ChatRTX: NVIDIA’s AI chatbot integrates new linguistic models to run locally on your PC


Camille Coirault

May 2, 2024 at 6:11 p.m.

1


NVIDIA strengthens its experimental chatbot ChatRTX by grafting new artificial intelligence models. He thus sees his arsenal expand and his capabilities evolve considerably.

Presented last February under the name “Chat with RTX”, ChatRTX was initially only a demo application. Concretely, the application creates a local chatbot server accessible from your browser. This allows you to feed the AI ​​with your documents and even YouTube videos, transforming your machine into a powerful research tool capable of summarizing your content and answering your questions.

A supercharged personal assistant, but still a little heavy

In terms of the minimum configuration to take advantage of ChatRTX, no change on the horizon. To take advantage of it, you will need to have an RTX 30 or 40 series graphics card with at least 8 GB of VRAM.

Initially capable of exploiting the Mistral and Llama 2 models, ChatRTX can now count on new models: ChatGLM3 (English and Chinese bilingual language model), OpenAI CLIP (capable of generating text descriptions from images, and vice versa) and Gemma from Google. The latter was also specially designed in collaboration with NVIDIA and works wonderfully on well-equipped PCs.

ChatRTX therefore centralizes all these new models and takes care of simplifying their local execution. The interface appears to be rather intuitive and it is possible to juggle between different modules according to your needs: analysis of your photos, online videos or document summaries for example.

To use it, just go to the official NVIDIA website and be a little patient. In fact, ChatRTX requires no less than 36 GB of free space to be installed!

  The interface of the February demo, where we easily recognize the NVIDIA graphic charter © Nerces

The interface of the February demo, where we easily recognize the NVIDIA graphic charter © Nerces

Voice, the next interface with AI?

We can imagine that the results are more convincing than the first demo, where ChatRTX struggled a little to perform certain tasks that were asked of it. NVIDIA has not only improved its model support, but also equipped it with a voice recognition system. How ? By integrating Whisper, an AI model designed by OpenAI, specially designed for voice recognition and transcription.

Thus, the user will be able to query ChatRTX, just using their voice. Could this be the new direction that is emerging for chatbots? Free yourself from the need for the keyboard and simply listen to answer questions? Hard to say, but it’s possible that this initiative will move the AI ​​industry in this direction.

The 5 best artificial intelligence chatbots (2024)

In 2023, the landscape of chatbots in French has expanded considerably, boosted by the rise of artificial intelligence. In the past, these assistants were rationed to predefined questions and answers. But now, thanks to advances like ChatGPT, it’s possible to ask any question and get relevant answers generated in real time.
Read more

Source : The Verge

Camille Coirault

Camille Coirault

Once I woke up in the boat arriving in Morrowind, I got my finger in the gear. Another of my fingers was also stuck between the pages of books by classic authors: Charles Baud...

Read other articles

Once I woke up in the boat arriving in Morrowind, I got my finger in the gear. Another of my fingers was also stuck between the pages of books by classic authors: Charles Baudelaire, Émile Zola, Choderlos de Laclos or Victor Hugo to name a few. Twenty years later, a few thousand hours of playing, reading, and here I am! My heart always balances between my passion for tech, video games and my immeasurable love for Letters. Spoiler: I didn't choose and it's not likely to happen anytime soon.

Read other articles



Source link -99