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

fix tasks

Browse Source
This commit is contained in:
yuanmengqi
2025-07-01 15:57:14 +00:00
parent b48c69a2fb
commit 2e3a4a5ba9
8 changed files with 35 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
desktop_env/evaluators/getters/chrome.py
View File

@@ -1444,6 +1444,8 @@ def get_active_tab_url_parse(env, config: Dict[str, Any]):
# change original key to new key, keep value unchange
value = extracted_params.pop(key)
extracted_params[config["replace"][key]] = value
if config.get("split_list", False):
extracted_params = {key: extracted_params[key].split(',') for key in extracted_params.keys()}
return extracted_params

10
desktop_env/evaluators/metrics/general.py
View File

@@ -322,8 +322,14 @@ def check_direct_json_object(result, rules) -> float:
expected_json = rules["expected"]
for key in expected_json.keys():
expected_value = expected_json.get(key)
if expected_value != result.get(key):
return 0.
if expected_json.get("ignore_list_order", False):
expected_value = sorted(expected_value)
result_value = sorted(result.get(key))
if expected_value != result_value:
return 0.
else:
if expected_value != result.get(key):
return 0.
return 1.0
else:
expected_json = rules["expected"]