Awesome
📔 Instill AI Cookbook
A collection of notebooks and guides that showcase some of the many AI solutions you can build with Instill Core and Instill Cloud.
To execute these examples, you’ll need an Instill Cloud account along with an Instill API Token. To set up your account, please refer to our quickstart guide. For generating your API Token, consult the API Token Management page. We suggest configuring your API token as an environment variable with the following command:
export INSTILL_API_TOKEN=********
Or, you can make a .env
file at the root of your repository with the content
INSTILL_API_TOKEN=********
.
Notebooks
Explore the following example notebooks and learn how our full-stack AI infrastructure can help you build and deploy real-world production-grade AI applications from the comfort of a Jupyter Notebook.
Notebook | Google Colab | Description |
---|---|---|
Hello World | Get started with Instill AI's core services: Instill VDP, Instill Model, and Instill Artifact. | |
Instance Segmentation with Instill Cloud | Perform instance segmentation on microscopic plant stomata images using a production-grade Instill VDP pipeline on Instill Cloud. | |
Generating Structured Outputs from LLMs | Evaluate OpenAI's Structured Outputs feature and other tools including Instructor, Marvin, BAML, TypeChat, and LangChain for generating structured outputs from LLMs, demonstrated with a task that involves both reasoning and structured formatting. | |
Producing Structured Web Insights | Create structured insights from Google Search or a Web URL using OpenAI's Structured Outputs within Instill VDP. | |
Parsing PDF Files to Markdown | Transform PDF files into high-quality AI-ready Markdown using Instill VDP. | |
Visualize the Semantic Content of Any Website | Crawl a website, extracting high-quality AI-ready Markdown text which is then chunked, embedded, analyzed using unsupervised ML methods, and visualized. | |
RAG with LLM-Based Evaluation Metrics | Build an API-ready financial RAG system and use LLM-based evaluation metrics to characterize and evaluate its performance. |
More examples coming soon - watch this space! 🚀