LangChain

✨ Today, we’re thrilled to announce ✨ - The general availability of LangSmith (no more waitlist!) - Our Series A fundraise led by Sequoia Capital - Our beautiful new homepage and brand We've worked hard over the past few months to add requested features and ensure LangSmith can operate at scale. We’re now confident in saying that it is the most complete platform for building production-grade LLM applications, whether or not you’re using LangChain. Learn more here: https://lnkd.in/gZxW8X_V and sign up here: https://lnkd.in/dwXZt_ZT Our series A round will give us the capital needed to grow our open source and platform offerings. Working with Sonya Huang, Romie Boyd, and the rest of the Sequoia team has been a privilege so far! https://lnkd.in/g8nw36_Z Finally, we’re excited to unveil our new homepage and brand. Dive into our new website at https://www.langchain.com/ to see the changes for yourself, explore the expanded resources, and discover what LangChain, LangSmith, and LangServe have to offer. PS — we’re hiring! Explore our careers page and reach out if you think you’re a fit for any of our open positions! https://lnkd.in/g9rXjrvC

We're hosting the Berkeley LLM Meetup at our offices in San Francisco on Tuesday, June 18th 🥳 Come learn about LangGraph -- our framework for building stateful, multi-actor applications with LLMs. You'll also learn more about exciting OSS projects being built at UC Berkeley! Sign up here ➡️ https://lnkd.in/gUSrurAm

LangChain Release Notes (Week of May 27) Check out our release notes to rewind the latest Lang news, including: 📄 Our new, versioned docs for LangChain v0.2 🤯 Contest with NVIDIA (note: GPU prizes at stake!) 🏙️ In-person meetups in NYC and SF in June ✂️ Dataset splits & repetitions for improved evaluation 🔧 Off-the-shelf eval prompts to catch bad retrieval and RAG hallucinations Read all about it here 👉 https://lnkd.in/g937RUpx

In SF on June 12th?🌉 Join our very own Harrison Chase (LangChain co-founder & CEO) and Edo Liberty (Pinecone founder & CEO) for a panel discussion on how to meet the near-term demand for #GenAI results while planning for sustained growth and innovation. The discussion will take place at Minna Gallery in San Francisco. Sign up here for June 12th: https://lnkd.in/gCVBrJ4c

🔁 Repetitions in LangSmith 🔁 You can now run multiple repetitions of your experiment in LangSmith. This helps smooth out noise from variability introduced by your application or from your LLM-as-a-judge evaluator, so you can build confidence in the results of your experiment. In the video below, learn how to evaluate on a dataset using repetitions. You can check the mean score across N repetitions, and also compare the outputs for variability across repetitions. 📽️ Video: https://lnkd.in/gFRGPAv5 📓 Evaluate on a dataset with repetitions: https://lnkd.in/gP-Mt_Hh 🌟 Try it out in LangSmith: smith.langchain.com

Excited to be part of the Milvus Lite launch! Together, we're simplifying the creation of powerful GenAI apps. Discover how our combined capabilities can elevate your AI projects. ✍️ Read the blog from Milvus: https://lnkd.in/gm7sqqwm 📓 Check out the docs: https://lnkd.in/g3fvCzBH

📊Dataset improvements in LangSmith📊 In addition to splitting a dataset, you can also now do the following for dataset examples in LangSmith: • Clone examples to another dataset • Edit metadata directly in the UI • Search for specific examples These actions can help you speed up in finding relevant information within your desired dataset. 📓 Editing example metadata (in UI): https://lnkd.in/gzjFDhir 📓 Filtering examples: https://lnkd.in/gFhxikEU 🌟 Try it out in LangSmith: smith.langchain.com

🌀Self-correcting code assistants with Codestral 🌀 Mistral AI just released Codestral-22B, a top-performing open-weights code generation model trained on 80+ programming languages with diverse capabilities (e.g., instructions, fill-in-the-middle) and tool use. Of course, we had to test it out! In the video below, we show how to build a self-corrective coding assistant using Codestral with LangGraph. Using ideas borrowed from the AlphaCodium paper, we show how to use Codestral with structured output / tool use, unit testing in-the-loop, and error feedback to self-correct from mistakes. 📽️ Watch the video: https://lnkd.in/gPpCwF5d 📓 Code added to our Mistral AI recipes: https://lnkd.in/g9ByxuVh 🗣️ Codestral blog post: https://lnkd.in/gB-ahVWr

Semantic search is great for extracting meaning, and Qdrant’s filters make it even better for specific searches! Here's how to effectively narrow search results and dynamically generate and execute queries in real time with the innovative self-querying retriever in LangChain. In this tutorial, Daniel Romero shows you: ▶️ When to use semantic search or directed lookups. ▶️ Detailed walkthrough on handling a large dataset of 130k wine reviews in Qdrant. ▶️ How to perform semantic searches and apply precise filters based on metadata like country, price, and rating to refine results. ▶️ The new concept of a self-querying retriever in LangChain that combines natural language queries with structured data filtering. 🎥 Watch the full video here! 👇 https://buff.ly/4bQSIXK Learn more about Qdrant filtering: https://buff.ly/3V0I3mi

✂️ Dataset Splits ✂️ With dataset splits, it's now easier to run evaluations on a subset of your dataset in LangSmith. You can tag examples with different split names, edit and add to splits, and filter on your desired criteria. Splits can come in handy when have a dataset with multiple categories that you'd like to evaluate separately. Split datasets also allow you to test new use cases that you may want to include in a dataset and evaluate in the future — by first adding examples to a separate set to test, while preserving your evaluation workflow. 📽️ See a video example: https://lnkd.in/gqb4znMj 📓 Create/manage dataset splits: https://lnkd.in/gQBWZGgZ 📓 Evaluate on a dataset split: https://lnkd.in/g7WcFWCU 🌟 Try it out in LangSmith: https://lnkd.in/gbbj5tf5