Yeaminul

icon location Dhaka, Bangladesh icon location | Passionate Engineer | Machine Learning Practitioner | Data Enthusiast | Fast Learner |

Results 1 issues of Yeaminul

Possibility of adding Swish/Hard Swish Non linearity in the earlier layers

Hi, Clean and excellent implementation. Can we have swish non linearity in place of the default leaky ReLU in the earlier layers of the feature extractor backbone?