LLMs

0 episodes

Hosted byDevSocial Live

How large language models such as GPT, LLaMA and Claude work, and how to build them into real projects, from the first concepts to running them in production. • Foundations: tokenization, embeddings, attention and transformers • Prompt engineering: zero-shot and few-shot prompting, and avoiding hallucinations • Fine-tuning a pre-trained model for your domain • Deployment on-premises, in the cloud, at the edge and hybrid • Tooling: LangChain, Hugging Face Transformers and RAG pipelines • Performance: latency, cost and prompt caching • Security and privacy