SleepTransformer model weights

This is the model weights of the SleepTransformer trained on the SHHS data described in the paper:

Huy Phan, Kaare Mikkelsen, Oliver Y. Chén, Philipp Koch, Alfred Mertins, and Maarten De Vos. SleepTransformer: Automatic Sleep Staging With Interpretability and Uncertainty Quantification. IEEE Transactions on Biomedical Engineering (TBME), vol. 69, no. 8, pp. 2456-2467, 2022

Model configurations:

• sequence length: 21
• number Transformer encoder blocks in the epoch encoder: 4
• number Transformer encoder blocks in the sequence encoder: 4

This model should be used in combination with the code available here: https://github.com/pquochuy/SleepTransformer