File size: 1,289 Bytes
754a956
 
90ff572
 
 
 
 
 
754a956
90ff572
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
---
license: mit
library_name: onnxruntime
tags:
- onnx
- qwen3
- browser-inference
- distributed-inference
---

# WebAI@Home Qwen3-0.6B ONNX shards

This repository contains the three ONNX shards used by the [WebAI@Home Worker](https://github.com/webai-at-home/worker) to run the Qwen3-0.6B language model across browser workers.

## Files

- `shards/shard-1.onnx` contains the token embedding and the first group of decoder layers.
- `shards/shard-2.onnx` contains the middle group of decoder layers.
- `shards/shard-3.onnx` contains the final group of decoder layers and the language-model output head.

The three files are static model assets. A Worker browser downloads only the shard assigned to its stage and stores the downloaded bytes in its browser cache for later use.

## Source and version

The shards were split from the Qwen3-0.6B ONNX model published as [`onnx-community/Qwen3-0.6B-ONNX`](https://huggingface.co/onnx-community/Qwen3-0.6B-ONNX). The WebAI@Home Worker uses this repository by immutable revision, not by the moving `main` branch.

This repository is part of the [WebAI@Home](https://github.com/webai-at-home/webai-at-home) project. It is intended to be downloaded by browser code through the Hugging Face Hub and does not contain executable application code.