Generating tokens with a transformer is inherently sequential: each token depends on all previous tokens, so you cannot generate token t+1…Continue reading on Towards AI »
From Machine Learning Fundamentals to Modern LLMsContinue reading on Medium »
This blog post discusses the details of what finetuning is, why it’s needed, and how we can finetune an LLM model with practical examples.The fine-tuning is what brings life to the LLM model. It’s a technique to make models adapt to a specific task, su…
How one research paper introduced Transformers and became the foundation of ChatGPT, Gemini, and the entire modern AI revolution.Continue reading on Medium »
Yann LeCun raised $1.03 billion for a bet that most AI is built on the wrong foundation. Three days later, a 15-million-parameter model…Continue reading on Artificial Intelligence in Plain English »
In Part 1, we built something powerful a rich, contextual representation of the sentence “How are you.” The encoder did its job…Continue reading on Towards AI »
這篇是我在學習Nvidia: Building Transformer-Based Natural Language Processing Applications的筆記之一，如有錯誤請多多見諒&…
For a decade, context length was the silent constraint shaping everything in AI. SubQ says it’s solved the math that made that constraint…Continue reading on Artificial Intelligence in Plain English »
LLaVA — Large Language and Vision Assistant is an end-to-end trained large multimodal model that connects a vision encoder and a LLM for…Continue reading on Medium »
Imagine a single knob. Turn it up, and an AI that was just trying to help you debug code starts threatening to leak your private data…Continue reading on Medium »