Skip to content

M
models

PaddlePaddle / models
大约 2 年 前同步成功

通知 232
Star 6828
Fork 2962
M
models
未验证 提交 225db53d 编写于 作者: L lvmengsi 提交者： GitHub
浏览文件
操作

fix gan memory (#2342)

上级 446243af
隐藏空白更改
内联 并排
Showing with 3 addition and 3 deletion
PaddleCV/gan/trainer/CGAN.py
浏览文件 @ 225db53d
...@@ -114,7 +114,7 @@ class CGAN(object): ...@@ -114,7 +114,7 @@ class CGAN(object):
### memory optim ### memory optim
build_strategy = fluid.BuildStrategy() build_strategy = fluid.BuildStrategy()
build_strategy.enable_inplace = True build_strategy.enable_inplace = True
build_strategy.memory_optimize = True build_strategy.memory_optimize = False
g_trainer_program = fluid.CompiledProgram( g_trainer_program = fluid.CompiledProgram(
g_trainer.program).with_data_parallel( g_trainer.program).with_data_parallel(
......
PaddleCV/gan/trainer/CycleGAN.py
浏览文件 @ 225db53d
...@@ -237,7 +237,7 @@ class CycleGAN(object): ...@@ -237,7 +237,7 @@ class CycleGAN(object):
### memory optim ### memory optim
build_strategy = fluid.BuildStrategy() build_strategy = fluid.BuildStrategy()
build_strategy.enable_inplace = True build_strategy.enable_inplace = True
build_strategy.memory_optimize = True build_strategy.memory_optimize = False
gen_trainer_program = fluid.CompiledProgram( gen_trainer_program = fluid.CompiledProgram(
gen_trainer.program).with_data_parallel( gen_trainer.program).with_data_parallel(
......
PaddleCV/gan/trainer/DCGAN.py
浏览文件 @ 225db53d
...@@ -110,7 +110,7 @@ class DCGAN(object): ...@@ -110,7 +110,7 @@ class DCGAN(object):
### memory optim ### memory optim
build_strategy = fluid.BuildStrategy() build_strategy = fluid.BuildStrategy()
build_strategy.enable_inplace = True build_strategy.enable_inplace = True
build_strategy.memory_optimize = True build_strategy.memory_optimize = False
g_trainer_program = fluid.CompiledProgram( g_trainer_program = fluid.CompiledProgram(
g_trainer.program).with_data_parallel( g_trainer.program).with_data_parallel(
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或