Add FLF2V-14B model support for video generation - #4

Open

zhangtao0408 wants to merge 7 commits into

Eco-Sphere:mainfrom

zhangtao0408:flf2v_br

Commits on Nov 1, 2025

Refactor rope_apply function to simplify tensor processing and improve performance
TaoZhang-Work
committed
e6ec8a5
Refactor rope_apply function to simplify tensor processing and improve performance
TaoZhang-Work
committed
2ffd232
Update attention.py to correct layout parameter from "BNSD" to "BSND" in attention function
TaoZhang-Work
committed
1a976fa
Add support for 1024x1024 video size in configuration and attention layer
TaoZhang-Work
committed
4df1477
Add FLF2V-14B model support for video generation with first and last frame inputs, including new example images and Gradio interface for user interaction.
TaoZhang-Work
committed
323069e
Add FLF2V-14B configuration and support for multi-image prompts in prompt extender.
TaoZhang-Work
committed
5f2379d
Update FLF2V Vae Patch Parallel Encode
TaoZhang-Work
committed
66e8fda