Skip to content

mfai documentation

mfai documentation

Installation
Quickstart
Use Cases
NamedTensor

Models
- ModelABC
- ArchesWeather
- BaseModel
- DeepLabV3
- DeepLabV3Plus
- HalfUNet
- IdentityModel
- BaseGraphModel
- BaseHiGraphModel
- GraphLAM
- HiLAM
- HiLAMParallel
- PanguWeather
- Segformer
- SwinUNetR
- CustomUNet
- UNet
- UNetRPP
- ViTClassifier
- VitEncoder
Losses
Transforms
Metrics
- CSINeighborhood
- FAR
- FNR
- FSS
- PR_AUC
Lightning
Reference
- mfai
  - encoding
  - http
  - pytorch
    - callbacks
    - dummy_dataset
    - lightning_modules
    - losses
    - lr_scheduler
    - metrics
    - models
      - archesweather
      - base
      - clip
      - deeplabv3
      - gan_dgmr
        
        blocks
        
        discriminators
        
        generators
        
        layers
        
        attention
        
        conv_gru
        
        coord_conv
      - half_unet
      - identity
      - llms
        
        cross_attention
        
        fuyu
        
        gpt2
        
        llama2
        
        llama3
      - nlam
        
        create_mesh
        
        interaction_net
      - pangu
      - resnet
      - segformer
      - swinunetr
      - unet
      - unetrpp
      - utils
      - vit
      - weather_projector
    - namedtensor
    - padding
    - transforms
  - tensorflow
  - tokenizers

About
Acknowledgements
Contributing

mfai.pytorch.losses¶

Copyright © 2026, Météo-France AI Lab

Made with Sphinx and @pradyunsg's Furo