Hello how are you
✍️️

A set of model files for the C++ implementation of Gemma 2B. It is easy to download, build, and tinker with. This is a compressed version of the weights for a 2B parameter model that has been instruction-tuned. These will load, run, and download more quickly.

Pre-trained (PT) models can be used as base models for further development, while instruction-tuned (IT) variants can be used for chatting and following prompts.