Problem downloading files from dataverse

[ennauata]

Hello,

Thanks for the great work. I'm trying to download the I3D keypoints and .tsv from the dataverse, but the download is really slow.

I wonder if there's a faster option for downloading the data.

Thanks.

[rabeya-akter]

Try downloading in collab .

from google.colab import drive
drive.mount("/content/gdrive")
!wget -c -r https://dataverse.csuc.cat/api/access/datafile/51543?gbrecs=true -O {"folder name/train.zip"}

this might fail 2/3 times. I could download the file in the third try.

Other files can be downloaded without any issues individually.

[ennauata]

Thank your for your response. I'm trying to download this one (https://dataverse.csuc.cat/dataset.xhtml?persistentId=doi%3A10.34810%2Fdata693). Would you know the right path for accessing through the API?

[rabeya-akter]

Thank your for your response. I'm trying to download this one (https://dataverse.csuc.cat/dataset.xhtml?persistentId=doi%3A10.34810%2Fdata693). Would you know the right path for accessing through the API?

No, I don't know about that. But the main problem happens while downloading the train.zip file. Other files can be downloaded individually quiet easily without any problem.

[ycmin95]

A fessible solution is using curl to split the large file into chunks, download them one by one, and merge them into the whole file. This solution works well for me:

curl --range 0-2000000000 -o part1 <download link for train.zip>
curl --range 2000000001-4000000000 -o part2 <download link for train.zip>
curl --range 4000000001-6000000000 -o part3 <download link for train.zip>
curl --range 6000000001- -o part4 <download link for train.zip>
cat part1 part2 part3 part4 > train.zip

The download results look like this: