简体繁体 English

在Anaconda中安装tensorflow-gpu，CUDA计算能力低于3.0

[英]Install tensorflow-gpu in Anaconda with CUDA Compute Capability less than 3.0

原文 2018-01-29 07:27:51 7 1 python/ tensorflow/ anaconda

I have been trying to install tensorflow-gpu in Anaconda on Windows 10 with a graphics card that has Compute Compatibility 2.1. 我一直在尝试使用具有Compute Compatibility 2.1的图形卡在Windows 10的Anaconda中安装tensorflow-gpu。 I keep getting the error below. 我不断收到以下错误。 Screenshot of issue 问题截图

pip install tensorflow-gpu works well but then I cannot import the library or create a tensorflow environment. pip install tensorflow-gpu效果很好，但是我无法导入该库或创建一个tensorflow环境。 Is there a workaround for this? 有没有解决方法？ Even if I get tensorflow-gpu installed, will I be able to run my model on the GPU without having Compute Compatibility 3.0? 即使安装了tensorflow-gpu，我是否也可以在没有Compute Compatibility 3.0的情况下在GPU上运行我的模型？

1 个解决方案

Unfortunately you won't be able to run any model on your GPU as Compute Compatility 3.0 is required (see here ). 不幸的是，由于需要Compute Compatility 3.0，因此您将无法在GPU上运行任何模型（请参阅此处）。

Your first attempt showed that you were missing some packages (numpy for instance). 您的第一次尝试表明您缺少某些软件包（例如numpy）。 If you installed tensorflow-gpu with pip and you can't import the library that means that your python interpreter does not see the path where pip installs your packages. 如果您使用pip安装了tensorflow-gpu并且无法导入该库，则意味着您的python解释器看不到pip安装软件包的路径。

Anaconda 中的 Tensorflow-Gpu 和 Cuda 驱动程序存在问题 - Problem with Tensorflow-Gpu and Cuda drivers in Anaconda

tensorflow-gpu 未使用 anaconda 安装 - tensorflow-gpu not install using anaconda

GPU 计算能力 3.0 但所需的最低 Cuda 能力为 3.5 - GPU Compute Capability 3.0 but the minimum required Cuda capability is 3.5

TensorFlow-gpu 安装与 Anaconda - Tensorflow-gpu installation with Anaconda

Tensorflow：Cuda 计算能力 3.0。最低要求的 Cuda 能力是 3.5 - Tensorflow: Cuda compute capability 3.0. The minimum required Cuda capability is 3.5

Kebuntu上的keras（-gpu）+ tensorflow-gpu + anaconda - keras(-gpu) + tensorflow-gpu + anaconda on Kubuntu

Tensorflow-GPU 不使用 GPU 和 CUDA,CUDNN - Tensorflow-GPU not using GPU with CUDA,CUDNN

如果我的计算能力低于3.0，可以通过GPU导入Tensorflow吗？ - Can I import Tensorflow with GPU if my compute ability is less than 3.0

尝试安装 tensorflow-gpu 但出现此错误：CUDA 驱动程序版本不足以用于 CUDA 运行时版本 - Trying to install tensorflow-gpu but got this error: CUDA driver version is insufficient for CUDA runtime version

tensorflow-gpu 2.3 安装了 cuda 并且 cudnn 无法检测到 GPU？ - tensorflow-gpu 2.3 install with cuda and cudnn can't detect GPU?

暂无

暂无

声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

相关问题 Anaconda 中的 Tensorflow-Gpu 和 Cuda 驱动程序存在问题 - Problem with Tensorflow-Gpu and Cuda drivers in Anaconda tensorflow-gpu 未使用 anaconda 安装 - tensorflow-gpu not install using anaconda GPU 计算能力 3.0 但所需的最低 Cuda 能力为 3.5 - GPU Compute Capability 3.0 but the minimum required Cuda capability is 3.5 TensorFlow-gpu 安装与 Anaconda - Tensorflow-gpu installation with Anaconda Tensorflow：Cuda 计算能力 3.0。最低要求的 Cuda 能力是 3.5 - Tensorflow: Cuda compute capability 3.0. The minimum required Cuda capability is 3.5 Kebuntu上的keras（-gpu）+ tensorflow-gpu + anaconda - keras(-gpu) + tensorflow-gpu + anaconda on Kubuntu Tensorflow-GPU 不使用 GPU 和 CUDA,CUDNN - Tensorflow-GPU not using GPU with CUDA,CUDNN 如果我的计算能力低于3.0，可以通过GPU导入Tensorflow吗？ - Can I import Tensorflow with GPU if my compute ability is less than 3.0 尝试安装 tensorflow-gpu 但出现此错误：CUDA 驱动程序版本不足以用于 CUDA 运行时版本 - Trying to install tensorflow-gpu but got this error: CUDA driver version is insufficient for CUDA runtime version tensorflow-gpu 2.3 安装了 cuda 并且 cudnn 无法检测到 GPU？ - tensorflow-gpu 2.3 install with cuda and cudnn can't detect GPU?

相关标签

粤ICP备18138465号 © 2020-2024 STACKOOM.COM