train_script.sh

WORK_DIR=~/Workspace/complex-networks-release   #repo directory
DATASET_DIR=/home_local/ekcole/knee_data        #where the knee dataset was downloaded to
MASKS_PATH=/home_local/ekcole/knee_masks        #where the masks were generated by BART

TYPE=complex        #real or complex convolution
ITERATIONS=4        #number of unrolled iterations
FEAT=128            #number of feature maps
ACTIVATION=relu  #activation function

LOG_DIR="f"$FEAT"_g"$ITERATIONS

# training
python3 $WORK_DIR/train_loop.py \
    --train_dir $TYPE"_"$ACTIVATION \
    --mask_path $MASKS_PATH \
    --dataset_dir $DATASET_DIR \
    --log_root $LOG_DIR \
    --shape_z 256 --shape_y 320 \
    --num_channels 8 \
    --batch_size 2 \
    --device 0 \
    --max_steps 10000 \
    --feat_map $FEAT \
    --num_grad_steps $ITERATIONS \
    --activation $ACTIVATION \
    --conv $TYPE

# testing
# python3 $WORK_DIR/test_loop.py --train_dir $TYPE \
#     --shape_z 256 --shape_y 320 \
#     --batch_size 1 \
#     --feat_map $FEAT \
#     --num_grad_steps 4 \
#     --mask_path $MASKS_PATH \
#     --dataset_dir $DATASET_DIR \
#     --device 0 \
#     --log_root $LOG_DIR \
#     --num_channels 8 \
#     --activation $ACTIVATION \
#     --gpu single \
#     --conv $TYPE