weh
weh
Hi, is there any reason why this change is never merged into the repo? I think from a readibility point of view, having YAML config and `IAM_POLICY_FILE` will be of...
Many thanks for the help! Just to confirm: does the 8 minutes stated in the paper refer to time taken to: 1) generate single 30 steps out forecast i.e (+360h)...
Thank you! I haven't yet seen an increase in run time tbh, but will re-run some experiments and see if I actually made mistakes, will get back to you!
@andrewlkd Would like to ask if precipitation inputs for Gencast will be set to NaN as in the case for Graphcast? or do both models actually take in different inputs?
Says in the doc here you will need ~60GB for inference on GPU: https://github.com/google-deepmind/graphcast/blob/main/docs/cloud_vm_setup.md#running-inference-on-gpu I managed to get it run on a H100, which has 80GB vRAM cap and can...
@alvarosg presumably `nan` on land is set through this function? or users still have to set the corresponding coordinates to `nan` by themselves? ``` from graphcast import nan_cleaning predictor =...
that is clear, thanks! @andrewlkd