At this moment relu_layer op doesn't allow threshold configuration, and legacy RELU op allows that.
We should add configuration option to relu_layer.

🚀 Feature

Enable training purely based on number of iterations instead of epochs

Motivation

This can be useful for certain training runs. Without this feature, the user must set an unreachably high value for max_epochs and set max_steps to the desired iteration count. With this setup, the trainer will break from the training loop based on max_steps since we'd never reach `max_e

Currently, when entering epic mode the README is frozen in the last level of the tower. When you're trying to fine-tune the score for a level other than the last one, it would be helpful if we had the README for that level available. The proposal is that when entering epic mode, the README is updated with all levels, one following the other.

Example:

# Starbolt - beginner

Documentation Is:

• Missing
• Needed
• Confusing
• Not Sure?

E