08 — LLM¶
Large language models: concepts, APIs, local inference, and deployment.
Scope¶
Cheat sheets for working with LLMs directly. Retrieval systems belong to 09-RAG, agent orchestration to 10-AI-Agents.
Cheat Sheets¶
| Sheet | Covers |
|---|---|
| llm-concepts | Tokens, context windows, temperature, sampling, prompting, model selection, quantization |
| ollama | Local LLMs: CLI, OpenAI-compatible API, Modelfile, GGUF models |
| huggingface | HF pipelines, chat templates, quantization, vLLM, Inference API |
| litellm | Unified API for 100+ providers, proxy server, fallbacks, cost tracking |
Contribute¶
- Template: CHEATSHEET_TEMPLATE.md
- Guide: CONTRIBUTING.md
- Planned: fine-tuning (LoRA, data prep), evaluation, prompt engineering deep-dive