hf-transformers-bot commited on
Commit
ecd42a9
Β·
verified Β·
1 Parent(s): ceb3d7b

Add tiny processor fixture for testing

Browse files
chat_template.jinja ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ {% for message in messages %}{% if message['role'] != 'system' %}{{ message['role'].upper() + ': '}}{% endif %}{# Render all images first #}{% for content in message['content'] | selectattr('type', 'equalto', 'image') %}{{ '<image>
2
+ ' }}{% endfor %}{# Render all text next #}{% if message['role'] != 'assistant' %}{% for content in message['content'] | selectattr('type', 'equalto', 'text') %}{{ content['text'] + ' '}}{% endfor %}{% else %}{% for content in message['content'] | selectattr('type', 'equalto', 'text') %}{% generation %}{{ content['text'] + ' '}}{% endgeneration %}{% endfor %}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ 'ASSISTANT:' }}{% endif %}
chat_template.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "chat_template": "{% for message in messages %}{% if message['role'] != 'system' %}{{ message['role'].upper() + ': '}}{% endif %}{# Render all images first #}{% for content in message['content'] | selectattr('type', 'equalto', 'image') %}{{ '<image>\n' }}{% endfor %}{# Render all text next #}{% if message['role'] != 'assistant' %}{% for content in message['content'] | selectattr('type', 'equalto', 'text') %}{{ content['text'] + ' '}}{% endfor %}{% else %}{% for content in message['content'] | selectattr('type', 'equalto', 'text') %}{% generation %}{{ content['text'] + ' '}}{% endgeneration %}{% endfor %}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ 'ASSISTANT:' }}{% endif %}"
3
+ }
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlavaForConditionalGeneration"
4
+ ],
5
+ "ignore_index": -100,
6
+ "image_token_index": 32000,
7
+ "model_type": "llava",
8
+ "pad_token_id": 502,
9
+ "projector_hidden_act": "gelu",
10
+ "text_config": {
11
+ "_name_or_path": "lmsys/vicuna-7b-v1.5",
12
+ "architectures": [
13
+ "LlamaForCausalLM"
14
+ ],
15
+ "max_position_embeddings": 4096,
16
+ "model_type": "llama",
17
+ "rms_norm_eps": 1e-05,
18
+ "torch_dtype": "float16",
19
+ "vocab_size": 32064
20
+ },
21
+ "tie_word_embeddings": false,
22
+ "torch_dtype": "float16",
23
+ "transformers_version": "4.36.0.dev0",
24
+ "vision_config": {
25
+ "hidden_size": 1024,
26
+ "image_size": 336,
27
+ "intermediate_size": 4096,
28
+ "model_type": "clip_vision_model",
29
+ "num_attention_heads": 16,
30
+ "num_hidden_layers": 24,
31
+ "patch_size": 14,
32
+ "projection_dim": 768,
33
+ "vocab_size": 32000
34
+ },
35
+ "vision_feature_layer": -2,
36
+ "vision_feature_select_strategy": "default",
37
+ "vocab_size": 501
38
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 32001,
6
+ "transformers_version": "4.36.0.dev0"
7
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 336,
4
+ "width": 336
5
+ },
6
+ "do_center_crop": true,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.48145466,
13
+ 0.4578275,
14
+ 0.40821073
15
+ ],
16
+ "image_processor_type": "CLIPImageProcessor",
17
+ "image_std": [
18
+ 0.26862954,
19
+ 0.26130258,
20
+ 0.27577711
21
+ ],
22
+ "processor_class": "LlavaProcessor",
23
+ "resample": 3,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "shortest_edge": 336
27
+ }
28
+ }
processor_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_token": "<image>",
3
+ "num_additional_image_tokens": 1,
4
+ "patch_size": 14,
5
+ "processor_class": "LlavaProcessor",
6
+ "vision_feature_select_strategy": "default"
7
+ }
tokenizer.json ADDED
@@ -0,0 +1,846 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "version": "1.0",
3
+ "truncation": null,
4
+ "padding": null,
5
+ "added_tokens": [
6
+ {
7
+ "id": 0,
8
+ "content": "<unk>",
9
+ "single_word": false,
10
+ "lstrip": false,
11
+ "rstrip": false,
12
+ "normalized": false,
13
+ "special": true
14
+ },
15
+ {
16
+ "id": 1,
17
+ "content": "<s>",
18
+ "single_word": false,
19
+ "lstrip": false,
20
+ "rstrip": false,
21
+ "normalized": false,
22
+ "special": true
23
+ },
24
+ {
25
+ "id": 2,
26
+ "content": "</s>",
27
+ "single_word": false,
28
+ "lstrip": false,
29
+ "rstrip": false,
30
+ "normalized": false,
31
+ "special": true
32
+ },
33
+ {
34
+ "id": 501,
35
+ "content": "<image>",
36
+ "single_word": false,
37
+ "lstrip": false,
38
+ "rstrip": false,
39
+ "normalized": false,
40
+ "special": true
41
+ },
42
+ {
43
+ "id": 502,
44
+ "content": "<pad>",
45
+ "single_word": false,
46
+ "lstrip": false,
47
+ "rstrip": false,
48
+ "normalized": false,
49
+ "special": true
50
+ }
51
+ ],
52
+ "normalizer": {
53
+ "type": "Sequence",
54
+ "normalizers": [
55
+ {
56
+ "type": "Prepend",
57
+ "prepend": "▁"
58
+ },
59
+ {
60
+ "type": "Replace",
61
+ "pattern": {
62
+ "String": " "
63
+ },
64
+ "content": "▁"
65
+ }
66
+ ]
67
+ },
68
+ "pre_tokenizer": null,
69
+ "post_processor": {
70
+ "type": "TemplateProcessing",
71
+ "single": [
72
+ {
73
+ "SpecialToken": {
74
+ "id": "<s>",
75
+ "type_id": 0
76
+ }
77
+ },
78
+ {
79
+ "Sequence": {
80
+ "id": "A",
81
+ "type_id": 0
82
+ }
83
+ }
84
+ ],
85
+ "pair": [
86
+ {
87
+ "SpecialToken": {
88
+ "id": "<s>",
89
+ "type_id": 0
90
+ }
91
+ },
92
+ {
93
+ "Sequence": {
94
+ "id": "A",
95
+ "type_id": 0
96
+ }
97
+ },
98
+ {
99
+ "SpecialToken": {
100
+ "id": "<s>",
101
+ "type_id": 1
102
+ }
103
+ },
104
+ {
105
+ "Sequence": {
106
+ "id": "B",
107
+ "type_id": 1
108
+ }
109
+ }
110
+ ],
111
+ "special_tokens": {
112
+ "<s>": {
113
+ "id": "<s>",
114
+ "ids": [
115
+ 1
116
+ ],
117
+ "tokens": [
118
+ "<s>"
119
+ ]
120
+ }
121
+ }
122
+ },
123
+ "decoder": {
124
+ "type": "Sequence",
125
+ "decoders": [
126
+ {
127
+ "type": "Replace",
128
+ "pattern": {
129
+ "String": "▁"
130
+ },
131
+ "content": " "
132
+ },
133
+ {
134
+ "type": "ByteFallback"
135
+ },
136
+ {
137
+ "type": "Fuse"
138
+ },
139
+ {
140
+ "type": "Strip",
141
+ "content": " ",
142
+ "start": 1,
143
+ "stop": 0
144
+ }
145
+ ]
146
+ },
147
+ "model": {
148
+ "type": "BPE",
149
+ "dropout": null,
150
+ "unk_token": "<unk>",
151
+ "continuing_subword_prefix": null,
152
+ "end_of_word_suffix": null,
153
+ "fuse_unk": true,
154
+ "byte_fallback": true,
155
+ "ignore_merges": false,
156
+ "vocab": {
157
+ "<unk>": 0,
158
+ "<s>": 1,
159
+ "</s>": 2,
160
+ "<0x00>": 3,
161
+ "<0x01>": 4,
162
+ "<0x02>": 5,
163
+ "<0x03>": 6,
164
+ "<0x04>": 7,
165
+ "<0x05>": 8,
166
+ "<0x06>": 9,
167
+ "<0x07>": 10,
168
+ "<0x08>": 11,
169
+ "<0x09>": 12,
170
+ "<0x0A>": 13,
171
+ "<0x0B>": 14,
172
+ "<0x0C>": 15,
173
+ "<0x0D>": 16,
174
+ "<0x0E>": 17,
175
+ "<0x0F>": 18,
176
+ "<0x10>": 19,
177
+ "<0x11>": 20,
178
+ "<0x12>": 21,
179
+ "<0x13>": 22,
180
+ "<0x14>": 23,
181
+ "<0x15>": 24,
182
+ "<0x16>": 25,
183
+ "<0x17>": 26,
184
+ "<0x18>": 27,
185
+ "<0x19>": 28,
186
+ "<0x1A>": 29,
187
+ "<0x1B>": 30,
188
+ "<0x1C>": 31,
189
+ "<0x1D>": 32,
190
+ "<0x1E>": 33,
191
+ "<0x1F>": 34,
192
+ "<0x20>": 35,
193
+ "<0x21>": 36,
194
+ "<0x22>": 37,
195
+ "<0x23>": 38,
196
+ "<0x24>": 39,
197
+ "<0x25>": 40,
198
+ "<0x26>": 41,
199
+ "<0x27>": 42,
200
+ "<0x28>": 43,
201
+ "<0x29>": 44,
202
+ "<0x2A>": 45,
203
+ "<0x2B>": 46,
204
+ "<0x2C>": 47,
205
+ "<0x2D>": 48,
206
+ "<0x2E>": 49,
207
+ "<0x2F>": 50,
208
+ "<0x30>": 51,
209
+ "<0x31>": 52,
210
+ "<0x32>": 53,
211
+ "<0x33>": 54,
212
+ "<0x34>": 55,
213
+ "<0x35>": 56,
214
+ "<0x36>": 57,
215
+ "<0x37>": 58,
216
+ "<0x38>": 59,
217
+ "<0x39>": 60,
218
+ "<0x3A>": 61,
219
+ "<0x3B>": 62,
220
+ "<0x3C>": 63,
221
+ "<0x3D>": 64,
222
+ "<0x3E>": 65,
223
+ "<0x3F>": 66,
224
+ "<0x40>": 67,
225
+ "<0x41>": 68,
226
+ "<0x42>": 69,
227
+ "<0x43>": 70,
228
+ "<0x44>": 71,
229
+ "<0x45>": 72,
230
+ "<0x46>": 73,
231
+ "<0x47>": 74,
232
+ "<0x48>": 75,
233
+ "<0x49>": 76,
234
+ "<0x4A>": 77,
235
+ "<0x4B>": 78,
236
+ "<0x4C>": 79,
237
+ "<0x4D>": 80,
238
+ "<0x4E>": 81,
239
+ "<0x4F>": 82,
240
+ "<0x50>": 83,
241
+ "<0x51>": 84,
242
+ "<0x52>": 85,
243
+ "<0x53>": 86,
244
+ "<0x54>": 87,
245
+ "<0x55>": 88,
246
+ "<0x56>": 89,
247
+ "<0x57>": 90,
248
+ "<0x58>": 91,
249
+ "<0x59>": 92,
250
+ "<0x5A>": 93,
251
+ "<0x5B>": 94,
252
+ "<0x5C>": 95,
253
+ "<0x5D>": 96,
254
+ "<0x5E>": 97,
255
+ "<0x5F>": 98,
256
+ "<0x60>": 99,
257
+ "<0x61>": 100,
258
+ "<0x62>": 101,
259
+ "<0x63>": 102,
260
+ "<0x64>": 103,
261
+ "<0x65>": 104,
262
+ "<0x66>": 105,
263
+ "<0x67>": 106,
264
+ "<0x68>": 107,
265
+ "<0x69>": 108,
266
+ "<0x6A>": 109,
267
+ "<0x6B>": 110,
268
+ "<0x6C>": 111,
269
+ "<0x6D>": 112,
270
+ "<0x6E>": 113,
271
+ "<0x6F>": 114,
272
+ "<0x70>": 115,
273
+ "<0x71>": 116,
274
+ "<0x72>": 117,
275
+ "<0x73>": 118,
276
+ "<0x74>": 119,
277
+ "<0x75>": 120,
278
+ "<0x76>": 121,
279
+ "<0x77>": 122,
280
+ "<0x78>": 123,
281
+ "<0x79>": 124,
282
+ "<0x7A>": 125,
283
+ "<0x7B>": 126,
284
+ "<0x7C>": 127,
285
+ "<0x7D>": 128,
286
+ "<0x7E>": 129,
287
+ "<0x7F>": 130,
288
+ "<0x80>": 131,
289
+ "<0x81>": 132,
290
+ "<0x82>": 133,
291
+ "<0x83>": 134,
292
+ "<0x84>": 135,
293
+ "<0x85>": 136,
294
+ "<0x86>": 137,
295
+ "<0x87>": 138,
296
+ "<0x88>": 139,
297
+ "<0x89>": 140,
298
+ "<0x8A>": 141,
299
+ "<0x8B>": 142,
300
+ "<0x8C>": 143,
301
+ "<0x8D>": 144,
302
+ "<0x8E>": 145,
303
+ "<0x8F>": 146,
304
+ "<0x90>": 147,
305
+ "<0x91>": 148,
306
+ "<0x92>": 149,
307
+ "<0x93>": 150,
308
+ "<0x94>": 151,
309
+ "<0x95>": 152,
310
+ "<0x96>": 153,
311
+ "<0x97>": 154,
312
+ "<0x98>": 155,
313
+ "<0x99>": 156,
314
+ "<0x9A>": 157,
315
+ "<0x9B>": 158,
316
+ "<0x9C>": 159,
317
+ "<0x9D>": 160,
318
+ "<0x9E>": 161,
319
+ "<0x9F>": 162,
320
+ "<0xA0>": 163,
321
+ "<0xA1>": 164,
322
+ "<0xA2>": 165,
323
+ "<0xA3>": 166,
324
+ "<0xA4>": 167,
325
+ "<0xA5>": 168,
326
+ "<0xA6>": 169,
327
+ "<0xA7>": 170,
328
+ "<0xA8>": 171,
329
+ "<0xA9>": 172,
330
+ "<0xAA>": 173,
331
+ "<0xAB>": 174,
332
+ "<0xAC>": 175,
333
+ "<0xAD>": 176,
334
+ "<0xAE>": 177,
335
+ "<0xAF>": 178,
336
+ "<0xB0>": 179,
337
+ "<0xB1>": 180,
338
+ "<0xB2>": 181,
339
+ "<0xB3>": 182,
340
+ "<0xB4>": 183,
341
+ "<0xB5>": 184,
342
+ "<0xB6>": 185,
343
+ "<0xB7>": 186,
344
+ "<0xB8>": 187,
345
+ "<0xB9>": 188,
346
+ "<0xBA>": 189,
347
+ "<0xBB>": 190,
348
+ "<0xBC>": 191,
349
+ "<0xBD>": 192,
350
+ "<0xBE>": 193,
351
+ "<0xBF>": 194,
352
+ "<0xC0>": 195,
353
+ "<0xC1>": 196,
354
+ "<0xC2>": 197,
355
+ "<0xC3>": 198,
356
+ "<0xC4>": 199,
357
+ "<0xC5>": 200,
358
+ "<0xC6>": 201,
359
+ "<0xC7>": 202,
360
+ "<0xC8>": 203,
361
+ "<0xC9>": 204,
362
+ "<0xCA>": 205,
363
+ "<0xCB>": 206,
364
+ "<0xCC>": 207,
365
+ "<0xCD>": 208,
366
+ "<0xCE>": 209,
367
+ "<0xCF>": 210,
368
+ "<0xD0>": 211,
369
+ "<0xD1>": 212,
370
+ "<0xD2>": 213,
371
+ "<0xD3>": 214,
372
+ "<0xD4>": 215,
373
+ "<0xD5>": 216,
374
+ "<0xD6>": 217,
375
+ "<0xD7>": 218,
376
+ "<0xD8>": 219,
377
+ "<0xD9>": 220,
378
+ "<0xDA>": 221,
379
+ "<0xDB>": 222,
380
+ "<0xDC>": 223,
381
+ "<0xDD>": 224,
382
+ "<0xDE>": 225,
383
+ "<0xDF>": 226,
384
+ "<0xE0>": 227,
385
+ "<0xE1>": 228,
386
+ "<0xE2>": 229,
387
+ "<0xE3>": 230,
388
+ "<0xE4>": 231,
389
+ "<0xE5>": 232,
390
+ "<0xE6>": 233,
391
+ "<0xE7>": 234,
392
+ "<0xE8>": 235,
393
+ "<0xE9>": 236,
394
+ "<0xEA>": 237,
395
+ "<0xEB>": 238,
396
+ "<0xEC>": 239,
397
+ "<0xED>": 240,
398
+ "<0xEE>": 241,
399
+ "<0xEF>": 242,
400
+ "<0xF0>": 243,
401
+ "<0xF1>": 244,
402
+ "<0xF2>": 245,
403
+ "<0xF3>": 246,
404
+ "<0xF4>": 247,
405
+ "<0xF5>": 248,
406
+ "<0xF6>": 249,
407
+ "<0xF7>": 250,
408
+ "<0xF8>": 251,
409
+ "<0xF9>": 252,
410
+ "<0xFA>": 253,
411
+ "<0xFB>": 254,
412
+ "<0xFC>": 255,
413
+ "<0xFD>": 256,
414
+ "<0xFE>": 257,
415
+ "<0xFF>": 258,
416
+ "▁▁": 259,
417
+ "▁t": 260,
418
+ "er": 261,
419
+ "in": 262,
420
+ "▁a": 263,
421
+ "en": 264,
422
+ "on": 265,
423
+ "▁th": 266,
424
+ "es": 267,
425
+ "▁▁▁▁": 268,
426
+ "▁s": 269,
427
+ "▁d": 270,
428
+ "at": 271,
429
+ "or": 272,
430
+ "an": 273,
431
+ "▁c": 274,
432
+ "is": 275,
433
+ "re": 276,
434
+ "it": 277,
435
+ "▁the": 278,
436
+ "ar": 279,
437
+ "le": 280,
438
+ "▁w": 281,
439
+ "▁p": 282,
440
+ "ou": 283,
441
+ "al": 284,
442
+ "▁f": 285,
443
+ "▁m": 286,
444
+ "ed": 287,
445
+ "▁o": 288,
446
+ "▁b": 289,
447
+ "om": 290,
448
+ "ion": 291,
449
+ "ing": 292,
450
+ "ic": 293,
451
+ "as": 294,
452
+ "el": 295,
453
+ "ent": 296,
454
+ "▁in": 297,
455
+ "▁h": 298,
456
+ "nd": 299,
457
+ "et": 300,
458
+ "▁l": 301,
459
+ "▁n": 302,
460
+ "st": 303,
461
+ "▁to": 304,
462
+ "ch": 305,
463
+ "▁I": 306,
464
+ "ro": 307,
465
+ "▁▁▁▁▁▁▁▁": 308,
466
+ "il": 309,
467
+ "▁of": 310,
468
+ "de": 311,
469
+ "ct": 312,
470
+ "▁(": 313,
471
+ "am": 314,
472
+ "▁C": 315,
473
+ "▁de": 316,
474
+ "▁S": 317,
475
+ "▁u": 318,
476
+ "▁A": 319,
477
+ "▁\\": 320,
478
+ "▁e": 321,
479
+ "▁and": 322,
480
+ "▁T": 323,
481
+ "ol": 324,
482
+ "▁v": 325,
483
+ "im": 326,
484
+ "ot": 327,
485
+ "ad": 328,
486
+ "ut": 329,
487
+ "▁g": 330,
488
+ "em": 331,
489
+ "ur": 332,
490
+ "id": 333,
491
+ "▁*": 334,
492
+ "ig": 335,
493
+ "ra": 336,
494
+ "▁re": 337,
495
+ "▁is": 338,
496
+ "qu": 339,
497
+ "ow": 340,
498
+ "▁M": 341,
499
+ "est": 342,
500
+ "▁y": 343,
501
+ "se": 344,
502
+ "ve": 345,
503
+ "ce": 346,
504
+ "ie": 347,
505
+ "un": 348,
506
+ "▁P": 349,
507
+ "▁B": 350,
508
+ "ag": 351,
509
+ "ul": 352,
510
+ "▁=": 353,
511
+ "he": 354,
512
+ "end": 355,
513
+ "ode": 356,
514
+ "ter": 357,
515
+ "ment": 358,
516
+ "os": 359,
517
+ "▁D": 360,
518
+ "if": 361,
519
+ "ation": 362,
520
+ "▁for": 363,
521
+ "▁r": 364,
522
+ "▁L": 365,
523
+ "▁you": 366,
524
+ "▁be": 367,
525
+ "ly": 368,
526
+ "ver": 369,
527
+ "ab": 370,
528
+ "te": 371,
529
+ "▁it": 372,
530
+ "▁on": 373,
531
+ "ri": 374,
532
+ "us": 375,
533
+ "▁\"": 376,
534
+ "▁wh": 377,
535
+ "▁con": 378,
536
+ "▁H": 379,
537
+ "▁st": 380,
538
+ "ir": 381,
539
+ "▁E": 382,
540
+ "▁F": 383,
541
+ "ck": 384,
542
+ "▁an": 385,
543
+ "th": 386,
544
+ "eg": 387,
545
+ "ay": 388,
546
+ "ith": 389,
547
+ "▁R": 390,
548
+ "ist": 391,
549
+ "and": 392,
550
+ "▁that": 393,
551
+ "▁al": 394,
552
+ "▁$": 395,
553
+ "▁#": 396,
554
+ "od": 397,
555
+ "um": 398,
556
+ "▁W": 399,
557
+ "ht": 400,
558
+ "code": 401,
559
+ "▁G": 402,
560
+ "ate": 403,
561
+ "ess": 404,
562
+ "▁N": 405,
563
+ "ere": 406,
564
+ "pp": 407,
565
+ "▁as": 408,
566
+ "▁se": 409,
567
+ "▁pro": 410,
568
+ "▁with": 411,
569
+ "pe": 412,
570
+ "▁k": 413,
571
+ "ers": 414,
572
+ "pt": 415,
573
+ ");": 416,
574
+ "lo": 417,
575
+ "▁▁▁▁▁": 418,
576
+ "▁com": 419,
577
+ "ame": 420,
578
+ "▁`": 421,
579
+ "▁Com": 422,
580
+ "ia": 423,
581
+ "ant": 424,
582
+ "▁la": 425,
583
+ "▁{": 426,
584
+ "▁en": 427,
585
+ "ction": 428,
586
+ "▁ex": 429,
587
+ "ld": 430,
588
+ "ub": 431,
589
+ "▁j": 432,
590
+ "la": 433,
591
+ "ue": 434,
592
+ "▁J": 435,
593
+ "ich": 436,
594
+ "▁do": 437,
595
+ "▁O": 438,
596
+ "▁qu": 439,
597
+ "iv": 440,
598
+ "ort": 441,
599
+ "art": 442,
600
+ "▁un": 443,
601
+ "▁##": 444,
602
+ "▁this": 445,
603
+ "ke": 446,
604
+ "▁ha": 447,
605
+ "▁-": 448,
606
+ "out": 449,
607
+ "▁The": 450,
608
+ "▁not": 451,
609
+ "▁ne": 452,
610
+ "ill": 453,
611
+ "▁le": 454,
612
+ "ci": 455,
613
+ "rom": 456,
614
+ "ine": 457,
615
+ "//": 458,
616
+ "op": 459,
617
+ "egin": 460,
618
+ "▁Comment": 461,
619
+ "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁": 462,
620
+ "begin": 463,
621
+ "ст": 464,
622
+ "ass": 465,
623
+ "iz": 466,
624
+ ").": 467,
625
+ "og": 468,
626
+ "▁п": 469,
627
+ "▁or": 470,
628
+ "▁was": 471,
629
+ "▁at": 472,
630
+ "our": 473,
631
+ "▁i": 474,
632
+ "ain": 475,
633
+ "▁K": 476,
634
+ "Π½Π°": 477,
635
+ "▁V": 478,
636
+ "ge": 479,
637
+ "▁su": 480,
638
+ "ap": 481,
639
+ "age": 482,
640
+ "ould": 483,
641
+ "ne": 484,
642
+ "av": 485,
643
+ "xt": 486,
644
+ "ore": 487,
645
+ "ile": 488,
646
+ "--": 489,
647
+ "▁в": 490,
648
+ "▁by": 491,
649
+ "li": 492,
650
+ "ath": 493,
651
+ "Ρ€Π°": 494,
652
+ "ber": 495,
653
+ "ach": 496,
654
+ "all": 497,
655
+ "▁Th": 498,
656
+ "ult": 499,
657
+ "▁": 500
658
+ },
659
+ "merges": [
660
+ [
661
+ "▁",
662
+ "th"
663
+ ],
664
+ [
665
+ "▁t",
666
+ "he"
667
+ ],
668
+ [
669
+ "▁",
670
+ "in"
671
+ ],
672
+ [
673
+ "▁",
674
+ "de"
675
+ ],
676
+ [
677
+ "▁a",
678
+ "nd"
679
+ ],
680
+ [
681
+ "▁",
682
+ "and"
683
+ ],
684
+ [
685
+ "▁",
686
+ "re"
687
+ ],
688
+ [
689
+ "▁",
690
+ "is"
691
+ ],
692
+ [
693
+ "at",
694
+ "ion"
695
+ ],
696
+ [
697
+ "▁f",
698
+ "or"
699
+ ],
700
+ [
701
+ "▁y",
702
+ "ou"
703
+ ],
704
+ [
705
+ "▁",
706
+ "it"
707
+ ],
708
+ [
709
+ "▁",
710
+ "on"
711
+ ],
712
+ [
713
+ "▁c",
714
+ "on"
715
+ ],
716
+ [
717
+ "▁",
718
+ "st"
719
+ ],
720
+ [
721
+ "▁",
722
+ "an"
723
+ ],
724
+ [
725
+ "▁th",
726
+ "at"
727
+ ],
728
+ [
729
+ "▁",
730
+ "al"
731
+ ],
732
+ [
733
+ "▁",
734
+ "as"
735
+ ],
736
+ [
737
+ "▁",
738
+ "se"
739
+ ],
740
+ [
741
+ "▁p",
742
+ "ro"
743
+ ],
744
+ [
745
+ "▁w",
746
+ "ith"
747
+ ],
748
+ [
749
+ "▁c",
750
+ "om"
751
+ ],
752
+ [
753
+ "▁C",
754
+ "om"
755
+ ],
756
+ [
757
+ "▁",
758
+ "la"
759
+ ],
760
+ [
761
+ "▁",
762
+ "en"
763
+ ],
764
+ [
765
+ "ct",
766
+ "ion"
767
+ ],
768
+ [
769
+ "▁",
770
+ "qu"
771
+ ],
772
+ [
773
+ "▁",
774
+ "un"
775
+ ],
776
+ [
777
+ "▁th",
778
+ "is"
779
+ ],
780
+ [
781
+ "▁T",
782
+ "he"
783
+ ],
784
+ [
785
+ "▁n",
786
+ "ot"
787
+ ],
788
+ [
789
+ "▁",
790
+ "ne"
791
+ ],
792
+ [
793
+ "▁",
794
+ "le"
795
+ ],
796
+ [
797
+ "eg",
798
+ "in"
799
+ ],
800
+ [
801
+ "▁Com",
802
+ "ment"
803
+ ],
804
+ [
805
+ "▁",
806
+ "or"
807
+ ],
808
+ [
809
+ "▁w",
810
+ "as"
811
+ ],
812
+ [
813
+ "▁",
814
+ "at"
815
+ ],
816
+ [
817
+ "ou",
818
+ "ld"
819
+ ],
820
+ [
821
+ "▁",
822
+ "▁"
823
+ ],
824
+ [
825
+ "▁▁",
826
+ "▁▁"
827
+ ],
828
+ [
829
+ "▁▁▁▁",
830
+ "▁▁▁▁"
831
+ ],
832
+ [
833
+ "▁▁▁▁",
834
+ "▁"
835
+ ],
836
+ [
837
+ "▁",
838
+ "▁▁▁▁"
839
+ ],
840
+ [
841
+ "▁▁▁▁▁▁▁▁",
842
+ "▁▁▁▁▁▁▁▁"
843
+ ]
844
+ ]
845
+ }
846
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": null,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "single_word": false,
9
+ "lstrip": false,
10
+ "rstrip": false,
11
+ "normalized": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "single_word": false,
17
+ "lstrip": false,
18
+ "rstrip": false,
19
+ "normalized": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "single_word": false,
25
+ "lstrip": false,
26
+ "rstrip": false,
27
+ "normalized": false,
28
+ "special": true
29
+ },
30
+ "501": {
31
+ "content": "<image>",
32
+ "single_word": false,
33
+ "lstrip": false,
34
+ "rstrip": false,
35
+ "normalized": false,
36
+ "special": true
37
+ },
38
+ "502": {
39
+ "content": "<pad>",
40
+ "single_word": false,
41
+ "lstrip": false,
42
+ "rstrip": false,
43
+ "normalized": false,
44
+ "special": true
45
+ }
46
+ },
47
+ "bos_token": "<s>",
48
+ "clean_up_tokenization_spaces": false,
49
+ "eos_token": "</s>",
50
+ "extra_special_tokens": {
51
+ "image_token": "<image>"
52
+ },
53
+ "image_token": "<image>",
54
+ "legacy": false,
55
+ "model_max_length": 1000000000000000019884624838656,
56
+ "pad_token": "<pad>",
57
+ "padding_side": "left",
58
+ "processor_class": "LlavaProcessor",
59
+ "sp_model_kwargs": {},
60
+ "tokenizer_class": "LlamaTokenizer",
61
+ "trust_remote_code": false,
62
+ "unk_token": "<unk>",
63
+ "use_default_system_prompt": false
64
+ }