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
OpenDAS
Megatron-LM
Commits
057a1fab
"vscode:/vscode.git/clone" did not exist on "e05f03ae41540123a99afdacff86d26170c1315d"
Commit
057a1fab
authored
Oct 01, 2020
by
Mostofa Patwary
Browse files
Fixed uninit communication gruop
parent
10c828ed
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
1 addition
and
1 deletion
+1
-1
megatron/mpu/mappings.py
megatron/mpu/mappings.py
+1
-1
No files found.
megatron/mpu/mappings.py
View file @
057a1fab
...
@@ -65,7 +65,7 @@ def _gather(input_):
...
@@ -65,7 +65,7 @@ def _gather(input_):
tensor_list
=
[
torch
.
empty_like
(
input_
)
for
_
in
range
(
world_size
)]
tensor_list
=
[
torch
.
empty_like
(
input_
)
for
_
in
range
(
world_size
)]
tensor_list
[
rank
]
=
input_
tensor_list
[
rank
]
=
input_
torch
.
distributed
.
all_gather
(
tensor_list
,
input_
,
group
=
group
)
torch
.
distributed
.
all_gather
(
tensor_list
,
input_
,
group
=
get_model_parallel_
group
()
)
# Note: torch.cat already creates a contiguous tensor.
# Note: torch.cat already creates a contiguous tensor.
output
=
torch
.
cat
(
tensor_list
,
dim
=
last_dim
).
contiguous
()
output
=
torch
.
cat
(
tensor_list
,
dim
=
last_dim
).
contiguous
()
...
...
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