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
DeepEP
Commits
d79b3cd1
Commit
d79b3cd1
authored
Jul 02, 2025
by
Chenggang Zhao
Browse files
Refactor the bench function
parent
85793dda
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
26 additions
and
34 deletions
+26
-34
tests/test_low_latency.py
tests/test_low_latency.py
+4
-7
tests/utils.py
tests/utils.py
+22
-27
No files found.
tests/test_low_latency.py
View file @
d79b3cd1
...
...
@@ -144,18 +144,15 @@ def test_main(num_tokens: int, hidden: int, num_experts: int, num_topk: int,
# Separate profiling
for
return_recv_hook
in
(
False
,
True
):
group
.
barrier
()
bench_outpu
t
=
bench_kineto
(
partial
(
test_func
,
zero_copy
=
True
,
return_recv_hook
=
return_recv_hook
),
dispatch_t
,
combine_
t
=
bench_kineto
(
partial
(
test_func
,
zero_copy
=
True
,
return_recv_hook
=
return_recv_hook
),
kernel_names
=
(
'dispatch'
,
'combine'
),
barrier_comm_profiling
=
True
,
suppress_kineto_output
=
True
,
duplicate_name
_period
=
2
if
return_recv_hook
else
None
)
suppress_kineto_output
=
True
,
num_kernels_per
_period
=
2
if
return_recv_hook
else
1
)
if
not
return_recv_hook
:
dispatch_t
,
combine_t
=
bench_output
print
(
f
'[rank
{
rank
}
] Dispatch bandwidth:
{
num_dispatch_comm_bytes
/
1e9
/
dispatch_t
:.
2
f
}
GB/s, avg_t=
{
dispatch_t
*
1e6
:.
2
f
}
us | '
f
'Combine bandwidth:
{
num_combine_comm_bytes
/
1e9
/
combine_t
:.
2
f
}
GB/s, avg_t=
{
combine_t
*
1e6
:.
2
f
}
us'
,
flush
=
True
)
else
:
dispatch_t
,
combine_t
,
detail_times
=
bench_output
print
(
f
'[rank
{
rank
}
] Dispatch send/recv time:
{
dispatch_t
*
2
*
1e6
:.
2
f
}
=
{
detail_times
[
"dispatch"
][
0
]
*
1e6
:.
2
f
}
+
{
detail_times
[
"dispatch"
][
1
]
*
1e6
:.
2
f
}
us | '
f
'Combine send/recv time:
{
combine_t
*
2
*
1e6
:.
2
f
}
=
{
detail_times
[
"combine"
][
0
]
*
1e6
:.
2
f
}
+
{
detail_times
[
"combine"
][
1
]
*
1e6
:.
2
f
}
us'
,
flush
=
True
)
print
(
f
'[rank
{
rank
}
] Dispatch send/recv time:
{
sum
(
dispatch_t
)
*
2
*
1e6
:.
2
f
}
=
{
dispatch_t
[
0
]
*
1e6
:.
2
f
}
+
{
dispatch_t
[
1
]
*
1e6
:.
2
f
}
us | '
f
'Combine send/recv time:
{
sum
(
combine_t
)
*
2
*
1e6
:.
2
f
}
=
{
combine_t
[
0
]
*
1e6
:.
2
f
}
+
{
combine_t
[
1
]
*
1e6
:.
2
f
}
us'
,
flush
=
True
)
return
hash_value
...
...
tests/utils.py
View file @
d79b3cd1
...
...
@@ -8,7 +8,7 @@ import os
import
sys
import
torch
import
torch.distributed
as
dist
from
typing
import
Optional
from
typing
import
Optional
,
Union
def
init_dist
(
local_rank
:
int
,
num_local_ranks
:
int
):
...
...
@@ -154,9 +154,9 @@ class suppress_stdout_stderr:
self
.
errnull_file
.
close
()
def
bench_kineto
(
fn
,
kernel_names
,
num_tests
:
int
=
30
,
suppress_kineto_output
:
bool
=
False
,
def
bench_kineto
(
fn
,
kernel_names
:
Union
[
str
,
tuple
]
,
num_tests
:
int
=
30
,
suppress_kineto_output
:
bool
=
False
,
trace_path
:
Optional
[
str
]
=
None
,
barrier_comm_profiling
:
bool
=
False
,
duplicate_name_period
:
Optional
[
int
]
=
None
):
num_kernels_per_period
:
int
=
1
):
# Profile
suppress
=
suppress_stdout_stderr
if
suppress_kineto_output
else
empty_suppress
with
suppress
():
...
...
@@ -175,7 +175,7 @@ def bench_kineto(fn, kernel_names, num_tests: int = 30, suppress_kineto_output:
# Parse the profiling table
assert
isinstance
(
kernel_names
,
str
)
or
isinstance
(
kernel_names
,
tuple
)
is_tuple
d
=
isinstance
(
kernel_names
,
tuple
)
is_tuple
=
isinstance
(
kernel_names
,
tuple
)
prof_lines
=
prof
.
key_averages
().
table
(
sort_by
=
'cuda_time_total'
,
max_name_column_width
=
100
).
split
(
'
\n
'
)
kernel_names
=
(
kernel_names
,
)
if
isinstance
(
kernel_names
,
str
)
else
kernel_names
assert
all
([
isinstance
(
name
,
str
)
for
name
in
kernel_names
])
...
...
@@ -199,29 +199,24 @@ def bench_kineto(fn, kernel_names, num_tests: int = 30, suppress_kineto_output:
break
break
if
duplicate_name_period
is
None
:
return
tuple
(
kernel_times
)
if
is_tupled
else
kernel_times
[
0
]
else
:
detail_times
=
extract_detail_times_from_prof
(
prof
,
kernel_names
=
kernel_names
,
duplicate_name_period
=
duplicate_name_period
)
return
tuple
(
kernel_times
)
+
(
detail_times
,)
def
extract_detail_times_from_prof
(
prof
,
kernel_names
,
duplicate_name_period
:
int
):
with
tempfile
.
NamedTemporaryFile
(
suffix
=
".json"
)
as
tmp
:
prof
.
export_chrome_trace
(
tmp
.
name
)
profile_data
=
json
.
loads
(
Path
(
tmp
.
name
).
read_text
())
ans
=
{}
for
kernel_name
in
kernel_names
:
name_matcher
=
f
'::
{
kernel_name
}
<'
events
=
[
e
for
e
in
profile_data
[
"traceEvents"
]
if
name_matcher
in
e
[
"name"
]]
events
=
sorted
(
events
,
key
=
lambda
e
:
e
[
"ts"
])
durations
=
[
e
[
"dur"
]
/
1e6
for
e
in
events
]
ans
[
kernel_name
]
=
[
list_mean
(
durations
[
i
::
duplicate_name_period
])
for
i
in
range
(
duplicate_name_period
)]
return
ans
def
list_mean
(
xs
):
return
sum
(
xs
)
/
len
(
xs
)
# Expand the kernels by periods
if
num_kernels_per_period
>
1
:
with
tempfile
.
NamedTemporaryFile
(
suffix
=
'.json'
)
as
tmp
:
prof
.
export_chrome_trace
(
tmp
.
name
)
profile_data
=
json
.
loads
(
Path
(
tmp
.
name
).
read_text
())
for
i
,
kernel_name
in
enumerate
(
kernel_names
):
events
=
[
event
for
event
in
profile_data
[
'traceEvents'
]
if
f
'::
{
kernel_name
}
'
in
event
[
'name'
]]
events
=
sorted
(
events
,
key
=
lambda
event
:
event
[
'ts'
])
durations
=
[
event
[
'dur'
]
/
1e6
for
event
in
events
]
assert
len
(
durations
)
%
num_kernels_per_period
==
0
num_kernel_patterns
=
len
(
durations
)
//
num_kernels_per_period
kernel_times
[
i
]
=
[
sum
(
durations
[
j
::
num_kernels_per_period
])
/
num_kernel_patterns
for
j
in
range
(
num_kernels_per_period
)]
# Return execution times
return
kernel_times
if
is_tuple
else
kernel_times
[
0
]
def
hash_tensor
(
t
:
torch
.
Tensor
):
return
t
.
view
(
torch
.
int64
).
sum
().
item
()
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