effusiveperiscope / PPPDataset

Various tools for manipulating Pony Preservation Project related data

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

What is this

Tools for manipulating PPP voice and text data and an attempt to make combined synthetic training data for pony-related LLMs, made by a codelet.

ppp.py

Point it at your Sliced Dialogue directory. Can be used to reformat datasets for training voice models (example is for PITS). Also has an updated version of horsewords.clean for ARPAbet substitutions.

FiMFiction Tools

Various tools for trimming/sampling the fimficOmegaV3 dataset.

fimfarchive.py

A tool for accessing and manipulating a downloaded copy of the fimfarchive archives.

Text training data

Tools for scraping sources like the FiMFiction wiki for episode summaries, episode titles/transcripts, Wikipedia episode synopsis, as well as experiments for using LLMs via oobabooga api to create synthetic training data.

About

Various tools for manipulating Pony Preservation Project related data


Languages

Language:Jupyter Notebook 92.6%Language:Python 7.4%