Disclaimer: not a programmer, just a user with a few notes
GPT-4 was boasting an incredible 25k words context, but it seems to me that this plugin has a way lower hard cap, why is that? Was it optimized for GPT-3 and is it waiting on an update for the 4?
Rolling context is a must, at least. At the time, I'm forced to request a summary of the convo or move the last part of it to a new note.
Is it really necessary to submit the whole convo every time? Does the API really work that way? I think it's extremely inefficient and costly, but not sure there's a workaround for this
Disclaimer: not a programmer, just a user with a few notes
GPT-4 was boasting an incredible 25k words context, but it seems to me that this plugin has a way lower hard cap, why is that? Was it optimized for GPT-3 and is it waiting on an update for the 4?
Rolling context is a must, at least. At the time, I'm forced to request a summary of the convo or move the last part of it to a new note.
Is it really necessary to submit the whole convo every time? Does the API really work that way? I think it's extremely inefficient and costly, but not sure there's a workaround for this