Skip to content
ModelVerifiedExternal
Download

Omnivoice Onnx Kv B1 FP16

This repository contains the model bundle used by OmniVoice Vulkan Bridge. The main diffusion model is the fp16, KV-cached, B=1 split ONNX export.

Sourcehuggingface.co/MarkShark2/omnivoice-onnx-kv-b1-fp16/tree/0111dd42766f32e6ebb91434ded76f0e697d834a
KindAudio modelBase modelk2-fsa/OmniVoiceVersionRevision 0111dd4Publisher@MarkShark2Cgrade

Model source

Kind
Audio model
File size
1.61–624 MB
Base model
k2-fsa/OmniVoice
Version
vRevision 0111dd4
Updated
Oct 2, 2026
Publisher
@MarkShark2
Downloads
0
Tasks
Audio
Source
Hugging Face

Source description

OmniVoice WebGPU FP16 KV B1 ONNX Bundle

This repository contains the model bundle used by OmniVoice Vulkan Bridge. The main diffusion model is the fp16, KV-cached, B=1 split ONNX export.

Bridge code: MarkShark2/omnivoice-vulkan-bridge

Model Notes

  • Base model: k2-fsa/OmniVoice
  • Main graph: fp16, KV-cached, B=1 split conditional/unconditional CFG passes
  • Decoder: WebGPU-compatible graph surgery for ONNX Runtime Web 1.25 prerelease
  • Encoder: WebGPU-capable voice-reference encoder with WASM fallback in the bridge

Attribution

Based on OmniVoice by k2-fsa and Xiaomi Corp.

The original inference code for this project was based on Magkino/vocoloco_tts.

Sources

1 source
Hugging FacePrimaryVersion 0111dd42766f32e6ebb91434ded76f0e697d834a
huggingface.co/MarkShark2/omnivoice-onnx-kv-b1-fp16/tree/0111dd42766f32e6ebb91434ded76f0e697d834a

Verified Oct 2

Model artifacts

3 artifacts

omnivoice-decoder-webgpu.onnx

onnx · 111 MB · SHA-256 12ddfca66e03…9cc2 · Hugging Face

Download

omnivoice-encoder-fixed.onnx

onnx · 624 MB · SHA-256 e6b492911cde…9ad0 · Hugging Face

Download

omnivoice-main-kv-fp16-b1.onnx

onnx · 1.61 MB · SHA-256 cab55e2fe02d…745e · Hugging Face

Download

Source excerpts

3 excerpts
ReadmeJul 11huggingface.co/MarkShark2/omnivoice-onnx-kv-b1-fp16

--- license: apache-2.0 tags: - text-to-speech - tts - onnx - onnxruntime - webgpu - vulkan - voice-cloning - fp16 - kv-cache language: - multilingual library_name: onnxruntime base_model: k2-fsa/OmniVoice --- # OmniVoice WebGPU FP16 KV B1 ONNX Bundle This repository contains the model bundle used by OmniVoice Vulkan Bridge. The main diffusion model is the fp16, KV-cached, B=1 split ONNX export. Bridge code: [MarkShark2/omnivoice-vulkan-bridge](https://github.com/MarkShark2/omnivoice-vulkan-bridge) ## Model Notes - Base model: `k2-fsa/OmniVoice` - Main graph: fp16, KV-cached, B=1 split conditional/unconditional CFG passes - Decoder: WebGPU-compatible graph surgery for ONNX Runtime Web 1.25 prerelease - Encoder: WebGPU-capable voice-reference encoder with WASM fallback in the bridge ## Attribution Based on OmniVoice by k2-fsa and Xiaomi Corp. The original inference code for this project was based on [Magkino/vocoloco_tts](https://github.com/Magkino/vocoloco_tts).

NoteJul 11huggingface.co/MarkShark2/omnivoice-onnx-kv-b1-fp16

Source context: 0 downloads · 1 likes · Pipeline text-to-speech · Library onnxruntime · Repo MarkShark2/omnivoice-onnx-kv-b1-fp16

NoteApr 26huggingface.co/api/models/MarkShark2/omnivoice-onnx-kv-fp16

Source context: 0 downloads · 0 likes · Pipeline text-to-speech · Library onnx · Repo MarkShark2/omnivoice-onnx-kv-fp16