Lately, I have been playing with Llama 3.1, the latest version of the Large Language Model powering Meta AI. I've been able to develop two key applications. I was using it for learning and research purposes until I found a great potential in it as an open source LLM with great capabilities.
The first application I created was a Trivia game on Nigerian Music. The app will ask you questions, after answering, it will award you points. However, I faced some challenges because of Llama 3.1 70B, the model I used paucity of knowledge on Nigerian Music and it's constant hallucination.
Nevertheless, I got my breakthrough application 3 days ago when I developed a Fact-checking application using the LLM. It gave incredible results with no hallucination. You can verify any claim on it, however, it won't be able to verify new claims after its release. Regardless, it is a solid application and I'm super proud of it.
The development of open source models like Llama opens us to infinite possibilities of building various AI applications without any cost or hassle. It's a game-changer in the arms race of AI development.
It's important to note that I am available for all your AI application development concerns. If you have any project in mind, I can help you bring it to fruition as an AI expert.
Find the link to the projects below:
TriviaVilla: Nigerian Music Trivia
Facto: A fast and accurate fact-checker
Top comments (0)