langchain-ai/langchain v0.0.167 on GitHub

What's Changed

Fix a typo in langchain/docs/modules/models/llms/integrations/anyscale.ipynb by @kylehh in #4526
Make BaseStringMessagePromptTemplate.from_template return type generic by @jonasalexander in #4523
fix typos in the prompts of LLMSummarizationCheckerChain by @yule-BUAA in #4518
Add on_chat_message_start by @vowelparrot in #4499
[pyproject.toml] add tiktoken when install langchain[openai] by @ofey404 in #4514
add invocation params as extra params in llm callbacks by @agola11 in #4506
Add Invocation Params by @vowelparrot in #4509
Optimize the initialization method of GPTCache by @SimFG in #4522
Add prestodb prompt by @yanghua in #4516
Enhance the prompt to make the LLM generate right date for real today by @yanghua in #4505
added arxiv retriever by @leo-gan in #4538
Dereference Messages by @vowelparrot in #4557
feat: Added class to support huggingface text generation inference server by @gsaivinay in #4447
add constitutional principles from paper by @hwchase17 in #4554
Add pytest --only-extended and --only-core options by @eyurtsev in #4494
update init.py for Anyscale by @kylehh in #4585
Improve error messages formatting in doc loaders by @andreakiro in #4586
added documentation on retrieving a PG vectorstore by @eRuaro in #4578
Provide get current date function dialect for other DBs by @yanghua in #4576
Improve vespa interface by @dev2049 in #4546
Release 167 by @dev2049 in #4589

Full Changelog: v0.0.166...v0.0.167