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');
Update manifest: add forced-language prompts (default: English)
Browse files- examples_mels/manifest.json +125 -3
examples_mels/manifest.json
CHANGED
|
@@ -41,9 +41,131 @@
|
|
| 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\
|
| 47 |
"audio_pad_id": 151676,
|
| 48 |
-
"eos_token_id": 151645
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
}
|
|
|
|
| 41 |
198,
|
| 42 |
151644,
|
| 43 |
77091,
|
| 44 |
+
198,
|
| 45 |
+
11528,
|
| 46 |
+
6364,
|
| 47 |
+
151704
|
| 48 |
],
|
| 49 |
+
"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\nlanguage English<asr_text>",
|
| 50 |
"audio_pad_id": 151676,
|
| 51 |
+
"eos_token_id": 151645,
|
| 52 |
+
"prompts": {
|
| 53 |
+
"auto": {
|
| 54 |
+
"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",
|
| 55 |
+
"ids": [
|
| 56 |
+
151644,
|
| 57 |
+
8948,
|
| 58 |
+
198,
|
| 59 |
+
151645,
|
| 60 |
+
198,
|
| 61 |
+
151644,
|
| 62 |
+
872,
|
| 63 |
+
198,
|
| 64 |
+
151669,
|
| 65 |
+
151676,
|
| 66 |
+
151670,
|
| 67 |
+
151645,
|
| 68 |
+
198,
|
| 69 |
+
151644,
|
| 70 |
+
77091,
|
| 71 |
+
198
|
| 72 |
+
]
|
| 73 |
+
},
|
| 74 |
+
"english": {
|
| 75 |
+
"text": "<|im_start|>system\n<|im_end|>\n<|im_start|>user\n<|audio_start|><|audio_pad|><|audio_end|><|im_end|>\n<|im_start|>assistant\nlanguage English<asr_text>",
|
| 76 |
+
"ids": [
|
| 77 |
+
151644,
|
| 78 |
+
8948,
|
| 79 |
+
198,
|
| 80 |
+
151645,
|
| 81 |
+
198,
|
| 82 |
+
151644,
|
| 83 |
+
872,
|
| 84 |
+
198,
|
| 85 |
+
151669,
|
| 86 |
+
151676,
|
| 87 |
+
151670,
|
| 88 |
+
151645,
|
| 89 |
+
198,
|
| 90 |
+
151644,
|
| 91 |
+
77091,
|
| 92 |
+
198,
|
| 93 |
+
11528,
|
| 94 |
+
6364,
|
| 95 |
+
151704
|
| 96 |
+
]
|
| 97 |
+
},
|
| 98 |
+
"chinese": {
|
| 99 |
+
"text": "<|im_start|>system\n<|im_end|>\n<|im_start|>user\n<|audio_start|><|audio_pad|><|audio_end|><|im_end|>\n<|im_start|>assistant\nlanguage Chinese<asr_text>",
|
| 100 |
+
"ids": [
|
| 101 |
+
151644,
|
| 102 |
+
8948,
|
| 103 |
+
198,
|
| 104 |
+
151645,
|
| 105 |
+
198,
|
| 106 |
+
151644,
|
| 107 |
+
872,
|
| 108 |
+
198,
|
| 109 |
+
151669,
|
| 110 |
+
151676,
|
| 111 |
+
151670,
|
| 112 |
+
151645,
|
| 113 |
+
198,
|
| 114 |
+
151644,
|
| 115 |
+
77091,
|
| 116 |
+
198,
|
| 117 |
+
11528,
|
| 118 |
+
8453,
|
| 119 |
+
151704
|
| 120 |
+
]
|
| 121 |
+
},
|
| 122 |
+
"japanese": {
|
| 123 |
+
"text": "<|im_start|>system\n<|im_end|>\n<|im_start|>user\n<|audio_start|><|audio_pad|><|audio_end|><|im_end|>\n<|im_start|>assistant\nlanguage Japanese<asr_text>",
|
| 124 |
+
"ids": [
|
| 125 |
+
151644,
|
| 126 |
+
8948,
|
| 127 |
+
198,
|
| 128 |
+
151645,
|
| 129 |
+
198,
|
| 130 |
+
151644,
|
| 131 |
+
872,
|
| 132 |
+
198,
|
| 133 |
+
151669,
|
| 134 |
+
151676,
|
| 135 |
+
151670,
|
| 136 |
+
151645,
|
| 137 |
+
198,
|
| 138 |
+
151644,
|
| 139 |
+
77091,
|
| 140 |
+
198,
|
| 141 |
+
11528,
|
| 142 |
+
10769,
|
| 143 |
+
151704
|
| 144 |
+
]
|
| 145 |
+
},
|
| 146 |
+
"korean": {
|
| 147 |
+
"text": "<|im_start|>system\n<|im_end|>\n<|im_start|>user\n<|audio_start|><|audio_pad|><|audio_end|><|im_end|>\n<|im_start|>assistant\nlanguage Korean<asr_text>",
|
| 148 |
+
"ids": [
|
| 149 |
+
151644,
|
| 150 |
+
8948,
|
| 151 |
+
198,
|
| 152 |
+
151645,
|
| 153 |
+
198,
|
| 154 |
+
151644,
|
| 155 |
+
872,
|
| 156 |
+
198,
|
| 157 |
+
151669,
|
| 158 |
+
151676,
|
| 159 |
+
151670,
|
| 160 |
+
151645,
|
| 161 |
+
198,
|
| 162 |
+
151644,
|
| 163 |
+
77091,
|
| 164 |
+
198,
|
| 165 |
+
11528,
|
| 166 |
+
16134,
|
| 167 |
+
151704
|
| 168 |
+
]
|
| 169 |
+
}
|
| 170 |
+
}
|
| 171 |
}
|