EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[bugfix]Fix no attribute 'data' when MLAPO is enable (#6601)

Browse Source 
### What this PR does / why we need it?

This PR fixes an `AttributeError: 'Parameter' object has no attribute
'data'` that occurs when MLAPO is enabled with vLLM v0.15.0.

The error is caused by a monkey-patch on
`MLAAttention.process_weights_after_loading` which is incompatible with
changes in vLLM v0.15.0. This is likely related to PyTorch's deprecation
of the `.data` attribute on `torch.nn.Parameter` objects.

This change makes the monkey-patch conditional, so it is not applied for
vLLM v0.15.0 and newer versions, resolving the crash.


- vLLM version: v0.15.0
- vLLM main:
d7e17aaacd

Signed-off-by: Meihan-chen <jcccx.cmh@gmail.com>
This commit is contained in:
meihanc
2026-02-10 09:04:32 +08:00
committed by GitHub
parent 905f0764e0
commit 5b8e47cb68
1 changed files with 8 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
vllm_ascend/ops/mla.py
View File

@@ -126,16 +126,17 @@ class AscendMultiHeadLatentAttention(MultiHeadLatentAttentionWrapper):
o_proj=mla_modules.o_proj,
)
original_process_weights = self.mla_attn.process_weights_after_loading
if not vllm_version_is("v0.15.0"):
original_process_weights = self.mla_attn.process_weights_after_loading
def wrapped_process_weights(act_dtype: torch.dtype):
from vllm_ascend.attention.sfa_v1 import AscendSFAImpl
def wrapped_process_weights(act_dtype: torch.dtype):
from vllm_ascend.attention.sfa_v1 import AscendSFAImpl
if not isinstance(self.mla_attn.impl, AscendSFAImpl):
original_process_weights(act_dtype)
self.mla_attn.impl.process_weights_after_loading(act_dtype)
if not isinstance(self.mla_attn.impl, AscendSFAImpl):
original_process_weights(act_dtype)
self.mla_attn.impl.process_weights_after_loading(act_dtype)
self.mla_attn.process_weights_after_loading = wrapped_process_weights
self.mla_attn.process_weights_after_loading = wrapped_process_weights
compilation_config = get_current_vllm_config().compilation_config
if prefix in compilation_config.static_forward_context: