Graphical User Interface (GUI)

Welcome to the GUI. This interface is designed to provide you with a convenient and straightforward user experience. Simply configure your user information and parameters within the GUI, and it will return the results for you. Please see the user manual below for detailed instructions.

Main Component

To run the GUI correctly, you must place the EasyCO project folder in the proper directory on your remote server. Then, execute the GUI’s Python script run_gui_test.py on your local machine (an operating system that can render a graphical interface, typically Windows).

• Menu Bar

  • github: Redirects to the EasyCO GitHub repository.

  • documention: Redirects to the EasyCO documentation page for user guides and references.

• Server Setup

• host: The server’s IP address or domain name to connect to.

• port: The specific port number through which the connection is established.

• username: The user identity for logging into the server.

• login method: The authentication method, either key-based or password-based.

• key: The path to the private key file used for secure login (only if key-based method is chosen).

• device: The target computation device on the server (e.g., CPU, GPU such as cuda:0).

• connect: The button to initiate the connection to the server with the provided settings

• Problem

• Problem: click the problem you want to solve in the Scroll Bar Frame.

• scale: If you tick the box on the left, it means you enable the vary scale function. Enter the range correctly(min-max). In this case, a dataset containing multiple scales will be provided to the solver, and the solver will filter out the instances within the specified range and solve them.

• capacity: If you check the box on the left, it enables the vary capacity function. Enter the range correctly(min-max). different capacity values will be provided to the solver, and the solver will select the instances satisfiy the specified range and solve them.

• Solver

💡Tips: To ensure Solver are displayed properly, please select Problem first.

• Solver: Click the solver you want to apply for the problem in the Scroll Bar Frame. Please select problem before you select solver.

• Save: Click Save button to view historical test results. A pop-up will appear where you can select the results you wish to visualize. The system will automatically create a local backup of the log files (located in the GUI/log_temp/ directory) for your convenience.
  insert picture here: tick the left frame that you want.

• Plot: Click Plot button to see the visualized results. insert picture here: In the Select Metric dropdown, you can choose the metric you want to visualize:Score, Augmented Score, Gap, or Augmented Gap. You can also customize the plotting parameters in the Plot Parameters sidebar.

• Parameter

💡Tips: Please select Problem and Solver first to display the correct Parameter.

• Parameter: The settings for specific Solver would be shown in this frame, at least you choose Problem and Solver properly. insert picture here: you can modified settings_parameters for Solver directly in GUI.

• Test Parameter

💡Tips: Please select Problem and Solver first to display the correct Method and Task.

• Method: Your selection of Method would be displayed here.

• Task: Your selection of Task would be displayed here.

• mode: This is a dropdown menu to select the mode. The options are Test and Training, with Test being the default setting.

• Batch_size: The number of instances in a batch.

• Episode: The total number instances in an epsiode.

• Decoder_Strategy: The Decoder Strategy, which includes Greedy Search and Sampling, determines how the next token is selected from the decoder’s output probability distribution to construct the final sequence, balancing between precision and diversity.

• Datasets: Click the Dataset button, select a file in the pop-up window, and the file’s location will be displayed in the entry field. When conducting a vary-scale test, you can select an entire folder instead of a single file.

• Checkpoint: Click the Checkpoint button, select a file in the pop-up window, and the file’s location will be displayed in the entry field.

• Start: Click the Start button to start the process.

• Stop: Click the Stop button to stop the process. > 💡Tips: Please click Stop button to terminate the last Start and reset settings before initiating the next Start.

• Epoch(only for training): The number of epoch for training process.

• Curve(only for training): Click the Checkpoint button, you can trace training process in visualized results. There are three options available in the drop-down box: loss score, score and eval_score.

• Log File Output

During the Test and Training processes, the terminal logs will be synchronously updated in the “Log File Output” area, facilitating process tracking.

• Summary Output

💡Tips: Detailed parameter information will be displayed in the Summary Output section, where you can verify whether your parameters are set correctly.

Source	Description
cfg	You can refer to the parameter in configs.yaml.
settings	You can refer to the parameter in xxx_settins.yaml.

Instructions for Quick Start

1. Connect the remote Server first. When a successful connection to the remote server is established, you can view the prompt message in the Log File Output section.

2. Choose the Problem you want to solve in Problem section. tick the box when you want select vary-scale or vary-capacity function. (💡 left: min value, right: max value)

3. Choose the Solver you want to apply in Solver section.

4. You can customize parameter settings in the “Parameter” section; alternatively, you can leave the default settings intact without making any adjustments.

5. modify Test Parameter or Training Parameter carefully. make sure the location of datasets and checkpoint is correct.

6. click start button to start process.

7. click stop process to terminate process.