smp-hub/resnext101_32x4d.swsl

Model card for resnext101_32x4d.

This repository contains the swsl pre-trained weights for the resnext101_32x4d model used as encoder in the segmentation-models-pytorch library.

Example usage:

1. Install the library:

pip install segmentation-models-pytorch

2. Use the encoder in your code:

import segmentation_models_pytorch as smp

model = smp.Unet("resnext101_32x4d", encoder_weights="swsl")

References

• Github: https://github.com/qubvel/segmentation_models.pytorch
• Docs: https://smp.readthedocs.io/en/latest/
• Original weights URL: https://dl.fbaipublicfiles.com/semiweaksupervision/model_files/semi_weakly_supervised_resnext101_32x4-3f87e46b.pth