Head swap with the BFS LoRA on Qwen-Image-2.1 (+turbo)
Discussions about the Inference Providers feature on the Hub