Hugging Face's logo

kaitchup
/
translategemma-4b-it-FP8-Dynamic

Safetensors
gemma3
compressed-tensors
Model card Files
xet
 Community
bnjmnmarie commited on
Commit
2e223ed
·
verified ·
1 Parent(s): 545846b

Upload folder using huggingface_hub

Browse files
Files changed (13) hide show
  1. .gitattributes +1 -0
  2. added_tokens.json +3 -0
  3. chat_template.jinja +641 -0
  4. config.json +312 -0
  5. generation_config.json +7 -0
  6. model-00001-of-00002.safetensors +3 -0
  7. model-00002-of-00002.safetensors +3 -0
  8. model.safetensors.index.json +0 -0
  9. recipe.yaml +6 -0
  10. special_tokens_map.json +33 -0
  11. tokenizer.json +3 -0
  12. tokenizer.model +3 -0
  13. tokenizer_config.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144
3
+ }
chat_template.jinja ADDED
@@ -0,0 +1,641 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set languages = {
2
+ "aa": "Afar",
3
+ "aa-DJ": "Afar",
4
+ "aa-ER": "Afar",
5
+ "ab": "Abkhazian",
6
+ "af": "Afrikaans",
7
+ "af-NA": "Afrikaans",
8
+ "ak": "Akan",
9
+ "am": "Amharic",
10
+ "an": "Aragonese",
11
+ "ar": "Arabic",
12
+ "ar-AE": "Arabic",
13
+ "ar-BH": "Arabic",
14
+ "ar-DJ": "Arabic",
15
+ "ar-DZ": "Arabic",
16
+ "ar-EG": "Arabic",
17
+ "ar-EH": "Arabic",
18
+ "ar-ER": "Arabic",
19
+ "ar-IL": "Arabic",
20
+ "ar-IQ": "Arabic",
21
+ "ar-JO": "Arabic",
22
+ "ar-KM": "Arabic",
23
+ "ar-KW": "Arabic",
24
+ "ar-LB": "Arabic",
25
+ "ar-LY": "Arabic",
26
+ "ar-MA": "Arabic",
27
+ "ar-MR": "Arabic",
28
+ "ar-OM": "Arabic",
29
+ "ar-PS": "Arabic",
30
+ "ar-QA": "Arabic",
31
+ "ar-SA": "Arabic",
32
+ "ar-SD": "Arabic",
33
+ "ar-SO": "Arabic",
34
+ "ar-SS": "Arabic",
35
+ "ar-SY": "Arabic",
36
+ "ar-TD": "Arabic",
37
+ "ar-TN": "Arabic",
38
+ "ar-YE": "Arabic",
39
+ "as": "Assamese",
40
+ "az": "Azerbaijani",
41
+ "az-Arab": "Azerbaijani",
42
+ "az-Arab-IQ": "Azerbaijani",
43
+ "az-Arab-TR": "Azerbaijani",
44
+ "az-Cyrl": "Azerbaijani",
45
+ "az-Latn": "Azerbaijani",
46
+ "ba": "Bashkir",
47
+ "be": "Belarusian",
48
+ "be-tarask": "Belarusian",
49
+ "bg": "Bulgarian",
50
+ "bg-BG": "Bulgarian",
51
+ "bm": "Bambara",
52
+ "bm-Nkoo": "Bambara",
53
+ "bn": "Bengali",
54
+ "bn-IN": "Bengali",
55
+ "bo": "Tibetan",
56
+ "bo-IN": "Tibetan",
57
+ "br": "Breton",
58
+ "bs": "Bosnian",
59
+ "bs-Cyrl": "Bosnian",
60
+ "bs-Latn": "Bosnian",
61
+ "ca": "Catalan",
62
+ "ca-AD": "Catalan",
63
+ "ca-ES": "Catalan",
64
+ "ca-FR": "Catalan",
65
+ "ca-IT": "Catalan",
66
+ "ce": "Chechen",
67
+ "co": "Corsican",
68
+ "cs": "Czech",
69
+ "cs-CZ": "Czech",
70
+ "cv": "Chuvash",
71
+ "cy": "Welsh",
72
+ "da": "Danish",
73
+ "da-DK": "Danish",
74
+ "da-GL": "Danish",
75
+ "de": "German",
76
+ "de-AT": "German",
77
+ "de-BE": "German",
78
+ "de-CH": "German",
79
+ "de-DE": "German",
80
+ "de-IT": "German",
81
+ "de-LI": "German",
82
+ "de-LU": "German",
83
+ "dv": "Divehi",
84
+ "dz": "Dzongkha",
85
+ "ee": "Ewe",
86
+ "ee-TG": "Ewe",
87
+ "el": "Greek",
88
+ "el-CY": "Greek",
89
+ "el-GR": "Greek",
90
+ "el-polyton": "Greek",
91
+ "en": "English",
92
+ "en-AE": "English",
93
+ "en-AG": "English",
94
+ "en-AI": "English",
95
+ "en-AS": "English",
96
+ "en-AT": "English",
97
+ "en-AU": "English",
98
+ "en-BB": "English",
99
+ "en-BE": "English",
100
+ "en-BI": "English",
101
+ "en-BM": "English",
102
+ "en-BS": "English",
103
+ "en-BW": "English",
104
+ "en-BZ": "English",
105
+ "en-CA": "English",
106
+ "en-CC": "English",
107
+ "en-CH": "English",
108
+ "en-CK": "English",
109
+ "en-CM": "English",
110
+ "en-CX": "English",
111
+ "en-CY": "English",
112
+ "en-CZ": "English",
113
+ "en-DE": "English",
114
+ "en-DG": "English",
115
+ "en-DK": "English",
116
+ "en-DM": "English",
117
+ "en-ER": "English",
118
+ "en-ES": "English",
119
+ "en-FI": "English",
120
+ "en-FJ": "English",
121
+ "en-FK": "English",
122
+ "en-FM": "English",
123
+ "en-FR": "English",
124
+ "en-GB": "English",
125
+ "en-GD": "English",
126
+ "en-GG": "English",
127
+ "en-GH": "English",
128
+ "en-GI": "English",
129
+ "en-GM": "English",
130
+ "en-GS": "English",
131
+ "en-GU": "English",
132
+ "en-GY": "English",
133
+ "en-HK": "English",
134
+ "en-HU": "English",
135
+ "en-ID": "English",
136
+ "en-IE": "English",
137
+ "en-IL": "English",
138
+ "en-IM": "English",
139
+ "en-IN": "English",
140
+ "en-IO": "English",
141
+ "en-IT": "English",
142
+ "en-JE": "English",
143
+ "en-JM": "English",
144
+ "en-KE": "English",
145
+ "en-KI": "English",
146
+ "en-KN": "English",
147
+ "en-KY": "English",
148
+ "en-LC": "English",
149
+ "en-LR": "English",
150
+ "en-LS": "English",
151
+ "en-MG": "English",
152
+ "en-MH": "English",
153
+ "en-MO": "English",
154
+ "en-MP": "English",
155
+ "en-MS": "English",
156
+ "en-MT": "English",
157
+ "en-MU": "English",
158
+ "en-MV": "English",
159
+ "en-MW": "English",
160
+ "en-MY": "English",
161
+ "en-NA": "English",
162
+ "en-NF": "English",
163
+ "en-NG": "English",
164
+ "en-NL": "English",
165
+ "en-NO": "English",
166
+ "en-NR": "English",
167
+ "en-NU": "English",
168
+ "en-NZ": "English",
169
+ "en-PG": "English",
170
+ "en-PH": "English",
171
+ "en-PK": "English",
172
+ "en-PL": "English",
173
+ "en-PN": "English",
174
+ "en-PR": "English",
175
+ "en-PT": "English",
176
+ "en-PW": "English",
177
+ "en-RO": "English",
178
+ "en-RW": "English",
179
+ "en-SB": "English",
180
+ "en-SC": "English",
181
+ "en-SD": "English",
182
+ "en-SE": "English",
183
+ "en-SG": "English",
184
+ "en-SH": "English",
185
+ "en-SI": "English",
186
+ "en-SK": "English",
187
+ "en-SL": "English",
188
+ "en-SS": "English",
189
+ "en-SX": "English",
190
+ "en-SZ": "English",
191
+ "en-TC": "English",
192
+ "en-TK": "English",
193
+ "en-TO": "English",
194
+ "en-TT": "English",
195
+ "en-TV": "English",
196
+ "en-TZ": "English",
197
+ "en-UG": "English",
198
+ "en-UM": "English",
199
+ "en-VC": "English",
200
+ "en-VG": "English",
201
+ "en-VI": "English",
202
+ "en-VU": "English",
203
+ "en-WS": "English",
204
+ "en-ZA": "English",
205
+ "en-ZM": "English",
206
+ "en-ZW": "English",
207
+ "eo": "Esperanto",
208
+ "es": "Spanish",
209
+ "es-AR": "Spanish",
210
+ "es-BO": "Spanish",
211
+ "es-BR": "Spanish",
212
+ "es-BZ": "Spanish",
213
+ "es-CL": "Spanish",
214
+ "es-CO": "Spanish",
215
+ "es-CR": "Spanish",
216
+ "es-CU": "Spanish",
217
+ "es-DO": "Spanish",
218
+ "es-EA": "Spanish",
219
+ "es-EC": "Spanish",
220
+ "es-ES": "Spanish",
221
+ "es-GQ": "Spanish",
222
+ "es-GT": "Spanish",
223
+ "es-HN": "Spanish",
224
+ "es-IC": "Spanish",
225
+ "es-MX": "Spanish",
226
+ "es-NI": "Spanish",
227
+ "es-PA": "Spanish",
228
+ "es-PE": "Spanish",
229
+ "es-PH": "Spanish",
230
+ "es-PR": "Spanish",
231
+ "es-PY": "Spanish",
232
+ "es-SV": "Spanish",
233
+ "es-US": "Spanish",
234
+ "es-UY": "Spanish",
235
+ "es-VE": "Spanish",
236
+ "et": "Estonian",
237
+ "et-EE": "Estonian",
238
+ "eu": "Basque",
239
+ "fa": "Persian",
240
+ "fa-AF": "Persian",
241
+ "fa-IR": "Persian",
242
+ "ff": "Fulah",
243
+ "ff-Adlm": "Fulah",
244
+ "ff-Adlm-BF": "Fulah",
245
+ "ff-Adlm-CM": "Fulah",
246
+ "ff-Adlm-GH": "Fulah",
247
+ "ff-Adlm-GM": "Fulah",
248
+ "ff-Adlm-GW": "Fulah",
249
+ "ff-Adlm-LR": "Fulah",
250
+ "ff-Adlm-MR": "Fulah",
251
+ "ff-Adlm-NE": "Fulah",
252
+ "ff-Adlm-NG": "Fulah",
253
+ "ff-Adlm-SL": "Fulah",
254
+ "ff-Adlm-SN": "Fulah",
255
+ "ff-Latn": "Fulah",
256
+ "ff-Latn-BF": "Fulah",
257
+ "ff-Latn-CM": "Fulah",
258
+ "ff-Latn-GH": "Fulah",
259
+ "ff-Latn-GM": "Fulah",
260
+ "ff-Latn-GN": "Fulah",
261
+ "ff-Latn-GW": "Fulah",
262
+ "ff-Latn-LR": "Fulah",
263
+ "ff-Latn-MR": "Fulah",
264
+ "ff-Latn-NE": "Fulah",
265
+ "ff-Latn-NG": "Fulah",
266
+ "ff-Latn-SL": "Fulah",
267
+ "fi": "Finnish",
268
+ "fi-FI": "Finnish",
269
+ "fil-PH": "Filipino",
270
+ "fo": "Faroese",
271
+ "fo-DK": "Faroese",
272
+ "fr": "French",
273
+ "fr-BE": "French",
274
+ "fr-BF": "French",
275
+ "fr-BI": "French",
276
+ "fr-BJ": "French",
277
+ "fr-BL": "French",
278
+ "fr-CA": "French",
279
+ "fr-CD": "French",
280
+ "fr-CF": "French",
281
+ "fr-CG": "French",
282
+ "fr-CH": "French",
283
+ "fr-CI": "French",
284
+ "fr-CM": "French",
285
+ "fr-DJ": "French",
286
+ "fr-DZ": "French",
287
+ "fr-FR": "French",
288
+ "fr-GA": "French",
289
+ "fr-GF": "French",
290
+ "fr-GN": "French",
291
+ "fr-GP": "French",
292
+ "fr-GQ": "French",
293
+ "fr-HT": "French",
294
+ "fr-KM": "French",
295
+ "fr-LU": "French",
296
+ "fr-MA": "French",
297
+ "fr-MC": "French",
298
+ "fr-MF": "French",
299
+ "fr-MG": "French",
300
+ "fr-ML": "French",
301
+ "fr-MQ": "French",
302
+ "fr-MR": "French",
303
+ "fr-MU": "French",
304
+ "fr-NC": "French",
305
+ "fr-NE": "French",
306
+ "fr-PF": "French",
307
+ "fr-PM": "French",
308
+ "fr-RE": "French",
309
+ "fr-RW": "French",
310
+ "fr-SC": "French",
311
+ "fr-SN": "French",
312
+ "fr-SY": "French",
313
+ "fr-TD": "French",
314
+ "fr-TG": "French",
315
+ "fr-TN": "French",
316
+ "fr-VU": "French",
317
+ "fr-WF": "French",
318
+ "fr-YT": "French",
319
+ "fy": "Western Frisian",
320
+ "ga": "Irish",
321
+ "ga-GB": "Irish",
322
+ "gd": "Scottish Gaelic",
323
+ "gl": "Galician",
324
+ "gn": "Guarani",
325
+ "gu": "Gujarati",
326
+ "gu-IN": "Gujarati",
327
+ "gv": "Manx",
328
+ "ha": "Hausa",
329
+ "ha-Arab": "Hausa",
330
+ "ha-Arab-SD": "Hausa",
331
+ "ha-GH": "Hausa",
332
+ "ha-NE": "Hausa",
333
+ "he": "Hebrew",
334
+ "he-IL": "Hebrew",
335
+ "hi": "Hindi",
336
+ "hi-IN": "Hindi",
337
+ "hi-Latn": "Hindi",
338
+ "hr": "Croatian",
339
+ "hr-BA": "Croatian",
340
+ "hr-HR": "Croatian",
341
+ "ht": "Haitian",
342
+ "hu": "Hungarian",
343
+ "hu-HU": "Hungarian",
344
+ "hy": "Armenian",
345
+ "ia": "Interlingua",
346
+ "id": "Indonesian",
347
+ "id-ID": "Indonesian",
348
+ "ie": "Interlingue",
349
+ "ig": "Igbo",
350
+ "ii": "Sichuan Yi",
351
+ "ik": "Inupiaq",
352
+ "io": "Ido",
353
+ "is": "Icelandic",
354
+ "it": "Italian",
355
+ "it-CH": "Italian",
356
+ "it-IT": "Italian",
357
+ "it-SM": "Italian",
358
+ "it-VA": "Italian",
359
+ "iu": "Inuktitut",
360
+ "iu-Latn": "Inuktitut",
361
+ "ja": "Japanese",
362
+ "ja-JP": "Japanese",
363
+ "jv": "Javanese",
364
+ "ka": "Georgian",
365
+ "ki": "Kikuyu",
366
+ "kk": "Kazakh",
367
+ "kk-Arab": "Kazakh",
368
+ "kk-Cyrl": "Kazakh",
369
+ "kk-KZ": "Kazakh",
370
+ "kl": "Kalaallisut",
371
+ "km": "Central Khmer",
372
+ "kn": "Kannada",
373
+ "kn-IN": "Kannada",
374
+ "ko": "Korean",
375
+ "ko-CN": "Korean",
376
+ "ko-KP": "Korean",
377
+ "ko-KR": "Korean",
378
+ "ks": "Kashmiri",
379
+ "ks-Arab": "Kashmiri",
380
+ "ks-Deva": "Kashmiri",
381
+ "ku": "Kurdish",
382
+ "kw": "Cornish",
383
+ "ky": "Kyrgyz",
384
+ "la": "Latin",
385
+ "lb": "Luxembourgish",
386
+ "lg": "Ganda",
387
+ "ln": "Lingala",
388
+ "ln-AO": "Lingala",
389
+ "ln-CF": "Lingala",
390
+ "ln-CG": "Lingala",
391
+ "lo": "Lao",
392
+ "lt": "Lithuanian",
393
+ "lt-LT": "Lithuanian",
394
+ "lu": "Luba-Katanga",
395
+ "lv": "Latvian",
396
+ "lv-LV": "Latvian",
397
+ "mg": "Malagasy",
398
+ "mi": "Maori",
399
+ "mk": "Macedonian",
400
+ "ml": "Malayalam",
401
+ "ml-IN": "Malayalam",
402
+ "mn": "Mongolian",
403
+ "mn-Mong": "Mongolian",
404
+ "mn-Mong-MN": "Mongolian",
405
+ "mr": "Marathi",
406
+ "mr-IN": "Marathi",
407
+ "ms": "Malay",
408
+ "ms-Arab": "Malay",
409
+ "ms-Arab-BN": "Malay",
410
+ "ms-BN": "Malay",
411
+ "ms-ID": "Malay",
412
+ "ms-SG": "Malay",
413
+ "mt": "Maltese",
414
+ "my": "Burmese",
415
+ "nb": "Norwegian Bokmål",
416
+ "nb-SJ": "Norwegian Bokmål",
417
+ "nd": "North Ndebele",
418
+ "ne": "Nepali",
419
+ "ne-IN": "Nepali",
420
+ "nl": "Dutch",
421
+ "nl-AW": "Dutch",
422
+ "nl-BE": "Dutch",
423
+ "nl-BQ": "Dutch",
424
+ "nl-CW": "Dutch",
425
+ "nl-NL": "Dutch",
426
+ "nl-SR": "Dutch",
427
+ "nl-SX": "Dutch",
428
+ "nn": "Norwegian Nynorsk",
429
+ "no": "Norwegian",
430
+ "no-NO": "Norwegian",
431
+ "nr": "South Ndebele",
432
+ "nv": "Navajo",
433
+ "ny": "Chichewa",
434
+ "oc": "Occitan",
435
+ "oc-ES": "Occitan",
436
+ "om": "Oromo",
437
+ "om-KE": "Oromo",
438
+ "or": "Oriya",
439
+ "os": "Ossetian",
440
+ "os-RU": "Ossetian",
441
+ "pa": "Punjabi",
442
+ "pa-IN": "Punjabi",
443
+ "pa-Arab": "Punjabi",
444
+ "pa-Guru": "Punjabi",
445
+ "pl": "Polish",
446
+ "pl-PL": "Polish",
447
+ "ps": "Pashto",
448
+ "ps-PK": "Pashto",
449
+ "pt": "Portuguese",
450
+ "pt-AO": "Portuguese",
451
+ "pt-BR": "Portuguese",
452
+ "pt-CH": "Portuguese",
453
+ "pt-CV": "Portuguese",
454
+ "pt-GQ": "Portuguese",
455
+ "pt-GW": "Portuguese",
456
+ "pt-LU": "Portuguese",
457
+ "pt-MO": "Portuguese",
458
+ "pt-MZ": "Portuguese",
459
+ "pt-PT": "Portuguese",
460
+ "pt-ST": "Portuguese",
461
+ "pt-TL": "Portuguese",
462
+ "qu": "Quechua",
463
+ "qu-BO": "Quechua",
464
+ "qu-EC": "Quechua",
465
+ "rm": "Romansh",
466
+ "rn": "Rundi",
467
+ "ro": "Romanian",
468
+ "ro-MD": "Romanian",
469
+ "ro-RO": "Romanian",
470
+ "ru": "Russian",
471
+ "ru-BY": "Russian",
472
+ "ru-KG": "Russian",
473
+ "ru-KZ": "Russian",
474
+ "ru-MD": "Russian",
475
+ "ru-RU": "Russian",
476
+ "ru-UA": "Russian",
477
+ "rw": "Kinyarwanda",
478
+ "sa": "Sanskrit",
479
+ "sc": "Sardinian",
480
+ "sd": "Sindhi",
481
+ "sd-Arab": "Sindhi",
482
+ "sd-Deva": "Sindhi",
483
+ "se": "Northern Sami",
484
+ "se-FI": "Northern Sami",
485
+ "se-SE": "Northern Sami",
486
+ "sg": "Sango",
487
+ "si": "Sinhala",
488
+ "sk": "Slovak",
489
+ "sk-SK": "Slovak",
490
+ "sl": "Slovenian",
491
+ "sl-SI": "Slovenian",
492
+ "sn": "Shona",
493
+ "so": "Somali",
494
+ "so-DJ": "Somali",
495
+ "so-ET": "Somali",
496
+ "so-KE": "Somali",
497
+ "sq": "Albanian",
498
+ "sq-MK": "Albanian",
499
+ "sq-XK": "Albanian",
500
+ "sr": "Serbian",
501
+ "sr-RS": "Serbian",
502
+ "sr-Cyrl": "Serbian",
503
+ "sr-Cyrl-BA": "Serbian",
504
+ "sr-Cyrl-ME": "Serbian",
505
+ "sr-Cyrl-XK": "Serbian",
506
+ "sr-Latn": "Serbian",
507
+ "sr-Latn-BA": "Serbian",
508
+ "sr-Latn-ME": "Serbian",
509
+ "sr-Latn-XK": "Serbian",
510
+ "ss": "Swati",
511
+ "ss-SZ": "Swati",
512
+ "st": "Southern Sotho",
513
+ "st-LS": "Southern Sotho",
514
+ "su": "Sundanese",
515
+ "su-Latn": "Sundanese",
516
+ "sv": "Swedish",
517
+ "sv-AX": "Swedish",
518
+ "sv-FI": "Swedish",
519
+ "sv-SE": "Swedish",
520
+ "sw": "Swahili",
521
+ "sw-CD": "Swahili",
522
+ "sw-KE": "Swahili",
523
+ "sw-TZ": "Swahili",
524
+ "sw-UG": "Swahili",
525
+ "ta": "Tamil",
526
+ "ta-IN": "Tamil",
527
+ "ta-LK": "Tamil",
528
+ "ta-MY": "Tamil",
529
+ "ta-SG": "Tamil",
530
+ "te": "Telugu",
531
+ "te-IN": "Telugu",
532
+ "tg": "Tajik",
533
+ "th": "Thai",
534
+ "th-TH": "Thai",
535
+ "ti": "Tigrinya",
536
+ "ti-ER": "Tigrinya",
537
+ "tk": "Turkmen",
538
+ "tl": "Tagalog",
539
+ "tn": "Tswana",
540
+ "tn-BW": "Tswana",
541
+ "to": "Tonga",
542
+ "tr": "Turkish",
543
+ "tr-CY": "Turkish",
544
+ "tr-TR": "Turkish",
545
+ "ts": "Tsonga",
546
+ "tt": "Tatar",
547
+ "ug": "Uyghur",
548
+ "uk": "Ukrainian",
549
+ "uk-UA": "Ukrainian",
550
+ "ur": "Urdu",
551
+ "ur-IN": "Urdu",
552
+ "ur-PK": "Urdu",
553
+ "uz": "Uzbek",
554
+ "uz-Arab": "Uzbek",
555
+ "uz-Cyrl": "Uzbek",
556
+ "uz-Latn": "Uzbek",
557
+ "ve": "Venda",
558
+ "vi": "Vietnamese",
559
+ "vi-VN": "Vietnamese",
560
+ "vo": "Volapük",
561
+ "wa": "Walloon",
562
+ "wo": "Wolof",
563
+ "xh": "Xhosa",
564
+ "yi": "Yiddish",
565
+ "yo": "Yoruba",
566
+ "yo-BJ": "Yoruba",
567
+ "za": "Zhuang",
568
+ "zh": "Chinese",
569
+ "zh-CH": "Chinese",
570
+ "zh-TW": "Chinese",
571
+ "zh-Hans": "Chinese",
572
+ "zh-Hans-HK": "Chinese",
573
+ "zh-Hans-MO": "Chinese",
574
+ "zh-Hans-MY": "Chinese",
575
+ "zh-Hans-SG": "Chinese",
576
+ "zh-Hant": "Chinese",
577
+ "zh-Hant-HK": "Chinese",
578
+ "zh-Hant-MO": "Chinese",
579
+ "zh-Hant-MY": "Chinese",
580
+ "zh-Latn": "Chinese",
581
+ "zu": "Zulu",
582
+ "zu-ZA": "Zulu",
583
+ }
584
+ -%}
585
+ {{ bos_token }}
586
+ {%- if (messages[0]['role'] != 'user') -%}
587
+ {{ raise_exception("Conversations must start with a user prompt.") }}
588
+ {%- endif -%}
589
+ {%- for message in messages -%}
590
+ {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
591
+ {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
592
+ {%- endif -%}
593
+ {%- if (message['role'] == 'assistant') -%}
594
+ {%- if message['content'] is none or message['content'] is not string -%}
595
+ {{ raise_exception("Assistant role must provide content as a string") }}
596
+ {%- endif -%}
597
+ {{ '<start_of_turn>model\n'}}
598
+ {{ message["content"] | trim }}
599
+ {%- elif (message['role'] == 'user') -%}
600
+ {%- if message['content'] is none or message['content'] is not iterable or message['content'] | length != 1 -%}
601
+ {{ raise_exception(
602
+ "User role must provide `content` as an iterable with exactly one item. That item must be a " +
603
+ "`mapping(type:'text' | 'image', source_lang_code:string, target_lang_code:string, " +
604
+ "text:string | none, image:string | none)`."
605
+ ) }}
606
+ {%- endif -%}
607
+ {%- set content = message["content"][0] -%}
608
+ {%- set source_lang_code = content["source_lang_code"] | replace("_", "-") -%}
609
+ {%- set source_lang = languages[source_lang_code] -%}
610
+ {%- set target_lang_code = content["target_lang_code"] | replace("_", "-") -%}
611
+ {%- set target_lang = languages[target_lang_code] -%}
612
+ {{ '<start_of_turn>user\nYou are a professional ' + source_lang + ' (' + source_lang_code + ') to ' +
613
+ target_lang + ' (' + target_lang_code + ') translator. Your goal is to accurately convey the meaning and '
614
+ 'nuances of the original ' + source_lang + ' text while adhering to ' + target_lang + ' grammar, '
615
+ 'vocabulary, and cultural sensitivities.\n'
616
+ }}
617
+ {%- if content["type"] == 'text' -%}
618
+ {{
619
+ 'Produce only the ' + target_lang + ' translation, without any additional explanations or ' +
620
+ 'commentary. Please translate the following ' + source_lang + ' text into ' + target_lang + ':\n\n\n' +
621
+ content["text"] | trim
622
+ }}
623
+ {%- elif content["type"] == 'image' -%}
624
+ {{
625
+ 'Please translate the ' + source_lang + ' text in the provided image into ' + target_lang + '. ' +
626
+ 'Produce only the ' + target_lang + ' translation, without any additional explanations, ' +
627
+ 'alternatives or commentary. Focus only on the text, do not output where the text is located, ' +
628
+ 'surrounding objects or any other explanation about the picture. Ignore symbols, pictogram, and ' +
629
+ 'arrows!\n\n\n<start_of_image>'
630
+ }}
631
+ {%- else -%}
632
+ {{ raise_exception("User role `content` can contain either `text` or `image` data.") }}
633
+ {%- endif -%}
634
+ {%- else -%}
635
+ {{ raise_exception("Conversations must only contain user or assistant roles.") }}
636
+ {%- endif -%}
637
+ {{ '<end_of_turn>\n' }}
638
+ {%- endfor -%}
639
+ {%- if add_generation_prompt -%}
640
+ {{'<start_of_turn>model\n'}}
641
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,312 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "dtype": "bfloat16",
7
+ "eoi_token_index": 256000,
8
+ "image_token_index": 262144,
9
+ "initializer_range": 0.02,
10
+ "mm_tokens_per_image": 256,
11
+ "model_type": "gemma3",
12
+ "quantization_config": {
13
+ "config_groups": {
14
+ "group_0": {
15
+ "format": "float-quantized",
16
+ "input_activations": {
17
+ "actorder": null,
18
+ "block_structure": null,
19
+ "dynamic": true,
20
+ "group_size": null,
21
+ "num_bits": 8,
22
+ "observer": null,
23
+ "observer_kwargs": {},
24
+ "scale_dtype": null,
25
+ "strategy": "token",
26
+ "symmetric": true,
27
+ "type": "float",
28
+ "zp_dtype": null
29
+ },
30
+ "output_activations": null,
31
+ "targets": [
32
+ "Linear"
33
+ ],
34
+ "weights": {
35
+ "actorder": null,
36
+ "block_structure": null,
37
+ "dynamic": false,
38
+ "group_size": null,
39
+ "num_bits": 8,
40
+ "observer": "minmax",
41
+ "observer_kwargs": {},
42
+ "scale_dtype": null,
43
+ "strategy": "channel",
44
+ "symmetric": true,
45
+ "type": "float",
46
+ "zp_dtype": null
47
+ }
48
+ }
49
+ },
50
+ "format": "float-quantized",
51
+ "global_compression_ratio": null,
52
+ "ignore": [
53
+ "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj",
54
+ "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj",
55
+ "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj",
56
+ "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj",
57
+ "model.vision_tower.vision_model.encoder.layers.0.mlp.fc1",
58
+ "model.vision_tower.vision_model.encoder.layers.0.mlp.fc2",
59
+ "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj",
60
+ "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj",
61
+ "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj",
62
+ "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj",
63
+ "model.vision_tower.vision_model.encoder.layers.1.mlp.fc1",
64
+ "model.vision_tower.vision_model.encoder.layers.1.mlp.fc2",
65
+ "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj",
66
+ "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj",
67
+ "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj",
68
+ "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj",
69
+ "model.vision_tower.vision_model.encoder.layers.2.mlp.fc1",
70
+ "model.vision_tower.vision_model.encoder.layers.2.mlp.fc2",
71
+ "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj",
72
+ "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj",
73
+ "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj",
74
+ "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj",
75
+ "model.vision_tower.vision_model.encoder.layers.3.mlp.fc1",
76
+ "model.vision_tower.vision_model.encoder.layers.3.mlp.fc2",
77
+ "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj",
78
+ "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj",
79
+ "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj",
80
+ "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj",
81
+ "model.vision_tower.vision_model.encoder.layers.4.mlp.fc1",
82
+ "model.vision_tower.vision_model.encoder.layers.4.mlp.fc2",
83
+ "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj",
84
+ "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj",
85
+ "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj",
86
+ "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj",
87
+ "model.vision_tower.vision_model.encoder.layers.5.mlp.fc1",
88
+ "model.vision_tower.vision_model.encoder.layers.5.mlp.fc2",
89
+ "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj",
90
+ "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj",
91
+ "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj",
92
+ "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj",
93
+ "model.vision_tower.vision_model.encoder.layers.6.mlp.fc1",
94
+ "model.vision_tower.vision_model.encoder.layers.6.mlp.fc2",
95
+ "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj",
96
+ "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj",
97
+ "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj",
98
+ "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj",
99
+ "model.vision_tower.vision_model.encoder.layers.7.mlp.fc1",
100
+ "model.vision_tower.vision_model.encoder.layers.7.mlp.fc2",
101
+ "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj",
102
+ "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj",
103
+ "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj",
104
+ "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj",
105
+ "model.vision_tower.vision_model.encoder.layers.8.mlp.fc1",
106
+ "model.vision_tower.vision_model.encoder.layers.8.mlp.fc2",
107
+ "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj",
108
+ "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj",
109
+ "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj",
110
+ "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj",
111
+ "model.vision_tower.vision_model.encoder.layers.9.mlp.fc1",
112
+ "model.vision_tower.vision_model.encoder.layers.9.mlp.fc2",
113
+ "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj",
114
+ "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj",
115
+ "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj",
116
+ "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj",
117
+ "model.vision_tower.vision_model.encoder.layers.10.mlp.fc1",
118
+ "model.vision_tower.vision_model.encoder.layers.10.mlp.fc2",
119
+ "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj",
120
+ "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj",
121
+ "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj",
122
+ "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj",
123
+ "model.vision_tower.vision_model.encoder.layers.11.mlp.fc1",
124
+ "model.vision_tower.vision_model.encoder.layers.11.mlp.fc2",
125
+ "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj",
126
+ "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj",
127
+ "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj",
128
+ "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj",
129
+ "model.vision_tower.vision_model.encoder.layers.12.mlp.fc1",
130
+ "model.vision_tower.vision_model.encoder.layers.12.mlp.fc2",
131
+ "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj",
132
+ "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj",
133
+ "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj",
134
+ "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj",
135
+ "model.vision_tower.vision_model.encoder.layers.13.mlp.fc1",
136
+ "model.vision_tower.vision_model.encoder.layers.13.mlp.fc2",
137
+ "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj",
138
+ "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj",
139
+ "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj",
140
+ "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj",
141
+ "model.vision_tower.vision_model.encoder.layers.14.mlp.fc1",
142
+ "model.vision_tower.vision_model.encoder.layers.14.mlp.fc2",
143
+ "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj",
144
+ "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj",
145
+ "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj",
146
+ "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj",
147
+ "model.vision_tower.vision_model.encoder.layers.15.mlp.fc1",
148
+ "model.vision_tower.vision_model.encoder.layers.15.mlp.fc2",
149
+ "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj",
150
+ "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj",
151
+ "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj",
152
+ "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj",
153
+ "model.vision_tower.vision_model.encoder.layers.16.mlp.fc1",
154
+ "model.vision_tower.vision_model.encoder.layers.16.mlp.fc2",
155
+ "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj",
156
+ "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj",
157
+ "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj",
158
+ "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj",
159
+ "model.vision_tower.vision_model.encoder.layers.17.mlp.fc1",
160
+ "model.vision_tower.vision_model.encoder.layers.17.mlp.fc2",
161
+ "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj",
162
+ "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj",
163
+ "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj",
164
+ "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj",
165
+ "model.vision_tower.vision_model.encoder.layers.18.mlp.fc1",
166
+ "model.vision_tower.vision_model.encoder.layers.18.mlp.fc2",
167
+ "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj",
168
+ "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj",
169
+ "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj",
170
+ "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj",
171
+ "model.vision_tower.vision_model.encoder.layers.19.mlp.fc1",
172
+ "model.vision_tower.vision_model.encoder.layers.19.mlp.fc2",
173
+ "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj",
174
+ "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj",
175
+ "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj",
176
+ "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj",
177
+ "model.vision_tower.vision_model.encoder.layers.20.mlp.fc1",
178
+ "model.vision_tower.vision_model.encoder.layers.20.mlp.fc2",
179
+ "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj",
180
+ "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj",
181
+ "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj",
182
+ "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj",
183
+ "model.vision_tower.vision_model.encoder.layers.21.mlp.fc1",
184
+ "model.vision_tower.vision_model.encoder.layers.21.mlp.fc2",
185
+ "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj",
186
+ "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj",
187
+ "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj",
188
+ "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj",
189
+ "model.vision_tower.vision_model.encoder.layers.22.mlp.fc1",
190
+ "model.vision_tower.vision_model.encoder.layers.22.mlp.fc2",
191
+ "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj",
192
+ "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj",
193
+ "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj",
194
+ "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj",
195
+ "model.vision_tower.vision_model.encoder.layers.23.mlp.fc1",
196
+ "model.vision_tower.vision_model.encoder.layers.23.mlp.fc2",
197
+ "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj",
198
+ "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj",
199
+ "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj",
200
+ "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj",
201
+ "model.vision_tower.vision_model.encoder.layers.24.mlp.fc1",
202
+ "model.vision_tower.vision_model.encoder.layers.24.mlp.fc2",
203
+ "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj",
204
+ "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj",
205
+ "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj",
206
+ "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj",
207
+ "model.vision_tower.vision_model.encoder.layers.25.mlp.fc1",
208
+ "model.vision_tower.vision_model.encoder.layers.25.mlp.fc2",
209
+ "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj",
210
+ "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj",
211
+ "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj",
212
+ "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj",
213
+ "model.vision_tower.vision_model.encoder.layers.26.mlp.fc1",
214
+ "model.vision_tower.vision_model.encoder.layers.26.mlp.fc2",
215
+ "lm_head"
216
+ ],
217
+ "kv_cache_scheme": null,
218
+ "quant_method": "compressed-tensors",
219
+ "quantization_status": "compressed",
220
+ "sparsity_config": {},
221
+ "transform_config": {},
222
+ "version": "0.13.1.a20260115"
223
+ },
224
+ "text_config": {
225
+ "_sliding_window_pattern": 6,
226
+ "attention_bias": false,
227
+ "attention_dropout": 0.0,
228
+ "attn_logit_softcapping": null,
229
+ "dtype": "bfloat16",
230
+ "final_logit_softcapping": null,
231
+ "head_dim": 256,
232
+ "hidden_activation": "gelu_pytorch_tanh",
233
+ "hidden_size": 2560,
234
+ "initializer_range": 0.02,
235
+ "intermediate_size": 10240,
236
+ "layer_types": [
237
+ "sliding_attention",
238
+ "sliding_attention",
239
+ "sliding_attention",
240
+ "sliding_attention",
241
+ "sliding_attention",
242
+ "full_attention",
243
+ "sliding_attention",
244
+ "sliding_attention",
245
+ "sliding_attention",
246
+ "sliding_attention",
247
+ "sliding_attention",
248
+ "full_attention",
249
+ "sliding_attention",
250
+ "sliding_attention",
251
+ "sliding_attention",
252
+ "sliding_attention",
253
+ "sliding_attention",
254
+ "full_attention",
255
+ "sliding_attention",
256
+ "sliding_attention",
257
+ "sliding_attention",
258
+ "sliding_attention",
259
+ "sliding_attention",
260
+ "full_attention",
261
+ "sliding_attention",
262
+ "sliding_attention",
263
+ "sliding_attention",
264
+ "sliding_attention",
265
+ "sliding_attention",
266
+ "full_attention",
267
+ "sliding_attention",
268
+ "sliding_attention",
269
+ "sliding_attention",
270
+ "sliding_attention"
271
+ ],
272
+ "max_position_embeddings": 131072,
273
+ "model_type": "gemma3_text",
274
+ "num_attention_heads": 8,
275
+ "num_hidden_layers": 34,
276
+ "num_key_value_heads": 4,
277
+ "query_pre_attn_scalar": 256,
278
+ "rms_norm_eps": 1e-06,
279
+ "rope_local_base_freq": 10000,
280
+ "rope_parameters": {
281
+ "full_attention": {
282
+ "factor": 8.0,
283
+ "rope_type": "linear"
284
+ },
285
+ "sliding_attention": {
286
+ "rope_type": "default"
287
+ }
288
+ },
289
+ "rope_scaling": null,
290
+ "rope_theta": 1000000,
291
+ "sliding_window": 1024,
292
+ "use_bidirectional_attention": false,
293
+ "use_cache": true,
294
+ "vocab_size": 262208
295
+ },
296
+ "transformers_version": "4.57.3",
297
+ "vision_config": {
298
+ "attention_dropout": 0.0,
299
+ "dtype": "bfloat16",
300
+ "hidden_act": "gelu_pytorch_tanh",
301
+ "hidden_size": 1152,
302
+ "image_size": 896,
303
+ "intermediate_size": 4304,
304
+ "layer_norm_eps": 1e-06,
305
+ "model_type": "siglip_vision_model",
306
+ "num_attention_heads": 16,
307
+ "num_channels": 3,
308
+ "num_hidden_layers": 27,
309
+ "patch_size": 14,
310
+ "vision_use_head": false
311
+ }
312
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cache_implementation": "hybrid",
3
+ "do_sample": true,
4
+ "top_k": 64,
5
+ "top_p": 0.95,
6
+ "transformers_version": "4.57.3"
7
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93bfd07670edd6cdefd942aec484e7c1da5c7d94cfb99b3dfb26ed0e4f1b52da
3
+ size 4989619792
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8c0501a88bcf371afb7a2593c008f68b1072476f77506a505155ba18580c77e
3
+ size 404067168
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
recipe.yaml ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ default_stage:
2
+ default_modifiers:
3
+ QuantizationModifier:
4
+ targets: [Linear]
5
+ ignore: [lm_head, 're:model\.vision_tower.*', 're:model\.multi_modal_projector.*']
6
+ scheme: FP8_DYNAMIC
special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<eos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
3
+ size 33384568
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff