New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

LitData Integration #11

Open

tchaton opened this issue Mar 25, 2024 · 1 comment

tchaton commented Mar 25, 2024

Hey there,

I believe it could be possible to remove lot of boilerplate and pain with an integration within LitData: https://github.com/Lightning-AI/litdata

Best,
T.C

Owner

weiji14 commented Apr 3, 2024

Hi @tchaton, yes, I've been keeping an eye on LitData (have packaged it for conda-forge at https://github.com/conda-forge/litdata-feedstock) 😉

The key part I'm trying to demo in this repo is the NVIDIA GPUDirect Storage integration via kvikIO (detailed in this recent blog post - https://developmentseed.org/blog/2024-03-19-combining-cloud-gpu-native), which avoids the CPU-to-GPU transfer bottleneck. I did see that LitData has a geospatial demo at https://lightning.ai/lightning-ai/studios/convert-spatial-data-to-lightning-streaming which reads from GeoTIFF, and will definitely evaluate this for some of my next projects!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment