Instructions to use OpenVoiceOS/nvidia-ca-conformer-ctc-large-onnx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use OpenVoiceOS/nvidia-ca-conformer-ctc-large-onnx with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("OpenVoiceOS/nvidia-ca-conformer-ctc-large-onnx") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Add ONNX export of nvidia/stt_ca_conformer_ctc_large for onnx-asr
Browse files- .gitattributes +1 -0
- README.md +33 -0
- config.json +6 -0
- model.int8.onnx +3 -0
- model.onnx +3 -0
- model.onnx_data +3 -0
- vocab.txt +129 -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 |
+
model.onnx_data filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
library_name: onnx-asr
|
| 4 |
+
tags:
|
| 5 |
+
- automatic-speech-recognition
|
| 6 |
+
- onnx
|
| 7 |
+
- nemo
|
| 8 |
+
pipeline_tag: automatic-speech-recognition
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# nvidia-ca-conformer-ctc-large-onnx
|
| 12 |
+
|
| 13 |
+
NVIDIA NeMo ASR model exported to ONNX for the
|
| 14 |
+
[onnx-asr](https://github.com/istupakov/onnx-asr) library and the
|
| 15 |
+
[ovos-stt-plugin-onnx-asr](https://github.com/OpenVoiceOS/ovos-stt-plugin-onnx-asr)
|
| 16 |
+
OpenVoiceOS STT plugin.
|
| 17 |
+
|
| 18 |
+
Converted from [nvidia/stt_ca_conformer_ctc_large](https://huggingface.co/nvidia/stt_ca_conformer_ctc_large) (cc-by-4.0).
|
| 19 |
+
Decoder: CTC (`model.onnx`) — `model_type: nemo-conformer-ctc`, features_size 80,
|
| 20 |
+
subsampling_factor 4.
|
| 21 |
+
|
| 22 |
+
```python
|
| 23 |
+
import onnx_asr
|
| 24 |
+
model = onnx_asr.load_model("OpenVoiceOS/nvidia-ca-conformer-ctc-large-onnx")
|
| 25 |
+
print(model.recognize("audio.wav"))
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
Or via OpenVoiceOS config:
|
| 29 |
+
|
| 30 |
+
```json
|
| 31 |
+
{"stt": {"module": "ovos-stt-plugin-onnx-asr",
|
| 32 |
+
"ovos-stt-plugin-onnx-asr": {"model": "OpenVoiceOS/nvidia-ca-conformer-ctc-large-onnx"}}}
|
| 33 |
+
```
|
config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "nemo-conformer-ctc",
|
| 3 |
+
"features_size": 80,
|
| 4 |
+
"subsampling_factor": 4,
|
| 5 |
+
"max_tokens_per_step": 10
|
| 6 |
+
}
|
model.int8.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:971761fec724bec2721c5ead77bc57f6dac79e3246cfd3589b9930b092e2b210
|
| 3 |
+
size 143924803
|
model.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ba999cc55b27c6f043008eb676228937a38d5f667dec0613a9178ff317805b6
|
| 3 |
+
size 646561
|
model.onnx_data
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5fcafac1ce31ce2169a84d2052fab24e80c82bbc767eac323b6b0af501a00bd
|
| 3 |
+
size 506408960
|
vocab.txt
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
<unk> 0
|
| 2 |
+
s 1
|
| 3 |
+
▁ 2
|
| 4 |
+
a 3
|
| 5 |
+
i 4
|
| 6 |
+
t 5
|
| 7 |
+
e 6
|
| 8 |
+
u 7
|
| 9 |
+
c 8
|
| 10 |
+
n 9
|
| 11 |
+
m 10
|
| 12 |
+
l 11
|
| 13 |
+
r 12
|
| 14 |
+
▁de 13
|
| 15 |
+
o 14
|
| 16 |
+
es 15
|
| 17 |
+
d 16
|
| 18 |
+
▁a 17
|
| 19 |
+
p 18
|
| 20 |
+
ar 19
|
| 21 |
+
' 20
|
| 22 |
+
v 21
|
| 23 |
+
g 22
|
| 24 |
+
b 23
|
| 25 |
+
▁d 24
|
| 26 |
+
er 25
|
| 27 |
+
▁la 26
|
| 28 |
+
en 27
|
| 29 |
+
▁p 28
|
| 30 |
+
or 29
|
| 31 |
+
re 30
|
| 32 |
+
▁f 31
|
| 33 |
+
▁el 32
|
| 34 |
+
al 33
|
| 35 |
+
ra 34
|
| 36 |
+
in 35
|
| 37 |
+
▁l 36
|
| 38 |
+
▁i 37
|
| 39 |
+
▁es 38
|
| 40 |
+
an 39
|
| 41 |
+
on 40
|
| 42 |
+
ri 41
|
| 43 |
+
▁va 42
|
| 44 |
+
ta 43
|
| 45 |
+
at 44
|
| 46 |
+
▁s 45
|
| 47 |
+
à 46
|
| 48 |
+
ci 47
|
| 49 |
+
▁t 48
|
| 50 |
+
f 49
|
| 51 |
+
▁en 50
|
| 52 |
+
ll 51
|
| 53 |
+
▁se 52
|
| 54 |
+
▁c 53
|
| 55 |
+
▁per 54
|
| 56 |
+
ant 55
|
| 57 |
+
x 56
|
| 58 |
+
ti 57
|
| 59 |
+
h 58
|
| 60 |
+
▁un 59
|
| 61 |
+
▁al 60
|
| 62 |
+
ment 61
|
| 63 |
+
▁b 62
|
| 64 |
+
it 63
|
| 65 |
+
q 64
|
| 66 |
+
▁del 65
|
| 67 |
+
▁no 66
|
| 68 |
+
▁re 67
|
| 69 |
+
ca 68
|
| 70 |
+
j 69
|
| 71 |
+
í 70
|
| 72 |
+
os 71
|
| 73 |
+
di 72
|
| 74 |
+
▁és 73
|
| 75 |
+
▁com 74
|
| 76 |
+
▁ca 75
|
| 77 |
+
ol 76
|
| 78 |
+
ó 77
|
| 79 |
+
▁con 78
|
| 80 |
+
li 79
|
| 81 |
+
ent 80
|
| 82 |
+
è 81
|
| 83 |
+
▁ma 82
|
| 84 |
+
vi 83
|
| 85 |
+
▁g 84
|
| 86 |
+
▁les 85
|
| 87 |
+
gu 86
|
| 88 |
+
▁que 87
|
| 89 |
+
la 88
|
| 90 |
+
ò 89
|
| 91 |
+
▁amb 90
|
| 92 |
+
tre 91
|
| 93 |
+
- 92
|
| 94 |
+
▁po 93
|
| 95 |
+
▁ha 94
|
| 96 |
+
▁els 95
|
| 97 |
+
▁una 96
|
| 98 |
+
ada 97
|
| 99 |
+
ro 98
|
| 100 |
+
é 99
|
| 101 |
+
z 100
|
| 102 |
+
eix 101
|
| 103 |
+
ació 102
|
| 104 |
+
▁pro 103
|
| 105 |
+
que 104
|
| 106 |
+
ter 105
|
| 107 |
+
▁co 106
|
| 108 |
+
ç 107
|
| 109 |
+
▁ser 108
|
| 110 |
+
y 109
|
| 111 |
+
▁aquest 110
|
| 112 |
+
▁mo 111
|
| 113 |
+
▁mi 112
|
| 114 |
+
▁hi 113
|
| 115 |
+
▁més 114
|
| 116 |
+
any 115
|
| 117 |
+
▁també 116
|
| 118 |
+
ú 117
|
| 119 |
+
k 118
|
| 120 |
+
· 119
|
| 121 |
+
w 120
|
| 122 |
+
ï 121
|
| 123 |
+
ü 122
|
| 124 |
+
á 123
|
| 125 |
+
ñ 124
|
| 126 |
+
— 125
|
| 127 |
+
– 126
|
| 128 |
+
ı 127
|
| 129 |
+
<blk> 128
|