google-deepmind / google-deepmind/weathernext
Erros with HRES: NaNs / Cannot reshape array
Open
- Dominant language
- Python
- Stars
- 7.7k
- Forks
- 986
- PR merge metrics
- No merged PRs in 30d
Description
Hi,
I am running with HRES data.
ai-models --retrieve-requests graphcast
Shows which request to use to download operational stream. This is successful, but downloads 2 files.
Run the model
ai-models --debug --only-gpu --input file --file 06.grib --file 12.grib --date 20240624 --time 1200 --lead-time 240 graphcast
This runs and creates:
Training_xarray.nc
Input_xr.nc
Forcings_xr.nc
Output.nc # full of nans
The error
NaN values found in field. Args=() kwargs={'template':GribField(2t,None,20240624,600,0,0),'step':6}
How should i run the command with operational HRES?
Thank you
Contributor guide
Assessment
This issue has not been assessed yet.