You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, can kontext be loaded quantitatively at present? Because I only have a 4090 with 24g video memory, the current fp16 loading method will cause OOM. Like flux, can it be loaded with torchao or gguf, so that this model can run on 4090?