This repository contains a collection of Jupyter notebooks demonstrating how to use the Writer Python SDK for various AI-related tasks. These notebooks provide practical examples and tutorials for working with our models and services, including model retrieval, text and chat completion, knowledge graph manipulation, and more.
You can run the cookbooks locally, or, if you prefer, copy and paste the code snippets from the notebooks into your own Python files.
- Prerequisites
- Installation
- Running the cookbooks
- Cookbook descriptions
- More resources
- About Writer
- Support
- Python 3.7 or higher
- Jupyter Notebook or JupyterLab
- pip (Python package installer)
- An AI Studio account
- A Writer API key. Follow this API Quickstart to get your API key.
To run the cookbooks locally, follow these setup steps:
-
Clone this repository:
git clone https://github.com/writer/cookbooks.git cd cookbooks -
Create and activate a virtual environment:
python -m venv my_env source my_env/bin/activate # On Windows, use `my_env\Scripts\activate` -
Once you have an API key, we recommend that you store it as an environment variable in a
.envfile like so:WRITER_API_KEY="{Your Writer API key goes here}"
-
Ensure your virtual environment is activated.
-
Start Jupyter Notebook or JupyterLab:
jupyter notebookor
jupyter lab -
Navigate to the notebook you want to run and open it.
-
Follow the instructions within each notebook to execute the cells and interact with the Writer SDK.
- Applications (
/applications)application_basic_usage.ipynb: Invoke no-code applications using the Writer SDK.application_graph_management.ipynb: Attach a Knowledge Graph to a no-code application via the Writer SDK.application_jobs_utilization.ipynb: Run no-code applications asynchronously and manage the async jobs.
- Completion (
/completion)text_completion.ipynb: Use the Writer SDK for text completion tasks.chat_completion.ipynb: Explores chat-based interactions.structured_output_cookbook.ipynb: Shows how to use structured output in Writer AI Studio with JSON Schema and Pydantic, for streaming and non-streaming responses, usingchat.chatandchat.parse.multimodal_chat_x5.ipynb: Demonstrates passing images and text to Palmyra X5 chat completions, using local or hosted images.
- Knowledge Graphs (
/knowledge_graph)knowledge_graph.ipynb: Introduces the basics of working with files and Knowledge Graphs.
- Models (
/models)model_retrieval.ipynb: Retrieve and list available Palmyra models.palmyra_creative.ipynb: Use Palmyra Creative for brainstorming and creative tasks.palmyra_fin.ipynb: Use Palmyra Fin for financial analysis.palmyra_med.ipynb: Use Palmyra Med for medical analysis.
- Tool calling (
/tool_calling)tool_calling_api.ipynb: Introduces the basics of tooling calling and how to use the Writer SDK for tool calling tasks.tool_calling_kg.ipynb: Query a Knowledge Graph during a chat.tool_calling_llm.ipynb: Invoke a different LLM during a chat with a Palmyra model.tool_calling_mcp.ipynb: Use the Writer MCP server to connect agents to Writer tools.tool_calling_streaming.ipynb: Use tool calling with streaming responses.tool_calling_math.ipynb: Solve a math problem using tool calling.tool_web_search: Demonstrates how to use the prebuilt web search tool in Writer. Covers setup, basic usage, and examples of customizing search parameters such as limiting results, filtering by date, and including raw source text in responses.tool_translation_chat.ipynb: Shows how to use the translation chat completions tool in Writer AI Studio, including setup and working examples for translating text in a chat.
- Tools (
/tools)medical_comprehend.ipynb: Analyze a medical document and extract entities.pdf_parser.ipynb: Parse a PDF file.
- Writer developer docs
- Writer Framework: An open-source Python framework for rapidly building AI apps
- Writer Framework sample apps: Example applications built using the Writer Framework
- Writer Python SDK
- Writer Node SDK
Writer is the full-stack generative AI platform for enterprises. Quickly and easily build and deploy AI apps with a suite of developer tools fully integrated with our LLMs, graph-based RAG, AI guardrails, and more. To learn more, visit our website.
If you encounter any issues or have questions, please file an issue on this repository.