kittyn commited on
Commit
50019c3
·
verified ·
1 Parent(s): c87dd5b

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,6 @@ 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
+ translation/nllb-200-1.3B/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ translation/nllb-200-distilled-1.3B/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ translation/nllb-200-distilled-600M/tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: mixed-see-below
4
+ tags:
5
+ - automatic-speech-recognition
6
+ - translation
7
+ - whisper.cpp
8
+ - ctranslate2
9
+ - vrstt
10
+ ---
11
+
12
+ # vrSTT models
13
+
14
+ Model weights for [vrSTT](https://github.com/kittynXR/vrSTT) — local speech-to-text
15
+ for VR and streaming. Everything vrSTT runs is here so the app can fetch the one
16
+ model a user chose, verify it, and never touch the network again.
17
+
18
+ **This repository redistributes third-party models. It trains nothing and
19
+ modifies nothing except file format.** Licences differ per directory and are
20
+ stated below; the upstream terms travel with the weights.
21
+
22
+ ## Layout
23
+
24
+ ```
25
+ manifest.json # what the fetcher reads first
26
+ whisper/ggml-<name>.bin # whisper.cpp ggml format, one file each
27
+ translation/<name>/ # CTranslate2 format, a directory each
28
+ ```
29
+
30
+ `manifest.json` carries a `sha256` and a byte count for every file. The fetcher
31
+ verifies both before handing a path to whisper.cpp or CTranslate2 — a truncated
32
+ ggml file otherwise fails at load with an error about magic bytes, which sends
33
+ you looking at the wrong thing.
34
+
35
+ ## What is here
36
+
37
+ **Speech recognition** — [whisper.cpp](https://github.com/ggerganov/whisper.cpp)
38
+ ggml conversions, byte-identical to
39
+ [ggerganov/whisper.cpp](https://huggingface.co/ggerganov/whisper.cpp). Mirrored
40
+ rather than linked so the bytes vrSTT downloads are pinned and hashed.
41
+
42
+ `tiny`, `tiny.en`, `base`, `base.en`, `small`, `small.en`, `medium`,
43
+ `medium.en`, `large-v1`, `large-v2`, `large-v3`, `large-v3-turbo`
44
+
45
+ vrSTT defaults to **`small.en` for commits and `base.en` for live previews** —
46
+ the split is deliberate, previews are allowed to be wrong and are replaced within
47
+ a second, so they should not spend the larger model.
48
+
49
+ **Translation** — [NLLB-200](https://ai.meta.com/research/no-language-left-behind/)
50
+ converted to CTranslate2 with `ct2-transformers-converter --quantization
51
+ float16`, from `facebook/nllb-200-*`.
52
+
53
+ `nllb-200-distilled-600M` (default), `nllb-200-distilled-1.3B`, `nllb-200-1.3B`
54
+
55
+ float16 rather than int8 because CTranslate2 can cast **down** at load time but
56
+ not up — one artifact serves both, and converting to int8 here would cap quality
57
+ for everyone. 600M is the default because vrSTT's stated minimum-VRAM target is
58
+ 8 GB, and a 1.3B model beside the ASR stages and a game does not fit
59
+ comfortably there.
60
+
61
+ ## Licences — read this before redistributing further
62
+
63
+ | Directory | Licence | Commercial use |
64
+ |---|---|---|
65
+ | `whisper/` | **MIT** (OpenAI Whisper; ggml conversions by Georgi Gerganov) | permitted |
66
+ | `translation/` | **CC-BY-NC-4.0** (Meta AI, NLLB-200) | **NOT permitted** |
67
+
68
+ **NLLB-200 is non-commercial.** Redistribution is allowed with attribution and
69
+ the licence must travel with it — which is what this section is — but neither
70
+ Meta's weights nor anything derived from them may be used commercially. If vrSTT
71
+ ever becomes a paid product, the translation models have to be replaced rather
72
+ than relicensed.
73
+
74
+ NLLB-200 also carries Meta's own caveats, which apply unchanged here: it is
75
+ research work, not certified for production translation, it was trained on
76
+ Wikimedia-domain text and does not target long-form or domain-specific input,
77
+ and translation quality varies widely by language pair. Mistranslations can
78
+ carry real consequences and vrSTT displays them live on a stream — the
79
+ `include_original` setting exists so the source text stays visible beside the
80
+ translation for exactly this reason, and it is on by default.
81
+
82
+ ## Citation
83
+
84
+ Whisper — Radford et al., *Robust Speech Recognition via Large-Scale Weak
85
+ Supervision*, OpenAI, 2022.
86
+
87
+ NLLB-200 — NLLB Team et al., *No Language Left Behind: Scaling Human-Centered
88
+ Machine Translation*, Meta AI, 2022.
manifest.json ADDED
@@ -0,0 +1,233 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "schema": 1,
3
+ "repo": "kittyn/vrstt-models",
4
+ "revision": "main",
5
+ "generated_by": "scripts/make-model-manifest.ps1",
6
+ "models": {
7
+ "whisper/base": {
8
+ "kind": "whisper",
9
+ "name": "base",
10
+ "files": [
11
+ {
12
+ "path": "whisper/ggml-base.bin",
13
+ "bytes": 147951465,
14
+ "sha256": "60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe"
15
+ }
16
+ ]
17
+ },
18
+ "whisper/base.en": {
19
+ "kind": "whisper",
20
+ "name": "base.en",
21
+ "files": [
22
+ {
23
+ "path": "whisper/ggml-base.en.bin",
24
+ "bytes": 147964211,
25
+ "sha256": "a03779c86df3323075f5e796cb2ce5029f00ec8869eee3fdfb897afe36c6d002"
26
+ }
27
+ ]
28
+ },
29
+ "whisper/large-v1": {
30
+ "kind": "whisper",
31
+ "name": "large-v1",
32
+ "files": [
33
+ {
34
+ "path": "whisper/ggml-large-v1.bin",
35
+ "bytes": 3094623691,
36
+ "sha256": "7d99f41a10525d0206bddadd86760181fa920438b6b33237e3118ff6c83bb53d"
37
+ }
38
+ ]
39
+ },
40
+ "whisper/large-v2": {
41
+ "kind": "whisper",
42
+ "name": "large-v2",
43
+ "files": [
44
+ {
45
+ "path": "whisper/ggml-large-v2.bin",
46
+ "bytes": 3094623691,
47
+ "sha256": "9a423fe4d40c82774b6af34115b8b935f34152246eb19e80e376071d3f999487"
48
+ }
49
+ ]
50
+ },
51
+ "whisper/large-v3-turbo": {
52
+ "kind": "whisper",
53
+ "name": "large-v3-turbo",
54
+ "files": [
55
+ {
56
+ "path": "whisper/ggml-large-v3-turbo.bin",
57
+ "bytes": 1624555275,
58
+ "sha256": "1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69"
59
+ }
60
+ ]
61
+ },
62
+ "whisper/large-v3": {
63
+ "kind": "whisper",
64
+ "name": "large-v3",
65
+ "files": [
66
+ {
67
+ "path": "whisper/ggml-large-v3.bin",
68
+ "bytes": 3095033483,
69
+ "sha256": "64d182b440b98d5203c4f9bd541544d84c605196c4f7b845dfa11fb23594d1e2"
70
+ }
71
+ ]
72
+ },
73
+ "whisper/medium": {
74
+ "kind": "whisper",
75
+ "name": "medium",
76
+ "files": [
77
+ {
78
+ "path": "whisper/ggml-medium.bin",
79
+ "bytes": 1533763059,
80
+ "sha256": "6c14d5adee5f86394037b4e4e8b59f1673b6cee10e3cf0b11bbdbee79c156208"
81
+ }
82
+ ]
83
+ },
84
+ "whisper/medium.en": {
85
+ "kind": "whisper",
86
+ "name": "medium.en",
87
+ "files": [
88
+ {
89
+ "path": "whisper/ggml-medium.en.bin",
90
+ "bytes": 1533774781,
91
+ "sha256": "cc37e93478338ec7700281a7ac30a10128929eb8f427dda2e865faa8f6da4356"
92
+ }
93
+ ]
94
+ },
95
+ "whisper/small": {
96
+ "kind": "whisper",
97
+ "name": "small",
98
+ "files": [
99
+ {
100
+ "path": "whisper/ggml-small.bin",
101
+ "bytes": 487601967,
102
+ "sha256": "1be3a9b2063867b937e64e2ec7483364a79917e157fa98c5d94b5c1fffea987b"
103
+ }
104
+ ]
105
+ },
106
+ "whisper/small.en": {
107
+ "kind": "whisper",
108
+ "name": "small.en",
109
+ "files": [
110
+ {
111
+ "path": "whisper/ggml-small.en.bin",
112
+ "bytes": 487614201,
113
+ "sha256": "c6138d6d58ecc8322097e0f987c32f1be8bb0a18532a3f88f734d1bbf9c41e5d"
114
+ }
115
+ ]
116
+ },
117
+ "whisper/tiny": {
118
+ "kind": "whisper",
119
+ "name": "tiny",
120
+ "files": [
121
+ {
122
+ "path": "whisper/ggml-tiny.bin",
123
+ "bytes": 77691713,
124
+ "sha256": "be07e048e1e599ad46341c8d2a135645097a538221678b7acdd1b1919c6e1b21"
125
+ }
126
+ ]
127
+ },
128
+ "whisper/tiny.en": {
129
+ "kind": "whisper",
130
+ "name": "tiny.en",
131
+ "files": [
132
+ {
133
+ "path": "whisper/ggml-tiny.en.bin",
134
+ "bytes": 77704715,
135
+ "sha256": "921e4cf8686fdd993dcd081a5da5b6c365bfde1162e72b08d75ac75289920b1f"
136
+ }
137
+ ]
138
+ },
139
+ "translation/nllb-200-1.3B": {
140
+ "kind": "translation",
141
+ "name": "nllb-200-1.3B",
142
+ "files": [
143
+ {
144
+ "path": "translation/nllb-200-1.3B/config.json",
145
+ "bytes": 233,
146
+ "sha256": "72901fbd8abd89fb5cf4a388f26fc681f5c4c58a1e1a88b30b879f107270e7ee"
147
+ },
148
+ {
149
+ "path": "translation/nllb-200-1.3B/model.bin",
150
+ "bytes": 2743417091,
151
+ "sha256": "5cd66fd6bebdd79c2d0d9736c66eeac2c6f48c50473d462568ad60e63ab8098b"
152
+ },
153
+ {
154
+ "path": "translation/nllb-200-1.3B/shared_vocabulary.json",
155
+ "bytes": 6177383,
156
+ "sha256": "768aa4170693765cb4c62fc485ea4fd954cd3e5a44bc645d11d83f41f2393776"
157
+ },
158
+ {
159
+ "path": "translation/nllb-200-1.3B/special_tokens_map.json",
160
+ "bytes": 3548,
161
+ "sha256": "992bd4ed610d644d6823081937bcc91bb8878dd556cea4ae5327f2480361330e"
162
+ },
163
+ {
164
+ "path": "translation/nllb-200-1.3B/tokenizer.json",
165
+ "bytes": 17331176,
166
+ "sha256": "e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665"
167
+ }
168
+ ]
169
+ },
170
+ "translation/nllb-200-distilled-1.3B": {
171
+ "kind": "translation",
172
+ "name": "nllb-200-distilled-1.3B",
173
+ "files": [
174
+ {
175
+ "path": "translation/nllb-200-distilled-1.3B/config.json",
176
+ "bytes": 233,
177
+ "sha256": "72901fbd8abd89fb5cf4a388f26fc681f5c4c58a1e1a88b30b879f107270e7ee"
178
+ },
179
+ {
180
+ "path": "translation/nllb-200-distilled-1.3B/model.bin",
181
+ "bytes": 2743417091,
182
+ "sha256": "33fb28bddc52840dab354abdd15dc4836829272d477fb2df275855d90b4e5eb3"
183
+ },
184
+ {
185
+ "path": "translation/nllb-200-distilled-1.3B/shared_vocabulary.json",
186
+ "bytes": 6177383,
187
+ "sha256": "768aa4170693765cb4c62fc485ea4fd954cd3e5a44bc645d11d83f41f2393776"
188
+ },
189
+ {
190
+ "path": "translation/nllb-200-distilled-1.3B/special_tokens_map.json",
191
+ "bytes": 3548,
192
+ "sha256": "992bd4ed610d644d6823081937bcc91bb8878dd556cea4ae5327f2480361330e"
193
+ },
194
+ {
195
+ "path": "translation/nllb-200-distilled-1.3B/tokenizer.json",
196
+ "bytes": 17331176,
197
+ "sha256": "e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665"
198
+ }
199
+ ]
200
+ },
201
+ "translation/nllb-200-distilled-600M": {
202
+ "kind": "translation",
203
+ "name": "nllb-200-distilled-600M",
204
+ "files": [
205
+ {
206
+ "path": "translation/nllb-200-distilled-600M/config.json",
207
+ "bytes": 233,
208
+ "sha256": "72901fbd8abd89fb5cf4a388f26fc681f5c4c58a1e1a88b30b879f107270e7ee"
209
+ },
210
+ {
211
+ "path": "translation/nllb-200-distilled-600M/model.bin",
212
+ "bytes": 1232266703,
213
+ "sha256": "795cc73c70446f8ba0cad530e8dda72930f725b73124bb6e8bf0ab838a8eb5f1"
214
+ },
215
+ {
216
+ "path": "translation/nllb-200-distilled-600M/shared_vocabulary.json",
217
+ "bytes": 6177383,
218
+ "sha256": "768aa4170693765cb4c62fc485ea4fd954cd3e5a44bc645d11d83f41f2393776"
219
+ },
220
+ {
221
+ "path": "translation/nllb-200-distilled-600M/special_tokens_map.json",
222
+ "bytes": 3548,
223
+ "sha256": "992bd4ed610d644d6823081937bcc91bb8878dd556cea4ae5327f2480361330e"
224
+ },
225
+ {
226
+ "path": "translation/nllb-200-distilled-600M/tokenizer.json",
227
+ "bytes": 17331176,
228
+ "sha256": "e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665"
229
+ }
230
+ ]
231
+ }
232
+ }
233
+ }
translation/nllb-200-1.3B/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_source_bos": false,
3
+ "add_source_eos": false,
4
+ "bos_token": "<s>",
5
+ "decoder_start_token": "</s>",
6
+ "eos_token": "</s>",
7
+ "layer_norm_epsilon": null,
8
+ "multi_query_attention": false,
9
+ "unk_token": "<unk>"
10
+ }
translation/nllb-200-1.3B/model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cd66fd6bebdd79c2d0d9736c66eeac2c6f48c50473d462568ad60e63ab8098b
3
+ size 2743417091
translation/nllb-200-1.3B/shared_vocabulary.json ADDED
The diff for this file is too large to render. See raw diff
 
translation/nllb-200-1.3B/special_tokens_map.json ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ace_Arab",
4
+ "ace_Latn",
5
+ "acm_Arab",
6
+ "acq_Arab",
7
+ "aeb_Arab",
8
+ "afr_Latn",
9
+ "ajp_Arab",
10
+ "aka_Latn",
11
+ "amh_Ethi",
12
+ "apc_Arab",
13
+ "arb_Arab",
14
+ "ars_Arab",
15
+ "ary_Arab",
16
+ "arz_Arab",
17
+ "asm_Beng",
18
+ "ast_Latn",
19
+ "awa_Deva",
20
+ "ayr_Latn",
21
+ "azb_Arab",
22
+ "azj_Latn",
23
+ "bak_Cyrl",
24
+ "bam_Latn",
25
+ "ban_Latn",
26
+ "bel_Cyrl",
27
+ "bem_Latn",
28
+ "ben_Beng",
29
+ "bho_Deva",
30
+ "bjn_Arab",
31
+ "bjn_Latn",
32
+ "bod_Tibt",
33
+ "bos_Latn",
34
+ "bug_Latn",
35
+ "bul_Cyrl",
36
+ "cat_Latn",
37
+ "ceb_Latn",
38
+ "ces_Latn",
39
+ "cjk_Latn",
40
+ "ckb_Arab",
41
+ "crh_Latn",
42
+ "cym_Latn",
43
+ "dan_Latn",
44
+ "deu_Latn",
45
+ "dik_Latn",
46
+ "dyu_Latn",
47
+ "dzo_Tibt",
48
+ "ell_Grek",
49
+ "eng_Latn",
50
+ "epo_Latn",
51
+ "est_Latn",
52
+ "eus_Latn",
53
+ "ewe_Latn",
54
+ "fao_Latn",
55
+ "pes_Arab",
56
+ "fij_Latn",
57
+ "fin_Latn",
58
+ "fon_Latn",
59
+ "fra_Latn",
60
+ "fur_Latn",
61
+ "fuv_Latn",
62
+ "gla_Latn",
63
+ "gle_Latn",
64
+ "glg_Latn",
65
+ "grn_Latn",
66
+ "guj_Gujr",
67
+ "hat_Latn",
68
+ "hau_Latn",
69
+ "heb_Hebr",
70
+ "hin_Deva",
71
+ "hne_Deva",
72
+ "hrv_Latn",
73
+ "hun_Latn",
74
+ "hye_Armn",
75
+ "ibo_Latn",
76
+ "ilo_Latn",
77
+ "ind_Latn",
78
+ "isl_Latn",
79
+ "ita_Latn",
80
+ "jav_Latn",
81
+ "jpn_Jpan",
82
+ "kab_Latn",
83
+ "kac_Latn",
84
+ "kam_Latn",
85
+ "kan_Knda",
86
+ "kas_Arab",
87
+ "kas_Deva",
88
+ "kat_Geor",
89
+ "knc_Arab",
90
+ "knc_Latn",
91
+ "kaz_Cyrl",
92
+ "kbp_Latn",
93
+ "kea_Latn",
94
+ "khm_Khmr",
95
+ "kik_Latn",
96
+ "kin_Latn",
97
+ "kir_Cyrl",
98
+ "kmb_Latn",
99
+ "kon_Latn",
100
+ "kor_Hang",
101
+ "kmr_Latn",
102
+ "lao_Laoo",
103
+ "lvs_Latn",
104
+ "lij_Latn",
105
+ "lim_Latn",
106
+ "lin_Latn",
107
+ "lit_Latn",
108
+ "lmo_Latn",
109
+ "ltg_Latn",
110
+ "ltz_Latn",
111
+ "lua_Latn",
112
+ "lug_Latn",
113
+ "luo_Latn",
114
+ "lus_Latn",
115
+ "mag_Deva",
116
+ "mai_Deva",
117
+ "mal_Mlym",
118
+ "mar_Deva",
119
+ "min_Latn",
120
+ "mkd_Cyrl",
121
+ "plt_Latn",
122
+ "mlt_Latn",
123
+ "mni_Beng",
124
+ "khk_Cyrl",
125
+ "mos_Latn",
126
+ "mri_Latn",
127
+ "zsm_Latn",
128
+ "mya_Mymr",
129
+ "nld_Latn",
130
+ "nno_Latn",
131
+ "nob_Latn",
132
+ "npi_Deva",
133
+ "nso_Latn",
134
+ "nus_Latn",
135
+ "nya_Latn",
136
+ "oci_Latn",
137
+ "gaz_Latn",
138
+ "ory_Orya",
139
+ "pag_Latn",
140
+ "pan_Guru",
141
+ "pap_Latn",
142
+ "pol_Latn",
143
+ "por_Latn",
144
+ "prs_Arab",
145
+ "pbt_Arab",
146
+ "quy_Latn",
147
+ "ron_Latn",
148
+ "run_Latn",
149
+ "rus_Cyrl",
150
+ "sag_Latn",
151
+ "san_Deva",
152
+ "sat_Beng",
153
+ "scn_Latn",
154
+ "shn_Mymr",
155
+ "sin_Sinh",
156
+ "slk_Latn",
157
+ "slv_Latn",
158
+ "smo_Latn",
159
+ "sna_Latn",
160
+ "snd_Arab",
161
+ "som_Latn",
162
+ "sot_Latn",
163
+ "spa_Latn",
164
+ "als_Latn",
165
+ "srd_Latn",
166
+ "srp_Cyrl",
167
+ "ssw_Latn",
168
+ "sun_Latn",
169
+ "swe_Latn",
170
+ "swh_Latn",
171
+ "szl_Latn",
172
+ "tam_Taml",
173
+ "tat_Cyrl",
174
+ "tel_Telu",
175
+ "tgk_Cyrl",
176
+ "tgl_Latn",
177
+ "tha_Thai",
178
+ "tir_Ethi",
179
+ "taq_Latn",
180
+ "taq_Tfng",
181
+ "tpi_Latn",
182
+ "tsn_Latn",
183
+ "tso_Latn",
184
+ "tuk_Latn",
185
+ "tum_Latn",
186
+ "tur_Latn",
187
+ "twi_Latn",
188
+ "tzm_Tfng",
189
+ "uig_Arab",
190
+ "ukr_Cyrl",
191
+ "umb_Latn",
192
+ "urd_Arab",
193
+ "uzn_Latn",
194
+ "vec_Latn",
195
+ "vie_Latn",
196
+ "war_Latn",
197
+ "wol_Latn",
198
+ "xho_Latn",
199
+ "ydd_Hebr",
200
+ "yor_Latn",
201
+ "yue_Hant",
202
+ "zho_Hans",
203
+ "zho_Hant",
204
+ "zul_Latn"
205
+ ],
206
+ "bos_token": "<s>",
207
+ "cls_token": "<s>",
208
+ "eos_token": "</s>",
209
+ "mask_token": {
210
+ "content": "<mask>",
211
+ "lstrip": true,
212
+ "normalized": true,
213
+ "rstrip": false,
214
+ "single_word": false
215
+ },
216
+ "pad_token": "<pad>",
217
+ "sep_token": "</s>",
218
+ "unk_token": "<unk>"
219
+ }
translation/nllb-200-1.3B/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665
3
+ size 17331176
translation/nllb-200-distilled-1.3B/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_source_bos": false,
3
+ "add_source_eos": false,
4
+ "bos_token": "<s>",
5
+ "decoder_start_token": "</s>",
6
+ "eos_token": "</s>",
7
+ "layer_norm_epsilon": null,
8
+ "multi_query_attention": false,
9
+ "unk_token": "<unk>"
10
+ }
translation/nllb-200-distilled-1.3B/model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33fb28bddc52840dab354abdd15dc4836829272d477fb2df275855d90b4e5eb3
3
+ size 2743417091
translation/nllb-200-distilled-1.3B/shared_vocabulary.json ADDED
The diff for this file is too large to render. See raw diff
 
translation/nllb-200-distilled-1.3B/special_tokens_map.json ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ace_Arab",
4
+ "ace_Latn",
5
+ "acm_Arab",
6
+ "acq_Arab",
7
+ "aeb_Arab",
8
+ "afr_Latn",
9
+ "ajp_Arab",
10
+ "aka_Latn",
11
+ "amh_Ethi",
12
+ "apc_Arab",
13
+ "arb_Arab",
14
+ "ars_Arab",
15
+ "ary_Arab",
16
+ "arz_Arab",
17
+ "asm_Beng",
18
+ "ast_Latn",
19
+ "awa_Deva",
20
+ "ayr_Latn",
21
+ "azb_Arab",
22
+ "azj_Latn",
23
+ "bak_Cyrl",
24
+ "bam_Latn",
25
+ "ban_Latn",
26
+ "bel_Cyrl",
27
+ "bem_Latn",
28
+ "ben_Beng",
29
+ "bho_Deva",
30
+ "bjn_Arab",
31
+ "bjn_Latn",
32
+ "bod_Tibt",
33
+ "bos_Latn",
34
+ "bug_Latn",
35
+ "bul_Cyrl",
36
+ "cat_Latn",
37
+ "ceb_Latn",
38
+ "ces_Latn",
39
+ "cjk_Latn",
40
+ "ckb_Arab",
41
+ "crh_Latn",
42
+ "cym_Latn",
43
+ "dan_Latn",
44
+ "deu_Latn",
45
+ "dik_Latn",
46
+ "dyu_Latn",
47
+ "dzo_Tibt",
48
+ "ell_Grek",
49
+ "eng_Latn",
50
+ "epo_Latn",
51
+ "est_Latn",
52
+ "eus_Latn",
53
+ "ewe_Latn",
54
+ "fao_Latn",
55
+ "pes_Arab",
56
+ "fij_Latn",
57
+ "fin_Latn",
58
+ "fon_Latn",
59
+ "fra_Latn",
60
+ "fur_Latn",
61
+ "fuv_Latn",
62
+ "gla_Latn",
63
+ "gle_Latn",
64
+ "glg_Latn",
65
+ "grn_Latn",
66
+ "guj_Gujr",
67
+ "hat_Latn",
68
+ "hau_Latn",
69
+ "heb_Hebr",
70
+ "hin_Deva",
71
+ "hne_Deva",
72
+ "hrv_Latn",
73
+ "hun_Latn",
74
+ "hye_Armn",
75
+ "ibo_Latn",
76
+ "ilo_Latn",
77
+ "ind_Latn",
78
+ "isl_Latn",
79
+ "ita_Latn",
80
+ "jav_Latn",
81
+ "jpn_Jpan",
82
+ "kab_Latn",
83
+ "kac_Latn",
84
+ "kam_Latn",
85
+ "kan_Knda",
86
+ "kas_Arab",
87
+ "kas_Deva",
88
+ "kat_Geor",
89
+ "knc_Arab",
90
+ "knc_Latn",
91
+ "kaz_Cyrl",
92
+ "kbp_Latn",
93
+ "kea_Latn",
94
+ "khm_Khmr",
95
+ "kik_Latn",
96
+ "kin_Latn",
97
+ "kir_Cyrl",
98
+ "kmb_Latn",
99
+ "kon_Latn",
100
+ "kor_Hang",
101
+ "kmr_Latn",
102
+ "lao_Laoo",
103
+ "lvs_Latn",
104
+ "lij_Latn",
105
+ "lim_Latn",
106
+ "lin_Latn",
107
+ "lit_Latn",
108
+ "lmo_Latn",
109
+ "ltg_Latn",
110
+ "ltz_Latn",
111
+ "lua_Latn",
112
+ "lug_Latn",
113
+ "luo_Latn",
114
+ "lus_Latn",
115
+ "mag_Deva",
116
+ "mai_Deva",
117
+ "mal_Mlym",
118
+ "mar_Deva",
119
+ "min_Latn",
120
+ "mkd_Cyrl",
121
+ "plt_Latn",
122
+ "mlt_Latn",
123
+ "mni_Beng",
124
+ "khk_Cyrl",
125
+ "mos_Latn",
126
+ "mri_Latn",
127
+ "zsm_Latn",
128
+ "mya_Mymr",
129
+ "nld_Latn",
130
+ "nno_Latn",
131
+ "nob_Latn",
132
+ "npi_Deva",
133
+ "nso_Latn",
134
+ "nus_Latn",
135
+ "nya_Latn",
136
+ "oci_Latn",
137
+ "gaz_Latn",
138
+ "ory_Orya",
139
+ "pag_Latn",
140
+ "pan_Guru",
141
+ "pap_Latn",
142
+ "pol_Latn",
143
+ "por_Latn",
144
+ "prs_Arab",
145
+ "pbt_Arab",
146
+ "quy_Latn",
147
+ "ron_Latn",
148
+ "run_Latn",
149
+ "rus_Cyrl",
150
+ "sag_Latn",
151
+ "san_Deva",
152
+ "sat_Beng",
153
+ "scn_Latn",
154
+ "shn_Mymr",
155
+ "sin_Sinh",
156
+ "slk_Latn",
157
+ "slv_Latn",
158
+ "smo_Latn",
159
+ "sna_Latn",
160
+ "snd_Arab",
161
+ "som_Latn",
162
+ "sot_Latn",
163
+ "spa_Latn",
164
+ "als_Latn",
165
+ "srd_Latn",
166
+ "srp_Cyrl",
167
+ "ssw_Latn",
168
+ "sun_Latn",
169
+ "swe_Latn",
170
+ "swh_Latn",
171
+ "szl_Latn",
172
+ "tam_Taml",
173
+ "tat_Cyrl",
174
+ "tel_Telu",
175
+ "tgk_Cyrl",
176
+ "tgl_Latn",
177
+ "tha_Thai",
178
+ "tir_Ethi",
179
+ "taq_Latn",
180
+ "taq_Tfng",
181
+ "tpi_Latn",
182
+ "tsn_Latn",
183
+ "tso_Latn",
184
+ "tuk_Latn",
185
+ "tum_Latn",
186
+ "tur_Latn",
187
+ "twi_Latn",
188
+ "tzm_Tfng",
189
+ "uig_Arab",
190
+ "ukr_Cyrl",
191
+ "umb_Latn",
192
+ "urd_Arab",
193
+ "uzn_Latn",
194
+ "vec_Latn",
195
+ "vie_Latn",
196
+ "war_Latn",
197
+ "wol_Latn",
198
+ "xho_Latn",
199
+ "ydd_Hebr",
200
+ "yor_Latn",
201
+ "yue_Hant",
202
+ "zho_Hans",
203
+ "zho_Hant",
204
+ "zul_Latn"
205
+ ],
206
+ "bos_token": "<s>",
207
+ "cls_token": "<s>",
208
+ "eos_token": "</s>",
209
+ "mask_token": {
210
+ "content": "<mask>",
211
+ "lstrip": true,
212
+ "normalized": true,
213
+ "rstrip": false,
214
+ "single_word": false
215
+ },
216
+ "pad_token": "<pad>",
217
+ "sep_token": "</s>",
218
+ "unk_token": "<unk>"
219
+ }
translation/nllb-200-distilled-1.3B/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665
3
+ size 17331176
translation/nllb-200-distilled-600M/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_source_bos": false,
3
+ "add_source_eos": false,
4
+ "bos_token": "<s>",
5
+ "decoder_start_token": "</s>",
6
+ "eos_token": "</s>",
7
+ "layer_norm_epsilon": null,
8
+ "multi_query_attention": false,
9
+ "unk_token": "<unk>"
10
+ }
translation/nllb-200-distilled-600M/model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:795cc73c70446f8ba0cad530e8dda72930f725b73124bb6e8bf0ab838a8eb5f1
3
+ size 1232266703
translation/nllb-200-distilled-600M/shared_vocabulary.json ADDED
The diff for this file is too large to render. See raw diff
 
translation/nllb-200-distilled-600M/special_tokens_map.json ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "ace_Arab",
4
+ "ace_Latn",
5
+ "acm_Arab",
6
+ "acq_Arab",
7
+ "aeb_Arab",
8
+ "afr_Latn",
9
+ "ajp_Arab",
10
+ "aka_Latn",
11
+ "amh_Ethi",
12
+ "apc_Arab",
13
+ "arb_Arab",
14
+ "ars_Arab",
15
+ "ary_Arab",
16
+ "arz_Arab",
17
+ "asm_Beng",
18
+ "ast_Latn",
19
+ "awa_Deva",
20
+ "ayr_Latn",
21
+ "azb_Arab",
22
+ "azj_Latn",
23
+ "bak_Cyrl",
24
+ "bam_Latn",
25
+ "ban_Latn",
26
+ "bel_Cyrl",
27
+ "bem_Latn",
28
+ "ben_Beng",
29
+ "bho_Deva",
30
+ "bjn_Arab",
31
+ "bjn_Latn",
32
+ "bod_Tibt",
33
+ "bos_Latn",
34
+ "bug_Latn",
35
+ "bul_Cyrl",
36
+ "cat_Latn",
37
+ "ceb_Latn",
38
+ "ces_Latn",
39
+ "cjk_Latn",
40
+ "ckb_Arab",
41
+ "crh_Latn",
42
+ "cym_Latn",
43
+ "dan_Latn",
44
+ "deu_Latn",
45
+ "dik_Latn",
46
+ "dyu_Latn",
47
+ "dzo_Tibt",
48
+ "ell_Grek",
49
+ "eng_Latn",
50
+ "epo_Latn",
51
+ "est_Latn",
52
+ "eus_Latn",
53
+ "ewe_Latn",
54
+ "fao_Latn",
55
+ "pes_Arab",
56
+ "fij_Latn",
57
+ "fin_Latn",
58
+ "fon_Latn",
59
+ "fra_Latn",
60
+ "fur_Latn",
61
+ "fuv_Latn",
62
+ "gla_Latn",
63
+ "gle_Latn",
64
+ "glg_Latn",
65
+ "grn_Latn",
66
+ "guj_Gujr",
67
+ "hat_Latn",
68
+ "hau_Latn",
69
+ "heb_Hebr",
70
+ "hin_Deva",
71
+ "hne_Deva",
72
+ "hrv_Latn",
73
+ "hun_Latn",
74
+ "hye_Armn",
75
+ "ibo_Latn",
76
+ "ilo_Latn",
77
+ "ind_Latn",
78
+ "isl_Latn",
79
+ "ita_Latn",
80
+ "jav_Latn",
81
+ "jpn_Jpan",
82
+ "kab_Latn",
83
+ "kac_Latn",
84
+ "kam_Latn",
85
+ "kan_Knda",
86
+ "kas_Arab",
87
+ "kas_Deva",
88
+ "kat_Geor",
89
+ "knc_Arab",
90
+ "knc_Latn",
91
+ "kaz_Cyrl",
92
+ "kbp_Latn",
93
+ "kea_Latn",
94
+ "khm_Khmr",
95
+ "kik_Latn",
96
+ "kin_Latn",
97
+ "kir_Cyrl",
98
+ "kmb_Latn",
99
+ "kon_Latn",
100
+ "kor_Hang",
101
+ "kmr_Latn",
102
+ "lao_Laoo",
103
+ "lvs_Latn",
104
+ "lij_Latn",
105
+ "lim_Latn",
106
+ "lin_Latn",
107
+ "lit_Latn",
108
+ "lmo_Latn",
109
+ "ltg_Latn",
110
+ "ltz_Latn",
111
+ "lua_Latn",
112
+ "lug_Latn",
113
+ "luo_Latn",
114
+ "lus_Latn",
115
+ "mag_Deva",
116
+ "mai_Deva",
117
+ "mal_Mlym",
118
+ "mar_Deva",
119
+ "min_Latn",
120
+ "mkd_Cyrl",
121
+ "plt_Latn",
122
+ "mlt_Latn",
123
+ "mni_Beng",
124
+ "khk_Cyrl",
125
+ "mos_Latn",
126
+ "mri_Latn",
127
+ "zsm_Latn",
128
+ "mya_Mymr",
129
+ "nld_Latn",
130
+ "nno_Latn",
131
+ "nob_Latn",
132
+ "npi_Deva",
133
+ "nso_Latn",
134
+ "nus_Latn",
135
+ "nya_Latn",
136
+ "oci_Latn",
137
+ "gaz_Latn",
138
+ "ory_Orya",
139
+ "pag_Latn",
140
+ "pan_Guru",
141
+ "pap_Latn",
142
+ "pol_Latn",
143
+ "por_Latn",
144
+ "prs_Arab",
145
+ "pbt_Arab",
146
+ "quy_Latn",
147
+ "ron_Latn",
148
+ "run_Latn",
149
+ "rus_Cyrl",
150
+ "sag_Latn",
151
+ "san_Deva",
152
+ "sat_Beng",
153
+ "scn_Latn",
154
+ "shn_Mymr",
155
+ "sin_Sinh",
156
+ "slk_Latn",
157
+ "slv_Latn",
158
+ "smo_Latn",
159
+ "sna_Latn",
160
+ "snd_Arab",
161
+ "som_Latn",
162
+ "sot_Latn",
163
+ "spa_Latn",
164
+ "als_Latn",
165
+ "srd_Latn",
166
+ "srp_Cyrl",
167
+ "ssw_Latn",
168
+ "sun_Latn",
169
+ "swe_Latn",
170
+ "swh_Latn",
171
+ "szl_Latn",
172
+ "tam_Taml",
173
+ "tat_Cyrl",
174
+ "tel_Telu",
175
+ "tgk_Cyrl",
176
+ "tgl_Latn",
177
+ "tha_Thai",
178
+ "tir_Ethi",
179
+ "taq_Latn",
180
+ "taq_Tfng",
181
+ "tpi_Latn",
182
+ "tsn_Latn",
183
+ "tso_Latn",
184
+ "tuk_Latn",
185
+ "tum_Latn",
186
+ "tur_Latn",
187
+ "twi_Latn",
188
+ "tzm_Tfng",
189
+ "uig_Arab",
190
+ "ukr_Cyrl",
191
+ "umb_Latn",
192
+ "urd_Arab",
193
+ "uzn_Latn",
194
+ "vec_Latn",
195
+ "vie_Latn",
196
+ "war_Latn",
197
+ "wol_Latn",
198
+ "xho_Latn",
199
+ "ydd_Hebr",
200
+ "yor_Latn",
201
+ "yue_Hant",
202
+ "zho_Hans",
203
+ "zho_Hant",
204
+ "zul_Latn"
205
+ ],
206
+ "bos_token": "<s>",
207
+ "cls_token": "<s>",
208
+ "eos_token": "</s>",
209
+ "mask_token": {
210
+ "content": "<mask>",
211
+ "lstrip": true,
212
+ "normalized": true,
213
+ "rstrip": false,
214
+ "single_word": false
215
+ },
216
+ "pad_token": "<pad>",
217
+ "sep_token": "</s>",
218
+ "unk_token": "<unk>"
219
+ }
translation/nllb-200-distilled-600M/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e316b82de11d0f951f370943b3c438311629547285129b0b81dadabd01bca665
3
+ size 17331176
whisper/ggml-base.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60ed5bc3dd14eea856493d334349b405782ddcaf0028d4b5df4088345fba2efe
3
+ size 147951465
whisper/ggml-base.en.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a03779c86df3323075f5e796cb2ce5029f00ec8869eee3fdfb897afe36c6d002
3
+ size 147964211
whisper/ggml-large-v1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d99f41a10525d0206bddadd86760181fa920438b6b33237e3118ff6c83bb53d
3
+ size 3094623691
whisper/ggml-large-v2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a423fe4d40c82774b6af34115b8b935f34152246eb19e80e376071d3f999487
3
+ size 3094623691
whisper/ggml-large-v3-turbo.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fc70f774d38eb169993ac391eea357ef47c88757ef72ee5943879b7e8e2bc69
3
+ size 1624555275
whisper/ggml-large-v3.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64d182b440b98d5203c4f9bd541544d84c605196c4f7b845dfa11fb23594d1e2
3
+ size 3095033483
whisper/ggml-medium.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c14d5adee5f86394037b4e4e8b59f1673b6cee10e3cf0b11bbdbee79c156208
3
+ size 1533763059
whisper/ggml-medium.en.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc37e93478338ec7700281a7ac30a10128929eb8f427dda2e865faa8f6da4356
3
+ size 1533774781
whisper/ggml-small.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1be3a9b2063867b937e64e2ec7483364a79917e157fa98c5d94b5c1fffea987b
3
+ size 487601967
whisper/ggml-small.en.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6138d6d58ecc8322097e0f987c32f1be8bb0a18532a3f88f734d1bbf9c41e5d
3
+ size 487614201
whisper/ggml-tiny.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be07e048e1e599ad46341c8d2a135645097a538221678b7acdd1b1919c6e1b21
3
+ size 77691713
whisper/ggml-tiny.en.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:921e4cf8686fdd993dcd081a5da5b6c365bfde1162e72b08d75ac75289920b1f
3
+ size 77704715