rixty/KoboldAI-Client
1
1
Fork 0
mirror of https://github.com/KoboldAI/KoboldAI-Client.git synced 2025-06-05 21:59:24 +02:00
Code Issues Projects Releases Wiki Activity

Model: Fix crash on full GPU load

Browse Source
This commit is contained in:
somebody
2023-02-25 18:19:46 -06:00
parent 465e22fa5c
commit d53d2bcc45
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
model.py
View File

@@ -2274,7 +2274,7 @@ class GenericHFTorchInferenceModel(HFTorchInferenceModel):
if utils.koboldai_vars.hascuda:
if utils.koboldai_vars.usegpu:
# Use just VRAM
model = model.half().to(utils.koboldai_vars.gpu_device)
self.model = self.model.half().to(utils.koboldai_vars.gpu_device)
elif utils.koboldai_vars.breakmodel:
# Use both RAM and VRAM (breakmodel)
if not utils.koboldai_vars.lazy_load: