-
Notifications
You must be signed in to change notification settings - Fork 0
/
run_experiment_by.sh
executable file
·36 lines (26 loc) · 1.27 KB
/
run_experiment_by.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
# #!/bin/bash
# # Seeds array
# seeds=(2222 42 43 44 45 46 47 48 49 50)
# # Output file
# output_file="experiment_output.txt"
# # Loop over seeds
# for seed in "${seeds[@]}"
# do
# echo "Running experiment with seed $seed"
# echo "Running experiment with seed $seed" >> $output_file
# /mnt/nas/share2/home/by/envs/hndna/bin/python train.py experiment='hg38/nt_ablation' wandb=null train.seed=$seed >> $output_file
# done
# echo "All experiments completed."
#!/bin/bash
# 定义一个数组,包含所有的数据集名称
datasets=("enhancer" "enhancer_types" "H3" "H3K4me1" "H3K4me2" "H3K4me3" "H3K9ac" "H3K14ac" "H3K36me3" "H3K79me3" "H4" "H4ac" "promoter_all" "promoter_non_tata" "promoter_tata" "splice_sites_acceptor" "splice_sites_donor")
# 循环遍历数组中的每个元素
for dataset in ${datasets[@]}; do
# 打印当前正在处理的数据集名称
echo "Processing ${dataset}..."
# 运行命令,并将输出重定向到以数据集名称命名的.txt文件中
nohup /mnt/nas/share2/home/by/envs/hndna/bin/python train.py experiment='hg38/hg38_bert' wandb=null dataset.dataset_name="${dataset}" train.seed=48 trainer.devices=4 dataset.batch_size=128 > "${dataset}.txt" &
# 等待当前命令完成
wait
done
echo "All datasets processed."