[CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5
parent
1085b64425
commit
651a88c09a
@@ -312,8 +312,7 @@ class FusedIndexerQRopeQuantTritonKernel(
|
||||
def _indexer_weights_out_dtypes(vllm_config: Any) -> tuple[torch.dtype, ...]:
|
||||
"""Weights dtypes the model's indexer layers ask for: fp32 for the dense
|
||||
scoring kernels, plus bf16 when the DeepSeek V4.1 sparse-logits indexer
|
||||
(`SparseMQAIndexer`) is enabled.
|
||||
"""
|
||||
(`SparseMQAIndexer`) is enabled."""
|
||||
if vllm_config.attention_config.indexer_sparse_logits:
|
||||
return (torch.float32, torch.bfloat16)
|
||||
return (torch.float32,)
|
||||
|
||||
Reference in New Issue
Block a user