From 741e26c3f8c08fc59e3d81bf92bdeb92072368b6 Mon Sep 17 00:00:00 2001
From: Timothyxxx <384084775@qq.com>
Date: Wed, 13 Mar 2024 23:35:04 +0800
Subject: [PATCH] Update

---
 experiment_screenshot.py  | 7 +++++++
 mm_agents/gpt_4v_agent.py | 1 +
 2 files changed, 8 insertions(+)

diff --git a/experiment_screenshot.py b/experiment_screenshot.py
index 0fb875d..58a2f94 100644
--- a/experiment_screenshot.py
+++ b/experiment_screenshot.py
@@ -1,3 +1,4 @@
+# todo: unifiy all the experiments python file into one file
 import datetime
 import json
 import logging
@@ -114,23 +115,29 @@ def run_one_example(example, agent, max_steps=10, example_trajectory_dir="exp_tr
 
     try:
         func_timeout.func_timeout(120, stop_recording)
+        # todo: make sure we got the video file, check the bug
     except func_timeout.exceptions.FunctionTimedOut:
         logger.info("Recording timed out.")
 
     result = env.evaluate()
     logger.info("Result: %.2f", result)
 
+
+    # fixme: change to write the result into a separate file
     with open(trajectory_recording_path, "a") as f:
         f.write(json.dumps({
             "result": result
         }))
         f.write("\n")
 
+    # todo: append the result to the wandb for visualization
+
     # env.close()
     logger.info("Environment closed.")
 
 
 def main(example_class, example_id, gpt4_model="gpt-4-vision-preview"):
+    # fixme: change all the settings like action_space, model, etc. to the argparser
     action_space = "pyautogui"
     gemini_model = "gemini-pro-vision"
 
diff --git a/mm_agents/gpt_4v_agent.py b/mm_agents/gpt_4v_agent.py
index d2d6a9e..7e9c400 100644
--- a/mm_agents/gpt_4v_agent.py
+++ b/mm_agents/gpt_4v_agent.py
@@ -30,6 +30,7 @@ from mm_agents.prompts import SYS_PROMPT_IN_SCREENSHOT_OUT_CODE, SYS_PROMPT_IN_S
     SYS_PROMPT_SEEACT, ACTION_DESCRIPTION_PROMPT_SEEACT, ACTION_GROUNDING_PROMPT_SEEACT
 
 import logging
+# todo: cross-check with visualwebarena
 
 logger = logging.getLogger("desktopenv.agent")