One of the key features of ReinforceUI Studio is the ability to load and analyze pre-trained models. If you have a log folder from a previous training session, you can load the model, visualize its policy performance, and observe the agent executing the task. This feature is especially useful for:
  • Reviewing a specific training session
  • Comparing different models
  • Refining and optimizing your models over time

ReinforceUI Studio Load Pre-Trained Model

​
How to Load a Pre-Trained Model

We recommend using log folders created by ReinforceUI Studio to ensure compatibility.
Follow these simple steps to load a pre-trained model in ReinforceUI Studio:
  1. Click on the “Load Pre-Trained Model” button on the main screen.
  2. Select the log folder containing the pre-trained model.
  3. Automatic Setup:
    • The system will search for the required config file and model file.
    • It will automatically create the environment and load the model.
    • If successful, you will see the platform, environment, and algorithm names in green.
    • If there’s an issue, an error message in red will indicate what went wrong.
  4. Press “Test Policy” to observe the agent executing the task.
That’s it! 🎉 You can now visualize the model’s policy performance, compare it with other models, and refine your training process efficiently.