Fix bug: default value of weights_only changed in PyTorch 2.6
This commit is contained in:
+1
-1
@@ -61,7 +61,7 @@ class fsModel(BaseModel):
|
||||
|
||||
# Id network
|
||||
netArc_checkpoint = opt.Arc_path
|
||||
netArc_checkpoint = torch.load(netArc_checkpoint, map_location=torch.device("cpu"))
|
||||
netArc_checkpoint = torch.load(netArc_checkpoint, map_location=torch.device("cpu"), weights_only=False)
|
||||
self.netArc = netArc_checkpoint
|
||||
self.netArc = self.netArc.to(device)
|
||||
self.netArc.eval()
|
||||
|
||||
Reference in New Issue
Block a user