How can Transformers handle variable length input?Exposing one of the most untapped properties of self-attentionMar 3, 20242Mar 3, 20242
PyTorch BatchNorm2D Weights ExplainedUnderstanding PyTorch BatchNorm2D and its weightsJul 17, 2022Jul 17, 2022
Published inTDS ArchivePytorch Conv2d Weights ExplainedUnderstanding weights dimension, visualization, number of parameters and the infamous size mismatchNov 26, 20211Nov 26, 20211
Published inTDS ArchiveHuman Priors in Object DetectionWhat we tell the AI determines how it learnsMar 9, 2021Mar 9, 2021
Published inTDS ArchiveTensorFlow 2.1: A How-ToThe keras mode, the eager mode and the graph modeApr 19, 20201Apr 19, 20201
Published inTDS ArchiveKeras Custom Training LoopHow to build a custom training loop in Keras at a lower level of abstraction, K.function, get_updates usage and other stuff under the hoodOct 19, 20193Oct 19, 20193
How to write a linux device driverA beautiful guide for the Hello World of the device driver programmingAug 20, 20175Aug 20, 20175