FSCIL-via-Entropy-regularized-DF-Replay This is the official implementation of Few-Shot Class-Incremental Learning via Entropy-Regularized Data-Free Replay.