tensorflow
_yuki_
一路打怪升级,直至见到大魔王
展开
-
解决 Could not create cudnn handle: CUDNN_STATUS_INTERNAL_ERROR
问题:升级 TF 容易引起Failed to get convolution algorithm. This is probably because cuDNN failed to initialize解决:import tensorflow as tf#-----------------------------------------------------------------------# physical_devices = tf.config.experimenta.原创 2020-07-17 15:07:01 · 736 阅读 · 0 评论 -
解决 Could not load dynamic library cudart64_101.dll 或 tensorflow 2.3.0 CUDA 10.0 问题
环境:python 3.6.8 tensorflow-gpu2.3.0rc0 CUDA 10.0 cudnn 7.4错误:Could not load dynamic library ‘cudart64_101.dll‘Could not load dynamic library ‘cublas64_10.dll‘Could not load dynamic library ‘cufft64_10.dll‘Could not load dynamic library ‘cur..原创 2020-07-16 16:25:01 · 2987 阅读 · 0 评论 -
解决 Could not load dynamic library libcudart.so.10.1
环境:python 3.6.8 tensorflow-gpu2.3.0rc0 CUDA 10.0 cudnn 7错误:Could not load dynamic library 'libcudart.so.10.1'; dlerror: libcudart.so.10.1: cannot open shared object file: No such file or directoryCould not load dynamic library 'libcublas.so.1..原创 2020-07-16 16:01:38 · 24364 阅读 · 2 评论 -
Jetson 系列 (六)Tensorflow
一、基础依赖系统:$ sudo apt-get update$ sudo apt-get install libhdf5-serial-dev hdf5-tools libhdf5-dev zlib1g-dev zip libjpeg8-dev liblapack-dev libblas-dev gfortran更新pip3:$ sudo apt-get install python3-pip$ sudo pip3 install -U pip testresources setup原创 2020-05-19 14:49:26 · 392 阅读 · 0 评论 -
解决:tensorflow have no attribute enable_eager_execution()
导致报错,你的tensorflow版本为多少呢?如果为2.0 或 2.0 以上,那么请看下文:2.0版本 移除 了直接调用 enable_eager_execution() 的方式。原因:新版本将 enable_eager_execution()放入了 compat.v1 内部。整理:# tensorflow >= 1.5tf.compat.v1.enable_e...原创 2020-04-20 13:01:34 · 11158 阅读 · 0 评论 -
解决:Please use tf.compat.v1.Session instead.
环境:tensorflow >=1.14错误:WARNING:tensorflow:From /Users/yuki//tf_test/test1.py:4: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.错误分析:为了使TensorFlow在2.0版中更...原创 2020-04-19 16:48:23 · 8175 阅读 · 0 评论 -
解决:'Tensor' object has no attribute 'numpy'
环境tensoflow >= 1.4制造错误:import tensorflow as tfm = tf.keras.metrics.Accuracy()m.update_state([1, 2, 3, 4], [1, 2, 3, 4])print('Final result: ', m.result().numpy()) 错误:Attribut...原创 2020-04-19 16:38:39 · 32447 阅读 · 5 评论 -
OpenVino (三) tensorflow yolo3 同步/异步 实战
环境:python3.6 / anacoda3 tensorflow>=1.2.0,<2.0.0 networkx>=1.11,<2.4 numpy>=1.12.0 defusedxml>=0.5.0 openvino (python) opencv4.0一、下载yolo3.weight与coco.names以coco数据集训练的权重...原创 2020-02-08 03:45:30 · 1964 阅读 · 0 评论 -
tensorflow 出现 ... np_resource = np.dtype([("resource", np.ubyte, 1)]) 解决办法
/home/yuki/.local/lib/python3.6/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, ...原创 2019-12-26 16:55:20 · 8547 阅读 · 4 评论 -
tensorflow 出现“cannot identify image file ‘/Users//.DS_Store‘”问题解决的办法
“cannot identify image file '/Users//.DS_Store'”在用mac跑深度学习tensorflow框架是制作图片格式向tfrecord格式转化中,读取文件夹中的图片出现系统文件时候报错,提示有.DS_Store系统文件。最简单的办法就是删除,之前在网上找了有些办法不太适用,要么是删除内存中所有的.DS_Store文件,显然在这里是小题大做了,采用的做法是,cd...原创 2018-05-24 11:00:45 · 41384 阅读 · 0 评论