running training num train images * repeats: 44 num reg images: 0 num batches per epoch: 44 num epochs: 57 batch size per device: 1 gradient accumulation steps: 1 total optimization steps: 2500
running training num train images * repeats: 44 num reg images: 0 num batches per epoch: 44 num epochs: 57 batch size per device: 1 gradient accumulation steps: 1 total optimization steps: 2500