使用 sudo 导入 Caffe 不起作用

使用 sudo 导入 Caffe 不起作用

当我使用 sudo 运行该文件时,它说找不到它:

veli@veli:~/deepdream$ sudo python video-convert.py
Traceback (most recent call last):
  File "video-convert.py", line 16, in <module>
    import caffe
ImportError: No module named caffe

当我在没有 sudo 的情况下运行它时,它确实找到了它,但不幸的是需要文件中稍后的权限:

veli@veli:~/deepdream$ python video-convert.py
Traceback (most recent call last):
  File "video-convert.py", line 81, in <module>
    open('tmp.prototxt', 'w').write(str(model))
IOError: [Errno 13] Permission denied: 'tmp.prototxt'

我已经跑完了export PYTHONPATH=/home/veli/caffe/python:$PYTHONPATH,所以我真的不知道现在还能做什么。

答案1

解决了!(部分)

我跑了:

cd
chgrp adm ./caffe
chmod g+rwx ./caffe

然后:

sudo chgrp adm ./deepdream
sudo chmod g+rwx adm ./deepdream

(^由于某种原因,它不允许我运行这些非 sudo ^)

那么现在就可以了:

cd deepdream
python video_convert.py

它成功运行直到我收到以下错误:

Traceback (most recent call last):
  File "video-convert.py", line 85, in <module>
    channel_swap = (2,1,0)) # the reference model has channels in BGR order instead of RGB
  File "/home/veli/caffe/python/caffe/classifier.py", line 26, in __init__
    caffe.Net.__init__(self, model_file, caffe.TEST, weights=pretrained_file)
RuntimeError: Could not open file /home/veli/caffe/models/bvlc_googlenet/bvlc_googlenet.caffemodel

答案2

(无法评论其他答案所以我必须在这里发布)

修理:

Traceback (most recent call last):
  File "video-convert.py", line 85, in <module>
    channel_swap = (2,1,0)) # the reference model has channels in BGR order instead of RGB
  File "/home/veli/caffe/python/caffe/classifier.py", line 26, in __init__
    caffe.Net.__init__(self, model_file, caffe.TEST, weights=pretrained_file)
RuntimeError: Could not open file /home/veli/caffe/models/bvlc_googlenet/bvlc_googlenet.caffemodel

下载bvlc_googlenet.caffemodelhttps://github.com/BVLC/caffe/tree/master/models/bvlc_googlenet并将其放入缺少的目录中。

相关内容