Upload folder using huggingface_hub
Browse files- README.md +1 -1
- adapter_config.json +2 -3
- adapter_model.safetensors +2 -2
- tokenizer.json +2 -968
- tokenizer_config.json +204 -445
README.md
CHANGED
|
@@ -199,4 +199,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
-
- PEFT 0.
|
|
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
+
- PEFT 0.12.0
|
adapter_config.json
CHANGED
|
@@ -3,8 +3,6 @@
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
| 6 |
-
"eva_config": null,
|
| 7 |
-
"exclude_modules": null,
|
| 8 |
"fan_in_fan_out": false,
|
| 9 |
"inference_mode": true,
|
| 10 |
"init_lora_weights": true,
|
|
@@ -13,7 +11,6 @@
|
|
| 13 |
"layers_to_transform": null,
|
| 14 |
"loftq_config": {},
|
| 15 |
"lora_alpha": 16,
|
| 16 |
-
"lora_bias": false,
|
| 17 |
"lora_dropout": 0.1,
|
| 18 |
"megatron_config": null,
|
| 19 |
"megatron_core": "megatron.core",
|
|
@@ -23,7 +20,9 @@
|
|
| 23 |
"rank_pattern": {},
|
| 24 |
"revision": null,
|
| 25 |
"target_modules": [
|
|
|
|
| 26 |
"v",
|
|
|
|
| 27 |
"q"
|
| 28 |
],
|
| 29 |
"task_type": "SEQ_2_SEQ_LM",
|
|
|
|
| 3 |
"auto_mapping": null,
|
| 4 |
"base_model_name_or_path": "Salesforce/codet5-base",
|
| 5 |
"bias": "none",
|
|
|
|
|
|
|
| 6 |
"fan_in_fan_out": false,
|
| 7 |
"inference_mode": true,
|
| 8 |
"init_lora_weights": true,
|
|
|
|
| 11 |
"layers_to_transform": null,
|
| 12 |
"loftq_config": {},
|
| 13 |
"lora_alpha": 16,
|
|
|
|
| 14 |
"lora_dropout": 0.1,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"k",
|
| 24 |
"v",
|
| 25 |
+
"o",
|
| 26 |
"q"
|
| 27 |
],
|
| 28 |
"task_type": "SEQ_2_SEQ_LM",
|
adapter_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:b91ed7832527de7906346574ac86b022b807fe1feacc57442209aa3f8a78187f
|
| 3 |
+
size 7117824
|
tokenizer.json
CHANGED
|
@@ -1,21 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
|
| 5 |
-
"max_length": 128,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
-
"padding": {
|
| 10 |
-
"strategy": {
|
| 11 |
-
"Fixed": 128
|
| 12 |
-
},
|
| 13 |
-
"direction": "Right",
|
| 14 |
-
"pad_to_multiple_of": null,
|
| 15 |
-
"pad_id": 0,
|
| 16 |
-
"pad_type_id": 0,
|
| 17 |
-
"pad_token": "<pad>"
|
| 18 |
-
},
|
| 19 |
"added_tokens": [
|
| 20 |
{
|
| 21 |
"id": 0,
|
|
@@ -61,906 +47,6 @@
|
|
| 61 |
"rstrip": false,
|
| 62 |
"normalized": true,
|
| 63 |
"special": true
|
| 64 |
-
},
|
| 65 |
-
{
|
| 66 |
-
"id": 32000,
|
| 67 |
-
"content": "<extra_id_99>",
|
| 68 |
-
"single_word": false,
|
| 69 |
-
"lstrip": true,
|
| 70 |
-
"rstrip": false,
|
| 71 |
-
"normalized": true,
|
| 72 |
-
"special": true
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"id": 32001,
|
| 76 |
-
"content": "<extra_id_98>",
|
| 77 |
-
"single_word": false,
|
| 78 |
-
"lstrip": true,
|
| 79 |
-
"rstrip": false,
|
| 80 |
-
"normalized": true,
|
| 81 |
-
"special": true
|
| 82 |
-
},
|
| 83 |
-
{
|
| 84 |
-
"id": 32002,
|
| 85 |
-
"content": "<extra_id_97>",
|
| 86 |
-
"single_word": false,
|
| 87 |
-
"lstrip": true,
|
| 88 |
-
"rstrip": false,
|
| 89 |
-
"normalized": true,
|
| 90 |
-
"special": true
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"id": 32003,
|
| 94 |
-
"content": "<extra_id_96>",
|
| 95 |
-
"single_word": false,
|
| 96 |
-
"lstrip": true,
|
| 97 |
-
"rstrip": false,
|
| 98 |
-
"normalized": true,
|
| 99 |
-
"special": true
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"id": 32004,
|
| 103 |
-
"content": "<extra_id_95>",
|
| 104 |
-
"single_word": false,
|
| 105 |
-
"lstrip": true,
|
| 106 |
-
"rstrip": false,
|
| 107 |
-
"normalized": true,
|
| 108 |
-
"special": true
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"id": 32005,
|
| 112 |
-
"content": "<extra_id_94>",
|
| 113 |
-
"single_word": false,
|
| 114 |
-
"lstrip": true,
|
| 115 |
-
"rstrip": false,
|
| 116 |
-
"normalized": true,
|
| 117 |
-
"special": true
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"id": 32006,
|
| 121 |
-
"content": "<extra_id_93>",
|
| 122 |
-
"single_word": false,
|
| 123 |
-
"lstrip": true,
|
| 124 |
-
"rstrip": false,
|
| 125 |
-
"normalized": true,
|
| 126 |
-
"special": true
|
| 127 |
-
},
|
| 128 |
-
{
|
| 129 |
-
"id": 32007,
|
| 130 |
-
"content": "<extra_id_92>",
|
| 131 |
-
"single_word": false,
|
| 132 |
-
"lstrip": true,
|
| 133 |
-
"rstrip": false,
|
| 134 |
-
"normalized": true,
|
| 135 |
-
"special": true
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"id": 32008,
|
| 139 |
-
"content": "<extra_id_91>",
|
| 140 |
-
"single_word": false,
|
| 141 |
-
"lstrip": true,
|
| 142 |
-
"rstrip": false,
|
| 143 |
-
"normalized": true,
|
| 144 |
-
"special": true
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"id": 32009,
|
| 148 |
-
"content": "<extra_id_90>",
|
| 149 |
-
"single_word": false,
|
| 150 |
-
"lstrip": true,
|
| 151 |
-
"rstrip": false,
|
| 152 |
-
"normalized": true,
|
| 153 |
-
"special": true
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"id": 32010,
|
| 157 |
-
"content": "<extra_id_89>",
|
| 158 |
-
"single_word": false,
|
| 159 |
-
"lstrip": true,
|
| 160 |
-
"rstrip": false,
|
| 161 |
-
"normalized": true,
|
| 162 |
-
"special": true
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"id": 32011,
|
| 166 |
-
"content": "<extra_id_88>",
|
| 167 |
-
"single_word": false,
|
| 168 |
-
"lstrip": true,
|
| 169 |
-
"rstrip": false,
|
| 170 |
-
"normalized": true,
|
| 171 |
-
"special": true
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"id": 32012,
|
| 175 |
-
"content": "<extra_id_87>",
|
| 176 |
-
"single_word": false,
|
| 177 |
-
"lstrip": true,
|
| 178 |
-
"rstrip": false,
|
| 179 |
-
"normalized": true,
|
| 180 |
-
"special": true
|
| 181 |
-
},
|
| 182 |
-
{
|
| 183 |
-
"id": 32013,
|
| 184 |
-
"content": "<extra_id_86>",
|
| 185 |
-
"single_word": false,
|
| 186 |
-
"lstrip": true,
|
| 187 |
-
"rstrip": false,
|
| 188 |
-
"normalized": true,
|
| 189 |
-
"special": true
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"id": 32014,
|
| 193 |
-
"content": "<extra_id_85>",
|
| 194 |
-
"single_word": false,
|
| 195 |
-
"lstrip": true,
|
| 196 |
-
"rstrip": false,
|
| 197 |
-
"normalized": true,
|
| 198 |
-
"special": true
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"id": 32015,
|
| 202 |
-
"content": "<extra_id_84>",
|
| 203 |
-
"single_word": false,
|
| 204 |
-
"lstrip": true,
|
| 205 |
-
"rstrip": false,
|
| 206 |
-
"normalized": true,
|
| 207 |
-
"special": true
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"id": 32016,
|
| 211 |
-
"content": "<extra_id_83>",
|
| 212 |
-
"single_word": false,
|
| 213 |
-
"lstrip": true,
|
| 214 |
-
"rstrip": false,
|
| 215 |
-
"normalized": true,
|
| 216 |
-
"special": true
|
| 217 |
-
},
|
| 218 |
-
{
|
| 219 |
-
"id": 32017,
|
| 220 |
-
"content": "<extra_id_82>",
|
| 221 |
-
"single_word": false,
|
| 222 |
-
"lstrip": true,
|
| 223 |
-
"rstrip": false,
|
| 224 |
-
"normalized": true,
|
| 225 |
-
"special": true
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"id": 32018,
|
| 229 |
-
"content": "<extra_id_81>",
|
| 230 |
-
"single_word": false,
|
| 231 |
-
"lstrip": true,
|
| 232 |
-
"rstrip": false,
|
| 233 |
-
"normalized": true,
|
| 234 |
-
"special": true
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"id": 32019,
|
| 238 |
-
"content": "<extra_id_80>",
|
| 239 |
-
"single_word": false,
|
| 240 |
-
"lstrip": true,
|
| 241 |
-
"rstrip": false,
|
| 242 |
-
"normalized": true,
|
| 243 |
-
"special": true
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"id": 32020,
|
| 247 |
-
"content": "<extra_id_79>",
|
| 248 |
-
"single_word": false,
|
| 249 |
-
"lstrip": true,
|
| 250 |
-
"rstrip": false,
|
| 251 |
-
"normalized": true,
|
| 252 |
-
"special": true
|
| 253 |
-
},
|
| 254 |
-
{
|
| 255 |
-
"id": 32021,
|
| 256 |
-
"content": "<extra_id_78>",
|
| 257 |
-
"single_word": false,
|
| 258 |
-
"lstrip": true,
|
| 259 |
-
"rstrip": false,
|
| 260 |
-
"normalized": true,
|
| 261 |
-
"special": true
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"id": 32022,
|
| 265 |
-
"content": "<extra_id_77>",
|
| 266 |
-
"single_word": false,
|
| 267 |
-
"lstrip": true,
|
| 268 |
-
"rstrip": false,
|
| 269 |
-
"normalized": true,
|
| 270 |
-
"special": true
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"id": 32023,
|
| 274 |
-
"content": "<extra_id_76>",
|
| 275 |
-
"single_word": false,
|
| 276 |
-
"lstrip": true,
|
| 277 |
-
"rstrip": false,
|
| 278 |
-
"normalized": true,
|
| 279 |
-
"special": true
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"id": 32024,
|
| 283 |
-
"content": "<extra_id_75>",
|
| 284 |
-
"single_word": false,
|
| 285 |
-
"lstrip": true,
|
| 286 |
-
"rstrip": false,
|
| 287 |
-
"normalized": true,
|
| 288 |
-
"special": true
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"id": 32025,
|
| 292 |
-
"content": "<extra_id_74>",
|
| 293 |
-
"single_word": false,
|
| 294 |
-
"lstrip": true,
|
| 295 |
-
"rstrip": false,
|
| 296 |
-
"normalized": true,
|
| 297 |
-
"special": true
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"id": 32026,
|
| 301 |
-
"content": "<extra_id_73>",
|
| 302 |
-
"single_word": false,
|
| 303 |
-
"lstrip": true,
|
| 304 |
-
"rstrip": false,
|
| 305 |
-
"normalized": true,
|
| 306 |
-
"special": true
|
| 307 |
-
},
|
| 308 |
-
{
|
| 309 |
-
"id": 32027,
|
| 310 |
-
"content": "<extra_id_72>",
|
| 311 |
-
"single_word": false,
|
| 312 |
-
"lstrip": true,
|
| 313 |
-
"rstrip": false,
|
| 314 |
-
"normalized": true,
|
| 315 |
-
"special": true
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"id": 32028,
|
| 319 |
-
"content": "<extra_id_71>",
|
| 320 |
-
"single_word": false,
|
| 321 |
-
"lstrip": true,
|
| 322 |
-
"rstrip": false,
|
| 323 |
-
"normalized": true,
|
| 324 |
-
"special": true
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"id": 32029,
|
| 328 |
-
"content": "<extra_id_70>",
|
| 329 |
-
"single_word": false,
|
| 330 |
-
"lstrip": true,
|
| 331 |
-
"rstrip": false,
|
| 332 |
-
"normalized": true,
|
| 333 |
-
"special": true
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"id": 32030,
|
| 337 |
-
"content": "<extra_id_69>",
|
| 338 |
-
"single_word": false,
|
| 339 |
-
"lstrip": true,
|
| 340 |
-
"rstrip": false,
|
| 341 |
-
"normalized": true,
|
| 342 |
-
"special": true
|
| 343 |
-
},
|
| 344 |
-
{
|
| 345 |
-
"id": 32031,
|
| 346 |
-
"content": "<extra_id_68>",
|
| 347 |
-
"single_word": false,
|
| 348 |
-
"lstrip": true,
|
| 349 |
-
"rstrip": false,
|
| 350 |
-
"normalized": true,
|
| 351 |
-
"special": true
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"id": 32032,
|
| 355 |
-
"content": "<extra_id_67>",
|
| 356 |
-
"single_word": false,
|
| 357 |
-
"lstrip": true,
|
| 358 |
-
"rstrip": false,
|
| 359 |
-
"normalized": true,
|
| 360 |
-
"special": true
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"id": 32033,
|
| 364 |
-
"content": "<extra_id_66>",
|
| 365 |
-
"single_word": false,
|
| 366 |
-
"lstrip": true,
|
| 367 |
-
"rstrip": false,
|
| 368 |
-
"normalized": true,
|
| 369 |
-
"special": true
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"id": 32034,
|
| 373 |
-
"content": "<extra_id_65>",
|
| 374 |
-
"single_word": false,
|
| 375 |
-
"lstrip": true,
|
| 376 |
-
"rstrip": false,
|
| 377 |
-
"normalized": true,
|
| 378 |
-
"special": true
|
| 379 |
-
},
|
| 380 |
-
{
|
| 381 |
-
"id": 32035,
|
| 382 |
-
"content": "<extra_id_64>",
|
| 383 |
-
"single_word": false,
|
| 384 |
-
"lstrip": true,
|
| 385 |
-
"rstrip": false,
|
| 386 |
-
"normalized": true,
|
| 387 |
-
"special": true
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"id": 32036,
|
| 391 |
-
"content": "<extra_id_63>",
|
| 392 |
-
"single_word": false,
|
| 393 |
-
"lstrip": true,
|
| 394 |
-
"rstrip": false,
|
| 395 |
-
"normalized": true,
|
| 396 |
-
"special": true
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"id": 32037,
|
| 400 |
-
"content": "<extra_id_62>",
|
| 401 |
-
"single_word": false,
|
| 402 |
-
"lstrip": true,
|
| 403 |
-
"rstrip": false,
|
| 404 |
-
"normalized": true,
|
| 405 |
-
"special": true
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"id": 32038,
|
| 409 |
-
"content": "<extra_id_61>",
|
| 410 |
-
"single_word": false,
|
| 411 |
-
"lstrip": true,
|
| 412 |
-
"rstrip": false,
|
| 413 |
-
"normalized": true,
|
| 414 |
-
"special": true
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"id": 32039,
|
| 418 |
-
"content": "<extra_id_60>",
|
| 419 |
-
"single_word": false,
|
| 420 |
-
"lstrip": true,
|
| 421 |
-
"rstrip": false,
|
| 422 |
-
"normalized": true,
|
| 423 |
-
"special": true
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"id": 32040,
|
| 427 |
-
"content": "<extra_id_59>",
|
| 428 |
-
"single_word": false,
|
| 429 |
-
"lstrip": true,
|
| 430 |
-
"rstrip": false,
|
| 431 |
-
"normalized": true,
|
| 432 |
-
"special": true
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"id": 32041,
|
| 436 |
-
"content": "<extra_id_58>",
|
| 437 |
-
"single_word": false,
|
| 438 |
-
"lstrip": true,
|
| 439 |
-
"rstrip": false,
|
| 440 |
-
"normalized": true,
|
| 441 |
-
"special": true
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"id": 32042,
|
| 445 |
-
"content": "<extra_id_57>",
|
| 446 |
-
"single_word": false,
|
| 447 |
-
"lstrip": true,
|
| 448 |
-
"rstrip": false,
|
| 449 |
-
"normalized": true,
|
| 450 |
-
"special": true
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"id": 32043,
|
| 454 |
-
"content": "<extra_id_56>",
|
| 455 |
-
"single_word": false,
|
| 456 |
-
"lstrip": true,
|
| 457 |
-
"rstrip": false,
|
| 458 |
-
"normalized": true,
|
| 459 |
-
"special": true
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"id": 32044,
|
| 463 |
-
"content": "<extra_id_55>",
|
| 464 |
-
"single_word": false,
|
| 465 |
-
"lstrip": true,
|
| 466 |
-
"rstrip": false,
|
| 467 |
-
"normalized": true,
|
| 468 |
-
"special": true
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"id": 32045,
|
| 472 |
-
"content": "<extra_id_54>",
|
| 473 |
-
"single_word": false,
|
| 474 |
-
"lstrip": true,
|
| 475 |
-
"rstrip": false,
|
| 476 |
-
"normalized": true,
|
| 477 |
-
"special": true
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"id": 32046,
|
| 481 |
-
"content": "<extra_id_53>",
|
| 482 |
-
"single_word": false,
|
| 483 |
-
"lstrip": true,
|
| 484 |
-
"rstrip": false,
|
| 485 |
-
"normalized": true,
|
| 486 |
-
"special": true
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"id": 32047,
|
| 490 |
-
"content": "<extra_id_52>",
|
| 491 |
-
"single_word": false,
|
| 492 |
-
"lstrip": true,
|
| 493 |
-
"rstrip": false,
|
| 494 |
-
"normalized": true,
|
| 495 |
-
"special": true
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"id": 32048,
|
| 499 |
-
"content": "<extra_id_51>",
|
| 500 |
-
"single_word": false,
|
| 501 |
-
"lstrip": true,
|
| 502 |
-
"rstrip": false,
|
| 503 |
-
"normalized": true,
|
| 504 |
-
"special": true
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"id": 32049,
|
| 508 |
-
"content": "<extra_id_50>",
|
| 509 |
-
"single_word": false,
|
| 510 |
-
"lstrip": true,
|
| 511 |
-
"rstrip": false,
|
| 512 |
-
"normalized": true,
|
| 513 |
-
"special": true
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"id": 32050,
|
| 517 |
-
"content": "<extra_id_49>",
|
| 518 |
-
"single_word": false,
|
| 519 |
-
"lstrip": true,
|
| 520 |
-
"rstrip": false,
|
| 521 |
-
"normalized": true,
|
| 522 |
-
"special": true
|
| 523 |
-
},
|
| 524 |
-
{
|
| 525 |
-
"id": 32051,
|
| 526 |
-
"content": "<extra_id_48>",
|
| 527 |
-
"single_word": false,
|
| 528 |
-
"lstrip": true,
|
| 529 |
-
"rstrip": false,
|
| 530 |
-
"normalized": true,
|
| 531 |
-
"special": true
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"id": 32052,
|
| 535 |
-
"content": "<extra_id_47>",
|
| 536 |
-
"single_word": false,
|
| 537 |
-
"lstrip": true,
|
| 538 |
-
"rstrip": false,
|
| 539 |
-
"normalized": true,
|
| 540 |
-
"special": true
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"id": 32053,
|
| 544 |
-
"content": "<extra_id_46>",
|
| 545 |
-
"single_word": false,
|
| 546 |
-
"lstrip": true,
|
| 547 |
-
"rstrip": false,
|
| 548 |
-
"normalized": true,
|
| 549 |
-
"special": true
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"id": 32054,
|
| 553 |
-
"content": "<extra_id_45>",
|
| 554 |
-
"single_word": false,
|
| 555 |
-
"lstrip": true,
|
| 556 |
-
"rstrip": false,
|
| 557 |
-
"normalized": true,
|
| 558 |
-
"special": true
|
| 559 |
-
},
|
| 560 |
-
{
|
| 561 |
-
"id": 32055,
|
| 562 |
-
"content": "<extra_id_44>",
|
| 563 |
-
"single_word": false,
|
| 564 |
-
"lstrip": true,
|
| 565 |
-
"rstrip": false,
|
| 566 |
-
"normalized": true,
|
| 567 |
-
"special": true
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"id": 32056,
|
| 571 |
-
"content": "<extra_id_43>",
|
| 572 |
-
"single_word": false,
|
| 573 |
-
"lstrip": true,
|
| 574 |
-
"rstrip": false,
|
| 575 |
-
"normalized": true,
|
| 576 |
-
"special": true
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"id": 32057,
|
| 580 |
-
"content": "<extra_id_42>",
|
| 581 |
-
"single_word": false,
|
| 582 |
-
"lstrip": true,
|
| 583 |
-
"rstrip": false,
|
| 584 |
-
"normalized": true,
|
| 585 |
-
"special": true
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"id": 32058,
|
| 589 |
-
"content": "<extra_id_41>",
|
| 590 |
-
"single_word": false,
|
| 591 |
-
"lstrip": true,
|
| 592 |
-
"rstrip": false,
|
| 593 |
-
"normalized": true,
|
| 594 |
-
"special": true
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"id": 32059,
|
| 598 |
-
"content": "<extra_id_40>",
|
| 599 |
-
"single_word": false,
|
| 600 |
-
"lstrip": true,
|
| 601 |
-
"rstrip": false,
|
| 602 |
-
"normalized": true,
|
| 603 |
-
"special": true
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"id": 32060,
|
| 607 |
-
"content": "<extra_id_39>",
|
| 608 |
-
"single_word": false,
|
| 609 |
-
"lstrip": true,
|
| 610 |
-
"rstrip": false,
|
| 611 |
-
"normalized": true,
|
| 612 |
-
"special": true
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"id": 32061,
|
| 616 |
-
"content": "<extra_id_38>",
|
| 617 |
-
"single_word": false,
|
| 618 |
-
"lstrip": true,
|
| 619 |
-
"rstrip": false,
|
| 620 |
-
"normalized": true,
|
| 621 |
-
"special": true
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"id": 32062,
|
| 625 |
-
"content": "<extra_id_37>",
|
| 626 |
-
"single_word": false,
|
| 627 |
-
"lstrip": true,
|
| 628 |
-
"rstrip": false,
|
| 629 |
-
"normalized": true,
|
| 630 |
-
"special": true
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"id": 32063,
|
| 634 |
-
"content": "<extra_id_36>",
|
| 635 |
-
"single_word": false,
|
| 636 |
-
"lstrip": true,
|
| 637 |
-
"rstrip": false,
|
| 638 |
-
"normalized": true,
|
| 639 |
-
"special": true
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"id": 32064,
|
| 643 |
-
"content": "<extra_id_35>",
|
| 644 |
-
"single_word": false,
|
| 645 |
-
"lstrip": true,
|
| 646 |
-
"rstrip": false,
|
| 647 |
-
"normalized": true,
|
| 648 |
-
"special": true
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"id": 32065,
|
| 652 |
-
"content": "<extra_id_34>",
|
| 653 |
-
"single_word": false,
|
| 654 |
-
"lstrip": true,
|
| 655 |
-
"rstrip": false,
|
| 656 |
-
"normalized": true,
|
| 657 |
-
"special": true
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"id": 32066,
|
| 661 |
-
"content": "<extra_id_33>",
|
| 662 |
-
"single_word": false,
|
| 663 |
-
"lstrip": true,
|
| 664 |
-
"rstrip": false,
|
| 665 |
-
"normalized": true,
|
| 666 |
-
"special": true
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"id": 32067,
|
| 670 |
-
"content": "<extra_id_32>",
|
| 671 |
-
"single_word": false,
|
| 672 |
-
"lstrip": true,
|
| 673 |
-
"rstrip": false,
|
| 674 |
-
"normalized": true,
|
| 675 |
-
"special": true
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"id": 32068,
|
| 679 |
-
"content": "<extra_id_31>",
|
| 680 |
-
"single_word": false,
|
| 681 |
-
"lstrip": true,
|
| 682 |
-
"rstrip": false,
|
| 683 |
-
"normalized": true,
|
| 684 |
-
"special": true
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"id": 32069,
|
| 688 |
-
"content": "<extra_id_30>",
|
| 689 |
-
"single_word": false,
|
| 690 |
-
"lstrip": true,
|
| 691 |
-
"rstrip": false,
|
| 692 |
-
"normalized": true,
|
| 693 |
-
"special": true
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"id": 32070,
|
| 697 |
-
"content": "<extra_id_29>",
|
| 698 |
-
"single_word": false,
|
| 699 |
-
"lstrip": true,
|
| 700 |
-
"rstrip": false,
|
| 701 |
-
"normalized": true,
|
| 702 |
-
"special": true
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"id": 32071,
|
| 706 |
-
"content": "<extra_id_28>",
|
| 707 |
-
"single_word": false,
|
| 708 |
-
"lstrip": true,
|
| 709 |
-
"rstrip": false,
|
| 710 |
-
"normalized": true,
|
| 711 |
-
"special": true
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"id": 32072,
|
| 715 |
-
"content": "<extra_id_27>",
|
| 716 |
-
"single_word": false,
|
| 717 |
-
"lstrip": true,
|
| 718 |
-
"rstrip": false,
|
| 719 |
-
"normalized": true,
|
| 720 |
-
"special": true
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"id": 32073,
|
| 724 |
-
"content": "<extra_id_26>",
|
| 725 |
-
"single_word": false,
|
| 726 |
-
"lstrip": true,
|
| 727 |
-
"rstrip": false,
|
| 728 |
-
"normalized": true,
|
| 729 |
-
"special": true
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"id": 32074,
|
| 733 |
-
"content": "<extra_id_25>",
|
| 734 |
-
"single_word": false,
|
| 735 |
-
"lstrip": true,
|
| 736 |
-
"rstrip": false,
|
| 737 |
-
"normalized": true,
|
| 738 |
-
"special": true
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"id": 32075,
|
| 742 |
-
"content": "<extra_id_24>",
|
| 743 |
-
"single_word": false,
|
| 744 |
-
"lstrip": true,
|
| 745 |
-
"rstrip": false,
|
| 746 |
-
"normalized": true,
|
| 747 |
-
"special": true
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"id": 32076,
|
| 751 |
-
"content": "<extra_id_23>",
|
| 752 |
-
"single_word": false,
|
| 753 |
-
"lstrip": true,
|
| 754 |
-
"rstrip": false,
|
| 755 |
-
"normalized": true,
|
| 756 |
-
"special": true
|
| 757 |
-
},
|
| 758 |
-
{
|
| 759 |
-
"id": 32077,
|
| 760 |
-
"content": "<extra_id_22>",
|
| 761 |
-
"single_word": false,
|
| 762 |
-
"lstrip": true,
|
| 763 |
-
"rstrip": false,
|
| 764 |
-
"normalized": true,
|
| 765 |
-
"special": true
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"id": 32078,
|
| 769 |
-
"content": "<extra_id_21>",
|
| 770 |
-
"single_word": false,
|
| 771 |
-
"lstrip": true,
|
| 772 |
-
"rstrip": false,
|
| 773 |
-
"normalized": true,
|
| 774 |
-
"special": true
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"id": 32079,
|
| 778 |
-
"content": "<extra_id_20>",
|
| 779 |
-
"single_word": false,
|
| 780 |
-
"lstrip": true,
|
| 781 |
-
"rstrip": false,
|
| 782 |
-
"normalized": true,
|
| 783 |
-
"special": true
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"id": 32080,
|
| 787 |
-
"content": "<extra_id_19>",
|
| 788 |
-
"single_word": false,
|
| 789 |
-
"lstrip": true,
|
| 790 |
-
"rstrip": false,
|
| 791 |
-
"normalized": true,
|
| 792 |
-
"special": true
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"id": 32081,
|
| 796 |
-
"content": "<extra_id_18>",
|
| 797 |
-
"single_word": false,
|
| 798 |
-
"lstrip": true,
|
| 799 |
-
"rstrip": false,
|
| 800 |
-
"normalized": true,
|
| 801 |
-
"special": true
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"id": 32082,
|
| 805 |
-
"content": "<extra_id_17>",
|
| 806 |
-
"single_word": false,
|
| 807 |
-
"lstrip": true,
|
| 808 |
-
"rstrip": false,
|
| 809 |
-
"normalized": true,
|
| 810 |
-
"special": true
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"id": 32083,
|
| 814 |
-
"content": "<extra_id_16>",
|
| 815 |
-
"single_word": false,
|
| 816 |
-
"lstrip": true,
|
| 817 |
-
"rstrip": false,
|
| 818 |
-
"normalized": true,
|
| 819 |
-
"special": true
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"id": 32084,
|
| 823 |
-
"content": "<extra_id_15>",
|
| 824 |
-
"single_word": false,
|
| 825 |
-
"lstrip": true,
|
| 826 |
-
"rstrip": false,
|
| 827 |
-
"normalized": true,
|
| 828 |
-
"special": true
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"id": 32085,
|
| 832 |
-
"content": "<extra_id_14>",
|
| 833 |
-
"single_word": false,
|
| 834 |
-
"lstrip": true,
|
| 835 |
-
"rstrip": false,
|
| 836 |
-
"normalized": true,
|
| 837 |
-
"special": true
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"id": 32086,
|
| 841 |
-
"content": "<extra_id_13>",
|
| 842 |
-
"single_word": false,
|
| 843 |
-
"lstrip": true,
|
| 844 |
-
"rstrip": false,
|
| 845 |
-
"normalized": true,
|
| 846 |
-
"special": true
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"id": 32087,
|
| 850 |
-
"content": "<extra_id_12>",
|
| 851 |
-
"single_word": false,
|
| 852 |
-
"lstrip": true,
|
| 853 |
-
"rstrip": false,
|
| 854 |
-
"normalized": true,
|
| 855 |
-
"special": true
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"id": 32088,
|
| 859 |
-
"content": "<extra_id_11>",
|
| 860 |
-
"single_word": false,
|
| 861 |
-
"lstrip": true,
|
| 862 |
-
"rstrip": false,
|
| 863 |
-
"normalized": true,
|
| 864 |
-
"special": true
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"id": 32089,
|
| 868 |
-
"content": "<extra_id_10>",
|
| 869 |
-
"single_word": false,
|
| 870 |
-
"lstrip": true,
|
| 871 |
-
"rstrip": false,
|
| 872 |
-
"normalized": true,
|
| 873 |
-
"special": true
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"id": 32090,
|
| 877 |
-
"content": "<extra_id_9>",
|
| 878 |
-
"single_word": false,
|
| 879 |
-
"lstrip": true,
|
| 880 |
-
"rstrip": false,
|
| 881 |
-
"normalized": true,
|
| 882 |
-
"special": true
|
| 883 |
-
},
|
| 884 |
-
{
|
| 885 |
-
"id": 32091,
|
| 886 |
-
"content": "<extra_id_8>",
|
| 887 |
-
"single_word": false,
|
| 888 |
-
"lstrip": true,
|
| 889 |
-
"rstrip": false,
|
| 890 |
-
"normalized": true,
|
| 891 |
-
"special": true
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"id": 32092,
|
| 895 |
-
"content": "<extra_id_7>",
|
| 896 |
-
"single_word": false,
|
| 897 |
-
"lstrip": true,
|
| 898 |
-
"rstrip": false,
|
| 899 |
-
"normalized": true,
|
| 900 |
-
"special": true
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"id": 32093,
|
| 904 |
-
"content": "<extra_id_6>",
|
| 905 |
-
"single_word": false,
|
| 906 |
-
"lstrip": true,
|
| 907 |
-
"rstrip": false,
|
| 908 |
-
"normalized": true,
|
| 909 |
-
"special": true
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"id": 32094,
|
| 913 |
-
"content": "<extra_id_5>",
|
| 914 |
-
"single_word": false,
|
| 915 |
-
"lstrip": true,
|
| 916 |
-
"rstrip": false,
|
| 917 |
-
"normalized": true,
|
| 918 |
-
"special": true
|
| 919 |
-
},
|
| 920 |
-
{
|
| 921 |
-
"id": 32095,
|
| 922 |
-
"content": "<extra_id_4>",
|
| 923 |
-
"single_word": false,
|
| 924 |
-
"lstrip": true,
|
| 925 |
-
"rstrip": false,
|
| 926 |
-
"normalized": true,
|
| 927 |
-
"special": true
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"id": 32096,
|
| 931 |
-
"content": "<extra_id_3>",
|
| 932 |
-
"single_word": false,
|
| 933 |
-
"lstrip": true,
|
| 934 |
-
"rstrip": false,
|
| 935 |
-
"normalized": true,
|
| 936 |
-
"special": true
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"id": 32097,
|
| 940 |
-
"content": "<extra_id_2>",
|
| 941 |
-
"single_word": false,
|
| 942 |
-
"lstrip": true,
|
| 943 |
-
"rstrip": false,
|
| 944 |
-
"normalized": true,
|
| 945 |
-
"special": true
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"id": 32098,
|
| 949 |
-
"content": "<extra_id_1>",
|
| 950 |
-
"single_word": false,
|
| 951 |
-
"lstrip": true,
|
| 952 |
-
"rstrip": false,
|
| 953 |
-
"normalized": true,
|
| 954 |
-
"special": true
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"id": 32099,
|
| 958 |
-
"content": "<extra_id_0>",
|
| 959 |
-
"single_word": false,
|
| 960 |
-
"lstrip": true,
|
| 961 |
-
"rstrip": false,
|
| 962 |
-
"normalized": true,
|
| 963 |
-
"special": true
|
| 964 |
}
|
| 965 |
],
|
| 966 |
"normalizer": null,
|
|
@@ -36813,10 +35899,6 @@
|
|
| 36813 |
"pr",
|
| 36814 |
"int"
|
| 36815 |
],
|
| 36816 |
-
[
|
| 36817 |
-
"#",
|
| 36818 |
-
"#"
|
| 36819 |
-
],
|
| 36820 |
[
|
| 36821 |
"W",
|
| 36822 |
"ith"
|
|
@@ -39937,10 +39019,6 @@
|
|
| 39937 |
"b",
|
| 39938 |
"ase"
|
| 39939 |
],
|
| 39940 |
-
[
|
| 39941 |
-
"##",
|
| 39942 |
-
"##"
|
| 39943 |
-
],
|
| 39944 |
[
|
| 39945 |
"R",
|
| 39946 |
"el"
|
|
@@ -44969,10 +44047,6 @@
|
|
| 44969 |
"Ġre",
|
| 44970 |
"trie"
|
| 44971 |
],
|
| 44972 |
-
[
|
| 44973 |
-
"####",
|
| 44974 |
-
"####"
|
| 44975 |
-
],
|
| 44976 |
[
|
| 44977 |
"Ġmult",
|
| 44978 |
"iple"
|
|
@@ -46625,10 +45699,6 @@
|
|
| 46625 |
"At",
|
| 46626 |
"om"
|
| 46627 |
],
|
| 46628 |
-
[
|
| 46629 |
-
"#",
|
| 46630 |
-
"{"
|
| 46631 |
-
],
|
| 46632 |
[
|
| 46633 |
"Ġro",
|
| 46634 |
"und"
|
|
@@ -54121,10 +53191,6 @@
|
|
| 54121 |
"Ġex",
|
| 54122 |
"plicit"
|
| 54123 |
],
|
| 54124 |
-
[
|
| 54125 |
-
"########",
|
| 54126 |
-
"########"
|
| 54127 |
-
],
|
| 54128 |
[
|
| 54129 |
"Service",
|
| 54130 |
"Response"
|
|
@@ -72209,10 +71275,6 @@
|
|
| 72209 |
"el",
|
| 72210 |
"em"
|
| 72211 |
],
|
| 72212 |
-
[
|
| 72213 |
-
"#",
|
| 72214 |
-
"'"
|
| 72215 |
-
],
|
| 72216 |
[
|
| 72217 |
"Ġcon",
|
| 72218 |
"versation"
|
|
@@ -72617,10 +71679,6 @@
|
|
| 72617 |
"ĠPre",
|
| 72618 |
"fix"
|
| 72619 |
],
|
| 72620 |
-
[
|
| 72621 |
-
"################",
|
| 72622 |
-
"################"
|
| 72623 |
-
],
|
| 72624 |
[
|
| 72625 |
"s",
|
| 72626 |
"ources"
|
|
@@ -84661,10 +83719,6 @@
|
|
| 84661 |
"Ġh",
|
| 84662 |
"el"
|
| 84663 |
],
|
| 84664 |
-
[
|
| 84665 |
-
"##",
|
| 84666 |
-
"#"
|
| 84667 |
-
],
|
| 84668 |
[
|
| 84669 |
"And",
|
| 84670 |
"Set"
|
|
@@ -100589,10 +99643,6 @@
|
|
| 100589 |
"Ġrule",
|
| 100590 |
"ValidID"
|
| 100591 |
],
|
| 100592 |
-
[
|
| 100593 |
-
"#",
|
| 100594 |
-
"\""
|
| 100595 |
-
],
|
| 100596 |
[
|
| 100597 |
"^",
|
| 100598 |
"\\"
|
|
@@ -116621,10 +115671,6 @@
|
|
| 116621 |
"ĠUn",
|
| 116622 |
"der"
|
| 116623 |
],
|
| 116624 |
-
[
|
| 116625 |
-
"#{",
|
| 116626 |
-
"@"
|
| 116627 |
-
],
|
| 116628 |
[
|
| 116629 |
"ĠAct",
|
| 116630 |
"or"
|
|
@@ -138005,10 +137051,6 @@
|
|
| 138005 |
"Queue",
|
| 138006 |
"Entry"
|
| 138007 |
],
|
| 138008 |
-
[
|
| 138009 |
-
"########",
|
| 138010 |
-
"####"
|
| 138011 |
-
],
|
| 138012 |
[
|
| 138013 |
"alan",
|
| 138014 |
"ces"
|
|
@@ -154045,10 +153087,6 @@
|
|
| 154045 |
"SETT",
|
| 154046 |
"ABLE"
|
| 154047 |
],
|
| 154048 |
-
[
|
| 154049 |
-
"#",
|
| 154050 |
-
"$"
|
| 154051 |
-
],
|
| 154052 |
[
|
| 154053 |
"C",
|
| 154054 |
"ores"
|
|
@@ -154449,10 +153487,6 @@
|
|
| 154449 |
"ENDI",
|
| 154450 |
"AN"
|
| 154451 |
],
|
| 154452 |
-
[
|
| 154453 |
-
"#",
|
| 154454 |
-
"__"
|
| 154455 |
-
],
|
| 154456 |
[
|
| 154457 |
">",
|
| 154458 |
"`"
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
| 7 |
"id": 0,
|
|
|
|
| 47 |
"rstrip": false,
|
| 48 |
"normalized": true,
|
| 49 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
}
|
| 51 |
],
|
| 52 |
"normalizer": null,
|
|
|
|
| 35899 |
"pr",
|
| 35900 |
"int"
|
| 35901 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35902 |
[
|
| 35903 |
"W",
|
| 35904 |
"ith"
|
|
|
|
| 39019 |
"b",
|
| 39020 |
"ase"
|
| 39021 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39022 |
[
|
| 39023 |
"R",
|
| 39024 |
"el"
|
|
|
|
| 44047 |
"Ġre",
|
| 44048 |
"trie"
|
| 44049 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44050 |
[
|
| 44051 |
"Ġmult",
|
| 44052 |
"iple"
|
|
|
|
| 45699 |
"At",
|
| 45700 |
"om"
|
| 45701 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45702 |
[
|
| 45703 |
"Ġro",
|
| 45704 |
"und"
|
|
|
|
| 53191 |
"Ġex",
|
| 53192 |
"plicit"
|
| 53193 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53194 |
[
|
| 53195 |
"Service",
|
| 53196 |
"Response"
|
|
|
|
| 71275 |
"el",
|
| 71276 |
"em"
|
| 71277 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71278 |
[
|
| 71279 |
"Ġcon",
|
| 71280 |
"versation"
|
|
|
|
| 71679 |
"ĠPre",
|
| 71680 |
"fix"
|
| 71681 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71682 |
[
|
| 71683 |
"s",
|
| 71684 |
"ources"
|
|
|
|
| 83719 |
"Ġh",
|
| 83720 |
"el"
|
| 83721 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 83722 |
[
|
| 83723 |
"And",
|
| 83724 |
"Set"
|
|
|
|
| 99643 |
"Ġrule",
|
| 99644 |
"ValidID"
|
| 99645 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 99646 |
[
|
| 99647 |
"^",
|
| 99648 |
"\\"
|
|
|
|
| 115671 |
"ĠUn",
|
| 115672 |
"der"
|
| 115673 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 115674 |
[
|
| 115675 |
"ĠAct",
|
| 115676 |
"or"
|
|
|
|
| 137051 |
"Queue",
|
| 137052 |
"Entry"
|
| 137053 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 137054 |
[
|
| 137055 |
"alan",
|
| 137056 |
"ces"
|
|
|
|
| 153087 |
"SETT",
|
| 153088 |
"ABLE"
|
| 153089 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 153090 |
[
|
| 153091 |
"C",
|
| 153092 |
"ores"
|
|
|
|
| 153487 |
"ENDI",
|
| 153488 |
"AN"
|
| 153489 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 153490 |
[
|
| 153491 |
">",
|
| 153492 |
"`"
|
tokenizer_config.json
CHANGED
|
@@ -1,955 +1,714 @@
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
-
"
|
| 4 |
-
|
| 5 |
-
"content": "<pad>",
|
| 6 |
-
"lstrip": false,
|
| 7 |
-
"normalized": true,
|
| 8 |
-
"rstrip": false,
|
| 9 |
-
"single_word": false,
|
| 10 |
-
"special": true
|
| 11 |
-
},
|
| 12 |
-
"1": {
|
| 13 |
-
"content": "<s>",
|
| 14 |
-
"lstrip": false,
|
| 15 |
-
"normalized": true,
|
| 16 |
-
"rstrip": false,
|
| 17 |
-
"single_word": false,
|
| 18 |
-
"special": true
|
| 19 |
-
},
|
| 20 |
-
"2": {
|
| 21 |
-
"content": "</s>",
|
| 22 |
-
"lstrip": false,
|
| 23 |
-
"normalized": true,
|
| 24 |
-
"rstrip": false,
|
| 25 |
-
"single_word": false,
|
| 26 |
-
"special": true
|
| 27 |
-
},
|
| 28 |
-
"3": {
|
| 29 |
-
"content": "<unk>",
|
| 30 |
-
"lstrip": false,
|
| 31 |
-
"normalized": true,
|
| 32 |
-
"rstrip": false,
|
| 33 |
-
"single_word": false,
|
| 34 |
-
"special": true
|
| 35 |
-
},
|
| 36 |
-
"4": {
|
| 37 |
-
"content": "<mask>",
|
| 38 |
-
"lstrip": true,
|
| 39 |
-
"normalized": true,
|
| 40 |
-
"rstrip": false,
|
| 41 |
-
"single_word": false,
|
| 42 |
-
"special": true
|
| 43 |
-
},
|
| 44 |
-
"32000": {
|
| 45 |
"content": "<extra_id_99>",
|
| 46 |
"lstrip": true,
|
| 47 |
"normalized": true,
|
| 48 |
"rstrip": false,
|
| 49 |
-
"single_word": false
|
| 50 |
-
"special": true
|
| 51 |
},
|
| 52 |
-
|
| 53 |
"content": "<extra_id_98>",
|
| 54 |
"lstrip": true,
|
| 55 |
"normalized": true,
|
| 56 |
"rstrip": false,
|
| 57 |
-
"single_word": false
|
| 58 |
-
"special": true
|
| 59 |
},
|
| 60 |
-
|
| 61 |
"content": "<extra_id_97>",
|
| 62 |
"lstrip": true,
|
| 63 |
"normalized": true,
|
| 64 |
"rstrip": false,
|
| 65 |
-
"single_word": false
|
| 66 |
-
"special": true
|
| 67 |
},
|
| 68 |
-
|
| 69 |
"content": "<extra_id_96>",
|
| 70 |
"lstrip": true,
|
| 71 |
"normalized": true,
|
| 72 |
"rstrip": false,
|
| 73 |
-
"single_word": false
|
| 74 |
-
"special": true
|
| 75 |
},
|
| 76 |
-
|
| 77 |
"content": "<extra_id_95>",
|
| 78 |
"lstrip": true,
|
| 79 |
"normalized": true,
|
| 80 |
"rstrip": false,
|
| 81 |
-
"single_word": false
|
| 82 |
-
"special": true
|
| 83 |
},
|
| 84 |
-
|
| 85 |
"content": "<extra_id_94>",
|
| 86 |
"lstrip": true,
|
| 87 |
"normalized": true,
|
| 88 |
"rstrip": false,
|
| 89 |
-
"single_word": false
|
| 90 |
-
"special": true
|
| 91 |
},
|
| 92 |
-
|
| 93 |
"content": "<extra_id_93>",
|
| 94 |
"lstrip": true,
|
| 95 |
"normalized": true,
|
| 96 |
"rstrip": false,
|
| 97 |
-
"single_word": false
|
| 98 |
-
"special": true
|
| 99 |
},
|
| 100 |
-
|
| 101 |
"content": "<extra_id_92>",
|
| 102 |
"lstrip": true,
|
| 103 |
"normalized": true,
|
| 104 |
"rstrip": false,
|
| 105 |
-
"single_word": false
|
| 106 |
-
"special": true
|
| 107 |
},
|
| 108 |
-
|
| 109 |
"content": "<extra_id_91>",
|
| 110 |
"lstrip": true,
|
| 111 |
"normalized": true,
|
| 112 |
"rstrip": false,
|
| 113 |
-
"single_word": false
|
| 114 |
-
"special": true
|
| 115 |
},
|
| 116 |
-
|
| 117 |
"content": "<extra_id_90>",
|
| 118 |
"lstrip": true,
|
| 119 |
"normalized": true,
|
| 120 |
"rstrip": false,
|
| 121 |
-
"single_word": false
|
| 122 |
-
"special": true
|
| 123 |
},
|
| 124 |
-
|
| 125 |
"content": "<extra_id_89>",
|
| 126 |
"lstrip": true,
|
| 127 |
"normalized": true,
|
| 128 |
"rstrip": false,
|
| 129 |
-
"single_word": false
|
| 130 |
-
"special": true
|
| 131 |
},
|
| 132 |
-
|
| 133 |
"content": "<extra_id_88>",
|
| 134 |
"lstrip": true,
|
| 135 |
"normalized": true,
|
| 136 |
"rstrip": false,
|
| 137 |
-
"single_word": false
|
| 138 |
-
"special": true
|
| 139 |
},
|
| 140 |
-
|
| 141 |
"content": "<extra_id_87>",
|
| 142 |
"lstrip": true,
|
| 143 |
"normalized": true,
|
| 144 |
"rstrip": false,
|
| 145 |
-
"single_word": false
|
| 146 |
-
"special": true
|
| 147 |
},
|
| 148 |
-
|
| 149 |
"content": "<extra_id_86>",
|
| 150 |
"lstrip": true,
|
| 151 |
"normalized": true,
|
| 152 |
"rstrip": false,
|
| 153 |
-
"single_word": false
|
| 154 |
-
"special": true
|
| 155 |
},
|
| 156 |
-
|
| 157 |
"content": "<extra_id_85>",
|
| 158 |
"lstrip": true,
|
| 159 |
"normalized": true,
|
| 160 |
"rstrip": false,
|
| 161 |
-
"single_word": false
|
| 162 |
-
"special": true
|
| 163 |
},
|
| 164 |
-
|
| 165 |
"content": "<extra_id_84>",
|
| 166 |
"lstrip": true,
|
| 167 |
"normalized": true,
|
| 168 |
"rstrip": false,
|
| 169 |
-
"single_word": false
|
| 170 |
-
"special": true
|
| 171 |
},
|
| 172 |
-
|
| 173 |
"content": "<extra_id_83>",
|
| 174 |
"lstrip": true,
|
| 175 |
"normalized": true,
|
| 176 |
"rstrip": false,
|
| 177 |
-
"single_word": false
|
| 178 |
-
"special": true
|
| 179 |
},
|
| 180 |
-
|
| 181 |
"content": "<extra_id_82>",
|
| 182 |
"lstrip": true,
|
| 183 |
"normalized": true,
|
| 184 |
"rstrip": false,
|
| 185 |
-
"single_word": false
|
| 186 |
-
"special": true
|
| 187 |
},
|
| 188 |
-
|
| 189 |
"content": "<extra_id_81>",
|
| 190 |
"lstrip": true,
|
| 191 |
"normalized": true,
|
| 192 |
"rstrip": false,
|
| 193 |
-
"single_word": false
|
| 194 |
-
"special": true
|
| 195 |
},
|
| 196 |
-
|
| 197 |
"content": "<extra_id_80>",
|
| 198 |
"lstrip": true,
|
| 199 |
"normalized": true,
|
| 200 |
"rstrip": false,
|
| 201 |
-
"single_word": false
|
| 202 |
-
"special": true
|
| 203 |
},
|
| 204 |
-
|
| 205 |
"content": "<extra_id_79>",
|
| 206 |
"lstrip": true,
|
| 207 |
"normalized": true,
|
| 208 |
"rstrip": false,
|
| 209 |
-
"single_word": false
|
| 210 |
-
"special": true
|
| 211 |
},
|
| 212 |
-
|
| 213 |
"content": "<extra_id_78>",
|
| 214 |
"lstrip": true,
|
| 215 |
"normalized": true,
|
| 216 |
"rstrip": false,
|
| 217 |
-
"single_word": false
|
| 218 |
-
"special": true
|
| 219 |
},
|
| 220 |
-
|
| 221 |
"content": "<extra_id_77>",
|
| 222 |
"lstrip": true,
|
| 223 |
"normalized": true,
|
| 224 |
"rstrip": false,
|
| 225 |
-
"single_word": false
|
| 226 |
-
"special": true
|
| 227 |
},
|
| 228 |
-
|
| 229 |
"content": "<extra_id_76>",
|
| 230 |
"lstrip": true,
|
| 231 |
"normalized": true,
|
| 232 |
"rstrip": false,
|
| 233 |
-
"single_word": false
|
| 234 |
-
"special": true
|
| 235 |
},
|
| 236 |
-
|
| 237 |
"content": "<extra_id_75>",
|
| 238 |
"lstrip": true,
|
| 239 |
"normalized": true,
|
| 240 |
"rstrip": false,
|
| 241 |
-
"single_word": false
|
| 242 |
-
"special": true
|
| 243 |
},
|
| 244 |
-
|
| 245 |
"content": "<extra_id_74>",
|
| 246 |
"lstrip": true,
|
| 247 |
"normalized": true,
|
| 248 |
"rstrip": false,
|
| 249 |
-
"single_word": false
|
| 250 |
-
"special": true
|
| 251 |
},
|
| 252 |
-
|
| 253 |
"content": "<extra_id_73>",
|
| 254 |
"lstrip": true,
|
| 255 |
"normalized": true,
|
| 256 |
"rstrip": false,
|
| 257 |
-
"single_word": false
|
| 258 |
-
"special": true
|
| 259 |
},
|
| 260 |
-
|
| 261 |
"content": "<extra_id_72>",
|
| 262 |
"lstrip": true,
|
| 263 |
"normalized": true,
|
| 264 |
"rstrip": false,
|
| 265 |
-
"single_word": false
|
| 266 |
-
"special": true
|
| 267 |
},
|
| 268 |
-
|
| 269 |
"content": "<extra_id_71>",
|
| 270 |
"lstrip": true,
|
| 271 |
"normalized": true,
|
| 272 |
"rstrip": false,
|
| 273 |
-
"single_word": false
|
| 274 |
-
"special": true
|
| 275 |
},
|
| 276 |
-
|
| 277 |
"content": "<extra_id_70>",
|
| 278 |
"lstrip": true,
|
| 279 |
"normalized": true,
|
| 280 |
"rstrip": false,
|
| 281 |
-
"single_word": false
|
| 282 |
-
"special": true
|
| 283 |
},
|
| 284 |
-
|
| 285 |
"content": "<extra_id_69>",
|
| 286 |
"lstrip": true,
|
| 287 |
"normalized": true,
|
| 288 |
"rstrip": false,
|
| 289 |
-
"single_word": false
|
| 290 |
-
"special": true
|
| 291 |
},
|
| 292 |
-
|
| 293 |
"content": "<extra_id_68>",
|
| 294 |
"lstrip": true,
|
| 295 |
"normalized": true,
|
| 296 |
"rstrip": false,
|
| 297 |
-
"single_word": false
|
| 298 |
-
"special": true
|
| 299 |
},
|
| 300 |
-
|
| 301 |
"content": "<extra_id_67>",
|
| 302 |
"lstrip": true,
|
| 303 |
"normalized": true,
|
| 304 |
"rstrip": false,
|
| 305 |
-
"single_word": false
|
| 306 |
-
"special": true
|
| 307 |
},
|
| 308 |
-
|
| 309 |
"content": "<extra_id_66>",
|
| 310 |
"lstrip": true,
|
| 311 |
"normalized": true,
|
| 312 |
"rstrip": false,
|
| 313 |
-
"single_word": false
|
| 314 |
-
"special": true
|
| 315 |
},
|
| 316 |
-
|
| 317 |
"content": "<extra_id_65>",
|
| 318 |
"lstrip": true,
|
| 319 |
"normalized": true,
|
| 320 |
"rstrip": false,
|
| 321 |
-
"single_word": false
|
| 322 |
-
"special": true
|
| 323 |
},
|
| 324 |
-
|
| 325 |
"content": "<extra_id_64>",
|
| 326 |
"lstrip": true,
|
| 327 |
"normalized": true,
|
| 328 |
"rstrip": false,
|
| 329 |
-
"single_word": false
|
| 330 |
-
"special": true
|
| 331 |
},
|
| 332 |
-
|
| 333 |
"content": "<extra_id_63>",
|
| 334 |
"lstrip": true,
|
| 335 |
"normalized": true,
|
| 336 |
"rstrip": false,
|
| 337 |
-
"single_word": false
|
| 338 |
-
"special": true
|
| 339 |
},
|
| 340 |
-
|
| 341 |
"content": "<extra_id_62>",
|
| 342 |
"lstrip": true,
|
| 343 |
"normalized": true,
|
| 344 |
"rstrip": false,
|
| 345 |
-
"single_word": false
|
| 346 |
-
"special": true
|
| 347 |
},
|
| 348 |
-
|
| 349 |
"content": "<extra_id_61>",
|
| 350 |
"lstrip": true,
|
| 351 |
"normalized": true,
|
| 352 |
"rstrip": false,
|
| 353 |
-
"single_word": false
|
| 354 |
-
"special": true
|
| 355 |
},
|
| 356 |
-
|
| 357 |
"content": "<extra_id_60>",
|
| 358 |
"lstrip": true,
|
| 359 |
"normalized": true,
|
| 360 |
"rstrip": false,
|
| 361 |
-
"single_word": false
|
| 362 |
-
"special": true
|
| 363 |
},
|
| 364 |
-
|
| 365 |
"content": "<extra_id_59>",
|
| 366 |
"lstrip": true,
|
| 367 |
"normalized": true,
|
| 368 |
"rstrip": false,
|
| 369 |
-
"single_word": false
|
| 370 |
-
"special": true
|
| 371 |
},
|
| 372 |
-
|
| 373 |
"content": "<extra_id_58>",
|
| 374 |
"lstrip": true,
|
| 375 |
"normalized": true,
|
| 376 |
"rstrip": false,
|
| 377 |
-
"single_word": false
|
| 378 |
-
"special": true
|
| 379 |
},
|
| 380 |
-
|
| 381 |
"content": "<extra_id_57>",
|
| 382 |
"lstrip": true,
|
| 383 |
"normalized": true,
|
| 384 |
"rstrip": false,
|
| 385 |
-
"single_word": false
|
| 386 |
-
"special": true
|
| 387 |
},
|
| 388 |
-
|
| 389 |
"content": "<extra_id_56>",
|
| 390 |
"lstrip": true,
|
| 391 |
"normalized": true,
|
| 392 |
"rstrip": false,
|
| 393 |
-
"single_word": false
|
| 394 |
-
"special": true
|
| 395 |
},
|
| 396 |
-
|
| 397 |
"content": "<extra_id_55>",
|
| 398 |
"lstrip": true,
|
| 399 |
"normalized": true,
|
| 400 |
"rstrip": false,
|
| 401 |
-
"single_word": false
|
| 402 |
-
"special": true
|
| 403 |
},
|
| 404 |
-
|
| 405 |
"content": "<extra_id_54>",
|
| 406 |
"lstrip": true,
|
| 407 |
"normalized": true,
|
| 408 |
"rstrip": false,
|
| 409 |
-
"single_word": false
|
| 410 |
-
"special": true
|
| 411 |
},
|
| 412 |
-
|
| 413 |
"content": "<extra_id_53>",
|
| 414 |
"lstrip": true,
|
| 415 |
"normalized": true,
|
| 416 |
"rstrip": false,
|
| 417 |
-
"single_word": false
|
| 418 |
-
"special": true
|
| 419 |
},
|
| 420 |
-
|
| 421 |
"content": "<extra_id_52>",
|
| 422 |
"lstrip": true,
|
| 423 |
"normalized": true,
|
| 424 |
"rstrip": false,
|
| 425 |
-
"single_word": false
|
| 426 |
-
"special": true
|
| 427 |
},
|
| 428 |
-
|
| 429 |
"content": "<extra_id_51>",
|
| 430 |
"lstrip": true,
|
| 431 |
"normalized": true,
|
| 432 |
"rstrip": false,
|
| 433 |
-
"single_word": false
|
| 434 |
-
"special": true
|
| 435 |
},
|
| 436 |
-
|
| 437 |
"content": "<extra_id_50>",
|
| 438 |
"lstrip": true,
|
| 439 |
"normalized": true,
|
| 440 |
"rstrip": false,
|
| 441 |
-
"single_word": false
|
| 442 |
-
"special": true
|
| 443 |
},
|
| 444 |
-
|
| 445 |
"content": "<extra_id_49>",
|
| 446 |
"lstrip": true,
|
| 447 |
"normalized": true,
|
| 448 |
"rstrip": false,
|
| 449 |
-
"single_word": false
|
| 450 |
-
"special": true
|
| 451 |
},
|
| 452 |
-
|
| 453 |
"content": "<extra_id_48>",
|
| 454 |
"lstrip": true,
|
| 455 |
"normalized": true,
|
| 456 |
"rstrip": false,
|
| 457 |
-
"single_word": false
|
| 458 |
-
"special": true
|
| 459 |
},
|
| 460 |
-
|
| 461 |
"content": "<extra_id_47>",
|
| 462 |
"lstrip": true,
|
| 463 |
"normalized": true,
|
| 464 |
"rstrip": false,
|
| 465 |
-
"single_word": false
|
| 466 |
-
"special": true
|
| 467 |
},
|
| 468 |
-
|
| 469 |
"content": "<extra_id_46>",
|
| 470 |
"lstrip": true,
|
| 471 |
"normalized": true,
|
| 472 |
"rstrip": false,
|
| 473 |
-
"single_word": false
|
| 474 |
-
"special": true
|
| 475 |
},
|
| 476 |
-
|
| 477 |
"content": "<extra_id_45>",
|
| 478 |
"lstrip": true,
|
| 479 |
"normalized": true,
|
| 480 |
"rstrip": false,
|
| 481 |
-
"single_word": false
|
| 482 |
-
"special": true
|
| 483 |
},
|
| 484 |
-
|
| 485 |
"content": "<extra_id_44>",
|
| 486 |
"lstrip": true,
|
| 487 |
"normalized": true,
|
| 488 |
"rstrip": false,
|
| 489 |
-
"single_word": false
|
| 490 |
-
"special": true
|
| 491 |
},
|
| 492 |
-
|
| 493 |
"content": "<extra_id_43>",
|
| 494 |
"lstrip": true,
|
| 495 |
"normalized": true,
|
| 496 |
"rstrip": false,
|
| 497 |
-
"single_word": false
|
| 498 |
-
"special": true
|
| 499 |
},
|
| 500 |
-
|
| 501 |
"content": "<extra_id_42>",
|
| 502 |
"lstrip": true,
|
| 503 |
"normalized": true,
|
| 504 |
"rstrip": false,
|
| 505 |
-
"single_word": false
|
| 506 |
-
"special": true
|
| 507 |
},
|
| 508 |
-
|
| 509 |
"content": "<extra_id_41>",
|
| 510 |
"lstrip": true,
|
| 511 |
"normalized": true,
|
| 512 |
"rstrip": false,
|
| 513 |
-
"single_word": false
|
| 514 |
-
"special": true
|
| 515 |
},
|
| 516 |
-
|
| 517 |
"content": "<extra_id_40>",
|
| 518 |
"lstrip": true,
|
| 519 |
"normalized": true,
|
| 520 |
"rstrip": false,
|
| 521 |
-
"single_word": false
|
| 522 |
-
"special": true
|
| 523 |
},
|
| 524 |
-
|
| 525 |
"content": "<extra_id_39>",
|
| 526 |
"lstrip": true,
|
| 527 |
"normalized": true,
|
| 528 |
"rstrip": false,
|
| 529 |
-
"single_word": false
|
| 530 |
-
"special": true
|
| 531 |
},
|
| 532 |
-
|
| 533 |
"content": "<extra_id_38>",
|
| 534 |
"lstrip": true,
|
| 535 |
"normalized": true,
|
| 536 |
"rstrip": false,
|
| 537 |
-
"single_word": false
|
| 538 |
-
"special": true
|
| 539 |
},
|
| 540 |
-
|
| 541 |
"content": "<extra_id_37>",
|
| 542 |
"lstrip": true,
|
| 543 |
"normalized": true,
|
| 544 |
"rstrip": false,
|
| 545 |
-
"single_word": false
|
| 546 |
-
"special": true
|
| 547 |
},
|
| 548 |
-
|
| 549 |
"content": "<extra_id_36>",
|
| 550 |
"lstrip": true,
|
| 551 |
"normalized": true,
|
| 552 |
"rstrip": false,
|
| 553 |
-
"single_word": false
|
| 554 |
-
"special": true
|
| 555 |
},
|
| 556 |
-
|
| 557 |
"content": "<extra_id_35>",
|
| 558 |
"lstrip": true,
|
| 559 |
"normalized": true,
|
| 560 |
"rstrip": false,
|
| 561 |
-
"single_word": false
|
| 562 |
-
"special": true
|
| 563 |
},
|
| 564 |
-
|
| 565 |
"content": "<extra_id_34>",
|
| 566 |
"lstrip": true,
|
| 567 |
"normalized": true,
|
| 568 |
"rstrip": false,
|
| 569 |
-
"single_word": false
|
| 570 |
-
"special": true
|
| 571 |
},
|
| 572 |
-
|
| 573 |
"content": "<extra_id_33>",
|
| 574 |
"lstrip": true,
|
| 575 |
"normalized": true,
|
| 576 |
"rstrip": false,
|
| 577 |
-
"single_word": false
|
| 578 |
-
"special": true
|
| 579 |
},
|
| 580 |
-
|
| 581 |
"content": "<extra_id_32>",
|
| 582 |
"lstrip": true,
|
| 583 |
"normalized": true,
|
| 584 |
"rstrip": false,
|
| 585 |
-
"single_word": false
|
| 586 |
-
"special": true
|
| 587 |
},
|
| 588 |
-
|
| 589 |
"content": "<extra_id_31>",
|
| 590 |
"lstrip": true,
|
| 591 |
"normalized": true,
|
| 592 |
"rstrip": false,
|
| 593 |
-
"single_word": false
|
| 594 |
-
"special": true
|
| 595 |
},
|
| 596 |
-
|
| 597 |
"content": "<extra_id_30>",
|
| 598 |
"lstrip": true,
|
| 599 |
"normalized": true,
|
| 600 |
"rstrip": false,
|
| 601 |
-
"single_word": false
|
| 602 |
-
"special": true
|
| 603 |
},
|
| 604 |
-
|
| 605 |
"content": "<extra_id_29>",
|
| 606 |
"lstrip": true,
|
| 607 |
"normalized": true,
|
| 608 |
"rstrip": false,
|
| 609 |
-
"single_word": false
|
| 610 |
-
"special": true
|
| 611 |
},
|
| 612 |
-
|
| 613 |
"content": "<extra_id_28>",
|
| 614 |
"lstrip": true,
|
| 615 |
"normalized": true,
|
| 616 |
"rstrip": false,
|
| 617 |
-
"single_word": false
|
| 618 |
-
"special": true
|
| 619 |
},
|
| 620 |
-
|
| 621 |
"content": "<extra_id_27>",
|
| 622 |
"lstrip": true,
|
| 623 |
"normalized": true,
|
| 624 |
"rstrip": false,
|
| 625 |
-
"single_word": false
|
| 626 |
-
"special": true
|
| 627 |
},
|
| 628 |
-
|
| 629 |
"content": "<extra_id_26>",
|
| 630 |
"lstrip": true,
|
| 631 |
"normalized": true,
|
| 632 |
"rstrip": false,
|
| 633 |
-
"single_word": false
|
| 634 |
-
"special": true
|
| 635 |
},
|
| 636 |
-
|
| 637 |
"content": "<extra_id_25>",
|
| 638 |
"lstrip": true,
|
| 639 |
"normalized": true,
|
| 640 |
"rstrip": false,
|
| 641 |
-
"single_word": false
|
| 642 |
-
"special": true
|
| 643 |
},
|
| 644 |
-
|
| 645 |
"content": "<extra_id_24>",
|
| 646 |
"lstrip": true,
|
| 647 |
"normalized": true,
|
| 648 |
"rstrip": false,
|
| 649 |
-
"single_word": false
|
| 650 |
-
"special": true
|
| 651 |
},
|
| 652 |
-
|
| 653 |
"content": "<extra_id_23>",
|
| 654 |
"lstrip": true,
|
| 655 |
"normalized": true,
|
| 656 |
"rstrip": false,
|
| 657 |
-
"single_word": false
|
| 658 |
-
"special": true
|
| 659 |
},
|
| 660 |
-
|
| 661 |
"content": "<extra_id_22>",
|
| 662 |
"lstrip": true,
|
| 663 |
"normalized": true,
|
| 664 |
"rstrip": false,
|
| 665 |
-
"single_word": false
|
| 666 |
-
"special": true
|
| 667 |
},
|
| 668 |
-
|
| 669 |
"content": "<extra_id_21>",
|
| 670 |
"lstrip": true,
|
| 671 |
"normalized": true,
|
| 672 |
"rstrip": false,
|
| 673 |
-
"single_word": false
|
| 674 |
-
"special": true
|
| 675 |
},
|
| 676 |
-
|
| 677 |
"content": "<extra_id_20>",
|
| 678 |
"lstrip": true,
|
| 679 |
"normalized": true,
|
| 680 |
"rstrip": false,
|
| 681 |
-
"single_word": false
|
| 682 |
-
"special": true
|
| 683 |
},
|
| 684 |
-
|
| 685 |
"content": "<extra_id_19>",
|
| 686 |
"lstrip": true,
|
| 687 |
"normalized": true,
|
| 688 |
"rstrip": false,
|
| 689 |
-
"single_word": false
|
| 690 |
-
"special": true
|
| 691 |
},
|
| 692 |
-
|
| 693 |
"content": "<extra_id_18>",
|
| 694 |
"lstrip": true,
|
| 695 |
"normalized": true,
|
| 696 |
"rstrip": false,
|
| 697 |
-
"single_word": false
|
| 698 |
-
"special": true
|
| 699 |
},
|
| 700 |
-
|
| 701 |
"content": "<extra_id_17>",
|
| 702 |
"lstrip": true,
|
| 703 |
"normalized": true,
|
| 704 |
"rstrip": false,
|
| 705 |
-
"single_word": false
|
| 706 |
-
"special": true
|
| 707 |
},
|
| 708 |
-
|
| 709 |
"content": "<extra_id_16>",
|
| 710 |
"lstrip": true,
|
| 711 |
"normalized": true,
|
| 712 |
"rstrip": false,
|
| 713 |
-
"single_word": false
|
| 714 |
-
"special": true
|
| 715 |
},
|
| 716 |
-
|
| 717 |
"content": "<extra_id_15>",
|
| 718 |
"lstrip": true,
|
| 719 |
"normalized": true,
|
| 720 |
"rstrip": false,
|
| 721 |
-
"single_word": false
|
| 722 |
-
"special": true
|
| 723 |
},
|
| 724 |
-
|
| 725 |
"content": "<extra_id_14>",
|
| 726 |
"lstrip": true,
|
| 727 |
"normalized": true,
|
| 728 |
"rstrip": false,
|
| 729 |
-
"single_word": false
|
| 730 |
-
"special": true
|
| 731 |
},
|
| 732 |
-
|
| 733 |
"content": "<extra_id_13>",
|
| 734 |
"lstrip": true,
|
| 735 |
"normalized": true,
|
| 736 |
"rstrip": false,
|
| 737 |
-
"single_word": false
|
| 738 |
-
"special": true
|
| 739 |
},
|
| 740 |
-
|
| 741 |
"content": "<extra_id_12>",
|
| 742 |
"lstrip": true,
|
| 743 |
"normalized": true,
|
| 744 |
"rstrip": false,
|
| 745 |
-
"single_word": false
|
| 746 |
-
"special": true
|
| 747 |
},
|
| 748 |
-
|
| 749 |
"content": "<extra_id_11>",
|
| 750 |
"lstrip": true,
|
| 751 |
"normalized": true,
|
| 752 |
"rstrip": false,
|
| 753 |
-
"single_word": false
|
| 754 |
-
"special": true
|
| 755 |
},
|
| 756 |
-
|
| 757 |
"content": "<extra_id_10>",
|
| 758 |
"lstrip": true,
|
| 759 |
"normalized": true,
|
| 760 |
"rstrip": false,
|
| 761 |
-
"single_word": false
|
| 762 |
-
"special": true
|
| 763 |
},
|
| 764 |
-
|
| 765 |
"content": "<extra_id_9>",
|
| 766 |
"lstrip": true,
|
| 767 |
"normalized": true,
|
| 768 |
"rstrip": false,
|
| 769 |
-
"single_word": false
|
| 770 |
-
"special": true
|
| 771 |
},
|
| 772 |
-
|
| 773 |
"content": "<extra_id_8>",
|
| 774 |
"lstrip": true,
|
| 775 |
"normalized": true,
|
| 776 |
"rstrip": false,
|
| 777 |
-
"single_word": false
|
| 778 |
-
"special": true
|
| 779 |
},
|
| 780 |
-
|
| 781 |
"content": "<extra_id_7>",
|
| 782 |
"lstrip": true,
|
| 783 |
"normalized": true,
|
| 784 |
"rstrip": false,
|
| 785 |
-
"single_word": false
|
| 786 |
-
"special": true
|
| 787 |
},
|
| 788 |
-
|
| 789 |
"content": "<extra_id_6>",
|
| 790 |
"lstrip": true,
|
| 791 |
"normalized": true,
|
| 792 |
"rstrip": false,
|
| 793 |
-
"single_word": false
|
| 794 |
-
"special": true
|
| 795 |
},
|
| 796 |
-
|
| 797 |
"content": "<extra_id_5>",
|
| 798 |
"lstrip": true,
|
| 799 |
"normalized": true,
|
| 800 |
"rstrip": false,
|
| 801 |
-
"single_word": false
|
| 802 |
-
"special": true
|
| 803 |
},
|
| 804 |
-
|
| 805 |
"content": "<extra_id_4>",
|
| 806 |
"lstrip": true,
|
| 807 |
"normalized": true,
|
| 808 |
"rstrip": false,
|
| 809 |
-
"single_word": false
|
| 810 |
-
"special": true
|
| 811 |
},
|
| 812 |
-
|
| 813 |
"content": "<extra_id_3>",
|
| 814 |
"lstrip": true,
|
| 815 |
"normalized": true,
|
| 816 |
"rstrip": false,
|
| 817 |
-
"single_word": false
|
| 818 |
-
"special": true
|
| 819 |
},
|
| 820 |
-
|
| 821 |
"content": "<extra_id_2>",
|
| 822 |
"lstrip": true,
|
| 823 |
"normalized": true,
|
| 824 |
"rstrip": false,
|
| 825 |
-
"single_word": false
|
| 826 |
-
"special": true
|
| 827 |
},
|
| 828 |
-
|
| 829 |
"content": "<extra_id_1>",
|
| 830 |
"lstrip": true,
|
| 831 |
"normalized": true,
|
| 832 |
"rstrip": false,
|
| 833 |
-
"single_word": false
|
| 834 |
-
"special": true
|
| 835 |
},
|
| 836 |
-
|
| 837 |
"content": "<extra_id_0>",
|
| 838 |
"lstrip": true,
|
| 839 |
"normalized": true,
|
| 840 |
"rstrip": false,
|
| 841 |
-
"single_word": false
|
| 842 |
-
"special": true
|
| 843 |
}
|
| 844 |
-
},
|
| 845 |
-
"additional_special_tokens": [
|
| 846 |
-
"<extra_id_99>",
|
| 847 |
-
"<extra_id_98>",
|
| 848 |
-
"<extra_id_97>",
|
| 849 |
-
"<extra_id_96>",
|
| 850 |
-
"<extra_id_95>",
|
| 851 |
-
"<extra_id_94>",
|
| 852 |
-
"<extra_id_93>",
|
| 853 |
-
"<extra_id_92>",
|
| 854 |
-
"<extra_id_91>",
|
| 855 |
-
"<extra_id_90>",
|
| 856 |
-
"<extra_id_89>",
|
| 857 |
-
"<extra_id_88>",
|
| 858 |
-
"<extra_id_87>",
|
| 859 |
-
"<extra_id_86>",
|
| 860 |
-
"<extra_id_85>",
|
| 861 |
-
"<extra_id_84>",
|
| 862 |
-
"<extra_id_83>",
|
| 863 |
-
"<extra_id_82>",
|
| 864 |
-
"<extra_id_81>",
|
| 865 |
-
"<extra_id_80>",
|
| 866 |
-
"<extra_id_79>",
|
| 867 |
-
"<extra_id_78>",
|
| 868 |
-
"<extra_id_77>",
|
| 869 |
-
"<extra_id_76>",
|
| 870 |
-
"<extra_id_75>",
|
| 871 |
-
"<extra_id_74>",
|
| 872 |
-
"<extra_id_73>",
|
| 873 |
-
"<extra_id_72>",
|
| 874 |
-
"<extra_id_71>",
|
| 875 |
-
"<extra_id_70>",
|
| 876 |
-
"<extra_id_69>",
|
| 877 |
-
"<extra_id_68>",
|
| 878 |
-
"<extra_id_67>",
|
| 879 |
-
"<extra_id_66>",
|
| 880 |
-
"<extra_id_65>",
|
| 881 |
-
"<extra_id_64>",
|
| 882 |
-
"<extra_id_63>",
|
| 883 |
-
"<extra_id_62>",
|
| 884 |
-
"<extra_id_61>",
|
| 885 |
-
"<extra_id_60>",
|
| 886 |
-
"<extra_id_59>",
|
| 887 |
-
"<extra_id_58>",
|
| 888 |
-
"<extra_id_57>",
|
| 889 |
-
"<extra_id_56>",
|
| 890 |
-
"<extra_id_55>",
|
| 891 |
-
"<extra_id_54>",
|
| 892 |
-
"<extra_id_53>",
|
| 893 |
-
"<extra_id_52>",
|
| 894 |
-
"<extra_id_51>",
|
| 895 |
-
"<extra_id_50>",
|
| 896 |
-
"<extra_id_49>",
|
| 897 |
-
"<extra_id_48>",
|
| 898 |
-
"<extra_id_47>",
|
| 899 |
-
"<extra_id_46>",
|
| 900 |
-
"<extra_id_45>",
|
| 901 |
-
"<extra_id_44>",
|
| 902 |
-
"<extra_id_43>",
|
| 903 |
-
"<extra_id_42>",
|
| 904 |
-
"<extra_id_41>",
|
| 905 |
-
"<extra_id_40>",
|
| 906 |
-
"<extra_id_39>",
|
| 907 |
-
"<extra_id_38>",
|
| 908 |
-
"<extra_id_37>",
|
| 909 |
-
"<extra_id_36>",
|
| 910 |
-
"<extra_id_35>",
|
| 911 |
-
"<extra_id_34>",
|
| 912 |
-
"<extra_id_33>",
|
| 913 |
-
"<extra_id_32>",
|
| 914 |
-
"<extra_id_31>",
|
| 915 |
-
"<extra_id_30>",
|
| 916 |
-
"<extra_id_29>",
|
| 917 |
-
"<extra_id_28>",
|
| 918 |
-
"<extra_id_27>",
|
| 919 |
-
"<extra_id_26>",
|
| 920 |
-
"<extra_id_25>",
|
| 921 |
-
"<extra_id_24>",
|
| 922 |
-
"<extra_id_23>",
|
| 923 |
-
"<extra_id_22>",
|
| 924 |
-
"<extra_id_21>",
|
| 925 |
-
"<extra_id_20>",
|
| 926 |
-
"<extra_id_19>",
|
| 927 |
-
"<extra_id_18>",
|
| 928 |
-
"<extra_id_17>",
|
| 929 |
-
"<extra_id_16>",
|
| 930 |
-
"<extra_id_15>",
|
| 931 |
-
"<extra_id_14>",
|
| 932 |
-
"<extra_id_13>",
|
| 933 |
-
"<extra_id_12>",
|
| 934 |
-
"<extra_id_11>",
|
| 935 |
-
"<extra_id_10>",
|
| 936 |
-
"<extra_id_9>",
|
| 937 |
-
"<extra_id_8>",
|
| 938 |
-
"<extra_id_7>",
|
| 939 |
-
"<extra_id_6>",
|
| 940 |
-
"<extra_id_5>",
|
| 941 |
-
"<extra_id_4>",
|
| 942 |
-
"<extra_id_3>",
|
| 943 |
-
"<extra_id_2>",
|
| 944 |
-
"<extra_id_1>",
|
| 945 |
-
"<extra_id_0>"
|
| 946 |
],
|
|
|
|
| 947 |
"bos_token": "<s>",
|
| 948 |
-
"clean_up_tokenization_spaces": false,
|
| 949 |
"cls_token": "<s>",
|
| 950 |
"eos_token": "</s>",
|
| 951 |
"errors": "replace",
|
| 952 |
-
"extra_special_tokens":
|
|
|
|
| 953 |
"mask_token": "<mask>",
|
| 954 |
"model_max_length": 512,
|
| 955 |
"pad_token": "<pad>",
|
|
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
+
"additional_special_tokens": [
|
| 4 |
+
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"content": "<extra_id_99>",
|
| 6 |
"lstrip": true,
|
| 7 |
"normalized": true,
|
| 8 |
"rstrip": false,
|
| 9 |
+
"single_word": false
|
|
|
|
| 10 |
},
|
| 11 |
+
{
|
| 12 |
"content": "<extra_id_98>",
|
| 13 |
"lstrip": true,
|
| 14 |
"normalized": true,
|
| 15 |
"rstrip": false,
|
| 16 |
+
"single_word": false
|
|
|
|
| 17 |
},
|
| 18 |
+
{
|
| 19 |
"content": "<extra_id_97>",
|
| 20 |
"lstrip": true,
|
| 21 |
"normalized": true,
|
| 22 |
"rstrip": false,
|
| 23 |
+
"single_word": false
|
|
|
|
| 24 |
},
|
| 25 |
+
{
|
| 26 |
"content": "<extra_id_96>",
|
| 27 |
"lstrip": true,
|
| 28 |
"normalized": true,
|
| 29 |
"rstrip": false,
|
| 30 |
+
"single_word": false
|
|
|
|
| 31 |
},
|
| 32 |
+
{
|
| 33 |
"content": "<extra_id_95>",
|
| 34 |
"lstrip": true,
|
| 35 |
"normalized": true,
|
| 36 |
"rstrip": false,
|
| 37 |
+
"single_word": false
|
|
|
|
| 38 |
},
|
| 39 |
+
{
|
| 40 |
"content": "<extra_id_94>",
|
| 41 |
"lstrip": true,
|
| 42 |
"normalized": true,
|
| 43 |
"rstrip": false,
|
| 44 |
+
"single_word": false
|
|
|
|
| 45 |
},
|
| 46 |
+
{
|
| 47 |
"content": "<extra_id_93>",
|
| 48 |
"lstrip": true,
|
| 49 |
"normalized": true,
|
| 50 |
"rstrip": false,
|
| 51 |
+
"single_word": false
|
|
|
|
| 52 |
},
|
| 53 |
+
{
|
| 54 |
"content": "<extra_id_92>",
|
| 55 |
"lstrip": true,
|
| 56 |
"normalized": true,
|
| 57 |
"rstrip": false,
|
| 58 |
+
"single_word": false
|
|
|
|
| 59 |
},
|
| 60 |
+
{
|
| 61 |
"content": "<extra_id_91>",
|
| 62 |
"lstrip": true,
|
| 63 |
"normalized": true,
|
| 64 |
"rstrip": false,
|
| 65 |
+
"single_word": false
|
|
|
|
| 66 |
},
|
| 67 |
+
{
|
| 68 |
"content": "<extra_id_90>",
|
| 69 |
"lstrip": true,
|
| 70 |
"normalized": true,
|
| 71 |
"rstrip": false,
|
| 72 |
+
"single_word": false
|
|
|
|
| 73 |
},
|
| 74 |
+
{
|
| 75 |
"content": "<extra_id_89>",
|
| 76 |
"lstrip": true,
|
| 77 |
"normalized": true,
|
| 78 |
"rstrip": false,
|
| 79 |
+
"single_word": false
|
|
|
|
| 80 |
},
|
| 81 |
+
{
|
| 82 |
"content": "<extra_id_88>",
|
| 83 |
"lstrip": true,
|
| 84 |
"normalized": true,
|
| 85 |
"rstrip": false,
|
| 86 |
+
"single_word": false
|
|
|
|
| 87 |
},
|
| 88 |
+
{
|
| 89 |
"content": "<extra_id_87>",
|
| 90 |
"lstrip": true,
|
| 91 |
"normalized": true,
|
| 92 |
"rstrip": false,
|
| 93 |
+
"single_word": false
|
|
|
|
| 94 |
},
|
| 95 |
+
{
|
| 96 |
"content": "<extra_id_86>",
|
| 97 |
"lstrip": true,
|
| 98 |
"normalized": true,
|
| 99 |
"rstrip": false,
|
| 100 |
+
"single_word": false
|
|
|
|
| 101 |
},
|
| 102 |
+
{
|
| 103 |
"content": "<extra_id_85>",
|
| 104 |
"lstrip": true,
|
| 105 |
"normalized": true,
|
| 106 |
"rstrip": false,
|
| 107 |
+
"single_word": false
|
|
|
|
| 108 |
},
|
| 109 |
+
{
|
| 110 |
"content": "<extra_id_84>",
|
| 111 |
"lstrip": true,
|
| 112 |
"normalized": true,
|
| 113 |
"rstrip": false,
|
| 114 |
+
"single_word": false
|
|
|
|
| 115 |
},
|
| 116 |
+
{
|
| 117 |
"content": "<extra_id_83>",
|
| 118 |
"lstrip": true,
|
| 119 |
"normalized": true,
|
| 120 |
"rstrip": false,
|
| 121 |
+
"single_word": false
|
|
|
|
| 122 |
},
|
| 123 |
+
{
|
| 124 |
"content": "<extra_id_82>",
|
| 125 |
"lstrip": true,
|
| 126 |
"normalized": true,
|
| 127 |
"rstrip": false,
|
| 128 |
+
"single_word": false
|
|
|
|
| 129 |
},
|
| 130 |
+
{
|
| 131 |
"content": "<extra_id_81>",
|
| 132 |
"lstrip": true,
|
| 133 |
"normalized": true,
|
| 134 |
"rstrip": false,
|
| 135 |
+
"single_word": false
|
|
|
|
| 136 |
},
|
| 137 |
+
{
|
| 138 |
"content": "<extra_id_80>",
|
| 139 |
"lstrip": true,
|
| 140 |
"normalized": true,
|
| 141 |
"rstrip": false,
|
| 142 |
+
"single_word": false
|
|
|
|
| 143 |
},
|
| 144 |
+
{
|
| 145 |
"content": "<extra_id_79>",
|
| 146 |
"lstrip": true,
|
| 147 |
"normalized": true,
|
| 148 |
"rstrip": false,
|
| 149 |
+
"single_word": false
|
|
|
|
| 150 |
},
|
| 151 |
+
{
|
| 152 |
"content": "<extra_id_78>",
|
| 153 |
"lstrip": true,
|
| 154 |
"normalized": true,
|
| 155 |
"rstrip": false,
|
| 156 |
+
"single_word": false
|
|
|
|
| 157 |
},
|
| 158 |
+
{
|
| 159 |
"content": "<extra_id_77>",
|
| 160 |
"lstrip": true,
|
| 161 |
"normalized": true,
|
| 162 |
"rstrip": false,
|
| 163 |
+
"single_word": false
|
|
|
|
| 164 |
},
|
| 165 |
+
{
|
| 166 |
"content": "<extra_id_76>",
|
| 167 |
"lstrip": true,
|
| 168 |
"normalized": true,
|
| 169 |
"rstrip": false,
|
| 170 |
+
"single_word": false
|
|
|
|
| 171 |
},
|
| 172 |
+
{
|
| 173 |
"content": "<extra_id_75>",
|
| 174 |
"lstrip": true,
|
| 175 |
"normalized": true,
|
| 176 |
"rstrip": false,
|
| 177 |
+
"single_word": false
|
|
|
|
| 178 |
},
|
| 179 |
+
{
|
| 180 |
"content": "<extra_id_74>",
|
| 181 |
"lstrip": true,
|
| 182 |
"normalized": true,
|
| 183 |
"rstrip": false,
|
| 184 |
+
"single_word": false
|
|
|
|
| 185 |
},
|
| 186 |
+
{
|
| 187 |
"content": "<extra_id_73>",
|
| 188 |
"lstrip": true,
|
| 189 |
"normalized": true,
|
| 190 |
"rstrip": false,
|
| 191 |
+
"single_word": false
|
|
|
|
| 192 |
},
|
| 193 |
+
{
|
| 194 |
"content": "<extra_id_72>",
|
| 195 |
"lstrip": true,
|
| 196 |
"normalized": true,
|
| 197 |
"rstrip": false,
|
| 198 |
+
"single_word": false
|
|
|
|
| 199 |
},
|
| 200 |
+
{
|
| 201 |
"content": "<extra_id_71>",
|
| 202 |
"lstrip": true,
|
| 203 |
"normalized": true,
|
| 204 |
"rstrip": false,
|
| 205 |
+
"single_word": false
|
|
|
|
| 206 |
},
|
| 207 |
+
{
|
| 208 |
"content": "<extra_id_70>",
|
| 209 |
"lstrip": true,
|
| 210 |
"normalized": true,
|
| 211 |
"rstrip": false,
|
| 212 |
+
"single_word": false
|
|
|
|
| 213 |
},
|
| 214 |
+
{
|
| 215 |
"content": "<extra_id_69>",
|
| 216 |
"lstrip": true,
|
| 217 |
"normalized": true,
|
| 218 |
"rstrip": false,
|
| 219 |
+
"single_word": false
|
|
|
|
| 220 |
},
|
| 221 |
+
{
|
| 222 |
"content": "<extra_id_68>",
|
| 223 |
"lstrip": true,
|
| 224 |
"normalized": true,
|
| 225 |
"rstrip": false,
|
| 226 |
+
"single_word": false
|
|
|
|
| 227 |
},
|
| 228 |
+
{
|
| 229 |
"content": "<extra_id_67>",
|
| 230 |
"lstrip": true,
|
| 231 |
"normalized": true,
|
| 232 |
"rstrip": false,
|
| 233 |
+
"single_word": false
|
|
|
|
| 234 |
},
|
| 235 |
+
{
|
| 236 |
"content": "<extra_id_66>",
|
| 237 |
"lstrip": true,
|
| 238 |
"normalized": true,
|
| 239 |
"rstrip": false,
|
| 240 |
+
"single_word": false
|
|
|
|
| 241 |
},
|
| 242 |
+
{
|
| 243 |
"content": "<extra_id_65>",
|
| 244 |
"lstrip": true,
|
| 245 |
"normalized": true,
|
| 246 |
"rstrip": false,
|
| 247 |
+
"single_word": false
|
|
|
|
| 248 |
},
|
| 249 |
+
{
|
| 250 |
"content": "<extra_id_64>",
|
| 251 |
"lstrip": true,
|
| 252 |
"normalized": true,
|
| 253 |
"rstrip": false,
|
| 254 |
+
"single_word": false
|
|
|
|
| 255 |
},
|
| 256 |
+
{
|
| 257 |
"content": "<extra_id_63>",
|
| 258 |
"lstrip": true,
|
| 259 |
"normalized": true,
|
| 260 |
"rstrip": false,
|
| 261 |
+
"single_word": false
|
|
|
|
| 262 |
},
|
| 263 |
+
{
|
| 264 |
"content": "<extra_id_62>",
|
| 265 |
"lstrip": true,
|
| 266 |
"normalized": true,
|
| 267 |
"rstrip": false,
|
| 268 |
+
"single_word": false
|
|
|
|
| 269 |
},
|
| 270 |
+
{
|
| 271 |
"content": "<extra_id_61>",
|
| 272 |
"lstrip": true,
|
| 273 |
"normalized": true,
|
| 274 |
"rstrip": false,
|
| 275 |
+
"single_word": false
|
|
|
|
| 276 |
},
|
| 277 |
+
{
|
| 278 |
"content": "<extra_id_60>",
|
| 279 |
"lstrip": true,
|
| 280 |
"normalized": true,
|
| 281 |
"rstrip": false,
|
| 282 |
+
"single_word": false
|
|
|
|
| 283 |
},
|
| 284 |
+
{
|
| 285 |
"content": "<extra_id_59>",
|
| 286 |
"lstrip": true,
|
| 287 |
"normalized": true,
|
| 288 |
"rstrip": false,
|
| 289 |
+
"single_word": false
|
|
|
|
| 290 |
},
|
| 291 |
+
{
|
| 292 |
"content": "<extra_id_58>",
|
| 293 |
"lstrip": true,
|
| 294 |
"normalized": true,
|
| 295 |
"rstrip": false,
|
| 296 |
+
"single_word": false
|
|
|
|
| 297 |
},
|
| 298 |
+
{
|
| 299 |
"content": "<extra_id_57>",
|
| 300 |
"lstrip": true,
|
| 301 |
"normalized": true,
|
| 302 |
"rstrip": false,
|
| 303 |
+
"single_word": false
|
|
|
|
| 304 |
},
|
| 305 |
+
{
|
| 306 |
"content": "<extra_id_56>",
|
| 307 |
"lstrip": true,
|
| 308 |
"normalized": true,
|
| 309 |
"rstrip": false,
|
| 310 |
+
"single_word": false
|
|
|
|
| 311 |
},
|
| 312 |
+
{
|
| 313 |
"content": "<extra_id_55>",
|
| 314 |
"lstrip": true,
|
| 315 |
"normalized": true,
|
| 316 |
"rstrip": false,
|
| 317 |
+
"single_word": false
|
|
|
|
| 318 |
},
|
| 319 |
+
{
|
| 320 |
"content": "<extra_id_54>",
|
| 321 |
"lstrip": true,
|
| 322 |
"normalized": true,
|
| 323 |
"rstrip": false,
|
| 324 |
+
"single_word": false
|
|
|
|
| 325 |
},
|
| 326 |
+
{
|
| 327 |
"content": "<extra_id_53>",
|
| 328 |
"lstrip": true,
|
| 329 |
"normalized": true,
|
| 330 |
"rstrip": false,
|
| 331 |
+
"single_word": false
|
|
|
|
| 332 |
},
|
| 333 |
+
{
|
| 334 |
"content": "<extra_id_52>",
|
| 335 |
"lstrip": true,
|
| 336 |
"normalized": true,
|
| 337 |
"rstrip": false,
|
| 338 |
+
"single_word": false
|
|
|
|
| 339 |
},
|
| 340 |
+
{
|
| 341 |
"content": "<extra_id_51>",
|
| 342 |
"lstrip": true,
|
| 343 |
"normalized": true,
|
| 344 |
"rstrip": false,
|
| 345 |
+
"single_word": false
|
|
|
|
| 346 |
},
|
| 347 |
+
{
|
| 348 |
"content": "<extra_id_50>",
|
| 349 |
"lstrip": true,
|
| 350 |
"normalized": true,
|
| 351 |
"rstrip": false,
|
| 352 |
+
"single_word": false
|
|
|
|
| 353 |
},
|
| 354 |
+
{
|
| 355 |
"content": "<extra_id_49>",
|
| 356 |
"lstrip": true,
|
| 357 |
"normalized": true,
|
| 358 |
"rstrip": false,
|
| 359 |
+
"single_word": false
|
|
|
|
| 360 |
},
|
| 361 |
+
{
|
| 362 |
"content": "<extra_id_48>",
|
| 363 |
"lstrip": true,
|
| 364 |
"normalized": true,
|
| 365 |
"rstrip": false,
|
| 366 |
+
"single_word": false
|
|
|
|
| 367 |
},
|
| 368 |
+
{
|
| 369 |
"content": "<extra_id_47>",
|
| 370 |
"lstrip": true,
|
| 371 |
"normalized": true,
|
| 372 |
"rstrip": false,
|
| 373 |
+
"single_word": false
|
|
|
|
| 374 |
},
|
| 375 |
+
{
|
| 376 |
"content": "<extra_id_46>",
|
| 377 |
"lstrip": true,
|
| 378 |
"normalized": true,
|
| 379 |
"rstrip": false,
|
| 380 |
+
"single_word": false
|
|
|
|
| 381 |
},
|
| 382 |
+
{
|
| 383 |
"content": "<extra_id_45>",
|
| 384 |
"lstrip": true,
|
| 385 |
"normalized": true,
|
| 386 |
"rstrip": false,
|
| 387 |
+
"single_word": false
|
|
|
|
| 388 |
},
|
| 389 |
+
{
|
| 390 |
"content": "<extra_id_44>",
|
| 391 |
"lstrip": true,
|
| 392 |
"normalized": true,
|
| 393 |
"rstrip": false,
|
| 394 |
+
"single_word": false
|
|
|
|
| 395 |
},
|
| 396 |
+
{
|
| 397 |
"content": "<extra_id_43>",
|
| 398 |
"lstrip": true,
|
| 399 |
"normalized": true,
|
| 400 |
"rstrip": false,
|
| 401 |
+
"single_word": false
|
|
|
|
| 402 |
},
|
| 403 |
+
{
|
| 404 |
"content": "<extra_id_42>",
|
| 405 |
"lstrip": true,
|
| 406 |
"normalized": true,
|
| 407 |
"rstrip": false,
|
| 408 |
+
"single_word": false
|
|
|
|
| 409 |
},
|
| 410 |
+
{
|
| 411 |
"content": "<extra_id_41>",
|
| 412 |
"lstrip": true,
|
| 413 |
"normalized": true,
|
| 414 |
"rstrip": false,
|
| 415 |
+
"single_word": false
|
|
|
|
| 416 |
},
|
| 417 |
+
{
|
| 418 |
"content": "<extra_id_40>",
|
| 419 |
"lstrip": true,
|
| 420 |
"normalized": true,
|
| 421 |
"rstrip": false,
|
| 422 |
+
"single_word": false
|
|
|
|
| 423 |
},
|
| 424 |
+
{
|
| 425 |
"content": "<extra_id_39>",
|
| 426 |
"lstrip": true,
|
| 427 |
"normalized": true,
|
| 428 |
"rstrip": false,
|
| 429 |
+
"single_word": false
|
|
|
|
| 430 |
},
|
| 431 |
+
{
|
| 432 |
"content": "<extra_id_38>",
|
| 433 |
"lstrip": true,
|
| 434 |
"normalized": true,
|
| 435 |
"rstrip": false,
|
| 436 |
+
"single_word": false
|
|
|
|
| 437 |
},
|
| 438 |
+
{
|
| 439 |
"content": "<extra_id_37>",
|
| 440 |
"lstrip": true,
|
| 441 |
"normalized": true,
|
| 442 |
"rstrip": false,
|
| 443 |
+
"single_word": false
|
|
|
|
| 444 |
},
|
| 445 |
+
{
|
| 446 |
"content": "<extra_id_36>",
|
| 447 |
"lstrip": true,
|
| 448 |
"normalized": true,
|
| 449 |
"rstrip": false,
|
| 450 |
+
"single_word": false
|
|
|
|
| 451 |
},
|
| 452 |
+
{
|
| 453 |
"content": "<extra_id_35>",
|
| 454 |
"lstrip": true,
|
| 455 |
"normalized": true,
|
| 456 |
"rstrip": false,
|
| 457 |
+
"single_word": false
|
|
|
|
| 458 |
},
|
| 459 |
+
{
|
| 460 |
"content": "<extra_id_34>",
|
| 461 |
"lstrip": true,
|
| 462 |
"normalized": true,
|
| 463 |
"rstrip": false,
|
| 464 |
+
"single_word": false
|
|
|
|
| 465 |
},
|
| 466 |
+
{
|
| 467 |
"content": "<extra_id_33>",
|
| 468 |
"lstrip": true,
|
| 469 |
"normalized": true,
|
| 470 |
"rstrip": false,
|
| 471 |
+
"single_word": false
|
|
|
|
| 472 |
},
|
| 473 |
+
{
|
| 474 |
"content": "<extra_id_32>",
|
| 475 |
"lstrip": true,
|
| 476 |
"normalized": true,
|
| 477 |
"rstrip": false,
|
| 478 |
+
"single_word": false
|
|
|
|
| 479 |
},
|
| 480 |
+
{
|
| 481 |
"content": "<extra_id_31>",
|
| 482 |
"lstrip": true,
|
| 483 |
"normalized": true,
|
| 484 |
"rstrip": false,
|
| 485 |
+
"single_word": false
|
|
|
|
| 486 |
},
|
| 487 |
+
{
|
| 488 |
"content": "<extra_id_30>",
|
| 489 |
"lstrip": true,
|
| 490 |
"normalized": true,
|
| 491 |
"rstrip": false,
|
| 492 |
+
"single_word": false
|
|
|
|
| 493 |
},
|
| 494 |
+
{
|
| 495 |
"content": "<extra_id_29>",
|
| 496 |
"lstrip": true,
|
| 497 |
"normalized": true,
|
| 498 |
"rstrip": false,
|
| 499 |
+
"single_word": false
|
|
|
|
| 500 |
},
|
| 501 |
+
{
|
| 502 |
"content": "<extra_id_28>",
|
| 503 |
"lstrip": true,
|
| 504 |
"normalized": true,
|
| 505 |
"rstrip": false,
|
| 506 |
+
"single_word": false
|
|
|
|
| 507 |
},
|
| 508 |
+
{
|
| 509 |
"content": "<extra_id_27>",
|
| 510 |
"lstrip": true,
|
| 511 |
"normalized": true,
|
| 512 |
"rstrip": false,
|
| 513 |
+
"single_word": false
|
|
|
|
| 514 |
},
|
| 515 |
+
{
|
| 516 |
"content": "<extra_id_26>",
|
| 517 |
"lstrip": true,
|
| 518 |
"normalized": true,
|
| 519 |
"rstrip": false,
|
| 520 |
+
"single_word": false
|
|
|
|
| 521 |
},
|
| 522 |
+
{
|
| 523 |
"content": "<extra_id_25>",
|
| 524 |
"lstrip": true,
|
| 525 |
"normalized": true,
|
| 526 |
"rstrip": false,
|
| 527 |
+
"single_word": false
|
|
|
|
| 528 |
},
|
| 529 |
+
{
|
| 530 |
"content": "<extra_id_24>",
|
| 531 |
"lstrip": true,
|
| 532 |
"normalized": true,
|
| 533 |
"rstrip": false,
|
| 534 |
+
"single_word": false
|
|
|
|
| 535 |
},
|
| 536 |
+
{
|
| 537 |
"content": "<extra_id_23>",
|
| 538 |
"lstrip": true,
|
| 539 |
"normalized": true,
|
| 540 |
"rstrip": false,
|
| 541 |
+
"single_word": false
|
|
|
|
| 542 |
},
|
| 543 |
+
{
|
| 544 |
"content": "<extra_id_22>",
|
| 545 |
"lstrip": true,
|
| 546 |
"normalized": true,
|
| 547 |
"rstrip": false,
|
| 548 |
+
"single_word": false
|
|
|
|
| 549 |
},
|
| 550 |
+
{
|
| 551 |
"content": "<extra_id_21>",
|
| 552 |
"lstrip": true,
|
| 553 |
"normalized": true,
|
| 554 |
"rstrip": false,
|
| 555 |
+
"single_word": false
|
|
|
|
| 556 |
},
|
| 557 |
+
{
|
| 558 |
"content": "<extra_id_20>",
|
| 559 |
"lstrip": true,
|
| 560 |
"normalized": true,
|
| 561 |
"rstrip": false,
|
| 562 |
+
"single_word": false
|
|
|
|
| 563 |
},
|
| 564 |
+
{
|
| 565 |
"content": "<extra_id_19>",
|
| 566 |
"lstrip": true,
|
| 567 |
"normalized": true,
|
| 568 |
"rstrip": false,
|
| 569 |
+
"single_word": false
|
|
|
|
| 570 |
},
|
| 571 |
+
{
|
| 572 |
"content": "<extra_id_18>",
|
| 573 |
"lstrip": true,
|
| 574 |
"normalized": true,
|
| 575 |
"rstrip": false,
|
| 576 |
+
"single_word": false
|
|
|
|
| 577 |
},
|
| 578 |
+
{
|
| 579 |
"content": "<extra_id_17>",
|
| 580 |
"lstrip": true,
|
| 581 |
"normalized": true,
|
| 582 |
"rstrip": false,
|
| 583 |
+
"single_word": false
|
|
|
|
| 584 |
},
|
| 585 |
+
{
|
| 586 |
"content": "<extra_id_16>",
|
| 587 |
"lstrip": true,
|
| 588 |
"normalized": true,
|
| 589 |
"rstrip": false,
|
| 590 |
+
"single_word": false
|
|
|
|
| 591 |
},
|
| 592 |
+
{
|
| 593 |
"content": "<extra_id_15>",
|
| 594 |
"lstrip": true,
|
| 595 |
"normalized": true,
|
| 596 |
"rstrip": false,
|
| 597 |
+
"single_word": false
|
|
|
|
| 598 |
},
|
| 599 |
+
{
|
| 600 |
"content": "<extra_id_14>",
|
| 601 |
"lstrip": true,
|
| 602 |
"normalized": true,
|
| 603 |
"rstrip": false,
|
| 604 |
+
"single_word": false
|
|
|
|
| 605 |
},
|
| 606 |
+
{
|
| 607 |
"content": "<extra_id_13>",
|
| 608 |
"lstrip": true,
|
| 609 |
"normalized": true,
|
| 610 |
"rstrip": false,
|
| 611 |
+
"single_word": false
|
|
|
|
| 612 |
},
|
| 613 |
+
{
|
| 614 |
"content": "<extra_id_12>",
|
| 615 |
"lstrip": true,
|
| 616 |
"normalized": true,
|
| 617 |
"rstrip": false,
|
| 618 |
+
"single_word": false
|
|
|
|
| 619 |
},
|
| 620 |
+
{
|
| 621 |
"content": "<extra_id_11>",
|
| 622 |
"lstrip": true,
|
| 623 |
"normalized": true,
|
| 624 |
"rstrip": false,
|
| 625 |
+
"single_word": false
|
|
|
|
| 626 |
},
|
| 627 |
+
{
|
| 628 |
"content": "<extra_id_10>",
|
| 629 |
"lstrip": true,
|
| 630 |
"normalized": true,
|
| 631 |
"rstrip": false,
|
| 632 |
+
"single_word": false
|
|
|
|
| 633 |
},
|
| 634 |
+
{
|
| 635 |
"content": "<extra_id_9>",
|
| 636 |
"lstrip": true,
|
| 637 |
"normalized": true,
|
| 638 |
"rstrip": false,
|
| 639 |
+
"single_word": false
|
|
|
|
| 640 |
},
|
| 641 |
+
{
|
| 642 |
"content": "<extra_id_8>",
|
| 643 |
"lstrip": true,
|
| 644 |
"normalized": true,
|
| 645 |
"rstrip": false,
|
| 646 |
+
"single_word": false
|
|
|
|
| 647 |
},
|
| 648 |
+
{
|
| 649 |
"content": "<extra_id_7>",
|
| 650 |
"lstrip": true,
|
| 651 |
"normalized": true,
|
| 652 |
"rstrip": false,
|
| 653 |
+
"single_word": false
|
|
|
|
| 654 |
},
|
| 655 |
+
{
|
| 656 |
"content": "<extra_id_6>",
|
| 657 |
"lstrip": true,
|
| 658 |
"normalized": true,
|
| 659 |
"rstrip": false,
|
| 660 |
+
"single_word": false
|
|
|
|
| 661 |
},
|
| 662 |
+
{
|
| 663 |
"content": "<extra_id_5>",
|
| 664 |
"lstrip": true,
|
| 665 |
"normalized": true,
|
| 666 |
"rstrip": false,
|
| 667 |
+
"single_word": false
|
|
|
|
| 668 |
},
|
| 669 |
+
{
|
| 670 |
"content": "<extra_id_4>",
|
| 671 |
"lstrip": true,
|
| 672 |
"normalized": true,
|
| 673 |
"rstrip": false,
|
| 674 |
+
"single_word": false
|
|
|
|
| 675 |
},
|
| 676 |
+
{
|
| 677 |
"content": "<extra_id_3>",
|
| 678 |
"lstrip": true,
|
| 679 |
"normalized": true,
|
| 680 |
"rstrip": false,
|
| 681 |
+
"single_word": false
|
|
|
|
| 682 |
},
|
| 683 |
+
{
|
| 684 |
"content": "<extra_id_2>",
|
| 685 |
"lstrip": true,
|
| 686 |
"normalized": true,
|
| 687 |
"rstrip": false,
|
| 688 |
+
"single_word": false
|
|
|
|
| 689 |
},
|
| 690 |
+
{
|
| 691 |
"content": "<extra_id_1>",
|
| 692 |
"lstrip": true,
|
| 693 |
"normalized": true,
|
| 694 |
"rstrip": false,
|
| 695 |
+
"single_word": false
|
|
|
|
| 696 |
},
|
| 697 |
+
{
|
| 698 |
"content": "<extra_id_0>",
|
| 699 |
"lstrip": true,
|
| 700 |
"normalized": true,
|
| 701 |
"rstrip": false,
|
| 702 |
+
"single_word": false
|
|
|
|
| 703 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 704 |
],
|
| 705 |
+
"backend": "tokenizers",
|
| 706 |
"bos_token": "<s>",
|
|
|
|
| 707 |
"cls_token": "<s>",
|
| 708 |
"eos_token": "</s>",
|
| 709 |
"errors": "replace",
|
| 710 |
+
"extra_special_tokens": [],
|
| 711 |
+
"is_local": false,
|
| 712 |
"mask_token": "<mask>",
|
| 713 |
"model_max_length": 512,
|
| 714 |
"pad_token": "<pad>",
|