|
Algorithm 1 Layer-wise PSM |
-
1:
Execution time on CPU or GPU per layer
-
2:
Data transfer time on FP and BP per layer
-
3:
Data transfer time on UP per layer
-
4:
for do
-
5:
-
6:
-
7:
if
then
-
8:
-
9:
for
do
-
10:
-
11:
if
then
-
12:
-
13:
-
14:
else
-
15:
for
do
-
16:
-
17:
if
then
-
18:
-
19:
-
return
|