How shamshad/llm-zoomcamp works
Project type and purpose
• This is neither a traditional web application nor a backend service; it is a course content repository for “LLM Zoomcamp.”
• It holds lecture notebooks, markdown lessons, homework assignments, reference code, cohort-specific materials, and links to external platforms (Slack, YouTube, Airtable).
• Its goal is to teach real-world applications of LLMs over six modules and a capstone project.
File structure
• Top level:
• README.md, project.md, asking-questions.md, awesome-llms.md – site-wide entry points and community guidelines
• directories 01-intro through 06-project-example – module-by-module content (notebooks, markdown, JSON data, PDFs)
• cohorts – year-and-module-specific workshops, homework, Docker setups, data files
• images – visual assets for README and slides
• No service code for a running application, but Dockerfiles & docker-compose.yaml appear under cohorts/2024/04-monitoring/app
• External-service demos: Elasticsearch in module 1, Qdrant in module 2, Grafana dashboards, etc.
README insights
• Describes course syllabus, external links (Airtable for signup, Slack, YouTube, Telegram)
• Enumerates modules and their focus (RAG, vector search, evaluation, monitoring, best practices, end-to-end project)
Be detailed but avoid over-segmentation
• Every essential element: authoring, CI/CD, hosting, student environment, runtime services, and communication channels