Vectara Hackathon Guide
Vectara Overview
Vectara is the trusted GenAI platform. It's designed to make it easy for you to build and deploy GenAI applications that can generate text-based answers using your particular data (this type of application flow is also known as RAG or retrieval-augmented-generation). You just ingest your data and then build apps using the Query and Summarization APIs. It's that simple.
Use-cases include:
- Question Answering: provide direct answers to specific questions based on the data.
- Conversational AI / Chat: chatbots that can hold a full conversation with a human, with back-and-forth exchanges.
- Semantic (Neural) Search: build applications powered by fast and powerful semantic search that finds documents by matching to the user intent.
Getting started with Vectara
You can get to know Vectara’s features in 5 minutes:
We have more resources to help you build apps with Vectara:
- The Quick Start guide shows you how to use the Vectara Console.
- Read the API recipes for common patterns with our APIs.
- Our API playground shows you how Vectara’s API requests and responses are structured.
- See “Additional resources”, below, for a comprehensive list.
General guidelines:
- Vectara provides a generous free plan (ideal for hackathon) with up to 50MB of text and 15K monthly queries.
- Use the Indexing API to ingest data into your Vectara corpus.
- Use the File Upload API to upload files such as PDF, PPT or DOC.
- Use the Search API to run queries against the data ingested.
- The Console provides a unified view of information about your Vectara account, corpora and their characteristics. You can also run example queries directly from the console for quick experimentation.
Common Questions
What should I do if I need to go over the free plan limits?
Vectara’s free plan provides a pretty generous offering of up to 50MB of text and 15K queries a month. This should be sufficient for many use-cases, including in a hackathon setting.
To get additional capacity, you can add a credit card to the account and buy additional bundles, or transition to our Scale plan.
Should I use RAG and not fine-tuning?
Our experience shows that “Fine-tuning is for Form, and RAG is for facts” as discussed here and here.
What is the Boomerang embedding model?
Boomerang is the name of Vectara’s newest embedding model. This model encodes text from your data as “vector embeddings” and is used to power the high performance retrieval process that is part of the RAG pipeline.
Read more about Boomerang, and the importance of using a good retrieval model for getting best results from RAG.
Where is my customer ID / corpus ID / API key?
Here’s some info on how to find your customer ID.
To get a corpus ID, view a corpus in the Console. The corpus ID is at the top of the screen, e.g. “Corpus ID: XXX”.
To create API Keys - follow this guide.
What is HHEM?
- HHEM stands for the “Hughes Hallucination Evaluation Model” and is an open source model created by Vectara, which can evaluate the likelihood a foundation model will hallucinate. It was named for Simon Hughes.
- Model card: https://huggingface.co/vectara/hallucination_evaluation_model
- Blog: https://vectara.com/blog/cut-the-bull-detecting-hallucinations-in-large-language-models/
- Leaderboard: https://huggingface.co/spaces/vectara/leaderboard
Additional resources
| API docs | https://docs.vectara.com/docs/ |
|---|---|
| API playground | https://docs.vectara.com/docs/rest-api/ |
| Getting help | Join our Discord server or Discussion forums if you have questions. If you have any feedback for us, we would be glad to hear it - please let us know in the forums or our Discord channel. |
| Open source | We have created two useful open source projects to help: |
| - vectara-ignest helps with data ingestion - crawling data sources and indexing them into Vectara. | |
| - vectara-answer is a user interface for question answering - demonstrates a UI concept. | |
| - React-search is a React package that allows you to integrate Vectara semantic search in any React app with just a few lines of code. | |
| - React-chatbot is a React package that allows you to integrate Vectara Chat in any React app with just a few lines of code. | |
| - Create-UI is a fast way to generate a Vectara-powered sample codebase for a range of user interfaces. | |
| Sample applications | We have quite a few "sample applications" hackers can take a look at https://vectara.com/demos/. Additionally, we published some sample code in these jupyter notebooks. |
| Blog posts | A reference architecture for RAG |
| What is vector search? | |
| RAG done right: chunking | |
| RAG done right: retrieval | |
| RAG done right: databases | |
| Vectara Chat | |
| Vectara and Airbyte | |
| Vectara and Unstructured | |
| Youtube | Flowise + Vectara Tutorial |
| Langflow + Vectara Tutorial | |
| Ask LangChain video | |
| More here | |
| Integrations | LangChain: https://blog.langchain.dev/langchain-vectara-better-together/ |
| LlamaIndex: https://vectara.com/llamaindex-vectara/ | |
| Vectara and Airbyte: https://vectara.com/blog/vectara-and-airbyte/ | |
| Startup program | https://vectara.com/startups |
Happy Hacking!
