Patrick Rho
Patrick Rho
We need the endpoint like how Perplexity provides oai compatible endpoint.
just create an alias then... there's a reason Claude made the flag dangerously long
related and dup #1555
I notice my cc doing this today as well, are you on Opus?
Thanks! By the way, the doc says "http://localhost/sse:9121" but it should really be http://localhost:9121/sse :)
Can we get Lora rank > 64 supported and merged? edit: im also curious if this was by design to support up to 64 rank, if so please let me...
I think they tried improving the streaming support and jammed everything into memory and causes overflow - i downgraded to 1.0.9 and disabled auto update using the env variable. I...
works for me, the ctx per read i think has been reduced to 25000 but works better