Observability for LLM Systems
You’ve built an LLM-powered app. It works great in your notebook. Then you ship it, and something breaks — but what? Was it the prompt? The retrieval? The model itself? Without observability, you’r...
fareeha
Personality Post-Training
How do you take a base LLM and give it a personality? Not just make it accurate — make it funny, empathetic, blunt, or sound like a specific character? That’s what post-training is for. This is a ...
Hugging Faces - selecting models
A practical walkthrough of how to find, evaluate, and use open-source models from the Hugging Face ecosystem — covering everything from model selection to running NLP, audio, and vision tasks with ...
Creating a 🧪 Longevity & Biohacking agentic RAG using LlamaIndex
LLMs have leveled up my self-care game, so I played around with building a longevity personal research assistant to summarize findings and answer questions. Here’s how you can built one too: I use...
Finetuning Bascis
What is Fine-Tuning? Fine-tuning is the process of adapting a general-purpose model like GPT-3 to better suit specific tasks or domains. For example, fine-tuning GPT-3 to ChatGPT, optimized for c...
Gradio Tutorial: Creating Interactive ML Demos
I’m a visual thinker and learner, so finding Gradio was a game-changer. With minimal code, you can launch a web-based UI in your browser — with both a local and publicly shareable link (active for...