atsuki-yamaguchi/Llama-3-8B-si-30K-lapt

Llama3 8B for Sinhala: No vocabulary adaptation

This model is built on top of Llama3 8B adapted for Sinhala using 30K target language sentences sampled from CC-100.

Model Details

• Vocabulary: This model has no additional target vocabulary. It retains the original vocabulary of Llama3 8B.

Model Description

• Language: Sinhala
• License: Llama 3 Community License Agreement
• Fine-tuned from model: meta-llama/Meta-Llama-3-8B

Model Sources

• Repository: https://github.com/gucci-j/lowres-cve
• Paper: https://arxiv.org/abs/2406.11477

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModelForCausalLM

model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Meta-Llama-3-8B"
)
model = PeftModelForCausalLM.from_pretrained(
    model,
    "atsuki-yamaguchi/Llama-3-8B-si-30K-lapt"
)
model = model.merge_and_unload()
tokenizer = AutoTokenizer.from_pretrained(
    "meta-llama/Meta-Llama-3-8B"
)

Citation

@article{yamaguchi-etal-2024-effectively,
    title={How Can We Effectively Expand the Vocabulary of LLMs with 0.01GB of Target Language Text?}, 
    author={Atsuki Yamaguchi and Aline Villavicencio and Nikolaos Aletras},
    year={2024},
    journal={ArXiv},
    year={2024},
    volume={abs/2406.11477},
    url={https://arxiv.org/abs/2406.11477}, 
}