frozenleaves/vllm
1
0
Fork 0
mirror of https://github.com/vllm-project/vllm.git synced 2025-10-20 14:53:52 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix] Fix k_proj's bias for whisper self attention (#12342)

Browse Source 
Signed-off-by: Isotr0py <2037008807@qq.com>
This commit is contained in:
Isotr0py
2025-01-23 18:15:33 +08:00
committed by GitHub
parent 8ae5ff2009
commit c5b4b11d7f
1 changed files with 18 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
vllm/model_executor/models/whisper.py
View File

@ -729,7 +729,22 @@ class WhisperForConditionalGeneration(nn.Module, SupportsMultiModal):
def load_weights(self, weights: Iterable[Tuple[str,
torch.Tensor]]) -> Set[str]:
loader = AutoWeightsLoader(self, skip_prefixes=["proj_out."])
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)
# add fake zeros bias for k_proj to state_dict
weights = _create_fake_bias_for_k_proj(weights)
return loader.load_weights(weights, mapper=mapper)
def _create_fake_bias_for_k_proj(
weights: Iterable[Tuple[str, torch.Tensor]]
) -> Iterable[Tuple[str, torch.Tensor]]:
"""
Create full zeros bias for k_proj weight in self-attention layers.
So that the bias for k_proj in qkv_proj can be initialized with zeros.
"""
for name, weight in weights:
if ".self_attn.k_proj.weight" in name:
bias = torch.zeros(weight.size(0))
bias_name = name.replace("weight", "bias")
yield from [(name, weight), (bias_name, bias)]
yield name, weight