Skip to content

Fine-tuning LLMs on medical text data of various functionalities.

Notifications You must be signed in to change notification settings

yxzwayne/Medilora

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MediLora

Langauge model fine-tuning with limited datasets and Q-LoRA in data-limited scenarios.

Code | Report

Model Checkpoints Considered

  • From Nous Research:

    • teknium/OpenHermes-2.5-Mistral-7B
    • NousResearch/Nous-Hermes-llama-2-7b
    • NousResearch/Nous-Hermes-Llama2-13b
  • From Microsoft:

    • microsoft/Orca-2-7b
    • microsoft/Orca-2-13b

Notes

  1. Mistral-7b on OpenHermes benched higher than on llama2-13b, which makes sense because mistral is a llama2 finetune.
  2. Orca-2-13b underperforms OpenHermes-2.5-Mistral-7B on BigBench, which is quite a reflection on its reasoning abilities.

Repo Referenced

About

Fine-tuning LLMs on medical text data of various functionalities.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published