#Text stops generating in the middle of a response
1 messages · Page 1 of 1 (latest)
so this happens because the LLM will limit the response in proportion to the amount of prompt tokens. the only way around this is to make the context window "smaller"
could enabling "message summarization" help in that situation?
since
it only happens in longer conversations
yeah it does
How do I enable message summarization? Is it in the .env?