深度学习报错Error #15: Initializing libiomp5md.dll, but found libiomp5md.dll already initialized.的解决方法

python进行深度学习时运行程序报错OMP: Error#15: Initializing libiomp5md.dll, but found libiomp5md.dll already initialized.OMP: Hint This means that multiple copies of the OpenMP runtime have been linked into the program.That is dangerous,since it can degrade performance or cause incorrect results.The best thing todois to ensure that only a single OpenMP runtime is linked into theprocess,e.g.by avoiding static linking of the OpenMP runtime in any library.As an unsafe,unsupported,undocumented workaround you cansetthe environment variable KMP_DUPLICATE_LIB_OKTRUE to allow the program tocontinueto execute,but that may cause crashes or silently produce incorrect results.Formore information,please see http://www.intel.com/software/products/support/.这个问题表示libiomp5md.dllOpenMP运行时库被重复初始化。根据报错信息最简单的处理方法是在程序前面添加:importos os.environ[KMP_DUPLICATE_LIB_OK]TRUE这样就跳过了问题程序可以成功运行。或者使用https://cloud.tencent.com/developer/article/2542156 里面的方法直接删除多余的dll。但是感觉这样只是治标不治本并没有解决问题于是一行一行地处理程序发现问题出现在调用matplotlib包的时候。看到有人说matplotlib影响那么可以尝试一下该命令conda install nomkl 对该包进行安装看是否能解决问题。查了一下nomkl 是 conda 包管理器中一个已弃用的元包‌主要用于控制 Python 科学计算库是否使用 Intel MKL 加速库不是普通的功能包而是一个切换开关类型的元包用来告诉 conda 在安装 NumPy、SciPy 等库时选择非 MKL 版本 。‌‌‌所以重新安装或更新涉及到 MKL 的库如 NumPy可以解决冲突问题。所以我更新了这个环境中的这两个库pip install --upgrade numpy matplotlib再次运行成功解决问题不再出现Error #15: Initializing libiomp5md.dll, but found libiomp5md.dll already initialized.。