Download Simulation/GelSight_mini/roughness_classify/rough/38.hdf5 from FVL-Lab/OpenViTac: direct link, hf CLI and curl.
- Browser
- Download file 118 MB
-
https://huggingface.co/datasets/FVL-Lab/OpenViTac/resolve/main/Simulation/GelSight_mini/roughness_classify/rough/38.hdf5
- Command line
-
hf download hf://datasets/FVL-Lab/OpenViTac/Simulation/GelSight_mini/roughness_classify/rough/38.hdf5
-
curl -L -o 38.hdf5 https://huggingface.co/datasets/FVL-Lab/OpenViTac/resolve/main/Simulation/GelSight_mini/roughness_classify/rough/38.hdf5
118 MB
- Xet hash:
- d5a46db60678d3b189c6890c6fdc4c229e90e0243946a20881d8299992d6967d
- Size of remote file:
- 118 MB
- SHA256:
- 5ca329711db3b59f398cccce50778a2fe1e1b02120379690619c6d7acf9032cf
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.