Datasets:
Download HiRISE/lAv.tar from WML05/MOMO-pretraining-data: direct link, hf CLI and curl.
- Browser
- Download file 49 GB
-
https://huggingface.co/datasets/WML05/MOMO-pretraining-data/resolve/main/HiRISE/lAv.tar
- Command line
-
hf download hf://datasets/WML05/MOMO-pretraining-data/HiRISE/lAv.tar
-
curl -L -o lAv.tar https://huggingface.co/datasets/WML05/MOMO-pretraining-data/resolve/main/HiRISE/lAv.tar
49 GB
- Xet hash:
- 754be6636a2167f546ee931bd5a0c56d808444be45d14e0bd4097f25f7fc2a91
- Size of remote file:
- 49 GB
- SHA256:
- 01027153684162facfd9205a83ee6e6c4200f4ac6ad48df917f0e0767566138c
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.