CogKit is an open-source project that provides a user-friendly interface for researchers and developers to utilize models from ZhipuAI, currently supports CogView (image generation) and CogVideoX (video generation) series. Users must comply with legal and ethical guidelines to ensure responsible implementation.
Visit our Docs to start.
-
Training Optimization: Includes pre-computation and caching of latents and embeddings, sequence packing, and various memory-efficient strategies to improve training throughput and reduce GPU memory usage.
-
Native Resolution Training Support: Seamlessly train models at original image resolutions for improved quality and consistency.
-
Easy-to-use Interface: Offers multiple easy-to-use inference options, including a CLI, OpenAI-compatible API server, and interactive Gradio-based UIs for both training and inference.
- Add support for CogView4 ControlNet model
- Docker for easy deployment
This project is licensed under the Apache 2.0 License.