Accelerated by another 5% using FP16/32 Batch-norm for Tensor Cores.

AlexeyAB

2018-04-16 9bae70b22549b68f5cdeece8b6c3b3de00c22714

1
2
3
4
5
6
7
8
classes=9418
train  = data/9k.train.list
valid  = /data/imagenet/imagenet1k.valid.list
leaves = data/imagenet1k.labels
backup = /home/pjreddie/backup/
labels = data/9k.labels
names  = data/9k.names
top=5