langchain-ai / langchain-google

MIT License
74 stars 78 forks source link

require that async client be initialized inside a running event loop #244

Closed vbarda closed 1 month ago

vbarda commented 1 month ago

This change fixes the issue with running ChatGoogleGenerativeAI inside a ThreadPoolExecutor and requires initializing the model inside an event loop to enable async client

vbarda commented 1 month ago

is there anything we can do to make this usable in a setting without an eventloop? Or does it need to be fixed in the upstream sdk?

I believe it would have to be fixed in the SDK

cc @lkuligin