The data structure for multimodal data · Neural Search · Vector Search · Document Store https://xitter.cf/t.co/1TMkliw3N5
Berlin, Germany
Joined November 2022
- Tweets104
- Following17
- Followers427
- Likes44
DocArray retweeted
🎉@docarray is now an Incubation Stage project!
Their dynamic schemas, vectorDBs integration, and seamless framework compatibility are fueling their journey of success.
Join us in celebrating their achievements! 🌟
Read more: hubs.la/Q01ZQdBk0
DocArray retweeted
We've updated Jina to 3.19 with new features:
👉 Support for all versions of @DocArray
👉 Hubble adapts to the local DocArray version
👉 Support for pulling from private registries using Kubernetes secrets
👉 Rewrote documentation for new DocArray
github.com/jina-ai/jina/rele…
DocArray retweeted
Finally we have managed to bring all the new benefits introduced with @docarray refactor into github.com/jina-ai/jina.
Now you can serve your models and applications with your own types and schemas. Jina adapts to your data instead of forcing you to adapt to Jina expected schema
We've updated Jina to 3.19 with new features:
👉 Support for all versions of @DocArray
👉 Hubble adapts to the local DocArray version
👉 Support for pulling from private registries using Kubernetes secrets
👉 Rewrote documentation for new DocArray
github.com/jina-ai/jina/rele…
DocArray retweeted
🎉 Excited to introduce DocArrayRetriever in @langchain!
Looking to link external data to LLMs? @docarray is your key to:
1⃣ Representing data flexibly
2⃣ Numerous data storage options
3⃣ Smooth integration into Langchain
Follow this thread for a step-by-step guide 🧵
1/5
DocArray retweeted
[1] @supabase integration: supabase.com/docs/guides/ai/…
Tair (@AlibabaGroup) integration: gpt-index.readthedocs.io/en/…
In-memory stores using @docarray:
gpt-index.readthedocs.io/en/…
gpt-index.readthedocs.io/en/…
DocArray retweeted
This is a big deal.
Now @jinaAI_ fully support @docarray v2 and allows to build multi-modal AI API with an experience similar to @FastAPI
Here is how it looks like now
We've updated Jina to 3.17, focusing on compatibility with @DocArray v2.
👉Flows and Deployments fully support DocArray v2
👉Support includes HTTP, gRPC, and WebSocket
Now you have the capacity and expressivity of DocArray v2 with all the power of Jina.
github.com/jina-ai/jina/rele…
DocArray retweeted
Thrilled to see @docarray in action within @DeepLearningAI_'s Langchain course, delivered by @AndrewYNg and @hwchase17!
DocArrayInMemorySearch really is an excellent foundation for kickstarting your own LLM applications with @langchain
DocArray retweeted
Explore cross-modal search🔍, by combining the power of @MetaAI's ImageBind model and @docarray!
My latest blog post provides an in-depth guide on implementing a cross-modal search system, with 3 modalities: 🔊 audio, 🖼️image and 📝 text
Link: jina.ai/news/cross-modal-sea…
DocArray retweeted
👏We're excited to announce the release of DocArray v2.
This library for storing and sending multimodal data, especially for #ML, has a unified API for vector databases, Pydantic-style syntax, and more.
Get the full scoop in our latest blog:
hubs.la/Q01PJHCC0
DocArray retweeted
DocArray v0.31.0 is out ! It is the first released after the big DocArray v2 announcement.
My favorite feature is that you can now work with optional nested document field. Making nested machine learning on multi-modal with DocArray even easier.
DocArray v0.31.0 was just released 🎉
New: An in-memory Document Index! ⚡️
Also, 11 bug fixes 😬🐞, 9 documentation improvements 📗, and 2 small breaking changes 💥
github.com/docarray/docarray…
DocArray retweeted
Who would have thought you don't need a fully fledge vector database to store 1000 vectors?
The good thing with DocArray is that you can start simple and switch to a vector database like @qdrant_engine @weaviate_io for production.
👀 integration in @langchain coming
We just released @docarray v0.31.0! 🎉
Biggest new feature is an in-memory Document Index, for exact search over small amounts of data.
This should be very useful LLM long-term memory use cases and for prototyping!
github.com/docarray/docarray…
DocArray retweeted
It's been a fun Pythony day. I decided to implement @SemantleGame (like Wordle, but based on semantics, not letters) in @docarray v2 using @OpenAI GPT embeddings and hints.
github.com/alexcg1/semantle-…
#Python #LLM #ArtificialIntelligence
DocArray retweeted
One super cool thing about @docarray is its unified API for ~all vector databases - we've seen the need for this pop up again and again, for example in @langchain or the ChatGPT retrieval plugins.
Supported at launch:
- @weaviate_io
- @qdrant_engine
- @elastic
- HNSWLib
DocArray retweeted
My first PR got merged in @docarray. Thanks much @atomicflndr!
Looking forward to contribute more.
#opensource
DocArray retweeted
Our team at @JinaAI_ is hiring for a dedicated OSS position!
You would spend most of your time improving @docarray, a library for multi modal ML, and work closely with myself and @samsja19.
If interested, you can reach out to me directly :) 👇
tinyurl.com/y5tvch5d
DocArray retweeted
📢 Project Spotlight! DocArray is a library for nested, unstructured, multimodal data in transit. Learn more here: hubs.la/Q01xNGCR0
#DocArray #DocArrayProject #AI #Data
@LFAIDataFdn
#linuxfoundation
DocArray retweeted
3 - Tensor shape type hint 🚀
Last but not least. We included features inspired from TorchTyping by @PatrickKidger
Our tensor type hint can contain shape information and do validation with @pydantic from @samuelcolvin !
Working with Pytorch and Numpy
DocArray retweeted
Super proud of this new alpha release of DocArray v2 with a focus on the Mutli-Modal ML side 🚀.
My favorite features in the thread 👇
Today we are releasing a new alpha version of DocArray v2! 🥳
Get it here: github.com/docarray/docarray…
Or try it out in colab: colab.research.google.com/dr…
What's new:
1. Enabling the use of DocArray inside of ML models
2. Closing the gap between model training and deployment
🧵👇
DocArray retweeted
Incredibly excited to work on this!
My personal highlights from our roadmap:
- "stacked tensor mode" for DocArray, making it suitable for use inside of ML models, not just as an intermediary data structure
- Disentangling DocumentArray from Document Store
We're happy to announce the first alpha version of a rewrite that we are kicking off - DocArray v2! 🎉
With this we are fully embracing a multi-modal interface, so that DocArray adapts to your data, instead of the other way around.
See the announcement: github.com/docarray/notes/bl…