Quantization made by Richard Erkhov.

youri-7b - GGUF

Model creator: https://huggingface.co/rinna/
Original model: https://huggingface.co/rinna/youri-7b/

Name	Quant method	Size
youri-7b.Q2_K.gguf	Q2_K	2.36GB
youri-7b.IQ3_XS.gguf	IQ3_XS	2.6GB
youri-7b.IQ3_S.gguf	IQ3_S	2.75GB
youri-7b.Q3_K_S.gguf	Q3_K_S	2.75GB
youri-7b.IQ3_M.gguf	IQ3_M	2.9GB
youri-7b.Q3_K.gguf	Q3_K	3.07GB
youri-7b.Q3_K_M.gguf	Q3_K_M	3.07GB
youri-7b.Q3_K_L.gguf	Q3_K_L	3.35GB
youri-7b.IQ4_XS.gguf	IQ4_XS	3.4GB
youri-7b.Q4_0.gguf	Q4_0	3.56GB
youri-7b.IQ4_NL.gguf	IQ4_NL	3.58GB
youri-7b.Q4_K_S.gguf	Q4_K_S	3.59GB
youri-7b.Q4_K.gguf	Q4_K	3.8GB
youri-7b.Q4_K_M.gguf	Q4_K_M	3.8GB
youri-7b.Q4_1.gguf	Q4_1	3.95GB
youri-7b.Q5_0.gguf	Q5_0	4.33GB
youri-7b.Q5_K_S.gguf	Q5_K_S	4.33GB
youri-7b.Q5_K.gguf	Q5_K	4.45GB
youri-7b.Q5_K_M.gguf	Q5_K_M	4.45GB
youri-7b.Q5_1.gguf	Q5_1	4.72GB
youri-7b.Q6_K.gguf	Q6_K	5.15GB
youri-7b.Q8_0.gguf	Q8_0	6.67GB

Original model description:

language: - ja - en license: llama2 datasets: - mc4 - wikipedia - EleutherAI/pile - oscar-corpus/colossal-oscar-1.0 - cc100 thumbnail: https://github.com/rinnakk/japanese-pretrained-models/blob/master/rinna.png inference: false model-index: - name: youri-7b 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: 49.06 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b 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: 74.89 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b 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: 42.22 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b 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: 36.03 source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b 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: 71.82 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b 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: 8.64 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=rinna/youri-7b name: Open LLM Leaderboard

`rinna/youri-7b`

Overview

We conduct continual pre-training of llama2-7b on 40B tokens from a mixture of Japanese and English datasets. The continual pre-training significantly improves the model's performance on Japanese tasks.

The name youri comes from the Japanese word 妖狸/ようり/Youri, which is a kind of Japanese mythical creature (妖怪/ようかい/Youkai).

Library

The model was trained using code based on EleutherAI/gpt-neox.
Model architecture

A 32-layer, 4096-hidden-size transformer-based language model. Refer to the llama2 paper for architecture details.
Continual pre-training

The model was initialized with the llama2-7b model and continually trained on around 40B tokens from a mixture of the following corpora
- Japanese CC-100
- Japanese C4
- Japanese OSCAR
- The Pile
- Wikipedia
- rinna curated Japanese dataset
Contributors

Benchmarking

Please refer to rinna's LM benchmark page.

How to use the model

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("rinna/youri-7b")
model = AutoModelForCausalLM.from_pretrained("rinna/youri-7b")

if torch.cuda.is_available():
    model = model.to("cuda")

text = "西田幾多郎は、"
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=200,
        min_new_tokens=200,
        do_sample=True,
        temperature=1.0,
        top_p=0.95,
        pad_token_id=tokenizer.pad_token_id,
        bos_token_id=tokenizer.bos_token_id,
        eos_token_id=tokenizer.eos_token_id
    )

output = tokenizer.decode(output_ids.tolist()[0])
print(output)
"""
西田幾多郎は、プラトンの復権を主張し、対する従来の西洋哲学は、近代の合理主義哲学に委ね、「従来の哲学は破 壊されてしまった」と述べている。 西田幾多郎は、西洋近代哲学の「徹底的な検討」を拒んだ。それは、「現代的理解の脆弱性を補う筈の、従来のヨーロッパに伝わる哲学的な方法では到底それができなかったからである」とい
"""

Tokenization

The model uses the original llama-2 tokenizer.

How to cite

@misc{rinna-youri-7b,
    title = {rinna/youri-7b},
    author = {Zhao, Tianyu and Kaga, Akio and Sawada, Kei},
    url = {https://huggingface.co/rinna/youri-7b}
}

@inproceedings{sawada2024release,
    title = {Release of Pre-Trained Models for the {J}apanese Language},
    author = {Sawada, Kei and Zhao, Tianyu and Shing, Makoto and Mitsui, Kentaro and Kaga, Akio and Hono, Yukiya and Wakatsuki, Toshiaki and Mitsuda, Koh},
    booktitle = {Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)},
    month = {5},
    year = {2024},
    pages = {13898--13905},
    url = {https://aclanthology.org/2024.lrec-main.1213},
    note = {\url{https://arxiv.org/abs/2404.01657}}
}

References

@software{gpt-neox-library,
    title = {{GPT}-{N}eo{X}: Large Scale Autoregressive Language Modeling in {P}y{T}orch},
    author = {Andonian, Alex and Anthony, Quentin and Biderman, Stella and Black, Sid and Gali, Preetham and Gao, Leo and Hallahan, Eric and Levy-Kramer, Josh and Leahy, Connor and Nestler, Lucas and Parker, Kip and Pieler, Michael and Purohit, Shivanshu and Songz, Tri and Phil, Wang and Weinbach, Samuel},
    doi = {10.5281/zenodo.5879544},
    month = {8},
    year = {2021},
    version = {0.0.1},
    url = {https://www.github.com/eleutherai/gpt-neox}
}

License

The llama2 license

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	47.11
AI2 Reasoning Challenge (25-Shot)	49.06
HellaSwag (10-Shot)	74.89
MMLU (5-Shot)	42.22
TruthfulQA (0-shot)	36.03
Winogrande (5-shot)	71.82
GSM8k (5-shot)	8.64