Feature/webcanvas integration

[han032206]

Pull Request: Integrate WebCanvas Key Node Evaluation and Mind2web-live Benchmark into BrowserGym

Description

This PR officially integrates the WebCanvas key node evaluation and the Mind2web-live benchmark into BrowserGym.

Core Features

• Key Node-Based Evaluation:

  • Implements a key node-based evaluation system to provide detailed assessments of web task processes.

• JavaScript Event Evaluation:

  • Utilizes page JavaScript events to deliver accurate evaluations independent of the action space.

• Debug Modules and Logging:

  • Includes various debug modules and logging functionalities to clearly display the evaluation process.

• Mind2web-live Dataset Integration:

  • Integrates the Mind2web-live dataset to support benchmark testing.

• Community Contributions:

  • Encourages further contributions from the community to expand and refine the evaluation framework and benchmarks.