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 16, 17:00 UTC
[Bugfix][Model] Fix Python 3.8 compatibility in Pixtral model by updating type annotations (#8490)
Chris committed
Sep 15, 2024 at 06:20 UTC
3724d5f6b59d9859e5b47c047535bb8edc124eab
1 file changed
+1
-1
vllm/model_executor/models/pixtral.py
+1
-1
@@ -454,7 +454,7 @@ class Transformer(nn.Module):
454
return x
455
456
457
-def position_meshgrid(patch_embeds_list: list[torch.Tensor], ) -> torch.Tensor:
457
+def position_meshgrid(patch_embeds_list: List[torch.Tensor], ) -> torch.Tensor:
458
positions = torch.cat([
459
torch.stack(
460
torch.meshgrid(