Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try...

最新推荐文章于 2026-06-19 13:33:59 发布
原创 最新推荐文章于 2026-06-19 13:33:59 发布 · 629 阅读 · 1 ·
本内容遵循CC 4.0 BY-SA版权协议
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
标签
#tensorflow #机器学习
在使用Inception-V3模型进行预训练时，遇到GPU初始化失败的问题。通过在retrain.py文件中添加代码，设置GPU动态增长，成功解决了“cuDNN failed to initialize”错误。

使用inception-V3批处理文件retrain.bat进行预训练时，发现报错“Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try…”于是打开retrain.py文件，在前面加上如下代码：
from tensorflow.compat.v1 import ConfigProto
from tensorflow.compat.v1 import InteractiveSession

config = ConfigProto()
config.gpu_options.allow_growth = True
session = InteractiveSession(config=config)
在cd到retrain.bat文件，成功运行。
参考文章：https://blog.csdn.net/qq_41868689/article/details/98503069

Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try…
01-06
使用inception-V3批处理文件retrain.bat进行预训练时，发现报错“Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try…”于是打开retrain.py文件，在前面加上如下代码： from tensorflow.compat.v1 import ConfigProto from tensorflow.compat.v1 import InteractiveSession config = ConfigProto() config.gpu_o
Failed to get convolution algorithm. This is probably because cuDNN failed..(TensorFlow和keras显存不足报错)
nohopenolove的博客
11-25 2246
在跑深度学习程序时，用到TensorFlow或者keras时候，经常会报一个错误： tensorflow.python.framework.errors_impl. UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize解决办法
牧野渔樵
11-08 904
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize解决办法 在网上看了很多说法，总结就两种情况： cudnn和tensorflow的版本不相符 GPU内存不足造成 第一种情况： 需要查出cuda、cudnn、tensorflow的版本对应关系，cuda版本只能是9.0,...
Error : Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so
胡侃有料的博客
03-05 627
every blog every motto: Brevity is the soul of wit 0. 前言 在用tensorflow 进行训练时，添加了如下代码： Error : Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to...
“Failed to get convolution algorithm. This is probably because cuDNN failed to initialize”错误的解决办法
热门推荐
史丹利复合田的博客
11-06 5万+
最近在使用TF2.0。运行程序出现以下错误。 Failed to get convolution algorithm. This is probably because cuDNN failed to initialize 一开始怀疑是CUDA和CuDNN配置错误（要求版本匹配）。反复试验后，还是有这个错误。 最后发现可能是GPU内存不足造成的。需要在程序前加以下一段代码： from tenso...
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try loo
weixin_42671891的博客
07-08 244
一，问题描述 在使用tensorflow-gpu时，出现下面的错误： tensorflow.python.framework.errors_impl.UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. [[node sequ
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try lo
weixin_45904768的博客
04-20 266
“Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. [Op:Conv2D]”错误解决 Failed to get convolution algorithm. This is probably because cuDNN failed to init
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try ...
敲代码的乔帮主
06-21 501
报错内容： tensorflow.python.framework.errors_impl.UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. [Op:Conv2D] Process finished with ex
ailed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try look
Terry_n的博客
06-04 514
添加代码 from tensorflow.compat.v1 import ConfigProto from tensorflow.compat.v1 import InteractiveSession config = ConfigProto() config.gpu_options.allow_growth = True session = InteractiveSession(config=config)
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
baidu_34172099的博客
12-31 2288
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. Could not create cudnn handle: CUDNN_STATUS_...
Failed to get convolution algorithm. This is probably because cuDNN failed to initialize【解决】
libertyhhn的博客
02-27 9696
已安装系统背景 1、tensorflow-gpu==1.12 2、cuda9.0 3、cudnn-9.0-linux-x64-v7.tgz 4、ubuntu系统 在运行代码的时候出现了报错 UnknownError (see above for traceback): Failed to get convolution algorithm. This is probably becau...
【ERROR】Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
这个人很蓝
09-06 728
1、有可能是显存不足，尝试减小batchsize 2、os.environ['TF_FORCE_GPU_ALLOW_GROWTH'] = 'true' 这个语句的意思是，TensorFlow 在分配显存的时候，设置为可以逐步分配，需要多少显存，就一点点往上添，直到够用为止。比如GPU显存为8GB，开始训练一个模型，TensorFlow 先分配100MB，然后200MB 、300MB……最后在700MB处停住了，因为实际700MB就够用了。参考自 ...
【深度学习】Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so t
m0_52357468的博客
03-09 1238
本人TensorFlow-gpu 2.0 电脑显卡1650ti。 有时候电脑明明gpu没任何占用，跑深度学习代码还是会出现以下这个报错： UnknownError: 2 root error(s) found. (0) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log messa
【深度学习】问题解决：Failed to get convolution algorithm. This is probably because cuDNN failed to initialize
qq_30885821的博客
06-11 975
目前本人TensorFlow 1.12 1.在import tensorflow as tf 语句后面，设置allow_growth： config = tf.ConfigProto（） config.gpu_options.allow_growth = True 2.把原来的sess = tf.Session(不管里面有什么)语句换成： sess = tf.Session(config = config) 3.重新运行，问题解决 PS：另外一个很麻烦的解决方案是使用以下方法将TensorFlow降级到
解决Failed to get convolution algorithm. cudnn failed to initialize...
Allen__Iverson的博客
11-13 2万+
CuDNN library compatibility error after loading model weights服务器环境错误原因如解决方案 服务器环境 Ubuntu16.04、cuda9.0、cudnn7.0、tensorflow1.12.0、python3.5、1080ti 错误 Failed to get convolution algorithm. cudnn failed to...
keras错误记录：tensorflow.python.framework.errors_impl.UnknownError: Failed to get convolution algorithm
cyz52的博客
08-15 2084
tensorflow.python.framework.errors_impl.UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was print...
Ubuntu 20.04 TensorFlow 2.10 环境基线构建指南
最新发布
weixin_33816821的博客
06-19 427
TensorFlow 是一个用于机器学习和深度学习的开源框架，其运行依赖严格的 Python 版本、系统库（如 glibc、CUDA）、以及第三方包（如 numpy、protobuf）的版本对齐。在 Ubuntu 20.04 这类长期支持（LTS）系统中，环境不稳定性往往源于系统 Python 与用户环境的耦合、pip 全局安装导致的依赖污染，以及 GPU 加速所需的 CUDA/cuDNN 版本错配。TensorFlow 2.10 作为最后一个内置 GPU 支持的稳定版本，要求 Python 3.9、pip
评论 1
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值