+
"grad_norm": 2.4163272380828857,
|
| 2340 |
"learning_rate": 1.2833333333333333e-07,
|
| 2341 |
+
"loss": 0.0768,
|
| 2342 |
"step": 7850
|
| 2343 |
},
|
| 2344 |
{
|
| 2345 |
+
"epoch": 0.984375,
|
| 2346 |
+
"grad_norm": 3.4462175369262695,
|
| 2347 |
"learning_rate": 1.075e-07,
|
| 2348 |
+
"loss": 0.0635,
|
| 2349 |
"step": 7875
|
| 2350 |
},
|
| 2351 |
{
|
| 2352 |
+
"epoch": 0.9875,
|
| 2353 |
+
"grad_norm": 2.5306878089904785,
|
| 2354 |
"learning_rate": 8.666666666666666e-08,
|
| 2355 |
+
"loss": 0.0679,
|
| 2356 |
"step": 7900
|
| 2357 |
},
|
| 2358 |
{
|
| 2359 |
+
"epoch": 0.990625,
|
| 2360 |
+
"grad_norm": 4.3514485359191895,
|
| 2361 |
"learning_rate": 6.583333333333333e-08,
|
| 2362 |
+
"loss": 0.1073,
|
| 2363 |
"step": 7925
|
| 2364 |
},
|
| 2365 |
{
|
| 2366 |
+
"epoch": 0.99375,
|
| 2367 |
+
"grad_norm": 5.388062953948975,
|
| 2368 |
"learning_rate": 4.5e-08,
|
| 2369 |
+
"loss": 0.1159,
|
| 2370 |
"step": 7950
|
| 2371 |
},
|
| 2372 |
{
|
| 2373 |
+
"epoch": 0.996875,
|
| 2374 |
+
"grad_norm": 3.7506585121154785,
|
| 2375 |
"learning_rate": 2.4166666666666668e-08,
|
| 2376 |
+
"loss": 0.1783,
|
| 2377 |
"step": 7975
|
| 2378 |
},
|
| 2379 |
{
|
| 2380 |
+
"epoch": 1.0,
|
| 2381 |
+
"grad_norm": 4.964086532592773,
|
| 2382 |
"learning_rate": 3.333333333333334e-09,
|
| 2383 |
+
"loss": 0.1408,
|
| 2384 |
"step": 8000
|
| 2385 |
},
|
| 2386 |
{
|
| 2387 |
+
"epoch": 1.0,
|
| 2388 |
+
"eval_loss": 0.21413126587867737,
|
| 2389 |
+
"eval_runtime": 423.6508,
|
| 2390 |
+
"eval_samples_per_second": 4.966,
|
| 2391 |
+
"eval_steps_per_second": 0.621,
|
| 2392 |
+
"eval_wer": 10.129008133121436,
|
| 2393 |
"step": 8000
|
| 2394 |
},
|
| 2395 |
{
|
| 2396 |
+
"epoch": 1.0,
|
| 2397 |
"step": 8000,
|
| 2398 |
+
"total_flos": 1.3063742816256e+20,
|
| 2399 |
+
"train_loss": 0.1566779458373785,
|
| 2400 |
+
"train_runtime": 20473.0101,
|
| 2401 |
+
"train_samples_per_second": 6.252,
|
| 2402 |
+
"train_steps_per_second": 0.391
|
| 2403 |
}
|
| 2404 |
],
|
| 2405 |
"logging_steps": 25,
|
|
|
|
| 2419 |
"attributes": {}
|
| 2420 |
}
|
| 2421 |
},
|
| 2422 |
+
"total_flos": 1.3063742816256e+20,
|
| 2423 |
"train_batch_size": 16,
|
| 2424 |
"trial_name": null,
|
| 2425 |
"trial_params": null
|