Datasets:

davidchan

/

anim400k

like 31

Thanks for taking a look! I was able to download and untar the files today correctly. Are you concatenating the part files before trying to untar:

# Either:
cat anim400k_audio_clips.tar.gz.part-* > anim400k_audio_clips.tar.gz
tar -xzvf ./anim400k_audio_clips.tar.gz
# Or:
cat anim400k_audio_clips.tar.gz.part-* | tar -xzv

It's not possible to untar each part file directly.

yes and it either gives me an error like this

 cat anim400k_audio_clips.tar.gz.part-* | tar -xz
tar: Skipping to next header
tar: Archive contains ‘k\a\336D\037h\026\371\244\225\377\373’ where numeric off_t value expected
tar: Archive contains ‘\356$\300\004\027a\252Tň\0\006’ where numeric off_t value expected
tar: Archive contains ‘k*I\260\003\206S\200\004vYZ’ where numeric off_t value expected
tar: Archive contains ‘\356<\004(\206\n\220\303\024\0F’ where numeric off_t value expected
tar: Archive contains ‘\326\304L9\323\037 L2;\377\373’ where numeric off_t value expected

or an error like this (when I tried doing the archive with 7z, I don't have the original tar error anymore, but it was similar)

./anim400k_video_clips/ef/ef005f3b-ba0e-4a6c-96b4-7beba3454e24.mp4
./anim400k_video_clips/ef/efa04739-04af-4fe6-a39e-8fd15c08c23e.mp4

gzip: stdin: invalid compressed data--crc error
tar: Child returned status 1
tar: Error is not recoverable: exiting now

It might be that something went wrong in the git clone process (the datasets python method also yields and error)

Could you maybe share the datasets, github, huggingface-cli and tar versions you're using?
Also some md5 checksums to verify against in case the archives just got corrupted in download on the client side would be amazing :)

Sure -
datasets - [not installed]
huggingface-hub - 0.26.0
tar - 1.3

For the audio files, the md5 hashes I have are:

75926bc81c0bf26866b3f3ac2a4aaa8d  ./anim400k_audio_clips.tar.gz.part-02
71c66eb47fef1ef49ef44805421623d1  ./anim400k_audio_clips.tar.gz.part-00
cba9a7553cdc8a5886f533d40a75b688  ./anim400k_audio_clips.tar.gz.part-05
1fed9d8667161f44013561378927d22f  ./anim400k_audio_clips.tar.gz.part-03
06dbc7ac91775707451987368b3b5d07  ./anim400k_audio_clips.tar.gz.part-01
f140b56f2666fead1225041400d8c809  ./anim400k_audio_clips.tar.gz.part-04

I ran the following command to download from Huggingface:

huggingface-cli download davidchan/anim400k --repo-type dataset --local-dir .