Created by: shenlan211314

开始是内存不足，调整batch_size为2后，内存应该是能满足（实际8G）。但是后面又报错了。 运行环境是win10，CPU，内存8G，paddlepaddle 1.4.1。

Theoretical memory usage in training: 3420.067 - 3582.927 MB warm validate or test Load pretraining parameters from MODEL_PATH/params. ParallelExecutor is deprecated. Please use CompiledProgram and Executor. CompiledProgram is a central place for optimization and Executor is the unified executor. Example can be found in compiler.py. W0506 20:38:55.854854 15620 graph.h:204] WARN: After a series of passes, the current graph can be quite different from OriginProgram. So, please avoid us ing the OriginProgram() method! Traceback (most recent call last): File "run_classifier.py", line 307, in main(args) File "run_classifier.py", line 201, in main main_program=train_program) File "D:\python3\lib\site-packages\paddle\fluid\parallel_executor.py", line 134, in init self._compiled_program._compile(place=self._place, scope=self._scope) File "D:\python3\lib\site-packages\paddle\fluid\compiler.py", line 307, in _compile scope=self._scope) File "D:\python3\lib\site-packages\paddle\fluid\compiler.py", line 278, in _compile_data_parallel self._exec_strategy, self._build_strategy, self._graph) paddle.fluid.core.EnforceNotMet: Fail to allocate CPU memory: size = 263382656 . at [D:\1.4.1\paddle\paddle\fluid\memory\detail\system_allocator.cc:56] PaddlePaddle Call Stacks: Windows not support stack backtrace yet.