Created by: shiyazhou121

这个报错是需要的显存大于可利用的吗？我应该怎么优化这个问题？尝试过FLAGS_fraction_of_gpu_memory_to_use = 0.8，没有效果。 试了fluid.memory_optimize，没有效果。

paddle.fluid.core.EnforceNotMet: Enforce failed. Expected allocating <= available, but received allocating:20998686233 > available:19587333888. Insufficient GPU memory to allocation. at [/paddle/paddle/fluid/platform/gpu_info.cc:170] PaddlePaddle Call Stacks: 0 0x7f22456da6a6p paddle::platform::EnforceNotMet::EnforceNotMet(std::__exception_ptr::exception_ptr, char const*, int) + 486 1 0x7f224707162ep paddle::platform::GpuMaxChunkSize() + 782 2 0x7f224706adaap 3 0x7f2280848a90p pthread_once + 80 4 0x7f224706a21dp paddle::memory::legacy::GetGPUBuddyAllocator(int) + 93 5 0x7f224706a505p void* paddle::memory::legacy::Allocpaddle::platform::CUDAPlace(paddle::platform::CUDAPlace const&, unsigned long) + 37 6 0x7f224706abb5p paddle::memory::allocation::LegacyAllocator::AllocateImpl(unsigned long, paddle::memory::allocation::Allocator::Attr) + 389 7 0x7f224706cafbp paddle::memory::allocation::Allocator::Allocate(unsigned long, paddle::memory::allocation::Allocator::Attr) + 27