Tripper

A pipeline that generates animations with SD1.5 (SDXL compatibility not tested). This work is inspired by deforum, check it out.

Features

• support loading lora models (safetensors)
• support prompt token size exceeding 75, plus negative prompt
• frame zoom in/out
• multiple methods to control the smoothness/creativeness between frames

Usage

Edit the config.json to configure the settings:

• model_path: path to your SD model safetensors
• generate_animation: set false to generate a batch of images to choose from as an initial image, then set true to switch to animation mode.
• nframes: total number of frames to generate
• scheduler: choose one from "euler", "euler a", "DDIM", "DDPM", "DPM++ 2M SDE Karras", "DPM++ 2M Karras"
• num_inference_steps: per image
• lora_dict: dictionary containing: {"[path_to_lora_safetensor]":[weight]}
• zoom: zoom in(<1) or out(>1)
• strength: roughly the similarity between two consecutive images (0~1)
• diffusion_cadence: how often the images get direct diffused, controls smoothness of the animation

then run with

python generate.py

Have fun