teaching_llm_applications

Things to cover

• token budget (you train until run out of money! )

• DAMPTP LLM course material see

• bluedotimpact practical

• Small language models (SLMs) for low resource languages

• SLMs finetuned on specific use cases for finance/healthcare

• SLMs/guardrails > train on curated datasets so that they remain on task remain on task do not hallucinate cannot be jailbroken. Use SLMs for these tasks (trained for a specific task and on curated datasets)

• Training AI models on private data on-premise (healthcare, NASA, space, government sectors, defense, TREs [trusted research environments] etc.)

Technical

• Paged attention Vizuara video

• KV cache Vizuara video

• [LLMs from scratch Vizuara]

• LLMs from scratch Karpathy

• LLMs from scratch NanoGPT

• The busy person’s intro to LLMs by Andrej Karpathy

• LLM from scratch and here

• ANNs from zero to hero

• deep dive into LLMs by Andrej Karpathy

Other topics

• Prioritize explainability and uncertainty over fancy visuals.

• Test early with real people and iterate.

• HCI lectures

K-V Cache

• K-V Cache

This site is open source. Improve this page.