use train_network.py to training lora how to get the remaining completion time in real time

when i use nodejs to run this script :
```
// Import the 'exec' function from the 'child_process' module
const { exec } = require('child_process');

// Execute the shell command and capture the stdout
exec('sh train.sh', { maxBuffer: 1024 * 10000 }, (error, stdout, stderr) => {
if (error) {
console.error(`Error executing the script: ${error}`);
return;
}

// Log any errors captured in stderr
console.log('Error stderr:', stderr);

// Log the stdout (script output) in real-time
console.log('Script output:', stdout);

// Clear the 'scriptProcess' variable if necessary
scriptProcess = null;
});
```
I only can get the all output and stdout in this callback for completion this training, how to get the progress in real time ?

Review train.sh and train_network.py, then inspect how Node.js child_process.exec handles output while the command is running. Determine whether the training script exposes progress or timing information before completion. Done means documenting a real-time progress approach and explaining how remaining time can be obtained, if available.

Written by the indexing model from the issue text.