Show HN: SynapseKit – Async-native Python framework for LLM pipelines and agents (opens in new tab)

(github.com)

4 pointsaminau25d ago2 comments

2 comments

aminauOP25d ago

https://www.reddit.com/r/LangChain/comments/1sm3g2q/i_tested...

Async is definitely the right approach for this, but how are you handling the event loop blocking if a local model takes way too long to generate a heavy sequence?

1 more reply

j / k navigate · click thread line to collapse

2 comments

aminauOP25d ago

https://www.reddit.com/r/LangChain/comments/1sm3g2q/i_tested...

grimmai14325d ago

Async is definitely the right approach for this, but how are you handling the event loop blocking if a local model takes way too long to generate a heavy sequence?

1 more reply

j / k navigate · click thread line to collapse