adambielski / siamese-triplet

Siamese and triplet networks with online pair/triplet mining in PyTorch
BSD 3-Clause "New" or "Revised" License
3.09k stars 634 forks source link

batch_size #69

Closed PangXuejiao closed 2 years ago

PangXuejiao commented 2 years ago

Why do I always lose the last batch_size when I use enumerate to load data from train_loader and test_Loader? (Originally batch_size=36, only loaded to 35, but can be divisible.) could you help me ?? thanks!!