run_train_07.sh

##!/usr/bin/env bash
#
#DATA="jd"
#loss="bin"
#checkpoints="/opt/intern/users/xunwang/checkpoints"
#r="_model.pth"
##echo 'Batch Norm before FC test'
#mkdir $checkpoints
#mkdir $checkpoints/$loss/
#mkdir $checkpoints/$loss/$DATA/
#
#mkdir result/
#mkdir result/$loss/
#mkdir result/$loss/$DATA/
#
#
#DIM="512"
#alpha_list="30 50"
#for alpha in $alpha_list;do
#    l=$checkpoints/$loss/$DATA/$DIM-Crop-alpha$alpha
##    pre=$l/70_model.pkl
#    mkdir $l
#    CUDA_VISIBLE_DEVICES=4,5,6,7   python train.py -data $DATA  -BN 1  -init random  -lr 1e-5 -dim $DIM -alpha $alpha -num_instances 2 -BatchSize 272 -loss $loss  -epochs 161 -checkpoints $checkpoints  -log_dir $l -save_step 5
#    Model_LIST="40 60 80 100 150 160"
#
#    for i in $Model_LIST; do
#        CUDA_VISIBLE_DEVICES=4,5,6,7  python test.py -data $DATA  -batch_size 280 -r $l/$i$r >>result/$loss/$DATA/$DIM-Crop-alpha$alpha.txt
##        CUDA_VISIBLE_DEVICES=1  python pool_test.py -data $DATA -r $l/$i$r
#    done
#done
#

DATA="shop"
loss="bin"
checkpoints="/opt/intern/users/xunwang/checkpoints"
r="_model.pth"
#echo 'Batch Norm before FC test'
mkdir $checkpoints
mkdir $checkpoints/$loss/
mkdir $checkpoints/$loss/$DATA/

mkdir result/
mkdir result/$loss/
mkdir result/$loss/$DATA/

DIM="512"
alpha_list="40"
for alpha in $alpha_list;do
    l=$checkpoints/$loss/$DATA/$DIM-Crop-alpha$alpha
#    pre=$l/70_model.pkl
    mkdir $l
#    CUDA_VISIBLE_DEVICES=4,5,6,7   python train.py -data $DATA  -BN 1  -init random  -lr 1e-5 -dim $DIM -alpha $alpha -num_instances 5 -BatchSize 260 -loss $loss  -epochs 161 -checkpoints $checkpoints  -log_dir $l -save_step 5
    Model_LIST="40 60 80 100 110 120 130 160"

    for i in $Model_LIST; do
        CUDA_VISIBLE_DEVICES=4,5,6,7  python test.py -data $DATA  -batch_size 280 -r $l/$i$r >>result/$loss/$DATA/$DIM-Crop-alpha$alpha.txt
#        CUDA_VISIBLE_DEVICES=1  python pool_test.py -data $DATA -r $l/$i$r
    done
done