Nuno Campos

icon company @langchain-ai icon location San Francisco icon location Helping people build stuff.

Results 249 comments of Nuno Campos

Add the `name` and `arguments` from `function_call` to `on_llm_new_token` when `streaming=True`

This is now available as the `chunk` arg to on_llm_new_token

RFC: Embedding -> EmbeddingModel

This is great!

Fix incorrect value of outputKeys on AnalyzeDocumentsChain

@vowelparrot I only caught this because I saw a similar mistake in JS while doing this https://github.com/hwchase17/langchainjs/pull/843

Fix incorrect value of outputKeys on AnalyzeDocumentsChain

If it's anything like the JS package I think this issue will be present in more chains

Accept message-like things in Chat models, LLMs and MessagesPlaceholder

That's message prompt templates, this is messages

JSONDecodeError: Invalid control character at: line 2 column 631 (char 632) when using langchain new agent

@glejdis hi, which llm are you using?

Add asyncio support for LLM (OpenAI), Chain (LLMChain, LLMMathChain), and Agent

Making the async methods live in the same class makes reusing helper methods across both implementations easier

Make logger instance local to each instance that uses it

@hwchase17 This is a first step towards being able to eg. run multiple independent agents in the same python process logging to different places.

Make logger instance local to each instance that uses it

Closing this

Add `verbose` option to `ConversationalRetrievalQAChain` and question answering chains

Thanks!