- Published on
This post provides a brief overview of the ideas behind LoRA and its quantized counterpart QLoRA, which enable efficient fine-tuning of LLMs
This is some of the stuff I recently learned and write about!
CraftOne was a lightweight and user-friendly website builder optimized for high performance. It featured AI-guided design assistance powered by a LLM.
Shipfast is a beautifully designed, feature-packed template built to streamline the process of launching SaaS applications quickly and efficiently.
A from-scratch, 50M-parameter implementation of Meta’s LLaMA 3 model. It supports RoPE (rotary positional encodings), grouped query attention (GQA), and a KV cache. Trained on the TinyStories dataset, the model is capable of generating short, coherent stories.