This commit is contained in:
wailovet
2024-08-13 00:11:28 +08:00
parent 56910da219
commit d812eb37ba
+1 -1
View File
@@ -30,7 +30,7 @@ def quantize_loader(model, state_dict, bits=4, device='cuda', includes=[]):
for name, module in model.named_modules():
if name in qkey:
print(f"Quantizing {name}")
# print(f"Quantizing {name}")
module = module.to(dtype=torch.float16)