Accurate prediction of the effects of genetic variation is a major goal in biological research. Towards this goal, numerous machine learning models have been developed to learn information from evolutionary sequence data. The most effective method so far is a deep generative model based on the variational autoencoder that models the distributions using a latent variable. In this study, we propose a deep autoregressive generative model based on a Temporal Convolutional Network architecture, which employs dilated causal convolutions and attention mechanism for the modeling of inter-residue correlations in a biological sequence. We show that this model is competitive with the variational autoencoder model when tested against a set of 42 deep mutational scan experiments. In particular, our model can more efficiently capture information from multiple sequence alignments with lower effective number of sequences, such as in viral sequence families, compared to the latent variable model. Also, we extend this architecture to a semi-supervised learning framework, which shows high prediction accuracy. We show that our model enables a direct optimization of the data likelihood and allows for a simple and stable training process.