GPT Index v0.4.4

GPT Index v0.4.4 introduces some exciting enhancements to our query capabilities to better interface with our data structures.

• Added a QueryBundle abstraction allowing you to separate the query string from the string that's used for embedding lookup.
• Added a QueryTransform class allowing us to transform queries within our data structures - first implementation is with HyDE

Some examples are given here: https://github.com/jerryjliu/gpt_index/blob/main/examples/query_transformations/HyDEQueryTransformDemo.ipynb

GPT Index v0.4.3

GPT Index v0.4.3 is out:

We've added a Knowledge Graph 🧠 index! Build a KG by extracting triplets, and leverage the KG during query-time.

Take a look at this nb for an example: https://github.com/jerryjliu/gpt_index/blob/main/examples/knowledge_graph/KnowledgeGraphDemo.ipynb

GPT Index v0.4.2

GPT Index 0.4.2 is out:

• Added caching to download_loader call! Use from cache instead of downloading from llamahub.ai every time. (thanks @emptycrown)
• Added pinecone kwargs to all index operations for pinecone index
• RSS reader fix (thanks @bborn)
• Fixed some shorturl/python bugs

GPT Index v0.4.1

GPT Index 0.4.1 is out:

• Added Azure OpenAI example notebook! (Thanks Rob)
• Added @github repository loader! (Thanks ahmetkca)
• Some nice response refactors (thanks Kevin)
• Updated retry policy for @openai (thanks Ajinkya)

• bug fixes/doc improvements! @lukawskikacper, Marco, Patrick

GPT Index v0.4.0

GPT Index 0.4.0 is out! We bumped the middle version up because we’ve now added

📄🪵 Full support for Python logger 🪵📄

Throughout the entire codebase. No more print statements!

HUGE shoutout to @mcminis1 for landing this awesome change!

NOTE  ⚠️: this is a breaking change. There are no more verbose parameters anymore.

See the following example for how to configure basic logging to stdout (similar to print statement).

GPT Index v0.3.6

GPT Index 0.3.6 is out:

• Added parser/reader for .mbox files - feed emails 📧 into GPT Index! (Thanks Markus)
• Hot fixed Qdrant index querying (thanks @lukawskikacper)
• Refactored openai embeddings enum (thanks Kevin)

We have a very exciting new release coming up tomorrow, stay tuned :)

GPT Index v0.3.5

GPT Index 0.3.5 is out.

• You don’t need to specify query_configs for a recursive query anymore. We’ll use default configs instead.
• 📁 Added save/load from string for indices/graphs - allows you to save/load from other sources besides disk!
• Fixed google reader import error (thanks ahmetkca)

GPT Index v0.3.4

• You can now “retrieve” 🔖 any index within a composed graph! See below 👇for an example walkthrough of how this works.
• Added headers to our markdown parser

GPT Index v0.3.3

GPT Index 0.3.3 is out:

• Added an epub parser! (Thanks Jimmy)
• Swapped out the CSV parser for a pandas-based parser (thanks Johanna).
• Fixed token too big bug

GPT Index v0.3.2

GPT Index 0.3.2 is out:

• Big update: Added qdrant index and reader! Thanks @kacperlukawski for implementing everything from scratch. Now you can use qdrant both as a data source and as a store for your index.
• Fixed Faiss bug, Document metadata bug