How can I detach or disable a specific LoRA, such as MoviiGen1.1, from the main model during inference?"
Great wrok! I would like to remove the LoRA model MoviiGen1.1 because it gives the output too much of a cinematic or movie-like look, which is stronger than I want.
Thanks! I'd like to ask: what's the difference between a native PNG and an outer PNG? Also, which LoRA is MoviiGen1.1?
np. -You have two WFs - native and wrapped. Native is.... native for comfyui - sampler and almost all other nodes are from Comfy Core. Wrapper uses WanVideoWrapper nodes to process generation.
Lora you seek is Wan21_T2V_14B_MoviiGen_lora_rank32_fp16.safetensors (used in T2V version).
Also remember: if you use fused model you need to replace it with plain one and load FusionX loras (like in WF made by vrgamedevgirl84).
Thank you! When using the WAN base model, should I choose the I2V model or the FL2V model? I'd like to use it with the I2V model.