File Scripts Project

A collection of tools for managing machine learning workflows, file organization, and environment setup.

Quick Links

• tools
  • Launch tools menu from any directory
• source ~/.bashrc
  • Reload shell after installation

Available Tools

Environment

• Setup (st)
  • Environment configuration and setup
  • Git synchronization
  • Command shortcuts installation
  • WANDB configuration (RunPod)

Model Management

• Config Manager (cm)

  • Training configuration management
  • Template-based configuration
  • Version control integration

• LoRA Mover (lm)

  • Process and organize LoRA models
  • Automated file organization
  • Version management

Cleanup

• Remove Configs (rc)

  • Remove configuration files
  • Batch cleanup support

• Dataset Cache (rd)

  • Clear dataset cache
  • Free up disk space

• Dataset JSON (rj)

  • Clear dataset JSON files
  • Reset training metadata

• Checkpoints (cp)

  • Delete .ipynb_checkpoints directories
  • Clean workspace structure

• Delete Models (dm)

  • Remove model files and data
  • Selective cleanup options

Utilities

• Download Configs (dc)

  • Sync configurations with Dropbox
  • Cloud backup integration

• Debug Crops (db)

  • Debug image preparation issues
  • Visual feedback system

Installation

# Clone repository
git clone https://github.com/rafstahelin/file-scripts.git /workspace/file-scripts

# Run setup
cd /workspace/file-scripts
python tools.py  # Select 'setup' or 'st'
source ~/.bashrc

Project Structure

/workspace/file-scripts/
├── tools/          # Tool implementations
├── docs/          # Documentation
└── tools.py       # Main menu system

Environment Support

• ✓ RunPod

  • Full support with cloud integrations
  • WANDB integration
  • Network volume support

• ✓ Container

  • Standard functionality
  • Local development features

• ~ WSL

  • Basic functionality
  • Future enhancements planned

Development

Tools follow a standardized pattern:

• Python module in tools directory
• Tool class with run() method
• Rich console interface
• Error handling and user feedback

Support

For issues and feature requests, please open an issue on the GitHub repository.