initial data cleaning

noticed some issues when loading the data file(s) from hf so made some basic cleaning improvements 🧽

What's Changed

• Eda & HF links by @pszemraj in #4
• Textsum v2 compat by @pszemraj in #6

Full Changelog: v0.1.1...v0.1.2

summary writing score with NLI

What's Changed

• Zeroshot writing qual by @pszemraj in #1

New Contributors

• @pszemraj made their first contribution in #1

Full Changelog: v0.1...v0.1.1

MWE compiled gauntlet

• initial functional compiled version of my summarization gauntlet
• there are ~19 input documents and ~1900 model-generated summaries of these documents, including summaries from GPT-4 via langchain
• data cleaning/processing/analysis, as well as better code and repo structure, to come later

compiled_gauntlet_v0p1_stats.txt

Full Changelog: https://github.com/pszemraj/SummComparer/commits/v0.1