
LANGCHAIN — New in LangSmith
Social media is not about the exploitation of technology but service to community. — Simon Mainwaring
In the latest LangChain release notes, there are several exciting updates and enhancements in LangSmith and the open-source LangChain Experimental. Let’s dive into the new features and improvements, and also explore some use cases that demonstrate the power of LangChain.
New in LangSmith
Streaming in the Playground
The playground now supports streaming! You can enter the playground by clicking on the “Open in Playground” button on the upper right-hand corner of select runs.
Infrastructure Changes
The backend infrastructure has been strengthened to support more users, aiming to accommodate those on the waitlist soon.
UI and Navigation Enhancements
Significant quality-of-life improvements have been made to the UI, especially around the run details page. The trace tree is now positioned on the left-hand side.
New in Open Source
LangChain Experimental
Selected chains have been moved into a separate package to make LangChain leaner, more focused, and safer. However, this change may induce some backwards incompatible modifications.
Docs Restructures
Numerous improvements have been made to the documentation. The skeleton has been ported over from Python JS, ensuring consistency. Additionally, the docs have been restructured to better distinguish between modules and use cases.
Web Retriever
A significant effort has been made to create a web research assistant, resulting in an interesting and novel retriever. The source code for this can be found here.
Llama 2 Support
Support and examples for Llama 2, including some Llama 2 specific prompts, have been quickly added.
In Case You Missed It
The LangChain team has collaborated with various entities on some insightful content. These collaborations include a webinar, diagnosing and fixing slow chatbots, transforming AI into a savvy CFO, and more.
Coming Soon
The LangChain team is working on bringing support for teams and organizations to share work in LangSmith, increasing the composability of chains, and enhancing the use case sections in the documentation.
Use-Cases We Love
Several compelling use-cases have been highlighted, such as CLAIRe, a conversational learning AI with recall, Q&A over data for interacting with Snowflake database using natural language queries, a personal financial advisor and copilot, and an AI companion.
With this latest release of LangChain, there are numerous enhancements and new features that can help developers and organizations to build powerful and efficient AI solutions. Whether it’s through improvements in the playground, infrastructure changes, or the introduction of new open-source functionalities, LangChain continues to evolve and provide valuable tools for the community. If you’re interested in exploring more updates and engaging with the LangChain community, consider subscribing to the newsletter for regular updates.
