jacqueline-weng

> Thanks @anxiangsir, I will try this out. Can you also share any insights you have on why ViT architectures do not converge when ArcFace loss is used? They seem...

I found DALI not accelerating the training speed either but it did stabilize the speed. I used to have speed fluctuating around 1000 to 2000 iter/s but with DALI it...

I implemented this function in my code. Make each dataset into the form of mxnet style train.rec and train.idx. When accessing to each index, convert the global index to local...

> I implemented this function in my code. Make each dataset into the form of mxnet style train.rec and train.idx. When accessing to each index, convert the global index to...