Popular repositories Loading
-
qwen-image-2.1-lora-12gb
qwen-image-2.1-lora-12gb PublicTrain a Qwen-Image-2.1 LoRA on a 12 GB GPU and 32 GB of RAM — NF4 weights kept in VRAM, no CPU offload, 500 steps in 12 minutes. Unofficial — not affiliated with Alibaba or Tongyi.
Shell 2
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.