fusorlabs/qwen3.5-2b-web

Qwen3.5-2B, compiled for the browser

Qwen's Qwen3.5-2B compiled into a fusor-ai bundle: the weights the browser runtime of qwen3-5-wasm loads to talk about text, images and videos on the device's CPU. Try it in the Space fusorlabs/qwen3-5-wasm.

• Source: Qwen/Qwen3.5-2B at revision 15852e8c16360a2fea060d615a32b45270f8a8fc, compiled with the q7 profile: every large matrix 7-bit (2517 MB).
• Format: manifest.json lists every tensor and part; each part is named by its BLAKE3 hash, which the runtime checks as it loads.
• Reproduce it: just model in qwen3-5-wasm downloads the checkpoint and compiles the same bundle.

This is not a PyTorch or safetensors checkpoint: it only loads in the qwen3-5-wasm runtime.

The model is Apache-2.0 (LICENSE).