Large Language Model (LLM)¶
{: .d-inline-block }
Since v20230805
Commerical License
MediaPipe4U provides a plugin that supports locally deployed large language models (LLMs). No Python code or external programs are required; start LLM inference directly in Unreal Engine.
CPU inference is currently supported, so you can use large-model AI capabilities even without an expensive GPU.
GPU inference will be provided in the future.
{: .warning}
LLM is not included in the free license.
LLM is a commercial feature. Only the Ultimate license can package this feature.
The free license can only use it in the Unreal Editor.
Road Map¶
- Local deployment
- LLaMA2 model support
- ChatGLM support
- GPU inference
- Baby LLaMA (self-trained small model)
