README.md

metadata

language:
  - ja
  - en
license: llama2
model-index:
  - name: ELYZA-japanese-Llama-2-13b
    results:
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: AI2 Reasoning Challenge (25-Shot)
          type: ai2_arc
          config: ARC-Challenge
          split: test
          args:
            num_few_shot: 25
        metrics:
          - type: acc_norm
            value: 57
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: HellaSwag (10-Shot)
          type: hellaswag
          split: validation
          args:
            num_few_shot: 10
        metrics:
          - type: acc_norm
            value: 80.89
            name: normalized accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: MMLU (5-Shot)
          type: cais/mmlu
          config: all
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 54.38
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: TruthfulQA (0-shot)
          type: truthful_qa
          config: multiple_choice
          split: validation
          args:
            num_few_shot: 0
        metrics:
          - type: mc2
            value: 40.43
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: Winogrande (5-shot)
          type: winogrande
          config: winogrande_xl
          split: validation
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 76.87
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard
      - task:
          type: text-generation
          name: Text Generation
        dataset:
          name: GSM8k (5-shot)
          type: gsm8k
          config: main
          split: test
          args:
            num_few_shot: 5
        metrics:
          - type: acc
            value: 27.29
            name: accuracy
        source:
          url: >-
            https://huggingface.co./spaces/HuggingFaceH4/open_llm_leaderboard?query=elyza/ELYZA-japanese-Llama-2-13b
          name: Open LLM Leaderboard

ELYZA-japanese-Llama-2-13b

Model Description

ELYZA-japanese-Llama-2-13b は、 Llama 2をベースとして日本語能力を拡張するために追加事前学習を行ったモデルです。 詳細は Blog記事 を参照してください。

Usage

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "elyza/ELYZA-japanese-Llama-2-13b"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,
    use_cache=True,
    device_map="auto",
    low_cpu_mem_usage=True,
)
model.eval()

text = "自然言語処理とは、"
token_ids = tokenizer.encode(text, add_special_tokens=False, return_tensors="pt")

with torch.no_grad():
    output_ids = model.generate(
        token_ids.to(model.device),
        max_new_tokens=256,
        pad_token_id=tokenizer.pad_token_id,
        eos_token_id=tokenizer.eos_token_id,
    )
output = tokenizer.decode(output_ids.tolist()[0], skip_special_tokens=True)
print(output)

ELYZA-japanese-Llama-2-13b Models

Model Name	Vocab Size	#Params
elyza/ELYZA-japanese-Llama-2-13b	32000	13.02B
elyza/ELYZA-japanese-Llama-2-13b-instruct	32000	13.02B
elyza/ELYZA-japanese-Llama-2-13b-fast	44581	13.14B
elyza/ELYZA-japanese-Llama-2-13b-fast-instruct	44581	13.14B

Developers

• Akira Sasaki
• Masato Hirakawa
• Shintaro Horie
• Tomoaki Nakamura
• Sam Passaglia
• Daisuke Oba (intern)

Licence

Llama 2 is licensed under the LLAMA 2 Community License, Copyright (c) Meta Platforms, Inc. All Rights Reserved.

How to Cite

@misc{elyzallama2023, 
      title={ELYZA-japanese-Llama-2-13b}, 
      url={https://huggingface.co./elyza/ELYZA-japanese-Llama-2-13b}, 
      author={Akira Sasaki and Masato Hirakawa and Shintaro Horie and Tomoaki Nakamura and Sam Passaglia and Daisuke Oba},
      year={2023},
}

Citations

@misc{touvron2023llama,
      title={Llama 2: Open Foundation and Fine-Tuned Chat Models}, 
      author={Hugo Touvron and Louis Martin and Kevin Stone and Peter Albert and Amjad Almahairi and Yasmine Babaei and Nikolay Bashlykov and Soumya Batra and Prajjwal Bhargava and Shruti Bhosale and Dan Bikel and Lukas Blecher and Cristian Canton Ferrer and Moya Chen and Guillem Cucurull and David Esiobu and Jude Fernandes and Jeremy Fu and Wenyin Fu and Brian Fuller and Cynthia Gao and Vedanuj Goswami and Naman Goyal and Anthony Hartshorn and Saghar Hosseini and Rui Hou and Hakan Inan and Marcin Kardas and Viktor Kerkez and Madian Khabsa and Isabel Kloumann and Artem Korenev and Punit Singh Koura and Marie-Anne Lachaux and Thibaut Lavril and Jenya Lee and Diana Liskovich and Yinghai Lu and Yuning Mao and Xavier Martinet and Todor Mihaylov and Pushkar Mishra and Igor Molybog and Yixin Nie and Andrew Poulton and Jeremy Reizenstein and Rashi Rungta and Kalyan Saladi and Alan Schelten and Ruan Silva and Eric Michael Smith and Ranjan Subramanian and Xiaoqing Ellen Tan and Binh Tang and Ross Taylor and Adina Williams and Jian Xiang Kuan and Puxin Xu and Zheng Yan and Iliyan Zarov and Yuchen Zhang and Angela Fan and Melanie Kambadur and Sharan Narang and Aurelien Rodriguez and Robert Stojnic and Sergey Edunov and Thomas Scialom},
      year={2023},
      eprint={2307.09288},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	56.14
AI2 Reasoning Challenge (25-Shot)	57.00
HellaSwag (10-Shot)	80.89
MMLU (5-Shot)	54.38
TruthfulQA (0-shot)	40.43
Winogrande (5-shot)	76.87
GSM8k (5-shot)	27.29