Hugging Face's logo

likhonsheikh
/
Sheikh-ABF

Text Generation
Transformers
Safetensors
Bengali
gpt2
Bangla
nlp
decoder-only
causal-lm
Lora
code-generation
agentic-ai
from-scratch
text-generation-inference
Model card Files
xet
Community
likhonsheikh commited on
Commit
3445b9d
·
verified ·
1 Parent(s): 8767317

Upload merged LoRA model and tokenizer

Browse files
Files changed (6) hide show
  1. config.json +39 -0
  2. generation_config.json +7 -0
  3. model.safetensors +3 -0
  4. special_tokens_map.json +6 -0
  5. tokenizer.json +1057 -0
  6. tokenizer_config.json +44 -0
config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 0,
8
+ "dtype": "float32",
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 1,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 512,
16
+ "n_head": 8,
17
+ "n_inner": null,
18
+ "n_layer": 8,
19
+ "n_positions": 1024,
20
+ "pad_token_id": 2,
21
+ "reorder_and_upcast_attn": false,
22
+ "resid_pdrop": 0.1,
23
+ "scale_attn_by_inverse_layer_idx": false,
24
+ "scale_attn_weights": true,
25
+ "summary_activation": null,
26
+ "summary_first_dropout": 0.1,
27
+ "summary_proj_to_labels": true,
28
+ "summary_type": "cls_index",
29
+ "summary_use_proj": true,
30
+ "task_specific_params": {
31
+ "text-generation": {
32
+ "do_sample": true,
33
+ "max_length": 50
34
+ }
35
+ },
36
+ "transformers_version": "4.57.3",
37
+ "use_cache": true,
38
+ "vocab_size": 32000
39
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 1,
5
+ "pad_token_id": 2,
6
+ "transformers_version": "4.57.3"
7
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83cb2653702e35987827ae35715f05fe107e13469fdf418b9d421b003622df5f
3
+ size 168523504
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<bos>",
3
+ "eos_token": "<eos>",
4
+ "pad_token": "<pad>",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer.json ADDED
@@ -0,0 +1,1057 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": "1.0",
3
+ "truncation": null,
4
+ "padding": null,
5
+ "added_tokens": [
6
+ {
7
+ "id": 0,
8
+ "content": "<bos>",
9
+ "single_word": false,
10
+ "lstrip": false,
11
+ "rstrip": false,
12
+ "normalized": false,
13
+ "special": true
14
+ },
15
+ {
16
+ "id": 1,
17
+ "content": "<eos>",
18
+ "single_word": false,
19
+ "lstrip": false,
20
+ "rstrip": false,
21
+ "normalized": false,
22
+ "special": true
23
+ },
24
+ {
25
+ "id": 2,
26
+ "content": "<pad>",
27
+ "single_word": false,
28
+ "lstrip": false,
29
+ "rstrip": false,
30
+ "normalized": false,
31
+ "special": true
32
+ },
33
+ {
34
+ "id": 405,
35
+ "content": "<unk>",
36
+ "single_word": false,
37
+ "lstrip": false,
38
+ "rstrip": false,
39
+ "normalized": false,
40
+ "special": true
41
+ }
42
+ ],
43
+ "normalizer": null,
44
+ "pre_tokenizer": {
45
+ "type": "ByteLevel",
46
+ "add_prefix_space": false,
47
+ "trim_offsets": true,
48
+ "use_regex": true
49
+ },
50
+ "post_processor": {
51
+ "type": "ByteLevel",
52
+ "add_prefix_space": true,
53
+ "trim_offsets": false,
54
+ "use_regex": true
55
+ },
56
+ "decoder": {
57
+ "type": "ByteLevel",
58
+ "add_prefix_space": true,
59
+ "trim_offsets": true,
60
+ "use_regex": true
61
+ },
62
+ "model": {
63
+ "type": "BPE",
64
+ "dropout": null,
65
+ "unk_token": null,
66
+ "continuing_subword_prefix": "",
67
+ "end_of_word_suffix": "",
68
+ "fuse_unk": false,
69
+ "byte_fallback": false,
70
+ "ignore_merges": false,
71
+ "vocab": {
72
+ "<bos>": 0,
73
+ "<eos>": 1,
74
+ "<pad>": 2,
75
+ "<think>": 3,
76
+ "</think>": 4,
77
+ "!": 5,
78
+ "\"": 6,
79
+ "#": 7,
80
+ "$": 8,
81
+ "%": 9,
82
+ "&": 10,
83
+ "'": 11,
84
+ "(": 12,
85
+ ")": 13,
86
+ "*": 14,
87
+ "+": 15,
88
+ ",": 16,
89
+ "-": 17,
90
+ ".": 18,
91
+ "/": 19,
92
+ "0": 20,
93
+ "1": 21,
94
+ "2": 22,
95
+ "3": 23,
96
+ "4": 24,
97
+ "5": 25,
98
+ "6": 26,
99
+ "7": 27,
100
+ "8": 28,
101
+ "9": 29,
102
+ ":": 30,
103
+ ";": 31,
104
+ "<": 32,
105
+ "=": 33,
106
+ ">": 34,
107
+ "?": 35,
108
+ "@": 36,
109
+ "A": 37,
110
+ "B": 38,
111
+ "C": 39,
112
+ "D": 40,
113
+ "E": 41,
114
+ "F": 42,
115
+ "G": 43,
116
+ "H": 44,
117
+ "I": 45,
118
+ "J": 46,
119
+ "K": 47,
120
+ "L": 48,
121
+ "M": 49,
122
+ "N": 50,
123
+ "O": 51,
124
+ "P": 52,
125
+ "Q": 53,
126
+ "R": 54,
127
+ "S": 55,
128
+ "T": 56,
129
+ "U": 57,
130
+ "V": 58,
131
+ "W": 59,
132
+ "X": 60,
133
+ "Y": 61,
134
+ "Z": 62,
135
+ "[": 63,
136
+ "\\": 64,
137
+ "]": 65,
138
+ "^": 66,
139
+ "_": 67,
140
+ "`": 68,
141
+ "a": 69,
142
+ "b": 70,
143
+ "c": 71,
144
+ "d": 72,
145
+ "e": 73,
146
+ "f": 74,
147
+ "g": 75,
148
+ "h": 76,
149
+ "i": 77,
150
+ "j": 78,
151
+ "k": 79,
152
+ "l": 80,
153
+ "m": 81,
154
+ "n": 82,
155
+ "o": 83,
156
+ "p": 84,
157
+ "q": 85,
158
+ "r": 86,
159
+ "s": 87,
160
+ "t": 88,
161
+ "u": 89,
162
+ "v": 90,
163
+ "w": 91,
164
+ "x": 92,
165
+ "y": 93,
166
+ "z": 94,
167
+ "{": 95,
168
+ "|": 96,
169
+ "}": 97,
170
+ "~": 98,
171
+ "¡": 99,
172
+ "¢": 100,
173
+ "£": 101,
174
+ "¤": 102,
175
+ "¥": 103,
176
+ "¦": 104,
177
+ "§": 105,
178
+ "¨": 106,
179
+ "©": 107,
180
+ "ª": 108,
181
+ "«": 109,
182
+ "¬": 110,
183
+ "®": 111,
184
+ "¯": 112,
185
+ "°": 113,
186
+ "±": 114,
187
+ "²": 115,
188
+ "³": 116,
189
+ "´": 117,
190
+ "µ": 118,
191
+ "¶": 119,
192
+ "·": 120,
193
+ "¸": 121,
194
+ "¹": 122,
195
+ "º": 123,
196
+ "»": 124,
197
+ "¼": 125,
198
+ "½": 126,
199
+ "¾": 127,
200
+ "¿": 128,
201
+ "À": 129,
202
+ "Á": 130,
203
+ "Â": 131,
204
+ "Ã": 132,
205
+ "Ä": 133,
206
+ "Å": 134,
207
+ "Æ": 135,
208
+ "Ç": 136,
209
+ "È": 137,
210
+ "É": 138,
211
+ "Ê": 139,
212
+ "Ë": 140,
213
+ "Ì": 141,
214
+ "Í": 142,
215
+ "Î": 143,
216
+ "Ï": 144,
217
+ "Ð": 145,
218
+ "Ñ": 146,
219
+ "Ò": 147,
220
+ "Ó": 148,
221
+ "Ô": 149,
222
+ "Õ": 150,
223
+ "Ö": 151,
224
+ "×": 152,
225
+ "Ø": 153,
226
+ "Ù": 154,
227
+ "Ú": 155,
228
+ "Û": 156,
229
+ "Ü": 157,
230
+ "Ý": 158,
231
+ "Þ": 159,
232
+ "ß": 160,
233
+ "à": 161,
234
+ "á": 162,
235
+ "â": 163,
236
+ "ã": 164,
237
+ "ä": 165,
238
+ "å": 166,
239
+ "æ": 167,
240
+ "ç": 168,
241
+ "è": 169,
242
+ "é": 170,
243
+ "ê": 171,
244
+ "ë": 172,
245
+ "ì": 173,
246
+ "í": 174,
247
+ "î": 175,
248
+ "ï": 176,
249
+ "ð": 177,
250
+ "ñ": 178,
251
+ "ò": 179,
252
+ "ó": 180,
253
+ "ô": 181,
254
+ "õ": 182,
255
+ "ö": 183,
256
+ "÷": 184,
257
+ "ø": 185,
258
+ "ù": 186,
259
+ "ú": 187,
260
+ "û": 188,
261
+ "ü": 189,
262
+ "ý": 190,
263
+ "þ": 191,
264
+ "ÿ": 192,
265
+ "Ā": 193,
266
+ "ā": 194,
267
+ "Ă": 195,
268
+ "ă": 196,
269
+ "Ą": 197,
270
+ "ą": 198,
271
+ "Ć": 199,
272
+ "ć": 200,
273
+ "Ĉ": 201,
274
+ "ĉ": 202,
275
+ "Ċ": 203,
276
+ "ċ": 204,
277
+ "Č": 205,
278
+ "č": 206,
279
+ "Ď": 207,
280
+ "ď": 208,
281
+ "Đ": 209,
282
+ "đ": 210,
283
+ "Ē": 211,
284
+ "ē": 212,
285
+ "Ĕ": 213,
286
+ "ĕ": 214,
287
+ "Ė": 215,
288
+ "ė": 216,
289
+ "Ę": 217,
290
+ "ę": 218,
291
+ "Ě": 219,
292
+ "ě": 220,
293
+ "Ĝ": 221,
294
+ "ĝ": 222,
295
+ "Ğ": 223,
296
+ "ğ": 224,
297
+ "Ġ": 225,
298
+ "ġ": 226,
299
+ "Ģ": 227,
300
+ "ģ": 228,
301
+ "Ĥ": 229,
302
+ "ĥ": 230,
303
+ "Ħ": 231,
304
+ "ħ": 232,
305
+ "Ĩ": 233,
306
+ "ĩ": 234,
307
+ "Ī": 235,
308
+ "ī": 236,
309
+ "Ĭ": 237,
310
+ "ĭ": 238,
311
+ "Į": 239,
312
+ "į": 240,
313
+ "İ": 241,
314
+ "ı": 242,
315
+ "IJ": 243,
316
+ "ij": 244,
317
+ "Ĵ": 245,
318
+ "ĵ": 246,
319
+ "Ķ": 247,
320
+ "ķ": 248,
321
+ "ĸ": 249,
322
+ "Ĺ": 250,
323
+ "ĺ": 251,
324
+ "Ļ": 252,
325
+ "ļ": 253,
326
+ "Ľ": 254,
327
+ "ľ": 255,
328
+ "Ŀ": 256,
329
+ "ŀ": 257,
330
+ "Ł": 258,
331
+ "ł": 259,
332
+ "Ń": 260,
333
+ "à¦": 261,
334
+ "à§": 262,
335
+ "Ġà¦": 263,
336
+ "া": 264,
337
+ "র": 265,
338
+ "à§ĩ": 266,
339
+ "à§į": 267,
340
+ "ি": 268,
341
+ "ন": 269,
342
+ "ত": 270,
343
+ "ল": 271,
344
+ "য": 272,
345
+ "ব": 273,
346
+ "à¦ķ": 274,
347
+ "¥¤": 275,
348
+ "।": 276,
349
+ "দ": 277,
350
+ "়": 278,
351
+ "ম": 279,
352
+ "à§ģ": 280,
353
+ "Ġস": 281,
354
+ "শ": 282,
355
+ "Ġà¦ķ": 283,
356
+ "Ġব": 284,
357
+ "Ġপ": 285,
358
+ "à§Ģ": 286,
359
+ "স": 287,
360
+ "à¦Ĥ": 288,
361
+ "প": 289,
362
+ "Ġà¦ı": 290,
363
+ "হ": 291,
364
+ "à¦Ł": 292,
365
+ "Ġà¦Ĩ": 293,
366
+ "Ġà¦ķর": 294,
367
+ "ড": 295,
368
+ "ণ": 296,
369
+ "Ġà¦ħ": 297,
370
+ "ধ": 298,
371
+ "à§ĭ": 299,
372
+ "াà¦Ĥ": 300,
373
+ "à¦ľ": 301,
374
+ "Ġà¦ī": 302,
375
+ "hi": 303,
376
+ "nk": 304,
377
+ "thi": 305,
378
+ "ষ": 306,
379
+ "à¦Ľ": 307,
380
+ "Ġন": 308,
381
+ "Ġম": 309,
382
+ "à§ĩ।": 310,
383
+ "think": 311,
384
+ "Ġà¦ľ": 312,
385
+ "à¦Ĺ": 313,
386
+ "à¦ļ": 314,
387
+ "Ġত": 315,
388
+ "Ġদ": 316,
389
+ "Ġহ": 317,
390
+ "থ": 318,
391
+ "à¦ĩ": 319,
392
+ "à¦Ń": 320,
393
+ "à§ĥ": 321,
394
+ "Ġà¦ıà¦ķ": 322,
395
+ "à¦ĸ": 323,
396
+ "à¦ł": 324,
397
+ "à§Ĥ": 325,
398
+ "Ġà¦ĵ": 326,
399
+ "রণ": 327,
400
+ "Ġà¦Ĩম": 328,
401
+ "Ġà¦ħন": 329,
402
+ "</": 330,
403
+ "Ġফ": 331,
404
+ "Ġর": 332,
405
+ "Ġশ": 333,
406
+ "়া": 334,
407
+ "Ġসà¦ķ": 335,
408
+ "Ġ'": 336,
409
+ "ফ": 337,
410
+ "à¦Ĩ": 338,
411
+ "Ġà¦Ĺ": 339,
412
+ "Ġà¦ļ": 340,
413
+ "Ġà¦Ń": 341,
414
+ "।</": 342,
415
+ "়à§ĩ": 343,
416
+ "়ি": 344,
417
+ "Ġপর": 345,
418
+ "Ġà¦ıব": 346,
419
+ "Ġহয": 347,
420
+ "Ġà¦ıà¦ķà¦Ł": 348,
421
+ "Ġ<": 349,
422
+ "à¦ı": 350,
423
+ "à§Į": 351,
424
+ "Ġয": 352,
425
+ "ি?": 353,
426
+ "যব": 354,
427
+ "দর": 355,
428
+ "Ġসম": 356,
429
+ "পর": 357,
430
+ "পন": 358,
431
+ "Ġà¦ıà¦ĩ": 359,
432
+ "Ġà¦Ĩপন": 360,
433
+ "à¦ľà¦¨": 361,
434
+ "Ġà¦īà¦ļ": 362,
435
+ "Ġহব": 363,
436
+ "à¦ĸত": 364,
437
+ "Ġà¦ĵà¦ł": 365,
438
+ "à¦Ĩম": 366,
439
+ "?'": 367,
440
+ "ĠA": 368,
441
+ "ĠB": 369,
442
+ "à¦ģ": 370,
443
+ "à¦ħ": 371,
444
+ "à¦Ļ": 372,
445
+ "Ġধ": 373,
446
+ "Ġà¦Ł": 374,
447
+ "া।": 375,
448
+ "রত": 376,
449
+ "রয": 377,
450
+ "রশ": 378,
451
+ "রহ": 379,
452
+ "à§ĩ?": 380,
453
+ "ি।": 381,
454
+ "তর": 382,
455
+ "তà¦ı": 383,
456
+ "লন": 384,
457
+ "à¦¯à¦Ł": 385,
458
+ "়।": 386,
459
+ "Ġসত": 387,
460
+ "Ġবল": 388,
461
+ "সব": 389,
462
+ "Ġà¦ıà¦Ł": 390,
463
+ "হল": 391,
464
+ "Ġà¦Ĩà¦Ľ": 392,
465
+ "Ġà¦Ĩà¦Ĺ": 393,
466
+ "ধন": 394,
467
+ "à¦ľà¦§": 395,
468
+ "Ġà¦īড": 396,
469
+ "Ġমরণ": 397,
470
+ "Ġà¦ľà¦¨": 398,
471
+ "Ġতথ": 399,
472
+ "Ġফল": 400,
473
+ "Ġসà¦ķল": 401,
474
+ "ফল": 402,
475
+ "তà¦ıব": 403,
476
+ "Ġমরণশ": 404
477
+ },
478
+ "merges": [
479
+ [
480
+ "à",
481
+ "¦"
482
+ ],
483
+ [
484
+ "à",
485
+ "§"
486
+ ],
487
+ [
488
+ "Ġ",
489
+ "à¦"
490
+ ],
491
+ [
492
+ "à¦",
493
+ "¾"
494
+ ],
495
+ [
496
+ "à¦",
497
+ "°"
498
+ ],
499
+ [
500
+ "à§",
501
+ "ĩ"
502
+ ],
503
+ [
504
+ "à§",
505
+ "į"
506
+ ],
507
+ [
508
+ "à¦",
509
+ "¿"
510
+ ],
511
+ [
512
+ "à¦",
513
+ "¨"
514
+ ],
515
+ [
516
+ "à¦",
517
+ "¤"
518
+ ],
519
+ [
520
+ "à¦",
521
+ "²"
522
+ ],
523
+ [
524
+ "à¦",
525
+ "¯"
526
+ ],
527
+ [
528
+ "à¦",
529
+ "¬"
530
+ ],
531
+ [
532
+ "à¦",
533
+ "ķ"
534
+ ],
535
+ [
536
+ "¥",
537
+ "¤"
538
+ ],
539
+ [
540
+ "à",
541
+ "¥¤"
542
+ ],
543
+ [
544
+ "à¦",
545
+ "¦"
546
+ ],
547
+ [
548
+ "à¦",
549
+ "¼"
550
+ ],
551
+ [
552
+ "à¦",
553
+ "®"
554
+ ],
555
+ [
556
+ "à§",
557
+ "ģ"
558
+ ],
559
+ [
560
+ "Ġà¦",
561
+ "¸"
562
+ ],
563
+ [
564
+ "à¦",
565
+ "¶"
566
+ ],
567
+ [
568
+ "Ġà¦",
569
+ "ķ"
570
+ ],
571
+ [
572
+ "Ġà¦",
573
+ "¬"
574
+ ],
575
+ [
576
+ "Ġà¦",
577
+ "ª"
578
+ ],
579
+ [
580
+ "à§",
581
+ "Ģ"
582
+ ],
583
+ [
584
+ "à¦",
585
+ "¸"
586
+ ],
587
+ [
588
+ "à¦",
589
+ "Ĥ"
590
+ ],
591
+ [
592
+ "à¦",
593
+ "ª"
594
+ ],
595
+ [
596
+ "Ġà¦",
597
+ "ı"
598
+ ],
599
+ [
600
+ "à¦",
601
+ "¹"
602
+ ],
603
+ [
604
+ "à¦",
605
+ "Ł"
606
+ ],
607
+ [
608
+ "Ġà¦",
609
+ "Ĩ"
610
+ ],
611
+ [
612
+ "Ġà¦ķ",
613
+ "র"
614
+ ],
615
+ [
616
+ "à¦",
617
+ "¡"
618
+ ],
619
+ [
620
+ "à¦",
621
+ "£"
622
+ ],
623
+ [
624
+ "Ġà¦",
625
+ "ħ"
626
+ ],
627
+ [
628
+ "à¦",
629
+ "§"
630
+ ],
631
+ [
632
+ "à§",
633
+ "ĭ"
634
+ ],
635
+ [
636
+ "া",
637
+ "à¦Ĥ"
638
+ ],
639
+ [
640
+ "à¦",
641
+ "ľ"
642
+ ],
643
+ [
644
+ "Ġà¦",
645
+ "ī"
646
+ ],
647
+ [
648
+ "h",
649
+ "i"
650
+ ],
651
+ [
652
+ "n",
653
+ "k"
654
+ ],
655
+ [
656
+ "t",
657
+ "hi"
658
+ ],
659
+ [
660
+ "à¦",
661
+ "·"
662
+ ],
663
+ [
664
+ "à¦",
665
+ "Ľ"
666
+ ],
667
+ [
668
+ "Ġà¦",
669
+ "¨"
670
+ ],
671
+ [
672
+ "Ġà¦",
673
+ "®"
674
+ ],
675
+ [
676
+ "à§ĩ",
677
+ "।"
678
+ ],
679
+ [
680
+ "thi",
681
+ "nk"
682
+ ],
683
+ [
684
+ "Ġà¦",
685
+ "ľ"
686
+ ],
687
+ [
688
+ "à¦",
689
+ "Ĺ"
690
+ ],
691
+ [
692
+ "à¦",
693
+ "ļ"
694
+ ],
695
+ [
696
+ "Ġà¦",
697
+ "¤"
698
+ ],
699
+ [
700
+ "Ġà¦",
701
+ "¦"
702
+ ],
703
+ [
704
+ "Ġà¦",
705
+ "¹"
706
+ ],
707
+ [
708
+ "à¦",
709
+ "¥"
710
+ ],
711
+ [
712
+ "à¦",
713
+ "ĩ"
714
+ ],
715
+ [
716
+ "à¦",
717
+ "Ń"
718
+ ],
719
+ [
720
+ "à§",
721
+ "ĥ"
722
+ ],
723
+ [
724
+ "Ġà¦ı",
725
+ "à¦ķ"
726
+ ],
727
+ [
728
+ "à¦",
729
+ "ĸ"
730
+ ],
731
+ [
732
+ "à¦",
733
+ "ł"
734
+ ],
735
+ [
736
+ "à§",
737
+ "Ĥ"
738
+ ],
739
+ [
740
+ "Ġà¦",
741
+ "ĵ"
742
+ ],
743
+ [
744
+ "র",
745
+ "ণ"
746
+ ],
747
+ [
748
+ "Ġà¦Ĩ",
749
+ "ম"
750
+ ],
751
+ [
752
+ "Ġà¦ħ",
753
+ "ন"
754
+ ],
755
+ [
756
+ "<",
757
+ "/"
758
+ ],
759
+ [
760
+ "Ġà¦",
761
+ "«"
762
+ ],
763
+ [
764
+ "Ġà¦",
765
+ "°"
766
+ ],
767
+ [
768
+ "Ġà¦",
769
+ "¶"
770
+ ],
771
+ [
772
+ "়",
773
+ "া"
774
+ ],
775
+ [
776
+ "Ġস",
777
+ "à¦ķ"
778
+ ],
779
+ [
780
+ "Ġ",
781
+ "'"
782
+ ],
783
+ [
784
+ "à¦",
785
+ "«"
786
+ ],
787
+ [
788
+ "à¦",
789
+ "Ĩ"
790
+ ],
791
+ [
792
+ "Ġà¦",
793
+ "Ĺ"
794
+ ],
795
+ [
796
+ "Ġà¦",
797
+ "ļ"
798
+ ],
799
+ [
800
+ "Ġà¦",
801
+ "Ń"
802
+ ],
803
+ [
804
+ "।",
805
+ "</"
806
+ ],
807
+ [
808
+ "়",
809
+ "à§ĩ"
810
+ ],
811
+ [
812
+ "়",
813
+ "ি"
814
+ ],
815
+ [
816
+ "Ġপ",
817
+ "র"
818
+ ],
819
+ [
820
+ "Ġà¦ı",
821
+ "ব"
822
+ ],
823
+ [
824
+ "Ġহ",
825
+ "য"
826
+ ],
827
+ [
828
+ "Ġà¦ıà¦ķ",
829
+ "à¦Ł"
830
+ ],
831
+ [
832
+ "Ġ",
833
+ "<"
834
+ ],
835
+ [
836
+ "à¦",
837
+ "ı"
838
+ ],
839
+ [
840
+ "à§",
841
+ "Į"
842
+ ],
843
+ [
844
+ "Ġà¦",
845
+ "¯"
846
+ ],
847
+ [
848
+ "ি",
849
+ "?"
850
+ ],
851
+ [
852
+ "য",
853
+ "ব"
854
+ ],
855
+ [
856
+ "দ",
857
+ "র"
858
+ ],
859
+ [
860
+ "Ġস",
861
+ "ম"
862
+ ],
863
+ [
864
+ "প",
865
+ "র"
866
+ ],
867
+ [
868
+ "প",
869
+ "ন"
870
+ ],
871
+ [
872
+ "Ġà¦ı",
873
+ "à¦ĩ"
874
+ ],
875
+ [
876
+ "Ġà¦Ĩ",
877
+ "পন"
878
+ ],
879
+ [
880
+ "à¦ľ",
881
+ "ন"
882
+ ],
883
+ [
884
+ "Ġà¦ī",
885
+ "à¦ļ"
886
+ ],
887
+ [
888
+ "Ġহ",
889
+ "ব"
890
+ ],
891
+ [
892
+ "à¦ĸ",
893
+ "ত"
894
+ ],
895
+ [
896
+ "Ġà¦ĵ",
897
+ "à¦ł"
898
+ ],
899
+ [
900
+ "à¦Ĩ",
901
+ "ম"
902
+ ],
903
+ [
904
+ "?",
905
+ "'"
906
+ ],
907
+ [
908
+ "Ġ",
909
+ "A"
910
+ ],
911
+ [
912
+ "Ġ",
913
+ "B"
914
+ ],
915
+ [
916
+ "à¦",
917
+ "ģ"
918
+ ],
919
+ [
920
+ "à¦",
921
+ "ħ"
922
+ ],
923
+ [
924
+ "à¦",
925
+ "Ļ"
926
+ ],
927
+ [
928
+ "Ġà¦",
929
+ "§"
930
+ ],
931
+ [
932
+ "Ġà¦",
933
+ "Ł"
934
+ ],
935
+ [
936
+ "া",
937
+ "।"
938
+ ],
939
+ [
940
+ "র",
941
+ "ত"
942
+ ],
943
+ [
944
+ "র",
945
+ "য"
946
+ ],
947
+ [
948
+ "র",
949
+ "শ"
950
+ ],
951
+ [
952
+ "র",
953
+ "হ"
954
+ ],
955
+ [
956
+ "à§ĩ",
957
+ "?"
958
+ ],
959
+ [
960
+ "ি",
961
+ "।"
962
+ ],
963
+ [
964
+ "ত",
965
+ "র"
966
+ ],
967
+ [
968
+ "ত",
969
+ "à¦ı"
970
+ ],
971
+ [
972
+ "ল",
973
+ "ন"
974
+ ],
975
+ [
976
+ "য",
977
+ "à¦Ł"
978
+ ],
979
+ [
980
+ "়",
981
+ "।"
982
+ ],
983
+ [
984
+ "Ġস",
985
+ "ত"
986
+ ],
987
+ [
988
+ "Ġব",
989
+ "ল"
990
+ ],
991
+ [
992
+ "স",
993
+ "ব"
994
+ ],
995
+ [
996
+ "Ġà¦ı",
997
+ "à¦Ł"
998
+ ],
999
+ [
1000
+ "হ",
1001
+ "ল"
1002
+ ],
1003
+ [
1004
+ "Ġà¦Ĩ",
1005
+ "à¦Ľ"
1006
+ ],
1007
+ [
1008
+ "Ġà¦Ĩ",
1009
+ "à¦Ĺ"
1010
+ ],
1011
+ [
1012
+ "ধ",
1013
+ "ন"
1014
+ ],
1015
+ [
1016
+ "à¦ľ",
1017
+ "ধ"
1018
+ ],
1019
+ [
1020
+ "Ġà¦ī",
1021
+ "ড"
1022
+ ],
1023
+ [
1024
+ "Ġম",
1025
+ "রণ"
1026
+ ],
1027
+ [
1028
+ "Ġà¦ľ",
1029
+ "ন"
1030
+ ],
1031
+ [
1032
+ "Ġত",
1033
+ "থ"
1034
+ ],
1035
+ [
1036
+ "Ġফ",
1037
+ "ল"
1038
+ ],
1039
+ [
1040
+ "Ġসà¦ķ",
1041
+ "ল"
1042
+ ],
1043
+ [
1044
+ "ফ",
1045
+ "ল"
1046
+ ],
1047
+ [
1048
+ "তà¦ı",
1049
+ "ব"
1050
+ ],
1051
+ [
1052
+ "Ġমরণ",
1053
+ "শ"
1054
+ ]
1055
+ ]
1056
+ }
1057
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<eos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "405": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<bos>",
37
+ "clean_up_tokenization_spaces": false,
38
+ "eos_token": "<eos>",
39
+ "extra_special_tokens": {},
40
+ "model_max_length": 1000000000000000019884624838656,
41
+ "pad_token": "<pad>",
42
+ "tokenizer_class": "PreTrainedTokenizerFast",
43
+ "unk_token": "<unk>"
44
+ }