Ernie Bot 文心一言

author: Jael

Description 描述

A LLM operator generates answer given prompt in messages using a large language model or service. This operator is implemented with Ernie Bot from Baidu. Please note you will need Ernie API key & Secret key to access the service.

LLM 算子使用大语言模型或服务，为输入的问题或提示生成答案。LLM/Ernie 利用了来自百度的文心一言。请注意，您需要文心一言服务的 API Key 和 Secret Key才能访问该服务。

Code Example 代码示例

Write a pipeline with explicit inputs/outputs name specifications:

from towhee import pipe, ops

p = (
    pipe.input('messages')
        .map('messages', 'answer', ops.LLM.Ernie(api_key=ERNIE_API_KEY, secret_key=ERNIE_SECRET_KEY, temperature=0.5))
        .output('answer')
)

messages=[
        {'question': 'Zilliz Cloud 是什么？', 'answer': 'Zilliz Cloud 是一种全托管的向量检索服务。'},
        {'question': '它和 Milvus 的关系是什么？'}
    ]
answer = p(messages).get()[0]

Write a retrieval-augmented generation pipeline with explicit inputs/outputs name specifications:

from towhee import pipe, ops


temp = '''根据以下材料回答最末尾的问题：

{context}

问题：{question}
'''


docs = ['你可以通过`pip install towhee` 安装 Towhee。']
history = [
    ('什么是 Towhee？', 'Towhee 是一个开源项目，可以将非结构化数据转换为向量。')
]
question = '怎么安装它？'

p = (
    pipe.input('question', 'docs', 'history')
        .map(('question', 'docs', 'history'), 'prompt', ops.prompt.template(temp, ['question', 'context']))
        .map('prompt', 'answer',
             ops.LLM.Ernie(api_key=ERNIE_API_KEY, secret_key=ERNIE_SECRET_KEY)
             )
        .output('answer')
)

answer = p(question, docs, history).get()[0]

Factory Constructor 接口说明

Create the operator via the following factory method:

LLM.Ernie(api_key: str, secret_key: str)

Parameters:

api_key: str=None

The Ernie API key in string, defaults to None. If None, it will use the environment variable ERNIE_API_KEY.

secret_key: str=None

The Ernie Secret key in string, defaults to None. If None, it will use the environment variable ERNIE_SECRET_KEY.

**kwargs

Other Ernie parameters such as temperature, etc.

Interface 使用说明

The operator takes a piece of text in string as input. It returns answer in json.

__call__(txt)

Parameters:

messages: list

​ A list of messages to set up chat. Must be a list of dictionaries with key value from "question", "answer". For example, [{"question": "a past question?", "answer": "a past answer."}, {"question": "current question?"}]. It also accepts the orignal Ernie message format like [{"role": "user", "content": "a question?"}, {"role": "assistant", "content": "an answer."}]

Returns:

answer: str

​ The next answer generated by role "assistant".