μFormer Encoder

This item primarily hosts the checkpoint and some sampled data for training μFormer (or Mu-Former, uFormer, MuFormer, for readability), a potent tool tailored for predicting the effects of protein mutations. It is configured to facilitate the replication of the models presented in the paper titled Accelerating protein engineering with fitness landscape modeling and reinforcement learning which can be accessed at this link. Please adhere to the CC BY-NC (Attribution-NonCommercial) license terms when using these files.