TensorChord builds tools for data scientists.

TensorChord

TensorChord | 176 followers on LinkedIn. Build tools for data scientists. Check out our product modelz.ai | At TensorChord, we have created impactful projects that have gained recognition and made a significant impact in the AI/ML community. Some of our notable projects include:

🔹 GitHub - tensorchord/openmodelz: Our one-click machine learning deployment platform, empowering organizations to effortlessly scale their AI models on any cluster, cloud platform, or even a single machine.
🔹 GitHub - tensorchord/pgvecto.rs: A powerful vector database plugin for Postgres, written in Rust, designed specifically for AI/ML applications.
🔹 GitHub - tensorchord/envd: A reproducible development environment for AI/ML, ensuring seamless collaboration and accelerating the development process.

https://modelz.ai is our commercial product that allows you to deploy your machine learning models effortlessly on a serverless infrastructure.

Building vector search in Postgres: pgvecto.rs

Ce Gao

CTO@TensorChord, Building pgvecto.rs

Jinjing Zhou

CS@XJTU

Mingzhuo Yin

TensorChord - Senior Software Engineer

Yuandong Xie

We have recently built and launched https://lnkd.in/gn7Sqn3d on GitHub, a tool aimed at evaluating the impact of reranking on search quality. By leveraging reranking techniques, we have observed remarkable improvements in the top-k candidates' quality. To visualize the results, we have integrated qtext with textual, a lightweight Python application framework. Whether you prefer dense, sparse, or text search, each method provides distinct insights into the query. We encourage you to experiment with different queries to identify the retrieval approach that best aligns with your specific requirements.

After the Japanese RAG, here’s my quick implementation using Mixtral 8x7B ( Mistral AI ) for the financial reports of Porsche AG in German, Hermès in French, and Brunello Cucinelli in Italian with great results. For this vanilla RAG, the main packages are LlamaIndex , BGM-3 of BAAI, pgvecto.rs of TensorChord , Transformers of Hugging Face , LlamaCPP, Mixtral-8X7B (quantized version), ngrok , and last but not least, Postgres. This open-source RAG enables independent developers or SMEs to leverage powerful MoE-based LLM architecture for European languages at a low cost all while respecting data governance. Please feel free to discover my latest notebook. -> https://lnkd.in/dn3xmVky

Akim Mousterou

We are seeking endorsement for the Conviction Embed Program. If you appreciate ai-ifnra.fun and pgvecto.rs , we kindly request your support. https://lnkd.in/gjgkiTtT

I created the AI Infrastructure Landscape https://ai-infra.fun to clarify the crowded space of Generative AI infrastructure. It categorizes tools, frameworks, platforms, and libraries, helping me and others understand the landscape better. It includes popular frameworks like LangChain , Dify , Flyte , along with platforms like Lepton AI , Modal Labs , Replicate , and Anyscale . Databases like The Milvus Project , Weaviate , Qdrant , and TensorChord pgvecto.rs are also featured. The landscape empowers researchers, developers, and AI enthusiasts like me to make informed decisions for generative AI projects. Check out https://ai-infra.fun for more products! The landscape also encourages contributions, so if you have something to add, your input is more than welcome: https://lnkd.in/g79d2_rx

I'm delighted to share that pgvecto.rs 0.2 has been released! 🎉 Experience the potential of pgvecto.rs 0.2, where vector search and relational queries come together for optimized efficiency. #Postgres #VectorSearch https://lnkd.in/gKvtwswd

📣 Exciting News! 🚀 We wanted to share with you a website for AI infrastructure usage statistics. If you're interested in AI infra products, you have to check it out! 📈 The website https://lnkd.in/gUZu8FtB provides insights into the latest trends. Two products, Chroma , #pgvector , and #vLLM , have been getting a lot of downloads lately. 🌟 Please let us know if you have any new AI infrastructure products that you'd like to include in the report. #AIInfrastructure

TENS

AI:TENSORCHORD

TensorChord