Sailor-7B/README.md

language, datasets, tags, license, base_model, model-index

language	datasets	tags	license	base_model	model-index
en zh id th vi ms lo	cerebras/SlimPajama-627B Skywork/SkyPile-150B allenai/MADLAD-400 cc100	multilingual sea sailor	apache-2.0	Qwen/Qwen1.5-7B	name results Sailor-7B task dataset metrics type text-generation name type XQuAD-Thai XQuAD-Thai name type value EM (3-Shot) EM (3-Shot) 57.88 name type value F1 (3-Shot) F1 (3-Shot) 71.06 task dataset metrics type text-generation name type TyDiQA-Indonesian TyDiQA-Indonesian name type value EM (3-Shot) EM (3-Shot) 60.53 name type value F1 (3-Shot) F1 (3-Shot) 75.42 task dataset metrics type text-generation name type XQuAD-Vietnamese XQuAD-Vietnamese name type value EM (3-Shot) EM (3-Shot) 53.81 name type value F1 (3-Shot) F1 (3-Shot) 74.62 task dataset metrics type text-generation name type XCOPA-Thai XCOPA-Thai name type value EM (3-Shot) EM (3-Shot) 59.00 task dataset metrics type text-generation name type XCOPA-Indonesian XCOPA-Indonesian name type value EM (3-Shot) EM (3-Shot) 72.20 task dataset metrics type text-generation name type XCOPA-Vietnamese XCOPA-Vietnamese name type value EM (3-Shot) EM (3-Shot) 72.20 task dataset metrics type text-generation name type M3Exam-Thai M3Exam-Thai name type value EM (3-Shot) EM (3-Shot) 30.00 task dataset metrics type text-generation name type M3Exam-Indonesian M3Exam-Indonesian name type value EM (3-Shot) EM (3-Shot) 32.88 task dataset metrics type text-generation name type M3Exam-Vietnamese M3Exam-Vietnamese name type value EM (3-Shot) EM (3-Shot) 44.10 task dataset metrics type text-generation name type BELEBELE-Thai BELEBELE-Thai name type value EM (3-Shot) EM (3-Shot) 41.56 task dataset metrics type text-generation name type BELEBELE-Indonesian BELEBELE-Indonesian name type value EM (3-Shot) EM (3-Shot) 44.33 task dataset metrics type text-generation name type BELEBELE-Vietnamese BELEBELE-Vietnamese name type value EM (3-Shot) EM (3-Shot) 45.33

Sailor is a suite of Open Language Models tailored for South-East Asia (SEA), focusing on languages such as 🇮🇩Indonesian, 🇹🇭Thai, 🇻🇳Vietnamese, 🇲🇾Malay, and 🇱🇦Lao. Developed with careful data curation, Sailor models are designed to understand and generate text across diverse linguistic landscapes of SEA region. Built from Qwen 1.5 , Sailor encompasses models of varying sizes, spanning from 0.5B to 14B versions for different requirements. We further fine-tune the base model with open-source datasets to get instruction-tuned models, namedly Sailor-Chat. Benchmarking results demonstrate Sailor's proficiency in tasks such as question answering, commonsense reasoning, and other tasks in SEA languages.

The logo was generated by MidJourney

Model Summary

• Model Collections: Base Model & Chat Model
• Project Website: sea-sailor.github.io/blog/sailor1/
• Codebase: github.com/sail-sg/sailor-llm
• Technical Report: arxiv.org/pdf/2404.03608.pdf

Training details

Sailor is crafted by continually pre-training from language models like the remarkable Qwen 1.5 models, which already has a great performance on SEA languages. The pre-training corpus heavily leverages the publicly available corpus, including SlimPajama, SkyPile, CC100 and MADLAD-400.

By employing aggressive data deduplication and careful data cleaning on the collected corpus, we have attained a high-quality dataset spanning various languages. Through systematic experiments to determine the weights of different languages, Sailor models undergo training from 200B to 400B tokens, tailored to different model sizes. The approach boosts their performance on SEA languages while maintaining proficiency in English and Chinese without significant compromise. Finally, we continually pre-train the Qwen1.5-0.5B model with 400 Billion tokens, and other models with 200 Billion tokens to obtain the Sailor models.

Requirements

The code of Sailor has been in the latest Hugging face transformers and we advise you to install transformers>=4.37.0.

Quickstart

Here provides a code snippet to show you how to load the tokenizer and model and how to generate contents.

from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model

model = AutoModelForCausalLM.from_pretrained("sail/Sailor-7B", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("sail/Sailor-7B")

input_message = "Model bahasa adalah model probabilistik" 
### The given Indonesian input translates to 'A language model is a probabilistic model of.'

model_inputs = tokenizer([input_message], return_tensors="pt").to(device)

generated_ids = model.generate(
    model_inputs.input_ids,
    max_new_tokens=64
)

generated_ids = [
    output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]

response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)

License

Sailor is distributed under the terms of the Apache License 2.0. No restrict on the research and the commercial use, but you should comply with the Qwen License, which means that you shall request a license from Qwen team if your product or service has more than 100 million monthly active users in your commerical scenarios, otherwise no need for further request.

Citation

If you find sailor useful, please cite our work as follows:

@inproceedings{dou-etal-2024-sailor,
    title = "Sailor: Open Language Models for South-{E}ast {A}sia",
    author = "Dou, Longxu and Liu, Qian and Zeng, Guangtao and Guo, Jia  and Zhou, Jiahui and Mao, Xin and Jin, Ziqi and Lu, Wei and Lin, Min",
    booktitle = "Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
    year = "2024",
}

Contact Us

If you have any questions, please raise an issue or contact us at doulx@sea.com or liuqian.sea@gmail.com.