Companion Notebooks¶
Interactive Jupyter notebooks that accompany the course notes. Each notebook covers one session's key concepts with runnable code — no setup required when opened in Google Colab.
Running in Google Colab
Click the "Open in Colab" link next to any notebook. You'll need to add your API keys as Colab Secrets (the key icon in the left sidebar) — OPENAI_API_KEY and ANTHROPIC_API_KEY. Never paste keys directly into notebook cells.
Running locally
Notebooks live in the notebooks/ directory at the repo root. Run pip install -r requirements.txt and jupyter lab to get started.
Week 01 Notebooks¶
| Notebook | Topics Covered | Open in Colab |
|---|---|---|
week01-day01-llm-fundamentals.ipynb |
Tokenization, context windows, temperature sampling | Open |
week01-day01-prompt-engineering.ipynb |
Zero-shot, few-shot, CoT, structured output | Open |
week01-day02-apis.ipynb |
OpenAI and Anthropic API calls, streaming, tool use | Open |
week01-day02-embeddings.ipynb |
Embedding models, cosine similarity, semantic search | Open |
week01-day03-rag-basics.ipynb |
Chunking, ChromaDB ingestion, end-to-end RAG pipeline | Open |
week01-day03-vector-databases.ipynb |
ChromaDB, Pinecone, Qdrant, hybrid search | Open |
week01-day04-evaluation.ipynb |
RAGAS faithfulness, answer relevancy, LLM-as-judge | Open |
week01-day04-responsible-ai.ipynb |
Prompt injection detection, OpenAI Moderation API, PII | Open |
week01-day05-huggingface.ipynb |
Hub navigation, pipeline(), Inference API |
Open |
week01-day05-local-llms.ipynb |
Ollama API calls, quantization comparison | Open |
Week 02 Notebooks¶
| Notebook | Topics Covered | Open in Colab |
|---|---|---|
week02-day01-langchain.ipynb |
LCEL pipelines, memory, async, streaming | Open |
week02-day01-advanced-rag.ipynb |
CrossEncoder reranking, HyDE, multi-query, compression | Open |
week02-day02-fine-tuning.ipynb |
QLoRA setup, SFTTrainer, evaluation, merge and save | Open |
week02-day02-function-calling.ipynb |
OpenAI tools cycle, Anthropic tool_use, Pydantic extraction | Open |
week02-day03-agents.ipynb |
ReAct from scratch, tool registration, memory strategies | Open |
week02-day03-langgraph.ipynb |
StateGraph, reducers, conditional routing, MemorySaver | Open |
week02-day04-llmops.ipynb |
LangSmith tracing, cost tracking, latency benchmarking | Open |
week02-day04-deployment.ipynb |
FastAPI + AsyncOpenAI, SSE streaming, semantic cache | Open |
Project Starter Notebooks¶
Each project has a starter notebook with data loading and client setup done. Exercise cells are left blank for you to complete.
| Project | Starter Notebook |
|---|---|
| RAG Q&A Chatbot | starter.ipynb |
| AI Writing Assistant | starter.ipynb |
| Document Summarizer | starter.ipynb |
| Function-Calling Extractor | starter.ipynb |
| LangGraph Research Agent | starter.ipynb |
| Fine-Tuned Classifier | starter.ipynb |
Notebooks vs course notes
Notebooks show working code in isolation. The course notes explain the why. Use notebooks to run and experiment, but read the notes to understand the design decisions. Don't skip the notes because the notebook runs — that's the difference between copying and learning.