Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
norm
vllm
Commits
4bd18ec0
Unverified
Commit
4bd18ec0
authored
Feb 26, 2024
by
Woosuk Kwon
Committed by
GitHub
Feb 26, 2024
Browse files
[Minor] Fix type annotation in fused moe (#3045)
parent
2410e320
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
4 additions
and
3 deletions
+4
-3
vllm/model_executor/layers/fused_moe/fused_moe.py
vllm/model_executor/layers/fused_moe/fused_moe.py
+4
-3
No files found.
vllm/model_executor/layers/fused_moe/fused_moe.py
View file @
4bd18ec0
...
...
@@ -2,7 +2,7 @@
import
functools
import
json
import
os
from
typing
import
Any
,
Dict
,
Optional
from
typing
import
Any
,
Dict
,
Optional
,
Tuple
import
torch
import
triton
...
...
@@ -137,7 +137,7 @@ def fused_moe_kernel(
def
moe_align_block_size
(
topk_ids
:
torch
.
Tensor
,
block_size
:
int
,
num_experts
:
int
)
->
(
torch
.
Tensor
,
torch
.
Tensor
,
torch
.
Tensor
)
:
num_experts
:
int
)
->
Tuple
[
torch
.
Tensor
,
torch
.
Tensor
,
torch
.
Tensor
]
:
"""
Aligns the token distribution across experts to be compatible with block size for matrix multiplication.
...
...
@@ -185,7 +185,8 @@ def invoke_fused_moe_kernel(A: torch.Tensor, B: torch.Tensor, C: torch.Tensor,
sorted_token_ids
:
torch
.
Tensor
,
expert_ids
:
torch
.
Tensor
,
num_tokens_post_padded
:
torch
.
Tensor
,
mul_routed_weight
:
bool
,
top_k
:
int
,
config
:
dict
):
mul_routed_weight
:
bool
,
top_k
:
int
,
config
:
Dict
[
str
,
Any
])
->
None
:
assert
topk_weights
.
stride
(
1
)
==
1
assert
sorted_token_ids
.
stride
(
0
)
==
1
...
...
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