OLMo 2 1124 13B Instruct for Amharic: GMT

This model is built on top of OLMo 2 1124 13B Instruct adapted for Amharic using 200M target language tokens sampled from MADLAD-400. The model is adapted using the GMT approach, a state-of-the-art dynamic selective parameter update approach that drops gradients of a pre-defined ratio (50% in this study for fair comparison with HFT and SSU) with smaller absolute values on the target data. This is based on https://ojs.aaai.org/index.php/AAAI/article/view/34621.

Model Description

Language: Amharic
License: Apache 2.0
Fine-tuned from model: allenai/OLMo-2-1124-13B-Instruct

Model Sources

Repository: https://github.com/gucci-j/ssu
Paper: https://arxiv.org/abs/2512.04844

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained(
    "ssu-project/OLMo-2-1124-13B-Instruct-am-gmt"
)
tokenizer = AutoTokenizer.from_pretrained(
    "ssu-project/OLMo-2-1124-13B-Instruct-am-gmt"
)

Citation

@misc{yamaguchi2025mitigatingcatastrophicforgettingtarget,
    title={Mitigating Catastrophic Forgetting in Target Language Adaptation of LLMs via Source-Shielded Updates}, 
    author={Atsuki Yamaguchi and Terufumi Morishita and Aline Villavicencio and Nikolaos Aletras},
    year={2025},
    eprint={2512.04844},
    archivePrefix={arXiv},
    primaryClass={cs.CL},
    url={https://arxiv.org/abs/2512.04844}, 
}