DiffusionGemma is a new open-weight Gemma model for faster text generation. This release offers up to 857 tokens/second inference speed, significantly improving text generation efficiency. It's available on Hugging Face and hosted by NVIDIA for free cloud API access.
Opening Kapyn…