Views
No views yet
onnxruntime (no GPU, no PyTorch needed at inference time).| Folder | Role | Size |
|---|---|---|
omnivoice_lm_int8_hq/ | Diffusion LM (INT8-HQ, audio head kept FP32) | ~611 MB |
audio_tokenizer_decoder_int8/ | tokens → 24 kHz waveform | ~82 MB |
audio_tokenizer_encoder_int8/ | reference audio → codec tokens (voice cloning) | ~378 MB |
tokenizer.json, config.json | text tokenizer + config | small |
Each model ismodel.onnx+model.onnx_data(ONNX external data) — download both files.
| Model | Input | Output |
|---|---|---|
| LM | input_ids[batch,8,seq] int64; audio_mask[batch,seq] bool; attention_mask[batch,1,seq,seq] bool; position_ids[batch,seq] int64 | logits[batch,8,seq,1025] f32 |
| Encoder | audio[batch,1,num_samples] f32 @24kHz | audio_codes[batch,8,num_frames] int64 |
| Decoder | audio_codes[batch,8,num_frames] int64 | audio[batch,1,num_samples] f32 |
infer_onnx.py). Example:python infer_onnx.py --variant int8hq --num-step 16 --only demo_voice_clone--num-step for higher quality (slower).omnivoice_lm_int8_hq/): derived from Qwen3-0.6B / OmniVoice — Apache-2.0.audio_tokenizer_*_int8/): derived from Higgs Audio 2 (Boson AI) —
Boson Higgs Audio 2 Community License (based on the Meta Llama 3 Community License); see
LICENSE. It includes an acceptable-use policy and Llama-style terms.other.
This is an ONNX export + quantization only — weights are unchanged; all rights belong to the
original authors (k2-fsa/OmniVoice, Boson AI, Qwen).