NVIDIA Updates Its Local Chatbot with Gemma and Voice Queries

IBL News | New York

NVIDIA updated this month its experimental ChatRTX chatbot app, which creates a local chatbot server that users can access from a browser and feed with docs, notes, images, YouTube videos, or other data, in order to get summaries and answers to questions.

The update expands the power of NVIDIA’s RTX-accelerated chatbot with additional features like video and photo/image search, as well as support for new models.

The chatbot, which runs locally on a Windows PC workstation, supports a growing list of AI models that include Google’s Gemma, ChatGLM3, and OpenAI’s CLIP, intended to search and interact with local photo data.

It comes with an integrated Whisper, an AI speech recognition system that let users  search data using their voice.

Available as a 36GB download from Nvidia’s website, ChatRTX also now supports ChatGLM3, an open bilingual (English and Chinese) large language model.