New to this! I have no idea what I am looking in here. Looking for explanation.

[moreAImore]

I found this project by searching for LORAS for Cogvideox (https://huggingface.co/ECNU-CILab/ExVideo-CogVideoX-LoRA-129f-v1/tree/main)

My question: What is this exaclty?

1. Can it run with a local UI? (A gradio to be installed with the code of this github repo?)
2. Is this repo adaptable to be a comfy ui custom node?
3. If it can only run locally, can we POINT to a path where are our models so we dont have to copy them twice? (I already have my flux and other models on another path) If yes, how?
4. How exaclty, step by step, could I use the "improved" version of SVD? For example?

I am not even sure where to start.

Finally, is there a discord or something? Thanks

[Artiprocher]

1. We didn't provide a gradio UI for cogvideox-lora.
2. This repo doesn't include comfy UI support, but you may search for other repo based on this repo.
3. You can use model_manager.load_models(["path_to_your_model"]) to load your own models. Our code will detect the model type automatically.
4. See here: https://github.com/modelscope/DiffSynth-Studio/tree/main/examples/ExVideo
5. We don't have discord.