You can deploy this model and start earning money today!

RichardErkhov/
elyza_-_Llama-3-ELYZA-JP-8B-gguf

Name: elyza_-_Llama-3-ELYZA-JP-8B-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama-3-ELYZA-JP-8B - GGUF

Model creator: https://huggingface.co/elyza/
Original model: https://huggingface.co/elyza/Llama-3-ELYZA-JP-8B/

Name	Quant method	Size
Llama-3-ELYZA-JP-8B.Q2_K.gguf	Q2_K	2.96GB
Llama-3-ELYZA-JP-8B.IQ3_XS.gguf	IQ3_XS	3.28GB
Llama-3-ELYZA-JP-8B.IQ3_S.gguf	IQ3_S	3.43GB
Llama-3-ELYZA-JP-8B.Q3_K_S.gguf	Q3_K_S	3.41GB
Llama-3-ELYZA-JP-8B.IQ3_M.gguf	IQ3_M	3.52GB
Llama-3-ELYZA-JP-8B.Q3_K.gguf	Q3_K	3.74GB
Llama-3-ELYZA-JP-8B.Q3_K_M.gguf	Q3_K_M	3.74GB
Llama-3-ELYZA-JP-8B.Q3_K_L.gguf	Q3_K_L	4.03GB
Llama-3-ELYZA-JP-8B.IQ4_XS.gguf	IQ4_XS	4.18GB
Llama-3-ELYZA-JP-8B.Q4_0.gguf	Q4_0	4.34GB
Llama-3-ELYZA-JP-8B.IQ4_NL.gguf	IQ4_NL	4.38GB
Llama-3-ELYZA-JP-8B.Q4_K_S.gguf	Q4_K_S	4.37GB
Llama-3-ELYZA-JP-8B.Q4_K.gguf	Q4_K	4.58GB
Llama-3-ELYZA-JP-8B.Q4_K_M.gguf	Q4_K_M	4.58GB
Llama-3-ELYZA-JP-8B.Q4_1.gguf	Q4_1	4.78GB
Llama-3-ELYZA-JP-8B.Q5_0.gguf	Q5_0	5.21GB
Llama-3-ELYZA-JP-8B.Q5_K_S.gguf	Q5_K_S	5.21GB
Llama-3-ELYZA-JP-8B.Q5_K.gguf	Q5_K	5.34GB
Llama-3-ELYZA-JP-8B.Q5_K_M.gguf	Q5_K_M	5.34GB
Llama-3-ELYZA-JP-8B.Q5_1.gguf	Q5_1	5.65GB
Llama-3-ELYZA-JP-8B.Q6_K.gguf	Q6_K	6.14GB
Llama-3-ELYZA-JP-8B.Q8_0.gguf	Q8_0	7.95GB

Original model description:

library_name: transformers license: llama3 language:

Llama-3-ELYZA-JP-8B

Model Description

Llama-3-ELYZA-JP-8B is a large language model trained by ELYZA, Inc. Based on meta-llama/Meta-Llama-3-8B-Instruct, it has been enhanced for Japanese usage through additional pre-training and instruction tuning. (Built with Meta Llama3)

For more details, please refer to our blog post.

Usage

python

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

DEFAULT_SYSTEM_PROMPT = "あなたは誠実で優秀な日本人のアシスタントです。特に指示が無い場合は、常に日本語で回答してください。"
text = "仕事の熱意を取り戻すためのアイデアを5つ挙げてください。"

model_name = "elyza/Llama-3-ELYZA-JP-8B"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto",
)
model.eval()

messages = [
    {"role": "system", "content": DEFAULT_SYSTEM_PROMPT},
    {"role": "user", "content": text},
]
prompt = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
token_ids = tokenizer.encode(
    prompt, add_special_tokens=False, return_tensors="pt"
)

with torch.no_grad():
    output_ids = model.generate(
        token_ids.to(model.device),
        max_new_tokens=1200,
        do_sample=True,
        temperature=0.6,
        top_p=0.9,
    )
output = tokenizer.decode(
    output_ids.tolist()[0][token_ids.size(1):], skip_special_tokens=True
)
print(output)

Developers

Listed in alphabetical order.

License

Meta Llama 3 Community License

How to Cite

tex

@misc{elyzallama2024,
      title={elyza/Llama-3-ELYZA-JP-8B},
      url={https://huggingface.co/elyza/Llama-3-ELYZA-JP-8B},
      author={Masato Hirakawa and Shintaro Horie and Tomoaki Nakamura and Daisuke Oba and Sam Passaglia and Akira Sasaki},
      year={2024},
}

Citations

tex

@article{llama3modelcard,
    title={Llama 3 Model Card},
    author={AI@Meta},
    year={2024},
    url = {https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md}
}

RichardErkhov/elyza_-_Llama-3-ELYZA-JP-8B-gguf