摘要

摘要流水线对文本进行总结。此流水线运行一个 text2text 模型，用于抽象地创建输入文本的摘要。

示例

以下展示了使用此流水线的简单示例。

from txtai.pipeline import Summary

# Create and run pipeline
summary = Summary()
summary("Enter long, detailed text to summarize here")

请参阅下方链接以获取更详细的示例。

笔记本	描述
构建抽象文本摘要	运行抽象文本摘要

配置驱动示例

流水线可以使用 Python 或配置运行。流水线可以在配置中使用流水线的小写名称进行实例化。配置驱动的流水线通过工作流或API运行。

config.yml

# Create pipeline using lower case class name
summary:

# Run pipeline with workflow
workflow:
  summary:
    tasks:
      - action: summary

通过工作流运行

from txtai import Application

# Create and run pipeline with workflow
app = Application("config.yml")
list(app.workflow("summary", ["Enter long, detailed text to summarize here"]))

通过 API 运行

CONFIG=config.yml uvicorn "txtai.api:app" &

curl \
  -X POST "http://localhost:8000/workflow" \
  -H "Content-Type: application/json" \
  -d '{"name":"summary", "elements":["Enter long, detailed text to summarize here"]}'

方法

此流水线的 Python 文档。

__init__(path=None, quantize=False, gpu=True, model=None, **kwargs)

源代码位于 txtai/pipeline/text/summary.py

def __init__(self, path=None, quantize=False, gpu=True, model=None, **kwargs):
    super().__init__("summarization", path, quantize, gpu, model, **kwargs)

__call__(text, minlength=None, maxlength=None, workers=0)

对文本块运行摘要模型。

此方法支持字符串或列表形式的文本输入。如果输入是字符串，返回类型为文本。如果输入是列表，则返回一个文本列表，其中每行对应一个文本块。

参数

名称	类型	描述	默认值
text		text|list	必填
minlength		摘要的最小长度	None
maxlength		摘要的最大长度	None
workers		用于处理数据的并发工作线程数，默认为 None	0

返回值

类型	描述
	摘要文本

源代码位于 txtai/pipeline/text/summary.py

def __call__(self, text, minlength=None, maxlength=None, workers=0):
    """
    Runs a summarization model against a block of text.

    This method supports text as a string or a list. If the input is a string, the return
    type is text. If text is a list, a list of text is returned with a row per block of text.

    Args:
        text: text|list
        minlength: minimum length for summary
        maxlength: maximum length for summary
        workers: number of concurrent workers to use for processing data, defaults to None

    Returns:
        summary text
    """

    # Validate text length greater than max length
    check = maxlength if maxlength else self.maxlength()

    # Skip text shorter than max length
    texts = text if isinstance(text, list) else [text]
    params = [(x, text if len(text) >= check else None) for x, text in enumerate(texts)]

    # Build keyword arguments
    kwargs = self.args(minlength, maxlength)

    inputs = [text for _, text in params if text]
    if inputs:
        # Run summarization pipeline
        results = self.pipeline(inputs, num_workers=workers, **kwargs)

        # Pull out summary text
        results = iter([self.clean(x["summary_text"]) for x in results])
        results = [next(results) if text else texts[x] for x, text in params]
    else:
        # Return original
        results = texts

    return results[0] if isinstance(text, str) else results