Skip to content

Fine-tuning Llama-3.2-3B on Marcus Aurelius' writings using QLoRA.

Notifications You must be signed in to change notification settings

mrochk/Llama-Aurelius

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Llama Aurelius

Fine-tuning Llama on Marcus Aurelius' writings using QLoRA.

Examples of generations:

PROMPT: 'The best reaction against the uncertitude of life is'
MODEL: 'to be content with the present, and to be satisfied with the things which are in our power.'
PROMPT: 'The way of life is'
MODEL: 'the way of the gods, and the way of the gods is the way of the universe.'
PROMPT: 'The best thing in life is'
MODEL: 'to be able to do what thou wilt, and to be able to do it at the right time.'

All the code can be found in the notebook.

The training takes approximately 12 minutes on a P100 gpu.

The model is also available on my HuggingFace account:
https://huggingface.co/mrochk/Llama-Aurelius

References:

About

Fine-tuning Llama-3.2-3B on Marcus Aurelius' writings using QLoRA.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published