Created by: super3kl

import paddle.fluid as pp

pp.install_check.run_check() Running Verify Fluid Program ... W0520 15:26:18.307366 24061 device_context.cc:252] Please NOTE: devic W0520 15:26:18.312017 24061 device_context.cc:260] device: 0, cuDNN V W0520 15:26:20.338177 24061 device_context.h:155] WARNING: device: 0.is 7.3, which may cause serious incompatible bug. Please recompile or Your Paddle Fluid works well on SINGLE GPU or CPU. W0520 15:26:31.293614 24061 dynamic_loader.cc:120] Can not find libra_PATH. /home/zhulei/anaconda3/envs/paddle/lib/python3.7/site-packages/paddlen. "The following exception is not an EOF exception.") 2020-05-20 15:26:31,316-WARNING: Your Paddle Fluid has some problem w

1. There is only 1 or 0 GPU visible on your Device;
2. No.1 or No.2 GPU or both of them are occupied now
3. Wrong installation of NVIDIA-NCCL2, please follow instruction on to test your NCCL, or reinstall it following https://docs.nvidia.com

Original Error is:

C++ Call Stacks (More useful to developers):

0 std::string paddle::platform::GetTraceBackString<char const*>(cha 1 paddle::platform::EnforceNotMet::EnforceNotMet(std::__exception_p 2 paddle::platform::dynload::GetNCCLDsoHandle() 3 void std::__once_call_impl<std::_Bind_simple<paddle::platform::dyint*)::{lambda()#1 (closed)} ()> >() 4 paddle::platform::NCCLContextMap::NCCLContextMap(std::vector<padd, unsigned long, unsigned long) 5 paddle::platform::NCCLCommunicator::InitFlatCtxs(std::vector<paddcclUniqueId*, std::allocator<ncclUniqueId*> > const&, unsigned long, 6 paddle::framework::ParallelExecutorPrivate::InitNCCLCtxs(paddle:: 7 paddle::framework::ParallelExecutorPrivate::InitOrGetNCCLCommunic 8 paddle::framework::ParallelExecutor::ParallelExecutor(std::vectortor<std::string, std::allocatorstd::string > const&, std::string coorpaddle::framework::Scope* > const&, paddle::framework::details::Eramework::ir::Graph*)

Error Message Summary:

Error: Failed to find dynamic library: libnccl.so ( libcudart.so.8.0: Please specify its path correctly using following ways: Method. set environment variable LD_LIBRARY_PATH on Linux or DYLD_LI For instance, issue command: export LD_LIBRARY_PATH=... Note: After Mac OS 10.11, using the DYLD_LIBRARY_PATH is impossible orm/dynload/dynamic_loader.cc:177)

Your Paddle Fluid is installed successfully ONLY for SINGLE GPU or CP Let's start deep Learning with Paddle Fluid now

使用学校服务器 提示无法找到libcudart.so.8.0 服务器装了 cuda10 和cuda 9 这俩有libcudart.so.9.0 libcudart.so.10.0 但就是没有8.0