Automatic Speech Recognition
ONNX
Transformers.js
onnxruntime
qwen3_asr
text-generation
onnxruntime-web
asr
speech-recognition
robust-asr
quantized
int4
int8
matmulnbits
gptq
on-device
browser
web
qwen3
qwen3-asr
mega-asr
Instructions to use Reza2kn/mega-asr-onnx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers.js
How to use Reza2kn/mega-asr-onnx with Transformers.js:
// npm i @huggingface/transformers import { pipeline } from '@huggingface/transformers'; // Allocate pipeline const pipe = await pipeline('automatic-speech-recognition', 'Reza2kn/mega-asr-onnx');
Add pre-computed mel features for the 8 VITW examples + prompt manifest
Browse files- examples_mels/distortion.mel.bin +3 -0
- examples_mels/dropout.mel.bin +3 -0
- examples_mels/echo.mel.bin +3 -0
- examples_mels/far_field.mel.bin +3 -0
- examples_mels/manifest.json +49 -0
- examples_mels/mixed.mel.bin +3 -0
- examples_mels/noise.mel.bin +3 -0
- examples_mels/obstructed.mel.bin +3 -0
- examples_mels/recording.mel.bin +3 -0
examples_mels/distortion.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1fd0124655e77b2f5b3b00a805abbb998a739d7d75ff7e707983d0ff32e4cc06
|
| 3 |
+
size 768000
|
examples_mels/dropout.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cef76aad0d3f3a3dad30362ded12b29c1f3b5c1b783955a14e8c44c5255dc2ae
|
| 3 |
+
size 768000
|
examples_mels/echo.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a4a8da6b18f950322c17cae4a0dac59180f7189f636f520c299c31861c7f2f3
|
| 3 |
+
size 768000
|
examples_mels/far_field.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6662d3ab65ce7f45171f92adbf10e9af76b1e04715d655041289e3e6e61eb84
|
| 3 |
+
size 768000
|
examples_mels/manifest.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"examples": {
|
| 3 |
+
"noise": {
|
| 4 |
+
"T_mel": 868
|
| 5 |
+
},
|
| 6 |
+
"far_field": {
|
| 7 |
+
"T_mel": 592
|
| 8 |
+
},
|
| 9 |
+
"obstructed": {
|
| 10 |
+
"T_mel": 880
|
| 11 |
+
},
|
| 12 |
+
"distortion": {
|
| 13 |
+
"T_mel": 736
|
| 14 |
+
},
|
| 15 |
+
"recording": {
|
| 16 |
+
"T_mel": 500
|
| 17 |
+
},
|
| 18 |
+
"echo": {
|
| 19 |
+
"T_mel": 748
|
| 20 |
+
},
|
| 21 |
+
"dropout": {
|
| 22 |
+
"T_mel": 668
|
| 23 |
+
},
|
| 24 |
+
"mixed": {
|
| 25 |
+
"T_mel": 508
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"prompt_ids": [
|
| 29 |
+
151644,
|
| 30 |
+
8948,
|
| 31 |
+
198,
|
| 32 |
+
151645,
|
| 33 |
+
198,
|
| 34 |
+
151644,
|
| 35 |
+
872,
|
| 36 |
+
198,
|
| 37 |
+
151669,
|
| 38 |
+
151676,
|
| 39 |
+
151670,
|
| 40 |
+
151645,
|
| 41 |
+
198,
|
| 42 |
+
151644,
|
| 43 |
+
77091,
|
| 44 |
+
198
|
| 45 |
+
],
|
| 46 |
+
"prompt_text": "<|im_start|>system\n<|im_end|>\n<|im_start|>user\n<|audio_start|><|audio_pad|><|audio_end|><|im_end|>\n<|im_start|>assistant\n",
|
| 47 |
+
"audio_pad_id": 151676,
|
| 48 |
+
"eos_token_id": 151645
|
| 49 |
+
}
|
examples_mels/mixed.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3faac13c9783584420fa2bf6a239bed39be999868b6ad42330c4259155b50406
|
| 3 |
+
size 768000
|
examples_mels/noise.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d93cc26eda4fce424b586f103a79e02a2846322842496fb143e0fdccd2d6ef61
|
| 3 |
+
size 768000
|
examples_mels/obstructed.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9adbb45399f3694af11740a2de26b113765eac0551a43a36354c591016911be0
|
| 3 |
+
size 768000
|
examples_mels/recording.mel.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92f4f3df7fe4b9c9c0cb121fd05dee34beeeb5c37ba40f1580b7fd285d61304f
|
| 3 |
+
size 768000
|