Open source function calling in `langchain-fireworks` integration Try out @FireworksAI_HQ
's new FireFunction-v1 model in our integration package, which also offers connections to Fireworks' many other features: HIPAA and SOC2 compliance, all-model JSON mode, and a first-party
@langchain
-
FireFunction-v1 Open Source Function Calling in LangChain Integration
By
–
-
LangSmith GA: LangGraph Examples and RAG Resources
By
–
LangChain Release Notes, week of 2/19 LangSmith Generally Available!!
Lots of cool advanced LangGraph examples
Some great retrieval and RAG resources
Awesome community created content Read all here: https://
blog.langchain.dev/week-of-2-19-l
angchain-release-notes/
… -
Monte-Carlo Tree Search: UCT Algorithm for Decision-Making
By
–
4/ Monte-Carlo Tree What? Monte-Carlo Tree Search (MCTS) is a heuristic search algorithm used in decision-making processes. In LATS, MCTS selects the best candidate using an Upper Confidence Bound for Trees (UCT).
This balances exploration and exploitation of a policy. -
Language Agent Tree Search: Self-Optimizing AI Algorithm
By
–
1/ Language Agent Tree Search (LATS) LATS is a unified algorithm that uses self-reflection (and additional external feedback) to self-optimize and efficiently adapt to new tasks and environments. It performs better than similar algorithms like Tree of Thoughts, ReAct, and
-
LangGraph LATS Implementation Code and Video Tutorial Released
By
–
2/ Links For those who are eager, you can check out the code and video tutorial at the links below: Youtube: https://
youtube.com/watch?t=625
Python: https://
github.com/langchain-ai/l
anggraph/blob/main/examples/lats/lats.ipynb
… Thanks to @hinthornw for the implementation. -
LATS: Monte-Carlo Tree Search for AI Decision Making
By
–
3/ What is LATS? 1. Select: the best scoring action thus far
2. Expand & act: Generate candidate actions and execute
3. Reflect & evaluate: Observe outcomes and score decisions
4. Backpropagate: update applicable trajectory scores So… Monte-Carlo Tree Search! -
LangSmith Improves Data Filtering for LLM Pipeline Analysis
By
–
Better Filtering Experience in LangSmith If you’re logging a good portion of your production traces to LangSmith, you probably have a lot of data to sift through. Filtering through data in a systematic way is important for understanding how your LLM pipeline is
-
LangSmith Tracing and Evaluation Tools Now Available for TypeScript
By
–
LangSmith Tracing and Evaluation in TypeScript It's now easier than ever to trace non-LangChain TS code and run evaluations in LangSmith! We've added a cookbook here: https://
github.com/langchain-ai/l
angsmith-cookbook/blob/main/typescript-testing-examples/traceable-example/LangSmith_TS_Demo-Traceable.ipynb
… For tracing, we've added – wrapOpenAI(new OpenAI()) allows you to call @OpenAI -
Natural Language to SQL with Local LLM Models
By
–
natural language -> sql, visualizations, analytics impressive stuff. uses a local LLM
-
FireFunction V1: Open-Weights Function Calling Model from Fireworks
By
–
FireFunction V1 – a new, open-weights function calling model from @FireworksAI_HQ Returning structured outputs is (a) hard, (b) incredibly useful Excited to highlight **4** different ways to use this model with LangChain #1: Function Calling with LangChain JS Cookbook: