Loading
Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
int8 ONNX export of FunAudioLLM/Fun-ASR-MLT-Nano-2512 (≈0.8B, SenseVoice encoder + Qwen3-0.6B LLM decoder), packaged for sherpa-onnx's OfflineRecognizer.fromfunasrnano. Runs on CPU with no PyTorch/transformers at...
Model source
Source description
int8 ONNX export of FunAudioLLM/Fun-ASR-MLT-Nano-2512
(≈0.8B, SenseVoice encoder + Qwen3-0.6B LLM decoder), packaged for
sherpa-onnx's OfflineRecognizer.from_funasr_nano.
Runs on CPU with no PyTorch/transformers at inference. Covers the 31 languages of the
multilingual Fun-ASR-Nano, including Chinese, English, Vietnamese, Indonesian, Thai, Malay,
Korean, and more.
Sources
1 sourceVerified Sep 22
Model artifacts
3 artifactsSource excerpts
2 excerpts| file | size | purpose |
|---|---|---|
encoder_adaptor.int8.onnx | ~238 MB | SenseVoice audio encoder + adaptor |
embedding.int8.onnx | ~156 MB | token embedding |
llm.int8.onnx | ~600 MB | Qwen3-0.6B decoder |
tokenizer/ | — | Qwen3-0.6B tokenizer (vocab / merges / config) |
Total ≈ 994 MB.
import sherpa_onnx # pip install sherpa-onnx>=1.13.3
rec = sherpa_onnx.OfflineRecognizer.from_funasr_nano(
encoder_adaptor="encoder_adaptor.int8.onnx",
llm="llm.int8.onnx",
embedding="embedding.int8.onnx",
tokenizer="tokenizer",
num_threads=4,
sample_rate=16000,
provider="cpu",
language="", # "" = auto language id; or a Chinese language name, e.g. 英文 / 印尼语 / 越南语
max_new_tokens=512,
)
s = rec.create_stream()
s.accept_waveform(16000, audio_float32_mono_16k) # numpy float32 in [-1, 1]
rec.decode_stream(s)
print(s.result.text)
language="" runs the model's built-in language identification (recommended for mixed /
auto-detect audio). To force a language, pass the Chinese language name the model was
trained on (e.g. 英文 English, 印尼语 Indonesian, 越南语 Vietnamese) — the string is
interpolated into a Chinese instruction prompt, so English names do not work.
model.pt was exported to ONNX (encoder-adaptor / embedding / LLM) and dynamic int8-quantized; the tokenizer is copied unchanged from the base checkpoint. No fine-tuning or weight changes beyond quantization.Apache-2.0, inherited from the base model. See the LICENSE and NOTICE files in this repo.
This is a redistributed derivative work — attribution above.
llm.int8.onnx
onnx · 572 MB · SHA-256 61361d839355…8859 · Hugging Face
--- license: apache-2.0 library_name: sherpa-onnx pipeline_tag: automatic-speech-recognition base_model: FunAudioLLM/Fun-ASR-MLT-Nano-2512 tags: - sherpa-onnx - onnx - int8 - fun-asr - speech-recognition - asr language: - zh - en - yue - ja - ko - vi - id - th - ms - ar - hi - pt - pl - cs - da - nl - fi - el - hu - ro - sk - sl - sv --- # Fun-ASR-MLT-Nano-2512 — sherpa-onnx (int8) int8 ONNX export of **[FunAudioLLM/Fun-ASR-MLT-Nano-2512](https://huggingface.co/FunAudioLLM/Fun-ASR-MLT-Nano-2512)** (≈0.8B, SenseVoice encoder + Qwen3-0.6B LLM decoder), packaged for **[sherpa-onnx](https://github.com/k2-fsa/sherpa-onnx)**'s `OfflineRecognizer.from_funasr_nano`. Runs on CPU with no PyTorch/transformers at inference. Covers the 31 languages of the multilingual Fun-ASR-Nano, including Chinese, English, Vietnamese, Indonesian, Thai, Malay, Korean, and more. ## Files | file | size | purpose | |---|---|---| | `encoder_adaptor.int8.onnx` | ~238 MB | SenseVoice audio encoder + adaptor | | `embedding.int8.onnx` | ~156 MB | token embedding | | `llm.int8.onnx` | ~600 MB | Qwen3-0.6B decoder | | `tokenizer/` | — | Qwen3-0.6B tokenizer (vocab / merges / config) | Total ≈ 994 MB. ## Usage (Python) ```python import sherpa_onnx # pip install sherpa-onnx>=1.13.3 rec = sherpa_onnx.OfflineRecognizer.from_funasr_nano( encoder_adaptor="encoder_adaptor.int8.onnx", llm="llm.int8.onnx", embedding="embedding.int8.onnx", tokenizer="tokenizer", num_threads=4, sample_rate=16000, provider="cpu", language="", # "" = auto language id; or a Chinese language name, e.g. 英文 / 印尼语 / 越南语 max_new_tokens=512, ) s = rec.create_stream() s.accept_waveform(16000, audio_float32_mono_16k) # numpy float32 in [-1, 1] rec.decode_stream(s) print(s.result.text) ``` `language=""` runs the model's built-in language identification (recommended for mixed / auto-detect audio). To force a language, pass the **Chinese** language name the model was trained on (e.g. `英文` English, `印尼语` Indonesian, `越南语` Vietnamese) — the string is interpolated into a Chinese instruction prompt, so English names do not work. ## Provenance & modifications - **Base model:** [FunAudioLLM/Fun-ASR-MLT-Nano-2512](https://huggingface.co/FunAudioLLM/Fun-ASR-MLT-Nano-2512) (Apache-2.0). - **Export tooling:** [Wasser1462/FunASR-nano-onnx](https://github.com/Wasser1462/FunASR-nano-onnx) (the pipe...
Source context: 0 downloads · 0 likes · Pipeline automatic-speech-recognition · Library sherpa-onnx · Repo luigi000/fun-asr-mlt-nano-onnx-int8