AdvancedDigitalSignalProcessingCourse-FinalTermProject/train.sh at main · swoonge/AdvancedDigitalSignalProcessingCourse-FinalTermProject · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
#!/bin/sh

# following is training
#time sudo python3 train.py -f tiny-imagenet-200/train/n12267677/images -g
#time sudo python3 train.py -f tiny-imagenet-200/train/n01443537/images -g
#sudo python3 train.py -f tiny-imagenet-200/train/n01443537/images -g
# sudo python train.py -f data/tiny-imagenet-200/test/images

# following is encoding
#sudo python3 encoder.py --model checkpoint/encoder_epoch_00000001.pth --input tiny-imagenet-200/test/images/test_9970.JPEG --cuda --output ex --iterations 16
#sudo python3 encoder.py --model checkpoint/encoder_epoch_00000001.pth --input kodim24.png --cuda --output ex --iterations 16
#sudo python3 encoder.py --model checkpoint/encoder_epoch_00000001.pth --input kodim10.png --cuda --output ex --iterations 16

# following is decoding
#sudo python3 decoder.py --model checkpoint/decoder_epoch_00000001.pth --input ex.npz --cuda --output output

python train.py --train /path/to/training/images -N 64 -e 1000 --lr 0.002 -f data/tiny-imagenet-200/test/images --random_seed 0
python train.py --train /path/to/training/images -N 64 -e 1000 --lr 0.001 -f data/tin       y-imagenet-200/test/images --random_seed 0
python train.py --train /path/to/training/images -N 64 -e 1000 --lr 0.0005 -f data/tiny-imagenet-200/test/images --random_seed 0