lzy/sci-gui-agent-benchmark
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

xiaochuan correct his bugs in multiapp examples, you can try it again now

Browse Source
This commit is contained in:
Jason Lee
2024-03-10 14:48:56 +08:00
parent 2291af394f
commit 775cef744f
7 changed files with 97 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
desktop_env/evaluators/getters/file.py
View File

@@ -86,7 +86,6 @@ def get_vm_file(env, config: Dict[str, Any]) -> Union[Optional[str], List[Option
if not config.get("multi", False):
paths: List[str] = [config["path"]]
dests: List[str] = [config["dest"]]
print(config)
if "time_suffix" in config.keys() and config["time_suffix"]:
if "time_format" in config.keys():
time_format = config["time_format"]
@@ -96,8 +95,6 @@ def get_vm_file(env, config: Dict[str, Any]) -> Union[Optional[str], List[Option
else:
paths: List[str] = config["path"]
dests: List[str] = config["dest"]
print(paths)
print(dests)
cache_paths: List[str] = []
@@ -121,9 +118,6 @@ def get_vm_file(env, config: Dict[str, Any]) -> Union[Optional[str], List[Option
cache_paths.append(_path)
with open(_path, "wb") as f:
f.write(file)
# debug
print("cache_paths")
print(cache_paths)
return cache_paths[0] if len(cache_paths)==1 else cache_paths

2
desktop_env/evaluators/metrics/__init__.py
View File

@@ -68,7 +68,7 @@ from .general import (
compare_time_in_speedtest_results,
is_included_all_json_objects,
is_gold_text_included_in_pdf,
check_csv_line_number,
check_line_number,
file_contains,
compare_terminal_and_txt,
fuzzy_place_math,

108
desktop_env/evaluators/metrics/general.py
View File

@@ -1,4 +1,5 @@
import csv
import os
import datetime
import difflib
import functools
@@ -91,23 +92,24 @@ def fuzzy_match(result, rules) -> float:
def fuzzy_place_math(result_file_path, rules) -> float:
if result_file_path is None:
return 0.
expect = rules["expected"] # a list of possible answers
# read list.docx, and get all texts out, overlook blank lines, remove blanks before and after each line
doc = Document(result_file_path)
words_list = []
for para in doc.paragraphs:
words_list.extend(para.text.split())
# Print out the list of extracted words
print("Your Answers: ")
print(words_list)
fuzzy_score_list = []
for word in words_list:
max_score = 0
for ans in expect:
score = fuzz.ratio(word, ans)
score = fuzz.ratio(word, ans)/100
max_score = max(max_score, score)
fuzzy_score_list.append(max_score)
return sum(fuzzy_score_list) / len(fuzzy_score_list)
if len(fuzzy_score_list) != 3:
return 0.
return sum(fuzzy_score_list) / 3
def check_csv(result: str, rules: Dict[str, List[Dict[str, str]]]) -> float:
"""
@@ -304,29 +306,40 @@ def check_direct_json_object(result, rules) -> float:
result = result.replace("'", '"')
# load json object
result = json.loads(result)
print("result: ")
print(result)
print("expected: ")
print(rules["expected"])
if result is None:
return 0.
try:
expect_in_result = rules.get("expect_in_result", False)
if not expect_in_result:
expected_json = rules["expected"]
for key in expected_json.keys():
expected_value = expected_json.get(key)
if expected_value != result.get(key):
return 0.
return 1.0
else:
expected_json = rules["expected"]
expect_in_result = rules.get("expect_in_result", False)
if not expect_in_result:
expected_json = rules["expected"]
for key in expected_json.keys():
expected_value = expected_json.get(key)
if expected_value != result.get(key):
return 0.
return 1.0
else:
expected_json = rules["expected"]
for key in expected_json.keys():
expected_value = expected_json.get(key)
if expected_value not in result.get(key):
return 0.
return 1.0
for key in expected_json.keys():
if isinstance(expected_json.get(key), List):
flag = 0
expected_value_list = expected_json.get(key)
for each_expected_value in expected_value_list:
if each_expected_value in result.get(key):
flag = 1
break
if flag == 0:
return 0.
elif isinstance(expected_json.get(key), str):
if expected_json.get(key) not in result.get(key):
return 0.
else:
logger.debug("check_direct_json_object: expected value type not supported")
return 0.
return 1.0
except:
logger.debug("check_direct_json_object: result is not a valid json object")
return 0.
def compare_time_in_speedtest_results(speedtest_result_path, time_diff):
if not speedtest_result_path:
@@ -397,24 +410,37 @@ def is_gold_text_included_in_pdf(pdf_file_path, gold_text_path):
def file_contains(file_path, config):
# file_path ends with .txt
if not file_path:
return 1
with open(file_path, 'r') as f:
file_text = f.read()
for text in config["expected"]:
if text not in file_text:
return 0
return 1
return 0.
try:
with open(file_path, 'r') as f:
file_text = f.read()
for text in config["expected"]:
if text not in file_text:
logger.debug(f"file_contains: {text} not found in {file_path}")
return 0.
except:
logger.debug("file_contains: file not found or not readable")
return 0.
return 1.
def check_csv_line_number(file_path, line_number):
# check file_path suffix
if not file_path.endswith('.csv'):
return 0
# check line number
with open(file_path, 'r') as f:
reader = csv.reader(f)
line_count = sum(1 for row in reader)
return 1 if line_count == int(line_number["expected"]) else 0
def check_line_number(file_path, line_number):
# check if file_path exists
if file_path is None or not os.path.isfile(file_path):
return 0.
timeRegex = "([01]\d|2[0-3]):[0-5]\d:[0-5]\d"
# check if the string that matches the timeRegex in this txt file equals to line_number["expected"]
try:
with open(file_path, 'r') as f:
line_count = 0
for line in f:
if re.search(timeRegex, line):
line_count += 1
# if line_count equals to line_number["expected"], return 1, else return 0
return 1 if line_count == int(line_number["expected"]) else 0
except:
logger.debug("check_line_number: file not found or not readable")
return 0.
def compare_terminal_and_txt(txt_file_path, terminal_output):

12
evaluation_examples/examples/multi_apps/2373b66a-092d-44cb-bfd7-82e86e7a3b4d.json
View File

@@ -1,7 +1,7 @@
{
"id": "2373b66a-092d-44cb-bfd7-82e86e7a3b4d",
"snapshot": "multiapps",
"instruction": "I want to understand the resource usage of my Ubuntu system under normal workloads. Please use the `sar` command in the `sysstat` toolkit to monitor the system CPU usage, evaluate the status once every second for 30 seconds, output the results to \"System_Resources_Report.txt\" under Desktop, and convert the txt to csv file with the same name.",
"instruction": "I want to understand the resource usage of my Ubuntu system under normal workloads. Please use the `sar` command in the `sysstat` toolkit to monitor system activity, evaluate the status once every second for 30 seconds, output the results to \"System_Resources_Report.txt\" under Desktop.",
"source": "author",
"config": [
{
@@ -17,7 +17,7 @@
"os", "calc"
],
"evaluator": {
"func": ["file_contains", "check_csv_line_number"],
"func": ["file_contains", "check_line_number"],
"result":
[
{
@@ -27,8 +27,8 @@
},
{
"type": "vm_file",
"path": "/home/user/Desktop/System_Resources_Report.csv",
"dest": "System_Resources_Report.csv"
"path": "/home/user/Desktop/System_Resources_Report.txt",
"dest": "System_Resources_Report.txt"
}
],
"expected":
@@ -36,13 +36,13 @@
{
"type": "rule",
"rules" :{
"expected": ["tps", "rtps", "wtps", "dtps", "bread/s", "bwrtn/s", "bdscd/s", "Average", "Linux"]
"expected": ["CPU", "%user","%nice","%system", "%iowait", "%steal", "%idle"]
}
},
{
"type": "rule",
"rules": {
"expected": "33"
"expected": "31"
}
}]
}

18
evaluation_examples/examples/multi_apps/7ff48d5b-2df2-49da-b500-a5150ffc7f18.json
View File

@@ -33,6 +33,24 @@
"chrome", "os", "writer"
],
"evaluator": {
"postconfig":[
{
"type": "activate_window",
"parameters": {
"window_name": "AllLocations.docx - LibreOffice Writer"
}
},
{
"type": "execute",
"parameters": {
"command": [
"python",
"-c",
"import pyautogui; import time; pyautogui.hotkey(\"ctrl\", \"s\"); time.sleep(0.5); pyautogui.press(\"enter\");"
]
}
}
],
"func": "fuzzy_place_math",
"result": {
"type": "vm_file",

8
evaluation_examples/examples/multi_apps/ce2b64a2-ddc1-4f91-8c7d-a88be7121aac.json
View File

@@ -20,7 +20,7 @@
"parameters": {
"files": [
{
"url": "https://drive.usercontent.google.com/download?id=1iquZNI4ktx2915srQ1MMlhcwCuXnmq5T&export=download&authuser=0&confirm=t&uuid=e0bfd86f-a0f1-4a9e-8344-0f645eb379dd&at=APZUnTWQnCd7bNtf8ZN4KfkQUaJ9:1709972645590",
"url": "https://drive.usercontent.google.com/download?id=1XLBhnopqQrTw3bxRWOyfTCslfawFyOaZ&export=download&authuser=0&confirm=t&uuid=f41a09c1-8319-4a50-b8c3-4fc58ba51c00&at=APZUnTV3uiv7wKiSvoCekJG_aVo3:1710051227101",
"path": "/home/user/Pictures/picture2.jpg"
}
@@ -68,9 +68,9 @@
"type": "rule",
"rules":{
"expected":{
"ec076282f61ba74642e94b5a6a1250c6988204d59d9b02936606b6b8ef1e4433": "Kilimanjaro",
"999957c8de835bfa420d198270e7a6b079ee20ff53a3f214491e8134768a7c0b": "Himalayas",
"79f45d40d8413d4e81f1b9734ea39e58622cafd79e12bab32959643fc245147c": "Hua"
"ec076282f61ba74642e94b5a6a1250c6988204d59d9b02936606b6b8ef1e4433": ["Kilimanjaro"],
"6ed4239ecc2be3ec15ad65a78c5c823b9004d640b8cc83a6a7af5930f354de91": ["Himalayas", "Everest", "Sagarmatha"],
"79f45d40d8413d4e81f1b9734ea39e58622cafd79e12bab32959643fc245147c": ["Hua", "hua"]
},
"expect_in_result": true
}

2
evaluation_examples/examples/multi_apps/da922383-bfa4-4cd3-bbad-6bebab3d7742.json
View File

@@ -1,7 +1,7 @@
{
"id": "da922383-bfa4-4cd3-bbad-6bebab3d7742",
"snapshot": "multiapps",
"instruction": "I browsed a lot of interesting blog articles today. I hope to store these articles in my local designated folder just like zotero stores papers. Please download the blogs you are opening now in pdf format to /home/user/Documents/Blogs.",
"instruction": "I browsed a lot of interesting blog articles today. I hope to store these articles in my local designated folder just like zotero stores papers. Please download the blogs opening now in pdf format and save them in their tile to /home/user/Documents/Blogs.",
"source": "authors",
"config": [
{