Wxy/opencua (#274)

* OpenCUA Agent code base

* update url

* debug, modify url input

* debug opencua

* show result

* debug agent history overlap

* modify opencua agent; add comment lines

* update parallel; clean code; use sleep 3s

* ui-tars-0717
This commit is contained in:
Xinyuan Wang
2025-07-20 15:52:23 +08:00
committed by GitHub
parent bec7129fff
commit e10dd9267c
5 changed files with 320 additions and 224 deletions

View File

@@ -172,9 +172,7 @@ def run_single_example_opencua(agent, env, example, max_steps, instruction, args
action_timestamp = datetime.datetime.now().strftime("%Y%m%d@%H%M%S")
logger.info("Step %d: %s", step_idx + 1, action)
obs, reward, done, info = env.step(action)
time.sleep(3)
obs = env._get_obs()
obs, reward, done, info = env.step(action, args.sleep_after_execution)
logger.info(f"Action {action} executed, reward: {reward}, done: {done}")
# Save screenshot and trajectory information