File size: 5,299 Bytes
c4ef4f9
5f30d29
c4ef4f9
 
 
 
bb690a1
 
c4ef4f9
 
 
 
 
5f30d29
bb690a1
5f30d29
c4ef4f9
bb690a1
05e4fe8
c4ef4f9
bb690a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c4ef4f9
 
bb690a1
5f30d29
bb690a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c4ef4f9
 
 
 
bb690a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c4ef4f9
 
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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
# import imp
from email.policy import default
import streamlit as st
import pandas as pd
import numpy as np
import time
from config import  update_device_func
from parse_code import parse_base_code


# title
st.title("Exp Command Generator")

# experiment mode
# exp_mode = st.sidebar.selectbox("Select Experiment Mode", ["OneExpOnecard", "MultipleExpOnecard"],key="MultipleExpOnecard")

## 检查框
debug = st.sidebar.checkbox("Debug: 选择则会串行地执行命令", value=True)
# st.sidebar.write(f"checkbox的值是{res}")

# setup = st.sidebar.text_area("Some setup of env at beginning.", """cd $(dirname $(dirname $0))
# source activate xai
# export PYTHONPATH=${PYTHONPATH}:/Users/apple/Desktop/workspace/research_project/attention:/mnt/yixin/:/home/yila22/prj""")

# exp_hyper = st.sidebar.text_area("Hyperparameters", """exp_name="debug-adv-training-emotion"
# dataset=emotion
# n_epoch=3
# K=3
# encoder=bert
# lambda_1=1
# lambda_2=1
# x_pgd_radius=0.01
# pgd_radius=0.001
# seed=2
# bsize=8
# lr=5e-5""")

## gpu 相关参数
gpu_list = st.sidebar.multiselect("multi select", range(10), [0,1,2,3,4,])
# print(gpu_list)
allow_gpu_memory_threshold_default=5000
gpu_threshold_default=90
total_gpu_memory = st.sidebar.number_input("单卡总容量", value=24564, min_value=0, max_value=30000, step=1000)
max_gpu_memory_gap = st.sidebar.number_input("最小单卡剩余容量", value=allow_gpu_memory_threshold_default, min_value=0, max_value=total_gpu_memory, step=500)
max_gpu_utilization = st.sidebar.number_input("最大单卡利用率", value=gpu_threshold_default, min_value=0, max_value=100, step=10)
sleep_time_after_loading_task= st.sidebar.number_input("加载任务后等待秒数", value=10, min_value=0,step=5)
# all_full_sleep_time = st.sidebar.number_input("全满之后等待秒数", value=20, min_value=0,step=5)
username = st.sidebar.text_input("用户名", value="yila22")
cpu_max_utility = st.sidebar.number_input("cpu最大利用率", value=77, min_value=0, max_value=100, step=1)
memory_max_utility = st.sidebar.number_input("内存最大利用率", value=80, min_value=0, max_value=100, step=1)
constrain_total = st.sidebar.checkbox("限制总资源", value=True)
constrain_mine = st.sidebar.checkbox("限制我的资源", value=False)
constrain_rate = st.sidebar.number_input("限制率", value=2, min_value=1, max_value=10, step=1)

# username_mine=root
# max_gpu_utilization=90
# total_gpu_memory=24564
# max_gpu_memory_gap=5000
# available_devices=( 0 1 2 3 4 5 6 7 8 9 )
# current_device_idx=-1
# sleeptime=30
# cpu_mean_max=77
# memory_rate_max=80
# constrain_total=true
# constrain_mine=false
# constrain_rate=2
gpu_list = " ".join([str(i) for i in gpu_list])
setup_for_gpu_utility = f"""
username={username}
max_gpu_utilization={max_gpu_utilization}
total_gpu_memory={total_gpu_memory}
max_gpu_memory_gap={max_gpu_memory_gap}
available_devices=( {gpu_list} )
current_device_idx=-1
sleeptime={sleep_time_after_loading_task}
cpu_mean_max={cpu_max_utility}
memory_rate_max={memory_max_utility}
constrain_total={constrain_total}
constrain_mine={constrain_mine}
constrain_rate={constrain_rate}
"""


base_code = st.text_area("Base Code", """##### setup
export CUDA_VISIBLE_DEVICES=2
source activate /data/yixin/anaconda/mib
exp_name="single_user"
#####

##### loop
for poison_method in char_basic word_basic sent_basic; do
for dataset_idx in 0 1 2; do
#####

##### main
  python single_user.py --dataset_idx $dataset_idx --trigger_size 1 --target 0 \
   --loc 0 --batch_size 16 --num_epochs 2 --poison_method $poison_method --lr 5e-5 --pattern 0 --exp_name $exp_name \
    --log_wb
#####

#####
done;done;
#####""", height=400)



g = st.button("Generate")
if g:
    st.success("Finished")
    contents = base_code
    gpu_utility = ""
    gpu_utility = setup_for_gpu_utility + "\n\n" + update_device_func 
    
    
    new_code = parse_base_code(contents, debug=debug)
    
    # create file for download 
    timestr = time.strftime("%Y%m%d-%Hh%Mm%Ss")
    import os 
    os.makedirs(f"./res/{timestr}", exist_ok=True)
    filename_script = f"./res/{timestr}/script.sh"
    with open(filename_script, "w") as f:
        f.write(new_code)
    filename_config = f"./res/{timestr}/gpu_utility.sh"
    with open(filename_config, "w") as f:
        f.write(gpu_utility)
    
    # zip them into one file
    # import shutil
    # shutil.make_archive(f"./res/{timestr}", 'zip', f"./res/{timestr}")
    # st.download_button(
    #     label="Download zip",
    #     data=f"./res/{timestr}.zip",
    #     file_name=f"{timestr}.zip",
    #     mime="application/zip",
    # )
    
    
    st.download_button(
        label="Download script",
        data=new_code,
        file_name=filename_script,
        mime="text/plain",
    )
    # after clicking i don't want the website to refresh
    st.download_button(
        label="Download gpu_utility.sh",
        data=gpu_utility,
        file_name=filename_config,
        mime="text/plain",
    )
    
    # st.markdown(f"### [Download script](./{filename_script})")
    # st.markdown(f"### [Download gpu_utility.sh](P{filename_config})")
    st.markdown("## script.sh")
    st.code(new_code, language="shell")
    
    
    st.markdown("## gpu_utility.sh")
    st.code(gpu_utility, language="shell")