需要使用pynvml库 官网:https://pythonhosted.org/nvidia-ml-py/ 下载文件地址:https://pypi.org/project/nvidia-ml-py/#history 现阶段pip安装的命令为: sudo pip install nvidia-ml-py 具体实例 import pynvml pynvml.nvmlInit() # 这里的1是GPU id handle
安装 pip install nvidia-ml-py from pynvml import * def nvidia_info(): # pip install nvidia-ml-py
简介对应的py库介绍主要是来自: nvidia-ml-py。 developer.nvidia.com/nvidia-management-library-nvmlDownload the latest package from: http://pypi.python.org/pypi/nvidia-ml-py return codes are raised as exceptions as described in the section below.安装pip安装python3 -m pip install nvidia-ml-py 安装包安装下载地址:nvidia-ml-py 12.555.43wget https://files.pythonhosted.org/packages/ee/c6/2348fc1fb776ff41a34635fb1f18010a6d6fd7ba6e57184dabd9d98ba9cf
https://pypi.python.org/pypi/nvidia-ml-py 同时需要执行如下命令安装NVML的Python库: pip install nvidia-ml-py 2.自定义监控配置
安装 直接通过pip安装: pip install nvidia-ml-py 或者根据所使用的python版本安装对应包: pip install nvidia-ml-py2 # python2
=gpu_name,power.draw,temperature.gpu --format=csv --loop=1 Python查询方法 借助pynvml,官方文档::: pip install nvidia-ml-py
• requirements 中替换 pynvml 为 nvidia-ml-py。 • CI 流程中增加构建前释放磁盘空间和长会话参数测试。 • 增加 prefixcache 功能及性能测试。
如果想进一步了解: https://discuss.pytorch.org/t/understanding-gpu-memory-usage/7160 https://pypi.org/project/nvidia-ml-py out-of-memory https://pytorch.org/docs/master/notes/cuda.html#memory-management https://pythonhosted.org/nvidia-ml-py
http://developer.nvidia.com/nvidia-management-library-nvml/ http://pypi.python.org/pypi/nvidia-ml-py/
http://developer.nvidia.com/nvidia-management-library-nvml/ http://pypi.python.org/pypi/nvidia-ml-py/ 下篇博文预告会讲关于python http://pypi.python.org/pypi/nvidia-ml-py/的使用以及调用方法。