run operagor

evaluation_examples/examples/chrome/9f935cce-0a9f-435f-8007-817732bfc0a5.json

@@ -57,5 +57,5 @@
       }
     }
   },
-  "proxy": true
+  "proxy": false
 }

evaluation_examples/examples/chrome/a728a36e-8bf1-4bb6-9a03-ef039a5233f0.json

@@ -56,5 +56,5 @@
       }
     }
   },
-  "proxy": true
+  "proxy": false
 }

evaluation_examples/settings/proxy/dataimpulse.json

@@ -2,8 +2,8 @@
     {
         "host": "gw.dataimpulse.com",
         "port": 823,
-        "username": "your_username",
-        "password": "your_password",
+        "username": "e750e5abb74376d28361",
+        "password": "e5ec245537e1e76a",
         "protocol": "http",
         "provider": "dataimpulse",
         "type": "residential",

monitor/.env

@@ -4,11 +4,11 @@
 # Monitor configuration
 TASK_CONFIG_PATH=../evaluation_examples/test_all.json
 EXAMPLES_BASE_PATH=../evaluation_examples/examples
-RESULTS_BASE_PATH=../results_all
+RESULTS_BASE_PATH=../results_operator_full_test_0713
 ACTION_SPACE=pyautogui
 OBSERVATION_TYPE=screenshot
 MODEL_NAME=computer-use-preview
-MAX_STEPS=150
+MAX_STEPS=100
 FLASK_PORT=80
 FLASK_HOST=0.0.0.0
 FLASK_DEBUG=true

run_operator.sh

@@ -0,0 +1,9 @@
+python run_multienv_openaicua.py \
+--headless \
+--observation_type screenshot \
+--model computer-use-preview \
+--result_dir ./results_operator_full_test_0713 \
+--test_all_meta_path evaluation_examples/test_all.json \
+--max_steps 100 \
+--num_envs 15 \
+--provider_name aws

show_result.py

@@ -68,4 +68,4 @@ def get_result(action_space, use_model, observation_type, result_dir):
 
 
 if __name__ == '__main__':
-    get_result("pyautogui", "gpt-4o", "a11y_tree", "./results")
+    get_result("pyautogui", "computer-use-preview", "screenshot", "./results_operator_full_test_0713")