All cookbooks
Granite Language
- Agentic RAG: Use Granite to answer complex queries using external information.
- Travel planner agent: Use Granite with LangChain to build a travel planner agent that can answer complex queries about planning a trip.
- Function-calling: Interpret customer queries, interact with APIs, and perform real-time data operations on a database through function-calling capabilities with Granite.
- Fine-tuning pirate style: Fine-tune Granite on a custom ‘pirate-talk’ dataset using the qLoRA (Quantized Low-Rank Adaptation) technique.
- PDL stand-alone or from a Python SDK: Get started with PDL and Granite.
- Docling RAG: Use Granite to build an AI-powered document retrieval system with Docling.
- LangChain RAG: Use Granite to build a RAG system with LangChain.
Granite Guardian
- Get started guide: Get started with Granite Guardian to identify potential risks in prompts and responses across dimensions.
- Detailed Guide: Explore in-depth various use cases that highlight different risk in prompts and responses across various risk dimensions of our taxonomy.
- HAP detection: Use Granite Guardian to detect hate, abuse, and profanity, either in a prompt, the output, or both.
- Usage governance workflow: Use Granite Guardian with IBM’s AI Risk Atlas to evaluate the risk profile of a potential AI use case.
Granite Time Series
- Get started guide: Get started with Granite Time Series running locally.
- Get Started with watsonx AI SDK: Use the watsonX SDK to perform inference calls against a model hosted remotely on watsonX.
- Few-shot fine-tuning and evaluation: Use Granite Time Series TTM for few-shot forecasting on an energy demand dataset.
- Zero-shot inference with exogenous data: Use Granite Time Series with zero-shot inference and fine-tuning to forecast bike sharing demand.
For more recipes, visit:
