ContainersNVIDIADeepSeek-R1NVIDIALlama-3.1-Nemotron-70B-InstructNVIDIAPyTorch
CollectionsNVIDIAOmniverse Kit (FB)NVIDIADeepStream SDKNVIDIAOmniverse Kit App Streaming
ModelsNVIDIAStyleGAN3 pretrained modelsNVIDIAPeopleNetNVIDIATrafficCamNet
ResourcesNVIDIARiva Skills Quick Start
Helm ChartsNVIDIAGPU OperatorNVIDIANVIDIA NIM Operator
NVIDIA
NVIDIA
Earth-2 CorrDiff US
Model
NVIDIA
NVIDIA
Earth-2 CorrDiff US

Corrector Diffusion (CorrDiff) US GEFS-HRRR is a generative downscaling model for the contiguous United States.

NVIDIA Earth-2 InferenceNVIDIA Developer ProgramNVIDIA AI Enterprise
NVIDIA AI Enterprise SupportedNVIDIA NIM
This model is backed by NVIDIA's Plus Plus (++) Promise
to learn more about the quality of the datasets used to train this model.
FieldResponse
Intended Application & Domain:Global Weather Forecasting
Model Type:Diffusion Model
Intended User:Climate and Weather scientists accelerating weather prediction with AI.
Output:Downscaled inputs.
Describe how the model works:Diffusion model progressively adds noise to interpolated input to recreate fine-resolution features.
Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of:Not Applicable
Technical Limitations:The model may perform poorly for data sources different from that in the ERA5 training dataset
Verified to have met prescribed NVIDIA quality standards:Yes
Performance Metrics:Accuracy, Throughput and Latency
Potential Known Risks:This model may mispredict fine-resolution features.
Licensing:NVIDIA AI Foundation Models Community License