Skip to content

[Bugfix] Update attention interface in Whisper #11784

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jan 7, 2025
Merged

[Bugfix] Update attention interface in Whisper #11784

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 11 additions & 13 deletions vllm/model_executor/models/whisper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -106,6 +106,7 @@ def __init__(
cache_config=cache_config,
quant_config=quant_config,
prefix=f"{prefix}.attn",
attn_type=self.attn_type,
)

def _init_qkv(
Expand Down Expand Up @@ -134,12 +135,7 @@ def forward(
qkv, _ = self.qkv_proj(hidden_states)
q, k, v = qkv.split([self.q_size, self.kv_size, self.kv_size], dim=-1)

attn_output = self.attn(q,
k,
v,
kv_cache,
attn_metadata,
attn_type=self.attn_type)
attn_output = self.attn(q, k, v, kv_cache, attn_metadata)

output, _ = self.out_proj(attn_output)

Expand All @@ -164,6 +160,7 @@ def __init__(
cache_config=cache_config,
quant_config=quant_config,
prefix=prefix,
attn_type=AttentionType.ENCODER_DECODER,
)

def _init_qkv(
Expand Down Expand Up @@ -207,12 +204,13 @@ def forward(
else:
k = v = None

attn_output = self.attn(q,
k,
v,
kv_cache,
attn_metadata,
attn_type=AttentionType.ENCODER_DECODER)
attn_output = self.attn(
q,
k,
v,
kv_cache,
attn_metadata,
)

output, _ = self.out_proj(attn_output)

Expand Down Expand Up @@ -734,4 +732,4 @@ def load_weights(self, weights: Iterable[Tuple[str,
loaded_weights = [(name, loaded_weight)
for name, loaded_weight in weights]
mapper = WeightsMapper({".fc1.": ".mlp.fc1.", ".fc2.": ".mlp.fc2."})
return loader.load_weights(loaded_weights, mapper=mapper)
return loader.load_weights(loaded_weights, mapper=mapper)
Loading