google-deepmind / google-deepmind/dmvr
Help on generating tfrecord file
- Dominant language
- Python
- Stars
- 68
- Forks
- 14
- PR merge metrics
- No merged PRs in 30d
Description
Dear developers of this code,
I am trying to follow the guideline you have provided but somehow fail to generate the tfrecord file. Any help would be much appreciated. Let me explain the error I am encountering:
I have successfully installed the requirements and created the csv file as instructed. The only difference is, inside my csv file, I gave the absolute path of the video. The rest is the same. i.e. "video_path,start,end,label,caption _somepath_\heic1608c.mp4,1.5,6.0,space,the view of the space from a telescope"
Then I created the folder to put the generated files.
Then, when I try to run "generate_from_file.py" from the console like this: "python generate_from_file.py input.csv generated_dataset"
where both the input.csv and generated_dataset folder is in the same directory, I get this error:
File "C:\Users\user\miniconda3\envs\ObjectVivit\Lib\site-packages\pandas\io\common.py", line 460, in _get_filepath_or_buffer
raise ValueError(msg)
ValueError: Invalid file path or buffer object type:
I have tried changing the direction of slashes in the csv file, doubling them etc. (just in case) but the error does not change.
I also tried to give any permissions in case it suffers from such an issue. But the folder "generate_from_file.py" is in has all the permissions as well.
I am out of ideas.
Any help is highly appreciated.
Thanks
Contributor guide
Assessment
This issue has not been assessed yet.