Add 'WAIT', 'FAIL', 'DONE' to the action space; Debug basic prompting-based GPT-4 and Gemini agents; Initialize experiments script;

mm_agents/gpt_4v_prompt_code.py

@@ -4,5 +4,8 @@ For each step, you will get an observation of an image, which is the screenshot
 
 You are required to use `pyautogui` to perform the action. 
 Return one line or multiple lines of python code to perform the action each time, be time efficient.
-Return `None` if you cannot perform the action.
+
+When you think you have to wait for some time, return `WAIT`.
+When you think the task can not be done, return `FAIL`.
+When you think the task is done, return `DONE`.
 """