Notes from projects I build and problems I investigate.
Posts
- ComfyUI on the RX 6800: fixing INT8, GPU VAE decode, and VRAM spill
- Qwen-Image-2.1 on Radeon 780M: faster images and RGBA sprites
- Qwen3.8-27B on Radeon 780M: inference speed and game repair
- A Local LLM Harness for Building and Testing a Browser Game on a Radeon 780M
- Generating bilingual Open Graph images locally with Qwen-Image
