Bard, ChatGPT, Copilot (formerly Bing Chat) are the best known and used chatbots in the world and they allow us ask all kinds of questions so that, depending on the content with which they were trained, they offer us an answer that is not always valid, due to the variety of data with which they were entered. A problem that we will not have with NVIDIA’s chat with RTX.
For a few months now, ChatGPT allows you to create personalized chatbots that base their responses on the data with which they have been trained, to create chatbots with very specific and specific topics like, for example, comics, cooking recipes, routines of training, musical and cinematic knowledge, and much more. more, further.
Create your own custom chatbot with NVIDIA
Chat with RTX by NVIDIA is an artificial intelligence chatbot for Windows powered by TensorRT-LLM. Unlike ChatGPT’s personalized chatbot creation solution, it is completely free for all users whose graphics card is an RTX 30 or 40 series with a minimum of 8 GB of video memory.
This new feature in NVIDIA Graphics is designed to create personalized chatbots with content that we have stored on our computer, both images and documents. But, in addition, we can also add links to web pages or even to YouTube to extract the text from a video and on which to ask all kinds of questions.
Once we have added all the sources we want to use to interact with this chatbot, it’s not necessary have an Internet connection, since the whole process is carried out locally as is the case with other chatbots from Google, Microsoft and ChatGPT.
So that the chatbot can recognize and analyze all the sources that we wish to use, it is necessary connect it to a big language model (LLM) like Llama 2 or Mistral. To interact with all the information we collected to create the chatbot, it’s time to interact with it.
You don’t need any knowledge to start interacting with Chat with RTX, since the user interface It is very simple. The first thing we need to do is select the AI model we want to use, then establish the sources we want to use, if we want to add additional sources that we took into account during the training process. Finally, we must ask the question in such a way that the appropriate answer is displayed based on the data we have entered, so that it is always much more reliable and truthful than if we use Bard, ChatGPT or Copilot from Microsoft.
If you want to create your own chatbot, you can download the app Chat with NVIDIA’s RTX directly from its website. This application is only compatible with Windows 10 and Windows 11, requires a minimum of 16 GB of RAM and logically requires the latest drivers currently available on the market.
This is not the first NVIDIA application that uses AI, NVIDIA RTX Voice being another example, an application that eliminates background noise from our environment using artificial intelligence.
