refactor: improve checkpoint naming and exit code logic

[TEGRAXD]

Removing hard-coded 2333333 in latest saved checkpoint model. This make the code more logical and informative.

Pull request checklist

• [x] The PR has a proper title. Use Semantic Commit Messages. (No more branch-name title please)

• [x] Make sure this is ready to be merged into the relevant branch. Please don't create a PR and let it hang for a few days.

• [x] Ensure you can run the codes you submitted succesfully. These submissions will be prioritized for review:

  Introduce improvements in program execution speed;

  Introduce improvements in synthesis quality;

  Fix existing bugs reported by user feedback (or you met);

  Introduce more convenient user operations.

PR type

• synthesis quality improvement

Description

This pull request refactors the checkpoint saving logic to replace the arbitrary value 2333333 with a more meaningful and current identifier. It also updates the exit code to 0 for successful completion.

Pull Affects

• Checkpoint Naming: Replaced the hardcoded value 2333333 in checkpoint filenames with the current value for clarity and informative issue.

• Exit Code: Changed the exit code from 2333333 to 0 to follow conventional practices for indicating successful completion of the training process.

[fatima-falconit]

Hey! Can you please tell me where to find a dataset that would fit this code? Or at least, what do I have to have in my dataset? some .wav files and a csv I guess? The model name is lulu... where is that????

I need help with this. Plus, I wanna train it for another language, do I just have to give a dataset in that language?

[blaisewf]

just put the path of a folder with audios on the webui

[RVC-Boss]

Automatically reading the last checkpoint during the next training is lost.