Qwen3.5-4B Q3_K_L WebGPU (moorejared97)

A community-maintained, Q3_K_L-quantized build of Alibaba's Qwen3.5-4B multimodal language model packaged for WebGPU inference. It enables browser-side and consumer-hardware deployment of the 4B-parameter vision-language model without a hosted backend. Source: supabase.