Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn to run Qwen Image and WAN 2.2 AI models locally on the Framework Desktop featuring AMD's Ryzen AI Max "Strix Halo" processor in this comprehensive 25-minute tutorial. Set up your system with proper BIOS and kernel configurations using a pre-built Fedora toolbox containing ROCm 7, Qwen Image Studio, and ComfyUI. Generate high-quality images using Qwen Image's state-of-the-art open-weight model, then create text-to-video and image-to-video content with Alibaba's WAN 2.2 model. Master LoRA adapters for accelerated generation times, experiment with FP8 scaled weights in ComfyUI for optimized performance, and navigate current ROCm stability challenges on Strix Halo systems. Understand the technical foundations of diffusion models including CLIP, MMDiT, VAE, and LoRA components through detailed explanations. Access complete setup scripts, toolboxes, benchmark logs, and a VRAM estimator through the provided GitHub repository, while leveraging up to 128 GB of unified memory available on Strix Halo systems for demanding AI workloads.
