End of training
Browse files- all_results.json +10 -10
- eval_results.json +6 -6
- nohup.out +22 -0
- runs/Jul23_08-44-43_tknadmin-System-Product-Name/events.out.tfevents.1690127180.tknadmin-System-Product-Name.1702135.2 +3 -0
- train_results.json +5 -5
- trainer_state.json +1095 -597
all_results.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 3.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer": 13.
|
| 8 |
-
"train_loss": 0.
|
| 9 |
-
"train_runtime":
|
| 10 |
-
"train_samples_per_second": 1.
|
| 11 |
-
"train_steps_per_second": 0.
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.06,
|
| 3 |
+
"eval_loss": 0.2200811207294464,
|
| 4 |
+
"eval_runtime": 1786.3982,
|
| 5 |
+
"eval_samples_per_second": 3.69,
|
| 6 |
+
"eval_steps_per_second": 0.461,
|
| 7 |
+
"eval_wer": 13.179958686054519,
|
| 8 |
+
"train_loss": 0.208335678424154,
|
| 9 |
+
"train_runtime": 23034.2525,
|
| 10 |
+
"train_samples_per_second": 1.216,
|
| 11 |
+
"train_steps_per_second": 0.304
|
| 12 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second": 3.
|
| 6 |
-
"eval_steps_per_second": 0.
|
| 7 |
-
"eval_wer": 13.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.06,
|
| 3 |
+
"eval_loss": 0.2200811207294464,
|
| 4 |
+
"eval_runtime": 1786.3982,
|
| 5 |
+
"eval_samples_per_second": 3.69,
|
| 6 |
+
"eval_steps_per_second": 0.461,
|
| 7 |
+
"eval_wer": 13.179958686054519
|
| 8 |
}
|
nohup.out
CHANGED
|
@@ -236516,3 +236516,25 @@ The progress bars may be unreliable.
|
|
| 236516 |
|
| 236517 |
|
| 236518 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 236516 |
|
| 236517 |
|
| 236518 |
|
| 236519 |
+
To https://huggingface.co/xezpeleta/whisper-medium-eu
|
| 236520 |
+
dbef9e0..dc6df58 main -> main
|
| 236521 |
+
|
| 236522 |
+
07/23/2023 17:16:31 - WARNING - huggingface_hub.repository - To https://huggingface.co/xezpeleta/whisper-medium-eu
|
| 236523 |
+
dbef9e0..dc6df58 main -> main
|
| 236524 |
+
|
| 236525 |
+
***** train metrics *****
|
| 236526 |
+
epoch = 2.06
|
| 236527 |
+
train_loss = 0.2083
|
| 236528 |
+
train_runtime = 6:23:54.25
|
| 236529 |
+
train_samples_per_second = 1.216
|
| 236530 |
+
train_steps_per_second = 0.304
|
| 236531 |
+
07/23/2023 17:16:34 - INFO - __main__ - *** Evaluate ***
|
| 236532 |
+
[INFO|trainer.py:2955] 2023-07-23 17:16:34,494 >> ***** Running Evaluation *****
|
| 236533 |
+
[INFO|trainer.py:2959] 2023-07-23 17:16:34,494 >> Num examples: Unknown
|
| 236534 |
+
[INFO|trainer.py:2960] 2023-07-23 17:16:34,494 >> Batch size = 8
|
| 236535 |
+
|
| 236536 |
+
[INFO|trainer_utils.py:689] 2023-07-23 17:16:37,002 >> The following columns in the evaluation set don't have a corresponding argument in `WhisperForConditionalGeneration.forward` and have been ignored: input_length. If input_length are not expected by `WhisperForConditionalGeneration.forward`, you can safely ignore this message.
|
| 236537 |
+
[INFO|trainer.py:2700] 2023-07-23 17:46:20,894 >> Saving model checkpoint to ./
|
| 236538 |
+
[INFO|configuration_utils.py:447] 2023-07-23 17:46:20,894 >> Configuration saved in ./config.json
|
| 236539 |
+
[INFO|modeling_utils.py:1693] 2023-07-23 17:46:25,011 >> Model weights saved in ./pytorch_model.bin
|
| 236540 |
+
[INFO|feature_extraction_utils.py:368] 2023-07-23 17:46:25,021 >> Feature extractor saved in ./preprocessor_config.json
|
runs/Jul23_08-44-43_tknadmin-System-Product-Name/events.out.tfevents.1690127180.tknadmin-System-Product-Name.1702135.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbc3dd6c8c812709323a7c5b349f90abe7c61c271352897d0a2c8dd63973992e
|
| 3 |
+
size 358
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second": 1.
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.06,
|
| 3 |
+
"train_loss": 0.208335678424154,
|
| 4 |
+
"train_runtime": 23034.2525,
|
| 5 |
+
"train_samples_per_second": 1.216,
|
| 6 |
+
"train_steps_per_second": 0.304
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 13.
|
| 3 |
-
"best_model_checkpoint": "./checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
-
"epoch": 0.
|
| 12 |
"learning_rate": 4.6000000000000004e-07,
|
| 13 |
"loss": 1.4182,
|
| 14 |
"step": 25
|
|
@@ -26,1245 +26,1743 @@
|
|
| 26 |
"step": 75
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
"learning_rate": 1.94e-06,
|
| 31 |
"loss": 0.7765,
|
| 32 |
"step": 100
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
"learning_rate": 2.4400000000000004e-06,
|
| 37 |
"loss": 0.7103,
|
| 38 |
"step": 125
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
"learning_rate": 2.9400000000000002e-06,
|
| 43 |
"loss": 0.6597,
|
| 44 |
"step": 150
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
"learning_rate": 3.44e-06,
|
| 49 |
"loss": 0.6657,
|
| 50 |
"step": 175
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
"learning_rate": 3.94e-06,
|
| 55 |
"loss": 0.5853,
|
| 56 |
"step": 200
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
"learning_rate": 4.440000000000001e-06,
|
| 61 |
"loss": 0.5273,
|
| 62 |
"step": 225
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0.
|
| 66 |
"learning_rate": 4.94e-06,
|
| 67 |
"loss": 0.5979,
|
| 68 |
"step": 250
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0.
|
| 72 |
"learning_rate": 5.4400000000000004e-06,
|
| 73 |
"loss": 0.5861,
|
| 74 |
"step": 275
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0.
|
| 78 |
"learning_rate": 5.94e-06,
|
| 79 |
"loss": 0.5085,
|
| 80 |
"step": 300
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
"learning_rate": 6.440000000000001e-06,
|
| 85 |
"loss": 0.4827,
|
| 86 |
"step": 325
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
"learning_rate": 6.9400000000000005e-06,
|
| 91 |
"loss": 0.4909,
|
| 92 |
"step": 350
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
"learning_rate": 7.440000000000001e-06,
|
| 97 |
"loss": 0.4651,
|
| 98 |
"step": 375
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch": 0.
|
| 102 |
"learning_rate": 7.94e-06,
|
| 103 |
"loss": 0.494,
|
| 104 |
"step": 400
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch": 0.
|
| 108 |
"learning_rate": 8.44e-06,
|
| 109 |
"loss": 0.4188,
|
| 110 |
"step": 425
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch": 0.
|
| 114 |
"learning_rate": 8.94e-06,
|
| 115 |
"loss": 0.3849,
|
| 116 |
"step": 450
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 0.
|
| 120 |
"learning_rate": 9.440000000000001e-06,
|
| 121 |
"loss": 0.4577,
|
| 122 |
"step": 475
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
"learning_rate": 9.940000000000001e-06,
|
| 127 |
"loss": 0.4415,
|
| 128 |
"step": 500
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"learning_rate": 9.
|
| 133 |
-
"loss": 0.
|
| 134 |
"step": 525
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch": 0.
|
| 138 |
-
"learning_rate": 9.
|
| 139 |
-
"loss": 0.
|
| 140 |
"step": 550
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch": 0.
|
| 144 |
-
"learning_rate": 9.
|
| 145 |
-
"loss": 0.
|
| 146 |
"step": 575
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch": 0.
|
| 150 |
-
"learning_rate": 9.
|
| 151 |
-
"loss": 0.
|
| 152 |
"step": 600
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
-
"learning_rate": 9.
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 625
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 0.
|
| 162 |
-
"learning_rate": 9.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 650
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"learning_rate": 9.
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 675
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"learning_rate": 9.
|
| 175 |
-
"loss": 0.
|
| 176 |
"step": 700
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch": 0.
|
| 180 |
-
"learning_rate": 9.
|
| 181 |
-
"loss": 0.
|
| 182 |
"step": 725
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch": 0.
|
| 186 |
-
"learning_rate": 9.
|
| 187 |
-
"loss": 0.
|
| 188 |
"step": 750
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch": 0.
|
| 192 |
-
"learning_rate": 9.
|
| 193 |
-
"loss": 0.
|
| 194 |
"step": 775
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch": 0.
|
| 198 |
-
"learning_rate": 9.
|
| 199 |
-
"loss": 0.
|
| 200 |
"step": 800
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch": 0.
|
| 204 |
-
"learning_rate": 9.
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 825
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"learning_rate": 9.
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 850
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"learning_rate": 9.
|
| 217 |
-
"loss": 0.
|
| 218 |
"step": 875
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
-
"learning_rate": 9.
|
| 223 |
-
"loss": 0.
|
| 224 |
"step": 900
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch": 0.
|
| 228 |
-
"learning_rate": 9.
|
| 229 |
-
"loss": 0.
|
| 230 |
"step": 925
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 0.
|
| 234 |
-
"learning_rate": 9.
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 950
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch": 0.
|
| 240 |
-
"learning_rate":
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 975
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 1000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"eval_loss": 0.
|
| 253 |
-
"eval_runtime":
|
| 254 |
-
"eval_samples_per_second": 3.
|
| 255 |
-
"eval_steps_per_second": 0.
|
| 256 |
-
"eval_wer": 28.
|
| 257 |
"step": 1000
|
| 258 |
},
|
| 259 |
{
|
| 260 |
-
"epoch": 0.
|
| 261 |
-
"learning_rate":
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 1025
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch": 0.
|
| 267 |
-
"learning_rate":
|
| 268 |
"loss": 0.3674,
|
| 269 |
"step": 1050
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 0.
|
| 273 |
-
"learning_rate":
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 1075
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"learning_rate":
|
| 280 |
-
"loss": 0.
|
| 281 |
"step": 1100
|
| 282 |
},
|
| 283 |
{
|
| 284 |
-
"epoch": 0.
|
| 285 |
-
"learning_rate":
|
| 286 |
-
"loss": 0.
|
| 287 |
"step": 1125
|
| 288 |
},
|
| 289 |
{
|
| 290 |
-
"epoch": 0.
|
| 291 |
-
"learning_rate":
|
| 292 |
-
"loss": 0.
|
| 293 |
"step": 1150
|
| 294 |
},
|
| 295 |
{
|
| 296 |
-
"epoch": 0.
|
| 297 |
-
"learning_rate": 8.
|
| 298 |
-
"loss": 0.
|
| 299 |
"step": 1175
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch": 0.
|
| 303 |
-
"learning_rate": 8.
|
| 304 |
-
"loss": 0.
|
| 305 |
"step": 1200
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch": 0.
|
| 309 |
-
"learning_rate": 8.
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 1225
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch": 0.
|
| 315 |
-
"learning_rate": 8.
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 1250
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"learning_rate": 8.
|
| 322 |
-
"loss": 0.
|
| 323 |
"step": 1275
|
| 324 |
},
|
| 325 |
{
|
| 326 |
-
"epoch": 0.
|
| 327 |
-
"learning_rate": 8.
|
| 328 |
-
"loss": 0.
|
| 329 |
"step": 1300
|
| 330 |
},
|
| 331 |
{
|
| 332 |
-
"epoch": 0.
|
| 333 |
-
"learning_rate": 8.
|
| 334 |
-
"loss": 0.
|
| 335 |
"step": 1325
|
| 336 |
},
|
| 337 |
{
|
| 338 |
-
"epoch": 0.
|
| 339 |
-
"learning_rate": 8.
|
| 340 |
-
"loss": 0.
|
| 341 |
"step": 1350
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch": 0.
|
| 345 |
-
"learning_rate": 8.
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 1375
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch": 0.
|
| 351 |
-
"learning_rate": 8.
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 1400
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"learning_rate":
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 1425
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"learning_rate":
|
| 364 |
-
"loss": 0.
|
| 365 |
"step": 1450
|
| 366 |
},
|
| 367 |
{
|
| 368 |
-
"epoch": 0.
|
| 369 |
-
"learning_rate":
|
| 370 |
-
"loss": 0.
|
| 371 |
"step": 1475
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch": 0.
|
| 375 |
-
"learning_rate":
|
| 376 |
-
"loss": 0.
|
| 377 |
"step": 1500
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch": 0.
|
| 381 |
-
"learning_rate":
|
| 382 |
-
"loss": 0.
|
| 383 |
"step": 1525
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch": 0.
|
| 387 |
-
"learning_rate":
|
| 388 |
-
"loss": 0.
|
| 389 |
"step": 1550
|
| 390 |
},
|
| 391 |
{
|
| 392 |
-
"epoch": 0.
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 1575
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch": 0.
|
| 399 |
-
"learning_rate":
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 1600
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"learning_rate":
|
| 406 |
-
"loss": 0.
|
| 407 |
"step": 1625
|
| 408 |
},
|
| 409 |
{
|
| 410 |
-
"epoch": 0.
|
| 411 |
-
"learning_rate":
|
| 412 |
-
"loss": 0.
|
| 413 |
"step": 1650
|
| 414 |
},
|
| 415 |
{
|
| 416 |
-
"epoch": 0.
|
| 417 |
-
"learning_rate":
|
| 418 |
-
"loss": 0.
|
| 419 |
"step": 1675
|
| 420 |
},
|
| 421 |
{
|
| 422 |
-
"epoch": 0.
|
| 423 |
-
"learning_rate":
|
| 424 |
-
"loss": 0.
|
| 425 |
"step": 1700
|
| 426 |
},
|
| 427 |
{
|
| 428 |
-
"epoch": 0.
|
| 429 |
-
"learning_rate":
|
| 430 |
-
"loss": 0.
|
| 431 |
"step": 1725
|
| 432 |
},
|
| 433 |
{
|
| 434 |
-
"epoch": 0.
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 1750
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch": 0.
|
| 441 |
-
"learning_rate":
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 1775
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch": 0.
|
| 447 |
-
"learning_rate":
|
| 448 |
-
"loss": 0.
|
| 449 |
"step": 1800
|
| 450 |
},
|
| 451 |
{
|
| 452 |
-
"epoch": 0.
|
| 453 |
-
"learning_rate": 7.
|
| 454 |
-
"loss": 0.
|
| 455 |
"step": 1825
|
| 456 |
},
|
| 457 |
{
|
| 458 |
-
"epoch": 0.
|
| 459 |
-
"learning_rate": 7.
|
| 460 |
-
"loss": 0.
|
| 461 |
"step": 1850
|
| 462 |
},
|
| 463 |
{
|
| 464 |
-
"epoch": 0.
|
| 465 |
-
"learning_rate":
|
| 466 |
-
"loss": 0.
|
| 467 |
"step": 1875
|
| 468 |
},
|
| 469 |
{
|
| 470 |
-
"epoch": 0.
|
| 471 |
-
"learning_rate":
|
| 472 |
-
"loss": 0.
|
| 473 |
"step": 1900
|
| 474 |
},
|
| 475 |
{
|
| 476 |
-
"epoch": 0.
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 1925
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch": 0.
|
| 483 |
-
"learning_rate":
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 1950
|
| 486 |
},
|
| 487 |
{
|
| 488 |
-
"epoch": 0.
|
| 489 |
-
"learning_rate":
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 1975
|
| 492 |
},
|
| 493 |
{
|
| 494 |
-
"epoch": 0.
|
| 495 |
-
"learning_rate":
|
| 496 |
-
"loss": 0.
|
| 497 |
"step": 2000
|
| 498 |
},
|
| 499 |
{
|
| 500 |
-
"epoch": 0.
|
| 501 |
-
"eval_loss": 0.
|
| 502 |
-
"eval_runtime":
|
| 503 |
-
"eval_samples_per_second": 3.
|
| 504 |
-
"eval_steps_per_second": 0.
|
| 505 |
-
"eval_wer":
|
| 506 |
"step": 2000
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch": 0.
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 2025
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch": 0.
|
| 516 |
-
"learning_rate":
|
| 517 |
-
"loss": 0.
|
| 518 |
"step": 2050
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch": 0.
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 0.
|
| 524 |
"step": 2075
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch": 0.
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 2100
|
| 531 |
},
|
| 532 |
{
|
| 533 |
-
"epoch": 0.
|
| 534 |
-
"learning_rate":
|
| 535 |
-
"loss": 0.
|
| 536 |
"step": 2125
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch": 0.
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 2150
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch": 0.
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 2175
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch": 0.
|
| 552 |
-
"learning_rate":
|
| 553 |
-
"loss": 0.
|
| 554 |
"step": 2200
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch": 0.
|
| 558 |
-
"learning_rate":
|
| 559 |
-
"loss": 0.
|
| 560 |
"step": 2225
|
| 561 |
},
|
| 562 |
{
|
| 563 |
-
"epoch": 0.
|
| 564 |
-
"learning_rate":
|
| 565 |
-
"loss": 0.
|
| 566 |
"step": 2250
|
| 567 |
},
|
| 568 |
{
|
| 569 |
-
"epoch": 0.
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss": 0.
|
| 572 |
"step": 2275
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch": 0.
|
| 576 |
-
"learning_rate":
|
| 577 |
-
"loss": 0.
|
| 578 |
"step": 2300
|
| 579 |
},
|
| 580 |
{
|
| 581 |
-
"epoch": 0.
|
| 582 |
-
"learning_rate":
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 2325
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch": 0.
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 2350
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch": 0.
|
| 594 |
-
"learning_rate":
|
| 595 |
-
"loss": 0.
|
| 596 |
"step": 2375
|
| 597 |
},
|
| 598 |
{
|
| 599 |
-
"epoch": 0.
|
| 600 |
-
"learning_rate":
|
| 601 |
-
"loss": 0.
|
| 602 |
"step": 2400
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch": 0.
|
| 606 |
-
"learning_rate":
|
| 607 |
-
"loss": 0.
|
| 608 |
"step": 2425
|
| 609 |
},
|
| 610 |
{
|
| 611 |
-
"epoch": 0.
|
| 612 |
-
"learning_rate":
|
| 613 |
-
"loss": 0.
|
| 614 |
"step": 2450
|
| 615 |
},
|
| 616 |
{
|
| 617 |
-
"epoch": 0.
|
| 618 |
-
"learning_rate":
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 2475
|
| 621 |
},
|
| 622 |
{
|
| 623 |
-
"epoch": 0.
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 2500
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch": 0.
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 2525
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch": 0.
|
| 636 |
-
"learning_rate":
|
| 637 |
-
"loss": 0.
|
| 638 |
"step": 2550
|
| 639 |
},
|
| 640 |
{
|
| 641 |
-
"epoch": 0.
|
| 642 |
-
"learning_rate":
|
| 643 |
-
"loss": 0.
|
| 644 |
"step": 2575
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch": 0.
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss": 0.
|
| 650 |
"step": 2600
|
| 651 |
},
|
| 652 |
{
|
| 653 |
-
"epoch": 0.
|
| 654 |
-
"learning_rate":
|
| 655 |
-
"loss": 0.
|
| 656 |
"step": 2625
|
| 657 |
},
|
| 658 |
{
|
| 659 |
-
"epoch": 0.
|
| 660 |
-
"learning_rate":
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 2650
|
| 663 |
},
|
| 664 |
{
|
| 665 |
-
"epoch": 0.
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 2675
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch": 0.
|
| 672 |
-
"learning_rate":
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 2700
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch": 0.
|
| 678 |
-
"learning_rate":
|
| 679 |
-
"loss": 0.
|
| 680 |
"step": 2725
|
| 681 |
},
|
| 682 |
{
|
| 683 |
-
"epoch": 0.
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss": 0.
|
| 686 |
"step": 2750
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch": 0.
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 2775
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"epoch": 0.
|
| 696 |
-
"learning_rate":
|
| 697 |
-
"loss": 0.
|
| 698 |
"step": 2800
|
| 699 |
},
|
| 700 |
{
|
| 701 |
-
"epoch": 0.
|
| 702 |
-
"learning_rate":
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 2825
|
| 705 |
},
|
| 706 |
{
|
| 707 |
-
"epoch": 0.
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 2850
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch": 0.
|
| 714 |
-
"learning_rate":
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 2875
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch": 0.
|
| 720 |
-
"learning_rate":
|
| 721 |
-
"loss": 0.
|
| 722 |
"step": 2900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
-
"epoch": 0.
|
| 726 |
-
"learning_rate":
|
| 727 |
-
"loss": 0.
|
| 728 |
"step": 2925
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch": 0.
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 2950
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"epoch": 0.
|
| 738 |
-
"learning_rate":
|
| 739 |
-
"loss": 0.
|
| 740 |
"step": 2975
|
| 741 |
},
|
| 742 |
{
|
| 743 |
-
"epoch": 0.
|
| 744 |
-
"learning_rate":
|
| 745 |
-
"loss": 0.
|
| 746 |
"step": 3000
|
| 747 |
},
|
| 748 |
{
|
| 749 |
-
"epoch": 0.
|
| 750 |
-
"eval_loss": 0.
|
| 751 |
-
"eval_runtime":
|
| 752 |
-
"eval_samples_per_second": 3.
|
| 753 |
-
"eval_steps_per_second": 0.
|
| 754 |
-
"eval_wer":
|
| 755 |
"step": 3000
|
| 756 |
},
|
| 757 |
{
|
| 758 |
-
"epoch": 0.
|
| 759 |
-
"learning_rate":
|
| 760 |
-
"loss": 0.
|
| 761 |
"step": 3025
|
| 762 |
},
|
| 763 |
{
|
| 764 |
-
"epoch": 0.
|
| 765 |
-
"learning_rate":
|
| 766 |
-
"loss": 0.
|
| 767 |
"step": 3050
|
| 768 |
},
|
| 769 |
{
|
| 770 |
-
"epoch": 0.
|
| 771 |
-
"learning_rate":
|
| 772 |
-
"loss": 0.
|
| 773 |
"step": 3075
|
| 774 |
},
|
| 775 |
{
|
| 776 |
-
"epoch": 0.
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 3100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
-
"epoch": 0.
|
| 783 |
-
"learning_rate":
|
| 784 |
-
"loss": 0.
|
| 785 |
"step": 3125
|
| 786 |
},
|
| 787 |
{
|
| 788 |
-
"epoch": 0.
|
| 789 |
-
"learning_rate":
|
| 790 |
-
"loss": 0.
|
| 791 |
"step": 3150
|
| 792 |
},
|
| 793 |
{
|
| 794 |
-
"epoch": 0.
|
| 795 |
-
"learning_rate":
|
| 796 |
-
"loss": 0.
|
| 797 |
"step": 3175
|
| 798 |
},
|
| 799 |
{
|
| 800 |
-
"epoch": 0.
|
| 801 |
-
"learning_rate":
|
| 802 |
-
"loss": 0.
|
| 803 |
"step": 3200
|
| 804 |
},
|
| 805 |
{
|
| 806 |
-
"epoch": 0.
|
| 807 |
-
"learning_rate":
|
| 808 |
-
"loss": 0.
|
| 809 |
"step": 3225
|
| 810 |
},
|
| 811 |
{
|
| 812 |
-
"epoch": 0.
|
| 813 |
-
"learning_rate":
|
| 814 |
-
"loss": 0.
|
| 815 |
"step": 3250
|
| 816 |
},
|
| 817 |
{
|
| 818 |
-
"epoch": 0.
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 3275
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 1.0,
|
| 825 |
-
"learning_rate":
|
| 826 |
-
"loss": 0.
|
| 827 |
"step": 3300
|
| 828 |
},
|
| 829 |
{
|
| 830 |
-
"epoch": 1.
|
| 831 |
-
"learning_rate":
|
| 832 |
-
"loss": 0.
|
| 833 |
"step": 3325
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 1.01,
|
| 837 |
-
"learning_rate":
|
| 838 |
-
"loss": 0.
|
| 839 |
"step": 3350
|
| 840 |
},
|
| 841 |
{
|
| 842 |
-
"epoch": 1.
|
| 843 |
-
"learning_rate":
|
| 844 |
-
"loss": 0.
|
| 845 |
"step": 3375
|
| 846 |
},
|
| 847 |
{
|
| 848 |
-
"epoch": 1.
|
| 849 |
-
"learning_rate":
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 3400
|
| 852 |
},
|
| 853 |
{
|
| 854 |
-
"epoch": 1.
|
| 855 |
-
"learning_rate":
|
| 856 |
-
"loss": 0.
|
| 857 |
"step": 3425
|
| 858 |
},
|
| 859 |
{
|
| 860 |
-
"epoch": 1.
|
| 861 |
-
"learning_rate":
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 3450
|
| 864 |
},
|
| 865 |
{
|
| 866 |
-
"epoch": 1.
|
| 867 |
-
"learning_rate":
|
| 868 |
-
"loss": 0.
|
| 869 |
"step": 3475
|
| 870 |
},
|
| 871 |
{
|
| 872 |
-
"epoch": 1.
|
| 873 |
-
"learning_rate":
|
| 874 |
-
"loss": 0.
|
| 875 |
"step": 3500
|
| 876 |
},
|
| 877 |
{
|
| 878 |
-
"epoch": 1.
|
| 879 |
-
"learning_rate":
|
| 880 |
-
"loss": 0.
|
| 881 |
"step": 3525
|
| 882 |
},
|
| 883 |
{
|
| 884 |
-
"epoch": 1.
|
| 885 |
-
"learning_rate":
|
| 886 |
-
"loss": 0.
|
| 887 |
"step": 3550
|
| 888 |
},
|
| 889 |
{
|
| 890 |
-
"epoch": 1.
|
| 891 |
-
"learning_rate":
|
| 892 |
-
"loss": 0.
|
| 893 |
"step": 3575
|
| 894 |
},
|
| 895 |
{
|
| 896 |
-
"epoch": 1.
|
| 897 |
-
"learning_rate":
|
| 898 |
-
"loss": 0.
|
| 899 |
"step": 3600
|
| 900 |
},
|
| 901 |
{
|
| 902 |
-
"epoch": 1.
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 3625
|
| 906 |
},
|
| 907 |
{
|
| 908 |
-
"epoch": 1.
|
| 909 |
-
"learning_rate":
|
| 910 |
-
"loss": 0.
|
| 911 |
"step": 3650
|
| 912 |
},
|
| 913 |
{
|
| 914 |
-
"epoch": 1.
|
| 915 |
-
"learning_rate":
|
| 916 |
-
"loss": 0.
|
| 917 |
"step": 3675
|
| 918 |
},
|
| 919 |
{
|
| 920 |
-
"epoch": 1.
|
| 921 |
-
"learning_rate":
|
| 922 |
-
"loss": 0.
|
| 923 |
"step": 3700
|
| 924 |
},
|
| 925 |
{
|
| 926 |
-
"epoch": 1.
|
| 927 |
-
"learning_rate":
|
| 928 |
-
"loss": 0.
|
| 929 |
"step": 3725
|
| 930 |
},
|
| 931 |
{
|
| 932 |
-
"epoch": 1.
|
| 933 |
-
"learning_rate":
|
| 934 |
-
"loss": 0.
|
| 935 |
"step": 3750
|
| 936 |
},
|
| 937 |
{
|
| 938 |
-
"epoch": 1.
|
| 939 |
-
"learning_rate":
|
| 940 |
-
"loss": 0.
|
| 941 |
"step": 3775
|
| 942 |
},
|
| 943 |
{
|
| 944 |
-
"epoch": 1.
|
| 945 |
-
"learning_rate":
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 3800
|
| 948 |
},
|
| 949 |
{
|
| 950 |
-
"epoch": 1.
|
| 951 |
-
"learning_rate":
|
| 952 |
-
"loss": 0.
|
| 953 |
"step": 3825
|
| 954 |
},
|
| 955 |
{
|
| 956 |
-
"epoch": 1.
|
| 957 |
-
"learning_rate":
|
| 958 |
-
"loss": 0.
|
| 959 |
"step": 3850
|
| 960 |
},
|
| 961 |
{
|
| 962 |
-
"epoch": 1.
|
| 963 |
-
"learning_rate":
|
| 964 |
-
"loss": 0.
|
| 965 |
"step": 3875
|
| 966 |
},
|
| 967 |
{
|
| 968 |
-
"epoch": 1.
|
| 969 |
-
"learning_rate":
|
| 970 |
-
"loss": 0.
|
| 971 |
"step": 3900
|
| 972 |
},
|
| 973 |
{
|
| 974 |
-
"epoch": 1.
|
| 975 |
-
"learning_rate":
|
| 976 |
-
"loss": 0.
|
| 977 |
"step": 3925
|
| 978 |
},
|
| 979 |
{
|
| 980 |
-
"epoch": 1.
|
| 981 |
-
"learning_rate":
|
| 982 |
-
"loss": 0.
|
| 983 |
"step": 3950
|
| 984 |
},
|
| 985 |
{
|
| 986 |
-
"epoch": 1.
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 3975
|
| 990 |
},
|
| 991 |
{
|
| 992 |
-
"epoch": 1.
|
| 993 |
-
"learning_rate":
|
| 994 |
-
"loss": 0.
|
| 995 |
"step": 4000
|
| 996 |
},
|
| 997 |
{
|
| 998 |
-
"epoch": 1.
|
| 999 |
-
"eval_loss": 0.
|
| 1000 |
-
"eval_runtime":
|
| 1001 |
-
"eval_samples_per_second": 3.
|
| 1002 |
-
"eval_steps_per_second": 0.
|
| 1003 |
-
"eval_wer":
|
| 1004 |
"step": 4000
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"epoch": 1.
|
| 1008 |
-
"learning_rate":
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 4025
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
-
"epoch": 1.
|
| 1014 |
-
"learning_rate":
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 4050
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
-
"epoch": 1.
|
| 1020 |
-
"learning_rate":
|
| 1021 |
-
"loss": 0.
|
| 1022 |
"step": 4075
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
-
"epoch": 1.
|
| 1026 |
-
"learning_rate":
|
| 1027 |
-
"loss": 0.
|
| 1028 |
"step": 4100
|
| 1029 |
},
|
| 1030 |
{
|
| 1031 |
-
"epoch": 1.
|
| 1032 |
-
"learning_rate":
|
| 1033 |
-
"loss": 0.
|
| 1034 |
"step": 4125
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
-
"epoch": 1.
|
| 1038 |
-
"learning_rate":
|
| 1039 |
-
"loss": 0.
|
| 1040 |
"step": 4150
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
-
"epoch": 1.
|
| 1044 |
-
"learning_rate":
|
| 1045 |
-
"loss": 0.
|
| 1046 |
"step": 4175
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
-
"epoch": 1.
|
| 1050 |
-
"learning_rate":
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 4200
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
-
"epoch": 1.
|
| 1056 |
-
"learning_rate":
|
| 1057 |
-
"loss": 0.
|
| 1058 |
"step": 4225
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
-
"epoch": 1.
|
| 1062 |
-
"learning_rate":
|
| 1063 |
-
"loss": 0.
|
| 1064 |
"step": 4250
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
-
"epoch": 1.
|
| 1068 |
-
"learning_rate":
|
| 1069 |
-
"loss": 0.
|
| 1070 |
"step": 4275
|
| 1071 |
},
|
| 1072 |
{
|
| 1073 |
-
"epoch": 1.
|
| 1074 |
-
"learning_rate":
|
| 1075 |
-
"loss": 0.
|
| 1076 |
"step": 4300
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
-
"epoch": 1.
|
| 1080 |
-
"learning_rate":
|
| 1081 |
-
"loss": 0.
|
| 1082 |
"step": 4325
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
-
"epoch": 1.
|
| 1086 |
-
"learning_rate":
|
| 1087 |
-
"loss": 0.
|
| 1088 |
"step": 4350
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
-
"epoch": 1.
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 4375
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
-
"epoch": 1.
|
| 1098 |
-
"learning_rate":
|
| 1099 |
-
"loss": 0.
|
| 1100 |
"step": 4400
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
-
"epoch": 1.
|
| 1104 |
-
"learning_rate":
|
| 1105 |
-
"loss": 0.
|
| 1106 |
"step": 4425
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
-
"epoch": 1.
|
| 1110 |
-
"learning_rate":
|
| 1111 |
-
"loss": 0.
|
| 1112 |
"step": 4450
|
| 1113 |
},
|
| 1114 |
{
|
| 1115 |
-
"epoch": 1.
|
| 1116 |
-
"learning_rate":
|
| 1117 |
-
"loss": 0.
|
| 1118 |
"step": 4475
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
-
"epoch": 1.
|
| 1122 |
-
"learning_rate":
|
| 1123 |
-
"loss": 0.
|
| 1124 |
"step": 4500
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
-
"epoch": 1.
|
| 1128 |
-
"learning_rate":
|
| 1129 |
-
"loss": 0.
|
| 1130 |
"step": 4525
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
-
"epoch": 1.
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 4550
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch": 1.
|
| 1140 |
-
"learning_rate":
|
| 1141 |
-
"loss": 0.
|
| 1142 |
"step": 4575
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
-
"epoch": 1.
|
| 1146 |
-
"learning_rate":
|
| 1147 |
-
"loss": 0.
|
| 1148 |
"step": 4600
|
| 1149 |
},
|
| 1150 |
{
|
| 1151 |
-
"epoch": 1.
|
| 1152 |
-
"learning_rate":
|
| 1153 |
-
"loss": 0.
|
| 1154 |
"step": 4625
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
-
"epoch": 1.
|
| 1158 |
-
"learning_rate":
|
| 1159 |
-
"loss": 0.
|
| 1160 |
"step": 4650
|
| 1161 |
},
|
| 1162 |
{
|
| 1163 |
-
"epoch": 1.
|
| 1164 |
-
"learning_rate":
|
| 1165 |
-
"loss": 0.
|
| 1166 |
"step": 4675
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
-
"epoch": 1.
|
| 1170 |
-
"learning_rate":
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 4700
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
-
"epoch": 1.
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 4725
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
-
"epoch": 1.
|
| 1182 |
-
"learning_rate":
|
| 1183 |
-
"loss": 0.
|
| 1184 |
"step": 4750
|
| 1185 |
},
|
| 1186 |
{
|
| 1187 |
-
"epoch": 1.
|
| 1188 |
-
"learning_rate":
|
| 1189 |
-
"loss": 0.
|
| 1190 |
"step": 4775
|
| 1191 |
},
|
| 1192 |
{
|
| 1193 |
-
"epoch": 1.
|
| 1194 |
-
"learning_rate":
|
| 1195 |
-
"loss": 0.
|
| 1196 |
"step": 4800
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
-
"epoch": 1.
|
| 1200 |
-
"learning_rate": 3.
|
| 1201 |
-
"loss": 0.
|
| 1202 |
"step": 4825
|
| 1203 |
},
|
| 1204 |
{
|
| 1205 |
-
"epoch": 1.
|
| 1206 |
-
"learning_rate": 3.
|
| 1207 |
-
"loss": 0.
|
| 1208 |
"step": 4850
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
-
"epoch": 1.
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 4875
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
-
"epoch": 1.
|
| 1218 |
-
"learning_rate":
|
| 1219 |
-
"loss": 0.
|
| 1220 |
"step": 4900
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
-
"epoch": 1.
|
| 1224 |
-
"learning_rate":
|
| 1225 |
-
"loss": 0.
|
| 1226 |
"step": 4925
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
-
"epoch": 1.
|
| 1230 |
-
"learning_rate":
|
| 1231 |
-
"loss": 0.
|
| 1232 |
"step": 4950
|
| 1233 |
},
|
| 1234 |
{
|
| 1235 |
-
"epoch": 1.
|
| 1236 |
-
"learning_rate":
|
| 1237 |
-
"loss": 0.
|
| 1238 |
"step": 4975
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
-
"epoch": 1.
|
| 1242 |
-
"learning_rate":
|
| 1243 |
-
"loss": 0.
|
| 1244 |
"step": 5000
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
-
"epoch": 1.
|
| 1248 |
-
"eval_loss": 0.
|
| 1249 |
-
"eval_runtime":
|
| 1250 |
-
"eval_samples_per_second": 3.
|
| 1251 |
-
"eval_steps_per_second": 0.
|
| 1252 |
-
"eval_wer":
|
| 1253 |
"step": 5000
|
| 1254 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1255 |
{
|
| 1256 |
"epoch": 1.34,
|
| 1257 |
-
"
|
| 1258 |
-
"
|
| 1259 |
-
"
|
| 1260 |
-
|
| 1261 |
-
|
| 1262 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1263 |
}
|
| 1264 |
],
|
| 1265 |
-
"max_steps":
|
| 1266 |
"num_train_epochs": 9223372036854775807,
|
| 1267 |
-
"total_flos": 2.
|
| 1268 |
"trial_name": null,
|
| 1269 |
"trial_params": null
|
| 1270 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 13.179958686054519,
|
| 3 |
+
"best_model_checkpoint": "./checkpoint-7000",
|
| 4 |
+
"epoch": 2.0582857142857143,
|
| 5 |
+
"global_step": 7000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
"learning_rate": 4.6000000000000004e-07,
|
| 13 |
"loss": 1.4182,
|
| 14 |
"step": 25
|
|
|
|
| 26 |
"step": 75
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
"learning_rate": 1.94e-06,
|
| 31 |
"loss": 0.7765,
|
| 32 |
"step": 100
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
"learning_rate": 2.4400000000000004e-06,
|
| 37 |
"loss": 0.7103,
|
| 38 |
"step": 125
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
"learning_rate": 2.9400000000000002e-06,
|
| 43 |
"loss": 0.6597,
|
| 44 |
"step": 150
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
"learning_rate": 3.44e-06,
|
| 49 |
"loss": 0.6657,
|
| 50 |
"step": 175
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
"learning_rate": 3.94e-06,
|
| 55 |
"loss": 0.5853,
|
| 56 |
"step": 200
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
"learning_rate": 4.440000000000001e-06,
|
| 61 |
"loss": 0.5273,
|
| 62 |
"step": 225
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
"learning_rate": 4.94e-06,
|
| 67 |
"loss": 0.5979,
|
| 68 |
"step": 250
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
"learning_rate": 5.4400000000000004e-06,
|
| 73 |
"loss": 0.5861,
|
| 74 |
"step": 275
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
"learning_rate": 5.94e-06,
|
| 79 |
"loss": 0.5085,
|
| 80 |
"step": 300
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.05,
|
| 84 |
"learning_rate": 6.440000000000001e-06,
|
| 85 |
"loss": 0.4827,
|
| 86 |
"step": 325
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.05,
|
| 90 |
"learning_rate": 6.9400000000000005e-06,
|
| 91 |
"loss": 0.4909,
|
| 92 |
"step": 350
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
"learning_rate": 7.440000000000001e-06,
|
| 97 |
"loss": 0.4651,
|
| 98 |
"step": 375
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 0.06,
|
| 102 |
"learning_rate": 7.94e-06,
|
| 103 |
"loss": 0.494,
|
| 104 |
"step": 400
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 0.06,
|
| 108 |
"learning_rate": 8.44e-06,
|
| 109 |
"loss": 0.4188,
|
| 110 |
"step": 425
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
"learning_rate": 8.94e-06,
|
| 115 |
"loss": 0.3849,
|
| 116 |
"step": 450
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 0.07,
|
| 120 |
"learning_rate": 9.440000000000001e-06,
|
| 121 |
"loss": 0.4577,
|
| 122 |
"step": 475
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.07,
|
| 126 |
"learning_rate": 9.940000000000001e-06,
|
| 127 |
"loss": 0.4415,
|
| 128 |
"step": 500
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.07,
|
| 132 |
+
"learning_rate": 9.966153846153847e-06,
|
| 133 |
+
"loss": 0.4614,
|
| 134 |
"step": 525
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.08,
|
| 138 |
+
"learning_rate": 9.927692307692309e-06,
|
| 139 |
+
"loss": 0.4283,
|
| 140 |
"step": 550
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 0.08,
|
| 144 |
+
"learning_rate": 9.88923076923077e-06,
|
| 145 |
+
"loss": 0.4486,
|
| 146 |
"step": 575
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 0.09,
|
| 150 |
+
"learning_rate": 9.850769230769231e-06,
|
| 151 |
+
"loss": 0.4434,
|
| 152 |
"step": 600
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.09,
|
| 156 |
+
"learning_rate": 9.812307692307694e-06,
|
| 157 |
+
"loss": 0.4245,
|
| 158 |
"step": 625
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.09,
|
| 162 |
+
"learning_rate": 9.773846153846154e-06,
|
| 163 |
+
"loss": 0.4254,
|
| 164 |
"step": 650
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.1,
|
| 168 |
+
"learning_rate": 9.735384615384616e-06,
|
| 169 |
+
"loss": 0.4143,
|
| 170 |
"step": 675
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.1,
|
| 174 |
+
"learning_rate": 9.696923076923078e-06,
|
| 175 |
+
"loss": 0.4115,
|
| 176 |
"step": 700
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.1,
|
| 180 |
+
"learning_rate": 9.658461538461539e-06,
|
| 181 |
+
"loss": 0.3656,
|
| 182 |
"step": 725
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.11,
|
| 186 |
+
"learning_rate": 9.620000000000001e-06,
|
| 187 |
+
"loss": 0.4004,
|
| 188 |
"step": 750
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 0.11,
|
| 192 |
+
"learning_rate": 9.581538461538462e-06,
|
| 193 |
+
"loss": 0.388,
|
| 194 |
"step": 775
|
| 195 |
},
|
| 196 |
{
|
| 197 |
+
"epoch": 0.11,
|
| 198 |
+
"learning_rate": 9.543076923076924e-06,
|
| 199 |
+
"loss": 0.3665,
|
| 200 |
"step": 800
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 0.12,
|
| 204 |
+
"learning_rate": 9.504615384615386e-06,
|
| 205 |
+
"loss": 0.3777,
|
| 206 |
"step": 825
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.12,
|
| 210 |
+
"learning_rate": 9.466153846153846e-06,
|
| 211 |
+
"loss": 0.4338,
|
| 212 |
"step": 850
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.12,
|
| 216 |
+
"learning_rate": 9.427692307692309e-06,
|
| 217 |
+
"loss": 0.4017,
|
| 218 |
"step": 875
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.13,
|
| 222 |
+
"learning_rate": 9.38923076923077e-06,
|
| 223 |
+
"loss": 0.3757,
|
| 224 |
"step": 900
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.13,
|
| 228 |
+
"learning_rate": 9.350769230769231e-06,
|
| 229 |
+
"loss": 0.4084,
|
| 230 |
"step": 925
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 0.14,
|
| 234 |
+
"learning_rate": 9.312307692307693e-06,
|
| 235 |
+
"loss": 0.3821,
|
| 236 |
"step": 950
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 0.14,
|
| 240 |
+
"learning_rate": 9.273846153846154e-06,
|
| 241 |
+
"loss": 0.421,
|
| 242 |
"step": 975
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 0.14,
|
| 246 |
+
"learning_rate": 9.235384615384616e-06,
|
| 247 |
+
"loss": 0.4203,
|
| 248 |
"step": 1000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.14,
|
| 252 |
+
"eval_loss": 0.4127572178840637,
|
| 253 |
+
"eval_runtime": 1802.5775,
|
| 254 |
+
"eval_samples_per_second": 3.656,
|
| 255 |
+
"eval_steps_per_second": 0.457,
|
| 256 |
+
"eval_wer": 28.265624367127064,
|
| 257 |
"step": 1000
|
| 258 |
},
|
| 259 |
{
|
| 260 |
+
"epoch": 0.15,
|
| 261 |
+
"learning_rate": 9.196923076923078e-06,
|
| 262 |
+
"loss": 0.4094,
|
| 263 |
"step": 1025
|
| 264 |
},
|
| 265 |
{
|
| 266 |
+
"epoch": 0.15,
|
| 267 |
+
"learning_rate": 9.158461538461539e-06,
|
| 268 |
"loss": 0.3674,
|
| 269 |
"step": 1050
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 0.15,
|
| 273 |
+
"learning_rate": 9.12e-06,
|
| 274 |
+
"loss": 0.3613,
|
| 275 |
"step": 1075
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.16,
|
| 279 |
+
"learning_rate": 9.081538461538463e-06,
|
| 280 |
+
"loss": 0.3904,
|
| 281 |
"step": 1100
|
| 282 |
},
|
| 283 |
{
|
| 284 |
+
"epoch": 0.16,
|
| 285 |
+
"learning_rate": 9.043076923076923e-06,
|
| 286 |
+
"loss": 0.3865,
|
| 287 |
"step": 1125
|
| 288 |
},
|
| 289 |
{
|
| 290 |
+
"epoch": 0.16,
|
| 291 |
+
"learning_rate": 9.004615384615386e-06,
|
| 292 |
+
"loss": 0.3677,
|
| 293 |
"step": 1150
|
| 294 |
},
|
| 295 |
{
|
| 296 |
+
"epoch": 0.17,
|
| 297 |
+
"learning_rate": 8.966153846153846e-06,
|
| 298 |
+
"loss": 0.3295,
|
| 299 |
"step": 1175
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 0.17,
|
| 303 |
+
"learning_rate": 8.927692307692308e-06,
|
| 304 |
+
"loss": 0.3207,
|
| 305 |
"step": 1200
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 0.17,
|
| 309 |
+
"learning_rate": 8.88923076923077e-06,
|
| 310 |
+
"loss": 0.2817,
|
| 311 |
"step": 1225
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 0.18,
|
| 315 |
+
"learning_rate": 8.850769230769231e-06,
|
| 316 |
+
"loss": 0.3218,
|
| 317 |
"step": 1250
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 0.18,
|
| 321 |
+
"learning_rate": 8.812307692307693e-06,
|
| 322 |
+
"loss": 0.3502,
|
| 323 |
"step": 1275
|
| 324 |
},
|
| 325 |
{
|
| 326 |
+
"epoch": 0.19,
|
| 327 |
+
"learning_rate": 8.773846153846155e-06,
|
| 328 |
+
"loss": 0.3432,
|
| 329 |
"step": 1300
|
| 330 |
},
|
| 331 |
{
|
| 332 |
+
"epoch": 0.19,
|
| 333 |
+
"learning_rate": 8.735384615384616e-06,
|
| 334 |
+
"loss": 0.3244,
|
| 335 |
"step": 1325
|
| 336 |
},
|
| 337 |
{
|
| 338 |
+
"epoch": 0.19,
|
| 339 |
+
"learning_rate": 8.696923076923078e-06,
|
| 340 |
+
"loss": 0.2607,
|
| 341 |
"step": 1350
|
| 342 |
},
|
| 343 |
{
|
| 344 |
+
"epoch": 0.2,
|
| 345 |
+
"learning_rate": 8.658461538461538e-06,
|
| 346 |
+
"loss": 0.3469,
|
| 347 |
"step": 1375
|
| 348 |
},
|
| 349 |
{
|
| 350 |
+
"epoch": 0.2,
|
| 351 |
+
"learning_rate": 8.62e-06,
|
| 352 |
+
"loss": 0.2756,
|
| 353 |
"step": 1400
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 0.2,
|
| 357 |
+
"learning_rate": 8.581538461538463e-06,
|
| 358 |
+
"loss": 0.294,
|
| 359 |
"step": 1425
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 0.21,
|
| 363 |
+
"learning_rate": 8.543076923076923e-06,
|
| 364 |
+
"loss": 0.3915,
|
| 365 |
"step": 1450
|
| 366 |
},
|
| 367 |
{
|
| 368 |
+
"epoch": 0.21,
|
| 369 |
+
"learning_rate": 8.504615384615385e-06,
|
| 370 |
+
"loss": 0.3237,
|
| 371 |
"step": 1475
|
| 372 |
},
|
| 373 |
{
|
| 374 |
+
"epoch": 0.21,
|
| 375 |
+
"learning_rate": 8.466153846153847e-06,
|
| 376 |
+
"loss": 0.2723,
|
| 377 |
"step": 1500
|
| 378 |
},
|
| 379 |
{
|
| 380 |
+
"epoch": 0.22,
|
| 381 |
+
"learning_rate": 8.427692307692308e-06,
|
| 382 |
+
"loss": 0.2917,
|
| 383 |
"step": 1525
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 0.22,
|
| 387 |
+
"learning_rate": 8.38923076923077e-06,
|
| 388 |
+
"loss": 0.2921,
|
| 389 |
"step": 1550
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 0.23,
|
| 393 |
+
"learning_rate": 8.35076923076923e-06,
|
| 394 |
+
"loss": 0.3724,
|
| 395 |
"step": 1575
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 0.23,
|
| 399 |
+
"learning_rate": 8.312307692307693e-06,
|
| 400 |
+
"loss": 0.3364,
|
| 401 |
"step": 1600
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.23,
|
| 405 |
+
"learning_rate": 8.273846153846155e-06,
|
| 406 |
+
"loss": 0.3661,
|
| 407 |
"step": 1625
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 0.24,
|
| 411 |
+
"learning_rate": 8.235384615384615e-06,
|
| 412 |
+
"loss": 0.3307,
|
| 413 |
"step": 1650
|
| 414 |
},
|
| 415 |
{
|
| 416 |
+
"epoch": 0.24,
|
| 417 |
+
"learning_rate": 8.196923076923078e-06,
|
| 418 |
+
"loss": 0.3201,
|
| 419 |
"step": 1675
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 0.24,
|
| 423 |
+
"learning_rate": 8.15846153846154e-06,
|
| 424 |
+
"loss": 0.2973,
|
| 425 |
"step": 1700
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 0.25,
|
| 429 |
+
"learning_rate": 8.120000000000002e-06,
|
| 430 |
+
"loss": 0.3013,
|
| 431 |
"step": 1725
|
| 432 |
},
|
| 433 |
{
|
| 434 |
+
"epoch": 0.25,
|
| 435 |
+
"learning_rate": 8.081538461538462e-06,
|
| 436 |
+
"loss": 0.3349,
|
| 437 |
"step": 1750
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 0.25,
|
| 441 |
+
"learning_rate": 8.043076923076923e-06,
|
| 442 |
+
"loss": 0.3244,
|
| 443 |
"step": 1775
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 0.26,
|
| 447 |
+
"learning_rate": 8.004615384615385e-06,
|
| 448 |
+
"loss": 0.3331,
|
| 449 |
"step": 1800
|
| 450 |
},
|
| 451 |
{
|
| 452 |
+
"epoch": 0.26,
|
| 453 |
+
"learning_rate": 7.966153846153847e-06,
|
| 454 |
+
"loss": 0.2723,
|
| 455 |
"step": 1825
|
| 456 |
},
|
| 457 |
{
|
| 458 |
+
"epoch": 0.26,
|
| 459 |
+
"learning_rate": 7.927692307692308e-06,
|
| 460 |
+
"loss": 0.316,
|
| 461 |
"step": 1850
|
| 462 |
},
|
| 463 |
{
|
| 464 |
+
"epoch": 0.27,
|
| 465 |
+
"learning_rate": 7.88923076923077e-06,
|
| 466 |
+
"loss": 0.2648,
|
| 467 |
"step": 1875
|
| 468 |
},
|
| 469 |
{
|
| 470 |
+
"epoch": 0.27,
|
| 471 |
+
"learning_rate": 7.850769230769232e-06,
|
| 472 |
+
"loss": 0.2965,
|
| 473 |
"step": 1900
|
| 474 |
},
|
| 475 |
{
|
| 476 |
+
"epoch": 0.28,
|
| 477 |
+
"learning_rate": 7.812307692307694e-06,
|
| 478 |
+
"loss": 0.256,
|
| 479 |
"step": 1925
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 0.28,
|
| 483 |
+
"learning_rate": 7.773846153846155e-06,
|
| 484 |
+
"loss": 0.2156,
|
| 485 |
"step": 1950
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 0.28,
|
| 489 |
+
"learning_rate": 7.735384615384615e-06,
|
| 490 |
+
"loss": 0.2418,
|
| 491 |
"step": 1975
|
| 492 |
},
|
| 493 |
{
|
| 494 |
+
"epoch": 0.29,
|
| 495 |
+
"learning_rate": 7.696923076923077e-06,
|
| 496 |
+
"loss": 0.2693,
|
| 497 |
"step": 2000
|
| 498 |
},
|
| 499 |
{
|
| 500 |
+
"epoch": 0.29,
|
| 501 |
+
"eval_loss": 0.3239877223968506,
|
| 502 |
+
"eval_runtime": 1817.7748,
|
| 503 |
+
"eval_samples_per_second": 3.626,
|
| 504 |
+
"eval_steps_per_second": 0.453,
|
| 505 |
+
"eval_wer": 22.052330997610273,
|
| 506 |
"step": 2000
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 0.29,
|
| 510 |
+
"learning_rate": 7.65846153846154e-06,
|
| 511 |
+
"loss": 0.2191,
|
| 512 |
"step": 2025
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 0.29,
|
| 516 |
+
"learning_rate": 7.620000000000001e-06,
|
| 517 |
+
"loss": 0.2437,
|
| 518 |
"step": 2050
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 0.3,
|
| 522 |
+
"learning_rate": 7.581538461538462e-06,
|
| 523 |
+
"loss": 0.2519,
|
| 524 |
"step": 2075
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 0.3,
|
| 528 |
+
"learning_rate": 7.544615384615386e-06,
|
| 529 |
+
"loss": 0.3198,
|
| 530 |
"step": 2100
|
| 531 |
},
|
| 532 |
{
|
| 533 |
+
"epoch": 0.3,
|
| 534 |
+
"learning_rate": 7.5061538461538465e-06,
|
| 535 |
+
"loss": 0.2708,
|
| 536 |
"step": 2125
|
| 537 |
},
|
| 538 |
{
|
| 539 |
+
"epoch": 0.31,
|
| 540 |
+
"learning_rate": 7.467692307692308e-06,
|
| 541 |
+
"loss": 0.265,
|
| 542 |
"step": 2150
|
| 543 |
},
|
| 544 |
{
|
| 545 |
+
"epoch": 0.31,
|
| 546 |
+
"learning_rate": 7.42923076923077e-06,
|
| 547 |
+
"loss": 0.2935,
|
| 548 |
"step": 2175
|
| 549 |
},
|
| 550 |
{
|
| 551 |
+
"epoch": 0.31,
|
| 552 |
+
"learning_rate": 7.390769230769231e-06,
|
| 553 |
+
"loss": 0.2595,
|
| 554 |
"step": 2200
|
| 555 |
},
|
| 556 |
{
|
| 557 |
+
"epoch": 0.32,
|
| 558 |
+
"learning_rate": 7.3523076923076935e-06,
|
| 559 |
+
"loss": 0.2613,
|
| 560 |
"step": 2225
|
| 561 |
},
|
| 562 |
{
|
| 563 |
+
"epoch": 0.32,
|
| 564 |
+
"learning_rate": 7.313846153846155e-06,
|
| 565 |
+
"loss": 0.2229,
|
| 566 |
"step": 2250
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 0.33,
|
| 570 |
+
"learning_rate": 7.275384615384616e-06,
|
| 571 |
+
"loss": 0.2814,
|
| 572 |
"step": 2275
|
| 573 |
},
|
| 574 |
{
|
| 575 |
+
"epoch": 0.33,
|
| 576 |
+
"learning_rate": 7.236923076923078e-06,
|
| 577 |
+
"loss": 0.2681,
|
| 578 |
"step": 2300
|
| 579 |
},
|
| 580 |
{
|
| 581 |
+
"epoch": 0.33,
|
| 582 |
+
"learning_rate": 7.198461538461539e-06,
|
| 583 |
+
"loss": 0.2531,
|
| 584 |
"step": 2325
|
| 585 |
},
|
| 586 |
{
|
| 587 |
+
"epoch": 0.34,
|
| 588 |
+
"learning_rate": 7.16e-06,
|
| 589 |
+
"loss": 0.2104,
|
| 590 |
"step": 2350
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 0.34,
|
| 594 |
+
"learning_rate": 7.121538461538462e-06,
|
| 595 |
+
"loss": 0.2294,
|
| 596 |
"step": 2375
|
| 597 |
},
|
| 598 |
{
|
| 599 |
+
"epoch": 0.34,
|
| 600 |
+
"learning_rate": 7.0830769230769235e-06,
|
| 601 |
+
"loss": 0.2556,
|
| 602 |
"step": 2400
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 0.35,
|
| 606 |
+
"learning_rate": 7.044615384615386e-06,
|
| 607 |
+
"loss": 0.2185,
|
| 608 |
"step": 2425
|
| 609 |
},
|
| 610 |
{
|
| 611 |
+
"epoch": 0.35,
|
| 612 |
+
"learning_rate": 7.006153846153847e-06,
|
| 613 |
+
"loss": 0.2585,
|
| 614 |
"step": 2450
|
| 615 |
},
|
| 616 |
{
|
| 617 |
+
"epoch": 0.35,
|
| 618 |
+
"learning_rate": 6.967692307692308e-06,
|
| 619 |
+
"loss": 0.2325,
|
| 620 |
"step": 2475
|
| 621 |
},
|
| 622 |
{
|
| 623 |
+
"epoch": 0.36,
|
| 624 |
+
"learning_rate": 6.9292307692307705e-06,
|
| 625 |
+
"loss": 0.2068,
|
| 626 |
"step": 2500
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 0.36,
|
| 630 |
+
"learning_rate": 6.890769230769231e-06,
|
| 631 |
+
"loss": 0.2233,
|
| 632 |
"step": 2525
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 0.36,
|
| 636 |
+
"learning_rate": 6.852307692307692e-06,
|
| 637 |
+
"loss": 0.2649,
|
| 638 |
"step": 2550
|
| 639 |
},
|
| 640 |
{
|
| 641 |
+
"epoch": 0.37,
|
| 642 |
+
"learning_rate": 6.8138461538461545e-06,
|
| 643 |
+
"loss": 0.2524,
|
| 644 |
"step": 2575
|
| 645 |
},
|
| 646 |
{
|
| 647 |
+
"epoch": 0.37,
|
| 648 |
+
"learning_rate": 6.775384615384616e-06,
|
| 649 |
+
"loss": 0.2847,
|
| 650 |
"step": 2600
|
| 651 |
},
|
| 652 |
{
|
| 653 |
+
"epoch": 0.38,
|
| 654 |
+
"learning_rate": 6.736923076923078e-06,
|
| 655 |
+
"loss": 0.1954,
|
| 656 |
"step": 2625
|
| 657 |
},
|
| 658 |
{
|
| 659 |
+
"epoch": 0.38,
|
| 660 |
+
"learning_rate": 6.698461538461539e-06,
|
| 661 |
+
"loss": 0.2553,
|
| 662 |
"step": 2650
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"epoch": 0.38,
|
| 666 |
+
"learning_rate": 6.660000000000001e-06,
|
| 667 |
+
"loss": 0.2138,
|
| 668 |
"step": 2675
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 0.39,
|
| 672 |
+
"learning_rate": 6.621538461538463e-06,
|
| 673 |
+
"loss": 0.2245,
|
| 674 |
"step": 2700
|
| 675 |
},
|
| 676 |
{
|
| 677 |
+
"epoch": 0.39,
|
| 678 |
+
"learning_rate": 6.583076923076923e-06,
|
| 679 |
+
"loss": 0.2137,
|
| 680 |
"step": 2725
|
| 681 |
},
|
| 682 |
{
|
| 683 |
+
"epoch": 0.39,
|
| 684 |
+
"learning_rate": 6.5446153846153846e-06,
|
| 685 |
+
"loss": 0.2558,
|
| 686 |
"step": 2750
|
| 687 |
},
|
| 688 |
{
|
| 689 |
+
"epoch": 0.4,
|
| 690 |
+
"learning_rate": 6.506153846153847e-06,
|
| 691 |
+
"loss": 0.1981,
|
| 692 |
"step": 2775
|
| 693 |
},
|
| 694 |
{
|
| 695 |
+
"epoch": 0.4,
|
| 696 |
+
"learning_rate": 6.467692307692308e-06,
|
| 697 |
+
"loss": 0.2327,
|
| 698 |
"step": 2800
|
| 699 |
},
|
| 700 |
{
|
| 701 |
+
"epoch": 0.4,
|
| 702 |
+
"learning_rate": 6.42923076923077e-06,
|
| 703 |
+
"loss": 0.219,
|
| 704 |
"step": 2825
|
| 705 |
},
|
| 706 |
{
|
| 707 |
+
"epoch": 0.41,
|
| 708 |
+
"learning_rate": 6.3907692307692315e-06,
|
| 709 |
+
"loss": 0.226,
|
| 710 |
"step": 2850
|
| 711 |
},
|
| 712 |
{
|
| 713 |
+
"epoch": 0.41,
|
| 714 |
+
"learning_rate": 6.352307692307693e-06,
|
| 715 |
+
"loss": 0.2192,
|
| 716 |
"step": 2875
|
| 717 |
},
|
| 718 |
{
|
| 719 |
+
"epoch": 0.41,
|
| 720 |
+
"learning_rate": 6.313846153846155e-06,
|
| 721 |
+
"loss": 0.2146,
|
| 722 |
"step": 2900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
+
"epoch": 0.42,
|
| 726 |
+
"learning_rate": 6.2753846153846155e-06,
|
| 727 |
+
"loss": 0.2058,
|
| 728 |
"step": 2925
|
| 729 |
},
|
| 730 |
{
|
| 731 |
+
"epoch": 0.42,
|
| 732 |
+
"learning_rate": 6.236923076923077e-06,
|
| 733 |
+
"loss": 0.2369,
|
| 734 |
"step": 2950
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"epoch": 0.42,
|
| 738 |
+
"learning_rate": 6.198461538461539e-06,
|
| 739 |
+
"loss": 0.2202,
|
| 740 |
"step": 2975
|
| 741 |
},
|
| 742 |
{
|
| 743 |
+
"epoch": 0.43,
|
| 744 |
+
"learning_rate": 6.16e-06,
|
| 745 |
+
"loss": 0.2228,
|
| 746 |
"step": 3000
|
| 747 |
},
|
| 748 |
{
|
| 749 |
+
"epoch": 0.43,
|
| 750 |
+
"eval_loss": 0.27374517917633057,
|
| 751 |
+
"eval_runtime": 1805.7724,
|
| 752 |
+
"eval_samples_per_second": 3.65,
|
| 753 |
+
"eval_steps_per_second": 0.456,
|
| 754 |
+
"eval_wer": 18.14370772408765,
|
| 755 |
"step": 3000
|
| 756 |
},
|
| 757 |
{
|
| 758 |
+
"epoch": 0.43,
|
| 759 |
+
"learning_rate": 6.1215384615384625e-06,
|
| 760 |
+
"loss": 0.2029,
|
| 761 |
"step": 3025
|
| 762 |
},
|
| 763 |
{
|
| 764 |
+
"epoch": 0.44,
|
| 765 |
+
"learning_rate": 6.083076923076924e-06,
|
| 766 |
+
"loss": 0.2085,
|
| 767 |
"step": 3050
|
| 768 |
},
|
| 769 |
{
|
| 770 |
+
"epoch": 0.44,
|
| 771 |
+
"learning_rate": 6.044615384615385e-06,
|
| 772 |
+
"loss": 0.196,
|
| 773 |
"step": 3075
|
| 774 |
},
|
| 775 |
{
|
| 776 |
+
"epoch": 0.44,
|
| 777 |
+
"learning_rate": 6.006153846153847e-06,
|
| 778 |
+
"loss": 0.1999,
|
| 779 |
"step": 3100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
+
"epoch": 0.45,
|
| 783 |
+
"learning_rate": 5.967692307692308e-06,
|
| 784 |
+
"loss": 0.2076,
|
| 785 |
"step": 3125
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 0.45,
|
| 789 |
+
"learning_rate": 5.929230769230769e-06,
|
| 790 |
+
"loss": 0.1999,
|
| 791 |
"step": 3150
|
| 792 |
},
|
| 793 |
{
|
| 794 |
+
"epoch": 0.45,
|
| 795 |
+
"learning_rate": 5.890769230769231e-06,
|
| 796 |
+
"loss": 0.2361,
|
| 797 |
"step": 3175
|
| 798 |
},
|
| 799 |
{
|
| 800 |
+
"epoch": 0.46,
|
| 801 |
+
"learning_rate": 5.8523076923076926e-06,
|
| 802 |
+
"loss": 0.2003,
|
| 803 |
"step": 3200
|
| 804 |
},
|
| 805 |
{
|
| 806 |
+
"epoch": 0.46,
|
| 807 |
+
"learning_rate": 5.813846153846155e-06,
|
| 808 |
+
"loss": 0.249,
|
| 809 |
"step": 3225
|
| 810 |
},
|
| 811 |
{
|
| 812 |
+
"epoch": 0.46,
|
| 813 |
+
"learning_rate": 5.775384615384616e-06,
|
| 814 |
+
"loss": 0.2724,
|
| 815 |
"step": 3250
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 0.47,
|
| 819 |
+
"learning_rate": 5.736923076923077e-06,
|
| 820 |
+
"loss": 0.2001,
|
| 821 |
"step": 3275
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 1.0,
|
| 825 |
+
"learning_rate": 5.6984615384615395e-06,
|
| 826 |
+
"loss": 0.2355,
|
| 827 |
"step": 3300
|
| 828 |
},
|
| 829 |
{
|
| 830 |
+
"epoch": 1.0,
|
| 831 |
+
"learning_rate": 5.66e-06,
|
| 832 |
+
"loss": 0.2297,
|
| 833 |
"step": 3325
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 1.01,
|
| 837 |
+
"learning_rate": 5.621538461538461e-06,
|
| 838 |
+
"loss": 0.2265,
|
| 839 |
"step": 3350
|
| 840 |
},
|
| 841 |
{
|
| 842 |
+
"epoch": 1.01,
|
| 843 |
+
"learning_rate": 5.5830769230769235e-06,
|
| 844 |
+
"loss": 0.21,
|
| 845 |
"step": 3375
|
| 846 |
},
|
| 847 |
{
|
| 848 |
+
"epoch": 1.01,
|
| 849 |
+
"learning_rate": 5.544615384615385e-06,
|
| 850 |
+
"loss": 0.1944,
|
| 851 |
"step": 3400
|
| 852 |
},
|
| 853 |
{
|
| 854 |
+
"epoch": 1.02,
|
| 855 |
+
"learning_rate": 5.506153846153847e-06,
|
| 856 |
+
"loss": 0.1844,
|
| 857 |
"step": 3425
|
| 858 |
},
|
| 859 |
{
|
| 860 |
+
"epoch": 1.02,
|
| 861 |
+
"learning_rate": 5.467692307692308e-06,
|
| 862 |
+
"loss": 0.1774,
|
| 863 |
"step": 3450
|
| 864 |
},
|
| 865 |
{
|
| 866 |
+
"epoch": 1.03,
|
| 867 |
+
"learning_rate": 5.42923076923077e-06,
|
| 868 |
+
"loss": 0.1688,
|
| 869 |
"step": 3475
|
| 870 |
},
|
| 871 |
{
|
| 872 |
+
"epoch": 1.03,
|
| 873 |
+
"learning_rate": 5.390769230769232e-06,
|
| 874 |
+
"loss": 0.1473,
|
| 875 |
"step": 3500
|
| 876 |
},
|
| 877 |
{
|
| 878 |
+
"epoch": 1.03,
|
| 879 |
+
"learning_rate": 5.352307692307692e-06,
|
| 880 |
+
"loss": 0.1743,
|
| 881 |
"step": 3525
|
| 882 |
},
|
| 883 |
{
|
| 884 |
+
"epoch": 1.04,
|
| 885 |
+
"learning_rate": 5.313846153846154e-06,
|
| 886 |
+
"loss": 0.1759,
|
| 887 |
"step": 3550
|
| 888 |
},
|
| 889 |
{
|
| 890 |
+
"epoch": 1.04,
|
| 891 |
+
"learning_rate": 5.275384615384616e-06,
|
| 892 |
+
"loss": 0.1507,
|
| 893 |
"step": 3575
|
| 894 |
},
|
| 895 |
{
|
| 896 |
+
"epoch": 1.04,
|
| 897 |
+
"learning_rate": 5.236923076923077e-06,
|
| 898 |
+
"loss": 0.1556,
|
| 899 |
"step": 3600
|
| 900 |
},
|
| 901 |
{
|
| 902 |
+
"epoch": 1.05,
|
| 903 |
+
"learning_rate": 5.198461538461539e-06,
|
| 904 |
+
"loss": 0.1545,
|
| 905 |
"step": 3625
|
| 906 |
},
|
| 907 |
{
|
| 908 |
+
"epoch": 1.05,
|
| 909 |
+
"learning_rate": 5.1600000000000006e-06,
|
| 910 |
+
"loss": 0.1143,
|
| 911 |
"step": 3650
|
| 912 |
},
|
| 913 |
{
|
| 914 |
+
"epoch": 1.05,
|
| 915 |
+
"learning_rate": 5.121538461538462e-06,
|
| 916 |
+
"loss": 0.1168,
|
| 917 |
"step": 3675
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 1.06,
|
| 921 |
+
"learning_rate": 5.083076923076924e-06,
|
| 922 |
+
"loss": 0.1095,
|
| 923 |
"step": 3700
|
| 924 |
},
|
| 925 |
{
|
| 926 |
+
"epoch": 1.06,
|
| 927 |
+
"learning_rate": 5.044615384615385e-06,
|
| 928 |
+
"loss": 0.1098,
|
| 929 |
"step": 3725
|
| 930 |
},
|
| 931 |
{
|
| 932 |
+
"epoch": 1.06,
|
| 933 |
+
"learning_rate": 5.006153846153846e-06,
|
| 934 |
+
"loss": 0.102,
|
| 935 |
"step": 3750
|
| 936 |
},
|
| 937 |
{
|
| 938 |
+
"epoch": 1.07,
|
| 939 |
+
"learning_rate": 4.967692307692308e-06,
|
| 940 |
+
"loss": 0.1263,
|
| 941 |
"step": 3775
|
| 942 |
},
|
| 943 |
{
|
| 944 |
+
"epoch": 1.07,
|
| 945 |
+
"learning_rate": 4.929230769230769e-06,
|
| 946 |
+
"loss": 0.093,
|
| 947 |
"step": 3800
|
| 948 |
},
|
| 949 |
{
|
| 950 |
+
"epoch": 1.08,
|
| 951 |
+
"learning_rate": 4.8907692307692315e-06,
|
| 952 |
+
"loss": 0.1163,
|
| 953 |
"step": 3825
|
| 954 |
},
|
| 955 |
{
|
| 956 |
+
"epoch": 1.08,
|
| 957 |
+
"learning_rate": 4.852307692307693e-06,
|
| 958 |
+
"loss": 0.1005,
|
| 959 |
"step": 3850
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 1.08,
|
| 963 |
+
"learning_rate": 4.813846153846154e-06,
|
| 964 |
+
"loss": 0.1036,
|
| 965 |
"step": 3875
|
| 966 |
},
|
| 967 |
{
|
| 968 |
+
"epoch": 1.09,
|
| 969 |
+
"learning_rate": 4.7753846153846155e-06,
|
| 970 |
+
"loss": 0.1289,
|
| 971 |
"step": 3900
|
| 972 |
},
|
| 973 |
{
|
| 974 |
+
"epoch": 1.09,
|
| 975 |
+
"learning_rate": 4.736923076923078e-06,
|
| 976 |
+
"loss": 0.1238,
|
| 977 |
"step": 3925
|
| 978 |
},
|
| 979 |
{
|
| 980 |
+
"epoch": 1.09,
|
| 981 |
+
"learning_rate": 4.698461538461539e-06,
|
| 982 |
+
"loss": 0.1218,
|
| 983 |
"step": 3950
|
| 984 |
},
|
| 985 |
{
|
| 986 |
+
"epoch": 1.1,
|
| 987 |
+
"learning_rate": 4.66e-06,
|
| 988 |
+
"loss": 0.1223,
|
| 989 |
"step": 3975
|
| 990 |
},
|
| 991 |
{
|
| 992 |
+
"epoch": 1.1,
|
| 993 |
+
"learning_rate": 4.621538461538462e-06,
|
| 994 |
+
"loss": 0.1002,
|
| 995 |
"step": 4000
|
| 996 |
},
|
| 997 |
{
|
| 998 |
+
"epoch": 1.1,
|
| 999 |
+
"eval_loss": 0.2553846836090088,
|
| 1000 |
+
"eval_runtime": 1819.1705,
|
| 1001 |
+
"eval_samples_per_second": 3.623,
|
| 1002 |
+
"eval_steps_per_second": 0.453,
|
| 1003 |
+
"eval_wer": 16.353436753209934,
|
| 1004 |
"step": 4000
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
+
"epoch": 1.1,
|
| 1008 |
+
"learning_rate": 4.583076923076924e-06,
|
| 1009 |
+
"loss": 0.1,
|
| 1010 |
"step": 4025
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
+
"epoch": 1.11,
|
| 1014 |
+
"learning_rate": 4.544615384615385e-06,
|
| 1015 |
+
"loss": 0.0925,
|
| 1016 |
"step": 4050
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
+
"epoch": 1.11,
|
| 1020 |
+
"learning_rate": 4.506153846153846e-06,
|
| 1021 |
+
"loss": 0.1064,
|
| 1022 |
"step": 4075
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
+
"epoch": 1.11,
|
| 1026 |
+
"learning_rate": 4.467692307692308e-06,
|
| 1027 |
+
"loss": 0.1127,
|
| 1028 |
"step": 4100
|
| 1029 |
},
|
| 1030 |
{
|
| 1031 |
+
"epoch": 1.12,
|
| 1032 |
+
"learning_rate": 4.42923076923077e-06,
|
| 1033 |
+
"loss": 0.1036,
|
| 1034 |
"step": 4125
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
+
"epoch": 1.12,
|
| 1038 |
+
"learning_rate": 4.392307692307693e-06,
|
| 1039 |
+
"loss": 0.091,
|
| 1040 |
"step": 4150
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
+
"epoch": 1.13,
|
| 1044 |
+
"learning_rate": 4.353846153846154e-06,
|
| 1045 |
+
"loss": 0.105,
|
| 1046 |
"step": 4175
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
+
"epoch": 1.13,
|
| 1050 |
+
"learning_rate": 4.315384615384616e-06,
|
| 1051 |
+
"loss": 0.1147,
|
| 1052 |
"step": 4200
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
+
"epoch": 1.13,
|
| 1056 |
+
"learning_rate": 4.276923076923078e-06,
|
| 1057 |
+
"loss": 0.0897,
|
| 1058 |
"step": 4225
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
+
"epoch": 1.14,
|
| 1062 |
+
"learning_rate": 4.238461538461539e-06,
|
| 1063 |
+
"loss": 0.1268,
|
| 1064 |
"step": 4250
|
| 1065 |
},
|
| 1066 |
{
|
| 1067 |
+
"epoch": 1.14,
|
| 1068 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 1069 |
+
"loss": 0.1129,
|
| 1070 |
"step": 4275
|
| 1071 |
},
|
| 1072 |
{
|
| 1073 |
+
"epoch": 1.14,
|
| 1074 |
+
"learning_rate": 4.161538461538462e-06,
|
| 1075 |
+
"loss": 0.1041,
|
| 1076 |
"step": 4300
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
+
"epoch": 1.15,
|
| 1080 |
+
"learning_rate": 4.123076923076924e-06,
|
| 1081 |
+
"loss": 0.1061,
|
| 1082 |
"step": 4325
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
+
"epoch": 1.15,
|
| 1086 |
+
"learning_rate": 4.084615384615385e-06,
|
| 1087 |
+
"loss": 0.1059,
|
| 1088 |
"step": 4350
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
+
"epoch": 1.15,
|
| 1092 |
+
"learning_rate": 4.0461538461538466e-06,
|
| 1093 |
+
"loss": 0.0818,
|
| 1094 |
"step": 4375
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
+
"epoch": 1.16,
|
| 1098 |
+
"learning_rate": 4.007692307692308e-06,
|
| 1099 |
+
"loss": 0.087,
|
| 1100 |
"step": 4400
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
+
"epoch": 1.16,
|
| 1104 |
+
"learning_rate": 3.96923076923077e-06,
|
| 1105 |
+
"loss": 0.0835,
|
| 1106 |
"step": 4425
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
+
"epoch": 1.16,
|
| 1110 |
+
"learning_rate": 3.930769230769231e-06,
|
| 1111 |
+
"loss": 0.0713,
|
| 1112 |
"step": 4450
|
| 1113 |
},
|
| 1114 |
{
|
| 1115 |
+
"epoch": 1.17,
|
| 1116 |
+
"learning_rate": 3.892307692307693e-06,
|
| 1117 |
+
"loss": 0.0792,
|
| 1118 |
"step": 4475
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
+
"epoch": 1.17,
|
| 1122 |
+
"learning_rate": 3.853846153846154e-06,
|
| 1123 |
+
"loss": 0.0799,
|
| 1124 |
"step": 4500
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
+
"epoch": 1.18,
|
| 1128 |
+
"learning_rate": 3.815384615384616e-06,
|
| 1129 |
+
"loss": 0.0755,
|
| 1130 |
"step": 4525
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
+
"epoch": 1.18,
|
| 1134 |
+
"learning_rate": 3.7769230769230775e-06,
|
| 1135 |
+
"loss": 0.0703,
|
| 1136 |
"step": 4550
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
+
"epoch": 1.18,
|
| 1140 |
+
"learning_rate": 3.7384615384615384e-06,
|
| 1141 |
+
"loss": 0.11,
|
| 1142 |
"step": 4575
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
+
"epoch": 1.19,
|
| 1146 |
+
"learning_rate": 3.7e-06,
|
| 1147 |
+
"loss": 0.0891,
|
| 1148 |
"step": 4600
|
| 1149 |
},
|
| 1150 |
{
|
| 1151 |
+
"epoch": 1.19,
|
| 1152 |
+
"learning_rate": 3.661538461538462e-06,
|
| 1153 |
+
"loss": 0.0757,
|
| 1154 |
"step": 4625
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
+
"epoch": 1.19,
|
| 1158 |
+
"learning_rate": 3.6230769230769236e-06,
|
| 1159 |
+
"loss": 0.0699,
|
| 1160 |
"step": 4650
|
| 1161 |
},
|
| 1162 |
{
|
| 1163 |
+
"epoch": 1.2,
|
| 1164 |
+
"learning_rate": 3.5846153846153845e-06,
|
| 1165 |
+
"loss": 0.0839,
|
| 1166 |
"step": 4675
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
+
"epoch": 1.2,
|
| 1170 |
+
"learning_rate": 3.5461538461538463e-06,
|
| 1171 |
+
"loss": 0.1029,
|
| 1172 |
"step": 4700
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
+
"epoch": 1.2,
|
| 1176 |
+
"learning_rate": 3.507692307692308e-06,
|
| 1177 |
+
"loss": 0.0806,
|
| 1178 |
"step": 4725
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
+
"epoch": 1.21,
|
| 1182 |
+
"learning_rate": 3.4692307692307698e-06,
|
| 1183 |
+
"loss": 0.0808,
|
| 1184 |
"step": 4750
|
| 1185 |
},
|
| 1186 |
{
|
| 1187 |
+
"epoch": 1.21,
|
| 1188 |
+
"learning_rate": 3.4307692307692307e-06,
|
| 1189 |
+
"loss": 0.0801,
|
| 1190 |
"step": 4775
|
| 1191 |
},
|
| 1192 |
{
|
| 1193 |
+
"epoch": 1.21,
|
| 1194 |
+
"learning_rate": 3.3923076923076924e-06,
|
| 1195 |
+
"loss": 0.0705,
|
| 1196 |
"step": 4800
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
+
"epoch": 1.22,
|
| 1200 |
+
"learning_rate": 3.353846153846154e-06,
|
| 1201 |
+
"loss": 0.0711,
|
| 1202 |
"step": 4825
|
| 1203 |
},
|
| 1204 |
{
|
| 1205 |
+
"epoch": 1.22,
|
| 1206 |
+
"learning_rate": 3.315384615384616e-06,
|
| 1207 |
+
"loss": 0.0721,
|
| 1208 |
"step": 4850
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
+
"epoch": 1.23,
|
| 1212 |
+
"learning_rate": 3.276923076923077e-06,
|
| 1213 |
+
"loss": 0.0806,
|
| 1214 |
"step": 4875
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
+
"epoch": 1.23,
|
| 1218 |
+
"learning_rate": 3.2384615384615385e-06,
|
| 1219 |
+
"loss": 0.0816,
|
| 1220 |
"step": 4900
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
+
"epoch": 1.23,
|
| 1224 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 1225 |
+
"loss": 0.078,
|
| 1226 |
"step": 4925
|
| 1227 |
},
|
| 1228 |
{
|
| 1229 |
+
"epoch": 1.24,
|
| 1230 |
+
"learning_rate": 3.161538461538462e-06,
|
| 1231 |
+
"loss": 0.0858,
|
| 1232 |
"step": 4950
|
| 1233 |
},
|
| 1234 |
{
|
| 1235 |
+
"epoch": 1.24,
|
| 1236 |
+
"learning_rate": 3.123076923076923e-06,
|
| 1237 |
+
"loss": 0.0948,
|
| 1238 |
"step": 4975
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
+
"epoch": 1.24,
|
| 1242 |
+
"learning_rate": 3.0846153846153847e-06,
|
| 1243 |
+
"loss": 0.0863,
|
| 1244 |
"step": 5000
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
+
"epoch": 1.24,
|
| 1248 |
+
"eval_loss": 0.23512449860572815,
|
| 1249 |
+
"eval_runtime": 1865.1318,
|
| 1250 |
+
"eval_samples_per_second": 3.534,
|
| 1251 |
+
"eval_steps_per_second": 0.442,
|
| 1252 |
+
"eval_wer": 14.787962250394912,
|
| 1253 |
"step": 5000
|
| 1254 |
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 1.25,
|
| 1257 |
+
"learning_rate": 3.0461538461538464e-06,
|
| 1258 |
+
"loss": 0.0899,
|
| 1259 |
+
"step": 5025
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.25,
|
| 1263 |
+
"learning_rate": 3.007692307692308e-06,
|
| 1264 |
+
"loss": 0.064,
|
| 1265 |
+
"step": 5050
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 1.25,
|
| 1269 |
+
"learning_rate": 2.969230769230769e-06,
|
| 1270 |
+
"loss": 0.0698,
|
| 1271 |
+
"step": 5075
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.26,
|
| 1275 |
+
"learning_rate": 2.930769230769231e-06,
|
| 1276 |
+
"loss": 0.0794,
|
| 1277 |
+
"step": 5100
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.26,
|
| 1281 |
+
"learning_rate": 2.8923076923076925e-06,
|
| 1282 |
+
"loss": 0.0758,
|
| 1283 |
+
"step": 5125
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.26,
|
| 1287 |
+
"learning_rate": 2.8538461538461543e-06,
|
| 1288 |
+
"loss": 0.0645,
|
| 1289 |
+
"step": 5150
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 1.27,
|
| 1293 |
+
"learning_rate": 2.815384615384615e-06,
|
| 1294 |
+
"loss": 0.071,
|
| 1295 |
+
"step": 5175
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 1.27,
|
| 1299 |
+
"learning_rate": 2.776923076923077e-06,
|
| 1300 |
+
"loss": 0.0555,
|
| 1301 |
+
"step": 5200
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.28,
|
| 1305 |
+
"learning_rate": 2.7384615384615387e-06,
|
| 1306 |
+
"loss": 0.0727,
|
| 1307 |
+
"step": 5225
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1.28,
|
| 1311 |
+
"learning_rate": 2.7000000000000004e-06,
|
| 1312 |
+
"loss": 0.0714,
|
| 1313 |
+
"step": 5250
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.28,
|
| 1317 |
+
"learning_rate": 2.6615384615384613e-06,
|
| 1318 |
+
"loss": 0.0631,
|
| 1319 |
+
"step": 5275
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.29,
|
| 1323 |
+
"learning_rate": 2.623076923076923e-06,
|
| 1324 |
+
"loss": 0.067,
|
| 1325 |
+
"step": 5300
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.29,
|
| 1329 |
+
"learning_rate": 2.584615384615385e-06,
|
| 1330 |
+
"loss": 0.0519,
|
| 1331 |
+
"step": 5325
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.29,
|
| 1335 |
+
"learning_rate": 2.5461538461538465e-06,
|
| 1336 |
+
"loss": 0.0629,
|
| 1337 |
+
"step": 5350
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 1.3,
|
| 1341 |
+
"learning_rate": 2.507692307692308e-06,
|
| 1342 |
+
"loss": 0.0787,
|
| 1343 |
+
"step": 5375
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.3,
|
| 1347 |
+
"learning_rate": 2.469230769230769e-06,
|
| 1348 |
+
"loss": 0.0524,
|
| 1349 |
+
"step": 5400
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.3,
|
| 1353 |
+
"learning_rate": 2.430769230769231e-06,
|
| 1354 |
+
"loss": 0.0618,
|
| 1355 |
+
"step": 5425
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.31,
|
| 1359 |
+
"learning_rate": 2.3923076923076922e-06,
|
| 1360 |
+
"loss": 0.0693,
|
| 1361 |
+
"step": 5450
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.31,
|
| 1365 |
+
"learning_rate": 2.353846153846154e-06,
|
| 1366 |
+
"loss": 0.0658,
|
| 1367 |
+
"step": 5475
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.31,
|
| 1371 |
+
"learning_rate": 2.3153846153846153e-06,
|
| 1372 |
+
"loss": 0.0764,
|
| 1373 |
+
"step": 5500
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1.32,
|
| 1377 |
+
"learning_rate": 2.276923076923077e-06,
|
| 1378 |
+
"loss": 0.0664,
|
| 1379 |
+
"step": 5525
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 1.32,
|
| 1383 |
+
"learning_rate": 2.2384615384615384e-06,
|
| 1384 |
+
"loss": 0.045,
|
| 1385 |
+
"step": 5550
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.33,
|
| 1389 |
+
"learning_rate": 2.2e-06,
|
| 1390 |
+
"loss": 0.065,
|
| 1391 |
+
"step": 5575
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.33,
|
| 1395 |
+
"learning_rate": 2.1615384615384614e-06,
|
| 1396 |
+
"loss": 0.0683,
|
| 1397 |
+
"step": 5600
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.33,
|
| 1401 |
+
"learning_rate": 2.123076923076923e-06,
|
| 1402 |
+
"loss": 0.0452,
|
| 1403 |
+
"step": 5625
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.34,
|
| 1407 |
+
"learning_rate": 2.0846153846153845e-06,
|
| 1408 |
+
"loss": 0.0574,
|
| 1409 |
+
"step": 5650
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.34,
|
| 1413 |
+
"learning_rate": 2.0461538461538462e-06,
|
| 1414 |
+
"loss": 0.0629,
|
| 1415 |
+
"step": 5675
|
| 1416 |
+
},
|
| 1417 |
{
|
| 1418 |
"epoch": 1.34,
|
| 1419 |
+
"learning_rate": 2.0076923076923076e-06,
|
| 1420 |
+
"loss": 0.0622,
|
| 1421 |
+
"step": 5700
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.35,
|
| 1425 |
+
"learning_rate": 1.9692307692307693e-06,
|
| 1426 |
+
"loss": 0.0717,
|
| 1427 |
+
"step": 5725
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.35,
|
| 1431 |
+
"learning_rate": 1.930769230769231e-06,
|
| 1432 |
+
"loss": 0.0792,
|
| 1433 |
+
"step": 5750
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.35,
|
| 1437 |
+
"learning_rate": 1.8923076923076924e-06,
|
| 1438 |
+
"loss": 0.078,
|
| 1439 |
+
"step": 5775
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.36,
|
| 1443 |
+
"learning_rate": 1.8538461538461541e-06,
|
| 1444 |
+
"loss": 0.0534,
|
| 1445 |
+
"step": 5800
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.36,
|
| 1449 |
+
"learning_rate": 1.8153846153846154e-06,
|
| 1450 |
+
"loss": 0.0583,
|
| 1451 |
+
"step": 5825
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.36,
|
| 1455 |
+
"learning_rate": 1.7769230769230772e-06,
|
| 1456 |
+
"loss": 0.0493,
|
| 1457 |
+
"step": 5850
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 1.37,
|
| 1461 |
+
"learning_rate": 1.7384615384615385e-06,
|
| 1462 |
+
"loss": 0.0566,
|
| 1463 |
+
"step": 5875
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.37,
|
| 1467 |
+
"learning_rate": 1.7000000000000002e-06,
|
| 1468 |
+
"loss": 0.0478,
|
| 1469 |
+
"step": 5900
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.38,
|
| 1473 |
+
"learning_rate": 1.6615384615384616e-06,
|
| 1474 |
+
"loss": 0.0477,
|
| 1475 |
+
"step": 5925
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.38,
|
| 1479 |
+
"learning_rate": 1.6230769230769233e-06,
|
| 1480 |
+
"loss": 0.0889,
|
| 1481 |
+
"step": 5950
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.38,
|
| 1485 |
+
"learning_rate": 1.5846153846153846e-06,
|
| 1486 |
+
"loss": 0.0553,
|
| 1487 |
+
"step": 5975
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.39,
|
| 1491 |
+
"learning_rate": 1.5461538461538464e-06,
|
| 1492 |
+
"loss": 0.0636,
|
| 1493 |
+
"step": 6000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.39,
|
| 1497 |
+
"eval_loss": 0.2250746488571167,
|
| 1498 |
+
"eval_runtime": 1815.1581,
|
| 1499 |
+
"eval_samples_per_second": 3.631,
|
| 1500 |
+
"eval_steps_per_second": 0.454,
|
| 1501 |
+
"eval_wer": 13.597148527684395,
|
| 1502 |
+
"step": 6000
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.39,
|
| 1506 |
+
"learning_rate": 1.5076923076923077e-06,
|
| 1507 |
+
"loss": 0.0399,
|
| 1508 |
+
"step": 6025
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.39,
|
| 1512 |
+
"learning_rate": 1.4692307692307694e-06,
|
| 1513 |
+
"loss": 0.0556,
|
| 1514 |
+
"step": 6050
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.4,
|
| 1518 |
+
"learning_rate": 1.4307692307692308e-06,
|
| 1519 |
+
"loss": 0.0568,
|
| 1520 |
+
"step": 6075
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 1.4,
|
| 1524 |
+
"learning_rate": 1.3923076923076925e-06,
|
| 1525 |
+
"loss": 0.0637,
|
| 1526 |
+
"step": 6100
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 1.4,
|
| 1530 |
+
"learning_rate": 1.3538461538461538e-06,
|
| 1531 |
+
"loss": 0.0391,
|
| 1532 |
+
"step": 6125
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.41,
|
| 1536 |
+
"learning_rate": 1.3153846153846156e-06,
|
| 1537 |
+
"loss": 0.0567,
|
| 1538 |
+
"step": 6150
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.41,
|
| 1542 |
+
"learning_rate": 1.2769230769230769e-06,
|
| 1543 |
+
"loss": 0.0456,
|
| 1544 |
+
"step": 6175
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.41,
|
| 1548 |
+
"learning_rate": 1.2384615384615386e-06,
|
| 1549 |
+
"loss": 0.0554,
|
| 1550 |
+
"step": 6200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.42,
|
| 1554 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 1555 |
+
"loss": 0.0638,
|
| 1556 |
+
"step": 6225
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.42,
|
| 1560 |
+
"learning_rate": 1.1615384615384617e-06,
|
| 1561 |
+
"loss": 0.0548,
|
| 1562 |
+
"step": 6250
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 1.43,
|
| 1566 |
+
"learning_rate": 1.1230769230769232e-06,
|
| 1567 |
+
"loss": 0.0488,
|
| 1568 |
+
"step": 6275
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 1.43,
|
| 1572 |
+
"learning_rate": 1.0846153846153848e-06,
|
| 1573 |
+
"loss": 0.0463,
|
| 1574 |
+
"step": 6300
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.43,
|
| 1578 |
+
"learning_rate": 1.0461538461538463e-06,
|
| 1579 |
+
"loss": 0.0616,
|
| 1580 |
+
"step": 6325
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.44,
|
| 1584 |
+
"learning_rate": 1.0076923076923078e-06,
|
| 1585 |
+
"loss": 0.0507,
|
| 1586 |
+
"step": 6350
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.44,
|
| 1590 |
+
"learning_rate": 9.692307692307693e-07,
|
| 1591 |
+
"loss": 0.0433,
|
| 1592 |
+
"step": 6375
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.44,
|
| 1596 |
+
"learning_rate": 9.323076923076923e-07,
|
| 1597 |
+
"loss": 0.0565,
|
| 1598 |
+
"step": 6400
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.45,
|
| 1602 |
+
"learning_rate": 8.938461538461539e-07,
|
| 1603 |
+
"loss": 0.0432,
|
| 1604 |
+
"step": 6425
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 1.45,
|
| 1608 |
+
"learning_rate": 8.553846153846154e-07,
|
| 1609 |
+
"loss": 0.0369,
|
| 1610 |
+
"step": 6450
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.45,
|
| 1614 |
+
"learning_rate": 8.169230769230769e-07,
|
| 1615 |
+
"loss": 0.0548,
|
| 1616 |
+
"step": 6475
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.46,
|
| 1620 |
+
"learning_rate": 7.784615384615385e-07,
|
| 1621 |
+
"loss": 0.0622,
|
| 1622 |
+
"step": 6500
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.46,
|
| 1626 |
+
"learning_rate": 7.4e-07,
|
| 1627 |
+
"loss": 0.0558,
|
| 1628 |
+
"step": 6525
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 1.46,
|
| 1632 |
+
"learning_rate": 7.030769230769231e-07,
|
| 1633 |
+
"loss": 0.065,
|
| 1634 |
+
"step": 6550
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.47,
|
| 1638 |
+
"learning_rate": 6.646153846153846e-07,
|
| 1639 |
+
"loss": 0.0607,
|
| 1640 |
+
"step": 6575
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.0,
|
| 1644 |
+
"learning_rate": 6.261538461538462e-07,
|
| 1645 |
+
"loss": 0.0577,
|
| 1646 |
+
"step": 6600
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 2.0,
|
| 1650 |
+
"learning_rate": 5.876923076923077e-07,
|
| 1651 |
+
"loss": 0.0713,
|
| 1652 |
+
"step": 6625
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 2.01,
|
| 1656 |
+
"learning_rate": 5.492307692307692e-07,
|
| 1657 |
+
"loss": 0.0697,
|
| 1658 |
+
"step": 6650
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 2.01,
|
| 1662 |
+
"learning_rate": 5.107692307692308e-07,
|
| 1663 |
+
"loss": 0.0452,
|
| 1664 |
+
"step": 6675
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 2.02,
|
| 1668 |
+
"learning_rate": 4.723076923076923e-07,
|
| 1669 |
+
"loss": 0.0579,
|
| 1670 |
+
"step": 6700
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 2.02,
|
| 1674 |
+
"learning_rate": 4.3384615384615384e-07,
|
| 1675 |
+
"loss": 0.05,
|
| 1676 |
+
"step": 6725
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.02,
|
| 1680 |
+
"learning_rate": 3.9538461538461537e-07,
|
| 1681 |
+
"loss": 0.0397,
|
| 1682 |
+
"step": 6750
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.03,
|
| 1686 |
+
"learning_rate": 3.569230769230769e-07,
|
| 1687 |
+
"loss": 0.0411,
|
| 1688 |
+
"step": 6775
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 2.03,
|
| 1692 |
+
"learning_rate": 3.184615384615385e-07,
|
| 1693 |
+
"loss": 0.0379,
|
| 1694 |
+
"step": 6800
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 2.03,
|
| 1698 |
+
"learning_rate": 2.8e-07,
|
| 1699 |
+
"loss": 0.04,
|
| 1700 |
+
"step": 6825
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 2.04,
|
| 1704 |
+
"learning_rate": 2.4153846153846155e-07,
|
| 1705 |
+
"loss": 0.0393,
|
| 1706 |
+
"step": 6850
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 2.04,
|
| 1710 |
+
"learning_rate": 2.0307692307692308e-07,
|
| 1711 |
+
"loss": 0.0412,
|
| 1712 |
+
"step": 6875
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 2.04,
|
| 1716 |
+
"learning_rate": 1.6461538461538462e-07,
|
| 1717 |
+
"loss": 0.0479,
|
| 1718 |
+
"step": 6900
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.05,
|
| 1722 |
+
"learning_rate": 1.2615384615384617e-07,
|
| 1723 |
+
"loss": 0.0387,
|
| 1724 |
+
"step": 6925
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.05,
|
| 1728 |
+
"learning_rate": 8.769230769230769e-08,
|
| 1729 |
+
"loss": 0.0453,
|
| 1730 |
+
"step": 6950
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 2.05,
|
| 1734 |
+
"learning_rate": 4.923076923076924e-08,
|
| 1735 |
+
"loss": 0.0255,
|
| 1736 |
+
"step": 6975
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 2.06,
|
| 1740 |
+
"learning_rate": 1.076923076923077e-08,
|
| 1741 |
+
"loss": 0.0271,
|
| 1742 |
+
"step": 7000
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 2.06,
|
| 1746 |
+
"eval_loss": 0.2200811207294464,
|
| 1747 |
+
"eval_runtime": 1807.6155,
|
| 1748 |
+
"eval_samples_per_second": 3.646,
|
| 1749 |
+
"eval_steps_per_second": 0.456,
|
| 1750 |
+
"eval_wer": 13.179958686054519,
|
| 1751 |
+
"step": 7000
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 2.06,
|
| 1755 |
+
"step": 7000,
|
| 1756 |
+
"total_flos": 2.857285499092992e+19,
|
| 1757 |
+
"train_loss": 0.208335678424154,
|
| 1758 |
+
"train_runtime": 23034.2525,
|
| 1759 |
+
"train_samples_per_second": 1.216,
|
| 1760 |
+
"train_steps_per_second": 0.304
|
| 1761 |
}
|
| 1762 |
],
|
| 1763 |
+
"max_steps": 7000,
|
| 1764 |
"num_train_epochs": 9223372036854775807,
|
| 1765 |
+
"total_flos": 2.857285499092992e+19,
|
| 1766 |
"trial_name": null,
|
| 1767 |
"trial_params": null
|
| 1768 |
}
|