Appendix: Advanced Topics
This chapter is in development.
Topics to cover:
- Temperature and sampling parameters
- Determinism vs creativity tradeoffs
- Tokens, tokenization, counting tokens, "Strawberry has how many R's?" example to explain how tokens are not characters.
- Token limits and how to work within them
- Cost optimization strategies
- Rate limits and retry logic
- RAG with Vectors
- RAG with Graphs
- Guardrails: tool call restrictions, content policies, monitoring
- File uploads: how to deal with large file uploads.
- File uploads: treating documents as images for better visual parsing (and its trade-offs)