Model Details

Granite-3.1-8B-instruct-Korean

Granite-3.1-8B-instruct-Korean is continued pretrained(fully fine-tuned) language model based on Granite-3.1-8B-Instruct.

This model is trained fully with publicily available resource at HuggingFace dataset hub, preprocessed Korean texts.

The train was done on A6000 48GB * 4.

Model developers Dongwook Min (mindw96)

Dataset beomi/KoAlpaca-v1.1a

Variations Granite-3.1-8B-instruct-KR comes in one size — 8B.

Input Models input text only.

Output Models generate text only.

Model Architecture Granite 3.1 is an auto-regressive language model that uses an optimized transformer architecture.

Model Release Date 02.01.2025.

Downloads last month
10
Safetensors
Model size
8.37B params
Tensor type
BF16
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for mindw96/Granite-3.1-8B-instruct-Korean

Finetuned
(2)
this model
Quantizations
2 models

Dataset used to train mindw96/Granite-3.1-8B-instruct-Korean