Fantastic tq I feel stupid because I’ve seen that for a while now. Do you know more about the autocompaction process? I could use some help there too with ghcp.

Fantastic tq I feel stupid because I’ve seen that for a while now. Do you know more about the autocompaction process? I could use some help there too with ghcp. I don’t see my context going up while abusing smaller context models with long inputs nor do I see a compaction indicator after turns.