[Spec Decode] Support hybrid attention models in extract_hidden_states (#39949)
Signed-off-by: mgoin <mgoin64@gmail.com> Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 4.6
parent
5794c65f8c
commit
2f821faeae
@@ -396,6 +396,13 @@ class MLAAttentionSpec(FullAttentionSpec):
|
||||
)
|
||||
|
||||
|
||||
@dataclass(frozen=True, kw_only=True)
|
||||
class HiddenStateCacheSpec(MLAAttentionSpec):
|
||||
"""Marker for hidden-state cache layers used by extract_hidden_states."""
|
||||
|
||||
pass
|
||||
|
||||
|
||||
@dataclass(frozen=True, kw_only=True)
|
||||
class ChunkedLocalAttentionSpec(AttentionSpec):
|
||||
attention_chunk_size: int
|
||||
|
||||
Reference in New Issue
Block a user