RichardErkhov/FlagAlpha_-_Llama3-Chinese-8B-Instruct-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama3-Chinese-8B-Instruct - GGUF

• Model creator: https://huggingface.co/FlagAlpha/
• Original model: https://huggingface.co/FlagAlpha/Llama3-Chinese-8B-Instruct/

Name	Quant method	Size
Llama3-Chinese-8B-Instruct.Q2_K.gguf	Q2_K	2.96GB
Llama3-Chinese-8B-Instruct.IQ3_XS.gguf	IQ3_XS	3.28GB
Llama3-Chinese-8B-Instruct.IQ3_S.gguf	IQ3_S	1.58GB
Llama3-Chinese-8B-Instruct.Q3_K_S.gguf	Q3_K_S	0.35GB
Llama3-Chinese-8B-Instruct.IQ3_M.gguf	IQ3_M	3.52GB
Llama3-Chinese-8B-Instruct.Q3_K.gguf	Q3_K	3.74GB
Llama3-Chinese-8B-Instruct.Q3_K_M.gguf	Q3_K_M	3.74GB
Llama3-Chinese-8B-Instruct.Q3_K_L.gguf	Q3_K_L	4.03GB
Llama3-Chinese-8B-Instruct.IQ4_XS.gguf	IQ4_XS	4.18GB
Llama3-Chinese-8B-Instruct.Q4_0.gguf	Q4_0	4.34GB
Llama3-Chinese-8B-Instruct.IQ4_NL.gguf	IQ4_NL	4.38GB
Llama3-Chinese-8B-Instruct.Q4_K_S.gguf	Q4_K_S	4.37GB
Llama3-Chinese-8B-Instruct.Q4_K.gguf	Q4_K	4.58GB
Llama3-Chinese-8B-Instruct.Q4_K_M.gguf	Q4_K_M	4.58GB
Llama3-Chinese-8B-Instruct.Q4_1.gguf	Q4_1	4.78GB
Llama3-Chinese-8B-Instruct.Q5_0.gguf	Q5_0	5.21GB
Llama3-Chinese-8B-Instruct.Q5_K_S.gguf	Q5_K_S	5.21GB
Llama3-Chinese-8B-Instruct.Q5_K.gguf	Q5_K	5.34GB
Llama3-Chinese-8B-Instruct.Q5_K_M.gguf	Q5_K_M	5.34GB
Llama3-Chinese-8B-Instruct.Q5_1.gguf	Q5_1	5.65GB
Llama3-Chinese-8B-Instruct.Q6_K.gguf	Q6_K	6.14GB
Llama3-Chinese-8B-Instruct.Q8_0.gguf	Q8_0	7.95GB

Original model description:

license: apache-2.0 tags: - llama3 - chinese

Llama3-Chinese-8B-Instruct

Llama3-Chinese-8B-Instruct基于Llama3-8B中文微调对话模型，由Llama中文社区和AtomEcho（原子回声）联合研发，我们会持续提供更新的模型参数，模型训练过程见 https://llama.family。

模型的部署、训练、微调等方法详见Llama中文社区GitHub仓库：https://github.com/LlamaFamily/Llama-Chinese

如何使用

import transformers
import torch


model_id = "FlagAlpha/Llama3-Chinese-8B-Instruct"

pipeline = transformers.pipeline(
    "text-generation",
    model=model_id,
    model_kwargs={"torch_dtype": torch.float16},
    device="cuda",
)


messages = [{"role": "system", "content": ""}]

messages.append(
                {"role": "user", "content": "介绍一下机器学习"}
            )

prompt = pipeline.tokenizer.apply_chat_template(
        messages, 
        tokenize=False, 
        add_generation_prompt=True
    )

terminators = [
        pipeline.tokenizer.eos_token_id,
        pipeline.tokenizer.convert_tokens_to_ids("<|eot_id|>")
    ]
outputs = pipeline(
    prompt,
    max_new_tokens=512,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.6,
    top_p=0.9
)

content = outputs[0]["generated_text"][len(prompt):]

print(content)