@KyleHessling1 for some reason, I am getting in lamma.cpp much better performance on my 5090 with Qwopus3.6-35B-A3B-v1-Q6_K - non MTP than with MTP
Any ideas what am I doing wrong?
@antigravity 90% of the time with GOOGLE AI ULTRA we get:
Our servers are experiencing high traffic right now, please try again in a minute.
and
Agent terminated due to error
You can prompt the model to try again or start a new conversation if the error persists.
Soon we may NOT renew!!