为什么 PAI DSW中一直无法使用GPU加速tensorflow，如何使用GPU加速.
并且会显示
Num GPUs Available: 0
Available GPUs: []
tensorflow/core/util/port.cc:111] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0.
2024-05-26 15:03:10.885007: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.
2024-05-26 15:03:10.914951: E tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:9342] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-05-26 15:03:10.914975: E tensorflow/compiler/xla/stream_executor/cuda/cuda_fft.cc:609] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-05-26 15:03:10.914996: E tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:1518] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-05-26 15:03:10.920456: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.
2024-05-26 15:03:10.920647: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropr
写回答
-
“Num GPUs Available: 0”，主要是因为环境未正确配置或缺少必要的CUDA和cuDNN驱动，导致系统无法识别或使用GPU。还有可能存在库冲突或重复注册的问题，检查是否有多余的库路径或环境变量设置，确保环境干净无误配置
2024-05-27 10:08:02赞同 2 展开评论 打赏 -
北京阿里云ACE会长cuDNN注册问题：错误信息Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered表明cuDNN工厂已经被注册，可能存在版本冲突或初始化问题。
cuFFT和cuBLAS注册问题：类似的注册问题也出现在cuFFT和cuBLAS上，这可能与TensorFlow试图加载的GPU操作库有关。
2024-05-27 08:04:31赞同 1 展开评论 打赏
版权声明：本文内容由阿里云实名注册用户自发贡献，版权归原作者所有，阿里云开发者社区不拥有其著作权，亦不承担相应法律责任。具体规则请查看《阿里云开发者社区用户服务协议》和《阿里云开发者社区知识产权保护指引》。如果您发现本社区中有涉嫌抄袭的内容，填写侵权投诉表单进行举报，一经查实，本社区将立刻删除涉嫌侵权内容。
人工智能平台 PAI（Platform for AI，原机器学习平台PAI）是面向开发者和企业的机器学习/深度学习工程平台，提供包含数据标注、模型构建、模型训练、模型部署、推理优化在内的AI开发全链路服务，内置140+种优化算法，具备丰富的行业场景插件，为用户提供低门槛、高性能的云原生AI工程化能力。
