vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 17:14:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

remove obsolete warning about flash attention on Vulkan (#1634)

Browse source
This commit is contained in:
Wagner Bruna 2025-07-03 05:57:03 -03:00 • committed by GitHub
parent f407aa3b8a
commit bc3e4c1197
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 0 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
koboldcpp.py
View file

@ -6911,8 +6911,6 @@ def kcpp_main_process(launch_args, g_memory=None, gui_launcher=False):
if not args.blasthreads or args.blasthreads <= 0:
args.blasthreads = args.threads
if args.flashattention and (args.usevulkan is not None) and args.gpulayers!=0:
print("\nWARNING: FlashAttention is strongly discouraged when using Vulkan GPU offload as it is extremely slow!\n")
modelname = os.path.abspath(args.model_param)
print(args)