Designing DNA and protein sequences with improved function has the potential to greatly accelerate synthetic biology. Machine learning models that accurately predict biological fitness from sequence are becoming a powerful tool for molecular design. Activation maximization offers a simple design strategy for differentiable models: one-hot coded sequences are first approximated by a continuous representation which is then iteratively optimized with respect to the predictor oracle by gradient ascent. While elegant, this method suffers from vanishing gradients and may cause predictor pathologies leading to poor convergence. Here, we build on a previously proposed straight-through approximation method to optimize through discrete sequence samples. By normalizing nucleotide logits across positions and introducing an adaptive entropy variable, we remove bottlenecks arising from overly large or skewed sampling parameters. The resulting algorithm, which we call Fast SeqProp, achieves up to 100-fold faster convergence compared to previous versions of activation maximization and finds improved fitness optima for many applications. We demonstrate Fast SeqProp by designing DNA and protein sequences for six deep learning predictors, including a protein structure predictor.