Resolved ALIBI bias regression due to porting flat PA #1323
Annotations
5 errors
|
Mypy:
vllm/attention/layer.py#L99
Too many arguments for "AttentionImpl" [call-arg]
|
Mypy:
vllm/attention/backends/hpu_attn.py#L279
Value of type "Any | None" is not indexable [index]
|
Mypy:
vllm/attention/backends/hpu_attn.py#L291
Item "None" of "Any | None" has no attribute "unsqueeze" [union-attr]
|
Mypy
Process completed with exit code 1.
|
Loading