Hugging Face's logo

OpenVoiceOS
/
stt-ca-es-conformer-transducer-large-coreml-int8

Automatic Speech Recognition
NeMo
Core ML
apple-silicon
ovos
Model card Files
xet
 Community
Jarbas commited on
Commit
a27bc3f
·
verified ·
1 Parent(s): 6ca6877

Add CoreML export: stt-ca-es-conformer-transducer-large-coreml-int8

Browse files
Files changed (13) hide show
  1. README.md +22 -0
  2. metadata.json +113 -0
  3. parakeet_decoder.mlpackage/Data/com.apple.CoreML/model.mlmodel +3 -0
  4. parakeet_decoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin +3 -0
  5. parakeet_decoder.mlpackage/Manifest.json +18 -0
  6. parakeet_joint_decision_single_step.mlpackage/Data/com.apple.CoreML/model.mlmodel +3 -0
  7. parakeet_joint_decision_single_step.mlpackage/Data/com.apple.CoreML/weights/weight.bin +3 -0
  8. parakeet_joint_decision_single_step.mlpackage/Manifest.json +18 -0
  9. parakeet_mel_encoder.mlpackage/Data/com.apple.CoreML/model.mlmodel +3 -0
  10. parakeet_mel_encoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin +3 -0
  11. parakeet_mel_encoder.mlpackage/Manifest.json +18 -0
  12. quantization_summary.json +17 -0
  13. vocab.json +258 -0
README.md ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ tags:
4
+ - automatic-speech-recognition
5
+ - nemo
6
+ - coreml
7
+ - apple-silicon
8
+ - ovos
9
+ ---
10
+
11
+ # stt-ca-es-conformer-transducer-large-coreml
12
+
13
+ CoreML export of [projecte-aina/stt_ca-es_conformer_transducer_large](https://huggingface.co/projecte-aina/stt_ca-es_conformer_transducer_large) for Apple Silicon.
14
+
15
+ Exported with [ovos-stt-plugin-coreml](https://github.com/TigreGotico/ovos-stt-plugin-coreml).
16
+
17
+ ## Usage
18
+
19
+ ```python
20
+ from ovos_stt_plugin_coreml import CoremlSTT
21
+ stt = CoremlSTT(config={"repo_id": "OpenVoiceOS/stt-ca-es-conformer-transducer-large-coreml"})
22
+ ```
metadata.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_id": "projecte-aina/stt_ca-es_conformer_transducer_large",
3
+ "model_type": "parakeet_rnnt",
4
+ "language": "",
5
+ "sample_rate": 16000,
6
+ "max_audio_seconds": 15.0,
7
+ "max_audio_samples": 240000,
8
+ "vocab_size": 256,
9
+ "blank_id": 256,
10
+ "checkpoint": {
11
+ "type": "pretrained",
12
+ "model_id": "projecte-aina/stt_ca-es_conformer_transducer_large"
13
+ },
14
+ "coreml": {
15
+ "compute_precision": "FLOAT32",
16
+ "quantization": "int8_per_channel_symmetric"
17
+ },
18
+ "components": {
19
+ "mel_encoder": {
20
+ "path": "parakeet_mel_encoder.mlpackage",
21
+ "inputs": {
22
+ "audio_signal": [
23
+ 1,
24
+ 240000
25
+ ],
26
+ "audio_length": [
27
+ 1
28
+ ]
29
+ },
30
+ "outputs": {
31
+ "encoder": [
32
+ 1,
33
+ 512,
34
+ 376
35
+ ],
36
+ "encoder_length": [
37
+ 1
38
+ ]
39
+ }
40
+ },
41
+ "decoder": {
42
+ "path": "parakeet_decoder.mlpackage",
43
+ "inputs": {
44
+ "targets": [
45
+ 1,
46
+ 1
47
+ ],
48
+ "target_length": [
49
+ 1
50
+ ],
51
+ "h_in": [
52
+ 1,
53
+ 1,
54
+ 640
55
+ ],
56
+ "c_in": [
57
+ 1,
58
+ 1,
59
+ 640
60
+ ]
61
+ },
62
+ "outputs": {
63
+ "decoder": [
64
+ 1,
65
+ 640,
66
+ 1
67
+ ],
68
+ "h_out": [
69
+ 1,
70
+ 1,
71
+ 640
72
+ ],
73
+ "c_out": [
74
+ 1,
75
+ 1,
76
+ 640
77
+ ]
78
+ }
79
+ },
80
+ "joint_decision_single_step": {
81
+ "path": "parakeet_joint_decision_single_step.mlpackage",
82
+ "inputs": {
83
+ "encoder_step": [
84
+ 1,
85
+ 512,
86
+ 1
87
+ ],
88
+ "decoder_step": [
89
+ 1,
90
+ 640,
91
+ 1
92
+ ]
93
+ },
94
+ "outputs": {
95
+ "token_id": [
96
+ 1,
97
+ 1,
98
+ 1
99
+ ],
100
+ "token_prob": [
101
+ 1,
102
+ 1,
103
+ 1
104
+ ],
105
+ "duration": [
106
+ 1,
107
+ 1,
108
+ 1
109
+ ]
110
+ }
111
+ }
112
+ }
113
+ }
parakeet_decoder.mlpackage/Data/com.apple.CoreML/model.mlmodel ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:740f6732917874e9466c8e655fa2883b65d85aaef8eb526b6c1f0f4e3fb42972
3
+ size 6439
parakeet_decoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87706721e27927ea18ee0b6553111f5ef25fea00fbc9e1842e7cb0ccfe901ba5
3
+ size 3476608
parakeet_decoder.mlpackage/Manifest.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "fileFormatVersion": "1.0.0",
3
+ "itemInfoEntries": {
4
+ "634E9466-22A3-4505-8804-019B9E263EF6": {
5
+ "author": "com.apple.CoreML",
6
+ "description": "CoreML Model Specification",
7
+ "name": "model.mlmodel",
8
+ "path": "com.apple.CoreML/model.mlmodel"
9
+ },
10
+ "BB8D5936-DC11-47C7-BC9D-CF75B26219EB": {
11
+ "author": "com.apple.CoreML",
12
+ "description": "CoreML Model Weights",
13
+ "name": "weights",
14
+ "path": "com.apple.CoreML/weights"
15
+ }
16
+ },
17
+ "rootModelIdentifier": "634E9466-22A3-4505-8804-019B9E263EF6"
18
+ }
parakeet_joint_decision_single_step.mlpackage/Data/com.apple.CoreML/model.mlmodel ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd49e0c984db4986ad60770ccb015aaeb2358065333c67a0c45eaa33a14c4105
3
+ size 5788
parakeet_joint_decision_single_step.mlpackage/Data/com.apple.CoreML/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a38eb85b7041c2f1ce3d130564d2245665735c34b2ff30de10a8785f74e0882c
3
+ size 915844
parakeet_joint_decision_single_step.mlpackage/Manifest.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "fileFormatVersion": "1.0.0",
3
+ "itemInfoEntries": {
4
+ "04D80882-3E90-4927-B1D6-505C9B1A3C92": {
5
+ "author": "com.apple.CoreML",
6
+ "description": "CoreML Model Weights",
7
+ "name": "weights",
8
+ "path": "com.apple.CoreML/weights"
9
+ },
10
+ "64362F47-583D-471C-84EF-F71CB809D657": {
11
+ "author": "com.apple.CoreML",
12
+ "description": "CoreML Model Specification",
13
+ "name": "model.mlmodel",
14
+ "path": "com.apple.CoreML/model.mlmodel"
15
+ }
16
+ },
17
+ "rootModelIdentifier": "64362F47-583D-471C-84EF-F71CB809D657"
18
+ }
parakeet_mel_encoder.mlpackage/Data/com.apple.CoreML/model.mlmodel ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37d4dedf5ce77233dd789767085832ee719d8674ea6259818a39648c559dc010
3
+ size 1336306
parakeet_mel_encoder.mlpackage/Data/com.apple.CoreML/weights/weight.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96286b01f166d55551752b279691441df172f25bb236a0cf83b1384dec746fb7
3
+ size 119110016
parakeet_mel_encoder.mlpackage/Manifest.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "fileFormatVersion": "1.0.0",
3
+ "itemInfoEntries": {
4
+ "14F3C370-AEB3-405E-93E9-01BB301C367C": {
5
+ "author": "com.apple.CoreML",
6
+ "description": "CoreML Model Weights",
7
+ "name": "weights",
8
+ "path": "com.apple.CoreML/weights"
9
+ },
10
+ "F7494086-7EFD-4A23-9517-EFA25F4DC32A": {
11
+ "author": "com.apple.CoreML",
12
+ "description": "CoreML Model Specification",
13
+ "name": "model.mlmodel",
14
+ "path": "com.apple.CoreML/model.mlmodel"
15
+ }
16
+ },
17
+ "rootModelIdentifier": "F7494086-7EFD-4A23-9517-EFA25F4DC32A"
18
+ }
quantization_summary.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "parakeet_decoder": {
3
+ "original_mb": 13.1,
4
+ "int8_mb": 3.3,
5
+ "compression": 3.96
6
+ },
7
+ "parakeet_joint_decision_single_step": {
8
+ "original_mb": 3.5,
9
+ "int8_mb": 0.9,
10
+ "compression": 3.92
11
+ },
12
+ "parakeet_mel_encoder": {
13
+ "original_mb": 449.8,
14
+ "int8_mb": 114.9,
15
+ "compression": 3.92
16
+ }
17
+ }
vocab.json ADDED
@@ -0,0 +1,258 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ "<unk>",
3
+ "▁d",
4
+ "▁e",
5
+ "▁a",
6
+ "▁p",
7
+ "es",
8
+ "▁l",
9
+ "er",
10
+ "▁de",
11
+ "▁c",
12
+ "en",
13
+ "▁s",
14
+ "ta",
15
+ "qu",
16
+ "ar",
17
+ "▁m",
18
+ "an",
19
+ "os",
20
+ "on",
21
+ "or",
22
+ "▁la",
23
+ "▁qu",
24
+ "ci",
25
+ "re",
26
+ "un",
27
+ "▁i",
28
+ "▁el",
29
+ "▁t",
30
+ "▁n",
31
+ "▁v",
32
+ "▁en",
33
+ "▁que",
34
+ "al",
35
+ "▁es",
36
+ "▁h",
37
+ "▁f",
38
+ "in",
39
+ "ra",
40
+ "ent",
41
+ "▁un",
42
+ "as",
43
+ "om",
44
+ "ti",
45
+ "ol",
46
+ "▁per",
47
+ "ad",
48
+ "is",
49
+ "▁se",
50
+ "ri",
51
+ "és",
52
+ "▁al",
53
+ "at",
54
+ "▁con",
55
+ "▁b",
56
+ "▁no",
57
+ "res",
58
+ "em",
59
+ "▁g",
60
+ "ro",
61
+ "▁y",
62
+ "ic",
63
+ "el",
64
+ "▁del",
65
+ "▁com",
66
+ "ció",
67
+ "▁re",
68
+ "mb",
69
+ "▁di",
70
+ "▁ha",
71
+ "am",
72
+ "id",
73
+ "ia",
74
+ "te",
75
+ "▁una",
76
+ "us",
77
+ "▁o",
78
+ "▁in",
79
+ "gu",
80
+ "▁és",
81
+ "il",
82
+ "▁par",
83
+ "ix",
84
+ "▁j",
85
+ "tat",
86
+ "le",
87
+ "la",
88
+ "it",
89
+ "▁ta",
90
+ "▁an",
91
+ "ec",
92
+ "▁su",
93
+ "ir",
94
+ "do",
95
+ "▁les",
96
+ "ques",
97
+ "▁va",
98
+ "ot",
99
+ "▁els",
100
+ "ab",
101
+ "▁des",
102
+ "▁pro",
103
+ "ran",
104
+ "tu",
105
+ "▁amb",
106
+ "ul",
107
+ "oc",
108
+ "ment",
109
+ "di",
110
+ "▁los",
111
+ "cion",
112
+ "▁por",
113
+ "va",
114
+ "ant",
115
+ "▁ca",
116
+ "da",
117
+ "▁si",
118
+ "era",
119
+ "ada",
120
+ "ob",
121
+ "ca",
122
+ "▁ser",
123
+ "mp",
124
+ "▁aques",
125
+ "▁ac",
126
+ "▁est",
127
+ "cia",
128
+ "et",
129
+ "ón",
130
+ "tan",
131
+ "od",
132
+ "ter",
133
+ "tes",
134
+ "tar",
135
+ "ur",
136
+ "si",
137
+ "tal",
138
+ "ía",
139
+ "ig",
140
+ "tra",
141
+ "▁na",
142
+ "ver",
143
+ "ue",
144
+ "ts",
145
+ "ma",
146
+ "ament",
147
+ "▁hi",
148
+ "de",
149
+ "ica",
150
+ "▁ex",
151
+ "ción",
152
+ "▁las",
153
+ "▁més",
154
+ "▁me",
155
+ "bl",
156
+ "▁ll",
157
+ "se",
158
+ "to",
159
+ "ru",
160
+ "▁mol",
161
+ "▁ten",
162
+ "ell",
163
+ "ost",
164
+ "ien",
165
+ "▁res",
166
+ "itat",
167
+ "▁esta",
168
+ "▁cu",
169
+ "im",
170
+ "ac",
171
+ "▁car",
172
+ "▁cons",
173
+ "▁ar",
174
+ "▁esp",
175
+ "ado",
176
+ "tor",
177
+ "▁tamb",
178
+ "▁tot",
179
+ "▁comp",
180
+ "eg",
181
+ "▁mar",
182
+ "▁vol",
183
+ "▁",
184
+ "e",
185
+ "a",
186
+ "s",
187
+ "n",
188
+ "r",
189
+ "i",
190
+ "o",
191
+ "l",
192
+ "t",
193
+ "d",
194
+ "u",
195
+ "c",
196
+ "m",
197
+ "p",
198
+ "b",
199
+ "v",
200
+ "g",
201
+ "q",
202
+ "f",
203
+ "h",
204
+ "ó",
205
+ "y",
206
+ "é",
207
+ "í",
208
+ "j",
209
+ "'",
210
+ "x",
211
+ "à",
212
+ "z",
213
+ "è",
214
+ "á",
215
+ "ò",
216
+ "’",
217
+ "ú",
218
+ "-",
219
+ "ñ",
220
+ "ç",
221
+ "k",
222
+ "·",
223
+ "w",
224
+ "–",
225
+ "ï",
226
+ "ü",
227
+ "­",
228
+ "3",
229
+ "2",
230
+ "1",
231
+ "+",
232
+ "0",
233
+ "”",
234
+ "“",
235
+ "5",
236
+ "4",
237
+ "9",
238
+ "8",
239
+ ")",
240
+ "7",
241
+ "6",
242
+ "â",
243
+ "ö",
244
+ "ä",
245
+ "‘",
246
+ "ã",
247
+ "`",
248
+ "ğ",
249
+ "(",
250
+ "¦",
251
+ "î",
252
+ "€",
253
+ "_",
254
+ "ë",
255
+ "č",
256
+ "ž",
257
+ ">"
258
+ ]