Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
vllm_cscc
Commits
9aabf7e7
Commit
9aabf7e7
authored
Mar 12, 2026
by
laibao
Browse files
feat(deepseek-mla):: 精简 fused RMS-RoPE concat 可用性判断
parent
cae53e46
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
0 additions
and
9 deletions
+0
-9
vllm/model_executor/layers/attention/mla_attention.py
vllm/model_executor/layers/attention/mla_attention.py
+0
-9
No files found.
vllm/model_executor/layers/attention/mla_attention.py
View file @
9aabf7e7
...
...
@@ -2216,17 +2216,8 @@ class MLACommonImpl(MLACommonBaseImpl[M], Generic[M]):
if
kv_cache_dtype_str
not
in
(
"fp16"
,
"bf16"
):
use_fused_rms_rope_concat
=
False
fused_inputs_provided
=
(
q_ori
is
not
None
or
key_normed
is
not
None
or
positions
is
not
None
or
weight
is
not
None
or
cos_sin_cache
is
not
None
or
epsilon
is
not
None
)
if
(
envs
.
VLLM_USE_LIGHTOP_RMS_ROPE_CONCAT
and
fused_inputs_provided
and
not
use_fused_rms_rope_concat
):
raise
RuntimeError
(
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment