A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 08:30 UTC
Fix an error in dummy weight loading for quantization models (#18855)
Signed-off-by: Chenyaaang <chenyangli@google.com>
Chenyaaang committed
May 29, 2025 at 03:07 UTC
f8977c233f453313c124dc0753dfc1669ef401f0
1 file changed
+1
-1
vllm/model_executor/model_loader/weight_utils.py
+1
-1
@@ -696,7 +696,7 @@ def initialize_dummy_weights(
696
# Note: We avoid using torch.rank_like as it doesn't currently
697
# support the generator argument.
698
param.copy_((high - low) *
699
- torch.rand(*param.shape,
699
+ torch.rand(param.shape,
700
generator=generator,
701
dtype=param.dtype,
702
layout=param.layout,