Training in progress, step 250
Browse files- config.json +9 -9
- model.safetensors +2 -2
- runs/Oct24_20-21-51_0b6bd6f5a84e/events.out.tfevents.1729801364.0b6bd6f5a84e.1229.0 +3 -0
- tokenizer_config.json +215 -213
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "distil-whisper/distil-
|
| 3 |
"activation_dropout": 0.0,
|
| 4 |
"activation_function": "gelu",
|
| 5 |
"apply_spec_augment": false,
|
|
@@ -13,17 +13,17 @@
|
|
| 13 |
],
|
| 14 |
"bos_token_id": 50257,
|
| 15 |
"classifier_proj_size": 256,
|
| 16 |
-
"d_model":
|
| 17 |
-
"decoder_attention_heads":
|
| 18 |
-
"decoder_ffn_dim":
|
| 19 |
"decoder_layerdrop": 0.0,
|
| 20 |
-
"decoder_layers":
|
| 21 |
"decoder_start_token_id": 50257,
|
| 22 |
"dropout": 0.0,
|
| 23 |
-
"encoder_attention_heads":
|
| 24 |
-
"encoder_ffn_dim":
|
| 25 |
"encoder_layerdrop": 0.0,
|
| 26 |
-
"encoder_layers":
|
| 27 |
"eos_token_id": 50256,
|
| 28 |
"forced_decoder_ids": [
|
| 29 |
[
|
|
@@ -44,7 +44,7 @@
|
|
| 44 |
"max_target_positions": 448,
|
| 45 |
"median_filter_width": 7,
|
| 46 |
"model_type": "whisper",
|
| 47 |
-
"num_hidden_layers":
|
| 48 |
"num_mel_bins": 80,
|
| 49 |
"pad_token_id": 50256,
|
| 50 |
"scale_embedding": false,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "distil-whisper/distil-small.en",
|
| 3 |
"activation_dropout": 0.0,
|
| 4 |
"activation_function": "gelu",
|
| 5 |
"apply_spec_augment": false,
|
|
|
|
| 13 |
],
|
| 14 |
"bos_token_id": 50257,
|
| 15 |
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 768,
|
| 17 |
+
"decoder_attention_heads": 12,
|
| 18 |
+
"decoder_ffn_dim": 3072,
|
| 19 |
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 4,
|
| 21 |
"decoder_start_token_id": 50257,
|
| 22 |
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 12,
|
| 24 |
+
"encoder_ffn_dim": 3072,
|
| 25 |
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 12,
|
| 27 |
"eos_token_id": 50256,
|
| 28 |
"forced_decoder_ids": [
|
| 29 |
[
|
|
|
|
| 44 |
"max_target_positions": 448,
|
| 45 |
"median_filter_width": 7,
|
| 46 |
"model_type": "whisper",
|
| 47 |
+
"num_hidden_layers": 12,
|
| 48 |
"num_mel_bins": 80,
|
| 49 |
"pad_token_id": 50256,
|
| 50 |
"scale_embedding": false,
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33de7ac2c2faa470d2f84715c0c836954334fd12713a26d6ffe1de3d0af031a2
|
| 3 |
+
size 664561848
|
runs/Oct24_20-21-51_0b6bd6f5a84e/events.out.tfevents.1729801364.0b6bd6f5a84e.1229.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2248cb0f61998006096fdd3b35fea91feab233e37d4001859f928f602ee9426f
|
| 3 |
+
size 9106
|
tokenizer_config.json
CHANGED
|
@@ -12,849 +12,849 @@
|
|
| 12 |
},
|
| 13 |
"50257": {
|
| 14 |
"content": "<|startoftranscript|>",
|
| 15 |
-
"lstrip":
|
| 16 |
"normalized": false,
|
| 17 |
-
"rstrip":
|
| 18 |
"single_word": false,
|
| 19 |
"special": true
|
| 20 |
},
|
| 21 |
"50258": {
|
| 22 |
"content": "<|en|>",
|
| 23 |
-
"lstrip":
|
| 24 |
"normalized": false,
|
| 25 |
-
"rstrip":
|
| 26 |
"single_word": false,
|
| 27 |
"special": true
|
| 28 |
},
|
| 29 |
"50259": {
|
| 30 |
"content": "<|zh|>",
|
| 31 |
-
"lstrip":
|
| 32 |
"normalized": false,
|
| 33 |
-
"rstrip":
|
| 34 |
"single_word": false,
|
| 35 |
"special": true
|
| 36 |
},
|
| 37 |
"50260": {
|
| 38 |
"content": "<|de|>",
|
| 39 |
-
"lstrip":
|
| 40 |
"normalized": false,
|
| 41 |
-
"rstrip":
|
| 42 |
"single_word": false,
|
| 43 |
"special": true
|
| 44 |
},
|
| 45 |
"50261": {
|
| 46 |
"content": "<|es|>",
|
| 47 |
-
"lstrip":
|
| 48 |
"normalized": false,
|
| 49 |
-
"rstrip":
|
| 50 |
"single_word": false,
|
| 51 |
"special": true
|
| 52 |
},
|
| 53 |
"50262": {
|
| 54 |
"content": "<|ru|>",
|
| 55 |
-
"lstrip":
|
| 56 |
"normalized": false,
|
| 57 |
-
"rstrip":
|
| 58 |
"single_word": false,
|
| 59 |
"special": true
|
| 60 |
},
|
| 61 |
"50263": {
|
| 62 |
"content": "<|ko|>",
|
| 63 |
-
"lstrip":
|
| 64 |
"normalized": false,
|
| 65 |
-
"rstrip":
|
| 66 |
"single_word": false,
|
| 67 |
"special": true
|
| 68 |
},
|
| 69 |
"50264": {
|
| 70 |
"content": "<|fr|>",
|
| 71 |
-
"lstrip":
|
| 72 |
"normalized": false,
|
| 73 |
-
"rstrip":
|
| 74 |
"single_word": false,
|
| 75 |
"special": true
|
| 76 |
},
|
| 77 |
"50265": {
|
| 78 |
"content": "<|ja|>",
|
| 79 |
-
"lstrip":
|
| 80 |
"normalized": false,
|
| 81 |
-
"rstrip":
|
| 82 |
"single_word": false,
|
| 83 |
"special": true
|
| 84 |
},
|
| 85 |
"50266": {
|
| 86 |
"content": "<|pt|>",
|
| 87 |
-
"lstrip":
|
| 88 |
"normalized": false,
|
| 89 |
-
"rstrip":
|
| 90 |
"single_word": false,
|
| 91 |
"special": true
|
| 92 |
},
|
| 93 |
"50267": {
|
| 94 |
"content": "<|tr|>",
|
| 95 |
-
"lstrip":
|
| 96 |
"normalized": false,
|
| 97 |
-
"rstrip":
|
| 98 |
"single_word": false,
|
| 99 |
"special": true
|
| 100 |
},
|
| 101 |
"50268": {
|
| 102 |
"content": "<|pl|>",
|
| 103 |
-
"lstrip":
|
| 104 |
"normalized": false,
|
| 105 |
-
"rstrip":
|
| 106 |
"single_word": false,
|
| 107 |
"special": true
|
| 108 |
},
|
| 109 |
"50269": {
|
| 110 |
"content": "<|ca|>",
|
| 111 |
-
"lstrip":
|
| 112 |
"normalized": false,
|
| 113 |
-
"rstrip":
|
| 114 |
"single_word": false,
|
| 115 |
"special": true
|
| 116 |
},
|
| 117 |
"50270": {
|
| 118 |
"content": "<|nl|>",
|
| 119 |
-
"lstrip":
|
| 120 |
"normalized": false,
|
| 121 |
-
"rstrip":
|
| 122 |
"single_word": false,
|
| 123 |
"special": true
|
| 124 |
},
|
| 125 |
"50271": {
|
| 126 |
"content": "<|ar|>",
|
| 127 |
-
"lstrip":
|
| 128 |
"normalized": false,
|
| 129 |
-
"rstrip":
|
| 130 |
"single_word": false,
|
| 131 |
"special": true
|
| 132 |
},
|
| 133 |
"50272": {
|
| 134 |
"content": "<|sv|>",
|
| 135 |
-
"lstrip":
|
| 136 |
"normalized": false,
|
| 137 |
-
"rstrip":
|
| 138 |
"single_word": false,
|
| 139 |
"special": true
|
| 140 |
},
|
| 141 |
"50273": {
|
| 142 |
"content": "<|it|>",
|
| 143 |
-
"lstrip":
|
| 144 |
"normalized": false,
|
| 145 |
-
"rstrip":
|
| 146 |
"single_word": false,
|
| 147 |
"special": true
|
| 148 |
},
|
| 149 |
"50274": {
|
| 150 |
"content": "<|id|>",
|
| 151 |
-
"lstrip":
|
| 152 |
"normalized": false,
|
| 153 |
-
"rstrip":
|
| 154 |
"single_word": false,
|
| 155 |
"special": true
|
| 156 |
},
|
| 157 |
"50275": {
|
| 158 |
"content": "<|hi|>",
|
| 159 |
-
"lstrip":
|
| 160 |
"normalized": false,
|
| 161 |
-
"rstrip":
|
| 162 |
"single_word": false,
|
| 163 |
"special": true
|
| 164 |
},
|
| 165 |
"50276": {
|
| 166 |
"content": "<|fi|>",
|
| 167 |
-
"lstrip":
|
| 168 |
"normalized": false,
|
| 169 |
-
"rstrip":
|
| 170 |
"single_word": false,
|
| 171 |
"special": true
|
| 172 |
},
|
| 173 |
"50277": {
|
| 174 |
"content": "<|vi|>",
|
| 175 |
-
"lstrip":
|
| 176 |
"normalized": false,
|
| 177 |
-
"rstrip":
|
| 178 |
"single_word": false,
|
| 179 |
"special": true
|
| 180 |
},
|
| 181 |
"50278": {
|
| 182 |
"content": "<|iw|>",
|
| 183 |
-
"lstrip":
|
| 184 |
"normalized": false,
|
| 185 |
-
"rstrip":
|
| 186 |
"single_word": false,
|
| 187 |
"special": true
|
| 188 |
},
|
| 189 |
"50279": {
|
| 190 |
"content": "<|uk|>",
|
| 191 |
-
"lstrip":
|
| 192 |
"normalized": false,
|
| 193 |
-
"rstrip":
|
| 194 |
"single_word": false,
|
| 195 |
"special": true
|
| 196 |
},
|
| 197 |
"50280": {
|
| 198 |
"content": "<|el|>",
|
| 199 |
-
"lstrip":
|
| 200 |
"normalized": false,
|
| 201 |
-
"rstrip":
|
| 202 |
"single_word": false,
|
| 203 |
"special": true
|
| 204 |
},
|
| 205 |
"50281": {
|
| 206 |
"content": "<|ms|>",
|
| 207 |
-
"lstrip":
|
| 208 |
"normalized": false,
|
| 209 |
-
"rstrip":
|
| 210 |
"single_word": false,
|
| 211 |
"special": true
|
| 212 |
},
|
| 213 |
"50282": {
|
| 214 |
"content": "<|cs|>",
|
| 215 |
-
"lstrip":
|
| 216 |
"normalized": false,
|
| 217 |
-
"rstrip":
|
| 218 |
"single_word": false,
|
| 219 |
"special": true
|
| 220 |
},
|
| 221 |
"50283": {
|
| 222 |
"content": "<|ro|>",
|
| 223 |
-
"lstrip":
|
| 224 |
"normalized": false,
|
| 225 |
-
"rstrip":
|
| 226 |
"single_word": false,
|
| 227 |
"special": true
|
| 228 |
},
|
| 229 |
"50284": {
|
| 230 |
"content": "<|da|>",
|
| 231 |
-
"lstrip":
|
| 232 |
"normalized": false,
|
| 233 |
-
"rstrip":
|
| 234 |
"single_word": false,
|
| 235 |
"special": true
|
| 236 |
},
|
| 237 |
"50285": {
|
| 238 |
"content": "<|hu|>",
|
| 239 |
-
"lstrip":
|
| 240 |
"normalized": false,
|
| 241 |
-
"rstrip":
|
| 242 |
"single_word": false,
|
| 243 |
"special": true
|
| 244 |
},
|
| 245 |
"50286": {
|
| 246 |
"content": "<|ta|>",
|
| 247 |
-
"lstrip":
|
| 248 |
"normalized": false,
|
| 249 |
-
"rstrip":
|
| 250 |
"single_word": false,
|
| 251 |
"special": true
|
| 252 |
},
|
| 253 |
"50287": {
|
| 254 |
"content": "<|no|>",
|
| 255 |
-
"lstrip":
|
| 256 |
"normalized": false,
|
| 257 |
-
"rstrip":
|
| 258 |
"single_word": false,
|
| 259 |
"special": true
|
| 260 |
},
|
| 261 |
"50288": {
|
| 262 |
"content": "<|th|>",
|
| 263 |
-
"lstrip":
|
| 264 |
"normalized": false,
|
| 265 |
-
"rstrip":
|
| 266 |
"single_word": false,
|
| 267 |
"special": true
|
| 268 |
},
|
| 269 |
"50289": {
|
| 270 |
"content": "<|ur|>",
|
| 271 |
-
"lstrip":
|
| 272 |
"normalized": false,
|
| 273 |
-
"rstrip":
|
| 274 |
"single_word": false,
|
| 275 |
"special": true
|
| 276 |
},
|
| 277 |
"50290": {
|
| 278 |
"content": "<|hr|>",
|
| 279 |
-
"lstrip":
|
| 280 |
"normalized": false,
|
| 281 |
-
"rstrip":
|
| 282 |
"single_word": false,
|
| 283 |
"special": true
|
| 284 |
},
|
| 285 |
"50291": {
|
| 286 |
"content": "<|bg|>",
|
| 287 |
-
"lstrip":
|
| 288 |
"normalized": false,
|
| 289 |
-
"rstrip":
|
| 290 |
"single_word": false,
|
| 291 |
"special": true
|
| 292 |
},
|
| 293 |
"50292": {
|
| 294 |
"content": "<|lt|>",
|
| 295 |
-
"lstrip":
|
| 296 |
"normalized": false,
|
| 297 |
-
"rstrip":
|
| 298 |
"single_word": false,
|
| 299 |
"special": true
|
| 300 |
},
|
| 301 |
"50293": {
|
| 302 |
"content": "<|la|>",
|
| 303 |
-
"lstrip":
|
| 304 |
"normalized": false,
|
| 305 |
-
"rstrip":
|
| 306 |
"single_word": false,
|
| 307 |
"special": true
|
| 308 |
},
|
| 309 |
"50294": {
|
| 310 |
"content": "<|mi|>",
|
| 311 |
-
"lstrip":
|
| 312 |
"normalized": false,
|
| 313 |
-
"rstrip":
|
| 314 |
"single_word": false,
|
| 315 |
"special": true
|
| 316 |
},
|
| 317 |
"50295": {
|
| 318 |
"content": "<|ml|>",
|
| 319 |
-
"lstrip":
|
| 320 |
"normalized": false,
|
| 321 |
-
"rstrip":
|
| 322 |
"single_word": false,
|
| 323 |
"special": true
|
| 324 |
},
|
| 325 |
"50296": {
|
| 326 |
"content": "<|cy|>",
|
| 327 |
-
"lstrip":
|
| 328 |
"normalized": false,
|
| 329 |
-
"rstrip":
|
| 330 |
"single_word": false,
|
| 331 |
"special": true
|
| 332 |
},
|
| 333 |
"50297": {
|
| 334 |
"content": "<|sk|>",
|
| 335 |
-
"lstrip":
|
| 336 |
"normalized": false,
|
| 337 |
-
"rstrip":
|
| 338 |
"single_word": false,
|
| 339 |
"special": true
|
| 340 |
},
|
| 341 |
"50298": {
|
| 342 |
"content": "<|te|>",
|
| 343 |
-
"lstrip":
|
| 344 |
"normalized": false,
|
| 345 |
-
"rstrip":
|
| 346 |
"single_word": false,
|
| 347 |
"special": true
|
| 348 |
},
|
| 349 |
"50299": {
|
| 350 |
"content": "<|fa|>",
|
| 351 |
-
"lstrip":
|
| 352 |
"normalized": false,
|
| 353 |
-
"rstrip":
|
| 354 |
"single_word": false,
|
| 355 |
"special": true
|
| 356 |
},
|
| 357 |
"50300": {
|
| 358 |
"content": "<|lv|>",
|
| 359 |
-
"lstrip":
|
| 360 |
"normalized": false,
|
| 361 |
-
"rstrip":
|
| 362 |
"single_word": false,
|
| 363 |
"special": true
|
| 364 |
},
|
| 365 |
"50301": {
|
| 366 |
"content": "<|bn|>",
|
| 367 |
-
"lstrip":
|
| 368 |
"normalized": false,
|
| 369 |
-
"rstrip":
|
| 370 |
"single_word": false,
|
| 371 |
"special": true
|
| 372 |
},
|
| 373 |
"50302": {
|
| 374 |
"content": "<|sr|>",
|
| 375 |
-
"lstrip":
|
| 376 |
"normalized": false,
|
| 377 |
-
"rstrip":
|
| 378 |
"single_word": false,
|
| 379 |
"special": true
|
| 380 |
},
|
| 381 |
"50303": {
|
| 382 |
"content": "<|az|>",
|
| 383 |
-
"lstrip":
|
| 384 |
"normalized": false,
|
| 385 |
-
"rstrip":
|
| 386 |
"single_word": false,
|
| 387 |
"special": true
|
| 388 |
},
|
| 389 |
"50304": {
|
| 390 |
"content": "<|sl|>",
|
| 391 |
-
"lstrip":
|
| 392 |
"normalized": false,
|
| 393 |
-
"rstrip":
|
| 394 |
"single_word": false,
|
| 395 |
"special": true
|
| 396 |
},
|
| 397 |
"50305": {
|
| 398 |
"content": "<|kn|>",
|
| 399 |
-
"lstrip":
|
| 400 |
"normalized": false,
|
| 401 |
-
"rstrip":
|
| 402 |
"single_word": false,
|
| 403 |
"special": true
|
| 404 |
},
|
| 405 |
"50306": {
|
| 406 |
"content": "<|et|>",
|
| 407 |
-
"lstrip":
|
| 408 |
"normalized": false,
|
| 409 |
-
"rstrip":
|
| 410 |
"single_word": false,
|
| 411 |
"special": true
|
| 412 |
},
|
| 413 |
"50307": {
|
| 414 |
"content": "<|mk|>",
|
| 415 |
-
"lstrip":
|
| 416 |
"normalized": false,
|
| 417 |
-
"rstrip":
|
| 418 |
"single_word": false,
|
| 419 |
"special": true
|
| 420 |
},
|
| 421 |
"50308": {
|
| 422 |
"content": "<|br|>",
|
| 423 |
-
"lstrip":
|
| 424 |
"normalized": false,
|
| 425 |
-
"rstrip":
|
| 426 |
"single_word": false,
|
| 427 |
"special": true
|
| 428 |
},
|
| 429 |
"50309": {
|
| 430 |
"content": "<|eu|>",
|
| 431 |
-
"lstrip":
|
| 432 |
"normalized": false,
|
| 433 |
-
"rstrip":
|
| 434 |
"single_word": false,
|
| 435 |
"special": true
|
| 436 |
},
|
| 437 |
"50310": {
|
| 438 |
"content": "<|is|>",
|
| 439 |
-
"lstrip":
|
| 440 |
"normalized": false,
|
| 441 |
-
"rstrip":
|
| 442 |
"single_word": false,
|
| 443 |
"special": true
|
| 444 |
},
|
| 445 |
"50311": {
|
| 446 |
"content": "<|hy|>",
|
| 447 |
-
"lstrip":
|
| 448 |
"normalized": false,
|
| 449 |
-
"rstrip":
|
| 450 |
"single_word": false,
|
| 451 |
"special": true
|
| 452 |
},
|
| 453 |
"50312": {
|
| 454 |
"content": "<|ne|>",
|
| 455 |
-
"lstrip":
|
| 456 |
"normalized": false,
|
| 457 |
-
"rstrip":
|
| 458 |
"single_word": false,
|
| 459 |
"special": true
|
| 460 |
},
|
| 461 |
"50313": {
|
| 462 |
"content": "<|mn|>",
|
| 463 |
-
"lstrip":
|
| 464 |
"normalized": false,
|
| 465 |
-
"rstrip":
|
| 466 |
"single_word": false,
|
| 467 |
"special": true
|
| 468 |
},
|
| 469 |
"50314": {
|
| 470 |
"content": "<|bs|>",
|
| 471 |
-
"lstrip":
|
| 472 |
"normalized": false,
|
| 473 |
-
"rstrip":
|
| 474 |
"single_word": false,
|
| 475 |
"special": true
|
| 476 |
},
|
| 477 |
"50315": {
|
| 478 |
"content": "<|kk|>",
|
| 479 |
-
"lstrip":
|
| 480 |
"normalized": false,
|
| 481 |
-
"rstrip":
|
| 482 |
"single_word": false,
|
| 483 |
"special": true
|
| 484 |
},
|
| 485 |
"50316": {
|
| 486 |
"content": "<|sq|>",
|
| 487 |
-
"lstrip":
|
| 488 |
"normalized": false,
|
| 489 |
-
"rstrip":
|
| 490 |
"single_word": false,
|
| 491 |
"special": true
|
| 492 |
},
|
| 493 |
"50317": {
|
| 494 |
"content": "<|sw|>",
|
| 495 |
-
"lstrip":
|
| 496 |
"normalized": false,
|
| 497 |
-
"rstrip":
|
| 498 |
"single_word": false,
|
| 499 |
"special": true
|
| 500 |
},
|
| 501 |
"50318": {
|
| 502 |
"content": "<|gl|>",
|
| 503 |
-
"lstrip":
|
| 504 |
"normalized": false,
|
| 505 |
-
"rstrip":
|
| 506 |
"single_word": false,
|
| 507 |
"special": true
|
| 508 |
},
|
| 509 |
"50319": {
|
| 510 |
"content": "<|mr|>",
|
| 511 |
-
"lstrip":
|
| 512 |
"normalized": false,
|
| 513 |
-
"rstrip":
|
| 514 |
"single_word": false,
|
| 515 |
"special": true
|
| 516 |
},
|
| 517 |
"50320": {
|
| 518 |
"content": "<|pa|>",
|
| 519 |
-
"lstrip":
|
| 520 |
"normalized": false,
|
| 521 |
-
"rstrip":
|
| 522 |
"single_word": false,
|
| 523 |
"special": true
|
| 524 |
},
|
| 525 |
"50321": {
|
| 526 |
"content": "<|si|>",
|
| 527 |
-
"lstrip":
|
| 528 |
"normalized": false,
|
| 529 |
-
"rstrip":
|
| 530 |
"single_word": false,
|
| 531 |
"special": true
|
| 532 |
},
|
| 533 |
"50322": {
|
| 534 |
"content": "<|km|>",
|
| 535 |
-
"lstrip":
|
| 536 |
"normalized": false,
|
| 537 |
-
"rstrip":
|
| 538 |
"single_word": false,
|
| 539 |
"special": true
|
| 540 |
},
|
| 541 |
"50323": {
|
| 542 |
"content": "<|sn|>",
|
| 543 |
-
"lstrip":
|
| 544 |
"normalized": false,
|
| 545 |
-
"rstrip":
|
| 546 |
"single_word": false,
|
| 547 |
"special": true
|
| 548 |
},
|
| 549 |
"50324": {
|
| 550 |
"content": "<|yo|>",
|
| 551 |
-
"lstrip":
|
| 552 |
"normalized": false,
|
| 553 |
-
"rstrip":
|
| 554 |
"single_word": false,
|
| 555 |
"special": true
|
| 556 |
},
|
| 557 |
"50325": {
|
| 558 |
"content": "<|so|>",
|
| 559 |
-
"lstrip":
|
| 560 |
"normalized": false,
|
| 561 |
-
"rstrip":
|
| 562 |
"single_word": false,
|
| 563 |
"special": true
|
| 564 |
},
|
| 565 |
"50326": {
|
| 566 |
"content": "<|af|>",
|
| 567 |
-
"lstrip":
|
| 568 |
"normalized": false,
|
| 569 |
-
"rstrip":
|
| 570 |
"single_word": false,
|
| 571 |
"special": true
|
| 572 |
},
|
| 573 |
"50327": {
|
| 574 |
"content": "<|oc|>",
|
| 575 |
-
"lstrip":
|
| 576 |
"normalized": false,
|
| 577 |
-
"rstrip":
|
| 578 |
"single_word": false,
|
| 579 |
"special": true
|
| 580 |
},
|
| 581 |
"50328": {
|
| 582 |
"content": "<|ka|>",
|
| 583 |
-
"lstrip":
|
| 584 |
"normalized": false,
|
| 585 |
-
"rstrip":
|
| 586 |
"single_word": false,
|
| 587 |
"special": true
|
| 588 |
},
|
| 589 |
"50329": {
|
| 590 |
"content": "<|be|>",
|
| 591 |
-
"lstrip":
|
| 592 |
"normalized": false,
|
| 593 |
-
"rstrip":
|
| 594 |
"single_word": false,
|
| 595 |
"special": true
|
| 596 |
},
|
| 597 |
"50330": {
|
| 598 |
"content": "<|tg|>",
|
| 599 |
-
"lstrip":
|
| 600 |
"normalized": false,
|
| 601 |
-
"rstrip":
|
| 602 |
"single_word": false,
|
| 603 |
"special": true
|
| 604 |
},
|
| 605 |
"50331": {
|
| 606 |
"content": "<|sd|>",
|
| 607 |
-
"lstrip":
|
| 608 |
"normalized": false,
|
| 609 |
-
"rstrip":
|
| 610 |
"single_word": false,
|
| 611 |
"special": true
|
| 612 |
},
|
| 613 |
"50332": {
|
| 614 |
"content": "<|gu|>",
|
| 615 |
-
"lstrip":
|
| 616 |
"normalized": false,
|
| 617 |
-
"rstrip":
|
| 618 |
"single_word": false,
|
| 619 |
"special": true
|
| 620 |
},
|
| 621 |
"50333": {
|
| 622 |
"content": "<|am|>",
|
| 623 |
-
"lstrip":
|
| 624 |
"normalized": false,
|
| 625 |
-
"rstrip":
|
| 626 |
"single_word": false,
|
| 627 |
"special": true
|
| 628 |
},
|
| 629 |
"50334": {
|
| 630 |
"content": "<|yi|>",
|
| 631 |
-
"lstrip":
|
| 632 |
"normalized": false,
|
| 633 |
-
"rstrip":
|
| 634 |
"single_word": false,
|
| 635 |
"special": true
|
| 636 |
},
|
| 637 |
"50335": {
|
| 638 |
"content": "<|lo|>",
|
| 639 |
-
"lstrip":
|
| 640 |
"normalized": false,
|
| 641 |
-
"rstrip":
|
| 642 |
"single_word": false,
|
| 643 |
"special": true
|
| 644 |
},
|
| 645 |
"50336": {
|
| 646 |
"content": "<|uz|>",
|
| 647 |
-
"lstrip":
|
| 648 |
"normalized": false,
|
| 649 |
-
"rstrip":
|
| 650 |
"single_word": false,
|
| 651 |
"special": true
|
| 652 |
},
|
| 653 |
"50337": {
|
| 654 |
"content": "<|fo|>",
|
| 655 |
-
"lstrip":
|
| 656 |
"normalized": false,
|
| 657 |
-
"rstrip":
|
| 658 |
"single_word": false,
|
| 659 |
"special": true
|
| 660 |
},
|
| 661 |
"50338": {
|
| 662 |
"content": "<|ht|>",
|
| 663 |
-
"lstrip":
|
| 664 |
"normalized": false,
|
| 665 |
-
"rstrip":
|
| 666 |
"single_word": false,
|
| 667 |
"special": true
|
| 668 |
},
|
| 669 |
"50339": {
|
| 670 |
"content": "<|ps|>",
|
| 671 |
-
"lstrip":
|
| 672 |
"normalized": false,
|
| 673 |
-
"rstrip":
|
| 674 |
"single_word": false,
|
| 675 |
"special": true
|
| 676 |
},
|
| 677 |
"50340": {
|
| 678 |
"content": "<|tk|>",
|
| 679 |
-
"lstrip":
|
| 680 |
"normalized": false,
|
| 681 |
-
"rstrip":
|
| 682 |
"single_word": false,
|
| 683 |
"special": true
|
| 684 |
},
|
| 685 |
"50341": {
|
| 686 |
"content": "<|nn|>",
|
| 687 |
-
"lstrip":
|
| 688 |
"normalized": false,
|
| 689 |
-
"rstrip":
|
| 690 |
"single_word": false,
|
| 691 |
"special": true
|
| 692 |
},
|
| 693 |
"50342": {
|
| 694 |
"content": "<|mt|>",
|
| 695 |
-
"lstrip":
|
| 696 |
"normalized": false,
|
| 697 |
-
"rstrip":
|
| 698 |
"single_word": false,
|
| 699 |
"special": true
|
| 700 |
},
|
| 701 |
"50343": {
|
| 702 |
"content": "<|sa|>",
|
| 703 |
-
"lstrip":
|
| 704 |
"normalized": false,
|
| 705 |
-
"rstrip":
|
| 706 |
"single_word": false,
|
| 707 |
"special": true
|
| 708 |
},
|
| 709 |
"50344": {
|
| 710 |
"content": "<|lb|>",
|
| 711 |
-
"lstrip":
|
| 712 |
"normalized": false,
|
| 713 |
-
"rstrip":
|
| 714 |
"single_word": false,
|
| 715 |
"special": true
|
| 716 |
},
|
| 717 |
"50345": {
|
| 718 |
"content": "<|my|>",
|
| 719 |
-
"lstrip":
|
| 720 |
"normalized": false,
|
| 721 |
-
"rstrip":
|
| 722 |
"single_word": false,
|
| 723 |
"special": true
|
| 724 |
},
|
| 725 |
"50346": {
|
| 726 |
"content": "<|bo|>",
|
| 727 |
-
"lstrip":
|
| 728 |
"normalized": false,
|
| 729 |
-
"rstrip":
|
| 730 |
"single_word": false,
|
| 731 |
"special": true
|
| 732 |
},
|
| 733 |
"50347": {
|
| 734 |
"content": "<|tl|>",
|
| 735 |
-
"lstrip":
|
| 736 |
"normalized": false,
|
| 737 |
-
"rstrip":
|
| 738 |
"single_word": false,
|
| 739 |
"special": true
|
| 740 |
},
|
| 741 |
"50348": {
|
| 742 |
"content": "<|mg|>",
|
| 743 |
-
"lstrip":
|
| 744 |
"normalized": false,
|
| 745 |
-
"rstrip":
|
| 746 |
"single_word": false,
|
| 747 |
"special": true
|
| 748 |
},
|
| 749 |
"50349": {
|
| 750 |
"content": "<|as|>",
|
| 751 |
-
"lstrip":
|
| 752 |
"normalized": false,
|
| 753 |
-
"rstrip":
|
| 754 |
"single_word": false,
|
| 755 |
"special": true
|
| 756 |
},
|
| 757 |
"50350": {
|
| 758 |
"content": "<|tt|>",
|
| 759 |
-
"lstrip":
|
| 760 |
"normalized": false,
|
| 761 |
-
"rstrip":
|
| 762 |
"single_word": false,
|
| 763 |
"special": true
|
| 764 |
},
|
| 765 |
"50351": {
|
| 766 |
"content": "<|haw|>",
|
| 767 |
-
"lstrip":
|
| 768 |
"normalized": false,
|
| 769 |
-
"rstrip":
|
| 770 |
"single_word": false,
|
| 771 |
"special": true
|
| 772 |
},
|
| 773 |
"50352": {
|
| 774 |
"content": "<|ln|>",
|
| 775 |
-
"lstrip":
|
| 776 |
"normalized": false,
|
| 777 |
-
"rstrip":
|
| 778 |
"single_word": false,
|
| 779 |
"special": true
|
| 780 |
},
|
| 781 |
"50353": {
|
| 782 |
"content": "<|ha|>",
|
| 783 |
-
"lstrip":
|
| 784 |
"normalized": false,
|
| 785 |
-
"rstrip":
|
| 786 |
"single_word": false,
|
| 787 |
"special": true
|
| 788 |
},
|
| 789 |
"50354": {
|
| 790 |
"content": "<|ba|>",
|
| 791 |
-
"lstrip":
|
| 792 |
"normalized": false,
|
| 793 |
-
"rstrip":
|
| 794 |
"single_word": false,
|
| 795 |
"special": true
|
| 796 |
},
|
| 797 |
"50355": {
|
| 798 |
"content": "<|jw|>",
|
| 799 |
-
"lstrip":
|
| 800 |
"normalized": false,
|
| 801 |
-
"rstrip":
|
| 802 |
"single_word": false,
|
| 803 |
"special": true
|
| 804 |
},
|
| 805 |
"50356": {
|
| 806 |
"content": "<|su|>",
|
| 807 |
-
"lstrip":
|
| 808 |
"normalized": false,
|
| 809 |
-
"rstrip":
|
| 810 |
"single_word": false,
|
| 811 |
"special": true
|
| 812 |
},
|
| 813 |
"50357": {
|
| 814 |
"content": "<|translate|>",
|
| 815 |
-
"lstrip":
|
| 816 |
"normalized": false,
|
| 817 |
-
"rstrip":
|
| 818 |
"single_word": false,
|
| 819 |
"special": true
|
| 820 |
},
|
| 821 |
"50358": {
|
| 822 |
"content": "<|transcribe|>",
|
| 823 |
-
"lstrip":
|
| 824 |
"normalized": false,
|
| 825 |
-
"rstrip":
|
| 826 |
"single_word": false,
|
| 827 |
"special": true
|
| 828 |
},
|
| 829 |
"50359": {
|
| 830 |
"content": "<|startoflm|>",
|
| 831 |
-
"lstrip":
|
| 832 |
"normalized": false,
|
| 833 |
-
"rstrip":
|
| 834 |
"single_word": false,
|
| 835 |
"special": true
|
| 836 |
},
|
| 837 |
"50360": {
|
| 838 |
"content": "<|startofprev|>",
|
| 839 |
-
"lstrip":
|
| 840 |
"normalized": false,
|
| 841 |
-
"rstrip":
|
| 842 |
"single_word": false,
|
| 843 |
"special": true
|
| 844 |
},
|
| 845 |
"50361": {
|
| 846 |
"content": "<|nocaptions|>",
|
| 847 |
-
"lstrip":
|
| 848 |
"normalized": false,
|
| 849 |
-
"rstrip":
|
| 850 |
"single_word": false,
|
| 851 |
"special": true
|
| 852 |
},
|
| 853 |
"50362": {
|
| 854 |
"content": "<|notimestamps|>",
|
| 855 |
-
"lstrip":
|
| 856 |
"normalized": false,
|
| 857 |
-
"rstrip":
|
| 858 |
"single_word": false,
|
| 859 |
"special": true
|
| 860 |
},
|
|
@@ -12984,5 +12984,7 @@
|
|
| 12984 |
"processor_class": "WhisperProcessor",
|
| 12985 |
"return_attention_mask": false,
|
| 12986 |
"tokenizer_class": "WhisperTokenizer",
|
| 12987 |
-
"
|
|
|
|
|
|
|
| 12988 |
}
|
|
|
|
| 12 |
},
|
| 13 |
"50257": {
|
| 14 |
"content": "<|startoftranscript|>",
|
| 15 |
+
"lstrip": true,
|
| 16 |
"normalized": false,
|
| 17 |
+
"rstrip": true,
|
| 18 |
"single_word": false,
|
| 19 |
"special": true
|
| 20 |
},
|
| 21 |
"50258": {
|
| 22 |
"content": "<|en|>",
|
| 23 |
+
"lstrip": true,
|
| 24 |
"normalized": false,
|
| 25 |
+
"rstrip": true,
|
| 26 |
"single_word": false,
|
| 27 |
"special": true
|
| 28 |
},
|
| 29 |
"50259": {
|
| 30 |
"content": "<|zh|>",
|
| 31 |
+
"lstrip": true,
|
| 32 |
"normalized": false,
|
| 33 |
+
"rstrip": true,
|
| 34 |
"single_word": false,
|
| 35 |
"special": true
|
| 36 |
},
|
| 37 |
"50260": {
|
| 38 |
"content": "<|de|>",
|
| 39 |
+
"lstrip": true,
|
| 40 |
"normalized": false,
|
| 41 |
+
"rstrip": true,
|
| 42 |
"single_word": false,
|
| 43 |
"special": true
|
| 44 |
},
|
| 45 |
"50261": {
|
| 46 |
"content": "<|es|>",
|
| 47 |
+
"lstrip": true,
|
| 48 |
"normalized": false,
|
| 49 |
+
"rstrip": true,
|
| 50 |
"single_word": false,
|
| 51 |
"special": true
|
| 52 |
},
|
| 53 |
"50262": {
|
| 54 |
"content": "<|ru|>",
|
| 55 |
+
"lstrip": true,
|
| 56 |
"normalized": false,
|
| 57 |
+
"rstrip": true,
|
| 58 |
"single_word": false,
|
| 59 |
"special": true
|
| 60 |
},
|
| 61 |
"50263": {
|
| 62 |
"content": "<|ko|>",
|
| 63 |
+
"lstrip": true,
|
| 64 |
"normalized": false,
|
| 65 |
+
"rstrip": true,
|
| 66 |
"single_word": false,
|
| 67 |
"special": true
|
| 68 |
},
|
| 69 |
"50264": {
|
| 70 |
"content": "<|fr|>",
|
| 71 |
+
"lstrip": true,
|
| 72 |
"normalized": false,
|
| 73 |
+
"rstrip": true,
|
| 74 |
"single_word": false,
|
| 75 |
"special": true
|
| 76 |
},
|
| 77 |
"50265": {
|
| 78 |
"content": "<|ja|>",
|
| 79 |
+
"lstrip": true,
|
| 80 |
"normalized": false,
|
| 81 |
+
"rstrip": true,
|
| 82 |
"single_word": false,
|
| 83 |
"special": true
|
| 84 |
},
|
| 85 |
"50266": {
|
| 86 |
"content": "<|pt|>",
|
| 87 |
+
"lstrip": true,
|
| 88 |
"normalized": false,
|
| 89 |
+
"rstrip": true,
|
| 90 |
"single_word": false,
|
| 91 |
"special": true
|
| 92 |
},
|
| 93 |
"50267": {
|
| 94 |
"content": "<|tr|>",
|
| 95 |
+
"lstrip": true,
|
| 96 |
"normalized": false,
|
| 97 |
+
"rstrip": true,
|
| 98 |
"single_word": false,
|
| 99 |
"special": true
|
| 100 |
},
|
| 101 |
"50268": {
|
| 102 |
"content": "<|pl|>",
|
| 103 |
+
"lstrip": true,
|
| 104 |
"normalized": false,
|
| 105 |
+
"rstrip": true,
|
| 106 |
"single_word": false,
|
| 107 |
"special": true
|
| 108 |
},
|
| 109 |
"50269": {
|
| 110 |
"content": "<|ca|>",
|
| 111 |
+
"lstrip": true,
|
| 112 |
"normalized": false,
|
| 113 |
+
"rstrip": true,
|
| 114 |
"single_word": false,
|
| 115 |
"special": true
|
| 116 |
},
|
| 117 |
"50270": {
|
| 118 |
"content": "<|nl|>",
|
| 119 |
+
"lstrip": true,
|
| 120 |
"normalized": false,
|
| 121 |
+
"rstrip": true,
|
| 122 |
"single_word": false,
|
| 123 |
"special": true
|
| 124 |
},
|
| 125 |
"50271": {
|
| 126 |
"content": "<|ar|>",
|
| 127 |
+
"lstrip": true,
|
| 128 |
"normalized": false,
|
| 129 |
+
"rstrip": true,
|
| 130 |
"single_word": false,
|
| 131 |
"special": true
|
| 132 |
},
|
| 133 |
"50272": {
|
| 134 |
"content": "<|sv|>",
|
| 135 |
+
"lstrip": true,
|
| 136 |
"normalized": false,
|
| 137 |
+
"rstrip": true,
|
| 138 |
"single_word": false,
|
| 139 |
"special": true
|
| 140 |
},
|
| 141 |
"50273": {
|
| 142 |
"content": "<|it|>",
|
| 143 |
+
"lstrip": true,
|
| 144 |
"normalized": false,
|
| 145 |
+
"rstrip": true,
|
| 146 |
"single_word": false,
|
| 147 |
"special": true
|
| 148 |
},
|
| 149 |
"50274": {
|
| 150 |
"content": "<|id|>",
|
| 151 |
+
"lstrip": true,
|
| 152 |
"normalized": false,
|
| 153 |
+
"rstrip": true,
|
| 154 |
"single_word": false,
|
| 155 |
"special": true
|
| 156 |
},
|
| 157 |
"50275": {
|
| 158 |
"content": "<|hi|>",
|
| 159 |
+
"lstrip": true,
|
| 160 |
"normalized": false,
|
| 161 |
+
"rstrip": true,
|
| 162 |
"single_word": false,
|
| 163 |
"special": true
|
| 164 |
},
|
| 165 |
"50276": {
|
| 166 |
"content": "<|fi|>",
|
| 167 |
+
"lstrip": true,
|
| 168 |
"normalized": false,
|
| 169 |
+
"rstrip": true,
|
| 170 |
"single_word": false,
|
| 171 |
"special": true
|
| 172 |
},
|
| 173 |
"50277": {
|
| 174 |
"content": "<|vi|>",
|
| 175 |
+
"lstrip": true,
|
| 176 |
"normalized": false,
|
| 177 |
+
"rstrip": true,
|
| 178 |
"single_word": false,
|
| 179 |
"special": true
|
| 180 |
},
|
| 181 |
"50278": {
|
| 182 |
"content": "<|iw|>",
|
| 183 |
+
"lstrip": true,
|
| 184 |
"normalized": false,
|
| 185 |
+
"rstrip": true,
|
| 186 |
"single_word": false,
|
| 187 |
"special": true
|
| 188 |
},
|
| 189 |
"50279": {
|
| 190 |
"content": "<|uk|>",
|
| 191 |
+
"lstrip": true,
|
| 192 |
"normalized": false,
|
| 193 |
+
"rstrip": true,
|
| 194 |
"single_word": false,
|
| 195 |
"special": true
|
| 196 |
},
|
| 197 |
"50280": {
|
| 198 |
"content": "<|el|>",
|
| 199 |
+
"lstrip": true,
|
| 200 |
"normalized": false,
|
| 201 |
+
"rstrip": true,
|
| 202 |
"single_word": false,
|
| 203 |
"special": true
|
| 204 |
},
|
| 205 |
"50281": {
|
| 206 |
"content": "<|ms|>",
|
| 207 |
+
"lstrip": true,
|
| 208 |
"normalized": false,
|
| 209 |
+
"rstrip": true,
|
| 210 |
"single_word": false,
|
| 211 |
"special": true
|
| 212 |
},
|
| 213 |
"50282": {
|
| 214 |
"content": "<|cs|>",
|
| 215 |
+
"lstrip": true,
|
| 216 |
"normalized": false,
|
| 217 |
+
"rstrip": true,
|
| 218 |
"single_word": false,
|
| 219 |
"special": true
|
| 220 |
},
|
| 221 |
"50283": {
|
| 222 |
"content": "<|ro|>",
|
| 223 |
+
"lstrip": true,
|
| 224 |
"normalized": false,
|
| 225 |
+
"rstrip": true,
|
| 226 |
"single_word": false,
|
| 227 |
"special": true
|
| 228 |
},
|
| 229 |
"50284": {
|
| 230 |
"content": "<|da|>",
|
| 231 |
+
"lstrip": true,
|
| 232 |
"normalized": false,
|
| 233 |
+
"rstrip": true,
|
| 234 |
"single_word": false,
|
| 235 |
"special": true
|
| 236 |
},
|
| 237 |
"50285": {
|
| 238 |
"content": "<|hu|>",
|
| 239 |
+
"lstrip": true,
|
| 240 |
"normalized": false,
|
| 241 |
+
"rstrip": true,
|
| 242 |
"single_word": false,
|
| 243 |
"special": true
|
| 244 |
},
|
| 245 |
"50286": {
|
| 246 |
"content": "<|ta|>",
|
| 247 |
+
"lstrip": true,
|
| 248 |
"normalized": false,
|
| 249 |
+
"rstrip": true,
|
| 250 |
"single_word": false,
|
| 251 |
"special": true
|
| 252 |
},
|
| 253 |
"50287": {
|
| 254 |
"content": "<|no|>",
|
| 255 |
+
"lstrip": true,
|
| 256 |
"normalized": false,
|
| 257 |
+
"rstrip": true,
|
| 258 |
"single_word": false,
|
| 259 |
"special": true
|
| 260 |
},
|
| 261 |
"50288": {
|
| 262 |
"content": "<|th|>",
|
| 263 |
+
"lstrip": true,
|
| 264 |
"normalized": false,
|
| 265 |
+
"rstrip": true,
|
| 266 |
"single_word": false,
|
| 267 |
"special": true
|
| 268 |
},
|
| 269 |
"50289": {
|
| 270 |
"content": "<|ur|>",
|
| 271 |
+
"lstrip": true,
|
| 272 |
"normalized": false,
|
| 273 |
+
"rstrip": true,
|
| 274 |
"single_word": false,
|
| 275 |
"special": true
|
| 276 |
},
|
| 277 |
"50290": {
|
| 278 |
"content": "<|hr|>",
|
| 279 |
+
"lstrip": true,
|
| 280 |
"normalized": false,
|
| 281 |
+
"rstrip": true,
|
| 282 |
"single_word": false,
|
| 283 |
"special": true
|
| 284 |
},
|
| 285 |
"50291": {
|
| 286 |
"content": "<|bg|>",
|
| 287 |
+
"lstrip": true,
|
| 288 |
"normalized": false,
|
| 289 |
+
"rstrip": true,
|
| 290 |
"single_word": false,
|
| 291 |
"special": true
|
| 292 |
},
|
| 293 |
"50292": {
|
| 294 |
"content": "<|lt|>",
|
| 295 |
+
"lstrip": true,
|
| 296 |
"normalized": false,
|
| 297 |
+
"rstrip": true,
|
| 298 |
"single_word": false,
|
| 299 |
"special": true
|
| 300 |
},
|
| 301 |
"50293": {
|
| 302 |
"content": "<|la|>",
|
| 303 |
+
"lstrip": true,
|
| 304 |
"normalized": false,
|
| 305 |
+
"rstrip": true,
|
| 306 |
"single_word": false,
|
| 307 |
"special": true
|
| 308 |
},
|
| 309 |
"50294": {
|
| 310 |
"content": "<|mi|>",
|
| 311 |
+
"lstrip": true,
|
| 312 |
"normalized": false,
|
| 313 |
+
"rstrip": true,
|
| 314 |
"single_word": false,
|
| 315 |
"special": true
|
| 316 |
},
|
| 317 |
"50295": {
|
| 318 |
"content": "<|ml|>",
|
| 319 |
+
"lstrip": true,
|
| 320 |
"normalized": false,
|
| 321 |
+
"rstrip": true,
|
| 322 |
"single_word": false,
|
| 323 |
"special": true
|
| 324 |
},
|
| 325 |
"50296": {
|
| 326 |
"content": "<|cy|>",
|
| 327 |
+
"lstrip": true,
|
| 328 |
"normalized": false,
|
| 329 |
+
"rstrip": true,
|
| 330 |
"single_word": false,
|
| 331 |
"special": true
|
| 332 |
},
|
| 333 |
"50297": {
|
| 334 |
"content": "<|sk|>",
|
| 335 |
+
"lstrip": true,
|
| 336 |
"normalized": false,
|
| 337 |
+
"rstrip": true,
|
| 338 |
"single_word": false,
|
| 339 |
"special": true
|
| 340 |
},
|
| 341 |
"50298": {
|
| 342 |
"content": "<|te|>",
|
| 343 |
+
"lstrip": true,
|
| 344 |
"normalized": false,
|
| 345 |
+
"rstrip": true,
|
| 346 |
"single_word": false,
|
| 347 |
"special": true
|
| 348 |
},
|
| 349 |
"50299": {
|
| 350 |
"content": "<|fa|>",
|
| 351 |
+
"lstrip": true,
|
| 352 |
"normalized": false,
|
| 353 |
+
"rstrip": true,
|
| 354 |
"single_word": false,
|
| 355 |
"special": true
|
| 356 |
},
|
| 357 |
"50300": {
|
| 358 |
"content": "<|lv|>",
|
| 359 |
+
"lstrip": true,
|
| 360 |
"normalized": false,
|
| 361 |
+
"rstrip": true,
|
| 362 |
"single_word": false,
|
| 363 |
"special": true
|
| 364 |
},
|
| 365 |
"50301": {
|
| 366 |
"content": "<|bn|>",
|
| 367 |
+
"lstrip": true,
|
| 368 |
"normalized": false,
|
| 369 |
+
"rstrip": true,
|
| 370 |
"single_word": false,
|
| 371 |
"special": true
|
| 372 |
},
|
| 373 |
"50302": {
|
| 374 |
"content": "<|sr|>",
|
| 375 |
+
"lstrip": true,
|
| 376 |
"normalized": false,
|
| 377 |
+
"rstrip": true,
|
| 378 |
"single_word": false,
|
| 379 |
"special": true
|
| 380 |
},
|
| 381 |
"50303": {
|
| 382 |
"content": "<|az|>",
|
| 383 |
+
"lstrip": true,
|
| 384 |
"normalized": false,
|
| 385 |
+
"rstrip": true,
|
| 386 |
"single_word": false,
|
| 387 |
"special": true
|
| 388 |
},
|
| 389 |
"50304": {
|
| 390 |
"content": "<|sl|>",
|
| 391 |
+
"lstrip": true,
|
| 392 |
"normalized": false,
|
| 393 |
+
"rstrip": true,
|
| 394 |
"single_word": false,
|
| 395 |
"special": true
|
| 396 |
},
|
| 397 |
"50305": {
|
| 398 |
"content": "<|kn|>",
|
| 399 |
+
"lstrip": true,
|
| 400 |
"normalized": false,
|
| 401 |
+
"rstrip": true,
|
| 402 |
"single_word": false,
|
| 403 |
"special": true
|
| 404 |
},
|
| 405 |
"50306": {
|
| 406 |
"content": "<|et|>",
|
| 407 |
+
"lstrip": true,
|
| 408 |
"normalized": false,
|
| 409 |
+
"rstrip": true,
|
| 410 |
"single_word": false,
|
| 411 |
"special": true
|
| 412 |
},
|
| 413 |
"50307": {
|
| 414 |
"content": "<|mk|>",
|
| 415 |
+
"lstrip": true,
|
| 416 |
"normalized": false,
|
| 417 |
+
"rstrip": true,
|
| 418 |
"single_word": false,
|
| 419 |
"special": true
|
| 420 |
},
|
| 421 |
"50308": {
|
| 422 |
"content": "<|br|>",
|
| 423 |
+
"lstrip": true,
|
| 424 |
"normalized": false,
|
| 425 |
+
"rstrip": true,
|
| 426 |
"single_word": false,
|
| 427 |
"special": true
|
| 428 |
},
|
| 429 |
"50309": {
|
| 430 |
"content": "<|eu|>",
|
| 431 |
+
"lstrip": true,
|
| 432 |
"normalized": false,
|
| 433 |
+
"rstrip": true,
|
| 434 |
"single_word": false,
|
| 435 |
"special": true
|
| 436 |
},
|
| 437 |
"50310": {
|
| 438 |
"content": "<|is|>",
|
| 439 |
+
"lstrip": true,
|
| 440 |
"normalized": false,
|
| 441 |
+
"rstrip": true,
|
| 442 |
"single_word": false,
|
| 443 |
"special": true
|
| 444 |
},
|
| 445 |
"50311": {
|
| 446 |
"content": "<|hy|>",
|
| 447 |
+
"lstrip": true,
|
| 448 |
"normalized": false,
|
| 449 |
+
"rstrip": true,
|
| 450 |
"single_word": false,
|
| 451 |
"special": true
|
| 452 |
},
|
| 453 |
"50312": {
|
| 454 |
"content": "<|ne|>",
|
| 455 |
+
"lstrip": true,
|
| 456 |
"normalized": false,
|
| 457 |
+
"rstrip": true,
|
| 458 |
"single_word": false,
|
| 459 |
"special": true
|
| 460 |
},
|
| 461 |
"50313": {
|
| 462 |
"content": "<|mn|>",
|
| 463 |
+
"lstrip": true,
|
| 464 |
"normalized": false,
|
| 465 |
+
"rstrip": true,
|
| 466 |
"single_word": false,
|
| 467 |
"special": true
|
| 468 |
},
|
| 469 |
"50314": {
|
| 470 |
"content": "<|bs|>",
|
| 471 |
+
"lstrip": true,
|
| 472 |
"normalized": false,
|
| 473 |
+
"rstrip": true,
|
| 474 |
"single_word": false,
|
| 475 |
"special": true
|
| 476 |
},
|
| 477 |
"50315": {
|
| 478 |
"content": "<|kk|>",
|
| 479 |
+
"lstrip": true,
|
| 480 |
"normalized": false,
|
| 481 |
+
"rstrip": true,
|
| 482 |
"single_word": false,
|
| 483 |
"special": true
|
| 484 |
},
|
| 485 |
"50316": {
|
| 486 |
"content": "<|sq|>",
|
| 487 |
+
"lstrip": true,
|
| 488 |
"normalized": false,
|
| 489 |
+
"rstrip": true,
|
| 490 |
"single_word": false,
|
| 491 |
"special": true
|
| 492 |
},
|
| 493 |
"50317": {
|
| 494 |
"content": "<|sw|>",
|
| 495 |
+
"lstrip": true,
|
| 496 |
"normalized": false,
|
| 497 |
+
"rstrip": true,
|
| 498 |
"single_word": false,
|
| 499 |
"special": true
|
| 500 |
},
|
| 501 |
"50318": {
|
| 502 |
"content": "<|gl|>",
|
| 503 |
+
"lstrip": true,
|
| 504 |
"normalized": false,
|
| 505 |
+
"rstrip": true,
|
| 506 |
"single_word": false,
|
| 507 |
"special": true
|
| 508 |
},
|
| 509 |
"50319": {
|
| 510 |
"content": "<|mr|>",
|
| 511 |
+
"lstrip": true,
|
| 512 |
"normalized": false,
|
| 513 |
+
"rstrip": true,
|
| 514 |
"single_word": false,
|
| 515 |
"special": true
|
| 516 |
},
|
| 517 |
"50320": {
|
| 518 |
"content": "<|pa|>",
|
| 519 |
+
"lstrip": true,
|
| 520 |
"normalized": false,
|
| 521 |
+
"rstrip": true,
|
| 522 |
"single_word": false,
|
| 523 |
"special": true
|
| 524 |
},
|
| 525 |
"50321": {
|
| 526 |
"content": "<|si|>",
|
| 527 |
+
"lstrip": true,
|
| 528 |
"normalized": false,
|
| 529 |
+
"rstrip": true,
|
| 530 |
"single_word": false,
|
| 531 |
"special": true
|
| 532 |
},
|
| 533 |
"50322": {
|
| 534 |
"content": "<|km|>",
|
| 535 |
+
"lstrip": true,
|
| 536 |
"normalized": false,
|
| 537 |
+
"rstrip": true,
|
| 538 |
"single_word": false,
|
| 539 |
"special": true
|
| 540 |
},
|
| 541 |
"50323": {
|
| 542 |
"content": "<|sn|>",
|
| 543 |
+
"lstrip": true,
|
| 544 |
"normalized": false,
|
| 545 |
+
"rstrip": true,
|
| 546 |
"single_word": false,
|
| 547 |
"special": true
|
| 548 |
},
|
| 549 |
"50324": {
|
| 550 |
"content": "<|yo|>",
|
| 551 |
+
"lstrip": true,
|
| 552 |
"normalized": false,
|
| 553 |
+
"rstrip": true,
|
| 554 |
"single_word": false,
|
| 555 |
"special": true
|
| 556 |
},
|
| 557 |
"50325": {
|
| 558 |
"content": "<|so|>",
|
| 559 |
+
"lstrip": true,
|
| 560 |
"normalized": false,
|
| 561 |
+
"rstrip": true,
|
| 562 |
"single_word": false,
|
| 563 |
"special": true
|
| 564 |
},
|
| 565 |
"50326": {
|
| 566 |
"content": "<|af|>",
|
| 567 |
+
"lstrip": true,
|
| 568 |
"normalized": false,
|
| 569 |
+
"rstrip": true,
|
| 570 |
"single_word": false,
|
| 571 |
"special": true
|
| 572 |
},
|
| 573 |
"50327": {
|
| 574 |
"content": "<|oc|>",
|
| 575 |
+
"lstrip": true,
|
| 576 |
"normalized": false,
|
| 577 |
+
"rstrip": true,
|
| 578 |
"single_word": false,
|
| 579 |
"special": true
|
| 580 |
},
|
| 581 |
"50328": {
|
| 582 |
"content": "<|ka|>",
|
| 583 |
+
"lstrip": true,
|
| 584 |
"normalized": false,
|
| 585 |
+
"rstrip": true,
|
| 586 |
"single_word": false,
|
| 587 |
"special": true
|
| 588 |
},
|
| 589 |
"50329": {
|
| 590 |
"content": "<|be|>",
|
| 591 |
+
"lstrip": true,
|
| 592 |
"normalized": false,
|
| 593 |
+
"rstrip": true,
|
| 594 |
"single_word": false,
|
| 595 |
"special": true
|
| 596 |
},
|
| 597 |
"50330": {
|
| 598 |
"content": "<|tg|>",
|
| 599 |
+
"lstrip": true,
|
| 600 |
"normalized": false,
|
| 601 |
+
"rstrip": true,
|
| 602 |
"single_word": false,
|
| 603 |
"special": true
|
| 604 |
},
|
| 605 |
"50331": {
|
| 606 |
"content": "<|sd|>",
|
| 607 |
+
"lstrip": true,
|
| 608 |
"normalized": false,
|
| 609 |
+
"rstrip": true,
|
| 610 |
"single_word": false,
|
| 611 |
"special": true
|
| 612 |
},
|
| 613 |
"50332": {
|
| 614 |
"content": "<|gu|>",
|
| 615 |
+
"lstrip": true,
|
| 616 |
"normalized": false,
|
| 617 |
+
"rstrip": true,
|
| 618 |
"single_word": false,
|
| 619 |
"special": true
|
| 620 |
},
|
| 621 |
"50333": {
|
| 622 |
"content": "<|am|>",
|
| 623 |
+
"lstrip": true,
|
| 624 |
"normalized": false,
|
| 625 |
+
"rstrip": true,
|
| 626 |
"single_word": false,
|
| 627 |
"special": true
|
| 628 |
},
|
| 629 |
"50334": {
|
| 630 |
"content": "<|yi|>",
|
| 631 |
+
"lstrip": true,
|
| 632 |
"normalized": false,
|
| 633 |
+
"rstrip": true,
|
| 634 |
"single_word": false,
|
| 635 |
"special": true
|
| 636 |
},
|
| 637 |
"50335": {
|
| 638 |
"content": "<|lo|>",
|
| 639 |
+
"lstrip": true,
|
| 640 |
"normalized": false,
|
| 641 |
+
"rstrip": true,
|
| 642 |
"single_word": false,
|
| 643 |
"special": true
|
| 644 |
},
|
| 645 |
"50336": {
|
| 646 |
"content": "<|uz|>",
|
| 647 |
+
"lstrip": true,
|
| 648 |
"normalized": false,
|
| 649 |
+
"rstrip": true,
|
| 650 |
"single_word": false,
|
| 651 |
"special": true
|
| 652 |
},
|
| 653 |
"50337": {
|
| 654 |
"content": "<|fo|>",
|
| 655 |
+
"lstrip": true,
|
| 656 |
"normalized": false,
|
| 657 |
+
"rstrip": true,
|
| 658 |
"single_word": false,
|
| 659 |
"special": true
|
| 660 |
},
|
| 661 |
"50338": {
|
| 662 |
"content": "<|ht|>",
|
| 663 |
+
"lstrip": true,
|
| 664 |
"normalized": false,
|
| 665 |
+
"rstrip": true,
|
| 666 |
"single_word": false,
|
| 667 |
"special": true
|
| 668 |
},
|
| 669 |
"50339": {
|
| 670 |
"content": "<|ps|>",
|
| 671 |
+
"lstrip": true,
|
| 672 |
"normalized": false,
|
| 673 |
+
"rstrip": true,
|
| 674 |
"single_word": false,
|
| 675 |
"special": true
|
| 676 |
},
|
| 677 |
"50340": {
|
| 678 |
"content": "<|tk|>",
|
| 679 |
+
"lstrip": true,
|
| 680 |
"normalized": false,
|
| 681 |
+
"rstrip": true,
|
| 682 |
"single_word": false,
|
| 683 |
"special": true
|
| 684 |
},
|
| 685 |
"50341": {
|
| 686 |
"content": "<|nn|>",
|
| 687 |
+
"lstrip": true,
|
| 688 |
"normalized": false,
|
| 689 |
+
"rstrip": true,
|
| 690 |
"single_word": false,
|
| 691 |
"special": true
|
| 692 |
},
|
| 693 |
"50342": {
|
| 694 |
"content": "<|mt|>",
|
| 695 |
+
"lstrip": true,
|
| 696 |
"normalized": false,
|
| 697 |
+
"rstrip": true,
|
| 698 |
"single_word": false,
|
| 699 |
"special": true
|
| 700 |
},
|
| 701 |
"50343": {
|
| 702 |
"content": "<|sa|>",
|
| 703 |
+
"lstrip": true,
|
| 704 |
"normalized": false,
|
| 705 |
+
"rstrip": true,
|
| 706 |
"single_word": false,
|
| 707 |
"special": true
|
| 708 |
},
|
| 709 |
"50344": {
|
| 710 |
"content": "<|lb|>",
|
| 711 |
+
"lstrip": true,
|
| 712 |
"normalized": false,
|
| 713 |
+
"rstrip": true,
|
| 714 |
"single_word": false,
|
| 715 |
"special": true
|
| 716 |
},
|
| 717 |
"50345": {
|
| 718 |
"content": "<|my|>",
|
| 719 |
+
"lstrip": true,
|
| 720 |
"normalized": false,
|
| 721 |
+
"rstrip": true,
|
| 722 |
"single_word": false,
|
| 723 |
"special": true
|
| 724 |
},
|
| 725 |
"50346": {
|
| 726 |
"content": "<|bo|>",
|
| 727 |
+
"lstrip": true,
|
| 728 |
"normalized": false,
|
| 729 |
+
"rstrip": true,
|
| 730 |
"single_word": false,
|
| 731 |
"special": true
|
| 732 |
},
|
| 733 |
"50347": {
|
| 734 |
"content": "<|tl|>",
|
| 735 |
+
"lstrip": true,
|
| 736 |
"normalized": false,
|
| 737 |
+
"rstrip": true,
|
| 738 |
"single_word": false,
|
| 739 |
"special": true
|
| 740 |
},
|
| 741 |
"50348": {
|
| 742 |
"content": "<|mg|>",
|
| 743 |
+
"lstrip": true,
|
| 744 |
"normalized": false,
|
| 745 |
+
"rstrip": true,
|
| 746 |
"single_word": false,
|
| 747 |
"special": true
|
| 748 |
},
|
| 749 |
"50349": {
|
| 750 |
"content": "<|as|>",
|
| 751 |
+
"lstrip": true,
|
| 752 |
"normalized": false,
|
| 753 |
+
"rstrip": true,
|
| 754 |
"single_word": false,
|
| 755 |
"special": true
|
| 756 |
},
|
| 757 |
"50350": {
|
| 758 |
"content": "<|tt|>",
|
| 759 |
+
"lstrip": true,
|
| 760 |
"normalized": false,
|
| 761 |
+
"rstrip": true,
|
| 762 |
"single_word": false,
|
| 763 |
"special": true
|
| 764 |
},
|
| 765 |
"50351": {
|
| 766 |
"content": "<|haw|>",
|
| 767 |
+
"lstrip": true,
|
| 768 |
"normalized": false,
|
| 769 |
+
"rstrip": true,
|
| 770 |
"single_word": false,
|
| 771 |
"special": true
|
| 772 |
},
|
| 773 |
"50352": {
|
| 774 |
"content": "<|ln|>",
|
| 775 |
+
"lstrip": true,
|
| 776 |
"normalized": false,
|
| 777 |
+
"rstrip": true,
|
| 778 |
"single_word": false,
|
| 779 |
"special": true
|
| 780 |
},
|
| 781 |
"50353": {
|
| 782 |
"content": "<|ha|>",
|
| 783 |
+
"lstrip": true,
|
| 784 |
"normalized": false,
|
| 785 |
+
"rstrip": true,
|
| 786 |
"single_word": false,
|
| 787 |
"special": true
|
| 788 |
},
|
| 789 |
"50354": {
|
| 790 |
"content": "<|ba|>",
|
| 791 |
+
"lstrip": true,
|
| 792 |
"normalized": false,
|
| 793 |
+
"rstrip": true,
|
| 794 |
"single_word": false,
|
| 795 |
"special": true
|
| 796 |
},
|
| 797 |
"50355": {
|
| 798 |
"content": "<|jw|>",
|
| 799 |
+
"lstrip": true,
|
| 800 |
"normalized": false,
|
| 801 |
+
"rstrip": true,
|
| 802 |
"single_word": false,
|
| 803 |
"special": true
|
| 804 |
},
|
| 805 |
"50356": {
|
| 806 |
"content": "<|su|>",
|
| 807 |
+
"lstrip": true,
|
| 808 |
"normalized": false,
|
| 809 |
+
"rstrip": true,
|
| 810 |
"single_word": false,
|
| 811 |
"special": true
|
| 812 |
},
|
| 813 |
"50357": {
|
| 814 |
"content": "<|translate|>",
|
| 815 |
+
"lstrip": true,
|
| 816 |
"normalized": false,
|
| 817 |
+
"rstrip": true,
|
| 818 |
"single_word": false,
|
| 819 |
"special": true
|
| 820 |
},
|
| 821 |
"50358": {
|
| 822 |
"content": "<|transcribe|>",
|
| 823 |
+
"lstrip": true,
|
| 824 |
"normalized": false,
|
| 825 |
+
"rstrip": true,
|
| 826 |
"single_word": false,
|
| 827 |
"special": true
|
| 828 |
},
|
| 829 |
"50359": {
|
| 830 |
"content": "<|startoflm|>",
|
| 831 |
+
"lstrip": true,
|
| 832 |
"normalized": false,
|
| 833 |
+
"rstrip": true,
|
| 834 |
"single_word": false,
|
| 835 |
"special": true
|
| 836 |
},
|
| 837 |
"50360": {
|
| 838 |
"content": "<|startofprev|>",
|
| 839 |
+
"lstrip": true,
|
| 840 |
"normalized": false,
|
| 841 |
+
"rstrip": true,
|
| 842 |
"single_word": false,
|
| 843 |
"special": true
|
| 844 |
},
|
| 845 |
"50361": {
|
| 846 |
"content": "<|nocaptions|>",
|
| 847 |
+
"lstrip": true,
|
| 848 |
"normalized": false,
|
| 849 |
+
"rstrip": true,
|
| 850 |
"single_word": false,
|
| 851 |
"special": true
|
| 852 |
},
|
| 853 |
"50362": {
|
| 854 |
"content": "<|notimestamps|>",
|
| 855 |
+
"lstrip": true,
|
| 856 |
"normalized": false,
|
| 857 |
+
"rstrip": true,
|
| 858 |
"single_word": false,
|
| 859 |
"special": true
|
| 860 |
},
|
|
|
|
| 12984 |
"processor_class": "WhisperProcessor",
|
| 12985 |
"return_attention_mask": false,
|
| 12986 |
"tokenizer_class": "WhisperTokenizer",
|
| 12987 |
+
"trust_remote_code": false,
|
| 12988 |
+
"unk_token": "<|endoftext|>",
|
| 12989 |
+
"use_fast": true
|
| 12990 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5304
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:89348f52306010c9939628e9f2b8015be02a830a896a5bb8e2b3c342cca6f734
|
| 3 |
size 5304
|