ArtificialGuyBR

Home / Projects / AxialView LoRA for Wan 2.2 T2V 14B

Stable Diffusion · LoRA

AxialView LoRA for Wan 2.2 T2V 14B

LoRA for Wan 2.2 text-to-video that generates top-down axial-view shots emphasizing circular patterns and symmetrical aerial compositions.

About this model

AxialView is a LoRA adapter trained for Wan-AI's Wan2.2-T2V-A14B text-to-video model. It steers generation toward axial (top-down / aerial) viewpoints with a focus on circular patterns and symmetrical compositions, producing dramatic wide shots that reveal geometric perfection in natural and man-made structures such as coral atolls, storm systems, stadium light shows, and industrial machinery.

The LoRA was adapted specifically for the Wan-AI ecosystem and is trained against the Wan2.2-T2V-A14B base model. GPU time for training was sponsored by reDMOND Ai. The model card lists a single trigger phrase: "AxialView". Authors recommend starting every prompt with "AxialView." followed by a description of a scene viewed from directly above, emphasizing circular patterns, symmetry, and geometric arrangements.

To use it in ComfyUI, download `AxialView-WAN22.safetensors` and place it in `ComfyUI/models/loras/`. Add a Load LoRA node after the Wan base model loader, connect its MODEL output to the KSampler, and set strength_model around 0.8-1.0 as a starting point. The LoRA is compatible with both text-to-video and text-to-image pipelines on the Wan2.2 base.

License: the model card declares cc0-1.0 (Creative Commons Zero).

Learn more

Read the in-depth guide about AxialView LoRA for Wan 2.2 T2V 14B

Project signals

  • 18 Hugging Face downloads
  • 1 Hugging Face likes

Topics

diffusers · lora · wan · text-to-video · axial-view · aerial · top-down · circular · geometric · cinematic

Explore the source