Hi all,
As per object, I’d like to know if there is a PyTorch equivalent of Tensorflow Records. I’m pretty sure that image loading (my images are high-resolution JPEG) consumes the most time during training, so I’d like to know if there are alternative ways to store my data which can lead to significant speedups. I’m using GitHub - open-mmlab/mmdetection: OpenMMLab Detection Toolbox and Benchmark
Thanks in advance for your support,
Andrea