-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrun_implicit.sh
executable file
·44 lines (37 loc) · 1.65 KB
/
run_implicit.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
#!/bin/bash
# Directory
dir="ImplicitEquiVSetFlax"
# Possible values for data_name
data_names=("gaussian" "celeba") #"amazon" "moons"
# Possible values for amazon_cat
amazon_cats=("apparel" "bedding" "carseats" "diaper" "feeding" "gear" "media" "bath" "health" "toys" "furniture" "safety" ) #
# Learning rates
learning_rates=("0.01" "0.001" "0.0001" "0.00001") #
# Number of repetitions
folds=5
fwd_tol=1e-6
fwd_maxiter=20
num_layers=3
# Loop over each data_name
for data_name in "${data_names[@]}"; do
if [ "$data_name" == "amazon" ]; then
# Loop over each amazon_cat
for amazon_cat in "${amazon_cats[@]}"; do
for lr in "${learning_rates[@]}"; do
for ((i=1; i<=folds; i++)); do
fold=$i
echo "Running $dir/main_flax.py with --data_name=$data_name --amazon_cat=$amazon_cat --lr=$lr --norm=fro --fold=$fold (Run $i) --fwd_tol=$fwd_tol --fwd_maxiter=$fwd_maxiter --num_layers=$num_layers"
(cd "$dir" && CUDA_VISIBLE_DEVICES=1 python main_flax.py --data_name="$data_name" --amazon_cat="$amazon_cat" --lr="$lr" --fold=$fold --norm=fro --fwd_tol=$fwd_tol --fwd_maxiter=$fwd_maxiter --num_layers=$num_layers)
done
done
done
else
for lr in "${learning_rates[@]}"; do
for ((i=1; i<=folds; i++)); do
fold=$i
echo "Running $dir/main_flax.py with --data_name=$data_name --lr=$lr --fold=$fold (Run $i) --fwd_tol=$fwd_tol --fwd_maxiter=$fwd_maxiter --num_layers=$num_layers"
(cd "$dir" && CUDA_VISIBLE_DEVICES=1 python main_flax.py --data_name="$data_name" --lr="$lr" --fold=$fold --fwd_tol=$fwd_tol --norm=fro --fwd_maxiter=$fwd_maxiter --num_layers=$num_layers)
done
done
fi
done