Youness Mansar

A practical tutorial with full code examples for building and running multi-tool agents Youness Mansar · Follow Published in Towards Data Science · 10 min read · 10 hours ago — Photo by Carter Yocham on Unsplash LLMs are remarkable — they can memorize vast amounts of information, answer general knowledge questions, write code, generate stories, and even fix your grammar. However, they are not without limitations. They hallucinate, have a knowledge cut-off that may

Simple techniques to alleviate LLM hallucinations using LangGraph Youness Mansar · Follow Published in Towards Data Science · 8 min read · 5 hours ago — Photo by Greg Rakozy on Unsplash If you’ve worked with LLMs, you know they can sometimes hallucinate. This means they generate text that’s either nonsensical or contradicts the input data. It’s a common issue that can hurts the reliability of LLM-powered applications. In this post, we’ll explore a few

Youness Mansar · Follow Published in Towards Data Science · 7 min read · 3 hours ago — A technical walk-through on leveraging multi-modal AI to classify mixed text and image data, including detailed instructions, executable code examples, and tips for effective implementation. Photo by Tschernjawski Sergej on Unsplash In AI, one of the most exciting areas of growth is multimodal learning, where models process and combine different types of data — such as images

Easy, extendable and fast LLM agent deployment template Youness Mansar · Follow Published in Towards Data Science · 8 min read · Just now — Photo by Jeremy Bishop on Unsplash Many tutorials show how to implement an LLM agent. However, resources on deploying these agents behind an API or a user-friendly UI are limited. This post addresses this gap with a step-by-step guide to implementing and deploying a minimal yet functional LLM agent. This

Generate new images based on existing images using diffusion models. Youness Mansar · Follow Published in Towards Data Science · 6 min read · 9 hours ago — Original image source: Photo by Sven Mieke on Unsplash / Transformed image: Flux.1 with prompt “A picture of a Tiger” This post guides you through generating new images based on existing ones and textual prompts. This technique, presented in a paper called SDEdit: Guided Image Synthesis and

Tutorial on enforcing JSON output with Llama.cpp or the Gemini’s API Youness Mansar · Follow Published in Towards Data Science · 6 min read · 11 hours ago — Photo by Etienne Girardet on Unsplash Large Language Models (LLMs) are great at generating text, but getting structured output like JSON usually requires clever prompting and hoping the LLM understands. Thankfully, JSON mode is becoming more common in LLM frameworks and services. This lets you define

An MLLM fine-tuning tutorial using the newest pocket-sized Mini-InternVL model Youness Mansar · Follow Published in Towards Data Science · 6 min read · 7 hours ago — Photo by Maarten van den Heuvel on Unsplash The world of large language models (LLMs) is constantly evolving, with new advancements emerging rapidly. One exciting area is the development of multi-modal LLMs (MLLMs), capable of understanding and interacting with both texts and images. This opens up a

Exploring possible use cases of Phi-3-Vision, a small yet powerful MLLM that can be run locally (with code examples) Youness Mansar · Follow Published in Towards Data Science · 7 min read · 9 hours ago — Photo by RoonZ nl on Unsplash Microsoft recently released Phi-3, a powerful language model, with a new Vision-Language variant called Phi-3-vision-128k-instruct. This 4B parameter model achieved impressive results on public benchmarks, even surpassing GPT-4V in some cases and

Augmenting LLM Apps with a Voice Modality Youness Mansar · Follow Published in Towards Data Science · 6 min read · 8 hours ago — Photo by Ian Harber on Unsplash Many LLMs, particularly those that are open-source, have typically been limited to processing text or, occasionally, text with images (Large Multimodal Models or LMMs). But what if you want to communicate with your LLM using your voice? Thanks to the advancement of powerful speech-to-text

Build custom web apps easily and quickly Youness Mansar · Follow Published in Towards Data Science · 8 min read · 3 hours ago — Photo by Alexander Schimmeck on Unsplash Meet NiceGUI, a simple Python-based UI framework that works smoothly with your web browser or as a desktop app. Whether you’re making small web apps, dashboards, or playing with robotics projects, NiceGUI makes it easy with its easy interface and many features. The goal