8000 GitHub - acesanderson/Sloth: finetuning
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

acesanderson/Sloth

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Purpose of this project

This is a simple ("for dummies") wrapper around unsloth and related libraries for the purpose of finetuning.

Current state of project:

  • Have the basic unsloth workflow simplified into something I can use.
  • In the middle of finetuning a full model on dataset (course titles + descriptions)

To do

  • create convenience functions for training and inference
  • do bulk training in a queue:
    • course titles -> course descriptions
    • course descriptions -> course TOCs
    • course descriptions|TOCs -> intro video
    • course descriptions|TOCs|intro video -> individual video scripts
  • add support for other models (llama3.2, mistral, gemma, phi, etc.)
  • determine which models are cleared for commercial use
  • add more detailed parameters

Architecture

Sloth/ └── datasets/ |── course_titles_and_descriptions.py |-- course_data_and_tocs.py |-- course_data_and_intro_videos.py |-- course_data_and_video_scripts.py ├── training/ └── inference/

About

finetuning

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

0