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

[MISC] Clean up useless env USE_OPTIMIZED_MODEL (#6618)

Browse Source 
Clean up uesless env `USE_OPTIMIZED_MODEL`

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

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2026-02-09 15:38:58 +08:00
committed by GitHub
parent b7aa511daa
commit 9c6d031797
2 changed files with 2 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
vllm_ascend/ops/mm_encoder_attention.py
View File

@@ -21,8 +21,6 @@ import torch.nn.functional as F
import torch_npu
from vllm.model_executor.layers.attention.mm_encoder_attention import MMEncoderAttention # type: ignore
import vllm_ascend.envs as envs_ascend
MIN_PAD_SIZE = 64 # min_size to pad weight
MAX_PAD_SIZE = 128 # max_size to pad weight
@@ -93,7 +91,7 @@ class AscendMMEncoderAttention(MMEncoderAttention):
# q, k, v: [b, s, head, head_dim] -> [b * s, head, head_dim]
q, k, v = self.reshape_qkv_to_3d(query, key, value, bsz, q_len, kv_len)
enable_pad = envs_ascend.USE_OPTIMIZED_MODEL and self.head_size > MIN_PAD_SIZE and self.head_size < MAX_PAD_SIZE
enable_pad = self.head_size > MIN_PAD_SIZE and self.head_size < MAX_PAD_SIZE
if enable_pad:
origin_shape = q.shape[-1]