hyunseoki's picture
Create README.md
fab5c9a verified
|
raw
history blame
572 Bytes
metadata
language:
  - ko
  - en
library_name: transformers
pipeline_tag: text-generation

This model has been developed by KAIST ALIN Lab and OMNIOUS.AI - HyunseokLee, TaeyoungKim

Input Models input text only.

Output Models generate text only.

Model Architecture ko-en-llama2-13b-aligned is an auto-regressive language model based on the LLaMA2 transformer architecture.

Base Model hyunseoki/ko-en-llama2-13b

Training Dataset Open dataset wiki and AIhub (English + Korean). Supervised Finetuned with Instruction Dataset and aligned with Human Preference Dataset using DPO.