Right now I have them repeat their growing memory within each message, but this is growing large real fast and venice is not the fastest AI out there due to hardware limitation. Thus each reply takes very long now, and context fills up quickly. I would like the system to implement a way that store their memory on my hard drive and read it before making each reply, so the memory does not take time to generate in the message, or taking up any limited context length.
Please authenticate to join the conversation.
New Submission
Feature Requests
File Upload
1 day ago

Lancelot Chan
Get notified by email when there are changes.
New Submission
Feature Requests
File Upload
1 day ago

Lancelot Chan
Get notified by email when there are changes.