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
d77a3d5b
Commit
d77a3d5b
authored
Jan 08, 2026
by
laibao
Browse files
fix:更新 Marlin W16A16 MoE 中支持的断言消息
parent
fb39e61b
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
1 addition
and
1 deletion
+1
-1
vllm/model_executor/layers/fused_moe/fuse_moe_w16a16_marlin.py
...model_executor/layers/fused_moe/fuse_moe_w16a16_marlin.py
+1
-1
No files found.
vllm/model_executor/layers/fused_moe/fuse_moe_w16a16_marlin.py
View file @
d77a3d5b
...
@@ -240,7 +240,7 @@ def fused_experts_impl_w16a16_marlin(hidden_states: torch.Tensor,
...
@@ -240,7 +240,7 @@ def fused_experts_impl_w16a16_marlin(hidden_states: torch.Tensor,
assert
hidden_states
.
dtype
in
[
torch
.
bfloat16
,
torch
.
float16
]
assert
hidden_states
.
dtype
in
[
torch
.
bfloat16
,
torch
.
float16
]
compute_type
=
hidden_states
.
dtype
compute_type
=
hidden_states
.
dtype
assert
use_lightop
,
(
assert
use_lightop
,
(
"only BW and set
LMSLI
M_USE_LIGHTOP=1 support Marlin W16A16 MoE"
)
"only BW and set
VLL
M_USE_LIGHTOP=1 support Marlin W16A16 MoE"
)
num_tokens
,
K
=
hidden_states
.
shape
num_tokens
,
K
=
hidden_states
.
shape
E
,
twoN
,
K_w1
=
w1
.
shape
E
,
twoN
,
K_w1
=
w1
.
shape
...
...
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