MARKDOWN_HEADER = (
    "# M-YASS\n<u>M</u>ultilingual - <u>Y</u>et <u>A</u>nother <u>S</u>ynergistic <u>S</u>core\n"
    "1. Filter: a subset of Language, Dataset and Model Type;\n"
    "2. Map Function: Each score will be mapped within each `<language,dataset>` pair. Can be used to get rank or normalize score;\n"
    "3. Reduce Function: An array of scores will be reduced to one. Can be used to get the average score;\n"
    "4. Group By: Language or Dataset. If `language` is selected, the score of all tasks within this language will be reduced by `Reduce Function`\n"
)

MARKDOWN_MAP_FUNCTION = (
    "Raw: `lambda x: x`\n"
    "Rank: `partial(pd.Series.rank, ascending=False, method='dense')`\n"
    "Normalize: `lambda x: (x - np.min(x)) / (np.max(x) - np.min(x))`\n"
)

MARKDOWN_REDUCE_FUNCTION = (
    "Mean: `lambda x: np.mean(x)`\n"
    "Median: `lambda x: np.median(x)`\n"
    "Max: `lambda x: np.max(x)`\n"
    "Min: `lambda x: np.min(x)`\n"
)

CITATION = """\
@inproceedings{lu-etal-2024-llamax,
    title = "{LL}a{MAX}: Scaling Linguistic Horizons of {LLM} by Enhancing Translation Capabilities Beyond 100 Languages",
    author = "Lu, Yinquan  and
      Zhu, Wenhao  and
      Li, Lei  and
      Qiao, Yu  and
      Yuan, Fei",
    editor = "Al-Onaizan, Yaser  and
      Bansal, Mohit  and
      Chen, Yun-Nung",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
    month = nov,
    year = "2024",
    address = "Miami, Florida, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.findings-emnlp.631",
    doi = "10.18653/v1/2024.findings-emnlp.631",
    pages = "10748--10772",
    abstract = "Large Language Models (LLMs) demonstrate remarkable translation capabilities in high-resource language tasks, yet their performance in low-resource languages is hindered by insufficient multilingual data during pre-training. To address this, we conduct extensive multilingual continual pre-training on the LLaMA series models, enabling translation support across more than 100 languages. Through a comprehensive analysis of training strategies, such as vocabulary expansion and data augmentation, we develop LLaMAX. Remarkably, without sacrificing its generalization ability, LLaMAX achieves significantly higher translation performance compared to existing open-source LLMs (by more than 10 spBLEU points) and performs on-par with specialized translation model (M2M-100-12B) on the Flores-101 benchmark. Extensive experiments indicate that LLaMAX can serve as a robust multilingual foundation model. The code and the models are publicly available.",
}\
"""