Daniel
Daniel
@crazywoola sorry) [issue demo.yml.zip](https://github.com/user-attachments/files/16601323/issue.demo.yml.zip)
Sorry for long reply @crazywoola I've tested it on gpt3.5 and the result is same so I don't really think that problem is in open source 7b models.
It works when i change reasoning mode to func call, but it's still strange that it doesn't work for any model in prompt reasoning when memory is turned on.
same with fetch server and logseq server