ERROR记录:Failed to get convolution algorithm. This is probably because cuDNN failed to initialize

GPU初始化错误:
用pycharm的话,只能一个程序用GPU跑,先把其他不跑的程序停止,重新运行,如果还有错误的话,加上如下代码:

import os
import tensorflow as tf

from tensorflow.compat.v1 import ConfigProto
from tensorflow.compat.v1 import InteractiveSession
os.environ['TF_FORCE_GPU.ALLOW_GROWTH'] = 'TRUE'
gpus = tf.config.experimental.list_physical_devices('GPU')
for gpu in gpus:
    tf.config.experimental.set_memory_growth(gpu, True)

IndexError: index 60 is out of bounds for axis 1 with size 52
YOLOv3错误,这个是标签位置与图片大小不对应,比如标签为1256,670,1308,760,0,而实际图片大小只有1100*1100…

猜你喜欢

转载自blog.csdn.net/weixin_45371989/article/details/108738745
今日推荐