Skip to content

ahwallace/llm-poet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 

Repository files navigation

llm-poet

This is an exercise to learn more about HuggingFace Transformers and efficient fine-tuning using quantisation (QLora) and PEFT. The aim is to fine-tuning GPT-Neo-1.3B on a dataset of English poetry using Google Colab's free tier GPU. Below is a sample of poems output by the model when given the prompt, "I taste a liquor never brewed, " (I taste a liquor never brewed (214) - Emily Dickinson):

I taste a liquor never brewed, I can't be tempted, but I feel Some things I like, and others, I like but can't think of.

I taste a liquor never brewed, And a liquor never drunk, The one of the year gone by, When I first met you, and drank your health. Tomorrow you'll be as I am.

About

Fine-tuning an LLM on English poetry

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published