Qureshimuhammadabubakar·8h agoBigger batch size is not always faster: fine-tuning Gemma 3 270M on four GPUsI recently fine-tuned Google’s Gemma 3 270M following Daniel Bourke’s Hugging Face full fine-tuning tutorial, and I got curious about one…
Brajendra Singh·1d agoLoRA Fine-Tuning Made Simple: Concepts and Math ExplainedUnderstand how LoRA works, why it needs fewer trainable parameters, and how lightweight adapters let one LLM handle multiple tasks.
Dineshk·1d agoWhen the Model Has to Change: Four Problems That Tell You It’s TimeFrom document extraction and computer vision to LLM fine-tuning: four engineering problems that taught me when to train, when to adapt, and…
Shivam Mishra·2d agoFine Tuning Vs RAG : Understanding when to use.As I continue learning GenAi , One question that i found perticularly intersting is :
InProduction ReadybyAnupam Sinha·2d agoFine-Tuning vs RAG vs Prompting — The Decision Nobody ExplainsThree ways to customise an LLM, one giant misconception that wastes budgets, and a decision framework that actually tells you which to…
Paul Hackenberger·2d agoWeight-Space Programming as a light-weight Approach to Recursive Self-ImprovementA reported research loop, a sleeping model that consolidates memories, and a lightweight way to test direct weight edits.
XB Software·2d agoRAG vs. Fine-Tuning: Why Your AI Application May Need BothLarge language models can generate convincing answers, summarize documents, and interact with users in natural language. But when…
Marcin Dudek·3d agoDoes Your Fine-Tuned LLM Sound Like You? How to Test ItI trained a small local model, Qwen3-4B, on my X replies and wanted to know one thing: does it sound like me? Answering that took seven…
Marcin Dudek·3d agoFine-Tune a Local LLM to Write Like You on a MacI fine-tuned a small local model, Qwen3-4B, on about 17,600 of my own X posts and replies, most of them paired with the post I was…
ElvisM89·3d agoA Small Model, Trained for a Specific JobHow we adapted a 0.6B language model to do one structured rewriting job reliably.A response icon1A response icon1
