Hugging Face's logo

tokyotech-llm
/
Llama-3.1-Swallow-70B-Instruct-v0.1

Text Generation
Transformers
Safetensors
English
Japanese
llama
conversational
text-generation-inference
Model card Files Community
1
Llama-3.1-Swallow-70B-Instruct-v0.1 / README.md
Taishi-N324's picture
Taishi-N324
Create README.md
8abaccb verified
preview code
|
raw
history blame
16.5 kB
metadata
language:
  - en
  - ja
library_name: transformers
pipeline_tag: text-generation
license: llama3.1
model_type: llama

Llama3.1 Swallow

Our Swallow model has undergone continual pre-training from the Llama 3.1 family, primarily with the addition of Japanese language data. The Instruct versions use supervised fine-tuning (SFT). Links to other models can be found in the index.

Model Release Updates

We are excited to share the release schedule for our latest models:

Swallow Model Index

Model Llama-3.1-Swallow Llama-3.1-Swallow-Instruct
8B Link Link
70B Link Link

logo

This repository provides large language models developed by Swallow-LLM.

Model Details

  • Model type: Please refer to Llama 3.1 MODEL_CARD for details on the model architecture.
  • Language(s): Japanese English
  • Library: Megatron-LM
  • Tokenizer: Please refer to Llama 3.1 blog for details on the tokenizer.
  • Contact: swallow[at]nlp.c.titech.ac.jp

Model Performance

Japanese tasks

Model Size JCom. JEMHopQA NIILC JSQuAD XL-Sum MGSM WMT20-en-ja WMT20-ja-en JMMLU JHumanEval Ja Avg
4-shot 4-shot 4-shot 4-shot 1-shot 4-shot 4-shot 4-shot 5-shot 0-shot
EM acc Char-F1 Char-F1 Char-F1 ROUGE-2 EM acc BLEU BLEU EM acc pass@1
モデル名 JCom. JEMHopQA NIILC JSQuAD XL-Sum MGSM WMT20-en-ja WMT20-ja-en JMMLU JHumanEval Ja Avg
---------- ------- ---------- ------- -------- -------- ------ ------------- ------------- ------- ------------ --------
Gemma 2 27B IT 0.9562 0.5413 0.5755 0.8832 0.1648 0.7000 0.2900 0.2500 0.6701 0.6293 0.5660
Phi-3.5-MoE Instruct 0.9321 0.4416 0.4920 0.9079 0.2255 0.7120 0.2575 0.2024 0.6447 0.4213 0.5237
GRIN-MoE 0.8606 0.4622 0.3943 0.8877 0.0302 0.6400 0.2300 0.1911 0.5696 0.4476 0.4713
KARAKURI LM 70B Chat v0.1 0.8847 0.5139 0.5668 0.9096 0.1369 0.2800 0.2526 0.2095 0.4648 0.2354 0.4454
Swallow-70b-instruct-v0.1 0.9231 0.5654 0.5751 0.9036 0.1861 0.4160 0.2619 0.2318 0.5727 0.2835 0.4919
Llama 3 70B Instruct 0.9419 0.6114 0.5506 0.9164 0.1912 0.7200 0.2708 0.2350 0.6789 0.6610 0.5777
Llama 3.1 70B Instruct 0.9482 0.6246 0.5781 0.9201 0.1772 0.7440 0.2805 0.2472 0.7323 0.6933 0.5945
Llama 3 Youko 70B Instruct 0.9526 0.6252 0.5853 0.9215 0.1983 0.7400 0.2633 0.2245 0.7170 0.6098 0.5838
Llama-3.1-70B-Japanese-Instruct-2407 0.9562 0.6466 0.6602 0.9187 0.1564 0.7480 0.2901 0.2410 0.7227 0.6274 0.5967
Llama 3 heron brain 70B v0.3 0.9660 0.6643 0.6817 0.9221 0.2611 0.7720 0.3093 0.2578 0.7077 0.6079 0.6150
Llama 3 Swallow 70B Instruct 0.9607 0.6188 0.6026 0.9236 0.1389 0.6560 0.2724 0.2532 0.6572 0.6000 0.5683
Llama 3.1 Swallow 70B Instruct 0.9598 0.6192 0.6605 0.9235 0.1938 0.7760 0.3123 0.2593 0.7117 0.4713 0.5887
Qwen2-72B-Instruct 0.9634 0.6268 0.5418 0.9210 0.1644 0.7840 0.2592 0.2327 0.7713 0.6909 0.5955
Qwen2.5-72B-Instruct 0.9696 0.5699 0.5811 0.7381 0.1706 0.8360 0.2269 0.2179 0.7899 0.6256 0.5726
Mixtral-8x22B-Instruct-v0.1 0.9053 0.5001 0.4609 0.9186 0.2060 0.6760 0.2327 0.2313 0.6094 0.5787 0.5319

English tasks

Model Size OpenBookQA TriviaQA HellaSWAG SQuAD2.0 XWINO MMLU GSM8K BBH HumanEval EnAvg
4-shot 4-shot 4-shot 4-shot 4-shot 5-shot 4-shot 3-shot 0-shot
Acc EMacc Acc EMacc Acc Acc EMacc CoTEMAcc pass@1
モデル名 OpenBookQA TriviaQA HellaSwag SQuAD2.0 XWIN MMLU GSM8K BBH HumanEval En Avg
---------- ------------ ---------- ------------ ---------- ------ ------ ------- ----- ----------- --------
Gemma 2 27B IT 0.4560 0.7660 0.6548 0.4012 0.9101 0.7624 0.8438 0.7876 0.6939 0.6973
Phi-3.5-MoE Instruct 0.4960 0.6746 0.6901 0.3174 0.8903 0.7872 0.8317 0.7618 0.5561 0.6673
GRIN-MoE 0.4660 0.7035 0.7046 0.3544 0.8976 0.7693 0.8287 0.7533 0.6841 0.6846
KARAKURI LM 70B Chat v0.1 0.4100 0.6873 0.6315 0.3677 0.9049 0.5941 0.3882 0.5724 0.2305 0.5319
Swallow-70b-instruct-v0.1 0.4440 0.7411 0.6567 0.3529 0.9166 0.6677 0.5095 0.6661 0.2835 0.5820
Llama 3 70B Instruct 0.4400 0.7999 0.6552 0.4024 0.9127 0.7992 0.9052 0.8326 0.7555 0.7225
Llama 3.1 70B Instruct 0.4300 0.8212 0.6621 0.3921 0.9157 0.8213 0.8764 0.8390 0.7915 0.7277
Llama 3 Youko 70B Instruct 0.4500 0.7973 0.6863 0.3914 0.9153 0.8055 0.8923 0.7814 0.6598 0.7088
Llama-3.1-70B-Japanese-Instruct-2407 0.4220 0.8104 0.6481 0.3744 0.9170 0.8071 0.8893 0.8228 0.7463 0.7153
Llama 3 heron brain 70B v0.3 0.4460 0.8107 0.6682 0.4085 0.9174 0.7898 0.8772 0.7586 0.6713 0.7053
Llama 3 Swallow 70B Instruct 0.4520 0.8174 0.6758 0.4050 0.9230 0.7883 0.8688 0.8152 0.6890 0.7150
Llama 3.1 Swallow 70B Instruct 0.4520 0.8148 0.6834 0.4012 0.9157 0.7855 0.8886 0.8486 0.5823 0.7080
Qwen2-72B-Instruct 0.4360 0.7588 0.6857 0.3913 0.9110 0.8391 0.8499 0.2436 0.6939 0.6455
Qwen2.5-72B-Instruct 0.4540 0.6764 0.7064 0.3550 0.8895 0.8478 0.9113 0.4027 0.6165 0.6511
Mixtral-8x22B-Instruct-v0.1 0.4540 0.8265 0.7074 0.3927 0.9222 0.7733 0.8324 0.8306 0.7348 0.7193

MT-Bench JA

Model Size coding extraction humanities math reasoning roleplay stem writing JMTAvg
Model coding extraction humanities math reasoning roleplay stem writing JMT Avg
------- -------- ------------ ------------ ------ ----------- ---------- ------ --------- ---------
Gemma 2 27B IT 0.5467 0.6752 0.8386 0.6246 0.7201 0.7916 0.6787 0.807 0.7103
Phi-3.5-MoE Instruct 0.5214 0.8106 0.647 0.4415 0.536 0.6712 0.5314 0.7304 0.6112
GRIN-MoE 0.5294 0.7224 0.5923 0.5467 0.499 0.603 0.538 0.6839 0.5893
KARAKURI LM 70B Chat v0.1 0.2804 0.5862 0.624 0.2934 0.4183 0.553 0.4859 0.5964 0.4797
Swallow-70b-instruct-v0.1 0.303 0.55 0.565 0.3483 0.305 0.542 0.4916 0.463 0.446
Llama 3 70B Instruct 0.5969 0.841 0.712 0.4481 0.4884 0.7117 0.651 0.69 0.6424
Llama 3.1 70B Instruct 0.5252 0.7846 0.7086 0.5063 0.6979 0.6888 0.6402 0.6653 0.6521
Llama 3 Youko 70B Instruct 0.6632 0.8387 0.8108 0.4655 0.7013 0.7778 0.7544 0.7662 0.7222
Llama-3.1-70B-Japanese-Instruct-2407 0.6267 0.7525 0.7938 0.575 0.559 0.7725 0.724 0.718 0.6902
Llama 3 heron brain 70B v0.3 0.3762 0.7892 0.7274 0.5589 0.507 0.6662 0.688 0.6996 0.6266
Llama 3 Swallow 70B Instruct 0.5269 0.725 0.569 0.4669 0.6121 0.6238 0.5533 0.5698 0.5809
Llama 3.1 Swallow 70B Instruct 0.5676 0.7859 0.749 0.5437 0.6383 0.687 0.6121 0.654 0.6547
Qwen2-72B-Instruct 0.5699 0.7858 0.8222 0.5096 0.7032 0.7963 0.7728 0.8223 0.7228
Qwen2.5-72B-Instruct 0.706 0.7866 0.8122 0.6968 0.6536 0.8301 0.806 0.7841 0.7594
Mixtral-8x22B-Instruct-v0.1 0.5061 0.7454 0.5978 0.4772 0.476 0.542 0.4679 0.6244 0.5546
Llama 3.1 405B Instruct (deepinfra API) 0.6464 0.8218 0.715 0.5313 0.6447 0.716 0.6737 0.677 0.6782
GPT-3.5 (gpt-3.5-turbo-0125) 0.6851 0.7641 0.7414 0.5522 0.5128 0.7104 0.6266 0.7361 0.6661
GPT-4o (gpt-4o-2024-05-13) 0.7296 0.854 0.8646 0.6641 0.6661 0.8274 0.8184 0.8085 0.7791

Evaluation Benchmarks

Japanese evaluation benchmarks

We used llm-jp-eval(v1.3.0), JP Language Model Evaluation Harness(commit #9b42d41) and Code Generation LM Evaluation Harness(commit #0261c52). The details are as follows:

  • Multiple-choice question answering (JCommonsenseQA [Kurihara et al., 2022])
  • Open-ended question answering (JEMHopQA [Ishii et al., 2024])
  • Open-ended question answering (NIILC [関根, 2003])
  • Machine reading comprehension (JSQuAD [Kurihara et al., 2022])
  • Automatic summarization (XL-Sum [Hasan et al., 2021])
  • Machine translation (WMT2020 ja-en [Barrault et al., 2020])
  • Machine translation (WMT2020 en-ja [Barrault et al., 2020])
  • Mathematical reasoning (MGSM [Shi et al., 2023])
  • Academic exams (JMMLU [尹ら, 2024])
  • Code generation (JHumanEval [佐藤ら, 2024])

English evaluation benchmarks

We used the Language Model Evaluation Harness(v.0.4.2) and Code Generation LM Evaluation Harness(commit #0261c52). The details are as follows:

  • Multiple-choice question answering (OpenBookQA [Mihaylov et al., 2018])
  • Open-ended question answering (TriviaQA [Joshi et al., 2017])
  • Machine reading comprehension (SQuAD2 [Rajpurkar et al., 2018])
  • Commonsense reasoning (XWINO [Tikhonov and Ryabinin, 2021])
  • Natural language inference (HellaSwag [Zellers et al., 2019])
  • Mathematical reasoning (GSM8K [Cobbe et al., 2021])
  • Reasoning (BBH (BIG-Bench-Hard) [Suzgun et al., 2023])
  • Academic exams (MMLU [Hendrycks et al., 2021])
  • Code generation (HumanEval [Chen et al., 2021])

MT-Bench JA

We used Japanese MT-Bench to assess the instruction-following capabilities of models. We utilized the following settings:

Usage 

pip install vllm

from transformers import AutoTokenizer
from vllm import LLM, SamplingParams

model_name = "tokyotech-llm/Llama-3.1-Swallow-70B-Instruct-v0.1"

tokenizer = AutoTokenizer.from_pretrained(model_name)
llm = LLM(
    model=model_name,
    tensor_parallel_size=4,
)

sampling_params = SamplingParams(
    temperature=0.6, top_p=0.9, max_tokens=512, stop="<|eot_id|>"
)


message = [
    {"role": "system", "content": "あなたは誠実で優秀な日本人のアシスタントです。"},
    {
        "role": "user",
        "content": "東京の紅葉した公園で、東京タワーと高層ビルを背景に、空を舞うツバメと草地に佇むラマが出会う温かな物語を書いてください。",
    },
]
prompt = tokenizer.apply_chat_template(
    message, tokenize=False, add_generation_prompt=True
)

output = llm.generate(prompt, sampling_params)

print(output[0].outputs[0].text)

Training Datasets

Instruction Tuning

The following datasets were used for the instruction tuning.

  • lmsys-chat-1m-synth-ja-wo-pii

    • Japanese translation of the lmsys-chat-1m dataset using DeepL, with synthetic instruction data created using the Llama-3.1-405B model.
    • 'wo-pii' indicates removal of personally identifiable information.

  • filtered magpie-ultra

    • Subset of the magpie-ultra dataset, containing samples rated as 'average,' 'good,' or 'excellent.'.

  • gemma-magpie

    • Japanese dataset.
    • Generated using prompts for specific category words.

Risks and Limitations

The models released here are still in the early stages of our research and development and have not been tuned to ensure outputs align with human intent and safety considerations.

Acknowledgements

We thank Meta Research for releasing Llama 3.1 under an open license for others to build on.

Our project is supported by the Large Generative AI Development Support Program of the National Institute of Advanced Industrial Science and Technology.

License

META LLAMA 3.1 COMMUNITY LICENSE

Authors

Here are the team members:

How to cite

If you find our work helpful, please feel free to cite us.

@inproceedings{Fujii:COLM2024,
   title={Continual Pre-Training for Cross-Lingual LLM Adaptation:
Enhancing Japanese Language Capabilities},
   author={Kazuki Fujii and Taishi Nakamura and Mengsay Loem and Hiroki
Iida and Masanari Ohi and Kakeru Hattori and Hirai Shota and Sakae
Mizuki and Rio Yokota and Naoaki Okazaki},
   booktitle="Proceedings of the First Conference on Language Modeling",
   series={COLM},
   pages="(to appear)",
   year="2024",
   month=oct,
   address={University of Pennsylvania, USA},
}

@inproceedings{Okazaki:COLM2024,
   title={Building a Large Japanese Web Corpus for Large Language Models},
   author={Naoaki Okazaki and Kakeru Hattori and Hirai Shota and Hiroki
Iida and Masanari Ohi and Kazuki Fujii and Taishi Nakamura and Mengsay
Loem and Rio Yokota and Sakae Mizuki},
   booktitle="Proceedings of the First Conference on Language Modeling",
   series={COLM},
   pages="(to appear)",
   year="2024",
   month=oct,
   address={University of Pennsylvania, USA},
}

References 

@misc{dubey2024llama3herdmodels,
      title={The Llama 3 Herd of Models}, 
      author={Abhimanyu Dubey and Abhinav Jauhri and Abhinav Pandey and Abhishek Kadian and Ahmad Al-Dahle and Aiesha Letman and Akhil Mathur and Alan Schelten and Amy Yang and Angela Fan et al.},
      year={2024},
      eprint={2407.21783},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2407.21783}, 
}