🦜🚀LangChain v0.1.0 After a year of development, we've released LangChain v0.1.0 Read the full blog here: blog.langchain.dev/langchain… After talking with the users and developers, we released this stable version and focused the library on a few core areas: 👀Observability: Building complex LLM applications is hard. In order to best debug, you need to know the exact steps that were taken and the input/output at each step. Through a tight integration with LangSmith, LangChain has best-in-class observability ↔️Integrations: With nearly 700 integrations, no matter what tech stack you want to use, LangChain supports it 🔗Composability: With LangChain Expression Language, it's easy (and fun!) to create arbitrary chains, bringing you all the benefits of a data orchestration framework 🎏Streaming: We've invested a lot in making sure that streaming is supported in a first class way for all chains created with LangChain Expression Language - including streaming of intermediate steps 🧱Output parsing: Getting the LLM to return information in a certain format is key for enabling it to take actions. 🔎Retrieval: adding advanced yet production ready methods for RAG, including text-splitting, retrieval, and an indexing pipeline 🤖Tool Use + Agents: collection of agents (decide what actions to take), collection of tools, easy way to define tools We will be highlighting these individually over the coming week. Each one has a lot of nuance to it, and is worth it's own deep dive Big thank you to our community for helping guide and build this with us!